ML20058L466
| ML20058L466 | |
| Person / Time | |
|---|---|
| Issue date: | 12/10/1993 |
| From: | Advisory Committee on Reactor Safeguards |
| To: | |
| References | |
| ACRS-T-1991, NUDOCS 9312170003 | |
| Download: ML20058L466 (100) | |
Text
W m
$5 i
N
^
Vggg?gigirp Ng)r
~~7p i;
g
@?8 y g gg; %
c 4
vyp w
hg r
n '& *edjy
.2 e.
V MJ a
i, g W SN
' % Si g e
r
- %%sBp
'f-%
j M vs " g,s k SM$hk }W b h,..
" ' g n,,
';n n
%V h"Y h
ahg; '
g-
!% g'ggggrW
.^as s
.s a
i y
g n,_gj C
E g
g:
for=
'd % "is Ng$ y =
~ w
" v 3
mJ "S
@!? 9MM S
V
' " - ~
KfV,;Q$KNMsO n a M
q g:MI na wh$Q;r ht x
.j f$bL umw n
AA A
- i As%nN reL /
~
1Pr yi n
- sg r y
p.
e r -
wa
_ _7nn p ewp r /( [5, gy, EfB= _
=
IM gg spie,
g Q
b l N d
' " - ~
W Y
w ^ ? YL
^
p & g:g$ilgg$,u n
' Q $QQ RQy@h f g [' 31
., 4*y Mj ggQh t'
=a Q ' -
?
j%
f a$
ef$yd h
hhhh @@hin$gdt?
"g$a.~f h ~ _ a $gj y
., kl w
W
$$d @{#3#h. h$b[*
i
~
!!n hlgh fb?
$ w. [k d
e 3D
~;
Mk%d&d%W 3A Y e L
hg4w g g A s$s ag }4 6 $ pW g
a#_ g%Q Y
$nggghg y
M" * ' (;
g
); y?
964%aght;W a%"'-
4
~
bwW V**
i
+
=
hM*W@%sq k dh~Y Wijh tji 3
%%A G
f
$s N$
i ^ $%y N %' f;f.
i
- k. e.mMb..
! EWRfr p
1 pa j
a
%dh a.
R mM a
LM gggg"]:'M bfpdf
~4f, F %.
+
L'
~
TS$ pb %
vu-C'
( dhysifh$ $7 qhffk}$$ s g $$$ $l
.d S -
h ghffe O M N 2
- .t -
7 hba?5N
$s
>4 Q b
1
?
hil %.
m y
a
(-
X 9312170003 931219 y ya M
0
- I T-1991 PDR k-NNEMEE$
[. La A.
2,
.+"A-2
.s sp.
Ac45 T:- /#/
ORIGINAL O
OFFICIAL TRANSCRIPT OF PROCEEDINGS Agency:
Nuclear Regulatory Commission Advisory Committee on Reactor Safeguards
Title:
404th ACRS Meeting l
Q Docket No.
^
f.OCATION:
Bethesda, Maryland DATE:
Friday, December 10, 1993 PAGES: 365 - 449 ar- = :;a._ - %
g I
d
}.I
,O f(Mth@!s,,IBSOECsmmittee
.e
.. e,,,
c.
't
,y gV
' 9'; 0 0' {"' ?
- VN RILEY & ASSOCMTES, LTD.
')
c
- s. J :
97 1612 K St.,N.W., Suite 300 0 -[
i t' i e.
Ih Washington, D.C. 20006 (202) 293 3950
)
1 PUBLIC NOTICE BY THE O
f UNITED STATES NUCLEAR REGULATORY COMMISSION ADVISORY COMMITTEE ON REACTOR SAFEGUARDS F
i DATE:
December 10, 1993 i
l O
The contents of this transcript of the proceedings 1
of the United States Nuclear Regulatory Commission's Advisory Committee on Reactor Safeguards, (date)
{
December 10, 1993
, as Reported herein, are a. record j
of the discussions recorded at the meeting held on the above date.
I This transcript has not been reviewed, corrected or edited, and it may contain inaccuracies.
l J
P ANN RILEY & ASSOCIATES, Ltd.
Court Reporters 1612 K. Street, N.W., Suite-300 Washington, D. C. 20006 es -
365 1
UNITED STATES OF AMERICA 2
l 3
NUCLEAR REGULATORY COMMISSION j
4 5
6 ADVISORY COMMITTEE ON REACTOR SAFEGUARDS-7
-t 8
9 404th ACRS MEETING 10 s
11 12 Nuclear Regulatory Commission 13 Conference Room P-110 j
14 7920 Norfolk Avenue (s
15 Bethesda, Maryland 16 l
17 Friday, December 10, 1993 18 l
l 19 8:30 o' clock a.m.
l 20 21 i
i 22 l
l 23 l
24 i
25 ANN RILEY & ASSOCIATES, LTD.
Court Reporters-1612 K Street, N.W.,
Suite 300 Washington, D.C.
20006 (202) 293-3950
I 366' l
1 PARTICIPANTS:
2 E. WILKINS, Chairman of the ACRS 3
J.
CARROLL, Vice-Chairman of the ACRS t
4 C. MICHELSON, Member of the ACRS l
5 C. WYLIE, Member of the ACRS 6
H.
LEWIS, Member of the ACRS 7
I.
CATTON, Member of the ACRS 8
T.
KRESS, Member of the ACRS 9
W.
LINDBLAD, Member of the ACRS 10 P. DAVIS, Member of the ACRS i
11 R.
SEALE, Member of the ACRS 12 W.
SHACK, Member of the ACRS i
13 J. LARKINS, Executive Director.of the ACRS 14 C.
ADER, NRC/RES O
15 J. MURPHY, NRC/RES
{
16 J.
FLACK, NRC/RES j
17 M. DROUIN, NRC/RES 18 R. HERNAN, NRC/NRR 19 i
i 20 l
21 22 f
23 i
24 25 i
ANN RILEY & ASSOCIATES, LTD.
Court Reporters
'l 1612 K Street, N.W.,
Suite.300 Washington, D.C.
20006 (202) 293-3950 i
367
)
1 PROCEEDINGS j
()
2
[8:30 a.m.]
3 MR. WILKINS:
The meeting will now come to order.
4 This is the second day of the 404th meeting of the Advisory.
5 Committee on Reactor Safeguards.
During today's. meeting, 6
the committee will discuss and/or hear reports on the 7
following: one, individual plant examination program; two, l
8 EPRI passive LWR requirements document; three,. safeguards-
{
9 and security requirements; four, preparation of ACRS 10 reports; five, future ACRS activities; six, election of l
11 officers and seven, reconciliation of ACRS comments and 1
12 recommendations.
13 Portions of today's meeting will be closed to 14 discuss safeguards and security information, classified 15 foreign intelligence information and information the release 16 of which would represent a clearly unwarranted invasion of 17 personal privacy.
18 This meeting is being conducted in accordance with 19 the provisions of the Federal Advisory Committee Act.
Mr.
20 Dean Houston is the designated federal official for the j
21 initial portion ot the meeting.
22 We have received no written statements or requests 23 for time to make oral statements from members of the-public 24 regarding today's sessions.
A transcript of portions of the 25 meeting is being kept and it is requested that each speaker f
ANN RILEY & ASSOCIATES, LTD.
4 Court Reporters 1612 K Street, N.W.,
Suite 300 Washington, D.C.
20006 (202) 293-3950
368 1
use one of the microphones, identify himself or herself and
()
2 speak with sufficient clarity and volume so that he or she 3
can be readily heard.
4 The only announcement I have is that there is an 5
ANS executive conference on policy implications of risk-6 based regulations in Washington at the Stouffer Mayflower 7
Hotel on March 13th through 16th.
8 It may be of some interest to the members of the 9
committee to know that one of the panel moderators is our 10 former chairman and friend, Dave Ward, and the meeting is 11 being organized by John Garrick whom.I think many of you 12 know.
13 Some of the speakers are Herb Kouts; Herschel 14 Specter, John Garrick, Vojin Joksimovich and Jack Heltemes..
15 The program will be distributed when it becomes available.
16 Are there any other general announcements or comments that 17 anyone wishes to make?
18
[No response.)
19 MR. WILKINS:
If not, we will proceed to the next 20 item on the agenda, item nine, the Individual. Plant 21 Examination Program and Pete, you are the subcommittee j
22 chairman.
i 23 MR. DAVIS:
Thank you, Mr. Chairman..On November 1
24 18th the IPE Subcommittee did meet and obtained from the 25 staff the latest information on the IPE progress and results-l ANN RILEY & ASSOCIATES, LTD.
Court Reporters i
1612 K Street, N.W.,
Suite 300 Washington, D.C.
20006 (202) 293-3950 i
)
i 369' i
1 to date.
We also heard a brief presentation on the (f
2 Individual Plant Examination for External Events program and 3
also a discussion of the Severe Accident Management program.
4 The subcommittee' decided that for the purposes of 5
the full committee meeting we would ret,*.rict our. attention i
6 to the IPE progress and put off the IPEEE and accident 7
management topics to a later time primarily because they are 8
in early stages now of development and we are still awaiting 9
results.
The agenda for this morning's presentation is page.
l 10 two of tab six.
l 11 We will hear the IPE program results and insights,
(
12 a discussion of the' database program and then finally a.
l 13 discussion of the generic issue resolution by the IPE 14 process.
The Subcommittee har decided to write a letter and 15 I have a draft that will be available for consideration 16 later.
17 I would like now to turn it over to the staff.
18 Mr. Flack, are you prepared?
19 MR. FLACK:
Yes.
20
[ SLIDE.]
21 MR. FLACK:
Good morning.
I am from the' Office of 22 Research and am the project manager of the~IPE program.
23
{ SLIDE.]
24 MR. FLACK:
What I would like to talk to you about 25 today is just briefly give an overview of.where we are right
()
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,
Suite 300 Washington, D.C.
20006
'(202) 293-3950
~
370 1
now with the program.
Basically I will provide a summary of t
2 the background, where the IPE came from very briefly and 3
then the status of where we are today and then look at the i
4 review progress that we are utilizing to review the IPE 5
submittals and what the objectives are and what we are 6
observing at this time with respect to licensees meeting 7
those objectives and then finally just some general 8
observations of the program.
9
[ SLIDE.]
10 MR. FLACK:
Basically the IPE program stemmed from 11 the Severe Accident Policy Statement which was issued in-i 12 1985 which stated that the nuclear power plants pose no i
13 undue risk to the public health and safety but at the same 14 time it also recognized the benefit of PRA.
s-)
15 This led to the integration plan for closure of i
16 severe accidents which was issued in 1988 and one of those 17 key elements was to have licensees do a form of PRA called I
18 IPE although a PRA was not required.
Later-on, objectives 19 would essentially require that licensees do PRAs.
I 20 The objectives of the key element IPE were laid 21 out in generic letter 88-20 which was also issued in 1988.
22 However, that did not initiate the process.
A workshop was l
23 held in the following year to more define what the staff 24 needed to look at, the information that the staff would need 25 to assure itself that the IPEs were being done properly.
F
)
ANN RILEY & ASSOCIATES, LTD.
Court Reporters i
1612 K Street, N.W.,
Suite 300 i
Washington, D.C.
20006 (202) 293-3950
371 1
1 On August 29, 1989 Supple: ment 1 to that generic
()
2 letter was issued with the submittal guidance document that i
3 requested a certain amount of information.
At that time the 4
IPE process was initiated.
5 Supplement 1 also contained CPI recommendations 6
that should be considered or that.needed to be considered as 7
part oZ the IPE for MARK I containments.
A second 8
supplement was issued in 1990 which contained accident 9
management strategies that were also to be considered during
+
10 the development of the licensee's IPEs.
11 Supplement 3 was issued in 1990 also and that 12 contained the CPI recommendations for other plant types ~
13 other than MARK I and finally Supplement 4 was issued in
~
14 1991 which initiated the IPEEE process.
i O
15
[ SLIDE. ]
l 16 MR. FLACK:
THE IPE submittal schedule, here is 17 how we are seeing the IPEs come in.
Basically, about half 18 of them have been submitted before the due date which was 19 the end of fiscal year 1992.
We see a lot of carry-over.
20 We a: e right here in the first quarter of fiscal year 1994 21 and we expect to get the rest of them.in by mid next year.
22 Most of this carry-over is due to a number of 23 things.
We have given extensions to those plantsLthat 24 decided that they were going to do full scope PRA, level 2 l
l 25 PkAs which were not required and that is about'42-percent of i
("N
( j/
1009 RILEY & ASSOCIATES, LTD.
I Court Reporters 1612 K Street, N.W.,
Suite 300 l
Washington, D.C.
20006 i
(202) 293-3950 t
372 1
plants, some of those that we gave extensions to.
()
2 Others are due to multi-unit sites where licensees 3
have chosen to do plants in sequence so we allowed 4
extensions on that but most of the slippage is due to the l
5 unexpected complexity of doing an IPE.
Many of them have f
l 6
come in and requested additional time because they see the j
7 need for additional analyses.
8
[ SLIDE.]
9 MR. FLACK:
Right now we are sitting actually with 10 64.
Fort Calhoun came in yesterday and I haven't been able l
11 to update the slide but essentially 80-percent of the 12 submittals are in.
13 We have initiated 35 reviews at this time.
j 14 Thirty-one of these are step 1 or are presently at step 1 or 15 have been completed as step 1 reviews and four have gone on 16 to step 2.
I will go over the review process.
Step 1 is a 17 submittal only review where step 2 requires a site visit.
18 At this time, 13 reviews are completed.
Eleven of 19 these have been completed under step 1 only.and.two'have 20 been completed under step 2 and there are a number of them 21 that don't show up under those numbers but these.have been 22 completed or in the very final stages but 13 at'least have
[
23 left the Office of Research.
24 MR. WILKINS:
Are those 13 part of the 35 or added 25 to the 35?
I ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,
Suite 300 i
Washington, D.C.
20006 (202) 293-3950
373-1 MR. FLACK:
No.
That is part of'the 35.
We are
)
2 still shooting for the end of.1995 as the completion date 3
for all reviews.
4 MR. WILKINS:
Are you prodding the complement of 5
the 63, that is 63 of 78, there are 15 missing orLdid I'miss 6
something?
7 MR. FLACK:
That are still to come in.
We expect 8
those in by mid next year.
9 MR. WILKINS:
There is no concern on your part 10 that people are dragging their feet?
11 MR. FLACK:
No, not at this time.
They have been 12 all granted extensions for one reason or another.
13 MR. FLACK:
Step 2 is a more detailed review.
As-14 I mentioned we do go to the site.
There-are four that have 15 entered into the step 2 process and the reasons for doing 16 so, I have listed here.
17 Turkey Point was the first plant we looked at 18 under step 2 and basically went to the site since it was the 19 first plant that came in that had not been' formally reviewed 20 by the staff before.
The other plants at that time had all 21 PRAs submitted to the staff that had been reviewed and we 22 had contractors' reports available so it was sort of-the 23 first plant that came in fresh.
24 FitzPatrick, we went to the site since there were l,
25 a number of issues that were surrounding that plant at the i
l i
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,
Suite 300
)
Washington, D.C.
20006 1
(202) 293-3950 i
3
-5
?
374 1
time and we wanted to better understand what was done as 2
part of the IPE so that led us into a step 2 review for 3
FitzPatrick.
4 McGuire, we did a limited step 2 to better i
5 understand the HRA process and that was the Duke Power i
6 plant.
They had a number of submittals and on first. reading j
7 over or through the submittal, it indicated more of a 8
cookbook approach and we wanted to understand more of the 1
9 details of how they did their HRA so for that part, we.did a 10 limited step 2 review on McGuire and this one is in the l
11 final stages right now.
12 Finally, a few months back we went to the Zion i
13 site.
We chose Zion.to some extent because of their l
14 methodology, to better understand their methodology but it l
O 15 is also a Commonwealth Edison plant and there-are a number l
16 of them coming in so we wanted to get as much information as 17 we can up front so that it could benefit our reviews-later i
18 on, j
19
[ SLIDE.]
20 MR. FLACK:
The process we a utilizing to review 21 the IPEs consist of a team basically from two to four i
22 members.
Some of the members have to be overlapped because 23 of lack of resources or the expertise lies with the team j
24 leader so we utilize his or her. expertise during the review.
i 25 But basically it is a team anywhere from two to-i ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,
Suite 300 Washington, D.C.
20006 (202) 293-3950 i
375 I
1 four members, the team leader heading up the team who is
()
2 responsible for getting the staff evaluation report out, a l
3 front-end system analyst, a back-end containment analyst and 4
one who is familiar with human reliability.
5 Under step 1 as mentioned,:it is a submittal only l
6 review.
