ML20093D719
| ML20093D719 | |
| Person / Time | |
|---|---|
| Site: | Comanche Peak  |
| Issue date: | 10/01/1984 |
| From: | Doyle J Citizens Association for Sound Energy |
| To: | |
| Shared Package | |
| ML20093D689 | List: |
| References | |
| OL, NUDOCS 8410110347 | |
| Download: ML20093D719 (82) | |
Text
- - - - - - - - - -
n -
o 4 .y
[
so l
^ ~
00CKETED UNITED STATES OF - AMERICA _ USMC NUCLEAR REGULATORY COMISSION .. !
BEFORE THE ' ATOMIC SAFETY AND LICENSI BOM 20 In the Matter of ETGF
'('
4 3cSECW".
j ; J.o: ,' ,
-TEXAS UTILITIES GENERATING i. Docket Nos'."50-445-1 COMPANY, et alc l and 50-446-1 (Comanche Peak Steam' Electric Station l Station, Units 1 and 2)- 1
. CASE'S FIRST MOTION FOR
SUMMARY
DISPOSITION e _
' ~ '
REGARDING CERTAIN ASPECTS OF THE IMPLEMENTATION 0F APPLICANTS DESIGN AND QA/QC FOR DESIGN 1
in the form of AFFIDAVIT OF CASE WITNESS' JACK D0YLE.
l
- 01." What is the purpose of this Affidavit?
A1. .Having stood back and taken a b ng look at-the overall picture of what is going on regariing the Applicants' Motions for Summar Disposition.
I perceived the need to file adCitional information for the Board pointing out and summarizing the following which outlines the failure by Applicants to comply with the provisions of:
10 CFR Part 50, Appendix A, specifically Criteria 1 and 4;
~10 CFR Part 50,' Appendix B, specifically Introduction, last
, . paragraph, which states, in part:
". . . Quality assurance includes taality control, which
. comprises those quality assurance actions related to the physical characteristic of-a material, structure, component, or system which provide a means to control the quality of the material, structure, component, or system to predetermined requirements." (Emphasis added.) *
. . . not manicuring of a fait accompli; e
8410110347 841006 PDR ADOCK 05000445 s O PDR , 1
r'
- . - , . 1- -.
., e: ,
,n -
^
Criterion I; i
2 Criterion II,'specifically last 17 lines, which state:
3 .
"The program shall take into. account the need for special controls, processes, test equipment, tools, and skills to 4
attain.the required quality, and the need for verification of quality.by inspection and test. The program shall provide 5
, for indoctrination and training of personnel performing
.a tivities affecting quality as necessary to assure that 6
suitable proficiency is achieved and maintained. The applicant shall regularly review the status and adequacy of 7
the quality assurance program. Management of other rganizations participating in the quality assurance program 8
shall regularly review the status and adequacy.of that part of'the quality assurance program which they are executing."
U*'***' " IIII 10 Criterion IV;
))
Criterion V; Criterion VI; Criterion VII; Criterion VIII; 15 .
Criterion Ix; Criterion X; Criterion x1, especially lines 8 through 14, which state:
"The test program shall e iude, as appropriate, proof tests
'19 prior to installation, p.soperational tests, and operational tests during nuclear power plant .-. . operation, of 20 structures, sy'tems, s and components." (Emphasis added.)
21 ' Criterion xV;
' 22 Criterion xVI; 23 Criterion XVII; and 24 criterion XVIII.
25- Applicants will also be sh6wn to be in non-compliance with:
2 4<,e -
Y '
k J_< l
,- 3
. :. N ,.i -
- . i +
7 .y i
Q-l fl '
/ ?!
r- .-
f.
W .Y. I: .
,34
.,3;-
M1 ANSI N45.2.11; , ')-
2' "~Y, - I ASME Section III;. . ',
'3
- D 'o
. AISCi (American Institne'of ,>
Steel Construction), to which ,
- ).-
4
{- Iv 4 Applicants are committed in their own Specification s
5 'x 2323-MS-46Af \ "Nuc, lear Safety Class Pipe Hangers and
' 6 '
4- Suports," at pages 3-15 through 3-20, specifically
'7 paragraph 3.3.h.; , f 8 s MSS (Manufacturers' Standardization Society), specifically A
9 .
. i; MSS .SP-58, " Pipe Hsngers and Support-Materials ant' -- t
.n 10 b!!' Design," end MSS SP-69, " Pipe Hangers and Supports-
-)) v. ,
SelectiobandApplication,"towhichApplicantsare3 s - . v, +
q Jcommitted in their own Specification 2323-MS-46A ac page
/[r > 4
- 6 . .$13 'I1~~l' 3-20, paragraph g.(1) and (2), respectively;
' e)4 - , \'i f'*
.10'CFR 50.34(a) and (b) . . . among others. , s ..
. / ,
p s i- !; i ,
15- e In short, Applicants' OA/QC program for,engineerine, hr< Comanche
'16 Pe'ak pot only has suffered, but currently is' suffering,'from a complete C
~
37~
'/
?, collapse,.thus allowing for fundamental engineering errors to be s
N
~'
~ 18 -
incorporated 1sto Comanche' Peak on such a masalve. scale that the hecith' '
.' . ,1;I->t
[39:h<a,,1. i
- b' .and safetylof'the public are at risk. ,
, 1 -
l t O :
gQ ,. L '- - v
' 02. Are the po, int's ivhit h you are about to make representative ot} c ,1.arge 1
+
t.? t ;
^
n perc5ntageofkheengineeeJngand[onstructionat. Comanche' Peak? #
, 4 "22 A2. .No! The-majority of ,the 7 points and allegations are based only on 23~ 'several hundred of thE more than 40,00G' p'ipe supports which are
- g. ' M'/ t s. .
.lt installed at Comanche Peak. One must keep in' mind that the typical qs .'c i
. ,, j
' twin nuclear f aci1Nty crht
, n's about 20 miH, ion elemental parts. For 4j; , E .%['
, .\Vr 3
.) -
t ) \
'( 'E h i k
I ) *' . =i. . .
.M_',,'. + ., , . - ._. - - . . .I,+ ),
<- <y .
, .y . .e -
r , , .
- f1 . example, the Davis Besse nuclear-power plant, one unit of only 900 MW
- , 124 capacity,:- contained, in part, approximately: _
h?3: :12,000 tons of structural steel; E4 .
50,000 cu. yards'of concrete; 1
5 12.million feet of copper wire;.
6: :100,000 ft. of' instrumentation pipe and tubing; P ':7!- 85,000 ft. of pipe _over 2'-1/2 in. diameter; y
8' 91,000 f t. of small bore pipe .(under 2-1/2 in. diameter;-
ithe' welding required for the piping was 20,000 girth' welds (appx.);
9L
$ L10 ~ -
- there.were over 12,000 pipe supports with hundreds of thousands of illI joint welds;
'.'l 2 : .the-numbers involved in the electrical, HVAC, and cable trays
- 13. rival the numbers cited for-the piping systems;
$14- this plant required over:10 million man-hours of manual labor and 15- over a-million' hours o'f non-manual labor.
s L 16 - 'As may be noted, it does not take much to reach 20 million 01 7 ielements.
[18 ' .Now'when'one is discussing a sample of components which may be
- q19 l : comprised of a few tens of-thousands of elements (welds, nuts, bolts,
~.:20; . plates, structural members, hardware, etc.) compared to 20 million 1217 elements, it cannot - be called -:a large sample. But when this small 122' ? sample indicates a complete failure to control fundamentals, then it is
' 'M 23 - ,
indeed fair to state that the safety of the complete facility is at
' 7 24- :best indeterminatecif not dangerous.
325: :
'03. Jou say your sample was relatively small. could you explain this?
4 s w . ~w - - . - , a--., ,,,..,,,-,--.,..,,-,c - , + , + , - . - . . . . --,...---.---c , , - - ,- . - - , . - , - ~ ~,, . . , , , . . - - . .
.. 3-il A3. Certainly. In my-original deposition, I offered documentation on 52 node points which contained 58 separate supports (X, Y, or Z
.3 -independent supports). .In addition to these supports, I have since had 4 the opportunity to' examine 10 calculations and drawings from the Cygna
.5 - Phases l'and 2 review. Further in relation to Cygna, I have reviewed 6 their findings on 22 main steam supports as outlined in Volume 3 of the L7 - Phase 3 segment of the review (" Final Report, Independent Assessment 8 Program of Comanche Peak Steam Electric Station (Phase 3)," Prepared by 2
9 Cygna Energy Services, July 16, 1984). This represents a total of some 10: -90' supports or so which I have examined to some extent (not including 11' the cursory review of the Phases 1, 2,~and 3 Cygna Reports). In 112 addition to_this, I have reviewed several of the Applicants' answers 13 (Motions lfor Summary Disposition, etc.) to generic allegations; for
- 14 example: (a)~ analysis on one-way constraints (U-bolts) which act as
'15 ' two-way_ constraints; and (b) generic stiffness study; to name two.
16 ' 04. . Taking these four items one at a time, what did you conclude from your
-17' analysis of the 58 individual supports. offered into evidence as CASE
- 18 Exhibit 669B (admitted at Tr. 3630)?
.19 - A4. To start with, the Nuclear Regulatory Commission (NRC) investigation of
~20~ the original allegations was remiss in its objectivity, to say the
. 21 least (see also discussion in CASE's 8/22/83 Proposed Findings of Fact 22 ~and Conclusions of Law (Walsh/Doyle Allegations), pages XXVII - 35 423- -through -39). .The SIT Report, which (coupled with the Affidavit of the 24 Staff's Witness Dr. Chen on open Walsh/Doyle items dated 10/14/83) was 25' your basic whitewash, arrived at two general conclusions:
5
,x :.
e l- (a) All of-the allegations were without merit with the exception of a.part of one allegation as related'to bending of Richmond
~
2;
- 3:
insert bolts (A-307 threaded rod); and
-4 '(b) .The Applicants were aware of any points of an allegation 5 which they could not talk away before Mark walsh or I
- 6. outlined the problem (stability,.for example).
7 :. On point (a), Dr. Chen would not accept his own finite analysis 8 :which showed an 18 per cent distribution of shear loads to the outer 9 _w eb of the tube steel which induced high stresses in the bolt (see i L10 . Staff witness Dr. chen at Tr. 6533-6536). Beyond this, Dr. Chen stated 11 that' Applicants also found'high bending stresses during their i 112 investigation-(see Tr. 6533-6536). Dr. Chen and Staff Witness Mr.
, .13 _Tapia preferred to accept the test data which was the result of testing 14 - with a flat plate and not a tube (see Mr. Tapia at Tr. 6515). This
~ 15' Procedure precluded the bending moments which would have resulted if 16 the t'est represented the actual field condition (the Applicants use of
.17 'a flat plate instead of the tube was not in compliance with 10 CFR Part 18 50, Appendix B, Criterion XI which (a) requires that tests be performed 19 prior.to installation and (b) that testing be representative and follow 20, a' written procedure which incorporates requirements and limitations).
21; After the_ completion of the' tests, no one could evaluate with any L22l finality the results as related to normal, upset or emergency 23 e nditions (see NRC Staff Witness Dr. Chen at Tr. 6542-6552).
_24. In effect, the-NRC Staff and the Applicants have not conceded that 25 there is any problem'with the supports at Comanche Peak and have taken 6
- . . , - . , . - . . . . . , , ..c ,..,-.~m..,--~~~.,.,-,-r- ,-----.,r-+--, ,----
j
.c .*:- ,
- .. -n.
I 1
1 1
6 1~ the position that all of the all'egations are without-merit.
jb The . facts are' that history and Applicants' activities have proved 3: beyond doubt that of the original 58 supports mentioned in CASE Exhibit l 14 - 669B, a large percentage have been subjected to major rework to bring
. 51 .them up to par as required by the law, the codes, basic fundamentals, 16- and common sense.
7- In enumerating the supports mentioned above, I am only addressing 8 those supports which I can document as having been changed, not the-9- others which I know have been revised but cannot cite why I know.
- 10 First, by Applicants' own testimony, all unstable supports which
- ll: . utilized a U-bolt in lieu of clamp were to be cinched up to prevent
-12 clamp rotation; this was also Board Chairman Bloch's understanding (see 13- Tr. 6646; see also Staff' Dr. Chen at Tr. 6718, where he states that
- 14. individually'these supports were unstable). Of the items contained in 15; CASE Exhibit 669B, the following fall into this category:
- 11 6 :(1) -MS-1-003-006-S72R (in design, this is similar for MS-1-001-117 006-S72R; MS-1-002-006-S72R; and MS-1-004-006-S72R);
-18 (2) _MS-1-001-005-S72R (see Cygna Report, Volume III, PS-071);
19 (3) SI-1-031-704-A32R (2 supports);
20 (4) CC-1-028-701-A33R.
'21, The use of U-bolts in this manner is a unique design feature, and 122_. is not in compliance with 10 CFR 50.34 (a)(8). In many cases, as may 23 ._ be noted in MSS'SP-58 and SP-69, the procedures used at Comanche Peak 24 -are not an industry practice.
-25 The Applicants also confirmed that the box beams would be 7
.=.. . _ . - . - _ - - . . _ - - - . - . - . . . - - , ..
cs .; e
- ,~ . .s -
- ] corrected to prevent' instability (see' Chairman Bloch, Tr. 6646, and Dr.
12- -chen, Tr. 6718). The following box beams required modification:
- 3- (1) cc-1-028-034-S33R (see below-for. change required for W6x12 4: . beam base attachment);
5 (2) cc-1-028-039-S33R (see below for change required for W6x12 6 beam base attachment);
7 -(3) cc-1-020-001-A33K; 78 (4) CC-1-159-010-S43R; 9 (5)- cT-1-008-007-S22K; 10 (6) cT-1-008-008-S22K.
11- The above eleven supports were part of the generic (to comanche 12 Peak) stability problem alleged in my August 1982 deposition 13 ~(Deposition / Testimony of CASE Witness Jack Doyle, CASE Exhibit 669 and 14- 669A, admitted at Tr. 3630).
- 15 ~As relates to the instability problem, Applicants stated that they
'l6; .had been aware of this since March'1981 (see Applicants' Witness Mr.
- 17 -Finneran, Tr. 4889/16-20,-Tr. 4890/1-4, and Tr. 4895; see also NRC SIT 18 Report,-Staff Exhibit 207, Page 4(h)). In reference to the awareness 19 .by Applicants of a problem with stability, they offer three memos as 20 Proof of this fact (see Affidavit of John C. Finneran, Jr. Regarding
?21f . Stability of Pipe Supports and' Piping Systems at pages 10 and 11, and 22 Attachments thereto, attached to Applicants' Motion for Summary 23: Disposition Regarding Stability of Pipe Supports):
,24 The first by cus Abele (ITT onsite lead engineer), dated 5/22/81 25 (Attachment A-1 to Finneran Affidavit, page 2) states:
8-
,, ---..,e - - - . -,--,-,n. . . , . , ., .,-.n, . - - . . . , . , - - - . -,, ,, ev-, . . . - -. ., - - - -e.--, ,- - --w.*, . . .,.
M . 3.m )
p i,; 1. ,
. \
'j E" Supports that utilize single or double trapeze assemblies (i.e. ~
' ~
struts or snubbers) and a pipe attachment consisting of a steel 2~
box frame to-restraint vertical and side loads. Technical
. Services require concrete guidelines and evaluation techniques to J etermine t e sta ility. f using box frames as pipe attachments
~3 -
. due to the-physical tendency of the box frame to rotate the pipe.-
Especially for vertical restraints. The governing variables are
~4 numerous: - pipe size, movement, existing clearance between pipe
.. and frame, single vs. double trapeze,-struts installed on skew,
"* location of adjacent supports, vertical.vs.' horizontal restraints, and'a special but common case of single or double struts and
'6 . snubbers from beneath pipe. Samples are enclosed. (Mark No.'s
'p, - CC-1-159-010-S43R1, CC-1-028-039-533R). Technical Services will hold all supports'using-box structures as pipe attachments until j written procedure.or approval from Providence is received."
The samples referenced in the third from last sentence in the 3-men were n t attached to' Applicants' Attachment A-1.
10 The second memo, by Gus Abele (Attachment A-2.to Finneran g Affidavit), dated 10/12/81, #RI-9, picks up most of the same wording as was used in Attachment A-1, and states: "(see attachments)" (emphasis ,
. : added). However, only one page was attached to Attachment A-2.
