ML20005B464
| ML20005B464 | |
| Person / Time | |
|---|---|
| Issue date: | 07/06/1981 |
| From: | Advisory Committee on Reactor Safeguards |
| To: | |
| References | |
| ACRS-T-0877, NUDOCS 8107080257 | |
| Download: ML20005B464 (72) | |
Text
-
j ;,? CSl?
U NUCI,IAR REGUI.ATORY CCMMISSICN f,.
~
, qsh
[p, s
' ;/-
3
,/
.$g V,j%
/~sj ff r[\\9(p.p)
$+
l
\\cf
,q /
\\
\\
s/\\
./ \\
4 Nr r i g \\ c -
o l
1 In the.v2.t::er of:
ADVISORY COMMITTEE ON REACTOR SAFEGUARDS SUBCOMMITTEE ON DECAY HEAT REMOVAL SYSTEMS CA".T:
July 6, 1981 pAggg:
1 - 65 f
A;;
Washington, D.
C.
l~
,/
_f,,~'
(Lb,
- 1, t
c&
/' ar m rua
.umsos O
F-Q 400 Virginia Are., S.W. Wasning: n, C.
C.
20024 O
1 O
i ux1TED STATES Or AMER 1cA 2
NUCLEAR REGULATORY COMMISSION O
3 ADVISORY COMMITTEE ON REACTOR SAFEGUARDS 4
SUBCOMMITTEE ON DECAY HEAT REMOYAL SYSTEMS g
6 7
Nuclear Regulatory Commission 1717 H Street, N.W.
8 Room 1046 Washington, D.C.
9 l
Monday, July 6, 1981 10 The meeting of the Subcommittee on Decay Heat 11 Removal Systems was convened, pursuant to notice, at 12 1s00 p.m.
13 j O atastaS rasSz"T=
14 D.A. WARD, Chairman 15 H. ETHERINGTON J.J. RAY 16 DESIGNATED FEDERAL EMPLOYEE:
17 R. SAVID 18 ACRS CONSULTANTS PRESENT:
19 E. EPLER 20 P. DAVIS 21 22 i
23 24 25 ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON D.C.20024 (202) 554 2345
2 1
E E 9 C. E E E I E E E 2
MR. WARD:
The meeting will now come tc order.
}
3 This is a meeting of the Advisory Committee on 4 Reactor Saf eguards, Subcommittee on Decay Heat Removal 5 Systems.
6 I am David Ward, the Subcommittee chairman.
The 7 0ther ACRS member present today is Harold Etherington.
We 8also have Mr. Epler and Mr. Davis, who are consultants to 9 the Committee, and another member has just arrived, Mr. Ray.
10 The purpose of the meeting is to discuss Draft 11 Task Action Plan TA-45, Shutdown Heat Removal Requirements.
12 This meeting is being conducted in accordance with 13 the provisions of the Federal Advisory Committee Act and the O
14 G ove rnmen t in the Sunshine Act.
D r. Richard Savio is the 15 designated federal employee for th e mee ting.
Rules for 16 participation in today 's meeting have been anncunced as part 17of the Notice for Meeting previously published in the 18 Federal Register on June 18, 1981.
I 19 A transcript of the meeting is being kept.
It 20will be made available by July 7, 1981.
It is requested 21 that each speaker first identif y himself and speak with 22 suf ficient clarity and volume so that he may be readily 23 h ea rd.
We have received no written comments or requests for 24 tis e i.o mak e oral statements from members of the public, so 1
25 we will proceed with the agenda as most of you have received O
l ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
3
()
l i t, and I will call on Mr. Marchese of the staff.
2 (Slide)
"N 3
MR. MARCHESE:
The main purpose of today's meeting (J
41s to discuss the ACRS Subcommittee's comments on Task 5 A-4 5.
What I would like to do is keep the meeting sort of 6 semi-inf orm al and have what I would call across-the-table i
7 dialogue on your review of the Draf t Task Action Plan.
8 I think we have the key people here today tha t 9 besides myself should be able to respond to your comments 10 and discuss them, and hopef ully by the end of the meeting we 11 will have some sort of ag re em e:' t on which comments should be 12 incorporated in the plan to make sure that some work is done 13 o n it.
O l
14 Before we get into your specific comments on the I
i 15 pla n, I would like to give you a brief update on what has 16 been going on in the last couple of months since we met on 17 M ay 5.
18 (Slidei 19 The first item, the Task Action Plan in draft 20 form, was issued on May 22, 1981.
Today we are here to 21 review the Subcommittee's comments on A-45.
The last couple l
22 c f weeks the plan has been receiving what I would call very l
23 intensive internal review, and I expect by the end of this f}
\\/
24 week we will have internal comments from the lir.e branches 25 that are going to be involved in one way or another in (1)
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
n.
I 4
~( f 1 carrying out the plan.
2 I hope in about a month to have the pir7 in final
)
3 form where it will be ready for management to cor.vur on 4 basically to commit the internal manpower and to improve the 5 technical assistance f unding that would be required for FY 6 82 technical assistance contracts.
7 Right now we are shooting for August 7 to issue 8 the plan in final f orm, which is about a month from now.
9 The next two items that I will te concentrating on 10 in the next several months is to discuss with various 11 national labs and private organizations, private firms their 12 capabilities to assist us in performing work that is 13 outlined in the Task Action Plan.
So during the ne x t few O
14 months I will be holding discussions with various labs and 15 firms to examine their capabilities.
16 After this is complete we will be writing requests 17 for proposals f or FY 82 technical assistance contracts, and 18! hope to have this complete probably by the end of the 19 summer.
Implementing contracts with national labs, I hope 20 to have this complete by the end of October; and 21 implementing contrccts with p'riva te firms, since it 1" a 22 more involved sort of thing because we have to solicit bids, 23 evaluate the bi ds, select probably several of the best bids 24 and do another evaluation, it is sort of an involved process 25 tha t takes at least thrce to six months by the time you are O
ALDERSON REPORTING COMPANY,INC.
400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345 x
5
()
linvolved with competitive bids.
2 So I am projecting that perha ps by the end of
()
3 January of '82 we will have all the technical assistance 4 work in place at least for FY ' 82.
5 MR. RAY:
Quention.
Is the terminal date still 61984?
7 MR. MARCHESE:
That is what we are shootir.g for.
8 MR. RAY:
Is tha t rtill the targe t?
9 HR. MARCHESE:
Yes.
10 MR. RAY:
Do you see any possibility to advance 11 that date?
12 MR. MARCHESE:
At this point I do not see it 13 ceally being available to advance that date.
O 14 MR. RAY Even that date, unless you put 15 sig nifican t manpower assignment on it, is a hard one to meet.
16 MR. MARCHESE:
There is a potentia' Tat that date 17 could slip, too, especially if the technical assistance 18 budget as outlined in the plan is not approved.
19 (Slide) 20 There are a few items that before we get inte your 21 comments, that I would like to bring to your attention.
I 22 h ad the opportunity about a week ago to tour the McGuire and 23 0co nee, wha t they reter to as their standby shutdown 24 f acilities, and the Oconee visit was particularly 25 interesting in terms of I believe it probably represents the
/~h (J
ALDERSON REPORTING COMPANY, INC.
400 VIRGINIA AVE., S.W.. WASHINGTON C.C. 20024 (202) 554-2345
i 6
()
1 closest thing we have right now to a dedicated shutdown 2 cooling system.
I 3
If time permits and if you are interested, I think s'
)
4 I can sort of highlight some of the key things that I saw a t 5 0conee I thought were particularly interesting.
So we might 6vant to come back to that later on.
7 MR. WARD:
Yes, I think we will, Andy.
That 8 sounds good.
9 MR. MARCHESE:
The next two items, I think the 10 third item, you can consider this as sort of a major comment 11 on the plan that sort of evolved since we last talked about 12 i t, and it is really the reason why I say that there may be 13 a very strong reluctance to issue 1.139 right now until we O
14 start going to work on A-45.
We may have to add a major 15 subtask to the plan which involves perf orming r. cos t-benefit 16 evaluation to determine if it is necessary for existing 17 plants - to have the capability to get to cold shutdown using 18 saf ety grade equipment.
19 As you know, 1.139 basically states that all the 20RHR systems to get from cold to tot shutdown should be 21 saf ety g ra d e.
Well, it turns out that most of the existb g 22 plants that are out there, probably equipment that is 23 required to get to cold shutdown, some of it may not ce l
N 24 saf e t y g ra d e.
25 Now, to do an evaluation to determine vnether or l
l ALDERSoN REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON. D.C. 20024 (202) 554 2345 l
_.~,,.
L
()
Inot it is necessary for these existing plan ts to have this 2 ca pabili ty would mean a fairly expensive level of effort,
}
3 and if this is added to the plan -- and we are still kicking 4 this around -- it could involve a major level of effort.
5The manpower identified in the plan would have to be 6 increased, as well as the technical assistance.
7 The plan righ t now ba sically sta tes that the 8 plants should have the capability to get to cold shutdown 9 using safety grade equipment.
So one of the things we are 10 kicking around is whether or not we should make that a firm 11 requirement, which means looking at the benefits, costs, 12 a ss essing or having someone assess what is the risk 13 contribution of getting from hot to cold shutdown, and is 14 tha t significant; how important is it to go in there and do 15 maintena nce and repair, which was brought out, I think, 16during Three Mile Island as one of the reasons why you would 1711ke to get to cold shutdown.
You would like to get in IP.there and do repair.
That is basically one of the major 19 problems in the plan so far.
20 So at this point I guess I would like to sort of l
21 turn it into what I would call an across-the-table kind of 22 dialogue and get your comments on the plan.
Hopefully, we 23have enough people here to respond.
f~)
l
'N /
24 I think I have one more chart here which I think, 25as f ar as getting the plan on one page, I think this figure
(
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S W WASHINGTON D.C. 20024 (202) 554 2345
. _. _ ~.. -
8 mU 1probably tends to do that.
2 (Slide)
(')
3 It outlines sort of in a logic network how the V
4different subtasks tie into one another.
Basically we have 5 developed interim acceptance criteria first feeding into 6 assessing the adequacy of existing decay heat removal 7 systems against those acceptance criteria.
We added another 8 major effort here, 2.5, which is the deterministic type of 9 evaluation using conventional engineering techniques, which 10I believe was a major comment that the Subcommittee had at 11 the last meeting.
This was not on the last chart.
12 We also have a block of effort outlined in the 13 eve nt that the probabilistic risk analysis basic approach 14does not turn out to be sufficient or completely useful.
We 15also have a major task tha t will be doing more conventional 16 engineering evaluations to see whether or not the shutdown 17 decay heat removal systems are adequate at the plar.ts.
18 Down at this point we get into naturally 19 e st ablishin g a feasibility of shutdown decay heat removal 20 system designs dedica ted, the f easibility in being ab3e to 21do the preliminary cost estimate.
At this point we compare 22the effectiveness of alternative means of reducing risk, 23 seeing how other areas of reducing risk such as improved 24 con tained design, vented filtered containment systems, 25 p os t-a ccid e n t core retention to see how their benefits s ta c k O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554 2345
9 lup against the cost-benefit ratios established from this 2 program.
3 At this point if there is a need to reiterate on 4the acceptance criteria because the costs are turning out to Sbe perhaps prohibitive, we can consider an iteration where Sve would perhaps have to revise the criteria.
And then 7 finally the plan or implementation.
