ML20127P769
| ML20127P769 | |
| Person / Time | |
|---|---|
| Site: | Comanche Peak  |
| Issue date: | 04/22/1985 |
| From: | NRC - COMANCHE PEAK PROJECT (TECHNICAL REVIEW TEAM) |
| To: | |
| Shared Package | |
| ML20127D716 | List: |
| References | |
| NUDOCS 8505240076 | |
| Download: ML20127P769 (91) | |
Text
t ORiGhAL UN11ED STATES NUCLEAR REGULATO~ R Y COMMISSION r.
IN THE MATTER OF: DOCKET NO: 50-445 6 TEXAS UTILITIES GENERATING COMPANY (Comanche. Peak Project.)
LOCATION: BETHESDA, MARYLAND PAGES: 1- 71 ,
DATE: MONDAY, APRIL 22, 1985 a
ACb-FEDERAL REPORTERS, INC.
Official Brycrte's 444. North Cacitol Street
. 8505240076 850500 **"NIUheo.50 (202) f PDR ADOCK 05000445 T PDR NA~IONWIDE COVERACE
- _ _ _ _ = . . . .
.'. 8 5 4 0' 0 0 0 1 1 1 JWRBwrb- 1 UNITED STATES OF AMERICA-2 NUCLEAR REGULATORY COMMISSION
-3 ' -----------------------------------:
F 4 .In the matter of: :
. .' 5 TEXAS UTILITIES GENERATING COMPANY : Docket No. 50-445 6 (Comanche Peak Project.) :
o 7 ___________________________________:
~
8 Room P-422, 9 Phillips Building,
, 10 7920 Norfolk Avenue, 11 Bethesda, Mary' land.
12 Monday, 22 April 1985 13 Technical Conference between Texas Utilities and
-(, ,
14 Comanche Peak Project Team, re: Applicants' U-bolt sample, 15 whether it is representative of Unit 1. ;
f 16 IN ATTENDANCE: -
17 For Texas Utilities:
18 Jack Redding, John C. Finneran, Jr.;
19 Robert C. Iotti, Ebasco (Tugco consultant);
'}
20 R. L. Cloud, R. L. Cloud Associates, 21 (Tugco consultant)
..e -
22 Nicholas Reynolds, Esq. and William Horin, Esq.,
23 Bishop, Liberman, Cook, Purcell and Reynolds 24 For Intervenor Citizens Association for Sound Energy:
25 Barbara Dinehart P
s,. i
.'8540 00:-.02-. 2 2 .. WRBwrb 'l~ For the Nuclear Regulatory Commission:
-2 Vincent S. Noonan, Director, Comanche-Peak 7..
'3- Project, _ Division of= Licensing 4 Charles Trammell: L. Shao; David Terao; J.- 5 W. Paul.Chen; Richard H. Vollmer; Robert Bosnak; 6 E. Holler.
o 7- Jane Axelrad, Esq., J. Lieberman, Esq.,
8 S. Treby, Esq., Geary-S. Mizuno, Esq.,
9 Office of Executive Legal Director, NRC.
-.- lo 11 -
12 13 .
.(; 14 ..
15 16 .-
17.
,. 18 19 P
20 21
.o - .
j 22 23 ;
24
, -('-
r 25 -
t
6 .. . - . _ . . - _ . _ . .
8540 01 01 3 l WRBagb 1 PROCEEDINGS 2 MR. NOONAN: Good morning. My name is Vince 3 Noonan, I'm the Director of the Comanche Peak project and we 4 are here to talk this morning with the Applicant on the
. 5 U-bolt sample.
. 6 For the record, I would like to read into the 7 record about 12 items that we plan to discuss in various 8 forms this morning. One is the sampling plan and 9 procedures. Number two is randomness, statistical versus 10 otherwise. Number three is Unit 1 versus 2 samples. We are 11 going to talk about the Table 2, the torque range,'the 12 average torque, that's the table of the summaries, .
13 representative tests, equal versus unequal torques, torquing
( 14 procedures, the practice for Unit 1 versus Unit 2 and the 15 turnover of personnel performing torquing on Unit 1 and Unit.
16 2. -
17 Also of some importance, but we will discuss this 18 as to how you want to handle it, is the SA-37, SA-36,
~
19 material similar or different mechanical performance, the 20 test of material, representative steels employed in the 21 plant, the statements on variability of A-36 composition,
.e 22 the test U-bolt configuration versus field configuration and 23 the use of cinched-up U-bolts industry practice.
24 With those items on the record I will ask people
(
25 to please identify themselves as they speak in this meetingt
8540 01 02 4 1 WRBagb 1 it is being reported and I will plan to send this to the 2 Board.
3 With that, Jack -- I am going to go ahead and 4 . turn it over to Jack Redding.
7 You know, we aopreciate the opportunity to talk 8 to you this morning to clarify what we think is simply a 9 misunderstanding. It has been alleged that we, Texas 10 Utilities, have made a misleading or false statement and 11 that possibly it's a material false statement. We' don't 12 believe this to be true.
13 (Slide.)
14 ,
On this chart, a shown statement, taken frcm the 15 original affidavit that is alleged to be false or 16 misleading. Let me read it: -
17 "To determine the range of torques 18 which exist in the field, Applicants inspected
~
19 the torque of a randomly selected representative 20 sample of cinched-down U-bolt supports."
21 I think that some of the misunderstanding occurs e
22 when you take the statement out of context and then just 23 look at how the sample was taken and what we want to do this 24 morning is clear up this misunderstanding.
25 (Slide.)
. . 1 _
0540 01 03 5 1 WRBagb 1 This chart is a preview of what we want to 2 discuss with you. We'll talk about the torque sampling work 3 and discuss why it was done.
4 I want to emphasize the purpose because I believe
. 5 that focusing on the purpose and remembering the commitment 6 that we made in our original affidavit, which is on pages 34 o
7 and 35 --
8 (Slide.)
9 -- of the original affidavit. And that 10 commitment was to retorque all of the U-bolt assemblies at 11 issue. I think that's going to clear up this focusing on 12 the purpose and this commitment will clear up a lot of the 13 misunderstanding and put this issue to rest.
14 ,
(Slide.)
15 We will tal,k about how the sampling was done, why 16 it was done that way and what was done with the data, which -
17 simply was to be used in a test program. We'll also talk 18 about the test program, the reasons for it, how in part it
~
19 was done, what the results were -- which again I think will 20 further help clear up this misunderstanding.
21 We also are prepared to briefly discuss some of 22 the other subjects that relate to this that Vince mentioned.
23 By the way, the information that we are 24 presenting has beer. provided previously and it's available 25 in documents filed with the Board.
i8
F .. . . . . .. . . _ .
8540 01 04 6 2 WRBagb 1 I would like to now introduce the participants in 2 this presentation who are here with us. You know, when we 3 found out about this alleged false statement, that it might 4 even be considered a material false statement, we were
, 5 really surprised. In fact, to the point of almost being 6 bewildered. When we reviewed the material, we just couldn't 7 understand how the question of a material false statement 8 could come up. So to make sure that we weren't overlooking 9 something, we went to an individual who is new with the 10 project and asked him to review this for us. He'll be 11 ,
making a statement later on. I think most of you know 12 Dr. Cloud of cloud and Associates.
13 Also we have Dr. Iotti, Vice-President with 14 Ebasco, who originally. worked on this issue. And John 15 Finneran, a Texas Utilities engineering manager who, at 16 Dr. Iotti's direction, had the field sampling done. .
17 Also with us to review with you where this matter 18 is with regard to the Board is one of our attorneys, Bill
~
19 Horin of Bishcp, Lieberman, Cook, Purcell and Reynolds --
e 20 and Nick Reyncids just walked in also of the same firm.
21 Bill will lead off c tr presentation.
22 MR. NCONAM: Before you get started I think what 23 I wanted to say here, when we finish the meeting today we're 24 not basically going to give you our answer today, but we 25 will draft up a response and we'll be addressing the
r _. __ ,
8540 01 05 7 1 WRBagb 1 question of randomness and representativeness and I will 2 draft the response and send it to you and I'll do that 3 probably by the end of the month.
4 MR. REDDING: I understand.
5 MR. HORIN: I just want to briefly set out for 6 those of you who aren't familiar with the issue where the 7 matter stands procedurally before the Licensing Board.
8 Dr. Cloud and Dr. Iotti and Mr. Finneran will explain the 9 technical issues to you.
10 The issue that gave rise to this meeting arose as 11 a result of an assertion by the Intervenor in this' 12 proceeding that a statement that Applicants made in the 13 affidavit accompanying their motion for summary disposition 14 regarding cinched-up U-bolts was a material false 15 statement. CASE made this allegation in November of last 16 year after receiving information which we had provided at 17 the request of the Board regarding the sample that was 18 generated and used for establishing some test values in the 19 original motion.
- 20 The Licensing Deard also issued a memorandum in 21 December questioning -- raising certain questions regarding O
22 this issue. Applicants provided on January 7 a motion for 23 reconsideration of that Licensing Board memorandum wherein
, 24 we explained fully our position regarding this and other 25 issues that the. Licensing Board raised.
t_ __ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ . _ . _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ . _ _ _ _ _ _ _ _ _ .
v . . . . . . - . - . . - . - , .
'8540 01 06 8 1 WRBagb 1 Shortly after we filed our motion for 2 reconsideration, the Licensing Board chairman, on January 9, 3 requested some additional information and before he ruled on 4 our motion for reconsideration. We are preparing that 5 additional information and it should be transmitted some 6 time in the not too distant future to the Board.
- 7 However the Board placed on hold its 8 consideration of this issue until it receives that 9 information and therefore there has been no further Board 10 action on the issue since this original memorandum.
11 Now I would like to turn the technical '
12 presentation over to Dr. Cloud, who will present the first 13 portion of our presentation.
14 DR. CLOUD Thanks, Bill.
15 What I would like to do is to give you my view of 16 the matter and Dr. Iotti and John Finneran will discuss the ,
17 whole issue on a blow-hy-blow basis. But the issue of a' 18 material false statement arose out of the discussion that L* 19 was presented in the summer of 1984 regarding work done in
, 20 response to the ASLB memo and order of December of '83.
21 New I was asked to review the work done, the 22 discussion, the context in which it was made and form my cwn 23 opinion on the issue, which I would like to chare with you.
24 I would say first in my review which, of 25 necessity, is ifmited to an extent, but in my review of the
Y- . . . ... .
8540 01 07 9
.1 WReagb 1 testimony, the transcript and the af fidavits of all the 2 parties, I found no false statements related to the torque 3 sampling, the randomness or the representativeness. But 4 having read all of that, I could see how a misunderstanding 5 could have developed. The basic thing that I noticed was 6 that -- it had to do with the way the data was used and let 7 me explain.
8 The issues of randomness, representativeness of 9 the U-bolt torque sampling data would be very important 10 indeed if any of the data had been used directly in an 11 empirical manner to qualify any of the piping or supports or 12 any of the other equipment. But as you will see in abundant 13 detail, that was not the case.
14 When I say "used directly," what I'm talking 15 about there is when you use the data itself as the basis for 16 the qualification of equipment or piping o;. whatever -- and 17 you know sometimes we do that, we did not do th'at in this 18 case. The results of the torque sampling were net used in 19 an empirical vay to qualify any of the equipment., As we w 20 will discuss ;n detail, the torque sampling effort was an 21 exploratory affair to gain an understanding of the range of
~
22 the torques t, hat existed in the field and it was used to 23 provide guidance in the development of a test program in an 24 analytical effort that was subsequently conducted.
25 The test pregram that followed the torque
- 0-
8540 01 08 10 1 .WRBagb 1 . sampling and the results of that were, in turn -- and this 2 is also impcrtant -- in turn were used to aid in the
.. 3 development of an analytical method which was applied and is 4 still under review but has been applied to the U-bolt
. 5 assemblies at the plant. And the test results provided 6 guidance in the development of the analytical method and it 7 provided confirmation that the method worked properly.
8 In summary, the torque sampling data obtained 9 from the exploratory program were not used in an empirical 10 way to qualify any equipment. Whether taken frem Unit 1 or 11 Unit 2 simply didn't matter. -
12 In addition, at another point that will be dwelt 13 on in some detail, the work had advanced to the point'that a 14 decision to retorque all of the U-bolt assemblies of the 15 type that,were in question had been made at the time the 16 a f fidavit in question was written. -
17 So it appears to me that, as has happened in the
. 18 past and undoubtedly will happen in the future, that a
~
19 misunderstanding has developed and we made a mountain out of 20 a molehill. That is my view of the issue.
21 With that, Dr. Iotti will discuss the work that 22 he did and tako us on a point-by-point basis through the 23 whole affair. '
24 DR. SHAO: Can you explain the safety 25 significance of*U-bolt tighteningt if it is too tight, what i 4
8540 01 09 11 1 .WRBagb 1 does that mean, if it is too loose, what does that mean?
2 How sensitive is it?
3 DR. CLOUD: I think Dr. Iotti will discuss that 4 in great detail. But one thing I can say is that the safety 5 significance of the torquing, exploratory torque measurement
- r. 6 effort was nil.
t "- 7 DR. SHAO: You have by a factor of two or three, l
8 what does that mean?
h- 9 DR. CLOUD: That's a separate issue. The thing 10' that counts for the present purpose is what was the value, L
11 what was the use -- what use was made of that data'.
l 12 DR. IOTTI: Good morning, my name is Robert 13 Iotti.