[15" e t ir sem , ate 2/82 (Attachment A-3 to Fineran Affidavit) was an answer to the previous request for information', and' I6 g was authored by Ron Wisniewski (ITT Grinnell horne office) to Gus Abele.
This document not only does not concur with the conclusions of Gus
- Abele but suggests the problem be discussed with Mr. Ed Eramian. This document was, as far as I an aware, the last discussion on the subject g; of instability at Comanche Peak until I revived the problem in about
- c ,,2 March of 1982. (It should be noted that CASE requested "any answer g that was made, any backup notes, any handwritten notes having anything to do with this particular matter in answer to attachment A3, and any subsequent information that went back and forth regarding this matter."
9 4 y- , - , , - - , - - - , + , , , .-.-.--,...w.,-.%, , yc-,..-- ,,.-,..,v,.,,w w-, ,.,ymm,,,-m...-.y.-,,-,v.,,-m , ,_w,--
4 ;a , '-
.i ,
,5 y j --See transcript.of 8/6/84 Applicants / Staff / CASE telephone conference
' call, at Tr. 56 (see also discussion at Tr. 49, line 24, through Tr.
- 3 .
83,;line 23; especially Tr. 51 through 65.) Applicants refused to j ' provide this information (Tr. 56-60, and Tr. 86, line 14, through 87, g lineL12).- But it.is my understanding from CASE President Juanita Ellis b that.on 8/22/84,.the Board Chairman ordered that Applicants provide the
'6; 7, .information requested. However, when she inquired on 9/28/84, g Applicants' counsel William Horin advised that there are no additional documents; Mrs. Ellis asked that he confirm this in writing.
~
. g'-
g For that matter, as recently as the September'1982 hearings, g ' Applicants' witness Mr. Finneran was stating that the subject of instability was still being studied at Comanche Peak and was an g indeterminate subject (Tr. 5268/10-13).
g; In relation to the use of U-bolts (such less cinched-up U-bolts),
g and still.on the subject of instability, the Applicants' procedures are in violation of standard industry practice for'a garbage processing
~ plant, much less for a nuclear plant, and therefore are in violation of 10 CFR Part 50, Appendix A, Criterion 1, as regards the adequacy qg .section-for nonconformance with standard practice.
The chain of codes and laws is as follows:
(1) In accordance with the provisions of ASME Section III, NA-3250 (PROVISION OF DESIGN SPECIFICATIONS), Applicants prepared a technical specification for nuclear safety-related equipment;
() The title of the document generated in compliance with L25 10
_e-- ,
e a - ~ , - , - ,-_
r
.~ .-
'l ASME Section III,'NA-3250, is NUCLEAR SAFETY CLASS PIPE 2 . HANGERS AND SUPPORTS, SPECIFICATION 2323-MS-46A; 3 (3)' ASME Section III, NA-3320(b) regarding Manufacturer's
-- 4 responsibilities, is directed to ASME NA-3340; and NA-5 3340, RESPONSIBILITY FOR COMPLIANCE WITH THIS SECTION, 6- states:
7 "The Manufacturer who completes or substantially completes any component, appurtenance, core support 8' structure, or component support required to be in complianc,e with this Eection has the responsibility 9 for the structural integrity using the Design Specifications as a basis.of design . . . "
10 (4) In the Design Specification (technical specification) 11 2323-MS-46A, page 3-15, Section 3.3, CODES AND 12 STANDARDS, the following is included:
13
" Design, fabrication, materials, certification,
.14 - code stamping, and testing requirements included in this specification shall be in accordance with the 15 edition and addenda of the follouing codes,
, legislation, regulations, and standards, i~n effect 16 on July 28, 1975, unless otherwise specified below-or authorized by the owner."
17:
On page 3-20 of the same document, the following arr 18 listed as codes with which Applicants shall comply:
19-
"g. ' Manufacturers' Standardization Society (MSS)
'20 a- "(1) MSS SP-58, Pipe Hangers and Support-Materials and 21 Design 22 "(2) MSS SP-69, Pipe Hangers and Supports-Selection and Application"
^23-(5) At the Foreword of MSS SP-69, the following is quoted:
24 "The requirements of this standard were developed 25 by a cooperative effort of representatives of pipe 11 ,
^
J. , : .; 'o c.s .-
hanger manufacturers. They are based on the best
-hl - practice current'at this time and on proven results of the research and experience of this industry."
I(6) At SP-69, Section 6, Selection of Hangers and Supports 3'
- - for Pipe Movement:
4 6.2. The last sentence states: "For-piping
'5' supported from below, bases, brackets, or
,atructural cross members'should be used."
.6.3. -The last sentence states: "Where there is
- 7. horizontal movement . ... For. piping supported from below, slides, rollers, or roller carriages should 8 be used."
9 (7) Table 1, page 3, of SP-69 does not recommend the use of U-bolts (type 24 is the designation for U-bolt in this
.10) material) for insulated lines classified as Hot A-1. And Hot A-1 is defined under 2.1 (Hot Systems) as being 120
~
'll. degrees F. to 450 degrees F. For Hot A-2 line, type 24 supports are not recommended whether the line is 12 insulated or not.
13- (8) SP-69 states at 12.2, " Riser clamps (Type 42) shall have a positive means of engagement between the pipe and the I4 clamp." (By inference, this would include type 24 clamps, which are U-bolts.)
(9) The practices used in the design and construction at 16: Comanche Peak proceed under a false premise; that is, one constructs the facility, then justifies the 17 construction as-opposed to justifying the procedure and then constructing.
(10) Comanche Peak practices evade the provisions of 10 CFR 50,-Appendix A, Criterion 1; (11) Also, Comanche Peak practices evade the provisions of 10 CFR 50.55a(s)('1), which states:
" Structures, systems, and components shall be 23 designed..' fabricated, erected, constructed, tested, and inspected to quality standards commensurate
_24 with the importance of the safety function to be performed." (Emphasis added.)
12
7 w .. ,; : ,;
I (12) Comanche Peak' practices evade the provisions of 10 CFR 2, - 50.34(a), which states that the minimum information to
^
3 be included in the Applicants' Preliminary Safety 4- ' Analysis Report (PSAR) was to have included:
5 -! . "(2) A summary description and~ discussion of the facility, with apecial attention to design and c6_ operating character' tics, unusual or novel design features, .-- principal safety 7 considerations." (Emphases-added.)
- 8 "(8) An identification of those structures, systems, or components . . . which require research 9- and development to confirm the adequacy of their design; and identification and description of the 10 research and development program which will be conducted to resolve any safety questions l _1 associated with such structures, systems or components; and a schedule of the research and
.12 development program showing that such safety questions will be resolved at or before the latest 13 date stated in:the application for completion of construction of the facility." (Emphases added.)
"(9) The technical qualifications of the applicant
-15 to engage in the proposed activities in accordance
,- with the regulations in this chapter." (Emphasis 16 added.)
17 10 CFR 50.34 (b) discusses the Final Safety Analysis
- 18 Report (FSAR) (Applicants' Exhibit 3), which is supposed 19' to include "information that describes.the facility, 20 .
presents the design bases and the limits on its 21; operation, and presents a safety analysis of the 22' structures, systems, and components and of the facility 23 as a whole . . . " and shall include information 24 updating that which was presented at the time of the 25- construction permit's issuance.
s 13 -
-r- w , 4, ,w., - , - . ,,r- .- .w-, ,.,-y-.-ve ..~%, --r--w.ye,w-y + ,,,..em y ..v-.,,w,.7,,--.,w-,--,,,yv---,,w.*..,yy--~~--,--
r
'c .
3 .
, '4; i.; -
l i
1 l i while on ehk subject of' instability, it appears that during the :
2; - May 1983 hearings, the testimony of the NRC Staff was either blind or-
- 3. - somewhat less than candid. As an afterthought, the Applicants, by_
1 4 -
. their silence in : the matter outlined below,' aust also be suspect.' ,
-5 Dr. Chen, on the matter of instability, stated that he looked at
'6' the exhibits and identified about 30 supports which were identified as l 7- - " suffering from this disease" (of instability), that he looked at some I (8 piping _ systems to see if any of those supports were all grouped
-. 9 together, and that he didn't see any cases of that in the piping stress il0 analysis which he looked at (Tr.-6718-6719). Mr. Walsh asked Dr. Chen 11 . whether there might be three or four unstable supports which were
.12 . adjacent-to each other (Tr. 6722-6727).= After some discussion, Dr. <
13 ;
Chen stated "I believe I saw one piping run where the supports had
. l'4 about three intervening supports" and that was the closest he found Ll5 them (Tr. 6727-6728). Dr. Chen elaborated on this and stated ,that the
[ .165, piping was not in the same direction in both locations, and therefore-b 17: the testability would not have been additive (Tr. 6728).
- s .
- 18
- At that point, we had nol documentation to prove otherwise, except 19 that I had seen the problem on four steam. lines in Unit 1: Ms-1-001; I
20- Ms-1-002; Ms-1-003; and Ms-1-004; and it may also exist on Ms-2-001, 21 .etc., for Unit 2. The statements made by Dr. Chen were therefore
~22- allowed to stand.
' 2 3.. However, the fact is that the main steam lines have five unstable 4
24: supports in a row, and in fact, the sixth support is unstable since it 25' is double pin-ended (clamp and verut) but all of the other supports i e
, 14 .
)
-,,.,_-m., ,,_.,-,y,, m~.~m,.- m.-.. ._.,.,.,--m~%,.-,-~- _,,m. ..-w
e-
.' e j
., e.
,j; cannot. assist.in giving the strut on the sixth support stability. But 2I neglecting the sixth support, there are still five unstable supports
- g-
.In a row at the critical main steam isolation valve and pressure relief
' 4 .-
valves. These five supports are as follows, as documented in the 5
referenced Portion of the Cygna Report, Phase 3, Volume 3:
- 6 MS-1-001-002-S72R -- instability not fixed (PS-068) 7 MS-1-001-003-S72R -- instability repaired (PS-069) g MS-1-001-004-S72R - instability repaired (PS-070) 9 MS-1-001-005-S72R - instability repaired (PS-071) 10 MS-1-001-006-S72R -- initial instability repaired (PS-072) jj .Beyond this, Applicants keep on insisting that there were less than two dozen unstable supports (and they are still reluctant to admit
'12 13 that there are any unstable supports - see 8/6/84 Applicants / Staff /
j4 CASE telephone conference call att Tr. 50, lines 3-4, 7, 13; Tr. 51, line 7; Tr. 52, lines 11-12; Tr. 53, lines 5-6; Tr. 60, line 21; Tr.
15 61, lines 21-22; Tr. 62, lines 1, 22-25; Tr. 63, lines 12-13; Tr. 65, 16 j7 lines 23-24; and Tr. 78, line 2). However, Applicants' position falls
)g apart since, with only the supports that I have seen and not including 19 Unit 2, I have listed 11 from my deposition, 20 from the Cygna Report 20 minus 2 duplicates for the main steam, plus 1 from the Phase 1 Cygna 21 Report (CASE Exhibit 928, formerly Exhibit 891, bound in following Tr.
22 9825) -- all of which equals an absolute minimum of 30 supports known t have been repaired to' correct stability problems.
23 Another area of proof that the initial allegations contained in 24 CASE Exhibit 669 (Deposition / Testimony of CASE Witness Jack Doyle, 25 15
\
y 1 admitted into evidence at Tr'. 36'30) were accurate, involves the repair 2 to supports to bring them into compliance with the fundamentals of 3 engineering and therefore the laws and the codes. Of these supports
'4 known to contain major rework and re-engineering, the following are 5 examples:
6 (1) 2323-SI-0538-07.(see CASE Exhibit 669B, items 90-9S) -- major
'7 member added, changed to truss support (see CASE Witness 8 *-
Doyle at Tr. 6227); l 9- (2) CC-1-028-034-S33R and CC-1-028-039-S33R - 6" wide flange 10 member; web failed; needed redesign (see 12/13/83 Affidavit 11 of NRC Staff Witness Dr. Rajan, attached to NRC Staff's 12 12/13/83 Motion to Reopen Record to Admit the Affidavit of 13 Dr. Jai Raj N. Rajan);
14 (3) CC-2-007-025-A43R - support changed to eliminate two-way 15 load (see Applicants' 5/23/84 Motion for Summary Disposition ,
16 on U-Bolts Acting As Two-Way Constraints, Table 3, Problem 17- AB-1-62E, Second page 2);
18 (4) CC-1-107-008-E23R -- major change (see 10/14/83 Affidavit of
-19 W. Paul Chen on Open Items Relating to Walsh/Doyle Concerns, 20- page 26);
21 (5) CC-1-116-038-F43R - 1/4 inch wall of 8-inch tube failed; 22 rebuilt (see Applicants'. Witness Finneran at Tr. 4793/4-8, 23 also CASE's 8/23/83 Proposed Findings of Fact and Conclusions 24- of Law (Walsh/Doyle Allegations) at pages XIII - 3 through -
25 6);
16 .
~. y
.- :o
~l (6) CS-1-239-007-A42R -- c' hanged by CMC 58004 (see SIT Report, 2, NRC Staff Exhibit 207,-page-42);
3 (7) CC-2-008-709-A43K -- the SIT uncovered a numerical error in 4, Applicants' preliminary calculations which resulted in 5 underestimation of bending stress and could result in an 6 overstress condition; Applicants indicated their subsequent 7 _ review identified an overstress condition which would be 8 _ rectified (SIT Report, pages 41-42).
9 'The above only includes those structural changes of which I have 10 personal knowledge.' The Applicants and SIT argue that the problems are
'll the result of "somewhat knowledgeable" engineers, but this would be a 12 violation of the skill provision of 10 CFR Part 50, Appendix B, 13 ~ Criterion II. Beyond this, the NRC Staff and the Applicants were well 14 aware of the lack of skill of the Comanche Peak staff when in 1981 they.
-15 had to commence a redesign of massive proportions, particularly at
~
16 . _ elevations 790 and 810, because the supports as designed would not fit
- 17 in the areas assigned (see NRC Staff Witness Taylor at Tr. 6666).
18 Further, the somewhat knowledgeable problem was no secret; everyone
'19 knows about it (see: for the NRC Staff: Taylor at Tr. 6403-6405 and
- 20 6665; Tapia at Tr. 6669; and for the Applicants: Finneran at Tr. 4955-
.21 4965; Reedy at Tr. 7164; and Vega at Tr. 7166).
!22 A second major excuse used by Applicants and NRC Staff was that
'23; every allegation that couldn't be buried was claimed to have been known 24 ._ earlier by the Applicants, but with no documentation to back up such a 25 statement. Finally, the NRC stated otherwise in several stipulations 17
ly .i c
- j. regarding'who first noted the problems (CASE Exhibit 848, admitted at 2
1 Tr. 8352). ~As just one: example:' At items 52 and 49, it was basically 13 8ti Pulated that Applicants did not recognize the problem or attempt to 4 deal with it. prior to my bringing it up, and that they had no mechanism-
-5 in P lace to keep the problem from recurring again and again.
L6 Of the original supports contained in my August 1982 allegations, 7 ' eighteen have been revised (and more than likely many more have also).
' 8l This represents 31% of the sample. one could add the two supports at
'g CASE Exhibit 669B, itene 2D and 2E, CT-1-137-701-S22R and CT-1-137-702-Il0 '
S22R, which were changed shortly before the hearings and were therefore
.jj-
. never'an issue. In reference to items 2D and 2E, I complained that 12' there were inadequate induced forces into the run pipe and were in
-13 Vi lati n of Appendix XI (see CASE Exhibit 669, pages 83/19-22, 84/10-
.j4- 24, 85/18-24, 101/16-25, 314/13-314/22).
If items 2D and 2E are 15' included a8 Part of the original problems which resulted in changes, 16 .the number of problem supports that I correctly identified is 20 out of.
j7 58 or 35 per cent. (See CASE Exhibits 669, 669A, and 669B, Doyle 18 Deposition / Testimony and Attachments, accepted into evidence at Tr.
.j g ; 3630.).
~20. APPlicanta have another loophole for problems which prove 21 embarrassing and that is to claim that since the particular support has 22 not been final vendor certified any allegation carries no weight. The 23 only time a Problem can be laid at the doorstep of Applicants is when 24 the problem can be showr. to exist after the nine iterative steps which 25 . lead t final vend r certificat.1 n (see SIT Report, NRC Staff Exhibit 18
,. 1 3 207, pages 14-16). Not only do Applicants stand by this postulate, but j
- 2. the concept is contagious. The NRC Staff keeps inserting this caviat
'for each problem which cannot be brushed aside by claiming that "it's 3
4 not vendor certified" (see 12/13/83 Affidavit of NRC Staff Witness Dr.
