ML20235H622
| ML20235H622 | |
| Person / Time | |
|---|---|
| Issue date: | 07/10/1987 |
| From: | Advisory Committee on Reactor Safeguards |
| To: | |
| References | |
| ACRS-T-1600, NUDOCS 8707150205 | |
| Download: ML20235H622 (142) | |
Text
FCRSF /600
\\
i
.tiWG M O
UNITED STATES NUCLEAR REGULATORY COMMISSION i
IN THE MATTER OF:
DOCKET NO:
ADVISORY COMMITTEE ON REACTOR SAFEGUARDS 327TH GENERAL MEETING O
LOCATION:
WASHINGTON, D.
C.
PAGES:
1 - 101 DATE:
FRIDAY, JULY 10, 1987 pp[$ ;'~ '
$o, e -
r N '%) L '
Jo Not Re:p: 1.::::: AJL;S DIfice ACE-FEDERAL REPORTERS, INC.
OfficialReporters 444 North Capitol Street 8707150205 070710 Washington, D.C. 20001 PDR AcRS (202)347-3700 T-1600 PI'R NATIONWIDE COVERACE
6310 01 01 1
t'lysimons 1 UNITED STATES OF AMERICA V
2 NUCLEAR REGULATORY COMMISSION 3
327TH GENERAL MEETING 4
OF THE 5
ADVISORY COMMITTEE ON REACTOR SAFEGUARDS 6
Nuclear Regulatory Commission 7
Room 1046 l
8 1717 H Street, N.W.
9 Washington, D.
C.
l 10 Friday, July 10, 1987 11 The committee convened at 1:15 p.m.,
Dr. William I
12 Kerr, Chairman of the Committee, presiding.
13 ACRS MEMBERS PRESENT:
14 W.
KERR, Chairman 15 J.
REMICK, Vice Chairman 16 J.
EBERSOLE 17 H.
LEWIS 18 C.
MARK j
19 C.
MICHELSON
{
20 D.
MOELLER 21 G. REED 22 P.
SHEWMON 23 C.
SIESS 24 D.
WARD (s'-)
25 C. WYLIE
/\\CE. FEDERAL REPORTERS, INC.
l
- oaa47xm Lionwiae cmerage sm-B6-6646
h 01 01 6-2 marysimons 1 ACRS COGNIZANT STAFF MEMBERS:
2-R.
FRALEY, ACRS Executive Director 3
NRC STAFF PRESENTERS:
4 E. JORDAN 5
T. NOVAK 6
M. WILLIAMS 7
J.
ROSENTHAL 8
M.
CHIRAMAL j
9 S.
ISRAEL 10 P.
LAM 11 R.
DENNIG O
\\/
12' K.
BLACK v
13 14 15 16 4
17 18 19 20 21 j
22 l
l 23 1
0 24 25 ACE. FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coserage 800 336 6646
O PUBLIC NOTICE BY THE t
UNITED STATES NUCLEAR REGULATORY COMMISSIONERS' ADVISORY COMMITTEE ON REACTOR SAFEGUARDS FRIDAY, JULY 10, 1987 The contents of this stenographic transcript of the proceedings of the United States Nuclear Regulatory Commission's Advisory Committee on Reactor Safeguards (ACRS), as reported herein, is an uncorrected record of the discussions recorded at the meeting held on the above date.
No member of the ACRS Staff and no participant at this. meeting accepts any responsibility for errors or inaccuracies of statement or data contained in this transcript.
O
6310 01 01 3
[ )D simons 1 y
PROCEEDI NGS m
2 Office for Analysis and Evaluation of 3
Operational Data Activities 4.
(Briefing / Discussion Regarding 1987 Case Studies 5
and Startup Plant Study by AEOD Representatives)
I 6
CHAIRMAN KERR:
We reconvene the meeting to 7
consider a report by the Office for Analysis and Evaluation 8
of Operational Data Activities, and Mr. Ebersole will make 9
whatever introduction is necessary.
10 Mr. Ebersole.
11 MR. EBERSOLE:
Let me refer you to Tab 11 about 12 the initiation of this meeting here which I hope will be a 13 first of a long and profitable experience we will have with 14 AEOD and I think one that is long overdue, 15 You can just quickly look over the agenda here 16 and find out that they are going to give us a sampling of 17 the kinds of events they study, the basis for looking at j
j 18 them and certainly one of the ones that is right on top 19 right now is the one we just left before lunch, which is 20 the incident at Diablo Canyon and the product of our last 21 session of an hour or so ago, which was that we saw no 22 particular reason to try to do anything or write a letter
\\
23 now, and we did have a few comments to add to the NRR i
24 tetter which was produced and still has not been processed 25 through the CRGR.
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 800-336-6646
a l
' 6310 01 01 4
. r; ysimons 1 You will see a copy of that coming around to 2
you.
It is going to be handed out shortly in the course of 3
this discussion with the AEOD.
You can look at that we 4
will comment a little bit later about what we should do in
]
5 this sort of closed-loop operation from us to NRR, a i
6 process of course that doesn't go through the Commissioners i
7 and we can talk about that.
8 We are going to hear from Ed Jordan and his 9
staff about their entire process with this particular event 10 being sort of a new regime of operational conditions, 4,
5 I
11 and 6, where there has been found a rather significant 12 potential for core melt and a long neglected area of 13 operation and perhaps a region in.which there are design 14 voids as well as operational lack of discipline.
1 15 It's your program, Ed.
16 MR. JORDAN:
All right, sir.
Thank you very 17 much.
18 We appreciate the opportunity to meet with the 19 ACRS and look forward to periodic briefings in alternate 20 months if that is the ACRS desire.
J 21 I would defer a generalized discussion of AEOD i
22 roles and functions to some separate meeting about the new 23 organization.
I won't discuss it all in this meeting.
You l
24 will get an idea of how the analysis of trends and patterns 0
25 functions through the discussion of the events that will be ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 800-336-6646
l 1
6310 01 01 5
)
)
()ysimons1 provided.
2 Of. course, our reviews are oriented toward those 3
longer-term operating issues.
Those are done by the 4
Division ot Safety Programs and Tom Novak is the Division 5
Director for the Division of Safety Programs and is here 6
with me now.
7 I would say just a couple of things about the 8
levels of reviews and the budget cycle we are in right 9
now.
The AEOD analysis of operational events is expected 10 to remain about constant in the oncoming years.
We are not 11 asking for nor getting a substantial increase in budget or 12 personnel that are being applied.
13 We are internally increasing our emphasis on the 14 human factors side of the analyses and the events in 15 looking and trying to provide some quantification of the 16 contribution of the events in the process of the analysis.
17 Then, finally and perhaps most importantly, we 18 are trying to find additional ways of communicating the 19 results to the industry so that the lessons out of these 20 analyses are not only available but are used in exploring 21 with industry ways of doing that and considering things 22 such as workshops on a periodic basis.
23 So with that I would like to turn it over to Tom 24 Novak.
O 25 MR. NOVAK:
Thank you, Ed.
ACE. FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 800-3364M6
~
1 1
i i
6310 01 01 6
r()ysimons1 I really don't have anything to add as far as 2
the organization, I think we ought to just get into our 3
technical discussion, and I will ask Jack Rosenthal, who is 4
the Chief of the Reactor Operating Analysis Branch to give 5
you a ahort dic'ussion on the activities of that branch.
j
'l I
6 (Slide.)
k 7
MR. ROSENTHAL:
ROAB is one of two branches l
8 within DSP and there is another group for non-reactor
{
9 events human factor issues.
10 I have put up this slide just for a moment, a j
11 rather simple slide.
12 There are two systems sections and an l
x 13 engineering section that were always in the branch.
We 14 have added the performance indicators to the branch, and I 15 would like to talk just a little bit about the differences 16 from before.
17 If one can imagine some sort of multidimensional 18 matrix, a narrow engineering issue might be an element in 19 that matrix.
A plane through that multidimensional surface 20 might be a systems study, and ROAD was always set up to do 21 that sort of work.
1 22 One can take an orthogonal plane r.0w through the
]
1 23 multiple number of studies and call that to some extent the 24 performance indicator program.
So it can look at both 25 systems and across systems at plants, and I hope to have a l
)
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 800-3364M6
i i
6310~01 01 7
(()ysimons1 healthy synthesis between the PIs which are definitely a 2
plant specific type activity and the systems and 3
engineering work which is a different focus.
That should 4
help us all, and we won't be discussing PIs today.
5 (Slide.)
6 The essential branch product is to review 3,000 7
LERs a year, and we turn out about 30 studies a year, tech 8
reviews, engineering evaluations, case studies and a 9
variety of products.
Everything else is in a sense 10 overhead.
11 MR. SHEWMON:
Will you define the difference 12 between those three?
13 MR. ROSENTHAL:
Yes, I will.
14 (Slide.)
15 I want to talk about the direction a little bit, 16 and I should say that this is not a radical cost change, 17 but only where I think that we are going to emphasize areas 18 or cause correction.
19 The branch was always involved in questions of 20 operator actions and procedures as well as hardware, but I 21 envision having a substantive amount of the resources 22 turning towards issues of human performance, procedures, 23 and procedure inadequacies, lack of procedures and issues t
J 24 of what should be in a procedure and what should be in the 25 knowledge base of people, and we will inevitably in the ACE. FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800-336-6646
'6310 01.01 8
()ysimons1 course of that work get involved in the man-machine 2
interface.
3 One of the specific studies you will hear about, 4
pressure control, really revolves around that issue, and we 5
will be working more in that.
6 We want to have an emphasis on quantitative 7
analysis.
First of all, we have the data.
Now that data 8
is shared for the agency, but it is convenient for us 9
within the office to get that data.
So that gives us a leg 10 up.
11 We want to use quantitative PRA to the extent 12 possible.
13 ASP, accident sequence precursor, fits on floppy 14 disks and will continue to be run at Oak Ridge, but we 15 brought it in-house and we will use it in-house in our 16 work.
\\
17 We have made SARA, the new RES tool, operational 18 in our shop, and we have got some other PC based tools so 19 that we can quantify what we believe is the safety 20 significance in our work.
That's a change.
21 MR. WARD:
SARA is for a particular plant, isn't 22 it?
23 MR. ROSENTHAL:
It's for the six NUREG 11-50 24 plants.
While ASP has a generic P and B which you then O
25 modify, we have got some other tools where we can do crude ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coserage 8(n336-6646
6310 01 01 9
t'~)ysimons 1 PRA estimates that are also custom fits.
2 You are not trying to do plant specific PRAs, 3
but you are trying to say, hey, I have a generic study, I 4
think it's important and you go into some sort of generic 5
model to try to get some handle on the incremental risk 6
you're talking about.
7 MR. WARD:
Okay, but you've got SARA operational 8
for those six plants then?
9 MR. ROSENTHAL:
Yes, the SARA package, right, 10 which right now goes up to core melt and will have added on 11 subroutines that will take you out to risk.
It's an RES 12 took and it's promised deliverable.
In fact, we have just 13 started to use it.
ASP has been around for a while and we 14 are using it.
Then there are some other tools that in time 15 we can tell you about.
16 MR. REED:
I notice you say emphasis on human, 17 procedure, operator and all these kinds of things, and 18 that's fine, but I hope that you're emphasis does not 19 exclude the finding of design vulnerabilities because I 20 think that sometimes these are much larger than have been 21 reported and are not considered and fed back appropriately.
22 MR. ROSENTHAL:
Two-thirds or three quarters of 23 the work, as you're saying, but let's spend some resources 24 on that first bullet, a quarter of the resources or a third 25 of the resources or something like that.
ACE-FEDERAL REPORTERS, INC, 202-347 3700 Nationwide Coserage W) 336-6646
1 j
6310'01 01 10
]
v"5ysimons 1 The last point that I wanted to pick up on, V
2 engineering analysis, I would like to become more 3
quantitative than we have to some extent been in the past, 4
and there are two ways to do it.
For example, we saw an 5
event at Songs, an overcooling event a couple cf weeks ago 6
that was of interest to us and we aren't quite sure of it.
7 So we are arranging to have it run on the CE simulator.
8 That's easy for us to do now that AEOD includes 9
the Tech Training Center and we've got lots of contacts.
10 So we can use the simulators as hybrid computers, and we're 11 looking around for some PC based products to also do some 12 work.
()
13 MR. EBERSOLE:
I would like to ask a question.
14 I would like to have you consider the nature of this group 15 and some of the charges that are handed to us by peer 16 advisers that we consider just global or cosmic, even 17 cosmic issues and we are heavy on academics.
Not too many 18 people here like to open the hood of a car and try and find 19 the fuel injection system or something.
20 So we have a charge to condense a lot of this 21 stuff into large, meaningful, heavy issues, and yet I find 22 in my mind that this is analogous to a model that I just 23 mentioned to Dr. Jordan that it's like the ---
24 CHAIRMAN KERR:
Mr. Ebersole, it's hard to hear A-25
- you, t\\CE FEDERAL REPORTERS, INC.
202-347-3HK)
Nationwide Coserage 8(Kk336-6M6
6310 01 01 11
-r'lysimons 1 MR. EBERSOLE:
The model is like the big game V
2 hunter who is going to defend himself in the African 3
woods.
So he takes this great, big rifle and walks in with 4
the lions, the tigers and the elephants, but he doesn't 5
wear shows and gets killed by the ants.
6 So how do we find the global, the cosmic pieces 7
that we are supposed to write about and stop and fix and 8
how do you help us find that?
9 (Laughter.)
10 MR. ROSENTHAL:
Everybody is searching for the 11 Holy Grail.
Why don't we show you the diversity of the 12 work products in just a few months and that would maybe O
13 help.
14 We want to emphasize systems, systems dynamics 15 and systems interactions.
That's where we'll look.
You 16 will hear about an aux feedwater study that was turned on 17 by a few events, but one event would have turned that study 18 on because it was a common mode failure that was of 19 interest to us, and we are going to push for timely 20 feedback and targeted feedback of our work.
21 (Slide.)
22 Now there is a diversity of work products.
23 One would be a tech review, and an example of 24 that was a -- actually this came on the morning calls that 25 we participated in.
It came over from NRR, and of course ACE. FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage Ro336-6646
6310 01 01 12 n tysimons 1 in' talking we saw an inadvertent turbine run-back, and the 2
question was asked is that system more trouble than it's 3
worth.
You know, you hung the gadget on the machine and is 4
it giving'you more trips than it's protecting you from.
5 We went in and did a quick LER search, a little 6
bit of analysis, a three page report with a conclusion of 7
no, the turbine run-back seems to be working pretty good 8
and it's not causing a lot of inadvertent trips.
It's 9
saving trips and the trips that are you have are more 10 gentle because of it.
So it belongs there.
It was a 11 quick study, in and out and checking a feature, and that 12 would be a tech report.
()
13-The next step would be an engineering 14 evaluation.
Engineering evaluations would take several 15 months to do.
It would also involve LER searches, some 16 analytic work and typically come out w' th suggestions.
17 Suggestions are things that we really think ought to be, 18 but are not going to pass to backfit criteria.
19 The aux feedwater study, which you will hear 20 about in just a few minutes is an engineering evaluation, 21 and that one, and I'm just talking about a diversity of 22 products, was kicked off by the observation of a common 23 mode failure in an aux feed system.
We looked around and 24 we saw some LERs and there were more of them and it paid to
()
25 look further.
It's an example of a study that is hardware AcsFEDERAL REPORTERS, }NC.
202 447 3700 Nationwide Coserage 8(G3364M6
6310 01 01 13 ysimons 1 oriented and is based on a modicum of data.
2 The pressure control study, which you will hear 3
about, is also an engineering evaluation.
That primarily 4
is in response to one event at Salem, and although it has 5
lots of hardware issues in it, it's more of a systems study 6
and a question of how does the man and the systems 7
interact.
8 Then there is a service water case. study.
That 9
one will have recommendations.
The recommendations that we 10 ought to have out of that ought to make 51-09.
There will 11 be a substantive report.
There will be about a man year 12 effort that goes into it and it will be a chunky piece of
()
13 work and you'll just hear how we get one of those studies 14 going.
15 Eo that represents the diversity of product.
16 Now I was asked to say just a' couple of words 17 about decay heat removal.
There was a 1985 case study on 18 loss of decay heat removal while in hot loop operation.
19 It's labeled as C-503.
That study is based on roughly 80-20 some-odd events of which eight or so were studied in 21 depth.
Recommendations came out of that study and it was 22 peer reviewed.
l i
23 Subsequent to the study there were at least 24 three substantive losses of decay heat removal, one at 25 Waterford, one at Songs and one at Diablo Canyon.
By the
/(CE. FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coserage 800-336- % 26
6310 01 01 14 1
time that the Diablo Canyon one came around it was clear rbysimons 2
that AEOD had done case studies, ins had been issued, SERs 3
had been written from INPO, SLERs from INPO had been 4
written and yet the frequency of events continued at too 5
high a level.
5 AEOD wrote a letter, a memo advising that action 7
be taken and that there be a generic communication to the 8
industry.
Now you have discussed the ensuing 50.54(f) 9 letter.
If you chose to have more of the technology behind 10 it, we can provide the report of course so you people would 11 have it and we could discuss it with the committee or 12 subcommittee as you desire.
()
13 I would like to say that one thing that now 14 comes out of it is we are trying to look ceyond loss of 15 decay heat removal in hot loop operation to the issue of 16 the prudence of operations in modes 4, 5 and 6, a little 17 bit bigger picture.
18 We go into the plant and we see in a regime in 19 which a lot of the tech spec requirements are gone and the 20 decay heat is down and the risk is surely down, and yet you 21 go into the plant and you see one diesel apart in pieces 22 and they are working on the aux feedwater pump and another 23 group is working on a starter transformer and you start 24 wondering what really is there.
I have nothing to report 25 out of it at this point, but we have initiated that work l
ACE FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coserage MG336-6M6
6310 01 01 15 t' )ysimons1 based on our experience with the mid-loop operation.
2 MR. EBERSOLE:
I would like to say at this time 3
that we do have this letter, but it was only partially 4
completed.
We had a look at it, you know, one of these low 5
loop type operations that we run between you and us that 6
really we get through the Commissioners.
7 We sort of elected, or I proposed that we take 8
this letter, which is still being written and it hasn't 9
been sent out yet and it hasn't gone to CRGR as far as I 10 know, and put a comment in to be considered just as a 11 consideration to do exactly what you say, to broaden the 12 scope to general safety matters in these modes and not just 13 this incident.
