ML20244C920
| ML20244C920 | |
| Person / Time | |
|---|---|
| Issue date: | 04/05/1989 |
| From: | Advisory Committee on Reactor Safeguards |
| To: | |
| References | |
| ACRS-T-1727, NUDOCS 8904210049 | |
| Download: ML20244C920 (114) | |
Text
ifcK3f~/ 787 !
OR G' N A_
O UNITED STATES NUCLEAR REGULATORY COMMISSION ADVISORY COMMITTEE ON REACT,OR SAFEGUARDS
^
In the Matter of
)
)
SUBCOMMITTEE ON INSTRUMENTATION
)
AND CONTROLS
)
)
)
O Pages:
1 through 84 Place:
Bethesda, M g
- f 19T90 e.Lhe_ I 10 0.[ M e L'I'* n r. +u m u Date
April 5, itil Y
l HERITAGE REPORTING CORPORATION O
opscialReporters 1220 L Street, N.W., Suite 600 Washington, D.C. 20005 (202) 6N 8904210049 e90403 PDR ACRS T-1727 PNU l
p
., s.
ik_).
1 PUBLIC NOTICE BY THE C
2 UNITED STATES NUCLEAR REGULATORY COMMISSION'S 3
ADVISORY COMMITTEE ON REACTOR SAFEGUARDS 4
1 5-6 7
The contents of.this stenographic _ transcript'of the 8
proceedings of the United States Nuclear. Regulatory 9
Commission's Advisory Committee on-Reactor. Safeguards-(ACRS),
]
r 10 as reported herein, is an uncorrected record of the discussions 11 recorded'at the meeting held on the above date.
12
'No member of the ACRS: Staff and no participant' at 13 this meeting accepts any responsibility for errors or n)
(_
14 inaccuracies of statement or data contained in this transcript.
15 16 17 18 4
19 20 21 22 23 24 25 O
Heritage Reporting Corporation f
(202) 628-4888
I
'u 1
i
/~S 1
UNITED STATES NUCLEARz REGULATORY COMMISSION'
(_)
2~
ADVISORY COMMITTEE ON REACTOR SAFEGUARDS 3
)
In the Matter of:
)
I 4
)
)
5 SUBCOMMITTEE ON INSTRUMENTATION
)
AND CONTROL SYSTEMS
)
i 6
)
7 Wednesday, April 5, 1989 8
Room P-110 9
7920 Norfolk Avenue Bethesda, Maryland 10 The above-entitled matter came on:for hearing, 11 pursuant to notice, at 8:30 a.m.
12 BEFORE:
DR. WILLIAM KERR
~-
g s_/
13 Professor of Nuclear Engineering University of Michigan 14 Ann Arbor, Michigan 15 ACRS MEMBERS PRESENT:
16 MR. CHARLES WYLIE Retired Chief Engineer 17 Electrical Division Duke Power Company 18 Charlotte, North Carolina 19 MR. JAMES C.
CARROLL Retired Manager, Nuclear Operations Support
[
20 Department Pacific Gas & Electric 21 San' Francisco, California 22 ACRS COGNIZANT STAFF MEMBER:
23 M.
El-Zeftaway
()
24 25 i
HERITAGE REPORTING CORPORATION -- (202)628-4888 Li__.________
.1
2
'i..
'l CONSULTANTS:
2 P. Davis L. Oakes
'3 W.
Lipinski 4'
NRC STAFF PPsESENTERS:
5 D. Basdekas K.
Kniel 7
8 9:
10 11 12' 13 14 15 16 17 l
18 19 20 21 22 23 24 L
25 l
HERITAGE REPORTING CORPORATION -- (202)628-4888
L 3
e-1 PROCEEDINGS.
ll 2
CHAIRMAN KERR:
.The meeting will come to order.
It.
i 3'
is a meeting of the Advisory Committee on Reactor Safeguards, l
4 specifically the Subcommittee on-Instrumentation and Control l'
5 Systems.
6 My name is Kerr.
I am Subcommittee Chairman.
Other 7
ACRS members.present are Mr.,
Carroll, Mr. Wylie.
We expect 8
Mr. Lewis ~and Mr. Michelson.
9 Our consultants today are Mr. Lipinski, Mr. Davis, 10 and I want to recognize Mr. Les.Oakes from Oak Ridge who has 11-recently been signed on as an ACRS consultant and is~with us 12 for the first time.
l().
13 The purpose of the meeting.is to review the proposed 14 resolution of Generic Issue 115 entitled, " Enhancement of 15 Reliability of the-Westinghouse Solid State Protection 16-System."
Mr. El-Zeftawy is the cognizant ACRS staff member 17 for the meeting.
18 The rules for participation were announced as part 19 of the notice of the meeting published in the Federal 1 Register 20 of Wednesday, March 22nd, of this year.
21 A transcript of the meeting is being kept and will 22 be available as stated in the Federal Register notice, and I 23 would ask that each speaker' identify himself or herself and
(}
24 use a microphone.
25 Are there any comments or questions from members of HERITAGE REPORTING CORPORATION -- (202)628-4888
__ _ _ _ - _ _ _ _ - _ _ - - - _ _ _ _ _ _ - _ - _ - _ _ - _ _ _ _ _ _ _ ~
- 4
.c 7e j9 1
.the Subcommittee that you want to give any particular k,/
2 attention?
3 MR. CARROLL:
I have none at-this time.
4 MR. WYLIE:
No.
5 MR. LIPINSKI:
In looking at options, the one option 1
6 that was not considered was to leave the existing systemLthere 7-and to add a diverse--
8 MR. WYLIE:
I thought they did.
9 MR. LIPINSKI:
They took one breaker out and put in 10 a contact in its place, but no option--but what if-they are 11 both in place and add a third diverse system?
And.I would 12 like that addressed.
(~)r.
13 CHAIRMAN KERR:
You would like to know why they--
(_
14 MR. LIPINSKI:
Didn't consider it.
15 CHAIRMAN KERR:
In your presentation; any other 16 comments?
17 I would like to comment, and in light of having 18 reviewed a resolution of another issue fairly recently, that I 19 have found an evaluation described in, published in NUREG 5197 20 a much more cogent and understandable analysis.
I thought the 21 preparers did--at least to me it was understandable and it 22 seems logical, and it did not go to common mode failures, 23 although the treatment of common mode failures still strikes
(}
24 me as smacking of witchcraft to some extent.
Nevertheless, 25 they used the witchcraft in an acceptable, in an accepted HERITAGE REPORTING CORPORATION -- (202)628-4888
5 g
l'.
manner.
>~/a
_ 2 MR.'MINNERS:
You forgot uncertainty.
3 CHAIRMAN'KERR:
Well,'I did--in reading 1.t, I was a 4~
bit curious and perhaps I' missed this.
If I did, I like to 5
think it may have it-identified.
There seems'to be no-
- 6 discussion of possible vulnerability of solid state system to
'7 electric spikes, the kind that might be generated by 8
lightning, or perhaps it was'there,1and I missed it.
9 Then it also, I did not see any comments on possible' 10 common mode failures that might.be due, for example, to i
11 overheating which.would be caused by failure of the air 12 conditioning system, and at least some solid state systems are
)
l I[
13 vulnerable to this sort of thing, and I really wasn't sure 14 whether I missed it or whether it is just common knowledge j
15 that it isn't vulnerable or something, so I would be i
16 interested in a comment.
I 17 Then in not the analysis itself, but in the letter j-18 that Mr. Beckjord wrote, there were some recommendations that 19 resulted from this study for existing reactors, some things 20 that might be done, not necessarily to be required by the 1
21 staff I gathered, but some suggestions, and I think this is, 22 this is a good idea because it would be unfortunate for all l-23 this effort to occur without some recommendations, and of j}
24 course, one recommendation that is logical is to leave things 25 alone, but at least the studies that were suggested may be 1
i HERITAGE REPORTING CORPORATION -- (202)628-4888
6 1
.make-sense, 2
However, I would also recommend that this staff not y
'3
' confine itself--in fact, I guess the. staff ~didn't confine 4
' itself to. existing reactors.
It did I think say-that this 5
could be applied both to existing and to future reactors, and if that was the case, I would suggest also that some 6
7 consideration be given to recommendation that future designs 8
be changed slightly so that ATWS is not a problem.
9 I'm convinced that you can certainly make the 10 consequences of an ATWS, of an ATWS much less severe than it 11 is in some existing systems by design changes that at least 2
12 people who know more about design than I do seem to think 13 might not be so traumatic, and it does not-appear to me that 14 in the new reactors that are being considered, I guess the 15 evolutionary ones, these changes are being considered 16 seriously.
17 Those are the. comments I had.
And let's see.
Mr.
18 Kniel it says is going to open things up.
Is that correct, 19 Mr. Kniel?
20 MR. KNIEL:
My name is Carl Kniel.
I am Chief of 21 the Reactor Plant Safety.
22 CHAIRMAN KERR:
Don't be mike shy.
Any of you who 23 would find it more convenient to sit here at the table to use
{)
24 the Mike, please feel free to do so.
25 MR. KNIEL:
My name is Carl Kniel.
I am Chief of HERITAGE REPORTING CORPORATION -- (202)628-4888 I
c.
7
('y 1-the Reactor Plant Service Issues Branch, the Division of
%)
2 Safety Issuo Resolution, and we are here this morning at the 3
invitation of the Committee to give you a presentation on the, 4
our resolution of Generic Issue 115, and this is,'we are doing 5
this in, as a follow-up to the agreement between our 6
management and the Committee as to that you would at your 7
option that you want to review issues for which we don't have 8
a new requirement, and as you are aware, this particular issue 9
doesn't have a requirement, and we sent it down to you at your 10 option to have a meeting, and you have elected to have a 11 meeting, and that's why we are here today, and Demetrius 12 Basdekas is our task manager, and he will give the
()
13 presentation on this issue.
14 CHAIRMAN KERR:
You didn't say we are delighted to 15 be here.
16 MR. KNIEL:
No, I didn't.
17 MR. CARROLL:
Are you?
18 CHAIRMAN KERR:
Don't answer.
l l
19 MR. MINNERS:
Take the Fifth!
20 MR. BASDEKAS:
My name is Demetrius Basdekas, and l
21 I'm with the Reactor Plant System Branch, Division of Safety 1
22 Resolution, and I'm the task manager for the resolution of 23 Generic Issue 115, which has been to develop the Westinghouse
()
24 protection system.
25 During my presentation I will try to give you some I
HERITAGE REPORTING CORPORATION -- (202)628-4888 1
g L
8-
]
I h
[
1; of the background that led to this establishment of this 2
issue.
I will give you the-objectives and scope that was i
3' established for this resolution, and I perfectly understand 4
the comments that were made earlier to the effect that perhaps 5'
this. scope might or might not have included some additional V,
6 options or some. additional evaluation that otherwise might 7-
'have been appropriate.
8 As you may recall from your previous experience, and 9
' as well,as from reading the package that was sent you 10 recently, this generic issue was established as a result of 11 the two ATWS events that have taken place at the Salem plant 12 in 1983.
Those events initiated a number of actions, notably
()
13 including the issuance of a Generic Letter 83-28, and as a 14 sequel-to that generic letter, another evaluation were U
15 performed by the staff as well as the industry, and the scope 16 of this generic issue was established at that time as one that 17 the staff had particular concerns regarding the reliability of 18 the undervoltage driver card in the Westinghouse plants which 19 had exhibited some. problematic behavior in terms of failures 20 that would have prevented possibly the action of the reactor 21 when needed, and also the problems Associated with 22 undervoltage tip device, the undervoltage coil in the reactor 23 assembly itself, so in other words, work will be addressed and
(])
24 have been, and/or are being addressed under Generic Letter 25 83-28, and the implementation that the staff has made for the HERITAGE REPORTING CORPORATION -- (202)628-4888
9 L
1 scope of this generic ~ issue is basically two options which j3 U. -
2 were basically centered around fixes of problems associated; 2
with the undervoltage driver card which showed problems either
~
4 by accepting the recommendations of Westinghouse which i
l'
!I submitted,-who sent a bulletin to.allHof the utilities-l l
6 outlining.the problem and proposing solutions both inI he near t
7 as.well.as in the long term as well as an additional staff I
l 8
recommendation of providing a redundant driver for the I
l 9
undervoltage coil that will be made of drivers, namely, relays H
J 10.
placed vis-a-vis the solid state component, components that l
1 11-were used by Westinghouse, but as the generic issue progressed 12 through its process of definition and subsequently more (3.
)'
13 importantly, through its process of the resol'ution phases, it'
-14 evolved somewhat and it evolved in the sense that the staff I
15 felt there were some additional options that ought to be J
16-considered, and the two increased to six, and then we had some 17 more comment, but I'll discuss these options just as we go 18 here.
19 (Slide) 20 MR. BASDEKAS:
As I mentioned, there were ultimately 21 six options that were evaluated that came out initially, but 1
22 as well as in the process of our evaluation of the initial two l
1 23 options as we were going through the evaluation of this issue 24 and as we, the staff and others were going through their
{}
r 25 evaluation of the licensee responses or compliance with the HERITAGE REPORTING CORPORATION -- (202)628-4888
10 7'j 1
~ requirements of Generic Letter 83-28, then some additional
%.)
2 options were suggested to us which we undertook also to I.
3 evaluate in a large initial scope of.this icsue.
L 4
The first option I already touched on briefly, was
~
5 a,
simply the replacement of the existing undervoltage driver 6
cards in the logic cabinets supplied by Westinghouse to avoid 7
the problem of shortened output of that card and damaging the' 8
transistor, the output transistor of that card that would have 9
resulted in undetected failure leading perhaps to failure of 10 the tip if it was, if both cards were involved, both things, 11 so Westinghouse had recommended a new design of a card that 12 simply incorporated the fusable, the output of that cards so
()
13 that the particular failure mode would be eliminated.
It 14 would still open up and not be able to, you know, to stop the 15 reactor, but nonetheless, it would have eliminated, does 16 eliminate the possibility of having undetected failure that 17 results in a failure of the breaker to trip.
18 Our survey showed that some of the licensees have 19 choseen to modify the operating procedures and introduce test-20 procedures that will avoid shorting of the output, but some 21 others, and I think they are the minority, chose to purchase 22 these new modified cards from Westinghouse and install them in 23 their system, so it is, it has been a mix, and--
.( )
24 MR. CARROLL:
One of the things that was not clear 25 to me in going through this material was what was the naturo HERITAGE REPORTING CORPORATION -- (202)628-4888
11 L
l)"4
.1 of or what really caused'these various failures?
M 2
MR. BASDEKAS:
It was during'the periodic testing 3
process that the. technician would short out the output and 14 damage the transistor.
5 CHAIRMAN KERR:
In one word, what--
6 MR. BASDEKAS:
Pardon me?
7 CHAIRMAN KERR:
I said in one word, what caused it?
8 MR. BASDEKAS:
Operator, maintenance testing.
9 MR. CARROLL:
The statement is made somewhere that 10 Option 1 would cure four of the five failure situations.
What 11 was unique about the fifth?
~
12 MR. BASDEKAS:
It would not cure all failures.
It
't I
13 will cure the possibility of having the, this particular card, 14 the transistor in particular, damaged in such a way that it l
l 15 will not trip when it'was supposad to and by fusable link--
I 16 MR. CARROLL:
But of the five actual failures, I 17 thought I read that Option 1 would have dealt with four of the i
18 five?
19 MR. BASDEKAS:
Yes.
20 MR. CARROLL:
What was unique about the fifth one?
21 I didn't find--
)
22 MR. BASDEKAS:
Okay.
The fifth one, Don, wou d you 23 comment on this one, please, what we have done?
Okay.
I will
(}
24 tell this later, but in any event, give a quick response.
25 MR. CARROLL:
I can wait if you are going to cover HERITAGE REPORTING CORPORATION -- (202)628-4888
i~
I 12 f
1 it later.
l
\\-
l:
2
'MR.'BASDEKAS:
.Go ahead.
l-I 3
MR. RENY:
I am Dan Reny from EG and G Idaho - The 4
fifth' failure was a random failure of component in the 5
undervoltage. driver : card unrelated to the other four f ailures 6
which caused the transistor switch problem, so the fix would 7
not affect that fifth failure.
8 MR. CARROLL:
On page 25 of the NUREG, which of 9
those failures is that?
10 MR. BASDEKAS:
Referring to NUREG or NUREG/CR?
11 MR. CARROLL:
NUREG/CR.
12 MR. BASDEKAS:
Okay.
)
13 DR. KERR:
Do you have a copy readily available, Mr.
14 Reny?
15 MR. RENY:
I do.
16 (There was a brief pause in the proceedings.)
17 MR. RENY:
The first one, the diode failure.
4 18 MR. CARROLL:
Okay.
That's what I guessed.
All 19 right.
Thank you.
20 MR. BASDEKAS:
We do have a--okay--details under the 21 specific options, and they are proposed, what the staff 22 outlined and we have examined here--we can cover later, and I 23 do have some other--the last six vugraphs are back-up
-( )
24 vugraphs, schematics of each one of the options, so let me go 25 quickly over the other options we have examined.
HERITAGE REPORTING CORPORATION -- (202)628-4888
13
/~s 1-Now Option.No.
2' involved, I referred to earlier ~as b
2 a diverse undervoltage driver, relays to be installed in 3
parallel with the existing solid state.undervoltage driver 4
cards.
5 Option No. 3 involves the diverse RTB actuation 6
mechanism that involved the incorporation of another current 7
type of circuitry that will burn up a fusable link in the same 8
line as the RTBs.
9 Option No. 4 involved the deployment of an extensive 10 diverse redundant trip logic system constructed of relays that 11 will be the redundant and diverse counterpart of the existing 12 solid state universal trip logic system that comes closer
( ).