5 Rajan, attached t NRC Staff's 12/13/83 Motion to Reopen Record to Admit the Affidavit of Dr. Jai Raj N. Rajan, page 2; see also, SIT 6
7 Report, page 41, lines 10 and 15, for example).
.'g The above~caviat "not vendor certified" cannot stand the test of logic, since using the caviat is itself an admission of.non-compliance with 10 CFR Part 50, Appendix B, Criterion VII, as relates to the 10 provision for adequacy of the document prior to release for use. And
))-
ne must keep in mind that these supports were fabricated, erected, 12 inspected, turned over to the utility, and many were in place long j3 g- before hydro testing.
I believe that this original segment of the allegations standing 15 i al ne (without the material which follows) is sufficient to prove two 16 points:- (1) The Applicants at best have suffered a massive breakdown 37
- )g in their QA/QC program, if such a program ever existed for the engineering discipline; (2) The NRC Staff has failed in their mission 19-to protect the health and safety of the public and has in effect become g
dependent on the' Applicants to justify the Staf f's position as the 21 industry regulator.
22 Bey nd this, it is apparent that the Applicants have reversed the 23 role of engineering from that of justifying structures to be used in g
Construction to justifying structures whlCh have been designed by 19
r;
- s; f. ' I + ' _ * :
}
-e *.
i Oli somewhat knowledgeable engineers. -The NRC Staff has been only too 2? eager to concur in the acceptance of this unique but dangerous 3 deviation from standard engineering fundamentals and sequence.
!4 ' 05.: Does this conclude your statement of status as relates to the original 5 . allegations?-
- 6! -A5. No! There are major areas of problems generic to Comanche Peak which 7 still' must be addressed. For example:
8- (1) Richmond bolts of A307 steel in bending; 9' (2) oversize holes for non-friction joints; fl0 (3) Coupling'and prying (Mx and Mz moments) Richmond / tube steel
~
~ll structures; l: 12 (4) Actual vs. generic stiffness; Il3 -(5) Couplingofmoments[fordoubleaxialrestraints; 114- (6) U-bolts which are actually two-way restraints but designed as
- 15) one-way restraints; abl6 (7) Cinching and thermal problems with U-bolts; 17: (8) _ Thermal problems with box frames with zero (0) clearance; '
18- (9) Thermal locking at anchors;
~19 (10) Problems with moment restraints and upper lateral restraint.
20 This is only to mention a part of a larger list of problems... some 21 of these will be discussed below.
i22. One must keep in mind that when 35 per cent of the allegations are 23 proven correct, it is not necessary to prove that esary allegation is 24 . legally accurate in order to show that a nuclear power plant is plagued i
25 .with problees which collectively indicate it is a dangerous structure.
i.
4 20 7 y w ' F= W y- yyy- m =%-eme,ggw-.,-+e.p-. -p g.p-Wy www-di,-.4%-=%,%9M + ,- gp ei s -w g g. g .y y.g. y p g -W _ M ew ggggg>+p-tw_ ,y y g ' rw
g . - - - .- . . -~ _ . --
6.: ' I= .
~
1 06.. What points do you wish to make in reference to the Phase 1 portion of
~2 . ' the review by Cygna Energy-Services?
i3 ' A6. - First -1 must restate' ay position on the credibility of Cygna as an
~
,4' ! independent reviewer. No person or organization can be effectively 5 independent in a review process if they are allowed to evaluate their
-6 own findings in a process'which is paid for by the reviewed 7-
~
organization -- that.is, not if they aver intend to perform another
!8 review under the same circumstances for another similar. organization.
!9 one is'only employed by private corporations as long as one satisfies
'10 their requirements; stated or unstated, 11 In the case of Cygna Energy Services, for Phase 1 of their review,
-12 the diligence of their gathering of raw data was somewhat less than 13' desired, as was pointed out in the February 1984 hearings and discussed 14 'in the material which appears below.
l15 ~ Among the basic problems which eluded Cygna, one of the more
- lG, critical problems involves the use of incorrect formulas for analysis.
- 17 .The fact that supports may later be correctly analyzed and fall within
~
-18 design 11mits is of no'value as an argument for tolerating such fl9 incompetence, since the fact that it occurred and passed through as 20 many as 10 checks (the original check plus 9 levels for vendor
.21 certification) proves that incompetence was rampant. As a result of
-22 this observation, no engineered element of Comanche Peak can be 23 considered safe until each element is rechecked by truly indsjendent 24 assessors.
L25 -Among the items not included in the review by Cygna, but which 21
. .* a s, . .. -
) were found by ne, brought out under cross-examination or delivered by B ard Order, the following are offered:
2 1 ' Stresses'and Deflections at Local Areas:
g In reference to the Applicants' ability to catch any. errors in
.' engineering during the Vendor Certification program, examination of the results of reviewing only 9 calculations during the evaluation of the 7
Cygna Phases 1 and 2 material will indicate the fallacy of that
-g thour's.
lor *X88ple, for calulation No. SI-1-079-001-S42S (CASE Exhibit 9
110 -
930), the calculation of a W4x13 which is less than 2" long is
. accomplished by flexibility analysis while in reality, with this beam
)) -
flexibility analysis is impossible. In addition, in using flexibility 12.
analysis, the stress in the.W4 is shown t be 1166 psi and for shear, 13
'3,000 psi. . The major stress is induced by the flange bending induced p
by the bracket and is in excess f 18,000 psi. This face was conceded 15 and in fact determined by Cygna when it was pointed out to them.
16 j7 The same impossible analytical process is also used for the 6" I 88 8 888 (888 8upport No. RH-1-024-011-S22A, CASE Exhibit 936), and 18 the 3-inch long, 6-inch diameter anchor (see support No. SI-1-030-003-39 S32A, CASE Exhibit 937).
20 On calculation RH-1-064-011-S22R (CASE Exhibit 934), on sheet 4 of 21 4 f r the welding calculation, only one component of torsion is 22 23; included; beyond this, the bending stress about the y axis was ignored.
On this same calculation, the bending stress inducing moments into the 24 anchor bolts was neglected (see sheet 3 of 4).
25-22
f .. .
- j For calculation RH-1-064-010-S22R (CASE Exhibit 931), sheet 4 of 8 2 indicates the strut angle to be 4.99 degrees, thus qualifying as less 3 than 5 degrees; however, a correct calculation would have shown that, 4 according to the analysis for the C to C dimension of 45.8125 with an 5
4" ffset, the c rrect angle is 5.01 degrees.
6 For support No. SI-1-325-002-S32R (CASE Exhibit 928), the clip 7 angle (item 15) which was installed t.o assure stability, fails. While 8 Cygna failed to mention this initially, when compelled to supply an 9 answer, they finally stated that they were aware of this problem. [1/
10 Fr a the above, the value of the Vendor Certified Drawing (VCD) jj program appears to offer little hope of correcting any problem and may 12 in act be introducing new probler.s.
- 13 Thi8 18 t 118t nly Partially the items involved in this problem.
j4 In reference to item 1 above, the eniculations were not " correctly 15 translated" as required by 10 CFR Part 50, Appendix B, Criterion III, fir 8C Paragraph.
16 j7 The previous (item 1) shows that of 9 vendor certified equations 18 which I reviewed, 6 were incorrectly done and one item completely jg_ failed.
20 In additi n t these six items, the course of cross-examinat. ion of 21 Cygna in reference to generic problems offers additional evidence that 22 the P stulate that items may be written ff by engineering judgement is not only fundamentally wrong, it is dangerous.
Ilf CASE Exhibits 927, 928, 930, 931, 934, and 936 have previously been 24 ~
supplied to the Board and all parties. Please advise if we need to 8end c pies.
25 23 m
P"~ . -
I 2. Double Axial Restraints (snubbers or struts):
2 (1) Either of the two supports (snubbers or struts), when input as a 3 moment' restraining system, will experience substantially higher 4 loads than are evident when the node point is input into the pipe 5 stress model as a single unit acting along the centerline of pipe 6- with the total load later divided equally to the two supports.
~
7 This is a recognized problem in industries associated with piping 8 systems; see, for example, ANSI B31.1, paragraph 121.3.1 (a) an'd 9 (b). The facts on this item are that after two years of resting 10 secure in the' knowledge that the Applicar.ts were convinced of the 11 problems in this procedure and had instituted corrective measures, 12 we learned that their knowledge only extended to several systems, 13 and other systems were still being accepted which were not being 14 introduced into the computer as two-way restraints (force and 15 moment).
16 (2) When pressed on this issue, Cygna performed analyses which 17 revealed that for the supports analyzed, the following was notedt 18 (a) When input as a single entity, the load on either snubber of 19 the system analyzed was 1328 lbs., but when both snubbers 20 were input into the computer excentric to the center 11tne of 21 pipe, the division of loading was no. longer 50/50.
22 (b) The most loaded snubber had a load of 3933 lbs., almost three 23 times the previous condition. This would also have an effect 24 on the base plate and anchor bolts. (See Tr. 12,851.)
25 (c) The facts involved in noncompliance in this case are as 24
~*
..
- I
- 1 follows:
l 2 (1) Since the configuration of the support is such that a de l 3 facto moment restraint exists, the manufacturer is 4 required by the ASME code to ensure the integrity of the 5 support. (See ASME Section III, NA-3340.)
6- (ii) Beyond this, the provisions of 10 CFR Part 50, Appendix 7 A, Criterion 1, require that procedures not be 8 restricted by the letter of the code nor the fact that 9 direct prohibitions do not exist within the code.
10 In reference to item 2, Cygns witness Ms. Nancy Williams, under 11 cross-examination, stated that Cygna did find that there was retation 12- (Tr. 12769). She further stated: " . . . yes, you will get a 13 rotational restraint there, and you will probably come up with a 14 slightly different distribution of loads on the struts or snubbers" 15 (Tr. 12773). And Ms. Williams attempted to justify the engineering 16 judgement used to assume that there is no significant effect due to 17 neglecting the restraint of rotation by referring to the fact that the la " pipe stresses go down" and one must balance the two (Tr. 12774).
19 But Ms. Williams is misleading the Board, since the stress levels 20 in the pipe and the loads on the struts (or stresses) are not equatable 21 but actually are totally independent. For example, if a procedure were 22 used that overloaded building columns but reduced the loading on beams 23 and girders, the building could still collapse due to that overload.
24 So the balancing of loading consideration is without merit.
25 Ms. Williams admitted that the restraint effect was taken into 25
~
N ..~
s:
l l
Laccount.on some systems at Comanche Peak (she believed), "perhaps due to
.l
- . the importance of that system or a particular configuration associated D 2 with given systems, purely at the judgment of the analyst" (Tr. .
"3 12,775).- Es. Williams stated that l'f two axial snubbers were modelled 4
as such (that is,' two restraint points), then anchor bolts could be 5
underdesigned (Tr. 13,040).. But most important, Ms. Williams confirmed 6
that Applicants modelled two snubbers as.if they were one snubber
-7 acting on the centerline of pipe (Tr. 12,770).
8 It can only be stated that in reference to item 2 above, the
- 9 procedure used does not comply with the " correctly translated"
- '10 provision of 10 CFR Part 50, Appendix B, Criterion III, first
.11 1'
paragraph. Nor is this procedure in compliance with 10 CFR Part 50,
- I.2
' Appendix A, Criterion 1, in reference to adequacy, since the procedure 13 does not comply with the provisions of ANSI B31.1 121.3.1(c), the code i 14 for non-nuclear pressure piping which states in parts 15' e .
"In addition to.the provisions of (b) above clamps to support 16 vertical lines should be' designed to support the total load on
.either arm in the event the load shifts due to pipe'and/or hanger
, 17. novement."
18 3,2 Failure to include masses on the run pipe as a result of i 19 supportins suspended et asnes of the pio supports:
^
20 (1) Box frames, U-bolt structural clamping devices, and other hardware supported by the pipe must be included in as dead weight in the 22 pipe stress analysis.
23 (2) The SIT found-that an assessment of this contribution to the
, 24 piping loads is made on a case-by-case basis . . . if the o
26 E
a
- 2. ~
.I . contribution is considered significant. (And the SIT considers 2 several hundred ib. loads suspended off the pipe not to be "3 significant.) (See SIT Report,' NRC Staff Exhibit 207, page 35.)
4 (3)' After Cygna Energy Services was challeuged on this point, a finite i .5 pipe stress analysis was executed with the following results (see
.;6 Tr. 12,571): The difference in the seismic load on the support 7- with the largest increase due to including masses supported by the 8: pipe was 24 per cent, plus other load increases of indeterminate
=9- magnitude. And this'is for a system of.16 supports with a total 10 mass for all supports of only 750 lbs., a maximum single support 11 suspended load of 120 lbs. and an average load per support point 12 of less than 50 lbs.
-13 (4) The facts are that the failure to include masses on the run pipe 141 -as a result of supporting suspended elements of the pife supports
.15 represents a gross concern for the design of Comanche Peak and is
- 16 in violation of the codes and laws to which Applicants are
. 17- committed.
~
18 In reference to item 3 above, the chain of codes and laws is
'l9' as follows:
12 0 (a)' In accordance with the provisions of ASME Section III, 1211 NA-3250 (PROVISION OF DESIGN SPECIFICATIONS), Applicants 22: prepared a technical specification for nuclear safety-23 related equipment; 24 (b) The title of the document generated in compliance with
'25 ASME Section III, NA-3250, is NUCLEAR SAFETY CLASS PIPE 27
l, :i a r . .. .
II L
j HANGERS AND SUPPORTS, SPECIFICATION 2323-MS-46A; 2 (c) ASME Section III, NA-3320(b) regarding Manufacturer's 3- re8ponsibilities, is directed to ASME NA-3340; and NA-
~
4- 3340, RESPONSIBILITY FOR COMPLIANCE WITH THIS SECTION, 5
states:
6 "The Manufacturer who completes or substantially completes any component, appurtenance, core support 7 structure, or component support required to be in como11ance with this Section has the responsibility for g the structural integrity using the Design Specifications as a basis of design . . ."
9 (5) In the Design Specification (technical specification) 2323-10 MS-46A, page 3-15, Section 3.3, CODES AND STANDARDS, the following 11 is included:
12
" Design, fabrication, materials, certification, code
-13 stamping, and testing requirements included in this specification shall be in accordance with the edition and 14 addenda of the following codes, legislation, regulations, and standards, in effect on July 28, 1975, unless otherwise 15 specified below or authorized by the owner."
16 On page 3-20 of the same document, the following are listed as j7 codes with which Applicants shall comply jg "g. Manufacturers' Standardizatior Society (MSS) 19 "(1) MSS SP-58, Pipe Hangers and Support-Materials and Design 20 "(2) MSS SP-69, Pipe Hangers and Supports-Selection and Application" 21 (6) At the Foreword of MSS SP-69, the following is quoted:
~
22 .
"The requirements of this standard were developed by a 23 C perative effort of representatives of pipe hanger manufacturers. They are based on the best practice current 24 at this time and on proven results of the research and experience of this industry."
25 28
..- w, l- (7) SP-69 3.4 for load calculations "where required by specifications, 2 calculations _shall give consideration to the following: ...
~3- weights of pipe, valves,-fittings, insulating materials, suspended 4 hanser components, and normal fluid contents." (Emphasis added.)
5 secause this suspended material reacts on the pipe, it must 16 be included in the pipe stress analysis, particularly in our case
~7 since the dynamic out-of-plane loads up and down stream can be 8 significant.
'9
.10 h cinched-Up U-Boles:
11 The extent of the problem may be noted from the following:
112 (a) In almost two years of attempts, the NRC Staff and the 13 Applicants have failed to show that cinching is no problem.
14- (b) cygna Energy Services, hired.as a consultant by the 15- Applicants for an independent assessment program, was 16 challenged on their uncontesting acceptance of this method of 17: clamping. They spent over two months in analyzing the
-18 problem using the most sophisticated and accurate means 19 currently available.
20 The results were that the stresses in the U-bolt and the 21 pipe under 5 foot pounds of torque exceeded the allowables of 22 the materials and with' higher torques would. exceed the yield
'23 of the materials (at 80 foot Ibs., in fact, in one analysis 24 the pipe stress was 80,000 lbs. per square inch) (see Tr.
25' 12,331 through 12,335). (If this were compared to the actual 29
.s___
g .- .