14 You know, in modes 4, 5 and 6 there are safety 15 problems and this is the outstanding one at the moment.
It 16 goes on to say "As a more general concern, it has come to 17 our attention that there has been little attention paid to 18 reactor safety when operating in modes 4, 5 and 6.
In 19 those the critical RHR systems are not safety grade 20 presumably on the basis that ECCS will safe the day.
21 However, the operation of the ECCS under these conditions 22 may not have been examined in sufficient detail."
23 I don't know how that's done.
So we'are acting 24 I think in sort of a productive mode by looking at this O
25 letter in its present temporary stage and we'll just go ACE FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coserage 800-336 6646
6310 01 01 16
([)ysimons1 back and' formulate these comments that the committee 2
approves.
3 This includes consideration later of the 4
possibility of open-cycle run-out, and you can just pump 5
the' water out on the floor all over the place if you'run 6
these things without consideration of exhausting your 7
coolant.
8 And, lastly, and maybe a too caustic comment, is 9
that we would prefer that the information generated by the j
l 10 operations be generated in the owner-operators by certainly 11 representation from the operations people, you know, like 12 the shift technical adviser or the shift engineers as well j
13 as Glenn calls the desk engineers so that we be sure that 14 they at that level which he calls in the trenches know what 15 the hell is going on.
16 These are suggestions that go in that letter as 17 an initial contribution from us to you all without going 18 any further at this time.
19 Does that sound appropriate?
Does the committee i
20 at large have anything to say about this.
21 CHAIRMAN KERR:
This is so far an Ebersole 22 position I think without a committee position.
23 MR. EBERSOLE:
Right.
They only had gotten this 24 copy here 10 or.15 minutes ago.
}
25 MR. REED:
Well, in addition to what Jesse has ACE FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coserage 800-336-6646
)
6310.01 01 17
/~T r,jysimons 1 said, I don't know whether you're going to have someone s
2 come up and discuss the Diablo incident or not.
Are you?
3 MR. ROSENTHAL:
No.
4 MR. REED:
Let me make a point.
I haven't 5
reviewed the material, but I do think you may have missed a 6
significant lead-in precursor to this incident, and that is 7
that this plant and many other plants with recirculating 8
steam generators do not have water box plenum chamber 4
11 9
' events and therefore they are required to go to half pipe, 10 or as they call it half pipe in order to drop the water out 11 of the steam generators, and that is a very healthy or easy 12 thing to do.
You get slugs and you're trying to get air to O
13 come down through the hot leg and run along the top of the 14 pipe and up into the steam generators.
15 Now the reason they have to do that is because 16 of a design deficiency, in my opinion, which is no water 17 box events.
However, some plants do have them.
Here is a 18 case where you might for the first time find a design 19 deficiency that would run the gamut of backfit and be cost 20 beneficial because it is a radioactivity reduction device 21 because you can vent the steam generators and get gaseous 22 radioactivity out before you open the man-ways and because 23 you can avoid this dangerous drain down to half pipe in 24 order to get access to the steam generators.
fs t
25 Did you follow me?
ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 800-3364M6
6310'01 01~
18 r'~3ysimons 1 MR. ROSENTHAL:
Yes.
V 2
MR. REED:
And I don't think you have mentioned 3
that in any of your work.
4 MR. EBERSOLE:
I don't think there has been any 5
mention of engineering voids at large like the absence of 6
any temperature.
You disconnect it.
Unless the RHR is 7
running, there is no transport mechanism to the 8
thermocouple and there is a need for consistent, 9
reasonably unambiguous temperature level instrument 10 protection through all modes.
11 MR. REED:
Well, I'm particularly concerned 12 about the design deficiency that leads the operator into a
(
13 Catch-22 trap.
He's trying to get the half pipe and get 14 the drainage out of his steam generators and you recall the 15 Ginna incident where'they dumped the steam generators after 16 they opened the man-ways.
This is an operator trap, and 17 once the trap is corrected for by design or modification, 18 then the procedures have to be corrected.
So you never do 19 go to half pipe unless there is some other reason way out 20 in the refueling event.
21 MR. EBERSOLE:
My problem with that a little bit 22 is so you are protected from going to half pipe, but 23 nevertheless you've still got to get down to the flange.
24 MR. REED:
There is nothing wrong in getting 25 down to the flange.
ACE. FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coserage 800-336-6646
6310 01 01-19 t'"Nysimohs 1 MR. EBERSOLE:
Well, I mean you've got to take V
2 the flange off or you've got to take the lid off.
3 MR. REED:
But you've got feedwater.
4 MR. EBERSOLE:
Yes, you've got feedwater but ---
5 MR. REED:
You have six or eight feet of water.
6 MR. EBERSOLE:
--- but if you're doing this 7
early on to catch the bad stick of fuel, you're still going 8
to boil out pretty quick.
9 MR. REED:
Well, you now have when you do raise 10 the lid, Jesse, and you must understand this, you have ---
11 MR. EBERSOLE:
Lots of water.
12 MR. REED:
--- many, many thousands of cubic
()
13 feet of water setting right at the lid, and as soon as the 14 lid comes up the water comes out.
15 MR. EBERSOLE:
Sure, that's true, but there are 16 exposure intervals in there where if you lose AC power or 17 something you're in trouble.
18 CHAIRMAN KERR:
Does this conclude your 19 presentation, Mr. Rosenthal?
20 MR. ROSENTHAL:
Yes.
21 Matt Chiramal, the Section Leader of the 22 Engineering Branch will now present the aux feedwater 23 study.
We have about three studies of about five minutes 24 each just to show you some details.
25 CHAIRMAN KERR:
May I ask a question about one ACE FEDERAL REPORTERS, INC.
202 347 3700 Nationuide Coverage 8@ 3364M6
6310 01 01 20 r~3ysimons 1 of the comments you made which I thought I understood to V
2 say you would make recommendations that you thought really 3
should be done but which would not be justified under the i
4 backfit rule, or something like that.
5 MR. ROSENTHAL:
If we think that there is an 6
issue which warrants attention and we know that it won't 7
pass the backfit rule, it will probably come out of our 8
shop as a suggestion.
If we make recommendations in 9
ge ne ra l, they should pass backfit.
10 There is an intermediate and undefined ground 11 that you do one at a time where judgment and prudence say, 12 hey, we think something out to be, whether the backfit rule
()
13 is justifie'd or not.
14 I really can't talk too much about them because 15 you grapple with those issues one at a time and custom fit 16 them.
17 CHAIRMAN KERR:
I would hope that if you 18 encounter a significant number of issues that you think 19 require attention but which don't pass the backfit rule 20 that you would make a recommendation that perhaps the 21 backfit rule be revisited.
I think it will take time 22 before you know that to be the case.
23 MR. EBERSOLE:
Can you or Ed or anybody make an 24 observation about what you consider the quality of response 25 to your recommendations?
I often consider you as kind of ACE. FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coserage 800-336-6M6
6310 01 01 21 I
l 1
(gj simons 1 an internal NTSP who may be heard or may not be heard.
y 2
MR. JORDAN:
Yes.
We have been described a 3
number of ways.
Our role really is to try to identify 4
issues only by intrinsic patterns or technical analyses i
5 that the Projects Office doesn't have the resources or time l
6 to spend on and to be persistent about our recommendations, 7
but to have a clear understanding of the safety 8
significance because se can spend an awful lot of agency 9
resources in making recommendations that didn't really have 10 a safety benefit, things that would be nice if they were, 11 but we do recognize resources that the agency has as well 12 as the utilities and we'll make that we are prioritizing O
13 our concerns and putting into the place that we have the 14 greatest ---
15 MR. EBERSOLE:
Well, are you getting the kind of 16 response that you think you deserve?
17 MR. JORDAN:
It's too early for me to say that.
18 I guess my view is on the decay heat removal issue that 19 things seem to be moving appropriately and I'm pleased with 20 that.
21 CHAIRMAN KERR:
Please go ahead.
22 MR. CHIRAMAL:
My name is Matt Chiramal.
I'm 23 the Section Chief of the Engineering Section.
24 (Slide.)
O 25 This is an example of an engineering evaluation ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 800-336-6M6
I 6310 01 01 22
- /~N r, jysimons 1 study done in my section by Chuck Shu who is not here 2
today, and it is entitled " Auxiliary Feedwater Pump Trips 3
Caused By Low Suction Pressure."
4 (Slide.)
5 This study was stated by an event that occurred 6
in January 1987 at Millstone 3.
The event occurred at 100 7
percent when they were testing the two motor-driven 8
auxiliary feedwater pumps, and one after the other they 9
tripped after they were started.
In fact, they tried to 10 restart the pump three times and it kept tripping under low-11 suction pressure, although at that time they did not know 12 the real cause of the failure and it took some O
13 investigation before they found out that it was the low-14 suction pressure because the alarms did not come in but the 15 pump tripped.
i 16 They tried pump B and it tripped once, and after 17 that they stopped the pumps and they went back to 18 investigate the event.
19 During the investigation the licensee determined 20 that the cause of trip was due to low-suction pressure that i
21 occurred while some oscillations were taking place on the 22 discharge side of the pump.
i 23 (Slide.)
24 Based on the common mode failure of two pumps 25 tripping on low-suction pressure, we decided to go in and ACE FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coserage -
800-336-6646
-6,330 01 01 23
()ysimons1 look for additional events and found three more in the data 2
base from 1981 to the present, I mean to January of
'87, 3
and those were at Trojan, D.
C. Cook and Zion 2.
4 In the Trojan and the Zion event both pumps 5
tripped on low-suction pressure.
I think Cook only had one 6
trip.
7 CHAIRMAN KERR:
One of those has a reference to 8
spurious low-suction pressure due to fluctuations.
9 MR. CHIRAMAL:
It's fluctuation, but there was 10 sufficient head in the condensate storage tank so that the 11 pressure itself was not really low.
It's just that the 12 fluctuations were low.
13 MR. EBERSOLE:
There are a couple of things that 14 come out of this.
One of them is just yesterday we heard
)
15 Northeast Utilities on Millstone 1, 2 and 3, and you have 16 No.
3.
17 MR. CHIRAMAL:
Right.
18 MR. EBERSOLE:
And out of that came an 19 expression of disgust on their part, which I share, and I 20 think the committee shares.
The Westinghouse design always 21 invokes aux feedwater to do its thing on every turbine 22 reactor trip.
So the challenge frequency is high on these 23 things, whereas on other plants you ran down the main feeds 24 on the GB and ---
25 CHAIRMAN KERR:
Jesse, I can't hear you.
J ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 800-336-6646
6310 01 01 24 Oveimoee1
~
sa 888aso'e=
series tais is em 2-extrapolation of what they said yesterday.
They were 3
disgusted with the design of the Millstone 3 aux feedwater 4
system in that they had to have it function every time they 5
had a reactor trip or turbine trip or generator trip.
'You 6
have to use these and the challenge frequency is very high i
7 compared to any other design, and I think it's an 8
intrinsic ---
9 CHAIRMAN KERR:
What is wrong with having the I
10 aux feed system work ---
11 MR. EBERSOLE:
What is wrong is the same thing 12 that is wrong with any safety system, you should minimize 13 the challenge frequency.
14 MR. REED:
I disagree.
15 MR. EBERSOLE:
Well, he says you exercise it, 16 but I say if you need it that is not an exercise, and they 17 need it.
Now they can recover main feed after a while, I 18 don't know, a half an hour or something.
19 MR. CHIRAMAL:
Well, in most of these cases once f
20 they found the root cause they were able to either bypass 21 the trip and shut the pump or throttle tne discharge and 22 start the pump.
1 23 MR. EBERSOLE:
Yes, sure.
It's not a smooth 24 translation into ---
O 25 CHAIRMAN KERR:
But I'm still curious about thi.s ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coserage 800-336-6M6
6310 01 01 25 rdysimons1 spurious low suction pressure.
^-
2 MR. CHIRAMAL:
The word " spurious" is used 3
because the actual suction pressure, the head that is 4
required to start the pump or run the pumps was sufficient, 5
but due to these transient effects there are some pressure 6
oscillations on the suction side that was sense by ---
7 CHAIRMAN KERR:
Well, pressure oscillation can 8
produce a low pressure, can't it?
9 MR. EBERSOLE:
Sure.
That's what it did and 10 that's why it tripped.
11 CHAIRMAN KERR:
Well, then it seems to me it was 12 not spurious ---
()
13 MR. CHIRAMAL:
Well, some of the corrective 14 actions give us some slight time delay to the system so 15 that it can override these transients and not trip.
16 CHAIRMAN KERR:
Okay, but it really was a low 17 pressure ---
j 18 MR. CHIRAMAL:
Oh, definitely, there was low i
19 pressure at that time.
20 MR. MICHELSON:
This problem you might have 21 found if you had looked up HPCI and RCIC turbines through a j
22 long-standing problem and put time delays on those NPSH 23 trips.
When the turbine starts and cranks up the pump the 24 pressure,goes down momentarily and trips it out again.
25 It's an easy fix.
ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coserage 800-336-6646
6310 01 01 26
()ysimons1 MR. CHI RAMAL:
There were many reasons for these 2
low-suction pressures.
In some cases it was because of 3
excess flow and some because of oscillations.
4 MR. EBERSOLE:
Another thing to notice from 5
yesterday is the fact that these pumps are protective by a 6
low-suction trip, whereas many critical pumps are not.
7 They are not protected against low-suction pressure so they 8
have to f ace the problem of having too low-suction pressure 9
and try to live through it and this leads to potential 10 d a. mage.
11 I have never found a consistent pattern of 12 identifying when you do and when you do not have a low 13 suction or low NPSH trip.
If you can find it for me, 14 great.
There is no standardization on when you protect a 15 pump against low NPSH.
16 MR. CHIRAMAL:
The safety significance of the 17 event is that spurious low-suction pressure trips can 18 constitute a common mode failure that can potentially 19 render the air system inoperable.
20 MR. EBERSOLE:
And that would apply to the steam 21 driven as well, wouldn't it?
22 MR. CHIRAMAL:
Sometimes the steam driven don't 23 have a trip, but some of them do have.
Ye s, it applies to 24 all types of trips, all aux feedwater pumps, both motor 25 driven and ---
ACE FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coserage 800-336-6646
6310 01 01 27
()ysimons1 MR. EBERSOLE:
And really that is not a costly 2
modification.
3 MR. CHIRAMAL:
No. it's not.
In fact, the 4
corrective action taken by licensees fall into two 5
different categories and one is that they provides a time 6
delay to override the transient low pressure and the second 7
one is they remove the low suction pressure entirely.
8 MR. EBERSOLE:
Were those all Westinghouse l
9 plants?
l 10 MR. CHIRAMAL:
Yes.
11 MR. EBERSOLE:
Okay.
There is a common mode, 12 Westinghouse.
13 (Laughter.)
14 MR. CHIRAMAL:
Although in the case of Cook they 15 replaced it with an alarm and operator action that if he 16 was to get a low pressure alarm the operator will trip the 17 pump and find the cause and restart it.
18 (Slide.)
19 These were the conclusions reached by the study.
20 One is that pressure oscillations or 21 fluctuations can occur in suction lines and create a 22 transient pressure drop despite suction available from a 23 source like this condensate storage tank, and these 24 momentary low pressures may trip the pumps.
i 25 Two, of course, is a low suction pressure trip 4
ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
1 6310 01 01 28 f()ysimons1 is provided to protect the pumps from cavitational damage, 2
and without the low-pressure trip or some sort of action to 3
trip the pump the pump may be damaged and the protection 4
may not be adequate.
5 MR. EBERSOLE:
No. 3 up there says you did' a 6
certain thing.
l 7
MR. CHIRAMAL:
Well, actually we have not, or 8
the study is not completed as yet.
We are recommending 9
that NRR issue an information notice to complete this 10 testing.
11 MR. EBERSOLE:
What I want to ask you is is 12 there a-follow-up to that?
O 13 MR. CHIRAMAL:
Well, the follow-up to that will 14 be we will go back in a year's time and review the data 15 again to see has there been a decrease in the number of 16 evente.
17 MR. EBERSOLE:
Yes, but will you follow up that 18 an information notice was issued?
19 MR. CHIRAMAL:
Yes, we will.
20 MR. EBERSOLE:
And.then will you get a response 21 from the industry sector?
l 22 MR. CHIRAMAL:
Yes.
23 MR. EBERSOLE:
And then you'll stand back and 24 see what happens?
O 25 MR. CHIRAMAL:
No, we don't get a response from ACE. FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800-336-6M6
6310 01 01 29
()ysimons1 the industry.
2 MR. EBERSOLE:
So it's open ended ---
3 MR. CHI FAMAL :
The resident inspectors and the 4
regional inspectors do look at the information notice and 5
see if the licensee has, you know, received it and gone 6
through his process of review.
7 MR. EBERSOLE:
There is some certification that 8
he got it anyway.
9 MR. CHIRAMAL:
Right, and some certification 10 that he has done something with it.
11 CHAIRMAN KERR:
What is the cause of pressure 12 oscillations?
Are they caused by the pump start up?
>]k 13 MR. CHIRAMAL:
In some case it's because of 14 excess flow of the discharge side or in some cases because 15 the governor controlling the turbine driven feedwater pump 16 was opened too fast or in some cases they started two pumps 17 exactly at the same time more or less together, that type 18 of pump.
19 And that's it.
20 MR. ROSENTHAL:
Thank you, Matt.
21 The next speaker is Sanford Israel.
22 This event was originally picked off the daily 23 report and the regional report.
It wasn't until that LER 24 was read that there was a greater appreciation for the O
25 event.
ACEJFEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800-336 6646
6310 01 01 30 A
f( jysimons 1 MR. ISRAEL:
I'm going to talk about a little 2
study that I'm involved in dealing with mostly pressure 3
control in PWRs.
4 (Slide.)
5 What ~I'm talking about here is really the 6
ability to depressurize the primary system in a PWR.
7 MR. REED:
I 'm ale rt.
So be careful.
8 MR. EBERSOLE:
It is automatically Combustion 9
Engineering.
10 MR. ISRAEL:
No. I'm only dealing with 11 Westinghouse, but some of these thoughts carry over to 12 other designs as well.
C) 13 (Slide.)
14 As Jack mentioned, the stimulus for this study 15 was the Salem event from last August.
I think the 16 committee was briefed on this particular event, and this 17 was the event where technicians shorted out a bus, an 18 instrument bus.
They got a false safety injection signal 19 of a plant trip, and they tried to load up the transformer 20 with the emergency loads and they started bouncing from one 21 transformer to another.