14 About a year and a few months ago I was asked to 15 help. Texas Utilities formulate a plan that would he 16 responsive to some of the issues that the Board had raised ,
17 in its memorandum of December 28, 1983. There were several 18 issues and amongst those there were issues concerning the 19 effect of cinching U-bolts and, because of the question that
.. 20 Dr. Shao has just raised, I will try and place the 21 presentation that I am going to make this morning in that 22 context, what is the safety significance also of the 23 cinching process of the U-bolts and all the possible 24 ramifications thereof.
..)
25 (Slide.)
a -
l -
'8540"Ol'10 .
i 12 1_ WRBagb. 1 The topics I intend to cover this morning are,
! 2 first of all, the origin of the issues and the Applicants
_ 3 program that resulted-from those issues, to present an
, 4 overview of the U-bolt testing and analysis. programs. Now, 5 this overview may be too simplified for those of you who.
6 have been involved in the review of the work that we have u.
7 done for the past, oh, at least a year." But I'm sure that 8 some of you have not been as deeply involved in this program.
9' and to understand better some of the statements that I will 10 be m.aking and, in particular, to assess in your own mind 11 whether anything we have done could have safety implication 12 it's important to understand what the program consisted of, 13 how it was performed and ultimately what conclusions were
(,...
14 derived ,from such a program.
15 Thirdly, what was the purpose of this exploratory 16 sample of torques, why did we take this sample to begin 17 with.
18 Finally, how we intended to use this sample and 19 how ultimately it was used, which are two different things.
}
20 I will jump ahead to the conclusion, the ultimate -
21 resolution of U-bolt cinching issues does not utilize any of 22 the information derived from the torque sampling. Once we 23 made the decision to retorque all of the U-bolts, they would 24 be cinched in the field. And let me place it in l --
25 perspective.
Y 8540 01 11 13
'l WRBagb -1 The decision was made that those U-bolts which 2 are on supports which utilize single struts or snubbers, 3 there are ether U-bolts in the field that are cinched but 4 our decision is strictly confined to those U-bolts which
. 5 presently are on single struts or snubbers, all of those
. 6 Unit 1, Unit 2 and Unit 1 and 2 in common are presently 7 being retorqued. We maintain that with ultimate use of the 8 sampling that at the point in time we made the decision to 9 retorque that the sample we had was certainly adequate for 10 the use that we made of it.
11 DR. SHAO: Why did you make such a decision to 12 retorque? -
13 DR. IOTTI: It would be difficult for me to 14 explain to you right now. If you permit me to go forward in 15 my presentation, it will become evident to you why that .
16 decision was made and why it was, in essence, unavoidable. -
17 Maybe I should ask you, unless you have 18 questions that are for purposes of clarification, let me go 19 through it because particularly for those of you who haven't
}
20 been following this program directly, you will probably get 21 the answer to your question as I go forward with my o
22 presentation.
23 This was one of the most complex of the issues 24 that we had to face and it was the one that resulted in the 25 most testing, the most analysis and perhaps the most
8540 02 01 14
'1 >WRBagb 'l elaborate of all of the affidavits that we submitted.-
2 (Slide.)
~
3 Let me start, first of all, with the origin of 4 the issue.
. 5 The origin of the issue, quite simply put, was
, 6 the Board memorandum of December 28, 1983 on quality 7- assurance for design. In its memcrandum the Board cautioned 8 -applicants that there was, in their opinion, inadequate a 9 evidence regarding the effects of cinched U-bolts.
10 Now let me state what
- hose effects are:
. one is 11 related to the stability of the assemblies. If there exist 12 insufficient preload in the U-bolts, then it is alleged that 13 the assembly itself will not maintain its position, . may 14 rotate ,around the pipe or may slide around the pipe and 15 therefore not be able to perform the function as intended by 16 the piping analyst. That has been defined as an issue of -
17 stability.
18 Perhaps most important at that point in time, the
~
19 ancillary issues were related to the effect, once you 20 cinched a U-bolt, even if you achieved the desired 21 stability of the assembly, could you possibly introduce 22 loads which would be unacceptable to the U-bolt, cause high 23 stresses in the U-bolt, because once you cinched it and the 24 U-bolt might have been cinched to a pipe that could be 25 thermally hot and be cycled between room temperature and its:
u
'8540 02 02 15 1 WRBagb -1 operating t'emperature, could that expansion process plus the 2 external' loads introduce stresses in the.U-bolts that might.
3 prove =.to be unacceptable and in parallel with the stresses 4 on the pipe itself caused by the cinching process thermal
. 5 expansion, pressure expansion and external load all added 6 together in some fashion, could that also result in 7 ' unacceptably.high stresses in the pipe, particularly local.
~
8 So those were the principal issues related to the 9~ cinching of U-bolts.
10 The Board suggested the Applicants develop a plan 11 to provide this additional evidence that, when Applicants 12 'have stated that U-bolt cinching would in fact work, was in 13 fact the case. Applicants responded with a plan which was 14 flied -- Bill, you will have to help me -- on February --
, 15 MR. HORIN: February 3, 1984.
16 DR. IOTTI: The plan addressed many issues. -
17 Amongst those issues, three directly related to the U-bolt 18 einching and I will read them for you. Part of our plan was 19 to "... perform independent review and assessment..." of the
}
~^
20 einching of U-bolt issue and to " provide evidence that the 21 use of U-bolt cinching is appropriate to eliminate 22 potential local instability without introducing
'23 adverse effects in the piping and U-bolt itself.
24 This evidence will be generated by a combination 25 of tests add analyses."
1 8540'02 03 16-2 WRDagb 1' We plan to " provide evidence that there 2 -are no adverse long-term effects from U-belts 3 caused by heatup and cooldown and related 4 friction on the pipe. This evidence will be
. 5 generated by the tests and analyses noted in.
6 Item 3 above."
7 And finally to " provide evidence of the 8 acceptability of stresses on pipes caused by 9 thermal expansion in local areas around 10 cinched U-bolts. This evidence will be 11 generated by the tests and analyses noted in 12 Item 3 above, supplemented by a field sampling 13 of torque values."
14 (Slide.)
.15 Now to understand how the program evolved, it is 16 necessary for you to understand how the test program was
- 17 ' formulated because that will' permit you to understand what 18 was in our mind as we formulated the plan. We obviously had 19 some opinion at the time. Amongst those opinions was that
}
~*
20 what was in the field would be adequate for'the stability.
21 We were more worried about the effects of the pipe stress 22 and the U-bolt stress than we were on the issue of 23 stability. We plainly felt that almost any torque out in 24 the field would assure stability, any preload.
25 Now'the tests were performed by Westinghouse L
r 8540 02 04 17 1 WRBagb 1 under our direction. They were performed for seleczed pipe 2 and U-bolt sizes. Obviously we knew from the very beginning 3 we couldn't test all configurations. We selected 4 configurations that were representative of the
- 5 configuration that exists in the field insofar as piping and 6 U-bolts are concerned. The decision was made to select the 7 ' pipe that would represent the small-bore and the smaller 8 sized pipes and the pipe selected was a four-inch Schedule 9 160 stainless steel pipe which would be subjected to 10 ccnsiderable cycling in temperature.during its operation. A 11 pipe was selected which would represent the intermediate 12 pipe sizes and we ultimately picked a ten-inch because there
- 13. was more cinched U-bolts en ten-inch sized pipes than there 14 are on 12-inch and 14-inch. ,W e finally also picked the 15 largest pipe size that has a cinched U-bolt on it, and this 16 .happens to be the 32-inch main steam. So the tests were
- 17 performed on three pipe sizes and U-bolt sizes corresponded l
18 to the pipe sizes that were selected.
19 Because naturally the test was limited to just 20 three specimens, we had to have from the very beginning a -
l 21 parallel effort, a detailed finite element analysis effort 22 that would permit us to do two thigns: number one, to 23 extract information frem the test and derive relevant 24 information on the condition of the pipe stresses, U-bolt 25 stresses, that Oculd not be readily available from the test e
./
8540 02-05 .
18 L 'l WRBagb 1 results themselves, thus provide information just at 2 selected points, and you need a tool to extrapolate 3 information to other points.
.4 But more importantly, through an extension to
>- 5 perhaps a more simplified analytical mathematical model 6 would permit us to extend the results of'the tests, as well 7 as the detailed finite element analysis to configurations 8 that-have not been tested or analyzed. We always knew from 9 the very beginning that ultimately we would require a model 10 that would permit us to extend results of tests and detailed 11 analysis to any other configuration in the field.
12 Now because of that, the test was not done as a 13 single-shot affair, it was done in a series of' tests, and I-14 will describe each o,f the tests fo,r you and why they were 15 done. ,
16 The test series involved a test that was to -
i 17 establish a correlation between the torque that supplied the ;
18 nut of the U-bolts to each of the legs of the U-bolts versus ,
19 the preload that results in the U-bolt. As we formulated 20 the test plan, we only had the Intervenors' say-so that a 21 certain torque would result in a certain preload. Clearly 22 correlations between torques and preloads have been 23 established in the literature before but we wanted to 24 establish what the correlation would be for our particular 25 configurations."
l
[e.
.~. _ . - _ _ . _ , _ , ._ . . . . - _ . _ _ _ _ _ _ . _ . , _ _ _ . , _ _ _ . _ _ _ - - . . _ _ , - . . _ _ - . . _ _ .
f] . . . .-.
8540 02 06 19 1 WRBagb 1 A second series of tests which was directly 2 related to the issue of stability was one that was to 3 establish for a given preload what was the slip force 4 required or the moment applied to the assembly that would 5 overcom'e the frictional resistance to the torsion of the 6 assembly around the pipe. Obviously if we established a e
7 certain threshold value and all of our torques were above 8 those values, then we wouldn't have an issue of stability 9 but we didn't know what the results would be prior to 10 performance of the test.
11 The test provided two results, one, the' moment 12 needed to overcome the frictional resistance to the assembly 13 rotation as a function of the preload applied to the 14 assembly and, secondly, it also provided information on the 15 friction co-efficient that would then be used in the finite 16 element analysis and any other analytical model. We needed 17 to know that because there could be substantial variation in 18 the frictional co-efficients.
19 A third test that was called a load distribution 20 test, it's an unfortunate misnomer, was performed to
,; 21 establish the joint stiffness characteristic.
22 Now I want to clarify that point. We did not 23 establish stiffness of the U-bolt or stiffness of the pipe, 24 we really wanted to determine the stiffness of the assembly per se, how it would behave as a part[ ally preloaded joint 25
- 7. _
8540 02 07 20 l' WRBagb 1 or a-fully preloaded joint. The reason the test was 2 conducted is that the Intervenor had alleged that all loads 3 would be strictly _ adequate. If you had a preload and then 4 subsequently a pressure expansion, thermal expansion load
. -5 and an-external mechanical load, all of those would 6 absolutely combine into a large load to which the U-bolt and 7 the pipe would be subjected.
~
8 We knew that that could not be the case because 9 once you preload an assembly there are certain preload
- 10. characteristics that an assembly will exhibit and the 11 purpose of this test was to find out those characteristics.
12 And then from the results of the test we would know how to 13 best model those characteristics both in the finite element 14 analysis model and, particularly, in-the.other analytical 15 models that would later on be used for other configurations.
~
16 A fourth test was devised to determine the effect -
17 of thermally cycling the assembly. We were interested in 18 two results from that test, one, to determine the thermal 19 expansion load and, secondly, to determine the long-term 20 loss of preload characteristics of the assemblies as they 21 were thermally cycled.
22 Now perhaps it is unfortunate that the word 23 " relaxation" has been used to denote this loss of preload.
24 " Relaxation" has a fairly straight definition among the 25 material scientists. But we have used it kind of
A 8540~02 08 21 1..WRBagb 1- indiscriminately to describe general loss of preload. So 2 whenever I refer to " relaxation," please bear in mind that 3 what I'm really. addressing is general-loss of preload, 4 however it might be caused in the assembly, some of which is
- 5 due to actual material relaxations, some of which may be due 6- to assembly. adjustments and so forth.
.e 7 A fifth test was to test whether creep could have 8 an effect to assess the long-term loss of preload and, as 9 you see there, the same mistake: I used the word-10 " relaxation," characteristics of the assembly under 11 long-term exposure to high temperature. I will get ahead of 12 myself.
'13 The results showed there was no creep effect 14 whatsoever, that the maximum operating temperature of these 15 assemblies are well below the temperatures at which creep 16 could have any influence on the results.
- 17 The final initially planned test was one which 18 was intended to assess the long-term loss of preload
~
19 characteristics of the assemblies under the normal 20 vibrational levels, the pipe experience was in the plant 21 because we could not simulate it for the life of the plant.
22 At its own vibrational level we ended up simulating it in 23 accelerated fashion by subjecting the assembly to a very 24 high load and sweeping the assembly through a set of 25 frecuencies sta'rting from five Hertz up to 200 Hert and D
8540 02H09 22 1 WRBagb 1 back and forth for a period of 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. But the total test 2 simulated more than 10 to the 6 cycle of vibration that the 3 assembly could experience and we ran the amplitude of the 4 cycle at close to the load that the assembly would
. 5 experience as a result of the operating basis earthquake.
6 That answered part of our question. We 7 recognized that the assembly could see higher cyclic loads 8 even though for a much shorter duration, so we extended the 9 test to a second test which we called a simulated seismic 10 test which tested the assembly under the worst expected 11 cyclic load that we could see for the particular assembly 12 tested, and that's the load corresponding to the safe 13 shutdown earthquake.