+ -.
1 l
allowable of 37,500 psi for pipe based on Sh plus Sa, the
.2 pipe stress far exceeds the allowables.)
Cygna's Dr. Bjorkman stated that the methods used in the
'4 modelling were not precise; therefore, the loads indicated were somewhat conservative (however, the model did not 6:
contain the mechanical loads). Dr. Bjorkman referred to this analysis as a scoping' study and stated that an exact solution 0
would require extensive effort (Tr. 12,337).
I In reference to Item 4 above, several points must be made. Cygna 10 Energy Services, which was responsible,for the review of systems at U Comanche Peaki either deceived the Board or totally lacked knowledge of 12 ,,,,,,1 1,po,c,,e g, ego,, ,,1,ggy, g,p , gin,,g ... sections which, while not adhered to at Comanche Peak, are required for even the most
~I4 insignificant facility. .See Tr. 12,369, where Ms. Williams states that 15 they never considered MSS SP-69, and Tr. 12,371-12,372, where Ms.
1 16' IWilliams states that she does not know about the MSS organization.
l7 Beyond this, while Applicants, NRC Staff, and Applicants' agents 18 continually refer to industry practice as the caviat for not performing II analysis or accepting procedures on the basis of judgement, Ms.
20
~
Williams admitted that the industry practice being referred to was the i 21 nuclear industry (see Tr. 12,367). Further, Ms. Williams relied on an 22, office survey to determine industry practice (see Tr. 12,954).
23 'In reference to the problem involved with cinched-up U-bolts, Ms.
24 Williams stated that Cygna had been considering torque for some time, 25 'that they were trying to address all of Mr. Doyle's questions, and
~
30
3, - .
7,
. ..
- 1
.- y
,j' .r. ,
L j i realized that the unc(ertaking was far larger than could be answered 'in ni ,
I kV. 2 that time period . . . (Tr. 12,406/0'-12).
t , ,
l
a 3: On this subject., after lengthy'reseaNh, Cygna's Dr. Bjorkman U
gj state'd that further study was required (Tr. 13,002).
5 #1"*II D beth Ms. Williams and Dr. Bjorkman admitted that they had '
_f .
3 _,
never seen the cinched-up U-bolt or box frame concept at other nuclear 6
7 . plants (see,Ms. Williams at Tr.13,027, and Dr. Bjorkman at Tr. 4 4 e g 13,02P). ,
I q
((. <
g Relative to the non-compliance of the above, the chain of codes and laws is as follows:
10 ,
jjj >i (a) In accordance with the provisions of ASME Section III, NA-
,)
3250, PROVISIONS OF DESIGN SPECIFICATIONS, Applicants 12 13_
Prepared {n technical specification for nuclear safety-related , ;
equipmer.t; 14-
!,' (b) The title f the d cument generated in compliance with ASME
- 15 _ f i y , o Section HI, NA-3250, is NUCUAR SAFETY CLASS PIPE HANGERS ,
y AND SUPPORTS, SPECIFICATION 2323'MS-46A;
, x gg- (c) ASME Section III, NA-3320(b) regarofng Manufacturcr's responsibilitie , is directed to ASMT! NA-3340; and NA-3340, q 19 ,. 4 20 .s
'k ' RESPONSIBI M FOR CO W ANCE WIM HIS SECTION, states:
t T
'i' "The ranufacturer who completes or stthstantially
' 21
' y"g 1 ,
completes any component, appurtenance,'coregsupport r
' 2g ' structure, or component su; sport required t3 he in
, , compliance with this Section has the responsibility.for
' t e structural inte ty using the Design Spec m cations 23 ,
as a basis of design ... ." *
(d) In the Design Specification (technical specification) 2323-
~ 25 ~
MS-46A, page 3-15, Se'ction 3.3, CODES AND STANDARDS, the
/.3, f, ,
cr . : l , ,
- $ 31' S -
W,.l&
bj - ..
q ,p ,+v
a 67 g 1 following is included:
~2 " Design, fabrication, materials, certification, code stamping, and testing requirements included in this 3 specification shall be in accordance with the edition and addenda of the following codes, legislation, 14 regulations, and standards, in effect on July 28, 1975, unless'otherwise specified below or authorized by the 5 owner."
-: e .
n' 6 - 'on page 3-20 of the same document, the following are listed 7- as codes with which Applicants shall comply:
i
! 8 "g. Manufacturers' Standardization Society (MSS)
- 9 "(1) MSS SP-58, Pipe Hangers and Support-Materials and Design 10
"(2) MSS SP-69, Pipe Hangers and Supports-Selection and 11 Application" II 12 (e) At the Foreword of MSS SP-69, the following is quoted:
13- "The requirements of this standard were developed by a cooperative effort of representatives of pipe hanger 11 4 manufacturers. They are based on the best practice current at this time and on proven results of the 15 research and experience of this industry."
16 (f) Table 1, page 3, of SP-69 does not recommend the use of U-17 bolts (type 24 is the designation for U-bolt in this
'18 material) for insulated lines classified as Hot A-1. And 19 Hot A-1 is defined under 2.1 Hot Systems as bieng 120 degrees 20 F. to 450 degrees f. For Hot A-2 line, type 24 supports are 21 not recommended whether the line is insulaced or not.
a 22 (g) SP-69 states-at 12.2, " Riser clamps (type 42) shall have a 23- position means of engagement between the pipe and the clamp."
~24 (By inference, this would include type 24 clamps, which are 25 U-boirs.)
32
'l-(h) The practices used in the design and construction at Comanche 2 Peak nuclear plant proceed under a false premise; that is,
.3 one constructs the facility, then justifies the construction 4' as opposed to justifying the procedure and then constructing.
5- (1) Comanche Peak practices evade the provisions of 10 CFR Part 6 50, Appendix A, Criterion 1; 7 (j)' Also, Comanche Peak practices evade the provisions of 10 CFR 8- 50.55(a)(a) regarding design to standards commensurate with 9 the safety function.<
10- (k) Comanche' Peak practices evade the provisions of 10 CFR 11 50.34(a)(8) on the requirement to prepare a plan for research 12 and development for unique designs in the PSAR.
i3 14 The next item for consideration is in reference to:
.=15- E Box Frames:
16 Box frames utilized as quasi-clamping devices present several 17 unique problems:
18- (1):. Stresses in the box frame and most particularly the welding in the 11 9 - box frame are indeterminate.
20 -(2) S:resses in the pipe due to the constraint of the thermal growth-21 of the pipe caused by the box frame are indeterminate.
J22 (3) We again have the problem of no positive means of engagement
'23 between'the pipe and the box frame.
24 ; .. (4) The use of box frame deviates from the provisions of SP-69 in that
-25' this is not'a rigid support in the direction of loading, whether i
- r .' 3 3
- 1. . * .
.e i
!1: ~
'from below or-regardless of direction in a seismic event.
'2 (5) The effects of thermal constraint on the box frame are as follows:
~
- 3 (a) Af ter attempting to justify the use of this unique support concept, the NRC Staff and the Applicants have failed to
'S produce evidence of'its adequacy. And in fact, the Staff and 6 --
the Applicants through their agent admitted that this box-7' ~
frame type arrangement and the U-bolts arrangement (double 8
pin-ended supports) are unstable. (See: Judge Bloch at-Tr.
9 6696; Staff's Dr. Chen, at Tr. 6697, 6698, 6721, and 6727.)
10' Cygna Energy Services, when challenged on accepting-this 11-design concept (see CASE Exhibit 928, expanded and renumbered
.12 '
version of CASE Exhibit 891, bound in at Tr'. 9825), spent 13 .
several months in analysis by the. finite element method. The
'14 results of their analysis showed that the stresses in the box L15 ,
. frame (welds not analyzed) were, and this does not include
~
e' 11 6 internal pressure or moments of the pipe,_ at about 70 per 17- .
cent of.the material' allowable which was assumed to be 3 Sm-11 8 (not the Sh plus Sa that we would have assumed for the pipe
'19 as required by technical specifications and the code).
20' Cygna's Dr. Bjorkman stated at Tr.12,669, "I said it L : 21' would need further consideration." Also, when asked if this
- 22
-problem should have been looked at by Applicants, he stated,-
'23=
, "Yes." At Tr. 12,724, Dr..Bjorkman stated that the stress in
. L24 - ~
the box frame was over yield by 1500 psi, but he said he was
'25- .
using A36 allowables and the actual allowable for A500 is 9
34
. . , , . . _ . . . . _ . _ _ _ . _ , - _ . . _ . . . _ _ _ , . . . _ _ . . _ _ . _ _ . . - - _ . . _ . ~ . . _ _
I 6,000 psi, better than A36; (however, the strain hardened 2 area on a 16 inch long steel member subjected to 100 inches 3- of welding has undoubtedly been annealed back to the A36 1
4 condition). Dr. Bjorkman stated that the constraint forces 5 induced in the sides of the box frame are about 70 to 75 6 kips (Tr. 12,714). He stated that while the analysis for the 7 box frame is nearly accurate, the analysis for the U-bolt 8 cinched-up on the pipe requires much more work and fine 9 tuning (Tr. 12,882).
10 The facts are that the problems with box frames represent a gross 11 . concern for the design of Comanche Peak and are in violation of the 12 codes and laws to which Applicants are committed.
'13 In reference to item 5 above, Dr. Bjorkman admitted that 14 engineering judgement to dismiss calculations for box frames was 15- inadequate (see Tr. 12,666 through 12,669). Beyond this, Dr. Bjorkman 16 stated that this problem should be looked at by Applicants, including 17 adding in the effects of pressure and pipe bending (see Tr. 12,669).
18 Ms. Williams confirmed that Applicants never looked into this 19- . stress _ problem (see Tr. 12,666), nor were the effects of such problem 20 added to the normal calculated l'oads for this support (see Tr. 12,666).
21 This problem indicates a non-conformance with 10 CFR Part 50, 22 Appendix A, Criterion 1, particularly as relates to the generally 23 accepted code section. Additionally, the procedure is not in 24 compliance with 10 CFR 50.34 (a)(8) relative to unique designs. Also, 25 - the support fails to consider the industry practice as outlined in Mss 35
c ..
1 'SP-69 at 12.2'(depending on ther' mal expansion is not a positive means 2 of engagement).
^3 07. What. conclusions do you draw from this phase of.the Cygna review?
~4' A7.' Of'the 9 support calculations reviewed, the following is apparent: 6 5 of the 9 calculations, or 67%, were truncated and/or incorrectly done
~6; and one :of the supports had to be reworked since a structural failure ,
.7 L was apparent (see clip angle to hold U-bolt, CASE Exhibit 928; see also 8 Ms. Williams where she agrees that the clip angle fails, see Tr.
9 12,699-12,706)'. Finally all of the above points were established by
-11 0 - *e, not Cygna.
11 08. What are your points to be made relative to Phase 3 of the review by
~
l2 Cygna Energy Services?'
el3 A8. Bef re proceeding, I would like to establish some' ground work. First, 14L as far as the contents of the' calculations review by Cygna, I have 15 .never seen them. I therefore am proceeding on the basis of the raw f
.16 material pro 0ided by Cygna in Volume 3 of the Phase 3 report starting with PS-067-through'PS-088.
~
fl7_ I am not referring to Volume 1 of the-l18 report, because the conclusions are not based on whether the design of 1g the plant was properly'done but rather now that ,the plant is complete.
20 .does Cygna think it is 0.K. . I am not concerned with the fact that the 21' money is safe because the thief didn't get it; I am interested in whether or not the chief was innocent because the money is safe.
, 22 23 .The following analysis of the raw data from Volume 3 of Phase 3 I2f f r.the main steam system will point out the real depth of failure as 25 - relates to the Applicants' 9-level final _ vendor certification program 36 ,
r-2-r - - . . . _ -- .
c.: :
'a= -
' ~
(the minor fin'aings are not-inciuded in this assessment but only those 1
'2' findings which caused Cygna to have to produce a calculation to qualify
'3 the support -- or equivalent). I didn't include errors relative to
.4 local omissions, although I don't agree with Cygna Note 3 because the
5' AISC requires a local check at all levels of moment connections (see 6 .AISC 4-88 through 4-99). The fact that another code does not mandate 17 '. this is irrelevant as a consequence of 10 CFR Part 50, Appendix A, 8 Criterion 1.
- 9' All of the following are from Volume 3 of the Phase 3 Cygna 10- = Report:
-11 PS-069, Item 2, pages 1 and 2 of 10:
~
12 " Items 16'to 23 (stability part of the structure) are not designed. -Acceptability of this support-is entirely based on
- 13. the review of the design of the structral components not relate'd to the stability part of the structure. (See' 14 Observat' ion PS-02).
-15L
- "The methodology used to check the weld between items 8.and 9 is' incorrect. However, the weld is acceptable (composite
~16: section).- See Cygna calculation 84042, 4-F, Set B3, Rev. O.
See Observation PS-07."
T17-PS-070, Item 2, page 1 of 9, and
- 18 . PS-070, Item 22, page 8 of 9:
19; Item 2: "There are no' design calculations for the bumper portion of this restraint (see Observation PS-02)."
120 Item 22: " Weld of T.S. to Item 5 is not consioered in 21 calculations, weld stresses acceptable per Cygna Calculation (File 4-F, 84042, Calculation Set B2, Rev. 0)."
22.
PS-071, Item 13, page 6 of 10. It should be noted that this
- 23- problem could be critical if the error were reversed and a column load were shown'as tension:
-24
" Analysis SA-4298: Signs for loads were changed twice while L25 taking the loads from STRUDL to input into base plate 37 -
l analysis. As a result, compression became tension on the baseplate. The approach is conservative.
-2'-
" Base plate model does not show any node points or grid
~
3 points, bolt locations, etc. (Ref. SA-4299.)"
4 The use of the word " approach" above indicates intent; this
.5 was an error, not an approach.
6 PS-071, Item 21, pages 8 and 9 of 10:
7 "On vendor certification cover sheet 1, Rav. I a note stating that ' warping of Items 3 and 7 as indicated in CMC 59142 not 8 considered or approved'. There is no indication in the calculation that the warping problem has been resolved. Also 9 on sheet 1 of drawing, Rev. 4, note 3 states that Items 3 and 7 may not warp. Per TUGCO's response, this matter was
-10 addressed by the site task force (see response in 6/8/84 letter, Question #42) and engineers were directed to increase 11' the section properties 5% (see Attachment.K to that letter). Support member is acceptable per Cygna Calculation 12 Set A9 (Binder 4F, 84042) with 5% modification. Per TUCCO response dated 6/22/84, all 12 affected supports have been or 13 will be reviewed to ensure that the 5% is accounted for."
l14 'It should be noted that this is not a legitimate procedure.
15 The residual stress caused by the warping must be calculated and ,
- 16. then added to other stresses.
-17:
'PS-072, Item 2, page 1 of 9, and 18l PS-072, Item 22, page 9 of 9:
19 Item 2: " Reaction forces at support points were not distributed correctly for both the base analyses. See 20 comments under Item 13."
'2:1. It should be noted that this is still in violation of Gibbs &
L22 Hill procedures in Specification 2323-MS-46A and results in lower
'23 ' stresses in the members.
24 Item 22: "The weld between Items 22 and 23 was not checked.
However, the weld is adequate based on Cygna Calculation Set
-25 B7.(84042, 4-F)."
l 38 e
L, ] . PS-073, Item 1, page 1 of'9, PS-073, Item 15, page 6 of 9,.and
?2 PS-073, Item 22, page 8 of 9:
- 3. "STRUDL model does not match as-built support drawings with respect to:
"4 1. Member Lengths;
- 2. Load-Application Point.
35 However, Cygna agrees that the discrepancies are acceptable per TUGCO's response (46a) dated June 8, 1984.
- 6
" Sheets 10 through 14 of Rev. 4 calculation are missing.
.7 Sheet 9 references these pages. These are modification sketches."
8 It should be noted that the June 18, 1984 (item 46(a)),
9 letter justifies a problem; it does not alter the fact that the 10 problem was missed by the complex vendor review system.
'11 Item 15: " Frame is attached to existing whip restraint
'12; structural members. Design has not addressed connection of support members to 1" Plate'or 1" Plate connection to pipe
- 13 - whip restraint members. Cygna agrees that the connections to the pipe whip restraint members are acceptable per TUGCO'e
.]4 ' June 8, 1984"
! .15 - Same comments apply as for Item 1 preceding.
