ML20113D238

Transcript of C Strokes 831207 Interview Re Seismic Restraints

ML20113D238
Person / Time
Site:	Diablo Canyon
Issue date:	12/07/1983
From:	Stokes C PACIFIC GAS & ELECTRIC CO.
To:
Shared Package
ML20105B784	List: ML20113D238 ML20113D319 ML20113D327
References
FOIA-84-21 NUDOCS 8501230022
Download: ML20113D238 (121)

Text

,

. . s l l

l CHARLES STOKES /12-7-83 s

DUKOFF: Good morning. Today's date is December 7,1983.

~

The time is approximately 8:25 a.m. This will be the interview of Charles Stokes, being conducted at the Diablo Canyon Nuclear

. Power Plant. Mr. Stokes is represented by Mr. Thomas Devine as his counsel and I. understand that you'are his counsel., Is that correct, Mr. Devine?

DEVINE: That is correct, sir.

DUKOFF: Thank you. For the NRC Technical St ff, present '

are Mr. Thomas Bishop, Mr. Jess Kruze, Mr. Esa Yin, Mr. Samuel i Reynolds, Jr., Mr. John Fair, Mr. Stuart Eonedder, Mr. Dennis Kirsch, and Mr. Marvin Mindonka. Representing the Office of Investigations are Mr. Eugene Power and myself Phillip DUKOFF.

At this point in time, I will. turn this over to Mr. Bishop, who will start the interview.

I have asked each of our staff members before t..ev make ,

their statements to provide their last name which I just didn't do, for the purpose of-if it's decided to transcribe these tapes, we would know who the speaker was. So, I'd just like to remind  ;

each of our members to do that. I think also tha't it's important that the principal speakers put up with the inconvenience of I

having to hold this microphone so that we can have a clearly i auditable record. The stated purpose of our effort this morning is to gather additional information. We have received your affidavit dated November, '83. We have had an opportunity to go

~

8501g0022BMW #

hDEs 21 l'

Stokes /12-7 1 l i-

through that and in the course of some of our follow-up work to this point, we do have some additional auestions we'd like to ask your assistance on. We would also seek any additional

. information that you might have that might help us in this regard and we'd appreciate anything along that line. Mr. Ebnedder is the team leader of this effort on the site. Do you have a particular format that you would like to follow as far as individuals go?

EBNEDDER: This is Ebnedder. Not particularly. We each have gone through the affidavit and have identified areas where we would like clarification. Some of these areas we would like you to also point out to us in the field. We will take you on a tour of Unit 1 and have you specifically identify areas that we can look at. I think we can just, we might go through the

, affidavit page by page, and then ask each staff member as we go by the page, if he has any questions in that particular area.

And I think that will be the easiest way to do it. Is that all right with you?

!  ??: Fine.

d

EBNEDDER

Staf f members, is that all right? Sam, Esa?

??: We don't all have copies of the ...

DEVINE:

While we're getting seated here there was one point that came up last night that I was a little concerned about and ,

i just wanted to check with you. The Mothers for Peace told me that they had received a call from the News Media asking if there would be a press conference surrounding Mr. Stokes' plant tour. ,

And we were a little bit concerned that the word of this plant i

Stokes /12-7 2

_ , - ._,-,._,e -. , , _ v.m. , __- ,.,,,. ,,_,. , c ._. .-c.,.-..r,-. _ . . _ - . - - . ,-.w. wy-1---.---

.. . . - - .. - - - _ = .. -

tour had.been announced in advance for the simple reason that

.he'd just as soon not have people ready for him when he comes in. And I wanted to inquire about that as well as whether the i affidavit has been shared with the applicant or Bechtel or any of the people that we'd be checking.

BISHOP ' The answer is no to both of those. We have not

provided any material-to any parties outside the NRC nor have we, I

to my knowledge, informed anybody other than security people just

! recently that we would like to have Mr. Stokes and yourself come 1

l on site and go on a tour with us. We've had no press discussion, 4

i no discuIssions with the press to my knowledge. I did discuss

! - this with Sandra Silver of the Mothers for Peace in trying to run

! you down over the last few days, but other than that, that's the i

only person outside the NRC in addition to the security people, which we had to line up for clearance to get on site.

DEVINE: This is Devine and you might want to check on that l then because we didn't and Ms. Silver was horrified that the

! press was asking her about it. And she assured me that she didn' t and asked the reporter.

He said he'd learned about it ll

from the NRC. So you might want to just check with how that t

! occurred. And I would refer you back to Ms. Silver.

BISHOP: Thank you.

l 1

CRUZE On this question of the affidavit, soon after we got I

here we did learn that the licensee had obtained a copy. I attehpted to find out how that might have happened. I understand they obtained it from media sources.

i i

i Stokes /12-7 3

- - - - - , - - - , , , - , , , ,,-,-ww- ,,,-4---,,,,- ,--~vw -,,,-wr .-.,-,v ,

w+--- es--,,n ~~ - - -w'-ww--* -- ,ne ---------v-- ---mver-~'w-~-w*~

i. DEVINE: I would think that they would have gotten it from the Department of Labor. That was my expectation. We only gave it to one media source and that was the day before the Commission received it and he told me that he wouldn't be sharing-it. That O-was kind of a condition of getting it. So, I would guess that f' they could've gotten it from the Department of Labor and that's '

unavoidable. I wasn' t sure how fast their system operated.

! BISHOP:. Okay, with that let's proceed into the technical session. Again, our general format we thought would be to go through the issues and ask our questions and seek any suppleme'ntary information you may have.1 Following the technical area the representatives from the Office of Investigation may 4

! have some additional auestions from their area or they may ask i

their questions while we're going on, whichever is most appropriate. Let me turn it over then to Mr. Ebnedder, Mr. Yin, .

and Mr. Reynolds. They will be the principal questioners and I'll turn it over.

1

??: Thank you. Let's go page by page and discuss topics as l we go. On page 1 is background on Mr. Stokes, his history and j qualifications as an engineer and lawyer so fortn. Does anyone l have any questions on page l? Page 2 is more data on Mr. Stokes, I

l background and the issue starts to appear of control documents.

That moves on into page 3 on control documents. Any questions on l

control documents?

??: Could I make a suggestion? If possible, you could sit in between Esa and me. It would be better. Then both of us I

could see that. l Stokes /12-7 4

- , , - ,,,&, , . - , . - . . . ~ , + . , . , - - . - . . .-,,,-.%%---c -,.v .-y- .-.,e,--.,,..-.e-

, ,---- y --... - - . ... . - . yw .- -

U BISHOP: This is Bishop again. If at any time we're talking 1 l

this morning and you feel you need a break or.want to get a cuo 4

of coffee, we do have some in the back on the tray there. Just feel f ree to say let's take a break. I recoonize it's stressful e  :

enough for you at this point without making things any worse. l i

1

??: Jess has instructed there will be no more -

interruptions. Any issues on control documents? Esa.

YIN: Esa Yin. On the affidavit, page 3, second paragraph. You stated the size of gaps between hardware controls the definition, such as whether an item qualifies as a restrain't. I would like Mr. Stokes to clarify what he means about this sentence.

i

! STOKES : This is Charles Stokes. By the size of gaps i between hardware controlling definition, what I was. intending to state was in every restraint, depending upon the amount of gap between the pipe and the restraint, it will be considered either i

j a restraint or a non-restraint. In the case of a lateral restraint, or dead weight support, if it's not bearing on the pipe, or within a sixteenth of an inch which is allowed in the l industry for thermal expansion, it's not considered a support.

l At Diablo Canyon if it was not within a sixteenth of an inch, it j was considered a non-restraint. In the case of anchors, it's a direct contact and it had to be welded. No gap, and therefore, l

! full restraint. That's what I meant by the size of the gaps l

l controlling the design or support conditions.

l YIN: Esa Yin. Mr. Stokes, do you are aware the present 1

l design of the seismic restraints allows a maximum of one-eighth inch gaps, and still qualify as a seismic restraint?

l

! S tokes /12-7 5

4 STOKES : That's true. This is ' Charles Stokes. Typically, that's supposed to justify one-sixteenth each side of the i

support. However, there are cases where the variance if plus or

, minus from zero on one side up to-an eighth maximum on the opposite' side. However, the overall gap on any one particular

. side supposedly under a seismic condition that could exist would

] be one-sixteenth on all sides and no more than an eighth 3 overall.

YIN: Esa Yin, again.- Based on the exoerience we see at the operation plants, during the initial thermal startup testings, not always the design movements matches the actual movements that j we see. One point. The second point is during the normal plant 1

j maintenance, when we check the snubber hot and cold position settings, those settings vary'from shakedown to shakedown. So as

! far as the design engineering is concerned, whether or not one 1

side is altogether one-eighth or one-sixteenth inch on each side, '

doesn't really make any difference as f ar as in calculation of the thermal movement as far as the design basis earthquake and l

operation basis earthquake is concerned. Do you agree?

STOKES: Um, it's true that the actual location within a support of a pipe will vary based on operation. The supports are considered installed at a zero operational time frame and l

installed with the tolerances that construction are instructed to install them to. Which typically, that it's sixteenth each i

side. Having known that this plant was operqting or had at least operated at hot testing, all the engineers here were aware that the actual gaps and the support would not be exactly one-d 4

i Stokes /12-7 6

. _ ._ . . . . - - . ~ - .

sixteenth or even close on any side. All the design that I worked on and anyone that I know of worked on here used an overall gap of an eighth. But if a gap already existed of an eighth according to the design manuals on this plant, that is not e

considered a restraint. In other words, we were considering our

work at a zero base in time design. Not after operation. If we had designed this plant strictly as though it was operating, then we would have taken an eighth. But we had to assume for design that it had never operated and that the gaps weren' t constructed according to the tolerances per the ESD 223. Which would have allowed'at most a thirty second over that. But we did not allow for that in the installation per our checks. The critical thinq f

! was to us, and as far as stress was concerned, if the pipe had moved during warm testing, and already had a new set, that is a permanent position assumed from that, then the new location justified where the hangar should have been to start with. And if the gaps were not correct at that time, to offer the

{ restraints per that condition, they weren't correct. L9 was the governing design document which said if the gap exceeded an eighth, or was an eighth, there was no restraint.

BISHOP: Just a point for our staff. I'd like to make sure our line of questioning does not take the frame of does Mr. Stokes agree or disagree with a particular technical point.

Our line of questioning should be along 'do you have further l information or clarification of what the statements mean?' We're

not here to ask him what his technical opinion is or put him in  !

that kind of pressure. It's primarily just to get additional j information. So, ... I l

i Stokes /12-7 7

4 STOKES : - This is Charles S tokes. I'd like to say somethina along those lines. When a I come on the plant site, the first controlling document I request in the design are control documents, design documents. Typically, any job I work on takes quality control two weeks to issue documents in a controlled manner to new personnel. Once I receive those documents I read through them, I determine'what codes and standards that the plant is supposed to be designed to. I also look at any project I supplied information to clarify. And any answer I give I will not give my personal feeling. I will alve you what I know to be hard fac'ts concerning the gaps, the design conditions, and other

things to consider which management did not want to consider.

DEVINE: I'd just like to add I appreciate your concern, l Mr. Bishop, but we don't feel any particular restraints on how intensively Mr. Stokes is questioned. The more that Mr. Yin puts him through the mill, the better chance we'll be that we didn' t miss anything today, no stone was left unturned. Every NRC i

engineer or investigator whose really been serious about the l assignment has just put my clients, pinned them up against the 4

wall, and they' ve passed the test. So we welcome all pressure he can offer.

l YIN: Esa Yin. If I understand correctly, what you are 1

l saying, Mr. Stokes, is the fact that what you have seen perhaps in some occasion that stress group was manipulating the stress level, trying not to take account of the thermal stress, thermal loadings rather. On some of the restr91nts, and also thermal j stresses on some of the piping. Because, in fact, if you really l

Stokes /12-7 8 I, _ . , , ~ _ . . _ _ _ _ _ _ _ , _ _ , _ _ _ , _ - _ _ _ . . _ . _ . ~ . _ , , - . _ _ _ _ . _ . _ . . _ _ , . _ _ _ , . . . . _ _ _ _ . _ _ . . _ _ _ _ . _ , . _ _ _ _ _ .

see the one-eighth inch now, you may change next time. So there's no way a conscientious-design engineer will factor the one-eighth inch that you are going to create for the restraints or you have seen it exist at the restraint. Is that correct a

statement?

STOKES : Yes, I'll agree with that. The gaps are primarily concerned with thermal loading. Seismicly, it doesn' t matter whether it's a sixteenth or an eighth. But the thermal loadings in many of the Class 1 supports far exceed the seismic. As a result of that I took several pictures of U-bolts, which, not only the U-bolt but the supporting structure was deformed based on the previous thermal loading. I tried to take pictures as close to vertical alignment as possible to make it very obvious that there was a large amount of distortion. A lot of these supports were Class 1 systems but they were in the Ox Buildina.

I happen to have a very limited time frame to take pictures. I used my own camera and I had a pass that was good for two days.

A friend of mine asked me to come down and take these pictures of the U-bolts and the supporting plate. But they do indicate that the thermal loading is very critical to these structures and that they are a zero gap on one side and already exist distortion to the support.

EBNEDDER: Mr. Stokes, this is Stu Ebnedder. Can you, on the plant tour, specifically identify wnere these supports are?

STOKES : Yes, it may take me a few moments of recollection. I may have to pull some copies of the plant organization or look at drawings, floor level drawings, but I do Stokes /12-7 9

~ . -- - . . . -. ._. - -. -. -

know where, in the general area, all the pictures I submitted exist. And as far as I know, up to the date I was terminated, these problems had not been corrected.

??: Dennis, do you have any questions on these pictures?

O You were talking about those, weren' t you?

KIRSCH

I was able to locate the accumulator 1-2 discharge line pictured on pictures on the far.right.- The other three pictures, or four pictures, I could not locate and any help that Mr. Stokes can alve me in that regard would be t

appreciated.

YIN': Esa Yin. I guess I don' t have any further questions In this area. I would like to observe the actual condition while ,

we were conducting the site tour.

3BNEDDER: Any other cuestions, this is Ebnedder. Any more questions on control documents? I think we will probably ...

l YIN: Yes, I do have an additional question on page 3. In j second paragraph, shown on page 3, it was stated, let me paraphrase it, and used by

were checked to being consistent with those on the i

j control documents. And specifically, Mr. Stokes mentioned the i -

4 assumption used in fact differ from those on control documents in  !

I i

! thirty percent of the cases. In this particular area, I have two 1 I-specific questions to ask. First, approximately how many calculations sere checked by Mr. Stokes? And second, can Stokes f provide some of the calculation numbers for us to check and i

! verify?

i i

Stokes /12-7 10

_ . - - - ~ . . - - - - - . . . - . - - .

1 STOKES : That's a big question. Typically, we were responsible as far as production here at the plant to produce in

, the neighborhood of one and a half supports a day. If production is checked, I believe I will _ be beared (sic) out that I was in j the top production of anyone here at the site. Which, in effect, l would say that over a period of year, I did hundreds of 4

} hangars. The exact number I can' t say. Because of the speed

, required by management, I did not make additional copies for myself on many of the designs. It was hard enough to get copies that they needed for the checking and the work that was actually required. I do have copies of many supports where the design i assumptions are incorrect. I do have a list of specifically, I had intended to request that these documents be located and made l copies for the NRC immediately on my arrival. I was very much afraid that if anyone on the site knew that I was comino, they would completely review certain areas of the calculations which had indeed caused them many hours of work and modify those hangars, clear up records, and correct things which I have cooles showing to be incorrect. I would like to turn over a list of l

approximately twenty or so supports, some of which I have the very first drafts of the cales showing them performed by one set of engineers. I believe when they locate the final draft of ,

these cales, they will find that the final set of cales was

! performed by another engineer. There was absolutely no mention or no record of the earlier engineer's work. The copies I have show that the cales failed, the supports failed for one reason or another. In some cases, the' gaps caused extreme load problems.

Stokes /12-7 11

--m_ _ - .- . -~._..-m - - - _ . _ . . _ , - _ - _ ~ . . _ , . _ - , . , _ . ~ . ~

And in other cases, I am aware that strudel was used quite deliberately in up to like seven runs, making modifications to

~ the design assumptions to try to get hangars to work. I also have a copy of the control log for these documents and in some cases they didn't completely update the log to sho., the final

, engineer's work and they still lef t the original engineer showing

' but the final calc shows someone else did it. I would like to turn this over to someone in the NRC, Mr. Bishop, and I would like, as quickly as possible, without disturbing our meeting, to have someone obtain those documents for them, because I am very much afraid they will be tampered with.

1 DUKOFF: This is Dukoff. Mr. Stokes has presented i Mr. Bishop with an eight and a half by eleven piece of yellow i

paper with a series of numbers on it. We'll be marking this as

j. Exhibit 1.

BISHOP: I don' t know, is it beneficial to have these l

! numbers recorded into the transcript or would we just attach this document?

l DUKOFF: If it's acceptable to Mr. Devine, we could just i

l mark it as Exhibit 1. And then we can keep it as an Exhibit to i

the interview. Is that acceptable to you, Mr. Devine? I'll be marking with my signature on the back that we received it from j you on this date.

BISHOP : Probably missed that. Bishop, turn it back over to

^

Mr. Yin.

i YIN: I have reviewed some of, this is Esa Yin again, I

have reviewed some of the calculations in both suoport group and Stokes /12-7-83 12 i

...-,.,--,,--...-,,,---,.-,+--,,.----,.,,.....e-.,_~.------.w , - - - - - , - ,,w - , ~ , , , - , . . - , ~ - , , . . , ~ - - ,--,-.,,c.--.

the stress group. Many of the calculations-that I reviewed show a number of revisions. In most cases, the first revision may be or may not be a rev number. Some cases have revs 0; some cases do not'show a rev number. Those are the ,onsiderations that, O

those are the hangars restraints and may not have received )

sufficient calculatic7. And subseauently, does require further review in providing the bac2up calculations. And rev 1 or rev 0 usually indicates the hangars, have been calculated based on preliminary assumed data, or loading-data. And subsequent revs,

^

they change the preliminary calculation to final calculation and the last' ref.normally showing the as built evaluation. My question to Mr. Stokes here is that particular exhibit, Exhibit 1, that was presented just now, what time frame, and what condition are those calculations? Are they all in as built condition or in various stages of upgrading the system?

- STOKES : I'm glad you asked that. I'd like to try to, maybe everyone here understands how this plant was designed in the first place. Let me point out that when I came to this plant, we were supplied no preliminary cales from existing work. The drawings from the hangars in many cases were well up past rev-0, they were up to rev 4, 5, 7, the number varied. I think the largest I saw was seven. Many of the drawings had been stamped

~

as built two or three times. ;They had supposedly been reviewed and reviewed. We, in effect, were instructed not to-assume

! anything we received was any good.- ife also were told that if any l calc existed it was at such a state of condition, that it could not be used to verify the support-as it existed. We did perform Stokes /12-7 13

. from November 8th through March what was considered a rev 0 calculation noting that there was some kind of calc somewhere on each of the supports possibly, in many cases none, but we did reference those cales. We never saw those cales. We did see, I O

saw or was shown at one time one of the cales. It was a

, combination stress and a combinaton hangar calc. It wasn' t what I consider a valid calc for either, but what I'm trying to tell everyone here is the rev 0 cales we were instructed to perform were supposedly under a reverification program. We were supposed to determine whether the support, as existing, was correct, could stand th'e loads given to us by stress, the gaps, the configuration, everything met code standards. We, the group of engineers and myself, we reviewed these supports strictly from an as built drawing, and with the personal lockdown of each in many cases. I won' t say that in all cases every hangar was physically looked at. To make a field trip, especially when we were never given security clearance, and we were required to come up with an escort, and to put out one and a half hangars a day documenting cover sheets, loads and everything else, made it impossible to look at some supports. Things that looked questionable, supports that looked beyond, I mean just the drawing looked questionable, we did look at. Very simple cantilevers that looked well over design, we didn' t l'ok at. We didn' t check the gaps on the bolts,- we didn' t have time. We were not given time. If we had, we wouldn' t have been here more than two weeks. The process here was they hired. thirty-five guys, they filled the trailer, they gave them preliminary calc and then they gave.them a set of cales Stokes /12-7 14

to do. They didn' t even have control documents. They were informed to start producing. If they didn' t produce in two weeks, they were terminated. They were replaced with more people. Production had to increase. It was simply a very fast paced approach. But I was very shocked when I came to this plant and I found that there were no control documents concerning the design. I've worked on plants all over the United States and it's true that in many cases, they are not as good as they are now, people tend to be more critical of plants designed in the last four or five years, than they were fifteen years ago. But I came f rdm an area, Atlanta, and I worked for a utility engineering firm there for five years af ter school. I did have l

associations with Bechtel. I found that Bechtel's document

, control from scratch on plants that they had started from day one to be exceptional. Their cross-reference system, their computer i

i printouts, giving-you an ability to research a hangar was just unbelievably accurate in my opinion. I did find problems on other plants where the as builting was what I considered up to par. I know why. Typically, Bechtel sends down designers, j draf tsmen, they don't send high level, high paid engineers to do as builting. They do not have instructional how to determine the weld. They look at a hangar and they say everything's a fill-in. In many ' cases, they' re not. And even if they weren' t, they l don't-have the experience of looking at a weld and determining whether or not it was prepped and how it was made. And . many defects from that kind of thing come into the welding. However, most-designers and draftsmen that have done work, that I've l Stokes /12-7 15 I

worked with, including this plant, are capable of taking dimensions down, noting gaps. As far as gaps on U-bolts, they've never been given filler gauges, the QC here were given little pieces of wire which supposedly worked to compliance with the 0

gaps and they were on a string for them to use and so typically gaps are just a visual thing. You estimate whether it's an eighth or more. In many cases, it's more than an eighth, but if you looked at, you give it the benefit of the doubt. What I'm saying is that things that were measured here that I've noted are beyond any doubt. I mean, they were gross. There's many things in this ' plant I don' t like because they weren' t adeauate equipment to do a good job doing their job. But many of things t

that I tend to overlook in my personal engineering knowledge, I-use my own judgment a little bit, everyone here does. Every engineer on this job site uses his knowledge, his engineering judgment, depending on his background, how he was born and raised, where he went to school, what kind of life he lived.

Most of them, their engineering judgment probably isn't worth 10 cents. Some of the guys, I value their judgment very much.

j In any case, any of the work we did here, our primary objective

- was to validate within the proceduces what existed as being acceptable. We were batting more than fifty percent failure on all areas of investigation. We only, and I'm not sure anyone here is knowledgeable of this, I've been- told they did a hundred percent seismic by a reporter. To my knowledge, _ up to the time I left, there was only and never was any more than a ten percent-sample done on this plant. The attitude of management was from i

Stokes /12-7 16

the beginning they would perform ten percent on a random basis supposedly, they would try to prove that ninety-five percent passed. There was a five percent failure which statistically they would be agreeable to buy off. From my statistics

, o background, that's an acceptable figure. I also contacted a Cal

, Poly professor of statistics in regards to one of my noted areas that I didn' t like, particularly the l'-bolt load allowable switch  ;

they wanted us to use on Unit 1. He agreed with me that ninetv- l five percent and the method was an acceptable procedure for validating the large mass of data. However, in the case of mv U-bolt p'rogram, or analysis in reviewing their program, they did not use a statistical mean, they used the average. It's very rough to say they complied with the Code Section which allowed them to do what they did. The Code says that they could have performed only one calc and reduced it ten percent and used that figure. They did use five load cases. Statistically, according to the professor, they would have had to do in the neighborhood of a hundred to come up with a reasonable accurate number and, based on the statistical average of the data I got in their analysis, when I did my cales, the best they could have done was just slightly over what ITT Grinnell supplied at 650 degrees.

They stated in their report it was only at a 120 degrees, which I, using the ability, or training I've had, I could see where they would get a. slight load increase based on temperature not t

being as high as it should be because the allowables are reduced at higher temperatures. I may be veering, and I'll.stop at this,

. but I. hope I did explain-the' problem with'the drawings, the Stokes /12-7 17

calculations, they were extended beyond the rev 0 stage. When they split the trailers, the cales we had performed were, we were told would be nothing more than a review for control document assumptions. However, the people they put in that trailer. They never questioned anything we did. I had friends who were

, transferred to that trailer that kept me informed. They did not just give a cursory review for accuracy and assumptions. In many-cases, they completely redid the cales and I can prove that.