That is actually 8vriting the documentation which the licensees would have to Gheed in terms of new requirements for dedicated systems.
10 So we could kind of use this as a guide to 11 commenting on the plan.
12 MR. WARD:
I notice in reading the draf t plan 13 there does not seem to be much discussion of the BWR as
. O 14o p posed to the PWR.
Can you tell me what your --
15 MR. MARCHESE:
I guess there is a limited amount 16of things that come to mind in terms of improving the decay 17 hea t removal systems for a BWR.
The first thing that comes 18to mind is the suppression, that that suppression pool 19 cooling is something we would not want to focus in on right 20 a wa y.
21 There is far less discussion on the boiler because 22it is not immediately apparent, other than the suppression 23 pool, what we might concentrate on in terms of operating 24their decay heat removal systems.
i-25 MR. WARD:
Is it that you do not see how to solve OO ALDERSoN REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
10
()
Ithe problem if there is one or that you do not see a problem 2vith the PWRs to the extent there is with BWRs, if there is
}
3vith BWRs?
4 MR. MARCHESE:
I cannot say that we see a problem 5there until we get in and develop the acceptance criteria 6and establish rome risk targets and then go in and assess 7various BWRs to see how they stack up.
It is hard to see 8how it is going to turn out.
9 We do mention some things that at least right at 10the beginning we felt we could look at.
I think we 11 men tioned suppression pool cooling and also maybe putting a 12 reflux condenser of f of the -- either right in the vessel or 13off the vessel in some form.
I think it is sort of similar 14to what Westinghouse has talked about in terms of their 15 advance decay heat removal system, but until we get into the 16 BWRs more, it is not apparent what other alternatives we i
17 mig h t want to look a t.
18 Now, Sandia in their report I think has mentioned l
19three shutdown dedicated systems that they are going to be 20 a ss essin g, I think three for boilers and six for the PWRs.
21 MR. WARD:
Well, do you think that the program as 22you have it outlined is balanced enough so that you will be l
l 23 making as good an assessment of the potential or problems in 24the BWR as you have in the PWR?
25 MR. MARCHESE:
I hope it is.
I think if we find O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S W., WASHINGTON, D.C. 20024 (202) 554-2345
,. _, _ ~ _, _.. _
11 4
,-')
1out, for example, there is a boiler out there that
(_
2 represents a high risk, a disproportionate risk, we should
(~)'s 3 find tha t out, hopefully in the next six to twelve months 4because there are a couple of these risk assessments for 5 boilers.
Limerick comes to mind right now and there are 6several others.
7 If we find out that the risk is high, we would 8have to focus more attention on the boiler right off the 9 bat, and I think that is in the plan.
If we find that some 10 plants have a disproportionate risk and we find that 11 improving their decay heat removal systems would lovar that 12 risk significantly, those plants would have a very high 13 priority in the plan in terms of getting immediate attention.
14 MR. WARD:
Do we have any comments?
m 15 MR. EPLER I have a question I have not 16 encountered any discussion of.
It is the issue of whether
[
17we should go in the direction of making more options l
18available to. the operator or whet :er we should minimize the l
l 19need for operator par 61cipation.
We have evidence, I think, 20 tha t the operator is the weakest link in the system, and by 21 m ak ing more options available we can thereby require 22 additional retraining.
23 I have not seen a discussion of the system that 24was intended to require the minimum operator participation.
25I think that is a major point that should be on the first oV l
ALDERSoN REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W. WASHINGTON, D.C. 20024 (202) 554-2345
12 O,s 1 page.
s_
2 MR. MARCHESEs As part of the development of a
}
3 dedicated system, I agree with you.
I think opera tor action 4should be a very vital part of that assessment.
Now, we did Snot intend in the plan to describe in detail, you know, a 1
6 specific dedicated system.
We put in we were going to be 7 assessing various dedicated systems.
But I agree that 8 minimizing the operator action ought to be a very important 9part.
I think that is an important aspect.
10 I might sention, and this gets back to the Oconee 11 visit I had, that their system is all manual and an operator 12vould have to run down to this dedicated building and turn 13on the system in a manual f ashion.
It is not automatic.
So 14 ve would have to assess tha t.
The cost implications and so 15forth would be part of the assessment.
16 What I heard down there is automatic initiation 17could be very expensive.
18 MR. EPLER:
Yes.
19 MR. MARCHESE:
But you would have to trade that t
200f f against other things.
21 MR. EPLER:
One additional question, if I may.
22The question of independence of the dedicated system I do 23 no t think has been adequately addressed.
One of the major 24 problems we now have is that the failure of vital equipment 25 or vital services will in turn cause the need for those
(%s ALDERSoN REPORTING COMPANY,INC.
400 VIRGINTA AVE., S.W., WASHINGTON. D.C. 20024 (202) 554 2345
13 (m
(,)
1 services or equipment; therefore, the dedicated system must 2 have no capability whatever of causing plant shutdown and
'~}
3 causing it itself to be needed.
4 I have not seen this called out anywhere as a 5 requirement.
6 MR. MARCHESEs That is another important point.
7 Again I mention we are not trying to spell out in the plan 8what the requirements are going to be.
We are planning to 9 develop those requirements and I think that is a very 10important point that definitely ought to be considered in 11 e stablishing dedicated systems, make them completely 12 ind ependen t.
13 MR. EPLER I'm doing so well, maybe I could ask a O
14 thi rd question.
15 MR. MARCHESE:
By the way, I think in the Sandia 16 d ocumen t on alternative decay heat systems, tha t is spelled 17 ou t, I believe.
18 MR. EPLER:
I think one of the Sandia proposals in 19 f act isolated existing systems when it went into action.
Ic 20 a ppea t s tv me that this would indeed cause, if it acted 21 s pu riously and isolated otner systems, it would lead to l
22 pretty serious dif fi ulty, wouldn't it?
l 23 MR. MARCHESE:
Taylor.
24 MR. TAYLCRs I do not think we have reached the 25 point where we can say exactly what all the logic would be D
i
/
v ALDERSON REPORTING COMPANY. INC.
400 VIRGINIA AVE., S.W., WASHINGTON. D.C. 20024 (202) 554 2345
~ - _ - _ _.,_
14
()
ton an alternate heat removal concept.
That is part of the 2 work that is ongoing right now.
We have called or exploring
)
3 the features of automation, certainly we have called for
(~J 4 criteria, independent criteria to keep the alternatives as 5 sepa rate as possible, separate from the existing normal 6 systems, self-independent, if you will, also self-powered, 7 what have you, their own AC power supplies ~.
8 But I do not think we have reached the point where 9 we can say right now that the logic would isolate all normal 10 systems or cause a shutdown.
But I would add one other 11 thing.
I do think we are kind of co'ncerned at this point in 12 time about a heavy reliance on automation.
I think we want 13 to leave the operator with a little flexibility to control 7,
14 these systems.
15 I also would like to point out that the word 16 " dedicated" by itself in our terminology implies automation, 17 tha t it is dedicated to a function independent of the 18 o pe ra to r.
So I woulu 11so like to correct the impression 19 t ha t we have not reached the point that we can say 20 d ed ica ted.
That is full automation.
21 MR. "ARDs Let's see.
Matt, the fact that 22 dedicated implies automation seems to be kind of a special 23 definition on your part.
Could you explain that a little
/'_N
()
24 bit more?
It is not clear to me why there cannot be a 25 dedicated system that would require initiation by the tN ALDERSON REPORTING COMPANY. INC, 400 VIRGINIA AVE., S.W WASHINGTON. D.C. 20024 (202) 554 2345
15
()
1 operator.
2 MR. TAYLOR:
That is true, but dedicated, at least 3 as we have looked at it, has meant -- every, time, I believe, 4 that I have seen this used in both European terminology as 5 vell as wha t my understanding is of concepts put forth by 6 Dr. Okrent and Dr. Ebersole -- in essence automated away 7 f rom the operator. The controls are es entially away from 8 the operator.
O MR. WARD 4 So you are saying every example that 10 has been put forward as a dedicated system --
11 MR. TAYLOR 4 As a dedicated system has been 12 autcr.ated in essence independent of the operator.
The 13 operator can everride but it is rather an excruciating job O
14 to do so, so I think I wanted to set that fine point 15 straight, that at least in my mind a dedica ted system, I do 16 no t think we ha ve concluded tha t a dedicated system, i.e. an 17 automated syst em, is necessary at this point in time.
18 MR. WARD 4 Well, would you be more inclined to 191ook f avorably on a dedicated system that was not automa ted?
20 MR. TAYLOR:
Yes, personally.
Personal opinion.
l l
21 MR. WARD:
Okay..
d 22 MR. MARCHESE:
Anyway, I think these two points on 23 operator action and independence are good points in terms of C
\\
24 their plan, developing the systems concepts.
They are two 25 important items tha t will have to be considered.
AV i
ALDERSON REPORTING COMPANY,INC, 400 VtRGINIA AVE., S.W WASHINGTON, D.C. 20024 (202) 554 2345
16
()
1 MR. SAY:
Since we are clarifying definitions, 2 there is no question but that dedicated mea ns it has its own 3 power supply, it is not dependent on of f site power.
4 MR. MARCHESE:
I think that is --
5
'MR. RAYS That is a basic requirement.
6 MR. MARCHESE Yes.
7 MR. WARD:
Let me ask you a question about Item 8 3.5 which you have added, and what you have said, that if 9 the probabilistic assessment does not lead to some definite 10 conclusion, it will go to a deterministic assessment.
11 MR. MARCHESE:
Ma ybe I think the plan right now, 12 the wording has to be changed in terms of we have it 13 inverted as a subtask.
I think it is going to have to start O
14 off right a t the beginning and go in parallel rather than 15 waiting at this point.
I think we are going to have to 16 s ta rt right at the beginning.
17 MR. WARD:
I guess what I am getting at is there 18seems to be some of the preliminary indications from 19 probabilistic numbers that have been genera ted so f ar that 20 there is n~ t much to gain by improving the decay heat 21 removal systems, improving their reliability.
That is, you 22 know, it is begging the question to an extent.
Are you 23 going to look at whether that is the answer o'; not or a more
)
24 con side r ed evaluation.
But if it does show tha t there is
(
25not much to gain probabilistically, do you think that is O
ALDERSoN REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
_ _ __ _ __ _._ _ ~
17
()
1 going to give you a firm enough position to say that nothing 2 further needs to be done in improving decay heat removal
}
3 systems?
4 MR. MARCHESE:
I think we talked the last time, 5 comparing plant to plant, that the reliability, say, of the 6 aux feedwater system -- I think this is NUREG-0611 that Matt 7 Taylor has worked on extensively -- the variability from 8 plant to plant on the reliability of the aux feedwater 9 system was like two orders of magnitude.
10 Now, we have taken steps to improve that 11 situation, but say we go through HBI and we find that same 12 sort of variability, two orders of magnitude, I think as 13part of this plan we would take steps to upgrade those that 14 are in the lower end of the spectrum.
I do not think, and 15this is my own personal view, tha t it is acceptable to have 16 that kind of range f rom plant to plant.
17 I think we would have to require some sort of 18 upg rading -- that is just an additional reaction -- as was 19done on the feedwater system, even if we find that upgrading 20does not reduce the risk substantially, and tha t may be true 21 on some plants, including the overall decay heat removal 22 system does not improve this, but still to bring their 23 systems with, say, the industry norm or whatever.
r
(_)\\
24 Matt, d. id you have any response to that?
Did you 25 v an t to add anything to that, because you worked extensively GU l
I ALDERSON REPORTING COMPANY, INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
~. _. _ _
18
()
l on auxiliary f eedwa ter?