116 = Item 22: "Many member joints and the connection to whip restraint beams are not evaluated in the design. See comment 173 under Item'15.
, .: ]g - "No check of base plate weld adequacy, as shown in Section K-l K. Weld stresses are acceptable per Cygna calculation B6 7]g' Binder 4/F.
l-
- 20 " Connections.between TS 6 x 8 x 3/8 not designed (Items'39 and 40). Weld stresses are acceptable per Cygna calculation
'21 B6 Binder 4/F."
L22l PS-074, Item 2,'page 1 of 9:
,' 23 " Punching shear was not checked; however, it meets the requirements per Cygna review.
" Weld between material Items 13 and 18 was checked without
- 25; e nsidering the additional eccentric moment due to the
-39
- j. _
l
!~
l
( ,
I
m, 1
location of weld center.of gravity for assymmetric weld pattern (Sheet 17 of 19). However, the stress is still 22 acceptable per Cygna's calculation, set A6 (84,042, 4-F).
- Sae Observation PS-05."
.3 .
'PS-075,' Item 2,.page 1 of 9, 14i PS-075, Item 8, page 3 of 9, PS-075,: Item 11, page 5 of 9, 3 5 .- PS-075, Item 21, page 8 of 9, and PS-075,. Item 22, page 8 of 9:
- 6 Item 2
- " Stresses in pad are not computed and are the
~7? . responsibility of NPSI (Secaucus). See Note 2 for more
'information.
^8-
"The component forces due to swing angle were not considered 9 in the design, but are acceptable since effect is small."
10~ Item 8: '" Snubber design calculation was not.provided.
However, it meets the requirements based on the Cygna review
- ll: Calculation Set No. A4."
12 Item 11: "The bolt interaction.was based on the allouable for 13-1/8" minimum embedment, but the as-built drawing
- 13 shows that the embedment length was only 13". However, the interaction'is still less than 1 if the allowable load for l14 13" embedment-length is used in the calculation (per Cygna review Calculation Set No. B1). The 1/8" difference is
.15~ minor." j.
-16 ' Items 21 and 22: " Web crippling and buckling of members-where concentrated load is applied are not checked.- However, 1 71 .all the members are within allowable."
J18 "The' welds of the composite. beam section are not checked between Material Items 21, 25 and 26. . However, they are
-19 acceptable per.Cygna Calculation Set No. Bl."
!20' PS-076,-Item 8, page 3 of 9, and PS-076, Item 13,-page 6 of 9:
21 Item 8: "The snubber setting and other catalog items were 122 not checked-in the design calculation. However, they are all acceptable based on Cygna calculation."
-23 Item 13: " Minimum spacing violations exist-(i.e. less than 24- 10") but since interaction value is very low (0.26),-the anchor bolts are still acceptable by inspection."
u :25 ..
~
40 f
. ~
, .~^
11 - It should be:noted, however, that the fact remains that the 2 . procedure was.not followed and the final vendor review missed it.
3- ?PS-077, Item 24, pages 8 of 9 and 9 of 9:
j4 " Buckling is not checked. Cygna checked total normal stress
.. against allowable stress. Ratio of faulted stress to normal
'5l ' allowable is small (24%). Therefore this is acceptable."
'6 It should be noted that, while the comment of Cygna may be '
7" true (although I am not agreeing with Cygna),.they had to perform
~8 'a scoping analysis'which Applicants failed to provide, and this 9 . was not found by the final vendor review. The ratio before
~
10- Cygna's. scoping was indeterminate and could have been serious.
11; PS-078,. Item 1, pages 1 of-l'0 and 2 of 10,
.PS-078, Item 2, page 2 of 110:
12 --
Item 1: " Input data calculations for STRUDL model shown on 31 3. sheet 1 to 8 Design Input calculation (dated 6/21/83) not checked. Moment of inertia and section modulus for members 5 14 -and 6 and 8 to 11 were not calculated correctly.
--15 '- -
" Members 5 & 6
. "Iy should be 359.95, but 681.51 used in STRUDL input Sy 16' should be 71.99, but 136.3 was used in design.
11 7 " Members 8 to 11 "Iy should be 642.73, but 1213.1 was used in design. Sy 18 should be 107.12, but 2.7.18 (sic) was used in design.
'19; " Reanalyses of the frame was required in order to obtain the correct forces & moment for designs. Per TUGCO's calculation L20; dated June 8,;1984,,the frame is re-analyzed with beam .(Item 22)~ boxed with 3/4" plates. The modified support is
. 21 - acceptable (Reference Cygna communications report, dated May 24, 1984, Question 41). See Observation PS-01."
122 Item 2: "The STRUDL analysis assumed fixed ends at nodes 4,
.23 : 8, 14 & 17. This should be evaluated since those joints can not take high moments and forces. No detail calculation was J 24 ; provided in the design for those joints.
Per TUGCO's calculation, dated June 8, 1984, the frame was 25 modified. .In the re-analysis Nodes 4, 8, 14 & 17 are modeled 41
~
.. 4e ,
l l
3
.as pinned connections. Cygna's review shows the support is
/2' acceptable'.""
'3 ,
It should be noted that this'is not precise; a release based
' ~
4 on stiffness would be the correct approach, not complete release.
-5 PS-079, Item 2, page 1 of 9:
67
'.'No calculations were performed for snubber and weld
' connection designs. 'However, they are acceptable per Cygna Calculation Set No. A4 (File 4-F, 84042)."
7
.g. ,
PS-081, Item 1, page 1 of'9, PS-081, Item 2, pages 1 and 2 of 9, PS-081, Item 9, page 4 of 9,
'9 PS-081, Item 11, page 5 of 9 PS-081, Item 13', page 6 of 9, and PS-081, Item 23, page 9 of 9:
Il-Item 1: " Applied forces.used in SA-4284 were-referenced to j ints 2 & 6 which did not match the corresponding location 12 of Item 20 in-the STRUDL frame analysis model. There was no corresponding Node No. assigned to the locations where the member forces were supposed to be used in SA-4284. Per p- TUGCO's June 8, 1984 response (Reference Question 44), nodes 2 and 6 are from a model for a similar support. The locations match closely to those of the actual connection.
Cygna's review of that support (MS-1-004-002-S72R) shows this t be 8n 8cceptable comparison."
' 16 ,
- g. It should be noted, however,-that Cygna could not review the
- )g . ' support without assistance.
Item 2: "No calculation provided to evaluate member stresses and weld capacities. Cygna's review of the siinilar support
,g. shows them to be acceptable."
lg It should be noted that the same assistance as above was g required,'since the calculation did not refer to similar~ support
.g but calculated PS-081. The reviewers for certification missed this p'roblem.
' Item 9: "No calculation provided for strut design. However, 25 -it is acceptable (see'Cygna Calculation Set. No. A4, Binder 4-F)."
42
).-
Item 11: ~ "The app' lied forces uad to qualify base plate shown on Drawing Section B-B were different from those shown 2 in the STRUDL output. (They are verified per TUGCO's June 8, 1984 response, Attachment L.) Bolt locations assumed in the 3~ analysis were different from the actual bolt locations shown on the drawing. However, the discrepancies are not 4 ,
significant and bolt and plate stresses are very low."
5~ .It-should be noted.that since the qualification for this 6 support is based on another support calculation, the material 7 contained in the calculations for this support are fictitious and
~ 6- yet there was no comment by the final review process. Beyond 9 this, even with the assistance of TUGCO, Cygna had to perform at 10- least two calculations.
11 Item 13: " Applied loads used to qualify Thru-Bolt did not match the STRUDL frame output. Also,_ analysis SA-3662, Rev.
12: 1 does not show the same values as used in sheet 4 of 8 the analysis SA-4282 calculation. Note: These comments are
- 13 . clarified per TUGCO's response (item 44) and Cygna's review of the original calculation at Grinnell's office in I4' Providence, RI, which confirms the acceptability of these i components."
., l 5 -
Same consents apply here as for all of the above.
- 16 . -
Item 23: "No cover plate provided at tube end moment joint
'17 (Material Item 18). .It is acceptable per Cygna Calculation set C2,' Binder 4-F. See also Note 3."
PS-082, Item 2, page 1 of 9, 19 PS-082, Item 2t., page 2 of 9, and PS-082, Item 10, page 5 of 9:
Item 2: " Incorrect'section modulus used for combined section
!- '2l. (Items 8 and 9). Values of section modulus cannot be added algebraically to get total section modulus (Sheet 3 of 9).
< 22 However, stress results are conservative, therefore acceptable."
Item 2A: "No design calculations were provided for the loads
. 24. or stresses on the members added for stability (see l Observation PS-02). This review covers only the remaining l '25 members." -
l l .
j 43 -
e
.a ?? l l
l1 Item 10: " Incorrect allowable used_but OK; final design load l s
is within ITT Grinnell LCD allowable."
12.
, PS-083, Item 1, page 1 of 9, !
E3x PS-083, Item 2, page 2 of 10, PS-083, Item 10,-page S-of 10, 4~ PS-083, Item 13, page 7 of 10, PS-083, Item 22, page 9 of 10, and 5: PS-083, Item 23, page 10 s' 10:
'6: Item.1: " Moment load on base plate is transferred into compression / tension loads by taking incorrect dimensions of
, 17; rear bracket. (See comment under Item 13.)
'8- " Incorrect moment of inertia for combined section of Items 30
& 31. However, it is acceptable'since stresses are low.
9
~
" Rear bracket dimensions used in design are taken from ITT 10 LCD-211,'Rev. 16. See Note 11.
-11. " Improper dimension from center-line of plate to gusset plate (C 1/8") in Section A-A."
- . 12 Item.2
- "The approach to check the weld between Items 30 &
13 31 was incorrect. Stresses acceptable per Cygna Calculation
.C4, Binder 4F. 'See Observation PS-07." '
14-JItem 10: "SA-4122, referenced in calculation for U-bolt c 15 design, shows the failure of the bolt. However, the bolt is qualified based on the ITT Grinnell IOC from_ Frank Birch to
'16 Roland Serino dated 12/15/82. This is further justified by the results of the ITT test program."
Item 13: " Larger dimensions were taken for rear bracket 11 8 while_ converting moment load on base plate into tension / compression forces at the four corners of welded i19f attachment. Therefore, actual loads are slightly different from the calculated ones. Discrepancy is not significant in 20l terms of anchor bolt loads and stresses in base _ plate."
(21 Item 22: " Incorrect section properties calculated for weld
.. (Items 31 & rear brackets and base plates). Weld i.s designed L22- based on larger dimensions of the rear bracket which do not physically fit'within TS 1/2" x 12" x 8". However, the 231 connections are acceptable with the smaller dimensions. (Per Grinnell LCD, Rev. 16 for Figure 211.)"
i- 24 Item 23: "Special analysis SA-4123 shows that the pad used
'25- between U-bolt and pipe fails due to bearing stress.
44
11 However, the pad is qualified based on a rapid letter from R. ,
B. Reslau to Nick Patsalides which does not show back-up -
j
-2 calculation. It is acceptable per TUGC0 calculation in their '
June 8, 1984 letter, Item 43."
-3 :
PS-085, Item 2, page 1 of 9, and 4 PS-085, Item 11, pages 4 and-5 of 9:
5' Item 2: "In STRUDL model, member 18 (Item 34) has member forces released.at joint 17. Releases should have been at 6 joint 20. However, since this is mainly an axial. force member, there is no significant impact on the design. The 7 stress is low."
8 Item 11: "For analysis of the. base plates (Items 31, 32, 44, 45), the NPSI Base Plate Program requires. loads to be input 9 at 5 load points. Generally these points are the center and corners of the attachment. No calculations were made to show
- 10 how these tensile or compressive loads input for these base plates were derived. Also note that on plates 31 & 32 the 11 sides of the welded attachments are not parallel to sides of the base plate, which is the assumption for this method.
However, they are acceptable based on the calculations shown in TUGC0 response-dated June 8, 1984. Also, see Cygna
~13 Calculation Set No. C3 for additional justification."
'14 PS-086, Item 3, page-1 of 9, and-PS-086, Item 11, page 5 of 9:
15 Item 3: "The load used in the calculation was 20194 lbs.
11 6. .The. load reported in the analysis AB-1-02, Rev. I was'20305
,1bs., a difference of 111 lbs.' This difference is negligible E171 because the support allows a 10% deviation in loads.
See Observation PI-00-06 "
~18
. _In reference to the final vendor certification review, see 19 Tr. 7153, where Applicants' Witness Mr. Vega states that Step 7 of 20 -~
the iteration compares loads and modifies them if required.
- /21.
"The base plate and anchor bolt were considered Item 11:
~ 22 - adequate by engineering judgments without any calculation.
However, the anchor bolts and plate are acceptable based on 23 the Cygna calculation set."
U 24 PS-087, Item 1, page 1 of 10, PS-087, Item 2, page 2 of 10, 25 ' PS-087, Item 6, page 4 of 10, i-45
'E4 ..-,,.-~,~----,,,,y, ,,,.,--m-. .w... .,w..,.,,, ,,-p -- - - ~ -.w-
l .
e
- l. PS-087, Item 17, page 8'of 10, PS-087, Item 18, page 8 of 10, 2 Ps-087, Item 19, page 9 of 10, and PS-087,. Item 22, page 9 of 10:
Item 1: "In gang support CH-1-030-004-C76R, which connects 4 to this support, loads were incorrect in load case 1 (Y load at node 6). Load Case 2 has correct loads (except friction) 5: .and.STRUDL results show little change for this support between Load Case 1 and 2; therefore, the load cases used are 6; adequate for design of thie support.
7 " Wrong angle section used in gang support bracing in STRUDL analysis - minimal effect on this support.
" Latest load Fy = -34676 lb. from piping analysis, design 9 used -34027 lb., which is a change only 2%. This does not impact design.
10' See Observation PI-00-06.
11 " Thermal displacement in Z-dir. (normal Min.) is -0.2'03" instead of 0.0" as noted in the load summary sheet. No 12 impact on design.
- 13 "On support drawing sheet-1 of 3 plan view at el. 899' 9-1/2" shows . structural tubes located at higher elevation 903' 7".
14 Minor drafting inconsistency."
Ll5 Item 2: " Tube end prying action on Richmond inserta,is not considered. See Note 6 for further details."
16 Note 6 is contained in Cygna Volume 2, Appendix J, pages 3
-17 and 4 of 9, and states:
18 "6. Tubesteel Prying on Richmond Inserts 19 "In the designs which employ tubesteel/ Richmond insert
. 20 . combinations, Cygna noted that the engineer released the rotation about the Y and Z axes (see, sketch) in the STRUDL 21 model for the frame. While release of the Y rotation is appr'priate o since the bolt is free within the tube, release 22 of the Z rotation assumes that the tube will not bear against the washer at point ' A' and create a load due to prying on J23 the bolt. TucCo has provided justification for this and
'other analytical assumptions such as: the bolt does not
, 24 carry any. load in bending and the effect of bolt hole offset on bolt load. Details of the justification may be found in 25 the TUGC0 letter dated 5/8/84 and in the ' Affidavit of John 46 -
[ ! a,
- a -
l I C. Finneran, Jr., Robert C. Iotti and R. Peter Deubler Regarding Design of Richmond Inserts and their Application to 2 Support Designs.' In the letter to Cygna, TUGC0 shows that ,
prying due to rotation about the Z axis is not present when L3 only vertical loads exist. When torsional moments (Mx)-
exist, the study done by TUGC0 shows that even with small 4 . amounts of torsion (1000 in-lb vs 40000 lb tension load), the
, effect of prying is due to torsion, with no contributio'n from 6' moments about the Z axis. For large torsional loads (4000 in-lb vs. 2000 lb tension), the same effect holds true.
6' Cygna then reviewed all tubesteel/ Richmond insert joints within their scope and determined that the configuration 7' analyzed tar TUGC0 (4 x 4 x 3/8 TS with 20" bolt spacing) is representative of the most flexible configurations and, 8 therefore, most conservative. As a result, Cygna finds the method used by TUGC0 to model these connections is 9 acceptable." (A drawing then follows.)
10 In my opinion, the contents of Note 6 appear to'be a desperate 11 attempt of one looking for a way out of a dilemma. But in any event, 12 the final review team missed the, problem.
-13 Item 6: " Swing angle of rod is not checked - OK per Cygna calculatiot. Refer to Binder 84042/4-F call calculation set 14 C1." -
15 Item 17: " Weight of the spring is not considered in STRUDL analysis. (Approximately 5%)"
16-It should be noted that if one can't increase the allowable
.17
'by 5%, one cannot do the reverse (reduce by neglect -- the dead 18 load).