BISHOP: Thank you, Mr. S tokes. Before we proceed further, I want to introduce a new member who just joined our group. It's Mr. Mark' Hartzman, oh excuse me, Paul. Mr. Bezler is with our f

Office of Nuclear Regulation in Bethesda. Paul, we are recording this session and we would ask that if you have a cuestion that you'd like the mike, and we'll pass it to you. And you state your last name and then proceed with your auestion. Also, Mr. Cruze has a few words to mention in reference to the list you 2 just.provided us.

CRUZE: Ah, yes, this is Jeff Cruze. I just wanted you to i

understand from the very outset here of our undertaking to retrieve the documents from Exhibit 1. The arrangements that  !

I' ve made, I have Mr. Marvin Mendonsa, who will accompany whoever  !

. I

! within the organization here that will be necessary to make a  !

l prompt retrieval of those documents. And Marvin is there just so that we know the process by which these documents are retrieved from their files. Is that agreeable to you? I think that will expedite getting this matter taken care of while we continue ~with l

l ~ our session here.

l j .S tokes /12-7 18 i

BEZ LER: At some point, this is Paul Bezler. At some point, I may like to see a number of drawings that are mentioned in this disclosure. -And since you're going to be getting documents, ...

CRUZE: This is Jeff Cruze. I only want to give some o

priority attention to the documents that Mr. Stokes has asked that we retrieve. I think for our purposes we'll give that whatever priority it deserves. [auestion from someone -

unclear] I'll deal with you separately on that. I just would like to confirm that the arrangements we've made are acceptable to Mr. Stokes and Mr. Devine.

STOKES : - They're acceptable to me.

EBNEDDER: Stu Ebnedder. Mr. Stokes, let me oursue just very briefly, I don' t want to spend a long time on it. On this statistical, could you describe for me your technique? Was it large sample size, small sample size?-

STOKES : I don't have with me a copy of the program they used here at the plant. I used the data supplied to me.in a report that concerned the U-bolt test program. It can be located down in the control office. The sample was five tests on each l

, particular design condition. That was established by PG&E personnel. The method for testing was PGsE personnel methods.

. All I simply did was review their data, their report, I noted things I questioned and I noted the way they used the data was the way I said. It was not a statistical, it was strictly an average.

EBNEDDER: I was hoping for, you said you did some

! statistical calculations and checked them with a professor. I S tokes /12-7 19

was interested in the technique you used and whether there was a

< \

specific ncne, normally,- there's about five different techniques l l

. you can use. Would you describe them?  !

! STOKES: Well, statistically, there are many methods. At the moment, boy, I got pinned down. I tend'to, when I get put

~

,under pressure, I tend to slip. Um, I do have copies that I can pull out of my formulas that I used. I didn' t have any statistical books here with me at the plant. The closest thing I had covering statistics was a new, recent World Book encyclopedia and the reason I called the professor was to substantiate the method I used out of all the methods was the reasonable way to use the data. It was a critical, oh God, I, well, not majoring statistics, at the moment I can' t remember the precise method.

Like I said, I can supply that.

DEVINE: There is undoubtedly going to be a number of questions that Mr. Stokes can't answer from memory and he's going to have to go back to his notes. I'll make a star of each of those and you all can too.and we'll check to see if we have the l same follow up questions.

EB NE DDER: That's fine. You say you can provide us with a copy of the statistics that you developed. We would like to see those for clarification purposes. Any other questions, we're on page 4.

YIN: I think we got some side track. This 'is Esa Yin l

here. Let me readdress my specific question and I appreciate the l

( general comments from Mr. Stokes but the specific question that I raised.previously has not been addressed by him. The first S tokes /12-7 20

I specific question let's assume they have a calculation Rev 0 shows no calculation, Rev 1 the calculation was performed on site. (end of side one, tape one) 1 BISHOP: We went off record to turn the tape over. We're l now going back on record at 9:27 a.m. During the interim, a l 1

, break was taken by all the participants in the interview. Bishop i

speaking. Before we proceed with the pages of the affidavit you supplied, we have one question from a statement.that you gave a little earlier in the morning that has to do with taking the photographs out in the plant. You mentioned that-a friend had asked.yo'u to take some photographs. A auestion that occurred to l

us was is this perhaps another individual who might have knowledge of the types of problems vou're talking about and could help us get an additional data base? Could you identify that individual for us and tell him, give us an idea if he might have i

knowledge of this area.

4 STOKES : In answer to that question, this is Charles Stokes, try to limit my answers to being exactly what's asked.- I shouldn' t have pointed out that I was asked to go by a f riend. l l

He does have knowledge, he does still work here.

And as far as

giving you his name, I'm not at= liberty to do that. At his wish, l l

he wishes to remain anonymous. He has agreed, if he could stay anonymous, to providing those conditions are met in advance, cive testimony. At least to me, and to my attorney. I would have to ask him personally if he's willing to give anything to the NRC.

.I'm not,. my whole attitude on this whole thing was. I never at-any time, wished to put anyone in jeopardy as f ar as what he does Stokes /12-7 21 I

as a living. I do know people who support me. They've, on a l personal basis, they' ve been very supportive. They do, however, have more-constraints upon them as far as earning a living than I do and I will not at any time, do anything that they're not O

acceptable to in advance.

,  ; DEVINE: We'll sure cooperate on trying to help you follow through on these issues within those limitations. And there's a lot of ways we can do it. For example, we could suggest to you that you all contact individuals who work in a certain area and just come in of your own volition. And, as a matter of course,

! you'll cover the right person and they won't be so much on the spot coming forward as an alleger but rather responding to your inquiries which they don't have any choice about. And we just thought we'd see how things went today and what you needed for followup. Trying to be as creative as possible to help you without threatening them.

j BISHOP: Bishop, I appreciate your- concerns for that. I do think it's very important if 'there is another individual who has knowledge of these or similar issues that we do have an opportunity to talk specifically with him. -And I would seek your cooperation in making him available to us or identifying his name. I can tell you that we would do our utmost to protect his j identity. We always have the possibility-of it falling through but we would seek to the best of our ability to protect his

)

identify. But I- think it is very important that we have an I i opportunity to talk with him if he has knowledge of what you feel i - 1 or what he feels are improper activities at the site.

i Stokes /12-7 22

.. . .. - . _- - , - - - , , . - 4 ,- , ,. - - , - . _ . - , J-, , - - - . , _ , , , , _ - - - - - - - , - . - ,

EBNEDDER: Any further ct'astions on these control documents and statistical techniques?

YIN: Mr. Stokes, I'd like to ask you two specific questions l l

regarding those calculations that you have checked. As I l indicated earlier that calculation, let's say, 8

, assume you got three or maybe four stages, say, let's assume Rev. O. That was with no calculation at all. And Rev. 1. You l have performed calculation based on preliminary loading assumption. And Rev. 2, you design these hangars based on final loading condition. And Rev. 3, is the as-built evaluation. Now which of these Revs. do you personally involve and check there was 30% to be improper?

STOKES : Okay, the Rev. O on this plant was not a no count. That was Rev. A, or ~ a dash. That's all it was in the control block on the calculation -- a dash. Any reference previous calculations, it was given no number.

Rev. O was, like he says, preliminary loading. However ,

Bechtel's procedure on this plant was to maintain a preliminary-status for this calculation's write-up until, to my knowledge, l

two weeks before I was terminated. When they finally decided to i sign those off as being final. They were the same calen. There I was, by that time, several revisions to some of the cales. To my knowledge there was no as-building _of them. At the same time

. -these cales were being finalized, preliminary loadings and l everything, they were at the same time in the field being constructed, modified, or, if there was any. I'm not sure there would be a 3 as-built. I'm sure it might happen sometime in the Stokes /12-7 23

future, but ... what we were told when we performed these cales is ... it's true that the Rev. 0 were preliminary. They were, and may be, the final calc. Rev. 0 ... may also be as-built.

The review that I stated was supposed to have been performed in March following our calculations on these reports was, as

, stressed, finalized their preliminary loadings. They were supposed to review the calc for loads to see what loads were used. Whether they were conservative, non-conservative, and at that time make them final. They were also supposed to review the i

changes that were required by M9. Because, by the time I left j here, M9' had 12 revisions. Each revision changed certain things i

in the design that were affected in each of the cales. To mv knowledge, they've never finalized M9. Technically, there is no 4

final calc, even at the moment, even though they've been finally issued as final. They're still subject to change, depending on

, any new reanalysis.

YIN: Based on my personal observation, this week and last week, they are final calculation provided following the preliminary calculation, and this also as-built evaluation that I personally reviewed. So based on the information that you have provided to me, I can only assume that during your employment at the site, you were performing and checking the preliminary l calculation based on the preliminary load assumptions. Is that correct?

STOKES : I would have to say yes.

YIN: The second specific question related to the questions l

l regarding documents. Is that the fact, when you i

Stokes /12-7 24

~ ._.. . . . - -- -_ --- . __ -

4

' talk about the assumption used in fact differ from those on document. Are we talking about the assumption preliminary loads? Are we talking about the incorrect use of

, formulas? Are we talking about confusion maybe based on the fact the allowable loads differ from one, the design document, to

,another? Can you be more specific on what is the load assumption j

whether it is before calculation, or af ter the calculation, or in l

the process of doing the calculation? Can you be more specific l than thst?

STOXES: I'll try. Some of the . problems I listed in my statemen't cover this. They are listed as separate items. One of the DRs thet I issued just before I was terminated concernina the bolt spacinq on the plates, I found out that we had never been supplied an M9 or any of our design manuals. The size of the shells for anchor bolts, concrete anchor bolts. The vendor catalogs did not show these dimensions; a few of the people may

?

have had information from other design sites. The allowables for ,

i the spacing on the bolts wan supposed to be based according to M9 Rev. 12, the diameter of the hole for the bolt ... or for the l

anchor. The reason I wrote that DR is, to my knowledge, I didn' t

have the shell size and unless someone brought it with them from 1

another, plant, he didn' t have it either. Also, wo were i instructed in M9 as far as a design assumption that we could check the support movement in the support direction to a .025-inch movement. Rather than doing a stiffness cale, and if it was within-that .025 inches it was acceptable for stiffness. The problem with that is in the U-bolt allowable loads, in their test Stokes /12-7 25 l

program they deflected the U-bolts to .025. If you add up all the displacements that they technically used in a U-port support, you'd have had .025 if the allowable load was pushed to the limit. A .025 maximum for the support, and if the gaps were not correct, an additional movement. M9 states that the total

, restraint movement was .025. Now I know for a fact on the U-bolt l supports that was not even close to being correct in many cases. That statement is primarily concerned with things that changed in M9. The methods for doing the base plate cales. We were instructed per a memo which stated our work here, that on small bo're Bechtel assumed that small bore plates did not flex.

There was no displacement from the base plate involved in a

. support. From my previous work on many stancheon supports that can lever off plates, there is sizeable movement affected with j the plate displacement. The reason they wanted to make that assumption is they assumed the loads were very small, they were 1

adequately braced, and that there would be very marginal i displacement. They did however ultimately end up deciding to do finite element plate analysis on the supports. They only checked stresses to the plate in the finite element. They did not ever a go back and include that displacement at the support point due to j

the plate movement to the supporting steel, which is added-to 1 these other movements that they would have allowed in the calculation. There were other cases and other points, but I'd l rather not make a complete list of all of them. I don' t think it'd be worth while at this point. It would take an awful lot of  !

I i

time.

l S tokes /12-7 26

-. . _. . -. - __ ~ - . _ - . - _ - -

i DEVINE: Would you like that Mr. Yin? Would you like him to prepare a list of other examples on this? l YIN: Yes, I do appreciate you can give me more information l

in that. You mentioned allowable bolt spacing, and the U-bolt

-that may be defect more than .025 inches, and the M9 base plate

, calculations. Those are not assigned in my areas. Look's like I  :

can go home now, but I do appreciate if you can provide me some additional information in regard to other areas that you have concerns, such as formulas, such as basic design allowables, and such as design of that was incorrect,.so in the area of' actual calculation other than the other areas that you

, just mentioned, I would appreciate you provide me more t

information.

STOKES : Let me give you one. Concerning formulas. Not knowing exactly what each person's area . . . if you'll ask me a question telling me what area you' re in, I'll try to provide j something. When you mentioned formulas, there's one formula

concerning the U-bolt allowable loads which was dictated on Unit 1. The U-bolt loadings were too high, which were part of M9 . They required that we used those loadings even though we f disagreed with them. They also forced us ... or wanted us to use a formula concerning the interaction of the U-bolt loadings.

~ That~ formula was based on~ a square of both terms ... in summation --

of those terms. ITT Rennell and they're ... let me stop for a second.

l DUKOFF: Time is approximately 9:44 AM and Mr. Stokes has stopped for a minute to change the tape.in his tape recorder.

l l S tokes /12-7 27 l

i .

.J - - , , .. - . - - _ .- . . _ . - - , - . _ _ _ .

e We'll be going off the record at approximately 9:44 as Mr. Stokes works with his tape recorder a little bit.

DUKOFF: This Dukoff. The time is approximately 9:46 AM.

. Going back on record. Mr. Stokes has his tape recorder running l- again. Mr. S tokes. . . .

STOKES

Concerning formulas. Typically that formula ...

and past experience has not been sguare by square. It tends to make the calc unconservative. The closer you get to one with any factor, it tends to put it away from one. The loads for the U-bolts on Unit 1 that they wanted us to use, which were in mv 4

~

opinion too high, also included that formula. It wasn' t i

something people brought with them. It was something that was in a controlled document at that - time, meaning the peoole didn' t

=

have that ... it was supplied in a drawing; the number is

. 049243. It was supplied by PG&E. It wasn' t part of M9 to my knowledge. I should note that I worked on both Unit 1 and 4

Unit 2. I only worked on Unit 1 up through March. I.did work in 5

the quick-fix fill area on Unit 1 from June to near the end of the summer, almost up to my termination, and I was given the i responsibility for modifications to snubbers in the field that j had problems in installation. But as f ar as providing formulas 1

) that people brought with them, and design assumptions from all l plants, all I can say is I don't have copies of those. I do know

'I was shown documents by other people from Susquehanna. I myself

,i

! had documents from Farley, Davis Bessy, ... they also were old .

Bechtel methods of calculations. Anchor bolt calcs, and so 1

forth. Due to the fact we were not supplied with M9s or guides S tokes /12-7 28

l l

l as to how they wanted us ... or methods they wanted us to perform the work, the people used what they had in their background.

That's . .. I'm sure many they still . . . many of these documents are still in the trailer over there.

EB NE DDER: Mr. Stokes, again I'm just gonna ask you, and you

..can say yes or no. Could you provide us with names of peoole who use their own personal documents?

STOKES : Well I could, but like I said earlier, I'd rather not without their approval.

EBNE DDER: Thank you. In relation to the U-bolt on page 5, you mention that ratings for U-bolts that we purchased in some I cases were one-third to one-fourth more stringent than those on drawing 049243. Could you give us some specific cases. You said

"some cases." I would like to know if you have any specifics.

If you don't have them now, could you supply them at some future

! date?

STOKES : Well, I' ve got copies of ITT Rennell catalogs.

Many of the people in the design group aise have copies of these catalogs. The catalog themselves supplied by the vendor gives a recommended load rating. It does state that there's a considered

factor of safety of five. It gives them the operational maximum l i i j temperature which the U-bolt is supposedly allowed to be rated  !

at, which is 650 degrees. I mentioned that earlier. The requirements for design were 120 in many of the areas, which is not very high. In the containment, there was a 300-degree operational temperature just from room conditions. Many of the pipe operational temperatures, however, exceeded 650 degrees. In 1 Stokes /12-7 29

the U-bolt load test program and the allowable computation that they supplied us for the U-bolts, they did not account for these differences in temperature due to pipe line temperature. They did the test program at 120 degrees. They did not statistically justify their results. Based on the ITT data, I took the test program and I statistically took their same data to determine how valid their numbers were. The calculations I did I never turned over to anyone. I couldn't find anyone in the design group willing to check them. They'v'e never been checked. I did, like I said, consult with the professor on my formulas and I redid the formul'as and the cales several times myself to make sure that what I was getting was valid. What I'm saying Is the loads per

049243 when compared to the load ratings for ITT catalog ratings, the numbers are 4 times greater in many cases.

EBNE DDER: Those statistical calculations are one you agreed 4

to supply to us previously ... these same ones. So let me move on on that. Would you be specific, on what lines you have ...

you have identified in your statement that there are-several lines out there that exceed 650 degrees, and in some cases 1,000

degrees. Could you be specific on what systems these lines are

, located on Mr. S tokes?

STOKES : Well, I' ve got a copy of the pipe vendor list. It gives supposedly operational temperatures, or standard design  ;

temperatures . I made that statement based on what was' written in that list of popping Isos systems. I do know that from experience doing the calculations, those numbers were low, and many times those actual operating temperatures of the line was in  !

Stokes /12-7 30

f ar excess of those numbers. I personally worked on supports doing M40 cales on a line that was in the neighborhood of 900 degrees .. 800 degrees. 6h boy. I may be able to go back in the room of that line. The number I can't give you without physically finding it in the plant again.

EBNEDDER: How about 650-degree lines. Can you identify any of those?

STOKES :- Well, I can give you a copy of that line list. It shows right on the list itself those lines.

EBNE DDER: The licensee system?

STOKES : That's something they should be able to give you in Document Control.

EBNE DDER: . Or a number pulled out.

STOKES : Right. I can' supply what numbers listed per that document. The document I have probably has been updated. It_ was constantly under revision. It's typically a line list.

YIN: Just to followup on the statement, you mentioned you

have performed some M40 calc. From what I understand, there was only about 6 people given the assignment in performing that particular task. This M40 calc, if I understand correctly, is )

based on hand calculation to determine the hangar loads-that will be imposed on the supports. Is that correct?

STOKES : Mr. Yin, that's correct. They were used to determine loads. The number of people were more'than 6.in many

, cases. Concerning Unit 1, before March, there was a group that I was part of, which was assigned to high. thermal and sam-tam systems which we were asked to do M40 cales, trying to justify Stokes /12-7 31 l

_ _-- . _ . - _ . _ . __ _ . _ . . - _ _. . _ _ _ . ..__, _--.1

the code break locations. The code hangar loads. I was the first person in that group to find a line which did not have adequate offset to be support under sam-tam conditions and thermally within acceptable procedures. My judgment and my ability was guestioned. The system I was talking about earlier that was very high temperature, I performed that same analysis on that system. I required 7 snubbers to allow it to move thermally without overstressing it, and at the same time supported seismically. They ran that same system on an ME101 computer analysis. They only removed 1 snubber. I felt fairly confident that that justifies my ability to do a hand cale that is very

.close to pushing computer abilities. A person with good experience, making valid decisions to judgments, and using

, accepted procedures within the industry can do an M40 cale on very high temperature, high seismic load case lines and be f airly accurate. I can' t say that you can't go and take another support l

out with a computer analysis, because it's much more accurate in the calculation, but a person can push computers and at the same time he can do it in a closer timeframe which makes it economical in many cases to put in that extra support, rather than try to stretch the impossible.

, YIN: If I understand correctly, M40, the basic use of the M40 is ' to determine the- 3-span peak loading on the cold break.

The non-safety piping that is connected to the safety cross-piping. The use of M40 to calculate the entire piping system, J

especially the high temperature ones, I'm not too sure that- that was applicable at the site. Can you be more specific whether or not that was the case?

1 4

S tokes /12-7 32

_ ._ =_. , _ _ . . _ . ._ _ _

l l

l l

- STOKES : Initially, ... what you're saying is true ... not initially; now. When they first started doing the review on Unit 1, they didn' t want running all the systems on computers.

They felt it was too expensive. They felt if they could buy off a line using an M40 hand calc that was done very quickly and roughly, they'd do that. The thing was, on the code break l I

systems which were high temperature in many cases and connected to large bore piping, ... the large pipe moved itself quite an amount, which required a lot of offset. They initially started us out doing M40 cales, but when we started failing those systems requiring that additional pipe be installed for adequate offset, they decided they would not use any of those M40 cales. They would rerun every one on an ME101 run. M40 calcs were used, not only for the 3-span load determination, which, you didn't need M40 for. The 3-span load determination was just a table length based on size and schedule of typical links between supports.

Typically 7-9 or 10 feet, and the 3-span load rating was just 3 times that length times the weight with water in it, assuming that it's hydrate. It was a load placed at the end of a seismic system just to give a large, non-seismic load coming into it.

That was an assumption based on management's decision as to how much load-down line on the non-seismic side could come to that support. If all the non-seismic supports fell down on that side or if it was a chain reaction, the end result would be you could have 100 feet on that support, not just a few .. 30 feet.

However, that was their decision. In my paper I didn't question their 30 feet or-the 3-span criteria. Whether or not I feel Stokes /12-7 33

that's a problem, I'm not going to say. But M40 was also used to determine preliminary loads when we couldn't get those from the ME101 group ... on gang hangers, when you get 5 and 6 lines on a support, and they're running one system, in many cases we held off on doing those cales until we had more than one line. We tended to wait until we had pretty near all the loads and then assume pipes with the similar configuration have similar loads to the ones we already were supplied. We didn't just pull numbers out of the air. We also, at times, did an M40 calc on the lines where there were like 2 supports. And we had one, but one was quite'different ....to determine the loads. Typically, those loads though did not vary very much from an M40 cale when it came l back f rom the computer analysis. In many cases they came back even larger. And that was a problem with those last reviews we were doing. The loads were going up. Not only because of the i f act ... our M40 calcs weren't very accurate, but the entire stress program here was dictated out of San Francisco. I had a problem early on when we started doing M40 concerning C28, C30, C17 ... the seismic data sheets coming from the City. We were

! not supplied control copies, and we didn't have the latest l

revisions. There was also some other guides which were supplied l telling us about the local conditions, the thermal expansion of l failure and containment, giving the displacements along certain points of the wall, which we were supposed to include from the anchor locations going through the wall. Initially the City

zeroed all the building movements at ground level. .Their

f. assumption was there was zero movement at ground level from all

.i S tokes /12-7 3 4 l

-_ -. . . . .- = - - - - -- - _ - . -

O those buildings. Through a discussion with the person who wrote the paper in San Francisco, I was told that it was going to be  :

)

released on a Friday. Two weeks later it had not been issued.

l That document ultimately came to us with movements all the way to  !

\

o their base level. There was building displacements even between

,the ... at rock level. Stress, on the other hand, when they used that data, their calculations were affected because they did not include those building movements correctly. Later on, when they got the corrected building movements which I had argued for, the loads went up based on that. Especially in between buildings.

In many ' cases there was no way the line could be supported between buildings without rerouting. They did not want to do that.

BISHOP: I'm sorry to change the subject for a second.

l

We' re having a . .. we' re in the process of copying the calculations from Exhibit 1 that you had provided us, and Mr. Cruze tells me that there is a very large volume of documents I associated with each of those numbers. Are there particular l

i records within that package?

l CRUZE: I just ... I want to make sure we're getting what you want, okay. And that we're doing that in an expeditious way. I understand we're not ooing to have these reproduced at this point in time. We want the originals in our possession.  !

A nd . . . in that basically what you want, number one, and you want ,

the calculations and the calculation log sheets for each of 1 I

those. And any other information in the files, you appreciate, i

l some of ' these packages are rather voluminous. But as I look l

I S tokes /12-7 35 g a- - - - - - - - - - - v,v., -- -

r- - ,. ,- , - ,

- ' l I

i 1

-things over there, it looks like the calculations are sort of in a separate folder, or at least, excuse me, at least the log sheets are. And so we're gonna have original documents and of course we want to make sure we return those complete to them, so a

we're not going to reproduce things at the moment, and would you

,ee s that that would-be necessary for any reason. Other than for our purposes, we may, for our purposes. Okay, Alec, maybe you

-can answer the question ... quite along with a series of questions perhaps.

STOKES : I'll try to make it short. I didn' t ask for i document's which I didn' t feel had some problem with them that 1

would be relevant to what I had to say. Many of the supports at the end of that list, I feel that if you review those particular 5

calculations, you'll find probably deficiencies in how the assumptions were made. Because those were the problem I

i supports. And they do come f rom sam-tam sampling, thermal i-

sampling, and from seismic sampling. They're varied. The first supports that I asked for, the reason I asked for those is I have the zero rail cales. The point I was trying to make in zeroes and the ones, twos and threes, is I have copies of cales that are Rev. O that are not included as Rev. O in the final pack. They

) completely discarded that preliminary calc. To my knowledge, in the industry, once a cale is performed on a support you don' t support.it. It's only added to, subsequently, when new information is provided. - You do not destroy it. They completely wiped out Rev. O that we did showing it failed, and the final pack, which ISO end is looking at showing Rev. O,1, 2 and 3 and S tokes /12-7 36

( - _ . _ -. -

as-built do not show those failure conditions, even if they were assumed incorrectly. And what I'm saying is they tampered with cales that, according to industry criteria, should not have been done away with. They should have only added to that pack. I do know t' hat the packs are very voluminous. They're computer runs in some cases . . . I'm not . .. you need those because the struddles ... a guy's assumptions on releases at a joint are very critical unless a guy's a struddle ... fairly knowledgeable in

. struddle. Knows how that you can model a release at a joint.