13 basically to what now Westinghouse plants would be required to1 14 do.
15 Option No. 5 involved the replacement of the 16 undervoltage driver, the undervoltage coil,.the undervoltage 17 trip device which turned out to be the most unreliable in the 18 particular part of the system, with a device of similar design 19 as the shunt trip device.
g 20 And Option No. 6 involved the replacement of one l
21 reactor trip breaker with a contactor, and that presently you 22 may recall we do have a configuration of one of the two--we 23 have two reactor trip breakers in series and two bypass
{}
24 breakers in series connected in parallel, so this is, we have l
25 replaced one reactor trip breaker and one bypass breaker with l
l HERITAGE REPORTING CORPORATION -- (202)628-4888 l
l
14 1
a contactor being a simpler device, device existing for 2.
reactor trip breakers, the best approach we.took.
3 MR..WYLIE:
Would this be a time to talk about'the 4
question about why other options were_not considered?
5 MR. BASDEKAS:
I'll be glad-to answer yiu.
6' MR. WYLIE:
Do you want to restate the question?
I 7
MR. LIPINSKI:
My question is why didn't you add a 8
third diverse. trip system?
Let me point out these two 9
mechanical breakers are the weak link in this entire' system.
10 Just scanning over rour data, in' fact your data factor under.
I 11 there is something like five times ten to the minus 3 12 probability of failure to demand on'both of them.
(
13 MR. BASDEKAS:
No, I think the numbers, we have 14 got--
15 MR. LIPINSKI:
You did your analysis but giving two 16 breakers in series, and if I.take their independent product 17 and use it the way the new data' compares,.you should end'up 18 with five times ten to the minus 3.
When you get to the 19 analysis, we will take a closer look.
20 MR. BASDEKAS:
Let's do that.
I think ue will be 21 delighted to do that.
It's an issue that has been debated 22 extensively.
It's a good legitimate question and we will 23 address it.
Can we take a few minutes as we go through it and 4
1
()
24 and come to a point where we can do that?
25 MR. LIPINSKI:
Sure, but the point is using my HERITAGE REPORTING CORPORATION -- (202)628-4888
15 y-1 contention that the.two' breakers are the weak lihkfin the Os 2
-system, you would have to go in with a third system that would 3
be diverse and not subject'to the common cause failure.
4 MR. BASDEKAS:
I understand your point.
5-MR. LIPINSKI:
Let's talk about common cause 6
failures.
You are looking at'a. single plant.
I am going to 7
postulate that Westinghouse came up with a new super-lubricantL 8
and then sent a letter out to.all the operators saying here is 9
this lubricant, install it in your breakers, but this-10 lubricant has a peculiar property.
It works nice for one 11 year, after one year, turns to a property like epoxy, so now I 12 not only have a common mode on the single plant, I have a
/'h
' _j 13 common mode on a population of plants due to this common
(
14 cause.
15 That type of analysis is not factored in, but given 16-a diverse system, I think you have a lot more to offer in 17 terms of going to a diverse trip.
18 MR, BASDEKAS:
Well, let me' defer responding in 19 detail to your question, and I think already in this case Don 20 will be also helping with the presentation because he will 5
21 give you, when we come to it, a detailed step by step i
22 description of how we have done this analysis, the assumptions 23 we made, if any, and the data we used, how we model the
{}
24 system, how we derive better factors for common cause 25 failures, and will give you the reasons why our data turned HERITAGE REPORTING CORPORATION -- (202)628-4888
16
(^
1 out to be based on what you cited for the.two breakers v}
2 combined, which is basically the trip function of,'the good
-3 part of trip function, of the trip function of the reactor-4 Litself, the unreliability of that.
Then we will show you that 5
our numbers are different markedly from yours by--
6 MR. LIPINSKI:
Let me ask you this question.
Your 7
report does not deal with sizes.
What is the interrupting 8
current and voltage that goes through a breaker?
9 MR. BASDEKAS:
Well, it is 480 volts.
I don't 10 remember the amperage, but it is certainly--Don, do you happen 11 to remember?
(
12 MR. RENY:
I believe it is 600 amps.
O
(/
13 MR. BASDEKAS:
So we have examined the existing, you 14 know, breakers, and we have examined operational experience.
15 We have examined the potential failure modes that may be 16 involved.
We have used an approach for figuring out the 17 common mode factors the other factors.
18 MR. LIPINSKI:
We will talk about common mode later?
19 MR. BASDEKAS:
Sure.
20 MR. LIPINSKI:
In terms of waiving--but diverse is 21 not coming into the picture other than in your one case.
22 MR. BASDEKAS:
We did provide in one case, you will 23 notice we provided a fusable link for both trains.
()
2(
MR. LIPINSKI:
That doesn't get rid of my mechanical 25 breaker common mode.
HERITAGE REPORTING CORPORATION -- (202)628-4888
E l
l?
E 17 L
. (']
1 MR. BASDEKAS:
But it-is a diverse way of
\\/
2 interrupting power, 3
MR. LIPINSKI:
It does not guarantee the two 4
' mechanical devices will function if I have done something l
5 wrong to them.
6 CHAIRMAN KERR:
Why don't you, as you said, look at 7
this in detail?
There is a more logical matter.
8 MR. BASDEKAS:
I will come to it and I will keep it 1
9 in mind to revisit the question.
10 The best approach we took here was to--
11 MR. OAKES:
Demetrius, I wanted to ask a question.
12 Perhaps it isn't the correct time to ask it, either, but maybe
()
13 we will'take the issue up later.
14 Are you going to discuss these options in more 15 detail, or is that the last time you are going to talk about
~
16 the options?
17 For example, Option 1, one's sensitibility is 18 offended by the fact that we have solved a, what appears to be 19 an incompetence problem in maintenance by adding a 20 non-testable device, a fuse.
21 MR. BASDEKAS:
No.
It is not--well, they, they have 22 provisions for considering the fuse and make sure that it is--
23 MR. OAKES:
What I was going to ask is--perhaps this
(}
24 isn't germane to our discussion, but have other options for 25 limiting the current been looked at, something that is HERITAGE REPORTING CORPORATION -- (202)628-4888
L l
.18 l
[
1 testable by current limiters, and does one know that the fuse 1
c f-2 will always blow before'the' resistor is' damaged in-this-case?
3 Are there data to support it?
4 MR. BASDEKAS:
It is a quality assurance question.
5 MR. OAKES:
Have tests actually been made to show 6
that it will or not in most cases blow before it damages the 7
resistor?
8 MR. BASDEKAS:
I'm not aware of any test performed 9
by Westinghouse or anybody else, but the description of.the 10 quality assurance is programs, they do address the: integrity 11 of the new design, notably the fuse.
12 CHAIRMAN KERR:
Mr. Reny, you were about to make a
()
13 comment?
14 MR. RENY:
Yes.
The fuse was added to prevent the 15 use of the undervoltage driver after a short circuit has been 16 applied to it such that the undervoltage driver would have to 17 be removed and tested prior to its use again, so the fuse 18 prevents the use of the cards in case the transistors have 19 been damaged before the fuse blew.
20 MR. LIPINSKI:
You are saying short, but you can 21 have various degrees of shorting which are not necessarily 22 zero ohms that draw current.
23 The question is can you draw the current through 24 that fuse damage, the output transistor, and not blow the
(}
25 fuse?
HERITAGE REPORTING CORPORATION -- (202)628-4888
l
' 19 '
r~
1 MR. BASDEKAS:
The size of fuse was determined on
<~
?l ):
N a
2-the basis for such as the--
3 MR. LIPINSKI:
Zero ohms, that is an assumption on 4
your part--they are very low in peak shorts that draw high I'
5
' currents.
6 MR. BASDEKAS:
That's a good point, but I think'it 7
has been provided as part of the design.
Faust Rosa?
8 CHAIRMAN KERR:
Faust, will you come to a mike, 9
please, sir?
10 MR. ROSA:
I am Faust Rosa, Chief of the Electrical-
.6 11 Systems Branch in NRR.
12 I recall the presentation that Westinghouse made on
(( )
13
.the subject of these driver cards some years ago, and as I 14 recall, the fuse is rated at about five milliamperes, and the 15 transistor can withstand up to around 20 milliamperes, so 16
-there is, there is very little change that a short circuit 17 will not blow the fuse before it damages the transistor.
18 MR. WYLIE:
Let me ask a question in that regard.
19 You know, they have a habit of opening when you don't want 20 them to open.
21 Do you know wiisther Westinghouse or anybody looked 22 at the increased probability of SCRAM rates due to the 23 addition of these fuses to opening when you don't want them to 24 open?
[}
25 MR. ROSA:
I'm not aware of any Westinghouse study HERITAGE REPORTING CORPORATION -- (202)628-4888
20 l'-
on that.
\\g3v).
]
2 MR. BASDEKAS:
However, they did make a provision,
'3 Charlie, that they do X-ray them in case there is some defect.
4 MR. WYLIE:
You know, fuses as they age, at least my 5
experience has been that strange things happen.
They.open i
6 when you. don't.want them to open.
They all are good when they j
7 are new, but generally-these are silver sand fuses or some 8
other type of fuse, and after some years, why they start 9
opening, and I wondered whether anybody looked at that as to 10 what the risk associated with increased SCRAM rates are by 11 adding fuses.
12 MR. KNIEL:
I think about--
()
13 CHAIRMAN KERR:
Microphone, please, sir, We record 14 these priceless words, 15' MR. KNIEL:
According to an INPO document, 25
.16 percent of the SCRAMS come from various malfunctions in the 17 reactor protection system at this stage, so--I think it is 22 18 peteent to be exact, so I don't think that this hypothetical 19 mechanism you are talking about is going to have very 20 significant impact on something that is already pretty, pretty 21 dominant.
22 MR. WYLIE:
You mean by increasing the scam rate?
23 MR. KNIEL:
Right.
And the SCRAM rate deficiencies,
{}
24 I mean the deficiencies from the SCRAM rate come from other 25 components in the reactor protection system, principally the HERITAGE REPORTING CORPORATION -- (202)628-4888
l C
i
'21 L
l 1
instrumentation and the, and the lines that bring in the i
2 instrumentation and'that-kind of stuff; doesn't come from the 3
reactor protection system, the SCRAM breakers, l'
4 CHAIRMAN KERR:
Let me say I wasn't sure I 5
understood what you said.
Twenty-five?
6 MR..KNIEL:
Twenty-two percent is-the number.
7 CHAIRMAN KERR:
Twenty-two percent of the false 8
SCRAMS, what other term we were using, come from malfunctions i
9 of the trip breaker?
10 MR. KNIEL:
No--come from malfunction of the reactor 11 protection system.
l 12 CHAIRMAN KERR:
Oh, okay.
)
13 MR. KNIEL:
Only a very sma11' fraction of'those come 14 from the breaker malfunctions.
'This is the existing data that 15 we have today.
16 CHAIRMAN KERR:
I don't know what 25 percent means.
17 MR. KNIEL:
I think right now what is it, like four 18 SCRAMS per reactor year?
Per plant; roughly in the ballpark, 19 so--
20 CHAIRMAN KERR:
If you had an additional SCRAM, it 21 would incrase things by 20 percent?
Twenty-five percent I 22 guess?
23 MR. KNIEL:
Yes.
24 MR. LIPINSKI:
Per year?
25 CHAIRMAN KERR:
Yes.
i HERITAGE REPORTING CORPORATION -- (202)628-4888
(_-_-_-
1 22 y
h
'(~N 1
MR. KNIEL:
What I am trying to point out is that u
L 2
'this'smal1 slight increase in possibility of the fuse failing
~
l-3 and SCRAMing the plant, what appears to me a microscopic.
14 number compared to what we have in the existing--
5 CHAIRMAN KERR:
The response to Mr. Wylie's question 6
is nobody has looked at it?
7 MR. KNIEL:
That is correct.
8 CHAIRMAN KERR:
Which is what he asked, if anybody 9
had looked at it, and the answer is nobody is?
10 MR. KNIEL:
Well, I am trying--
11 CHAIRMAN KERR:
Perhaps for good reason.
12 MR. KNIEL:
I am trying to give you the facts as I
( ).
13 understand today as reported in the INPO report, and the 14 implication is that the mechanism that suggested would have a 15 very minor--
16 CHAIRMAN KERR:
He is not the sort of person that-17 goes around making implications.
He just asked a question.
18 MR. KNIEL:
Okay.
I am trying to give the--
19 MR. BASDEKAS:
Let me reiterate what I said earlier.
20 It is a concern, good one.
It is a valid one shared by 21 Westinghouse.
They provided, they have a provision as part of 22 their quality assurance program to X-ray these fuses before 23 installation.
()
24 Now whether or not this calls for priority X-ray, I 25 am not in a position to say, but that we will find out.
I HERITAGE REPORTING CORPORATION -- (202)628-4888
23 q'T 1-will let you know, Charlie, so that we close the-loop.on this V
2-question, but as I said, it is a concern that is shared by 3
Westinghouse and the staff and to address it at least in a 7
4 practical, meaningful way, will'be to have the X-ray at least 5
before.insta11aticn.
I'm not sure if that entails.also 6
X-raying priority thereafter.
'7 Going back to the--
8 MR. CARROLL:
I have one thing that was troubling 9
'me.
I guess I heard that Westinghouse' rates the fuse, what 10 was it, 5 milliamps?- And'the transistor is quote, good for 20 11' milliamps, and that really doesn't tell the whole story.
12 The fuse blows, it doesn't heat, which is there is
()
13 some relationship between current and resistance and time, and 14 the transistor is damaged because of heat which is some other 15 relationship involving current and resistance and heat and 16 time.
17 I'm not sure a 5 amp, quote, 5 amp fuse protects a 18 quote, 20 amp rated transistor on the face of it.
19 MR. BASDEKAS:
Well, I do not have that direct 20 knowledge, but I think based on my understanding of what the 21 process of sizing design in this particular modified card, we 22 went to the sizing of the fuse.
23 I cannot--Faust--
()
24 MR. CARROLL:
Did Westinghouse run tests that would 25 make me feel better if they did?
HERITAGE REPORTING CORPORATION -- (202)628-4888
24-
. -41 1
MR. BASDEKAS:
I cannot answer the question.
2-MR. ROSA:
Yes, Westinghouse did run tests, and the 3
addition of this, this fuse in the output driver card is their 4
recommendation, which they have transmitted to all of their 5
licensees.
6 MR. LIPINSKI:
Let me pursue that question because I 7
would venture probably the test that Westinghouse ran was a 8-steady state test, but you put a transistor test on, you have 9-a momentary short and releaseHit; did they look at it as a 10 function of time, zero out in terms of mode seconds 11 application of the short?
12 MR. ROSA:
I have no idea of the details of the O
(,j 13 test.
14 MR. LIPINSKI:
Then it is a race between the heating 15 of the fuse and the heating of the transistor as to who goes 16 first.
17 MR. BASDEKAS:
I think the fuse will go first.
18 MR. LIPINSKI:
But the point is if I only have a 19 momentary and then heat the fuse long enough to open it, but I 20 could blow the transistor which is faster responding 21 temperature-wise--
l 22 MR. BASDEKAS:
Your point is well taken, but we do i
l 23 have fuse and fuse, we have quick response fuse and slow I
l
\\
-()
24 burning fuses.
1 25 MR. LIPINSKI:
Right now we are discussing the l
HERITAGE REPORTING CORPORATION -- (202)628-4888 L __ _ ___ _.___
H
K 25 l
l 1
critical pointLin the decision, o
t
.w 2
MR. WYLIE:
What kind--
3 MR.. ROSA:
Demetrius, about seven or eight 4
Westinghouse plants.have installed these driver cards. -They
~5 have been installed for some five years.
6 MR. BASDEKAS:
It varies.
Not necessarily; some 7'
perhaps as long as five, some perhaps as long as one or more.
8 MR. ROSA:
We haven't'had any reports of a failure 9
of one of those fuse scausing a spurious SCRAM.
10 MR. CARROLL:
But the other variation of that is to 1
11 disable the driver cards and don't know it.
In other words, j
12 making this mod seems'to me to give people a false sense of
( ).
13.
security.
14 MR. BASDEKAS:
If the fuse, if the fuse burns or if 15 the fuse does not burn, it does not burn, you are no worse 16-than what you have with the fuse.
17 MR. ROSA:
The fuses in this series is with the, 18 with the trip coil, the undervoltage trip coil, so if the fuse 19 blows, you get a SCRAM, period.
20 MR. CARROLL:
But the transistor is damaged.
l 21 MR. ROSA:
Then you must rely on your periodic 22 testing which occurs every two months, every two months now.
23 MR. BASDEKAS:
Highly unlikely occurrence, Mr.
)
i
(' }
24 Carroll.
25 MR. LIPINSKI:
Let's get back to the testing mode, f
f HERITAGE REPORTING CORPORATION -- (202)628-4888
26.
i ~
r's 1
Are both--
b.
2 MR. BASDEKAS:
Of what?
3 MR. LIPINSKI:
Undervoltage drivers; are they both 4
tested at the same time, or tested on alternate months?.
5 MR. BASDEKAS: ' Alternate months.
6 MR. LIPINSKI:
So they can't be subject to the-same 7
failure at the same time?-
8.
MR. BASDEKAS:--That is correct.
The basic approach 9
we took in resolving of this issue, as you know, it is 10 basically that one deploys, of course, the cost / benefit the 11 methodology, and to accomplish that we perform a reliability 12 analysis of the reactor trip system.
We started with the base O
- (_)
13 case and we have done the same thing for the options as 14 modified based on the proposed modifications.