After about 30 seconds basically 22 they got a loss of of f site power and in ef fect they 23 stripped all the loads off the emergency buses and then 24 loaded up a new set of loads, which is loss of offsite O
25 power plus safety injection signal safety loads on the ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 800-336-6M6
6310 01 01-31 Ar,j simons 1 emergency diesels.
y s
2 I'm sort of interested in the backend of this 3
event and I have pursued the electrical portion.
That was 4
being pursued by the region with a great deal of 5
interaction between the utility and the region on that 6
particular aspect.
l 7
Uhat I was interested in is when we got our l
8 first safety injection signal during this event obviously 9
the high-pressure safety injection pumps came on and they 10 obviously were putting water into the primary system 11 through the cold leg here, and they were also providing 12 seal cooling to the reactor coolant pumps which were still
()
13 operating at this time.
14 Subsequently the loss of offsite power signal 15 came in and they stripped everything off and they reloaded 16 the diesels.
Now Salem happened to have three diesels, or 17 has three diesels, but only two were operable.
One was out 18 for maintenance.
19 At this particular plant when they reloaded the 20 diesels with combined loss of offsite power and a safety 21 injection signal, they don't load component cooling water 22 onto the diesels.
Those are not automatically loaded onto
(
23 the diesels.
So the other safety systems were loaded on 24 except for the component cooling water.
O 25 The reactor coolant pumps were now operating ACE FEDERAL REPORTERS, INC.
202-347-3700 Natio_nwide Coter_ age MG336-6646
6310 01 01 32 l
t()ysimons1 without any component cooling water.
The operators 2
discussed it amongst themselves and they said, gee, our i
3 training has told us that it is not really a good idea to 4
keep the reactor coolant pumps running more than five 5
minutes without component cooling water.
So they turned 6
off the reactor coolant pumps.
7 MR. REED:
Clarify that, please.
If you had a 8
loss of offsite AC, you're telling me that the main reactor 9
coolant pumps are running?
10 MR. ISRAEL:
This was a false loss of offsite 11 power.
12 MR. REED:
Oh, excuse me.
O 13 MR. ISRAEL:
So what happened was that they 14 loaded their emergency loads on the diesels, but they still 15 didn't have offsite to drive the reactor coolant pumps.
16 CHAIRMAN KERR:
Component cooling water was 17 apparently not conndeted to offsite power.
18 MR. IS RAEL:
Component cooling water is a safety 19 system, and it's loaded on an emergency bus ordinarily.
It 20 would be ordinarily driven by offsite power obviously 21 during normal operation through the transformers, the i
22 switch on transformers.
I 23 But when I have this, what I will call false 24 loss of offsite power, they n'ow stripped off the component O
25 cool water off the emergency bus, they stripped off
/\\CE. FEDERAL REPORTERS, INC.
1 202-347 3700 Nationwide Coverage 800-336 6646
(
6310 01 01 33
_()ysimons1 everything off the emergency bus and they started loading 2
up with specific components related to a safety injection 3
signal plus the loss of offsite power.
4 MR. EBERSOLE:
I'm being to suspect something.
5 You just said component cooling was a safety system.
6 MR. ISRAEL:
That's right.
7 MR. EBERSOLE:
Now the component cooling system 8
may have safety systems or they may not, and they may do 9
anything or virtually nothing.
They can be or cannot be.
10 You can even get tid of them and many plants don't have 11 them.
But here you say it's a safety system.
Now let me 12 ask you a question.
O 13 CHAIRMAN KERR:
Jesse, let me complete my 14 question, please.
Why did the component cooling water pump 15 not load on the diesels?
Is it not supposed to or was it 16 an error?
17 MR. ISRAEL:
I'm sorry, I didn't hear you.
18 CHAIRMAN KERR:
Why did the component cooling 19 water pump not load on the diesel?
20 MR. ISRAEL:
It's not automatically loaded 21 because I presume they have not sufficient capacity on the 22 diesels to ---
l 23 MR. EBERSOLE:
I was coming precisely to that.
24 That sounds like a compromised design allowed by regulatory O
25 which shouldn't have been allowed.
Somebody said, oh, my i
l l
ACE FEDERAL REPORTERS. INC.
202-347 3700 Nationwide Coserage 800-33 4 6646
6310 01 01 34 A
r,_jysimons 1 God, I haven't got enough power and we'll start it later on 2
by hand if we have enouah cower.
That's a terrible 3
compromise.
4 MR. ISRAEL:
This plant may be unique in terms 5
of this particular set-up.
Most plants obviously 6
automatically load the component cooling water.
7 CHAIRMAN KERR:
Does this imply that component 8
cooling water is normally not required when one looses 9
offsite power?
10 MR. ISRAEL:
With the simple loss of offsite 11 power, the component cooling water would be loaded on the 12 emergency buses.
We are now talking about loss of offsite O
13 power in quotes, plus a safety injection signal.
There is 14 a whole new set of components that are now loaded on the 15 emergency buses.
16 MR. EBERSOLE:
What does component cooling do?
17 MR. ISRAEL:
The component cooling water 18 provides cooling for various ECCS pumps, it provides 19 cooling for your containment spray ---
20 MR. EBERSOLE:
Stop there.
21 MR. ISRAEL:
Okay.
22 MR. EBERSOLE:
Don't we now have in full view a 23 severely compromised design due to generosity on the part 24 of NRC?
25 CHAIRMAN KERR:
Jesse, please let me understand ACE. FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coterage 800-336-6646
L 6310 01 01 35
()ysimons1 what.is going on before drawing conclusions, and I still do 2
not understand why if you need component cooling water for 3
the emergency systems you don't load it onto the emergency 4
diesel.
5 MR. EBERSOLE:
He just said you haven't got 6
power enough.
7 MR. ISRAEL:
That was a decision that was made 8
obviously when the plant was designed.
The rationale for 9
this was that the diesel CS pumps, if in fact I had a LOCO 10 or a safety injection signal in a loss of offsite power 11 where I would need the ECCS pumps, they had indications 12 from the manufacturers that these pumps would operate O
13 satisfactorily during the safety in'jection phase where they 14 were using cold water.
15 This component cooling water here is only 16 cooling the seals and is not used for the oil coolers on 17 the pumps.
They are only used for cooling the seals.
18 MR. WARD:
You don't really need it for the 19 safety injection pumps until recirculation; is that it?
20 MR. ISRAEL:
That's right.
That's what they are 21 saying is that I don't need it until recirculation.
By the 22 time I get to recirculation I will strip off some of the 23 other loads that I may not need and I will now load on the 24 component cooling water.
O 25 MR. REED:
You must bear with us.
You'll find ACE-FEDERAL REPORTERS, INC.
202 4 47-3700 Nationwide Coserage 800-336-6646
6310 01 01 36
()ysimons1 there is a great deal of similarity between this committee 2
and the Congressional committee trying Colonel North.
3 (Laughter.)
4 MR. NOVAK:
Remember, we are not here just to 5
discuss component cooling, and you are going to lose your 6
punch line.
7 (Laughter.)
8 MR. IS RAEL:
What has happened now is once I 9
have turned off my reactor coolant pumps I no longer have 10 my normal spray system which requires large Delta P between 11 the cold legs and the hot legs in order to pump gater.
12 Even so, with this event, one of the normal O
13 spray lines was isolated and had been isolated I think for 14 about three months because of a leaky valve.
15 What they also noticed was if they kept pumping 16 in water into the primary system they had normal ECCS 17 injection also through the reactor coolant seal line.
They 18 decided to fill the pressurizer and they started opening up 19 the PORV.
The plant has two PORVs.
One PORV had been 20 isolated previously because the PORV was leaking.
21 They were now challenging the remaining PORV 22 because it kept pumping water into the primary system, and 23 they went to isolate the normal ECCS injection which was 24 giving them the highest flow, but they couldn't do that O
25 because it happened to be that the valve, the isolation ACE. FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coserage 800-336-6646
i 6310 01 01 37 t()ysimons1 valve happened to be on that power supply related to the 2
diesel generator that was out of service.
So they had to i
3 keep pumping in water through here.
4 They couldn't turn off the pumps themselves 5
because they had to provide seal cooling to the reactor 6
coolant pumps or end up with a seal LOCA.
7 MR. EBERSOLE:
Even though they were static?
8 MR. ISRAEL:
That's a problem that is being 9
worked on and it ties into the station blackout situation.
10 so we now have a situation, and this is what 11 caught my eye, where I no longer have normal spray, and I 12 couldn't actuate the alternate spray which comes off the 13 high pressure pump because this valve also happened to be 14 off that power supply to the diesel that was out of 15 service, and the only other remaining depressurization 16 component I had was one of the two PORVs.
17 I looked at this event and what caught my eye 18 was a concern that there were a number of circumstances 19 where depressurization is important, and here we have 20 provided in essence three different systems to provide 21 primary system depressurization, but in this one event many 22 of them disappeared.
23 MR. REED:
Let me ask a question about the 24 PORVs.
Normally if you isolate a PORV train, it is capable 25 of being reopened automatically or manually remote simply ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage MX)-336-6646
6310 01 01 38
)ysimons1 by opening the block valve.
Are you saying both the block i
2 valve and the air operated valve were closed and could not 3
function?
4 MR. IS RAEL:
No.
The block value, I can't say 5
that it was not functioning.
We probably could have opened 6
up the block valve to actuate the other PORV.
7 MR. REED:
But the block valve was not on 8
automatic pressure reduction?
9 MR. ISRAEL:
No, that is manual.
The block 10 valve is on manual.
11 MR. REED:
The other thing I note is I'm a 12 little surprised to see the alternate spray line coming off O
13 off the high pressure safety injection pump unless that 14 pump has ---
15 MR. ISRAEL:
That's a charging pump.
16 MR. REED:
Oh, well, those combined ---
17 MR. ISRAEL:
I simplified the diagrams to make 18 life easier for us.
19 (Slide.)
1 20 MR. REED:
Well, that would not be a very common 21 Westinghouse piping system design because I don't think 22 they have combined the pump duty all that much.
Are the 23 centrifugal on?
24 MR. ISRAEL:
It depends on the plant.
Those 25 plants that have centrifugal charging pumps would in effect ACE FEDERAL REPORTERS, INC.
202-347 3N)0 Nationwide Coserare 800-336 6646
)
6310 01 01 39 iO r,_jysimons 1 have this type of system, and there are those plants that 2
only have the centrifugal for the HPCI system and there are
?
3 other plants that have the centrifugal plus the ---
4 MR. REED:
What they would normally do is take 5
the two pipes from two loops and bring them through the 6
normal spray nozzle and bring the high pressure safety 7
injection one through and tie it into an alternate line.
8 MR. ISRAEL:
You're talking about they may be 9
tied together.
That was my rendition.
10 MR. REED:
Well, we're happy you mentioned the 11 points of pressure reduction.
We find that there are 12 reactors out there, CE System 80's that don't have much for O
13 that.
14 MR. ISRAEL:
My concern, as I say, is to support 15 your local reactor operator, especially since there are 16 events that have safety significance, particularly the 17 steam generator tube rupture which challenges the defense 18 in depth concept, fails the primary system and it also a 19 potential for failure the containment pressure boundary as 20 happened in Ginna when they had the steam generator tube 21 rupture and it opened up the steam generator safety outside 22 of containment.
23 We are also interested in depressurization for a 24 natural circulation cool-down.
If you have an extended 25 loss of offsite power you only have a limited amount of ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 8(0 336-6646
R l
6310 01 01 40
()ysimons1 condensate storage tank water and ultimately you have to 2
get down onto your RHR, and you also have feed and bleed 3
which is a backup to total loss of main feedwater.
4 MR. EBERSOLE:
Why is depressurization needed 5
for natural circulation cooling through the steam 6
generators?
7 MR. ISRAEL:
Because I really want to.get back S
down onto my RHR system.
9 MR. EBERSOLE:
Eventually.
10 MR. ISRAEL:
Eventually.
11 MR. EBERSOLE:
But you're telling me I can't get' 12 down by simply rejecting heat from the secondary side C:)-
13 becau'se it takes too long, right?'
14 MR. ISRAEL:
I have to cool off my pressurizer 15 which is initially at 650 degrees, and I have an upper head 16 that's at a high temperature.
17 MR. EBERSOLE:
That's the CE problem that's out 18 there and you're never going to ---
19 MR. MICHELSON:
Why do you have to cool off your 20 pressurizer?
21 MR. EBERSOLE:
Because that is what makes the 22 temperature ---
23 MR. MICHELSON:
No, no, but there is nothing 24 saying that you've got to do this in a few hours either.
O 25 MR. EBERSOLE:
Well, that's the issue, how long ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Cmerage 8(0-336 66.16
l i
6310 01 01 41 f()ysimons1 is it.
2 MR. MICHELSON:
Well, there is nothing saying 3
that you have to for this case.
4 MR. REED:
Well, then you risk pushing water 5
through the tube rupture continuously during that time.
6 MR. MICHELSON:
If steam tube rupture is the 7
accident that he is dealing with, that's different.
Is 8
that the one you are trying to ---
9 MR. IS RAEL:
You need depressurization for all 10 of these events.
11 MR. MICHELSON:
For tube rupture you do.
12 MR. ISRAEL:
Well, but you also need it -- well,
/')
13 when I say you also need it, obviously one can improvise as 14 an the event ic going along.
I always like the concept 15 that we have procedures as a preferred way of doing rather 16 than challenging the operator to improvise in the middle of 17 the event.
18 If you have an extended loss of offsite power, 19 you have a limited condensate storage tank supply, and what 20 you really want to do is to get down on RHR.
When I'm 21 talking about loss of offsite power, I'm considering that 22 the diesels work so that you have your heat sink, your 23 ultimate heat sink and all the other pumps are working.
24 So you really want to get down onto RHR, and in
~
25 order to get down to RHR you have to depressurize cooling ACE-FEDERAL REPORTERS, INC.
i 202.M7-3700 Nationwide Coserage 800 336-6M6
i 6310 01 01 42
-G tsjysimons 1 and depressurize the primary system.
Cooling is no problem 2
in terms of the steam generators being available 3
presumably, but you still have to depressurize and you have 4
to provide that function.
5 (Slide.)
6 To wrap this up, what we are planning to do with 7
this is we drafted an information notice that we will send 8
over to the Generic Communications Branch over in NRR to 9
process.
We will also transmit our study to the tech spec 10 group people over at NRR to fold into their ongoing tech 11 spec improvement programs as appropriate.
12 Similarly we will send the study over to the
()
13 NRR.
There is a group over there that is still responsible 14 for standard review plans even though they aren't any new 15 plants on line, but at least it gives them an opportunity 16 to codify a position in terms'of steam generator tube 17 rupture and depressurization systems.
18 And, lastly, we'll send the study over to 19 Research.
The Accident Managen.ent Program Group in the 20 Division, whatever the division is, they have an ongoing 21 program where they are concerned with accident management, 22 with accidents of less than core melt, design basis 23 accidents, and there were some corollary or some subsidiary 24 items that are in my report dealing with the use of 25 alternate spray, pressurizer sprays and concern for upper ACE. FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage RX)-3.W6646
6310 01 01 43 r()ysimons I
head temperatures with regard to depressurization.
2 MR. EBERSOLE:
In the large or generic context, 3
this opens up a view into I guess the general area of where 4
are there designs where it is allowed a critical service to 5
be denied adequate power during a preliminary interval, but 6
it has to be manually energized later when you can afford 7
to cut off the other thing that is using too much energy.
l l
8 MR. ISRAE L:
I did not focus on that, Jesse.
I 9
think this may be unique, but I'm sure the staff reviewed 10 that when the plan came up for analysis.
11 MR. EBERSOLE:
I can imagine other situations 12 where you are denied a full-scale parallel energization of O
13 critical equipment and then an operator has to dance 14 through this in sequence.
15 MR. REED:
Well, it all depends on in your 16 initial design of whether or not the system like a 17 component cooling system is considered to be a safety 18 system.
19 MR. EBERSOLE:
Well, here it was anyway.
20 MR. REED:
In many of the earlier PWRs component 21 cooling was not considered to be a safety system and 22 neither is the air systems and we are debating ---
23 MR. EBERSOLE:
Well, clearly it was here because 24 you've had to run these systems.
25 MR. IS RAEL:
It was clearly used here because ACE FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverace Rott136 6M6 l
6310 01.01 44 r)ysimons1 you used this as a heat sink during the ---
2 MR. EBERSOLE:
It's-only after the pumps are run 3
up that you've got to have it.
4 MR. IS RAE L:
Aside from a support to normal ECCS 5
pumps you have to ---
6 MR. REED:
You're saying that component cooling 7
is needed for main coolant circulating water pump ---
8 MR. EBERSOLE:
For ECCS pump operations.
9 MR. REED:
Well, I would question that.
There 10 are many, many ECCS pumps that run without component pump 11 cooling.
12 ME. EBERSOLE:
He said there were.
O 13 MR. ISRAEL:
I'm talking about the heat sink in 14 the recirculation phase post-LOCA.
15 MR. REED:
That goes to the residual heat 16 exchangers?
17 MR. ISRAEL:
That's right.
18 MR. EBE RSOLE :
I think you should extrapolate to 19 the broader issue.
20 MR. IS RAEL:
Pardon me?
21 MR. EBERSOLE:
I think you should extrapolate to 22 the boarder issue.
What other plants have potentially in
' hey have to program these critical loads 23 their design that t
24 as time demands them?
25 MR. ISRAEL:
I did sort of a cursory study, and ACE. FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserase 800-336-6646
I
\\
6310 01 01 45
()ysimons1 I kept coming up ---
2 MR. REED:
I can't resist this.
Many plants d
3 have reliable diesel generators, Jesse, because that 4
affords a smaller size.
5 (Laughter.)
6 MR. WYLIE:
Well, in this case here the diesel l
7 is out of service, one of them.
8 MR. ISRAEL:
That was another contributory ---
9 MR. WYLIE:
It didn't have an N plus 2 design.
10 (Laughter.)
11 MR. ISRAEL:
Ye s, they had three diesels.
12 MR. WYLIE:
Three?
~
13 MR. ISRAEL:
Yes.
14 MR. NOVAK:
Sandy, thank you.
15 I think your comments are well taken.
What we 16 were trying to do by this study was to ascertain whether 17 the status of the plant initially plus the events that 18 occurred put the operator in a position that made pressure 19 control difficult.
Had that initiating event been 20 something like a stream generator tube failure, it would 21 have made more headlines clearly.
So your comments are 22 well taken.
23 I think this ends up being an engineering 24 evaluation now where we take the " suggestions" and we will O
25 attempt to point them more specifically.