It's very easy to change load conditions to a plate by assuminq

.no moments. They did that. And I'm aware of that. One support l in particular, I have Rev. O, the guy that did that calc, the only reference to him in the final calc is the drawings showing 4

the hangar. He went out and as-built it; his name's on the as-built sketches in this package. The Rev. O's credited to some i other guy at a future date. That particular calc they didn't get I

the catalog adjusted on it, I don' t think. It still shows him doing it, but the final never shows him at any point in time as having worked with that calculation. That's what I was trying to j point out to Mr. Yin that those documents are primarily addressed  :

i to him, and with what-I know ... in other words, he could look at '

cales for 4 days, but unless he knows what calc to look at, the odds are he will never see what I notice. i l

,  ??: Why don't we get at least 3 of those things completely i done and get over here so that Mr. Stokes can show Mr. Yin exactly what he's talking about.- But there's so many of them

there that it might take a long time, but let's get at least 3 here before Mr. Stokes leaves.

Stokes /12-7 37 q ,

STOKES : First one on that sheet has a one, zero being destroyed, and it is also a hangar that had an awful lot of difficulty in getting it to work, and I'm' pretty sure that if I pull not only those drawings but the ISOs, the construction  ;

i dates 'for all the supports on the ISOs, that particular support had a support located within 6 inches of it to suck up the load

-' l

. because it was failing. The final calc I don' t believe shows it  ;

failing. And that's only one calc.

i CRUZE: That was precisely the point I was getting around to, Stu. Perhaps we can get a couple ... get us started, make sure we've got the right information you're seeking, and then

~

we'll continue with the effort to obtain them all.

STOKES : I do think you should get the regionals, because I don't think they should be lef t over there in that trailer for 4

them to .. . once we pull one drawing, they' re gonna know what we' re looking at. I mean ... see a friend of mine did the

! struddles for me on those supports that were problems. He did not agree with what they were doing, but all he did was the j struddles. He helped program in and run it. He didn' t have to sign off the cale or anything or he wouldn't have done it. He j supplied me with the list of a lot of those drawings, and that's ,

l the reason I know the ones that are critical to what they did as l far as making false assumptions. The first one on that sheet 1

though, I monitored from the day it was first done, because the l guy that I mentioned earlier that asked me to take these pictures is the one who did Rev. O. He kept a copy of that. And .I got a l

copy of it, so you will know his name once you see the calc.

S tokes /12-7 38

_ _ _ , _ . _ _ _ _ , - . . . , _ _ . . , _ _ , - ~ .

BISHOP:- We will pull all those calculations of all those packages.. All the originals. And then we will assess what we need to do with them. Whether we should be making copies of them for our retention or not. One thing I wanted to make clear is that we will not be providing copies of those documents outside the NRC. Hope you appreciate that.

STOKES : Well I expect to be providing you with extensions of those cales, rather than you providing me with anythina.

BISHOP : Thank you. Mr. Bezzler has a comment.

BEZZ LER: I just wanted to get one thing clear. The calcul-ations that you say disappeared, the preliminary or

earliest calculations, did these in fact disappear for all supports or just for those that had some problem, in your opinion?

STOKES : To my knowledge they only disappeared on the ones that failed. The ones that I have lists of f ailed, and they do not exist now. I know because I' ve already . . . I managed to xerox some of those calculations myself before I left the site.

It was very difficult because I was not allowed in that trailer without supervision, however, when I was .placed in quick-fix on nights I was given pretty much free rein of the plant because I was the lead engineer, the only engineer responsible for design. There was one other guy assigned to help me a little later on. During that time when the maintenance people were in that trailer at night, I managed to get in there, without control, and xerox controlled documents. Without tampering with them, I want to say. I did not adjust anything. All I did was Stokes /12-7 39

xerox what they had finalized to see how it compared to what I already had copies of.

DEVINE: This question is to Mr. Cruze. I just wanted to clear up what the prooer standards were. To see that we have a consensus. Is it improper for there to be two versions of Rev. O with the first version missing? Is that something that's not supposed to happen?

STOKES?? : I don' t think we can answer that.

CRUZE: My experience in this area is limited and I'm goina to have to disqualify myself in making a specific judgment or stating our position on that. Perhaps Mr. Yin might be able to add to that. He has extensive experience where I do not in this particular area.

YIN: Now let me say something. We're not making much progress. I'd like to get through the affidavit, and this issue 4

is a big issue, and we can discuss it at length. But let's try and identify other areas so that we can at least get through the affidavit. And then ... this area is a big one for Mr. Stokes and NISA's following up on it, and then we can do more limited discussion, particularly when we get the package and Mr. Stokes can show us exactly what he's talked about on the Revs. Is that all right Mr. Devine?

DEVINE: Mr. Yin, not to belabor the point, but is there a simple answer to that question?

YIN: Probably not.

I BEZZ LER: Mr. S tokes, you allude that they' ve destroyed some preliminary calculations and yet you say that at times they would Stokes /12-7 40

add new supports adjacent to old to accommodate the loads, or what have you. It seems to me they're making adjustments or corrections to the system by adding new supports. Why did they choose not to own up that if one support was bad and redesign it,

-rather than add a new. support? In other words, they seem to be

-doing what they have to do in any case, but....

STOKES?? : Can I go of f record, Mr. S tokes, we' re going to run out of tape.

DUKOFF: This is Dukoff. We're going off record at 10:15 to i

change the tape.

DUKOFF: The time is approximately 10:22 a.m. We have changed the tape and a break has been taken by the participants. Mr. Stokes will now respond to the last question asked.

i STOKES : The last question was how could a new support be added without relevance to the redesign program? It seems that that'would make the plant safer. That's correct. The problem is the management here submitted a list of supports to be reviewed under the sampling program to the NRC staff. Those supports

~

existed at that date. When we started encountering failures, their fi st impulse was to code', to shorten the code break i- locations and cut out failing supports. They could not come uo with a reasonable decision or answer on how to explain not doing th5 analysis on those supports which they-had already given the list to the NRC. Their decision then was not to limit the calculations but they decided if they f ailed they would camouflage them in calculations and at a last resort, in certain k ~

Stokes /12-7 41

, , . . , - .,.-..-..,c ,. , - , . , , , ,.n , - ,.,. , . , ,

hangars they had problems they couldn' t make them pass under any conditions with any amount of load other than a pound. They added supports within six inches, which, according to stress cales, they, depending on how the gaps were modeled, they could assume that support to suck up all the load. That support had not existed previously, had no count number, no documentation, was not given to the NRC. They did not plan on giving the NRC adequate information for them to determine that hangar had been

-added as a coverup of a failing support. The ISOs which were modified to show the addition of that support, the basis for the engineering changes on that drawing that I was told by Pullman OC was quickly a note stating engineering judgment changes. There was no record that they added a support. The only way you could look at the ISO and determine the support is pull that ISO, take down a list of the supports, go to the support issued to field copies and find out when they were constructed, and you would find that one particular support was added. Otherwise, you'd never know it. It's a very intentional type act that they didn' t expect anyone in the NRC to go through enough to ever determine.

BISHOP : Mr. B ishop. Mr. Stokes, you have made f recuent reference to "they". Can you be more specific as to who was providing you this in- truction and what form that instruction took?

STOKES : Well, I put it in my document. The person l responsible for trying to make this all work and to my knowledge r

the one who was in all the meetings was a Mr. Marvin Lepke. From-Stokes /12-7 42

the people that I was working with and all instructions came f rom Mr. Lepke on how we would do this. Most of the things that I didn' t work that they had to modify, the person who was in charge, Mr. Leo Maagoba, didn't ever stand up for any of this, o When we first started working in November through March, he did support me when I was requesting control documents in that one of his group leaders, Mr. Kevin Wah, had a letter typed up by the secretary and had that list of all the people in our trailer submitted to management, which was Mr. Lepke and Mr. Bob Omen, who was the Project Manager here at that time. I was told that lettet went to one of their desks and was put in a drawer. When we followed through on it, we found out that they decided, Mr. Lepke told me personally he didn' t feel I deserved control documents. And, as f ar as cordiality between me and Mr. Lepke and Mr. Bob Omen, Bob was f riendly to me, he acknowledged my presence. Mr. Lepke, on the other hand, never even admitted that I existed, even if I spoke to him first. His hatred for me pushed, in my opinion, the points which I was bringing up and he could not stand me to question his authority for a decision. And I'm sure he was predominantly in control of making those decisions. I have raised questions of who was in management and who was the actual control leader. Bob supposedly was the Project Engineer. Mr. Lepke made the decisions. I don' t know if he ever consulted with him. I wasn't aware of it. He was very prone to just tell you what he felt right on the point. He didn't ask to delay it and talk to Bob. He was very instrumental in doing what he wanted to. In this case, I know for a f act that i

Stokes /12-7 4 3 i

he was the one who thought up all this rigamarole for covering up deficient supports.

BISHOP : Bishop. Thank you. I'm sure our investigative staff will have some more questions for you on that topic but at this. point let me turn it over to Mr. Ebnedder to continue through the affidavit.

~

EBNE DDER: This is Ebnedder. Let me ask you something on page 10 and 11 very briefly. And I just need some answers to the fact that you can supply us with information to verify them. It-has something to do with falsification. On page 10, you said the

~1ogs had been f alsified and at the bottom of that and the tor of page 11 you said you would share the falsification of these i records with the NRC. Mr. S tokes, do you have actual documents l that will verify f alsification that you can provide to us?

STOKES : I have Rev 0 copies which were performed. I'also

have final copies which I made before I was terminated. Whether l or not they've been modified, I'm not sure. I can provide the 5 large sheet showing that one guy did it initially. The final drawing shows that he had nothing to do with it. I can also show certain cales that failed. The log sheet has been modified to show final calc and final cale shows that it was done by someone else other than that.

i EBNEDDER: Ebnedder. These are related then to the packages l we were discussing with Mr. Yin so that problem could also be resolved with the ... The OI people should definitely be ~

involved when they look at... Let me move on to Sam' Reynolds area in the flare bevel and the weld conditions. . Sam?

Stokes /12-7 44

j. REYNOLDS : Reynolds. The first question I'd like to ask '

> concerns'a hangar attachment. Mr. Stokes, did you supply the

hangar welded attachment we received?

STOKES: (unclear)

REYNOLDS : Did you supply the hangar welded attachment we received? And if you did, can you indicate where this was received from and under what conditions?

STOKES : Sir, can you describe the hangar welded attachment that you received?

5 EB NE DDER: This is Ebnedder. We have a hangar support, a large.U-bolt that has some laminations. Tom Bishop has just brought it in. I guess what we're really interested in where did that particular support come from in the plant?

STOKES : The particular item you referenced as a U-bolt is a i

spring can supporting end bracket. It's a structural beam attachment. It's a U-shaped item. I can' t say that I supplied

it. I'll have to plead the Fourteenth on that. I will say I know where it came from. I can show you the support that that

~

was to be installed on and I was hoping to lead into that without

! having to tell you that I would plead the Fourteenth. I probably could be pressed for thef t if I admitted taking that and I won' t.

i

??: That's fine Mr. Stokes. We're interested in what system it came from, the location in the plant. So you can supply that, though? Sam? .

REYNOLDS: Mr. Stokes, are you aware of the practice of i meeting the qualification requirements stated in the last l sentence of D 1.1 paragraph 5.2 by performing procedure qualifications in accordance with ASME Section 97 -

Stokes /12-7 45

- _ _ . - - . _ . _ _ . _ . _ _ . _ _ _ _ - , _ . . _ _ - . _ - _ _ _ _ _._, - - - _ . a

??: What do you want to know?

REYNOLDS : I just want to know if he's aware of what- it says 4

- in paragraph ...

STOKES

Could you quote what it is? I do not know the Code

'- by heart. I have a copy of D 1.1. I have a copy of the Symbols _

section and the Code. I have read it multiple times, and I have

~ redded out and yellowed in many of the sections. I am sure that h I have looked at that particular section but right off the top of my head I can' t tell you that I' ve ever seen it. I can't auote it. I do think I am f amiliar with what you' re saying as concerns testing, destructive testing. I wouldn' t say that for sure. I do know that you do not have to comply with AWS if you' re willing 4

1 to do your own test program and justify your own procedures.

Those have been done adequately here on the plant. The procedures for the welds, Pullman's weld procedures, were written. I have a copy that I got from QC department. The first statements of all the procedures state they were written for piping, pipe supports, pipe attachments.- They were used for pipe supports, not attachments to pipes. The pipe supports, on the

other hand, per what we were instructed, were governed under AISC 4

and they did not even want to admit AWS was included in AISC.

They instructed us to use precualified partials and full pan

! welds per AIC. Those same welds go right back to AWS. It's just i

a different way of expressing the call outs. In the AISC they t

give you a local diagram, explaining exactly what is required.

) AMS gives you a chart with multiple call outs and multiple weld procedures. Whether or not that answers your question, I can't I

Stokes /12-7 46

-. _ - -. .. . _ _ . ._ _ - . ~

[

say. But I spent five years trying to become more knowledgeable 4.

on welding.- And I'm still learning, but... I was used by the management group that I was in to write a weld paper on hon to handle flare bevels and flare B welds and how to preo them, how to call them out so that we:would get what we asked for as far as design. I'm familiar with what they asked for as far as

~

design. I'm f amiliar with AWS limitations on penetration, as f ar

. 1 I'm also aware as installation of weld material in the throats.

I of the angle requirements. The weld procedures that Pullman had here'that they used on support steel did not meet the minimums per AIEC, prequalified ' minimums per AISC, pre-aualified minimums or full pans. On top of that these various DRs which were written against the piping. The co-hydro systems were U.T. the i

^

weld joints, full pan. There is DR written against those joints, e

showing that they're deficient. They didn' t get adeouate throat

! penetration. That was on a pipe weld joint per a pine j procedure. They didn' t get it on the pipe. They didn' t get it on the supporting steel. They used the same thing on both.

EB NE DDER: This is Ebnedder. I'm somewhat curious, j Mr. Stokes, which weld are you referring to that were U.T.'d, Phillip, flare bevels, or pipe welds. We were talking about flare bevels, and those are rather difficult to U.T. But I'm just curious if you could identify which ones failed by U.T.

STOKES: Well, when I mentioned the DR, I was talking in

! reference to the co-hydro pipe systems. They U.T.'d the full l

penetration call out welds on those systems to determine full i

penetration. A U.T. exam will show a deficiency in the throat Stokes /12-7 47

due.to an echo. Those particular welds are flare V penetration welds, partial groove welds. [someone asks cuestion, unclear).

"No, that's on pi. Okay, the flare bevel welds, per Pullman weld procedures, were handled under fillets. They were shown because

e of the management's decision as full pens. Our group, however, the paper which I was requested to write, which was sent to San

Francisco and never used, recommended that these welds be called i

,out with an S and E per AWS considering the drawings to be working drawings. They were also. shot drawings. The materials were fabricated. We used, in my group, an S call out based on 2T, ba. sed on a manual I was supplied by the American tubing manufacturers, telling me what the minimum radius was they would

roll to. The E call out was based on AWS minimum reauirements of 5/16ths or well, anyway, there's a table in there that tells you, for the radius there's a minimum required weld which is allowed. So we put down those effective weld sizes.

??: Let's clarify that. The T dimensior. was f rom American l Tubing Manufacturers.

STOKES

That's in my group. That's all covered in the weld DR. I think it pretty well explains that. I do know where a support is, or did, I worked on it. It was a Westinghouse i support which was crossing tube structures. All welds at that joint are flare bevels. The OC, all they had to do was'look at l

the material. They did not check the radius of the material.

The City in the design group was using 3T for that calculation.

i We used 2T. We found out here through Pullman requisition that steel had a minimum radius of 1.5T which meant that the cales Stokes /12-7 48  ;

1

1 coming up were defective per AWS for that radius, were not conservative because we all used too high a number. On top of l that QC did not have to inspect the joint for adequacy in any way. They didn' t have to measure. The biggest criteria was that l

. it was filled out flush, which should be adeauate for the Code.

The problem comes in in the documentation. The design calc shows

in the case of the trailer I worked in, a 2T E calculation in l most cases, San Francisco at 3T. The other trailer over here,

)

i probably 3T. But then, there's no guarantee that the effective throat, as specified, per the minimum, is there. The AWS specif,ies even that effective, there's two quick ways that effective can be determined. It's the minimum per the throat 4 size for design, but then the minimum installed is a minimum based on thickness. The required well for installation was not, l there's no procedure for doing that. There's no QC inspection for it. The design was based on something different from what

{ they got because the material is not the same. It's a very, in other words you can' t insure the designer got installed an l adequate well.

BISHOP: Bishop, I'm sorry to interrupt again, quick chance j in subjects. Mr. Cruze has come in with one of the documents.

He wants to make a statement about that and then he'll leave.

CRUZE: This is Jess Cruze. Actually, what I have in my possession, which I' ve had in my personal possession' for about

, fif teen or twenty minutes, are the complete packages, calculations and calculation logs for the first one which you seemed to have special interest in. I just went and assured that i Stokes /12-7 49 l

. . - - - - . . -.- ,r-_--____.. -- - - - _ . - . - _ _ - . .- -

we had that as a top priority. I have that in my possession. I J

have the second one which Marvin has pretty much assured that that was together. And then I have an additional one which is the eleventh one on Exhibit 1. And so, I just want the record to o show we have those now. It's our intent to maintain these in our j possession for the foreseeable future.

REYNOLDS : Reynolds. Mr. Stokes, would you be satisfied if the flare bevel welds are full welded with a slight reinforcement on the outside, that they would meet the 5/16th OI reauirement?

STOKES : On that particular point, I would be satified if projec.t management, there are several possibilities on how they can solve that particular weld problem. I've discussed that with Mr. Ken Palmer and also Jeff van Klomptenberg who was my superior when I wrote that weld DR. Both of those two people are mentioned in my DR as having talked with Mr. Curtis in San Francisco. They either need to review each of those welds in calculation based on 1 1/2T to determine whether the weld is valid or they need to add additional welds. They don' t want to add additional weld, it means additional work for the field.

They, more than likely, don't even want redo the cales. I'm not sure which way management would decide to do that. I' d be satisfied with any procedure for assuring the adequacy of the joint, period.

REYNOLDS: I think the ouestion is if they'r'e full welded with a reinforcement, are you satified that that is in accordance with Dl.1 rules and if not, how would you propose that they add more weld metal af ter it's full welded with reinforcement?

S tokes /12-7 50

._ - . . . .. .. . - - - . - _ _ - . . . .- -. -. - --.. ~ - -

4 DEVINE: Mr. Reynolds, if we' re going to be basing his answers on D.l.1, paragraph 5.2, could he have a copy of that to 1

j be looking at?

l REYNOLDS : We got of f that nurber. i DEVINE: Okay, okay. Thank you.

4 l ,

STOKES : I think I'm getting at what you're saving. If it's flush welded per the weld code, it's considered effective as a

-prequalified weld on flare bevels. That's not a problem as far i

i as I'm concerned. The welds exist._ The problem is in the linking between the design cales showing that that weld joint is

~

adequate for the loads that are imposed on that joint. There's a breakdown in that documentation on that particular weld. It doesn' t exist. The joint is valid per AWS as installed because i even if it was welded as a fillet, but you need to know what weld 2

size to make sure that the penetration was effective, there

should have been a procedure established. The procedure was for

a fillet. Fillet, there is a big difference in how the joint f angle is limited. The joint configuration, the procedure that exists for when it was done. I'm sure that the joint's flush.

Most cases that's the way welders weld it anyway, _they filled it 4

up. But whether they started with a quarter inch rod, an eighth inch rod, a tig weld, depends on how much penetration they got.

And they were required, or are required, to provide the minimum

!l effective weld, per the code. Based on thickness. They have no procedure for establishing that minimum. It may be flush, it may

tm adequate. But how can you tell me it is? You don' t even know i

how deep the weld goes. It may be full of slag. Maybe it's just l

Stokes /12-7 51

a superficial cover weld. I don' t know that. I can' t make that

decision. I don' t, I didn' t QC the joint. I didn't have access

-to a procedure they installed it with. I didn' t supervise the

, weld size, the rod size he used, the voltage he used. All these i ..

things aff ect how much penetration he got. I can' t guarantee

! ,that the effective minimum that was required for that joint per the code was established. Because they didn' t .have the j paperwork, the controlling factors controlled to make it right.

I don' t know if that's answering your question. This is a very

... It took me an awfully long time to get this in my head l straight'. Because it's not just an installation thing. It's an i

installation, it's a design, it's a QC, it's, in other words, I'm

b responsible as an engineer, or the project engineer is responsible for every design from the design point through the installation of what he asked for, to insure that he gets what he

, wanted. There is no documentation on that on these welds. There is no assurety that he got anything that he asked for. There's i no simple valve on these joints. I personally think each needs

to be U.T.'d to determine the minimum effective that exists. And j then the cale needs to be rechecked to see what the loads

! require. I never will agree to put this simple code minimum on l any joint out there without reference to loads. I'm not going to make a determination that the minimum is always going to buy off l the joint. In many cases, that's true. But there are supports l out there where the minimum wasn' t even close to buying off the 1

i joint.

l 4

l S tokes /12-7 52

. _ - . . _ _ _ _ . . _ - _ _ _ _ __ . _. _ . _ _ __ _ __ _ --_ _ . _ . _ - . _ _ _ _ _ _ _ _ . . ~ .

. ~ . _ . . _ _- _ _ _ _ ___ _ _ . _ _ - . - _ _ _ _ _ _ _ _ _ _

s l

BISHOP: Bishop here. I.just want to make a point and how

'do I put this diplomatically? You were asked the question of f

something to the effect of would this satisfy you? I don't think, ...

STOKES: I just said no to that. 'I just told you what would

! , satisfy. On those particular joints, I know the procedures weren't there for getting the weld that was asked for. The design group didn't have adequate information on the joint material being used for the joint, to insure that a conservative

];

joint was installed. And, to just put a flesh weld and put a little e"xtra, no, I won' t buy that.

f

??

I understand your position. What I wanted to say in that regard is that again, we are not here to, we're here primarily to get more information. We will not be askina you in i this process to approve or disaporove the conclusions we draw.

So I would ask our staff again, in asking our questions, let's f not put you in that position. We' re primarily interested in getting more technical information and when things wrap up, we

! may agree on issues, or we may agree to disagree from a l

l professional standpoint. But, that won' t be the topic of today's j discussion.

l EBNEDDER: This is Ebnedder and I totally agree with that.

l l

We're here to We're not here to satisfy you or the licensee.

l determine that the plant is safe. And that's our sole I i i

j. objective. And your information will clarify that for us but the l 1 staff should not be going over what we have seen or not seen or any other information at this point because everything is l Stokes /12-7-8? - 53 l l 1  :

l _ - . , . . . . - , , - - - _ - - . . . .

preliminary. So, again, Mr. Reynolds, do not put Mr. S tokes on

, the spot and do not make any premature decisions on where we're at.

DEVINE: If I could, as we said, we don' t mind being put on a

the spot at all but I agree that it raises a little question in t

,my mind to have a check on what an acceptable solution would be

before you all flushed out all of the additional information. I i sort of thought that that would probably come up in follow up 7

l interviews to resolve Mr. Stokes' allegations af ter you all finished your homework which he is here to help you on.

EBNEDDER: This is Ebnedder. That's true, any follow up would be through Region V, Mr. Bishop. But right now, we're just l

looking for clarification. So who has the next question? We do have the package here. Do you have any more on this flare bevel, i

Sam, do we want to pursue anything on that?

REYNOLDS : No. Mine were general, many of mine have been answered. For example, names, and just for the record, you've indicated you do not want to supply names, for other reasons, at j this point. I think the major issue, Paul indicates, do you want f

to discuss something different, Paul, or you want to get back on this design package with Esa?

PAUL Back on the design package.

REYNOLDS: Okay, let's get back to this major issue,

Mr. Stokes, and Esa Yin, and Paul Bessler. We do have the first package over here. Now, on the back table, it's on a separate three packages, it's on a separate table, we do not want those papers to leave that table. They're originals. All discussion.