Basically it comes in, a team is assigned, they 7
review the submittal, generate questions about any 8
peculiarities about the analysis, the findings and the 9
results and we have sent questions to the licensee and given 10 usually about 60 days to respond.
i' 11 MR. LINDBLAD:
Excuse me, Mr. Flack.
How many 12 such step 1 teams are there or how many people are involved 13 in review teams?
I 14 MR. FLACK:
Right now, we are running about'six
.s 15 and a half people, around six to seven people that are 16 actually involved in the reviews themselves.
l 17 MR. LINDBLAD:
So there :bs really at any given 18 time --
19 MR. FLACK:
We don't have different teams.
We 20 don't have a specifically assigned team.
So we will have a 21 team leader on several plants at one time.
)
1 22 MR. LINDBLAD:
But the total pool you are drawing
)
i 23 teams from are six to seven?
24 MR. FLACK:
Around six people, yes, six to seven.
25 MR. CARROLL:
These are all staff people or do ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,
Suite 300 Washington, D.C.
20006
)
(202) 293-3950
376 1
they include contractors?
I 2
MR. FLACK:
These are staff people.
Now 3
contractors do support but the staff, of course,'has the 4
responsibility of writing the SER so that is the team's 5
responsibility.
I 6
MR. CARROLL:
Are contractors typically members of I
7 review teams?
8 MR. FLACK:
.The way we worked the contractors 9
generally is we will have the team and now we are moving 10 into this mode where the submittal is sent to a contractor i
11 for review and the team interacts with the contractor.
l 12 Then the contractor will come in and give a 13 presentation and at that time, the questions that'need to be i
14 raised to the licensee would be generated and from that O
15 moment on, the contractor is out of the picture.
At least j
)
16 that is the way we have been working it so we have 17 capitalized on the contractor's insights _and then we go-18 forward with the rest of the review.
l 19 Because of limited resources we are planning on.
j i
20 utilizing the contractors more in evaluating the response 21 from the licensee.
Up to now, they have not been involved j
22 in that process.
I 23 MR. CARROLL:
How many contractors have been used 24 and who are they?
25 MR. FLACK:
The contractors are SEA for the front-f ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,
Suite 300 Washington, D.C.
20006 (202) 293-3950
~~ ~
l 377 i
1 end, Scientech for the back-end and Concord Associates for
()
2 the HRA.
So we have three separate contractors assigned to 3
the three different areas.
4 MR. CARROLL:
Thank you.
5 MR. FLACK:
Upon receipt of the licensee's l
l l
6 response to the questions the team needs to make a decision 7
on whether to go to step 2 based on the information they 8
have.
The decision is generally based on an understanding l
t 9
of the process, whether they can close it at that. time, they 10 have seen enough information, they understand what was done 11 and we can close it or if not, we have to go further and 12 that would require a site visit or if the plant had some i
13 unique characteristic that needed to be explored a little-14 bit further.
O-t 15 The site visit will involve contractors from each t
16 of the areas.
They will go to the site with the team and I
17 then following the site visit and resolution of any issues
{
f 18 that might come out of that, the staff will then write its 19 staff evaluation report.
l l
l 20
[ SLIDE.]
21 MR. FLACK:
Now the objectives of the review as j
22 well as the objectives of the program itself actually center i
23 around the generic letter 88-20 objectives.which-I have i
24 listed here and which I would like'to relate each of these 25 objectives back to what we are finding as far as.what the ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,
Suite 300 Washington, D.C.
20006 (202) 293-3950
l l
l 378 I
1 licensees are doing actually.
()
2 To achieve an overall appreciation of the severe 3
accidents and this was generally to' bring licensees into I
4 thinking beyond the design basis, to ask the question "what-j 5
if" and, of course, the staff's point of view is that they 6
are complete with respect to addressing the phenomenologies.
i 7
The second objective, of course, is to identify 8
the dominant sequences which is probably one of the most i
l 9
important outputs of the IPE but not only to identify and 10 understand the dominant sequences but to understand what is 11 contributing to those sequences and we will talk about that.
12 The third objective was to gain an quantitative
[
13 understanding of core damage ~and radioactive release.
14 Quantify, of course, is a part of the IPE process to O
15 understand, for example, what'is the importance of a system 16 with respect to core damage and how that could be extracted l
17 out of the analysis and that is an important part of that.
I 18 Finally, to reduce the probability of core damage f
i 19 and release of radioactivity and-that is through the
{
't 20 understanding of what a vulnerability is, the definition of 21 vulnerability or at what threshold do licensees enhance.
22 their plant.
These are the objectives.
23 MR. CARROLL:
You mentioned the definition of a 24 vulnerability.
Apparently you now know what it is.
25
. ix. FLACK:
It varies.
i l
O.
I ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,
Suite 300 i
Washington, D.C.
20006 (202) 293-3950 l
379 l
1
[ Laughter.]
)
2 MR. LINDBLAD:
Mr. Flack, do I understand it is 3
intended that the licensee is the receptor of this 4
appreciation and understanding, is that right?
5 MR. FLACK:
That's right.
6 MR. LINDBLAD:
It is not a training ground for~the 7
staff so much as for the licensees?
i 8
MR. FLACK:
Well, the emphasis.was on the 9
licensees, that's right.
But I will now go back and relate-10 back to these objectives and what we are finding with 11 respect to licensees' processes.
12
[ SLIDE.]
13 MR. FLACK:
With respect to the appreciation of-14 severe accidents, the first as'I have mentioned.before, they l
~
15 are all doing at least a level 1 PRA except Susquehanna 16 which has done some other methodology which is pretty much a 17 PRA only they use substantial deterministic reasoning which 18 moves them away from that arena somewhat but basically they 19 are all daing at least a level 1 PRA and 42-percent are j
20 doing a level 2 or level 3 PRA which goes beyond the scope i
21 or the requirements.
j 22 Licensees are involved.
We see quite a bit of 23 involvement with the licensee which gives us the' feeling 24 that this information is being transmitted to the license 25 where it should be.
Roughly 50 percent on the average is ANN RILEY & ASSOCIATES, LTD.
Court Reporters i
1612 K Street, N.W.,
Suite 300 Washington, D.C.
20006 (202) 293-3950
)
i I
380 1
licensee involvement.
Od 2
We see a number of implementations of safety 3
enhancements and this is not based on simply the' definition
'i 4
of what a vulnerability might be.
They are taking action 5
based on understanding the output of the analysis rather 6
than establishing a threshold where'if it is above this j
7 number, we will fix it; if it is below, we won't.
We don't 8
see that.
We see them making enhancements are they.are f
9 going through their plants.
l 10 Most important is we are seeing this evolution'of
{
11 living documents which was somewhat kind of unexpected.
To-12 date all of the plants that we have reviewed, the licensees l
13 have committed to living programs which is a very good sign-i 14 that the program will be long-lasting and possibly_ validate 15 some of these assumptions that they are making in theiriIPEs I
16 in time, 17 MR. CARROLL:
Why did you find that surprising?
l 18 MR. FLACK:
Well, it was.not a' requirement and I
.i 19 guess early on in the program when we went through Fort i
20 Worth with the workshop and so on there seemed to have been 21 some resistance to the effort and we didn't know exactly.
22 Maybe it was just people who were outspoken at the 23 time but we didn't know exactly how they would.do it.
They l
24 may simply do it as a one-shot deal-and just. walk away from I
i 25 it but we are not seeing that.
So we are somewhat surprised
'l i
l I
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,
Suite 300-Washington, D.C. 20006 j
(202) 293-3950
-j l
381 1
to see'them all going into the living program to date.
j 2
MR. KRESS:
John, for those people who chose 1to go 3
to a level 2 analysis, what was the basis'for that decision?
4 Is there something that characterizes those that is 5
different than the other plants?
6 MR. FLACK:
I guess I never asked that licensees.
i 7
that question as to why they went to a level'2 but my own 8
feeling would be that well certainly for some plants that I
9 are different in nature.and-they believe that they want a i
10 more accurate description of their plant from a r
11 probabilistic perspective rather than using other analyses 12 to fill in have committed to doing that.
Maybe they see a 13 benefit by doing that.
14 of course, with the level 3's now they do use the
\\-
15 cost / benefit to some extent as a criteria'for ycaking fixes I
i i
16 so they get that part of it and I guess about half of them t
17 have gone to level 3's so you have that part.
i l
18 But I don't know.
I have questioned them and 19 asked them, "How much more does it take to do a level 2 over i
29 what you are already doing?"
It is not really that much-23 more so some went the complete distance by doing the 22 structural analysis and may have gained some better insights 23 as to where the containment is going to fail, for example.
24 MR. SEALE:
This terminology of a living document 25 is a fairly subjective term.
I imagine there-are probably i
O ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,
Suite 300 Washington, D.C. 20006 (202) 293-3950
.}
t 382 1
two or three times as many definitions as there are people 2
in this people depending on the time of day.
Do you have 3
some criterion for deciding what constitutes a'partrof the'
{
4 evidence that these documents live?
What do you mean by 5
that?
6 MR. DAVIS:
There isn't any evidence yet.
7 MR. FLACK:
Right.
8 MR. SEALE:
But it is an attribute that you have 9
assigned.
[
t 10 MR. DAVIS:
Well, they say they are doing to do it 11 and that is all the evidence we have, f
12 MR. FLACK:
If it is not a one-shot deal which 13 means they just want to do it and be done with it and not 14 think about it any more, I guess we would put them all in C.g$
15 the latter category where they intend to at least go back-l 16 and update it with data.
17 MR. SEALE:
But in the long run, there has to be a 18 more specific understanding of what that means.
I 19 MR. FLACK:
Yes.
I think personally that this 20 area is something that should be focused on in the future as 21 to what we mean by living and what requirements might be in 22 place that would constitute it as a living document 23 certainly.
I think that is a whole new area that is 24 growing.
25 MR.. DAVIS:
But you didn't have any requirements O
ANN RILEY & ASSOCIATES, LTD.
.j Court Reporters i
1612 K Street, N.W.,
Suite 300 Washington, D.C.
20006 (202)'293-3950-1
1~
l s
1 3831 1
to maintain a living document.
()
2 MR. FLACK:
No,.that's right.
This is something 3
that goes beyond the requirements.in the generic letter.
4 MR. SEALE:
It is not surprising.
5 MR. CARROLL:
But we do have evidence that they 6
are living.
We recently received.either a NUMARC or an INPO.
7 report cautioning people about how PRA has been misused in i
8 looking at plant operating situations.
9 MR. SEALE:
- Yes, I just wanted'to know what his l
10 definition is.
l 11 MR. CARROLL:
I know.
12 MR. SEALE:
I have a definition.
i 13 MR. DAVIS:
Jay, you brought up an interesting 14 point and it leads to my question.
In some of the IPEs that N
15 I have looked at the licensee is actually taking action to 16 institute improvements on the basis of.the IPE and, I guess,.
l 17 I had a mild concern that somebody should be looking at what 18 these changes are to make sure that they.indeed reflect what 19 the PRA is saying and indeed are improvements because'almost 20 every action has some downside involved with it.
21 MR. FLACK:
Sure.
i i
l 22 MR. DAVIS:
Is somebody looking at those things?
23 At what point does it trigger some regulatory review, the l
24 actual changes that are being proposed or made?
25 MR. FLACK:
I think if it affects the licensing ANN P.ILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,' Suite 300 Washington, D.C.
20006
{
(202) 293-3950
i 384 1
basis, of course, they would have to come in under 50.59
()
2 which would address that concern.
3 MR. DAVIS:
Right.
4 MR. FLACK:
Now if it is a procedural enhancement, i
5 I don't believe that would need to come in under 50.59.
~
6 MR. CARROLL:
Who typically reviews the annual r
7 50.59 rubmittals?
-l 8
MR. FLACK:
That is sort of outside.my scope.
'I j
9 don't know if anyone from NRR is here who could address i
10 that.
[
11 MR. CARROLL:
Is it the region?
12 MR. DAVIS:
Does the region do that?-
l 13 MR. CARROLL:
Here comes somebody who'knows.
I 14 MR. HERNAN:
I would like'to talk to that.
I am 15 Ron Hernan.
I am the lead project manager'for the IPE 16 program-for NRR.
Reviews of the 50.59 determinations and i
17 evaluations are typically done by'both the project manager 1
18 for that plant and the resident inspectors.
It is done 19 usually about once a year and there is an inspection' module 1 20 for that purpose.
21 MR. CARROLL:
That tells people what to look for.
i 22 MR. HERNAN:
Yes.
23 MR. CARROLL:
In the course of these reviews how:
{
24 often do these two reviewers buck something up for 25 consideration at a higher level?
i i
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,
Suite 300 Washington, D.C.
20006 (202) 293-3950
385 i
1 MR. HERNAN:
I can only speak for my plant and I l
()
2 have Three Mile Island Unit 1, we haven't had one that had 3
to be elevated in the six years that I have been with'the i
4 project.
j l
5 MR. LINDBLAD:
Excuse me.
Mr. Flack, when you
-i i
6 spoke of improvements and when we have been discussing them, 7
I first thought of capital improvements and then you 8
mentioned procedural enhancements but does it also include 9
changes in the maintenance practices, the frequency 10 periodicity of doing inspections or calibrations?
11 MR. FLACK:
It may.
I can't say it does or it 12 dcasn't at this point.
I haven't seen it as a strong 13 contributor to risk so I don't know.
We will talk a little 14 bit more about enhancements later on.
O, i
15 MR. CARROLL:
I did think, Pete, that document i
16 that we have been talking about, was it NUMARC or INPO?
17 MR. DAVIS:
INPO.
18 MR. CARROLL:
It really focused on a classic type 19 of misuse of risk analysis.
I thought it was excellent that 20 it came out.
I hope it is a wake-up call to some people who l
21 think they understand what PRAs mean.
22
[ SLIDE.]
23 MR. FLACK:
In the next two slides I just briefly-24 touch upon some of the key. areas of the front-end and back-25 end IPEs or PRAs in general.
Firet of all, the level 1 is ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,' Suite 300 Washington, D.C.
20006 (202) 293-3950
L d
386 r
1 the analysis that takes you from the initiating event or the
()
2 demand for plant trip to core damage and some of the key
?
3 areas are the initiating events and the associated event 9
4 trees.
5 We are seeing as the initiators loss of'offsite 6
power as being one of the most important initiators equally 7
for both pressurized water reactors as well as boiling water 8
reactors.
LOCAs seem to be more important for "P's" and i
9 ATWS being more important for the BWRs.
10
" However, there is a class of initiating events 11 called transients that can really dominate, any one of which i
12 could dominate the risk profile at any plant-and this comes 13 from the system dependencies which is really at the heart of i
14 the IPEs, the PRAs, is understanding what.your systems 15 depend on and how loss of a support system can propagate 16 through the plant.
17 For example, the component cooling dependencies on 18 charging and the pump seals on component. cooling water,1for 1
19 example, is an important dependency that certainly has come 20 out of many IPEs.
I 21 Another key area is the success criteria and f
22 understanding what you need to mitigate an initiating event l
23 for each initiating event.
Data analysis and common cause 24 failure again is an important area.
Basically we are seeing 25 generic data utilized from 1150 and common cause failure i
ANN RILEY & ASSOCIATES, LTD.
s Court Reporters 1612 K Street, N W.,
Suite 300 Washington, D.C.
20006 (202) 293-3950 I
i 387 1
being treated pretty much consistently with 4780 with is the
)
2 multiple Greek method.
{
3 Internal flooding has been important in a couple 4
of plants.
This, of course, can be a vulnerability, a 5
source of vulnerability.
It was at one. plant and'the 6
switchgear room is generally the area.that is susceptible or 7
could be susceptible to flooding.
8 Of course, the decay heat removalLfunction is an 9
important area looking for not only dependencies which'can 10 compromise a decay heat removal system but also putting in-11 diverse means of decay heat removal, for example,-feed and 12 bleed in PWRs and secondary side depressurization and 13 condensation for PWRs and venting and using the fire water i
14 system for BWRs.
So'that was another key area in the' level 7_
15 1 analysis.
s-16
[ SLIDE.]
j 17 MR. FLACK:
Level 2 stems from the core' damage to 18 containment failure and these key areas are, first is-the 19 level 1 or the carry-over dependencies from the front-end 20 analysis to the back-end or level 2 analysis.
)
21 This is treated in different ways.
Some plants 22 tend to bin and end up with plant damage states and then 23 they propagate the plant damage states with the containment-24 analysis and others have used one large tree that will not 25 discern a difference between a level 2 and level 1.
i ANN RILEY & ASSOCIATES, LTD.
Court Reporters i
1612 K Street, N.W.,
Suite 300 l
Washington, D.C.
20006 (202) 293-3950 l
i
(
388 1
It will go right through.
