ML20205Q518
| ML20205Q518 | |
| Person / Time | |
|---|---|
| Issue date: | 04/16/1986 |
| From: | NRC COMMISSION (OCM) |
| To: | |
| Shared Package | |
| ML20204B851 | List: |
| References | |
| REF-QA-99900403 NUDOCS 8704030553 | |
| Download: ML20205Q518 (91) | |
Text
.
UlN11EU STATES
- NUCLEAR REGULATORY COMMISSION IN THE MATTER OF: DOCKET NO:
INVESTIGATIVE INTERVIEW OF SAM A. MILAM III LOCATION: SAN JOSE, CALIFORNIA PAGES: 1- 86 DATE: WEDNESDAY, APRIL 16, 1986 L
ACE-FEDERAL REPORTERS, INC.
oftmalw 444 North CapitolStreet Washington, D.C. 20001 (202) 347-3700 OD04 $20$$ @ jjjg mmmmma 99900403 PDR
1 UNITED STATES l 2 NUCLEAR REGULATORY COMMIESION 3 _
4 )
In the Matter of: )
5 )
INVESTIGATIVE INTERVIEW )
6 )
(CLOSED MEETING) )
7 ---------------------------------)
Room 422, S Holiday Inn Park Center 9 San Jose, California Wednesday, April 16, 1986 10 11 An inves t :.i; -
- ive interview :29 conducted 12 thic data with % "
.. . .44, III, commc:ncing at 5:00 p.m.
13 Pi 14 For .ne Nuclear ne,ul .T Cor.utission:
15 ROBERT L. PETTIS, JR.
RICHARD P. MC INTYRE, and 16 PETER J. PRESCOTT Office of Inspection and Enforcement 17 Washington D. c. 20555 IS For the Government Accountability Project:
19 LOUIS CLARK, Esq.
1555 Conneciticut Avenue, N. W.
20 Washington, D. C. 20036 l
21 l
22 i
- n
! Jamsas W. HMMNNS
! VU4ETN AEPORTER 37 STATES STREET #23 SAN FAANoSCO. CAUFORNIA 94114 (415] 621-2460
2-I S Q E T E E T E-2 WITNESS:
'I SAM A. MILAM, III PAGE 4
Examination by Representa*.ives'of the NRC 4 5
I FOR IDEN- IN EXHIBITS: TIFICATION EVIDENCE 7
(None)
S 9
10 11 12 IS 14 15 16 17 18 19 20 21 22 25 24 25
3 1 PEQgEERINEE 2 6:20 p.m.
I MR. PETTIS: On the record.
4 I would like to introduce myself. My name 5 is Bob Pettis, with the Office of Inspection and Enforce-6 ment, Vendor Program Branch, located in Bethesda, 7 Maryland.
8 This is Richard Mc Intyre (indicating). and 9 he is also with the Vendor Program Branch; and Pete 10 Prescott, who is also with the Vendor Program Branch.
11 What I would like to do is, ask Mr. Milam 12 some general questions concerning his work activities 13 during the time of his employment from March 1978 to April 14 1982, which correlates to his work record; and basically 15 focusing 'on that five-year period rather than his total 16 ten-year employment at GE.
17 The NRC has in its possession documents 18 furnished by the Government Accountability Project. Those I9 records include a report prepared by Mr. Stokes for the 20 Government Accountability Project, which is entitled 21 " General Electric's Record in Nuclear Safety." This 22 report summarizes your work record over that period of 23 five years: from March 1978 to April 1982.
24 Also in our possession is a package of 25 Engineering Review Memoranda and Engineering Change
m ___
4 I Notices that were changed after your signature and while 2 you were working under Mr. Cobler from the period March I 1980 to October 1980. They are approximately one-half 4 inch thick.
5 We also have several thousand pages of your 6 work record, which cover your employment from March 2, 7
1978 through April 30, 1982. These are grouped in six 8 categories spanning that five-year time span.
' In addition, we also have a package of 10 Deferred Verification Memoirs, which were written by Mr.
I 33 Milam; and, your report entitled " Deferred Verifications,"
12 which is approximately one-quarter inch thick. That was 33 generated, I believe, in July 1983.
14 Whereupon, 15
- SAM A. MILAM, III 16 was called as a witness by and on behalf of the NRC and, 17 having been first duly affirmed was examined and testified 18 as follows:
II EXAMINATION 20 BY THE REPRESENTATIVES OF THE NRC:
21 BY MR. PETTIS:
22 am just describing those documents to you 0 I 23 so that you will understand what we have. This has all 24 been furnished to us by the people at GAP.
25 One of the questions we have is based on
q -
3; - 9-
.c Y 'v ;- 5 c
~
A
_2 .
I ' that information, based on the work records, .and other ,
2 information -I just ' described.
. 8 Didlyou then, or do you now, have any safety
'4 concerns that you would-like to discuss with us?.
If .you~
- 5. -
could, please be specific.'
.j.
6 A My, answer to that might well lay some groun-
~
I dwork 'for the rest-of this discussion. I have tried to 8 make this clear to everyone I have talked to since this
' situation began.
10 I am not prepared to say that some specific 33 -piece of equipment has some particular safety function.
. 12 What I am claiming is: There was -- and possibly there II still is -- a design control system, a design change 14 control system, that is, in some way, deficient. I am 15 talking about generic inadequacies in a design control 14 system."
17 It is my opinionthat that, at least 38 potentially, is' a safety problem in its self. I am not I' prepared to point to some specific piece of hardware and 20 say that hardware has a safety defect.' I think it is kind 21 of important that we understand the nature of the thing II that I am trying to address here. We might or we might 28 not have safety problems in a particular piece of 24 hardware.
25 The point is: I don't believe -- and when I
r s 6
I say "we", I am going back in time now and speaking as if I 2 were s
at GE. I don't believe that we can prove how we I
design something-- I don't think we can prove what we did 4
or did not do by way of verification or design change, 5
because I think a lot of information was lost, design 6 verifications that may not have happened, did not get 7
recorded; or the record, if it was made, was lost. I just 8
don't think that we have a good design history or good design record. I don't think that we can prove that 10 something is safe or is not safe, certain and was done a II way.
12 That I think is the problem. It is a I3 generic malfunction is the design change control system, a 14 design verification system. I think that is a very 15 different kind of allegation than pointing to a pump and 16 saying that that pump won't start when it should.
II BY MR. MC INTYRE:
I8 When you say " lost," do you mean physically Q
I' lost or lost in the tracking system?
20 A I mean both. There were tracking systems 21 that were not used, or that did not work when someone 22 attempted to use them; and there was information that 28 never actually was put into a retrievable storage 24 location. Information that maybe was embodied only in a 25 memo rather than a design record file, and was stored on
7 1
he corner of someone's desk, rather than in a 2
documentation system.
So, I mean that in both senses, it was lost.
4 I personally tried to track down a record of
' The verification and ended up in exactly that situation.
6 It was fairly common.
information just became a deadend.
BY MR. PETTIS:
I which Q Was there any specific time frame in
' In other words, these concerns may be more predominant?
10 or later on would they be earlier on in your work record, II in your work record?
12 My awareness of the problem grew. The A
II reason I was not aware of it during my first five years --
'* which you have rightly chosen not to look at -- is, that, my first five years was rather academic. I did a lot of 16 scoping analyses and computer programming, and things that 17 did not relate to anything that was necessarily being 18 built right then.
19 When I moved into C&I, then I was in an area 20 where we were designing a product that was being built and 21 it was a different type of environment. There was much 22 more immediacy to the design and to the need to verify 25 things because they were being built. In fact, they were 24 being shipped even as we designed them.
25 But when I arrived at C&I, I was not aware
8 I even that most of.the rules existed._ I-was aware, really, 2
only of design record files. I did not know much about I
ECNs. I did not know what an ERM WAS; I did not know that 4- the nuclear safety related drawings had to say nuclear 5 related safety on them. No one told me that. I learned 4 all of those things gradually about' it: just simply I encountering the requirement in one way or another.
8 Of course, the more I learned the more I
' realized there were things that I did not.know. It took a 10 period of time to learn that. That process continued II throughout my last five years. Eventually, the things 12 that I became aware of were, to me, sufficiently important 13 that I guess that is why things just did not continue to 14 i work out between me and GE. But it was a continuous 15 process of learning, becoming aware of things that I 16 perceived as wrong.
II So there isn't any one period where I would II be m o r.'. concerned than any other period. Just I'
continuously, through that last five years, I became o increasingly concerned with the nature of things.
I I would like to point out that, the first Q
five years of your work record, we did not choose to pass 23 over. The first five years of your work record, we don't 24 have.
25 Very probably, then, I chose not to give it A
9 I to1 GAP because, at the time, I probably decided-it wasn't 2 really of any significance to this issue.
3 Thinking back on it, that first five years 4 probably only occupies one or two binders. At that time, 5 I was mostly recording information that would say what 6 project I had worked on. I don't believe I was recording 7 any relevant details of the work itself. And, at the 8 time, probably -- thinking back on it , the reason I kept 9 that record was just so I would be able to tell on my 10 weekly reports what projects I had worked on. There 11 wasn't anything particularly controversial or significant.
12 That changed a lot in C&I. I believe you ,
13 can see that if you know to look for it. As you go 14 through that first volume, where I worked with Chuck Hart, 15 I went back and read through it recently, in preparing for 16 this meeting, and, when I read through it this last time, 17 I was_ aware of a, I suppose you would say, an improvement 18 in my perception of what was important to record. Between 19 the beginning of that first volume, and the end of it, I 20 think that the relevance of what I recorded improved 21 through that first volume.
12 After the first volume, the nature of what I 23 recorded wasn't based so much on a changing awareness of 24 what was important; but just on the changing awareness of 25 what was there throughout the remaining volumes.
10 I
Q I would like to go over, one more time, what 2
it is we do have just so it is clear in your mind.
- I We have six separate folders of your work 4
history, which began March 1978 and end at April 1982.
5 A Let me say that that would be Chuck Hart, O
Lee Cobler, Bob Reghitto. I divided it up by the manager I
at the time. I believe after Bob Reghitto was Arnie 8
Koslow, that is four. Tom Wortham was the last one. I don't believe I left out anybody. Probably the other one 10 would be the Reactor Mode Switch Volume. I am trying to II remember what is there.
II _
Q These are six specific time frames--
I3 A George Stramback, that is the one I left 14 out.
15 Q So, George Stramback--
16 A George Stramback fit between Bob Reghitto 17 and Arnie Koslow. That would be the six periods of time II and six different managers during that span of five years.
II Q And you are saying that the last--
20 A The last one covered a period of 30 days, 21 Tom Wortham. Of all of those managers, he possibly was the best. But I only knew him for 30 days. So it is 23 really hard to say for sure.
O Based on the work record that we have, and 25 the documents that I mentioned earlier, which comprise the
i i
11 1
l deferred verification memoirs and the reports of deferred 2
verifications cased on that, do you have any other I in addition to those already furnished to GAP, documents or contained in your work record, that may affect the safe
' of a nuclear power facility either domestic of operation foreign?
A I have one volume that you did not mention 8
that deals with the reactor mode switch, and it is an
' extraction-from thost'six volumes of pages from the record to that dealt specifically with that issue.
II We have that. I failed to mention that; Q
12 but, yes, we have that. We have the reactor mode switch.
II A I can't recall right now any other volumes I*'
that they did not mention. They mentioned the ERMs and 15 ECNs that were changed; they mentioned the deferred 16 verifications, reactor mode switch.
O And a report by Mr. Stokes.
IE I believe that is everything.
A 19 your present Q Could you briefly describe 20 position?
' A I am an auditor. That is brief.
22 Do you want me to be not quite so brief?
25 You can be a little more MR. MC INTYRE:
24 specific.
25 manufacturing areas THE WITNESS: I audit
12 I and supporting areas of the manufacturing in the 2
semiconductor industry. I' verify compliance with military I
specifications with our corporate policy documents, and 4
with the areas own controlling documents.