15 We did core damage frequencies for each one of them.
16 We proceeded to the consequences based on a site, typical 8
17 Midwestern site characteristics.
We did perform a cost 18 analysis for each option, and we did perform a scientific 19 analysis to show us the leval of confidence we can attach to 1
(
20 the results we are getting and to see which parts of the 21 system are the most contributing to risk in this particular 22 system.
23 And' finally, we, and I see a mistake here which on i
l
()
24 the slide we have a discussion with the decision rationale and i
25
-the recommendation we developed as a result of this analysis.
1 I
HERITAGE REPORTING CORPORATION -- (202)628-4888 L
K.
27
!.X 1
(Slide) lu Q)-'
l 2
MR. BASDEKAS:
.I will try to~ discuss next the steps-l 3
we went.through to have the, this process. accomplished, and j
l 4
the first thing we have done was to evaluate the, the base 5
case six options.
(!
6 The base. case includes the automatic shunt trip L
7 function whichswas added after the 1983 events at Salem.
That 8
particular addition alone contributes significantly.to the 9
improvement of the trip reliability.
10 We have made some assumptions here that we find them l'
11 reasonable, and.they are namely, we assume that all electrical l
12 power sources were available to allow the reactor protection
( ).
13 system to function properly, and we have also assumed'that 14 once the trip is initiated, it will go to completion as far as
.15 '
movement of the control rods.
In other words, our scope of 16 work did not include the machanical part of the actual rods 17 per se, move it into the core.
18 The next vugraph shows the functional diagram and 19 one that we used--the reactor protection system, and one that 20 we used to model it for purposes of analyzing its reliability.
21 (Slide) 22 MR. DAVIS:
Excuse me.
I have a question on your 23 previous slide.
{}
24 MR. BASDEKAS:
Previous slide, yes, sir.
25 MR. DAVIS:
Did you assume that if SCRAM failed, you HERITAGE REPORTING CORPORATION -- (202)628-4888 i-I
28 jg 1
would always.'get core damage?
V 2
.MR.
BASDEKAS:
No.
3 MR. DAVIS:
Because I believe for Westinghouse 4
plants, they can-ride through it and not guarantee it is 5
unfavorable.
6 MR. BASDEKAS:
That is correct.
We took that into 7
account, and we can cover it as we go down the line in the 8
consequences and the calculations.
9 MR. DAVIS:
You may want to revise the last 10 paragraph, Roman 12, of your NUREG 1341 because it says there 11
'that SCRAM must take place within one minute to avoid core 12 damage.
)
13 MR. BASDEKAS:
I see--in most cases.
14 MR. DAVIS:
I'm not sure it is in most cases.
15 MR. BASDEKAS:
At least in selected cases.
16 MR. MINNERS:
Don't rewrite it now.
17 MR. BASDEKAS:
It is intended to show it is 18 important as far as the operator action goes.
It is part of 19 the operator take some action within one minute.
20 DR. KERR:
No matter what it is intended to show, if l
21 it is incorrect, I expect you will want to look at it again.
22 MR. BASDEKAS:
We will certainly look at it.
23 MR. DAVIS:
Westinghouse has made a big case for 1
()
24
' installing the AMSAK system that will allow them to ride 25 through the SCRAM failure, i
HERITAGE REPORTING CORPORATION -- (202)628-4888
-_a
l
-29 yw.
1-MR. BASDEKAS:
That is exactly-right.
We will
- k_)
2 discuss that in the Executive Summary-as well as in--
3 MR. MINNERS:
What was the page again?
-4l MR. DAVIS:
Roman 12, the last' paragraph; thank you.
5
.You have answered the. question.
6 MR. BASDEKAS:
Sure'.
Thank you.
7 (Slide) 8 MR. BASDEKAS:
Okay.
This document shows the best 9-portions of the reactor protection system'are on the basis 10 of--which our contractor developed the reliability model that 11 was implemented in the IRIS computer code.
That stands for 12 Integrated Reactor Risk Analysis System, and this has'been
()
13 very useful in being able to analyze the various options as 14 well as the sensitivity of the system to various parameters 15-that are involved here.
16 Basically as you see here, as you have seen before, 17 is the channels, and the number of them varies from design to 18 design.
Some, they have three.
Some may have four, or even 19 two in some instances.
20 Then you have the input relays and universal logic, 21 part of which is the undervoltage driver cards on both trains, 22 and finally, the two reactor trip breakers.
23 As you notice there, the operator enters and can 24 override or compensate for a failure in the system starting
(}
25 with the input relays up to and including the undervoltage HERITAGE REPORTING CORPORATION -- (202)628-4888
30 s
l' driver: cards.
This was part of the considerations we had to,
- C 2
we did make and. carry a number of these calculations starting 3
with the particular ability core damage frequency consequences 4
and cost benefits with or without operator action.
We felt 5
strongly_that operator error here can have some significant 6
impact, and we have carried this throughout.
7 The next--
8 MR. CARROLL:
With regard to the operator action, 9
9 I'm sure it doesn't apply or doesn't have a major effect, I
-10 guess I would point out that in projects with a high seismic 11 design, the operator ain't there.
He is on~the floor.
12 MR. BASDEKAS:
For most cases--
(~%
(_,j 13 MR. CARROLL:
Is not available to intervene and 14 initiate a SCRAM, 15 MR. BASDEKAS:
Sure.
That is correct.
That's, for 16 those cases, we have calculation performed without operator 17 action, and the attendant numbers, string of numbers that we 18 have calculated, so the sensitivity of operator _ error of 1
19 omission or just inability to perform a safety action, it's 20 reflected in our calculations.
j l
21 (Slide) l 22 MR. BASDEKAS:
Moving on here, I think the next J
i 23 slide will show an apportionment of the unreliabilities of the
- ( )
24 reactor trip system.
As you see here, common mode failures 25 dominate the risk contributions.
The reactor trip break is J
HERITAGE REPORTING CORPORATION -- (202)628-4888
31 f-c 1-common cause failure rate ten to the minus 5 and represents 40 hf 2
. percent of the total, and so does the analog channel common 3
mode ~ failure distribution, and only 20 percent is represented' 4
by others, principally independent failures as well as some 5
common cause failures that are rather weakly linked in the 6-
' contributions.
7 The reactor trip and unreliability results-are, for 8
the1 base case, are calculated in the performance slide.
I am 9
just going through the slides quickly just for'the sake of-10 completion.
You have seen them tabulated in the reports 11 before, and I don't want necessarily to be repetitious of what 12 is in the report because you have seen them or-you would have
()
13 seen them I assume.
14 And the contributions, here we are talking, well, 15 the unreliabilities of the base case before everything, any of 16 the options were incorporated there and analyzed, without 17 operator action were five times ten to the minus 5, and with 18 operator action, 2.5 times ten to the minus 5, so the operator 19 action there accounts for a factor of two in the total 20 unreliability of the function.
21 (Slide) 22 MR. BASDEKAS:
The next slide shows the same numbers 23 for each of the options, and I would like to draw your
/^\\
24 attention to the bottom line there, the total, and most L) 25 specifically the delta from the base case, the change from the HERITAGE REPORTING CORPORATION -- (202)628-4888
32
/y 1-base case which these numbers here are, the minus. sign denotes
\\_)'
1 2
a reduction in the unreliability, and you will see here that 3
allEoptions have a minus sign in front of them except for 4
Option No.
5, which have an' increase in risk.
5 Well, that's an eye opener for us because the 6'
situation for that option is that if we replace the 7
undervoltage trip device which had proven to be the most 8
troublesome in the reactor trip breaker failure rate, the 9
studies, if we replace that with a shunt trip device which 10 turned out to be something like ten times more reliable than 11 the undervoltage device, then we felt perhaps we are moving 12 the reactor and therefore we increase the reliability of the 13 system.
14 Well, it turns out by removing the undervoltage 15 device and replacing it with a shunt trip, we were removing an 16 important diversity from the system, so the common cause--
17 CHAIRMAN KERR:
Diversity which you quantified by 18 guess?
19 MR. BASDEKAS:
Pardon?
20 CHAIRMAN KERR:
And a diversity which the 21 contribution you quantified by a guess?
22 MR. BASDEKAS:
No.
Was not quantified by guess.
23 Was quantified by rational experience.
(}
24 CHAIRMAN KERR:
Operational experience in terms of?
{
25 MR. BASDEKAS:
Yes, operational experience, the same HERITAGE REPORTING CORPORATION -- (202)628-4888 m_____----
33 es 1
event and the benefit' factor.
l 1,
~ \\.f 2
CHAIRMAN KERR:
The same event, you were saved by j
3 having a-shunt trip.
4 MR. BASDEKAS:
Well, let me put it in perspective.
5
.There was no shunt trip in the same event because the shunt 6
trips were still late, so it was at undervoltage devices.or 7
except for manual.
8 CHAIRMAN KERR:
I thought you were going to convince 9
me of the shunt trip.
You are convincing me now that the 10 undervoltage trip is unreliable, and I don't have to be 11 convinced of that.
I am already convinced of that.
12 MR. BASDEKAS:
We removed the undervoltage and
()
13 replaced it with the shunt which was more reliable.
j 14 CHAIRMAN KERR:
That makes the whole thing less 4
1 1
15 reliable, and I am saying that conclusion is reached by, based 16 on the contribution that you give'to diversity, and that is 17 unquantifiable.
18 MR. BASDEKAS:
Well, we used persistant sets of data 19 as far as, as far as the benefit factors of the--
20 CHAIRMAN KERR:
Consistency can be both wrong and 21 right.
You know, you could be consistently wrong.
110 22 MR. BASDEKAS:
Right, and we can be consistently I
23 right, which we hope that was the case, and since you brought 24 up the subject here, it may be a good time to ask Dr. Reny
}
25 very shortly to give his brief presentation of our, exactly HERITAGE REPORTING CORPORATION -- (202)628-4888
s'h' 34 f3 1
what data wereLused'for components.
In other words, the
%);
2-undervoltage, the' shunt trip, and the overall breaker 3
. reliability, you know, as well, one by one based on 4
operational experience were. extracted.from whatever source you 5
could get, including the NPRDS data, so let me finish this 6
.here and point out the fact that okay, all options showed a 7
favorable result here except one, and I do appreciate your 8
statement that the benefit factor is something that had some 9
uncertainty associated with it, but to offset it in more 10 concrete and specific terms, I would like to ask at this time 11 Don Reny to give you a brief discussion of what they have done 12 for us in this area, and hopefully we will answer your
()
13 questions in a way that I believe will be an intelligent and 14 defensible conclusion.
15 Go ahead.
16 MR. LIPINSKI:
The data factors,were confirmed 17 statistically?
18 MR. BASDEKAS:
Why don't we let Don--
19 MR. MINNERS:
Let's finish up your presentation, 20 okay?
21 MR. BASDEKAS:
We will make the presentation.
Fine.
22 MR. MINNERS:
Why don't you finish up and then we 23 can do this sidebar stuff at the end?
I l
'()
24 MR. BASDEKAS:
Okay.
The next step in the process 25 applied earlier was to perform core damage frequency results, HERITAGE REPORTING CORPORATION -- (202)628-4888
L 35 1
l
(~y 1
and these analyses included directions of the NTC and the V
2 overall thermal hydraulic behavior of Westinghouse plants, and
'3 the results that we'got here are, do show, the minus sign here 4
denotes an increase in CDF, and you see Option 5 reflects the 5
same type of behavior that we showed earlier in the 6
reliability calculations.
7 (Slide) 8 MR. BASDEKAS:
The consequences--okay.
The 9
cost / benefit evaluation methodology involves specific 10 analysis.
This is basically summarizing these specific steps.
11 The ATWS event. consequences, the core damage 12 analysis, the generic consequences analysis which I referred
~
()
13 to earlier as a set of characteristics for a typical 14 Mid-western site, the proposed analysis for each proposed 15 option, and the proposed options cost / benefit results--the 16 approach here was taken in an analytical sense, it was to vary 17 the reactor trip reliability and cost for each option and hold 18 the rest of the parameters constant, and finally evaluate the 19 option changes from the base case.
20 MR. DAVIS:
I have a question on your consequence 21 analysis.
I realize this is a difficult one, but I noticed 22 that your consequences varied quite a bit and you had to 23 select a number, I think i t was 2.4 ten to the minus 6 or ten
()
24 to the plus 6 person-rem per event or something like that.
25 MR. BASDEKAS:
All right.
Yes.
HERITAGE REPORTING CORPORATION -- (202)628-4888
J 36 l/~y 1
MR. DAVIS:
And'that was based on an analysis done LJ 2
with the CRAC code.
Of course, now the last couple of years 3
that code.has been replaced by the MACKS code which actually-4
-produced higher consequence because of the revised lung dose 5
models.
6 MR. BASDEKAS:
Right.
7 MR. DAVIS:
I am wondering if you looked at, for 8
example, NUREG 1150 results to see if there is a big 9
-difference between what you use and what is now beingL 10 calculated for that accident?
11 MR. BASDEKAS:
Your observations with respect to the 12 CRAC code are correct, and we had our Idaho people follow us
()
13 as close as practically, as practical, what analysis were 14 done, as part of the 1150 work were done.
Would you like to 15 comment some specifically, if it is a quick response?
16
-Otherwise we will cover it later.
17.
MR. RENY:
Yes.
There is a wide spread.
However,.
18 the two low points of the four consequence data points I used.
19 here were taken from NUREG 1150 for the Surry and Sequoyah 20 plants, consequence analysis for ATWS events that led.to core 21 damage, so the NUREG 1150 studies actually showed lower 22 consequence results for ATWS events that led to core damage.
23 MR. DAVIS:
I think those were the previous system
(}
24 results and not the most recent.
Maybe they haven't changed 25 that much, but the reference you gave for those two cases was HERITAGE REPORTING CORPORATION -- (202)628-4888
37 1
1 a report by Benjamin I believe, which is 1986, so it must have 2
been the original NUREG 1150 calculations.
I don't want to 3
belabor it now, but I think the concern is that you can get 4
substantial variation in consequences, and that has a lineer 5
input on your cost / benefit analysis.
6 MR. MINNERS:
Is that something new, Pete?
The last 7
time I talked to the people who I though knew something about 8
variance he said there wasn't that big a difference between 9
the two.
10 MR. BASDEKAS:
In the ATWS at least category, that 11 was our impression.
Now we will check it out.
12 MR. DAVIS:
You could be right.
I have not myself
,s
()
33 seen the NUREG 1150 results for ATWS, for Westinghouse plants.
14 MR. BASDEKAS:
We made the point since a 15 particularly--because of the group that was doing the work for 16 us were sitting next door to the ones working 1150 and the 17 successful thing to do at least keep an eye and talk to one 18 another as to what was going on, and I think what Dr. Reny 19 says basically is that--and what Warren Minners also said, 20 they were not too far off in the ATWS.
21 MR. DAVIS:
My impression is the number you used is 22 possibly conservative, which would--
l 23 MR. BASDEKAS:
As a matter of fact, it is.
There f
(^/\\
24 ara--
i x_
I 25 MR. DAVIS:
Tend to support your overall conclusion.
1 l
l l
I HERITAGE REPORTING CORPORATION -- (202)628-4888
)
m-
38 1
MR. BASDEKAS:
That is exactly right, because we 2
knew over the 25 rem, you know, limitation, placed under the l
3 CRAC calculations, we are mindful of that, but we will make 4
allowances to compensate so that we have, you know, a more l
5 conservative outcome, but your point is well taken.
6 Here we show briefly the cost results.
I'm not sure 7
if we need to go through number by number under the U
option-by-option basis.
9 (Slide) 10 MR. BASDEKAS:
We did get some estimates that we 11 call the best, but there are some things and some, some people 12 can argue that well, it is too high, some people, it is too lk 13 low.
We took the respondents' view into account, and our own 14 independent judgment, and came up with a low, best, and high, 15 and it was one distribution, a triangular, linear type 16 distribution that was used later on in a separate analysis.
17 The next, I think the next few tables are just 18 reductions of what was in the main two documents that we have 19 seen, but I think it will be worthwhile to go selectively over 20 some of them, and what we see on this particular table is the 21 cost / benefit calculations we received both in terms of 22 perron-rem reduction without operator action, and with 23 operator action, so as I iudicated earlier, we made the point (l) 24 to carry this throughout and observe the sensitivity of 25 operator action, you know, as we went down the line basis HERITAGE REPORTING CORPORATION -- (202)628-4888
39 q:
y"j 1
.rather than just do it once and forget about it.
~ (./.
2 MR. DAVIS:
I have a question on that table.
The 3
numbers don't seem to agree with the table on page 12 of your 4'
NUREG 1341 report in the cost / benefit per' person-rem column.
5 MR. BASDEKAS:
With or without operator?
Because 6
there perhaps we give only--I don't remember.
- Okay, e
7 MR. DAVIS:
Neither one of them agree, although the L
.8 with operator SCRAM is much closer.
I don't know if I am L
9 comparing it right or these tables are directly comparable or 10 not, but you might want to check on that.
11 MR. BASDEKAS:
Okay.
Let me make a note and check 12 on this because we have different, we presented the results in
()
13 different ways, and what may be the page number what--you said 14 page what?
15 MR. DAVIS:
Roman Numeral 12.
16 MR. BASDEKAS:
Okay.
17 MR. MINNERS:
Table 10 on page 21 is the same 18
- number, 19 MR. BASDEKAS:
Later on--okay.
=
20 MR. MINNERS:
We just screwed it up in the Executive 21 Summary.
22 MR. EASDEKAS:
Well, I'm not prepared to say.
Let's 23 not~ resolve this right here and now.
If it is a mix-up of
',}
24 tables, we can fix this up, but they came fron the same
-25 source, and as Warren points out, on page 21 and this slide, f
MERITACC REPORTING CORPORATION -- (202)628-4888
40
/"y 1
they are consistent.