2 MR. TAYLOR:
Well, I would add, just to back up
}
3vhat And y said, we did find two orders of magnitude 4 variation in the auxiliary feedwater system for 33 PWR Sunits.
This does not translate into a risk variation but it 6does translate into perhaps a need for the regulatory people 7 to assure a little less variation in a system that is 8 demanded with such frequency, several times a year.
9 Also we found seven units that had a complete 10 dependence on AC.
These were, again, if you look at the 11 possible risk imblications associa ted with just those 12 d e p en de n c,ie s, those dependencies could assure a very high 13 risk at a plant, so there are things that can be done, 14 including alternative systems, to improve the reliability of 15 the auxiliary f eedwater system, and I think we went with the 16 existin g plants as far as we could to, let's say, improve 17 f ac tors of, let's say, 2 to 10 in the system on 18 una vaila bili ty.
19 But we did not necessarily approve the overall 20 risk or the core melt probability with anywhere near that l
21 v alu e.
So what we have already seen across the population 22of light-water reactor designs is that the core melt 23 probabilit y is indeed domina ted by the transients and the m
\\_)
24 small LOCAs.
We have seen this across a fairly good 25 population of designs.
bq l
I ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554 2345
l 19
()
1 We also have seen the risk, with the exceptior, of 2 one PWR design, f rom, oh, I would say six or so PWRs, have
}
3 seen the risk influenced by the total loss of AC power 4 scenario.
Also we have seen the risk influenced by such 5 things as the unit f acing system.
That is entirely separate 6from the question of decay heat removal.
One needs to 7 tackle those other problems.
If one is striving to improve 8the risk level, one needs to improve those risk sequences 9somewhat independently of the decay hea t removal system.
10 So whether one gains just a factor of 2 may mean 11 tha t you could gain substantially more than a factor of 2 or 12substantially more than a f actor of 5 if you jointly 13 operated on several sequences.
.s d
14 MR. MARCHESE:
One other point I might add is on 15 page 11 of the plan under subtask 1 where we are talking 16about developing interim acceptance criteria.
We outlined 17 the ACRS proporal, and specifically Item 4 of the ACRS 18 proposal is utiliring the ALARA principle, as low as 19 rea so nably achievable.
7 think that also would be a t
20 mechanism of even though the risk improvement, including the 21 RHR, is not substantial, you would have to base it on the 22 ALARA principle to see how it stacks up in terms of 23 cos t-be n efi t.
24 HR. DAVIS:
Yes. I have a question.
About a year 25and a half ago I was involved in a reliability study on O
ALDERSoN REPORTING COMPANY,INC,
' " " ' " ' ^ ^ " ' ' * * ^ * " ' " " ' " -
i
20
()
1 auxiliary feedwater system to try to quantify the failure 2 probability, and as part of that study I visited a plant
()
3which was operating and had a discussion with the operator.
%./
4I asked him what he would do if he had a transient in the 5 plant, lost main feedvater and could not get auxiliary 6feedwater to operate; and his answer was, well, we have a 7 brand new procedure for that, and he pulled it out and I Bread it, and the procedure was as.follows.
9 Ihe first thing to do is depressurize the primary 10 b y opening the pressurizer relief valve and try to get the 11 pressure down to the low injection system injection 12 pressure, which is maybe 200, 250 psi.
Then the procedure 13went on to say if tha> does not work, select a steam 14 generator ti.at does not have tube leaks, depressurize the 15 secondary side until you get down to the condensate pump 16 discharge pressure, which is about 200 psi, and then use 17 tha t to remove heat through the secondary.
18 It occurred to me at the time and I guess from 19 wha t I have seen since it is somewhat reinforced that that 20 procedure may not be the optimum if you lose auxiliary 21 f ee d wa te r.
It seems to me that you would try the secondary 22 option first before you deliberately lost your primary 23 system inven tory.
I am getting to the question.
p)
(.
24 I am wondering now, and I do not see anythina in 25your Task Action Plan, how many other procedures like that O
N-ALDERSoN REPORTING COMPANY,INC, 400 VIRGINTA AVE., S.W., WASHINGTON, D.C. 20024 (2(2) 554 2345
21
()
1are out there that really are not workable.
I have a 2 concern that some procedures are detrimental and have not
}
3really been thought out and analyzed in depth and been shown 4 or verified to be effective.
5 Do you have any plans for re-evaluating these 6 procedures, some of which appear to be hastily developed as 7a result of TMI-2, and making sure that these backup 8 operator intervention procedures are optimum and effective?
9 NR. MARCHESE:
They have what I call last ditch 10 procedures spelled out for removing decay heat. I agree, I 11do not think we have that in the plan.
Does anybody have 12any reaction to that?
I think it is a good point.
Probably 13ve ought to reflect that in the plan.
That is a good point.
O 14 MR. CAVE:
In principle tha t poin t should be 15 covered by the probabilistic risk assessment of an 16ind ivid ual plant, but whether in f act the risk assessment 17tha t is going on now to cover that, I am not sure.
But I 18 tak e your poin t.
I has to be covered somehow.
19 MR. MARCHESE:
Okay.
20 MR. WARD 4 Well, the plan does include you have a 21 basis f or continued operation in licensing pending I
l 22 completion of the task.
23 MB. MARCHESE:
Right.
G(/
24 MR. WARD 4 And I should think that a systematic 25 evaluation of procedures that had been developed would be ALDERSON REPORTING COMPANY,INC, j
400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554 2345
22
()
1part of th a t.
2 MR. MARCHESE:
Let me ask one of our people on the 3SEP program, Safety Evaluation Program, at th e 11 plants.
4 Are they doing anything in terms of reviewing procedures?
MR. LOV ELL :
In the SEP reviews that were done, 6one of the most irportant parts of the review was reviewing 7 the plant procedures, and we visited plants, all 11 plants 8and talked to the shift supervisors and operators and 9 engineers a t the plant, and wa had a joint task force with 10 ICE, and the ICE people did most of that review but the NRR 11 people were there with them.
So that was looked at pretty 12 carefully for those plants.
13 But those are the only ones that have had a O
14 detailed look.
For the plants that have been done, where at 151ea st draft SERs have been written, that was an important 16 par t of the review.
17 MR. MARCHESE:
Thank you.
18 HR. WARD:
So Pete, I guess I could ask you was 19this particular plant you had r eference to one of the 11 20 older plants that is under the SEP program, do you know, 21 o r --
22 MR. DAVIS:
No, I do not knoi.
I have not checked 23 tha t, Dave.
24 MR. WARD:
Does that mean, Andy, that there is a 25 cap in the evaluation here, tha t the 50 or so newar plants O
ALDERSoN REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
i 23
()
1have not been evaluated in this way?
2 MR. MARCHESE.
I really cannot answer that.
I am
('~')
3not sure to what extent we have reviewed all the procedures, V
4 you know, to get the hot and cold shutdown.
5 MR. LOVELL:
I think this sounds like a new 6 post-TMI-type procedure.
Even if it is an SEP plant, it may 7not have been in existence at the time of the reviews.
8 MR. MARCHESE Do you know if that plant had feed 9 and bleed capability?
10 MR. DAVIS:
It is a Westinghouse plant.
11 MR. MARCHESE Half of the Westinghouse plants 12 have tha t ca pability, half don't.
13 MR. DAVIS.
I do not know what the disch'rge a
O^
14 pressure is on the HPI as for that plant.
I do not recall 15right now.
But feed and bleed was not included in the 16 emergency procedure as an alternate, and maybe it should 17 have been.
Maybe that is another reason that these 18 procedures need to be looked at.
19 MR. MARCHESE You mentioned turning on the LPI 20 p u m p, getting down to that level.
I think you would want to 21try feed and bleed if you had the capability.
22 MR. DAVIS:
I agree.
23 MR. MARCHESE:
I think this point is an important
)
24 one. I think we will try and reflect it in the plan.
25 MR. EPLER:
I might raise another issue.
This
. O t-ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345 L
24
()
1 discussion has desit with probabilistic assessment, risk 2 assessment, and with operator procedures. It is clear to me 3that there is something more than risk involved here.
It is 4 the public's perception of risk, and what we do may 5 ultimately be dominated by that rather than the real risk, 6which may be quite adequate as things now stand.
But 7 certainly the public 's perception of risk does give us a 8 feeling.
9 Now then, when it comes to reactor shutdown, the 10 operator does not struggle, he never struggles to shut down, 11 and in the large break LOCA he never struggles because it is 12 an improbable events but we have two very highly probable 13 events, residual heat removal and keeping the core covered,
~
14 and in both those cases the operator has on several 15 occasions struggled, in the public view, ultimately with 16some measure of success but not successful in instilling in 17 the public that we know what we are doing.
18 Now then, it may be that we do not improve the 19tisk by some of these measures, but we certainly can improve 20 our image, and that may be ultimately the dominant 21 consideration to stay in business.
Now, this says to me 22that we must not only attempt to get the heat out with the 23 minimum of highly visible opera tor participation; we must 24 also keep the core covered with the minimum of visible 25 operator participa tion, and these, I think, both transcend 1
O(J l
l.
ALDERSON REPORTING COMPANY. INC.
400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
25
()
1 the isrue of risk assessment.
Maybe it is more important to 2 improve our image than to improve the risk.
I do not see a
(~}
3 dicacussion of this.
V 4
MR. MARCHESE:
It is a difficult thing to reflect Sin the plan.
I think it is part of doing the cost-benefit 6 evaluations on dedicated systems.
Hopefully those 7 evaluations will flesh this out.
Because I will tell you 8 one thing I gained an appreciation for is the cost involved 91n dedicated systems when I went down to Oconee.
10 You know, we are talking about a $34 million 11 effort down there that started in 1978, and I guess that is 12 no t all the cost.
So we need to develop these systems, 13 assess their costs bef ore we can, you know, make a decision 14 whether or not it is worth it, even if it does not reduce 15 the risk, e ven if it still requires such a system to, you 16 know, enhance public confidence.
17 What you are saying may be right, but I really 18 can not say at this point that we would require, you know, a 19 multi-million dollar system just to increase the public 20 p e r ce p tion, just enhance the public perception.
l 21 MR. DAVISs I have a related question, Mr.
22 Chairma n.
I think that using a cost-benefit a pproach is 23 cor rect, but I am not sure how you are going to neasure i
)
24 ben efit.
Presumably there will be a risk reduction as the 25 benefit.
l O ALDERSoN REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
26
()
1 MR. MARCHESE:
Right.
2 MR. DAVIS:
But how do you quantify what that risk 3 reduction is worth?
I do not know whether you were going to 4use life-shortening or genetic effects or what you were 5 going to use, but how do you know if something is worth it 6or not, depending on what the risk red uction is.
7 MR. MARCHESE:
It is part of subtask 1 in terms of 8 developing criteria.
The way the plan is organized, right Sup front we develop, you know, the decision rules, 10 seceptance criteria, you know, to make these hard decisions.
11 I think we are going to be talking about perhaps major 12 expr 1ditures. I mean you have to have some targets right up 13 f ron t.
If the plants do not meet the goals, we are going to O '
~'
14 require an upgrade to make sure the plant meets the goal.