5' That is, in a nutshell, my charter at the 6
present.
7 BY MR. PETTIS:
8 Q Have you reviewed the report prepared by Mr.
I Stokes for the Government Accountability Project?
10 A I reviewed it.
II O It summarizes your work record during that 12 five-year period.
13 A I reviewed a draft of the report. I did 14 receive another copy that I expected was going to be a 15 final version. But, I believe, it turned out that I i6 somehow ended up with another copy of the draft.
II MR. MC INTYRE: There is no date on the copy II that we have.
II THE WITNESS: Look at the title. Does it 20 say Control and Implementation Department?
21 BY MR. PETTIS:
22 Q Is this the report that you reviewed?
23 (The document was proffered to the witness.)
- A Yes. This isn't final because this isn't 25 the correct name. I called the FSAR on this issue. It
13 I should Controlling Instrumentation Department. I found in 2 this version that there was some typos, misspelled words I
and things. I reported back the ones that I found to 4
Billy. And I have seen no other version since then.
5 BY MR. MC INTRYRE:
" Q But basically, the format of the report is 7
the same?
I Let me thumb through it.
A
' (After a pause.)
to It looks very much like the report that I.
II saw. I mean, not reading every word, but, in general, it 12 does look very much like the same report based on the 60s 33 second glance at it.
14 If it were drastically different, I am sure Q
15 you would r'ecognize that.
16 A Yes.
I BY MR. PETTIS:
II Q Could you describe your job capacity or job I'
description while employed at General Electric, San Jose 20 during the time period before and, also, during your work 21 record?
22 Well, as I mentioned earlier, the time A
23 before the work record, the first five years, I period worked in a part of the organization that was doing some advanced fuel designs. They were not designing fuel that
14 I
was presently being built; they were examining such things 2
as the benefits of using a nine by nine fuel array instead of an eight by eight fuel array. Whether the fuel channel should be 120 mills thick, or 100 mills thick; that sort of thing. And, as I said, it was pretty academic. We did 6
a lot of computer analysis. We took a lot of numbers out 7
of one program and put them into another and we kept 8
' design record files. But that was really the only kind of 9
documentation that we h4d to worry about. We were not 10 changing drawings in the sense of assembly drawings or II connection diagrams, or ECMs, or anything. We weren't 12-issuing any drawings so we did not use ERMs.
II That is essentially the entire first five I*
years I spent doing that kind of work.
" When I went to C&I, my first assignment was 16 in an organization that was doing, I guess you could call 17 it the mechanical design of the control panels. We were 18 not designing the circuits. We had people called systems 19 engineers here who did that. But we took their elementary 20 diagrams and, with the help of our drafting department, 21 produced panel designs that included assembly drawings 22 that showed the layout of instruments and switches on the 23 panels hopefully arranged according such as to satisfy a .
24 divisional separation requirements, connection diagrams, 25 wire lists, that sort of thing. We did both design and t
15 I design verification of those designs.
2 My second assignment -- which, incidentally,
' I changed while I was working for the same manager -- about 4
half way through the Chuck Hart binder, my assignment 5
changed and I went to, what they called, the production I
following. That job I was assigned to be responsive to problems that arose during the manufacture of the panels.
8 I think that was where I really hit the steep part of my learning curve on things that concerned me. I was helping to people who were trying to build the balance. I was trying II to document changes that I had to make during the assembly 12 process.
33 I was encountering problees that, in many I4 cases, had not been solved when unit one sas built. And 15 it appeared that I had to solve them over again on unit 16 two, and I was puzzled for sometime oser why that 37 happened.
' But, at any rate, I was providing all kind of engineering support during the assembly of the panels.
That included the welding of the structures together,
' installing the instruments, installing the wiring; doing 22 all of the QA inspecting, during the time when the QA 23 inspecting was happening, clear up to shipping and through 24 the final touch-up operation on the paint. I was 25 providing engineering support for that entire range of
l 16 I
operations. And, in fact, I was involved in occasionally 2
changing purchased part drawings, which dealt with things I
that were relevant before we even welded it together, buying some of the pieces that it took.
5 So I provided engineering support for the whole range of assembling the panels. The next assignment was in the technical licensing unit. I went through my I
remaining four supervisors in that assignment.
Originally, that unit was supposed to 10 provide technical licensing support to the licensing II organization; primarily, the review of FSARs against the 12 design documents and responses to customers and NRC II questions. It turned out that we also had some occasional I4 temporary design and design verification responsibilities
" that I thought we should not have had. But that was apart
' of the job.
That set of responsibilities occupied my L last, I guess, three years. I am not sure exactly how 19 ~
many years that represents. And that was the last 20 assignment.
21 Q When you mentioned the word " panels," are 22 these panels that would be eventually shipped to a nuclear 23-facility?
24 A Yes.-
25 0 Would they be control room panels?
i I
17 1 A Yes, control room panels and some local 2 panels, or remote panels; whatever phrase you use. But 3 they were primarily controlling panels and they would 4 house control systems or safety systems. I worked on both 5 kinds. Some PGCC and some non-PGCC panels--
6 Q Could you describe the PGCC concept?
7 A It stands for Power Generation Control 8 Complex. It was a different way of hooking the panels 9 together. Before PGCC, which is why we call it a non-10 PGCC, we did not make a distinction, then. We did not 11 have a name for the distinction before we made the 12 distinction.
13 But, before non-PGCC panels were set up in 14 the control room separately, the wires were installed 15 point to point, wire by wire, by hand. The PGCC concept 16 was that the wires were not run from point to point on 17 terminal strips. They were run to connectors. So we 18 would have wire bundles instead of individual wires. We 19 would fabricate the wire harnesses, or the wire bundles, 20 in the same facility where we made the panels. We would 21 ship the panels and the wire bundles and just simply plug 12 it in. It reduced the amount of work that had to'be done 23 at the utility installing the thing. It also enabled us, 24 in theory, to stage the whole control room and test it all 25 at once easier by not having to actually wire everything
J --
. 18 I just simply plug it all in over in a up. We could 2
building, which we called PGCC.
I But that was the main difference. The panels themselves were not all that different except, 5
instead of terminal strips, there were connectors. And we had wire harnesses that went with the panels.
I Q You referred earlier to unit one and unit I
two. Can you expand on that? ,
A I suppose, in theory, unit one and unit two, 10 on a particular project, were supposed to be identical.
II The reason I referred to it, the reason it is relevant to 12 me now in this discussion, is that, so often they were not II so, often I would encounter some problem identical, and, I4 during assembly that would require the change in the 15 drawing in order to make the panel buildable.
Interference problems or marker plate holes what were a
" different distance apart than the holes in the panel where 18 the marker plate mounted. I puzzled over how they built 19 unit one for a long time.
20 It turns out, there are all kinds of ways, I
' guess, that they can get unit one shipped out without 22 changing the drawings. But, what I would have to do is to 23 call the field, call the site, then have somebody tell me 24 what unit one looks like so that I could change the 25 drawing and make unit two look the same way.
19 1 -It was a little bit embarrassing, as an 2 engineer, to have to do that. But the fact is, unit one 3 and unit two did not always look alike. It was a 4 situation where I knew something had to have been changed 5 because they could not have built it the way it was 6 designed. Maybe there'would be a unit struck channel and 7 a stiffener that interfered with each other. And when 8 they got ready to put the unit struck channel on the
'- panel,_they couldn't because the stiffener was in the way.
10 So what they are doing in a unit one, I don't know. The II drawing did not indicate any kind of a change.
12 BY MR. MC INTYRE:
13- Q They did not make any changes back in the 14 home office on the drawings?
15 A No. Well, of course,. that particular 16 assembly problem would have been resolved while they were 17 building the panel. So, when it got to the field, nobody II was the wiser.
II But, when we got ready to build unit two, 20 and we encounter'the same problem. That was when I became 21 puzzled about what we did on unit one.
22 I remember one example of the utility outlet L
25 that was supposed to be on the back of an enclosure. I l
24 think it was a local panel. I don't remember the number.
25 And there was no cut out for it. The assembly drawing
20 I
showed the utility outlet; the cutout drawing did not --
2 what did they call it? The enclosure drawing did not show
~ I a cut out. Now, different things happen on different 4
drawings. But the enclosure drawing show the cut outs; 5
the assembly drawings showed where to put the hardware.
6 You had to have a cut out to put the hardware.
And, I said, "Well, I don't know. Maybe I
they did not even install the utility outlets.
So, I called the site. The guy went out and 10 looked and said, "Yes; there is a utility outlet there."
II I said, "Can you measure where it is for 12 mgyn IS And he measured and I said, "I am going to I4 put the cut out on that location on that enclosure 15 drawing and put it on the same place on unit two."
16 That is just an example of how I know they 37
_had to have done something to the enclosure when they made it, but they did not change the drawing.
I' BY MR. PETTIS:
0 Q Was this within your area of responsibility 21 while working at GE, to make these changes?
22 A Yes; it was my job to resolve engineering 23 The l
problems during assembly. fact that some of the 24 problems should have been resolved earlier did not change l 25 the fact that it was still my job to resolve them.
l
21 1
Q Could you' describe the training afforded you 2 during the period covered by your work record?
I A Well, a lot of it involved learning things 4 by doing . them wrong and finding out about it somehow.
5 There was,-I guess, an unspoken resistance among the 6 engineers to going to classes. Most of the classes were 7 sort of tedious. Most of the managers did not really 8 - encourage engineers- to go to classes because it took them 9 away from their jobs for a certain period of time. And 10 there wasn't a whole lot of incentive to take any classes.
II And, initially, I did not take any. Initially, I avoided 12 them like everybody else. But, eve ntually, I came to II realize that there were a lot of things that I ' did not 14 know.
15 BY MR. MC I'NTYRE:
16 So, when you went over to C&I, you said Q
37 there were a lot of systems and drawings that were new to II You did not have any formal training you: ECNs and ERMs.
I' when you first moved over?
20 A No; there wac none at all. In fact, at one Il point, I told George Senn, who was the manager of the 22 whole manufacturing organization -- just happened to meet 23 him in the hallway -- that it was dangerous to do it. And 24 I got into a discussion and I told him I learned how to
' And it was the gospel fill out an ECN by doing it wrong.
f.
22 1
' truth. That is the way I learned.
2 Now, there were classes available that were I not offered .very often, and they were over at the _ Tenth 4_ Street Facility. But C&I did not offer classes on ECN or 5 ERM, or design record files, or design verification, any 6 of those things.
7 g go you had to go' after the classes?
8 A E v e n tu ally , I started' going after the
'~- classes, and I started requesting them. But that was 10 probably, like, my last year, maybe my last year and.a II half.
12 There was one class that I never was able to 13 -get into. That was the one _on, I guess it is 10 C.F.R. 14 21, Potentially Reportable Condition type things. I never 15 really got 'into that. I guess I did not last long enough.
16 But they training wasn't common.
17 At one point, when I was in the technical I8 licensing organization, we had, I don't know, eight or I' nine engineers, I guess -- most of whom were not long-term 20 GE people; they had been hired from outside.- 'There was j- 21 only me and one other fellow who had been_ with GE long 22 -enough to understand very much about how the documentation 23 system was supposed to work. AND these guys were writing 24 ECNs and opening design record files not really knowing 25 how to do it.
s
-23 1
At one time, I complained to Dave Reigel, 2
wno was the section, subsection manager. I don't remember I
the name of the subsection, but it was electronic, or the 4
elementary design,~ elementary diagram people. I told them 5
the engineers did not know how to follow these GE 6 procedures. He said, "These are all engineers with 7 degrees. We have confidence that they know how to do 8 things.