14 (Slide.)
15 Having gone through this very simplified 16 description of the tests, let me get on to the relations of -
17 the field torque, the exploratory sampling of the field 18 torque to those tests.
19 The first three tests needed a range of torques.
20 I mean, when you test the correlation between the torque 21 that you apply and the preload that results, you don't want 22 to run that test to an infinite range of torques, you need 23 to establish some minimum torque and some maximum torque at 24 which you want to run the test.
25 Likekise, the test for the friction or the slip, L
c 8540 02 10 23 1 WRBagb l- we needed to have this range of torques.
2 Similarly, the test that was done for what I call 3 the stiffness characteristics of the joint or the prelcading 4 condition of the joint needed to have a certain range of
. 5 torque established. We could have used a totally arbitrary
. 6 torque range and, as a matter of fact, when I wrote the 7 initial test specification on January 9, which was almost 8 two or three weeks before we took any samples, that was what 9 was specified, an arbitrary torque range. However we always 10 felt it wculd be more appropriate if that torque range were 11 consistent with what exists in the field. And that should
. 12 precondition you as to why we took the torque sample the way 13 we did, we made it relative to the pipe size because the 14 test was done on different pipe sizes. We didn't do a 15 torque sample relative to U-bolt size or other 16 characteristics. '
17 Now for the thermal cycling and creep tests, you 18 would ideally like to conduct those tests at the high 19 preload value, the high torque value, because this maximizes
~~
20 any effect that might result frem the test. Again you could 21 conduct those tests at some arbitrarily high torque value; _
22 we felt it would be more reasonable if we could have seme 23 idea as to what the maximum torque values would be for the 24 pipe sizes that we were testing in the field.
25 On the other end of the spectrum, when we ran the w _--_ - - - . _ - - - _ -_ . _ . _ _ . _ - _ _ _ . _ _ _ _ _ . _ _ _ . - _ - _ - _ _ _ - _ . _-_ _ . _ . . _ _
c--
8540 02 11 24 11 WRBagb 1 accelerated vibration and the seismic tests, we wanted to 2 run those tests at about the minimum torque values that we i-3 could expect.
i 4 As it turns out, that is where there might have
. 5 been, Larry, some safety implication. Had we been 6 successful in those tests and had we found that the assembly 7 behaved stably at the torque values which were corresponding 8
to the low values we had gotten from our sample, that could 9 have had a safety implication. But as I will show you, the 10 test result immediately told us that wouldn't work so 11 -ultimately the tests were run at values which we.
12 progressively established would be required for the 13 stability issue and those values had nothing whatsoever to 14 do with what we sampled.
~
15 DR. SHAO: I don't get you.
16 DR. IOTTI: Let me go back. .
17 These tests are there to determine whether the 18 assembly loses load over long-term or whether the assembly 19 behaves stably under the maximum conditions of cyclic 20 loading, okay? Ideally you like to run these at the minimum
.21 values of torque that you can expect to have and if they 22 show stable then we say ch, this is great, you knew, the 23 assembly will work. Now had that been the case, there could 24 have been a. safety implication and we continued to use the
~
25 field data that we had acquired but we didn't and the reason u
8540 02 12 25 1 WRBagb 1 we did is the tests showed that was not the case.
2 I'm getting ahead of myself and I will have a 3 slide showing to you-the tests showed that at those low 4 values the assembly walked around, okay, and therefore we
. 5 kept increasing the torque values until we found the value
. 6 at which the assembly would stay put and that's the value --
7 at that point in time is when we made the decision to 8 retorque all of the U-bolts in the field. I have a slide 9 that takes you through --
10 DR. SHAO: So you had trouble with the lower 11 end. How about the upper end?
12 DR. IOTTI: I'll get to it shortly also and I 13 will explain to you why the upper end we never use the 14 values from the field either.
15 liR . TRAMMELL: For clarity, why don't you tell us 16 what the safety implications would have been had you done -
17 that?
18 DR. IOTTI: The safety implication, at least in 19 my mind, is the following: suppose that we had taken the 20 field sample, okay, and suppose that the field sample had 21 shown the minimum torques that we find anywhere is five 22 foot-pounds. I want to make it clear that I am now 23 speculating in order to answer your question.
24 MR. TRAMMELL: I understand but I just don't know 25 what you meant khen you said --
8540 02 13 26 1 WRBagb 1 DR. IOTTI: Let's further assume that the tests 2 that we ran at Westinghouse had shown that the assembly
-3 behaves stably at five foot-pounds. Now we could have gone
~4 forth from there and said See, what we told you or what we
, 5 had surmised was correct, the field data confirms that 6 everything out~ there is fine. Now from that standpoint the 7 sample that-was taken would not have been adequate enough to 8 make that conclusion. We would not have had enough 9 assurance that there were not torques out there possibly 10 below five foot-pounds, okay, and we would have been led to 11 a_ conclusion that yes, all assemblies were stable. But 12 there may have been some, however few out, there that would 13 not have been in a stable fashion -- would not have behaved 14 in a stable fashion. And at least without having done.any 15- analysis of what the effect of that lack of stability would 16 have been during a seismic event, to that extent there could -
17 have been scme safety implications.
18 Did I explain myself sufficiently well?
19 As it turned out, the test results completely 20 negated any possible idea that we may have had in regard to 21 that approach.
22 DR. SHAO: Where do you get the original torque 23 range values?
24 DR. IOTTI: Let me go on to the historical 25 svolution so yoc will knew how we got the torque values.
8540 02 14 27 1 WRBagb 1 MR. FINNERAN: I would like to interject one
'2 point: Dr. Iotti has used the term " behaved stably." He 3 did not mention that the test we observed on these cinched 4 U-bolts at the low torque ranges that the assembly did move 5 .around.or walk around on the pipe and that's kind of a 6 concept.that we have come to identify with Mr. Terac's 7 definition of stability. But the assembly that we tested at 8 low torques, which did walk around and did continue to carry 9 its load, we noticed that also in the test. So the movement 3
10 that we observed we decided was unacceptable but it was 11 still carrying the load in the test.
12 DR.-IOTTI: At that point in time we really had 13 two choices --
1,4 MR. NOONAN: Let me ask a question: when you say 15 " motion," how much motion are you talking about?
16 DR. IOTTI: You could not predefine what the .
17 mot' ion would be. It would behave in an erratic fashion, it 18 was not suitabl,e for prediction and that's why we considered 19 it unacceptable. At that point we had two options: we
.- 20 could have tried to continuously lower the torque until the 21 point in time where the assembly would no longer be able to 22 carry load, that would be one approach. We preferred not 23 going that way and we opted to torque.
24 DR. SHAO: You said you had trouble on the low 25 end but -- .
'8540 ! O2 15 28.
1 WRBagb 1 DR. IOTTI: I'll get to the high end.
2 MR. CHEN: When you say that the assembly behaved 3 stably, I think you indicated that you mean by that that it 4 was capable of taking the load, is that correct?
, 5 DR. IOTTI: No, let me clarify that. We have 6' always used in our own mind the definition of stability as
~
7~ one where the assembly carries the load that it's intended 8 to carry. We recognize that other people would not share 9 that opinion. Therefore the definition of stability that we 10 had in mind at the time when we opted to retorque was one 11 -that would cause the assembly.not to move.
12 13 14 .
15 ,_
16 .
17 18
-.~
19 A 20 21
-22 23-24
~
25- .
9
8540 03101 29
.1 WRBeb 1 MR. CLOUD: Excuse me. I would like to say 2 something. ,
3 We are here to discuss the use of the torque 4 sampling data in the test, and I don't think we are here to 5 discuss whether or not any given gecmetry is good, bad, or 6 indifferent, or the capability of that but, rather, to 7 discuss the consequences that ensued from the statements
-8 about the use of the torque data.
9' And I think it would be-- I just don't want us 10 to get off the track. That's all.
11 MR. BOSNAK: Bob, excuse me.
12 I think we got off on it because John Finneran 13 introduced, you know, his definition of what stability or 14 instability means. So that's why we got off on this.
.15 -
MR. CLOUD: Yes. Okay.
16
- MR. IOTTI: I guess'since we are digressing I 1,7 will digress for a few minutes more. I need to clarify that '
18 because in our affidavit we have made a statement that is 19 directly relevant to the issue.that is being raised here,.
20 and I don't want anybody to leave this meeting with any
}
21 confusion.
22 At the point when we wrote our affidavit, we 23 were using our own definition of stability. A redefinition 24 of that stability issue did not come forth until after we 25 had met with the Staff. And once we had agreed that a O,
. 8540 03 02 30 3 WRBeb 1 different definition would be used, then we have used it 2 consistently.
3 Nevertheless, even by our own definition of 4 stability, we already knew that we would get ourselves into 5 trouble if we tried to fight that battle, and we chose not
- - 6 to fight it.
7 We chose the easy way out, that of retorquing 8 every U-bolt configuration in the field. As I said, we did 9 have the option of trying to prove to everybody that 10 regardless of the icw torque, the assembly could carry 11 -load. We chose not to do so.
12 MR. SHAO: A lot of the NRC Staff are-not 13 familiar with the problem at all.
14 MR. TRAMMELL: I appreciate that, John, because 15 instability to most people means something co'llap,ses, and 16 what I hear you telling me is that the thing walked around a 17 little bit but it still held the pipe. Instability to me 18 means collapse.
19 MR. IOTTI: Let us all agree on the fact that
~
20 presently as we have redefined torques that need to be 21 applied to the U-bolt configurations in the field, those i 22 assemblies will not walk around. Okay?
23 And then I will address Larry's concern at the 24 upper end of the torque and the possible effects on pipe 25 stresses, U-bolt stresses and the like separately.
g-.
w_.
8540:03 03 31 1 WRBeb 1 (Slide.)
2 Perhap particularly for those of you who are not 3 familiar with the whole history as to how this issue has 4 evolved, let me go through an historical perspective. And 5 I have separated out the historical perspective on two
. 6 sides.
. 7 There is a tendency of confusing tests with 8 samples. On your right-hand side is any action that we took 9 that directly relates to sampling, the exploratory sampling 10 of the torques in the' field.
11 On the left side I have listed the actions that 12 pertained to the tests and finite element analyses conducted 13 by Westinghouse under our direction, so that there is no 14 confusion between the two.
15 MR. CLOUD: Chronologically? , ,
16 MR. IOTTI: Chronologically.
17 Now the very first thing that took place after -
18 the Board memo and order and even before the applicant's
- 19 plan- to respond to the memo and order was- in fact formulated-
~
20 and sent to the Board, I had already generated a test 21 specification on January 9th, and in that test 22 specification, Larry, I actually used an arbitrary torque 23 range. I had no idea what I would find in the field, but I 24 used-- I told the people that would bid on the test 25 specification that we would have a torque range somewhere h
, -8540LO3 04 32.
1 WRBeb. 1. 'between 5 and 125sfoot-pounds for all pipe sizes.
2: However, we also cautioned that some torque L
~3 levels would be maximum for certain tests and minimum for 4 other tests, and that we.would define those later on.
5- Now implied verbally was the fact that even
'. 6
.though this was the total torque range, for certain pipe 7 sizes'the torque range may be smaller; for other ones it
'8 might be larger. All of those were to be defined later.
9 We took the first' field samples to get a better 10 idea of what existed in the field on the 21st and 28th o'f 11 January 1984. This exploratory field sample was to provide 12: the-data to use in the program,.to provide the' range that
~
13 would substitute for this arbitrary range.
14 The data that came back to us showed considerable 15- . scatter of the torques'as they existed in the field. It
{ 16. also showed the presence of low torques, which was 17
~
surprising to us.
- 18 I'll get ahead of myself and I will ask 19' Mr. Finneran~later on to address this issue. But as we took
~
l -20. the field samples, we had also issued an instruction:to the j 21 people collecting the f,ield samples to continue sampling.
22 additicr.1 U-bolts as these U-bolts were being cinched. And-23 as I get to that point in time, I will ask Mr. Finneran to 24' speak to that.
25, MR. SHAO: Five to 125, that's a big range.
4
C . ;. .. . _ . . _ _ _ . _ . . . _ . : . , .a - ..n .-
t
.8540 03 L05 33
.2 :WRBeb 1 MR. IOTTI: Well,-believe me, the field sample 2 came back with a larger range, a larger range. The field
'3 sample ended up from-five to 160 -- I'm quoting from memory 4~ -- or highercthan 160.
5 MR. VOLLMER: What did installation procedures
_.: 6 call for?
.. 7 MR. IOTTI: We will have slides that address
-8 'that, Mr. Vollmer.
9 At'that point, considering the scatter of the 10 data and considering that-the data already showed a higher _
11 range than I had picked arbitrarily, I felt it was no longer 12 necessary to take any further samples. As it turned out, we 13 .got_a littl'e bit more data later on because of the 14 instruction issued to the' field forces that was never
, 15 rescinded until we started getting some of that information 16 back.
17 But the-decision to take further sampling was -
18 made at that time because I had all of the necessary data, 19 in my opinion, to establish the test parameters, the range
~
20 of torques that I would use in the test.
{s
' r- 21 MR. CHEN: Before you go on, Bob, do the samples 22 show with that ycu have quite a spread on the two legs of 3
23 the U-bolts?