15 MR. CAVE:
Well basically we think that the type 16 of criteria spelled by the ACRS in the trial goals will be 17 suitable for the existing plants.
They consider 18 cost-benefit on three diff erent levels: early deaths saved 19 and delayed cancer deaths saved, and also property damage l
20 saved.
So this does give you one way of looking at the 21 cos t-ben efit.
But that approach is overshadowed, as they 22 say themselves, by the uncertainty in the probabilistic risk 23 a sse ssme nt.
24 There is another way I think we could possibly 25 tackle this, working up some ideas on this, that you can 1
ALDERSON REPORTING COMPANY,INC.
400 VIRGINIA AVE., S.W WASHINGTON D.C. 20024 (202) 554 2345
_ _ ~ _. _ - _ ~...,. - - _., - _.,, -. _. _. _ _. _.. _
27
()
1 take an existing plant, and its continued use has got some 2 very definite value to the community.
Some proportion of 3 tha t benefit might be allocated to improving the plan t.
In 4 fact, the real risk or, as Mr. Epler says, the proced.ure S risk is too high to be accepted.
6 For example, if a plant has ten years to run, it 71s probably worth between $100 million and $1000 million in 8 terms of potential fuel costs, in terms of other means of 9 f uel genera tion.
So even if you take 20 percent of that to 10 assure its continuation f or another 20 years, then you would 11 have a usef ul sum in which you might set aside to improve 12 the plan t.
13 MR. ETHERINGTON.
In looking at the reliability of O
14 RHR systems I would like to echo the chairman's feeling that 15 the PWR should not be too much cited in the early stages of 16 tha t program.
In a PWR the RH system is essentially a 17 o ne-p u r p os e system.
In the BWRs it is a multipurpose 18 sys te m.
There are many connections, both on the suction 19sile and the discharge side.
It is used for removing low 20 tem pera ture decay heat in the usual way.
It is used for 21 cooling the pool water, and in some systems it has a steam l
22 condensing mode in the shell.of the heat exchanger.
23 Now, a good many years ago the Committee raised a
[~)
\\_e 24 question as to the possibility of water hammer going from l
25 the water in the shell to the steam in the shell mode or l ()
L i
i ALDERSoN REPORTING COMPANY. INC, 400 VIRGINIA AVE., S.W., WASHINGTON. D.C. 20024 (202) 554 2345
28
()
1vice-versa, but I think that was before the feedwater water 2 hammer incidents that occurred in PWRs.
I never heard 3 any thing but a kind of pooh pooh of the possibility.
For V
4 example, I guess it can be shown that there is no 5 possibility of water hammer, the steam condensing mode might 6be prohibited.
This is just for an example.
7 You know, I think you should look at it early and 8not wait until you have got all the PWR problems more or 91ess under control.
I know you cannot do everything all at 10 once, but I just suggest you do not leave it too long.
11 MR. MARCHESE:
Okay.
I believe the Sandia people 12 are focusing on RHR for the boilers; is that correct?
That 131s one of the three alternatives?
Maybe you want to say 14 som ething about what is being done on the dedicated system 15 f or boilers.
16 VOICE:
At Sandia we have looked at and are 17 currently looking at in more detail three alternatives, 18 con ce p ts.
We are looking at three alternative-type concepts 19 f or BWRs.
As Andy indicated earlier, we feel there is not a 20whole lot of flexibility in the types of systems or concepts 21 tha t we look a t, especially in terms of retrofit, so we 22 selected three concepts that work similar to the current 23 sup premsion pool cooling systems and RHR cooling systems of
()
24 B WR s.
25 The concepts we are looking at are concepts that O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
1 29 i
()
1could ~ wo rk in one of three types of modes.
One would be 21ooking at a high pressure closed loop system that would not 3 depressurize the plant.
Another concept would be to look at 4 the duplicating or replicating the current depressurization 5 systems in BWRs where you use a low pressure injection after 6you depressurize.
7 The third mode would be something where you had a 8 controlled depressurization which would work together with a 9 controlled injection that would vary from high pressure to 101ow pressure.
All three of thase concepts, though, would 11 req uire the secondary cooling loop that would eventually 12 remove heat to an ultimate heat sink.
13 In doing our work we have chosen a base case power O
14 pla n t that has steam condensing capability in its RHR 15 sys te m.
We are allowing ourselves to vary the reliability 16of th at existing base case system to include highly reliable 17 condensing capability or unreliable condensing capability.
iCIn that way we can, in our opinion feel that we can get an 19 assessment of the sensitivity or the reliability of existing 20 systems and changes in that reliability as a result of 21 adding other systems.
22 MB. ETHERINGTON:
I think I probably did not make 23myself clear.
I was not talking about concepts; I was I
)
24 talking about appraisal of present systems.
25 MR. BARRY:
We do not intend to do an engineering O
V ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W WASHINGTON, D.C. 20024 (202) 554-2345
30
()
1 appraisal of existing systems, but we feel, in the same way 2you have indicated, that the condensing mode of operation 3may not work because of some physical reason.
4 MR. ETHERINGTON:
It may be a hazard.
5 MR. BARRY:
We feel we can look at that from the 6probabilistic standpoint by varying its reliability from 7zero to one and seeing if that really means anything in 8 terms of the overall risk to the power plan t.
It may be 9 that an unreliable condensing mode operation may have little 10 imp act on the overall reliability of the plant.
11 MR. MARCHESE:
Thank you.
12 MR. WARD:
Andy, I have the feeling, and I guess 13not much more than a f eeling, that when all the
,_()
14probabilistic assessments are made and the numbers are there 151a ying in f ront of you, you might come to the conclusion 16that the systems as they are are pretty good, maybe a little 17 bit of specific fix-ups, particularly if auxiliary feedwater 18 systems are fixed up in existing plants, but the numbers 191ook pretty good.
But there is still going to be a strong 200pini^n or strong something coming from some people that an 21ind epend en t and dedicated system is desirable.
22 I look at, I guess, the evolution of shutdown 23 s ys te m s.
We seem to take it as an article of f aith tha t a i
m l
kj 24 scram system should be independent and dedicated to that one i
25 single function and not have anything else hanging on it. I O
%)
ALDERSON REPORTING COMPANY,INC, 400 VIRGINTA AVE., S.W., WASHINGTON, D C. 20024 (202) 554-2345
31 1
O
(_,1 1am not sure that probabilistic assessments have ever shown 2that tha t is necessary from a cost-beneficial standpoint.
U]
3Maybe they have, I do not know.
Maybe you can tell me
/
4whether they have or not.
5 But the safe management of a plant in an emergency 6or abnormal situation seenc to involve a two-headed attack:
7ene is to scram it and the other is to remove the decay heat 8after you have scrammed it.
For some reason we have never 9taken it as an article of faith that this second half of 10these two necessary things has to be independent and 11 dedicated as the scram system does.
12 Have you got anything to say about that, why there 13seems to be a difference in the attitudes there?
O 14 MR. MARCHESE:
I do not have anything to say about 15 t h a t.
I think the subtask III.5, thbconventional 16 engin eering approach, would hopefully answer, I think, what 17you are raising, plus we are going to be addressing things 18you can quantify f rom a probabilistic sense.
We are also 19 going to be looking at the Sandia work.
20 Go ahead, Dennis.
21 MR. BARRY:
Dennis Barry, Sandia.
22 With regard to your comment about the two 23f unctions -of scramming and decay heat removal, in our Sandia O) 24 program we had originally focused and still are focusing
\\_
25 mos t of our attention on decay heat removal following O)
\\~
f ALDERSoN REPoRTIM COMPANY,INC, 400 VIRGINIA AVE' J.W., WASHINGTON, D.C. 20024 (202) 554-2345
32
()
1 scram.
However, during our risk assessments and benefit 2 analyses related to alternative systems, we find that it may 3 be necessary for us to look at the reliability of scram 4 system in conjunction with the decay heat removal system, Sand whether or not an alternative decay heat removal system 6aar in fact need to have some scram-rela ted capability.
7 For instance, boron injection could help handle 8certain scenarios where the scram system itself may have 9 failed, so we are looking at, for instance, an alternative 10 sys te m that would have high pressure injection capability to 11 handle small LOCAs but at the same time could be injecting, 12 sa y, boron to gain you something from the standpoint of an 13 ATW S-an ticipa ted t ra nsien t without scram condition.
O 14 We have not done the full analysis and found out 15whether we can gain anything, but in f act we are looking at s
16both sides of the picture of the scram-related systems and 17 the non-scram-rela ted systems.
18 MR. WARDS Frank, are you ceady?
19 MR. ROWSOME:
Yes.
20 MR. WARD:
Maybe we would interrupt you for the 21 moment and come back to you later -
particularly I think we 22would like to hear your description of the Oconee system --
23 a nd let Frank rovsome give his presentation.
()
24 MR. ROWSOME:
My name is Frank Rowsome, from the 250f fice of Research, Division of Risk Analysis. I was invited O
ALDERSoN REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON. D.C. 20024 (202) 554 2345
s3
()
1here to talk briefly about what has come to be called the 2 paramedic criterion because I gather it was referred to in
}
3 the NRR program plan.
About a year or so ago, Bob Bernero 4and I published a recommended criterion for use in making a 5 decision on the urgency with which a detected safety 6 vulnerability should be dealt with in a plant.
It was based 7 exclusively on projected accident frequency: crudely, risk.
8 The objective we had in mind was the criterion 9that paid no attention to need for power, that paid no 10 attention to whether or not, for example, there was a 11 technical violation of the terms of the license in the 12 pla n t.
Both dealt merely wi th the discovery of a 13 vulnerabilit y for which a rough estimate of a probability of O
14 core melt could be generated.
15 It did not attempt to resolve severity of 16 c on se que nce s.
We know very well that some core melts can 17 produce a so urce term that is many, many orders of magnitude 18 higher than. other core melts, but the scenario we had in 19 mind was the discovery on relatively short notice of, 20vhoops, I just discovered a vulnerability in this plant.
21 And what do we do about it?
Do we shut the plant down?
Do 22we wait for a week, do we wait for a month, do we wait for a 23 yea r, do we wait for ten years?
On what urgency should we 24 deal with it?
25 So the scenario we had in mind was a quick l
ALDERSoN REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON. D.C. 20024 (202) 554-2345
34 f[ )
1 decision, should we shut it down or not, based on risk.
In 2 that context we did not think it made a lot of sense to
/"S 3 worry about trying to resolve consequence severity, is this
(/
4 going to be the kind of melt that will give me a release 5 category I, or is this the kind of melt that will give me a 6 release Category VI or something of that kind?
7 At the time we made this recommendation we were 8 seeing from Battelle and Sandia preliminary results from th e 9 Reactor Saf ety Study Methodology Applications Program, which 10had in it some very conservative treatments for hydrogen 11 burns threatening even large, d ry PWR containments, so that 12 r ou ghly one out of five of all core melt sequences could 13 b rea k the containment and produce a nasty release category.
14 So tha t we did not want to place a lot of faith in our 15 ability to judge from the nature of the vulnerability how 16 b ad the consequences would be.
17 So we wanted simply a probabilistic criterion of 18 j ud ging how long to go, and we came up with a number which I 191argely drew out of the air, but the re was a memorandum
{
20 attached to the transmittal that showed sore of the thinking 21 und erlying that number.
The number was one-tenth of one l
22 percent chance that the problem gets you between the time 23 you discover it and the time you deal with it.