Stokes /12-7 54

t should be done around the table without moving papers back and forth. If you want the blackboard over there to sketch on, we will move it back there. If you want the tape recorder down there, we will move the tape recorder down there. If you feel o

1 you want to record every paper shuffle, that's fine. But is

,everybody agreed to start looking at these packages? Let me ask Tom, first, do you have any questions that you want to ... So, this seems to be the big issue, let's get on that...

YIN: I would rather ask all my general questions and then

, go into the specific calculations that were brought to our j attentio'n, if that's okay with everybody. Get back to what we have left over. I left a short while ago talking about the M-40 4

calculation. I wasn' t too sure what the point Mr. S tokes was i making. Is there any problem with the M-40 calculation or is this concern that MU101, the computer program, should not be run, j to the M-40 calculation is done? What is the specific point of the problem that he wants to bring to the

, attention of the NRC?

! STOKES : Esa, what I was trying to point out, is that M-40 l cales had been done previous to computer cales on these

, supports. And that the computer cales even though they were done subsequent to that cannot be automatically assumed to be better i

than the M-40 cales and any subsequent computer cale any better than the one run previous because the assumptions that were being fed to the M-40 groe,o from San Francisco for the local conditions and other things, the seismic load data, they didn't input data, it came f rom a computer table. The computer saw it, they never 4

4 Stokes /12-7 55

l l

saw it, Do you, I mean, I don' t even know that they saw l anything. What I'm saying is you can't say just because Rev 0 is preliminary stamped on the drawing, Rev 1 is preliminary and

'O Rev 2 is final, and Rev 3 is as built, that Rev 3 is really any better than Rev 0. You have to look at what they put into each

,one of those. And each one of those is an awful lot of ifs, because the stress group were constantly being updated with locas 4

. and the seismic load data. We also were being constantly i

updated. The problem was we weren' t being supplied revisions.

4 At one time, when I questioned that, I had it assigned to a Mr. Kevin Wah. He is new the one who's assigned in the trailer that I was in the control seismic updates. It took an awful lot 1

of work to get him to agree to be signed up. But he was a Bechtel direct. But what I am trying to say with that, I think I just made it clear. Just because they put on a calc cover sheet i that one is preliminary, two is final, hell, you don't know if that's invalid, you don' t know if two is a valid calc justifying

that hangar. The stress conditions, the modeling, if it's all incorrect, and then the loads given the hangars are incorrect,

!, and then those guys are being forced to model things on strudel incorrectly. The whole end result is you've brought off a hangar i that never looked good. It's never been good. But it looks good unless you can do a really deep analysis. That's what I was i

j trying to point out. Sure, I know they' ve got four revision on some of these. I've seen them. I know they've been changed.

??

Is your badge up on the hill, Esa, or down on the bottom?

i Stokes /12-7 56

4 l

YIN: What's that?

??: Your badge, the one for the plant.

YIN: Down.  ;

??: I want to make sure so that when we go out on the tour, everybody's together. . .

STORES : I probably won' t be allowed on the tour.

l  ??: Yes you will. That's why I wanted to make sure.. .

YIN: This is Esa Yin again. Let me see if I can summarize here your specific concerns. For the M-40 series calculation, i

for the three span peak cobric calculation, you have specific span to work with, and you have a specific dynamic load factors that you can input in all three directions of seismic l calculation. For'the system line itself, you have the C series joints to work with. And subsequently, they summarize that, and 1

i pick the maximum, yeah, pick the OBESSE and G

, values so you can use them. So you feel comfortable with what i you were doing in regard to the M-40 calculations. As far as the l present program, the ME101 calculation, the combination of various response back from input, is specifically assigned to be i

done in San Francisco office. And the whole package is performed 4

and is not available for you to check review at the site. And l you are questioning whether or not the response back, the sets of response back , for various calculations was indeed i adequate to perform the calculations. Is that correct summary of l

what you have stated?

STOKES : Let me break here before I answer that and change the tape in my tape player.

S tokes /12-7 57

__. _ _ . _ _ _ . . . __ _ . _ , . . . _ _ . _ _ . , _ - . . _ , _ . _ _ _ . _ _ . _ . _ _ . , , _ _ _ _ , _ _ - . _ _ . _ , . . _ _ _ . . _ , . _ - _ _ ~ ,,1

I

+.  :

DUKOFF This is Dukoff. The time is approximately 10:57 a.m. and we'll be going off record so Mr. Stokes can change his tape.

DUKOFF This is Dukoff. The time is 11:30 a.m. We're now i ,

returning from the break. We're going to go to the other side of

,teh tape so we can have an uninterrupted recording.

! DUKOFF: This is Dukoff and we will be continuing with the interview after a break. Mr. Stokes will be answering the next 4

question.

STOKES : I'll try to answer Era's (sic) auestion which was a summary'which was quite lengthy. Um, he asked me if I was comfortable in my doing of the M-40 cales. And the answer to that is yes. But I don't think that that is the question that j should be asked. What I think he was trying to get at, and I tried to make some notes, as f ar as the seismic data from San Francisco, which they had control of, that was supposedly done by Blum, the seismic company responsible for the data. All that San Francisco supposedly did was to route copies of that in a controlled circumstance to certain people. As far as their control of input, the only input where they were completely i

unsupervised that I know of was in the tables which they l

l generated for the stress people to use in the stress cales.

Those tables were already placed in the computer. No one here at site saw what those tables were. We, I don' t question those, I-never saw them. But he also stated that we didn' t, I don' t i think, saw those seismic tables. We did in pipe stress receive C-17, C-28 and C-30 and we did use those tables daily, Stokes /12-7 58

, . - . - - , . , --,---,m. . _ , . - , . .- , - , - - - - , ,_,ye- - ., . ,

initially. In Unit 1, the stress group was required to do a stiffness cale or use or stiffness deflection in the line analysis of 15 hurts. Pipe supports were required to do their justification of 20 hurts. Based on this, when we were doing M-40 cales, I and Mr. Victor Chang worked up a design table for l all the areas in the plant within a reasonable affected category. And we based that table on a 15 hurts stiffness, because stress, under M-40 cales, was not peak in all cases. It was based on 20 hurts, or 15 hurts. It was just like a computer analysis. We didn' t use people. We used 15 because, for the

hangar c'alc, because it was conservative slightly. We did not,

( and management did not want us to sit down and generate _two 1

tables. It tookme and Mr. Chang roughly four days. He generated some, I checked, I generated some, he checked. We ended up with tables which the support people could go to and reference in their cale to select the seismic accelerations to be used in their cales without having them to go to C-17, 28 and 30 because those tables required an awful lot of time to go into every time that you did the same thing over. It was redundant for one l

person to look at one table, determine peak and then determine i

the fif teen-twenty hurts readings. And so, we did it all at one time to increase the production of all the people in the group.

San Francisco did not supply anything other than the curve data. They did not but we had problems locating the latest revisions to that. We did not want to do it to an earlier revision and then turn around and find out we had to go through l all our cales. We wanted to do our work on as late an edition of Stokes /12-7 59 i

information as possible. The peaks of the seismic curve data were only used in the three-span, code break calculation. They were specifically specified. They were specifically, the loadings were specified. We were dictated to a certain span of

. I length, aof pipe, which was governed by the schedule of the pipe jwhen it filled with water, that can be determined from almost any pipe data anywhere by any company. So for the three span loads, that was spelled out, no question. I have no comment on whether I agree or disagree. But, for the support steel and for the M-40 calcs, we were at liberty to decide what the accelerations were to be us'ed in our cales. On Unit 1 I had an instrumental, I was

instrumental in insuring that conservative numbers were used, and not overly conservative. Management agreed with that.

Ultimately, those numbers could be reduced slightly and probably were. I think that pretty much covers what I thought he was

asking me and I hope it does but if not...

r YIN: Esa Yin again. So, in summary, I don' t leave the I neighborhood to extract any of your specific concerns. So, unless you provide us more information on your specific concerns, then there is no way we can follow up on this particular area.

STOKES : I didn' t have any specific concerns with this, other than the fact that I used this as an example along with the loca data that was supplied to us by the city, which they ultimately changed their decision on, as far as zeroing the movements at ground level. I was attempting to state that since we had no control over the review of this stuff, or the seismic data in' the tables for stress, that I would at least like to i

Stokes /12-7 60

review the determination of that, but I'm. not going to auestion it, - that it was signed off and done by someone that I suppose,

.which did it. correct. I do know that stress used data supplied to them by the City that was incorrect. Based on the city's i assumptions, it came from upper management, and not only I, but j ,Mr. Kim, who was, at ' the initiation of our M-40 group, he was our

) lead. He also agreed with my decision that the locas, or the 1

building movements should not have been zeroed at ground level.

i He also supported me in pushing to make sure that they didn' t j have that crummy information. I wasn't trying to make a specific point to bring up. The points I'm bringing up I pretty well j specified in my paper and I can provide information.  ;

YIN

Esa Yin again. On your affidavit, page 4, the first paragraph, you stated that you have alerted your group leader, Leo Mangova, regarding of lack of control procedure for performing the calculations. And in December, I presume it's j 1982, you sent a memo, uh, he sent a memo to Lepke and Omen. I j talked to Mr. Mangova on December 6. That was yesterday and he

denied that he had ever written that memo. Do you in fact know the memo was there and you actually seen the memo and you have a i copy of that particular memo?

STOKES : The section in question in my statement doesn' t actually say Leo wrote the meno. It says he sent it. I

mentioned earlier in en earlier answer that the memo, as far as I <

know, was written by Mr. Kevin Wah, who was Mr. Mangova's i

assistant at that time. He, with Leo's blessing, could not send 4 that memo above Leo's head without his approval. Whether or not

S tokes /12-7 61

-.-._ _ _ . - _ _ _ _. _ __ - _ _ _ ~ . _ _ . . . . _ . _ . _ ._- . . _ . _ _ _ . . . _ _ - . _ .

, he sent it, I mean I didn't say that he initiated it, I just say that he let it go above his head. Intercompany policy is that you go through channels. You cannot go above your supervisor's

- head normally. But there are occasions when you can. In this

i. case, he did support it. I didn' t say he wrote it. The memo, I

, don' t know if Kevin has a copy or not, Mr. Wah. But the memo was '

sent. And it was sent to Mr. Marvin Lepke and Mr. Omen.

Mr. Mandova's entire, let me explain his relevance to this entire position. Leo's intent in every asset of his existence to this 1

plant has been to do whatever was necessary on his part to make himself 'look good. He would do whatever upper management wished as far as trying to make himself look good. He would comply with their any whim. He was not a no person. He was strictly a yes man. He tended to be close-lipped, he very seldom made any l

announcements in the trailer. If someone brought up something, j he'd have them do it. He didn't do it himself. He wrote very j few memos where he signed his name, pinning himself down. Not '

only he but the group leaders, Mr. Alex Schusterman, they were very deliberate and they would not put anything in writing that I

j. questioned. I was specifically told by Mr. Schusterman that he wouldn' t ask me to do anything and put it in writing. He would l only ask me, and if I was willing to do it,-then it was up to me. But he more than once informed me he was aware of the i

$10,000 fine and 10 years in jail penalties for altering and i affecting cales. And the possibility of him being punished.  ;

l Like I said, they just picked people who did not question it.

They didn' t have to push me. They didn' t have to give me a memo 1

4 i

I Stokes /12-7 62 '

.. .-.. . .- - - - .-. - . . - - . _ _ - -. . ~ . . - - - _ _ _ - _ - - - - - _ _ _ _ _ _

to get me to do it. And I was not the only one who told them they'd have to put it in writing.

YIN: However, this is Esa Yin, however, there was a memo, according to your affidavit, that was sent to Lapke and Omen.

Whether it was written from Kevin Wah or it was written from f(r. Mangova doesn' t really concern me one way or the other. Have you actually seen that particular memo? If you do, do you have a copy of it?

STOKES : I saw it when it was written. It was a cover letter and attachment sheet listing all the employees in the group. 'I did not get a copy of it. There should be a copy in '

the ... there's a file of interoffice memos. The secretary maintains it. Having questioned him on that, it could very easily have been removed. As I said, what I was pointing out, if you ask Leo, there's no way you can prove that he would admit that existed because he didn' t sign it. Even if you can find a copy, you can' t prove he ever knew about it. Because he didn't sign it or initial it. He could very easily say it was done without his knowledge. Questioning him on it, you' re not going to make him say he did. That's what I am saying.

YIN: I'm going to jump to the next question. On the same i

page, page 4 of your af fidavit, quote "Within two days, however, before I went to the NRC, I received the necessary control document. In this case, the M-9 design guide for engineering."

I have two specific questions, would appreciate you answer it.

The question number one, you didn't mention the time you receive the M-9. Is that in January,1983, you receive that M-9 Stokes /12-7 63

d

}

l document? The second question is, is it true only M-9 was received by you as the sole document for support design?

STOKES : I can' t say for absolute that I got the documents in January. It should be possible to verify that through the QC depar tment. However, they may only keep the latest revision dates and not when they first issued that to the person. It was i

either the end of the December or the first of January, I will say that. Initially, I didn't really get documents that were worth a crap. The M-9 I got was three to four revisions each of the sections out of date. What I did get was that copy assioned to me and issued to me in my name and a cover sheet listing the

')

latest revisions that I should record in my possession for i update. And once I filled out that cover sheet which everything on it was out of date, it was mailed to San Francisco. Within a i

i few days, I started receiving up to date sections of M-9. Other

than Mr. Wah, Mr. Alex Schusterman, Mr. Leo Mangova, and maybe 1 l one or two other guys, Bechtel people, that I can' t remember l their names at the moment, I can't say that anyone else other l than the Bechtel people and myself up till well past March, had

! M-9 control documents. What they had done when we first were i

! employed in November was to xerox only M-9 and a few specific i i

sections out of the appendices for M-9 and give these to the unployees. The control numbers and the peoples name that they

! had been assigned to that they zeroxed they blacked out before I

they zeromed it because they didn' t want to get them in ,

l Technically, those documents are never supposed to be trouble.

xerozed by anyone other than the controlling department. They're Stokes /12-7 64

, , ,--.~.m,,._e . , . - - . + - _ . - _ - , , _ . - . _ . . . - .

.m..--.___.

. - . . _ . - -- - , , - . _ , - - _ - - ,,----.-.,--.~--,_,.-..-..4 . , , - . - . , - _ _ , . _ _ - - -

i I

only supposed to be issued as a control document and that zeroxing and control goes through Oh0C program. I do know that in July there were several people who still had not received control documents. Those documents came to them while I was in f .

3 the quick fix group assigned to Unit 2. And I know that when I

, returned to the trailer two weeks subsequent to my beino terminated, the QC guy on site came around and he had a problem he was trying to determine who had not gotten revision 12 to their latest documents. They still were havino problems making a

j sure people had them. But by that time, they were trying to make sure everyone had it. This was nearly one year following our j original work started.

DEVINE

Excuse me, this is Tom Devine. Charles, you mean i two weeks before you were terminated?

i i STOKES : Yes, two weeks before I was terminated, the QC guy

! came to the trailer and he had a problem that he had three people who, according to his records, had not received revision 12. And he was getting all over my supervisor of the trailer, Mr. Roland ,

f Mannaplatt, at that time. And they issued an ultimatum, just 1

told everyone in the trailer to stop whatever they were doing and see what revision they had. It was that important to make sure that they had it. But subsequent to that, for nearly, for well I over half a year, I was the only person other than the direct

, with Bechtel, in my trailer, there were no PG&E directs lLn that 1 l

trailer. I'm sure Mr. Lepke probably has a copy because he was

! assistant project engineer and PG&E direct. But what I'm saying

is that people who were assigned to do the calculations, who l

Stokes /12-7 65 l

needed to look in these documents all during the calculation process, did not have control documents. They did not come ask me.

YIN: I think you addressed the first question. The second O

question I want to know is whether or not the M-9 document is the

. sole document for the support design. The reason for me to ask the question is I have reviewed M-9 and it doesn' t tell you how to design a lot of things. It doesn' t tell you how to check and verify the calculations. It doesn't tell you how to control the interfacing. Now, ...

i ?it Excuse me, Esa. . .

1 YIN: So it is important for me to know, have they provided additional documentation in the area of like procedures, and control design criterias, and acceptance criteria and that kind of thing to perform the work.

STOKES : I didn' t get that question when I took down what you were asking. This is Charles. Initially, when we went into the plant, we were given a xerox copy of M-9. I said also appendices. Occasionally, we were also supplied design guides that Bechtel had issued in their office. Those were also xerox copies. We did not have control copies of those. I did. When I was given that update sheet, I was given an update sheet for all the C items, P items, some M documents, I can't remember any other areas that were concerned. But these were Bechtel inner design guides, which management had decided were applicable to Diablo Canyon. Initially, the only ones we were shown and given to use were some concerning a seismic stiffness calculations and Stokes /12-7 66

. ~ . . , . _ . . -_.

a table simplifying how to come up with an equation for the stiffness.. It's just inverted deflection. But it had certain

, cases shown which were applicable, made it very cut and dried, easy to do. That was zeroxed. When I finally started getting my 0

, control documents on all these areas, other than the Bechtel directs,-I was the only one who had all the guides. Not only that, I was the only one that was sure to have the latest f edition. Even the Bechtel supervisors in some cases didn't have l the latest. The other thing is the people we had been working i

j for here, the group engineers, Bechtel supervisors, had been f supply'ing us these documents from their catalog of what they j had. Once I got the cover sheet, indicating what was valid for i

this plant, in many cases some of these documents had been voided i

4 by upper management. They were still being pumped into the group to be used. That upset me too.

YIN

Esa Yin. Does this Ps and M documentation, are those control documents? What I'm saying is I know M-9 has a control s

[ number assigned to you so that they can keep track whether you have received the new, the latest revision and so on. Are those I l

, Cs, Ps and Ms, are they having the same control number assigned to you?

f l STOKES : Yes, they did.

f YIN - Next question, on page 10. Let me repeat again, I i

forgot to turn on the mike. This is Esa Yin. On page 10, of l

I

! your affidavit, you stated most of the affected engineers object l

. to modifications of prior engineering calculations by a new staffer. I have two questions. First question,_can you name i Stokes /12 7 67 i

l some of the affected engineers? Second, who are those new 1 staffers, assuming there are many staffers? How long have they been working at the site?

STOKES : Okay, yes I can name some of the people who

objected and I don' t mind doing that because I don't think that's, I'm not going to pin them' down. One of the guys was

.Eddie Rasinella, another was Gary, God, I...

1  ??: I have a list, maybe you can...

STOKES : Yes, that may be easier.

4 DUKOFF: Mr. Esa Chin (?) has just given me a document dated November 29, 1983, PGEE construction on loan. I'll be giving this to Mr. Stokes so he can identify the names for us.

STOKES : Okay, this list is current. To be accurate, Ira

(?) needs to get copies of the people who were here that have been laid off or have subsequently lef t since March. Many of the people left here have never objected to doing anything. Many didn't work on Unit 1 during that first initial review from i

November to March. There are one or two, however, who are here, t.nd wer affected by the last statement, concerning to objecting i

! to modifications of their calculations with -them still being 1

present. Mr. Gary Katcher, K-a-t-c-h-e-r. Boy, there's not too many left on this list. I'm not on the list. Mr. Gary Katcher is the only one. I believe, well, Mr. Howard Bernstrum, was not there during the initial phase. I believe if anyone questioned him concerning my alteration of cales he performed, he would say he would object. Even though I don' t think he was there to have objected at that time. Boy, wish you had the first names on S tokes /12-7 68

, these guys. Ed Rasinilla (?) was another. There are some people here that are listed in the stress group which had nothing to do with this. I knew most of these people by first names. With initial and last name it makes it very impossible, almost

impossible unless I knew them very well, which I didn't. But I do have three here. The ones that aren' t listed on this sheet was Victor Chang, ...

YIN: I'm going to give you another list, with all the names.

DUKOFF: This is Dukoff. Mr. Yin has given me another list we' ll 'txt showing to Mr. Stokes and it's a four-page document. At the top is written OPEG Personnel Roster dated 11/23/83.

Mr. Stokes.

STOKES : Well, based on the date of this, if this is the latest listing of who works here, there's still a very large gap. It's dated 11/23. Many people lef t way back in January, not January, but June. Mr. Victor Chang lef t in the middle of the summer. He's not on the list. I'm not on the list. I don' t believe. I'd be surprised. Let's see. No, I' ve been excluded already. Should have been by 11/23. Caesar Orbeda is not on the list. He lef t roughly in June.

??: Excuse me, is ho someone who objected?

STOKES: Yes, 0-r-b-e-d-a. I' ve been in contact with him.

He told me personally he would be willing to sign an affidavit.

Yes, you want it on tape. He's located in the Richland, j Washington state area. He does not want a lot of publicity. .But I he said he would be willing to sign a statement as to anything S tokes /12-7 69 I-l

l that was true. Many of the people that objected did not object .

too strenuously but these people were more vocal. ,

l DUKOFF: Mr. Devine, the reading of these names into the '

record, is that sufficient with you, that we don't need to mark these as exhibits?  !

DEVINE: That's fine with you, okay, great.

YIN: Well, this is Esa Yin. The second question indicates that the second question was, are those new employees, seems to be implied that they may know what they were doing. They may know what they were supposed to do, or maybe not, probably traine~d, indoctrinated and so on and so forth. So can you give me more information on them?

STORES : Yes, I'm sorry I didn' t answer that part Hof it.

1 When I said new staffers in my paper, I also stated that in March, the trailer was split. Initially,- there had been one trailer, one twelve wide by' seventy feet. There was thirty-five engineers in it plus group leaders and Mr. Leo Mangova.

Following the date, there's a trailer established roughly right off the front road. Mr. Mangova was going to transfer himself to.

that trailer. The original staff of thirty-five was being split l 1

into two groups. Basically, the people that were being I transferred to his trailer had not objected which, with anything t i i they had asked them to do. Those people were going to be I assigned, continually assigned to Unit 1. To finish out that

. railer, which he moved to, they hired new staffers, people that had no past experience wi'h the plant. The people that had i

initially. worked on Unit 1 had already gone through the Minnesota -

- Stokes /12-7 70 1

s Multiphasic Test for security clearance to Unic 1. They had gone through the applicable security screening as far as what they've done for the last five years. I myself had been badged for Unit 1. I had completely free access to go anywhere I wanted a

without having to request someone to accompany me. This was a very critical thing to engineers working in Unit 1 because they would be asking them .to review hangars and it takes an awful lot l of time to do your work when you don't have that freedom. The people they were hiring to replace us, to my knowledge, never were tested. And even if they were, it was additional cost which was not necessarily necessary when they had people who had already been tested and security screened. However, they didn' t object to what they were being told to do. And the people that had objected were not asked to go to that other trailer. This was relevant, sort of, later on, when I was put in quick fix.

Because I was given Unit 1 snubber substitution on nights. There was also a Unit 1 quick fix group. You would think they would >

have someone in their group doing it to keep me away from Unit 1 but I did have security clearance and was allowed to go back into b

Unit 1. During this time is when I wrote those DRs. Because, when I wrote the angle DR, which this is getting away from the point, but I'd like to put it in, when I wrote that DR on a preliminary basis and it went over to Omen and Lepke's office it came back to me because I had not included a support which was governed by that DR. They asked me to supply them one hangar. I was afraid that when they wrote up the DR they would only issue i

it against that one hangar. I went out during my quick fix

' S tokes /12-7 71 i

abilities to go into Unit 1 under my original security screening alone,.and wrote down two pages of supports that were affected by that and attached it to the DR. Once that went over it was ultimately signed off as a generic. No one has asked me about

~

those three papers and the f act that management accepted them as l

. valid design problems, I'm not sure anyone is concerned with that. But it shows .that having management giving them control documents, control numbers, that management finally accepted that I had brought up points that were very critical. The problem is they didn't want to go through with doing anything about it.

YIN: Thank you. I'd like to ask the next question. On page 11 of your affidavit. You stated another technique was to 8 assume joint releases for rigid connections, which means the welds which were in place were assumed to be non-existent. The question: do you imply that the local yielding of restraints of was utilized in the basic elastic analysis?

STOKES : What was used and I have a specific example of.

this. There was a support which had completely welded joints. '

There was wide flange sections. And typically, industry requires that that joint be considered fixed.- It's definitely not pinned in.any direction. However, for this particular joint the plate was cantilevered off of concrete. The bolts were'way back away from the edge. The plate was failing due.to the moments because of the cantilevered section. To make that plate work, they-l simply released the joint assuming that it was pinned. Because l

they assumed-the plate failed. Which it did,.but if it failed

'then-that was a failure. They assumed the failure but they Stokes /12 7 72

assumed it before they did the calc. By assuming it failed and I

that it was yielded, it allowed them to release that moment.