We will check it out.
\\_/
l-2 MR. MINNERS:
Hust be a different table.
On page:14 3
they.have got the same numbers.
4 MR. BASDEKAS:
It probably is, but Warren, let me 5
take your advice seriously and say let's go on with the
'6 presentation and we will resolve this later.
7 MR. CARROLL:
Now on the thousand dollars per 8
- man-rem basis, doesn't that suggest Option 3 at least--
12 9
MR. BASDEKAS:
It shows that it would be cost 10 beneficial, but there is a good reason why we decided that it 11 was not.
This was based on point estimates that we' received 12 from a single source, the source being the proprietary of the
()
13 particular option, when performing a sensitivity analysis, 14 receiving independent cost estimates in writing, and we made 15 our own judgment.
16 It turned out both in terms of say new point 17 estimates, more particularly the uncertainty mean numbers we 18 received or we got from the low and high numbers of the range, 19 of course, for a particular option showed it not to be cost 20 effective ultimately, and I believe that we have a discussion, 21 decision rationale for that particular option and point out 22 why we chose a different number-than the one that appears in 23 this particular table..
1
()
24 MR. CARROLL:
I read all that.
Then I see you are 25 presenting this table again, i
HERITAGE _ REPORTING CORPORATION -- (202)628-4888 i
i
I 41
/4 1
MR..BASDEKASi Okay.
As I said, it is a progression 2
~ of things we'have done, and here this table' historically 3
speaking did not include an uncertainty analysis, not yet.
j 4
MR. CARROLL:
All right.
5 (Slide) 6 MR. BASDEKAS:
Now we will come to the uncertainty 7
analysis.
I think the uncertainty here is two basic 8
components that draws the set from--that's the model, the 9
model that we started earlier' used for reliability 10 calculations for the system, and'the other is the ATWS events 11 sequence model that we used as consequence analysis, that we 12 do have some uncertainties.
'( )
13 Then the other group of uncertainties come from the 14 data that we used such as operational experience.
The failure 15 rates were calculated and so forth, so there is uncertainty.
16 Uncertainty was calculated for each option based on the base 17 case risk uncertainty and on the option risk uncertainty.
18 The cost uncertainty steps I said, described earlier 19 from the cost data we have received from various sources, some 20 better than others, of course, and these were evaluated 21 quantitatively, and the cost / benefit uncertainty is the cost 22 uncertainty divided by the risk uncertainty.
K l
2?
Now the data uncertainty, and I don't want to--but L
1
{}
24 it is important to describe, that is the basic process we went 25 through here.
HERITAGE REPORTING CORPORATION -- (202)628-4888 r
42 l /"sj l'
(Slide)
V 2
MR. BASDEKAS:
In calculating the data uncertainty; 3
there were three basic steps, and each of the steps involved a 4-distribution of a given parameter, namely, step A 5
included--and I have an attachment here for'the lack of 6
anything better or anything, Monte Carlo. calculations txt the 7
two mode distributions to come up with a third distribution, 8
and the UNC standards for uncertainty in brackets throughout 9
this particular slide'are intended for whatever operations 10 were necessary in the analysis, statistical analysis, the 11 data--I took them up with it basically, so the uncertainty.of 12 the core. damage frequency was calculated by performing an
()-
13 uncorrelated Monte Carlo random sampling of the distribution 14 for the reactor trip unreliability, and for the initiating 15 event sequences.
16-There was a log normal distribution assumed for the 17 reactor trip unreliability and almost normal distribution for 18 the initiating event sequences resulting in a distribution and 19 attendant uncertainty on the right-hand side of the equation 20 of the core damage frequency.
21 The next step involved a single operation here 22 involving the core damage frequency distribution wh1ch we 23 derive from A, the consequences distribution times the reactor
{}
24 years.
We figure ic out in terms of for all plants, and that 25 was the uncertainty and the risk, and finally, step number C HERITAGE REPORTING CORPORATION -- (202)628-4888
43 y3 1
included a similar operation, but this one here was a
%.)
2 correlated Monte Carlo random sampling because both theLbase.
3 case and the option case risk, associated risks.were not 4
independent.
They were correlated, so that was the basic 5
analytical. process used here to calculate uncertainties of'the 6
various, of the various steps of analysis.
7 (Slide) 8 MR.'BASDEKAS:
Here we show the base, the results of 9
the base case uncertainty results with and without operator 10 action, again emphasizing the system sensitivity in operator 11 errors, and just summarizes the bottom line data in terms of 12 the mean values, the 5th and 95th and 50th percentiles.
These
()
13 were used later as you will see in the cost / benefit 14:
uncertainty analysis, and the.results of that analysis are 15 shown in the next slide.
16 (Slide) 17 MR. BASDEKAS:
This vugraph shows both.with and 18 without operator action on the cost / benefit in terms of 13 19 dollars per person-rem reduction for each option along with 20 the percent probability to be between zero and a thousand 21 dollars and probability of to be more than a thousand dollars, 22 and you can see that we have a wide variation in this, the 23 distribution parameters in here.
24 On the next slide we make comparison of the point Al c
25 estimates and the uncertainty means.
Here you will see we do
44 rs 1-have a--well, okay--here we have point estimate to mean, but
%.)
2
'to take it as the point estimate, the mean is'there just for I 3
guess,- I think.it slipped into our tabulation to indicate that 4
we are going to make the point estimate, you-hope you are 5
close to the mean, but talk basically in one column we have 6
the point estimate, in the other, the mean derived from the 7
distributions, and the statistical analysis you perform on the 8
data data that we analyzed, so in responding to or follow-up 9
on Mr. Carroll's statement, it was this step of the'results-10 that prompted us to make judgments well, on all six, but I 11 believe you referred to Option 3 that exhibited a point 12 estimate favorable cost / benefit ratio, so based on the results
. A(_) '
13' we have seen, both in the statistical analytical sense, but 14 equally important, if not more so, in the judgments we had and 15 we did make, based on incights we have gained during this 16 process of going through step by step and perform these 17 analyses, we have concluded that there were no backfit 18 requirements based on the requirements of the backfit rule or 19 the guidelines, the guidance of the backfit rule, 20 However, as I indicated, a PRA analysis is useful in 21 ways other than just the' ability to come up with a set of 22 numbers, is the opportunity that it offers you to gain 23 insights of how the system works, where the sensitivities are, j
24 and how you can best fix it if you please to get the most i
25 benefit for the money, and based on this collective body of HERITAGE REPORTING CORPORATION -- (202)628-4888
45 1
-work:and' knowledge we derived from that, as I said earlier, we-fs G
2 have come to the conclusion that no backfit requirements are 3'
warranted.
~
i 4
Okay.
That's fine.
And I think we have gone l
5 through some agonizing' discussions--well, not very agonizing, 6
but nonetheless, some, you know, extensive discussions as to 7
why we should accept and what numbers we should accept as 8
valid numbers, and where are those conclusions?
9 However, in the process of doing so, and based upon 10 the insights we have gained, we felt that even though we can 11 not impose regulatory requirements, it would have been 12 wasteful if we did not make our insights available to other
()
13 parts of the Agency, and indeed others for that matter, so 14 this last vugraph here called further work, it is something 15 that goes beyond the, you know, the strict resolution of this 16 issue.
17 This issue we consider resolved which by taking, by 18 proposing no new regulatory requirements.
However, there is 19 some related to this issue, activities going on within the 20 Agency that we have been involved in helping formulate their 21 actions, and/or vice-versa, getting inputs from them, namely 1
22 and most specifically, Mr. Rosa here, and his people, and as j
23 well as another group within NRR dealing with revising i
i 24 technical specification test requirements and this type of
(}
25 thing, and too, the insights that we have gained prompted us l
i i
l HERITAGE REPORTING CORPORATION -- (202)628-4888 q
46 1
and it will bring to the attention and as you probably read in 2
the draft' document that'we are submitting to'EDO and the 3
Director of NRR, we are bringing to their attention the fact 4
that it will be worthwhile to have further activities that we 5
have ongoing within the, within the Agency to consider these 6
things that I list here, namely, the first one I believe that 7
it will be very useful and perhaps productive to consider the, 8
decreasing the reactor trip test frequency in conjunction with 9
the addition of a trip function in the M/G set breaker or 10 breakers as the case may be, either the breaker or the 11 generator or, the output breaker or the generator for the M/G 12 set, and this compensates for the increase in the interval
()
13 between tests of the RTBS and perhaps at the same time, the 14 14 increase of, somewhat the increase of the time out of service i
15 interval because of the industry has been making the point and 1
16 the staff has been presented with the fact that if there is 17 not enough time available for technicians to perform the tests 18 and they are rushing, they are increasing the likelihood of 19 making an error, and in the process, you know, causing a 20 problem.
21 MR. LIPINSKI:
If M/G trip causes a shutdown, you 22 are going to do that once per refueling?
23 MR. BASDEKAS:
As far as the--
()
24 MR. LIPINSKI:
Testing.
25 MR. BASDEKAS:
Not necessarily; you have this HERITAGE REPORTING CORPORATION -- (202)628-4888 I
47
/^g 1
option.
That's one, an option, but not necessarily because as V
2' a matter of fact, after-the main study was completed, we asked 3
Daniel to take a quick look'at this and we have been provided 4
with bypass arrangements.
It will be. costlier, but it can be-5
.done,Tean test it during operation.
l 6
MR. LIPINSKI:
How are you going to bypass- '
)
7 MR. BASDEKAS:
Put in parallel breakers-and 8
parallel--and pretty much the same way you bypass the reactor 9
trip breakers; very comparable arrangement.
i 10 MR. LIPINSKI:
Okay.
11 MR. BASDEKAS:
But it will cost more.
12 CHAIRMAN KERR:
Is that the same, roughly the same
()
13 sort of breaker that was being used in the existing trip 14 circuit?
15 MR. BASDEKAS:
No.
They are different.
Some of 16 them use contractors, and if there isn't a--we have not come 17 out and said here is Option 7 and we have done this and we 18 have given one for a long time that, it was that--would not 19 have at least for the short time that was available to us, and 20 this was--as a matter of fact, during the last meeting we had l
21 to run off the resolution of this issue.
Somebody came from 22 Europe and pointed out that some European plant was using i
l 23 this, this approach to enhance the reliability and also extend
/~T 24 the life of the HTGs.
V 25 CHAIRMAN KERR:
What is the current, site current of l-HERITAGE REPORTING CORPORATION -- (202)628-4888
'48' geg 1
the generator?' Is it',
it is not 600 amps?
%-)
2 MR. BASDEKAS:
No.
That is very small.
3.
MR. WYLIE:
Twenty maybe I guess.
4 CHAIRMAN KERR:
It would.be different?
5 MR. BASDEKAS:
It is a different, all together L
6 different kind and different size.
7' CHAIRMAN KERR:
Different breaker, not necessarily 8
any more reliable, just depending on diversity here?-
9 MR. BASDEKAS:
That is exactly right.
10 MR. LIPINSKI:
11 MR. BASDEKAS:
Yes, sir, for recirculation pumps.
12 MR. LIPINSKI:
They didn't want to. interrupt the
()
13 main current and they said gee, we can interrupt the field 14 current, and that's where they stand.
15 MR. CARROLL:
But in the case of Westinghouse, is 16 this quick enough?
How fast output voltage delay?
17 MR. BASDEKAS:
That's one of things we have not been 18 able to do yet.
We did not go into this because we are 19 getting--we decided to prolong the resolution of this issue 20 and schedule, but they are not sacred, but we are getting to l
21 the point where we will be designing the system for a utility 22 and we decided to, to look at the main pictures if you please 23 or say shortcomings as the case may be of this particular
/~T 24 scheme that at least one European plant has used, and bring it
%)
25 to the attention of our people in NRR and the industry, and we l
l HERITAGE REPORTING CORPORATION -- (202)628-4888 l
l.
.49 f-j-1 thought that would be a. good' example of: industry initiatives
_V 2
to do more of the nitty-gritty type of work that along the-l l
3 lines you are asking your questions,'and to come back with a 4
proposal to the staff and say here'is what we are proposing to 5
do.
6 MR. LIPINSKI:
Westinghouse plants can withstand--
7 so whether it is a minute or not, immediate effect.
8 MR. BASDEKAS:.The time will be short, but I don't 9
know.
10 MR. LIPINSKI:
You would like it to be as fast as 11 your system if you have got a redundant diverse system.
I 12 don't know that we are talking about having it happen in 13 milliseconds.
14 MR. CARROLL:
Well, Westinghouse plants, based on 15 ATWS considerations, can have a delayed SCRAM, but there are 16 other things that we consider, like damage or overpower 17 transients that result in--
18-MR. BASDEKAS:
Mr. Carroll, I think the reason that 19 we felt, the additional reason we felt more comfortable is we 20 received unofficial, informal information, and that's why we 21 are not presenting this part of our resolution package, okay.
22 It is up to--the Europeans have gone through and given, tested 23 the system.
[}
24 MR. WYLIE:
Basically what you are recommending is 25 diverse--let me see if, let me state what I think you are HERITAGE REPORTING CORPORATION -- (202)628-4888
50 L
1 recommending.
r-
- ( f
,~'
! r L '
2
-You are recommending a' diverse system?
~
3 MR. BASDEKAS:
No, we are not recommending.
'15 4
MR. MINNERS:
We don't use the word recommend.
5 MR. BASDEKAS:
Bring it to the attention of-to 6
consider.
7-CHAIRMAN KERR:
Please let Mr._Wylie finish his-8 st'atement before you respond to it.
9 MR. BASDEKAS:
All right, sir.
10 MR. WYLIE:
It says conclusion and recommendations 11 in your--
.12 MR. CARROLL:
As we have the'following (f
13 recommendations.
14 MR. WYLIE:
I assume you are recommending it.
15 MR. KNIEL:
In this slide it says further work.
We 16 neglected to cross out the recommendation in the other.part.
17 MR. WYLIE:
What I am--as I understand it, what you 18 are, you are recommending is a consideration of-a diverse 19 means of removing power to the control rods by the use of 20 existing equipment in the plant.
21 MR. BASDEKAS:
With some modifications.
22 MR. WYLIE:
But it is the existing equipment, 23 breakers or contactors or whatever it happens to be, between
(}
24 the field circuits or output breakers or M/G sets or whatever, 25 but there are ways of doing that also like driving the voltage HERITAGE REPORTING CORPORATION -- (202)628-4888
51 1
regulator to.zero, for example.
s_)
2 MR. BASDEKAS:
I'm not going to dispute there are 3-ways.
4 MR. WYLIE:
Yes, but I mean rather than prescribing 5
a way to do it, it seems like to me you would prescribe the 6
criteria you are trying to achieve, which is as I stated it, 7
diverse means of removing the power.from the control rods 8
using the existing equipment as much as possible and let the 9
applicants or the licensees come in with recommendation in 10 which way they prefer to do it.
11' MR. BASDEKAS:
Absolutely.
12 CHAIRMAN KERR:
This apparently implies that either A1f 13 the number currently used for trip breaker reliability is 14 inadequately large or it is too large, or else the uncertainty 15 is too great or something.
Otherwise you wouldn't need to do 16 anything.
17 MR. BASDEKAS:
Well, the--okay.
I'm sorry.
18 MR. MINNERS:
That is what we are proposing is to do 19 nothing, okay, so you, I don't understand your statement.
We 20 are not proposing to do anything.
21 CHAIRMAN KERR:
It says NRR--and that's not you I 22 realize--NRR ie considered decreasing the test frequency in 23 conjunction with--
0) 24 MR. MINNERS:
Correct.
4
%d 25 CHAIRMAN KERR:
All right.
HERITAGE REPORTING CORPORATION -- (202)628-4888
ff f:-
c.
52.
p l'
1 MR. MINNERS:
Correct.
So if people.want, it is my 2
~ understanding that people, licensees want to come in and 3
increase their test interval, okay, as a compensating feature,.
4 people would say hey, put some more reliability into your 5-system.
,6 CHAIRMAN KERR:
This is just for information?
It.
7 says NRR is considering this?
8 MR. MINNERS:
That's right.
9 CHAIRMAN'KERR:
Sort of an oh,.by the way?
10 MR. MINNERS:
That's right.
11 MR. BASDEKAS:
Additional insights.
12 MR. MINNERS:
That's the message we are trying to y~.
13 give you.
(_j.
14 MR. BASDEKAS:
The bottom line as far as the 15 resolution of this issue goes, you know, we said the 16 conclusion was that no backfit requirements are warranted, and 17 we are away from that.
18 Now having said, that we are sharing some of the 19 insights we have gained because we have seen documents within 20 the Agency in the tech specs as well as the advanced reactor, 21 light water reactor activities that these insights may have a 22' constructive bearing on, and if it weren't, for instance, to 23 prolong the useful life of the RTB if a utility wants to do
(}
24 that and they feel it is desirable to extend the test
{
25 frequency in attempt for that, we have to do something else,
"" "^ " " -- <
2>e28-4888
- - """i"""""
,i-:
W l 53 1,
and we are saying--
2 CHAIRMAN KERR:
Let's suppose that one concludes
'3-that decrease in, increase in test frequency, in testing
/
. interval, made the breakers more reliable.
Would you still 5
.want to do something?
6 MR. BASDEKAS:
I understand, but I.think this is 7
perhaps that--
8
-MR.
MINNERS:
Ask NRR.
9 MR. BASDEKAS:
The NRR people are in the process of 10 developing the bulletin along these lines, and perhaps Mr.
11 Rosa, may wish to--
l 12 CHAIRMAN KERR:
If they are in the process, I'will
'()
13 wait.