You fix it.
s
(,)
24 You either shut the plant down or you have made an 25 improvement.
l ALDERSoN REPORTING COMPANY,INC, i
l 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
35 O)
(
1 Why that number?
Well as I say, it is largely 2 hand waving, but it seemed to us in doing sensitivity
'}
3 studies on the return intervals for core damage and core 4 melt accidents based on what we knew, we believed it would 5 be an improvement over what may be out there.
If there are 6 outliers, if there are other high vulnerability sequences, 7 unpleasant suprises lying in wait like TMI and the Browns 8 Ferry almost-ATWS and the Browns Ferry fire and Rancho Seco 911ght bulb, that one-tenth of a percent chance of getting 10 clobbered before you make a fix would be an improvement.
11 Second of all we thought it is about the best you 12can do; that if you set as your goal a much more stringent 13 criterion than that, you spand a lot of your time chasing 14 minor contributionc to the risk and some of the major ones 15 may get by you.
I do not believe that we can, for example, 16 survey the population of plants that are out there for 17 sus ce p tibility to unpleasant surprises like TMI or the 18 Rancho Seco light bulb with studies like IRAP in some time 19 tha t we could be sure of getting much less than a tenth of a i
l 20 percent chance of losing any one plant before we have 1
21 accomplished that certainly.
l 22 So I do not think you can do very much better; 23 that you run into diminishing returns in the sense that if 24you strive f or a much higher standard, you will sp(nd a 25 disproportionate fraction of your resources chasing minor j
l ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON. D.C. 20024 (202) 554-2345
.... - -. - -, -.-. ~
e 36 O
'contria tote ene ree ce=1e de causat with roet vente aown-2 So somewhere between the bound on one side of the best you 3 could reasonably do, and on the other hand the bound of what 4 would really constitute an improvement and deal adequately 5 with the problem, we came up with a tenth of a percent 6 chance of getting caught.
7 Now, you have to make a probability estimate when 8 you tind the vulnerability.
Very often this is a crude 9 hack-of-the-envelope exercise.
I know Professor Okrent has 10 been critical of our number, in part on the grounds that it 11 seems non-conservative and in part on the grounds that wt do 12 not have levels of confidence.
We have not told prospective 13 users of our criterion whether they should be doing a best O
14 estimate or 95 percent confidence level or something of that 15 kin d.
16 The scenario we had in our mind was that we would 17 have a very crude estimate when you first discover the 18 v uln er ability, that you will not have refined your analysis, 19 tha t you have to make a decision about whether to shut this 20 plant down or to give the licensee a month in which to fix 21 1t or a year in which to fix it.
You have to make that 22 decision much quicker than you could take the time to refine 23 a statistical analysis, but enough to achieve a distinction 24 between a best estimate and a 95 percent confidence estima te.
25 It is not intended that this criteria would serve O
ALDERSON REPORTING COMPANY. INC.
400 VIRGINtA AVE., S.W., WASHINGTON. D.C. 20024 (202) 554 2345
37
()
1 as an ideal acceptable risk criterion.
It is not intended 2 that this would be the kind of criterion you would want to
'N 3 use in the context of a degraded core rule, for example, or k'/
4 that this would be generated by the rulemaking on safety 5 goals.
This is not a goal of what we would like to see.
6 This is a goal of what we can live with in the short term 7 when you are f aced with saf ety deficiencies in plants 8 sta ring you in the face.
9 You just discover through a precursor event or 10 something of that kind that saf ety grade piping is not 11 seismically qualified, that there is a hydraulic design 12 d eficiency in the scram discharge volume or here is an 13 auxiliary system control bus whose f ailure can melt the O
1/. c o r e.
How long do you give the licensee to deal with it?
15It is that kind of quick panic, got to make a judgment now, 169ot to determine whether this is an immedia te shutdown order 17 or shutdown order in a month or deal with it in a year or l
18 deal with it in ten years.
It is that kind of context that 19 ve proposed our one-tenth of a percent chance that the 20 acciden t would happen to you before you get it fixed.
21 MR. WARD:
Frank, you say a tenth of a percent 22 chance you do not think would be appropriate for the 1
l 23 degraded core rule aking.
Is that because you do not like i
24 th e tenth of a percent as a criterion or because you would 2511ke tha t to be a 95 percent confidence number or something (T)
ALDERSoN REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON. D C. 20024 (202) 554-2345 l
38
()
liik e that?
2 MR. ROWSOME:
Well, the whole concept was how big
/~
3a chance would you run before you make a fix?
If you are V}
4 dealing with the lifetime of a plant, you may discover a Snumber of unpleasant surprises, more than one.
You may want 6to make a quick fix and then go for a while and then make a 7 permanent fix f or a problem, first install control grade 8 actuation system, later install better.
That kind of phased 9 thing.
10 If you are talking about what society can live 11with, what the agency's goals are, if you are talking abcut 12 v ha t we would set as an objective for a plant that is just 13on the drawing boards today, no, I don't think that would be O
14a meaningf ul approach.
The whole thing was predicated on a 15 quick decision when you just have discovered something that 16has alarmed.
17 dB. EPLER:
As you described this, Frank, I was l
18ha ving trouble reconciling in my mind the stated objective j
-6
-5 l
19 of 10 for core melt, the safety study of 5 x 10
-3 20with your number of 10 which then caused me to think.
l l
21This discovery that you have made of a very bad situation 22does not have a probability itself of 1 but it has another 23 p ro b abil ty, and that you did not address, and it could be 24almost anything, such that I think not only have you not 25 addressed the consequences, but you have not addressed the i
i ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
_,. -. - _, _, _ _. _ _ ~ _. _ _ _ _ _ _ _ -. _..
39
()
1 probability of this thing, which leaves me some questions.
2 MR. RDWSOME:
We know that discoveries are P.ade on
()N 3the order of six months to one years, centers of safety v
4 pro blems.
For example, I cited several the scram discharge 5 volume deficiencies revealed by the incident at Browns 6 Ferry, the discovery that a number of PWRs had auxiliary 7feedwater systems that were not seismically qualified, the 8 discovery of the errors in the seismic qualification code 9 tha t led to the shutdown of five Stone and Webster plants 10some months before TMI.
11 Every six months or a year or so e concern rears 12its ugly head at NRR.
Here is a safety issue, here are a 13 number of plants which we have just discovered today have a 14saf ety deficiency that we did not know about yesterday.
Do 15we shut them down?
Do we take a month to deal with it?
Do 16ve take a year to deal with it?
Do we take ten years?
It 171s a quick rule of thumb for getting an order of magnitude 18 sense of urgency about the problem?
19 MR. EPLER:
I would like to focus just a moment 20 longer on the scram discharge problem.
G.E. has responded
-9 21 tha t the probability of a piping failure would be 10 per 22 reactor year.
That ain't gonna happen, is it?
23 MR. ROWSOME:
No.
()
24 MR. EPLER4 On the other hand, the probability of 25somebody doing maintenance on a valve at the time of scram l
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
40 I
()
1has an entirely different probability. I am not sure I l
2 believe either one of these probabilities.
I do not think
/"]
3 we have looked at the problem long enough to make a judgment.
V 4
MB. ROW 50ME:
Yes.
In terms of interpreting what 5 the expected frequency of core melt should be given that you 6 discovered a safety deficiency, the problem that we have 7vith Professor Okrent about levels of confidence, you are 8not going to have a great deal of faith in the first 9probabilistic estimate you come up with.
And in practically 10 any real world example, you can find people who say, given 11 tha t somebody has done a back-of-the-envelope calculation 12and somebody has come up with a number, you can probably 13 come up with some people that say that it is conservative by 14 two orders of magnitude and somebody else who says it is 15nonconservative by two orders of magnitude.
16 The uncertainties associated with the estimate are 17 going to be larger than the difference between confidence 181evels based on statistics on a data bank.
That is almost 19 ce r tain.
20 MR. W1.RDs Any other questions?
21 (No response.)
22 MR. ROWSOME4 As long as I have the floor, I was 23 abo u t to comment on the prior work if you are done talking 24 a b o u t --
25 h..
WARD:
Let me make one comment.
I guess this l
OV ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WAL. 'NGTON, D.C. 20024 (202) 554-2345 I
_. _,, _ _ _ ~., -. _... _..
41
()
lis a question to NRR, but from your description of this 2 system, of the method of evaluation, it does not seem at all
/~s 3 suitable for the evaluation that you want to make under Task 4 Action 45.
At least that is my reaction to it.
5 Andy, do you have a comment on that?
Are you 6 planning on using this approach?
7 MR. MARCHESEs We were just noting those written 8 documents that exist.
I think we referenced three documents 9 that are out where an attempt has been made to publish 10 quantitative action or probabilistic goals.
We are not ti endersing one over the other; we are just saying these three 12 documents exist.
It is part of performing subtask 1.
The 13 s t arting point would be the review and consideration of n
14 these three documents, but I cannot say any more than that.
15 We are going to be using one over the other, I 16 think.
From what I have seen, the ACBS proposal is the most 17 com prehensive and that probably would be the basis or the 18 st ar tin g point for us.
19 MR. WARD.
Frank, from what you said, I guess your 20 recommendation to NRR would be not to use the tenth of a 21 percent criterion in the Task Action Plan.
Is that right?
22 MR. ROWSOME:
If they find an outlier, if ther l-I 23 fin d -- well, this is a complex situation, but if in the
()
24 course of reviewing plants they discover one that is really 251n bad shape, the question is do you shut it down or not?
(3 N) l ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
42
()
1 Then I would say maybe so, yes, then use the tenth of a 2 percent criteria.
If, on the other hand, you are looking 3 for a permanent fix, one that you can live with for the rest 4of the life of the plant, if you are looking for a standard 5 that is intended to be acceptable for all time, it is not a 8 band-aid fix.
Then I would look for something a little more 7 conservative than the tenth of a percent chance approach and 8 do a proper value impact analysis.
9 MR. WARD:
So you are saying only if the 10 investigations as part of the Task Ac' ion Plan turn up a nev 11 six to twelve-month vulnerability issue, only in that 12 instance would your tenth of a percent be a useful techniquc.
13 MR. RDWSOME:
Yes.
O 14 MR. WARD:
Anything else on this particular point 15before -- wh y don ' t you go ahead with your other comment.
16 MR. ROWSOME:
The first I saw of the A-45 program 17 plan was about ten days ago when a copy was sent to me by l
18 Carl and Andy, and I was kind of disturbed when I saw it.
I I
19 talked to Bob Bernero, have not yet gotten back to NBR on l
20it.
They seem to be pretty well a ware of Matt Taylor's 21 program to look at alternate decay heat removal system 22 d es ign s, but it seems to have been created in increments of 23 the work we were doing to set up a research program to deal 24 with the degrade core cooling rulemaking, which incorporates 25 virtually everything that is in that program plan in the tQ U
ALDERSoN REPORTING COMPANY,INC, 400 VIRGINfA AVE., S.W., WASHINGTON. D C. 20024 (202) 554 2345
43
()
11arger context of looking at mitigation as well as 2 prevention systems 3
I do not want to cast any blame here, either at 4 Andy or Carl, because I think we in Research have as big a 5 piece of that for that share of the blame for not being 6 clearer in what our scope of work for the degraded core 7 rulemaking has been.