19 Item 18: " Additional dead load has been added; however, no 20 frequency analysis was performed - an approximate inertial load was applied, though no explanation or justification is
, 21 provided. See Note 7."
12 2 Item 19: " Friction load is not considered on a related gang support CH-1-030-004-C76R. See Observation PS-08."
'23-Item 22: " Additional moments due to the eccentricity in 3 24 sided welded connection between Items 31 and 45 are not considered, but acceptable since the stresses in the weld are 25 low. See Observation-PS-05."
47 i
,, + - - - - - , , . . -g - , - - - ,r,,,- , m, e- n . - - - - -, , , a- - -
1 .PS-088, Item 2, page 1 of'9',
-PS-088, Item 8, page 3 of 9, and 12 -PS-088, Item 12, page 5 of 9:
3 Item 2: " Stress in pad was not checked, in this Revision but will be reviewed by NPSI (Secaucus). See Note 2."
- 4 Item 8: '" Snubber setting was not checked. However, it is 5 acceptable as per Cygna Calculation C6."
6 Item 12: " Prying action of tubes due to bending is not considered. See Note 6."
7-8 09: Does that conclude your comments on Phase 3 of the Cygna Report?
9 A9: Yes, except to summarize the material presented in the last two
-.10 answers. But first a point must be made relative to the QA/QC program 11 and its recent history.
i l2 On numerous occasions, we have he.ard the Applicants and NRC Staff 13 deny the validity of allegations based on the argument that the support 14: being questioned'was not final vendor certified and therefore not open
~15 to critique.
16- on an equal number of occasions we have heard the Applicants and
.17 NRC Staff state with equal persuasiveness that the errors would all be 18 caught in the final vendor review.
.19 Beyond this, in testimony we listened to Applicants' Witnesses
.20 Messrs. Finneran, Vega, Krishnan, and others extol the virtues of a 21 review system based on 9 levels of checking that was so tight that even
- 22. time couldn't. slip by (see Tr. 7140-7200, especially 7189 and 7192).
23 But now the time has come to silence these oracles of levity "24 because their flawless system is based on the dreams of "somewhat
-25 knowledgeable" or incompetent individuals.
48
y -- _
,; ' J, .
I' (1 For starters, as was shown'in my earlier answer, for the nine
_ - supports I reviewed'for Phase 1 of the Cygna Report, which were all
. l3^ .vendot. certified, one had a fatal structural failure and required 4 modifications, three were analyzed by procedures which are not 5 -applicable to the structures in question ~and one other contained
'6: incomplete analysis; the final support contained a deceptive equation
<7 to avoid the provis,lons of the code. In other words, 2/3 of this
.8 sample were not in accordance with to CFR Part 50, Appendix B, 9 Criterion III requirements of:
10; " Measures shall be established to assure that applicable regulatory requirements and the design basis . . . for those
.)) structures, systems, and components to which this appendix applies are correctly translated into specifications, drawings, 12 procedures, and instructions. . . " etc.
-13 ~ As for the problems discussed in my previous answer on the 22 main-14 steam supports reviewed as Phase 3'of the Cygna Report, I offer the
- 15; following summary. . of the 22 supports reviewed by Cygna, which are
.16 all Class 2 supports, 18 contained gross errors, a portion of which 17 include:
(
J18 i Support required major modification to meet the code
.19; requirements due.to calculational errors used to obtain the
' 20 8*Ction properties for 2 members (see Phase 3 Cygna Report,
~21 Volume 3, PS-078).
2 other supports had errors in the calculations for the section 22
23 properties (see PS-082 and PS-083).
' 24 ' 4 Eupports had. calculational errors in the weld analysis (see
- 25 PS-069, PS-074, PS-08,3, and PS-087).
49 .
7- -,
4., .
' supports have calc'ulations missing for structural members
~
l .-
12 (see PS-069, PS-070, PS-073, and PS-082).
- 3 1 support has a calculation for section property of a weld I4- which is incorrect (see PS-083).
5 2 . supports have no base plate A/B calculations (see PS-073 and 6 PS-087).
73 5 supports have missing weld calculations (see PS-070, PS-072,
'8' PS-073, PS-075, and PS-079)'.
79 1 support has severe warping problem while note on drawing
.10 ' prohibits warping (see PS-071).
11 ~ 1 support failed to include friction loads (see PS-087).
12 2 supports had calculations which used the wrong loads (see PS-13- 073 and PS-086)..
14 -I support incorrectly used the NPSI base plate computer program 11 5 -(see PS-085). ,
16 1 STRUDL model assumed to apply' to a suport did not - for 17 answers to calculation problems another support calculation
. 18 had to be consulted (see PS-081).
19 1 support required check for buckling but calculation not done
'20- .(see PS-077).
21 I support required check for crippling at concentrated loads 22 but this was not done (see PS-075).
'23- 4 supports did not have calculations for the snubbers (see PS-l =24 075, PS-076, PS-079, and PS-087).
- 25. I support calculation used the wrong allowables (see PS-082).
l 50 -
. ._.,.,,-.m r.-. .m- - - .,.-n-e.--3 ---. ,-e-- - -
r,-
. .- * ?.
'l 1 support had no calculation for the struts (see PS-081).
'2 '1; support had 2 calculations which indicated a component 3 -failure which were written off by menos, not calculations
- 4 (see PS-083).
5 2 supports had no calculation for pads at trunnion to pipe 6 interface (see PS-075 and PS-087).
7f While these 36 calculational errors are not the total number of
-8 errors found in these 22 supports, it is indicative of the extent of 9 the OA/QC breakdown.
10 Sumaing up for the last two answers, a total of 9 calc'ulations and
-11. drawings from Cygna Report Phase 1 and 22 supports reviewed in Volume 3 12 of Phase 3 have.been discussed. These supports were final vendor 13- -certified, and the total is 31.
14 of the 31 supports in question, 24 have gross errors which
? 15 escaped detection through 9 levels of review prior to final vendor
-16 certification. This represents a 77.5 per cent error rate, which is 17 fatal for any discipline and a health and safety concern of monumental e-proportions.
19 -010. Do you have any idea of how many of these 31 supports have been 20 modified as a result of the Walsh/Doyle allegations?
21 A10. As a matter of fact, I have a good guess as a result of reviewing the
-2?. drawings in CASE Exhibit 6698 (Attachment to Doyle Deposition /
23- . Testimony, admitted at Tr. 3630) with the drawings in Cygna Report
-24_ Phase 3, Volume 3.
- 25) -I found that'the following suports have been modified:
51
2- a
'E 8 1
MS-1-001-006-C72K -- as a result of section property 12 _MS-1-001-003-S72R -- stability
- 3 MS-1-001-004-S72R -- stability -
4 MS-1-001-005-S72R -- stability 5' MS-1-001-006-S72R -- stability 6 MS-1-002-001-S72R -- stability
- 7. - MS-1-002-003-S72R -- stability 8 MS-1-002-004-S72R -- stability 9 _ MS-1-002-005-S72R -- stability 10 Of the 89 supp~ orts reviewed in the three previous answers, 11 therefore, a total of 30 have been rebuilt to prevent their collapse, 12 and this represents 34 per cent of the sample -- obviously a fatal 13 level of failure and a threat to'the public health and safety.
14 011. Do you have any further statements in reference to the Cygna Reort?
15 LAll. I have in this regard only one general statement on the notes in Volume 16 1.of the Phase 3 Report, and in this I will only address Note 16 17 because the points would be similar for all of the excuses offered by 18 the Applicants.
19' The box frame was analyzed by Cygna's Dr. Bjorkman for the May 20- 1984 hearings, and he found the procedures which he used to be close to 21 reality.- For that matter, the fine tuning required pertained to minor 22- points which may be noted in what Dr. Bjorkman said (Tr. 12882/5-8):
- 23 "So in this particular problem, I don't think the level of refinement is great. It's to incorporate additional loads in the 12 4 problem, the right allowables, check welds, et cetera."
'25 _The highly sophisticated finite element procedure used by Dr.
52
, e:
1- Bjoriaan indicated overstress conditions (see Dr. Bjorkman at Tr.
2 12,710-12,712). The Applicants have decided to discard the more accu-
~3 rate finite approach in favor'of hand calculations with overlapping 4 assumptions. Having chosen this process, Applicants proceed to manipu-5 late the numbers to show a favorable answer. One of the manipulations 6; used by Applicants in calculation SI-1-325-002-S32R (see Attachment A l7 to Affidavit on Local Displacements; especially pages 1 and 2) is to
.8 average on a straight line asaumption which is inaccurate in Appli-9 cants' favor.- Having assumed an incorrect average temperature for the 10 box frame, Applicarits assume (again incorrectly) that the aver' age at 11 that point is constant around a square so that the average temperature 12' at the corners of the square frame are the same as the average at a 13 point on the tube opposite the contact point with the pipe.
14 By their gyrations, Applicants reduced the lpads found by finite 15 methods to 1/2 of their values and thereby manage to make the support
~
16 appear to be what might be termed "somewhat acceptable". But looking 17 .at equation 11, page 8, the stress level is about .9; therefore, any 18L increase would fail the pipe. At this point, I won't become involved 19 with the erroneous use of the equation shown on pages 5 and 6, because 20 the principal error involves the average temperature manipulations, 21 shown above, which cannot be manipulated on a computer.
22 obviously, the pipe'is in serious trouble and so is the somewhat 23 - knowledgeable individual who tried to justify this problem with Mickey
, 24 Mouse gyrations which result in several scrious errors in fundamentals
.25 for these equations.
53
y -
.j Applicants and their agent's, when confronted with a critique of 2
the construction resulting from designs by somewhat knowledgeable engineers, rather than offering the documentation which led to the 3
- c nstruction, persist in coming up with all types of methodologies to 4
5 3"8tifY
- f*It E" "Pll*
What is occurring at Comanche Peak undoubtedly will have an 6
adverse impact on the health and safety of the public.
7 012. What do you'have to say on the material submitted by Applicants as 8-M ti ns f r Summary Disposition?
-9 g -A12. While I could say enough to stock a library, I won't. However, I g) believe I must make at least a few points in order to keep the Applicants and NRC Staff from getting the false impression that they 12.
are 811 PP i ng their trash past me. This might give them the idea that 13 j4 I am, like they are, "somewhat knowledgeable."
When examining any material supplied by Applicants and NRC Staff, 15 one factor to keep in mind.is that the material being supplied is not offered as evidence to do something, but is rather a frantic effort to p
)g _ salvage a 5 billion dollar public menace. Secondly, as has been shown conclusively, the Applicants' ability to produce accurate calculations
)g is n t nly questi nable, it has been proven to be impossible -- even 20
.with 9 levels of review plus the original check.
21 A fact r n t generally considered in these hearings is the intent 22 of the' law as relates to engineering for these facilities. By devisive i g- tactics, Ap'plicants would have the Board believe that accuracy is with ut Privilege as long as some Mickey Mouse manipulation of numbers 25 54 -
l would ultimately indicate that a structure would not fall on the floor 2 the first day-of operation. Beyond_this, Applicants' philosophy is 3- 'that if the first calculations prove fatal, it is acceptable to try new 4 assumptions and machinations until, by whatever Machiavellan postulates 5 are successful, the problem goes away. This is best noted in three 6 incidences:
7; (1) The flip flop of equations and assumptions for the upper 8 lateral restraint (see CASE Witness Doyle, Tr. 6018-6032; NRC 9 Staff Witness Chen, Tr. 6051; and Applicants' Motion for 10 Summary Disposition Regarding Upper Lateral Restraint B'eam).
11 (2) When finite element analysis proved that serious problems 12 existed for-the pipe and structure in the box beam
-13 arrangement (see Tr. 12,710-12,712, where Cygna's Dr.
14- Bjorkman discusses the problems), Applicants reverted to ..and
-15 calculation which could be easier manipulated (see Attachment 15 A to Applicants' Motion for Summary Disposition relative to 17; box frame: with zero inch gaps).
1g' (3) Finite element analysis of cinched-up U-bolts indicated fatal 19 problems with this approach to clamping. While Cygna's Dr.
20 Bjorkman did not accept the results of this analysis, he
-21 suggested that refinements to the model were required (see 22 Tr. 12,882). The Applicants again chose to displace the 23 highly sophisticated and extremely accurate finite element
- 24 procedure with hand calculations which could be more easily
-25 manicured (see Applicants' Motion for Su mary Disposition of 55
0~ ,
]
. 1
. . . ~ .-
- T ': ,
Ll ? -CASE's Allegations Regarding Cinching Down of U-Bolts; see
~
c2 also calculation No. CC-1-028-007-S33R in Applicants' Memo to l3) Cygna Energy Services dated June 18, 1984).
- 4. The fact is'that regardless of whether or not the support or 5 procedure will ultimately prove acceptable, the inptallation and 6' . acceptance of indeterminate structures in nuclear power plants is a 7 crime under the law.
18- This fact is reflected in the federal laws as codified in the Code 9 of Federal Regulations. See, for example, 10 CFR Part 50, Appendix B, il 0' ? Criterion III, as quoted on page 49 herein. In addition to the law, 11 certain codes and standards to which Applicants are committed are 12- required-by law; see 29 CFR 1910, the William Steiger Action 1970; see
-13 also 10 CFR Part 50, Appendix A, Criterion 1, which states, in'part:-
. 11 4 "Where generally accepted codes and standards are used, they shall be identified and evaluated to determine their applicability, il5 L adequacy,'and sufficiency and shall be supplemented or modified as necessary to assure a quality product in keeping with the required
...15 safety function."
17 'Among the codes which are applicable to' Comanche Peak is ANSI
- 18: N45.2.11, Applicants' Exhibit 148, which states at page 12, item 3.1:
4 19 "The design input requirements shall be specified on a timely basis' and to a level of detail necessary to permit the design s '20. activity to be carried out in a correct manner." (Emphasis added.)
?
l 22. The' design activity can be carried out in a correct manner only if
- 23. the design criteria is established in advance of the design, not when 24 the criteria is established after construction.
.25 Further, N45.2.11 states, in part, at page 25, item 6.3.1:
56 T
- . . _ . - _ . - - , _ , . , . . - . , , _ _ . . _ . . . . . , . _ . , _ - ~ . _ - . , , _ . . _ . - .
.. ~. C )
3 . . .
" Design reviews are critical reviews to provide assurance that 2
design documents such as drawings,' calculations, analysis or specificationa are correct and satisfactory." (Emphases added.) '
The above includes only a portion of the codes which mandate accuracy and correctness. From the material which follows, it will I ~
become clear that at. Comanche Peak accuracy is an unknown quantity that 0- is replaced'by the simpler unwritten Comanche Peak code: " Defend the bottom line utilizing the doctrine of plausible deniability."
0-The Applicants, to defend the support systems which have been
-9
~ installed at Comanche Peak, are not offering documentation which 10 preceded the fabrication, installation, and ' inspection, but are ll offering an. iterative series of numerical and verbal manipulations.
12- [The Applicants hope that the Board will accept as proper their original 13' : position of allowing "somewhat knowledgeable" engineers to attempt to.
d4 produce the required' calculations qualifying these pipe supports. In
.15 short, the Applicants are attempting to hide the tirees with a forest.
^16- -Applicants' attempts to deceive the Licensing Board are summarized M 'as follows:
18 1. Cinched-Up'U-Bolts:
19' .The first of these attempts to deceive may be found in Applicants'
.20 Motion for Summary Disposition offCASE's Allegations Regarding Cinching
- 21. Down of.U-Bolts, but more precisely in the June 18, 1984 memo to Cygna
.22' Energy services.
- 23 ' : As an example, with reference to the June 18, 1984, meno to Cygna,
'24- in attempting to account for. thermal expansion of pipe constrained by
- -25 .U-bolts,.page 8 of 9 and 9 of 9'of the calculation, CC-1-028-007-S33R, 57
,,-uv~..e me_,,, . - e,.,.. ,s-,....,m.m--.-. ,y. ., . . , ,,,----..~..e.,-,... -
, . . ...n
7;q , 's-
~,. :.
-]J
. Applicants only show what they'want and how they want, usually falling 12; short of.-a realistic analysis.