You will start with an f( )
2 initiating event and go right on through to the releases.
3 But it is important that the dependencies.are carried over 4
and we see this in the analysis.
5 Containment failure modes and timing, of course, 6
is one of the most important outputs of the level 2 and in 7
there we have things like by-pass sequences and 8
overpressurization from containment and they are pretty much 9
consistent with what we have specified in NUREG-1335 as the 10 modes that need to be considered for the various plants.
l 11 The containment event trees are.the trees that 12 essentially represent the phenomena as well as things that 5
13 would be important in containment performance and we have-14 seen, of course, the staff in their review of IDCOR 15 commented that that methodology should not be used because 16 the CETs were very narrowly focussed.
It only had a'few f
17 tops and we haven't seen any IDCOR trees coming in.
The 18 licensees have been developing reasonable containment event 19 trees to represent containment performance.
20 Release categories, may have utilized the t
21 flexibility of using source terms from other analyses.
This 22 was granted them as part of the effort.
23 MR. LEWIS:
Did you learn to spell "catagory" that 24 way while you were with ACRS or somewhere else?
25
[ Laughter.]
ANN.RILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,
Suite 300 Washington, D.C.-20006
.(202) 293-3950-
}
389 1
MR. FLACK:
I wonder if I picked it up there.
(
)
2 Right, typo, 3
MR. LEWIS:
Fix it!
4 MR. FLACK:
Will do.
Uncertainties, of course, is 5
an important area.
Without them, the' treatment of 6
uncertainties one could end up with an unsupported technical 7
basis so treating uncertainties is an important part of the' 8
back-end analysis so that all phenomena is considered.
9 Finally, consideration of CPI recommendations 10 which is again another part of the back-end level 2 analysis 11 in which licensees consider things like water injection into 12 containment for BWRs and venting and hydrogen burns for 13 PWRs.
14 So in general we are seeing these things addressed O
\\--
15 as part of their IPE and meeting the requirement, the 16 licensees appreciate severe accidents.
17
[ SLIDE.]
18 MR. FLACK:
In addition to that we see a lot-of 19 follow-on activities which go beyond the scope of IPE which 20 again indicates that these analyses will be long living.
I 21 have summarized them here on these few bullets.
22 Accident management, of course, is a follow-on 23 work that the IPE will feed into with many licensees but 24 also this area, supporting licensee action which involves 25 possible tech spec changes and allowed outage time.
This is ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,
Suite 300 Washington, D.C.
20006 (202) 293-3950
i l
l l
390 t
l 1
not, of course, a part of the scope of IPE but.certain-
)
2 licensees are planning on using IPEs for that.
3 Evaluating safety issues whether.they are. generic; j
4 or plant specific as they come up is an area that'they will 5
be using them in.
Evaluating new design changes and what 6
that means with respect to core damage is another area that 7
the licensees are planning on using IPEs as well s
.l l
8 prioritizing proposed improvements or which improvements 9
would be the best as far as safety, bang for the buck type.
j 10 of thing, and finally, operator training by identifying i
11 areas which are important at the plant to train operators 12 on, scenarios, as well as actions that operators would need t
13 to take.
14 MR. SEALE:
I hope that operator is-a term here 15 which is not restricted to licensed operators.
Maintenance 16 training and that sort of thing as well, I would hope is-17 included.
18 MR. FLACK:
It is a generic term.
Now how they f
19 are using that, we don't really pursue that too much since 20 it is beyond the scope.
21 MR. SEALE:
If you do go into it, I think that is 22 a distinction you want to follow.
23 MR. FLACK:
All right.
24 MR. CARROLL:
You probably ought to stop talking J
25 about operator training and start talking about training of
)
L 1
U ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street,TN.W., Suite 300 Washington, D.C.
20006 (202) 293-3950 l
390 1
not, of course, a part of the scope of'IPE but certain f~
(_j) 2 licensees are planning on using IPEs for that.
i 3
Evaluating safety iscues whether they are generic 4
or plant specific as they come up is an area that they will 5
be using them in.
Evaluating new design changes and what t
6 that means with respect to core damage is another area that 7
the licensees are planning on using IPEs as well s f
8 prioritizing proposed improvements or which' improvements 9
would be the best as far as safety, bang for the buck type 10 of thing, and finally, operator training by identifying 11 areas which are important at the plant to. train operators s
12 on, scenarios, as well as actions that operators would need 13 to take.
14 MR. SEALE:
I hope that operator is a term here 15 which is not restricted to licensed operators.
Maintenance 16 training and that sort of thing as well, I would hope is 17 included.
18 MR. FLACK:
It is a generic term.
Now how they 19 are using that, we don't really pursue that too much since 20 it is beyond the scope.
21 MR. SEALE:
If you do go into it, I think chat is 22 a distinction you want to follow.
23 MR. FLACK:
All right.
l 24 MR. CARROLL:
You probably ought to stop talking:
25 about operator training and start talking about training of-Os ANN RILEY & ASSOCIATES, LTD.
Court Reporters i
1612 K Street, N.W.,
Suite 300 Washington, D.C.
20006 (202) 293-3950
f 391 i
i plant personnel.
A()
2 MR. SEALE:
That's right.
3
[ SLIDE.]
r 4
MR. FLACK:
The second objective, of course, was i
5 to understand the dominant sequences and certainly the 6
sequences as I have mentioned are an important' output of.the
-i 7
analyses and you will be hearing more about how we are 8
capturing those and putting those in the database at
'I 9
Brookhaven later on.
j 10 The licensees are required to report the top 100 l
11 sequences or 95-percent of core damage estimate so we do get 12 that information in but what is more important is how they 13 evaluate the contributors to those sequences.
I i
14 MR. WILKINS:
I guess I don't understand that, j
I
>\\>
15 They are required to report the top 100 sequences?'
16 MR. FLACK:
Yes.
.t 17 MR. WILKINS:
What is this "or 95-percent?"
If 18 there are fewer than 100 sequences altogether,'then they' i
19 report 95 percent of them?
20 MR. FLACK:
That's right.
There are different 21 methodologies.
For 100 sequences, you may only end up with I
22 65 percent of the core damage sequences so.95-percent is for 23 the functional type, different methodologies.
24 Licensees have been utilizing different' ways of 25 understanding how the contributors to these dominant
)
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,
Suite 300-Washington, D.C.
20006' (202) 293-3950:
[
n cw -v s
392 1
sequences by either performing a formal uncertainty analysis 2
or doing sensitivity studies and this is-calculating =
3 importance measures as a form of sensitivity study is
'l 4
'probably one of the most important areas in the IPE.
It was l
5 not required by generic letter 88-20 but it certainly puts 6
in perspective the significance of systems.
7 What that does is it will go into the analysis and 8
it set certain components equal to one and then see~what 9
that does with respect to core damage estimate and that is 10 called risk achievement worth or risk reduction worth,' going 11 in and placing zero in as a. probability of failure,-
[
12 conditional probability of failure of a system, for example, 13 and then seeing how that affects the core damage estimate,-
14 by doing that, by getting an appreciation'for how 15 significant those systems and components are.
I 16 We see that on many submittals' They have gone 17 and performed these importance measures which are really 18 beyond the scope of the program-but they do get at.an-19 understanding of what contributes.to those. sequences by 20 doing that.
21 (SLIDE.]
22 MR. FLACK:
The third objective was to obtain.a 23 quantitative understanding of core damage and here, of 24 course, when doing the probabilistic safety assessment they 25 do develop probabilistic estimate of core damage but what is
'l I
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,
Suite 300 Washington, D.C.
20006 (202) 293-3950
b I
i 393 l
I most important is understanding the basis for the estimate.
2 The numbers themselves are only as good as what 3
supports the numbers.
There are many assumptions that go 4
into these things, these analyses, that can drive the 5
numbers one way or the other and I will talk about that in a 6
minute but it is generally understanding what it is doing to 7
the estimates.
8 Of course, being sensitive to the analytic 9
limitations and, of course, as I have-mentioned licensees i
10 are not just basing their enhancements on numbers, bottom 11 line numbers that meet the threshold, they are making l
12 enhancements and they are sensitive from what we could tell 13 from our reviews to the limitations in the analysis.
14 MR. KRESS:
Is there consistent use of what is O'
i 15 meant by core damage in the IPEs like, for 4.nstance, having 16 the core uncovered for a fixed amount of
..r some such 17 thing as that?
1 18 MR. FLACK:
There is no generic definition.
Many 19 will use what 1150 used.
Again it depends, I guess where it 20 becomes important is how much they look at.in-vessel 21 recovery and how much-credit they are taking for that and i
22 that varies from plant-to-plant.
There is no generic 23 definition at this time that'I know.
24 MR. CARROLL:
But 1150 had a different definition 25 if I recall correctly for BWRs and PWRs.
ANN RILEY & ASSOCIATES, LTD.
Court Reporters i
1612 K Street, N.W.,
Suite 300
-Washington, D.C. 20006 i
(202) 293-3950
G l
394 l
1 MR. FLACK:
Yes, with respect to water level, I 2
believe, yes.
3 MR. KRESS:
How do you compare the numbers then 4
that you get?
5 MR. FLACK:
I don't know if the definition itself' 6
has a major impact but I was going to talk about some of the 4
7 assumptions that do.
8
[ SLIDE.]
9 MR. FLACK:
The next slide which indicates the 10 spread in the probabilistic estimate of core damage and all-11 I did was to bin them over a range of 2.5 on this histogram 12 and you can see that there is quite a range here on the PWRs 13 and on the BWRs.
This is the one with the other methodology 14 which we talked about, Susquehanna, down here.
i 15 By looking across here, we see that there is quite 16 a range of CDF coming out of the IPEs.
Some of it is due to 17 the design.
Having redundancy and diversity certainly will' 18 change the frequency estimates.
19 MR. CARROLL:
How will that curve be affected if q
20 the BWRs or the distribution be affected if the BWRs all had j
1 21 a diverse level indication in the reactor vessel?
)
22 MR. FLACK:
BWRs?
23 MR. CARROLL:
I am being facetious.
24 MR. FLACK:
Can anybody calculate that for me over 25 there?
1 l
ANN RILEY & ASSOCIATES, LTD.
Court Reporters
)
1612 K Street, N.W.,
Suite 300 Washington, D.C.
20006 (202) 293-3950
395 1
MR. CARROLL:
Ignore me.
2 MR. DAVIS:
Go ahead, John.
3 MR. CARROLL:
Ignore me.
4
[ Laughter.]
5 MR. CATTON:
Knock out the two top ones.
6
[ SLIDE.]
7 MR. FLACK:
Some of the important assumptions that t
8 we are seeing that could drive these numbers around are 9
certainly the treatment of operator actions and how much i
10 credit they are taking for recovery in the analysis will 11 impact the numbers.
Some will be conservative and some will P
12 go the extra yard and look for recovery actions and then 13 take credit for those.
14 Treatment of common cause failure and how much s
15 that is perceived to be important in the analysis can affect 16 the results.
The interpretation and applicability of data 17 and how data is treated certainly can affect the results, 18 whether they are using generic versus plant specific.
19 Success criteria can play an important role.
Some 20 will just use a conservative, what is known as the 21 conservative, the FSER, which has more-than enough to 22 mitigate initiating events and others will develop their own 23 success criteria by running codes such as MAAP and that can 24 affect the results as well as modeling.
For example,.the 25 reactor coolant pump seal model can have an effect on the ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,
Suite 300 Washington, D.C.
20006 (202) 293-3950 T
i 396 1
results.
()
2 MR. CARROLL:
What typically have PWR owners used 3
as a reactor coolant pump seal model?
I guess I got the-4 impression that the Westinghouse plants were being very.
5 pessimistic and the Combustion plants, the B&W plants, I 6
guess, were being very optimistic about coolant pump seal 7
leakage.
8 MR. FLACK:
Yes In fact we just reviewed Palo 9
Verde, for example, which has the Byron Jackson pumps and 10 they don't claim to have a very substantial leak. rate, a 1
r 11 very low probability of seal failure or resulting core l
I 12 damage whereas those that have_used 1150,'the seal model or 13 the Westinghouse seal model, have higher rates.
The whole 14 issue on the pump seal is an important issue.
[)
\\/
15 MR. CARROLL:
'Yes, it is.
16
[ SLIDE.]
17 MR. FLACK:
In fact the next slide'and let me just l
18 put the next slide up, you can see the' contribution from 19 different plants and what you say'is very true.
Of course, 20 you see Palo Verde at the bottom, less than one percent 21 contribution and you see the other plants just scattered 22 around.
23 MR. CARROLL:
Surry is low.
Why is that?
24 MR. FLACK:
Now Surry'had come in with a 25 resolution for generic issue 23.
The pump seal is being AV ANN RILEY & 'SSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,
Suite 300 Washington, D.C.
20006 (202) 293-3950
i 397 1
addressed as a separate generic issue.
(k 2
MR. CARROLL:
We know all that.
3 MR. FLACK:
All right.
4 MR. CARROLL:
That is why we are asking-the 5
questions.
6 MR. FLACK:
They take a lot of credit, I think, I
7 from what I remember, it has been a while since I looked at 8
surry, from tying in other systems and sharing and because 9
of the units and the ability to_ tie in other systems,
{
10 component cooling water and so on, they are able to reduce 11 the contribution of pump seals to core damage, I believe.
- 3 12 Now I don't know if anyone here~is more familiar with this
{
13 case than I.
14 MR. CARROLL:
I think that.is probably right but 15 your colleagues that are worrying about coolant pump seals 16 are saying deterministically you have ten minutes to get-17 cooling back on a seal after. loss of power and if you don't, 18 you have a 480 gallon-a-minute leak times four for four 19 pumps on a four-loop plant and any of the fixes'that you can i
20 propose are going to take more than ten minutes to.get the 21 water back on the seals.
22 MR. DAVIS:
Jay, I think that Surry has already 23 replaced their seals with the upgraded material.
24 MR. CARROLL:
That still doesn't solve the 25 problem.
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,
Suite 300 Washington, D.C.
20006 l
(202) 293-3950
.i 398 j
1 MR. DAVIS:
Not altogether but it does give them 2
additional time and I don't know why kind of credit they I
3 have taken for it.
4 MR. CARROLL:
The ten minutes is just how-long it 5
takes for the front of hot water to reach the seal.and then 6
if you want to view the world as some people on the staff
?
7 are, it is all over.
1 8
MR. MURPHY:
Joe Murphy from the staff.
At least 9
in the NUREG-1150 modeling in the plant as it existed when 10 we did that study, Surry had an independent charging pump 11 cooling water system that was unlike.anything I have.seen in 12 other plants.
13 MR. CARROLL:
Independent charging pump?
14 MR. MURPHY:
So you did not lose the charging.
.i fs e
15 Even if the seal failed, you did not lose the charging pumps 16 so you lost the mode of a component. cooling water or service 17 water failure that then led to both the loss of the seals 18 and the loss of the ECCS system.
19 There still was a contribution associated'with'the 20 station blackout type situation but in terms of the other 21 modes, they were covered by this charging pump cooling water 22 system that I haven't seen elsewhere.
23 MR. CARROLL:
I think a lot of people have put a i
24 system like that in or have one ready to put in once they 25 figure out where the staff is coming down on some of this.
l l
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,
Suite 300 Washington, D.C.
20006 (202) 293-3950
I i
399 j
1 MR. FLACK:
Yes.
The dependency on component
()
2 cooling water is clearly indicated here with the Catawba and i
l 3
McGuire analyses.
You will see that Catawba has that
-1 4
dependency on component cooling water for charging as well j
i 5
as pump seal cooling where McGuire does not.
So we see the-i i
1 l
6 difference.
Although they are very similar plants there is i
7 a difference in the contribution just based on that-one i
8 thing, dependencies.
9 MR. KRESS:
When you see CDFs of greater than 10 10 to the minus four, does that trigger any concern at.all?
I-11 notice maybe 20 plants on this chart you have-come-in with i
l 12 CDFs of greater than 10 to the minus four.
I am thinking in j
t 13 terms of the safety goals although they are not intended to 14 look at specific plant values, I realize.
15 MR. FLACK:
Not safety goals.
It is more of a 16 generic overview of the plants.
In fact, it goes back to-
}
i 17 our regulations.
We don't use it.as any criteria during our 18 review.
There may be political implications that are looked f
19 at elsewhere but not as a part of.our review.
20 MR. KRESS:
Is that considered as a grand overall 21 vulnerability?
22 MR. FLACK:
No.
Although some licensees have 23 defined using 10 to the minus four as a criteria for 24 vulnerability but those that have generally don't have them.
)
25 So the ones that are above 10 to the minus four use l
ANN RILEY & ASSOCIATES, LTD.
Court Reporters l
1612 K Street,' N.W.,
Suite 300 Washington, D.C.