What we are seeing here is a spin-off S of a case of grotesque ma1 communication, miscommunication 9 within the agency and within the office, and which has been 10 com plica ted by seas territorial squabbles between divisions 11 within the Office of Research as to who would manage the 12 decay heat -- the degraded core cooling research.
13 But in fact almost everything that is in this 14 program plan has already been contracted for out of the 15 0ffice of Research, and I would tuggest that NRR move toward 16 reconstituting their program into one of liaison and 17 t ra ckin g for office policy and the like on what we already 18 h ave going.
It is unfortunate that it go this far before we 19 discovered the parallelism, but those things happen.
I l
20 think the blame is as much ours as it is theirs.
I 21 HR. WARD:
Do you have any response to that?
22 MR. MARCHESE:
It turns out I have been working 23 very closely with people in NRR who are involved with (Q,/
l 24 degraded core rulemaking, namely, Jim Myer.
We almost share 25an office together.
And Jim Myer and Bernero had this plan 1
ALDERSoN REPORTING CCMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
?
44
()
1from day one, and this is the first time that I have heard in our plan has already been accounted for 2 that everything This is the first 31n the plan on degraded core rulemaking.
4 time I have heard that.
5 So we will have to, yo1 know, confirm that to see 61f that is in fact the case.
I kind of doubt it, but I do 7 not want to get into an argument at this point.
We 8certainly do not intend to duplicate, but I would be very 9 surprised if all of this work is going on in the systematic 10 basis that we have outlined.
We are not talking, by the 11 way, a generic evaluation.
We are talking about assessing out that 12 quite a number of plants because we are finding 13 there is quite a variability from plant to plant.
It is one O
14 thing to do this on a generic basis; it is another thing to 15 do it on a plant-to-plant basis.
That is basically what we 16 an e talking about.
that a great deal can 17 MR. ROWSOME:
I am very sure 18be done to make a profitable merger of what you have in your 19 plan and what we have in our plan, and I look forward to 20 working with you to sort that out.
The extent to which we 211n the degraded core cooling rulemaking will get into 22 pla n t-to-pl a nt research is open-ended and will be delineated 23 by the preliminary results of the research itself.
We have 24 nct pre-planned that, but you may well in f act be intending 25 to do more on a plant-to-plant basis.
O ALDERSON REPORTING COMPANY,INC, j
400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345 gi i
45
()
1 On 'the other hand, I do not think you can make a 2 decision about whether to add on a prevention system, core 3 ael t prevention system without also dealing with the 4 question of what you are going to do in mitigation.
One 5really needs a wholistic view of plant safety to evaluate 6what are the most cost-ef f ec tive, sa f e ty-ef f ective, with due 7 def erence to PR-effective improvements in the safety of the 8 plant.
9 MR. WARD:
Frank, at a meeting of the ACRS Class 9 10 Accident Subcommittee, we asked Mark Cunningham of Research 11 -- is Mark in your organization?
12 MR. ROWSOME:
Thu. is right.
13 MR. WARDS We asked there about the relationship O
14between the programs in Research and the Ta sk Action Plan, 15and we were told that we would hear about that.
Is your 16 remark the answer to that?
l l
17 MR. ROWSOMEs The first time in a public forum we l
18 talked extensively about what we wanted to do was for the 19 ACRS in a budget context about a month ago when we had a 20 presentation both to Okrent's subcommittee on our division, 21 s ub-unit type decision units, and the following day to the 22 Class 9 Subcommittee, in which we described the degraded 23 core rulemaking research, the systems-oriented research we (D
(_)
24 were doing, which entailed looking at quite a wide variety 25of add-on mitigation systems and add-on decay heat removal I
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON D.C. 20024 (202) 554 2345
46
()
1 systems and other prevention systems tailored to classes of 2 accident sequences and to do value impact analysis on the
/'%
3 d esirability, using in part risk assessment and in part 4 sensitivity studies, the return intervals for the kind of 5 accident scenarios that we cannot quantify very well, like 6 sabotage, to look at the trade-offs between costs and 7 benefits in adding on a decay heat removal system and so 8forth.
9 MR. WARD:
Do you have any further questions of 10 M r.
Rowsome?
11 (No response.)
12 Thank you, Frank.
13 I think we will take about a ten-minute break and O
14 then we will come back and ask Andy to tell us about the 150conee system and then see where we go from there.
16 Thank you.
17 (Recess.)
18 MR. WARD:
We will reconvene, and I think we will 19sta rt by asking Mr. Marchese to describe the Oconee system 20he indicated was interesting.
21 MR. MARCHESE:
Thank you.
22 I do not have any Vu-graphs or pictures of the 230conee system.
I tried to get some but there just was not
_ )
24 eno ugh time available for the Duke Power Company people to 25 respond to get me what I needed.
But let m. just briefly ALDERSON REPORT'NG COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
47
()
1 describe wha t I heard and what I saw.
2 We were there about ten days ago to talk with the 3 Duke Power Company in terms of getting a description of the 4 standby shutdown facility, what they refer to it as, in S terms of actually what is its design criteria, what is its 6 intended functions to perform, and also get a feel of the 7 hardware that is involved.
8 The Oconee system has its own separate building 9vhich is a seismic Category I building, reinforced 10 con cre te.
The walls are approximately three feet thick.
11The dimensions are about, I would say, 90 feet long by 12 perhaps 60 f eet wide.
It has its own separate diesel s
13 generator power supply.
It is about 3-1/2 megawatts, a very 141arge diesel, so it has its own power and control.
15 The system makes up both the primary and the 16 secondary side.
They are providing makeup, motor driven 17 mak e up pumps to the primary side, and also a motor-operated 18 pum p to the secondary side that can make up coolant to all 19three units, qteam generators, all those three steam 20 generators that are there, by the waye 21 MR. WARD:
Is this a single building serving all 22three reactors ?
23 MR. MARCHESE:
Yes, exactly.
Now, the components
()
24 and systems as f ar as the piping, pumps, valves are safety 25 class in that they are designed to ASME Section 3, Class 2, O
ALCERSON REPORTING COMPANY, INC, 400 VIRGINIA AVE, S.W, WASHINGTON, D.C. 20024 (202) 554-2345
48
()
land ASEE Section 3, Class 3 for the makeup to the secondary 2 side.
So it is a complete safety graded system, Seismic 3 Category I.
It also is environmentally qualified for 4 conditions in which it operates.
5 They started construction approximately 18 months 6ago when they received preliminary approval from the NRC, so 7they are 18 months into construction and they estimate that 8they have about another 18 months before they complete the 9 facility. ' A7Darently most of that time in the next 18 10aonths probably is going to be due to tying into the primary 11 system, tha t they are estimating about a 15-week period to 12 tie in to th-primary system, and they are only going to do 13 tha t when they shut down the unit for refueling or for some O
14 other outage.
15 I requested that they furnish me with a cost 16 breakdown of the structure and the major components, a 17 schedule and this sort of thing because I think it provides 18us with a poof. opportunity to get some pretty good numbers, 19 cost numbers on wha t an actual system is -- a structure l
20vould actually cost; but I sort of got the feeling that this l
21was kind of an of f-the-cuff estima te.
l 22 The costs right now are running about $34 23 million.
It probably, as I mentioned oefore, represents the 24 closest thing to a dedicated system that probably is out 25 there in existence.
A Y
s l
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON. D.C. 20024 (202) 554 234S
49
()
1 Now, the spstem is only designed to achieve and 2 maintain hot shutdown.
Okay, it does not get you to cold 3 shutdown.
They are designed tha t they can maintain hot 4 shutdown for 3-1/2 days, and in that time they will have Svhat they call controlled damage procedures, that is, 6 procedures and spares such that they could repair any 7 equipment that has been damaged, say, due to a fire, that 8they can repair that equipment within three days.
9 In terms of the design criteria -- I should have they were trying a 10 started with this at the beginning 11 year and a half ago or, no, more than that, probably several 12 yea rs ago, they instituted a task force at Duke Power 13 Com pany in which they wanted to take care of a lot of 14 problems tha t they saw coming in the future in terms of 15bei ng able to meet new requirements on fire protection, 16 sabotage, missile protection, floods.
So they were trying 17to design a facility to take care of a lot of different 18 problems and not just, for example, fire, to meet the new 19 fire protection requirements.
So I think it is a system l
20 t ha t, you know, takes care of a lot of things.
21 Now, comparing it to McGuire, which they also 22 built a standby shutdown f acility, that facility was just 23mainly designed to meet the new fire protection O)
(_
24 req uirem en ts, Appendix R, and it is a nonseismic Ca -
'y I 25 building, non-safety grade.
It provides a small diese.
s ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE S.W., WASHINGTON, D C. 20024 (202) 554 2345
50
()
1 generator of about 700 kilowatts and in which power is a 2 makeup pump to the primary side which is inside the and also provides power to an existing 3 containment 4 turbine-drive auxiliary feedwater pump.
It has its own shutdown panel inside this S
In 6 building.
It is a lot smaller than the Oconee plant.
7 terms of costs, they were talking roughly of about $3 so you can see there is like an order of magnitude
- 8million, 9 difference in terms of going to seismic safety grade and 10providing pumps to make up to both the primary and the 11 secondary systems.
12 MR. WARD:
The $34 million system serves three 13 r e a ctor s.
What is a t McGuire?
MR. MARCHESE:
Two units there.
14 steam generator, the 15 MR. ETHERINGTON:
Is it a 16 second ar y, or is it water --
MR. MAhCHESE4 The Oconee reactors are BCW units, 17 18 steam generators.
19 MR. E1hERINGTON:
I as talking about these new 20eme rgency -- these new shutdown systems.
What is the 7.1 secondary system?
dR. MARCHESE:
They make up both the primary side 22 23 and the secondary through the steam generators.
MR. ETHERINGTON:
Through the eristing steam
()
24 25 g en era to rs?
O I
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON. O C. 20024 (202) 554 2345
51
()
1 MR. MARCHESEs Yes.
In the Oconee system they 2 take, for the stesm generators, they take makeup from an 3 existing condenser cooling water pipe, which apparently runs 4 tc Lake Kiwi, I think it is, nearby, takes suction from that Slake that will inject in the steam generators as a 61ast-ditch effort.
7 dR. ETHERINGTON:
What is in the new tuilding, 8 then ?
9 MR. MARCHESEs In the Oconee, building, okay, you 10ha're a 3-1/2 megawatt diesel, you have a motor-driven pump 11 to make up to the secondary side, you have, of course, all 12 the controls, cables.
There is also going to be a security 13 station there.
In fset, we have someone here who can 14 address some of these sabotage-related things.
There is 15also a separate room for that.
16 There is a fuel oil suoply for the diesel.
I 17fe get how many days f uel oil supply they have.
So you have 18the pumps piping valves, of course, diesel generator.
- Gene, 19 a ny thing else?
20 MR. KNIEL:
I thought there was a high pressure 21 in j ectio n.
I do not think you mentioned that.
22 MR. MARCHESE:
Yes.
You are making up to the 23 p rima ry side, but the makeup pump is not in this facility.
O(_)
24 It is going to be inside containment.
25 MR. KNIEL:
It is powered by the diesel generator.
O ALDERSON REPORTING CCMPANY,INC, 400 VIRGINIA AVE., S.W., WASdlNGTON. D.C. 20024 (202) 554 2345
52
(}
1 MR. MARCHESEs It is powered by the diesel 2 genera to rs, right.
3 MR. WARDS Is that dedicated high pressure r3 4 injection pump or is this just another source of power to 5the system?
6 MR. MARCHESEs It is on an existing pump.
It is 7not like -- boy, they are actually adding a new pump for 8 both the primary and the secondary sides, but the pump for 9 the primary side will be inside existing containment.