3 :If we assume that Applicants' end premise is accurate, in that 4- ' pipe deflections (due to expansion) result in forces in the U-bolt
-5 relative to the formula P = Delta k, then the process must also be 6 ' capable of working in reverse for this force (obtained by Applicants as 7 :331 lbs.) or'in: fact the force due'to pretorquing which is given by 8 - Applicants'as 2908 lbs.- Therefore,.the equivalent deflection of the
', . 9" pipe due to the total U-bolt load would be 3240 divided by 331 times
- 10 approximately 1/64" ='the' total diametrical constraint, or
] ] approxima'tely 3/16". As is obvious, the postulate does not work.
12 -I believe.that all of the pencilwhipping in the world cannot l13 eliminate the fact that Applicants undertook this method of approaching 14 ' pipe supports without any knowledge of the consequences, which is
- 15l evident from the fact (as noted above) that at this time they still
/16 m don't know the consequences of cinching and thermal constraint.
$l'7. . : The facts are that cinched-up U-bolts represent a gross concern 118; for the design of Comanche Peak and are-in violation of the codes and
~
19- -laws to which Applicants are committed._ ,
20 men Applicants attempt to answer one area of concern, they fail
- - 21; to notice that their answer almost invariably indicates the result of
- 22 2 pencilwhipping. For example, in an effort to show there is no problem
'23 associated with constraint of thermal' growth offered by cinched-up U-24 bolts, the Applicants' own procedure proves that a severe problem 125 exists for cinching, and that is the status of this approach is still 58 i
- 4.
- g s
9
. \ '
1 /~
.7 indeterminate.- But one thing is certain - the stress levels in the U-t.. ,
, - s e,r f 2' > 04
( bolt and run pipe are far higher than originally believed by i
3 l"- Applicants,.who (along with the NRC Staff) assumed that the additional
,,e 4
loads due to cinching and thermal constraint were negligible (see 5; - '~
D Applicants' Dr.. Chang, Applicants' Exhibit 142F, page 5, Question and x Answer No. 15, where he states that the temperature of the pipe and the :
U- olt are the same). Perhaps Dr. Chang was try n Ao avoid the 8~
provisions of ASME Section III, Subsection NF-3272.1 (CASE Exhibit 710, 9
admitted into evidence), which would require analysis of the effects of 10 ". . .Lthiusts, moments and other' loads imposed."
II'
'7n the SIT Report (NRC Staff Exhibit 207) at page 33, the report J.:
'12 states, in part:
13
" Alternately, since the maximum temperature differential between the-U-bolt and pipe in uninsulated is expected to be less than 50 4' -14 '
-degrees Fahrenheit, calculations performed by the Special , 3
^
3 ' I"*P***I " T*** I"di"***d 'h** *h* *** "I***d ** "d**Y *****
.l 15 ~
-and loads re negligible relative to ASME Code allowables."
w :16 .y Dr. Chen, testifying for the NRC Staff, stated (Tr. 6742/13-24):
~17- "Have'the stresses, due to the pretensioning of the MR. WALSH:
~
[. .g bolts, been added to the normal expansion'of the pipe?"
JUDGE BLOCH: "First, is that addresied in the SIT' report?"
39 -
WITNESS'CHEN: ."As f re !..s the load combination -- the addition of 0 stresses that Mr. C.1v nas just identified, no. But I believe l .g ,
that an assess * @ cs ade of the stresses resulting from
~
- <- preloading we'e cut .c was made, andiwas writtenxof 3 'on the g basis of enginorins ;mgment and usual industry practice."
2 ' The NRC Staff again states. that the effects of tuernal constraint 24 .are'negligibleonU-boltsandpipes,[butthistimethestatementisin 5 reference to insul'ated lines. Seo SIT Report, Jage 32, where^1t
,3 4
- .y '
59 I
. , . . ., , I- -- - ~, m ..---,. .: ,.,3---- -,-w- , ,y- --c. _ . - , - . _ -
I ,
- I states, in part:
2 "In the case of insulated piping (e.g., main steam, feedwater, and residual heat removal piping), the temperature difference between 3- the U-bolt and the pipe will-be negligible because the U-bolt is In thermal contact with the pipe and the insulation is installed
_4 over both the U-bolt and the pipe "
5~ I could go on, but why kill a dead horse.
6- As may be noted from the above, the Applicants and the NRC Staff
-7 argued vociferously and adamantly that'neither cinching of U-bolts nor 0 . thermal constraint had any significant effect on the U-bolt or the
- 9 hip.,
I However, as the Cygna finite element program pointed out, the
'A <ll- stresses exceeded 3 times the allowable. Beyond this, the deceptive 12 . mathematical manipulations by Applicants still resulted in stresses
- 13 which were more than merely significant.
6 '
, .l 4 Obviously from the above, it is quite clear that neither the x15. . Applicants nor the NRC Staff had or now has the foggiest notion of the 16- stress levels present in the pipe or the cinched-up U-bolt.
17- I could go into box frames at this point, but the points and
'18 arguments as well as the legal requirements would-be approximately the
, 19 same as those which are outlined above for the cinched-up U-bolts. ,
720- 2. Two-Way U-Bolts Acting As One-Way Restraints:
4 21 The second attempt to deceive which I will address has to do with
.22' two-way U-bolts acting as one-way restraints. Information regarding
'23- .this concern was contained in CASE Exhibit 669B (Attachment to Doyle
~
"24 Deposition / Testimony),Litems 3D, 3E, 3F, and 3G.
~ 25' The fact is that Comanche Peak has failed to consider two-way 60
- g. a..
e Sloading,'which is not.in compliance with the technical specification or -
2$ .
-the.. manufacturer's specification for interacting loads.
3 . The effects ~of constraint on a node not considered in the pipe.
4 stress introduce a nonconservatism'in the output of loads used in the
. 15. ' design o'f.all supports in the system.
0-
-' Applicants' view of this problem is that:
(1)' Analysis'was not required out-of-plane since these install'ed 0
' supports were only in the preliminary phase.of construction
- # -and justification would come later. (See pages 3 to 4 of
-Applicants Motion for Summary Disposition of CASE's II Allegations Regarding U-Bolts Acting As Two-Way Restraints.)
12 - (2) Applicants, in line with their premise that all construction 13 at Comanche Peak is preliminary and finding that their.
Ik initial premise was incorrect, removed a number.of th'ese
-15!
1 -
supportsandredlacedthemwithnewsupports. (See page 5,
.l6 0 - paragraph 2, of Applicants' Motion for Summary Disposition of.
- 17- CASE's-Allegations Regarding U-Bolts Acting As.Two-Way.
- -
- 18 ~
Restraints.)
E191 , of the 8_U-bolts initially' analyzed in Applicants'
.20 - Motion for Summary Disposition, all 8 failed to act as p[21 assumed.
' 22 (3) Applicants commissioned further tests to determine the
, 323 4
.gg,eg, og two-way loads on U-bolts. These tests, which are i
24- . Ipso facto, are irrelevant to qualify erroneous decision
!25- which were made-and resulted in unqualified construction.
y i
61 P
4 I
- u. ,p .,4 ,, ,e.... .,_~4-. L- . - , , , ,. 4 ,,v ( ,e-ep.wyey re.+,,.--.,w.-<.,--.ve..-.9mw, -rm.,w-y-a,,.4>e,e-y-.- .- -,,, yn - e e -t w. 4 -..e r --
1
'l '
Beyond this, the term " failure" is not.used by me as meaning 2 collapse, break, etc.; it is rather a generic term used in 3: the industry indicating a failure to meet criteria to which 4 Applicants are committed.
5 ,(4) In the affidavit by Messrs. Finneran and Iotti at page 3 (of 6 Applicants' Motion for Summary Disposition),'they state 7 without qualification that the clearance for U-bolts to pipe
- 8. is 1/16". But I must mention that if the pipe is 4" or less, 9- the clearance is only 1/32" and these clearances are always 10- assuming' precision construction with no tolerances. The
= .1 l affidavit at page 6 by Messrs. Finneran and Iotti states that 12 of the 70 supports of this type (U-bolts), most of them are 13 on lines 6" and under.
14- (5) The affidavit of Messrs. Finneran and Iotti, at page 8,
-15 states that a rerun of the problems with U-bolts input as 16 two-way constraints produced no appreciable change in loads; 17- however, if one-studies the tables that resulted from these 18 reruns, the problems become obvious, as discussed in the 19 following:
- 20 . Table 3 (Attachment to Finneran/Iotti Affidavit),
21~ problem AB-1-62E, second page, support No. CC-1-007-025-22 A43R, shows that the load increased 29%.
23 . Page 4 of this Table, for Support No. CC-1-007-039-A43R,
- 24. -shows that the load increased 31%.
-25 . ? age 5 of Table 3, for support CC-1-007-704-A43R, shows 62
I 'that the load' increased 28%.-
2 . Table 4, problem AB-2-63B, at page 3, support No. CC 3 126-010-F43R, shows the load increased 26%.
4 .. -CC-2-126-007-F43R, shows a load increase of 61% (2200
-5' lb. increase).
6
- Page 3, CC-2-126-005-F43R, shows a load increase of 21%
~
plus a lateral load not existing before of 952 lbs.
0' .- Support CC-2-126-006-F43R, shows a load increase of 25%
9 plus a 318 lb. lateral load not existing before.
10- . CC-2-126-011-F43R shows.a load increase-of 11% plus a 61 Il ib. lateral load not existing before.
12 . On page 4 of Table 4, support CC-2-164-407-A63K, shows a 13 load increase of 27%.
-l4 The supports with increases selected above only included 15 those with loads that increased 20% or more or where the 16 increase was 10% and lateral loads were indicated that had l7- not existed previously. As may be noted, nine supports
'l8 suffered significant increase, and additionally, at least one l9 support absolutely failed (quite possibly three, since
.20 heavy-duty U-bolts are generally not used when a PUS type U-
~
21 bolt would only be loaded to approximately 1/3 of its 22 capacity for the old loads issued 1/17/83).
23 It should be noted that these significant increases and 24 the failure and potential failures were not mentioned in 25 Applicants' affidavits. For that matter, Dr. Iotti and John 63
... y
- ~,
- 1 -- Finneran. time after time in their affidavit assure us we were 2 not even approaching an area of problem.
f:3? These supports may have been replaced under the guise of 4 conservatism, as may be found in Applicants' 5/23/84 Motion 5- '
for summary Disposition, at page 5, second paragraph, where 6' -it states:
T -"Despite Gibbs & Hill's re-evaluation which established that the system as-built would have been acceptable, 18: Applicants decided.to replace all U-bolts on rigid frames initially considered as one-way restraints where 9 piping thermal movements were computed to be equal to,
. . . or exceed, 1/16".in the original analysis."
'10 11 However, in the three supports which either fail or may "12 fail as -indicated above (cc-2-126-011-F43R, CC-2-126-005-
- 13. :F43R, CC-2-126-006-F43R), the maximum thermal movement L14' according to Table 1 was only .008. My assumption for the il5 above is-based on the fact that not only are this failure and fl 6 - these potential failures not mentioned in the text, but in n17 . Table 4, page=3, where these supports are listed and 18' '
asterisked, no mention of per cent change is indicated.
~
.19 In reality, the reason for replacing these U-bolts was 20 that the U-bolt was not qualified to accept a 952 lb. lateral 21 load for normal upset for support cc-2-126-005-F43R, a 318 h
22 lb. lateral load combined with the normal load for support 123 cc-2-126-006-F43R, and a 61 lb. lateral load for normal upset
- 24 for support cc-2-126-011-F43R (I am assuming that the 25 change resulted since this was a light-weight U-bolt -- in
. c 64
3_ .s ;.
[=n p
61' ' any event, there is no doubt that one support absolutely
~
~2
, failed to meet the criterion, CC-2-126-005-F43R), even
- 3 assuming;that this was a heavy-duty U-bolt for this 12" line.
14-
.See CASE Exhibit 669B,-items- 13-0, 130, and 13R.
- 5 ~
Also, from Table 2, Problem AB-1-65, the following
-6; supports did not ' react in'the manner assumed. CC-X-013-012-17- -
A43R picked up a lateral load of 232.lbs. at normal upset load conditiens. CC-X-023-014-A43R picked up a 71 lb.
}9 lateral load, which did not exist previously. CC-X-023-015- ,
- 30 -
3 A43R picked up an 88 lb. lateral load. These lateral loads 11I represent a significant ratio of the lateral capacity of the 12- "
U-bolt.
13 In addition, another. support, CC-X-023-011-A43R,
~
=
indicates that an additional load of 2300 lbs. was picked up.
T 11 5 '
~
in the Fx direction,which did not exist in the previous
~~16 i s
analysis.
il7~
--In no case are the' effects of friction on these U-bolts-
$18 mentioned in the analysis by Applicants, which would increase
- 19.
the effects of interaction. Therefore, even this analysis is 120: '
non-conservative.
21; (6) This is yet another example of where the Applicants attempt 8
. 22- to prove a point invariably results in opposite conclusions.
=23' For example, in the cases above, Applicants actually prove
~ L 24
~that when node point restraints are properly input, 13% of-125 the supports analyzed are, at best, non-conservatively 65
- .' 3 1: designed.
2 Again, Applicants and NRC Staff have gone to school at the expense of, and under the tutorage of, CASE witnesses.
~
3 4 The attitude of the Applicants will illuminate the purpose for 10
.5 CFR Part 50, Appendix B, Criterion XVI, which states, in part:
6 " Measures shall be established to assure that conditions adverse to quality, such as failures, malfunctions, deficiencies, 7 devistions, defective material and equipment, and nonconformances are promptly identified and corrected. In the case of significant
'8l conditions adverse to quality, the measures shall assure that the cause of the condition is determined and corrective action taken
, 9. to preclude repetition."
' 10 ' -
See also 10 CFR 50.55(e)(ii), which states, in part:
'11
- .. . the holder of the permit shall notify the Commission of each deficiency found in design and construction, which, were it 12- to have remained uncorrected, could have affected adversely the safety of operations of the nuclear power plant at any time
,13 throughout the expected lifetime of the plant, and which represents: ...
-14 ~
~
"(ii) A'significant' deficiency in final design as approved and
- 15. released for construction such that the design does not conform to the criteria and bases stated in the safety analysis report or 15 "
construction permit. ..
17 'See also CASE's 8/22/83 Proposed Findings of Fact and Conclusions 118- - of' Law (walsh/Doyle Allegations),Section XXIX. Reportability of Non-19' Conforming Conditions.
-20 Also, ANSI N45.2 (CASE Exhibit 687, admitted at Tr. 6414), to
'21 which Applicants are committed, states that adverse items must be 22- ~ promptly identified and corrected as soon as practical.
~23' The Applicants' inability to conform with the above laws may be
'24 note,d from the following: At Question and Answer 16 in Applicants'
=25 Exhibit 142 (Applicants' Prefiled Testimony for September 1982 66
~.
1
- hearings), page 5, Mr. Finneran, Applicants' Pipe Support Engineer for 2
.the Pipe Support Engineering (PSE) Group, stated "the t.ue constraining E3 nature of the U-bolts is accounted for in the 'as built' piping
'4 analysis and' subsequent support review program. Y.odifications are 5
being made where this review indicates they are warranted." (But this
.6 means that corrective action cannot occur until after construction'of 7
the system.) Mr. Finneran's failure to comply with the requirements of 8-the law and codes is understandable. At Tr. 5004/9-12, Mr. Finneran
-9 conceded that he is not an expert in the codes. .
10 The NRC Staff, on this matter, plays a game of semantics by
~11 stating that none of the Walsh/Doyle allegations represented a 12 significant deficieacy (see NRC Staff Witnesses Tapia, Taylor, Burwell, Chen, and Rajan at Tr. 6682-6685). However, when 35 per cent of the 14 original supports cited in my original testimony (September 1982) had 15 to be rebuilt, the deficiency was significant.
On the matter of U-bolts designed as one-way constraints when
'17 actually operating as two-way constraints, the NRC Staff has the .
18 following to say on the subject (see SIT Report, Staff Exhibit 207,
=19:
page 30, fifth full paragraph):
20 "Furthermore, in the subsequent discussions with the Applicant,
' 21 - seismic displacement data at selected one-way U-bolt restraints were presented to the Special Inspection Team. These data 22 indicated that the8e di8 placements were less than about 1/32 inch.
Loads associated with these displacements are also negligible."
~
The NRC Staff continued (SIT Report, bottom of page 30 continued 24 top'of page 31):
5
"(1) Piping stresses due t'o restraint of up to 1/16 inch of this 67
~ -,
<:c -*
~s l
l type of thermal expansion movement are negligible for all piP e sizes.