14 MR. BASDEKAS:
It is still a pre-decisional insight 15 type of conversation still going on.
16 CHAIRMAN KERR:
It is also true I think that NRR is 17-looking at the question of should testing and power be 18 decreased if feasible?
And I certainly think that's a wise 19 move.
20 MR. BASDEKAS:
I'm sure that's part of it.
21 MR. WYLIE:
What is the test rate now on the 22 breakers?
23 MR. BASDEKAS:
It's alternating every month.
Every
(}
24 two months a breaker is tested.
25 MR. WYLIE:
It is about six times a year?
HERITAGE REPORTING CORPORATION -- (202)628-4888
54
/- -
1 ifR. BASDEKAS:-
A' year.
1 1m 2
MR. OAKES:
Demetrius, maybe I missed it in your I
3 write-up, but have you mentioned anywhere how the Europeans
]
4 make the transition from the analog signals to making the 16 5
interruption of the field coils or the output of the M/G, the 6
parallel with'the current UV cards or put in?
)
7 MR. BASDEKAS:
Undervoltage, it is my understanding 1
8 that they use the same undervoltage drivers to accomplish both t
9 trips if you please in parallel.
10 MR. OAKES:
If that's the case, how does this change 11 the basic reliability problem introduced by the unreliable UV-12 card?
()
13 MR. BASDEKAS:
It will make up--well, well, okay.
14 Speaking of unreliability of UV card, we are saying that it 15 will be prudent to consider it, implementation of Option 1, 16 namely, the installing the Westinghouse cards that will be 17 more reliable.
And secondly, the, primarily thrust of adding 18 of this parallel trip function to the M/G set breaker, okay, 19 is to compensate for what the expectations will be to have a 20 decrease in the reliability of the RTBs by increasing 21 substantially perhaps the test interval.
22 MR. OAKES:
Basically it looks like a good idea to 23 me, and I would think it would go some distance in meeting
{}
24 Walt's earlier comment.
25 MR. BASDEKAS:
That's why we, I said wait until the HERITAGE REPORTING CORPORATION -- (202)628-4888
55-1 end and we can get down tx> some of ' the nitty-gritty 2
discussion if necessary.
3-So basically, with this last slide, which we include 4
our insights beyond the conclusion, which is basically the 5-resolution of the issue of no backfit requirements; basically 6
that's the extent of what we have done, and then some, okay.
7 for resolving Generic issue 115.
8 This is basically the extent of the formal 9
presentation, Dr. Kerr, and there is still some desire to hear 10 the specifics of what we have done in analyzing common mode 11 failures and the like for the reactor trip breakers.
12 CHAIRMAN KERR:
You are going to tell me whether you A
t )
13 looked at a common mode failure or--
s 14 2.
BASDEKAS:
Yes.
Indeed I think that's the, the 15 subject I think that is contained in a brief presentation that 16 we asked Doctor, I meaning Mr. Reny to prepare.
17 CHAIRMAN KERR:
I would certainly be interested in 18 that.
I don't know about my colleagues.
19 MR. MINNERS:
Do you want to have that now?
l 20 CHAIRMAN KERR:
If that is tile only thing lef t, yes.
21 CHAIRMAN KERR:
Let's take about a 15 minute break 22 at this point.
23 (A brief recess was taken.)
(}
24 CHAIRMAN KERR:
Shall we continue?
We lost the 25 principle actor?
)
HERITAGE REPORTING CORPORATION -- (202)628-4888
q i
i 56-
)
i i
vs 1
(There was a brief pause in the proceedings.)
e G.
2 CHAIRMAN KERR:
Whenever you are re ady Mr. Reny.
3-MR. RENY:
Okay.
The first question I think you 4
wanted to have answers on was the reactor trip breaker 5
reliability, and what we did with that in our analysis.
j 6
This is a simplified schematic of the Westinghouse 7
system.
There are two trip breakers that do the job.
Each 8
breaker has an undervoltage device,-and the shunt trip device 9
to actuate them.
These devices work quite differently.
The 10 undervoltage device is a mechanical spring loaded type'of 11 device which is held back with a solenoid type of latch such 12 that the 48 volts trip signal here is holding this device
()
13 open, and when the signal is removed, the latch releases and 14 the spring force causes the breaker to trip.
15 The shunt trip is different in the fact that it's a 16 solenoid actuated device where 125 volts is applied through a 17 relay which causes the shunt trip to actuate and trip the 18 reactor.
19 The relay that actuates the shunt trip device is j
20 powered off of the undervoltage driver cards, the same 48 l
21 volts signal that supplies power to the undervoltage device.
22 (Slide) 23 MR. RENY:
We took a look at the data and we took a
()
24 look at it for the three devices--the undervoltage trip device 25 on the breaker, the shunt trip device, and the actual HERITAGE REPORTING CORPORATION -- (202)628-4888
57 I
l l
gs l'
remainin'g parts of.the breaker, the mechanical actuation u.)
2.
- devices, i
3 The NUREG 1000 study from the Salem events has 26 4
failures to open on demand and 6,000 estimated demands for the 5
undervoltage trip device.
l 6
Now this was actually the entire reactor trip 7
breaker failure data for Westinghouse because the undervoltage 8
trip device at that time was the only device actuating the 9
breaker, and all the failures were attributed to.that device.
10 After the Salem ATWS events, we collected NPRDS data.
-17 5.1
'from 1984 through 1987 and we were able to find eight more 12 events in approximately 4,000 estimated demands of
( )-
13 undervoltage device failures.
Now these were, were actual 14 undervoltage device failures, but were not necessarily reactor 15 trip breaker failures because of the fact that the shunt trip 16 may or may not have been available during these breakers to l
L 17 actuate.
l 18 So if we combine these two sets of data, prior to l
19 1943, and then *84 to
'87, we come up with an undervo]tage 20 device failure rate here of 3.4 times ten to the minus 3.
21 Rather new we took a look at shunt trip devices and prior to 22 the Salem, evait there was not any recorded history on shunt 1
23 trip devices because they were not considered safety devices
()
24 and information was not reported.
25 We did collect it for the same time period that we HERITAGE REPORTING CORPORATION -- (202)628-4888 a
58 l
73 1-did for the UV device, from '84 to '87, and we found two V
2 failures and 4,000 demands, and associated failure rates.
We 3
collected other types of reactor trip breakers not associated 4
with.the UV device or the shunt trip device, and found an 5
additional two failures in 4,000 demands that prevented the 6
breaker from operating, and so we have a failure rate for 7
that.
8 MR. CARROLL:
How would those failure rates change 9
as a result of the improved testing and maintenance that has 10 gone on in the industry since the time of the Salem event?
11 MR. RENY:
Well, in the undervoltage trip device, 12 which was what most of the direction of the testing and
(
13 maintenance was geared for, was for preventing that failure, 14 we see a slight improvement in reliability at that particular 15 device.
16 As far as the shunt trip or mechanical components, 17 we have nothing to, no previous history really to base that on 18 except for the fact that undervoltage trip device was by far I
19 the dominant and probably the only failure mode from breakern 20 such that it f ailed so of ter that maybe it masked other types 21 of failures the breaker might have had if it was able to l
22 opr: rate longer, and we see some mechanical failures down here 23 which were not present in data previous to
'83.
()
24 MR. LIPINSKI:
What were those mechanical failures?
25 MR. RENY:
The McGuire event, which was a shaft weld HERITAGE REPORTING CORPORATION -- (202)628-4888
n '
.]
59 1
failure which caused mechanical binding of the shaft, it was 2
.an event where both the undervoltage and shunt trip devices
')
3
' worked, but the shaft bound up'and it never tripped.
4 MR.~ LIPINSKI:
And the other one?
5 MR. RENY:
The other one was a problem with a,.a-6-
contactor in the trip breaker.itself such that when the 7
breaker is pushed into the cubicle, it makes contact with all I
8 of its electrical signals such that it tells the operators
'l 9
that everything is go,.that it is sitting in position and it 10 should work correctly, whereas in fact there was'a mechanical 11 failure in there with one of the, one of the contacts which 12 prevented the breaker from operating, and was not shown up O
13 enzwhere unt11 ther ectue111 hed e demend eed the breeker 14 didn't work, so this was another type of mechanical failure, 15 but actually not due to the binding or the actual operating 16 mechanism, but due to the breaker itself.
17 MR. LIPINSKI:
How frequently are these breakers 18 reacting?
19 MR. RENY:
Quite frequently.
There are 280 some odd 20 maintenance reports or events in NPRDS which have to do with 21 reactor trip breakers for all PWR plants over a three or 22 four-year period, so any -and that's what hns been reported, 23 so each one of those actions means that breakers react in or 24 out, worked on one way or another.
25 MR. LIPINSKI:
The event you have observed here HERITAGE REPORTING CORPORATION -- (202)628-4888
60 3
1-could be repeated.as these breakers: age?
2 fMR. RENY:
Definitely.
3 HR. WYLIE:
Which event'are you talking about?
4 MR. LIPINSKI:
Business.of racking it out, not 5
having the contacts meet up;after it is racked back into--
6 CHAIRMAN KERR:
That turned out to be a mistake in 7
construction as I remember, isn't that correct?
8 MR. RENY.:
I believe there was a OA problem with the 9
breaker.
10 MR. WYLIE:
You are talking about the two at 11 McGuire?
12 MR. LIPINSKI:
I am talking about the last one he 18 )
(;
13 talked about, racked back in and it didn't make up-.
14 MR. WYLIE:
I don't know about that.
15 CHAIRMAN KERR:
The same one I saw; it turned out 16 that they built something onto the enclosure that they weren't 17 supposed to have built and it was interfering with the breaker 18 activity, and that was corrected.
19 MR, LIPINSKI:
It is not a question of a ro'; tine 20 activity that says do this often enough 9.nd you can expect to l
El see this again?
22 CHAIPMAN KERR:
If you are thinkir g of the same one 23 I am--I think that's the case.
I
(}
24 MR. LIPINSKI:
If it was a design deficiency, then j
25 it was corrected, that's one thing, but if we are seeing an HERITAGE REPORTING CORPORATION -- (202)628-4888
- _ =
- t..
~
61, event that. occurs'from racking inn and out, you can see it 1
2 occur again in the future.
3 MR..WYLIE:
Also the McGuire event as I recall was a 4
quality control problem in the. manufacturing'of the breakers 5
originally, wasn't it, where the shaft.was out of round and 6
caused a binding of the breaker?
That was my recollection.
7 CHAIRMAN KERR:
I'm sure you are right.
I. don't 8
know whether you can avoid those or not.
9 MR. WYLIE:
Well, yes.
Well, well, as I recall the 10 history of the thing, Westinghouse had switched their breaker 11 manufacturing operation from Pittsburgh to Puerto Rico.
They:
12 had a new factory start up.
OA, QC was not that good, and
()
13 they shipped a' lot of breakers that had that problem that were 14 poorly manufactured, and then when they ran into the problem, 15 they then improved all of that, and they are now manufactured l
16 back at Pittsburgh on one of the reactor trip breakers that I l
17 think has taken care of that problem.
18 MR. BASDEKAS:
It was a weld prob 3an 1 think that l
l 19 caused--
20 MR. WYLIE:
Tha enes I saw was the shafts that were j
l F
'il out of round and somebody had taken 6 file to it, that kind of 22 thing.
I 23 MR. RENY:
To continue, so we took the data that we l
~
24 found and plugged it into our model.
25 (Slide) i HERITAGE REPORTING CORPORATION -- (202)628-4888 L
62 4
1_
MR. RENY: ;Our model currently shows the jA 2
undervoltage trip coil and shunt' trip coil in parallel such 3
that either of the. devices can actuate a breaker, and then the 4
mechanical components in series there, these are the failures.
5 that regardless.of the actuation mechanism, still cause 6
failure of the breaker.
7 And we see here we have independent cut sets which 8
are very small, which is current,'which is consistent with the 9
redundant model.
We have common cause failure cut sets and 10 the dominant one here being between the mechanical components, 11 common cause failure of McGuire type event or the other type 12 of event, here one times ten to the minus 5th; prior to I) 13 Generic Letter 83-28, we only had the undervoltage trip coils 14 actuating here on an automatic signal.
The shunt trip was 15 only actuated on a manual signal, so in that situation the 16 undervoltage coils then were the dominant common cause l
17 failure.
18 Now with the shunt trip diversity in thers,
- t. hey are l
19 not the common cause dominant fai]ure.
The mechanical I
20 components are, and our model shows that because you needJa 21 common cause failure of both shunt trip devices and UV coil 22
. devices to have system failure now; the component failure data 23 here from the previous chart, beta factors that were used for' l
\\
()
24 common cause failure.
{
t 25 MR. LIPINSKI:
Where did you get the beta factors?
l f.
I HERITAGE REPORTING CORPORATION -- (202)628-4888
a 63 L
/~
1 MR. RENY:
The-beta factors came primarily from the.
1 2
ATWS rulemaking analysis SECY document.
I can't recall the 3-number, and current PRAs, Seabrook being the latest one that
'4 had Westinghouse reactor trip breaker component failure 5
history with a beta factor.
6 MR. LIPINSKI:
Where is the justifications for the 7
two others?
You are saying you are lifting.it from another 8
report?
j l
9 MR. RENY:
There is:no justification besides the l
10 fact that that was what was derived and used in the ATWS 11 rulemaking.
12 MR. MINNER.S:
Have we ever had a common cause
'( )
13 mechanical failure of breakers?
14 MR. RENY:
No.
15 MR. MINNERS: -We have had what, sir?
We have had 16 6,000 operations?
17 MR. RENY:
Ten thousand or more.
18 MR. MINMERS:
Ten thousand cperations?
19
'tlR. WYLIE:
You know, Warren, that's a good 20 question, but going back to the McGuire ovent now, the McGuire 21 event was reported because the plant was starting up and then-22 operational.
23 Now after that event, and they shipped some of those
(}
24 things back to Pittsburgh, they found the same thing on a lot 25 of breakers and they corrected it, but they didn't get into HERITAGE REPORTING CORPORATION -- (202)628-4888
I' L
64 1F w 1
'NPRDS because it wasn't in an ope 2ating plant, so I mean the 3
v.
2 potential is there, but that again is due to the-manufacturing 3
quality control, which--
4 MR. MINNERS:
But they did catch it before they went l
~ 5
.into the plant.
6 MR. WYLIE:
Well, it was during the investigation of 7
the problem at McGuire that they looked at other breakers that 8
they had on site and' shipped them back up to Pittsburgh, they.
l 9
found the same' thing, so the potential for common cause 1
10 failure was there.
1 11 MR. CARROLL:
This is other DB 50s that weren't 12 used.
p.
(_)
13 MR. WYLIE:
Breakers at McGuire and Catawba hadn't 14 been used yet.
15 MR. CARROLL:
They were spares for reactor trip 16 breakers?
17 MR. WYLIE:
You have got two units.
One unit 18 started up and the other unit was sitting there.
19 MR.'MINNERS:
GE wasn't as lucky when they shifted 20 their operation to Puerto Rico, and glued their breakers 21 together on Montieelio, you know, they had a, well, that plant 22 hadn't started up, either.
That was during testing, but 23 that's what operating experience means.
I mean however you
()
24 catch it, that gets it included in operating experience.
25 Agreed the potential is there, but there are already a lot of HERITAGE REPORTING CORPORATION -- (202)628-4888
f 65 Ag 1
'other~contr'ols and things going on and operating experience
.g-2
~that just people--
3 CHAIRMAN KERR:
And the potential doesn't count.
4 MR. MINNERS:
No, I wouldn't say it doesn't count.
I 5
CHAIRMAN KERR:
It.doesn't count in beta.
Potential 6
doesn't.
p 7
MR. MINNERS:
If it didn't count, we wouldn't have 8:
.02 here.
9 CHATRMAN KERR:
But that is not from beta.
10 MR. MINNERS:
Correct.
11.
MR. WYLIE:
I think the point here, though, is that 12 you found two failures out of so many from NPRDS, but the
-( )
13 thing it didn't show was the number of failures that that same 14 family of breakers had that they caught because of that 15 failure that they did catch in operation.
16 MR. BASDEKAS:
NPRDS were not the only source of 17 data.
We had the same for bulletins going out and this was 18 another source of information that was going to these people 19 to make sure that we are catching all relevant information as 20 far as operations cc.
21 MR. WYLIE:
But I suspect in your data base is net Z2 the failures they caught when they were doing the testing of R23 these Pittsburghs, that they didn't report t heta.
24 MR. BASDEKAS:
That is correct.
For those, that is
{
25 correct.
. HERITAGE REPORTING CORPORATION -- (202)628-4888
66 1
CHAIRMAN KERR:
Why don't you concinue, Mr. Reny?
I 2
think we need a complete session some day on beta factors.
3 MR. RENY:
The current method of deriving beta 4
factors is somewhat uncertain.
It is based on the potential 5
for common cause, and that potential can be taken from a 6
review of operating history failures.
7 For example, in the undervoltage trip device, we 8
have a large number of failures, 30 or 40 different failures 9
to look at, and we do have a common cause failure of the Salem 10 events.
That's a fairly high number.
We have other types of 11 devices like the shunt trip and the mechanical components, 12 breakers that have very few failures to look at, one or two or
()
13 so failures, and the potential has to be derived from looking 14 at that experience here.
15 The current industry accepted method, though, has 16 been to put some type of potential on there for a conservative 17 appraisal or estimation of what the common cause contribution 18 could be.
l l
19 MR. DAVIS:
I would like to say a couple of words 20 about that.
One of the numbers that gets used for beta 21 factors frequently is
.1, and that in fact was used in the 22 Seabrook PRA as their generic beta factor.