9 I said, "I am sure they are great engineers, 10 but I don't know how our documentation system works.
11 And he says, " Well, there are classes that 12 they can take over at Tenth Street.
13 And I said, "Yes, maybe once every six 14 months, or maybe once a year they will offer a class.
15 But, in the' meantime, they don't know what to do.
16 And he said, " W ell, I don't know."
17 You know, he j ust kind . of shrugged 'it off.
18 Q Would these by job shoppers?
19 A No; these were permanent GE-type engineers.
20 They were the ones who were providing the technical 21 responses to the licensing unit ' that the licensing unit 22 was then taking to the customers and to the NRC.
23 BY MR. PETTIS:
24 Q Could you discuss the method used by GE to 25 control deferred verifications?
24 1 MR. MC INTYRE:- I know it is confusing, 2 but--
3 THE WITNESS: That is real hard question to 4 -answer. Because, while I was there, I did not see any 5 evidence that any method was being used. I discovered 6 that there was a method eventually. I started coming in a 7 little bit early and going into the boss's office and 8 reading the procedures; manuals. I discovered a lot of 9 things doing that' that I had not known before. I 10 discovered something called a design verification status II change notice that I had never heard of before, DVSCN.
12 I discovered that, whenever you did a design 13 change with a verification, you were supposed to fill out 14 one of these forms and send it over to Ken Dawley's people 15 at Tenth St'reet and they were supposed to be keeping track 16 of these things.
17 I went in and asked the secretary for one II and she did not know what it was; she had never heard of 39 it before. That makes me think that probably no one else 20 had ever asked her for one. I never saw anyone else use
. 21 one.
22 So I got the EOP manual and showed her the 23 picture and showed her what it is. And she did not know 24 how to get it.
25 l
i l
I
25 1 BY MR. MC INTYRE:-
2 'O You did not have your own copy of the EOPs?
- 3 A No; there was one in the boss's of fice.
4 There was no particular incentive for any of us to use it.
5 There were a lot of procedures manuals other than EOP.
6 Q So you were doing design verification and 7 were involved with them, but had never read the EOP on 8 independent design verification?
9 A Only to the extent that I went and looked at 10 it myself. There would be references to it in letters.
11 People would say we are going to do such and such per EOP 12 often without any reference to which EOP number. And it 13 was about two volumes about so thick (indicating) . That 14 was another of the disagreements I had during this time in 15 the technical licensing unit. -I told you that we had 16 occasional temporary design responsibilities that I felt 17 we should not have.
38 They would . write ~ a letter saying, for a 19 certain period of time, and for a certain systems, we are 20 going to -- and for a certain project -- the design 21 responsibility will be transferred from component 9012 --
22 I don't remember what component ours was at the time --
23 per EOP. And I complained to my manager at the time that 24 should have at least said which EOP since there were quite 25 a number of them.
I
26 I MR. MC INTYRE: That makes sense.
2 THE WITNESS: That was the kind of use that I It was something to refer was typically made of the EOP.
4 to when you wanted the authority to do something.
5 BY MR. PETTIS:
6 So what method did you use to control Q
deferred verifications?
I W ell, e ve ntu ally, I started trying to use A
' I first had DVSCNs. But there is more to it than that.
10 to learn how to tell if my source document had a deferred II verification. It wasn't visible.
12 The drawings had a certain little bitty 33 letter right on the corner called an ICER CODE. And I 14 believe ICER stands for Incomplete, Complete, and two 1
" other words that I don't remember, now, what they were.
16 But it dealt with the completion stage of the drawing.
I7 And I eventually--
BY MR. MC INTYRE:
Q The design drawing, you are talking about?
A Yes; the elementary diagrams, conne ction
' diagrams, assembly drawings, enclosure drawings, marker plate drawings. The all had the ICER CODE right down ~at 25 the top of the revision block.
24 I eventually learned that that had nothing 25 The fact that the drawing to do with verification status.
_ .27 I h'ad a "C" on it, did not mean that the design was
.2 verified. It simply meant that that drawing was . complete I
and you' had to ,look somewhere else to find out about the 4
verification status. . ' And that was an important piece of 5 knowledge because I- wasn't doing any deferred 8
ve rifications before. I had no real reason to . d'o a.
7
. deferred verification. I would simply go get my changes 8
verified. by somebody. But, when I' realized that the ICER CODE did not tell me anything about the ve rific'atio ns
.10 status of the design, I looked further and I discovered
'II
' that in EIS they had something called a SUN Code. That 12 stood for, S-U-N, I don't remember what they "S" stood I3 'for; and I don't remember what the "N" stood for; but the 14 "U" stood L for unverified. You could not tel'. by looking-15 at the drawing whether it was verified or not.- You had to 16 go run, I think it was, a DI program in EIS--
17 Q So this is the other--
II A This is the computerized engineering system, I'
which isn't, well, it isn't a retrievable document system 20 in the sense that it is revision controlle d. It is a 21 c urrent. information. And, when some particular piece of 22 data in the EIS is updated, the old data isn't retained.
23 To my knowledge, it is just simply updated. So it is a 24 current type data base, and it tells you the drawing
' 25 application, drawing revision, unincorporated ECNs, that
t 28 3" sort of thing. -It is very useful. It is a wonderful 2 system as long 'as it isn't misused.
I But, I learned - that you could go look in EIS 4 and it was a "U" in the drawing, in the design status 5 - column of that output, then you knew that there was on i record some incompletion in the verification. So I began 6
7 to consider those document, that they were' unverified, to
~
8 be, in f act, unverified. And, when I used one of those as 9 a source, the change that I made, using that source, I 10 would make as an unverified change. My theory being that II I could use an unverified source to make a verified change It matched 12 no matter who looked at it and verified it.
13 what I started with. Because what I started with, was
- 14 unverified.
15 I was hoping to, by then. escalate this 16 deferred verification thing, or snowball it, to the point 17 where they could not keep ignoring it. And, single l
II handedly, it was kind of hopeless, looking back on it.
l I' I eventually realized that even EIS wasn't a 20 completely reliable record. Because a lot of the deferred 21 ve rifications never made it into that record. And, 22 eventually, what I learned, and that I had to do, was 23 request every ECN -and every ERM that had ever been written 24 against that drawing, and look individually at each one;
' look at the verification statement on that ECN, or on that
(
29 I ERM, and see if 'that particular change had been properly 2
verified. And I found, then, deferred verifications for I which there was no record at EIS probably because nobody 4 had filled out a DVSCN and sent it to Ken Dawley, whose I- job it was top get the record into EIS. So it turned out 6 that there were very high proportion of drawings that had deferred verifications that were not only not visible, but 8 were sort of hidden, hard to find, took a lot of work to
' request all of those ECNs and wait for them to come back
~
10 and, then, look through them.
II Af ter that, I tried to use DVSCNs. That was 12 near to the end of my time at GE. It was one such episode II that, in my opinion, sort of contributed to the end of.
I4 _ things. I had changes that I was making on a particular 15 set of drawings that included a source that had a deferred I verification somewhere in its history. I completed my DVSCN. I did not put, I was making a change in 18 application, I guess, as well as-- Okay. I was issuing 19 new drawings, making a change in application based on 20 another set of drawings for another project. The other
' set of drawings for the.other project had a~ deferred 22 verification. So I wanted to make sure that my new set of drawings also was known to be unverified. I am trying to 24 snowball it.
25 You have seen the picture of the little
30
-1 mouse flipping off of the Owl. , That is what it was, I 2 guess.
3 MR. PETTIS: We will take a recess at this 4 point.
5 Off the record.
6 (Whereupon, a brief recess was taken.)
7 MR. PETTIS: On the record.
8 Where did we leave off?
9 THE WITNESS: I was just telling about my 10 efforts to use DVSCNs.
11 BY MR. PETTIS:
12 Q I just want to ask you one thing: Is there 13 a specific time that we are talking about right now when 14 you first encountered this deferred verification problem?
15 A That probably began within my last year at 16 GE. It was fairly near to the end that I started 17 realizing the situation with- deferred verifications. it 18 was while I was doing some of this temporary design and 19 design verifications assignments.
20 One thing I should mention is: When we, in 21 the technical lic e nsing unit, received a temporary 22 assignment of design r e s p o n sibility , that design 23 r e s p on sibility would often be farmed out to a 24 subcontractor immediately. We would retain the 25 verification responsibility.
A
31 I The whole thing, to me, at the _ time, seemed 2 kind of " flaky."
I Was this the onset--
Q 4 A Yes; that was -- No; I don't believe it was
- 5. C. F. Braun that was involved in that. C. F. Braun was 6 involved earlier in some C&V design work, I believe.
7 I was aware of.that back when I was working 8 for Chuck Hart. I am not aware of C. F. Braun doing any
' subcontract work for the technical licensing unit. But 10 Omtec was one of the companies that did, but I don't II remember the other names of the companies right offhand.
12 But it was-during that period of time that I became aware II of deferred verifications as I now perceive them. And, 14 during the last year, I learned more about them and became 15 more deter' mined in my efforts, well, not only to use the 16 system as it was; but to try to improve it.
17 There was some problems with DVSCN even if II For one thing, it did not have any kind of it were used.
I' an issue cycle; and it did not have any kind of a number 20 and it wasn't retrievable in any way. When it went to Ken l
21 Dawley, it was his job to log them into EIS and, then, 22 And, whether or not you could ever have file them.
23 retrieved a particular one, I think is very questionable.
BY MR. MC INTYRE:
' You would log it into EIS but it did not Q
32 I
have a number on it?
A I think he gave it some kind of number just in his own filing system. But I never tried to retrieve a 4
DVSCN, so I don't know if it is possible. But they did not have any kind of an issue cycle like any other kind of 6
a design document. It required a section level signature.
7 But they would usually rubber stamp it if you had initialed it.
9 That leads me to the next part of my story.
10 As I am going through this effort to change things, I guess it is inevitable that a few- people would 12 get mad at me. And it happened that, once while Reigel
'I was out .of town, Bob Reghitto was acting for him.
Reghitto was one of the people who was made at mew because 15 of another problem that had been going on.
16 W hile he was the a c tin g s e c tio n level 17 manager, I happened to finish this particular project that 18 I was going to tell you about a moment ago that included a 19 deferred verification. I wrote a DVSCN. I did not put a 20 design record file reference on the ERM because the ve rification was not complete. There was no work to 22 support the verification in the design record file. One 23 of the things we historically had problems with was .
24 referring to something in the design record file that 25 .
turned out not to be there when the auditors went looking
33 l
l I for it. So I was careful about that. So there was no 2
design record file reference, but there was a DVSCN.. That I-was in accordance with he original ' requirements as 4
specified in GE's procedures.
5 I handed the package to my manager' and it 6
was routed to Reghitto. Because of the DVSCN it had to 7
have Ba rr e ntine 's signature. Reigel normally just 8
initialed them and gave them to Barrentine, and' Barrentine signed them. But Reghitto did not. Reghitto drew a big 10 diagonal line across the DVSCN and said "Not required, per II EOP 42-6.00 -- I don't remember for sure the EOP number --
12 and sent it on back to me. He also included an 33 instruction to add a design record file number to the ERM.
14 I, to a certain extent, lost my cool and 15 sent a note back to him and said, "Yes; in fact, a DVSCN 16 is required. In the future, either read the procedures 37 more carefully before you mark up my work, or, you know, I'
so that I do not have to do it over again." And, I guess, I'
that just made him madder than he already was.
20 But I wrote a new DVSCN; I started over. I 21 instructed my manager that I did not want a design record 22 file opened. I did not want a reference to an existing 23 one put on the ERM because there wasn't any supporting 24 calculations or information of any sort in any design 25 record file that would back up what we were doing. I gave
34 I it to my manager again and he sent it to Reghitto again.
2 It did not come back. I started watching, along the way, 3 in the issue cycle and I sv.; the work go by the drafting 4 department without the DVSCN and with a DRF number on it.