20006 i
(202) 293-3950
'/
I something els 2
e, I
que tion.
s gues.
3 s
I don't MR. KRESS:
know if 4
\\
that 400
[ SLIDE.]
- Yes, answ
'\\
5 er is.
s your it 6
MR. FLACK:
\\
cor e damage is th
- Finally,
\\
7 us ba k e
to other c
r duc
\\
e to the defi obje tiv e the pr babili 8
ha c
s be impr en delegated nition and e
o 9
of what a ba ically tha ty of s
ov 10 ements to the lic v lner bility i t brings u
ar a
[ SLIDE. )e being m de i and what which ensee 11 s
a n respons plant 12 v lner bilityMR. FLACK:
e to that.
u a
13 Basic lly wit into thr the definiti a
hr 14 e, global e
ons espe t a
c ccidents definitio a tually can b c
to a 15 or being vulne equences which a lic categorized all ns s
e 16 v lner bilitir bilitie a
u or s ens which w ay the top si onsider ee a
s c
17 es x
all as ould sequ imply that a
18 Then gene ence rl a
s numerical e hav kind of all a
w as pla ts hav e tho 19 c iteria like Ise that definitio n
four.
r e
are n.
20 If it mor is hav e
vulner bility a
e 10 mentio specific abov e
1 to the
- ned, and s
10 use But mo :
minus fo to the ort of an s
which means tanalysis them if ur, it minus of or is not a
proces all in the framaking info s to define them it of ce it ork of whatrmation from th use some ew a
r du e
v ln u
e an lysis r bility a
regardles fixes e
a s
That of what might be do and u ing s
is basically the definiti ne that can mo t on is.
s of the plants ANN use RILEY some 1612 Cou & ASSOCIATES rt K Str Repo t
, LTD.
Washingto, N. W., er et r
e s
(202) n, D. C. Suite 300 293-395020006
-/'y-
-' y'p/'
i 400 j
1 something else, I guess.
I don't know if that answers your-1
()
2 question.
3 MR. KRESS:
Yes, it is.
i 4
[ SLIDE.]
5 MR. FLACK:
Finally, to reduce the probability of 6
core damage is the other objective and basically that. brings 7
us back to the definition of what a vulnerability is which 8
has been delegated to the licensee and what plant j
l 9
improvements are being made in response to that.
i l
10
[ SLIDE.]
11 MR. FLACK:
Basically with respect to a i
12 vulnerability, the definitions actually can be categorized 13 into three, global definitions which a licensee consider all l
i l
14 accidents or all sequences or say the top six sequences as 15 being vulnerabilities which would imply that all plants have 16 vulnerabilities as a general kind of definition.
17 Then we have those that are more specific and use 18 a numerical criteria like I'have mentioned, 10 to the minus I
i 19 four.
If it is above 10 to the minus four, it is a 20 vulnerability.
But most of them if not all of them use some l
21 sort of an analysis or process to define a vulnerability l
l 22 which means taking information from the analysis and using 23 it in the framework of what fixes might be done that can 24 reduce it regardless of what the definition is.
25 That is basically most of the plants use some
]
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,
Suite 300 Washington, D.C.
20006 (202) 293-3950 i
401 1
process in evaluating what the fix is which then is based on 2
cost as well as benefit.
3 For those that have defined vulnerabilities, I 4
guess one of the major areas which I mentioned before the 5
vulnerabilities lie in are in the dependencies in the. plant, 6
dependencies on HVAC or component cooling water whereas-7 some, like Surry, had an environmental effect which is 8
flooding which lead to a vulnerability so this is basically 9
from what we have seen as most people define a vulnerability 10 as something very significant stem from those areas.
11
[ SLIDE.]
12 MR. FLACK:
As far as improvements are concerned, 13 safety enhancements, I have listed them here, the ones that 14 we have been seeing performed on plants.
The first one, of 15 course, utilizing the fire water system for equipment 16 cooling and containment cooling is pretty widespread.
There 17 has been a source of water available at the plant and now 18 they are recognizing the benefit of having that water-19 available and utilizing it, for' example, in cooling charging
~
20 pumps, for example, to break that dependency of cooling on 21 the component cooling water system using fire water as well 22 as getting water into the containment on the BWRs through 23 either the drywell spray or the injection _ system.
24 Implementation of system and unit cross-ties is 25 again an area that they have benefited on by being able to ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,
Suite 300 Washington, D.C.
20006 (202) 293-3950
402 1
bring on power from either an adjacent unit or another bus 2
in order to pick up equipment is an area where.they have 3
taken credit and found improvements.
4 Protection against internal flood, making. walls 5
water tight, operator response to flooding and training is 6
an area that they are enhancing.
Adding alternate AC power 7
supplies, this could be adding in turbine generators or 8
black start diesel generators and so on is an area that they 9
have been improving in and upgrading DC power supplies by 10 extending battery life capacities from one to four hours, 11 for example, or putting in portable battery chargers.
12
[ SLIDE. ]
13 MR. FLACK:
Providing makeup to the reactor water 14 storage tank is another improvement that some licensees have
~
15 benefited on to extent out the time that injection is 16 available.
17 MR. DAVIS:
I think that is refueling water 18 storage tank.
19 MR. FLACK:
Refueling water storage tank.
What 20 did I say?
4 21 MR. DAVIS:
Reactor.
I 22 MR. FLACK:
Oh, yes, refueling water-storage tank.
]
.1 23 Enhancing RCP seal cooling and that is the one we were just 24 talking about where they used the fire water system to pass 25 through the charging pump coolers to cool them for once-ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,
Suite 300 Washington, D.C. 20006 (202) 293-3950'
i 403 1
through cooling and therefore be able to have some seal
(
2 injection during loss of component cooling water or service 3
water for that matter.
4 Enhancing room cooling capability,-this has been i
5 important in a number of plants where HVAC has been e
6 important, generally putting in procedures to open doors or 7
alarms that indicate-to the control room that there is loss l
8 of cooling in a room containing important electronic-l 9
equipment.
10 Reducing asymmetries or balancing buses, you may l
l 11 have the situation in some plants where the one bus is 12 loaded down with equipment and the other one is not so if F
13 you lose that one bus, you are taking down more than half a i
14 plant and so what they will do is to balance the loads or I
/,,
t
\\
15 reduce these asymmetries when they can.
I 16 MR. CARROLL:
That is specifically what you mean f
17 by that statement?
18 MR. FLACK:
Yes, balancing the buses.
19 MR. CARROLL:
Reducing asymmetries could mean a j
20 lot of things.
21 MR. FLACK:
Right, and adding diversity to DHR t
22 which would be like PWRs feed and bleed or depressurization 23 and using the condensate water.
So these are the types of
)
24 improvements we are seeing made not necessarily in response 25 to any vulnerabilities as defined but just means by which.
I
()
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,
Suite 300 l
Washington, D.C.
20006 (202) 293-3950
404 i
1 they are improving their plant's safety capacity.
j
)
2 MR. DAVIS:
John, there was one other one that I 3
found and I don't know how pervasive it is but there was 4
some changes made at Salem in the emergency operating 5
procedures due to a scenario where you have an RHR system I
6 failure and the operator could be misled into thinking he 7
had a LOCA inside the containment.
You may not recall that 8
one.
l 9
MR. FLACK:
I-have heard of the scenario.
I 10 didn't know it was Salem though.
I think that came up in a 11 simulator exercise, it was also flagged, where they thought
?
12 they had a LOCA inside containment and it turned out to be 13 an ISLOCA.
14 MR. DAVIS:
Yes.
15 MR. FLACK:
You are right, good.
We didn't get to 16 Salem yet though so we will be looking at that one.
17 MR. DAVIS:
Thank you.
18
[ SLIDE.]
19 MR. FLACK:
So general observations at;this time 20 with regard to the program is we have seen the benefits 21 exceeded our expectations.
We thought at the time based on l
22 our earlier discussions that licensees were going to do this 23 as a one-shot deal and they would only fix things unless-l 24 they becams defined as a vulnerability and we see a lot of 25 fixes and enhancements being'made and we see the evolution i
i ANN RILEY & ASSOCIATES, LTD.
)
Court Reporters i
1612 K Street, N.W.,
Suite 300 Washington, D.C.
20006 i
(202) 293-3950
+
I 405 1
of a living PRA quoted from somewhere.
My own definition is
[
h 2
just that they are planning on updating their PRAs at future
]
3 dates.
r 4
We know that the studies are very complex, that 5
they require a lot of resources.
Licensees are putting in 6
more resources than what was originally envisioned or-
.{
7 postulated in the beginning of the program, generally _two to 8
three million dollars is the going price with 50-percent of f
9 staff involved.
10 The assumptions that are being made in the i
11 analysis are very important-and they could drive the numbers-12 so they have to be understood to understand what_the number 13 represents.
So looking at the bottom lineEis not going to 14 tell you much-unless you understand what assumptions were f
15 made that made the bottom line what~it is.
16 As I have mentioned before, there were a number of 17 enhancement opportunities we have gone through that we are 18 seeing and there seems'to be a strong potential for future L
19 applications and licensers appear to be indicating that that 20 is exactly what they are planning, that there is going to'be 21 future applications of the IPE and PRA.
22 MR. CARROLL:
I think that is part of the 23 definition of a living PRA that you are using and not just i
24 keep it updated.
25 MR. FLACK:
Yes.
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1
1612 K Street, N.W.,
Suite 300 Washington, D.C.
20006 (202) 293-3950 i
i j
i 1
406 1
MR. SEALE:
Based on what you say it would seem 2
that a utility list like this might subsume several of those 3
comments in a bullet that said " enhanced investment 4
protection."
(
5 MR. FLACK:
They probably would.
6 MR. SEALE:
And that is why they are responding.
7 MR. FLACK:
Yes.
They would see it as a benefit 8
directly to their operations and costs.
9 MR. SEALE:
Yes.
10 MR. DAVIS:
Thank you, Dr. Flack.
Are there any i
11 questions?
3 12 MR. LINDBLAD:
Yes.
Dr. Flack, I picture 80 or so I
13 of these projects underway with all the licensees using 14 perhaps six to 12 engineers in each of their groups and 15 coming together with the submittals being read in detail by 16 six or seven NRC reviewers.
17 I guess my question is going to be how'are your 18 ideas and your review going to propagated through the NRC f
19 staff because I can picture a communications problem in the 20 future where licensees have' identified what they think are i
21 the important issues and many issuer. being identified as not 22 very safety significant but I can also picture.the NRC staif f
23 continuing to pursue certain areas that the licensees aren't 24 going to deem very important.
Will more people be absorbing 25 this information than just your review group?
^
i O
l l
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,
Suite 300 1
Washington, D.C.
20006 (202) 293-3950 1
i l
I
i 4
407 t
1 MR. FLACK:
My personal opinion is, I hope so.
()
2 MR. LINDBLAD:
The individual project managers, 3
are they very involved in reading what the licensees are
[
4 submitting?
f 5
MR. FLACK:
It varies.
Of course all project i
6 managers are required to now write a small synopsis of, very 7
brief, it is only a few pages, a synopsis of the' findings 8
that come out of the PRAs and we get copies of those.
So.
9 they do go a certain level.
Of course there are a lot of i
10 things on their plates constantly so how much they actually 11 get into the PRA itself and try to understand what it is all 12 about, I really don't know.
Some, of course, are more 13 fluent with it than others and it is going to vary from.
i 14 plant to plant.
15 How the staff uses it is a very interesting i
16 question and I think that is something that a-lot of people 17 are working on right now and all we can do at this point is 18 to try to document what our findings are and-we do that in i
19 an evaluation report which is about 15 or 20 pages.
l 20 MR. LINDBLAD:
I am sorry.
You said a lot of 21 people are working on it right now.
I thought previously 22 you told me the pool of people were six or seven.
23 MR. FLACK:
Working on coming to grips with'the i
24 IPE in general with respect to the regulatory process.
25 MR. CARROLL:
You focused on the project manager.
ANN RILEY & ASSOCIATES, LTD.
Court Reporters i
1612 K Street, N.W.,
Suite 300 Washington, D.C.
20006 (202) 293-3950 t
408 1
I guess my focus would be on the resident inspectors.
Are
()
2 you getting them or making efforts to force them to really 3
start getting a feel for PRA and its usefulness?
4 MR. FLACK:
To be honest with you, it is outside 5
our scope and I can't really respond to that but I do know 6
that the question has been raised before.
7 MR. CATTON:
You need a workshop or something.
8 MR. MURPHY:
If I can add a thought on this.. We 9
are just now as you will notice-that while we have a number 10 of IPEs in, we only have a relatively few on which we have' 11 written the SERs.
The SERs are drawing a lot of attention 12 in NRR.
They are being read by the appropriate management 13 level of the NRR and I think they are trying to understand 14 what it says.
15 We are trying to capture not only a statement that 16 they meet the requirements of the generic letter but also 17 the particular insights that are coming out of this and put 18 them in our SERs in a way that they are easily recognized by 19 somebody that doesn't have the time to really read the whole-20 thing.
21 Beyond that, of course, as you know there is a 22 major effort afoot right now in the area of PRA planning in 23 the agency largely in response to initiallyfthe PRA. working 24 group and the number of comments the committee has made over-25 the years and you have all seen the letter that was signed (f
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,
Suite 300 Washington, D.C.
20006 (202) 293-3950
l 409 1
by the four office directors.
(
2 We hope to have a plan that will be going forward l
3 for discussion at the senior management meeting in January l
4 and I believe it is scheduled to be discussed with the 5
committee in February.
At that time I think we will be able 6
to answer that.
That plan is still jelling but at that time 7
I think we can get into this a bit more.
8 Clearly there is an initiative with NUMARC that is 9
going on.
There was a meeting this week'but unfortunately I-10 was out of town so I can't relate what' happened at it but 11 there is a lot of activity going on and a lot of emphasis il going on.
I think you can be relatively sure that the 13 agency is going to' pay attention to what we are saying.
i 14 By the same token I have to say, is there_ going to C /}
s 15 be somebody in the agency that doesn't read it and doesn't 16 make use of the probabilistic insights that are coming from 17 it, the answer to that is probably yes but hopefully the-l 18 right levels of management will be attuned to give a proper l
19 response if the utilities find that we are not making use of.
f 20 the information that is provided to us.
l l
21 MR. CARROLL:
I think you are at a point where i
22 there is a wonderful opportunity to involve project. managers 23 and resident inspectors in the IPE reviews and'the results 24 of the reviews and really start making headway towards l
25 getting some PRA literate people throughout the agency.
l l
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street, N W.,
Suite 300
)
l Washington, D.C.
20006 (202) 293-3950
i 410 1
MR. MURPHY:
I was not surprised but happy to see
)
2 and where I was this last Monday was at a course on reactor 3
safety given at our technical training center at Chattanooga 1
i
)
4 and I went down to monitor.the first day of that class and i
5 there were several residents in the class.and they were very j
1 6
interested in how all of this affected them and how they i
7 should use it.
So I think that backs up what you are 8
saying.
9 MR. LINDBLAD:
I guess I would also be concerned 10 that the people involved in programmatic efforts look and 11 see if their continuation of the programmatic effort really 12 would have much effect on safety.
13 MR. CARROLL:
Yes.
14 MR. KRESS:
I hate to throw a damper on what I l
15 consider one of the nicer and finer things that NRC has'done 16 recently but it strikes me as being a program that is sort 17 of halfway done in that we never required a level 2 PRA and 18 only about half of the plants are.doing that.
19 I think that was a mistake from the start.-
I 20 think the level 2 is what would complete the job and make 21 the information a great deal more useful because I think you 22 can go from a level 2 to risk very quickly and very easily 23 and the reason for this comment is I think that ought to be 24 1, '. rf the things the staff ought to be thinking about, how 25
's we convert this to level 2 PRA without burdening the ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,
Suite 300 Washington, D.C.
20006 (202) 293-3950
l 411 1
individual plants.
, ()
2 I think maybe that might be a task the NRC itself 3
could think about taking on.
That would make them more 4
consistent and complete and useful.
Is there any thought 5
given to that aspect of it?
6 MR. MURPHY:
I am trying to think how to 7
characterize it, there is a rather simple method that we are 8
trying to develop now for use in regulatory analysis as part 9
of the regulatory analysis handbook.
10 If you recall, for many' years we had a table that 11 was developed out of WASH-1400 and it had the nine release i
12 categories from WASH-1400 and then it had person rem per 13 year for each one of those categories that was developed for-
~
14 the WASH-1400 sites.
15 When we did NUREG-1150 we had a lot more source 16 terms and we complicated matters immensely and it made it 17 difficult to pull something similar out of that directly.