24 MR. IOTTI: Yes, in some instances it showed that 25 one leg of the U-bolt was torqued to a higher value than the I
O e
m 8540 03 06 34 2 WRBeb 1 other leg. And we had a few, not many, where-in some 2 instances one leg of the U-bolts was torqued at twice the 3 value of the other leg.
4 I will get ahead of myself here again. But you 5 have to bear in mind that torque is not what is important 6 here; preload is what is important. And there is a
- -- 7 correlation between torque and preload but the correlation 8 has a variable coefficient. So just because you may in fact 9 have two torques which are vastly different and end up with 10 the identical preload on both legs simply because of the 11 variability of the correlations. And I will get into it a 12 little bit later. '
13 (Slide.)
14 Now Westinghouse-- The people were asked to 15 respond to the request for proposals. And 1;astinghouse 16 responded on the 17th of February. This is purely for 17 historical perspective. At that point in time, the test '
18 range had yet not been communicated to them. They still
.- 19 were' going on the basis of the arbitrary range.
20 We selected Westinghouse to perform the tests,
~
21 and I don't recall the precise date but it was early in 22 March of '84. And Westinghouse then requested formally that 23 we inform them of which torque range to use for each of the 24 ' specimens that they were going to test.
25' .We provided the torque ranges verbally seme time e
G '- . _. . . . . . . _.-
8540 03.07- .
35 1 WRBeb 1 in March. And the.first evidence that you find of the 2- torque ranges that were actually used in the test,-which are-3 completely. correlated to the field. sample that we got, was 4 on the 27th'of March where the first test procedure from 5 Westinghouse came out.
6 In that test procedure you will find that the 7' torque range for the four-inch pipe was to be be' tween five 8 foot-pounds and 60 foot-pounds.
9 Now if you were to look at Ta.ble 2 of the 10 affidavit you will find that_the minimum torque -- and we 11 quoted average torques, the average of the two'l'egs -- was 12 between seven and a half and 60, so we chose to lower the 13 low end to five. So we ran the test from five to 60.
14 For the ten-inch pipe, .the field sample had given 15 us evidence that the average torque ranged from ten to 85, 16 and we chose to run the test from ten to 100 foot-pounds.
17 For the 32-inch pipe, the evidence was scanty.
. 18 .W e didn't even have a particular sample of the 32-inch 19 line. We had-' samples from 40-inch linesi 30-inch lines, and 20 large lines. But we looked at that sample and decided that
. 21 a good. range for the 32-inch line would be from-20 22 foot-pounds to 160.
23 Westinghouse came back in the test procedure and
. 24 had actually given us a sample,-a range of torques from 20 25 to 260, so we immediately commented back to them that we
$ s L...
8540'03 08 36 2 WRBeb 1 felt that the 260 was too high. We told them to lower it to t 2 160.
3 Subsequently we rethought the upper range of the 4 torque and decided to disregard anything that the field 5 sample had given us and actually go to 240 foot-pounds. And 6 the 240 foot-pounds quite simply was determined by the 7 testing' practice -- by the torquing practice that was in 8 existence at Comanche Peak which calls for cinching the i ..
l 9 U-bolt by making it snug tight, " snug tight" being defined i
10 'as the full force of a man using a torque wrench.
11 We thought for a 32-inch line, because of the 12 size of the nut, you would use a three-foot torque wrench 13 and therefore, three foot times the full force of the man is 14 80 times three which gave us 240 foot-pounds. That's where 15 we got to the upper limit.
16 MR. CHEN: But none of your large size--
17 MR. IOTTI: None of them got that high, none in '
18 the field sample.
19 MR. SHAO: What was original piping analysis that 20 this analysis was based on? When you did.the original
}
~
21 piping analysis, what--
22 MR. IOTTI: The original piping analysis did not 23 address these issues other than incidentally by assuming 24 that the assembly would carry load and would function the 25 way that the analyst assumed it to function. There was no 9
s e 8540 03 09 37
- 1. WRBeb 1 preload imposed on the pipe at that point. It just is a 2 simple restraint, a rigid restraint at that point.
3 MR. CUEN: Defore you go on, you are saying you 4 based the 240 foot pounds based on this three-foot wrench.
5 MR. IOTTI: Right.
. 6 MR. CHEN: How about for the smaller diameter
. 7 pipes?
8 MR. IOTTI: For the smaller diameter pipes we 9 used the upper range of what we saw from the field, perhaps 10 augmented to make it-- Like on the ten-inch, we went up to 11 100 rather than stop at 85. For the four-foot--
12 MR. CHEN: Why wouldn't.you use the same argument 13 there?
14 MR. IOTTI: Pardon?
l
,15 MR. CHEN: Why wouldn't you use the same 16 argument thre?
17 MR. IOTTI: We could have. But you have to
18 actually look at the field data. For instance, for the
,. 19 four-inch, most of the data was around 25 foot-pounds.
20 There were one or two that went up to around 60. In fact,
. e 21 one of those had a 50 foot-pound on one leg and 70 on the 22 other, and I averaged them and we ended up with 60, and 23 that's what I used.
24 But if you were to plot a curve -- and I am not 25 representing here that there was statistical significance to l
r a 8540 03~10 38 1 WRBeb 1 that sample. But if you were to plot a curve to get the 2 view of the range, you would yourself pick the low end of 3 five, which is very few, up to 60.
4 MR. CLOUD: The fundamental point that is being 5 made is that the torque sampiing results were used only as a
. 6 guide. They were not the basis for the test.
. 7 MR. TRAMMELL: Bob, I haven't been that close to 8 this but is all this pertinent really? If we are going to 9 confine this meeting to whether or not the U-bolt sample l 10 regarding Unit 2 was representative of Unit 1, this stuff --
11 this discussion seems like it has gone far afield from that.
12 MR. IOTTI: tiot really, because that is~one of 13 the concerns, but the Board has introduced some other 14 ancillary concerns. For instance, was the sample,
, 15 regardless if taken in Unit 1 or 2, was it random,,and was 16 it representative?
17 MR. TRAMMELL: Well, let's talk about stuff like -
18 that ther,.
19 MR. HORIN: For those of you who have not been
~
20 involved in the process from the beginning, I think Bob is
- 21 presenting a detailed discussion but I think it is important 22 for people like Larry and for everybcdy here to understand 23 exactly what the sample was used for.
24 MR. IOTTI: I am trying to put it in-- I am 25 trying to answer his question.as it relates to the safety e
O 9
8540 03 11 39 1 WRBeb 1 significance. Now I think I have gotten myself ahead last 2 time in answering the. low end. I still haven't answered his 3 question, what about the high end? Why does the sample have 4 no importance at the high end?
5 And if you will permit me I will go through that 6 also.
7 MR. NOONAN: I would like to ask a question here.
8 It is kind of hot in here and we need to give the Recorder a 9 break point, so when you are at some place where you think 10 you can take a break--
11 MR. IOTTI: I think a good time would be as soon 12 as I have finished the historical perspective which I will 13 go through quickly.
14 MR. TERAO: A point of clarification.
15 When you said you're taking the U-bolt torque 16 values from the nuts, could you explain just for 17 clarification which out you are taking the torques from, or 18 the procedure you took? There are two nuts apparently.
19 MR. IOTTI: Yes. We took it.from both.
20 MR. TERAO: Two nuts on each U-bolt leg?
~
21 MR. IOTTI: Oh, no. There is an initial nut 22 which is the one that is being torqued, and then there is a 23 lock nut. We took it from the one closest to the 24 cross-piece, the one that provides the interface between the 25 cross-piece and the first nut.
<8540.03 12 40
^1 WRBeb 1 MR. TERAO: All right. So.first you would..have 2 to take off the lock nut. -
3 MR. IOTTI: You take off the lock nut, right, and 4 then you examine the other nut. Okay?
4 .
5 (Slide.)
.- '6. Well, just very briefly,-the first result that
- - 7 directly invalidated any idea we might have had of using the l 8- field sample data for any conclusion was the accelerated l' 9 vibration test which showed at icw torque levels the 10 assembly to move about. Even though it carried load, we ,
11 considered that this motion was not acceptable.
12 And at that point in time we decided'to torque 13 the U-bolts to certain torque levels that would be 14 established by both the test, the finite element analysis,
(
'15 and the other analytical methods. Okay? At that point in.
16 time, the field sampling became no longer of importance.
17 At the upper end, the decision as to what is -
18 acceptable is not predicated on the field sample; tit is
. 19 predicated on what the finite element analysis shows us to 20 be leading to acceptable U-bolt stresses, pipe stresses, 21 cross-piece stresses, and from the finite element analysis 22' to the other analytical method that is used to extend those-23 results to all configurations in the-field.
24 Sometimes those results can show torques to be 25 lower -- the torques that we choose to apply will be lower 1
l
8540 03 13 41
'l WRBeb 1 than actually the torques that will be acceptable from the 2 standpoint of pipe stresses, U-bolt stresses, cross-piece 3 stresses.
4 DR. SHAO: What allowable do you use?
5 MR. IOTTI: That can lead us into a long,
. 6 elaborate discussion. We have used two-- Well, frem the
- 7. U-bolt standpoint, -- okay? -- we used an allowable that 3 8 ultimately says if the U-bolt is stressed to less than half 9 of yield it will not relax. From the piping standpoint, we 10 had to develop a criteria because the code provides no 11 guidance. The criteria are clearly spelled out in the 12 affidavit.
13 We used a dual approach, -- okay? -- but I will 14 just get ahead of myself because this is really going to
^
15 take us out from what we are discussing now. Maybe the best 16 bet would be after we finish this is for those of you who 17 are interested, we can continue the discussion and get into .
18 the nitty-gritty details of the--
19 DR. SHAO: You are talking about the allowable f
. 20 for the upper end?
}
- . 21 MR. IOTTI: Right. What allowable we chose for 22 pipe stresses, for U-bolt stresses, those are the upper-end a
23 allowables.
24 DR. SHAO: You say the code doesn't ecver that?
25 MR. IOTTI: The code doesn't provide guidance
8540 03.14 42 1 WRBeb 1 for local stresses, other than it cautions you against
-2 excessive flattening. So you really can't use that per se.
3 You have to' develop a quantitative acceptance 4 criteria. We did, and we explained it in the affidavit.
5 After the meeting or even during the meeting but at the end
. 6 of this we can get together and discuss it. Okay?
7 Now this was the comment I made:
8 Even though we felt there was no longer any use 9 for the field sample, still some data kept filtering in to 10 us because of the directives.that had been given to_the
- 11 field forces way at the beginning that hadn't gotten turned 12 off. As soon as we realized that we turned them o'ff. And 13 so the data that was acquired like in April and May is very 14 little, and we included it in our affidavit, but in effect, 15 we almost wished they hadn't taken it.
16 MR. BOSNAK: Bob,'before you leave that, what is 17 the threshold for the vibration torque value? Did you -
-18 establich vae?
19 MR. IOTTI: Yes, we do have them. As a matter of
~
20 fact, each configuration in the field now has a specific i 21 torque value to which it will be torqued, and they are 22 presently torquing each of the U-bolts.
23 Even though we understand that you haven't 24 finalized your conclusion, that we may have a lot of 4
25 questions go answer, we are proceeding at some risk to O
8540 03 15 43
-2 WRBeb. 1 'retorque all of the cinched U-bolts or single stresses 2 numbers in the plant right now.
3 MR. BOSNAK: But you feel you have a lower limit 4 on vibration?
5 MR. IOTTI: That's correct. We are applying a
.- 6 torque value that ensures that the assembly is stable and 7- still does not cause undue stresses in the U-bolt, in the 8 pipe and the cross-piece. Le me get ahead of myself.
9 A few of the assemblies didn't pass all of our 10 criteria so we will be modifyi.ng some of the assemblies.
11 It doesn't mean that every.one of the -- how many? How many 12 did we have? -- 380-some U-bolts-- There were a $ew that 13 required modifications.
14 MR. BOSNAK: What kind of lower range and upper 15 range are you talking about?
16 MR. IOTTI: It depends on the U-bolt sizes, as 17 - well as the pipe size. Let me give you a for intance. -
18 For a Schedule 164-inch pipe, the torque of 24 19 foot-pounds will satisfy the stability condition and not 20 cause any problem with stresses. This is with a half-inch 4 21 U-bolt. If the U-bolt is a larger size, you will apply a 22 proportionally higher torque.
23 If you have a pipe which has a thinner schedule, 24 that torque may cause you a problem with stresses in the 25 pipe, and since that particular assembly may not require as
8540 03 16 44 1~ -WRBeb 1 high a torque from the stability standpoint, you apply a 2 torque of ten foot pounds.
3 The point is each assembly, each specific 4 assembly in the plant has been reviewed by an analytical 5 method which correlates very well with the finite element 6 anaiysis results.
7 MR. SHAO: What's the upper range?
- 8 MR. IOTTI: It doesn't matter because you can 9 apply higher torques and not still have problems, but we 10 don't need them.
11 MR. SHAO: You say the upper end is no problem?
- 12. MR. CLOUD: We're not communicating. Bob is 13- .saying that there is-- The analytical method being used to 14 establish the right basis predicts a single value, and he 15 gave an example of that. ,
16 Now your question is about ranges. What is the 17 question?
18 MR. SHAO: Do you have any results?
19 MR. IOTTI: It's a single number.
] 20 MR. CLOUD: That comes from the analytical method
- 21 which was the end result of all of this work for setting the 22 torques for each assembly in the plant.
23 Do you follow that?
24 MR. SHAO: No. You have a single number for 25 torque?
6 -
8540 03'17 45 2 .WRBeb 1 MR. CLOUD: No, each unique assembly has a unique 2 value.