23 Now they at the same time used different numbers
(}
24 where they had good data or where they thought they had good 25 data, but their data base was developed by Lowe and Garrick HERITAGE REPORTING COR.PORATION -- (202)628-4888
67 1
and it's proprietary so I don't know that anyone has been able
.,v 2
to really see it.
3 MR. RENY:
I could shed some light on that.
That.1 4
was derived as an average of the beta' factors scaled for 5
mainly mechanical components used in a typical BWR PRA, and 6
the dominant contributors to that beta factor were high
'7 pressure injection pumps, RHR pumps, reactor trip breakers, 8
and mechanical components of that type, so in my view, in this 9
analysis, the
.1 was not applicable to electrical components 10 as such here, but more applicable to large mechanical 11 components, somewhat applicable to electro-mechanical 12 components.
()
13 MR. DAVIS:
Well, my only point was there is a 20 14 reasonable base to argue for a
.1, On the other hand, I'm not 15 disagreeing that.02 is also reasonable.
16 Did you examine in your uncertainty analysis a range 17 of beta factors?
18 MR. RENY:
Yes.
19 MR. DAVIS:
What was that range?
20 MR. RENY:
We looked at a range I believe that went 21 from.1 I believe at the top end to.005 at the low end.
22 MR. DAVIS:
Thank you.
23 MR. RENY:
I'm not exactly sure.
I would have to
(~)N 24 refer back to the report.
25 MR. DAVIS:
That would be a reasonable range.
HERITAGE REPORTING CORPORATION -- (202)628-4888
r i
68 l
1 CHAIRMAN KERR:
Please continue.
2 MR. RENY:
Well, basically that was now we modeled.
3 The conclusion out of this was that prior to Generic Letter 4
83-28, the dominant failure mode was the undervoltage trip 5
device.
A single RTB had a failure rate in the order of-four 6
times ten to the minus 3.
This number here for single RTB-7 prier to Generic Letter'83-28 was reported as.five times ten 8
to the minus 3, and the ATWS study and the Salem study was on 9
the order of 4.6 or 7 times ten to the minus 3 and the 10 Seabrook PRA, so four to five times ten to the minus 3 is 11 about the range for an undervoltage trip device.
12' With that the only device actuating the breaker, we
()
13 had a high common cause of failure of the system due to both
~ 14 RTB failing due to the undervoltage trip device.
After 15 Generic Letter 83-28, the mechanical bind in the shaft 16 mechanism and other types of failures dominate the failure 17 mode.
The single RTB failure rate now has increased by order 18 of magnitude.
By adding the automatic shunt trip device, the 19 common cause failure of both RTBs is now in the order of ten 20 to the minus 5th rather than ten to the minus 4th, 21 MR. LIPINSKI:
Were they both used with the.02?
22 MR. RENY:
This one was used with undervoltage trip 23 device beta factor, and this one was used with the mechanical 24 device beta factor, so this one has the.02 and this one hau
(}
25 the.145.
HERITAGE REPORTING CORPORATION -- (202)628-4888
69 p
1 MR. LIPINSKI:
But in the top one, y6u also had to fg
(,_f I
2 include a common cause for the trip breakers,-too, ist order to 3
do your analysis dominated by the other'one I assume.02 then, 4
and then swamped by the undervoltage.
l S.
MR. RENY:
That is correct.
That would'be included I
l 6
in here except for prior to Generic Letter 83-28 that 7
' mechanical failure mode was not exhibited.
8 CHAIRMAN KERR:
Now if you take off your PRA hat and 9
.put on your, some other hat, you really believe that ten to 10 the minus 5?
11 MR. RENY:
Pardon me?
12 CHAIRMAN KERR:
Do you believe that ten to the minus
'(T
'_/
13 5?
14 MR. RENY:
Ten to the minus 5?
15 CHAIRMAN KERR:
Yes.
16 MR. RENY:
Current state of the knowledge, yes.
17 With the automatic shunt trip device, what I believe is the 18 relative difference between the numbers and not the absolute 19 value.
If you believe that the undervoltage trip device in 20 the breaker prior to the automatic shunt trip device was this 21 number, then I believe that after the automatic shunt trip 22 device, it is this number, 23 MR. LIPINSKI:
That says you have gotten rid of 24 everything and you are just left with the mechanics.
25 MR. RENY:
Basically, yes.
I HERITAGE REPORTING CORPORATION -- (202)628-4888
70 l'
CHAIRMAN KERR:
The reason I asked is because we
-~
2 demand quite a lot of these trip systems.
We demand that they 3
be more reliable than any other system in the reactor.by at
~
4 least an order of magnitude, and yet you tell me you are 5
unwilling to comment on what you think the absolute number is 6
or that's the number you used to arrive at these conclusions.
7 I am not being critical of you.
I am simply saying 8
that number sticks out it seems to me if you look'at risk 9
contributors to the reactor operations, not that it is 10 necessarily wrong, but it places a very high reliability on.a 11 fairly important system, much higher than any other important 12 system I know of.
- ( )'
13 MR. RENY:
I agree, and there is some uncertainty 14 included within all of these numbers.
These are the point-15 estimate numbers of which there are uncertainty. bounds.
I 16 CHAIRMAN KERR:
And I say this in context which I 17 believe that something could be done about this in the next 18 generation of reactors.
I think you can design them so you 1
19 don't have to depend on this system, but that's a side issue 20 to this discussion.
21 MR. LIPINSKI:
Take the mechanistic approach, 22 failure of the single active component, calculate what you 23 have left, it would be single breaker, forgetting about l
/~T 24 reliability.
O 25 CHAIRMAN KERR:
I would hope that we abandon the HERITAGE REPORTING CORPORATION -- (202)628-4888
71-f--y 1
single--okay.
.Q) 2
. Continue, please.
3 MR. RENY:
Well, this was all I had to present on 4
reactor trip breakers.
5 CHAIRMAN KERR:
That's interesting.
I'm glad to see 1
6 what is there.
I asked if somebody would comment'on whether 7
we looked at what I call at least the common mode failure such 8
as overtemperature in these solid state devices causing the l
1 9
failure of air conditioning systems.
1 10 MR. RENY:
We applied a beta factor'for common cause
.)
11 failure across the logic components, and within the 12 application of that beta factor would be included' common cause
()
13 effects such as common air conditioning. failures.and-other 14 type of environmental impacts.
15 A beta factor for a common cause includes all of the 16 postulated common effects that could affect both of those 1
17 components, so within that application, yes, the beta factor 18 does account for possible environmental conditions affecting 19 common cause failure, both logic components.
20 CHAIRMAN KERR:
I assume you would give me the same 21 answer if I asked you lightning and electrical surges?
22 MR. RENY:
That is correct.
23 CHAIRMAN KERR:
You wouldn't feel good about that
(}
24 answer, though, would you?
l 25 MR. RENY:
No.
A beta factor is a catchall.
l HERITAGE REPORTING CORPORATION -- (202)628-4888 1
72
'l CHAIRMAN KERR:
Has there been'any serious work on
. r -).
'Q
'2 the affect of lightning on'some of the solid state components 3
generally, not restricted to this?
I ask.out of ignorance.
4 Has the staff--
5 MR. ROSA:
This is Faust Rosa again.
6 CHAIRMAN KERR:
I don't mean. lightning direct 7
strikes, but lightninginduced transients.
8 MR. ROSA:
The.best study we have is the EMP study, 9
and that indicated that lightning, well, that lightning that 10 strikes out in the switch yard or on the distribution system 11 from a switch yard to the first transformer is.less severe 12 from the standpoint of the transient input into the plant than
()
13 an'EMP pulse, and therefore, the EMP study indicated that we 14 had nothing to fear from the EMP transmitted into the plant on 15 the power system.
16 Now those failures of instrumentation channels that 17 have occurred due to lightning strikes were primarily due to 18 strikes on a containment building which somehow or other 19 reached the penetration areas and directly impinged on the 20 circuits in those instruments.
21 CHAIRMAN KERR:
So they don't count?
22 MR. ROSA:
They count yes, sir, but every time it 23 happens, the licensees are required to upgrade their lightning l
l
.( )
24 protection system on that particular building, and that's it.
25 MR. RENY:
I might add that the logic channels are l
HERITAGE REPORTING (202)628-4888 L
___-___ _ __ ______----__-__ CORPORATION --- -
73 f3 1
isolated from the analog instrumentation such that that pulse V
2 could not be propagated through.
3 CHAIRMAN KERR:
It is the position of the staff then.
4 that the solid state circuitry, generally solid state-5-
components are no more vulnerable to-lightning-induced pulses, 6
analog instrumentation or whatever?
7' MR. ROSA:
Given the surge protection devices that 8
are in series with the power circuits yes.
9 CHAIRMAN KERR:
Are there further questions?
10 MR. CARROLL:
What has been the utilities' 11-experience in using solid state transmission protection 12 devices as far as lightning is concerned?
()
13 Has that been generally favorable, or has that been 14 looked at in terms of lightning?
15 MR. BASDEKAS:
I don't believe we have an answer for 16 that, Mr. Carroll, but we will try to get you one.
17 As Mr. Rosa outlined earlier, there is, there are 18 requirements for such protection throughout the plant.
I do 19 recall there,ere some operational experiences involved in the i
20 use of walkies-talkies in control room proximity, and those 21 problems were resolved many years ago as part of the FFTF l
22 start-up experience in other plants I'm sure, but nonetheless, 23 I think these are very good, you know, technical problems to
'( )
24 work on, but as far as the scope of this particular issue, as 25 I was careful to point out at the beginning, it was rather 1
HERITAGE REPORTING CORPORATION -- (202)628-4888
-=-.
74 1
limited-to. address 70 percent that stemmed from the Salem
.<:)
2 events and subsequent operational experience with the drivers, 2
1 3-but nonetheless, your concern is well taken.
4 CHAIRMAN KERR:
Further questions?-
5 MR. CARROLL:
Just for my edification Bill--I don't 6
-know if.others want to hear it or not, but I guess-I would 7
like'to understand a little better the situation on the.
8 Combustion and B&W plants, just for background purposes.-
9 CHAIRMAN KERR:
Insofar as the solid state device is 10 concerned?
They don't use'them, do they?
11 MR. BASDEKAS:
I'm_sure they use them, yes.
I 12 MR. WYLIE:
I think B&W originally used' extensive j
()
13 solid state logic in'the protection systems.
14 CHAIRMAN KERR:
Did they use this kind _of card?
15 MR. WYLIE:
I don't know about this specific card.
16 CHAIRMAN KERR:
This is a rather specialized--
17 MR. BASDEKAS:
Westinghouse plants specifically I 18 think; as far as Westinghouse plants go, but to the best of my 19 knowledge--
20 MR. CARROLL:
You are happy with what both 21 Combustion and B&W have in their circuitries similar to this?
22 MR. BASDEKAS:
No, we are not saying that, Mr.
23 Carroll.
Personally I cannot voice an opinion on the relative
()
24 merits of other designs.
Perhaps someone else here might shed 25 some light, but this issue was oriented specifically toward
. HERITAGE REPORTING CORPORATION -- (202)628-4888
[.
75 L
1 Westinghouse plants.
73
()
2 MR. DAVIS:
I think the other designs with, by ATWS 3
rule to have a diverse mechanism--
4 MR. BASDEKAS:
For installing a totally' diverse 5
parallel system and that will take away a lot of the--
6 MR. DAVIS:
Westinghouse was exempted from that 7
requirement on the basis of--
8 MR. BASDEKAS:
Yes.
9 MR. CARROLL:
Exempted on the basis?
10 MR. DAVIS:
On the basis of their enhancability to 11 ride through it like a SCRAM and installation of the automatic 12.
feedwater--
/~'s
(,)
13 MR. BASDEKAS:
And turbine trip.
14 MR. WYLIE:
B&W plants originally-had a driver trip
)
15 system as their original design.
16 MR. ROSA:
That's right.
17 MR. LIPINSKI:
The Combustion Engineering plant at 18 Arkansas has the protection system and it is very similar and 19 the equipment was in the same room, had an air conditioner 20 oversize link, were getting spurious trips.
21 MR. BASDEKAS:
Some events as recent as last month 22 or so continue to exist for that type of system.
23 CHAIRMAN KERR:
Further questions or comments?
Mr.
f}
24 Reny, did I understand correctly that the staff did look at 25 the possibility of replacing or that you did perhaps, the HERITAGE REPORTING CORPORATION -- (202)628-4888
f..
76
.(g 1
undervoltage coil and another shunt coil and have concluded-(/
l-2 that because of common mode failures, that that'would actually 3
decrease reliability?
I' 4
MR. RENY:
That is correct, and I wanted to address 5
that.
The model for that option looked like this except there J
6 was a shunt trip here and a shunt trip here, double redundant i
7 shunt trip devices on each breaker.
_j 8
Now if you follow the beta factor methodology, you 9
would have to assume a common cause failure mode between those-10 four shunt trip devices where, whereas with an undervoltage l
11 and a shunt trip coil those devices, being diverse, you would 12 assume no common cause failure modes between those two.
()
13 Therefore, the assumption of the common cause failure mode 14 between like devices there vould have an additional 15 contribution to common cause failure over here, which made it 16 a slight increase in unreliability.
17 CHAIRMAN KERR:
I just wanted to to make sure I 18 understood the conclusion and the basis for it.
19 MR. MINNERS:
Do you use the.145?
I 20 MR. RENY:
For shunt trip?
We used the.065 for l
21 shunt trip.
However, that number was used for the likelihood j
22 of failure of two shunt trips.
We used the number slightly 1
1 23 smaller by a factor of ten for the likelihood of failure of
( }-
24 three or more, four shunt trip devices, so it followed what is 25 called the multiple Greek letter method, which is t.n extension HERITAGE REPORTING CORPORATION -- (202)628-4888 l
___A
l 77 1
of beta factor to higher redundancy systems.
2 CHAIRMAN KERR:
Suppose instead-of-using it to
-3 control the reactor, you were using it say to control'your.
4 furnace-at home?
Which'of:these two systems would you refer?
5 The one with four shunt trips, or one with two undervoltage 6
trips and two shunt trips?
7 MR. RENY:
I would prefer the diverse system, this 8
one undervoltage and shunt.
9 CHAIRMAN KERR:
You are consistent anyway.
10 MR. RENY:
I think the diversity has a lot of merit 11 as far as preventing common cause failure, failure mechanisms,
,l 12 and at this point in time, that seems to be the extent of the n(j 13 defense for common cause is its diversity.
14 CHAIRMAN KERR:
I agree with you wholeheartedly.
15 The diversity can prevent certain kinds of common cause 16 failures, but my objective is reliability rather than 3
17 diversity, and I'm not sure that they increase the 18 reliability, but maybe it does.
19 Any further questions?
Well, that concludes the 20 staff's presentation.
I want some comments from some of you, 21 but I'm sorry that we took up most of Mr. Kniel's and his l
22 colleagues this afternoon, but it has at least been 23 educational for us, and perhaps we have a better understanding i
24 of the logic.
25 Personally I have no quarrel with the decision that HERITAGE REPORTING CORPORATION -- (202)628-4888 a
78-l1 was reached-on_this rather narrowly defined generic issue on b
l I
.2 the basis of-wh'at occurred, but--an'd I think that's enough 3
' recording.
Thank you ma'am.
'4 MR.' WYLIE:
Someone-mentioned that'the staff was 1
5 prepared to tell us something about the recommendation-I i
6 regarding some. testing of the reactor trip breakers?
~
7 Somebody?
Anybody in'Research or something?
8 MR. ROSA:
We had a research review group meeting 9
'last, two weeks ago during which part'of the--well, on aging.
10 It was a three-day meeting, and one of the presentations 11 described the life testing that was being performed on some-12 SCRAM breakers.
They were Westinghouse DS type breakers.
I l
13 forget exactly what the number, but that's intended to, to 14 arrive at a, at a life test based-on how many cycles the 15 breakers can withstand, given periodic mechanical maintenance.
16 That test will not,.will not~take into account the 17 undervoltage trip attachment or the shunt trip attachment 18 reliability.
19 MR. WYLIE:
What is the purpose of the test?
20 MR. ROSA:
Just to arrive at a, a number of life 21 cycles for the mechanical portions of the breaker.
22 MR. CARROLL:
So you are simulating the actual 23 surveillance test on the breaker where you are not
)
24 interrupting current?
25 MR. ROSA:
That's essentially it, yes.
Current I
l HERITAGE REPORTING CORPORATION -- (202)628-4888
]
'79
.i
(~wl 1
interruption capability has never been a problem on reactor i
V 2
trip breakers to my knowledge.
3 MR._WYLIE:
These breakers.
4 MR. ROSA:
The actual contact.
5 MR. WYLIE: -It is a very large breaker current-wise.
6 As I recall, 1200 amps,.something like that; only seeing what, 7
less than six.I guess?
8 MR. ROSA:
Yes.
9 MR. LIPINSKI:
Is this a single breaker?
10 MR. ROSA:
I believe they are testing three 11 breakers, two or three types.
12 MR. LIPINSKI:
And they came out of the shop, the
.' )
13 best QA?
14 MR. BASDEKAS:
No, no.
They had some already for 15 their own use at laboratory and pulled them out, spares or 16 whatever, and tested them and DB 50s DS 3416s and otherwise.
17 MR. LIPINSKI:
Random sample; it is not the best.
18 MR. WYLIE:
It is interesting they are testing only 19 the breakers itself rather than undervoltage device which has 20 been really the culprit.
21 MR. ROSA:
Well, I believe it is considered that the 22 undervoltage device unreliability has been pretty well 23 established.
()
24 MR. WYLIE:
I don't know whether it has with age or 25 not.
It may get better with age.
I don't know.