18 We are well along in the process of. creating a'similar set 19 of tables for the 1150 plants.
20 What this will allow you to-do is to take 21 categories of accident sequences like station blackout and 22 see what kind of level 3 results you get at the five 1150 23 plants and perhaps we can be a little more generic about 24 that but I won't know until I see the results.
25 That is not quite what you are asking for but.it ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,
Suite 300 Washington, D.C.
20006 (202) 293-3950
l l
412 1
is a step along the way.
()
2 MR. KRESS:
It is getting close.
3 MR. MURPHY:
What it does is it gives me sort of a 4
gross characterization of what is important to risk and what j
5 isn't and that will be very helpful but I-don't know whether 6
I can parse it down at the level you want.
It will be a 7
step in that direction at least.
8 MR, DAVIS:
Is there anything else?
l 9
MR. LINDBLAD:
It-is a very good presentation.
?
10 MR. DAVIS:
Yes.
Thank you, Dr. Flack, very good.'
11 I think now we will move to the database program and hear 12 from Mary Drouin.
I 13
[ SLIDE.]
14 MS. DROUIN:
My name is Mary Drouin with the O
15 i
Office of Research and the Severe Accident Issues Branch.
I 16 am the section leader responsible for the IPE database and 17 potential insights program.
18
[ SLIDE.]
19 MS. DROUIN:
Today I am going to try to talk to
~
20 these two issues focusing in on the database, what 21 objectives we were trying to achieve with the' database and.
22 thereby, what structure we-had built for the database and 23 potentially where we are going to go with this database and 24 the insights that can thereby be derived.from this database.
1 25 MS. DROUIN:
The major point to understand about l ()
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,
Suite 300 Washington, D.C.
20006 (202) 293-3950-
413 1
this database is that we were just not' arbitrarily just' f)\\'
2 trying to collect information from the different IPEs and
(_
3 throw them into these files on a database.
4 We really wanted some foresight here in'the sense 5
that what we were trying to do is understand plant features l
l 6
and how they could potentially impact core damage frequency l
7 perhaps, how they could impact containment performance.
l 8
So we were looking at if.you go through.and you 9
look at these IPEs and you start seeing class of plants 10 where the results the same or.where the'results are 11 different and trying to understand that. ~So.we wanted to 1
12 build a database that was going to capture that type of 13 information so that when we went through and looked at it we 14 could come in and do that type of examination.
15
[ SLIDE.]
16 MS. DROUIN:
So we wanted to be able to look at 17 plant design information, core damage frequency information 18 and containment performance.
In order to do this we needed-t 19 to have a database that was going to gather the information 20 in a very structured and uniform manner and that.was also 21 going to look at this information across plants and within t
22 plants.
23 So we have these files here which I am just.trying 24 to pictorially show you that as we gather dependence 25 information, we gather information on success criteria, on'
)
ANN RILEY & ASSOCIATES,.LTD.
Court Reporters 1612 K Street, N.W.,
Suite 300 Washington, D.C.
20006 (202) 293-3950
414 l
1 accident sequences, on the back-end information,-that these
()
2 files are cross-tied within the plants and across the 3
plants.
This is the real beauty of_this database.
4
[ SLIDE.]
5 MS. DROUIN:
Once the structure was established, 6
the real heart of the database is going through 1
7 systematically each of the IPEs and extrapolating this t
8 information and putting them into these files.
So we are 9
_ going through and gathering all thel system information cx1 10 all the front-lines and the support systems that are in the i
l 11 IPE.
12 We are going through the same thing'on the 13 dependencies, looking at the_ front-line and the support-i 14 systems so that you are looking at supports;to the front.
O 15 line and supports to supports in looking at the 16 dependencies.
17 Sequence information that we are gathering from 18 the IPEs, we are looking at both the success paths-and-the 19 failure paths.
We are looking at the causes and also the 20 frequencies.
Success criteria information, we are looking 21 at from a functional perspective and a systemic perspective l
22 what the strategies are for core damage prevention.
23 Then on the back-end, we are going from the plant 24 damage state to releases.
We are gathering-information on
(
25 how they are defining their plant damage state parameters.
I ANN RILEY & ASSOCIATES,.LTD.
Court Reporters 1612 K Street, N.W.,
Suite 300 Washington, D.C.
20006 (202) 293-3950
)
I l
415 1
We are building a containment matrix in the database and
)
2 looking at the level 2 analysis parameters and source term 3
information.
j 4
MR. WILKINS:
Excuse me.
With reference to the l
5 bullet at the top of the right there on core damage 6
prevention, do you consider core damage mitigation at all?
l i
7 MS. DROUIN:
In this one here?
8 MR. WILKINS:
Yes.
9 MS. DROUIN:
No.
This is still on the level 1.
j 10 MR. WILKINS:
All right.
l 11 MS. DROUIN:
This is still-on the level 1.
12 MR. LEWIS:
You can't prevent core damage.
13
[ SLIDE. ]
14 MS. DROUIN:
I juot want to show you a' couple and t
[_l k-15 I am not going to go through all the fields and records.but 16 I just want to show you a couple to give you a kind of a 17 sense of the depth-and breadth of the information that is.
18 being collected in the database.
19 Here is an example of a data sheet that will be 20 filled out and from that data sheet, this is what is entered
[
21 into the database.
Here is for a BWR system.
We will go i
22 through and pull out from the IPE the differentLfunctions.
23 For each of those functions, we will identify the different
-i 24 systems that have been identified to support that function.
f 25 Now when I put " system" here, we have normalized
(
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,
Suite 300-Washington, D.C.
20006 (202) 293-3950 l
l
i 416
^
1 this to our nomenclature so every where you see ARI or SLC,
)
()
2 it always means the same thing in our terminology.
We will 3
also put in what the plant specific nomenclature is.
4 This is an important part of the database because 5
everybody calls something differently.
You might have the 6
same thing but they all call it something different.
So 7
when we have come in and defined things in the database, we 8
have tried to normalize it to our definition but we haven't 9
changed their information though.
10 We will look at how many trains are in the system, 11 the number.
If these systems are being used for other 12 functicas, then this will be defined over here, whether it 13 has the potential to cross-tie to another unit and so this 14 is just a high level example of the information that is 15 being brought in on the system information.
16
[ SLIDE.]
17 MS, DROUIN:
Here-is an example-of the level of 18 information for dependencies.
This is for support systems, t
19 All the support systems that are modeled in the IPE, the 20 functions will be entered in.
One, two,-three,-four, five, 21 six, I didn't have room to put in the system names here.
.i 22 There just wasn't enough space but this will be the 23 different systems.
24 Then into each of the blocks will be the type of 25 dependency-that you see in the plant..Whether.it is an ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,
Suite 300 Washington, D.C.
20006 (202) 293-3950 l.
\\
1 417 l
1 absolute dependency, a primary or secondary or a partial 2
dependency is the type of information we are attempting to 3-capture.
~
4 MR. KRESS:
On that item labeled "D" at the 5
bottom, will.you attempt to put in what the beta' factors 6
were for that particular system?
4 7
MS. DROUIN:
No.
This is just noting.
8 MR. KRESS:
Just noting.
4 9
MR. WILK',
3:
You decided evidently that it wasn't 10 necessary to have a fifth category which I would call 11
" irrelevant."
Let me just pursue that.for a moment if'I 12 may.
13 MS. DROUIN:
Okay.
14 MR. WILKINS:
If you had such.a category that 15 would at least give some assurance that.you had asked-the 16 question.
e 17 MS. DROUIN:
That's right.
+
18 MR. WILKINS:
If you don't'have the' category and 19 something is missing, you are not quite-sure whether it is l
20 missing because you forgot about it.
21 MS. DROUIN:
Good point.
I 22 MR. WILKINS:
I don't know how-important a point 23 it is.
I 24 MS. DROUIN:
We could always go back.
On all of I
25 these we have space for notes or comments and if there is an j
i i
ANN RILEY & ASSOCIATES, LTD.
Court Reporters i
1612 K Street, N.W.,
Suite 300 Washington, D.C.
20006 (202) 293-3950 i
F
~,,
-w
4 i
s
]
418 i
1 interesting observation that the analyst sees, there is the 2
space for that and that might be something to consider, 5
3 putting it in there for indication of completeness.
4
[ SLIDE.]
5 MS. DROUIN:
At the seque11ce leve? each sequence 6
is entered, what potential plant damage state it got binned l
7 into, its core damage frequency, what initiators associated 8
with that sequence, what lost supports, what failed 9
functions, causes, attributes and functions / systems.
10 on each of these, here is the information again 11 that the analyst will use in entering on~each of these 12 columns and rows.
Again, the point of this is to normalize 13 it so that when you go to any single plant'when you read 14 something you can always make the same interpretation from O
15 the nomenclature in the information there.
16 So for each of these fields and records there will 17 always be a list of choices that you can select in entering 18 that information.
19 MR. KRESS:
Will you use the WASH-1400 to identify i
20 the sequences, the nomenclature there?
i J
21 MS. DROUIN:
Did we consciously think about WASH-
]
22 1400?
23 MR. KRESS:
No.
I am talking about.just 24 identifying which sequences.
Will it be standardized?
25 MS. DROUIN:
The sequences are the sequences that' ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,
Suite 300 Washington, D.C.
20006 (202) 293-3950
^
i 419' 1
are in the IPE but the nomenclature we use to identify the 2
sequence in terms of describing it --
3 MR. KRESS:
That is what I meant.
4 MS. DROUIN:
But each sequence is-the sequence as 5
modeled in the IPE.
6
[ SLIDE ]
7 MS. DROUIN:
Here is an example on'the back-end, f
~
8 one of the fields of information that is being entered into 9
the database for each release class what the containment i
10 failure mode is, the failure location, your containment 11 failure size, your oxidation in vessel, your_ amount of core.
12 in your core concrete interaction, your disposition, your 13 vessel failure mode, whether or not you have suppression 14 pool by-pass and this is a BWR example, sprays available and-(~l) 15 whether or not the analysis gave any credit for i
16 decontamination.
17 MR. KRESS:
How will you arrive at a value for 18 that containment failure size?
19 MS. DROUIN:
We don't derive it..We pull it from 20 the IPE, so if the IPE, whatever the value they assessed for 21 it.
22
[ SLIDE.]
l 23 MS. DROUIN:
Status of the database, it has been I
24 built with DBaseIV.
Right now we have about 45 IPEs in the 25 database.
I think we have 13 BWRs and 25 PWRs, somewhere i
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,
Suite 300 l
Washington, D.C.
20006 (202) 293-3950
---a.
w w-,
+
y-4 o
..m e,,~g y
r
-mys,
420-1 along that order.
()
2 It is covering all reactor types and all 3
containment types.
So right now we feel that we'have a real 4
good representation where we can start standing back and 5
trying to understand maybe what some of this-information is 6
telling us in looking for some significant insights.
I 7
The database is being expanded to include IPEEE 8
information and our goal ultimately when the~IPEEEs are done.
9 is that we have now a single database that captures the full t
10 risk at least at full power.
So it will include the 11 internal and external events in the database.
12 We are also currently developing a users manual.
I 13 There is a draft manual that is undergoing review and' i
14 comment by the staff and we anticipate having the user.
15 manual ready for distribution probably sometime in-mid-i 16 Spring.
I 17 MR. WILKINS:
Who do you see your principal users t
18 as being?
19 MS. DROUIN:
Excuse me?
20 MR. WILKINS:
Who do you think will be the l
21 principal users of the database?
i i
22 MS. DROUIN:
Well, we have gotten amounts of 23 requests for the database use from all the offices within l
l 24 NRC.
We are getting requests-from outside of the agency for 1
25 the database.
A primary single user, I hope everybody uses l
()
ANN RILEY & ASSOCIATES, LTD.
. Court Reporters i
1612 K Street, N.W.,
Suite 300
}
Washington, D.C.
20006 (202) 293-3950 i
i
~...,
+
l l
l
.421 l
t 1
it.
)
2 MR. WILKINS:
That is a good answer.
There is 3
nothing wrong with that answer.
i 4
MS. DROUIN:
The important part about.this users 5
manual is what we are attempting to do is to be able to-walk
[
6 the user through and not just show him what the files are.
7 but actually show him how to use it in setting up queries 8
and being able to ask questions and look at-this-and e
9 determine what kind of insights can be derived.
10 MR. WILKINS:
Will it extend to instructions that I
i 11 tell the user how to extract-information and sort it in the 12 way that he wants it sorted'or will he have to do that-13 separately?
14 MS. DROUIN:
We will have some examples in there O
15 for him but I can't potentially think of every single i
16 permutation.
[
(
17 MR. WILKINS:
I know you.can't.
f 18 MS. DROUIN:
But we do plan to have some examples r
19 and some example queries to show them as an aide.
i 20 MR. WILKINS:
Here you are about to address ~that 21 right now, aren't you?
22
[ SLIDE.]
23 MS. DROUIN:
This is an example of some of the 24 information that you can get out of the database and what we
{
l 25 are going to be pursuing in'looking at the insights and-i
\\
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,
Suite 300 Washington, D.C.
20006 (202) 293-3950 t
ic
+
422 1
trying to understand what this information means.
2 If we look at the database with the 45 plants that 3
are in there right now we can see from the database what 4
mean core damage frequency based on 45 plants and we also 5
have extrapolated a median value.
l 6
Now we can start asking ourselves questions.
Why l
7 is this the way it is?
Are these differences being driven 8
by design?
Are they being driven by data?
Are they being 9
driven by the assumptions?
Is there a single class of l
10 plants?
When I look at the differences between the PWRs, is l
l 11 this inherently just because of the differences between-the
(
12 "P's" and the "B's?"
Is it different because of the data?
l 13 When I start taking these values and breaking them i
14 and looking at them in terms of the different classes of 15 plants, am I going to see the same mean for a general set of 16 plants?
If there are differences or commonalties, why.are 17 they there?
Are they there by cc acidence or are they there 18 by design?
19 These are the types of things that we are going to 20 be looking at over the next year as we now have sufficient 21 information in the database.
The same thing on the back-22 end, when this is just an example of early containment.
23 failure.
24 MR. KRESS:
Are these real data that you have 25 extracted from the IPEs?
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,
Suite 300 Washington, D.C.
20006 (202) 293-3950
I j
423 r
1 MS. DROUIN:
Yes.
This is their data.
)
2 MR. KRESS:
This is real?
{
3 MS. DROUIN:
This is real.
This is what the l
4 industry in a sense is telling us.
We haven't done anything 5
to change these numbers.
i 6
MR. KRESS:
These are not just your idea of what I
7 you are going to get out?
It is actual real data.
l 8
MS. DROUIN:
No.
This is actual information.
So
[
9 now we are going to cry to understand what this information I
I 10 means and when you start looking at the early containment-11 failure probabilities, you are seeing if you look over here 12 on the PWRs, what are we seeing, an order, a half difference 13 in magnitude across the plants of these 25 PWRs?
14 MR. DAVIS:
Ma ry ', that particular one is of I
l 15 concern to me.
I noticed as I mentioned'in our meeting on l
l 16 the 18th that there is quite a distribution of containment 17 failure numbers and I became convinced that part of the l
18 problem is because of how it is defined by the-various 19 submitters.
I 20 In fact for FitzPatrick as I recall they actually.
21 defined a failure as if they deliberately vent the i
22 containment and that is certainly a different kind of 23 failure than an early over-pressure failure-in terms of r
I 24 consequences.
25 MS. DROUIN:
That's correct.
f
()
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,
Suite 300 Washington, D.C.
20006 (202) 293-3950
424 l
1 MR. DAVIS:
In fact, some used basemat melt-
)
2 through as a containment failure whereas that would not be a 3
particular concern as far as public health and safety or it 4
may not be.
So I think there needs to be a further 5
dissection of the containment failure information.
6 MS. DROUIN:
Oh, yes.
This particular graph here 7
i r
8 MR. DAVIS:
This could be misleading.
You might i
9 find a plant that has a high fraction of containme nt 10 failures but very low risk anyway.
11 MS. DROUIN:
Yes.
12 MR. KRESS:
That is the reason I asked my question 13 as to whether it was real because it doesn't agree with what 14 was found in NUREG-1150 particularly for the containment 15 failure part of it.
s-
{
16 MS. DROUIN:
That is correct.
l 17 MR. KRESS:
I had the same comment that Pete had.
i l
18 MS. DROUIN:
This is just all of the plants put 19 together.
We haven't tried to dissect it yet.
We are l
20 starting to do that.
21 MR. LINDBLAD:
But that is one of the things to be 22 focused on.
Part of the disparity is different 23 professionals developing the judgments, in-some cases 25 PWR 24 teams looking at things differently and the focus has to be~
25 on this review by the NRC staff is basically just reading j
()
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,
Suite 300 Washington, D.C.