We want to go ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 8 % 336 6646
l 6310 01 01 46 t()ysimons 1
back to the licensee and the region to see what they have 2
done to evaluate the status of the plant that they permit and we want to communicate these kinds of 3
to operate r 4
things to the owners group and to share this information.
4 5
I think the first blush by this agency in 6
looking in at the event focused on the electrical side.
7 It's only when you have a chance to look at the whole event 8
and put it together and maybe start out a little 9
differently that the pieces look different.
This is one of 10 the drives of the event, and I think your comments on 11 component cooling water are very valid.
12 Peter Lam is going to talk about the case study 13 that we are just starting out with.
14 Go ahead, Peter.
15 MR. LAM:
Thank you.
16 I would like to share with you very briefly, 17 perhaps in five-minutes time if I can march you through 18 this, on both the process as well as the product of a AEOD 19 case study.
20 As Jack said to you earlier, an AEOD case study 21 is a priority office work product for which we commit a 22 significant amount of resources to do.
We intend to spend 23 approximately one to one and a half man-years on the study 24 before it is finished.
O 25 (Slide.)
6 ACE FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Cowrage 800-336-6646
I 6310 01 01 47
()ysimons1 Let me share with you about our process.
Since 2
it is a study that requires a great deal of our resources, 3
we would spend great amount of time to plan for this study 4
and then spend a significant amount of time in the 5
collection and analysis and evaluation of the data.
6 The first two bullets are involved with the 7
collection of the data, and after collecting the data we 8
would do an analysis of that data in the area of the causes 9
and sequence of events.
10 After the analysis, we would perform an 11 evaluation of the data in terms of the safety 12 significance.
We intend to do both a qualitative and a 0
13 quantitative evaluation.
14 Somewhere along the way before the study is 15 completed, we would get an AEOD higher management 16 briefing.
At that time the potential direction of the 17 study would be finalized.
18 Then the step before finalizing the report would 19 be a peer review process for which we would send the 20 report, a preliminary draf t of the report with Mr. Jordan's 21 signature to all the program offices, to the regional 22 offices and to INPO, NSAC, the owners and perhaps the 23 vendors.
24 We allow them typically 30 to 60 days to respond 25 to us, and at that time we would incorporate their comments ACE FEDERAL REPORTERS, INC.
202 347-3700.
. Nationwide Coseras:e 800-336-6646
6310 01 01 48
(}ysimons1 to the extent feasible and practicable into our report'and 2
then we would finalize the report.
3 As Jack said to you earlier, we would develop 4
the recommendations only if we believed the recommendations 5
would pass the 50-109 backfit rule or which a vigorous cost-6 benefit analysis may be required.
7 MR. MOELLER:
Do the other types of studies, the 8
technical reports and the engineering studies which we 9
heard mentioned earlier, undergo peer review and the plan.
10 is approved and so forth?
11 MR. LAM:
No.
12 MR. MOELLER:
Well then why the difference in a
(
13 case study and those types of reports?
14 MR. ROSENTHAL:
I think it's simply 15 proportionate to the amount of effort you plan on spending 16 to modulate your management of the report.
17 MR. MOELLER:
So a case study is a pretty big 18 deal.
Earlier you said that some of them could be three or 19 four pages long and done in and out.
20 MR. ROSENTHAL:
An engineering evaluation could 21 be a few months and then there is some normal interaction 22 between the people doing the job.
23 MR. MOELLER:
And you said a man-year or a 24 person-year or more on a case study?
25 MR. ROSENTHAL:
Right.
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 800-336-6646
6310 01 01 49 r()ysimons1 MR. LAM:
To shed more light on the scope of the 2
study, we are looking at perhaps more than a thousand 3
events over the past five years encompassing all the 4
vendors.
That may be you some idea of the scope of the 5
study.
6 MR. MICHELSON:
Which studies result in 7
recommendations that you expect at least to get an action 8
or replies on?
9 MR. ROSENTHAL:
Case studies primarily, although 10 when we do an engineering evaluation and have a suggestion 11 and it's something that we think has merit, then we'll do 12 targeted distribution.
)
~
13 MR. MICHELSON:
But it's only a suggestion, 14 isn't it?
15 MR. ROSENTHAL:
Right.
i 16 MR. MICHELSON:
That's the difference between a 17 case study and a technical review is, one, you suggest 18 things that might be considered, and iti this case you are 19 really actually making firm recommendations for which 20 actions have to be taken one way or the other.
21 MR. LAM:
And, furthermore, the AEOD case study 22 recommendations require a formal response from the 23 recipient, and we would be put on at what we call an AEOD 24 recommendation tracking system which is formally tracked 25 every six months as to its status in the contacts.
So it ACE. FEDERAL REPORTERS, INC.
J 202 3.87 3700 Nationwide Coverve Mn116_fM6
6310 01 01 50
- ()ysimons1 is a priority item.
2 MR. EBERSOLE:
Don't pull that away.
The first-3 three ballets are missing something and it's interesting 4
that thoy follow each other in sequence and then they are 5
difforent from the rest up there.
6 MR. LAM:
Are you talking about the first slide?
7 MR. EBERSOLE:
No, that one right there, the 8
first three bullets, biofouling of pumps, sediment 9
deposition and corrosion / erosion of components.
10 What's missing is the absence of the 11 consideration that failure may not be easy and progressive 12 and detectable like a leak before break.
It may be that 13 these systems are caught in the biofouling or in the 14-sediment deposition context or corrosion / erosion or sudden 15 onset of any of these phenomena to absolutely plug i
16 everything there is in sight or the pipes to fail in a 17 seismic event or the sediment to come down around your ears 1
18 and plug all uptakes because of rains or floods or 19 earthquakes.
20 So this should be a subset of statements up 21 there that says the consideration is firmly a sudden 22 stoppage of all water.
23 MR. LAM:
I agree.
Mr. Ebersole, this slide was 24 prepared an an illustration to the committee as to wha.t O
25 type of mechanism we are dealing with.
/\\CE FEDERAL REPORTERS, }NC.
202-347 3700 Nationwide Coserage 800-336-6646
.6[V 10'02 02 51 marysimons 1 MR. EBERSOLE:
I think the rest of them are all 2
sudden anyway.
3 MR. LAM:
Yes.
As the committee is fully aware, 4
you know all of these.
They are up here as an illustration 5
with the thousand event data base, and we are slicing the 6
data which plan is planned by ---
7 MR. EBERSOLE:
Good.
Great.
8 MR. LAM:
And the safety significance really of 9
the study can be summarized in one sentence because of the 10 fact that we are dealing with a system failure has an 11 impact on multiple systems.
As you are aware, it had an
(])
12 impact on RHR heat exchangers, cooling water heat 13 exchangers, the diesels, everything.
14 MR. EBERSOLE:
The whole works.
15 MR. LAM:
And because of it, we are seeing many 16 actions that have been taken in the past five to ten years 17 from both the licensee, the NRC and various industry 18 groups.
We are talking about information notices, 19 bulletins, AEOD case studies, INPO SOER and INPO SER, EPRI 20 symposiums, General Electric as well Westinghouse 21 information notices.
22 What we intend to do is take a step back now to 23 do what we call a more integrated and coherent evaluation
()
24 of the system as to how the system may fail as demonstrated 25 by operational data, why did it fail and what the ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Cowrage 800-336-6646
)
6310 02 02 52 h
mdrysimons 1 corrective action has been and what they should be.
2 MR. EBERSOLE:
I think out of this may fall a 3
kind of a summary conclusion that you would be better off 4
if you had air to water heat exchangers and you didn't have 5
any service water to mess with, which sometimes have to do 6
when they are out in the boon docks or deserts, and a lot 7
of this would go away.
8 MR. LAM:
Yes.
9 That's all I have.
10 MR. NOVAK:
Thank you, Pete.
11 Uhat I would like to do now is have Martin 12 Williams who is.the Chief of the Trends and Analysis Branch 13 give you a quick summary of a different type of analysis.
14 MR. WILLIAMS:
The Trends and Patterns Analysis 15 Branch picks up the activities that were formally in the 16 Program Technology Branch of AEOD.
Among those activities 17 were the maintenance of the agency's data bases for 18 operating experience, and throughout the reorganization 19 those activities have been preserved and there has been no 20 perturbation of the data bases and continuity prevails.
21 (Slide.)
22 The trends and patterns activities include, as I 23 said, the management of the data bases and trends and
{}
24 patterns analysis which consists of a systematic screening 25 of the data bases routinely and generally we publish annual ACE. FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800-336-6646
)
6 10 02 02 53
'drysimons 1 reports on scrams and ESF actuations and the like.
nt 3
2 Then we use the data base to answer specific 3
questions, and we'll have an example later on with Bob 4
Dennig on new plant activities in one of these areas.
5 (Slide.)
6 The data management activities specifically 7
include the collection, screening and storage of the 8
information, and then the screening of information and the 9
feedback of information also.
He look for abnormal 10 occurrences and report those to ' ngress, and we also 11 report to the NEA for IRS reporting.
12 We include in those activities the generation of 13 the data base which means the coding of the information, 14 support access by the other offices to the data bases and 15 then we formally evaluate the quality of our data bases, 16 both SCSS and the industry's data base.
The Nuclear Plant 17 Reliability Data System is also the subject of a formal 18 evaluation by our branch semi-annually.
19 We also occasionally revisit the adequacy of the 20 need for changing of the reporting requirements.
Right now 21 we are looking at the 50.72 and 73 reporting requirements 22 to see if we need any changes there.
23 (Slide.)
(
24 On the trends and patterns side again in a 25 little more detail.
We continued the analysis of all ACE FEDERAL REPORTERS, INC.
202 347-3700 Natiorinide Coserage 800-336 6646
6310 02.02 54 O
twarysimons 1 reportable events, both 50.72 for short-term reporting, and 2
then after we get our licensee event report in, we go with 3
a 50.73 analysis.
4 We normally trend all scrams, ESF actuations, 5
tech spec violations and safety system failures and the 6
other reportable events.
We are also doing more component 7
analysis, specifically the trending of failures and the 8
component reporting system of the NPRDS.
Right now we have 9
a study essentially completed on feedwater regulating 10 valves and bypass valves and we are studying other BOP 11 equipment right now based on trends and patterns work of 12 the NPRDS.
So those studies will be coming out shortly.
13 MR. EBERSOLE:
Out of this squeezing process and 14-filtration process where is the I guess I would say the I
15 product that you would call at least related to cosmic or 16 global issues that we ought to look at real hard?
17 MR. WILLIAMS:
I have an example.
Across the l
18 industry we look at right now the initiators of a lot of l
19 events.
Routinely we look at the initiators of scrans, and 20 I have some example charts here that we can look across the 21 industry and it's a bit more global, and other issues such 22 as the startup of new plant which is a bit more global.
23 MR. EBERSOLE:
You mentioned initiator of 24 scrams.
Do you look at scrams in the context that a scram 25 ought to be a move in the direction not of hazard but of ACE FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coserage 800-3364M6 t.
)
i l
67L0 02 02 55 V
marysimons 1 safety, namely, you stop the fission process and don't 2
invoke the need for critical reserve equipment.
You just i
3 ride right down main feedwater with the service system 4
constant, and you just stop the fission process period.
5 The rods go in and that's the only think you ask to work.
6 Shouldn't that be a design fundamental requirement, that on 7
a scram you should move to a safe state and not to one in a 8
challenge condition to safety equipment?
9 MR. WILLIAMS:
Well, that's true, and we have 10 taken ---
11 MR. EBERSOLE:
See, Westinghouse doesn't do
()
12 that, which is my complaint.
They don't do that.
.And any i
13' other system that asks emergency equipment to jump up and 14 do their thing, I don't like that and I don't think it's 15 right.
16 MR. WILLIAMS:
I think the staf f shares that 17 view on challenges to the plant.
1R MR. RBERSOLE:
It should be benevolent.
19 (Slide.)
20 MR. WILLIAMS:
As an example of some of the 21 scram work, I have put in the handouts two slides.
The 22 slides are fairly provocative regarding questions.
He 23 included them as simple examples to give you an idea of the
()
24 depth that we trend and pattern, the information.
25 The cognizant engineer for this activity is Mr.
ace-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserace luK)-336-6646
6310 02 02 56 (fnurysf..auns 1 Lawrence Bell who is the audience with us.
This is again 2
just an example, and we could spend some time talking about 3
all these different categories, but tais is an example of.
j 1
4 the balance of plant hardware that caused trips above 15 5
percent power in 1986.
6 You can see that we cross-cut the information by 7
the system where it was initiated.
l 8
CHAIRMAN KERR:
Why did you restrict yourself to 9
the above 15 percent power?
10 MR. WI'LLIAMS:
Well, we didn't restrict 11 ourselves other than for synchronization past the early fg 12 startup phase.
We have all of the information on all trips V
13 and we have cut it a lot of different ways, and Mr. Bell 14 has certain statistics that we would be happy to share with 15 you on any ---
16 CHAIRMAN KERR:
I wasn't being critical.
I was 17 just curious as to why you chose 15 percent power as a cut-I 18 off point.
The 15 percent is somewhat arbitrary.
19 MR. BELL:
It is arbitrary.
Looking at the data 20 for three years, nothing really changed as far as what was 21 causal, hardware, personnel or whatever.
So I decided to 22 take and cut it at a lower level as to what was causing the 23 hardware failures and I just picked 15 percent as an
(
24 arbitrary value to start with.
25 CHAIRMAN KERR:
Thank you.
)
1 t\\CE. FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 800 336-6646
'6310 02 02 57 da'ysimons 1 MR. EBERSOLE:
Does that by any chance suggest r
2 that we have grounded systems that ought to be ungrounded 3
in a.large context?
You have to have two grounds to make 4
something fall out for many grounded systems.
To me that j
5 says there is a message, that I've got too many grounds 6
systems.
7 Charlie, what do you think?
8 MR. WYLIE:
Well, it depends.
Are you talking l
9 about power systems.
10 MR. EBERSOLE:
I'm talking about control 11 systems.
12 MR. WYLIE:
Well, control systems, you don't
)
13 normally ground them.
14 MR. EBERSOLE:
You're talking about power 15 grounds.
Do you know what these were?
16 MR. WILLIAMS:
I don't know specifically what 17 they were, but Larry does and we could spend a lot of time 18 on this chart I'm sure.
o 19 Do you have a quick answer?
20 MR. BELL:
A lot of them are just people 21 grounding things while they work.
22 MR. EBERSOLE:
Screwdriver grounding.
23 (Laughter.)
l 24 MR. WILLIAMS:
It's a level of detail and a 25 level at which we can cross-cut the data.
We have had this ACE-FEDERAL REPORTERS, INC.
202-387-3700 Nationwide Coserage 800-336 6646
.\\
6310 02 02 58
- (
ruarysimons 1 kind of level of detail when we've looked at it for various 2
problems for two years.
We are trying to look back towards 3
1984.
1985 and '86 we have at this level.
The LER rule 4
went into effect in '84 and we are trying to get back to 5
that level.
6 (Slide.)
7 MR. MICHELSON:
Do you have anything based on a 8
key word index that you can pull the key words out and get 9
an automatic computer plot then, or did you have to develop 10 this manually?
11 MR. BELL:
I developed it and programmed.it.
12 MR. MICHELSON:
But is it pulled right off the 13 sequence coding data base or are you doing this by hand?
14 MR. WILLIAMS:
No.
15 MR. BELL:
This is coming directly from the LERs 16 as something to look at.
17 MR. MICHELSON:
You mean you're reading them one 18 at a time and loading them in that way?'
19 MR. BELL:
Right.
l 20 MR. WILLIAMS:
This is coming from a program 21 where the LERs are read and cause coded specifically to the 22 root cause, and it's separate and apart from the SCSS, and 23 we also use the SCSS.
f
{
24 MR. JORDAN:
Before you take that one down, j
25 these were all balance of plant initiated trips, and I
/\\CE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 800-336-6646
6310 02 02 59 U
marysimons 1 think that balance of plant set for 1986 was 30 percent of 2
the total numbers of scrams.
Of that list none of those 3
were desired scrams.
They were not because the plant need 4
to come down.
They were because a balance of plant 5
malfunction caused the plant to come down.
So if there 6
were some predominant feature in here, either predominant 7
from the standpoint of the system that's causing it or from 8
the root in terms of vibration or age, if one of those was 9
predominant, then we would have a handle to try to do 10 something to reduce the trends in that area.
So that was 11 the reason for trying to cut it back.
()
12 MR. WILLIAMS:
The BOP in, fact is a very large 13 initiator.
This is hardware, and that is not all the cause 14 categories that are in the BOP.
15 MR. EBERSOLE:
The vibration I suspect is on the 16 main turbine and that raises a question.
When you stop and 17 go to the main turbine, you'll find on the main stop valves 18 this contact multiplication and you have multiple trips to 19 protect scram against single failure.
Then you proceed on 20 out into the turbine complex and you find that you revert I
i 21 to single trips in the indirect context of a vibrometer out 22 there some place, and if it gives a false signal it shuts 23 the turbine down and the whole thing comes crashing down
)
24 around you.
25 Do you find any advantage in considering ACE FEDERAL REPORTERS, INC.
a _ _ __ ___ _ _.__ __ _____ _ _ i _ _ __ _ _ _ _ __ __ __ __ _ ______ _ 202 -34 M 7 00 Nat.ionwide Cose_ rare
_J00-336-6646
j 6310 02 02 60 b
sarysimons 1 redundant or trip devices out beyond say the main stop 2
valves into the turbine complex?
3 MR. WILLIAMS:
I think so,-especially when you 4
get into the BOP area and especially we have recently 5
looked at new plants and we see this kind of a problem 6
where you have a large benefit.
7 MR. EBERSOLE:
A classic example was I remember 8
there was an activity detector that came out of the stack 9
at Browns Ferry and there was only one activity detector 10 that shut all three units down, one control device cascaded 11 to all three units.
12 MR. REED:
I'm somewhat pleased to see that only 13 31 percent of the trips, total trips came from balance of 14 plant.
15 MR. WILLIAMS:
Only the hardware trips.
I think 16 the number for 1986 for all the cause codes for BOP would 17 be somewhere in the neighborhood of 55 to 65.
18 MR. REED:
You mean human error and all the 19 others put in.
20 MR. WILLIAMS:
Right.
21 MR. REED:
I of course am opposed to extension 22 of NRC activities into balance of plant review and 23 prescriptions.
I think that the NRC has its hands full
(}
24 doing a real workmanship job on safety related and nuclear 25 parts.