Which the loading bin to that joint would never show that moment l

\

and the plate then would look like it was okay. But you'd have '

to look at the design assumptions to decide that they decided it failed before they run the calc. It was a very tricky, in other  :

words, they're not going to put a statement in their calculation that I'm going to assume this joint pinned here because the plate failed in the previous calc. They just did it. You have to look at that to know whether or not. In other words, if you add the

old calc in the loads, you could run a cale on the plate and know that it failed. They ran that as a preliminary. They looked at it. They didn' t like it. They ran it again, using a pin I

connection. That came out very favorable. They put that in the calc. They didn' t put the assumption.

BEZLER: What you're saying then is they assumed the joint release for a beam af ter first seeing that when it was a moment connection it exhibited failure. You said you had a specific example. Do you have a number or it's in this package over here?

[ STOKES: Yes, it's the primary calculation . . . I think it's l

180-12. I'm not sure that's exactly the number.

l BEZ LER: It's the first one on the list. Thank you.

YIN: On page 12 of the affidavit, you mention new supports were added within six inch (sic) of the first support. And you I stated the f ailing support could handle the small load that was left and assuming the gaps accurate. Question. What is the implication? Does it mean different gaps in these restraints S tokes /12-7 73 l

. . -c . ., - , - . ,

could cause some of the larger gap restraints, not neeing the design loading conditions?

STOKES : That same hangar that has the moment release has this. What I was trying to point out, let's see, um, what happened is there may not have been adeguate gaps at the existino

. support to justify assuming a non-restraint direction. But because the support was f ailing from load coming to it f rom that direction, because it was already failing, if it failed then it displaced enough that it didn' t take the load. If the load f ailed, therefore it couldn' t take it. Therefore, it had to ao somewhere else. By placing a support six inches away, it allowed them to make it stiff enough that if they assumed the other support deflected, even if they run an ME101 line analysis and if 1

it justified the displacement of the support, in otherwords they in placed the stiffness, then that hangar that was not adeauate i before, would only see a very marginal amount of the total load based on its stiffness compared to the support right next to it if it was very stiff. The other thing that I was trying to sav l

l is there may not have been an input similar to that in the l

stretch run. They may not have used the stiffnesses to do what I just said. They may have just on the other hand assumed the gaps of that existing support to be more than enough to justify a non-restraint and designed the new support with correct gaps so that it would be a restraint. Which, what I'm saying is the as built won' t be correct because it doesn' t look exactly the way it's interpreted in the calculation. Just looking at the old hangar l i

drawing,-it looked like a valid hangar with correct gaps. Look l l

l I

Stokes /12-7 74

., ._ - - . ._. -.- - - = - .

s at the new hangar-drawing right next to it. It looked just the sunne. Unless you see the ISO showing them six inches apart, you don' t know they' re six inches apart. You don' t know one's taking all the load and the other one's not. All you know is this looks like a good drawing. But when you take the whole thing into l

consideration, the whole plan system to make the hangar look bad. And you looked at all the cales, all the assumptions, you can see that it.was an intentional series of planned, forethoughted (sic) events, which were done to make that hangar look. good, in the eyes of the NRC, which as f ar as I know, had only been supplied those hangars that already existed, not the new one.

BEZ LER: 'I just want to be clear with the use of caps in the l supports. You' re saying that at times, depending on gap, you would assume a support did not exist in a calculation. I would think if all the supports have proper gaps, they should always be modeled in all analyses. Is that an incorrect statement?

YIN: The specific question just raised already taking, you know, into the review area. We have some problems and issues identified. So we don't to elaborate in that area. This is Esa Yin. Page 12 of your affidavit again, this issue has been asked little' bit earlier, but since I'm new here, I don't know the full story about independent design modification program, and the requirement of sending those allocations and support location to NRC and that kind of a requirements, so let me ask the same guestion again. Perhaps I already asked to you before. The new supports did not have control numbers or document number. - What i

Stokes /12-7 75

is the specific implication. What is the actual requirement.

Did the NRC ask the licensee to provide all the support numbers? Was the independent verification program required that all of the new support numbers be provided for their evaluation? Can you give me some background story on that?

STOKES: Well I can' t guarantee that I'm completely aware of what was done to the NRC or given to the NRC. I can only give what I was told through my relations with management here at the site, as to what they were giving the NRC and why they were doing. That's how I know why they did what they did. They told me they had only supplied the NRC with a sample of ten percent of the supports. This list was strictly a list of supports. They did not list what, they may have listed the ISOs but it did not (end of tape 2)

DUKOFF: This is Dukoff. We went off the last tape and ran

out of tape at approximately 12

21 hours. We' re now going to take a short break.

DUKOFF: This is Dukoff. The time is approximately 12:24.

We'll be going back on record. Mr. Stokes will be answering the l

I last question.

STOKES : Okay, I'll try to cover it. I hope I can cover it. Um, we were in. . . I was informed that the NRC had only been j given a list of the sample as I started to say. They were not I

given a complete list of all the supports. Many of the supports l in the plant didn' t even have adequate numbers. If they had been given a complete list and the ISOs they went with, and then later given the same list with all the current hangars on those ISOs.

S tokes /12-7 76 i

It would be very easy for the NRC to check one list against the other and determine a new support. The reason for adding a support on that was the f act I was told they never gave the NRC -

that hangar number. It had not previously existed. It had no previous record. It wasn' t that it wasn' t a control document and issued a number. It j ust .wasn' t issued . it .until it was built.

They could not have given it to the NRC before they decided to put it in. And I'm sure there was no reauirement that they do supply that but I think ...

BISHOP : Bishop, we' re taking a few second break while the fella brings in a pot of coffee.

YIN: Esa Yin. On your affidavit, page 17-18, you stated that walk through Unit 1 for an hour, you found 100 out of approximately, approximately 100 out of approximately 300 frames, checked to be in noncompliance with AISC bending and buckling design requirements. I guess you will show us where those i locations are. Correct?

STOKES : That's concerning -the DR I wrote up on angle concerning the bending allowable section of AISC. That section was overlooked. The reason it was overlooked was M-9 did not specify as a design guide that we should be checking that. They specified Calavar (?). They also said check to 200/T, the length I

versus 200 over T. They didn' t include that specific section on bending allowable length requirement that should have been-covered. I did see an awful large number of supports.. I was looking forward to being able to walk through Unit 1 containment and point out the f allacies of the system. People who supported S tokes /12-7 77

l l

me and agreed that that section should be complied with after I brought it to their attention. There's one support in particular, the guy took out the existing support and put in tube steel, it's a very monstrous support because the pipe is like eight to ten feet off the floor and it's kind of funny because he l carried me out and showed me through the plant and that same i support is right next to another support which was failing the

~

same way and they only did a very minor change to it but it still doesn' t comply to that particular section. But I do look forward to showing you as many as possible. Whether or not I can show you al-1 of them within this af ternoon, as large a group as we have, I'm not sure.

YIN: In the same question, is your determination based on the assumption that this angle frames will see the maximum AISC capacity. loadings?

STOKES : It's not an assumption that they will assume the

. maximum capacity loading.. AISC gives a criteria for using FB I

allowable of .6FY in the analysis. And that is that the angle be no longer than a certain length and be placed in bending irregardless of the stress being placed on it. What this DR ,

l revolves around is that the calc may f ail interaction equations i

because the correct FB allowable that is used in that section of the calc is not valid.because the angle exceeds the maximum

length that it should be installed at. I'm not saying that if they ran a calc showing what the correct FB was and compared it l 4

to the actual stress, then the hangar is going to fill without  !

l looking at each one individually, I can' t say that. If you look i d

S tokes /12-7 78

at -the angle DR I wrote and the including statements which are xerox copies of the '78 edition of AISC, '79 edition, the copies f rom the commentary state, by the people who wrote AISC manual that it is notLimpossible to calculate those FB allowables for angles beyond those critical lengths. They do state, however, it's beyond the average design office's ability. Bechtel's not

} average design office. They could assign someone to doing an 4

extensive computer analysis at many, many variances of lengths I

for each of these angles, and determine those FBs in a table.

That would be much more efficient use, or calculation procedure than having each person assign a support and the formulas and allow it to do. One of the guys in my trailer had done such calculations on a previous job. He informed me that it was a very extensive process and that just to do that one FB allowable calc would take in the neighborhood of three or four days. For Bechtel to want us to do a cale and a half a day, there was no way they were going to let us do it the other way. But to make the cales correct, in other words in this particular area, for the calc to be correct, that FB is not correct. It would have to i

be determined when it exceeds those lengths per that specific f section. Either that, or per that specific section, the change, the angle be changed to a shape which does not have that limitation. In most jobs that I've worked on and other people in

. the group, the. simplest procedure is when the angles are in l excess of those lengths. Per the code, they just simply replace l

them with tube steel irregardless of whatever the stress is on l

the structure because most places assume or agree with the fact l

Stokes /12 '7 79

that it's too much trouble to try to calc that FB out and prove that the hangar is okay. It's easier just to cut it out and put )

l in something that they know is okay. I do have a copy of NPS l data which a friend of mine gave to me who worked for NPS, a 1 design organization out of LA. It's a listing of sheets l

concerning every available and obvious type of structural l l member . It includes the pertinent structural steel sections and code sections per B31 and whatever. That section could be checked under for allowable stresses per any use on a support.  ;

B31 could apply, AISC could apply under various conditions, ASME  :

could apply, depending on which job you' re on, what the FSAR says. This plan I was told, the ANS doesn' t apply really.

l That's not exactly true but that's what I was told by management. But, in any case, I do have a copy of the NPS sheets which they give to their personnel as to what they should check in doing each one of these members. It shows the same equation out of AISC which shows that they were obvious to the section i

that they expected their personnel to do that calculation, to _

comply with that calcuation, and what I'm saying by that, it's not something out of the ordinary for a design office to do. We did it when I worked for Southern Company-Services five years I

ago. It took us a few weeks to decide that we should do it t

because we hadn' t been doing pipe supports at that time.

Grinnell, I worked with them, they were informed that they should comply with that. The only obvious way to solve angle f rames j that are nine feet long with two inch angle per this section is to brace it every 25.3 inches which in effect makes it a box Stekes/12-7 80

structure, or replace it with tube steel, which does not have that limitation on the compression-zone. There were many debates at that time in our group concerning this section and whether and h'ow we should comply with it. We ended up on the plants we were working on taking out the angles and replacing them with tube

, steel when they didn't ca'aply with these lengths, when it was very excessive. When it was only a minor problem, we simply added braces. I hope that clears up.

YIN: Esa Yin. Would you disagree-or agree that if sufficient engineering calculation is performed, that 25 inch and 50 inch and so on, can be exceeded?

i STOKES: Yes, I think I just said that'. I do agree that if you do sufficient cales, you can exceed 25.3 inches for a 2 inch angle.

??: Mr. Yin, maybe we can turn Mr. Stokes tape over before

, DUKOFF: This is Phil Dukoff and the time is 12:44. We' ll be going off record so Mr. Stokes can change the tape in his

! machine. #

DUKOFF: This is Dukoff. The time is approximately 12:46 and we're going back on the record. Mr. Chin will be asking a question.

YIN: Mr. Yin. (laughter) Esa Yin, again. Just to clear, i

1 at least clear my own mind that-you will not disagree to any I larger link that can be used on unbraced angle- f rames, as long as 1

those angle frames, no matter how many of them and what different sizes, has been previously either qualified or subsequently ii Stokes /12-7 81

_ _ _ _ ~ _ _ - _ _ _ _ _ . . . .___ _ _ _ _ -.. _ _ _ _ _ _ _ ._ .__

l l

l l

calculated, they are suitable for that particular loading. That l is, if you load increase, then obviously your link of the angle beam will be decreasing. If the loa'd is so small, then perhaps

-the longer frame will be able to handle the job.

STOKES : The answer to that is yes, they can exceed those links. The thing that's critical, that I would like to point out is beyond 25.3 inches for a 2 inch angle, the FB allowable decreases, regardless of whether the loading is decreasing, depending on how long the support is the load, well from previous cales, I know it can get down to an FB allowable of around 7.

There is a theoretical low point which it don' t go any lower, but that's not, it's around 4 KSI, I think. But there is a possibility of using the angles longer than these accepted lengths, or the lengths that AISC requires you to check them to,

in bending. But you have to check what the FB is to be able to do an ' interaction equation checking that with the sheer to know if the section is valid. Those calculations were not performed.

YIN: Is that true that M-9 requires you to design the

support to 20 hurts in terms of the fraction is 25 mils. So if l

l you satisfy that, the fraction criteria, will you also satisfy the unbraced angle beam?

STOKES : Yes, M-9 does require the 20 hurts and 25 mils. l You would think that if it meant that, that it would be acceptable. But when you run a strudel program, there is one-i thing strudel is deficient in. Actually, there are several. But l

in this particular case, there is one. Strudel does not account-for the warping effect caused to an angle concerned in the area I

( S tokes /12-7 82

~. __ _ , . _ , _ _ , . . _ - . - - . . , _ . . _ . . _, . . -I

_ _ . _ _ ~~ _ ._

where bending is critical. It does not account for the f act that the angle, instead of remaining 90 degrees as it started out, for the fact that the angle opens up. It becomes a flat plate. It l

does not account for that. Because strudel does not account for j that,'there is no way to adequately justify that support. Even if the strudel says that it met the deflection and the stiffness

~~ criteria. The angle still could f ail if you did the FB allowable calculations and did an interaction with the correct sheer values because the FB decreases, the allowable number decreases. I t' s theorectically possible to do strudel because of the deficiency in this area, on say,120 inch 2 inch angle, and 'the things still meet all the displacements, because you're only checking the i ends, they' re probably braced, you don' t check the middle.

That's where the bending is going to be critical. Strudel doesn' t adjust for the warping effect to the angle. That's why that section is NAISC. They don't want you putting an angle and 4 bending. They come straight out and tell you that in commentary. But, it's possible for the cale to look good. I t' s just not good.

YIN: This is Esa Yin. I don' t have any further cuestions.

DUKOFF: This is Dukoff. The time is 12 :51 p.m. We'll be l

going off the record for the site tour.

l I

BISHOP : This is Tom Bishop at 2:05 p.m. on the seventh.

During the tour, we're at the 65 foot elevation waste gas compressor, 1-1. Mr. Stokes just identified a quarter inch U-bolt that he feels has been spread to be able to fit under the support. The U-bolt -support is part of a support immediately downstream of Valve RV-228.

l S tokes /12-7 83 i

The second reported situation is in the waste gas compressor room 0-1. It's the equivalent U-bolt described in Room 1-1. In addition to that, Mr. Stokes has identified a second photograph that he has taken was of ... Mr. Stokes identified where each of the photographs were taken which were contained in Room 0-1 in

the waste gas compressor. The tour has moved to the Unit 1 steam driven speedwater pump area, elevation 105. Mr. S tokes i identified his, one of his high temperature lines as ASW EP 11 STM SUP trap drain L443. Mr. Stokes &lso identified the angle bracing in this area supporting the line which I described the number of as exceeding the code allowable on length of the angle a

member. Correction on that last elevation reading, it was at-the 100 foot elevation. 85 foot elevation of the pipe penetration -

area just outside containment, Mr. Stokes described this line as similar to the one which he was involved with where the movement

of the auxiliary building compared to that of- the containment

! during a seismic event would create a problem for the piping which penetrates both those buildings due to the location and type of supports. The line was identified as CBCSI-8100 RCP 11-14, seal- No.1 outlet isolation valve line. Mr. Stokes 4

identified the area where the bracket had been removed from and other brackets had been found defective outside the main steam l

isolation area, identified by support No.1029 5CS.

BISHOP: This is Bishop. The inspection tour was concluded j at 3:23.

l l DUKOFF: This will be a continuation of the interview of Charles Stokes. Today's date is December 7,1983. .The time is S tokes /12-7 84

approximately 3:46 p.m. Present for the continuation of the J

interview are Mr. Charles Stokes and his counsel, Mr. Thomas Devine. For the NRC technical staff, Mr. Thomas Bishop and

~

Mr. Esa Yin. Also present is Mr. Paul Bezler, Brookhaven

. National Laboratories, and for the NRC Office of Investigation Eugene J. Power and myself, Philip Dukof f. Mr. Bezler will start

~~ with the questioning.

BEZ LER: Charles, I'm just going to go through the

-document. The first is with the U-bolts you claim on this drawing to have an allowable of 2000 pounds, whereas Grinnell would . state that the allowable for those same U-bolts would only be 485 pounds. Is that a correct statement?

STOKES : Yes, uh...

BEZ LER: How did they then substantiate these 2000 pounds?

STOKES : I have a copy of the U-bolt test program which was done here in I think '78, '79. Another copy, which I may, I obtained my copy from is down in the office of, right outside Mr. Lepke's of fice now, it used to be Mr. Lepke and Mr.- Omen's.

The large loads, they ran a load case test program here, based under the ASME section which allows them to do so, to give a piece a load rating, do their own testing. They ran-five load l cases and side loading, forty-five degree angle and tension. And il l.

they took the average of their failures and they pushed.the U-bolts to .025 deflection before they considered f ailure. That is where earlier I mentioned the .025 is also involved in the hanger itself, was allowed to deflect that much. I agree .025 is not too much. But when you add up those displacements you're l

l

! S tokes /12-7-83 -- 85 1

\

pushing your criteria that you establish for yourself. You' re omitting things, overlooking things. Um, like I said, they came up with these load ratings based at'a, or according to the report, it says normal room conditions. The report itself gives j o

certain deficiencies which they listed. They tried to, however, in a summary point out that these things make the results more

~

conservative. They tended to do that in a lot of cases. They also, well, I think that answers the question as to where I got it.

BEZ LER: So essentially they had test results presumably substantiating the use of 2000 pounds. This interaction formula, you said that they said you should take the square of one component, add it to the square of the second component and use that as a resultant load. Or that quantity had to be below one.

,. STOKES : Below one. That's an interaction equation, side load plus tension.

BEZ LER: You didn' t take the square root of that sum?

STOKES : No, I didn' t.

BEZ LER: And that is stated right on the-drawing reference?

STOKES : That's on 049243, on that particular station.

BEZLER: Again, for U-bolt, no modification of the allowed load was taken for temperature effects. Would that have been done with a Grinnell U-bolt? No, I can answer my own'auestion, you don' t have to bother. Grinnell qualifies it at 650.

STOKES : Well, even at 650, there should be consideration when you use a U-bolt on a high temperature line, above that.

There would definitely have been more consideration except they Stokes /12-7 86

used normal room temperature and it should have been considered on every inspection.

BEZLER: Hartzman asked me to ask in the U-bolt testing you mention the fact that they use heavy wall pipe and then you refer e -

to a f ailure of the pipe itself. I assume you refer to crushing failure of the pipe?

~

STOKES : Local f ailure. They specify in their test program they use the 160 schedule pipe. I notice now in the field just now there are now many cases where it's used on a

. And they try to shim that, typically, to protect the local area? But even with the shim, it's not getting anywhere close to what 160 schedule pipe can take? And they didn' t derate the loads for the pipe. They didn' t do that on 40 schedule. In other words, if the load came across on the computer analysis that it's that much, they assumed the pipe locally could takt that much on the U-bolt. I, when I worked i there on a pre 71ous job, we did a calculation of local bearing f ailure from a U-bolt considering point contact because of the circular shape. And it doesn' t take very much load to fill the pipe locally because of .that point line contact and we ended up removing the U-bolt and putting on a U-shaped strap where the load was spread out over an inch wide, because of that bearing failure. Here it wasn' t considered at all.- And even though we made, I made statements that we had replaced the U-bolts on other jobs, they ...

BEZ LER: The limit of .25 inches, I think you said before that that was used in a way to determine stiffness of'the S tokes /12-7 87 i

support, that one could determine the stiffness by calculating deflection?

STOKES : Yes, you take that .25 deflection, you can use that same replacement as a relative magnitude that's given one's sort of topical to the other.

BEZ LER: Is the 025 provides a support with a stiffness that has higher than 20 hurts.

STOKES : Actually the number comes out for 20 hurts, I think, .02497. It's under .25. 25 is just a little over.

EEZ LER: But then the number comes about because it sort of defines the stiffness of the support. It doesn't really relate to a strength characteristic of the U-bolt. On page 8, I guess it's page 8, yo J start to talk about the Code break calculations and you talk about not having enough offset. Could you describe what you mean there to me? I was unclear.

STOKES : Um, maybe I should draw . The Code break locations were all containment , in many

cases, both sides of the penetration. On the inside, at times, about located out here on this. Typically, that distance was far enough from the wall, in the neighborhood of 6, 7, 8 - f eet, that l

this support on the line out in here, would come off the j

auxiliary steel, and the inside of the angle of steel. That j steel connected to the inner concrete wall on the inside. The displacements to that steel were based on that concrete movement because it was attached to that concrete. This wall, on the other hand, was allowed to move out under loca. It also had its own seismic movement. To do that particular calc for that I

i l

S tokes /12-7 88  ;

1 I

i l

.= - . - .- . - - .

section of pipe, that pipe had to be supported in a condition which would allow this wall to move out its max and all this field to move in its max without overstressing the pipe. It also had to be able to move in the opposite direction. What happened 4 o is these valves were mass loads on the pipe. They reouired that you put in a rigid support to support them, in most cases. Now,

" the pipe usually came straight out and turned. If they had them putting these valves right next to the penetration, then they could have built the support out past the valve off the penetration and protected the valve. And they could have let that 1.ine break if they had wanted and still protected the ,

valve. However, the line still had to be able to move because it was inside on the seismic side. So they couldn' t let the inside fail. The inside had to be at the last splay. In many cases, i

when these came through the wall, there was not adequate offset i

sideways for that movement to be taken out without overstressing l the line. On the opposite side of the line, a similar thing i happened between the auxiliary building which has a big boot around it, and the line coming through on the opposite. That was what I was looking for and I couldn' t find it, the section of pipe that I was looking for, that I worked on. The - first M-40 calc I did on one of those outside systems that could not work period. But it came through the wall and the valve was located about eight feet off the all. The acceleration vertically for the support to protect that valve because it was non-seismic off that end. What made it impossible to design a hangar off the wall. The other thing is we had been instructed, we couldn' t attach to that wall.

f

~l S tokes /12-7 89

~. . - . . -

BEZ LER: But then what'is your contention as to what they

'did or didn't do regarding these lines? I accept that they had a

problem with these lines.

STOKES : Okay, well, if these lines had been installed in

. . this particular case, the problem was, unless engineering had directed the layout, they probably wouldn' t have gotten the

~

inside correct under any conditions because there was this combination seismic and loca and the seismic had changed. On the outside, however, there was another problem that happened. I f, ,

I well, the outside it was still a problem in this particular i case. , There were other cases, however, where we had a large line and a small line coming off of it. These small lines had code break valves. In these particular cases, this valve was too close to the large line. Those particular lines made it additional offset so that- the valve would be placed

, far enough downline that you could support individually and the snubbers on the first sect!on, the spring cans to allow this kind i

of line to move. Because these had a lot of high movement. They f did not want to have these additional line sections on these

! lines. In this case, if they had an instructed construction, the people who installed the lines, in the region of the, and this l

was admitted by Mr. Lepke, if they had told him that actual line-operating temperatures of these lines, and given them the adequate directive, these. valves could have been off set enough that they could have been supported without rerouting. Mr. Lepke l

told me himself that these lines probably would have been worked had they been told correct information or the vendors had been S tokes /12-7 90

_________,,____.,,______m -____7m.. - - e.,m.. -  %%__-, y,, 7-e., _

instructed correctly. These, on the other hand, comina off the table walls, I didn' t feel that they would have ever been built correctly unless engineering had directed it, and didn' t begin with the new changes from the , I didn't expect that

, they would probably still work. But they would have come closer to work than they did.

BEZ LER

Yeah, but I'm still confused now as to what your contention is. I agree that it was poor design what they did, but I presume eventually they corrected it in some f ashion. Or have they corrected it in your opinion.

STOKES : Well, I know they didn' t want to admit there was a problem. They ignored our M-40 cales. They ended up runnina them on ME101. I am not sure that they rerouted the lines.

BEZ LER: You presume they've since shown that they aualified lines by one fashion or another?

STOKES

One way or another.

??: I think what he is saying that they describe the condition, and Paul is saying yes, that condition is bad. To

say, you knew it would not have worked, if engineering had not i redirected it?

BEZ LER: The inference is that they have done something to correct these problems?