-HERITAGE REPORTING CORPORATION -- (202)628-4888
80~
l'.
l
/g_
1 MR. CARROLL-Make it smarter in terms of---
.Q) 2 MR. LIPINSKI:
Whether they are physically tripping l
3 the breaker; they have got to get in there somehow.
4 MR. POSA:
I believe'they have rigged up some sort 5
of electro-mechanical trip device;'probably just a shunt trip.
6 MR. LIPINSKI:
Something special; isn't normal' 7
tripping.
8 MR. WYLIE:
They were using the shunt trip.
That 9
ought to give you some better indication of the reliability of 10 the shunt trip, which I would suspect is much better than the 11 numbers we have seen here frankly because when this problem 12 came up with the reactor trip breakers, I had never heard of
(')
13 these type of breakers ever failing to trip.
14 MR. ROSA:
The age research program is rather 15 extensive.
Each element of it is going to produce a separate I
16 NUREG report so we will have all of that data when they get 17 finished.
18 MR. CARROLL:
DB 50s have a bad name in fossil 19 plants.
20 MR. WYLIE:
I guess depending how they maintain--
21 MR. OAKES:
What is the environment?
22 MR. ROSA:
Oh, probably just a shop environment.
I 23 don't know any of the details.
They do intend to perform the 24 periodic maintenance that they expect a good utility would
{}
25 perform while they are doing this.
HERITAGE F.EPORTING CORPORATION -- (202)628-4888
81 ei 1
MR. CARROLL:
How do you define that?
We1are going a
.U 2
to talk about the maintenance rule tomorrow.
~
3 MR. ROSA:
I don't know.
We will have to wait for 4
that research report.
5 MR. OAKES:
What is the definition of a failure?'
6 The time response becomes too long, or has that been 7'
established?
8 MR. ROSA:
The presentation did cover that.
I would 4'
9 expect just an immediate opening would be all that's required.
10 You don't expect any, any delay in the actuation.
It is like 11 you get with the undervoltage trip attachment when it gets 12 hung up.
I couldn't give you the details on that.
A()
13 MR. OAKES:
Are they expecting to do any incipient 14 failure diagnosis on the test?
15 MR. ROSA:
I don't know.
The periodic examinations 16 that will be. conducted during the maintenance periods during 17 the tests will look for incipient failures like cracking welds 18 or cracked welds or things that nature.
19-MR. WYLIE:
While we are talking about breakers, I 20 talked to the maintenance people at Oconee and they had 21 problems with those.
Now those are GEA 25 breakers, which is 22 smaller breakers than these DB 50s, but they had considerable 23 problems with them, and their problem was just the breaker 24 itself.
The frame was not substantial enough.
When you took
{}
25 it out and you calibrated it on the test bed, that would test HERITAGE REPORTING CORPORATION -- (202)628-4888
82 l'
perfectly.
Then you'put it back into the breaker rack, it l
~~
2 would just warp enough_it would change.the setting, and they.
l 1
-3 had--I' don't know whether they still have a problem.
They had
~
I 4
extensive problems with those breakers because of that.
5 MR. ROSA:
I might add'one other bit:of information.
6 You know, NRR contacted Westinghouse management some months 7
ago about GA problems in their,.in their breaker manufacturing 8
facilities, and Westinghouse came in and made a presentation 9
and they described how breakers coming off the, their regular 10 production line in Puerto Rico or wherever before being sold 11 oto a nuclear utility for application in the safety-related 12 circuit go through a special inspection and test at
()
13 Westinghouse right on the outskirts of Pittsburgh, so to that 14 extent, that QA problems for safety-related breakers are being 115 corrected by Westinghouse.
16 MR. CARROLL:
There is a potential gap in all of 17 that.
Does the utility industry know that there are two kinds 18 of DB 50s?
The one that has got special nuclear treatment and l
l 19 the ones that they might have bought for a fossil plant and i
20 have urgent need at a nuclear plant and said a DB 50 is a DB l
21 50?
22 MR. WYLIE:
That has an N stamp on it.
l l
23 MR. CARROLL:
Not necessarily from what he 1
24 described.
}
25 MR. MINNERS:
I don't think that--
"" "^ * " --
<2
>e28-4888 E-
E r
'83-f- -
1-CHAIRMAN KERR:
Any further questions relating to
-2 the issue at hand?
3 We have scheduled an hour for discussion of'this 4
issue at the Full Committee meeting.
Do we need to ask the 5
staff for a presentation or can the Subcommittee handle tbe 6
discussion?
7 MR. WYLIE:
I would recommend the staff might have a 8
presentation,-45. minutes or whatever.
9 CHAIRMAN KERR:
Is the staff willing to do that?
10 Okay.
I think with not a great deal of condensation if those 11 present today will keep quiet at the Full Committee meeting, 12 you can cover most of the material that you covered in your
(
13 presentation.
I cant promise that they will.
14 MR. CARROLL:
I wish Lewis had been here before you 15 made that assessment.
16-CHAIRMAN KERR:
Lewis will have different questions.
17 They won't be the enes that they raised'today.
Do we need to 18 cover anything else?
Okay.
19 MR. BASDEKAS:
For the purchase of the presentation 20 to the Full Committee meeting, you normally allow 50 percent 21 of the time for questions and discussion and 50 percent of the 22 time for presentation, j
l 23 Would it be then appropriate to gauge say half an 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> presentation and half an hour discussion or thereabouts?
i 25 CHAIRMAN KERR:
I would say about 35 minutes of 4
l-HERITAGE REPORTING CORPORATION -- (202)628-4888 L--
84 1
presentation, and hope.
<-)-
%.)
2 MR. BASDEKAS:
We may have to condense some of the 3
slides that we have.
4 CHAIRMAN KERR:
I think so.
5 MR. BASDEKAS:
Thanks.
l 6
CHAIRMAN KERR:
Again, we thank you for coming, and 7
for the presentation, 8
MR. BASDEKAS:
Thank you.
9' CHAIRMAN KERR:
No more recording.
10 (Whereupon, at 3:30 p.m.,
the recorded portion of 11 the meeting was adjourned.)
12
()
13 14 15 16 17-18 19 20 21 22 23
,}
24 25 HERITAGE REPORTING CORPORATION -- (202)628-4888
i 1
CERTIFICATE L
j 2
L 3
This is to certify that.the attached proceedings'before the 4
United States Nuclear Regulatory Commission in the matter..
5 of:
ADVISORY COMMITTEE ON REACTOR SAFEGUARDS' E
6
.Name:
SUBCOMMITTEE ON INSTRUMENTATION AND CONTROL SYSTEMS-7 8-Docket Number:
9 Place:~
Bethesda, Maryland 10 Date:
April 5, 1989 11 were held as herein appears,1 and that this is the original 12
. transcript thereof for the file of the United States Nuclear 13, Regulatory Commission taken stenographically by me and, I
14 thereafter reduced to typewriting by me oriunder the 15.
direction'of the court reporting. company, and that the 16 transcript is a true and accurate record.of:the' foregoing-17 proceedings.
18
/s/
^
19 (Signature typed) :
CATHERINE S. BOYD 20 Official Reporter 21 Heritage Reporting Corporation 22 23 24 25 Heritage Reporting Corporation (202) 628-4888 1
J
!l
!li L
/
O S
M M
5 E
EE 1
T ET 1
S TS E
Y TY U
S HC S
SF N I
I CO MON C T C YIT NITS E
OROS RLE AUE I
I RLR CT D
EI T SBO E
BNE R N B Y
O A
S EA R KME R UOTA GL P EER O I
SCTU E
DT T
I F
S S ER R E9 EDM G O
T89 A Y UA E
E A 1,
B S SL H NMF N
SU HT
.TI O
T A O A NOTS 5 Y
G L
B LNYE S
OI O NC UOI R O LE TFDI SAT R T OY R L
LP PFR I
NI R O O TOA R
T EN O AO C RMS MA O C E E F L E
ITTS A U
E R EU D O NN AN VE I
D C E
O TIO TE DR S N H CSO F
NMA O A G AIVE E
E S U PH N RDI C I
ONI ER RET FF R T P
O S
P S E
E W
N H
AI T
E N
HT O
O
~
o L
L A
8A S
I 2
T E
T N
3 N
C N
N 8G E
N E
O RS M
A T
I I
E E
H O
P P
T TI M V N
P TR O
EE R
E TI ETL R
OAE I
L R
T S
I S
R A RI NDD E
PD E
NM OI N
L D
EOWN TO A I
E GTO M P S )S 5
UH O
D ASM WP o
1 N
S 1
N O
E ANWC A H (S I
E E
E T TM N U
SVN TI R
R O
I E F E
S NG SW C OTT I
S LD SC EVI RD FY Y U I
CAN D
I C
S CA NI NE L
D WP I
RF O B OR R
I AI N
TR E F
ITI TK E
U AT V A E
O MTIRST V
U R R
E I
E I
E R
N E
D T
L T
G G
LU C
C AE OR K
AHVR E
VHRO C
SSUN J
ETPF o
A B
B ee*e O
e ll lll
H P
C MR A
I E
S T
E R
A I
S N
V U
R A
U O
H N
F O
E H
T D
N O H V
C I
E N
O W
I I
TG R
D A N D
M A
IT)
B CI C
AL T
E T
E N
IF S G
O G
UI R
I T
O I
A O
T E
D E I
A OWT T
C C L
N A
AV O
L U
MY O
U P
P U F
I L
D V
T R
B I
O A
RD R
C T
RG O
E A
TN T
AE S
V N
CD D
B P
TC N
I E R E
O N
N T
S NA R E U
R S N UL MO I
T T
E S
P V ME E
SI S E C
T L (R A A N
R R
I I
R R
R T
C L
C D O E
E E
A B P N O
S C
V V
VE I
I I
I T
D UR D
D DF DR RC P
O O
N N
1 2
3 4
5 6
O IT O
P O
S S
N I
S E
Y O
I S
L T
A A
A C
N D
E A
N S
Y E
A H
Y T
M B
I C
G V
M
(
I A
O Y
T O
L T
S S
C I
O O
E I
I L
R D
S N
R I
O P
O B
Y E
/
L S
E P
H A
A L
I D
A A
T L
N E
E N
N A
C M
E R
S A
O I
I S
T P )S C
S Y
T I
I A
F RN N
Y T
A I
L N
R B
E TO E
A I
N U.
A N
I E
RT N
B OP O.
T O
A R
TO E
I
/T C
S T
E S
S A D N
S C
IV O
EN O
O N
I C
RA C
C U
D O
e e
e e
e e
SN S
S I
E O
D N
S IT U
O I
P L
IT P
Y O
C R
R R
I T
N E
E T
I 6
I L
W S
D N
I O
N R
B N
O O
I A
AI P E E
TP L
DR T I
C C
L E
AI B
OU R
A E
S R
RA RL I
R E
O AU T TL I
P C
G T CI LAR A
OF N
E F
I I
R E
F LV R OI U
T S
N EA T
H N (N )N A
O S L E R
B C
L O
O L
R O
C A A I
C UT D
EI T
D FU D E
ST D S S B E C
A A CM MC MER T E
U R
L EO UR UCT A I
AST SU SCNT V
BA AS ASCI O
e G
e
O 1
P
~
IRS TR
'A E
R K
K O A B
C T
T E
C R
R O
A B L
E R
B E S Y
RG R T
EAE A
DTV I
LI V
L N OR U
I U V D BM A
R A
ON I
L T O LR A
A O
E I
G SC A
R T RI E C R
A EG PA L
VO O
P L
U I
I D
N I
U RT R
S T Y
'A O
U A P L R
T NE C
R I
I A
~
ER N
O L
L L
GUI E
E E
T N
N N
OR A LT N'
Nl Nl l
l AS T O
N N N A
A AI E
H H
H I
A M
C C
C
/5 E5 2
=
H E
S P T R A I
O C IRW F
E MS T SN UA E
S R
RTO F
U C
L OLI TU T S
G C IA R
OL F
CS C LE%
E6 AN0 E
AEA H%
4 S
TE0 NN U
O 0. 2 AA A
E R R L 5 H
C RYO L
C A
N E TT O
SLA M
I FC5 M
AI R C%
O B E B00 C
E C AP 4
T
=
1 R
F EILO C
S E C
AR B NU
R ON 5
7 5
7 8 6
5 HT O TAI E
E E
E E
E E
R T 0
3 0
3 5
1 5
I WE C 1
7 1
4 7
4 2
PA Y
O T
I T
R L
U)
U ON T
5 5
5 6 6 5
5 I
T O OD O AI E
E E
E E
E E
B HN H R T 0
0 0
1 1
3 0
A.
TA T E C 1
1 1
6 1
1 5
I L
WM WPA I
I O
E E
DD RS N
S L
R L
A NT AE E
T P
N O
L H
N T
U O
U TS A
)s I
E F H S
PS WR S
C C D
E Y U R
C R
I N
A
(
R R T L O
O C
S S
S T
E T
D R
R I
I T
LA I
U R
A R
E E
B (F B
U T
I L
N A
K V
R A N R
A E
C A
S I
I I
R O
LO T
F MC R
D R
E E
I I
B E
U T
EI N
E U G
R T S
R L
O P G S C
C O
I U
T L
A Y A
I DA C
A S
R T
A F
A E
E C N L
T L
L A
R R
O E
I E
TR S
N G S
E R O V
R AO A
O O T
R H
R LT C
ML V
C E T T
E A
A D
U O
UA M
I P
N N E N
L CR E
N L
O A
U R
U LE S
C A
I AP A
CO B
O e
E E E E E E
E 6
0 0 3 3 5 8 5
0
)
N 1
7 4 7 3 1
1 S
I O
T 7 5 7 8 6 5
6 T
5 C
E E E E E E E
E 5
L A
0 3 1
3 5 5
9 0
1 7
1 4 7 8
2 4
U R
+
O T
5 5
8 6
5 6
S A
4 E 0 E 0 E E
E E
R 0
~
0
~
5 0
3 0
E E
1 1
7 3
2 2
P R
O 5 7 7 8 6 5
5 H
E E E E E E
E 3
0 T
0 3 3 5 5 5
1 Y
W I
1 7
4 7 4 1
1 T
(
5 7 5 8 6 5
6 S
E E E E E E
E I
N 2
0 L
O 0 3 0 5 0 4
0 1
7 1
7 4 2
1 I
I T
B )D P
O 5 7 5 8 8 6
5 7
E E E E E E
E E
1 A'
0 3 0 6 5 1
5 4
1 7
1 8 7 4
2 5
T O
ILN SL E
L E
N A
O S
T N
R N
O A
C
)s T
F H
S O
C C
D P(
I C
R T
N A
(
I U
S O S S C R
I E
T D R R E
B R
A R E E
S T
U T
I R
L N A K V
A T
A E C A S
C I
I E
R E
F M C R D E
R N
R S
I B
E U
A E
U G O
O S
R O P G L
S B
U T
I L
A Y A
I C
A S
R T A F M 5
)
T C
N L T L R
O E A
L S
R O E R 5 I
C N
O O
T R H
N G S E
R O V R F 2 (
A V C E T T
A' E
A O
M L
A D U O M
I P
I N N E N
L E
T O
A U R U
N L
I P
C A
R O
o
N 6
4 4
6 O
E E
E E
S I
1 8
9 1
2 T
1 P
8 1
2 2
T O
F E
R L
5
(
7 R
N 5
4 4
U E
A
)
O E
E E
R 0
E T
1 1
9 I
9 Y
A P
1 1
5 S
E O
R O
Y T
E 4
C R
N 4
4 4
7 A
R O
O E
E E
E E
I R
T T
8 1
6 0
/
P 4
5 4
4 S
C O
T Y
A N
E C
E 3
V R
N 4
4 4
6 E
/
O E
E E
E 5
N I
S T
5 4
9 1
E P
7 2
2 2
0 T
E N
O 1
S E
2 I
U V
N 4
4 6
7 RC E
O E
E E
E N
O.