3 MR. IOTTI: Which is high enough to assure 4 stability and low enough not to cause problems with 5 stresses.
. 6 MR. SHAO: Well, but you don't have a range any 7 more?
8 MR. CLOUD: No.
9 MR. SHAO: It doesn't--
10 MR. IOTTI: That's correct. The range no longer .
11 matters because we have a specific torque to which people 12 can-- -
13 -MR. SHAO: Your twenty-five 25 foot-pounds 14 satisfies the stability and also satisfies the load; is that 15 correct 16 MR. IOTTI: That's correct.
17 MR. BOSNAK: One more question on the icwer range' 18 again:
19 You are confident that your test program has
~
20 enveloped the vibration amplitude that might be present all
'* 21 over with the combination of sizes? In other words, you 22 have that envelope?
23 MR. IOTTI: I am very confident on that. And 24 what makes me so confident is when I correlate the finite 25 element results with the test, I have good agreement.
'O
+
8540 04'01- 46 l' WRBwrb- 'l. When I correlate the analytical method that is 2 used to. extend those results to all other configurations, I 3 also have very good. agreement.
4 So I feel very comfortable that we have 5 encompassed all possibilities in the field.
6 MR. CLOUD: That's true. And I think a lot of 7 people feel that it is a very good method. But still and
. 8 all there is, as you know, a program that is under 9 development, and much'of this will be re-confirmed as well.
- 10. MR. BOSNAK: I guess I didn't know that.
11 MR. IOTTI: Let me conclude the historical.
12 Westinghouse completed the testing in June of 13 '84. We correlated the finite element analysis to the 14 tests, the tests'were evaluated,' conclusions were drawn, and 15 we prepared the affidavit.
16 We began the effort on this analytical method 17 that would extend those results to all configurations. The ,
18 analytical method is complete. We have established now a 19 torque level for each~ individual configuration in the 20 field. And, as I told you, some configurations did not pass.
, 21 all of the criteria, and we will be modifying them.
22 Now, you have not seen this analytical method, it
'23 has not been submitted.
- All you have seen is a commitment 24 on our part to establish the historic levels in our 25 affidavit. But you haven't seen this.
~
6540 04 02 47 l WRBwrb. 1 The reason.you haven't seen it is, presently that 2 method is being verified by an independent organization.
3 This is Bob Cloud Associates. I gave you my opinion as to 4- how confident I am with those methods. Someone else has to 5 share my opinion.
, .6 In-any case, we elected to start torquing the
, 7 U-bolts in the plant, even though we are doing so at'some 8
~
risk. Bob Cloud may come back later on and say "I'm not 9 satisfied with this." The Commission may come back later on 10 and tell us they're not sa ti s'fi ed ,
11 This is a good time, Vince, to break, if you want 12 to break. -
13 MR. VOLLMER: I thought during the historical 14 part you answered a question I had, and that is: Did 15 installation procedures specify this torquing?
16 MR. IOTTI: Yes. I will get into more detail as 17 to the specific procedures used. -
18 I1 prefer starting with the Conclusions after the 19 break, because that will summarize all of this.
20 MR. NOONAN: Let me ask one more question.
'. 21 On this analytical method, do you plan to--
22 MR. IOTTI: The analytical method will be
~a 23 submitted _as part of a reply to the Board. The Board asked
. 24 us by memorandum to give them reasons as to why the 25 configurations that were tested were, in fact, 9
8540 04'03 48 1 . WR2wrb 1 representative-of all the configurations in the field.
2 What we have done is, in fact, looked at.every 3 single U-bolt configuration in the field that belonged to 4 the category.of cinched U-bolts on single struts and 5 snubbers. We have looked at all the dimensions of all the
-6 configurations, all the attributes, and we have prepared 7 tables that actually compare-- Well, I don't want to show
,_ 8 the slide right now: I have one. Part of the reply will 9 address the method that is being used to establish all of 10 the torques.
11 We will not give to the Board as part of our 12 reply the actual torques because the' method has not been 13 fully verified yet. But the actual method itself that has 14 been used to arrive at the torques will be part of our
- 15 submittal. ..
16 MR. CLOUD: In the final analysis, you will 17 obviously have to know the entire basis for the ,
18 qualification of these.
19 MR. NOONAN: I understand.
20 Let's go ahead and break for a few minutes.
}
21 (Recess.)
22 MR. NOONAN: Let's resumed.
> 23 (Slide.)
24 MR. IOTTI: A good place to start is to summarize 25 the conclusions from the historical evolution that I have O
-8540 04 04' 49
.l ,WRBwrb 1 presented. Originally the expectation was that the field 2 torques would prove adequate from the stability standpoint, 3 and that's why we oriented the applicants' plan the way we 4 did, if you look at the words of the applicants' plan.
5 This, of course, required confirmation by test, o 6 so that the original field sample, the exploratory sample 7 .that we took, was for test purposes alone, and was 8 incidental to the adequacy of verification until the test
^
9 'results became available. So'we collected that raw data to 10 set reasonable parameters =for the U-bolt testing program.
11 And that was its sole purpose.
12 The test results as well as the scatter'of the 13 data-that we got from the sample prompted us to decide to 14 re-torque all of the U-bolts in the field, those on single
, 15 struts and snubbers, but for the sample adequacies, for 16 purposes other than intended it became moot.
17 (Slide.) -
18 Let me address a few other issues that the Board 19 has pointed out.
~
~ 20 As'I said, we were very, very. surprised, almost 21 bewildered, at this allegation of having made.a potentially-22 material false statement. We think the words that we have.
23 used.after the fact; when we used those words we had already 24 -filed our affidavit, or we were filing our affidavit, we'had 25 already decided to re-torque the U-bolts. Sample was the
..- r . . - . - . . . - - - . . . . . . .. .. ,-- . .
8540 04 05 50 1 WRBwrb 1 furthest thing.from our minds, the adequacy. So we used 2 words that.in our opinion got misinterpreted.
3 For instance, " random." We didn't use random 4 sampling in any statistical sense. First of all, we said 5 " randomly selected." And it was never our intent to 6 indicate that this was a random sample in the way that a 7 statistician would define " random sample."
Randomly
. 8 selected.
9 MR. LIEBERMAN: What was it?
10 MR. IOTTI: It was randomly picked. I mean, even 11 today, trying to come up with a different word than 12 " random," I have difficulty.
13 A statistician has a well-defined meaning for 14 the word " random," a certain thing that it has to mean.
, 15 That was not o,ur intent. We randomly picked U-bolts, and as 16 Mr. Finneran will tell you later, he will tell you the 17 instruction he issued to the field force. ,
18 MR. CLOUD: It's perfectly clear in the context 19 of the statement.
~~
20 MR. IOTTI: Let me define what I mean by
, 21 " randomly picked." To me, " randomly picked" is a picking 22 that was done without any intentional bias. That doesn't 23 mean there wasn't bias, but there certainly was none 24 intentional.
25 Considering the actual use of the data, which was h
~
'8540 04'06 51 1 WRBwrb 1 to provide information for test parameters, a statistically 2 significant sample was not required, and as long as that was 3 not required you certainly didn't need a written plan or a 4 procedure or anything of that sort.
5 Had we, of course, intended to use the sample for
, '6 other purposes, we would have evaluated the same adequacy in 7 t, hat other context but we didn't, we had no other context.
8 Now, we were also picked on the word 9 " representative." Well, the torques that are applied to the 10 U-bolts depend on the torque application practices, and they 11 depend on the configuration of the assembly. Now, if the 12 configurations are the same, or similar, and the torquing 13 practices used on those samples are the same, then to me 14 it's obvious that it's representative. I don't know how
. 15 anyone can quarrel with that.
15 (Slide.)
17 Now, I'd like to ask John Finneran, since he is -
18 the person who actually caused the data to be collected, to 19 tell you how he issued the instructions to the engineers.
20 John.
- 21 MR. TREBY: Could I ask a question?
22 You said you were going to be re-torquing all the 23 bolts, all the U-bolts.
24 MR. IOTTI: On single struts and snubbers.
25 MR. TREBY: Even those that are painted?
m --
8540_-04:07 52
.1 WRBwrb- 1 MR. IOTTI: Even those that are painted.
2 MR. CLOUD: Especially those that are 3 painted. ;
4 MR. IOTTI: We're going to have to-- 'Well, we'reI 5 going to have to have written procedures for cleaning up the.
6 threads and re-torquing them, so that the paint will have no 7 -influence. And then, of course, they will be repainted
,- 8 after the process has'been undertaken.
9 MR. TREBY: I asked the question because paint-10 apparently was one of the considerations.
11 MR. IOTTI: We'll get to that,i Absolutely.
12 MR. FINNERAN: For those of you who don' t know 13 me, I'm John Finneran. I work for TUGCO, and I am the 14 project pipe support engineer at Comanche Peak.
15 Dr. Iotti requested that we take a sample of
-t 16 torqued U-bolts in the field for our testing, and I was 17 responsible for getting that sample. As indic ted on the-C 18 previous slide, there were no written procedures or 19 instructions to gather the sample data. The instructions 20 were verbal, and they were issued by myself.
21 I instructed the engineers to go into Unit 2 --
22 and I'll explain to you in the next slide why Unit 2; to go-r 23 into Unit 2 and find all;of the cinched-down U-bolts they 24 could find, unpainted, l'n the two days that constitute the 25 majority of the sample that was taken.
F t
'e
,f
e- a
^2. - di- ,
8540 04]O8. 3. 53
- _1. -WRBwrb- - 1 They were to. test thefterque on each leg of the
- l lt (J l2 -U-bolt'and record'that torque reading. They were to report-
]nqj .
(( 3 also 1 the size 'of- the ' pipe they were 'taking the torque 4: reading from.
- s. t y[
i 5
i
.To reiterate,Iit'was all the samples that they-r.
- 6 could. find in wandering through the plant.
7 MR. NOONAN: Let me ask you a question before you
- 1 8 .go any farther.-
- 19 .This transcript:is going to the Board, and I.want-
~
10 to make sure'that you tell us'why you went-into Unit 2. I 11- want to make sure that that's on the record very clea'rly.
12 And if you're going to talk about paint'I want' 13- 'you to describe for the r ec'ord why you can't use -- why you
' li 14 could not do what you wanted to do with the painted U-bolts.
15 MR. FINNERAN: Okay. .
16 (Slide.)
- 17 One of the important issues, c f course, is the' .
K 18 representativeness of the. sample that we took from-Unit-2..
19 The first thing I want to talk'about is-why I 20 directed the engineers to~go to Unit 2. At the point in
]
-- (21 time we'were taking the sample,'we knewLthat virtually all <
HD .22 the supports in Unit I had been painted. And it was my
'23 direction that'they go to Unit 2. And the reason was-that 24 I didn't think that any sampling data that we took from
.25 torqued U-bolts that had been painted would be that valid, S
k
.* y
8540 04 09'- 54 !
1.~WRBwrb- 1 because you never know, ence you get the torque reading, how 2 much of that torque was due to the original preloading of i 3 the' joint prior to painting and how much of that torque 4 reading would be due to the adhesion of the paint between 5 the mating surfaces -- the mating surface and the nut that 6 it'was cinched down on. You would never know what that 7 impact would be. We still don't quite know what it would
, 8 be. But we felt like that it would have an adverse impact
, 9 on the torque readings that we took.
10 And, as I will explain to you later, I really 11 didn't have any problems in my own mind in taking a sample 12 from Unit 2, because I had confidence that that was 13 representative of the plant.
14 Well, let's talk about construction practices 15 now.
16 The construction practice is the same in both 17 units. There is not a unique construction procedure for 18 Unit I and then a separate construction procedure for Unit 19 2. There is one common procedure, and it's followed by the 20 crews that work in either Unit 1 or Unit 2.
21 So the construction procedures are the same.
22 Now, while I'm on construction procedures, I
.* 23 would like to talk about a point that has been at issue in 24 these proceedings, and that .s the modification that 25 occurred in October of '82 to the construction procedures.
_ , . . ._.s f.8540 04 10! ~
55 2l' WRBwrb. l- There-'have been implications'that U-bolts ~ torqued 2 prior to October '82~and U-bolts torqued after october.'82 3 were' torqued.by.different means because of the revision of 4 the construction procedures. So I would like to make a few
~
51 comments ab'out that.
15 The construction procedure was revised, first of 7 all, due to a verbal request from the re,sident nuclear ~
- 8 inspector, NRC inspector, that we-change the procedure. And-
- 9- the reason for changing it was to document the way that we
~10 were currently torquing those U-bolts, and the way that we 11 had been torquing;them all along. So the revision of the
~
12 construction procedures did not represent a new wa9-of doing i 13- business as far as torquing the-U-bolts, it just represented 14- .the documentation of the way they.had been torqued all 15 along,.so that there would.be some words in the construction 16 procedure related to what.was occurring with these things.
17; Now, in order to verify this for ourselves-in ,
18 _ support of one of our affidavits that we filed later on, Sne i
. 19 examined the installation construction packages for over a 20 hundred of these U-bolts that we sampled in Unit 2, and,we
]-
.- 21 'were able,to identify that 33 of those packages actually had 22 the U-bolts torqued prior to the revision to the 23 construction procedures, prior to October the 8th, 1982.