20006 (202) 293-3950 1
i
425 l
1 them and not normalizing them.
There has been no effort to f
)
2 normalize them to a common set of rules.
3 MS. DROUIN:
No, but that is what we are going to j
4 attempt to do here is to go and look at this and see where l
I 5
the differences are and when you do norma'ize them, now what 6
do you see.
7 MR. SEALE:
In the words that you have given us
{
8 which have set forth the details of the information sorting-
.l l
l 9
in your database, will the differences of the containment i
10 failure mode, say basemat melt-through versus deliberate.
}
11
. venting versus overpressure failure, will those. differences.
I 12 be in the database as you enter the data or is that l
I 13 something that somebody will have to come back and recover 1
14 from the individual IPE submittals after the fact.
'l
\\
+
15 MS. DROUIN:
Yes.
16 MR. SEALE:
I said "or."
j 17
[ Laughter.]
i 18 MS. DROUIN:
We are attempting to do both in 19 parallel right now.
4 20 MR. SEALE:
But you hope to have that distinction a
21 in the database so it won't be necessary to go back.to
{
22 scratch.
23 MS. DROUIN:
That's right.
Currently we are in 24 the process right now.
25 MR. SEALE:
To try to recognize those nuances.
4 1
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,
Suite 300 3;
Washington, D.C.
20006 (202) 293-3950 l
t 426 l
1 MS. DROUIN:
Yes, and correcting that but I can't j
()
2 say that that is completely done.
It is not an easy thing
[
3 to do and sometimes information is not there and we just 4
have to go with it.
5
[ SLIDE.]
l 6
MS. DROUIN:
I think we have probably talked about l
7 this slide already but basically we are going to be looking I
l 8
at this information and trying to understand what is the 9
importance of these plant features when you look at your l
t 10 core damage frequency and you look at your containment l
l 11 performance.
What is driving it?
Is it data?
Is it the i
12 boundary conditions?
Is it the different methods?
Is it 13 the different assumptions?
These are the things that we are 14 going to try to pursue and understand as we move forward and 7)
(
I 15 to try to understand are things being driven by a single i
1 16 plant?
17 Is there just an outlier out there?
Can we say j
18 that there is a global set of insights for a single class of 19 plants or a multitude of different classes of plants?- These 20 are the things that we are going to try and understand and 21 again ultimately one of the things that the database can be' i
22 used for is in regulation.
r 23
[ SLIDE.]
I 24 MS. DROUIN:
The final status of where we are, as 25 I pointed out we have 45 IPEs entered into the database.
We 1
t O
ANN RILEY & ASSOCIATES, LTD.
Court Reporters j
1612 K Street, N.W.,
Suite 300 i
Washington, D.C.
20006 l
(202) 293-3950
427 i
have all reactor and containment types so you will look in
()
2 it and there will be a large dry, there will be an ice 3
condenser, there is your MARK I, you MARK II and MARK III, I 4
believe.
5 We have drafted our program assumptions for 6
contractor assistan:e and that is in the concurrence chain' 7
right now.
8 We do anticipate having an initial draft report in 9
the fall of 1994 looking at the global examinations for 10 significant safety insights and looking at some of these 11 plant improvements or these insights and how they impact the 12 core damage frequency and containment _ performance and 13 perhaps how you can use these IPEs'in the regulatory i
14 process.
i 15 So right now we are planning on having a draft 16 report in the fall of 1994 and then in 1995 issuing some 17 type of NUREG and then in 1996 to include the IPEEEs and a 18 final report in 1998.
19 MR. DAVIS:
Thank you.
20 MS. DROUIN:
I kind of rushed through this rather 21 quickly.
22 MR. WILKINS:
What is the likelihood that by 1998 23 some of these original IPEs will have been replaced by j
second passes through the process?
24 25 MS. DROUIN:
I am sorry.
I didn't hear your i
l ()
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,
Suite 300 Washington, D.C.
20006 (202) 293-3950 l
428 1
1 question.
(
2 MR. WILKINS:
You have laid out a schedule that j
?
3 goes quite a way in the future and we have talked about 4
these IPEs as being living documents.
My question was L
5 basically the IPEs that you have gotten~in 1993 may be if 6
not obsolete, perhaps they have even been replaced, that is, 7
the utility will have done another one, a completely fresh
[
8 one, in that timetable.
f 9
MS. DROUIN:
Yes.
l l
10 MR. WILKINS:
What will your 1998 report address, 11 the new one or the old one?
l 12 MS. DROUIN:
Somewhat both.
The reason I answer 13 it that is because, of course, it will address the IPEs as 14 they are now.
One of the things we do want to try to.look O
15 at is what impact the plant improvements that have been.
16 identified by the licensees, what impact that would have on i
17 your damage frequencies and your containment performance.
I i
18 Now I am making somewhat a leap of faith here 19 thinking that the plant as it changes is going.toiimplement 20 these plant improvements that they have identified. 'So we 21 are going and assess what impact that would have and so'that l
22 would also be in this document.
Now the plant could do 23 something entirely different.
These are issues that we will 24 have to consider and look at.
25 MR. DAVIS:
This is a living database.
I i
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W, Suite 300 Washington, D.C.
20006 (202) 293-3950
~... _ _,
~.
t 429 1
MS. DROUIN:
This is living database, yes, it
)
2 is.
+
3 MR. CARROLL:
What mechanism do you have to get 4
the plants to provide you with this updated information?
5 MS. DROUIN:
I am talking about using the plant 6
improvements that they have right now identified in their 7
IPEs.
8 MR. CARROLL:
I think Ernest was asking a little i
9 broader question.
If they re-do their IPEs, do you have a 10 mechanism of capturing that in your database?
11 MS. DROUIN:
We are going to have to pursue'that.
12 MR. CARROLL:
All right.
Let me add just a little bit.
This is 14 Charlie Ader from staff.
When we are reviewing and we are
\\
15 finding that now, we have gotten an IPE in six months or.a 16 year ago and as we review them and we ask questions, 17 sometimes licensees have come back.and said that since we 18 sent you the original one we have implemented this or we 19 have decided not to implement this or we have updated it.
20 So we get some additional updated information back 21 through the review process and a next step with the database 22 would be to go back and try to input or revise the 23 information,
we receive it.
But at this point in time and i
24 Mary said that we may want to pursue it.
j 25 MR. WILKINS:
You are saying that it is your.
'l i
l 70RJ RILEY & ASSOCIATES, LTD.
Court Reporters 1
1612 K Street, N.W.,
Suite 300 i
Washington, D.C.
20006 l
(202) 293-3950 l
)
t 430 i
l 1
intention to do that but it is a little early to say'that it
()
2 has been done.
1 8
3 MR. ADER:
No.
As we ask questions on a submittal i
l 4
during our review, we do in some cases get updated l
5 information and we will go back then and input that into the 6
database.
7 MR. CARROLL:
But five years from now.
8 MR. ADER:
That is a good point.
9 MR. LINDBLAD:
But once they publish it there will l
10 be letters to the editor from the licensees who will send' i
11 corrections if they stick out like a sore thumb.
l 12 MR. WILKINS:
The ones at the top of the list, for 13 example?
14 MR. LINDBLAD:
Yes.
15 MR. CARROLL:
It is a shame that that happens but 16 it will.
17 MS. DROUIN:
There is just one thing that I might 18 add.
We are focusing in nn this 1998' data and that is l
19 really to capture the IPEEE but we are intending on l
l 20 publishing something in 1995.
l l
I 21 Now granted even in 1995 there could be some IPEs 22 that are out of date who made significant amount of changes l
l 23 to their plant that maybe their IPE as submitted doesn't 24 represent.
25 MR. WILKINS:
We don't need to belabor this point.
1 ANN RILEY & ASSOCIATES, LTD.
Court Reporters-1612 K Street, N.W.,
Suite 300 Washington,.D.C. 20006 (202) 293-3950
l i
i 431 i
1 MS. DROUIN:
The only point I want to make is that
()
2 there will be a NUREG issued before 1998.
This will be a j
3 revision.
4 MR. CARROLL:
Pete, when we discussed this subject 5
and concluded we ought to have a status update,-one of the 6
issues that. prompted us to ask for the update was.the fact 7
that we are seeing a lot of generic issues being resolved by 8
Minners and his Merry Band as, "Oh, this is taken care of by 9
the IPE."
10 MR. DAVIS:
That is the subject of the next 11 presentation.
12 MR. CARROLL:
Ah, hah!
13 MS. DROUIN:
What a great lead-in!
14 MR. DAVIS:
Thank you, Mary.
That was very good.
15 MS. DROUIN:
Thank you very much.
16 MR. DAVIS:
If there are no other questions, why 17 don't we move on to the next subject which is the generic 18 issue resolution.
19 MR. CARROLL:
I am a good straight man.
20 MR. DAVIS:
Dr. Ader.
21 MR. WILKINS:
I really didn't want to say this to 22 you particularly but this is about the fifth one now and I 23 thought I would say something to somebody.
Please when you 24 prepare these, put your name on them and a cover page and a 25 title and a date.
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,
Suite 300 Washington, D.C.
20006 (202) 293-3950
t 432 l
t l
1 MR. CARROLL:
And page numbers.
We did put out a 2
document not too long ago that explained how we would like 3
to see handouts prepared.
1 4
MR. WILKINS:
I don't want you to' feel particular 5
badly because we are saying this to you.
You just happened 6
to be there when I erupted on this particular point.
7 MR. ADER:
I plead guilty.
The page numbers are t
8 on there.
These slides I took from the last presentation t
9 where we had a cover on them and since I wasn't giving an i
10 overview I took the cover off.
11 MR. WILKINS:
That may not even be a bad answer.
12 MR. CARROLL:
You do have page numbers, just not 13 on page one.
14 MR. ADER:
It is from page two on.
i O 15
[ SLIDE.]
I i
16 MR. ADER:
I guess being part of Minners' Merry 17 Band I should be up here.
l r
18 MS. DROUIN:
Which one are you?
l 19
[ Laughter. ]
l 20 MR. ADER:
The issue has come up, it came up in a i
21 couple of letters that the Committee has sent on the generic i
22 issue resolution process in discussions about how they are 23 being resolved in the IPEs or not being' resolved in the 24 IPEs.
25 One point I want to make up front on this and this
! ("%
(,)
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,
Suite 300 Washington, D.C.
20006 (202) 293-3950 f
I
[
t l'
433 II i
1 is going to be a difficult discussion for me because we are 2
talking a lot of generalities and each of you may have a i
3 particular issue in mind and I may have a different issue in 4
mind and we may characterize them differently and that is-5 one of the problems we have.
l 6
But in the resolution process especially and in j
7 talking the IPEs and more, the IPEEEs, there have been l
8 studies done that looked at the risk from these issues.
I 9
was involved on the tail end of several-of.them, the 153 j
10 service water; 105, the ISLOCA issue and there were risk 11 studies done and they concluded that the risk, tnat there 12 was no generic risk.
13 They could not find a generic problem that had a i
14 generic fix and the issues were resolved on the risk 15 significance.
But in that resolution process, staff i
16 recognized that there is this IPE process out there that was 17 put in place for licensees to look for vulnerabilities in i
i 18 their plants.
I 19 In fact, they are doing PRAs but initially it was i
20 envisioned they could do something other than_a PRA but they
?
21 are doing at least level 1 PRAs, some of them level 2 and
{
i 22 some of them even into level 3.
23 Some of.the people doing the resolutions didn't 24 want to ignore that work that was ongoing and on any generic i
25 issue there is always potential for a very. unique plant i
I t
ANN RILEY &' ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,
Suite 300 Washington, D.C, 20006 (202) 293-3950 l
434 1
specific configuration.
A(,)
2 That was the premise of why the IPE and the IPEEE 3
program were started to begin with back out of the Severe 4
Accident Policy Statement, the recognition that the plants, 5
the view was they do not pose an undue risk.
However, there 6
may be a unigye configuration.
7 So staff in the resolution process based on the 8
generic work they did and not finding-anything that f
9 warranted a generic backfit, not finding any unique 10 configuration that they could go out and look at a l
11 particule* lant that had that unique one would resolve an
[
12 issue wiu.
_'urther staff action and maybe we got caught r
13 because we also wanted to take' credit to provide some 14 additional assurance.
I O
i 15 Here is this IPE program going out there that l
16 should catch anything major that would be a vulnerability on 17 a particular issue like service water.
DC power was one of 18 the other ones I wasn't involved in but I'went back and 39 locked through some of the resolutions.
Intersystem LOCA-j 20 was specifically called out in the IPE process.
l 21 MR. CARROLL:
That is an interesting perspective l
l 22 on it but I guess my perspective is that, for_ example, on 4
23 152 we suggested that no additional funding be spent on 1
24 looking at that issue, i
25 MR. ADER:
I am sorry, which issue?
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,
Suite 300-Washington, D.C.
20006 (202) 293-3950
.t i
435' l
1 MR. CARROLL:
The service water, 152.
f
. ()
2 MR. ADER:
Service water was 153.
3 MR. CARROLL:
Okay, 153 and the answer we got is, e'
4 "Oh, no, we can't do it that way because the utilities'are 5
doing the IPEs and they will all be different and we can't l
')
6 really make anything out of it."
So you did spend the.
}
l 7
taxpayers money on some useless studies on service water.
+
8 At least I thought they were given that any decent PRL is 9
going to really look at service water in very close detail.
10 Perhaps it is merely a difference of perspective, Charlie.
]
t 11 MR. ADER:
As I said, I came in'at the tail end'of.
[
12 those resolutions so I came in as that one was being written 13 up.
14 MR. CARROLL:
People argued with us strongly that i
s C-15 IPE wasn't the way to look'at it and the other point we-made 1
16 and the one you made is service water systems are so plant j
17 specific you just can't look at two or three and draw any i
18 real conclusions and after all, you are doing an IPE and
~
19 that is where vou are going to find out about the plant 20 specific things.
All right.
21 MR. ADER:
Again, not having the full history on i
22 it, I came in at the resolution process and I won't take l
23 credit or blame or what have you.
24 But basically the IPE,'the generic letter and the
)
25 framework we are working under in the review process, it was i
~
ANN RILEY & ASSOCIATES, LTD.
Court Reporters i
1612 K Street, N.W.,
Suite 300 Washington, D.C.
20006 (202) 293-3950 i
436 1
issued in 1988 in generic letter 88-20 and then finally the j
()
2 licensees were told in 1989 to proceed with their IPE.
3 In that generic letter they were required to l
4 addres:- one issue which was USI A-45, the shutdown risk from 5
internTl events.
They were also to look at~ internal 6
flooding which had come out of some other issues but'A-45 i
7 was the only one that was specifically called out to be 8
addressed.
.9 TLay weren't required to address any other generic 10 issues.
They were given the option if they wanted to'to f
11 propose a resolution of a generic issue the staff was 3
12 working on.
Several licensees have done that.
Issue 153 13 was one that a licensee proposed they thought their IPE did i
14 resolve.
But most of them are not proposing resolutions of 15 specific issues.
16 Where they have though such as 153, we did review i
17 that issue in more detail-but it generally required an 18 additional staff person would get involved and we would send i
19 out an additional set of questions to focus in much more.on.
20 the details than we do on the general IPE review.
21 As several of you mentioned, as part of.an IPE, as 1
22 part of doing a PRA, licensees are. going to look at those i
23 things that were kind of the basis for a lot of these' safety 24 issues.
On service water, theyfare going to look at system 25 dependencies, as a front-line system dependent on DC power rs r
Id ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,
Suite 300 Washington, D.C.
20006 (202) 293-3950
4 437 4
1 or AC power and service water.
Is there some ventilation 2
dependencies?
3 They look for plant specific initiators and try'to 4
model them in.
They will look at their own data to see-
.j i
^
5 whether there is something unique they should model in.
i i
6 They will look to see if there is common cause that they a-7 have had at the plant or they will look through the generic i
8 data and model in common cause.
They look at operator 9
actions and their impacts.
l 10 Now they do it in different ways and to varying _
11 degrees but as we review it, those are the types of things l
12 we would also look fe to see are they doing a decent i
13 examination for plant specific vulnerabilities.
14 As I said, at least the issues that I have seen at f
O, l
15 the tail end and in talking to some'of the people who have 16 looked at them and prioritized them, I think people have 17 recognized what in the IPE process the licensees are 18 required to do and Warren has said that he generally wants 19 someone to go to the generic letter or the submittal l
20 guidance which is NUREG-1335 and identify a place where a 21 licensee would be required to look at the type of issue i
22 before they are resolved with no further action.
23
[ SLIDE.]
i 24 MR. ADER:
Briefly, we have not talked about the.
)
25 IPEEE this morning but I know that was also one'of the i
ANN RILEY &' ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,
Suite 300 Washington, D.C.