5 So I just took it and -- I don't remember if I did that 6 again or not. I may have done it again. I think I made a 7 new DVSCN, started over with a clean ERM; went back and 8 got the signatures on it, without the DRF number on it,
' and put it into the cycle again. A second time, I had to 30 retrieve it in draf ting when I saw it go back with a DRF II number added and without a DVSCN on it.
12 At that point, I misfiled it in my file 33 cabinet and, then, waited. On Wednesday of.the week when 14 the work was due, my boss came around and asked me if the 15 package ha'd ever been issued. And I said, "No."
16 He said, "Well, are you going to issue it?"
17 And I said, "No."
II He said, "Why not?"
II And I said, "Because of what is happening 20 here."
21 And he said, "But it is due Friday."
22 And I said, "Yes, I know. I took my three 25 months to do it. You can start with a new engineer or do 24 it my way. And you have two days to decide."
25 Frida y, they came around and said that we
35 l
I will do it your way. And they did it my way. But the 2
first of the next month I got laid off. So I did not ever 3
get to finish the battle of the DVSCNs. That was the last 4
one I ever tried to use.
5 BY MR. MC INTYRE:
6 Q And your way was following the procedures--
7 A My way was following the requirements for 8
both DVSCNs and design record files. And the design
' ~
record file was not required for every design change. A '
10 design record file was intended to be used for anything II that was not otherwise retrievable. You could put your 12 entire support right on the he ECN, if you wanted to.
II And, in fact, it was bad to use'the design record files 14 too much, because they were not conveniently retrievable.
15 And ECN w a's retrievable by its subject. You could just 16 request from the drawing room by ECN number and get.that 17 ECN.
I8 The design record file, sometimes they were I'
used. And entire file cabinet could be one design record 20 file. But a design record file was only retrievable as an 21 entire design record file. You could not retrieve a page 22 of it. So it' was not really a very convenient way to 23 store something that you might want to get back again 24 unless you opened a new design record file for every 25 particular thi~ng. And, as I said, this procedure, EOP 42-g y _ . . . _ _ . -
36 1
10 - I am kind of guessing; I am kind of guessing. I am 2 not sure. I remember those numbers.
3 But the procedure said that the design 4 record file did not necessarily have to be used. I wasn't 5 required to use a design record file by the procedure. I
/
6 was required to - use a DVSCN by the procedure and that is 7 what I tried to do, and it wasn't very successful.
8 BY MR. PETTIS:
' Are you aware of any components or activity, Q
10 manufacturing activity, within General Electric which may II involve a high number of deferred verifications; is there 12 one area that maybe more prone to having more deferred 33 verifications?
14 A Systems Engineering Design or ga niz a tion, 15 under -Dave' Reigel, they might be kind of hard to find; 16 they are not very visible. But I think there are a lot of 17 them there. I think most of the elementary diagrams will, II somewhere in their background, have at least one deferred I'
verifict. tion that was never finished.
20 When you say " Deferred verification that was Q
21 not finished--
22' A I mean an ECN.
23 You don't have any, you can't recall any Q
24 closecut mechanism for that deferred verification; or is 25 it just your 6nderstanding that it hasn't been closed?
37
' A Well, _I can give you an example of one that 2
I tried to track down once. And I can't remember the
~ I drawing number, or the system number; I just remember that
- ~
I tried to track down this particular verification and see 5
if it had ever been done.
This particular change had been made by --
maybe his name will come to me in a minute. It was one of 8
the people who was involved in doing some support for production. He worked for Bob Parker. Larry Chamberlain, 10 He had made this change.
his name was Larry Chamberlain.
II He had not shown it to the system engineer who was 12 responsible for the system. At least there was no U evidence .that he had shown it to this system engineer.
" There was no' system engineer signature on it. W ell,
" normally a' system engineer would not have, would have 16 verified the change. I am not sure that that is proper 17 either, but that is the way it was done. The system 18 engineer would have verified the change to the production 19 support engineer that it had been made to the system. But
- it was a deferred verification. There was no signature there.
22 The lead system engineer over at Tenth 25 Street had not signed it either, where he would normally have signed it in " A". Larry Chamberlain had signed it.
25 So, really, he and his boss were the only
38~
I one who had ever seen ' this change. It was quite a long 2 BCN and it made some technical changes to the panel. And 3 I wondered if a deferred verifica tion had ever been 4 cleared up.
5 So I went to him and asked him, and he said, 6 "Oh, yes; we had a design review meeting that verified the 7 whole system." .c 8 I said, "Was there any kind of a record made
of that meeting?"
10 He said, "Yes; we issued a memo." ->
II And I said, "Do you know where it is, how I 12 can get hold of it?"
II .
And*he was a ble to find ' one on his desk 14 somewhere in the stack of stuff and showed it to me. And 15 it was a short memo with an attachment that said that we, 16 the design review team, "We have reviewed this system."
II and it listed some action items. One of these action II items are closed and the system is considered verified."
I' I said, "Were the action items closed?"
20 He said, "I don't know. They were all 3
21 assigned to different people." And the ,' list was on the 22 back.
23 'i I looked, and;there were about 52 action
. isems,;I believe. And I checked the first one and it had 25 been done; the thing that needed to be done had been done.
(t
39 I
I checked the second one, and it had not been done. And I 2 ~
-did not check anymore after that. It was just a deadend.
I But it had sufficed -to convince whoever was in authority that needed to be convinced that the thing had been 5
ve rifie d. T' hey had a design review meeting and they_
issued a memo. And, if the . action items never got followed u p, well, that was the fault of the people who 8
did not follow them up. It wasn't the design review committee's fault.
10 That kind of illustrates how difficult it II can be to track down whether or not a deferred 12 verification was ever closed, especially when there isn't II any good record of the deferred veri:- ations. If you can I*
find them, and find out what particular changes had a
" deferred ve rification, it is difficult, maybe impossible, to go back and find if they were ever closed. Because 17 there isn't any kind of traceability anywhere with respect 18 to deferred verifications. There is n; traceability to 19 DVSCN if one was used.
20 The ECN is really the only specific record 21 of a pa rticula r deferred ve rifica tion. And, if the 22 verification was ever completed, there won't be any 25 traceability from that ECN to that completion, wherever it 24 might be. Because, if the completion is done, the ECN is 25 isn't modified. At one point, I made a suggestion that
40 I the DVSCNs should be issued with the same number as an 2 ECN; and, when the verification was closed, that the DVSCN 3
go to the documentation area and be attached to the 4
original ECN as an extra page. At least, that'way, when 5
you requested the ECN, you would get the DVSCN with it; 6
that would have helped. But they did not want to do that for some reason. But there was a la ck of anyway to I
retrieve that information, assuming it is there.
BY MR. MC INTYRE:
10 You could not find that in the EIS that you Q
II had mentioned earlier, that you could look in that system 12 to find--
II A EIS would give you the current revision of I4 the document and it would notify you of the existence of 15 any unincorporated ECNs, but that is all.
16 0 It would not include a DVSCN?
A No, no; I never saw a DVSCN listed in the EIS. They did not even list ERMs -- and that was another 19 problem.
20 An ERM could be used to make a change in
' application, which meant that you might have a parts list 22 calling out a particular drawing for a particular project, when, in fact, there had been an ERM issued that changed 7
the application. So that that drawing did not apply and 25 another one did. And the ERM might not be incorporated l
l
41 1 into the drawing yet. So . you might have a drawing that 2 ' listed a number that was an incorrect number and you would 3 not know that the ERM existed because the ERMs were not 4 listed. An unincorporated ERM was not listed in the EIS.
5 And I tracked down the reason for that.
6 The reason for the ECM numbers, consisting 7 of two letters and five numbers; and ERM numbers consisted 8 of three letters and four numbers. And the format in the
' EIS was set up to take two letters and five numbers, not 10 three letters and four numbers. And they did not have the 11 - resources to have a programmer go change that to some sort 12 of algorithmic format. So they just did not include ERMs 13 in EIS. That is a typical reason.
14 0 And you never saw any documents listed in 15 the EIs system that would be consistent, verified?
16 A W ell, I am sure there were documents that 17 were probably verified. I don't mean to imply that all of 38 them were unverified. But I think a lot of them, maybe I' most of them, were unverified, maybe, depending on how 20 hard you look for the deferred verification. There might 21 also have been cases were you might question whether a 22 verification was in fact a verification or not, depending 23 on the statement. There were things like an ECN might say 24 something like : verified by review to appropriate 25 'And you wonder: Is that really a strong documents.
- - - - - - -. , - . _ - - __m, . ,3 - -
-v
$2 1
cnough verification statement to convince somebody that it 2
was really verified. It might be questionable.
~
BY MR. PETTIS:
4 Q Do you have any knowledge of any component, 5
specifically that was shipped to a domestic or foreign 6
nuclear power plant, which contains a known defect.
7 A Are you also interested in foreign plants?
8 BY MR. MC INTYRE:
9 Q If it would be a generic concern.
10 A It is hard to say. I remember one panel, 11 but I am not going to be able to tell you the project or 12 the panel number. It might be possible to dig it out ~of 13 my work record. But I may not have it recorded it. I 14 don't know.
15 I do remember one panel where there was a 16 ling of CR2940 switches, horizontal row of them, that was 17 above a horizontal stiffener and they were too close to 18 the stiffener. And, if they rotated slightly in the hole, 19 the screws could touch the stiffener. I wanted to cut the 20 stiffener out and weld it in a little lower. But I was 21 being transferred to the technical licensing unit at about 22 that time. So I did not have the opportunity to follow 23 through with that and one of the other engineers took it 24 over.
25 I heard a rumor that, what they did was,
i 43 i I just tape a layer of mylar to the top of the stiffener and 2 shipped 'it. So that might an example of what you are I talking about. . But I don't remember the panel number.. I 4 did ask _ if they. thought to include the mylar on the parts 5 list for the benefit of the next unit. And they said no.
8 I mean, that at least would have been the minimum amount 7
of documentation; .at least change the drawing to show that -
I you have to tape a strip f mylar to the top of the stiffener.
10 So, as far as you know, they shipped that Q
II panel?
12 -A As far as I know, they shipped it with the II strip of mylar taped to the top of the stiffener.
14 And that would be the fix?
Q 15 A That would be the fix to keep the screws 16 from shorting out. I never saw it; I was transferred out-I7 of the job. But that was what I heard from someone who II worked down there.
II BY MR. PETTIS:
20 Could you explain the term " bootleg" and 0
21
" mortgage"?
22 Those are interesting terms. We kind A Yes.
23 of made those up.
24
" B o o tle g " means that, done without authorization'and without any record. An example would be
44 1 - marker plate holes in the " fab shop." Well, marker plate 2 -holes in the assembly shop.
I When it came time to install a marker plate, 4 the holes in the panel were not the right distance apart.
5 What you do is: you go off somewhere, and you borrow an 6 electric d rill, and you look both ways and make sure 7 nobody is watching; you drill a hole, you put the drill 8 back, and then you install the marker plate. And it is a 9 bootleg fix. That is what I call a bootleg fix.
10 Q Even though there were procedures that Il probably prohibited that type of activity?
12 A W ell, I considered that to be a de facto 13 design change m y s elf. The panel doesn't match the 14 -drawing.
15 Q Were these manufacturing personnel at GE on 16 the manufacturing floor taking it upon themselves?
17 A Yes; that sort of thing would happen.
38 BY MR. MC INTYRE:
19 Q And they would not document these changes?
20 A No, no.
21 Did this happen quite often?
Q 22 A W ell, it happened often enough that I 23 worried about it. That was just an example. It is easy 24 to remember.
25 I remember another example that involved a
45 I guy named Bernie Gennetti, and a woman named Bubiak. I 2 am guessing at the name. I think it was-Bubiak. She was I a QA. Inspector and she was worried that some surface -- I 4 don't remember what the surf ace was; but'she was worried 5 that it was mounted too close to -.the top of the power 6 srpply ' and was going to interfere with the air flow, the'
'7 cooling. It was a convection cooling type thing through 8 this power. supply.