STOKES : Well, I recommended they add additional offset to keep from overstressing. They took it out of our hands. I t's l

like everything they did. We told them their hangars were bad, they put them in somebody else's hands. In this case, they .

l l didn' t like our M-40 cales, they put it in ME101. They may have Stokes /12 '7 91

--- ---y- ev- -,

l been able to juggle out some of the ME101 if they assumed gaps on supports that didn't exist. They did not want to reroute those

~ lines. They'd been hydro'd. They didn' t want to have to rehydro them. To me, all the schedules had been met, I'm not sure thev

. rerouted very many of them. I'm not sure they rerouted any.

BEZLER: But, presumably, either through calculation or rerouting, they' ve qualified those lines. And this particular

, code break is a break from Class 1.5 or Class 2 or seismic to non-seismic?

STOKBS : Right. Seismic to non-seismic.

BEZ LER:

~

So in the end you don't know how these conditions i

were resolved.

STOKES : I don' t know if they provided the offset or just modeled out the problem.

BEZ LER: On page 12, you talk about taking a reverse order of things, that is, the support load carrying capacity was specified to the piping designers and then presumably somehow you make the statement, they changed their minds about the models and resulting loads. I could infer from that they redid calculations assuming those stiffnesses possibly rerouted the lines, and what is your impression they did? I'm not quite sure I understand I

what you meant by they changed their minds about the models and resulting loads.

STOKES : Can I ask you which paragraph?

BEZ LER: The second paragraph, the one that almost starts at

the top. In the middle of the paragraph, you have that statement. In other words, the supports could only withstand a Stokes /12-7 92

certain load, they didn' t want to change the support. So they told the pipe designers ...

l . STOKES : Yes, that's exactly correct. Initial.t y, we were given loads to experience. Typically, that's the way all the

, hangars are done, stress tests, the ISO cales, by loads based on the configuration of the system, irregardless of what supports can take. The supports are designed to take a load. In this case, here, in the review program, that's the way they started

. out, the way you normally expect. However, when hangars f ail, 1 they resulted in a reverse turnaround. They requested that we do j a calculation to determine max load in each direction. It's okay

) if it's a one-directional support. It's f airly easy to do a one-i

! directional load cap for max. But in the case of an anchor, it

, was almost impossible. Because you have variance magnitudes in any one particular section of the loading that it would affect it. But what they did was they asked us to determine the max loading. Then they carried that back to . stress and gave that max

loading condition to stress and asked stress to do their best tx) lower the load to that allowable. They knew the hangar would j take it because they had already performed the calc that way. If stress came back and said they couldn' t, which they did do to start with, then they let them go as failures. Because they were i

working up to the 95% of the 10% sample. They had a f ew they could live with. The thing was there was so many that they couldn' t - get to work. And stress wouldn' t lower the, some of the stress guys wouldn' t jeopardize what they were doing. They wouldn' t agree to lower it beyond the reasonable amount. Then, 1

l S tokes /12-7 93 i

l 1

later on, .they went back and those same guys that objected were moved to another group. The stress group people that were doing the.cales when I first came here, a lot of those guys were transferred to other groups. There has been such a turnover in

,- the stress group that the lead over there has changed roughly I - four times in the period I was here. The people doing the cales, l

1 their f aces have changed equally. And, I personally, doina the hangar cales, have discovered defects in their assumptions and l have gone back to stress .and reauested they change the loads that they supplied to me.

BEZ LER:

Okay, are you implying then that the stress people would then use the same models but some different assumptions to 4

try to qualify rather than rerouting the pipe or redesigning the

! system?

i i STOKES : There are some gaps. I wasn' t in the stress 1

trailer. I do know things they did. They, in one case, when I was in quick fix, the guy who had done the calculation, had i assumed because the pipe was resting on a piece of unistret electrical support that it was a support. Because, - he either had  ;

! to remove it or show it. The computer analysis, ME101,- if you 1

put in one direction seismic, it assumed the opposite. Well, it l ended up coming back showing two directional seismic loading.

I j Rather than put down on the covcr that that should have been moved, removed, and the analysis done without it in there which would have shifted load slightly, he instructed that the support -

was supposed to be modified. It wasn' t a support. It ended up

- almost being a support. The person who removed the piece of l Stokes /12 '7 94 4

1

unistret in the fill called me, and because of the proximity to all the supports, it was questionable to him. He asked me if it looked valid to me. I told him I'd check on it and be back within 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. When I chased it all the way back through the

, stress group.to the guy who did the stress calc, he told us that yeah, it was just an unintentional restraint, it shouldn' t have been there. He agreed to help me even change the cover sheet.

He gave me a new copy, completely taking off that hangar as a hangar, told me I could delete the support. But he didn' t redo the calc at that moment. Technically, all the support loads would have varied some amount because of the change in the i

model. He didn't change it at that time. And as f ar as I know, I'm not sure they've redone all the cales on the other hangars with the updated loads. But, they did either intentionally, unintentionally, or just out of, well, maybe, because they were i

instructed to, make erroneous assumptions in their modeling and

that in itself, affected the loads to our supports.

BEZ LER: Okay, could you describe what you mean by erroneous assumptions?

DUKOFF: This is Dukoff. We' re going to go off record at approximately 4:11 to flip the tape.

DUKOFF: It's approximately 4 :12 - p.m. Back on the record with a new tape. Mr. Stokes will be responding to the last question.

STOKES : Um, well, I had several friends working in the stress trailer. I still do. I was aware that in many cases to i

lower loads they would do similar things to what we had done or Stokes /12-7 95 i

L i

some of the guys in the new trailer was (sic) doing to our supports and assuming gaps which didn't exist. Assuming no gap I

where a gap did exist.

BEZ LER: How does assuming a gap help pipe stress 4

o calculations? .

STOKES : Well, they had the ability in doing the modeling to

-' allow the pot to displace before it started loading a support.

We did that same thing in an M-40 calc. If we could thermally, if we determined the displacements from thermal, then we could specify a gap to allow the pipe to move that direction before the

! pipe support started being loaded from thermal. It still caught all the seismic loading technically, or theoretically, but the thing was if you displaced a pipe a half inch and you put a gap there of three-eighths of an inch, then you could calculate a reduced load from thermal, based on that smaller displacement.

By doing that, they could reduce the load to the support. If they assumed - a gap which didn' t exist, then they could lower that J

load. They were supposed to go out and check the gaps as they

exis'ted and model only in critical cases, what actually existed.

1 I do know that to make some of the hangars work.they had to do more than what was there. That was supported by an earlier disclosure by someone who did work in the stress trailer by another report which had been turned in. Matter of fact, that paper 'came out roughly the next week af ter I was terminated and from my f riends here at the site, which I still maintain contact

with, Mr. Lepke and all of management knew, they personally felt I'was the one who had submitted that paper. However, it was anonymous. I, however, had not had the direct contact with. . .

Stokes /12 '7 96 i

i

~ ~ --' ^""-

BEZ LER: What paper are you referring to, the 55-page document? I STOKES : Right. I had not, however, had been assigned to the stress trailer enouah to be able to obtain those documents. 1

, I had. trouble getting the hangar documents I got for my own group. Because I did stir up problems and every time I went to look at a hangar calc, they didn' t want me to look at it. But that person, I don't know who he was, supported what I had been told. I had talked to one of the ex-stress trailer people, who had been reassigned. He was aware of that hangar being added six inches away f rom one f ailing.

I' can' t give you his name without asking him if he'd be willing to say something about it. .But we did discuss it and he admitted that he was aware of it. You know, it's very hard to prove that management's intent is to cover up stuff because they move people around every five days, i

Bechtel sent all their people down here for one year maximum tour of stay. Some of the people were sent down here for three weeks, and then went back in three' weeks. But there was an awful lot,of

, movement of people f rom one group to another .when they had been 1

doing the job and doing it good. It didn' t make sense -for them i

to reassign people to a new task other than the fact they were doing it because they didn't want people to- have an extended stay and learn too much. H opef ully, that...

BISHOP: Bishop. Mr. Stokes, this~ was' the second time, I believe, you' ve identified a person .who knows of 'a problem and

. you're not comfortable .in giving us his - name. Is this a different individual or is this the same individual as the first?

Stokes /12-7 97 l

r STOBES : I' d like_ to'say in all the documents that I' ve already submitted, the person's name exists. And he went from stress group to a supervisor position. You do have his name.

But I won't tell you what person that actually is without having

, asked.him if I can. But it wouldn' t be too hard if you questioned all the people in it that I've given you names for you e to probably locate him.

BEZ LER: In any case, then, you've stated that one method of changing a model.was to take credit for gaps, whether real gaps or assumed gaps. And this was done both in the pipe stress area as wel1 as in the pipe support design area. Where there any other mechanisms to change one's mind about the pipe stress model? Was there any other technique to adjust the model? Well, an I had already said, if they couldn' t lower the loads, through taking account of gaps or configuration changes within what was there, to the point of lowering the loads for support, they ultimately added a support. That was the last thing they would do. Everything else was very subtle. In other words, in the i calc package, your assumptions are, they don't ever really spell

, them out. It's between you and your checker. To. find those assumptions af ter the fact is a very deep involved type thing.

.Is 's very easy for them to cover those kind of things. Both in hangara and in stress. The adding of the hangar, on the other-hand,--is something that's not that easy to cover up. It was the l

absolute end result used to make an existing support pass. When, t .

well, in particular, one of the supports that I hope I' have a '

copy of the calc, or I had it in my list, one of the f riends of t

S tokes /12-7 98

l l

mine in that trailer of Leo's, he instructed me in the problems ,

they had making the hangarnwork.

They had done all the things I said so far except adding the support. They had ultimately, the pipe stress lead from San Francisco came down and they were

, instructed to do the' analysis with one pound load and each of the restraint directions individually. Not all at once. It was a b i token loading. They ended up adding the support following that, to take all the rest of the load. Excuse me.

DEVINE : Cnarles, it's 4:20. Is this too late to try and call your friend? When will he be off his shif t?

STOKES : Um, he ' s not on . . .

??: Shall we ago off record Mr. Stokes?

STOKES : Well, we already know his name, we mentioned it in j the field. I think he goes off at 4 :30. It might wise to try to catch him.

t YIN: Well, most people in the stress and the hangar group working long overtime?

STOKES : Well, he's not in the stress or the hangar group.

Or he wasn' t. He was in a feasibility group and his primary function was to walk out and review things and suggest designs.

The percon I know, he may be scheduled for a lot of ' overtime, he used to keep a very liberal schedule. In other words, he might leave at 3 o' clock today and work' till midnight tomorrow. He had more freedom in that group because the leader of that group allowed him that freedom. We didn't have that freedom ourselves. I'd rather try to get him now or someone in the

- group.

o Stokes /12-7 99

_ _ . . ~ . _ . , . _ . . , - __ _ _ - _ . _ - . . .

T s-

r 4

DUKOFF ' This is Dukof f. The time is approximately 4:22.

We'll be going off the record. Mr. Stokes will be making a phone I

call. l DUKOFF: The time is approximately 4:28 p.m. We'll be going

,- back on the record.

BISHOP : ' Bishop speaking. I forgot my question. I'll turn it back ~ over to Mr. Bezler.

BEZ LER: I guess I have really just one last guestion. Back with the U-bolts and this matter of calculating deflections. - I l take it the normal procedure to calculate a deflection would be to consider the entire support except for the plates, or were we referring just to the U-bolt type supports? In other words, you took the deflection of the U-bolt but not of the plate, that's what you inferred, is that correct? You were told that's the way to do the calculation.

I STOKES : How to do the calc is generally lef t up to the engineer. Especially in my case. If I do a cantilevered-support, where the base plate has an effect on the displacement so the upper support point, then I include the plate if.they allow me to do a finite element ~ analysis to determine .that l

4- displacement. And I have done that. In a rigid frame, or a brace f rame, it's not as critical to worry about the plates because it's strictly an upward most of the time.

It's an engineering judgment decision whether or not to include the plate and myself I usually include it on a cantilever ;of

(

Short cantilevers I don' t f eel it's critical.

extreme length.

It also has the relative loading. It's always, in every job I've i

, , Stokes /12-7 100 i

worked, a requirement that you check it from the plate to the support point. As f ar as I know, all the engineers perform displacement cales in that method. I was not the only one though questioning the fact that in many cases the plate was critical and had' been omitted.

BEZ LER: I'm a little confused, though. I had the impression you said it was at the discretion of the desian

engineer to consider plate flexibility or not. But then I get 4

the inference that you were told not to consider it, now, which f is it?

STbKES : Well, we were definitely instructed on this job not to. We, even if I wanted to, they didn't want me to do it. In other words, if I had of done it, if they didn' t like the calc, they probably would have changed it in the other trailer-subsequently. From looking at the calc logs that I have copies of, I should have been in the ~ calc log as f ar as performing cales probably in the neighborhood a minimum of one every thirty-five and more than that. However, when I look through the log ,I can

only count five cales that I've been credited with out of a couple of hundred. Undoubtedly, the ones I, I had felled an i

awful lot of supports and the only adjustment thing I can see is they didn' t like me f ailing them and they redid them. And when I i

check out a support, I checked it out based on enaineering judgment thoroughly. And if the plate had more contribution to

displacement than the cantilever, then I include the plate. When I was working.at a previous job, we were told not to include base plate flex, as far as the displacement. -And I got an eloht foot S tokes /12-7 101

cantilever tube steel. I ran an ANSIS calc on it, took the displacements f rom one side of the plate based on the ANSIS calc, and the displacement of rotation of the plate over the eight foot was more critical than the tube steel attached to the support. i i I It passed before you include the plate, and it f ailed by two

.. inches after including the plate. They told me not to do it.

I They told me here not to do it. We were, initially, we were told the plates the rigid, even in considering the bolt cales, we were' 4

instructed not to include any plate flex. Regardless of the configuration, we couldn' t even make the decision on our own 5

whether it was critical. We had plates there in that plant that

! were three-eighths inch thick, eight feet long and one foot wide, with multiple supports on it. They didn' t even want us to, and the bolts were really random configuration. But they didn' t want us to even look at that plate and any other thing than rigid.

That plate wasn' t rigid over no eight feet. Not three-eighths plate. It wouldn' t have been rigid if it was a twelve by twelve plate. Depending on the loads. But the thina was they cut out

that part of our abilities, constrained us. They didn' t want us to check on it very thoroughly is what I am tryina to say. Thev-wanted us to overlook things. Stretch our engineering judgment  !

l as far as. If we were willing to sign off every calc we did, i

i just strictly with engineering judgment decisions, they was j l

j perfectly in agreement with that. They didn' t care if we did a l l

calc.

BEZ LER: Okay, I think I'm just about done. Let me just summarize then your calculations if you are basing it on a l Stokes /12-7 102

I f

deflection criteria, you would calculate the deflection of the beam members and of the U-bolt in determining whether the 1

structure was below the .025 limit.  !

STOKES : They didn' t consider U-bolt the spring.

BEZLER: Okay, but then just the beam...

STOKES : The U-bolt in the calculation was assumed rigid.

However, the load allowables allowed a maximum .025 displacement. There was no displacement at all considered. The hangar was allowed to go to .025 which was failure according to their criteria. If you had added the marginal amount of displacement at a minor load, it still would fail. If you added -

base plate flex in a lot of cases on top of that and failed even worse, it's a series of things they looked at one at a time but then they didn' t combine them.

BEZ LER: And you said before the .025 that corresponds also to the, essentially 2,000 pound load limit on U-bolts, is that t

correct or roughly correct?

STOKES : Well, it doesn' t correspond to the loading. That was the displacement they used to consider the U-bolt failed. It was like the hangar restraint condition according to their criteria was that it's supposed to be equivalent stiffness of 20 hurts. They carried the U-bolt to displacement comparable to

that. And then they ignored that when they did their supporting steel. They did that to the 20 hurts. And then they ignored the plate. If you added it all together, it was a very flexible i support.-

l l

Stokes /12-7 103

,..y -7 y ..-4y y--- y - ___-y ~ ~-w. g9.--, g,-%-  %. - --.T--wr-'n-'-mpa' --

W' ' -

• BEZ LER: Okay, what you're saying then is the .025 deflection limit would be placed on just beam members in the support structure, ignoring the flexibility of the U-bolt or plate?

STOKES : Right.

1

.- BISHOP : Bishop. You mentioned that there were the base plate flexibility problem occurred with you once before at a

{

different reactor site. I gathered it was a reactor site, where I

there too you were told not to consider base plate flexibility.

Would you identify that site, please?

l STOKES : That job was with Bechtel Gatesburg in 1980 and the plant site was Davis Bessie.

BISHOP : Thank you. Bishop again. On the modifications of your transcript, your affidavit, page 12, you, and in-your testimony a few minutes ago, you mentioned that the ultimate solution to these problems was to add a support to make'the pipe l l

l l pass the design criteria. Is is, are you aware of any supports  !

l l that were added to small board pipings in the time frame that you '

)

l were here for reasons other than to pass the seismic qualification concerns?

l STOKES : We did, oh boy. On Unit 1, other than-the support i

used to justify an existing support, I'm not aware, except that within the percentage of allowed f ailures of any corrections or

! additional hangars being added.

i BISHOP : I guess one could... The reason I asked that question.was for example, if we went,out and' looked at all new supports that were added to small bore piping in the eleven month Stokes /12-7-83 ~ 104

period you were here, from your perspective anyway, you would i believe all those new supports were added to justify other, the adequacy of other small bore systems as opposed to some other conditions, that new supports wouldn't have been added for any other reason.

STOKES : The only other condition we added supports that I'm aware of.that were added, were code break locations. They did allow us to justify code break locations. Seismicly, they didn' t

! want to do a complete seismic reanalysis. We did check out the code break. As far as I know, all the code break lines, there L are not that many, and they initially didn't want us to add any l supports. They didn' t want to reroute any lines but they didn' t mind adding supports. Their recommendation for management was to i add a single anchor following the valve if possible. ' And if that

- wasn' t feasible in many cases, there were up to two laterals reauired and one axial to protect that valve. They didn' t object

! to us modifying those particular supports. They did not want Haver (?) to modify the line itself. Other than code break locations, and on a limited basis those, I'm not aware of a

!i i

l hangar being modified at all in seismic, samtam, thermal, or any other system.

BISHOP : Thank you. I think we' re ready now to proceed with looking at the actual data package. And we probably want to go off record while we set up for that.

i DUKOFF: Time is approximately 4 :40 p.m. We'll be goina of f l record to set up to look at the calculations.

l

[ S tokes /'12-7-83 ' - 105

. - . . . . . , - , ~ . - - . - - . . - - - , . . - . .- - .--

l I

Today's date is December 7th, 1983. The time is  !

approximately 5:10 p.m. This will be a continuation of the interview of Mr. Charles Stokes. Present with Mr. Stokes is his

' counsel, Mr. Thomas Devine. Present for the NRC Staff will be Mr. Thomas Bishop, Mr. Esa Yin, and Mr. Paul Bezler, NRC

. Contractor f rom the Brookhaven National Laboratories. Also present f rom the Office of -Investigation will be Mr. Eugene J.

Power and Philip Dukoff. And Mr. Bishop will start the_ session.

BISHOP : Bishop. I guess we'll turn it over to Mr. Stokes to start the discussion. We have the documents we' re looking at, a document called "The On-Site Engineering Small Bore Pipe Support Calculation Log Unit 1". It's currently turned to page 60. We' re looking at the original document and we will take it from there.

l STOKES : Okay, this is Charles Stokes. What I have here, and what I am going to try to show is that the calculations that were originally.done were destroyed. The calc log that is here, there's two versions, the original, there's also a listing of Samtam, Thermal, and Seismic sampling. Those were done by someone separately from the original logs which were filled out by the individuals who did the cales. We' re going to be lookina at a calculation number for pipe support 100-132. It's a part of l the Samtam sample. I have a copy which I'm going to provide to the NRC as a reference. It was dated 2/7/83, no 2/1, yeah 2/7/83. The calculation was done on this particular hangar by G. D. Ketcher. It shows that it was Rev 1, 0 revision referring to the GC calc. log, or calc file, as all the other cales did. It S tokes /12-7 106

was signed off as performed and checked. This particular calculation, if you look at the results or the summary of the results, on page 1 of 34, it says, under conclusions " field modification reouired for the following reason, base plate failure overstressed and anchor bolt failure, see pp. 20-21, and )

ait gives the plate stress as 91.8 ksi versus allowable of 25.4 ksi. A bolt interaction equal to 2.3 greater than 1." I would like to note that in the first or master section of the On-Site Engineering Small Bore Log Calc No. MP988 for Hangar 100-132 shows that Rev 0 has been performed by a G.R. Shaw dated 3/1/83, '

1 checked by K. V. Cotcham, 3/5/83, approval 3/6/83. Those particular names aren' t ref ersnced in the calculation I have as being the Rev 1 or Rev 0. If we can pull that calculation which has been supplied to us, and we turn to the cover sheet which, okay, looking at the calculation that exists as a final for MP988 hangar 100-132, and looking at the record of revisions and the

work performed on this calculation, this calculation shows that

! the calc was done Rev 1 by G.R. Shaw and K.V. whatever per the

! log. It does not reference the revision which I have showing Mr. Ketcher as -having cerformed it on 2/1. I don' t know what happened to that particular calculation. It should not have been destroyed. If anything, any future revisions should have been only added to. As does the cover sheet exposes, as it's already up to Rev. 4 for final support evaluation. If we were to thumb through this calculation package, and try to find out the history i

of the calc that is shown as final, it shows here, there are several versions, first one revision which is shown, the 4 Stokes /12-7 107

conclusion is no field modifications. That was dated, that's 1

Rev 4. -There was new loads. Rev 3 shows no modifications.

Rev 1, which was originally performed by Mr. Shaw, no field modifications. There should be another revision here, Rev 1, showing Mr. Ketcher recommending modifications. Then maybe Rev 2 would show no modifications, but somewhere alona the line, somebody decided to destroy that cale showing that it could have f ailed. To my knowledge, this particular support not only, well, i let me point out something else. In the Samtam log, which is at the end of this master file, for hancar 988, if you' ll look, it

~

does give Mr. Ketcher credit for having done this hancar on 2/7/83, checked by Mr. Patel 2/9/83, but somehow he act mis-taken of f of the cover sheet of the calc and his calc got destroyed. It shouldn' t have been done that way. I've never seen this kind of thing done. It may look like it's strictly an accident but I will attempt to show throuch some other cales, which I don' t know if they' ve acquired yet, that every calculation that I have a copy of the original where it f ailed, that calculation no longer exists. In some cases, the calc log has been changed both in the original and in the final. S amtams sampling sheets, and in some cases it hasn' t. At this time, I' d like to ask anyone here if this, if they have any auestions concerning what I have to show them at this time.

l

> YIN: Esa Yin. I would like Mr. Stokes to show us perhaps a couple more examples, if he has the documents available. 1 i

STOKES : Well, before we do that, I think I'll take a break )

to get a little organization because I'd like to get all the f documents in_ front of me before we proceed.

l S tokes /12-7 108

. ..~ - ,--- . - . - -- - . ._

l DUKOFF ' At approximately 5:20 we'll be going of f the

record.

DUKOFF: This is Dukof f. The time is 6 :4 6. p.m. We' re going back on the~ record now. During the time we've been off record, Mr. Stokes has had some informal discussions with the technical staff regarding some calculations and two points have been

> brought up that we' re going to put on the record. Mr. S tokes could you please repeat for me the comments you made in reference' l

to Mr. Gray and also your feelings on the seismic versus non-seismic lines for the record, please?