ACE FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 800-336-6646
63.. 0 02 02 61 (marysimons 1 MR. EBERSOLE:
I would like to mention something 2
that I think you ought to look at as kind of a special 3
case.
I know of one case -- and I'll just one that I can 4
recall -- I know of a coal where a domestic water level 5
switch in a coal burning steam plant cascaded the entire 10 6
unit station and jeopardized the grid.
That was one 7
domestic water flip switch, the reason being that domestic 8
water was used for a critical function such as steam 9
protection, and if you lost the water, you lost the entire 10 service function like component cooling.
11 So you might look for those in these studies
(])
12 like this.
13 (Slide.)
14 MR. WILLIAMS:
On the human side, we didn't want 15 to leave out the human side.
So we brought an example 16 cross guide of the kinds of things we cross-cut personnel 17 errors with.
It's the same type of a display just for 18 simplicity, although the charts are cut a bunch of 19 different ways.
20 Again, on the human side you can see that some 21 people were taken by the inattention category, and that is 22 somewhere less than 20 trips total because this is only 26 23 percent of all the trips in 1986 plotted by personnel n()
24 errors to begin with.
But inattention does account for 25 part, and these cause categories are taken directly from ACE FEDERAL REPORTERS, INC.
- i...
202-347 3700 Nationwide Cmerage 8m-336W46
6310 02 02 62 mdrysimons 1 the lice 1see's root cause identified.
2 CHAIRMAN KERR:
What does inattention mean?
3 MR. WILLIAMS:
In this particular case 4
inattention could be a lack of awareness of the operator 5
configuration of the plant, or it could be a lack of 6
awareness of an operator of an evolution.
7 MR. EBERSOLE:
Is there out there some place 8
something that says ---
9 CHAIRMAN KERR:
Jesse, could he please answer my 10 question.
11 MR. EBERSOLE:
Oh, I thought he had.
Had you 12 not?
13 CHAIRMAN KERR:
I thought he was interrupted in 14 mid-sentence.
Maybe not.
15 MR. EBERSOLE:
Are you done?
16 MR. SIESS:
He defined inattention as ignorance.
17 MR. WILLIAMS:
It would be a lack of awareness, 18 or actually defined inattention to the plant status or to 19 the evolution that's occurring at the time in the plant or 20 something that the operator should have been aware of had 21 attention to detail been in effect.
22 I read a number of LERs myself to try to get a 23 handle on that because it's difficult to grasp.
The
(}
24 licensee actually in these cases pinned the root cause on 25 lack of awareness or not attention to detail of the plant ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage S( 4 336-6646
6310 02 02 63
()
Larysimons 1 condition.
2 MR. EBERSOLE:
That was what I was about to 3
mention when I interrupted you.
That means that the 4
operator when he takes the cover off of an instrument, he 5
doesn't know if his probe slips 3/8ths of an inch and he's 6
going to have a million dollar trip.
There is no 7
instructional process that tells him that before he goes 8
and opens the cover; is that correct?
9 MR. WILLIAMS:
That's right.
Some of these 10 things, the accessibility of them is very difficult to 11 begin with.
So it's complicated.
12 MR. EBERSOLE:
There must be some mechanism that 13 we need here to tell them that, that they have a million 14 dollars in their hands and a one-day shut down of a plant 15 plant.
16 MR. WILLIAMS:
We found the same thing, and we 17 can get into that in more detail later on, too, in the 18 training.
19 MR. EBERSOLE:
That ought to go back to the fact 20 that the cost of those things goes back to the ratepayers 21 instead of the corporate profits.
22 (Slide.)
23 MR. WILLIAMS:
This is an example of the types
(~N 24 of details.
Some of the changes we are making in the d
25 trends and patterns activities are not really changes, but ACE FEDERAL REPORTERS, INC.
202-34L3%O Nationwide Coserage 800-336-(646
6310 02 02 64
'b ritarysimons 1 they are improvements.
2 One of the things we are trying to do is to 3
increase the feedback of our programs to actually improve 4
operational safety, and that takes, as Ed Jordan said, 5
improved coordination with the industry and within our own 6
staff, and we are trying to interact a little bit better 7
with the industry and with our staff and communicate a 8
little bit more effectively.
9 We are also trying to apply operating experience 10 now that we have a better data base since the LER rule got 11 put into effect and there is more complete data.
We are 12 trying to use that data base to look at some of the 13 corrective actions of the industry, and as an example with 14 the scram reduction where there arc various scram reduction 15 measures in place by plant specific efforts and vendor 16 efforts, and we are trying to see if the operating 17 experience indicates that those scram reduction efforts are 18 aimed at the right places.
19 For example, if you had a training problem, is l
20 there a scram reduction effort on maintenance technician 21 training in dark areas for test probes.
It comes down to 22 simple things like that, does the operating experience 23 indicate that the scram reduction measures are on target.
(}
24 So this is one of'the things that we are thinking of 25 getting into.
f ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coserage 800-336-6646
6 0 02 02 65 marysimons 1 Ue are also starting to develop a data base 2
where we can try to look at plant specific things a little 3
bit better.
We are just getting into this area.
4 An example of the work that I will discuss, or 5
Bob Dennig will discuss in detail is applying the trends 6
and patterns effort to study the new plant startup 7
reportable events.
8 MR. LEWIS:
Before you leave that one viewgraph, 9
I'm sort of curious.
The first bullet which says increase 10 the use of operational experience to improve operational 11 safety, surely there has been nobody ever who denied that
()
12 one should use operational experience to do that.
So why 13 is this a new initiative?
Is it because you think it 14 hasn't been done enough?
15 MR. WILLIAMS:
I think that's true.
16 MR. NOVAK:
Well, I think we talked about this 17 in the beginning of this meeting where we had seen many 18 events where people had gone into this particular operation 19 and had difficulties and we had issued information notices, 20 bulletins and INPO issued safety evaluations, but the 21 events continued to repeat.
22 So somehow you've got to get the message across 23 in manner that becomes effective.
Sometimes they are
()
24 successful.
I think when we talked to INPO just a week or 25 two ago, they think they have nailed down the biofouling ACE FEDERAL REPORTERS, INC.
I 202 347-3700 Nationwide Coserage 800-336-6646
6310.02 02 66
)
marysimons l_
phenomena related to certain blockage phenomena in the heat 2
exchangers.
3 MR. LEUIS:
So if I understand what you're 4
saying, you're saying not that you don't think it's been 5
done enough, but you don't think it has been done very 6
well.
Did I misunderstand you?
7 MR. NOVAK:
I think that's a fair 8
characterization.
9 MR. LEWIS:
I see.
So it isn't to increase the 10 use, but it's to make it done well.
I'm putting words in 11 your mouth.
12 MR. NOVAK:
That's the way I would like to put
{~)
13 it.
14 MR. WILLIAMS:
Hell, let me give you an 15 example.
On new plants, which Bob will discuss in detail, 16 we found that in general when plants are constructed and 17 then turned over to operations, the process makes use of 18 the 50.55(e) deficiencies and the things they encounter 19 during construction in their site.
Then they start getting 20 information from engineering and the independent safety 21 engineering group at there at corporate.
22 One of the last things the new plants seem to l
23 do, the good ones do it early and there are a lot of good l
24 ones, but some of the ones that are a little bit behind
]
(}
25 seemed to have reached out a little bit late to try to find ACE. FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 800-336-6646 j
l 6310 02 02 67.
I')
Lsrysimons 1 out if other similar plants have had problems with their 2
feedwater bypass system and the turbine control system and 3
exactly what kind of problems they've had, including valves 4
and the like.
5 So some of the things, it's exactly what you 6
said, it's improving the operational feedback and it seems 7
as though the industry is in a different state of evolution 8
plant specifically and across the vendors.
Everybody is 9
using it, but it's just a different degree that they are 10 using.
11 MR. LEWIS:
I'm just trying to close in in my 12 own mind on the character of your statement.
You report I 13 guess to EDO, don't you?
I 14 MR. NOVAK:
That's correct.
15 MR. JORDAN:
Yes, sir.
I do report to the EDO.
16 The concern I think we have is that the feedback has not 17 been as successful, whether it's the fact that the agency 18 hasn't communicated as well to the industry as it should or 19 whether industry has been listening when we communicate, we 20 really don't have the bottom line, but we are going to try 21 to improve the communication process as our way of trying 22 to solve the problem and monitor each of these issues to 23 see that there is in fact a change in the rate of 24 occurrence.
25 So that has to be our methodology, and if the ACE. FEDERAL llEPORTERS, INC.
20 M 47-3700 Nationwide Coserage 8(G336-6646
6310 02 02 68 rysimons 1 frequency of continued problems with the decay heat removal 2
system failures does not change, then ther7 is still 3
something ' wrong with the process and we have to go back to 4
it again, despite the fact that we've made previous 5
recommendations.
So we are going to be persistent in 6
trying to solve it.
I would like to see a problem simply 7
go away because it's solved, and one that I hope in fact 8
works out that way is the biofouling.
9 There is indication that enough industry 10 attention has been placed on it now that biofouling is a 11 past problem and there is a continued process by industry 12 to make sure that doesn't reoccur in plants.
We'll see and 13 we'll watch the data and try to establish that it in fact 14 has gone away.
We would like that one solved and worry 15' about then other oroblems rather than continuing to repeat 16 the same problems year after year.
17 MR. LEWIS:
I'm just trying to close in in my 18 own mind on the question of whether you feel free to call a 19 spade a spade when you see one.
20 MR. JORDAN:
Absolutely.
21 MR. LEWIS:
I would love to see some evidence, 22 and I don't doubt that such things exist, but calling a 23 spade a spade doesn't mean proposing a large program to re-24 explore the subject, but it means saying to the EDO, you 25 guys are doing a lousy job on this and you've got to shape ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coserage 80)-336-6646
6310 02 02 69 ysimons 1 up, not necessarily in those words, and I would love to see 2
some examples in which that has happened.
I'm sure there 3
must be some, but I haven't seen one.
4 MR. JORDAN:
I will provide the ACRS with the 5
correspondence on the decay heat removal.
That would be a 6
good example.
7 MR. LEWIS:
Just for my education which after is 8
in the public interest.
9 MR. JORDAN:
Yes.
10 CHAIRMAN KERR:
I don't want to cut this 11 conversation short, and I won't, but I simply found out 12 that our allocated time for this topic has expended, but we 13 could go on as long as you want to explore.
14 Did you want to comment, Mr. Ward?
15 MR. WARD:
Well, I just wanted to say, Ed, that 16 that seems to be the problem that you've been analyzing 17 things and pointing out things for five or six years and 18 not much has changed.
Today we haven't heard much about 19 your new marketing program.
We've heard about things 20 you're analyzing and how you are going to analyze things 21 differently, but it seems like that the essential change 22 that is needed in your program is how you market your 23 product and get people to react to it, and we haven't heard 24 anything about that today.
25 MR. JORDAN:
That's right, and I promise that ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coserage 800-336-6646 i
6,
'02-02 70 marysimons 1 for a later session.
2 CHAIRMAN KERR:
Ed, at a later session I want to 3
ask you if you have thought about looking at the risk 4
increase or decrease that might accrue if instead of doing 5
testing during plant operation, you put testing off until i
6 the plant is down, as I understand the Japanese do, but I 7
won't ask that at this session.
8 Are there further comments or further questions?
9 (No response.)
10 Are you finished with your presentation?
11 MR. WILLIAMS:
We have two more speakers.
rf 12 MR. NOVAK:
We have two more items that we would 13 just like to cover.
14 Why don't you go ahead.
15 (Slide.)
16 MR. WILLIAMS:
Okay.
One of the things again is 17 that we looked at the effects of corrective actions in the 18 data and that's a key item.
19 (Slide.)
i 20 I would like to get on with it and let's talk 1
21 about key plants because I think you are interested in j
l 22 this, and we wanted to talk to you a little bit about how 23 we are improving our strategy to get information into the 24 operating arena a little bit more effectively.
25 The Commission's policy and planning guidance ACE-FEDERAL REPORTERS, INC.
m m
l 6
02 02 71 marysimons 1 endorsed the view that the staff should dedicate more 2
attention to the first two years of operation of new plants 3
as they start up.
4 In fact, at one of our senior management 5
meetings the subject of new plants came up and that 6
resulted in an October 1986 letter from the EDO to NUMARK l
7 which essentially requested some assistance to improve new 8
plant startup, to look at the startup of new plants and 9
reduce the frequency of reportable events.
10 (Slide.)
11 AEOD had initiated a study a long time ago that
()
12 was published about a year ago basically with count of 13 events at new plants and it showed that there were a high 14 number of reportable events.
15 Subsequent to that, just around September or 16 October of '86 AEOD initiated a follow-on study which 17 superseded the earlier one that was to look at the root 18 causes of events at new plants.
19 (Slide.)
20 This is NUREG 1275, and the Commission requested 21 a briefing from the staff on their new plant activities 22 back in November of 1986 and we explained the approach we 23 were pursuing with NUREG 1275, which was a lessons learned
(
24 approach where we were going to review the data, see what 25 it could tell us about the causes of initiators of events ACE FEDERAL REPORTERS, INC.
.Nn h rraon ROfL M
6310 02 02 72 mirysimons 1 at new plants and then try to develop lessons learned to 2
avoid these initiating events and feed those lessons back 3
to the industry and the staff.
4 So that is what we are off doing and we are well 5
along in it.
The draft report is out and it has been out 6
for peer review for a couple of months and it has been 7
review by all the utilities whose data we analyzed to 8
validate the results.
9 (Slide.)
10 Again, the purpose of the study was to 11 characterize the potential trends at new plants.
This
{)
includes contrasting the events at new plants to old 12 13 plants, to look at the initiating systems, the causes and 14 the frequencies to try to understand why new plants have 15 more scrams and engineering safety feats.me actuations.
16 CHAIRMAN KERR:
What is the difference between a 17 potential trend and a trend?
18 MR. WILLIAMS:
A potential trend and a trend, 19 well a potential trend were the questions that we asked, 20 and the trend were the things that we would find, or at 21 least that is how I thought of it when I wrote this slide.
22 We tried to identify correlations between the 23 plant attributes and the performance, and Bob will get into
)
l
()
24 this in some detail among the classes of plants and the 25 vendors and the like.
ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserace 800-336-6646
6310 02 02 73 radrysimons 1 Now let me try to identify improvement areas 2
that would avoid these new initiators of events in new 3
plants and identify lessons learned that could be fed back 4
to the new plant licensees to avoid the events.
5 (Slide.)
6 When we initiated our study, which is now 7
essentially complete, a number of activities started.
1 8
Staff and industry programs had been moving along for the 9
last eight months or so.
Our report is essentially 10 completed in draf t and is undergoing review in-house right 11 now.
INPO also initiated a program that they reported to
)
the Commission in May of 1987.
The INPO program was the 12 13 response of NUMARK to improving the performance of new 14 plants an'd they had picked up their activities and adopted 15 a number of different exchange ---
16 CHAIRMAN KERR:
What is the difference between 17 your initiative and INPO's?
Are they similar and hence 18 provide redundancy or there significant difference between 19 the two?
20 MR. WILLIAMS:
Well, they are similar in'a lot 21 of respects and they are different in others.
Both we and 22 INPO -- well, we have already developed a lessons learned 23 document and INPO is developing a lessons learned
(}
24 document.
INPO's programs go to the cross-training of 25 management talent from plants that have started up to ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Cmerage 8tn336-646
~
6/AD 02 02 74 d
marysimons 1 plants that are starting up through their new plant 2
manager's course and other initiatives that they had.
It's 3
a cross-pollina. ion effort to some degree.
i 4
INPO also uses its evaluation team and its i
5 inspection resources for their visits to new plants and 6
they are trying to give some up-front help to the new i
7 plants and then they will look at them post-commercial 8
operation.
9 From our initiatives we are probably going to be 10 working with the new plants during their start-up and early 11 commercial operation which at least in the INPO program is
()
12 a period of absence where the plant is left,along unless 13 they request assistance from INPO which might.
14 So the programs are to a large degree 1
15 supplemental.
They look at different things, although each 16 of them has lessons learned and each of them will have 17 probably good practices and items to avoid initiating 18 events.
19 (Slide.)
20 Also, NRR is modifying its readiness review 21 activities slightly and we are working with them and 22 participating in that process to improve those readiness 23 review activities.
()
24 Then there is a commission briefing scheduled in 25 a few weeks to discuss the staff's new plant initiatives.
ACE: FEDERAL REPORTERS, INC.
___ _ ______:- _ _ ____ m _ _ ___ _ _ __ _ _ m m. rr a___ _ ______ w oanm u w r a m _ -
6310 02 02 75 rysimons 1 Now with that Bob Dennig is in charge of the 2
Trends and Patterns Section and he is going to go ahead and 3
discuss some of the data and some of the things that we 4
have.
5 MR. JORDAN:
I understand that we are running 6
over.
7 CHAIRMAN KERR:
I'm told by the Subcommittee 8
Chairman that he is not going to use as much time for later 9
events as was scheduled.
So continue.
We have 15 more 10 minutes if you want to take that.
11 MR. NOVAK:
Okay.
j 12 MR. DENNIG:
Mark has already given you a great 13 deal of the background in what led up to doing this study.
~
14 I would just like to say that this effort represented an 15 incremental increase in the level of effort in this 16 particular topical area.
17 (Slide.)
18 The purpose was to characterize the early 19 startup period, and the method that was pursued was using 20 the in-house information LER data and milking that for all i
i 21 it was worth and categorizing and codifying it and so on 22 and so forth for a body of plants and drawing preliminary 23 conclusions about problem areas and trouble spots and then
(}
24 sort of a new wrinkle in the trends and patterns area, an 25 extensive move out to plant size to discuss these ACE FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coserage 800-336 6646
6310 02 02 76 ysimons 1 preliminary kinds of findings with the utilities that had 2
gone through startups and get feedback from them on some of 3
the things that might further be underlying the situations 4
that they were faced with.
5 CHAIRMAN KERR:
How many plants are likely to be 6
in startup mode beginning today?
7 MR. DENNIG:
It's not a large number.
There are 8
several at the current time, but further down the line it 9
thins out very quickly.
10 MR. WILLIAMS:
Also, a lot of these new plants 11 are within their first two or three years of operation.
So
)
12 there is a large population of those plants right now.
(~/
w 13 (Slide.)
l 14 MR. DENNIG:
Quickly some major findings, and 15 l hopefully these all comport with common sense.
16 More focus on operations readiness using 17 operating experience before you get your OL, and what this 18 really gets at is again say kind of a common sense notion 19 that if you put yourself in an operational mode post-fuel 20 load that it's a different world.