' She asked Bernie about it. Bernie said that 10 it was mounted on the unit struck c'hannel and to just II slide it .up a couple of inches and do it that way; and, 12 then, when you inspect it, it won't match the print.
' accept as 33 Write up an inspection report and we will say:
34 is.
15 And I said, "Not on my panel, you don't.
16 Let us change the design." But that was kind of the way II they worked.
II The thing may not be a real important design
" change; but, then again, it may be. The fact is: it is a 20 departure from the documented design. And the intent from 21 the very beginning was that the drawing was not going to 22 be brought up to date.
23 It was when I saw things like that happening
' that I started to realize how it was that I was fixing problems on ' unit two that had to have been accommodated
1 46 1
somehow bn unit'one; but the drawing was not' changed. It 2 was " bootleg" ~ fixes.
'I .BY MR. PETTIS:
4 Q
~
Did you feel obligated to raise this with 5' - your management, that this activity was going on?
6 A Yes. Mr. Cobler really did not vant to' hear 7
things 'like that. He would usually sit there and sort of 8
keep reading something while I explained it to him. He I would then .say: W ell, thank you, Sam; you are doing a 10 good job. And that was the end of it.
II Q Would you explain the term " mortgage"?
12 A .That came from what we called MRB. It dealt II mostly with either, w ell, purchased things. MRB dealt I4 almost exclusively with purchase things, either equipment .
that we bought, parts, meters, switches, power supplies, II or enclosures that were made at farmed out fab shops.
' The biggest' problem was the enclosures..
l They would come back with variations from the design. A good example is a case that I recall where there was a built in interference between a channel and a stiffener.
' They had to go past each other and the designer just did I
22
- not notice that, actually, they had to go through each 23 other with the dimensions that he had provided. And we 24 had three enclosures built to the same design, and I think 25 they had been built by two different subcontractors.
L 1.
l
47 I In one case, the subcontractor had chosen to 2 cut the stif fener; ti.e other case they chose to cut the I
channel. So, what we had was two dif f e rent bootleg s' o lutions to the prot,lem. And I was puzzling over what to 5
do about it. But, for that sort of thing, because of the 6
intense pressure t.o get things through receiving as quickly as possible., there wasn't time to write an ECN.
8 Usually the engineers would, on the inspection report, say something like there is not, accept form, fit or function; 10 And they would send it right through and accept as is.
II not. change any drawings at all.
12 But, occasionally, they would do something II This means they that they would call mortgaging an ECN.
I*
would say this problem will be corrected by ECN number, 15 and they would write an ECN number that they got off the 16 log. The ECN was not written yet. The intention was that 17 maybe tomorrow they will get around to righting the ECN 18 because the number is reserved and mortgaged. And, if 19 they were real conscientious, they might write the ECN and 20 maybe they would not. Because, once the panel was on the 21 through the receiving, inspection, then, the pressure was 22 off to do anything about it and the pressure was on to 23 solve the next problem.
24 So I suspect as lot of mortgaged ECNs 25 probably did not get written and there would be no way
o 48'
' that anybody would know about it. - .Because, the inspection 2
~
report was never attached'to the hardware. It just went
~3 into a file and the record center and would never be seen
'4 again. There was normally no record made of the' 5
-inspection report number on any piece of d esig n 6 So the inspection report was just an documentation.
vehicle for getting ~ discrepant ' things through receiving 8
and inspection, and filing the inspection report. It was 9
essentially not a retrievable document because nothing 10 Although, it had a number; it referred to it by number.
11-could have been retrievable if they had, somewhere on the 12 indicated drawing modified by inspection report drawing, 13 number, or something; you know, panel accepted with 14 discrepancies. But they never referenced anything, so it 15 became a nonretrievable document. .
16 Also, when it was used to make what I
! thought of as a de facto design change, .where it allowed a 18 departure from the design to~ happen, I considered that, in 19 effect, to be a design change because the panel was 20 When it was used for that, I considered it to different.
11 be a design change that did not have a proper review cycle 22 like an ECN. We were required to verify design changes 28 commensurate with the design under 10 C.F.R. 50, Appendix i
\
24 B. And I considered these irs to be occasionally used to 25 l i make design changes without verifications. Because they 4
1
,- , - - - , - , ....,._,y,,- -
,_,,,,,,,,,,,._,,__-,,-.,,-,._,_wy.,,
49 1 did not have the same kind of scrutiny that an ECN would 2 have. They did not go through anywhere near the same 3 issue cycle, a review cycle, that an ECN would go l 4 through. Yet, they would accommodate a change in the 5 hardware. And I complained about that to no avail. l 6 I only lasted a week on MRB and, then, I i 7 went back to my other assignment.
8 Q What is "MRB"?
9 A Material Review Board.
10 0 Where was QC all of this time?
II A Well, that-- I don't know. I guess there is 12 a lot of ways to answer that, and probably I would have to 13 give all of the answers before it would r e ally be the 14 whole answer.
15 Part of the problem, I think, is: the 16 manager of QA reported to the manager of manufacturing. I 17 think that was the problem. Ferguson reported to George 18 I think he should have reported to somebody on a Senn.
' I9 level lower than George Senn, or I. I don't think QA 20 really had the authority that it needed because of the way II the organization was structured. I think, also, QA may 22 have been worn down , in some respects, to a state of 23 apathy just because they got so little engineering 24 support.
25 There were a lot of little ways in which
50 I
that was manifested. One kind of amusing example was the 2
fellow named Bill Trader, who was one of the production engineers with me at the time I worked for Cobler. He and 4
Cobler went back a long way together. He, Cobler and 5
Gennetti, in fact, went a long way together with GE. They were old buddies.
But Bill Trader had an interesting technique for dealing with the CARS from QA. He had a whole list in his head of quick answers to CARS like: Does not affect 10 form, fit or function; use as is. Or, check for current II revision; or, he had a whole bunch of them. And when he 12 received a CAR from QA, he would put down one of these II answers, not always, but often; then shoot it back to QA I*
as resolved.
A couple of days later, another CAR would 16 come back from QA with the same problem on it referencing 17 the previous CAR, and saying that your answer is 18 unacceptable; this problem really does exist.
In the m e a n tim e , Bill would have been 20 researching the problem. So, when the second CAR arrived, 21 he had the answer and he would put the answer on it and 22 send it back. And he got credit for very fast response to 23 twice as many CARS as I could handle. But that did not 24 endear him to QA.
25 QA did not get very much support from
, . . ~ . . -- . .,
51 I
engineering. . . In . fact, . I discovered, at one point, that
.2 thel fab shop wasn't getting any support at. all from 3
engineering. I sort of took it upon myself to provide 4
,-ome support. But I had to do it under cover because I 5
had not been told-by my manager to support the fab shop.
6-So I had-to kind of sneak in the support. But I had I
discovered ' that there had grown into place a gentlemen's 8
agreement between the fab shop people and fab shop QA l- where, if they. encountered a problem in the design, they 10 would just simply fix it and then QA would accept it and II sign it off. . Because they despaired of getting . any a
12 support from engineering and there was really nothing else I3 they could do.
14 I started giving them some support and 15 t
making some' changes in the drawings that they needed for 16 awhile.
II Q All done under the procedures of the GE?
II A Well, the changed that I made, I made using II ECNs. I did it they way they should have been done. At 20 one point, somebody told me that the manager of the fab 21 shop QA people had told his ' people that, if they had a problem, not to go to Cobler and try to get anything 25 through the system, not to try to get help through system; "
but just go straight to me because I would help them. But ,
25 it was all so'rt of unauthorized. But they needed some f
b
52 I engineering support.
2 But that is part of the answer, like, where
' I was QA7 QA was there,-but, for a variety of reasons, QA 4 maybe wasn't very effective.
5 Q Well, QA, QC; I was more interested in QC 8 since they may have had the responsibility to follow I through some type of a traveler system and know where 8 these panels were at any stage during the manufacturing.
' They did not know where the panels were. I A
10 don't know if you notice that in record. But there was II one sequence of time where I explored that issue and I II discovered that production control, who were the people 13 who supposed to know where the panels were, did not know 14 where the panels were. I would ask them where the panel 15 was and I 'would go look there and it was not there. I I8 would search for it and maybe, eventually, find it; and II maybe not.
38 There were cases where they told me that a I' And I would find it in panel had not been started yet.
20 There were cases where they told the paint shop finished.
21 me that unit one had been shipped and unit two had not 22 And I found both of the panels both labeled been started.
25 But I don't know what to say. But it is u nit one.
documented in the work record. And, for those panels, I 25 have panel numbers written down.
53 I
Q That are in your work -records?
2 A They are in the work record, yes.
I Q Were you' aware, during this time frame, that
'4 there was such a provision called 10 C.F.R. Part 21 where 5
you could report defects or noncompliance direct to the NRC?
A Even now, I am not aware that, as a.GE 8
employee, I can report directly to the NRC. GE procedures prov'ide that I report it through the GE chain of command.
10 I think it would have been.very dangerous for me to try to II report something directly to the NRC.
12 Q So you used the mechanism within GE to II report?
I4 A Only one time did I do a PRC. That was on the reactor mode switch. And it took three months before I'
they could deliver it to the licencing organization.
Q Were you satisfied with the response for the
~PRC?
A No; I wasn't satisfied with anything about it.
I felt like that the way they handled that PRC was in itself a PRC. But I did not report that one. I was in 22 enough trouble already. But I thought they had lost it SS for a while. Ba rrentine said they he had sent it to 24 licensing, and licensing had received it; and nobody knew 25 where it was. I don't know where it was. Eventually it
54 turned up and got sent to licensing.
2 Q You are talking about the request?
I A My report, that I felt that this was a potentially reportable condition, had to go to my section level manager. And it was his job to report it to the licensing people, whose names I don't remember any more.
O Could you briefly explain the circumstances I
surrounding the mode switch issue?
A w ell, there were a variety of things wrong N with the mode switch. And, in my opinion, some of them II were kind of little things. But, the thing that prompted 12 me to write the PRC was: When I discovered that it was II possible to desynchronize the contact unit so that the switch selector handle in some specified position, you did
" not get th'e contact closure that you thought you were 16 getting. And the way that you a c hie v e d that 17 desynchronization was very similar to the steps you had to
' go through to in assembling the switch. So it is entirely 19 feasible that an operator, messing around with getting it i
put together, could desynchronize those contacts and not 21 l know it. And there was no functional test for that 22 switch, for that function, at any time that I could find 23 in any test instruction.
24 I considered that to be a PRC. And, in 25 combination with the other less important things that I
55 I felt were wrong with the switch, I decided that was enough 2 to write a PRC. So I did, actually.
I BY MR. MC'INTYRE:
4 O That issue was part of the evaluation in the j 5 PRC, wasn't it, with the desynchronization?
8 A Yes; they decided that the reactor mode 7
switch did not have any safety significance. I think that 8
is the way they, in a nutshell -- they had a two- or three-page letter that made the case for that. I had to 10 accept it because those fellows know a lot more about the 11 systems than I did, about the control systems and what 12 they controlled.
II Were you aware at the time that MR. PETTIS:
14 there were procedures in effect such that, if you were not 15 sa tisfied with the results of the PRC evaluation, you 16 could, then, resubmit it to the safety and licensing?
I7 A No; I don't believe I was aware of that. I
'I am not sure that GE had any such procedures. Maybe they I'
did, but I was not ware of them and I am still not. I do i
20 know that, at the time, I was struggling real hard with a
' lot of these issues, and, you know, I just did not have the resources lef t to pursue that any further. In fact, I 23 got a bad appraisal for pursuing it after I should have 24 stopped pursuing it. The issue was still open when I was 25 transferred from production following up to technical
55 1
licensing.
2 One of the comments that Reghitto made on my 5
first appraisal was that, I was, I think he used the word 4
" preoccupied" or " overly occupied."