20006 (202) 293-3950 i
438 l
1 committee's concerns and briefly the process is the same for 2
l 3
The generic letter that initiated IPEEE was issued i
L 4
in 1991.
There was a NUREG that went with it.
In the case 5
of the IPEEE there were two issues that. licensees were told l
6 to specifically address, again, A-45 for external' events and i
1
+
1 7
8 GI-131 which was and I don't remember the exact title, it-9 was in-core instrumentation issues.
10 In addition, they were also requested to'look at
[
11 the findings from the Sandia-siting study which were the 12 fire / seismic interaction and that was really the basis of 3
13 generic issue 57.
We were down earlier in'the year on the 14 resolution of that.
That is-one that is specifically called
()
15 out to look at the findings out of the NUREG; j
l r
l 16 Again the prioritizations, the ones that havefcome 17 through recently, a lot of them were old SEP issues that f
18 said people should look at flooding or should look at 19 seismic events, should look at industrial' hazard,.should 20 look at high winds.
These things are the basis of the.IPEEE i
l 21 and they are being reviewed under the context of the generic 22 letter.
{
23
[ SLIDE.]
24 MR. ADER:
As far as tracking, I think the 25 committee is familiar with the process.
When an issue is in I
k.
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,
Suite.300 Washington, D.C.
20006 (202) 293-3950 1
439 1
a resolution stage, it is tracked through the SIMS or GIMICS
()
2 as a subset of that for the high, medium and r.early resolved 3
issues and that tracks it through to a resolution.
4 Once'it is resolved it is documented in 0933, the 5
basis of the resolution.
In addition, when: issues are 6
prioritized low or drop the basis of the prioritization is l
7 documented in 0933.
8 Periodically, low issues are revisited to see if 9
there is any new information coming from whatever source 10 that would say that these issues ought to be elevated and 11 reprioritized or resolved.
12 So there is this process in place that will track-13 and document issues.
I think the committee has-had a 14 concern that some of the generic issues, I guess in my mind ~
' b,)l
\\-
15 there has been some confusion at least with the ones that I i
16 have been involved in when we said that a piece of the 17 problem is being taken care of in the IPE reviews'or the 18 IPEEE reviews, we were taking credit for an existing program 19 knowing the level of review that we go into.
It was'not 20 envisioned that we would do a more 21 in-depth review.
l 22 MR. DAVIS:
Thank you.
Any questions?
23 MR. MICHELSON:
Yes.
I have a question.
One of 24 the issues of particular interest to us is A-17 which deals l
25 with system interactive effects and in the case of the
)
ANN RILEY & ASSOCIATES, LTD.
i Court Reporters 1612 K Street, N.W.,
Suite 300 Washington, D.C.
20006 l
(202) 293-3950 1
l 440 i
j 1
boiling water reactors when your licensees are looking at
()
2 pipe breaks particularly in the reactor water clean-up 3
system, are they looking at such things or going back to 4
review their calculations on things like sub-compartment
+
5 pressurization and so forth to see first of all are they up-6 to-date in terms of current knowledge of the problem and are 7
they looking at their plant from that particular viewpoint 8
because it is very important in the PRA to make-sure that 9
you get the extent of the event.
I 10 MR. MICHELSON:
Unfortunately the staff member, 11 Dr. Burdick, who was involved in the resolution of generic 12 issue 105.who was here earlier had'to leave.
That was 13 looked at in 105, the reactor water clean-up system and the 14 environmental conditions.
i 15 MR. MICHELSON:
GE is looking at it very carefully 16 for ABWR and finding some very interesting things.
When you 17 sit down and you calculate it correctly and you look at'what 18 happens, indeed you get very high pressures, you blow doors i
19 off, you do things like that and the environment, in fact,
20 extends throughout the secondary containment of the ABWR.
21 So now we are anding up qualifying everything 22 inside of secondary containment for 15 pounds and 248 23 degrees Fahrenheit.
I don't think the' boilers out there now 24 even understand this problem because they don't put'a harsh 25 environment qualification in secondary containment.
It is t
()
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,
Suite 300 Washington, D.C.
20006 (202) 293-3950 sr-r
=,-
441
'I 1
thought to be a mild environment.
So I. assume when they are 2
doing their IPEEEs, they go back and' review the situation 3
and see if they really are in shape or not.
.l t
4 MR. ADER:
A lot of them will take credit in the 5
harsh environments for the work that was done under.EQ.
t i
6 MR. MICHELSON:
Yes, but you have to make sure 7
that the equipment that was purchased for inside of 8
secondary, not primary, secondary containment for the 9
boiling water reactors is qualified for whatever the i
10 environmental effect is of the rupture of that large reactor 11 water clean-up pipe.
12 If they have it, that's great-.
That is what IPE 13 is all about.
You postulate the h, i.
You look at the 14 sub-compartment pressurization.
You see if it is too great s
15 for the doors and whatever and if it is, then you chase it 16 until it finally equalizes into the railding and then you 17 look at that pressure and temperature condit an'and ask if 18 the instruments and so forth are okay.
I~ thought that was 19 IPE or IPEEE, I should say.
20 MR. DAVIS:
IPE.
J 21 MR. MICHELSON:
It should be IPE as well but it-22 depends on what you call these kind of events', internal.or 23 external.
24 MR. ADER:
In 105 as I remember because I did talk 25 with Dr. Burdick on it, in that case at least for the f
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,
Suite 300 Washington, D.C.
20006 l
(202) 293-3950
442 i-1 studies that were done there, they found the probability of j
f-(,]f 2
the pipe break in the blow down outside containment 3
sufficiently low to begin with.
5 4
MR. MICHELSON:
You had better look at.that one 5
closely too because again we looked at the-GE analysis that i
6 first tried to conclude that.and we found the analysis 7
significantly flawed.
8 When they went back and did the analysis right,.
j 9
then it began to get into a much more interesting range.
So.
10 if they claim that is done away'on a probability basis, l
11 first of all make sure that'they really have analyzed 12 probability of failure of the isolation valves to close I
13 under dynamic loading conditions and not under static 14 conditions like was done in WASH-1400.
15 If you do it'that way, sure, two valves in a 16 series always look good but if the valves can't'close under 17 the flow, then two valves in series is no better than one or 18 three or four.
They just won't close.
They;will just'go 19 out on overload and that is it.
t 20 MR. ADER:
I guess I am struggling somewhat.
'I l
21 have not been involved in the analysis that they have done l
22 of the advanced reactors but we were down in the spring I 23 know on 105 and went through some of_those issues with the 24 committee and I think some of those same questions came up 25 and went out to Idaho and looked at it.
)
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,
Suite 300 Washington, D.C.
20006 (202) 293-3950 i
m 443 1
MR. MICHELSON:
Now we are asking how.are they l
()
2 actually handled in the IPEs as they come inland you look at 3
them and what do you find out about it.
I guess you looked-l 4
at some of them.
I know you sort of sample some of_them at 5
least.
f 6
MR. ADER:
That was the case on 105 where the i
7 staff's resolution found the risk was low.
They did include 8
in that analysis the environmental effects and still found 9
the risk was low.
They resolved the issue needing no l
10 further action.
11 MR. MICHELSON:
We had extensive discussions with 12 them about how they did that and.I don't think that was ever 13 closed.
14 MR. DAVIS:
Is there anything else?
O\\
r l
15 MR. LINDBLAD:
Yes.
Dr. Ader, as I have tried to 16 review in my own mind the problems I have had with this 17 particular issue I guess it comes down to the precise-18 wording used in resolution of generic. issues and I really 19 believe that over the last several months as I have seen 20 generic issues referred to the IPE that the words really are 21 acknowledgement of error in the original classification of 22 the issue as being generic when actually it is plant 23 specific.
24 MR. ADER:
That is a discussion on an awful lot of 25 issues.
The generic issue process will look to see if there
(
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,
Suite 300 l
Washington, D.C. 20006 (202) 293-3950 l
i e
N
i k
444 l
1 is something at a class of plants that you could-fix.
2 MR. LINDBLAD:
It really has been a 3
reclassification of the issue as being a plant specific 4
issue rather than a generic issue.
It has hardly been a j
5 resolution of it is what I was trying to say.
6 MR. ADER:
In that process and I will jump-into 7
IPEEEs slightly, in generic issue 57 there was a specific 8
configuration that if a plant had and it dealt with a deluge 9
spray into a diesel generator intake and'it dealt with I
10 unqualified seismic relays that would actuate that.
11 There was consideration of what plant specific 12 configuration would be a high risk and there one was 13 identified and they went forward separately to see if there 14 were any plants that had that specific configuration.
l l
15 In a lot of these there is consideration of 16 different configurations of plants and what type of 17 configuration might result in a high risk and they do it 18 different ways as you have seen.
They will take 1150 and 19 they may adjust it and they may take others to try to cover 20 a range.
So there is consideration of plant specific l
21 differences.
l 22 MR. WILKINS:
Let me try to say something Bill 23 said in a somewhat different way.
NUREG-0933 will say a 24 generic issue is resolved and then there will be language 25 that says being handled by IPE.
Is that correct?
l i
ANN RILEY & ASSOCIATES, LTD.
Court. Reporters 1612 K Street, N.W.,-Suite 300 Washington, D.C.
20006 (202) 293-3950
I i
-445-1 MR. ADER:
In the ones that I remember and this is i
2 the problem again, there are a lot of them and they have l
3 been said differently and I looked.back at some.
4 MR. WILKINS:
Let me assume that my statement is j
5 correct then, all right, because I think we have been' told 6
that it is occasionally correct and I think the committee l
7 has expressed concerns that once.it gets into the IPE the l
i 8
reviewers of the IPE do not necessarily look to see if the i
l 9
plant, in fact, addresses -- the; plant IPE, in fact, l
l 10 addresses that issue and we have expressed some concern that 11 maybe there should be some, I don't know, moderately formal-12 process for making sure that when something goes-from.the 13 generic issue list and is handed off to the IPE-or'the'IPEEE 14 that it then doesn't get lost.
O 15 MR. SEALE:
It could evaporate.
16 MR. WILKINS:
Or sublimes.
i 17 MR. ADER:
I think some of the confusion because I 18 have gone back and I have talked to'some of the people 19 involved and like I have said, I have been involved in the 20 tail end of some of the resolutions, the wording may have 1
21 been poor.
22 We took credit for an on-going program to provide-23 some additional assurance that if there was something plant-I 24 specific it would generally be identified but if we knew of 25 a configuration like we did in 57, we went and looked for i
s/
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,
Suite 300 Washington, D.C.
20006 (202)1293-3950' i
446 1
specifically.
We didn't rely on the IPEEE.
Os 2
We did a survey of plants and I think they are 3
down to potentially one plant that might have that 4
configuration.
Generally these have been resolved on the 5
low risk significance that was found in the studies.
Then 6
there was this other program ongoing and we didn't want to 7
ignore that.
8 MR. CARROLL:
Let's look at 153.
It seems to me 9
that the resolution at least suggested that service water 10 systems had a big potential for problems but we are going.to 11 deal with these in the IPE since it is such a plant specific-12 issue.
13 Now I think to follow on with Ernest, is there a 14 check list someplace, for example, for'the guys who do the 15 IPE review to take the issues that'were'in 153 and make sure 16 each IPE has dealt with them.
17 MR. ADER:
There is not a checklist to say 153, 18 here are all the facets of 153, yes, it dealt with every 19 facet.
Here is 76, check, it has been dealt with in all l
20 facets.
But the safety significance that was judged at the 21 time in the generic issue process, 153 was r.n example and 22 105 was an example, told staff from the perspective that I 23 am aware of that here was something that we didn't see the 24 significance to go beyond the IPE process that was there and i
25 we knew what sort of review was being done.
l ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,
Suite 300 Washington, D.C.
20006 l
(202). 293-3950 i-
447 j
1 Service water in my mind is an easy one because a O)
(,
2 reviewer is clearly going to look, do they have service 3
water dependencies.
Have they addressed them?
4 MR. CARROLL:
Except, you see, in 105 you guys-and
(
5 I am going on memory but you guys hadLproduced~a list of j
l
\\
l 6
potential vulnerabilities that you had seen based on 3
7 operating experience out there and I. guess my question is or 8
my view is'there should be a checklist that the IPE reviewer 9
is looking at individual IPEs and satisfying himself that 10 they have dealt with these kind of issues.
f 11 MR. ADER:
And 105 is another easy one because i
12 there is specific guidance in the generic letter that they j
l j
13 report sequences intersystem LOCA and reviewers, the last 14 time we were down we left a copy of a draft review guidance
\\--
15 document that we are expanding and updating so it.has some 16 holes in it, the copy that we left.
17 But reviewers will go through and say did they 18 consider intersystem LOCAs and how did they treat them.
Did 19 they consider dependencies, service water DC power?
But 20 they do not go through and check this issue or that issue.
r l
21 MR. WILKINS:
Pete, I think I have gotten an 22 answer to my question.
23 MR. DAVIS:
Yes, I think so.
24 MR. MURPHY:
Let me try to expand that a little 25 bit on what has been said.
Obviously in an IPE that we ANN RILEY & ASSOCIATES, LTD.
Court Reporters i
1612 K Street, N.W.,
Suite 300 Washington, D.C.
20006 (202) 293-3950
448 1
ordered up in 1988 you may not find directly a generic issue
)
2 which was identified in 1992 or 1993.
3 What we do look for in our IPE process are things-4 and service water is an easy one as Charlie said because we-1 5
look for the dependencies, did they model the dependencies l
6 correctly, have they looked at. things.
We look at the t
7 boundary conditions in the type of review we do and that r
8 review is explained in the guidance that we are now 9
developing.
10 There are two things that we are trying to do a
~
11 little better than we have done before and these are in 12 recognition of the comments the committee has already given y
13 us. One is we are trying to disseminate the information that 14 comes out.of these generic resolutions a little more widely-.
i N
15 This will be more true of the IPEEEs because they 16 are still in progress.
Most of the IPEs are essentially 17 done as you saw from John's chart and even the ones that 18 aren't here yet I still would expect the analysis is pretty 19 well completed.
20 But as we resolve generic issues we are trying to i
21 put out information letters and we have done it a couple of 22 times now and we will do more of that to disseminate the l
23 information better than we may have in the past.
24 There is a series of discussions going onfright 25 now between Research and NRR to see if there is some;way-ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,
Suite 300 l
Washington, D'.C-20006 (202) 293-3950 l
I
'449 1
that we can strengthen the process in recognition of the 1
()
2 committee's concerns.
3 We haven't reach a conclusion-on that or a 4
4 consensus as to:how to go forward but I suspect'we will be 5
doing something.
What the nature of that.is I can't'really-6 tell right now.
That is still in the' course of 7
deliberations.
8 MR.' DAVIS:
Thanks, Joe.
Anything else?
9
[No. response.]
10 MR. DAVIS:
I will hand it back to you, Mr.
11 Chairman, on time I might note.
l l
12 MR. WILKINS:
In f&ct,.if I had been~a little less I
I I
13 verbose, it might have-been a minute or so early.
Thank you 14 very much, Mr. Ader and the members of your staff.
We will.
I
!O 15 take a 14-minute break and reconvene at 10:45.
l 16
[Brief. recess. ]
17 MR. CARROLL:
Let us reconvene.
Charlie, this is-18 a session on the EPRI Utility Requirements Document for l
l 19 passive plants.
20 MR. WYLIE:
Right.
l 21 MR. CARROLL:
What we are going to do is to i
22 consider a letter you have drafted so we have no need for a l
23 record of this so we may excuse our reporter for the day.
24
[Whereupon, the reported portion of the above -
25 entitled ACRS meeting was concluded at 10:53 a.m.)
i i
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1612 K Street, N.W.,
Suite 300 Washington,-D.C. 20006
-(202) 293-3950
1 i
j 1
I.
i
)
J REPORTER'S CERTIFICATE This is to certify that the attached proceedings before the United States Nuclear Regulatory i
Commission in the matter of:
i j
NAME OF PROCEEDING 404th ACRS Meeting 1
i i
DOCKET NUMBER:
r l
1 PLACE OF PROCEEDING:
Bethesda, MD i
j were held as herein appears, and that this is the original transcript thereof for the file of the 4
United States Nuclear Regulatory Commission taken
~
by me and thereafter reduced to typewriting by me or under the direction of the court reporting j
co'mpany, and that the transcript is a true.and
.(}
accurate record of the foregoing proceedings.
t 4
'D ?.n J. n - W[D official Reporter
/
i Ann Riley & Associates, Ltd.
~
i' i
l 1
k k[$)
4 4
4.-.
y,
O O
O Individual Plant Examination {IPE?