T 9
0 8
0 I
(
E a
P 7
2 4
2 O
E S
O TH E
Y R
C N
4 4
6 7
I 1
B D
F N
O E
E E
E E
E T
3 6
9 0
H I
P 8
1 4
1 S
U O
IL E
Q B
E E
A G
S T
R A
4 6
ES F
C A
E E
L E
9 0
A E
S 9
5 O
G A
M G
B A
S M
N E
E W
A A
O G
G T
I A
A A
T D
C M
N M
E D
E AM A
O A
H M
I D
T D
T RA C A R
A R E
N OR E
E O
TC R
C R
O AS O
RS O
S C
C Y O
C Y I
R R Y C
Y C
R T L E L NN AL NN A1
)
P L O
E E
P4 I
I OA R A 3
E U E U E U M1 T U GQ P
GQ UN NE O
NE OG N
AR O
R HM HF HM HF R U T
C T
C O
O O
FN I
IWT WT
S D
E S
N H
I N
S I
O Y
YF LT I
I L
Y ONM S
I LE T
A Y
T A
N E L AN HAO I
L TR R
ANE A
D I
UY O
NAB B
SF Y
/
N N LG T (C ATT A
S A O A AO I
SS L
ILE E
VL C
C OO EN C T I
O EO N
N CC ROS L B
E A
I T D ESESS PT R D P E I
I LUI UNN R
T N S
O EQ Y QOO T O IF E
R O
EH ELEI TT R HMI I
PSAS PP O CA T NT SI N
P E E ERTN OO TAR BM STNA O E) CDD CEA OE EE AR PES
/
Y R VE C SS TA L
S E OF T
A OEGI T
R AC S
- YT TL S O
A AWM NOOS H RS SA A C
CETAERRE C AOEV RADGPPR A VCRE B IF O
IC*e ee*
Re E
P O
P P
S A
H 6
8 9
6 5
1 6
9 7
9 7
3 G
1 2
3 2,
4 5
I H
2
)K S
P T
0 1
2 4
1 3
T S
5 8
3 8
0 4
0, 2
2 L
T E
1 N
B 1
U A
S L
O E
P R
R W3 E
T P
3 2
5 9
0 S
O 3
4 7
9 0
3 7
1 1
T L
O S
C OC N
O I
1 2
3 4
5 6
TP O
O
F E O N E R DI N N 6 3 3 0
- 62 T N 1
9 3 4 6,
D E
E O
6, 8,
8 1,
I B
.E S 7 9 8
1 SR T R 4
TC E
S E L
S O P U
T a
S T C /
I EF
(
RE R
)
N T
E N
E M N R
S E
I
)
F B E
O O M T 3 2 6 3 3 2 N -
R I
H T V
O T
N 9 4 3 7 0 1
E T
-N C T A A A 1 2 0,
6 8,
1, B S R R L 4
1 4
O U WE C P T CO I
I S D A
T P S S
R E 0
E R O
3 OL P
CA
(
C N
EI G
HM SU T O
)
b S
NT K T SN
$ N 0 4 0 8 7 0 E
D 7 9 6 9 5 9 GM T A 4,
3, 3,
3, 7,
6, L
NE ID S S
1 2 3 2 5 6 O
3 AR P
E C 0 H N O R 3
C O
(
)
L T M OP O CEM I
/
F E C 0 E R 0
E N N 0 4 9 0
- 8 T 0, 1
E 2 6 8 0 3
I 1
4 NTE O 3 4 6 1,
0, S$
3 B
S NE 1
3 1
T R R
OH G
S E I I T
E O P D
S-EO R
C /$
T NN TT U
(
R N
E E V F
I VI O
MN AT FOO
)
R S
O T E
A 5
R I
U O M T 9 4 0 0 9 4 E
E S
L T
N 8 4 0 0 6 0 HE D
T O
-N C A A A 1,
9, 7,
0, 8,
5, TR N
H R R L 3 3 4 0 1 6 O U T A
N R
S D I
E C P 1
GN S
NO 4
WP R E 0
O 3
I S
E R I
E E
T T
(
P CI E
S L
A B
B P
O R
A P
)
b S
T T
/
K T BR
$ N 0 0 0 0 1
0 UI M
T 2 3 4 0 5 5 E
T A 0,
8, 2,
0, 7,
7, S
H O
L S
1 1 3
1 5 6 FT R
S P
(
O 3
O F
C 0 E
3 TG T
O S
C S
L N
(
O U A C
1 2 3 4 5 6 E H E G N N N N N N R C U N OO O O O O ETL I
o I
T T T T T H OA S T
I I
I I
I P P P P P P TNV U N OO O O O O b
4 I
O YTN IATR ECN
)
D U
ET K
A S
D
)
S U
I I
L R
ET A
A A
S V
U E
L Y
L Y N
=
A E
L O V
L E
D L E
I O E V T Y
A A
KT Y
I MD T C S N L
O A
E N
I I
I Y M T F R A V
I T
I I
S T T T
T A
S N N N R I
I L E Y B Y
A A OE A
C T
I I
T N L U U TC I
E A Q Q P N T
N A Y
OU N
IL U N N
A Y
I O
T T R
(
A E Q A Y
Y T U
T I
E T
E P S E D
Q N N R
N N
I
(
C E
K A
I C R T N A S A
A T I
I S Y T
T R I
N T N E
T U I
U R
R T A
A R
U E
E A
N E
R V E
C D
C G O E Q C T E
I T
N T E
A S A N
N T
N C S S D
AT S
U U I
R A WN U
C R L
O K
E L
E T
E T O A L E C C
S S D
E R A C T A S N L
O R I
O A
A U L
C C
M e
e e
D B
A A
N U
Y Y
Y T
T T
N N
N I
I A
A A
I T
TY T
T I
R FT R
R E
EN E
E C
NI C
C A
N E
N N
T U
B U
U
/
R T
TE K
K S
SC O
S S
O ON I
I R
R C
CU 1
ll l'
S I
R
)
YR C
N CA N
NE U
O
=
E Y U R I
QO E T
_S T
R CA R )S F
A A T E E E
GR Y NA L
A R RL ME OP U
A P T
L DS CL T
AA C
E N
E(
R E
R L
O V CE A
(
X C
R S E E P C S Y
~
N M )T S
E O
T R )R R
E E
U N
C QE E A ENV N
G N P E
E SO E N USY NS I
I TQ D R OR A
A EN O CE ITSAT P
(
TMC I
T NEA X
N I
EDE R
' V( R E
YC E
N R E
C X
U TO QC N
EA R F F
)
R R
)
U D
E EA N
G PE L A A
Y S
B M MT A.
A E AN D
DE T
V RR E E NE R(
A U (P OC
~
D C
O N
C U
N U
A B
I1
.l
0 0
H 0
0 T
1, 6,
N 5
9 8
9 0
5 S
O )S 1
I T
T Y
M N
U T
A B
E R
0 L
I H
0 N
0 0
P R
N T
0 S O O
7, 9,
T I
S 5
0 A
3 S
1 I
D R L
S T
L L
Y E R
A A
T (P 0
I R
N H
0 0
F E
O T
AT T
0 S
I S
6 1
C E
0 TL 1
F 0
R U NS 0
I E S L
3 C E UT G
H N R O
L N
T U
A N
1 1
3 6
KU N
W I
I 8,
5, E
6 8
SS A
Y M
I 1
2 E
M T
IR E
N N
R R
A O
I E
R T
IT R
N C
S O
E F
O A
C A
I D
N TC R
C E
U T
A OT A
O R
A E
L L
U R
O R
S C
A T
E A
P A
L C
A R
O E
B C
E T
T P
U K
N O
S O
O H
H I
O R
M T
T I
I e
W W
0 YT 0,
S I
1 L $
I B N T
A A 4 0 2 6 0
0 0 9 0 7 8 9 9 0
0 0 9 0 B H 1
1 1
1 L
O T R E U
P R
%O S
M E
0 Y 0 R
T 0,
I 1
L $
IB O A T 6 0 8 4 0
0 0 1 0 Y
B 2 2 0
)
N O $
T R
O P M IT O
N C
%RF U
I D
A E
0 0 0 0 0
0 0 0 0 R
0 0 0 0 0
0 0 0 0 H
2, 3,
6, 0,
0, 0,
0, 0,
0, T
M S
T 6 8 5 7 5
9 9 1 0 5
2 3 8 6 0
7 4 0 3 E
R 9
1 5
5 6 3 0,
R R
E 2
T O
E O
M N
E 0 0 0 0 0
0 0 0 0 S
A H
8 2 1
0 0
0 0 0 0 R
R T
2, 3,
0, 3,
9, 0,
6, 9,
0, C
T T 0 4 0 3 1
E A
O 2 3 6 4 I
l P
P S
1 f 7 4 5 3 7 F 4 1
E E
N
/
N N
N O
E E
(
I 1
8 0 0 E 0 0 0 0 0 B 0 T
T H
0 7 0 0 0
3 0 0 0 U
I 2 2 6 3,
E 1,
4 6,
4, 3,
E F
U T
E B
S 1 V 7 4 4 3 7 V 4 I
1 I
N T
T I
R T
E I
I T
S S
B S
N 0 0 0 0 O 0 0 0 0 0 O 0
/
I 5 2 0 0 0
0 0 0 0 I
T D
A 7,
0, 1
2, 0,
0, 0,
0, 0,
P P
F S
E 6 0 2 2 O 1
6 9 8 6 O 7 1 2 4 3
4 5 7 2 O
M N
1 1
1 5 N E
C N
N O
I E
T N
CA O
B IT R
C O
A N
T T
O A
R T
R 1 2 3 4 5 6 O
1 2 3 4 5 6 I
S P
E T
P O
A O
R O
E T
P U
O O
O C
H H
T T
I I
W W
A E
)
N M T
T O
F F
I I
N 0 0 0 0 E 0 0 0 0 0 E 0 I
T O 5 0 0 0 0
0 0 0 0 C
7, 2,
1, 2, N 0,
0, 0, 2, 0, N Y
I E
E U T 6 0 2 2 1
6 9 8 6 D U 1 2 4 B 3 4 5 7 2 B 2 T
1 1 1 5
1 E B O
O I
R R N
N N
T M
S I
E AN I
R D TO N
O S
RS R
E N
EI P
R A
/
C
$ E
(
M A
T T
T N
E I
I I
F P
E T
F F
N A
0 4 9 0 E 8 6 3 3 0 E 6 U
1 3 8 4 N2 2 6 8 0 N3
/M E
M 3 4 6 1,
E 0,
6, 0, 8
1, E
B 3 B 1 7 9 8 B 1 I
O T
4
/
EO T
S O
O S E N
N T C O T C N A
I O
MS P
T IT L N
SU O
C E
A O
T R IT R
C N
O A
N T I
O A R
T R 1 2 3 4 5 6 O1 2 3 4 5 6 I
O P E T
O A
P P
O R
E T
P U
O O
H H
O T
T I
I W
W
STN E
M E
R N
I O
U D Q
I E
S E
T U
R N O
LC TA I
N F
R O
KR CA C
AW B
E OR NA O
ET T
I S
S R
E P
W I
T U
L I
P HR QI A
BT TTA E R R T R
I RWR N R
FO E
EI 1
E H NK S
K T
D N K N
OA A
E O
O R
GI E
E NT T
I I
R P T R T
O C
I P B P
S B O O
W A N OT EUTD E
E O R R J EA FS V
E C N S O G O
H DOG NT E
l B
/
C/ G N M A
N O
G T
G I
R NN MSA F
I EL TO N
U I
I I
E Y O (D P A
R F
N R
F T
E O
DC N
NI D
SN N N I
A EIT IS NEOO E MD N
I I
OUT TP D
O E
C QI CO LA C
RE D NR P R R
R R D UU MO R
NFAFE F
N I
e e
e O
w
=
CL c::"
ct:
C
"!K LiJ
~
O m
D M
U c::C CQ O
N)
O N E
E I
O G
G TI AI AO A
T R
TTT T
CD LL E L N
I A
VV I
L F D OO O
A A V D
R N R
O OE 8 EG S
4 I
E M
D S D
U TE N
N I
E(
U U
F AS C UO o_
II IF H
I I
DG ON I
u gII j
MT I
S D E Y
R R
T)
T TL W
I I
8 I T I
L A
U U O
P N A
C C V
G R R U I O
CY 4
S C C R H E C V
B I T R
I W
R D
S D
E R
G R ED A O A
T T VE C
OS L S I
VN IR D R
RA E
R E
D T N(
D N V
U I
R E
D VM E
U G
ST M
A UP T
IN 1
L O
O R
P I
NC V
T R
R O
OE T
E C
D A
R I
E N
R T
U P
O O
L L
A A
NR NR G E E GE E I
K C I
E E V EN E
G RE R
GA BD A BD TP E
TPE E
E E
E L
I G
T L
T C
C ORI AI OR A AI A
E O
D TV A P D T A P NCR I
E T
U YN YN O O O O AU V T AU V T L
L D
H H
8 E
8 E
R S 4
R S 4
+
+
V U
'j, 4,8 O
O p
t.
I-p
&I, 1 I
N E
1 N
g" d S
A B
O R
N N
r I
I A
A E
R R
T
,II l I.
T l l g]
V E
E R
_ G E
G R
I
_ A V
_ A EV I
_ L S
T T
R P
Y PY
_ Oli R
L S -
D D
_ OIT Y
D EI
_ VUA _
R R VUA _
EI R
G T
_ RCL G
RCL TI E R E T
A U
_ DI R A
_ E R E DI R TD TD
_ N C LR C
_ N C G
LR G
OAI UP N _
R 2
O A I I VCC I I UP N ER S VC ER S R
_ ST R
_ ST U _
U E
R E
R D
_ E D
E N
N V
N V
U I
U I
D D
rl lIL pl1 I L
O
+
+
I k
I lg l!L T
jl III P
I T
I U S
S P
P T RU R
N RI RI OL OL T A TA C N C N A
A G
E I E IG RS RS O
l ll
N O
IT B"
TB R*
A B "A L
T ORS R"
3I t
TT NE
~
U
~
S O
i CG DM T
O
. i 1
g E
R S
C l
F U
O H
S A
M T
R O
gl J_
-l S
B C N C
g O
R A O N
S T
TT.
T
- NC E
TI R
a Cj UA.
U L B A
g;N HT U P
- SN
.RN I
C C
O NR O
h1 AT N
L M
3 J. l EA SH I
e K
A
)
d 7!
RC A
8 l
10 v
>1 IN 5
K>
A 2
EE I
ll I
B A
1 C
1 R
E E
N S
R0, NT 2
I T
E AS V M A
R H=
R A T
W TE K
M I
F A
I S
D I
E
(
G P
A R)
T T
TC N
L
.H W R
O N
D E DG O
3 O
V I
LH R
R U Y U C R NUL TC E
H A OT U UOB D A D
S C OR L E U
WIWG RH O L
LSNAGT T U R N
A I EOWSWH O
E E
RS N
T T T
OO G
A OU H I TL ERLF CP O
I T H SLTWT SCC F
RO EI P AECTDHHRI I
RN TDI U
.D T O) L S E
ECYU EOSNN T
G T AO RWUE A AP AN LWAT CO S E
P T
D L UDH H NEDO O OH. C G R N T R V SCTU R ABE O
T 1
T S AET N ARET S PHR O
EE F
HP A FAXH R N TO( OPETOI
D L
O N
RS TT N E B"
S O
TB A
CG R~
DM O
T -
R V
I NL C
U UO HC S
I I d G
O C
O N l
N g g!! 1 N
b C
N S
l N
LO iI C-O U
,I N-C I
Bl N
,ll ST I
~
P R
PC I
O c
lL SN N
dV 52 E
OS U 1
G P
AD TNRI Y L AT O
A EF V Y T L RA NE LUR E
S E
D EH G
P N RS R P A
RI U
T R T
+
LE I
OV T"
I N'
V R ER U
R D E
H
)A D
S N
VT U
I A
A qI lIl I_
B R I
N I
I S NT A
A DV RA R
T YD T E AN S M
L AL 4U A
A E
RP S
(
I N
DRG NTI S AE N
GP C
I I AR GLT OO O
ILAG A DH T
LV N
DEC TI Y RE U I
OTO L
SSL EN S T
RU P
S O
l+
l L I
Y
+
A L
E R
l1 T
I li I I o
U l
I II
,II I L P
N I
8#
P RI t
T
T A L
R" ORS TT
%)
NES OCG DM O
R T
T2 T2 S '^
SA B
SB R
P B O S TC N A O O
O O
I E T
!jg Tl N8 l
l I
!II !Ial N
N
!jl RT R
U L B
_O l [
lg
[N O
O O
A N
N' 8
l,g IlIII t
N U P TR I
NR AT M
u 2
TH A
2 2
A B
1lC 1=
..C 1
e' A
N A
iN B
NC I
y UA R
O T Ac T
B c T I d
d N
N I U HE I V A5 R
AI V C C I
C f
C T1 AC S R C
OLO R
P T2 I
c I
Y U
I EOT d AF U
TV U
G R F h/ I OEN5 C RV C
[( I I TGU2 yI R
U N I DYI S RA AS F DE H O RI TE C G O VD O S S
TOE N
O
[-I R R D
I C C P O
[( l AVI G TO A
T O
AL R
P S
P O V A R I N A A E
CP NTO R
C T1 EC 5I VA A
E U
T GS L
L A
A N A T A B LN NA I
I OG N GN I
A VIS A R
R PR EP OU I
T T
DIR T
N T U
IT T P C O A I
LAG EC IDE C TI G
TO OTO
' SL SSL r
8 S
t L
o
^
A "U
NT GU I
P
$ N S N P
PI M
I RT
B Y
B O
i Nb F
F O
p f
C E
i8I C
_N III 1
1 O SH C
S C C
N O
L L
B.
TA AT
_ O RS R
PI R
~TT YW
_ TM O
N NE BS
_ NE S
O
[ O T C
C CG R
D b
D R
_ O O
i gC O
F A
_ R R
R
~
II O T A O Ig '
l l llggll ll8 l
II IM R
O S T
B TN T
R U R
NA L B Te A
S S
U P
~
g d
NRI ig!I llgl ggg1 I
lIIlI ET
~~
AT
~
M Ap Bp R
c R
MN C q [_
c I
C-L H
d d
A TC V
V 5
C ST 5
2 EI 2
EO OTW 1
1 L
S OC C
+e o
o N
)
N A A I
I N O G
T S
O C
SR C
O T
I I
A G
O A
E O
O T E
U T
- L I
A L
T" A A T
I TA B" L U
~
T" C A T " T C C T
~ T CB N
N N S
N A~
N A*
S N
PH I
L E
U U E
S N D
I D F N I
D SN D'l I
D A RIC G N L
R C I
D A O
T O
T R R E T R R A T I
N A.
S I
AT S
N I
N U
O R R O
U G
C OE C
G RI I U F
E TT E
F N P m
A 6W A
e S
O M S
M O
(
+
C NB T
N E
OT M
H C
R A
T I
T R
A E
T e
h C4 I
E K
P A
R E
Y P
A TI RRA C
E VB EL S
E KE L
D G
P DI AR O
E N
A S
TP RE P S N T
LI TRI SNDS)A BR E ANR. C OR R CHAE CA V T T
OS T TT RT TSTAA OC STE N ENCAN R
NO MS.O DUAPUN TR I
P C
NHEYHOO o
NMN N D
B A U AE EA L
L SR TI P R( T S BBR F
T I
VTTYTRB S
B USRBSCCB N
I l