24 A review of those torques indicates quite clearly 25 that there is no real difference in the torque ranges O
i
~ 8540 04-11' 56
~1- WRBwrb l' achieved prior to that date and after that'date. And on the 2 next slide I will show you we have a graphical illustration 3 .of that that I think will make it even more clear.
4~ MR. LIEBERMAN: Did you get the data after'the
7' MR. FINNERAN: We have to know what affidavit
-,; 8 'we're talking about. We're not talking about.the original
. 9 affidavit in support of'the motion for summary disposition, 10 it's one of the following af fidavits that we filed in answer 11 to the questions from the Board. Or it might have.been in 12 answer to CASE's motion concerning the material false 13 statement. I don't recall exactly which one.
14 But we reviewed those construction packages-prior
, 15 to filing that particular affidavit. But it was after we S'
16 had filed our original affidavit in support of the motion 17 for summary dispo.sition. .
18 MR. LIEBERMAN: So when the terms " randomly 19 selected" and " representative" were first used, what
~
20 knowledge did you have to form the view that they were 4
I 21 representative?
22 MR. FINNERAN: Myself?
23 MR. LIEDERMAN: Right.
, 24 MR. FINNERAN: I have personal knowledge of all 25 these points that I'm making for you. The review that we 6
_ . . _. a-i-
, 8540.04'12 57 1 -WRBwrb 1 did-was to offer some other supportive evidence of that. I 2 myself knew at that point.in time when we wrote the original '
. 3 affidavit why the'NRC inspector had asked us to change the 4 procedures. I nlso was aware that the practices before and
- 5. after were the same. But the review of'the construction 6 packages was done in order to gather some additional facts, 7 independent type facts, not just my personal opinion.
. 8 MR. VOLLMER: How was -hat procedure changed, ,
9 would you go over that briefly again?
.10 MR. FINNERAN: There were some words added to the t i
11 procedure that said when a pipe support, a U-bolt pipe 12 support called for zero clearance between the U-bolt and the-13 pipe, that it was to be installed snug-tight. And there 14 were already words in the procedure defining " snug-tight."
15 And I think that Dr. Iotti touched on that.
15 " Snug-tight" was. defined as the full effort of a 17 man with a stud wrench. I think he said " torque wrench," ,
18 but I believe the procedure says " stud wrench."
19 Also I think it also says "or.a few impacts with 20 an impact wrench."
21 MR. VOLLMER: So there were no specific torque
'22 values given for any specific hangers?
23 MR. FINNERAN: No, no specific torque values in.
I 24 the construction procedures.
e 25 MR. HORIN: Wher,e did the definition of l
IT -
\ T I Y
'8540 04 13 58 2 WRBwrb: 1- " snug-tight" come from?
2 MR. FINNERAN:' That definition of " snug-tight," I 3 don't remember the exact manual that it comes frem. But 4 it's a fairly common definition in the industry, the 5 construction industry. I have-seen 1,t in a textbook, and I-6 can't recall the exact textbook. But it is the full effort 7 of a man with an ordinary stud wrench.
- - 8 MR. REDDING: When he talks about changing 9 procedures, he's just talking about changing the written 10 procedures. Because the practice itself is the same.
11 MR. LIEBERMAN: You said that every accessible 12 bolt was sampled in Unit 2. Is that what was selected--
13 MR. FINNERAN: The instructions to the engineer
-14 who went and took the sample was, in the period of time that 15 he had to do the sample he was to find every cinched U-bolt 16 he could find and take a torque reading on that U-bolt.
17 MR. CLOUD: Wait a minute. But all the U-bolts ,
18 in Unit 2.
19 MR. FINNERAN: There were only about 160 U-bolts 20 in our sample, and the list of those is now in some of the
- .. 21 subsequent filings that were made.
22 MR. CLOUD: Is that clear?
23 MR. LIEBERMAN: What's going through my mind -- I l
24 hate to belabor the point, but I guess we're here to look at '
25 these words; that if the instructions were to look at all 6
F-
- i. -
'8540 04 14 59 4 WRBwrb 1 bolts that were cinched, what is the random aspect of that?
2 MR. FINNERAN: The random aspect, as Dr. Iotti 3 indicated, is that there is no intentional bias or biased 4 instructions given.to him. He wasn't told to look at 5 particular U-bolt' sizes er pipe sizes or anything, just to
,. 6* find all cinched U-bolts he could find. Cinched U-bolts is 7 the issue here: to find all the cinched U-bolts he could 8 find, unpainted. And that's really the only specific 9 instruction he was given, he was given no other bias than 10 that.
11 MR. CLOUD: But "all" in the context you just 12 used it does not -- it doesn't mean complete, not hvery one 1] in the plant. That is used in the vernacular.
14 MR. FINNERAN: There are more cinched U-bolts in 15 Unit 2 than were sampled.
16 MR. VOLLMER: Unpainted?
17 MR. FINNERAN: Oh, yes. ,
18 MR. VOLLMER: I guess I can't understand Bob's 19 clarification of "all."
~
20 MR. CLOUD: I'm saying it wasn't all: that's the
.* 21 point I'm trying to make.
22 MR. FINNERAN: The question was, were all the 23 unpainted cinched U-bolts in Unit 2 sampled? That's not the 24 case.
25 MR. VOLLMER: You gave him so many hours to go
~
l
+ . . .* ........ ..- ..
{
8540 04 15 60 2 WRBwrb 1 find all he could; is that right?
2 MR. FINNERAN: Yes; the bulk of the sample was 3 taken in two days. And you can see that if you look at the 4 list of the sample.
S- MR. IOTTI: 'Maybe that's a good definition of 6' " random:" as you run across them, take them. Through your 7 walk-down, whenever you see a cinched U-bolt that is
. 8 unpainted, go sample it.
. 9 MR. LIEBERMAN: And that tells you what l 10 percentage of of all of the U-bolts in the plant were--
11 MR. FINNERAN: I don't have the information to 12 answer that question. But we'll try to get that.f6r you if 13 it's important.
14 MR. LIEBERMAN: Was it 10 percent or 60 percent?
l 15 MR. FINNERAN: I hate to try to give you a figure-16 off the top of my head. We have sometimes done that in the 17 past and have suf fered for it. so I don't want to do that 18 again.
19 If you want a more specific answer I'll give it
~
20 to you. But I would like to go back and get it and give it
. 21 to you. !
'22 MR. LIEBERMAN: Fair enough.
23 MR. FINNERAN: There has also been a lot of
. 24 discussion in these proceedings about the people involved--
1 25 MR. HORIN: Before we go ahead: Is that a r
l i
k
8540.04 16 61 1 WRBwrb 1 request, that we go back and do that?
'2 MR. LIEBERMAN: No, no.
3 MR. TRAMMELL: It might have had scme bias. But, 4 you know, any sample-is biased. You gave them two days to 5 to do that; is that what I heard? Was it a two-day effort?
6 MR. FINNERAN: It was basically a two-day 7 effort. And it was all that they could get in those two
- . 8 days.
9 MR. TRAMMELL: Obviously they were going to get 10 the ones that were easiest to get to. That's bias, but I 11 don't know that that is necessarily significant.
12 MR. FINNERAN: I don't know if they got'the ones 13 that were easier to get to or not.
14 MR. TRAMMELL: I would suspect that they did.
15 MR. FINNERAN: Well, I dpn't know if that would 16 have some bias.
17 MR. TRAMMELL: I don't know either. I say they ,
18 probably got those that were easiest to get to.
19 MR. FINNERAN: They got to as many as they could
~
20 get to.
. 21 MR. IOTTI: Let's not lose sight of the fact that 22 we were' to look for a range under which the tests were being 23 performed. We couldn't care less whether they were all easy 24 to get access to or not. We might have a problem if the 25 information all came back and had nothing at all in the high N
O
.'8540 04 17 62 l' WRBwrb 1 pipe sizes, which we knew we were going to test. But as 2 long as there was information, it wouldn't bother us.
3 MR. TRAMMELL: Well, what I mean is, I understand t 4 your statement, but it just might be that all the ones that 5 are easy to get to are all the ones that were very nicely 6 torqued. Who knows?
7 MR. IOTTI: That could be.
-. 8 MR. TRAMMEL: That's not necessarily true,
. 9 either.
10 MR. FINNERAN: I can't think of any reason why 11 that would be the case. The construction procedure would be 12 the same regardless of where they are. .
13 Let's raove on. ,
14 Another issue in these proceedings, as I said 15 before, has been that the people involved in torquing the 16 U-bolts in the construction crews were somehow different in 17 Unit 1 than in Unit 2.
18 There appears to.be some misconception. That 19 construction on Unit 1 took place in some span of time, and 20 then construction on Unit 2 took place in kind of a separate
,, 21 span of time, with maybe a small overlap. In actuality, 22 just the opposite is true. The installation of supports in 23 Unit 2 started in late 1977, the installation of supports in 24 Unit I was not complete until about 1983. So there was 25 about a six-year span of overlap there of on-going work in 26 both Unit 1 and, Unit 2.
I
. .. -- - .-......a....
- 8540 05 01 63 1- WRBeb 1 My own personal knowledge is that there is a 2 movement of construction crews in that ecmmon span there 3 between the units, because the work efforts are not always 4 the same in each unit. That effort kind of ebbs and flows, 5 and the majority of the crews move back and forth between 6 units as work needs dictate.
7 To further verify this fact, and again in support 8 in one of our later affidavits that we filed, we looked 9 again at the construction packages that we had reviewed, as 10 I mentioned earlier, and we were able to identify about 45 11 or so crew foremen, lead men associated with those 12 construction packages, and about 27 or 28, I think'it was, 13 of those crew foremen were still at the job site.
14 We' personally interviewed them and we were able 15 to determine that 25 of those guys had worked i'n both Units 16 1 and 2, so this further supported our feeling that the 17 sample in Unit 2 was representative of the torques in the ,
18 plant.
19 (Slide.)
~
20 I think I indicated to you that I would show you
- 21 a graphical representation of the terque values that were 22 taken prior to the revision to the construction procedure in 23 October of '82 and those afterwards.
24 The vertical scale on this chart is the number of 25 torques, and the horizontal scale represents the torque I
k
t "8540 05 02 64
'l WRBeb 1 range'for those number of torques.
2 You can see here right at this point is 50 s 3 foot-pounds, and this chart here indicates that in our total 4 sample there were some 18 supports taken that were in the 50 5 foot-pound range, and there were t. bout-- It looks like
. 6 about six here supports previous to October of '82 that were 7 in that same range. And these supports are included in this 8 total sample, this total here.
9 So what you can see here is kind of a graphical 10 representation of the range of torques,.and this is for all 11' pipe sizes. And there was no real difference in the range 12 of these torques for the ones before the construction I 13- procedure changed and the ones after. You see about the 14 same distribution in the range.
15 ,
MR. CHEN: Excuse me, John. Are those averages?
16 MR. I,OTTI: No. Those are actual for each nut, 17 for each size. ,
18 MR. FINNERAN: I have other charts that are 19 broken down by pipe size which I can show you if you want.
20 They show about the same type of distribution.
! . 21 MR. VOLLMER: This difference is after the 22 procedure change which specified snugging?
23 MR. EINNERAN: Well, the implications were that 24 the procedure change resulted in a different way of f
25 installing cinched U-bolts. Our point was that they were l .
t i
t
'8540 05 03 65 ,
1 WRBab 1 always installed that way, and they weren't installed any 2 differently afterwards, that there was just now 3 documentation in the construction procedure indicating how 4 they.always had been done.
5 MR..VOLLMER: Well. let me ask a further 6 question. .
~
7 Do you draw any conclusions as to why there is i t
. 8 such a gross spread in the torquing values for a given --
- . 9 quote -- procedure? <
l 10 MR. FINNERAN: Well, we have talked about that,
'll and the only conclusion we can come to is that people are 12 different, and how they do things is sometimes different.
13 But our bott5m line conclusion and why we didn't discuss it 14 any further was we had made the decision that we were going 15 to go back,and torque these to specific measured torques, so 16 it kind of lost meaning for us. I 17 MR. VOLLMER: Well, I don't understand the graph 18 too well, but it looks like before '82 is better than a f ter 19 '82, like you have more low torques after '82. You have a
- 20 lot of low torques after '82.
~
21 MR. FINNERAN This is prior to '82.
22 MR. VOLLMER: Yes, prior to '82 seems to be a >
e 23 little better.
24 MR. FINNERAN: Now wait a minute. This is not 25 the torque reading over here. The reading is down here.
l*
8540 05 04 66 1 URBeb 1 MR. VOLLMER: The torque reading is the other 2 way.
3 MR. FINNERAN: So the distribution of torques is 4 the same for-- You have a high number over here and a low 5 torque range, and a low number over here in the higher 6 torque range, both before '82 and after '82. It's the 7 relative distribution of this line down here to this line.
, 8 MR. VOLLMER: What I see is you have a lot of low
. 9 torques after '82.
10 MR. IOTTI: You also had it before '82.
11 MR. CLOUD: To a certain extent it is distorted 12 by the size of the sample. The sample after '82,is a much 13 bigger sample. And all this is really saying is that most 14 of the torques are low.
15 MR. IOTTI: There's an answer for it, too. The 16 absence of the low torques for the prior to '82 is that not 17 a single one of the one-inch, one-and-a-half-inch and 18 two-inch pipe were -- that fell within our samples fell into 19 the prior to '82. And that is what fills this part. Okay?
- 20 MR. SHAO: Okay.
[ 21 MR. IOTTI: The point is there is no statistical 22 trend that would lead you to believe that there is really 23 any difference.