Overview and Status Advisory Committee on Reactor. Safeguards i
December 10,1993
[
i Presented by: Drs.-John H. Flack Office of' Nuclear Regulatory Research U.S. Nuclear Regulatory Commission 4
O o
o Overview
- Background
- Review Process and Objectives
- Preliminary Review Findings
- General Observations i
i i.
I
O o
o
Background
- SEVERE ACCIDENT POLICY STATEMENT 4
August 8,1985 f
- INTEGRATION PLAN FOR CLOSURE OF SEVERE ACCIDENT ISSUES (SECY 88-147)
-May 25,1988
- GENERIC LETTER 88-20 November 23,19.88
- Supplement 1.(Internal Events)
August.- 29, :1989
- 1 1
- Supplement 2 -(AM< Strategies) l
-April 4,1990 i
- Supplement 3-(CPI Recommendations) f July ' 6,1990.
- Supplement 4-(External-Events.lPEEE).
i June 27,1991' t
e 4
e
--r--
wer
,we,
-e-v-s u
- - -. +
s.
r v-
,,ooe.w sz-me,v.
---~w-
- - ~,-
--e-a m--s-
8 O
o IPE Submittal Schedule (Current Status)
TOTAL IPEs SUBMITTED PER QUARTER l
1 25 i
^
20-DUE DATE PER GENERIC LETTER 1
l 15-l 78 SUBMITTALS (113 UNITS) 10-f i
5-l
._._._..yl LRL_,FY95 FY 90 FY 91 FY 92 FY 93 FY94 Dated: 12/10/93
O O
O L
Review Status k
- 63 of 78 Submittals Received (80%)
- 35 IPE Reviews initiated
- 31 Step 1 4 Step 2 i
l
- 13 Reviews Completed
- 11 Step 1 2 Step 2.
I 1
i l
- Completion Target 12/95
Step 2 Review i
Rationale for Review
- Turkey Point:
First IPE submittal based on l
a PSA not formally reviewed by the Staff l
- FitzPatrick:
To obtain a better understanding of the IPE process and its relationship to identified issues
- McGuire (Limited Step -2):
To obtain a better understanding of the licensee's HRA process I
i
- Zion:
To obtain a better understanding of the licensee's IPE methodology i
i
O o
o' NRC Review Process Review Teams (2-4 Members):
- Team Leader
- Front-End System Analyst
- Back-End Containment Analyst
- Human--Reliability Analyst Step 1 Review.
- Submittal Only Review
- NRC/ Licensee Interaction Decision to Proceed to Step 2
- Understanding' Licensee's Process a
l
- Unique Characteristics-Step 2 Si.e Visit and Audit t
Staff Evaluation Report I
O o
o Ob ectives
- To Achieve an Overall Appreciation of Severe Accident Behavior i
- To identify Dominant Sequences
- To Obtain a Quantitative Understanding of Core Damage and Radioactive Release
- Appropriately Reduce the Overall Probability of Core Damage and Release of Radioactivity i
Generic Letter 88-20 (11/23/88)
J
O O
O To Appreciate Severe Accidents Performance of a Level 2 Analysis Licensee involvement Implementation of Safety Enhancements Evolution of "Living" Documents i
1 1--
O O
O Level 1 (Front-end}
Key Areas
- Initiating Events and Associated Event Trees
- System Dependencies
- Success Criteria e
i
- Data Analysis / Common Cause Failure I
i
- Internal Flood
- Decay Heat Removal Function i
i l
O o
o 1
Level 2 (Back-end) Analysis Key Areas
- Level 1/ Level 2 Dependencies
- Containment Failure Modes and Timing q
- Containment Event Trees
- Release Catagories
- Uncertainties 4
- Consideration of CPI Recommendations i
i l
. ~.
O o
o i
Licensee Follow-on Activities Accident Management t
Support Licensing Action 4
- Evaluating Safety issues i
l
. Evaluate Design Changes
- Prioritizing Proposed Improvements t
Operator Training i
e
{
i l
l 2
e
...-......~,....s.,_
O o
o To Identify Dominant Sequences
- Licensees Report Top 100 (or 95%)
l
. Evaluate Dominant Contributors 1
l
- In. Addition, Many Licensees Perform:
i
- Uncertainty Analyses i-
- Sensitivity Studies l
- Calculate importance Measures I
u
O o
o To Obtain a Quantitative Understanding 1
1
- Develop a Probabilistic Estimate of Core Damage
- Understand the Basis for the l
Estimate
- Sensitive to the Analytic Limitations i
i i
i i
!t
O O
O Probabilistic Estimate of Core Damage (Internal Events)
Number of Units 20 X::g 15-6 E
g g
a$ Il
.E 10-x.
w 2
5 a?
+3 3
si g
g i
{w h
{y p
5-l g
g B
n n
P y%
r:
?
m g
I LR E
,L.- EN i
i i
i i
9 __
l i
i 1
,10[-7]/ yr 10[-6]/y r 10[-51/yr 10[-4]/yr I
1 i
I l
4 M swRs I?MI PWRs 1
80% of Units, 12/10/93
]!
i t
O o
O Important Assumptions
- Treatment of Operator Actions
- Treatment of Common Cause
- Interpretation and Applicability of Data
- Success Criteria
- Modeling (e.g., RCP Seal Model) i e
O O
O RCP Pump Seal LOCA Contribution Plant Name CDF (x10-4/yr)
Contribution Watts Bar 1, 2 3.3 70.0%
0.4 60.0%
Catawba 1, 2 Turkey Point 3, 4 1.0 56.3%
Beaver Valley 1, 2 1.9 53.4%
Diablo Canyon 1, 2 0.9 42.9%
0.4 24.0%
i McGuire 1, 2 Oconee 1, 2, 3 0.2 17.0 %
Millstone 3 0.7 10.0 %
0.7
<1.0 %
Surry 1, 2 Palo Verde 1, 2, 3 0.9
<1.0 %
4 i
i
\\
O o
o' l
l To Reduce the Overall Probability
- Definition of Vulnerability
- Plant Improvements i
i l
l
}
l I
O O
O Vulnerability DEFINITION:
o Global - Accident Sequences Specific - Numerical Criteria I
o L
o Cost / Benefit.- Process SOURCES OF POTSATIAL VULNERABILITY:
o System Dependencies o
Environmental Effects
O O
O Licensee Plant Improvements (Safety Enhancements)
- Utilize Fire Water for:
- Equipment Cooling
- Containment / Core Cooling
- Implement System / Unit X-Ties i
i
- Protect Against Internal Flood
)
- Add Alternate AC Power Supplies I
- Upgrade DC Power Supplies
(
l'
O O
O Plant Improvements (Continued?
- Provide Makeup to RWST
- Enhance RCP Seal Cooling
- Enhance Room Cooling Capability i
- Reduce Asymmetries
- Add Diversity to DHR Capability-i i
L --
O O
O Genera Observations
- Benefits Exceeded Expectations
- Evolution of "Living" PRAs
- Complex Studies, Simple Insights
- Assumptions important to Findings
- Many Enhancement Opportunities
- Potential Future Applications i
'+
O O
O IPE DATABASE AND; INSIGHTS M. T. DROUIN
[
U.S. NUCLEAR REGULATORY COMMISSION OFFICE OF RESEARCH DIVISION OF SAFETY sSSUES RESOLUTION i
SEVERE ACCIDENT ISSUES BRANCH PRESENTED TO l
ADVISORY COMMITTEE ON REACTOR SAFEGUARDS 4-DECEMBER 10,1993 J
e i
ACR5/12-ICL 93/Pnge 1 of13 i
O O
O l
' OUTLINE I
Data Base Objectives t
Database Structure l
i i
Possible Future Uses and Insights i
i l
I ACKS/l2-1693/Page 2 of13
..m
O O
O OBJECTIVE OF THE DATABASE IS NOT JUST TO GATHER IPE INFORMATION How Do Plant Features Factor Into the Core Damage Frequency And Containment Performance?
9 If Two Plants In Basically Same Class Have Markedly Different Core Damage Frequencies Or Containment Performances, What Is Responsible?
If A Class Of Plants Seems To Share A Particular l
Contributor To
- Risk, What Plant Features Are l
Responsible?
i l
AQts/I2-to93/Fage 3 ofI3 4
.___...._...._..m._
._m.
___.--..._..m._.
.._._...m.
_....m......
O O
O DEVELOP A DATABASE THAT STORES INFORMATION IN A UNIFORM AND STRUCTURED MANNER AND THAT CROSS-TIES PLANT INFORMATION SYSTEMS A B C D B F G H:iJ K L M DEPENDENCE TABLE sst ss2 m
m m
m m
o EUNCTIONS IBl IBl IE2 IE2
~ SUCCESS PATHS STRATEGIES 5
CONT. EAILURE CHARACIERISTICS PDS1 181 PDS2 IE2 COmmi,T MAN m
m I
ACCIDENT SEQUENCES
~
ACK5/l2-10 93ftsge 4 ef13
O O
O CERTAIN INFORMATION ALSO NEEDED TO BE EXTRACTED FROM EACH IPE Success Criteria Information System Information
=
Front-line Functional Strategies For Support Core Damage Prevention Mapping From Core Damage Dependence Information Front-line dependencies Plant Damage State To Releases PDS Parameters Support dependencies Containment Matrix Level 2 Analysis Parameters.
Sequence Information Success Paths Source Term Failure Paths Causes Frequencies ACK5il2-l#939ege 3 ef13
O O
O EXAMPLE INFORMATION ENTERED FOR BWR SYSTEM FIELD OR RECORD FUNCTION SYSTEM NO ENC NOTTSSOURCE UNIT CROSS-TIE REACTIVITY RPS CONTROL SIL CRDS RECIRC RY-INTEGRITY HIGH f
PRESSURE INJECTION i
e l
l l
l ACR3/12-16 93/Page 6 efl3
O O
o EXAMPLE INFORMATION ENTERED FOR DEPENDENCE FIELD OR RECORD i
i ACRONYM FUNC110NS m TING FUNCTION A FUNCTION B FUNCTION C NOTE & COMMENTS
.4 11
." 14111 i
FxrRIES --
i Absolute (Failure of System 2 Causes Failure of System 1) l A
=
Primary (Fialure of System 2 Cuase Failure of System 1 Unless Backup Is Brought P
=
Into Play) j.
S
- Secondary (System 2 is a Backup to Some Primary ~ Support of System 1)
=
l D
Partial _(Failure of System 2 Increase the Probability of Failure of System 1)
=
l ACR$U2-lS93.%ge 7of13
O O
O EXAMPLE INFORMATION ENTERED FOR SEQUENCE FIELD OR RECORD seqUsucs ros ciw is msr FAIU m CAUSES ATrMBUTES MMCHOMSYSTEMS NOTES l
j ENTRIES -
S1, S2, S3, A, V, Tioop, Trx, Ttt, Tatws, Tuhs, Trecire,.....
=
LOST SUPPORTS AC, ACBUI,..., DC, EAC, HVAC, NSW,.....
=
RCS-BOR, RCS-INT, RCS-DEP, HPI, IIPR LPI......
FAILED. FUNCTIONS
=
IFL, FIRE, CCF CAUSES
=
ATWS, BYPASS, TIL, SBO, HUM ATTRIBUTES
=
ACKSil2-1093/Fece 8 efI3 s
e
. ~.. -.
O O
O EXAMPLE INFORMATION ENTERED FOR LEVEL 2 PARAMETERS FIELD OR RECORD l
RFIEASE CONT.
FAILIJRE CONT.
ZR AMotMT CCI
- VIWISEE, SUPP.
SFRAY Dt9 CONT.
CLARS FAILURE LOCATION FAILURE OXIDATION OFCORE DESPOSITI FAILURE FOOL AVAII.
CREDIT
- bootW, SIZE GN VHISF14 IN CCI ON
>RMW, BYFAAB mmmm mm I
m
.I k
ACKSil2-lo93)Page 9 of13
+
O O
O STATUS OF DATABASE I
Data Resides In DBaseIV Data From Half Of The Expected IPEs Entered i
Database Being Expanded To Include IPEEE Information i
4 Draft. User Manual Complete l
l i
j i
AOtS/22 IS03/Page 10 of13 ie
O O
O EXAMPLE OF IPE INFORMATION THAT CAN BE EXTRACTED FROM DATA BASE PRELIMINARY ESTIMATES IB-3 _
15 3 -
E IE -
-[
s a
g e
i ni usAN u
,ad e usmAN
[
^h' f
~
h;: j g'3
~
m
??
155 i
-!P " -
7 g
ll i
g kl3
[
C3 NEAN
=-
"k
- 3 I
3 h
154 -
. i....
~
1 3
8 j
uBMAN 3'g T
' h.:
l s.
h.
~
A IB-7 9
F g
7 gbl 2
i
~
M I
IbI -
l s
All.
BWRs FWRs g
g F1AN13 45 FtANTS -- 14 BWRs AND 31 FWRs 39 FLANTS -- 14 BWRs AND 25 FWRs I.
ACKZ/12-10-93/Page il $13
O O
O POTENTIAL USES OF THE IPE DATABASE INCLUDE 1
1 Determine The Importance of Plant Features At The System And Function Level t
Insights For A Single Plant Or Classes OF Plants Regarding Potential Significance Of Possible Generic Issues i
Provide Risk-Based Information From The Systems i
Standpoint To Assist In Issues (E.G., Maintenance Rule Implementation) i i
h AQtsu2-1093%ge 12 of13
1 O
O O
STATUS AND SCHEDULE OFIPE/IPEEE INSIGHTS PROGRAM t
IPE Database Currently Has 45 IPEs (All Reactor And Containment Types - 14 BWRs and 31 PWRs) i Draft Program Assumptions For Contractor Assistance Initial Draft Report Based On 50 IPEs Fall 1994 r
Revised Report Based on All IPEs 1995 l
l Revised Report To Include First Set Of IPEEEs 1996 r
n Final Report Issuance 1998
(
l AQts/12-10 wtage U of U
!w
~
O O
O
~
RESOLUTION OF GENERIC ISSUES VS. IPE/IPEEE i
i GUIDANCE FOR PERFORMING IPES ISSUED IN 1988 (GL 88-20) AND 1989 o
(NUREG-1335).
LICENSEES REQUIRED TO ADDRESS USI A-45, " SHUTDOWN DECAY HEAT REMOVAL l
.o REQUIREMENTS" AS PART OF IPE.
o LICENSEES NOT REQUIRED TO ADDRESS OTHER GENERIC ISSUES IN THEIR IPES, ALTHOUGH THEY WERE GIVEN THE OPTION TO PROPOSE RESOLUTION TO OTHER GENERIC ISSUES.
i o
HOWEVER, AS PART OF THEIR IPES, LICENSEES DO LOOK AT A NUMBER OF AREAS RELATED TO RECENT GENERIC ISSUE CONCERNS.
SPECIFICALLY, LICENSEES IPES ADDRESS:
- SYSTEM DEPENDENCIES, E.G. SERVICE WATER, DC POWER 4
- PLANT SPECIFIC INITIATORS AND COMPONENT DATA i
I
- COMON CAUSE FAILURES
- OPERATOR ACTIONS 1
o GENERIC ISSUE RESOLUTION AND PRIORITIZATION PROCESS-RECOGNIZES THIS j
ONGOING ACTIVITY AND ALSO RECOGNIZES-THE REVIEW PROCESS FOR THE IPES.
i s
E O
O O
/
RESOLUTION OF GENERIC ISSUES VS. IPE/IPEEE (CONT.)_
n 4
t o
IPEEE GUIDANCE ISSUED IN 1991 (GL 88-20, SUPLEMENT 4 AND NUREG-1407).
i o
IPEEE SPECICALLY ADDRESSES USI A-45 (EXTERNAL EVENTS PORTION) AND GI-131.
o LICENSEES SPECIFICALLY REQUESTED TO ADDRESS' ISSUES RESULTING FROH 1
L THE SANDIA FIRE RISK SCOPING STUDY (NUREG/CR-5088).
THIS WAS RECOGNIZED IN RESOLUTION OF GI-57.
o RECENT PRIORITIZATIONS RECOGNIZED THAT RELATED ISSUES ARE BEING ADDRESSED AS PART'0F IPEEE.
u i
i i
t i
l.
2 l
~~
~
O o
o RESOLUTION OF GENERIC ISSUES VS. IPE/IPEEE ICONT.).
o RESOLUTION OF HIGH, MEDIUM, AND NEARLY-RESOLVED GENERIC ISSUES IS TRACKED IN SAFETY ISSUES MANAGEMENT SYSTEM (SIMS) AND DOCUMENTED IN NUREG-0933 WHEN COMPLETE.
o WHEN AN ISSUE IS PRIORITIZED LOW OR DROP, THE BASIS IS ALSO DOCUMENTED IN NUREG-0933.
3 4
4
-