We've got tech specs, 21 we've got surveillance we have to do and we've got a 22 different level of detail that we have to adhere to.
23 If you get yourself in that mentality before you I')\\
24 get the license, then you are not going to go through any
'x-25 kind of a learning process after your license or you will l
/\\CE. FEDERAL REPORTERS, INC.
l 202 347-3700 Nationwide Coserage 800-336-6M6
)
6 0 02 02 77 marysimons 1 curtail that.
2 New plant event frequencies was fairly well 3
recognized.
4 CHAIRMAN KERR:
Are you suggesting that by 5
reading this report and talking to the NRC that one can 6
expect that a brand new plant will have the same number of 7
incidents as a plant that has been operating five years?
8 MR. DENNIG:
Not at all.
9 MR. JORDAN:
There is a learning curve for each 10 of these plants.
So it's recognized that new plants will 11 have a higher rate of events during early life, but the
(~J')
12 rate can be lower than many of the plants have sustained.
13 So what we really want to make sure is that the experience 14 of the plants in a given reactor type or utility are fed 15 back to the new plants in a way that they don't have to 16 learn from their own errors, but by somebody else's 17 experience.
18 We see evidence that plants that start up really 19 quite smoothly.
Generally each the plants has more events 20 than the mean of the already operating plants in that time 21 frame, but it's a matter of fact that some of the plants 22 have many times more events, enough to make that learning 23 curve smooth, (m) 24 MR. DENNIG:
The next few points.
Early e
25 startup performance tends to correlate with early ACE. FEDERAL REPORTERS, INC.
. Nationwide Coserage MX)-3364M6 Z3#3HXI
6310 02 02 78 marysimons 1 commercial operations, and what this gets at is that in 2
some cases when you look at it the learning curve does 3
extend beyond the end of the formal startup period and 4
things are not down to what you might consider a mature 5
plant level while the plant is moving into 100 percent 6
power operation and early commercial operation.
7 The last point, as Mark has already mentioned, 8
this study has been widely peer reviewed by the owners 9
groups and the plants-that were included in the study, and 10 generally what we have found is agreement with the findings 11 and the conclusions and the lassons that would tend to 12 improve things.
{}
13 CHAIRMAN KERR:
In looking at this did you see 14 any correlation between type of plant or it more nearly i
15 type of organization or is there anything you can 16 correlate?
q l
17 MR. DENNIG:
No, sir.
When we looked at the i
18 data for the 22 plants that we specifically focused on in 19 this study we looked at it in terms of NSSS and whether 20 this was their first nuclear power plant or whether this 21 was the second unit at the site, how long the startup 22 period was and a couple of other things.,
23 In general those things don't play much of a
()
24 role in explaining the variation from place to place and 25 how many difficulties they go through.
ACE FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coserage 800-336 6646 l
7
6310 02 02 79 ysimons 1 MR. WILLIAMS:
We had looked for those 2
correlations statistically informally, and we attempt to 3
correlate the type of plant and a number of other factors, 4
and the correlations came out not to be there or were not 5
statistically significant.
There were a couple that 6
followed cover, but we looked exactly down those lines and 7
there were a number of lines we looked down, but we really 1
8 found out that plant specific management seemed to be the 9
overriding factor, although that in and of itself is a 10 difficult thing to prove and we fund it kind of by looking 11 everywhere else and finding negative correlations and then 12 going out in the field to some degree.
}
13 MR. JORDAN:
I would like to emphasize that the 14 recommendations that this report will contain are those 15 that we obtained through interviews with the utilities 16 themselves who said that for the plants that did well in 17 their startups, these were things that they did, and for 18 those that did poorly, these were the things they wished 19 they had done.
20 So what we are really doing is reflecting back 21 what the utilities did, put it out for peer review, refine 22 it and then provide it to all of the utilities.
So our 23 statistics were very interesting and there was a lot of (v)
24 data to work through, but it didn't tell us at all how to 25 solve the problem.
We couldn't narrow it.
ACE FEDERAL REPORTERS, INC.
202 347 3~(0 Nationwide Coserage 800-336 4 46
6310 02 02 80
.I) imons 1 marys MR. REED:
You must have concluded that a second 2
similar or identical unit at the same site, the second in 3
chronological order certainly had a much better record in 4
LERs and trips than the first.
5 MR. DENNIG:
No, not in general.
6 MR. REED:
You couldn't conclude that?
7 MR. DENNIG:
In one particular area ---
8 MR. WILLIAMS:
We can name cases where that was 9
clearly the case.
10 MR. REED:
I can name lots of cases.
11 MR. DENNIG:
We can name cases in the other
()
12 direction.
13 MR. WILLIAMS:
It's not strong enough 14 statistically to be as overwhelming as you would like it to 15 be.
16 MR. REED:
Then was there different management 17 of the two units, were they separated or what?
There must 18 be something like that that would influence it, because 19 ordinarily the second unit identical or similar has a much 20 better first or second year second year record.
21 MR. DENNIG:
From a statistically analyzed 22 viewpoint there was one area that did stand out where the 23 second unit had the advantage, and that was in the area of
()
24 tech spec violations.
Early in commercial operation the 25 second unit at the site was much better than the first unit ACE-FEDERAL REPORTERS, INC.
l 202 347-3700 Nationwide Coserage 800-336-6M6
6310 02 02 81
(~)
nucysimons 1 at the site or much better than the overall population.
2 MR. WILLIAMS:
Some of the reasons for the 3
variability was the compression of the schedule for the 4
second unit.
The schedule for the second unit, for
{
i 5
example, Diablo 1 and Diablo 2, Diablo l's schedule was 6
obviously very drawn out with a lot of hot functional 7
testing and a lot of in-plant work, and the second unit's 8
schedule was somewhat compresses compared to the first 9
unit.
10 And in the addition to that, there~were BOP 11 differences that might be feed pump controllers, and there 12 might be little differences that seemed to play a part, 13 although it's very difficult to pin that down.
14 CHAIRMAN KERR:
Mr. Shewmon.
15 MR. SHEWMON:
I had heard a few years ago that 16 somebody had done a study like this and their conclusion 17 was that people who had other plants has less trouble than 18 utilities that this was their first plant.
You didn't 19 mention that, and that is not borne out by your study or 20 did I miss it?
21 MR. DENNIG:
We did look at that, and I thought 22 I did mention it, that whether this was the first nuclear 23 unit for a utility or not was one of the things that we 24 looked for correlation with, and it was not particular 25 atrong.
ACE FEDERAL REPORTERS, INC-202 347-3700 Nationwide Coserage 800-336-6646
63L0 02 02 82 b
marysimons 1 MR. SHEWMON:
Thank you.
2 MR. EBERSOLE:
Is each line up there a given 3
plant?
4 MR. DENNIG:
Each line is a track.
5 MR. EBERSOLE:
What I would like to see is a 6
correlation between the two bottom lines and the two top 7
lines in the context of characterization of upper 8
management.
What can you see about that?
9 (Laughter.)
10 MR. DENNIG:
I don't know how we would have 11 quantified upper management.
()
12 MR. EBERSOLE:
This is a human factors problem.
13 CHAIRMAN KERR:
Jesse, if a plant started up 14 smoothly, they concluded that it had good upper management.
15 MR. EBERSOLE:
Well, yes, but there are other 16 characterizations of upper management besides just whether 17 it runs or not, and how did they get that way.
18 Glenn says you don't need it, and I say you do.
19 (Laughter.)
20 MR. NOVAK:
Well, it all depends on whether 21 Colonel North is allowed to ---
22 (Laughter.)
f 23 MR. EBERSOLE:
It would be interesting.
(
24 MR. JORDAN:
I have to say again that I think 25 this is a case where the data simply defies your solution
. ACE FEDERAL REPORTERS, INC.
1 202-347-3700 Nationwide Coverage 800-336-6646
6310 02 02 83 niurysimons 1 in terms of providing the useful benefit, and by going back 2
to the utilities in a subjective fashion and using their 3
experience to then convey to the other plants, the 4
correlations just did not come out of this data.
5 MR. EBERSOLE:
Which were those two better 6
plants, do you remember?
7 M...
DENNIG:
If.I recall correctly, one of them 8
is McGuire 2 and one of them is St. Lucie 2.
I believe l
9 that's the case.
I think that I have stared at this long 10 enough that I know who those guys are.
11 MR. REED:
I heard you say two and two.
12 (Laughter.)
O,.
13 MR. EBERSOLE:
That's another factor, isn't it.
14 (Slide.)
15 MR. DENNIG:
This is another figure like the one 16 we've just seen, only this one has to do with unplanned 17 reactor scrams for a subset of the plants that we looked 18 at.
19 The interesting thing I guess in this display is 20 that there is somewhat of a need abcut this point, about 21 2,500 critical hours into operation, which is roughly about 22 six months that things start to turn over in general and in 23 most cases.
Before that time the scram rate is very high
(}
24 compared to mature plants, and then about that time things 25 start to roll over.
ACE FEDERAL REPORTERS, INC.
202-347 3700 Nationwide CoseraFe 800-336-6646
i 63)so 02 02 84' k.
marysimons'1 MR. MOELLER:
Now all of these are cumulative, 2
both this and the previous graph?
3 MR. DENNIG:
That's correct.
4 MR. MOELLER:
It's not a range or not within 5
this slope.
6 MR. JORDAN:
Why don't you identify the slope 7
corresponding to the industry mean at this time.
8 MR. DENNIG:
It's not on the figure.
We would 9
just have to take a line.
From 5 up to 5,000 roughly would 10 give you the mature plant rough accumulation of scrams over 11 5,000 critical hours and it would put them about there.
(]}
12 (Slide.)
13 One last picture.
This has to do with looking 14 at the issue of whether or not the scram experience is 15 somehow skewed or localized in one particular system that 16 is peculiar to new plants as opposed to old plants, and the 17 answer is no, that is not the case.
18 We see as a relative distribution that's the 19 same, during startup experience, early post-commercial and 20 then a mature plant is the dark bar.
What we see is i
21 roughly the same distribution across systems as far as 22 scram initiation, but with vastly higher rates.
23 So there isn't some peculiar skewed area with
()
24 one particular system.
That's a peculiar problem for a new 25 plant that's not also a problem for an old plant.
ACE. FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 800-3364M6
6310 02 02 85 I) mdrysimons 1 MR. EBERSOLE:
Did you see any difference 2
between motor and turbine main feedwater systems that was 3
interesting?
I should think the motor driven would be a 4
lot better.
5 MR. DENNIG:
We didn't control for that in what 6
we did.
So I really can't-get at your question.
I can't 7
give you a good answer.
I'm not even sure offhand to what 8
extent these guys have motor driven main feed pumps.
9 MR. EBERSOLE:
Well, that's a design decision 10 and most of the time it goes ---
11 MR. DENNIG:
Unless these guys are turbine 12 driven main feed pumps.
13 MR. WILLIAMS:
The feedwater configuration is 14 something that we're looking at in plants like Beaver 15 Valley 2.
Beaver Valley 1 didn't have a startup feedwater 16 pump, and Beaver Valley 2 has trimmed some colors also has 17 a startup pump so the BP across the valves is lower.
So 18 these are features in some of the later plants that they i
19 made when there was enough time ---
20 MR. EBERSOLE:
I'm talk about main feedwater 21 pumps where the usual pitch is that it's cheaper to run 22 turbine pumps, but they have a lot more trouble.
23 MR. WILLIAMS:
They do.
{}
24 (Slide.)
25 MR. DENNIG:
Major lessons that we wanted to ACE FEDERAL REPORTERS, INC, 202 347-3700 Nationwide Coserage 800-336-6646
1 j
I 6310 02 02 86 O
ractysimons 1 translate to industry were to begin the operations 2
mentality or mode as soon as you can, to train for routine 3
operations, train I&C well for surveillance activities and 4
the last one is understand your equipment characteristics 5
well, and that gets at focusing a little bit more on the 6
BOP side on some of the peculiarities and particularities 7
of your plant design.
8 Even if I can't say that my design is exactly 9
like your design, the chances are pretty good that I've got 10 some feature in'the same general area like the Frey Greg 11 valve that is a peculiar feature to me but I should pay 12 attention to that.
13 (Slide.)
14 Then, lastly, for the NRC is something we 15 derived from our analysis was that the root cause process 16 obviously has a great impact on damping things down and 17 preventing repetition, and the folks that do that well get 18 that damped down and folks that don't do that need some 19 encouragement.
20 Lastly, in looking at the plant's status for 21 licensing that, in addition to looking at complete 22 construction and close-out of licensing issues, one needs 23 to look at the actual readiness of that plant for routine
(}
24 operations.
25 MR. EBERSOLE:
That process came out first in ACE FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 8043364/>46
6310 02 02 87 O
tfu.lysimons 1 the Georgia plant, whatever the name of it is --- Vogel?
2 MR. REED:
Hatch?
3 MR. EBERSOLE:
No, no, the other one.
- Anyway, 4
there was a lot of controversy about that being being 5
costly, the readiness review.
6 MR. NOVAK:
Oh, that was for construction, 7
Jesse.
8 MR. EBERSOLE:
I was just going to ask because 9
that turned out to be a very good thing to do, a late -
10 coming readiness review, am I really ready to go.
11 MR. NOVAK:
There are two kinds of readiness.
12 MR. EBERSOLE:
Well, this is the kind of O-l 13 readiness I'm talking about.
14 MR. NOVAK:
The one that was done at Vogel was 15 very substantial and was part of the construction process, 16 that is where they following construction deficiency 17 reports.
So there was an overview of that construction as 18 part of the readiness of the plant being completed.
19 Now when we talk about a readiness review here, 20 we are talking about readiness for operation, are the 21 people there trained, do you have your HP trained and are 22 you doing the surveillance.
23 MR. EBERSOLE:
Uasn't that in Vogel, too?
24 MR. NOVAK:
No.
They had to cut that short.
25 MR. EBERSOLE:
Did they.
ACE FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coserage 804336-646
6310 02 02 88
/~%
()
marysimons 1 MR. NOVAK:
Yes, and that was not a module that 2'
was done in detail.
3 MR. EBERSOLE:
Okay.
4 CHAIRMAN KERR:
We've talked about 15 more 5
minutes, and I know it's our fault that this is not 6
proceeding any more rapidly, but I would remind you.
7 MR.'NOVAK:
Kathleen Black has about five 8
minutes.
9 CHAIRMAN KERR:
I certainly did not want to 10 eliminate her presentation.
11 MR. NOVAK:
I would let us end the meeting
("T 12 without Kathleen's presentation.
U 13 Kathleen is the head of our Non-Reactor Safety 14 Group and the individual safety programs, but wa have kind 15 of had her and some of the people working for ner spend 16 some time on the human factors.
17 MR. MOELLER:
This is radiographer and medical 18 administration.
19 MR. NOVAK:
She does that, too, but in her spare 20 time she works on human factor problems.
21 MS. BLACK:
I just wanted to say that that was 22 one area where we were very effective with feedback.
23 MR. MOELLER:
Well, we appreciate your reports.
()
24 I read them very thoroughly.
l 25 (Slide.)
ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coserage 14)0-336 6646
6310 02 02 89 U
marysimons 1 MS. BLACK:
I would just like to briefly, very 2
briefly to discuss two studies.
3 One is a completed study which is now rather 4
old, which is events involving wrong unit or wrong train or 5
wrong component and to give you some information about a 6
study that is currently in progress, events involving 7
procedures for a two-year period, 1984 to 1985.
8 (Slide.)
9 I do want to say something before I go into the i
10 story about wrong unit, wrong train or wrong component 11 events.
()
12 We derived all these data from counting and 13 reading and appraising LERs.
If an event occurred 14 involving the wrong unit or wrong train or wrong component 15 that was not reportable to us, we don't have it in our 16 statistics.
17 A special study was published in January 1984 18 and that followed or was requested by NRR following an 19 event at Turkey Point in which the AFW was out for a rather 20 long period of time.
There were missed independent 21 verifications.
i 22 As a result of the study, an information notice j
23 was published and this is one area in which AEOD and NRR
()
24 had a very close cooperation in following up with 25 suggestions.
These were not recommendations because this ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 800-3364M6 j
6310 02 02 90 k.
marysimons 1 was_not a pier review study.
There were a series of site 2
visits by people from both NRR and AEOD to investigate the 3
contributing factors, and as a result of these plant 4
visits, a document, NUREG 1192 was published, "An 5
Investigation of Contributors to Wrong Unit or Wrong Train 6
Events."
7 Now following these plant visits and NRR reports 8
on the subject, General Issue 102, " Human Error Events 9
Involving Wrong Unit or Wrong Train" was created.
10 Now as a part of the solution, INPO was to 11 include the reviews of these types of events and their
(}
12 evaluation and the generic issue is still open.
13 We have continued after our first study to 14 provide updated what amount to data dumps to NRR for their 15 use.
16 (Slide.)
17 The next slide presents ---
18 MR. REED:
Kathleen ---
19 MS. BLACK:
Yes, sir.
20 MR. REED:
In your pursuit of these human 21 aspects related to wrong unit and wrong train, did you ever 22 give consideration to whether or not the people involved 23 had been aptitude tested for the work they were doing?
()
24 MS. BLACK:
You know, that's a question I had 25 myself from a very personal viewpoint.
We did not approach ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336-6646
6310 02 02 91 cysimons 1 that in this study.
The human factors group may have in 2
other ways and may have published other documents, but they 3
have never been correlated, to my knowledge, against this 4
particular data set.
5 MR. REED:
Well, I know of one case of a very 6
serious wrong unit and wrong train where the aptitude of 7
the person involved was in question.
8 MS. BLACK:
I could give you a very personal 9
observation stemming from the non-reactor side, but in view 10 of the time I won't, but that is a very essential question 11 that I don't know whether industry has really well "3
12 addressed.
- U 13 MR. REED
I'm happy to hear you say that.
14 MR. MICHELSON:
Could I ask you just a brief 15 question on the resolution of the event at Zion involving a 16 wrong unit and wrong train and the resulting flooding of 17 the diesel compartments.
18 MS. BLACK:
Certainly.
19 MR. MICHELSON:
Are you doing any technical j
20 review of that you might publish on that?
21 MS. BLACK:
That has not been initiated as yet, i
22 no.
i 1
1 23 MR. NOVAK:
That was a very recent event.
(}
24 MR. MICHCLSON:
Yes.
25 MS. BLACK:
Is this the one at Calvert Cliffs?
ACE FEDERAL REPORTERS, INC.
202 347 3NO Nationwide Coserage 800-336W46 4
1
6310 02 02 92 A)
( ~
marysimons 1 MR. MICHELSON:
No, no.