24 MR. VOLLMER: I guess what you're saying is -- if 25 you will pardon my use of the word -- it is randem in both 0
0
8540 05 05 67 1 WRDeb 1 cases.
t 2 MR. IOTTI: You are absciutely right.
3 MR. FINNERAN: I would like to turn it back to 4 Dr. Iotti for further comments.
5 MR. IOTTI: Well, I guess there is one more issue 6 that I would like to address before I present conclusions,--
7 (Slide.)
. 8 -- why we used the average torque for the two legs. This is
. 9 something else that had perturbed the ASLB Board.
10 The reason in my mind is rather simple. We had 11 always intended to use those torque samples for the test. I 12 knew that I would apply equal torques during the test to 13 both sides of the leg. Now this does not necessarily mean 14 that the same preload will result on each side, but I would 15 use the same torque on both sides of the leg.
16 And the reason for that is that insofar as U-bolt 17 stresses are concerned, insofar as the pipe stresses, 18 thermal expansion loads, cross-piece stresses, it is the 19 total load at the interface between the U-bolt and the pipe
. 20 and the pipe and the cross-piece which matters, and that in
',, 21 turn depends on the total load, you know, a combination of 22 the two legs, not on each individual leg.
- 23 Likewise for stability reasons, it is the 24 frictional resistance to rotation which matters, not 25 individual leg resistance versus the other.
0
F . j 8540 05 06 68 1 WRBeb 1 That is why we summed up the test that way, and i
2 that is why I felt it appropriate to average the torques. !
l
! 3 Now ultimately the issue again is irrelevant. We 4 had decided to torque all of the U-bolts. The way we 5 decided to torque them is to apply an equal torque to both 6 sides. We are not going to go torque one side higher than 1 7 the other. Again that practice may result in unequal
, 8 preload but the torques will be the same. I
. 9 And bear in mind that during the test, even 10 though we started with equal torques, some tests did in fact 11 start with unequal preloads. And you can see evidence in l 12 the test reports of Westinghouse which are attached to our 13 affidavit, we applied the same torques but some legs were 14 preloaded higher than other legs.
15 So the test did address this effect to some l 1 l
16 extent, certainly to the extent that we intended.
17 (Slide.) ,
18 Now conclusions. Here are the conclusions.
19 Because of the decision to retorque, the field l
. 20 torque sample and whether the sample was random or
~
, 21 representative is moot.
22 However, there was no question ever in our mind
- 23 that the sample was random enough or representative for the l 24 use that we would make of it. You may disagree with the 25 choice of the words, but it is our choice of words, of 1
l O
I i
8540 05 07 69 1 WRBeb 1 random or representative, and what we intended to mean.
2 In retrospect, for completeness perhaps we should 3 have mentioned all the things that we mentioned today, such 4 as sampling only on Unit 2, the averaging of torques. That 5 we didn't even consider it may have been an oversight which, 6 from my point of view, was really caused by'the 7 insignificance of the sample to the ultimate conclusion that 8 we presented.
9 That's it I guess.
10 Jack, if you want to add any further words to 11 this?
12 MR. REDDING: I would like to add to your 13 conclusions, Bob.
14 (Slide.)
15 Again we want to emphasize and ask you to 16 remember what the purpose of the torque sample was and that 17 is the torque sampling work was done for exploratory 18 purposes only, not to shcw adequacy.
19 The data from the torque sampling work was used
- 20 as an aid in the establishment of an in-depth experimental
[, 21 and analytical effort. So for the purposes of the use of 22 the data, the torque samples were sufficiently random and
. 23 representative.
24 The results of the experimental and analytical 25 effort led to the decision to reinspect and retorque all of L
. _ - - - _ - - , . - . m. .
8540 05 08 70
'l WRBeb 1 the U-bolt assemblies at issue, and further, this decision i 2 was made prior to and communicated in the af fidavit of June 3 29th, 1984, that led to this issue.
4 The torque sampling data was not relied upon in 5 any way to demonstrate the adequacy of any Comanche Peak 6 hardware'.
7 MR. TRAMMELL: I.have a question, and that is was 8 the document in which you stated that the sample was 9 randomly selected and representative -- or maybe those 10 aren't the exact words you used. But is that the same 11 document in which.you said you were going to retorque 12 everything? -
13 MR. CLOUD: Yes.
14 MR. TRAMMELL: What type of document was that?
15 MR. FINNERAN: That was in the affidavit that was i
16 filed in support of our motion for summary disposition 17 regarding cinched U-bolts, and that was filed June 29th, 18 1984.
19 MR. TRAMMELL: Thank you.
20 MR. NOONAN: Are there any further Staff i -
'l 21 questions at this time?
22 (No response.)
- 23 Basically what I will do is ask the Staff to 24 write up their answer as to the randomness and 25 representativeness of the samples, and I will communicate
8540 05 09 71 1 WRBeb' I that back to the utility within in about -- oh, probably 2~ next week some time. Before the first of May I plan to get 3 it to you.
4 With that, I guess I don't have'any further 5 questions, so I guess we will call the meeting to an end.
6 And I thank you.
7 For'the record, since this is a public meeting, I
, 8 wanted to offer the CASE representative an opportunity for
. 9 public comment. She is not here at the present time but she 10 told me earlier she would have no comments, so let the
-11 record show that.
12 (Whereupon, at 11:00 a.m., the meeting was 13 concluded.)
14 15 ,,
16 17 18 .
19 e
20 s- 21 22
~
23 24 25 D
6__ __i______.___1______..
CERTIFICATE OF OFFICIAL REPORTER This is to - certify that the att, ached proceedings before the UNITED - STATES NUCLEAR REGULATORY COMMISSION in the matter of:
NAME OF PROCEEDING: TEXAS UTILITIES GENERATING COMPANY (Comanche Peak Project.) ,
DOCKET NO.: 50-445 PLACE: BETHESDA, MARYLAND DAT'E :
. . MONDAY, APRIL 22,' 1985
.were held as herein appears, and that this is the original -
transcript thereof for the file of the United States Nuclear Regulatory Commission.
e (sigt //
(TYPED)
WILLIAM R. BLOOM Official Reporter Ace-Federal Reporters, Inc.
Reporter's Affiliation t.,j e
i .
' ~
OVERVIEW OF PRESENTATION REGARDING U-BOLT TOROUE SAMPLE t
ORIGIN OF ISSUES AND APPLICANTS' PROGRAM OV3RVIEW OF U-BOLT TESTING / ANALYSIS PROGRAM PURPOSE OF EXPLORATORY SAMPLE OF TORQUES, I . E. WHY SAMPLE OF FIELD TORQUES WAS UNDERTAKEN ORIGINALLY INTENDED USE OF SAMPLE ULTIMATE RESOLUTION OF U-BOLT CINCHING ISSUE DOES NOT UTILIZE INFORMATION DERIVED FROM TORQUE SAMPLE, I . E. ,
DECISION TO RETOROUE (PRESENTED IN 6/29/84 MOTION FOR
SUMMARY
DISPOSITION) MOOTS QUESTIONS REG ARDING SAMPLE CHARACTERISTICS
('
SAMPLE ADEQUATE (APPROP'IATELY R RANDOM AND REPRESENTATIVE)
FOR ULTIMATE USE, AS DESCRIBED IN MOTION FOR
SUMMARY
DISPOSITION O
e 4
e 9
9 O
L
.. .. - ~ _ . _ . . . ._.
ORIGIN OF ISSUE BOARD MDf0RANDLH AND ORDER (QUALITY ASSURANCE FOR DESIGN),
. DECEMBER *28, 1983 .
T .. .' '
INADEQUATE EVIDENCE REGARDING EFFECTS OF CINCHED
- , U-BOLTS SUdGESTION FOR DEVELOPMEh7 PRESEhTATION OF ADDITIONAL EVIDENCE APPLICAhiS' PLAN TO RESPOND TO SOARD MEMORANDUM AND ORDER INDEPENDEhT REVIEW AND ASSESSMENT OF SEVERAL ISSUES PLAN REGARDING CINCHED U-BOLTS
"(3) PROVIDE EVIDENCE THAT TE USE OF U-BOLT CINCHING IS APPROPRIATI TO ELIMINATE POTEhTIAL LOCAL INSTABILITY
, WITdOUT INTRODUCING ADVERSE ETTECTS IN TEE PIPING AND s
c U-BOLT ITSELT. THIS EVIDENCE WILL BE GINERATID BY A
( .
COMBINATION OF TESTS AND ANALYSES.
- (4) PROVIDE EVIDENCE THAT TdERE ARE NO ADVERSE LONG- .
TERM ETTECTS TROM U-BOLTS CAUSED BY HEATUP AND C00LDOWN
', AND RELATED TRICTION ON THE PIPE. THIS EVIDENCE WILL
~
BE GENERATID BY TE TESTS AND ANALYSES NOTED IN ITEM
.. (3) ABOVE.
- (5) PROVIDE EVIDENCE OF TE ACCEPTABILITY OF STRESSES
' ~
ON PIPES CAUSED BY THERMAL EXPANSION IN LOCAL ARIAS AROUND CINCHED U-BOLTS. TdIS EVIDENCE WILL BE GENERATED BY THE TESTS AND ANALYSES NOTID IN ITEM (3) ABOVE, SUPPLEMEhTED ,BY A TIELD SAMPLING OF TORQUE VALUES."
8 L
SUMMARY
OF TEST / ANALYSIS PROGRAM ,
WESTINGHOUSE'S SERIES OF TESTS WAS PER70RMED FOR SELECTED PIPE Aht U-BOLT SIZES.
o -
TESTS NECESSITATED A PARALLEL ANALYTICAL EF70RT (7EA) TO AID IN TEST RESULT INTERPRETATION AND IN EXTENSION TO OTEER CONFIGURATIONS NOT COVEPJD BY TEST.
TEST SERIES INVOLVED:
^
TORQUE VERSUS PPILOAD - TO ESTABLISH CORRELATION SLIP FORCE VERSUS TORQUE (PRELOAD) - TO ESTA3tISH AS TUNCTION OF PRELOAD
. A. MOMDIT NEEDID TO OVERCOME TRICTIONAL RESISTANCE TO
( -
ASSEMPLY ROTATION B. TRICTION COEFFICIENIS LOAD DISTRIBUTION - TO ESTABLISH J0ENT STITFNESS CHARACTERISTICS AND BEHAVIOR AS PRELOADED JOINT
~
THEPJfAL CYCLING - TO ESTABLISH
,, A. THIF. MAL EXPANSION ET7ECTS
- 3. LOSS OF PRELCAD (RELAXA* ION) BEHAVIOR
~
CREEP - TO ESTABLISH LONO-TERM RILAXATION BEHAVIOR -
ACCELERATED VIBRATION - SAME SIMULATED SEISMIC - VERIFY BEHAVIOR UNDER WORST CYCLIC LOAD EZPECTED a
n _. _ _. ._ ._ ____ -_ __ _ . _ _ _ _ __ _. . _ _ _ _ _ _ _ _ _ . _ _ _ . _ . _ _ - .
, USE OF TOROUE SAMPLE DATA IN TEST 7
e FIRST THFIE TESTS NEED A RANGE OF TORQUES TO DEFINE TEST 30CNDARIES -
RANGE COULD HAVE BEIN ESTA3LISED WITHOUT SAMPLES GIVEN TF2 OVERALL
, PROGRAM OF TESTS AND ANALYSIS. WE E T IT WOULD SE M0?2 APPRC?RIATE
, IF IT WERE CONSISTINT WITH RANGE EXISTING IN FIELD.
THERMAL CYCLING / CREEP TESTS - TO BE CONDUCTED AT EIGH PRILOAD LEVELS.
USED VALUE EQUAL TO OR HIGHER THAN 03 SERVED IN FIELD SAMPLE. IOULD HAVE USED ARBITRARY VALUE. FEA/ TESTS CORRELATID EVALUATION MITFt0D Ui.TIMATELYDETERMINEDACCEPTABLEVALUE.
ACCELERATED VI3 RATION / SEISMIC - WERE TO HAVE USED MINIM lM VALUES
~
0F TORQUES SEEN IN THE SAMPLE. FIRST RISULTS INVALIDATED THIS APPROACH. VALUIS UN?ILATED TO SAMPLE WERE USED.
e e
e 4
- -e e
4 9
u.
. . . . . . ~ . . .
HISTORICAL. EVOLUTION OF SAMPLING AND TEST / ANALYSIS PROGRAM 4
TEST / ANALYSES TOROUE SAMPLE *
- TEST SPECITICATION (REQUEST TOR PROPOSAL) , ISSUED 1/9/84
, A. TORQUE RANGE SPECITICATION i
. SET AT 5-125 TT-LB'.
- 3. TORQUI LEVELS FOR CERTAIN TESTS TO BE MAXIMLH OR MINIMUM VALUES TO BE DETINED LATER FOR TESTED PIPE SIZES.
- FIRST TIELD SAMPLES TAKZN ON
(
1/21,28/84. EXPLORATORY SAMPLE TO PROVIDE DATA POR USE IN PROGRAM SHOWS SCATTER OF VALUES AND PRESENCE OF LOW TORQUES.
- TURTHIR SAMPLING NOT WARRANTED.
SAMPLE TO 3E USD To ESTABLISH TEST PAPA M ERS.
- m. _.
_ .. ., e.- _.
4
-W, _
ITT RESPONSE TO RFP l
(2/17/84). PRECISE TEST ~
RANGE NOT YET COMMUNICATED.