I think he said 5
_ preoccupied with concerns outside of the unit. And I 6
asked him what he meant by that. He indicated that I 7
should quit- pursuing this mode switch thing and - get on 8
with the job that I had been assigned to do in the new 9
job.
So I was under some pressure from my current manager 10 to leave it alone.
11 I did pursue it af ter we had a design review 12 meeting, a couple of problem review meetings, and done 13 some design changes on the way the switch was assembled 14 and they way it was supported that I did not think were 15 adequate, necessa rily. But I was outvoted by the design 16 review team. So there wasn't too much more that I could do beyond that.
18 BY MR. MC INTYRE:
19 Q Were you aware of the fact that, in the end, 20 they did make some modifications?
21 A
22 Yes; as I recall, they used a thicker gauge of metal. Are you familiar with how the mode switch goes 28 together?
24 It is assembled, it purchased 25 is as something that they call a CAM switch, a nonsafety related
57
' unit. And then it ~ gets assembled into a box, a metal box.
2 And they. call that assembly the mode switch. And that assembly' is considered to be a nuclear safety related 4
assembly.
One of the things they changed was the 6
thickness of the metal that is used on one portion of that 7
box. Because there were little fasteners called avdels.
8 It is sort of like a routed nut. You crimp it into the 9
metal and that gives you some threads to put a screw in.
10 But the metal gauge that we were using was a little thin and these avdels had a tendency to spin. They 12 were not crimping the metal. One of the changes was to use thicker metal.
14 They also made a change in how the box was 15 held onto the switch. And I don't remember the details of 16 what they changed. But the result was that it was no 17 longer necessary to remove the screws from the front of 18 the switch. Originally, the front surface of the metal 19 box was attached with the screws that held the switch 20 together. And that was the problem with the 21 desychronization. Because you had to take these screws 22 out, which loosened the front of the switch and we would 23 put the box on and you would put longer screws back into 24 the switch.
25 The way you desynchronize the contacts was
- r. :
I to loosen the screws, pull -the f ront surf ace a little, and '
2 click it; then, put the surface back on. And that would I cause the contacts on the inside to not be in the right 4 position anymore. That sequence is very similar to the 5 sequence that might happen wh'en you were putting the front 4 of the box on the switch. Because you take the same 7 screws loose. - That is why I was concerned about it.
8 So one of the changes they made was to
' redesign the box so that you did not have to take ' the 10 screws loose in order to put the front of the box on. But II I don't remember how they made that change. That was a 12 good change.
13 Q They did this on -- what, present and all 14 future designs?
15 A Yes. I don't know _if there was any 16 in te ntio n of retrofitting it to all designs; although, 17 that is one of the things that I said that I thought 38 should happen. They changed the way they supported the I' rear of the box, also, on the newer designs. They did not 20 change it on Kuo Sheng and Cofrendes. 'And I think they 21 should have.
22 The reason is: the neck of the switch, 23 where it went through the panel, was a rather thin truss 24 section. And I saw one switch broken during the assembly 25 process. The assembler had it struck through the hole i
59 1 and they just kind of lifted the back and snapped the 2 necks. It was fairly f ra gile. And that neck was being 3 used to support the. front of the switch. I did not think 4 it was seismically a good idea. And they changed the way 5 they supported the switch.
6 But, even af ter they made the change, they 7 did not change the way they supported it on the Kuo Sheng 8 and Cofrendes. I don't remember where they are. I think 9 the Cofrendes is in Spain, and the Kuo Sheng somewhere in 10 the Far East. I don't remember now.
11 But I wasn't entirely sa tisfied with the 12 resolution of the mode switch problems. But I had to drop 13 it because I was in a new job and the new boss did not 14 want me to keep pursuing things from the old job.
15 //
16 //
17 //
18 //
19 //
20 //
21 //
22 //
25 //
24 //
25 //
..-w ,- e- ~,,----,,
c.
4 60 1 BY.MR. McINTYRE:
2 Q -So in the end, moving orer to the new group 3 really. killed any effort that you had in trying to resolve 4 that.
5 A I think that backfired on them. They moved me 6 out of that project support job into what I believe they 7 considered would be a safe place for me. I think they were 8 trying to get me away from where the panels were being
-9 manufactured because I was creating a lot of ruckus there.
10 The reason it backfired was that they put me into an area 11 that was just as volatile because we were providing 12 technical support through the licensing people and legally
. 33 it was no better off than the panel shop I'd been. So I 14 think they tried to mo ve me into a safe location 15 and just did not foresee how it was going to work out.
16 That move wasn't my idea by the way. They suggested that I 17 move. What happened was -- it is attributed to some of 18 the friends that I had. A lady I knew named Sophia Renda 19 came to me one day and she said that they thought they were 20 going to find a new job for me because somebody she knew
. 21 heard Bill Marklein talking to --
in the hallway -- and 22 said that we ought to find Milam a new shop. Review needs 23 a new job and to see if Reghitto would like to have him.
24 And she said based on that little thing that we overheard 25 that maybe they were going to suggest that you go work for
61 1 Reghitto. And.I said -- well, that is interesting. And 2 about three days later Lee Cobler came to me and said --
3 Sam, I think we've found ' a real opportunity for you. There 4 is a' job opening up with Reghitto in the licensing organi-5 zation; I think it really would fit your skills well. And 6 I said -- well; and I tried to pin him down; I tried to get 7 him to tell me whether he wanted me to leave or not, and he
-3 would not say whether he wanted me to leave or not. He~just 9 said he thought it was an opportunity for me, so I left.
10 But it was arranged; I never even applied for that job .
l gg through the posting system. It was just offered to me.
12 And I had been applying for jobs and hadn't got anyone that 33 I applied for, and then that one fell in my lap.
34 BY MR. PETTIS:
l 15 0 Are you aware of any institutes in which a J 16 potentially reportable condition was not properly evaluated 17 by G.E. a s required under 10 C.F.R. 21, and also G.E.
1 13 procedures?
19 A I think we could probably make a case fo.r 20 that since, as I understand it, the potentially reportable 21 condition is not just a safety defect and a piece of 22 hardware, but also a violation of any rule or regulation 23 procedure or requirement, et cetera; it's all required.
24 Remember the exact word ing of it, but it says something to 25 that effect. I think using an unverified or -- well, I
s 62 I think _he's using unverified documents to support the FSAR, 2 and the support things.that we sent to customers are things
, 3 that we sent to licensing might.be sufficiently fraudulent 4 to be considered a reportable condition.
5 Q Could you be a little more specific?
6 A When we compared FSAR's to --
that were 7 designed -documents -- and found some disagreement, we 3 needed to resolve that disagreement. And depending on 9 which document needed to be changed, we would either to propose a change to the text or the FSAR, or we would make 11 a change to our design documents. The way that we would 12 make a change to the design document was with an 13 engineering change request. It was written on an ECN form.
14 We would send the engineering change request to the 15 responsible. engineer as request for a change. We would 16 send a copy of it, or maybe just the number of it, to the 17 licensing people with our statement that the design 13 documents did, in fact, support the FSAR. The licensing 19 people would take that statement and the engineering change-20 request and go to whoever they needed to go to in the NRC, 21 or a customer, and say -- ye s , the design documents do 22 support the FSAR.
23 Meanwhile, the engineering change request was 4
24 not tracked by any kind of system -- did not go anywhere 25 retrievable -- and just went to the engineer, and he could
~
r w
63 1 either. do it, or do some of it, or throw it away. And we 2 never told licensing that. Now, I told Bob ~ Strong that 3 once in licensing, but it was unofficial; it was just a 4 conversation I had, and-he was perplexed by it, but I am
~
5 not aware that he ever followed up on it, so I think that 6 is the case potentially -- making a false statement, saying 7 that the design supported the FS AR when, in f act, the 3 design change may never have happened and we did not follow
-9 up on it. Now, I don't know, maybe you can make a case of go that being a reportable condition, I don't know.
gg There were times with the answers to custom-12 ers' questions and FSAR refused where we used NEDO-type 33 documents to support things. Now, those were like public-14 documents or licensing typical reports -- that sort of 15 thing. They were not controlling documents; they were 16 descriptive documents. Nothing was made according to those 17 documents, or inspected according to those docunients. They 33 did no t have any kind of defined system of ef ficient 19 control or any kind of review or issue cycle, but they were 20 used. Licensing typical report was a good example. They 21 were used to say that things were a certain way, that 22 certain design features were there, that certain statements 23 in the FSAR were supported, and I don't think that was 24 proper either. I know that if we used an unverified 25 document -- and by that I mean a document with a deferred
g .
64' 1 ve rification -- if we used an unverified document in 2 support '- of any of these statements, we did not. tell the t.
3 licensing that.it was unverified.. We just let them assume l
4 that iti was , in fact, a verified design; and I don't think L
5 that that was right either.
b .6 I am-not sure that these circumstances are the 7 kind of answers you wanted to that question. If they are 3 not, ask the question again and I will try to give you a
, different kind of answer.
10 Q I was just hoping that you could be more l
specific with the particular instance or activity, or 33 12 s mething a little bit more definitive, g3 A Could you give me a little more help abotft 14 what kind of thing you are looking for?
15 0 Something that might be related to a specific
'e 16 plant, or something that may be related. t8 alc;pecific
(
17 safety-related component or piece.of equipment.
33 A It is real difficult to remember which problem 19 went with which plant. I was involved with a lot of plants 20 -- a whole bunch of plants -- and the problem is that the 21 design documents all looked very much alike from one plant /
- 22 to the next. All o f the problems that existed on one 23 existed on several because they had all been made from some 24 earlier different one, and after this length of time I just 25 can't remember which plant or which system had which
p - s 65 A
There were too many different onei I worked with
~
1 problem.
2 and most ways they were too similar. So I just can't do
~
3 that without actually going back into the record and try to 4 find some particular example that I might have recorded.
5 That would be the only way I can remember that. I can only 6 remember the kinds of things that happened in a generic 7 sense.
8 MR. PETTIS: Do you have anything further, Mr.
9 McIntyre?
10 MR. McINTYRE: No, not without getting into 11 specifics, but that would be hard for you to recall. Were 12 we to do that, I would have to spend some time searching 13 into the work record to find whatever specific thing you 14 were asking about, and it might be reasonable to do that, 15 but I don't happen to have my work record with me right 16 now.
17 MR. CLARK: Are you willing to do that?
18 THE WITNESS: I don't mind doing it if it is 19 useful. I am not sure that it would be useful since what I 20 am talking about is the generic misbehavior of the system.
21 I told Al I never intended to try to point to some specifiu 22 instance of a defect and say that it was in any way more 23 important than any other instance of a similar defect. I 24 am looking at it from the point of view that the system did 25 not work right to me -- that is what is important. The
..m -
a
'66
- ) 1 1 5,:
1 system did not work right; it routinely did not work right:
2 for years and years it did not work right, and'nobody was 3 able to change that. To me, that is what is important; and 4 that fact that my efforts to change it and, my opinion, 5 resulted in me losing my job there and ending my career in 6 the-nuclear industry, which is an industry I care for quite 7 a lot. I think if you want specific examples, pick a plant 3 -- pick a system -- and you will probably find specific 9 examples.
10 MR. PETTIS: Well, we would need something 11 from you that was highlighted in your work record that 12 resulted in a specific plant panel involved in a safety-13 related-type application. We are basically concerned with 14 only safety-related-type equipment.
15 THE WITNESS: Well, not all of my work was 16 with safety-related-type equipment. But the fact is the 17 same design control system was used regardless of whether is it was safety or i. *-safety-related equipment. So any 19 generic melfunction in the system would affecte either 20 kind of equipment. We did not have a separate design 21 folder for non-safety-related equipment. So I would 22 assume that any malfunction that I was aware of would 23 eventually come to bear on some safety-related system or 24 component.
25 //
67 1 BY MR. PETTIS:
2 O Would it be correct to say that your concerns
. 3 are basically general?