2 MR. NOVAK:
It was a shifting where you A come 3
in and B come in ---
4 MR. MICHELSON:
It's a classic.
I would like to 5
some day read the details, and I was wondered if you.were.
6 MS. BLACK:
That has not yet been looked at.
7 The next sheet merely gives you the data from 8
the time that the reporting requirements were changed, and 9
they have been normalized, two events per reactor year, 10 wrong unit events, wrong train events and wrong component 11 events, and we have looked at the data rather than doing
/3 12 any substantial correlation because these are aggregated V
13 and we just believe that the rate of occurrence doesn't 14 show any major change with time.
It bounces up and it 15 bounces down and bounces around.
16 Again, this is no new information or is not even 17 unexpected, that experienced plants have generally lower 18 rates than new plants.
19 MR. MICHELSON:
Are you making any kind of an 20 effort to determine what contribution to risk really comes 21 from this specific problem of wrong unit and wrong train?
22 MR. NOVAK:
We looked and it's a small 23 contribution.
()
24 MR. MICHELSON:
This one in particular is small, 25 but as a whole it's significant?
Ace FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800-336-6646
6310 02 02 93
[)
rumrysimons 1 MR. NOVAK:
I would say it's less than five 2
percent.
3 MR. MICHELSON:
But as a whole you think it's a 4
significant problem?
5 MR. NOVAK:
Well, on the surface it seems like 6
something that can be avoided.
You would think that with 7
the proper training and the proper management control that 8
those kinds of events could be eliminated.
9 MR. MICHELSON:
So with a little imagination I 10 can get a plant in vast trouble going with the wrong unit 11 and wrong train scenario.
12 MR. JORDAN:
Yes.
I think the thing that 13 worries us about some of these wrong unit and wrong train 14 things are that they really end up involving the safety 15 systems and not just a train in a sequence.
16 MR. MICHELSON:
That's right.
17 MR. JORDAN:
And we have had a number of those.
18 MR. MOELLER:
Excuse me, the two years 19 experience, is that of the people involved?
20 MS. BLACK:
That's the plant.
This is the age 21 of the plant.
22 MR. MOELLER:
And why two years, and why not 18 23 months?
24 MS. BLACK:
Well, because the Commission or 25 someone wrote a number down saying why don't you look at ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 800-336-6646
6310 02 02 94
.h marysimons 1 plants in their first two years of operation.
I'm sure 2
it's a totally arbitrary number, but perhaps based upon 3
observations about knees of curves or shakedowns or 4
something like that.
5 MR. NOVAK:
Kathleen made a point earlier that 6
we can only look at the point that is reported, and somehow 7
you feel and I think we all feel that this may be just a 8
fraction of the time that these things happen and they are 9
not reported.
So that's the difficulty that we have, too.
10 My personal guess is that this is less chan -- I don't 11 think this represents 10 percent of the time this happens,
(])
12 but if it doesn't result in a basis for reporting, we don't 13 know about it.
So we are sort of stymied from that point 14 of view.
I would caution my response to Carl, that had all 15 events been report, it might be a much more significant 16 review.
17 MR. REED:
Well, I'm disturbed by your 18 conversation.
I don't know whether you want a trip to the 19 wrong water closet to be reported or not.
20 MR. NOVAK:
A guy could go and do something to 21 an RHR pump or an ECCS pump, and it didn't cause the system 22 to become inoperable ---
23 MR. REED:
However, there would be a procedural
()
24 reportable LER.
25 MR. NOVAK:
No.
ACE FEDERAL REPORTERS, INC.
202 317-3700 Nationwide Coserage 800-336-6646
6310 02 02 95 amarysimons 1 MR. REED:
It's in the procedure and would be 2
reportable.
3 MR. NOVAK:
That's the problem.
4 MR. WILLIAMS:
There is not a deportability 5
requirement for a train being taken out of service 6
accidentally by human error, but there is for a system.
7 MR. EBERSOLE:
It has to be a total functional 8
failure, Glenn, and that a terrible part of the new system.
9 MR. NOVAK:
It may be other things that occurred 10 in the even that let us know about this.
In other words, 11 that particular action by itself with nothing else 12 happening may not have been reportable, but other things 13 associated with that particular event requirei it to be 14 reported and then we picked up this ---
15 MR. MICHELSON:
Well, there is the problem of 16 the near misses that they caught just in time ---
t 17 MR. EBERSOLE:
On a broader basis we still now
{
18 are vulnerable to the fact that we don't know of single q
l 19 train failures, do we, that is the NRC doesn't, and we are 20 still hung with that void in our information program, 21 whereas INPO is supposed to pick these up in their 22 reliability program.
23 Now how is that working out?
24 MR. JORDAN:
I think that would be a basis for 25 another discussion.
ACE-FEDERAL. REPORTERS, INC.
202-347-3700 Nationwide Coterage 8t&3 364M6
6310 02 02 96 U
marysimons 1 (Laughter.)
2 MS. BLACK:
A third meeting.
3 (Slide.)
4 Again, briefly, this is a study that is in 5
progress.
We looked at a subset of data base from 1984 to 6
1985 and identified about 100 events that involved 7
procedures, and we looked at them in somewhat more detail 8
and the study is still in progress, but I'll share some 9
preliminary observations.
10 There is a wide variation in the frequencies 11 from plant to plant.
About 12 percent of the plants
({])
12 reported 44 percent of the events.
13 About as many human e' vents involving procedures 14 occurred when the reactor was shut down as when the reactor 15 was critical.
16 The types of procedures involved in the events, 17 and these are not only numbers, but they are percentages 18 because the base is essentially 100.
There were 28 in 19 operations, 22 in maintenance, 21 in testing, 17 in 20 administrative and management, and that's things like 21 control of shift changes and such things as that, 22 modification, installation and other were again maybe 13 23 percent.
()
24 This is a common finding, that there were only 25 three events that were ascribed to personnel failing to ACE FEDERAL REPORTERS, INC.
I w.wxw um anum
6310 02 02 97 tr.arysimons 1 follow the procedure, and procedural content or 2
presentation was a dominant problem that was mentioned.
3 MR. MICHELSON:
Have you correlated this in any 4
way to the SALP ratings for the particular plants and so 5
forth?
6 MS. BLACK:
We did not look at the distribution i
7 of events per plant against the SALP rating, no.
That can 8
be added as something to look at.
9 MR. MICHELSON:
It's one of the few indicators 10 that we have of plants that are maybe a little ---
11 MS. BLACK:
I can tell you that one of the 12 plants that had the most events was also under edict by NRC 13 to do a procedure improvement program.
So people had
'14 gotten there before AEOD had.
15 Thank you.
16 MR. WARD:
The twice as many events, let's see, 17 when the reactor was shut down as when the reactor was 18 critical?
19 MS. BLACK:
No, about the same I think.
There 20 were about as many events.
21 MR. WARD:
Oh, about as many events.
22 MS. BLACK:
Thank you.
23 CHAIRMAN KERR:
Is there a trend within the 24 regulatory office or whatever it is to encourage more and 25 more written procedures to cover almost everything?
My ACE FEDERAL REPORTERS, INC.
202 347 37(K)
Nationwide Coserage 8(XL336-6646
i 631,0 02 02 98 U
marysimons 1 perception from a limited contact with the agency is that 2
is occurring.
3 MS. BLACK:
Sandy may be willing or able to give 4
you a brief feedback on an event in which he made a similar l
.5 observation which he was visiting recently.
I don't know 6
whether it's appropriate or not.
7 MR. JORDAN:
I think maybe.we just want to 8
indicate that it's a trend that utilities' procedures in 9
bulk increases with time and the regulatory impact does 10 cause that increase rather than a decrease.
So I think in 11 general you assertion is correct.
(])
12 CHAIRMAN KERR:
What happens is that an 13 inspector sees something and he isn't quite sQre he likes 14 it or not.
So he says you really ought to have a written 15 procedure for this, and the reason is because if you have a 16 written procedure and you don't follow that procedure to 17 the letter, whether is important or not, you could be cited 18 for it.
If you don't have a procedure, that is difficult 19 to cite.
20 I recognize that there is a need for written 21 procedures, but there isn't a need for a written procedure 22 for everything one does.
i 23 MR. JORDAN:
I would say that when plants are
()
24 licensed they get a very detailed procedure review and the 25 procedures are then ascertained as being adequate at the ACE-FEDERAL REPORTERS, lNC.
I 202-347-3700 Nationwide Cmerage S00-336-6646
1
.I 6310 02 02 99 (ruarysimons 1 time of licensing.
2 Subsequently as the plant goes through its-3 process and the utility sees other procedures they need and
]
4 they increase in response to an event or a problem, the NRC 5
identifies that a utility needs to change a procedure and 6
prove a procedure.
7 So there is a continuing growth in the volume 8
and quantities of procedures and I certainly hope in the 1
9 quality of procedures, but they do increase in bulk, and I 10 think Glenn Reed would agree with that.
11 CHAIRMAN KERR:
I would encourage people to look 12 at this and to look at the inspectors who go into the field 13 and interact with plants because I recognize that the 14 problems that the inspectors have, and they have a lot of 15 restrictions on them, but they are also in a position where 16 they can put pressure on a licensee to come up with 17 something that the licensee may feel is not the best thing 18 to do.
19 This is not meant to be a criticism of the 20 agency or individuals, but I've just seen this sort of 21 thing happen in my own experience, and I can see it would 22 happen if there isn't resistance to it and in a way which I f
23 think in some cases can have a negative influence on what
(}
24 is going on.
25 MR. JORDAN:
Perhaps Kathleen in her reviews 3
ACEJFEDERAL REPORTERS, INC.
1 202-347 3700 Nationwide Cmeraye 800-336-6646
63,1.0 02 02 100 marysimons I would look for complexity of procedures or the size of 2
procedures as being one of the potential causing factors.
3 MS. BLACK:
Yes, certainly that's the case.
We 4
also questioned whether or not we found procedures to be a 5
problem because people said they revised the procedure, and 6
whether the utility might have taken some much more 7
expensive longer activity but did not want to tell the 8
Commission that they were going to do so.
So it may be an 9
excuse rather than a real thing.
10 MR. EBERSOLE:
A particular kind of a procedure 11 is a check list.
What can you say about the use of check
(}
12 lists?
I endorse check lists because it rejects the thesis 13 that you memory is always good.
What do you find about the 14 use of check lists, are they overused or underused?
15 MS. BLACK:
I don't know that we have gotten 16 into that yet.
I've spoken to the people at INPO and they 17 were interested in that aspect of it, but we have not 18 gotten to that level of detail.
19 MR. EBERSOLE:
Well, for one thing you can leave 20 a record, if nothing else.
21 MS. BLACK:
That's right.
They did want -- I 22 mean they pointed out the idea that wnen you reach a 23 certain step in a procedure you sign off or initial so that
()
24 the check list is integral with the procedure.
25 MR. NOVAK:
That completes our formal
/LCE-FEDERAL REPORTER $, INC.
202 347 3NK)
Nationwide Coserage 800-336-6646
16310 02 02 101 L1ysimons 1 presentation, Dr. Kerr.
2 Again, we appreciate the opportunity to come 3
down.
As Ed said, we would like to come down.in two 4
months.
We'll look at this transcript and try to capture 5
some of the topics that have been suggested and get them 6
into the agenda.
7 MR. EBERSOLE:
I think these interchanges are 8
invaluable, and there is a cross-flow between both of us 9
here which is more formal.
From my point I think that's 10 great.
We get a lot from you, and whether or not you pay 11 any attention to what we say is your prerogative.
At least 12 it's not in a letter or anything.
13 CHAIRMAN KERR:
We thank you for your 14 presentation.
15 We'll take a 15-minute break.
We will convene 16 again at a quarter to four.
17 (Whereupon, at 3:45 p.m.,
the reported portion 18 '
of the proceedings concluded.)
19 20 21 22 23 C:)
25 ACE FEDERAL REPORTERS, INC, i
202-347-3700 Nanonwide Coserage mh3.46646
CERTIFICATE OF OFFICIAL REPORTER O
~
This is to certify that the attached proceedings before 1
the UNITED STATES NUCLEAR REGULATORY COMMISSION in the matter of:
NAME OF PROCEEDING:
ADVISORY COMMITTEE ON REACTOR SAFEGUARDS 327TH GENERAL MEETING DOCKET NO.:
PLACE:
WASHINGTON, D.
C.
O l
DATE:
FRIDAY, JULY 10, 1987 were held as herein appears, and that this is the original I
transcript thereof for the file of the United States Nuclear Regulatory Commission.
l l
(sigt)
/Jt/
J (TYPED)
MARY C.
SlfONS I
Official Reporter f
ACE-FEDERAL REPORTERS, IUC.
I Reporter's Affiliation O
nU AE00 STAFF PRESENTATION TO THE ACRS REVIEW 0F OPERATING EVENTS BY AE00 DIVISION OF SAFETY PROGRAMS O
1 Office for Analysis and Evaluation of Operational Data 1
July 10, 1987 O
m O
"eIl ea 5z A.
OZl i
a o
g 8
lsg O
- 8
(']
PRODUCTS SCREEN 3000 LER/YR PERFORM 30 STUDIES /YR TECHNICAL REVIEWS ENGINEERING EVALUATIONS CASE STUDIES O
i e
O
l@!P.ECTION EMPHASIS ON ISSUES OF HUMAN / PROCEDURE / TRAINING /VNOWLEDGE/
MAN-MACHINE INTERFACE EMPHASIS ON QUANTITATIVE ANALYSIS DATA USE OF PRA ENGINEERING AN/. LYSIS EMPHASIS OR SYSTEMS DYNAMICS INTERACTION
. O EMPHASIS ON MORE TIMELY FEEDBACK O
('*<
OVERVIEE SAMPLE OF WORK PRODUCTS AFW HARDWARE ORIENTED COMMON MODE FAILURE PRESSURE CONTROL SYSTEMS / HUMAN PERFORMANCE BASED PRIMARILY ON A SINGLE EVENT SW SYSTEM INTERACTION BASED ON MULTIPLE EVENTS O
n\\)
s l
O I
l AUXILIARY FEEDWATER PUMP TRIPS CAUSED BY LOW SUCTION PRESSURE O
O
lO 0
THE STUDY WAS INITIATED BY AN EVENT THAT OCCURRED AT MILLSTONE 3 ON JANUARY 29, 1987, 0
LWITH UNIT AT'100% POWER, MOTOR-DRIVEN AFW PUMPS AaB TRIPPED IMMEDIATELY AFTER BEING STARTED DURING OUARTERLY SURVEILLANCE TESTING CONDUCTED ON JANUARY 27 AND 29, 1987.
PUMP A TRIPPED THREE TIMES AND PUMP B TRIPPED ONCE, O
TRIPS DETERMINED TO BE CAUSED BY PRESSURE OSCILLATION IN THE SUCTION LINES THAT:RESULTED IN SPURIOUS LOW SUCTION PRESSURE TRIP SIGNALS.
.0 THREE ADDITIONAL EVENTS WERE FOUND IN THIS REVIEW TROJAN:
ON JULY 20, 1985, LER 85-09 O
D.C. COOK 1:
ON OCTOBER 27, 1985, LER 85-058 ZION 2:
ON DECEMBER 11, 1981, LER 81-033 0
A NUMBER OF CAUSES OF PRESSURE FLUCTUATIONS HAVE BEEN POSTULATED b
O 0
SAFETY SIGNIFICANCE OF THE EVENTS SPURIOUS LOW SUCTION PRESSURE TRIP CONSTITUTES A COMMON MODE FAILURE THAT CAN POTENTIALLY RENDER THE AFW SYSTEM INOPERABLE, 0
CORRECTIVE ACTIONS TAKEN BY LICENSEES TWO DIFFERENT SOLUTIONS WERE TAKEN BY THE LICENSEES, 1.
PROVIDE TIME DELAY S0 THAT MOMENTARY LOW PRESSURE DOES NOT ACTUATE TRIP.
TROJAN AND ZION 2.
O 2.
REMOVE THE LOW SUCTION PRESSURE TRIP MILLSTONE 3 AND D.C. COOK 1 i
D.C. COOK 1 REPLACED TRIP WITH AN ALARM /0PERATOR ACTION COMBINAT!0N 1
0
I 1
1 0
CONCLUSIONS 1.
PRESSURE OSCILLATIONS OR FLUCTUATIONS CAN OCCUR IN THE SUCTION LINES AND CREATE A MOMENTARY PRESSURE DROP DESPITE SUFFICIENT AVAILABLE SUCTION HEAD.
M0MENTARY LOW PRESSURE MAY TRIP THE PUMPS.
2.
LOW SUCTION PRESSURE TRIP IS PROVIDED TO PROTECT PUMPS FROM CAVITATION DAMAGE.
WITHOUT LOW PRESSURE TRIP OR USING ALARM /0PERATOR ACTION COMBINATION, PUMP PROTECTION MAY NOT BE ADEQUATE.
3.
AE0D HAS RECOMMENDED THAT NRR ISSUE AN INFORMATION NOTICE O
I 4
O
O LOSS OF PRESSilRE CONTROL i
i O
O
SALEM EVENT (AUGUST 20, 1986)
INADVERTENT SAFETY INJECTION ACTUATION LOSS OF STATION TRANSFORMERS i
COMPONENT COOLING WATER NOT LOADED ON EMERGENCY BUS OPERATORS TRIP REACTOR COOLANT PUMPS l
CHARGING PUMPS PROVIDE HIGH PRESSURE INJECTION AND RCP SEAL COOLING NORMAL PRESSURIZER SPRAY LOST CONE TRAIN OUT OF SERVICE FOR O
3 MONTHS)
AUXILI ARY SPRAY NOT AVAILABLE IN CONTROL ROOM (VALVE ON OUT OF SERVICE DIESEL)
ONE OF THE PORVs OUT OF SERVICE CANNOT ISOLATE HIGH PRESSURE INJECTION (VALVE ON OUT OF SERVICE DIESEL)
REMAINING PORY CHALLENGED SEVERAL TIMES BECAUSE OF CONTINUED HPSI NUMEROUS SECONDARY PIECES OF EQUIPMENT OUT OF SERVICE OR FAIL DURING THE EVENT O
O-
,A POR,y,.g REFUEUNG p
WATER STORAGE TANK ALTERNATE SPRAY LINE (1)
NORMAL SPRAY _LINE 2)
.h.t..m..