- SELECTION OF E'FOR U-30LT CINCEING TESTS.(3/ /.84). ...
l .
- - W. REQUESTS TORQUE RANGES TROM l
TUCCO (3/9/84). - TUCCO TRANSMITS TORQUE RAN5ES TO W VER3 ALLY.
~
- TIRST W TEST PROCEDURE (3/27/84) INCLUDES TORQUE RANGES (TEESE ARE CONSISTENT
( WITH TIILD SAMPLES. SEE , *
- TABLE 2 of 6/29/84 AFTIDAVIT). - TUCCO COMMENTS ON 32" MS UPPIR LIMIT - TROM FIELD SAMPLING (3/28/84). ,
I
- TUGC0 FURTHER REVISES 32" MS UPPER i LIMIT OF TORQUE RANGE (VER3AL
~'
,, COMMUNICATION APPROXIMATELY 4/13/85) -
240 FT-L3 PREMISED ON TULL TORCE ..
_ 0F MAN WITH 3 FT WRENCH (NOT TROM .,
a ,
SAMPLE). ,
i L
--. - r _ . _ . . . . _ . _ . . _ . _ . _ . . . . . . _
TEST RESULTS TROM ACCELERATED VIBRATION SHO*a'S ASSEMBLY ,
MOVDChT AT LO*=* TORQUE VALUE.
ASSEMBLY MOVES'UP TO TORQUIS
~
A30VE MINIMUM. VALUES OBSERVED ..
IN SAMPLE.
FIELD SAMPLE NO LONGER 07 IMPORTANCE.
APPLICANTS DECIDE TO RITOROUE ALL CINCEID U-30LTS ON SINGLE STRUTS OR SNUBBERS IN UNITS 1 AND 2 AND COMMON AFI.AS. (APRIL-MAY, 1984)
ADDITIOXAL FIELD S.49LES
.t , (4/14 - 5/24/84) .
. 'IHERE WAS A CONTINUING EFFORT TO SUPLE A FEK CINOED U-EOLTS AS Trs' 3 WERE ' INSTALLED, hHICH WAS '1TM!IRATED c ANTR DE DECISION TO RETORQUE ALL U-BOLTS AT ISSUE.
- e
- ed 4
h 4 v
e 4
\
.i A
d y ?
g j k '
e e
6 w- : , ._
f t
W TESTS ARE CGIPLcir.u (JUNE 1984)
FINITE ELEFENT ANALYSES ARE CORRELATED 'ID W .
e -
- 4 TESTS. TESTS ARE EVALUATED. CONCLUSIONS XRE . .
DERIVED. AFFIDAVIT IS FREPARED. #
,9 i i- . .;
~
EFFORT BEGINS ON ANALYTICAL METHOD TO SET -
TORQUES LEVELS FOR ALL' CONFIGURATIONS INCLUDING U-
, . . BOLTS, FIFE, CROSS FIECE STRESS EVALUATION AND CONSIDERAT]QN OF RELAXATION., . . .
ANALYTICAL FElliOD CG!PLeit. INDIVIDUAL TORQUES SET FOR ALL CONFIGURATIONS. FEK REQUIRE FDDI-FICATION.
ANALYTICAL hETriOD BEING VERIFIED BY INDEPENDEhT ORGANIZATION (RLCA). U-BOLTS BEING TORQUED IN PLANT AT SGE RISK (UNDERKAY FRESENTLY).
b O
e 6
. 9-
,f e
.g e
4
. e
. . . . . . . . . , _ . . . . . . _ . _ .. .._i._ _.. . _
N CONCLUSION FROM HISTORICAL EVOLUTION i I ) ' --
ORIGINALLY, THE EXPECTATION WAS THAT THE FIELD TORQUES WOULD BE ADEQUATE t
THIS REQUIRED CONFIRMATION BY TESTS
~~
-ORIGINAL FIELD ~ SAMPLE WAS FOR TEST PURPOSES AND WAS INCIDENTAL TO ADEQUACY / VERIFICATION UNTIL TEST RESULTS BECAME AVAILABLE COLLECTED RAW DATA TO SET REASONABLE PARAMETERS FOR THE U-BOLT' TESTING' PROGRAM
~
TEST RESULTS AND SCATTER OF DATA FROM SAMPLE PROMPTED ULTIMATE DECISION TO~RETORQUE
. p SAMPLE ADEQUACY FOR PURPOSES OTHER THAN INTENDED (SET PARAMETERS FOR TEST) BECAME MOOT
(. .
o
/
8 h
. e e4 I*
- ee 9
I .
w*
l t
-9 6
- . . - - - .e , , - , , ,, ,--m.,, ., - -
EdAT WAS MEANT 3Y " RANDOMLY SELECTED. REPRESENTATIVE" SAMPLE 7 TERMS ACCURATE AND APPROPRIATE IN CONTEXT: TERMS APPLIED TO SAMPLE ONLY IN CONTEXT OF AND AF~ER ULTIMATE DECISION TO RETORQUE A. RANDOM '('IRANDOMLY SELECTED")
" RANDOMLY SELECTED" WAS NOT INTENDED TO INDICATE " RANDOM SAMPLE' IN STATISTICAL SENSE RANDOMLY SELECTED MEANS PICRED "AT RANDOM" WITHOUT ANY "INTINT10 rial" BIAS ' - - --
CONSIDERING ACTUAL USE OF DATA (PROVIDE INFORMATION FOR TEST PARAFJ.TERS) STATISTICAL SIGNIFICANT SAMPLE WAS NOT REQUIRED, HENCE, NO WRITTEN PLAN OR PROCEDURE FOR SAT. LING
- HAD WE USED SAMPLE FOR OTER PURPOSES, WE"WOULD HAVE EVALUATED SAMPLE FOR ADEQUACY IN THAT OTHER CONTEXT MIGHT HAVE BEEN (E.G., IF USED TO SUPPORT ADEQUACY OF VdAT WAS IN FIELD (7dICH IT WAS NOT)
(, WOULD HAVE REQUIRED SUPPLEMENTATION) BUT SUCH EVALUATION NOT REQUIRED GIVEN ACTUAL USE OF SAMPLE B. REPRESENTATIVE TORQUES APPLIED TO U-BOLTS DEPEht ON
~
TORQUE APPLICATION PRACTICES
- C0hTIGURATION OF ASSEMBLY IF CONFIGURATIONS ARE SAME OR SIMILAR AND TORQUING PRACTICES USED ON SAMPLES ARE IDENTICAL, THEN SAMPLES ARE REPRESEN'"ATIVE 9
e
=
. METHOD OF COLLECTING DATA t
8D VER3AL INSTRUCTIONS TO ENGINEERS RECORD TORQUES OF BOTH NUTS ON Ah7 (SAFETY OR NON-SAFETY)
CINCED U-3OLT FOUND UNPAINTED IN UNIT 2
_ TORQUES CECKED BY TIGHTENING NUTS USING CALIBRATED TORQUE WRENCES
.( .
e
.S 6
13 e
amm .
G
- = g
.e4 4
9 4
b REPRESENTATIVENESS OF UNIT 2 TOROUESr e
s.
. NECESSARY BECAUSE ONLY UNPAINTED U-BOLTS WERE IN UNIT 2 CONSTRUCTION PRACTICES ARE SAME IN BOTH UNITS PROC EDURE,S ,,,
PROCEDURE DOCUMENTED PRACTICE AT REQUEST OF NRC RESIDENT INSPECTOR
' RESULTING TORQUES AFTER PROCEDURE SHOWS NO DIFFERENCE FROM PRIOR TORQUES .
PEOPLE OVERLAP OF UNIT 1 AND 2 SUPPORT INSTALLATION CREWS SHIFTED BETWEEN dNITS O
e
'e t.
m g-#
e
'O e
S r w .--n ,---.w , ,- y ,, - - , , - - ~-- - - - - , -- or - - -- -- -
.w---
1 1
1 4 .
q e
\ I s
e
.1 4 , ,s; x,
q -
. -- A s
N N
. 1 O
- 1 1
gs ss s_s s,* s. s.ss s*,
,\\ ' i
'_) 4
,\ j
,\ l
,s hl s
,.. . \sss
(' , w\\
k s,
N
\.
%g .%% g\ %
. r ,,s s - ,s . s,,
\
\ '
s'
- \ i
' s s % % s N s s s '.
- s. ..,,,s ss .
1
" A W
g %% s %ss%\\ '.gs% %
. ss s - ,s e ,, . . s s
- **', x . gk I
'. s us.m ,%. .s d
' ,ssss . sssg \
>s <
n.
~
. ,s u s sss
'# ). s s ss% . ss s
. s q
..s. .s . .. .s, ...,, ,
.,,s.s.ss.. .s..
,s. . , . ,Q f e
. e sQ O o m %
$$hW/$[
M,'
.m m ema4 w e, e m ,> g a S
9 0
e
=
D. s. o 1 *
-.n m
. .._ M
- s
= . m --
- D t,
N %
.. a cc T
k s
w M
sssss
- s. m 3
N s'sN' o
,, 3=
'S o
f'e &
p
,sss ss.
O
- ssss s 9 I
g
% ks \ \%
%% ss s \
l-
- N
\ sNNNNN t . ld , e I' %'A r
I t
1 osw e' i
.._u 1 %
i i
nnom 10 -#
e m
-~ - . -. . . . . . . . .. .. ,
- / //
b- /4 Pitt.s 9
10 ~
l n -
/ /
/
/
s f s/ /
4 e
r s
t e -
a R,
W,'
is ,
l l s
/ 'rasiss a < -u-i s ls a
's Yn im;
- t s s
r aa, t- ,
s s
< ' p,^M -
s C - ~ ~ ~ ~ '
s *
, /0 40 50 70 90 l 'O /3 O
( -
13ssucsa rec.y u-200 as (;'r-2) e .
a
> /4 Ncs J
w-i 4 g-
.4
.-s .
.s.
.s
- ? -
c-l
- -, m, m -
Abe ut ea F1
. .,,, ,,,, s w - - ,,o r -
m ; c-5 pi i I! - , , . , ,H,
.. b ,
. xy, , , ,
c,' ., , , , , l.
~ SO' fO /Og /SO
~ '
,'- , 76devk"//J EACH UEetr ll6 &Ylh )
,.. l'd nice n ip/sr.-
l-r
. - .- .. e..
. k7.Y AVERAGE TOROUES OF ~WO LECS? r
- ALWAYS HAD INTENT OF USING TORQUE SAMPLE FOR TEST TEST WOULD .. . . _
APPLY EQUAL TORQUES ON BOTH LEGS - THIS DOESN'T MEAN .
_e.
SAME PRELOAD (PRELOAD CANNOT BE READILY DETERMISTD IN FIELD)
TOTAL PRELOAD OF U-30LT IS THE IMPORTANT PA.MMETER FOR ASSESSING PIPE STRESSES, THER$AL EXPANSION LOADS, CROSSPIECE STRESSES.
THAT IS WHY TEST WAS SET UP WITH EQUAL TORQUES ON ECTH LEGS EFFECT ON "STA3ILITY" 0F UNEQUAL TORQUES IN TWO LEOS IS IRRELEVANT
.BECAUSE RETORQUING PROGRAM WILL ENSURE THAT THIS DOESN'T HAPPEN
-- EVEN WEEN TORQUES ARE SAME, PRELOAD CAN BE DI ::.AENT. THESE 7
\ .. . DIFFERENCES WERE OBSERVED IN THE TESTS e
e 9
4 m
4 l 6 s
-b- D a = = e. amN ese-, se a e- ey er on ee.e mm .esew. w e ==, e ..eewe ****p seem e ow.e * *w -q=. .- *- - eem m e s ,
, , A*.'. .
=
CONOLUSIONS
- BECAUSE OF DECISION TO RIIORQUE, FIELD SAMPLE AND L.:.i m SAMPLE RANDOM 01 RIFRESDMIVE IS MOOT
- C05SIDERINC REPRESDiTA"IVENESS OF SLv2LE AND ITS INSICHIFICANCE g TO ULTD'. ATE DETERMINATIONS AND CONCLUSIONS , TEE NEED FOR ur. :- <
DISCUSSION OF SAVFLE LMS k'AS NOT CONSIDE2ED
- ~
(
e W
~^ ~ ~
'E CONCLUSIONS i
THE TOROUE SAMPLING WORK WAS DONE FOR EXPOLORATORY PURPOSES,
, NOT TO SHOW ADEQUACY.
f^
THE DATA FROM THE TORQUE SAMPLING WORK WAS USED AS AN AID IN THE ESTABLISHMENT OF AN IN-DEPTH EXPERIMENTAL AND ANALYTICAL EFFORT.
THE RESULTS OF THE EXPERIMENTAL AND. ANALYTICAL-EFFORT LED TO .
THE DECISION TO REINSPECT AND RETORQUE ALL OF THE U-BOLT ASSEMBLIES AT ISSUE. FURTHER, THIS DECISION WAS MADE PRIOR TO AND COMMUNICATE] IN THE AFFIDAVIT OF JUNE 29, 1984 THAT k- LED TO THIS ISSUE.
~
THE TORQUE SAMPLING DATA WAS NOT RELIED UPON IN ANYWAY TO DEMONSTRATE THE ADEQUACY OF ANY COMANCHE PEAK HARDWARE.
L- .
. . . . . . . . . . . . . _ __ .. . . . . . ._ ._ . ...