4 A- Generic. I guess general is a good word.
5 BY MR. McINTYRE:
6 Q Generic within the GE system you are talking 7 about?
8 A I'm not speaking for any other participant in 9 the nuclear industry. I'm speaking only for GE and only 10 for the San Jose facility. -Maybe in Wilmington they have a 11 wonderful system -- I don't know. The only place I ever 12 worked was the San Jose facility and in fact it seemed that 13 C& I was probably worse than 10th Street people. I don't 14 know why but, before I went to C & I, I had heard of the 15 Place and it had a reputation as being the most screwed up 16 part of the whole company. And when I went there I came to 17 believe that that was, in fact, true. That was the way I 18 perceived it after I had been there awhile, but I don't 19 know why.
20 Q They were more production-oriented?
21 A It would be a shame to think that being 22 production-oriented necessarily means you have to be 23 screwed up.
24 M R. PETTIS: Well, if there is no other 25 comments or questions from any of the other members of the
7_
68 1 team --
2 MR. McINTYRE: Do you have anything you want 3 to ask, Mr. Prescott?
4 BY MR. PRESCOTT:
5 Q The only problem I-have ---did they tell you 6 that -- did a team come back and tell you that this PRC for 7 the reactor mode switch was closed out? Were you aware 8 that they had closed that out formally, and what they had' 9 done to close it out?
10 A I was notified by letter from Barrentine even-11 tually that it had been closed out. I don't remember the 12 date of the letter, but it was a good many months after I 13 wrote the PRC. I, perhaps naively at the time, thought 14 that a PRC had to be answered within a matter of days.
15 There is some provision in the Code of Federal Regulations, 16 I guess, and also in the GE procedures. I don't remember; 17 I think it is something like two to three days. But it 18 turned out they had a way of interpreting that. I guess 19 that they had to respond within a certain number of days 20_ after having decided that it was reportable or something 21 and they could take years to decide that if they wanted.
22 There was some kind of an interpretation that gave them, 23 what appeared to me, a loophole that was in conflict with 24 the spirit of the thing, but they did not seem to feel that 25 it was necessary to respond within any defined length of
- ~ . . -
4 69 1 time after'I sent my letter.
2 Q Were you aware that they made some design 3 changes to the actual switch?
4 A Yes; I was in on the design review team that 5 made those changes.
6 Q Remember making that design change part of 7 your work record showing?
8 A' Yes; it's there.
9 Q It is in your work record?
10 A My thoughts on the meeting are a part of the 11 record. I wrote that down in some detail. The easiest 12 place to find it would probably be in the . volume of the 13 reactor mode switch, but you could find it in -- near 14- the end of'the Lee Cobler record, and the beginning of the 15 Reghitto record , if you wanted to look in those volumes.
16 But you would find it a lot easier to follow the history of 17 that issue if you looked in the mode switch volume because 18 I have taken all of the pages from the record, put them in 19 one binder and drew a circle around the item number on each 20 page of the item that is a part of the mode switch'chron-21 ology. It would just be a lot easier to follow because, 22 often, one item on a page of half a dozen or so items would 23 relate to the mode switch.
24 BY MR. PETTIS:
25 0 .Do you feel the mode switch is the most pm ._ . _ _ . _ _ _ _ ..---_--.__ __ -
w .-
70
- g. significant part of your work record?
A No;'I.think the deferred' verifications.is, and 3
.that the' deferred' verifications affects everything. ~It is
~
3 -
4 either the philosophy or deferred verifications L got down ~
5 into the production area too, but I did not recognize it at g the time because I had not thought about' deferred verifica-7 tions.then, but we had things we call conditional shipping g releases' that we were supposed to us.e. I gues s I wa s
-, becoming aware of_ deferred verifications even while I was 10 in the production area, and I began to worry about whether gg or not we needed a conditional shipping list for things 12 that were one _way or.another incomplete a'nd did not ~have g3 something finished that they needed to have finished. But 14 I did'not have it as well-defined . in my head then, but 15 lo king back on it I can see that the doctrine' of not 16 completing something before you ' ship it was, in effect, 17 down in that a ssembly shop as well.
33 The doctrine is somewhat pervasive. You don't '.
39 have to complete anything before you are finished with it.
20 You just do what you have time to do, or what you feel like 23 doing, or what you are capable of doing -- whatever the l
33 circumstances that causes it to not be complete -- and then 28 issue it or ship it just to meet the schedule. The billing
- i. 24 schedules were very important, that is an issue.
25 Dates were very important with the drawings. But it is the i
I
~
V _
71 1 same doctrine -- to meet the schedule.
2 BY MR. McINTYRE:
3 0 What did you call that -- conditional? ,
4 A Conditional shipping release. It was supposed 5 to define what was wrong, or what wasn't there,. or what 6 wasn't complete. It was another one of those systems that 7 you show an auditor to demonstrate that you are controlling 3 something. I imagine the DVSCN was probably generated for 9 the benefit of some auditor to show that we were control-10 ling deferred verifications and then just never used. That 13 was kind of a typical way of doing business. In fact, 12 indeed, there would already be a procedure in place, but 33 somebody did not know-about it somewhat, or would catch 14 them doing something they should not do, so they would just 15 wr2te a procedure, and then they would have people 16 violating two procedures instead of one.
17 BY MR. McINTYRE:
13 Q- I have one other question going back to your 19' early years in I & C, where you said you did not 20 have your own copy of the operating procedures -- the EOPs.
21 I was just wondering -- was that common practice for all of 22 the engineers?
23 A' Certainly. Very few people had any EOPs. In 24 fact, there were a lot of other procedures manuals. I 25 never was able to learn of all of the ones that existed.
, - . - ~ - . -
72 1 Toward the end I started .trying to. collect pr'ocedures 2 manuals.. There was a quality assurance procedures manuali
, 3 manufacturing procedures manual and: manufacturing standard 4 procedures. There was an engineering internal procedures 5 document, a drafting procedures manual and EOP, an NEBG 6 procedures manual, and I don't know what all else. I would 7 occasionally'see one on somebody's bookshelf and ask what
~
3 it wa s. I f ound out once that it wa s some licensing
-9 procedures manual and I tried to get a copy of that, but I 10 don't believe I ever did. There were procedures manuals
- Il everywhere.
12 Q But the engineers did not have control 13- procedures of?
14 .A Not at their desk, n o. They were usually be 15 one of the. areas. somewhere -- usually in the manager's 16 office. -
17 0 'When you first would come on-board with this is unit, would you be required to sit down and read through 19 all of these procedures and, say,-sign off on them?
20 A No; never. In fact, when I moved into the 21 technical licensing unit, I went looking for the EOP and
- 22 f ound it in the bos s's of fic e. And if you look in the 23 first - probably the first dozen or so pages of the 24 Reghitto record -- maybe the first twenty pages of the 25 Reghitto record, you are going to find where I made a note t
,-- f -,, -
4 73
~
1 of how many revisions of~the EOP were missing in his. It 2 seems to me that there were-probably about ten revisions of 3 it that were not-there -- had not been inserted -- and I 4 had to get them myself and insert them into the manual for 5 him. So nobody was even paying any attention to it. Nobody
~
6 was even. bothering to insert revisions. I made a record of 7 that. Eventually I volunteered to take over maintaining 3 that manual, which was basically a secretarial job.
p Normally it was 'some clerk or secretary who was delegated 10 the responsibility for just inserting revisions when they 33 came around, but nobody was doing it; nobody cared, and I 12 needed an up-to'date manual so I just told them I would be 33 glad to do this if you want me to, and they said okay. So 14 after that they routed all of the updates to me, and I kept 15 their manua1s up-to-date for them.
16 MR. PETTIS: Did you two have anything else 17 that you would like to discus s tha t maybe we haven't is covered that is foremost in your work record?
19 MR. McINTYRE: What you feel is important.
20 THE WITNESS: I don't know. I guess we could 21 Probably sit and talk about this for hours. There were all 32 kinds of incidents that happened, but I think they probably 23 revolve around the same general principles that we have a
24 already covered. In five years a lot of things can happen.
25 //
74 1- BY MR. McINTYRE:
2 Q Could you see any positive changes within the 3 five years in this system?
4 A Not really.
5 THE WITNESS: Do I need to be careful about 6 quoting four-letter words here?
7 MR. CLARK: No; feel free.
3 THE WITNESS: I don't know if I related to you 9 what Barrentine said to me in the hallway one day, did I?
10 MR. McINTYRE: No.
13 THE WITNESS: It is just an amusing statement 12 to look back on I guess.
33 BY MR. McINTYRE:
14 Q Barrentine is your --
l i 15 A He is the section level manager about three 16 levels above me I guess. I think I told you this story 17 after I -- Diane Taylor and I had written to PRC jointly 13 and simultaneously gave it to our managers--
19 Q On your reactor mode switches are you talking 20 about?
23 A Yes. She was in QA, I was in engineering, and 22 we weren't really supposed to be cooperating, but s'he knew 23 a lot about the QA record center and test instructions that 24 were relevant to the issue, and I knew a lot about the 25 engineering documentation. So on this particular issue,
u 75 1 our knowledge was sort of complementary to one another. So 2 we' jointly researched this in each general-area of 3 expertise and we.put together this PRC , and wrote it. .
4 But a few days later, Barrentine stopped me in the hallway.
5 and I came close to quoting what he said' -- maybe not 6 exactly, but he said, " Sam, you have demeaned yourself as 7 an engineer by even allowing these-QA pricks to be involved 3 in your work," and that is. sort of the attitude that was 9 coming down to us I guess. That! ' because I cooperated with 10 a woman in Q A to write a PRC. It can ha ve a chilling 11 effect on a lot of things -- like cooperating with QA.
12 BY MR. PETTIS:
13 Q There was a statement in your work record that 14 indicated testing of panels to elementary diagrams.
15 A Th a t 's an interesting area. I'm glad you 16 mentioned that.
17 Q Would you kind of expound on that a little is bit.
19 A Well, it was fairly common to have very 20 similar elementary diagrams for different projects. And 21 in order to save draf ting man-hours, make one connection 22 diagram to satisfy more than one elementary di.agram. Now, 23 the connection diagram shows where to run the '< ires point 24 to point within the panels in order to achieve the circuit 25 functions that are specified on the elementary diagram. So
76 1 if you'had elementary diagrams with variations on the same 2 circuit, you end up with~ a connection diagram with 3 variations on the wiring. And you can't show that on one 4 connection diagr am.
5 I found one particular example of a connection 6 diagram that -applied to TVA and Grand Gulf which, when you 7 take it all together, there is at least half dozen Grand 8 Gulfs and a half-a-dozen TVAs, and they had revised 9 this . connection diagram back and forth depending on which 10 plant they happened to be building at the time so that it 11 would satisfy the appropriate elementary diagram. In other 12 words, it had to agree with the TVA one month, and then 13 then would revise it to agree with the Grand Gulf elemen-14 tary so they could build Grand Gulf and then they would 15 revise it back to agree with TVA again because they had one 16 connection diagrara and two elernentaries. Well, there were 17 other consequences of this.
18 One consequence to this is that if the connec-19 tion diagram -- well, there was at least one case where I 20 know that they tested a panel to the wrong elementary 21 diagram because of the panel, passed all of the functional 22 tests, and the connection diagram did not agree with the 23 elementary diagram for the panel; it agreed with the 24 elementary diagram for the other panel, f or the other 25 project. So the fact that the panel passed meant that the
~
.1 l
'l
- l 77
)
I wiring agreed with the one they tested it to, right? So 2 they could not have tested - it to the proper elementary
. 3 diagram or it would have. failed. I found the opposite --
4 that on one panel they could not get the panel to passnand 5 that it turned out they were testing it to the wrong 6 elementary diagram, and they did not-know it. The tester 7 did not know he was using the wrong elementary diagram. I 8 don't know what you say about a tester . who doesn't know 9 which element or diagram to use -- I guess he did not check 10 the number or something.