'HIGH RESSURE'
~ '
PRESSURIZER E#
pggp muss mW48 4 I
4 4
O V
RCP SEAL WJECT)ON uar a
/
1 u
In i n
O_ A
\\
~
j n
y ECCS INJECTION UNEM)
{
gg gg PUMP t
Rgure 1 SIMPUFIED DIAGRAM OF RbCTOR COOULNT SYSTEM
]
1
)
I
.O
'.:.};. -_L-
\\
'.; : q,. ;. - 4...,..
I SAFETY SIGNIFICANCE 1
SGTR ACCIDENT SEVERELY CHALLENGES DEFENSE IN DEPTH CONCEPT FAILS PRIMARY PRESSURE B0UNDARY
{
POTENTIAL FAILURE OF CONTAINMENT l
PRESSURE BOUNDARY DEPRESSURIZATION NEEDED FOR NATURAL CIRCULATION C00LDOWN AND FEED / BLEED SITUATIONS O
d O
1 F0ll0WUP.
INFORMATION NOTICE TECHNICAL SPECIFICATION IMPROVEMENT PROGRAM STANDARD REVIEW PLAN ACCIDENT MANAGEMENT PROGRAM (RES)
ALTERNATE PRESSURIZER SPRAYS UPPER' HEAD TEMPERATURES O
,p l
l j
SERVICE WATER CASE STUDY PROCESS l
i
.O REVIEW OPERATING EXPERIENCE - SEARCH DATA BASES
]
REVIEW PRIOR NRC/ INDUSTRY ACTIONS ANALYZE OPERATIONAL DATA CONDUCT SITE. VISITS EVALUATE CORRECTIVE ACTIONS
- LICENSEE
- NRC
- INDUSTRY ACTIONS O
ASSESS SAFETY SIGNIFICANCE AE00 MANAGEMENT REVIEW DEVELOP RECOMMENDATIONS PEER REV!EW/ FINAL REPORT l
l O
i
NUMEROUS DEGRADATION / FAILURE MECHANISMS O'
BIOFOULING OF PUMPS, VALVES, OR HEAT EXCHANGERS SEDIMENT DEPOSITION CORROSION / EROSION OF COMPONENTS DEPENDENCY ON AIR SYSTEMS INADEQUATE SEISMIC RESTRAINT i
FLOODING OF PUMP HOUSES l
DESIGN / INSTALLATION ERRORS INADEQUATE TEST, SURVEILLANCE OR MAINTENANCE PRACTICES
'O PERSONNELERRORS/EQUIPMENTFAILUP.sS O
...I
l IMPACT ON MULTIPLE SYSTEMS / SAFETY SIGNIFICANCE 3
(
CCW HEAT EXCHANGERS
.RHR HEAT EXCHANGERS HPSI AND SI PUMP LUBE OIL COOLERS REACTOR COOLANT PUMP SEAL CONTROL ROOM COOLING FIRE SUPPRESSION SYSTEM CONTAINMENT SPRAY HEAT EXCHANGERS O
CONTAINMENT FAN COOLERS 1
AFW PUMP LUBE Oll COOLERS l
1 l
O
1 PREVIOUS NRC/ INDUSTRY ACTIONS
)
. GENERIC ISSUES 36, 51, 52 AND ]30 IEB 79-14 (SEISMIC ANALYSIS)
IEB 80-24 (PREVENTION OF DAMAGE DUE TO WATER)
I
- AE0D/C202 (FLOW BLOCKAGE BY MOLLUSKS AT ANO-1)
AE0D/C204.(SWS EQUIPMENT FAILURES AT SONG-1)
'l IEIN 86-11 (BYRON RISK STUDY)
IEIN 86-96 (MUD AND SILT BUILDUP AT MCGUIRE)
REGIONAL INSPECTIONS.
EPRI SYMPOSIUMS ON BIOFOULING INPO DOCUMENTS W AND GE SILS O
1 TRENDS AND PATTERNS ACTIVITIES DATA MANAGEMENT 1
TRENDS AND PATTERNS ANALYSIS i
i O
)
O
O.
DATA MANAGEMENT COLLECT, SCREEN, STORE A0's, IRS REPORTING DATA BASE GENERATION, ACCESS, AND EVALUATION (SCSS, NPRDS, INFORMATION RETRIEVAL, AND REPORTING REQ'TS)
O O
i O
TRENDS AND PATTERNS l
{
CONTINUE TREND ANALYSIS OF REPORTABLE EVEN'TS (50.72 AND 50.73) TYPES AND FREQUENCY TREND CAUSES AND INITIATORS OF - SCRAMS, ESFS, T.S.,
.SSF - EXAMPLE COMPONENT TRENDING AND ANALYSIS AND FAILURE HISTORIES 0F CASE SPECIFIC EVENTS (NPRDS)
O O
O B7? S:IE.lTA"D "US A":?':E A3:E :5%
CONTRIBllTES 31% OF ALL TRIPS YEAR 1986 ROOT ELD FAILlRE E
INSlLATIONFAIL 5
DNIRON COM)
E IMPROPER COMP GROSSFAIllRE MOIST INTRUSION CORROSIW DEBRIS E
EROSION
/
AGE L
DIRTYCONTACTS I
O E01 BDOING M
POOR ELEC C0 5 M
PLUGGED MWrE EAR
?E_ m Y1BRATIW w v ::::. ' / :.*: ; e DESIGN l c ve'
'////?i.
%gg gn
--- mi X X X X R1 SETPOINT N
~
GR0l2
.v.v/t..e.*,* * * *
- i RAM)0M FAIllRE 4EPE!IiBIE3PWRNTTffM MM E__ _ _in x"x'(xY"2---- xxxxrxxr1m-d n.........,.........,...................,
0 5
18 f5 3
PERCENTAGE LEGDO: SYSTEM FEENATER
' MAIN STEM
+
...., 00ESSATE Dumm M
. *
- MAIN GEERATOR O
P O-FR'XR "I)S I"1"l BY :31EEEH 26% OF Al.i. TRIPS FROM CRITICALITY TO 1981 POER YEAR 1986 RTCAulE l
TESTEQUIPIOli 3
-,u. -
0-==
man PuxamE l!B. N
- a. wn n)...x. -
a-O
.rMg t fu......
C09ETIVE s
T==
- RMHis MRaBiW u n rAcToa W;. - >w a.:..: ::..:.,, -r:::::::; m '.,,,$,
ADGNCTR.
INATTENTION
~ *;S. : M. $+i. M,. d3E.:E.:E.:E:ii. f.$E.8}:ES...'..l.h.!..[e a f
.s 0
5 le 15 28 25 38 PENBITAGE IDE2: PERSGI LICDtlED P3 EEE llLEBtSED (PS
.,.... TEQECIM E E E PL M $TAFF
... tegmel
.. CGfTRACTOR O
O RECENT IMPROVEMENT INITIATIVES INCREASE 0.E. FEEDBACK TO IMPROVE OPERATIONAL SAFETY INCREASE USAGE OF 0.E. TO EVALUATE CORRECTIVE ACTIONS (E.G., SCRAM REDUCTION) 1 INCREASE INDEPENDENT PLANT SPECIFIC EYALUATION EXAMPLE OF RECENT WORK O
a O
H..
6 OPERATING EXPERIENCE FEEDBACK REPORT - NEW PLANTS HISTORY 1986 COMMISSION PPG GUIDANCE OCTOBER 1986 EDO LETTER TO NUMARC AE0D STUDY (NUREG-1275) INITIATED COMMISSION BRIEFING NOVEMBER 1986 - ENDORSED LESSONS LEARNED AND APPROACHJ REQUESTED FOLLOW-UP BRIEFING i
O
1 e
PURPOSE i
CHARACTERIZE POTENTIAL TRENDS IN OPERATIONAL EVENTS AT NEW PLANTS IDENTIFY CORRELATIONS BETWEEN PLANT ATTRIBUTES AND PERFORMANCE IDENTIFY IMPROVEMENT AREAS AND PROVIDE FEEDBACK THAT FACILITATED IMPROVEMENT O
I' O
O PROGRESS SINCE STUDY INITIATION STAFF AND INDUSTRY PROGRAMS HAVE BEEN FURTHER FORMULATED AND IMPLEMENTED TO IMPROYE NEW PLANT PERFORMANCE AE0D REPORT (NUREG-1275) DRAFTED AND PEER REVIEWED INP0 LETTER TO UTILITIES OF MARCH 17, 1987 AND INP0 COMMISSION BRIEFING OF MAY 13, 1987 NRR READINESS REVIEW INITIATIVES COMMISSION BRIEFING PLANNED FOR AUGUST 4, 1987.
O
4 b
NUREG-1275 PURPOSE i
TO CHARACTERIZE THE LEARNING CURVE PHENOMENA AND FEEDBACK LESSONS TO IMPROVE STARTUPS, METHODOLOGY THE MAJOR ROOT CAUSE CATEGORIES OBTAINED FROM THE ANALYSIS g 2400 REPORTED EVENTS WERE DISCUSSED WITH LICENSEES TO W ENTIFY EFFECTIVE CORRECTIVE ACTIONS AIMED AT REDUCING THE NUMBER OF REPORTABLE EVENTS AT NEW PLANTS.
I O
t O
MAJOR FINDINGS OF NEW PLANT STUDY MORE FOCUS ON OPERATIONS READINESS WITH OPERATING EXPERIENCE FEEDBACK PERSPECTIVE WILL IMPROVE. PERFORMANCE, NEW PLANT EVENT' FREQUENCIES ARE 4-5 TIMES HIGHER.
CAUSE PROFILES ARE SIMILAR.
EARLY STARTUP PERFORMANCE TENDS TO CORRELATE WITH EARLY O
C MMERCIAL OPERATIONS, NUREG-1275 LESSONS ARE VALID.
O
i S
m N
O 4
~
+,[x c*
n-yW 2
~
A 2
)
o.o,2 U
+
o.o T
f.-
e.p,, 2 0
+
m z C
g
_7,-
=
e o.
A l
- y8c
~f
~f a
E o.
1 n e
m.
F
+-
6 u O
f, g
m.
- e. -
_v.
s S
1 E
, s.
a.-
_g a
e E
r.
l
'o /
'v j
=
_q4 e3 L
1 Y
w e.
,y O
-lff
. 2
/
e..,
D
/ e...
q E
e 1
9 c
- y..
q N
/
- 0n g
h i
1 t
_-e f
e f.[:.:..x.
-q N
S n
s o
g M
/.
i:'
A
/
Y o.
q 8s 1
Y
. : : +.._-
h
=
q t
e P
a
/*
n t
,6
?
o R
/
/
/
N
.e.
t M
4f#
4 n
/
a U
l f
T il!
p P
m e
i.:'
/[
ru
/..
/
t yf
+-
2 a
M O
5 89 s
O 1
0 0
n, 0
o 0
0 o
o 0
0 o
g o
3 2
0 s
8 7
s n
4 3
n 1
1 1
wco:_OoY< 2 MoEa2 I
O 0
2 0
eo 0
o 6
i 1
e m
s ma J"
r H
c C
S
/
000 1
/
0 r
x a
1 0s o
0r
=
i u
t e
0 t
o c
a 1
R h
a t
l n
a e
a c
l R
P it i
e r
x r
u c
t O
d a
e g-M x
v e
5 i
8 t
n 9
a 1
l J,s u
n 0m a
0u i
l 0C p
6 n
U e
t i
s o
p m
o LO C0 0
0 0
o 4
3 2
1 O
aE
$._. S 3 s
4
O
, m r
=
.m 2
D r
r m N
.=
M m
E G
2 E
E T
9 TmL 1
Ae Rt 6
s My S
A a
r Ry u
o CB H
S l
O
/m a c
n it i
Do rC Es 0
i Zr 0
Ap 0
a I
L
/
1 e
m Mm
- o r
Ro c
w-S OC w__
o N
N w_
sum E
r
_".E m
N m%
1 v_
==
O
~
mu gx
- .c_
j l
I 1
4 O.
MAJOR LESSONS LEARNED FOR INDUSTRY BEGIN OPERATIONS MODE PRIOR TO FUEL LOAD TRAIN FOR ROUTINE OPERATIONS TRAIN IaC WELL FOR SURVEILLANCE ACTIVITIES
' C-UNDERSTAND EQUIPMENT DESIGN CHARACTERISTICS BEFORE STARTUP O
O LESSONS LEARNED FOR NRC_
CLOSER TRACKING 0F CORRECTIVE ACTIONS IMPROVE READINESS ASSESSMENT O
i l
O
O TN S
E E
N I
O D
P U
M T
O S
C R
G O
N O
T O
C R
A W
5 F
/
8 N
9 N
I 1
A A
M R
U T
H 4
G 8
N 9
O 1
/
T SE I
N R
U U
D G
E N
C O
O R
R W
P G
G N
N I
I V
V L
S L
O S
O Y
V E
V D
N R
N U
I G
I T
O S
S R
S T
P T
D N
N E
E N
E O
T V
V I
E E
E L
Y P
D M
U O
T C
S
O S
E E
E T S T U
4 R N R I
S 7
8 Y
E E O S E S 8
9 T
W V T R D I 9
1 E N E U E O U 1
F I
D B H L C Y
A A O F I
T T C I
D S
R S R E O R I
N R N
T A
T A
T F N
I E
A N
U F
S N O U
N E
N O G D E O O E V
A N
N P C S G T G 6
E J
T O A Y T
N 8
A R T E L O S 9
T N
E W R
H U R A 1
N I
F R E T S W W S E
E N S E
N N
D D T E F R G O O 5
O E
M H O N P I
8 I
P H
T N O T E
N T 9
I M
S N U R
N H VI A 1
O I
E F O O T L
U C
L T G T
I O ; L B
R N L
T S V N A 4
G U
E O E S A T N O V 8
N P
V R N R G N I
I E
9 O
D W N O I
E T
1 R
S A
O T T S S U R W
T N G S C S E T L I
N
/
N I
N R A E R N O E I
N E
I E F V P E S H O
V V
P N
V T
D I
A E
1 L G
I E F E
R 5 O F N S
O N H
T N
- V O
I N T N
S I
I 4 N T A N T
I I
G A
8 I
S U E
R S L
N R
M B V
R A T B
O T
N R D A
I 2 E O P N d
R I
O E E R 9 R
E F
W G
R H T T1 N R S V
/
N
, R S N 1 I
E A E
)
T O
Y E I
Y O - A S
I R
D L
B C G R N D F E
N W
U N B E T A E O I
U
/
T A U S E R M T R
T S M P T HU G U A S A
G I
U T N N H E E M
I N
N E H S S
O R P M
O U
H A
I E ; R C Y U
R T Y W V T) W T
S W
G B
A N 2 S N
F E G R
0 A E A
I O
O D M T
A O 1 W S T
I R
E E I
T R E
't.
W T S T S S T T E ",
H N D
L U S E /
I U
T E
(
F U A Y F V T N S N P
0 S C S O N U
S I
F O S
I E
I N
I A O E
Y-R N G S
U G R
L I
D O N E O N
CT S L D
U E
I O
I T G O S W
U I
I T
H T R R
N R T R G E T T
S T C W E E O W I
E N S
S I
N S D R S
N O V L
S U R A W O I
E R E S D
A A F O A M( T V GW R E
I I
T C
A E
D P
P O
S U
D 6
66 9 8 9
8 00 23 31 9
1 00 00 01 R
AEY R
O TC 5
07 85 5
A 8
1 0
2 25 1
E 9
R 1
00 00 01
/
S E
S T
M T
N I
N E
T E
V V
E H
S E
T T
4 4 3 4 4 4 8 N
I T
8 03 21 1 7 W
A N
9 L
E 1
00 00 00 E
P N
G O
6 N
W P
8 A
E M
9 H
N O
1 C
C N
/
R A
N O
O H
I 4
J T
A 8
A R
9 M
S T
1 E
/
O T
T N
A I
R N
S U
W R
O E
G H
W N
S O
O L
R S
W E E E E E E T
Y C C C C C C N
L N N N N N N E
L E E E E E E V
A I
I I
I I
I E
R R R R R R R E
E E E E E E T
N P P P P E P P W
E X X X X T X X
/
G E E E E A E E U
R W
E S S E S S S S V
E R R T R R T R R F
A T A A A A A N A A O
H A E E R E E E E E R Y Y Y Y V Y Y E
S T
E C
T T2 2 N 22 22 N
N N
E T
E A
E N N V N N N N N R
L V A A E A A E A A R
P E H H H H N H H U
T T N T T O T T C
D T
P C
E I
I E S A E S M E S O
C N R S R R S O R S N
U O E T O E C O E F
E O
I G
G G
R N
N N
E E
O O
O T
P m
R R
R A
X W
W W
R E
,I l
L
O TU O
B A
TU H
E T
S R
N U
A S
D L
A E
P W
C O
O R
R T
O P
T T
C A
N A
A E
W S
L R
O M
P L
E E
L L
M H
O B
O T
F S O
R M
R F
N O E P
E T L S
H B
L T
W G O A
S N
E N R R
E E
D R
I P
U 6
R V
E E
L D
8 U
E R
W I
T E
9 D
R A N C
1 E
H S U
S F A O
C C T C
T N
R O
U N C
N L
I P
R S E O
E E M O
4 P
V V
82 1 7 76 N O G
8 F E S
E 2 22 1 N D N
9 G
O E
O I
1 N
E R
E S E V
I Y H U
H R R L
V C T D L T
E E O
L N
E A P W V
O E F C C N
N V
U O O
O N I
I I
N Q
R T T O I
E %
P I
D I
S R 4 R
E D T T
S F 4 G C V
T N E A N
T N
L N O B T E
N N T I
S O
E I
I N
V E
U V A V
M T R E I
E V
O L W E A C S lr E
N B O
I G L S E O A V R A L A R 0
I N O S
N A P
0 T D T
E A T E
I 1
A E C
R M S R R T
S A U
/ N E O I
Y S
R R T E D
E I W
L N
A O N R E
V/
T E
O V P E
C I
N S N T
I E
V E O
E T O T E A
T E R E H R
S C A
N T I
M A
D T
P N N R T E N V
I S
Y O A T A V O I
X S
R W T N N F
I N G S C E C 0
S E
N A E O
T E N I
I R
E S
A A M H A T I
N F R E L P
R B
L W
S R N T I
I E
E A P
G O
S P S
E E
I S M D H R R A
O A S P
P A E D O T H U I
R Y
F A
Y OMTA MO T D D
P R
E O T
T E
E A
R U N Y C I
N N
E %
O W E
L O F
I I
H 2 B O H
N R O
M T1 A D T
O P I
T Y
I N
D L
E U
E D
T R
I S
P
._