10 At McGuire they added a new pump to make up to the 11 primary side, but for the secondary side they are just using 12 tha turbina-driven -- existing turbine-driven pump and 13 providing a separate power source.
By the way, I thought it 14 was interesting f rom a close comparison standpoint, in terms 15of, you know, an order of magnitude difference.
You raised 16 the point about, you know, we are supplying three units 17 varsus two, but they kept emphasizing that they are going to 18 a seismic Category I structure to make it completely safety 19 3rade, environmen tally qualified.
It adds considerably to 20 the cost.
21 MR. WARDS Frank -- Andy, is this a system that 22 Duke started the planning and development before Three Mile 23 Island ?
()
24 MR. MARCHESEs Yes.
25 MR. WARDS Primarily to meet the fire protection.
O ALDERSON REPORTING COMPANY, INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554 2345
53
()
1 MR. MARCHESEs Fire protection, sabotage, and also 2 they had a problem with flooding the turbine building in f^T 3 which if they flooded the turbine building, apparently it U
4could overflow to the auxiliary building where there is 5 safety grade equipment e.nd knock out safety grade 6 equipmen t.
So apparently -- I got the impression that in 7 terms of making it safety grade, that comes from the turbine 8 building flooding problem.
9 The fire protection rule specifically calls out it 10 does not have to be a saf ety grade system, or it does not 11 have to be a seismic Category I from the standpoint of just 12 fire protection.
13 Those are the highlights.
I guess I would 14 recommend -- in f act, I invited the Duke people up here but 15 they were reluctant on such short notice to come up and talk 16 bef ore the subcommittee.
I would recommend that you can do 17 either of two things: go down there and let them give you a 18 briefin g and a tour, or ha ve them come up here and give you 19 a presen tation.
I think it would be interesting.
I would 20 recommend, though, at some point in time you actually walk 21 through the facility.
It gives you a pretty good feel for 22 wha t we are talking about here.
23 It might be a little early because there was a lot
()
24 of equipment they have not installed yet.
Perhaps in 6 25 m on t hs, 12 months it might be a better time to go down O
ALDERSoN REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2346
54
()
i there.
You know, they have a lot of construction people 2 down there working on it.
3 MR. WARDS As far as you know, this is rather 4 unique in the country right now.
5 MR. MARCHESE:
Eight.
I think it is.
6 MR. EPLER4 A question.
I get the impression that 71n addressing the issue of the problem we may have doubled 8 the probability of overpressurization by this additional 9 HPSI.
Is that t e case?
10 MR. TAYLOR:
I believe I mentioned May 5 about the 11 concerns that one has about any of these add-on systems in 12 eff ect being contradictory to safety, and I think if one is, 13 for example, manually actuated, I do not believe you 14 necessarily have tha t risk of chilling the vessel.
If it is 15 au toma tic, then you also may have that risk.
So I think 16that is one of the trade-of f s.
I believe that is what you 17 are referring to that we are considering.
18 MR. RAY Andy, have you any disposition to 19 consider the Oconee installation in the light of the 20 requirements of Task A-45 as fulfilling the requirements?
21 MR. MARCHESE:
Yes.
I think a couple of things.
220ne of the things which I have not checked on to see, if the 23 risk assessment that is coming out on Oconee includes this
/~,()
24 new standby shutdown facility.
I got the feeling that the 25 RSM AP report on Oconee does not include consideration of the l
l l
ALDERSoN REPORTING COMPANY,INC, 400 VIRGtNIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554 2345
55
()
1 shutdown facility but that EPRI has a program that will 2 include the risk reduction associa ted wi th this f acility.
/~'
3 MR. TAYLORs I cannot answer that specifically.
I C;
4 believe the RSM AP plant or risk assessment did not 5specifically include this new add-on.
6 MR. MARCHESE:
Because I asked them that 7 q ue stion.
A pparently Duke has a program with EPRI in which 8 they are going to assess the risk reduction of t.".s standby 9 facility.
10 MR. TAYLOR:
The Oconee units were dominated by 11 such things as the small LOCA and the core melt probability 12 and the interf acing system LOCA, so this system would have 13 no impact relative to those dominant sequences, or minimal 1
14 im pact.
15 MR. DAVIS:
Andy 16 MR. WARDa What is the justification f or putting 171t in?
18 MR. MARCHESE4 To meet fire protection, the new 19 tule on sabotage protection, to solve the turbine buildirc 20 floodin g problem, and I guess missile protection also is 21part of the design.
22 MR. WARD:
I am probably not asking the right guy, 23 but why doesn't that show up as an improvement in the RSMAP?
()
24 MR. TAYl0R:
Well first off, it was not considered 25 -- when RSHAP was done the design was not installed on the
(
ALDERSON REPORTING COM?ANY,INC.
400 VIRGINIA AVE., S.W, WASHINGTON. D.C. 20024 (202) 554 2345
56 l
()
10conee plan t.
As RSMAP, the risk study was completed, yes, 2 then the design came into being.
We did not go back and
/~T 3 reevaluate that.
However, I have looked at the results of 4 0conee and I am f amiliar with what did dominnte the risk on 5 0conee, and from a risk reduction standpoint, there would be Osma11 impa:t by this add-on.
7 Now, this add-on does not have small LOCA 8 capability, for example.
It has a makeup, a limited leakage 9 makeup capability. In other words, in a tight primary system 10in cold shutdown conditions -- not into cold shutdown but in 11 hot shutdown, so that would have limited benefit given the 12 sequences that did dominate core melt.
13 MR. WARD In other words, if you were running 14 Duke you would have spent the $34 million on something else.
15 MR. TAYLOR:
Not necessarily.
I cannot speak for 16 Duke, but f or myself I believe there is a need to look at 17all of these requirements, sabotage, fire, those things that 18you can quantif y like the risk from small LOCA, the risk 19 f rom AC blackout and all the other risks, and pull this 20 together in an integra ted package.
21 If one is going to put an add-on system on, they 22 Wan t to do it for as broad a reason, as broad-based a reason i
i 23 to get the biggest return from that investment rather than j
(m_)
24 just one piece of a saf ety problem like fire and/or sabotage l
i 25 and/or flooding.
I think all of those reasons, small LOCA, OO l
ALDERSON REPORTING COMPANY,INC, 4C0 VIRGIN 1A AVE., S.W., WASHINGTON. D.C. 20024 (202) 554-2345
=--
. = _.
57
()
1;C blackout, all of those reasons should be pulled 2together.
This is what the program we have kicked off some 3 yea rs ago has been trying to do, is to pull together those 4 things that we can quantify and those things that we cannot Squantify to pr0 vide this integrated base.
6 MR. RAY.
Is the fundamental point here, then, 7 tha t the pumping capacity that is provided in this facility 81s not large enough to take care of the LOCAs?
Is that the 9whole point here?
10 ER. MARCHESEs Ma tt is right. On the primary side 111t is designed or 26 gpm, just for leakage.
12 ER. RAY:
It is too small.
13 ER. WARD:
Okay.
Well, leaving the sabject of he 140conee, I think Mr. Davis has some more general questions.
15 HR. DAVIS:
Yes. I had three quick questions on 16the Task Action Plan 45 plan.
On page 12 you indicate that 17 the criteria for core melt will be more than 30 percent of i
l 18the oxide f uel becoming molten.
I doubt very much if our 19 analysis is sophistica ted enough to determine f or a given 20 accident sequence whether 29 percent or 31 percent of the 21 core is going to melt.
Can you tell me where that came from 22 a nd exactly how you were going to use it?
It is the first i
l 23 fu11 paragraph.
l
/~
l k
24 MR. HARCHESE:
That basically came out of the ACRS 25 pro p os al.
ALDERSON REPORTING COMPANY,INC, 400 VIRGIMA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
58
()
1 MR. DAVIS:
Which proposal?
2 MR. MARCHESE:
The NUREG on quantitive action fS 3 crite ria.
That is a direct sl 4
MR. CAVE:
That is a quote from 0739.
5 MR. MARCHESE:
Large-scale fuel melt 61s defined in that document as more than 30 percent of the 7 oxide fuel becoming molten.
8 MR. DAVIS:
Okay.
On page 16 you defined existing 9and future plants.
There seems to be a gap in that 10 definition.
You define existing plants as those which are 11 already operating or are under construction and have reached 12a stage where the majority of the plant items are in place.
13 Future plants have not yet started construc tion.
There 14seems to be a group in which construction has started but 15 they have not reached the stage where the majority of the 161tems are in place.
Maybe that is a fine point, but it 17seems like there may be a few plants in that category.
18 Where would you put those?
19 MR. MARCHESE Future, I would believe.
Future.
20I think you are right.
I think there might be a gap in 21 t h e re.
They might f all into the "f uture" category.
22 MR. DAVIS:
All right, last question, page 41.
l l
l 23You list some problem areas that may compromise your ability
()
24 to achie ve the coals of Task Action Plan A-45.
Item D there 25 confused me a little bit, obtaining necesary operating (h
s_s I
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON. D.C. 20024 (202) 554 2345
59
()
1 experience on shutdown cooling systems.
2 Doesn't that information exist?
Why is that a 7-3 problem arsa at this point?
V}
4 MR. MARCHESE:
Well, we are hoping and intending Sto get as much help f rom licensees a s we can on this 6 program, particularly in subtask III.5 where we are doing 7more conventional engineering assessments, which would 81nvolve soliciting relevant PNIV documents, full schematics, 9 interface drawings on how the systems tie in to one another, 10and I guess the point I am making here, it is going to take 11 cooperation from the licensee to make sure they furnish to 12cs on a regular basis those drawings and documents that we 13 n e e d.
I think you are raising the question maybe we already
(~)%
(
14 have them.
That could be true, but I am not sure.
We may l
l 15need some initial information.
16 HR. DAVIS:
This just says operating experience, 17and it is my belief that they are required through the LER 18 system to report any malfunction of this equipment, but if 19you include also design information, then I can understand l
20why there may be a difficulty.
21 Ihat is all I had, Mr. Chairman.
Thank you.
22 MR. ETHERINGTONs I Lid one question.
Another 231ss u e.
We tried to clean up the nomenclature on RHR and so l ()
24 f or t h.
I think the different terms came about because one l
25manuf acturer refused to copy the nomenclature of another 1
m ALDERSON REPORTING COMPANY,INC, 40') VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
60
()
1rather than any basic difference.
Is this a unilateral 2 attempt on your part or do you have a committee somewhere on 3
3 nomenclature?
Are you making matters worse by choosing your 4own terminology?
5 MR. MARCHESEs I think it is probably unilateral.
6We felt the need to really right at the beginning set down 7some definitions so we know wha t we are talking about.
8 HR. ETHERINGTON:
The difference between decay 9 heat, residual heat in my mind is extremely vague.
Sensible 10hea t and decay heat, that I understand.
I guess we will 11 toddle along with confusion.
12 MR. WARD:
I want to ask a question on page 39.
13You say you are going to attempt to ask the contractor to 14 a tt empt to sort all the pisats in the groups based on the 15six plants that have been part of the RSMAP program, and I 16do not know how those six had been selected, but are ther 17 selected to be representative of six s.vu,;'
I mean is it MR. TAYLOR 4 The selection of those six plants was 19a joint operation of Research and the regulatory people some 20few years ago.