3 STOKES : ' Well, I'll be glad to expound on Mr. Gray's, uh, what I said about him. That's okay. I'm out of tape. The j

comments concerned a drawing which we just happened to be lookina at, 2171-16 which I couldn't believe that I did a Rev 0 calc when I rev 0, it says that's done per ASM and Dave Carr. But it does give me credit for Rev 1. At which case, having thumbed through the cale and looking at my sketch, I noted that my sketch, this particular hangar was one which I had recommended a fix for of the installation of an additional U-bolt and a short section of

. angle because the U-bolt fell side load. I had added a second

{

U-bolt and an angle and used both U-bolts strictly for the tension loading to which case both were adeauate. This brought f back memories of things that had slipped my mind. Mainly, when

. we did these calculations, we supplied Pullman a sketch of our proposed fix. These went through a Mr. Dick Gray, who was 4

supposedly only supposed to originate a cover and route these to Pullman and handle the documents. However, what happened in this i

S toke s/12 83 - - 109 l

I particular hangar's situation is shortly, when the sketches came back to me for review and signatures as far as issuing to the

~

l field, I discovered that the sketch had no semblance to the l l

sketch that I had included in my calculation. I The duolicate l

a U-bolts had been replaced with a boxed-in configuration. I traced this down through Pullman to find out what they had been,

! the sketches they had been submitted, what they looked like,

because I couldn't believe that my drawings had been chanced, I thought Pullman was responsible. They showed me sketches which had come through Mr. Gray to them with cover sheet from Mr. Grav, showing it boxed in. I, at that time, went back into my trailer and discussed this with Mr. Leo Mangova and Mr. Wah. I told them that one or the other was incorrect that the sketch needed to be changed to my configuration or I needed to change my calc. They said, well, it's obvious it's conservative, there's no need to

change your calc and at the same time I complained because t

l Mr. Dick Gray obviously the only person between us and the

Pullman people, had changed my sketch. He had done this without i doing any calculations. Strictly on his own accord, without an independent checker or any other person looking at it. This, to my knowledge, was extremely wrong. I told him I didn' t like

{

i it. I went in with Kevin and told Mr. Bob Omen what was 4 happening and recommended that he stop Mr. Gray from being allowed to review our drawings if he was going to change them without having authority to do so. He was removed f rom that i

I position. However, he has never been reprimanded to my i

o knowledge, and he's still at the site in a new position as lead I

Stokes /12-7 110 4

, .--- - - . . , , . . , - . - . - , - . . - . . - , , ,, - - , -----.-,.v. - .,.n. , . . , . , ..-, ,n-.

of the pipe stress group. I don' t feel he has the, or should be in that position.

Also, concerning the second part of the question, I don' t usually make comments on a personal nature. I tend to stick only to design criteria, design considerations, per codes. In this

, case, it's something I think should have been considered in a code but somehow it's been overlooked. That's the f act that in actual installation in the field of pipe, many non-seismic pipe

, systems are installed right over seismic lines. When the seismic lines are evaluated, they do not consider the other oices as falling on them. In many cases in the plant, there are extremely large non-seismic sit lines which are over Class 1 safety-related systems that have not been evaluated. The supports probably are inadequate if checked against the critical load condition. And I do know f rom past experience and personal knowledge that I have seen very large twelve inch lines over the control drive ride systems on other plants. And these lines were never analyzed to the same standard that the seismic were and could be very much a safety related problem. I will say that.

BISHOP : Bishop. Going back to point one, the document that Mr. Gray changed, in the record package that we' re lookina at, could you state what revision of the record package that you reviewed and could you tell us if there are subsequent revisions to that whether the package as of today's date, in your estimation, is proper, or whether it still reflects calculations which don't reflect the actual construction?

l i

l S tokes /12-7 111

l l

l l

STOKES : - Well, there was on this particular calc an existing l

calc rev 0 which had been performed earlier. However, I was i .

asked to reevaluate the hangar. And I reviewed what was Rev 1. I There are subsequent revisions to my revision. Rev 2 includes a l i

sketch of the as built conditions but no cales on the chance in configuration from my cale to it. It does have a kind of a, well, it has a sheet, page 2, and there's a few comments here

) .under deviation from engineering design, under minor, minimum I difference in dimension, 1 1/4 inch max from design, and it says plate design used rather than continuous angle in design modification. The first one was rated insignificant. The second one says design mod as as built, plate desian is preliminary for

U-bolt, primary for U-bolt f ailure, as built is still okay. I t's -

t basically only engineering-judgment decision on that. There's no cales. And then there's a subsequent revision, which is strictly a final load condition, looks like. Compares loads used in calc l with final loads. And, looking at that, well, it says that the f

! final loads are lower, boy, this is really neat, it lists the l loads here, it shows the D, E loads to be in the Z direction.

i 0 loads, both loads were 100 plus or minus, Y and Z. The final l

loads on the new one show 63 in Y and 125 in Z. Just a little bit less in one case but higher in one case. It shows D, E loads and 2 to be higher. It shows the Y loads to be higher, 32 versus

! 2.3. It shows the dead weight to be a little less than what I f

originally used of 40. It shows the high loads to l

l be, that I used to be 214.5 and the new hiah loads l

to be lower. It states the final loads are less than loads used Stokes /12-7 112 i

, , - - _ . ,_. _ ..--,_......,..,_ ...-_,. ,~ . .-. __-..,,__.,,_ - . _ _

...,_m . - - - - , - - _ . . , , _ _ _ , . . , _ _ , , , , _ - .

. . - _ ~ . . _ . - . . - . . - . _ . _ - . _ ~ _ . __ .

{

1-in analysis and no reanalysis required. However, unless the high

' loads were governing, it looks to me like the final loads are as high or higher in one case and lower in one case,

~ depending on which direction is the most critical, may not be less'than the original in any case. So I'm not sure that there's rx> reanalysis required on that but it would definitely take cuite a bit of looking into other than just looking at comparing those

. two load cases. I hadn' t noticed that. previ%usly. [more -

talking, but not intelligible)

BISHOP : Okay this is Bishop. I think we can take that i

package with it. Let me list just for purposes of the record the documents we have collected from you this evening and made copies of. I'll just do this very quickly. These are pipe support i

calculations. The first one is for hangar No. 100-132 revision The next one'is 2182-81 one. The next one is 98-82, revision 1.

revision 1, 2182-74, revision 1, 2182-81 revision 1, 2182-94, i revision 1, 4614, revision 1, 2182-66, revision 1, 2182-64, revision 1, and there may be one repeat in this entire process but I'll list them all anyway. 2182-93, revision 1, 2180-26, revision 1. And here's the repeat I believe, 2182-63, i

j revision 1. And yes, there are two of those. We also received today f rom a separate individual, Mr. Gary Ketcher, a document 8

that carries the number T60551-1A. It is hangar number 100-32, it's a computer program printout. And it is, I'm looking for a date on it. The rev number is blank, the date is blank. Input l by Ketcher, dated 1/31/83. [I thought there was a second part to l

this. Oh, here it is.] The other document from Mr. Ketcher is l S tokes /12-7 113 l'

t

also a computer program and it carries the number T70916-4 A. It also is for hangar no.. 100-132 followed by a Y. It has a input date by G.R. -Shaw, dated 2/26/83. We have a third document i received f rom Mr. Ketcher this af ternoon. It's a computer

. program also. It carries the number T70916.

DUKOFF: The last tape ran out at approximately 7 :02 p.m.

4 I t is now 7 :15 p.m. Mr. Bishop is going to finish reading the i document numbers into the file. During the time that we were off record, Mr. S tokes and Mr. Yin had a' discussion regarding the strudel programs.

BISHOP : Bishop. The last document that was provided to us i

by Mr. Gary Ketcher today is a computer program. It carries the number T70916-1A. It's in reference to hangar no. 100-132Y. It i was inputted by G.R. Shaw on 2/26/83. I think what I'd like to do at this point, also while we were of f the record, we've compared the document, the calculation package for hangar no. 100-132 rev. 1 that Mr. Stokes provided us with another document that's in the official PGGE record that carries the same

! revision and hangar number. We've also gone through hangar j package 2171-16 which Mr. Stokes described his concern with on that package. I'm wondering would it be beneficial to go through

, some of the other official record packages while we're here? Are there other types of errors that we haven' t discussed today that it might be helpful to have you identify for us?

STOKES : Well, other than the documents I gave Mr. Bishop so far which were concerned with the destruction of original calculations that showed a f ailing hangars, I gave him a list of f

S tokes /12-7 114

-- - - - - . - - ~.- . _ - - - - . . .. - _ _ . - .- -

T~

supports, or calculations to obtain earlier in the day and earlier on in the tape. The numbers that are not supplied by the copies I provided, those particular hangars are critical to the points I made earlier in the defective assumptions being made, erroneous assumptions, the juggling of loads by stress to make those hangars look good, and I'm f airly sure that within that list, one of those supports is the one that has a support located within six inches. I can' t guarantee there's more than one, but I do know there is one. I do think it is more than likely the one Gary did but I would focus my attention there first. But I do feel that within that list, that_ hangar does exist next to one of those supports. I would give, if at all possible, a very detailed look at the calculations on those supports. E specially under the final load conditions. And probably, if I had it to do myself, I'd do my own calculation with the final loads, rather than look through the cale package and try to pick out where some guy made an assumption. I would make my own assumptions under my own knowledge and see if I come up with his conclusions. Per M-9, that was an accepted design procedure for us here in the field. We did not have to check what the guy did. We had the option of performing our own cale and coming up with our own decisions and I feel it's just as reasonable for the NRC to do the same and then look and see what was turned out as a final print here agrees with their engineering judgment. That probably would be a lot easier than trying to weed through all the paperwork. 'But, other than that, I can' t really say what the other calculations may show on an individual basis. I do know Stokes /12-7 115 1

l

[- those calculations, like you said, are the hardcore problem areas

that gave people here at the site in management extreme difficulty. And f rom what we' ve seen of the cales, they all show no modification.

??: You mentioned again the calculation package that

! Mr. Ketcher did and that's on hangar no.100-32. I'm looking at an ISO I guess of that hangar, . and I'm comparina that with the Rev 1 ISO in the official record package. Does that show any obvious changes or examples?

STOKES : Yes, Gary, had already... In our discussion this

afternoon, Gary had already looked at these two packages. And he 4

explained something that I know that I consider critical. To the l average person, it might not be so obvious. He told me that his i

j model was roughly the same. But that there were f ewer nodes.

The critical thing that I can tell you immediately upon 4

inspection, without any further evaluation other than this

drawing and this calc, is that Gary modeled in the eccentricities i I of the attaching members, based on the f act they did not attach at the centroidal lines. These eccentricities seem minor on the drawing. They range f rom .5 inches to 3.5 inches. Let's see, l dimension, 4, it's obvious and these are very short seaments.

i l They are items eleven, ten, six, and twelve. The end result is i if you model out the eccentricities of the joints, you completely, you can definitely change the stress to the

!i members. Anyone that models the hangars should try to model' the hangar as closely as possible to the actual configuration constructed. There's an awful lot of people who think minor S tokes /12-7 116

F eccentricities like .5 are worth considering. On a very marginal hangar that may pass or may f ail, modeling it correct will f ail it. However, not modeling it correctly, it will pass it. It may be that in this package _the hangar f ailed and because it f ailed under Mr. Ketcher's calculation which was accurate, they didn' t like that. They remodeled it, taking out the eccentricities, a very simple. The other thing is they also noted the load point here, item 13 had an eccentricity from where the final cale shows it. He didn' t show any eccentricity on where the load was applied. That probably was accounted for in Mr. Ketcher's calc because the load was applied through a U-bolt. He modeled the load out the U-bolt location. The final calc, the ouy modeled it right at where the U-bolt attached, thereby cutting out moment, if there was sideload, and reducing stress. I would say that if you review this hangar in the field it probably should have been modeled like Mr. Ketcher's and not like Mr. Shaw's. And like I said, in some hangar conditions, that probably wouldn' t be critical, but in a failing hangar or very marginal hangar, being accurate will fail it. I've done that in many cases, I personally don't give a hangar a very thorough analysis if it looks like it will pass with flying colors. If it looks like it's going to be marginal, I give it the fif th degree. I give it everything that I know in my ability that is reasonable. If the hangar can' t take that, then it's not any good. This hangar per Gary's calc, . failed. And I feel it probably still does.

BISHOP : Bishop again. I want to ask our two technical i

experts here whether they have any specific questions they would i

i Stokes /12-7 117

- .. . . ._ - _ . - _ - ._ _=-- . .- _

T like to pursue further. Do you see a need to go through any other document packages with?

! BEZ LER: [ unclear]

STOKES : I wish I could tell you. I don' t have any knowledge of the number. I do know it exists. I am not the only person that knows that. We have not determined that it's next to

" Gary's because we do not know what's next to Gary's. I would look at that possibility first. There's two ways to go about, i you can pull the ISO, look at the ISO, or you could go to the field, using that drawing location to find your way there and to i

see what's there. The hangar may not, however, be within six inches. It could be easily downstream. If Gary's hangar was '

i i f ailing f rom an actual load, say, it's easy to put a hangar at the 90 degree location within six inches of the elbow and per stress assumptions, assume that support takes that axial load.

l And thereby cut Gary's hangar load out. But I was told by the i

people that had problems with one hangar that it was added six i inches away. I had a discussion with the stress guy who was working stress trailer at the time and I mentioned this same thing. He told me that he renembered that hangar being added.

He no longer works in the stress groups. I would like to ask him j before I give anybody's name if he would be willing to discuss I that beyond my and his discussion. But I will say that without j naming him point blank, that the names that I have mentioned in 1

j .the documents that I have, the names I have given here today, his 1

name is one of those people. If ouestioned, he should testify.

l l

I I Stokes /12-7 118 l

BEZ LER: I have one last question. I hope it's the last question. Your discussions relate to small bore hancars. Do you feel the same procedures were followed for large bore piping supports? Was there the same effort to cover up or show passino

' 1

, supports by various methods?

STOKES : The only program I was assigned to is small bore. l

'The only relationship I had concerning large bore was when I was in quick fix, I had ample opportunity because I was the ouick fix engineer for both Westinghouse, San Francisco, Bechtel and our on-site group to evaluate Westinghouse drawings, Bechtel drawings. And based upon the drawings, I could draw conclusions 1

i as to how the cales were made and defective things that were overlooked. And the weld DR I wrote, I specifically pointed out i

that I ouestioned Bechtel. [ Tape is blank for a f ew seconds) . I do feel there are problems that are related to the large bore group that I have mentioned that I know are a problem with small

bore. I can' t say that all the problems exist. I can' t say that the problem exists concerning the destruction of failing hangars,

I'm not sure that they didn' t fix all the large bore hangars that 4

failed. There's a lot smaller number of large bore hangars. I do know they did not want to do a complete re-evaluation of all the seismic small bore or safety-related stuf f here because thev l felt it would definitely postpone the operation of the plant and they did bend over backwards not to.

YIN: This is Esa Yin. The issue just raised by Mr. Stokes, relative to modeling of loadings, imposing at the actual locations, and also modeling of system geometries in accordance S tokes /12-7 119

+ r - v. - , - - - - , *-,y- -**e-*=eurw w w =s.m *-= =~e . . . + - , % --- < - - , - - -,- . - -

with actual physical frame conditions is understood and we will take a look in those areas.

BISHOP : B ishop. We've filled out a document receipt for i the documents we talked about earlier that we'd copied, borrowed, l

• copied, and returned to you and I'll keep a copy of this receipt. Is there anything from the office of OI? Anythina you

' gentlemen would like to talk about? Let me hold off on that question. I note Mr. Stokes you had some other documents there. They look like log sheets, or something like that? Xerox copies of log sheets?

STQKES : Well, these are copies of Samtam and the Thermal sample of the seismic sample sheets. I had reauested these and I ,

have been supplied them by Mr. Wah and all, these are my cooles, you have in the log sheet, the log book, in the ,back, identical copies to what I have. These were used by me to determine that i they had done intentional coverups and that the logs didn't agree with the cales. And I was hoping to stop the management from going through the calc log and correcting their problems by acting very promptly and requesting that those documents be .

removed from their authority. And now that NRC has them, and sees them, and has seen what I pointed out, I'm perfectiv comfortable with where I stand on that.

BISHOP : Okay, Bishop again. Do you have any other documents that might be useful in evaluation of your concerns?

DUKOFF: The time is approximately 7:32. We' re going of f the record. Mr. Stokes is going to review some of the other documents he brought with him.

S tokes /12-7 120

.-- . . -. . . -_ - ~ - . - - . - -.- -- . -. -. - -- - - . ._-. -

1 DUNOFF:

The time is now 8:09 p.m. and this interview is l

being' concluded. i wl 1

t 6 um I

1 I i 1

J

)

1 1

1 4

t i

i j ,

\

i 1

i i

i i

i s

i 1

6 T

j

)

Stokes /12-7 121

--..v=- -- , --_.---..----,,,y--s-,-.,,<=m- .-ww.~t. -

.--r-,-,-. .-.%,--, %.-%. -,., .r---- we,---.- , . . _ - . . . . - ,tr.-w-#, ,. ..--'g - --~

• l

...r .

N

)

1

_-fM, *"

10-20-83 I

[ -

2191 Lariat Dr.

Los Osos, CA 93402 1 Mr. T. W. Bishop, Resident ,

. i Keactor Pro,) ect s and Engine'ering Fogra~rns ~

.. .% g . v . i . . .-

- 1450 Marsa Lane -

.E ac + ,; Wa lnut . Creek,6.CA - 9 459&em c.c u.~v r e. 4- .6.i.# Maw r n. M.ev.w.a.m e _ ... .

. -O. w- w ~ ~ney. ,**e.. r . .p r4,. ..,

,. u : n r: : . . .w r ~:49. :.~' t.1.vms.*:??pg. v.p tm-wo

.....n...-. ~ .s . . :. .s .:- : . . - . . .: .. N .3 . .: '.9 e . . . - te d - ..

-W 4. .. m. 6 De ar - Mr.- B i sho p'e y9.m-&.W ~ i s t " h- W "*' + '? w'% *?'^3- H . @ % W .C -'-

i.v.Str.jgp- iske.12k"NM 9M'ttn,g Q'RHR.l'RfsT at$pr. gespgo g tgg0,,,gj!g4l e.ga gg ons concerE ng ual, beat emova M .sy.st ern, at D a a51"o,

~ Canyon. Your office. seerns to have spent- quite a bit of time on.

this investigation, and I appreciate that fact. I am concerned, however, that much of the information which I gave to Mr. powers during the interview in August of 15E2 was apparently not passed on to t he Si t e Rc;::recent at ive at Diablo Canyon. The inspector who ;araphrased" rny allegations for Inspection Report No. 50-275/E2-42 seerns to have "rai ssed the coint" on several of them.

I would like to take this opportunity to restate my position en the probl erns at Diiblo Canyon as I previously described them to Mr. Powers. I have attached a ec.py of " Allegations Regarding the Diablo Canyon Residual, Heat Rernova l Sys t ern" beginning on page 5 of Inspection Report No. 50-275/02-42, and will comment on these allegations paragraph by paragraph -

(a) I did not claim that there were no control and i nt er i c.ck circuit drawings for valves 6701 and 8702 as your inspection report stated. In fact, I provided Mr. Powers with excerpts from logic diagram 458840, and electrical schematics 437592 and 458846

~

explaining how this circuit functioned. I poin*ed out that it was rct'cleerly shown on any of them that ceraoving the power from the SSpS(Solid State protection System) out put relays we.u!d'eause the RHR suction valves to fail cicsed. I said that the power source

'c r the 55?S relays in this circuit should be s.hown on electrical Senerna t i c 437592. The omission of this infortnation from the

. electrical schematics at Diablo Canyon led to personnel error .

causing the inadvertent closure of valve 8701 and the isolation of the RHR pumps suction with a pump running (see NpPR DC1-81-Op- --

P1057 dated 3-29-81).

(b) I cid not, at any tinne during the interview with Mr.

Ocwe"s. ve tion the Mvj ed, routing of the RHR cont rol . cir-

.uitry. 1 stated that neither the Senior Control Operator or the 1&C Feronah w.?re aware that the RH R. c ont **c1 circuitry w.at routed

n r a. g. . . *nr 5SP5 and that removing the power f ror.) the SSPS out put 3.... .rel ays . woulci cause.. valves. 87 01. andt 8702> t o .,f.a i l + elosed. **Th i s . fact m n -it wM comonstrated on September 29th as rnentioned in the previous Cara ;r a DO.

3# (OMegd .

t,,

J

(e) This section contains several errors which'I have listec

. below:

1)The SSPS.does .r.igi amplify, or in any way change the signal from the Westighouse "hagan" racks to the auxiliary relay cabinets. I explained t hi s c.l e ar.l y t o Mr Powers, and

.,_,  ; :. v c ' . i . . * :. . .. .' . - . . . . _ . . . . ... .' '. . . .;, p r t i p ; t'- = ~ 7 1 et -

system function for your, office to revieaa. The person

.w w4. #.i .tg. GG &,,E, ,do. .,t clj , y.o u t h a t .. tge gog.,.i'ci s.-,,;

-- system cornpletes the l og ic f unct ion and' gt_ata,, generates prot aeet. ion,.q,. ,%..

larger . .

W=,e %+rr. w 6utfp'Ot' 'sTgKaYGriip3T')'.".'i'" Ta'd#c lis'sF..

ro.'vi sw hiWaW1.ci.*. Y. .*"'w.'M. . ." %

,,, g., ; . ; . . . ...,.....%.

.v.3..., n

. 4 .;;.~ - . . , , . -

el.eetrg.ca .t.y .. . S.*m.g~ 6 ,s .. g,.

. .- si.,,, ,t . 3 .

2)That th'is'is a ~ standard Westinghouse design'is true, which explains why the. Westinghouse reactors are cone to 44'R3M:5#t'h5%AR%Yc* tait:tnWiSWiF*f6195fQ Mei:fTn'iMVEWa . .FAM W /M'WdN

<3)-That the PG&E managernent is " unaware of .a'ny problems-with this arrangement" is curious. On 1-16-82, I submitted Design Change Request No. DCO-GE-2518-Rev.1 to PGaE Engi-reering. This DCR requested tha.t the RHR system at Diablo Car. yon he mocified to prevent reoccurrance of the i nc i d erns of 3-E?-81. Attached to it were copies of 16 Licensee Event Reports describing sitni.lar 1osses of Residual Heat Rernoval capability at other oower plants in the Country. This DCR was approved by D.A.Rockwell and R.D.Etzier, both of whom' were pr esent during the NRC investigation into these al.lega-tions. Copies of these LER's were also provided to Mr.

Powers.

(4) The raa,)or point which 1 reade to Mr. Powers was that the ssp 3 relays in this system perform ng htrdion @atsg-gve , reduce the relie.bi-lity 6f the RHR system, cause a ~

potential for damage to the RHR pumps, and should be rernoved, regardless of the fact that this is a " Standard Westinghouse Design".

(d) I pointed out to Mr. powers that several at ternpts had been made to "get ric of this p- oblem", and gave.hira a.. copy of decign change request No. DCO-GE-2518-Rev.1 w h.i c h . I described in paragrech (3) above. I also gave him a copy of Plant Tesign Cor..m e nt No. 559 which concerned valves 8701 and 8702. Apparently, your Site Representative confused a " drawing change recuest" for a " design change request".

~

. - . (e) I showed Mr. Powers that one portion of the liiablo Canyon FSRR claimed that valves 8701 and 8702 would close auto-r, : t i c s. l l y on an overpressure /overtemperature condition, w .ile another portion claimed that the power would be removed from these valves durint operation. Obviously, a valve can' t auto-

.t;:sily close wncn the power is c'ernov e d f rorc it s rec or

<..;.,..., J.ge t* a t et,J s,, t,o,,,t,h,e,Jf 8,C,,,f. i.nd,ig ,$ h at , .ty,,i.s ..,c on$ r,ad i,ct i o0, .

y ,1,,n. .$ h e ., . n ,. .,

FSAR present s no " noncompliance with regulateiry ~ requirement s",

may I call your attention to 10 CFR 50.71, paragraph (e), which recu.res that the FSAR be updated periodically (no less than

1. row..elly) enc shall " reflect all ch a r.g e s up to a maximum of 6

c. t.c ont h s prior to the date .of filing. " .

2 h.

e

- (f) PG&E rea t r.t a i ns that the spurious closure of a motor-operated valve is essent ially irnpossibl e. As'previously tre nt : or.ed, 'ed.p i'es of t he - f ol l owi rig Licensee Event Reports, all

~

instances of spurious.RHR suction valve closures, were given both to Mr. Powers and to PG&E'EngineeWng: LE R -369-81072, McGuire-1;

.i36-7 W 4, hurin %rina i ; Wo , 66, r e r a s y : 4 i. ,M 70010 T. v,,, ...

- 248-50377, Farley-1; 316-80060, Dav:s-Besse-1;344-76000, Trojan;

?I;.T . 346-80e50-1, .-Davis. Besse-1.; 316=Sc0

• 3,s.J Dav.is.Jas .,cqCryst

i. 33S-80001,%% ,.

e .n.,s . .Nortn, Anna .2;3,46-77000,,,Q,ayiz, Besse- 1,;: 0P.k-80015-1,. ~

al.. - . .

i?" Riv'er-3 ; 369-81729[,1 '.McGdire '1' ; l.51Y--74002i,ICaTver f t".CI'i ff s%,43kBD3+!#

' W ~~r800SO,Marley--F;'c318-7903BrCalverts.-Cli f.fs.lWhy pG&E. continues to". ; n , /,.

ignore this evidence is beyond me.