11 Q Did they have a procedure written by GE?
12 A They had a test instruction. They had a test 13 instruction that listed the number of elementary diagrams.
14 This particular test instruction was to both numbers -- did 15 not say which number was for which project. And he just 16 happened to use the first'one on the list, .and it just 17 happened to be the wrong one. And he'd read down on the is title block where it gave the name of the project on the 19 drawing. He could not tell that from the test instruction, 20 but he could have told it from the drawing but he did look 21 at the title block; he just started testing. So that.
22 situation led to a lot of interesting consequences.
23 Q So what you are saying is that a panel for 24 TV A could have been tested to a Grand Gulf elementary 25 diagram?
t
% 5 78 1 A Yes.
l l
2 Q -And it would have been apparent'if the tester j l
3 had maybe looked down at the bottom of the title block and
-4 saw that that elementary diagram was for the wrong plant?
5 A Yes. Of course, maybe he did not know which 6 Plant the panel went with. I saw instances where panels 7 had the wrong tags on them and production control did-not 3 e ve n -- in f act, that was other game that production 9 control would play. They had something called a 10 inventory transfer notice that I supposed had originally 33 been intended to transfer bolts from one stock number to 12 another stock number when the bolts were identical and they 33 had been bought from different supplies or something. It 14 was for transferring that kind of inventory. They were 15 using this inventory transfer list to transfer panels from 16 one party to another just because both panels happened to
. 17 be a G36P2. If they happened to need a Hope Creek G36P2 33 and they happened to have another one sitting there in the 19 fabshop, the y would just do an inventory transfer notice 20 because it was a G36P2. But they may not look the same.
21 They may have different cutouts or different arrangements 22 or things, but they would do it anyway, and it would become 33 a problem downstream tha: I would have to solve. It took 24 me a long time to realize that that was happening. It 25 explains a lot of strange problems I encountered where 4
e w -'-
y -es- e + - , . -- w -
m a ,- rr
c -
79 1 things did not add up right and I could see no reason why 2 it did not match the enclosure drawing. Things were not 3 done in the right place, and I understood eventually.it was 4 because they probably did not start out as a Hope Creek 5 panel. It was proved to be an enclosure drawing to some 6 different project, so that was another peculiar thing that 7 used to happen and, as I told you, production control often 8 did not even know.where the panels were.
9 I re, member one panel -- I don't remember which 10 project it was -- they told me that it was in assembly and 11 paint shop for a final touch up, and I finally found it on 12 the fab shop, laying on its front being welded together, 13 and I don't know what panel they thought -- what panel had 14 actually gone through into the paint shop had been shipped, 15 but it wasn't that one.
16 Q Could you explain what panel we are talking 17 about?
13 A It could have been any panel -- any control 19 panel or any local or remote panel. It just -- the 20 inability to know where a panel was had nothing to do with 21 what kind of panel it was ; it seems to just be the way 22- their system worked.
23 Q Just for our discussion here, are we talking 24 about safety-related panels?
25 A Well, it might have been. Like I said, any
=
80 1 aspect of this system would apply equally well to either a 2 safety or a non-safety; in fact, most panels had something 3 in them that was safety-related. I would say almost any 4 panel would have something in it because we had -- well, we 5 had divisional-wiring in almost every panel. I don't know 6 if there's any panel that had only one division of wiring, 7 so there was always at least divisional separation require-3 ments, and probably at least one nuclear safety-related 9 something on the panel. It was very -- a lot af what we 10 had was nuclear safety-related.
11 0 So what you are saying is that the panels 12 manufactured were all manufactured under the same QA, QC 13 program?-
14 A Right.
15 Q So there was no identification as to -- if a 16 Panel was on its way to a fossil power plant versus nuclear 17 power plant?
13 A Well, all we made was nuclear panels, but 19 there wasn't a separate system of control for whether it 20 was a safety-related panel or not. They were all treated 21 Pretty much the same. At l e a s .t I am not aware of any 22 dif ference in the system; not any documented difference in 23 the way that they were handled.
24 Q So it is your understanding that the 25 activities conducted during this time were strictly for
81~
-1 nuclear power. plants?
2 A Well, the only thing we made was nuclear and-3 nuclear panels. There was one point where George Senn was 4 talking about getting into some other kind of industry, but 5 I don't. believe it ever happened. The only thing I am
-6 aware of that we ever made was nuclear panels.
7 Q So there is a possibility that other panels, 3 non-nuclear related, are manufactured at a different GE 9 facility?
10 A Not by the nuclear energy division. The only yy business we were in was the nuclear business, and we were 12 not making washing machines or anything. It was strictly a 13 nuclear energy division, or the nuclear energy business 14 group, depending on the point in time the name changed; but 15 it was always under the nuclear busin'ss. And that's all I 16 did; that's all we did -- nuclear work.
17 BY MR. McINTYRE:
13 Q What kind of a system did they have in place 19 down on the floor to follow the panels where you said that 20 they would not know where they were?
21 A I'm not to familiar with how production 22 control kept up with where things were. I don't know how 23 they did that.
24 Q How would you go -- when you said you would 25 locate -- how would you go about finding them?
l 82
-1 A Well, I would just go a round phys ically 2 searching. If I could not find one, I would go look for
, 3 them.
4 0 So you would not go through a tracking system?
5 A No; I would just -- the panels had little 6 tags, little metal tags on the bottom down on the inside 7 that usually told you the name and project of the panel, 3 but sometimes they were wrong. And there would also be 9 people working on the side of the panel, or a label stuck to on the' panels somewhere. Any you eventually got where you 33 recognized the panels. I know that during my la's t month 12 when I was -- after I had received my layoff notice and I 33 was spending most of my time over in the Placement Center 14 on Pullman Way, one day I was tired of writing resumes and 15 filling out. job applications, and I went walking around through 16 the back lot of the building, just to relax, and they 17 had a bunch of old panels stored back there, and I was 13 walking through this row of panels and suddenly I stopped 19 and I said I recognize that panel, and I went around, I 20 looked in the back and I said yes, there are things I 21 change; this is one of my petals, and you got familiar with 22 them -- you learn to recognize your own panels. They were 23 each individual, so you know that is how you find them; you 24 just know what you'are looking for. I almost cried when I 25 saw that panel there because I worked so hard getting ready
~- -
83-1 to ship, and there it was'in storage. They terminated the 2 project or something,-and'it never'got shipped.
3 0 Were you ever involved with. any internal GE 4 audits in, say, the years of your work record?
-5 A No; I never had the opportunity to do any 6 auditing except just on my own. I was never part of any 7 kind of audit team or audit function.
8 Q so you don't recall any audits of your areas 9 turning up any other concerns that you had?
10 A There were occasional audits; I don't remember 11 any of them turning up. And of my particular concerns I do 12 remember once, toward the end, some fellow from the Tenth 13 Street QA Group called me. He had been referred to me by 14 Al Barkley under whom I'd taken a class -- I -don't remember 15 the guy's name now, but he was going in the auditing of the 16 licensing organization and he heard of me and he wanted to 17 talk to me about things. And he called me and asked me a 18 lot of questions and I told him a lot of things, and I 19 never heard from him again. . I don't know if he went over 20 and ordered a licensing or not, but that was the only time
-21 I ever came close to being a part of an audit.
22 Q That was licensing, and you don't recall any 23 in the C & I?
24 A The problem is that everybody knew what the 25 problems were. It would not have been a matter of an audit
84 I uncovering something nobody knew about.
2 .Q It would be documented though.
3 A Well, yo u s ee, those guys are experts in--
4 providing responses-to that sort of thing. That is'how a 5 lot of ' those procedures came to exist. They were just 6 written to resolve an audit finding, but it doesn't' change 7 the way you do business. Your' right o f procedure to-8 resolve an audit finding and you promise to follow it and 9 the auditor'goes away, and the system of procedures is so 10 complex and-there are so many of them that it is almost it impossible for anyb'ody.to go in an audit and really prove 12- anything. You can always go into-a procedure that you are 13 following -- that you are not following.
14 GE is really an expert at surviving 15 audits, snowing auditors, blowing smoke in front of them.
16 They are really very good at it. If they were as good as 17 that as -- if they were as good at running a design control 13 system as they are at surviving audits, there would not be 19 any reason to audit them.
20 Q So your biggest concern was the design to 21 def erred verifications systems I gues s if you had to 22 pinpoint it down.
23 A I guess that would be my biggest concern --
24 that, and the fact that it sort of graphically illustrates 25 a philosophy or a doctrine of doing only what you have to
-85
~
1 do to get the thing issued or shipped, pretending that you 2 follow up and do the rest later, and then dividing some 3 complex system for allowing it to get lost. That way of 4 doing th'ings ' kind of . recurs throughout the whole design 5 control system. And it is going to be very difficult.for 6 somebody from,the outside to be able to penetrate that dock 7 and document it.
-3 0 '
~It is a complex system?
9 A .And do.anything about it. And the way they do 10 those things isn't the way they will tell you. Barrentine 11 used to have all hands meetings and talk to us about how it 12 was nececsary that we followed procedures. But I' guess 13 nobody took him seriously.
What they say and what they do 14 are two -- well, rather different things.
15 MR. PETTIS : I would like to thank.Mr. Sam 16 Milam and his legal counsel, Mr. Lewis Clark from 17 Government Accountability Project, and the other members of 18 the NRC team for their contribution during this interview.
19 We would like to reserve the right in the future to have a 20 follow-up meeting if necessary.
21 MR. CLARK: Can I add something to the record.
22 Just as we discussed prior to the beginning of this -- what 23 was said -- at the beginning of the session before it was 24 on the record, we would like to request that the transcript 25 of this meeting not be sent to GE.
86 1 .THE WITNESS: And'I guess:we want:a' chance.to.
-2 review it..
-3 MR. CLARK: And for--th'e record we would'also 4 like an opportunity to see the complete transcript and'make 5 any-corrections that might be necessary, and make it more 6 accurate.
7 M R. PETTIS: Then that concludes the
'S interview.
9 10 11 .
12 13 14 15 16 17 18 19 20 21 D
23 24 25
g This is to certify that the attached procosdinga bofore tho
. UNITED STATES NUCLEAR REGULATORY COMMISSION in the matter of:
NAME OF PROCEEDING: INVESTIGATIVE LNTERVIEW (CLOSEQ fiEETING)
DOCKET NO.: NONE
' PLACE: sam JOSE, CALIFORNIA . ,
16 APRIL 1986
,. 'DATE:
waro held as herein appears, and that this is the original transcript thereof for the file of the United States Nuclear ,
~
-Regulatory Commission. '
I (sigt). 3 , N J (TYP,ED) }AMES W. H15G/9S)
Official Reporter l
l Reporter's Affiliation l
S e
4 g
- e 6
9 e
. V
'l ENCLOSURE 6 LIST OF PLANTS MENTIONED
- / .
Y ENCLOSURE 6 List of Plants Mentioned Listed below are the foreign and domestic nuclear power plants mentioned by Sam A. Milam III in his work record for the time period March 1978 through I April.1982 while employed at GE, San Jose. The references by Milam to these -_.
nuclear power plants do not identify specific defects in safety-related or non safety-related systems or equipment.
Foreign Nuclear Plants Alto Lazio Cofrentes CNY 1 & 2 Kuosheng 1 & 2 Leibstadt Domestic Nuclear Plants Allens Creek Black Fox Clinton 1 & 2 Fermi 2 Grand Gulf Hanford Hartsville Hatch Hope Creek 1 & 2 LaSalle Limerick Monticello Nine Mile Point 2 Perry 1 & 2 River Bend 1 & 2 Shoreham Skagit 1 & 2 Standard Plant Susquehanna 1 & 2 TVA (x17 - x22)
WPPSS 2 Zinner
i ENCLOSURE 7 1983 NRC (REGION IV) INSPECTION REPORT J