The purpose was, number one, on differences 211n vendors, number two, on differences in containment design 22because containment is, let's say, one of the safety 23 f eatures that tends to be most important in risk.
In other
(
24 Wo rds, the different behavior of the core meltdown sequences 25would be a little different in a MARK I as opposed to an ice N
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON. D.C. 20024 (202) 554 2345
-, _ - _ - _... - _ - ~ _ - _. _.
61
()
1 condenser as opposed to a large,' dry containment.
2 So it was selected based on a spectrum of designs 1
/"
3to look at design variations, not necessarily decay heat V}
4 removal variations.
5 MR. WARD:
So it mTy or may not fit the decay heat 6 removal variations.
7 MR. TAYLORa That is entirely corr
-t.
Now, I do 8 think some of the best indication, at least for the PWRs, 911es with NUREG-0611 and 0635, where we did look at the 10 existing population of Westinghouse and CE plants to find 11 out what their artillery feedwater systems were.
So I think 12that is some indication of the variability that one might 13 find an important part of the decay heat removal system.
14 Now, there is anoti.*r important part, which is the 15high pressure injection and/or: the chemical volume and 16 con trol system. That variation is not known as well as the 17 variation in aux feed right now.
18 MR. WARD:
Let's see.
19 MR. MARCHESEs Let me just add a little.
-20 MR. WARD:
Sure.
21 HR. MARCHESE:
We felt we need to attempt to try 1
22 and group the plants because if we cannot group the plants 231nto some manageable number, a half-dozen or a dozen, then
()
24this task is probably going to grow by leaps and bounds into 25having to do what we are talking about for 70 or so plants.
l\\
l ks) i l
l ALDERSON REPORTING COMPANY, INC, I
l 400 VIRGINIA AVE., S.W., WASHINGTON. D.C. 20024 (202) 554-234E
62
()
1 Basically the way we are going to approac:1 that is 2 to use the six plants that are going to be assessed in RSMAP 3 and W ASH-14 00, and we will have full-blown risk assessments 4 coming out as well as some additional plants -- Zion, Indian 5 Point and Limerick are also in our risk assessment -- and 6then compare all the other plants in terms of looking at the 7 design characteristics that are used for decay heat removal 8and see how they stack up against these plants that will 9have had that risk assessment.
That is one way of doing it 10 and we are going to attempt to move on that basis.
11 MR. WARD:
Do we have any other questions from the 12 committee 7r the consultants?
Comments or questions?
13 (No response.)
14 Well, you have gotten some perhaps rather random 15 comments on the Action Plan from us It appears you have a 1611ttle homework to do in addition to resolving yours with 17 the Research Office program, and I think we would like to 18 hear where you co9e down.
19 What is your plan as far as -- this is a draft.
I 20 guess it tells me in here what your schedule is for issuing 21 the final Action Plan.
Could you summarize that?
22 MR. MARCHESE.
Well, we were planning within the 23 n ex t month to finslize the plan in terms of considering both
()
24 the comments we heard today and I think probably this week 25 we are going to be getting some internal comments from staff ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
63
()
1 people and incorporate them into the plan, but I think today 2I heard something that was new to me in terms of there may 3 be a major overlap between programs in Research and NBR, 4which va are going to have to get straigh tened, obviously.
5 Whether or not that changes anything, including the plan on 6 sch edule, I just cannot say righ t now.
I have to pursue 7 this and see really we will get together with Research 8and try and iron things out.
9
?ut I an hoping, you know, to get this plan out in 10 approximately a month if I can.
11 MR. WARD 4 Well, at the present time, let's see, 12 we had, I guess, Mr. Savio discuss with you the possibility 13 of a presentation before the full committee later this week, 14 and I do not think that would be appropriate right now.
We 15 will discuss it with the full committee and perhapz get some 161np u t there, and we can maybe organize what we have told you 17 today and add to that anything we get from the full 18 com mittee.
19 I guess the next step may be we might want to hear 20briefly from you in a month or so on just how you resolved l
I 21 the program and what the final Action Plan looks like.
22 Maybe a briefer presentation to the full committee would i
23 perhaps be all tha t is necessary.
In fact, I guess that
()
24 might be a good way to think about it.
You said you would 25 be ready in about a month with a resolution of some of the
(~\\
()
l ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON D.C. 20024 (202) 554-2345
64
()
1 differences here.
Why don't we at least tentatively plan on 2 a f airly brief presenta tion to the full committee in August?
3 MR. MARCHESE:
Okay.
4 MR. WARD:
Unless --
5 HR. ETHERINGTON:
Could I raise one more 6 question ?
You are going to have a dozen or more 7 con tractors, and if similar programs, past programs are any 8 guide, you are going to have 11 of those 12 describing in i
Ggreat detail what the problem is.
10 Is there any way to discourage contractors from 11 sta rting way back and describing Three Mile Island, what was 12 vrong and how they should avoid it, or do you have to read 13 those ad nau s'eu m ?
14 MR. MARCHESE4 It would certainly make my job a 151o t easier if we just had one contractor.
16 MR. ETHERINGTON:
Do you have the same kind of
(
17 f eeling ?
18 MR. MARCHESE:
I think yes, probably what you are 19 saying is true to a large extent.
20 MR. ETHERINGTON4 Can't you tell the contractor i
l l
21 please d on ' t describe the problem?
22 MR. MARCHESEs Tell us how you are going to solve 23 1t.
1 f( )
24 MR. ETHERINGTON:
Right.
Perhaps somebody should 25 describe the problem once.
l
()
ALDERSON REPORTING COMPANY, INC.
400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024(202) 554 2345
65 O
an a^acarst-
> rt or tae eve 1u tioa or or oo 1=
2 when we send out an RFP.
I agree with you entirely.
A lot 3 of the proposals we get in basically pantomime a description 4 of the problem and do not propose firm methods of attacking 5 the problem that could hopefully lead to some solutions.
6 But that is part of the proposal evaluation process that you 7go through.
Hopefully in the process we will iron it out.
8 MR. ETHEBINGTON:
Okay. I am probably d rea ming.
9 MR. WARDS Is there anything else?
10 (No response.)
11 Well we will call it a day, then.
12 (Whereupon, at 3:10 p.m.
the meeting wa s 13 adjourned. )
14 l
15 l
l 16 17 18 19 l
20 21 22 23 0
24 25 l O I
l ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE. S.W., WASHINGTON, D.C. 20024 (202) 554-2345
2 i
NUC:2AR REGUMTORY CO.5EC/'.SSICN This is to certify that the attached proceedings before the ih in the ca.tter of:.
ACRS/Subcommitt ee on Decay Heat Removal Systems Date of Proceeding:
July 6, 1981 Docket Nuinber:
Place of Proceeding:
Washington, D.,C.
were held as herein appears,,and. that this is the criginal. transcM ;:t -
therecf for the file-of the Coc=ission.,
Das S.
Parker Official. Ieporter (Typed)
.m
~
l
.Lu k (SIGF.AWP.E OF.:.IPC?.C7.)
i 1
t O
4 e
s
O O
O NRC STAFF STATUS REPORT ON UNRESOLVED SAFETY ISSUE (USI) - TASK A 145 l
" SHUTDOWN DECAY HEAT REMOVAL (SDHR) REQUIREMENTS" FOR THE 1
l ACRS SUBCOMMITTEE ON DECAY HEAT REMOVAL SYSTEMS i
JULY 6,1981 j
p i
A. R. MARCHESE GENERIC ISSUES BRANCH l
DIVISION OF SAFETY TECHNOLOGY i
l i
i yQ I
i
_ - ~ _ _ -.-
c i
MAIN PURPOSE OF TODAY'S MEETING
- O l
0 DISCUSS ACRS SUBCOMMITTEE'S COMMENTS ON DRAFT TASK ACTION PLAN FOR A-45 i
e l
4 6
4 1
e 5
lO l
i 4
4 4
I a
l E
!O F
i
'f 2
-,-ve%. -, *
.w~..
e--y-
,-,w-w.-.-,ww+-we.-wee.,-,wr,
. 3m-m
,,we w w ww,,m -,,w-vey,,,,
,.,p__-m-.,,-
~
UPDATE ON TASK A-45 SINCE ACRS MEETING OF MAY 5, 1981 0DRAFTTASKACTiONPLANFORA-45ISSUEDONMAY 22', 1981 O
I ACRS SUBCOMMITTEE REVIEW AND COMMENT ON JULY 6', 1981 8 INTERNAL REVIEW, COMMENT AND COMMITMENT OF MANPOWER SHOULD BE COMPLETE BY JULY 10', 1981 0 FINALIZE AND APPROVE. PLAN BY AUGUST 7, 1981 0 CONDUCT DISCUSSIONS WITH NATIONAL LABS AND PRIVATE FIRMS TO DETERMINE THEIR CAPABILITIES TO PROVIDE TECHNICAL ASSISTANCE -
COMPLETE BY SEPTEMBER 1981 0 WRITE RFPs FOR FY82 TECHNICAL ASSISTANCE CONTRACTS - COMPLETE DURING JULY THROUGH SEPTEMBER 1981 8 IMPLEMENT FY82 CONTRACT WORK WITH NATIONAL LABS BY OCTOBER 1981 O
9 IMPLEMENT FY82 CONTRACT WORK WITH PRIVATE FIRMS BY JANUARY 1982
\\
O
4 OTHER ITEMS OF INTEREST WITH RESPECT TO TASK A-45 O
e TASK MANAGER TOURED STANDBY SHUTDOWN FACILITY FOR MCGUIRE AND OCONEE NUCLEAR STATIONS 8 INCREASED RELUCTANCE TO ISSUE REVISION 1 TO R. G. l'139 ON
~
RESIDUAL HEAT REMOVAL SYSTEMS 8 MAY NEED TO ADD A MAJOR SUB-TASK TO A-45 TO PERFORM COST /
BENEFIT EVALUATIONS TO DETE MINE IF IT IS NECESSARY FOR-EXISTING PLANTS TO ACHIEVE AND MAINTAIN COLD SHUTDOWN USING SAFETY GRADE EQUIPMENT
' O t
' O
-Figure 1.
Inter-Relation of Sub-Tasks in Task Action Plan A-45 Develop Means Develop Interim Accepta$
DevelopAcceptanceCriteM for Criteria for Existing for Future Plants (Risk O
tmProvia9 sowas eieats oa aisk Basis Terset ^'iocatioa)
Categorize All Plants As " Existing" or " Future" Y
V 1]
GroupingofExistMh Assess Adequacy cf AssessAdequach E
Plants for Assessment+
SDHRS in SDHRS in Future of Adequacy of SDHRS Existing Plants on Risk Plants Basis Assess Adequacy of tid SDHRS in Existing Plants on Deterministic Basis (As Necessary) i i_.
Design and Cost ProposaN Cesign a'd Cost #
n for Improved SDHRS for laproved SDHRS for Selected (High Risk)
Selected Future Plants p
Plants
\\r OtherTasb Compare Cost Other Studies Relating +
Effectivene.;s of Alternative <
to Risk Reduction Means for Reducing Risk ReconsiderationoftheAdequaN of the Proposed Acceptance Criteria for SDHR Systems Y
! evelop Plan for Implementink D
New Requirements Legend:
IX X - Identifies Sub-Task Number
~,