((e aikk'R 1 Nr YuY81N5M M U0C-2ED

" ^' ~

cont rol . room. ' OnfoFt'u'riat'ely', ' Yhe p'uinpI'rE.,t'cF orily' tri pss .afperthe"" 7 ~

pump has'been damaged by overheating due to lack' of flow; Yes, the reorsitor light boxes show RHR suction valve pcsition, but only dering 4;eident concitions, r.ot curing norreal operation. Whan valve E 7 ^.,1 went closed spuriously on 9-29-81, the Control Roore Operator would have remained unce.are of the fact until RHR pumD fa,ilure had not a consciencious painter who was working near the pump called the contro1 room due to the loud banging noises the pump was making.I sta*nd on rny original allegation: During nortr.al operation there is no control reem annunciation that an RRR suction valve is in the closed position, and there should be one to prevent darnage to the RHR pumps. .

(h) ( 3) As I explained to Mr. '

powers, Nue1 Ear Plant problem

eport No. DC1-81-OP-P1057 was i ni t d at ed 'on 9-29-81, but s i g r'ed off as complete with_gt,a_nv,. plant management review. When I ,

becarce eware of this, I contacted Juanito Diamonen, the head of.

the GC departraent at the time. NpPR DC1-81-Op-p1057 was

~

resurrected fr om the " closed" fi'.es and signed off by Jim S.xton,

u t' classified as "non-rep'orteble" and without any follow-up eetion such as an RHR pump inspection or invest 1gaticrn~into the ceuse of the incident. I alleged that both the loss of Residual

'e s t syov.>l Capability and the failure to report it were repor-table; The forraer under 10 CFR SC.72 " Notification of significant

. events", which states that: " personnel error or procedural inade-quacy which, during normal operations, anticipated operations, occurences, or accident conditions, prevents or could prevertt , by ._

~

itself, the fulfillment of the safety funcien of those strue-tures, systems, and components ir portant to safety that are needed to. . . (ii) remove residual heat following reactor. shut-down..." must be reported to the NRC.

(i) 1 ara a n a r e th t the Dicd c Canyon F5RR elains that the

i. ,,;-r . w.s.- RHR r pump, suct ion .f rom et h ee.RCS (Reaetor,, Coo.laot,;.Sy,st e4.84te, Legs,..i3 ,w.on. .,.,

not safety related, but my cuestion is wh! This systern is certainly necessary to mitiga e the consequences of an accident of tne sr. sell break LOCA type, so why is it not safety related? In the newer We s t i n gh o s.t s e and Cemeustson Engineering cesigns this syst ern is considered saf'ety related a n'd is totally redundant, so why not at Diablo Canyon?

5

(k) That the NRC Site Reoresentetive consicers it " r.o t of safety significance" that a problem report h'as been open, unre-solved, and unreviewed for 3 years is puzzling to me.

(1) DGl.E cl aims no awareness.Tof t he incorrect alarm listing

.. . . 'J - 1 .::. ** . Of th: ."1 : nt ~ ~ . :  ;. .. .~..- .u. 1;. . ' , . .

, .y, considering that I se ret rnemoranduras on thi s .sub,]ect to both Mr.

s.yy. . 1. M.~Gi sc.l or+, th e . Power .R ant *Eng i.ocer., e nd ..tbe:.second.,to. .Mr.w.: . . . . i,. s- . e C...T

....-.;,.. R . .-

these hor.nber.r.,y,.

. ..t.h e. . Pl a ndu.

t, . ring

.M.a the.

n. a ginterview.

e.r. ), .I, ,g.av, .v e. Mr.. . P.. ow e.rs, c. o pi e. s, of. ,4. .k t.wo snemorandums. . -.

.. .c .. . ...- '

_A.M_ ,y J, .5. Althou~g h;..yours.Qo,f fice ;fhas .gone <t,hr ou gh' the mot ions - of 'an4S- 5.- -

investigetion into these problems, it seems to me that the

, _ gu pGt

'r%,estions

% $9da which  % *n wserg}@

3 M N 6't6nc5 asked of_.3nN.E 8ib bwege tg.vgagor

'corsey'd'd .ineqqct T TMr. 5

• In ' sddi't' ion', i t ~a rip'e'iirs ih'at PG &E' s answers tr'even Tho'se were ' "' '

accepted without question'or follow-up. Furthermore, in the inctaneos where I provided Mr. Power with documents proving my allege.t enc, the documents were not en d e s.vailable to your site Eer;resentat ive or were ignored. .

I can only hope this and the 14 month delay in answering ray concerns can be explained merely.as a lack of communication between the. Office of . Inspection and Enforcement, your office, and the Site Representative at Diablo Canyon.

9 em= .=

+=

~ Your's Truly', "

John H.

/ .

Cooper

-m

.w-

.4 M.

3g yJ Q*a st .*** 4 *.F ar.?.A' * *'.# ..%-*.*r * * . 4. * *,4.47 t# *Jf P?

• ti. ** .%' # &* $ ' * * "J *".? 'O .*5. $*8 * * -* " "'? * ~

• d #E; %?'.* * * *** ' I*U# *I'* D *N 4

c 1y N J

-(

~

.-- j ~ 1 h- ) *4. ' (.

\! '

completed the as-built audit and has generated tee'nty-nine Open Iten Reports (01R). So f ar, the licensee has dispositioned eighteen of the OIR's. The ,

inspectors will complete the review-in this area wh'en the remainder of the 0]F.'s are dispositioned (82 42-02)., ,7 . .. -

y

. No ite=s of noncompliance or deviations were identified.

avsn ..< >.;.uv..- . . +. ee w.~ i : ow.%+--w .w-+w-s. w r".-4w.wv. Mne . ea 8,... ,. Open I tems Fol1owup ,,,,. ,,...,p...,.,........,%.,... . . . .. g. . Q.,. ., .g ., .; , j e ., e. ..

~ . . +.o. .,.

.. ._ , . . s . s . .;_e.. -. ::.w . .. ;

. C451. and D756' have been p c~

. .. . .& ;. - so .urepared to as

4. .- 4.r ~.fla,nt Admin'istrative:procedur..m . .c.es-reinstatement of- Environmental Qualif.ication conditions after maintenance

%A.4W# M[MM*NN ONMM.k.dN. -I N b C N TC M W g'l - Al TejatTons" R' Epa'EdinO t6 D'ia$To^CaGoK RI'sMaT'He'aT hem $vil System ' ' " ' " "

On December 2,1982 the inspector met with licensee representatives to discuss allegations regarding the Diablo Canyon residual heat removal (RNR) system.

The'se allegations had also previcusly been examined at the jebsite and documer.ted in Region V inspection reports 50-275/52-26 and 50-323/82-13. The following paragraphs paraphrase the allegations, summarize the inspection, and state the findings of the ingpector.

(a) Allegedly there were no control and interlock circuit drawings for motor operated valves 8701 and 8702 (RHR hot leg suction isolation

. valves). The inspector examined PGLE drawings 437592 " Residual Heat Removal Flow Control valves", and 103058 " Circuit Schedule 480 Volt for Busses F, G, H" circuitrH19?00 through HISP12 and G25F00 through G25P13. The inspector observed that these drawings describe the pcwer, control, and interlock circuits for the subject valves. The allegation was not substantiated.

(b) Allegedly no one k[lew how these circuits were routed in the plant.

Licensee project engineering personnel stated that. in. add.i. tion:to the drawings described above, the race,<ay schedule depicts circuits in a particular conduit, the conduit drawinqs show conduit locations in the plant, and the circuit schedule itsnizes the pull data for each wire in the plant. They also stated that the drawings and schedules were available to the plant staff through the site document contr61 center if this meterial was not available in the control room _The - _

inspector had previously verified that this type of documentatio^n was properly controlled and.readily available to the plant staff. This

• allegation was not substantiated. ,

(c) It was alleged that the !bgn was, r,o good in that the control / interlock circuits are r:uted f ro: ue "hagen' racks via the solid state prctection 57.5,t ich. i eengineers

.< w.,;y.# cm.that.em to.j,he,r,e.1,ay,s tfits was a standar ah'd Ve,shu.t, stinghouse th,e dEfign141ves.,,

and that}.ic,ensee.,

the "hagen" racks app) took low level analogue signals and (in this case) .used bistables to P. .

r - .,-- .-

~

. generate signals in the mi lliamp range. The s'olid state prote: tion system comple~tes th'e logic function and generates a larcer output signal (a=ps.) which in turn actuates relays in the auiiliary IBgic cabinet. ]

They explained that theyfwere not. in a position to change _this arrangement

>--- tsince it is a wes 6 nignune ce:, gn; aru 6nc6 cney er= uneu,c . .. ,

problems with this arrangement. The inspector examined the location

{n.wev.n.m of the components-of .the .RHR~ isolation-valve-control-end . inter, lock .ow.m .s

.:.wr w - . circuits..to. ver.ify,the. l.icensee'.s statenentsy... The.alleg~ation .was ,substantjated,,

:-yv. :i . - : to the . extent. that the circuits were as ellegedphowever.-there .vts.. b j J .. . .

~ ~

.7

.3,7 -

no appa'rerit'deviat' ion 'from 're'gulatory~ requirements or ' safety" criteria.4 ; -

>r.equeskd g~,f;a,e(d}Mikwas,:ca}ichdesigd.={.i.e..RHR

.s._; . . . . . a . . .. 1981 to get 'lr.id, of .that..systsn".. ho%R)psyk;.$Qti:aboAfgbrmy%9 t . .eg . suction isoiation . . ; 1.T ,

interlocks) has never been acted'upon by PGLE. The inspector verified

~*

.that there were no outstanding DCRs on PGLE drawing 437592 (which depicts the system in question) and that none were originated from or arrived at the Diablo Canyon project. The site Resident Insoectcrs verified that no DCRs were outstandi~ng for this dr' awing at the jcbsite. This allegation could not be substantjated.

(e) It was alleged that 'the FSAR, Chapter 5, paragraph 5.7, pages 37b and 38 as well as Chapter 7, paragraph 6.2, pages 3 and 4 describe the automatic high pressure /high temperature isolation of the RHR system from the reactor coolant' system, and that this is inconsistent with the technical specifications section 3.4.9-3 which requirres At to be re..oved from the associated valves (8701 and 8702-) thereby disabling the automatic isolation features. Therefc*re the FEAR should be amende'.-

1.icensee representatives showed the inspector Table 6.3-10 of the FSAR which shows that the valves are to be shut and racked out at power and open and racked out during shutdown cooling mode. This is in accordance with NRC direction.- The licensee representatives also stated that the entire FSAR would be updated (with inconsistencies removed) in September 1983 in accordance with 10 CFR 50. The allega.tlon was partially substantiated, but no saf ety problem or noncomp.liance with regulatory re:;uirements was identified.

. (f) The alleger stated that the FSAR section 3.1.3 states that spurious closure of normally open/f ail open valves is not considered as either

'a passive or active f ailure and is_not analyzed for..at all which is.

a problem, l.icensee engineers e:plained that there were no reasonable -

failure modes which would cause normally open/ fail open or normally closed /f ail closed valves to change state. The enly possibility they could imagine was a " copper octopus" which caused selective shorting.

This issue had been dealt with in -the Fire Protection Review and was cne reason that certain valve circuit breakers were racksd out after

. ..n , .m.. ...m the valve.was placed .ingthe edesired -position.. As far. as sontrol.,c.ircuits. . ., ..

are concerned, any short with 120 volts or higher would cause the locic circuits to go to a f ail safe condition due to the' overwhelming signal strer.gth (normal sige.als are 4 to 20 milliamps). The allegai. ion could

,, not be substantiated. , ,

(g) It was ' alleged that there was no low flow alarm f or the RRR system ar.d that there should be one. The inspector verified that an MR pu.o trip is annunciated, that shut RHR, suction valves are indicated, and that the subccoling meter was available to ensure adequate c re cooling. Licensee representative, pointed out that the RHR pumps have

-a miniflow recirculation'to maintain some flow, and that the monitor light box indicates valves or circuits in the incorrect state.- The inspector concluded that the 'allecation was to'rrect in..that there was no " low

.l.,./ .". i , but cli: h.:'udad t.%t th:: y ,,y, eJ '. '. ~..i. . :;uir:mer+

, or necessity to have one. .

.w s . ~~ ,. . .

' ~

..w s. .. .w. ..er  ;.ans.i~,---v.w +. . . .-~ sev 4 h - J<.w.nw;ww- . .<.w.h

.~ . af- %

m(.h) .- .It was alleged that an RHR pump ran without flow for 5 minutes in September.

1981i and ~th atthIP~eTent" Q as nit".i e'p?Irild"is' Ye'qYi red 'bfidini n is tra t fv'd 7

'. .M.. ..*, ' 2"procedu'rs 'C-12. 'and IDTCFR'. 50.72.. The site'iisident inspect'o( vetified P.' . ./f s ~

'that 'a Nuclear ~ Pla'nt' Problem Report (DCI-81-0.P P1057) and the associ'ated r@WMemWP **

M M EWEW e . afpg ) . It ~was 'sil'eged th'at the 'RHR' hot' l'e'g~suctiori ' doe's not"medt' the' single #

.. f ailure criteria for function (suction from reactor coolant system

' hot leg), that newer plants had this feature, and that this portion of the system should be redundant to . meet .10 CFR 50 Appendix A Design Criteria. The inspector verified that this function was not safety related in the Diablo Canyon plant design by examining the FSAR. The inspector observed that the suction from the containment sump and fr,cm the refueling water stcrage tank were both safety related and arranged to meet regulatory' requirements for redundancy. The inspector also o'o served that spme other plants did have two RHR suction lines but that thef a plants used a dif ferent' nuclear steam supply system vendor.

_ The inspector concluded that the all'egatiortwas correct in that the RHR suction line was redundant only for the purpose of reactor coolant system isolation, but that t.here was no apparent safety prcblem et c'eviation frc.? regulatory requirements associated with this design.

. (j) It was alleged that nuclear plant problem reports (NPPR) were not getting management review which is a violation of administrative procedure

~ C-12 and that HPPR DC 1-61-0? PIO57 had been signed off after this .

shortccming was identified to management. Other NP?Rs..sh.ould be examined.

The Resident Inspe: tors observed that other NFPRs were being given apprcpriate management review and resolution. The allegation was not substantiated.

(k) It was alleged that NPPRs DCD 79 TI P0005 and 79 TI P0117 are still

~

open after three years and should be closed. The Resident Inspectors observed that response te NPPR P0006 was complete and.that response

~

~

to F0117 was underway. The allegation was substantiated, but ne particular-safety or regulatory significance could be attached to this situation.

(1) It was alleged'that a change to the Plant Manual Volume 16, reactor coolant pump "lo oil level" alarn shculd have been changed to "lo-hi r.,il 'evel" t>. t had not been corrected eicht months af ter the correction c e....,

. p r.had.been ,submittede ..The:Res.ident, Inspectors identifiedfthis al1egationtgu w ~

to the licensee. The licensee ' initiated a NPPR {DCI-83-TN-P0001) and

'the problem is to be resolved. The licensee personnel that were interview.ed, were not previously aware of this problem. The allegatien was substantiated.

c. The inspector cencluded that the allegations. were partially correct but that ihase had no apparent safety significance or deviations from reSulatory requirements. -

A

r. . t, -  :

. . .: l,  ; -

.. , c .-  :

• i, t i t .-

.,' . o' 1

r *. M

. J -er

.i. i 1 s ( .

A.): .

.m. ,

.. 5 r .i t -

it . /, .- 4; .

4 ..:. > -

>. .*s ~

. f* }. o * . s er

'g

s. .

4 d

y e

f d .( . *

( *-' . j Le$- ' g A JollN 11. C H H'Eli 4 ( . $ / '. -  ;- /: '-

2 N t.4HIAf blH R lf y

• as j . I T

,,, k,,QO,Q ..,,, $

((' )@, j, a :?

r=sEs T,0 ===E

'/,'l,,'?"' '"' . s .:' / 4N

.ff )

s h 27, I'M l $  ?. 5 %.un t

[o .

t . 1 ..T .-, ,

? i.s 3

d.- -.- i e 21 'i c s','

,  ; r[t ..')O I IL 3 -

%[ L@.a N- 2(.A a.

it. ':: .

dti. 4 '4 59~VS; kt' g, ~ "

The.d'R*Cc0

- . . ' . .? 'l. I g f993 su .

[ittPI * .'- *)

.  ; 1 g i 4

iiv. - A -l.

).

y ,n. .g

.y( ,f. .. - '; . v.

'i M r.

. .T, ,

T. W. Bishop, Resident '

i

, i, {' 3. ,,i . Reactor Projects and Engirseering Prograriis i' Reyior V .

j j'. ' ,I g 14'i,0 Maria Lane , E

'jlf.  ? Walnut Creek, CA 94596 )

- y '. y - , - '

i e . / - , ,4

f. .!'."-

. i .

<! ? .. ;.

• 4< - - . -

(

q. 'v -,'. (0.g ,

l% ( ' e ., ',#

t

. .$c 1 .

, 0 c.

- - 2

.k >

1;i A.;

s

.g  : c -.- :*. I

{'. ;,,-

\

f, $. -

'g' P.

j '. > y:, .

. 3 t . r, 1/ . N.

'f i  ?

'f y k  ; -

a I -

i,

.?

L e O f

o' /$

<- 4 8 - t, 6.- ,a a

'q' . 'l. ., * -

• 's-6, ' g "i

.6 ' t .. 8

. t . 4' k I!!

g j $' I a8,'

I ll , }

L 6.. ' *

..: ,m <

,  : p

. , r

't - t .y

. .. : ,y' l . .e

- . .-y- .

. h.7 .-

(

l 'l .

- .i . $' '~ . .

v.

tg . ,', y J .  ;,

y > '

.* 9..t

. - $,* f, .

s

.} pE}

.. *t .

'  ?

Jotm :.unneEst '4) ( 3 , .' - ; .e; em igem 2 s*f t t.4Hl A f idtlH: f .

• 1 *

.; ,gptf,

• p;ph)M

• =

u ns w n.t 4:. s . g g #s c :: .

ci m -

). :4, 7 o u ; .{

g.;r ,

p . .m:. .

t e,:-  :

($ ,,imx, a ste' m *gm'{.gg,gsg,,.y3.p

. ' . .;. i,1 : . 3 :. .

g .

E .

4 u .o $"> -i 3)-

A .

. .. t 4> > '

t

.c

. ; * . .'s.

s. . .'. .

'y .  : Mr. T. W. Distiop, Resident 8 lt. p l' .

'f , *,  ; Reactor Projects and Engineering Prograriis j Region V . .

/j j '," ,I  %) ,;. g

- 1450 Maria Lane , I.'

j .e.  !. ~$

~

Walnut Creek, CA 94596

(

.! .' . r  ; - o, 1 . '. ,' -

./

.T ) . .

• * , e

-

• f.

- . 9. .y.*..4 .,y K y .,_ w. .

t p .:. ,3 c.

.- - i >

v.. -

.,s. :

s 4 a .}

.'l . .k, i,i.'

~ .

5 i(, h . , -

9

.y. .  ;. .

t 3.' 5b' . , , Q.

. f..'!., ;s

.f. g .

s' .

, ye  :

,r.:-

.I ,

4 s, -

a

?,. . y . t;-

s. , .

q.  : ,.

9

.  ?

p. , ' ;jr l/. ' i . .it .

i. ,

t Q. .

e p-. .. * < : .

.i

• Uf;:TED STt.TES

.i Y:, c .fl' ^; NUCLEAR P.EGULATORY COffef.!ESION

. . ;. % -- -e ,' . E

>t REGION v

.m

• N 3..r/

y** f c'

145c MARIA LANE. SUITE 210

% *a. **

' W AL NUT CR E E K. CA LIF oD Nf A Sir.5M

.. .OCT 261.983

_ . .[ V.20uNDt*M FOR: .D. G. Eisenh,ur, DirecEcz, Diy'iiic'n - of Licensine. NRR TROM: T. *d Bishop, Director, Division of Resident, Reactor

}-- ;!.g2. 5.r. . vW .

..'drejdets"and -Engineering ' Programs?RVMJ* 'M*d'l.W** . . . . . .

.rm.v-% . .w; ua s.a e.

SUBJECT:

.- . ;.r y .:s--.

A1.=":ww *Ln..WAML*w.&. 7O .m L.Weint?.&.~~k' a W->~ 'W%'.'hiW-n r . . 7- * :

LEGATIONS. CONCERNING THE RESIDUAI.' RRAT REMDVAL SYSTEM, 7..w fre ,.; A.puf.*DIABL'0*CANYOVGNITS YAND'VDOCIETdOSM504275^ AND*-" "^MW"'N"

• 50-323. ALLEGATTON NO. RV-83-A-0047 sepRtWWEeGNWENli fviEE#VMTtJF@is~'tMSisidE(8IMMNE-dM*M-[%MNM'Y'N M

-~

- Region V recently received correspondence (Attachment 1) from an~ alleger .-

concerning the Resideal IIeat Re= oval Syste= at Diablo Canyon Units 1 and 2.

Teese allegseions were previously addresse<* by Region V in NRC Inspection Repert No. 50 ~:75/E2-l.2 in respense te n11c gntiens received en August 20, 19C. A copy of the apnrepriate paragr.;phs of rcycrt no. 50-275/22-42 are previded vlth Attach =ent 1.

Evidently the alleger feels that Regien V f ailed to adequately address his

, original cencerna. Accord'ingly, Reginn ? rec.uests NRR assistance in resolving the concerns identified in paragraphs (c)(4), (e) , (f) , (g), and (1) . These concerns were disebssed with Mr. George Knighton by telephone on October 25, 1983. It is suggested a Task Interf ace _ Agreement be _ initiated on this subject. -

The felieving Inf or:ation is provided te aid in your evaluations.

Paragraph (c)(4): Evidently, removing power from the Solid State Prctection System causes the RER suction valves (8701 and 8702) to fail closed. The alleger f~ eels that rezoval of the suction head fro = the runni:5 RRR pumps vill caus~c pu.p failure. Thus, the alleger considers that such an arrangenent degrades the relisbility of the TG.R"syste:

belev an acceptable level with the attendant result of RER pu=p dan ge.

Therefore, NRR is recuested to evaluate the acceptability of this "S t andarc *destingtheuse Design".

Paragraph (f): NRR is requested to evaluate the potential for RRR suction valve closure, and the efIects thereof, in light of the LERs. -

identified by the alleger.

Pcragraph (g): NRR is requested to evaluate the acceptability. cf using G) the RRR Punp Trip nnnunciatien in Iten of a lov i2cv ala: i and (2) valve nenitor light fer indiccticn enly durins accident cc: iiticns and not daring aornal operation.

~, a' '. ..,. .. v. . .~ . p. e .;. on. .e :c ~. .w ~.~ '.r m .. * % . * .m- ., . ,m.sm w e .. n.% * <W n.e N ""4~~%' *"

. c Paracraph (1): 'NRR is requested to evaluate the con.,.,guraticn o.a qe

-iR ict ler sucticn f er co=pliance with 10 CFR 50. Appencu A. and evaluate tee .nacceptability of classifying the hot leg suctien as not m ety reAatec -

• y,, y g e/

d #

21 -

u. c. u sennue

_ 007 261983.

Picase docunent the resulns cf your evalu2:icns and for ard these to Regien V in order that thcsc m.y be incluccd in app nprihte-inspecticn docu:entatica

- sne, tnus, clumec th Ior:nni manner. luut proep. e66.n6Aun ou snA= = 1.= a

- vould be appre.ciated. -

g. r > .- :, ..+ r,; ... ,.ws. ...,.nw.w..s.w .- <+~ t . e > :.s w- -- . . + +. . m . <+=. . -

1 y.D6 D .f.5.4=E9.er..rtc3 Send 69et co"."M*Vii

. . 4Jh8 4.*HS ^PP5 P.ii"M.A"^'d+c+ Si. *w.- -

. '~,_ ' ' - -

~

-;; *. uutifications. . ., .. .. _ . g4 . ,

.,.}l.;-r .. ,.. .: e . . 1*.s.ya..n.5s,4 4. 4.4. . , . c. .: A . , -; ' * ~.s : ' - -

..~ .

.

• W -?-

.+ *  : ~ - d .~. - .c -.

l @ .=" $ :A % W & 4ti W.ASWW -*. W W }W W W W $

' p,' .' . '

,'* I. W. '31 shop ,' Dire c tor.. . . . . . ... .

Division of Resident, Reactor.

Proj ects and Engineering Prograts, Region V cc:

L. Chendler, ELD -

G. Knighton, h*RR -

H. Schierling. KRR

• ~

D. F. Iirsch, KV

?. J. F.or:ill, RV ,

ew 6

e. . . -

..g

• -" .e M-

,g 3

,M. . -

k7 8. . - * . ' - -

f. e

• L