2-
"(2) Lateral loads on the U-bolts due to thermal expansion movements of this' type of up to 1/16 inch are negligible for 4, .all pipe sizes when the relative flexibilities of the pipe
'and U-bolts are considered."
15-zHere again we have Applicants and the NRC Staff qualifying '
6-supports based on engineering judgement when in reality they had no concept of the structural system behavior or the load paths. History 8-
-.has proven that:a problem in fact existed and that this concept of
' 'I supporting pipes for one-way loads is in violation of the codes, the
- 10. 1,,,.and sound engineering fundamentals.
Because of the mass of facts contained in this affidavit, I will
- 12 - only present one more document by the Applicants to prove their case.
3 But'again, this report by Applicants did not precede the construction but.was generated to justify the construction. This last example has 115 to do with:
16 '3. Generic vs. Actual Stiffnesses in Piping Analysis a
<17 The Applicants' Generic Stiffness Study which was performed for 41 8 two as-bu.lt systems considered the effects of inputting the actual 19 stiffness of the supports on the loads at the node points. (See 120' . Applicants' Motion for Summary Disposition Regarding Use of Generic 21~ .Stiffnesses Instead of Actual Stiffnesses in Piping Analysis; see
- 22. .also Affidavit of W. Paul Chen on Open Items Relating to Walsh/Doyle 23" Concerns filed 10/14/83; and Attachment (Applicants' 8/17/83 Additional
~24 - Pipe Support ceneric Stiffness Study).
- L25- The Applicants' analysis df the actual stiffness was not complete, t
4 68 i:-
_f- e I since it failed to consider the effects of the anchor bolt, base plate, 2 and in some cases the U-bolts. Cygna's Dr. Bjorkman, in his testimony 3 during the April 1984 hearings stated (Tr. 12,865/16-21) that according 4 to actual tests:
5 ". . . for their particular test and their.particular configuration, a particular value of L that they chose. The 6- baseplate contributed no more than half to the total displacement.
In other words, it was approximately 50 percent or less to the 7 total."
8 Further, the Applicants failed to include such eccentricities as 9 are indicated on drawing N . oCC-2-011-001-A73R, CC-2-011-003-A73R, and
.10 CC-2-011-005-A73R -- this support also has about a 12 degree kick 11- angle. These supports are unstable structures which depend on the 12 : torsional stiffness of the pipe to establish the ultimate stiffness of 13 the support.
14 One fact that does come through loud and clear in the analysis 15 .(even though it-is incomplete) by the Applicants is that the generic 16 stiffness fact'rs o were not-represented in the installed supports, and 17 in fact, for the 6" line, the actual stiffness varied from 3.6 times
-18 the generic value to 1/70 of the generic value, and for that case 75%
19 .of the loads increased and were therefore nonconservative. The maximum 20 increase for one support was 200%; 20% of the supports exhibited load 21 changes greater than 25%. And it must be mentioned one more time, 12 2 these stiffness values did not include the effects of U-bolts, base 23 plates, anchor bolts, gaps, etc. The greatest load increase in terms 24 of actual load was one support, the load of which went from 824 lbs. to 25 1371 lbs. At one anchor, force and one moment also increased more than 4
69 b
1 Y
s A ' 25%. At the other anchor, all moments and two forces increased, and it must be recalled that this was for a system with only one support at
- 1/70 of the generic stiffness, and as shown above, the possibilities for many systems with one or more soft support (1/70) exists.
~
See W.
Paul Chen Affidavit on 0 pen Items Relating to Walsh/Doyle Concerns
~
0
- (under cover letter of 10/14/83), page 24, and Attachment (Applicants'
. 8/17/83 Additional Pipe Support Generic Stiffness Stud'y).
Recalculation'of support loads was also done for support No. CC 'I 107-008-E23R, which was a support that had a generic stiffness of 1/360 0
of the generie stiffness. The load increase in this particular support N was over 600% and'resulted in a redesign of the support by CMC 94130, 12 - July 29, 1983 (see 10/14/83 Chen Affidavit, pages 25 and 26; see also
~
3 dr wing, CASE Exhibit 669B, Attachment to Deposition / Testimony of Jack 4' Doyle, Item 11TT).
15 The fact that the effects of U-bolts, struts, etc., influenced the
- Ib' '
' dynamics of systems, while apparently not a concern of the Applicants or N NRC Region IV, is an express concern of the Nuclear Regulatory IO Commission. See.the Board's 12/28/83 Memorandum and order (Quality 9
Assurance for Design) at page 38, quoting Board Notification 82-105A,
~ 20 ty, p.g.,4.and 5, which states:
2[ "The dynamic-interaction between the pipe and pipe clamp is a 4 i,'22 complex design problem. From a design standpoint, there are many
. uncertainties that could affect the actual system response such as consideration of total support system flexibility, mechanical non- -
-- 23 itnearities,.' construction and installation tolerances, and
. uncertainties:in the dynamic loading itsdif. It is beyond the
. 24 , cop of this report to discuss the clamp-to piping responses to 4
these various factors. However, the report will focus on those 25 '. local dynamic effects on the piping that can be attributed i
70
- t
]
o ; ,
I primarily to the clamp attachment that, in general, are not explicitly evaluated by piping designers."
Therefore, the seismic analysis is rendered nonconservative due to the fact (as shown above) that the generic stiffness values are not representative of the supports as used at Comanche Peak.
The facts are that the use of generic stiffnesses represents a gross concern for the design of Comanche Peak and is in violation of the codes and laws to which Applicants are committed. On this item, while the NRC Staff attempts to derate the results of the generic sciffness analysis, it must also be pointed out that it was the NRC Staff who requested the study in order to justify the use of generic
- deflection in design. For ti.is they are to be congratulated.
The Applicants, on the other hand, engaged in a stiffness vs.
13-W frequency word game (see Applicants' Witness Vivirito, Tr. 7070/20- '
14 7071/10). In reading this section of the testimony, much verbage is devoted to comparing stiffness with frequency but no words are spoken which indicate the effects of the actual stiffness on loads - which 17 was the real issue. Applicants' diversionary tactics indicate clearly an unwillingness to be candid with the Board.,
Applicants' diversions continued with the statement by Mr.
.20 -
Vivirito at Tr. 7067/7-11: '
21
,t
". . . We have performed the study and it is a similar study to 22 -
what everyone has performed, and they came to the same conclusion, and so are we: that support stiffness variations have very little
'23 effeet on the dynamic response."
24 The Applicants have admitted that they employed "somewhat 25 knowledgeable" engineers. In my opinion, this'is a gross exaggeration '
71 s
d i4 . 4
O I - c N, of the Applicants' engineering qualifications. I say this because if 2: Applicants' engineers have any knowledge, they have failed to display 3 any tendency to exhibit such quality.
4 013. Do you have any further comments on the Comanche Peak QA/QC program?
'5 .A13. What-QA/QC program? -At Tr. 7192, Applicants' Witness Finneran stated 6: that for the 9-level' final vendor review program, only their most
(-
7 qualified personnel would participate. And yet with the oversight of
.<8 ;Mr.'vega and the most qualified engineers on site, 77-1/2% of the main
-9 steam support' calculations contained aross engineering errors, with 10J most of the calculations containing a multiplicity of errors. In fact, Il one support of the total of 22 main steam supports had such a gross 12 error that it failed and had to be rebuilt.
13 of the three major nuclear accidents that threatened the public I4 health and safety, all were initiated by minor incidents:
15 (1) Fermi 1 accident, caused by failure of a zirconium deflector 16 which wasn't really required in the first place.
17 (2) Brown's Ferry accident, caused by non-fireproof materials 18 ignited by a worker's candle.
19 (3) Three Mile Island accident, caused by failure of a 3/4 inch
-20 air' valve.
- 21. In view of the facts as stated above and the facts that:
- 22. (1) At least 6 of 9 of the calculations that I reviewed from 23 Phase 1 of the Cygna review had gross errors in the final
'24 vendorcertihiedcalculationswhichwerenotnotedbycygna.
'25 g. And again, 1 of the 9 supports structurally failed and r
72
+
- m. m ,
t .2 ' r e-
- o. tsl H,
[1 required _ redesign.
- 2 - -(2) Of the 89 supports (31 with calculations), 31 had to be 3 modified to prevent collapse.
4 When I proved that a fatal erre. existed in the upper lateral
.restrai , NRC Staff' Witness Dr. Chen stated at Tr. 6058 that he did 6- not consider one error in four calculations "to be that significant."
.77 I wonder:if Dr. Chen considers a maltiplicity of errors in each of 18 8 .- out of 22 supports to also be insignificant.
- 9 ~I-say again -- what 0A/QC program?
014.-Did you select the' main steam system to discuss because there are more
~
11 0 11 problems wich'it than with other systems?
'12: A14. No.. In a cursory review of the component cooling system (see Volumes 2 13 .and 3 of the Phase 3 Cygna Report, PS-001 to PS-066), I found that 7J[
?14 per cent'of the calculations contained aross ensineerins errors. I 15 also found that two (PS-017 and PS-026) of the supports would have 16- structural failures (same as the failure in. CASE Exhibit 928, Support
'l7 No.:SI-1-325-002-S32R). While this number is slightly less than found 18: in the main steam system, it is still very significant.
' 19 - -I~ chose the main steam becaue it was a more critical system and L20 was smaller (22 supports vs. 66 supports) and therefore easier to 121 detail in this affidavit.-
22 Beyond that, I didn't second-guess Cygna on generic calculational
.23 procedures which they wrote off (cinching, instability, double axial 24 restraints, etc.) since these issues are being handled in CASE's 25 Answers to Applicants' Motions for Summary Disposition. The only 73
- g. - e<
- . + \
.l errors I' cite here are'those cited by Cygna in the Phases 1, 2,'and 3 2: 'Cygna Reports.
31 015.LAre.there any problems which are of an interface nature on which you
-4' 1would like to comment?
5 .'A15. Yes. In Volume 1 of the Phase 3 Cygna Report (Section 1, page 4) under (6- PI-00-06, the observation is noted that in numerous instances (18 i
- 7' supports) for the main steam inside containment, the support loads used I8' in design did not match the values obtained by examining the computer 9- ' output. In one case (Support No.'MS-1-004-005-C72K), the snubber is 10: overloaded by using the method of analysis. historically used by Gibbs &
1.1 - .- Hill. .
12 -I have two statements to make on this point. First, the gravity 13 of the erroneous use of loads is not reflected in the pipe support 1l4 review, and second, this is not-in keeping with the intent of the final-15 : vendor, certification program as stated by Applicants' Witness Vega at 16c iTr. 7153 where he states that step 7 of the iteration compares loads
.t 117L and modifies them-if required.
=l8 _In all, at'least four supports are known to have required rework l9; and'one required reanalysis after the final design review for vendor
+ 20' certification was completed. This is in addition to the 31 which.were 21 redesigned as a result of the original Walsh/Doyle allegations. I must .
22 note that-this does not include the U-bolts which were replaced during
- 23- the reanalysis.for U-bolts used as one-way supports but acting as two-24 way constraints.
25 ;016.'could you briefly summarize the' main points contained in your 74 i
l'
.y' C V
~
=1 affidavit?
2 -'A16.-I'll try.. In'this affidavit, I have attempted to bring together the 3- waft and-weave wh'ich covers more then two years of hearings to present-4' a tapestry of Applicants' (generally with the approval of the NRC 5~ Staff) incompetence, evasion, and practices which are dangerous to the 6 health and safety of the public.
7' This affidavit covers four major phases that have occurred during 8 _these hearings:
9 (1) The filing of my deposition / testimony, September 1982.
~
~ 10 (2) Phases l'and 2 of the independent design review by Cygna 11 Energy Services.
12 =(3) Phase 3 of the independent design review by Cygna Energy 13 Services.
.14 (4) Applicants' filings of Motions for Summary Disposition.
15 The items shown above, taken individually or collectively, draw a 16 clear portrait of massive incorporation of errors in pipe supports
' 17 ' critical to the safety of Comanche Peak and, as a result, are directly 18 related to the health and safety of the public.
19 As a result of the condition of the engineering calculations for 20 the pipe supports after the final review certilcation, anyone but.the 21 "somewhat knowledgeable" must recognize that the status of engineering 22 at Comanche Peak at the time of the filing of the original allegations 23 was nothing less than a disaster.
24 The conclusions which I have drawn, based on documented evidence 25 (primarily supplied by Applicants, the NRC Staff, or Cygna Energy 75
,e -
~I ' Services) can be generally summarized as follows:
2 1. .The iterative design program which exists at Comanche Peak 3
consists of a design and check program followed by a nine-
- 1evel review program.
5 2. The allegations originally stated by me (and supported by 6
documentation) contained 52 node points involving 58 supports
.7 and related to the status of Comanche Peak to the last day of 0 my employment.
9- 3. Applicants' attempts to show that no problem exists relative
- 10- to CPSES~ generic concerns raised by Mark Walsh and me (the II- Walsh/Doyle allegations) actually prove the Applicants had no
'l2' knowledge of the level of problem prior to construction.
11 3- 4. Applicants' final vendor certified review program was l4 developed to insure that the contents of the design and L 15 analysis would be error-free.
16 5. Phases-1.and 2 of the review by Cygna Energy Serv. ices failed I7 to note at least 6 major engineering fundamental errors which l8 existed on 9 of the pipe support calculations reviewed by I3' CASE.
.20 6. Cygna Energy Services failed to note that one support of the 21 9 mentioned above suffered a structural failure on the 22 modification incorporated to insure stability.
23 7. In the Phase 3 review by Cygna Energy Services (main steam 24 supports), 77.5 per cent of the main steam support 25 calculations contained gross calculational errors.
76
n .
_, +
1 8. In the Phase 3 review by Cygna Energy Services, one support 2 had such gross calculational errors that the support failed
- 3' ~and was redesigned.
4 9. When 77.5 per cent of the calculations for one area and 67 5 per cent of the calculations in another have gross errors, 6 the accuracy of the remaining plant systems is suspect.
'7 10. When 31 per cent of the original supports alleged to have 8 serious faults are rebuilt.and 6-1/2 per cent of the 31 9 reviewed (by Cygna and later by CASE) had to be rebuilt, the
.10 safety of the plant is indeterminate.
I1 11. There is no doubt that the independent review process is 12 neither independent nor reliable.
13 12. The NRC Staff has failed in its assigned mission to insure
~14 the health and safety of the public.
15 13. The OA/QC program at Comanche Peak has not only failed, it is 16- , non-existent.
-17 14. Evasion of the truth by Applicants and their agents and the 18 NRC Staff has been rampant in the hearings before the Atomic 19 Safety and Licensing Board.
20 15. The review of three systems at Comanche Peak (residual heat t :21 removal, main steam, and component cooling), all of which
.'22 were final vendor cert'ified, resulted in a calculation error
-23 rate of 6 + 18 + 46 divided by 9 + 66 + 22 - 72 per cent.
24 (An error in this case could be 'one error or a multiplicity 25 of errors, because only the support was counted as an error 77
., .m
'I regardless of the number of errors.)
2 16. The U-bolt cross-bar/ clamp analogy used at Comanche Peak is 3' not standard industry practice.
4
- 17. Practices followed by engineering at Comanche Peak are in 5
non-compliance with the law.
6 18. The preceding analysis of the Cygna review proves beyond 7
doubt that there are major problems involving calculational 8
errors, including the fact that no less than four supports 9 required modification due to errors in the final vendor
- 10 certified calculation . . . identified by either Cygna or (in Il the event Cygna missed it) by CASE. Had CASE had access to 12 all of the calculations for Phase 2 of the Cygna Report, it
~13 is more than likely that we would have found more errors and I4 possibly more support failures, as was the case when we 15 reviewed the nine supports from Phase 1.
16- From the preceding, the Comanche Peak facility can only be 17 considered indeterminate at best and dangerous as a definite IO possibility. The fact that at a minimum 72 to 78 per cent of the 19 calculations which I reviewed contained gross errors -- even after some 20 ten levels of checking by Applicants, and that at a minimum four of 21 these supports required rework to prevent" failure, shows that 22 regardless of how many erroneous calculations can later be shown to be 23 within allowables, at least some of the about 90 calculations (or 24 neglecttodocalculations)provedfataltocomplyingwitit the intent 25 of the codes.
78
.m ,
.e n
yw I' The safety of this plant is therefore in doubt unless a 100 per
'2 cent reinspection - by parties not dependent on the nuclear industry
- 3. for their livelihood - identifies the massive errors which have been
-4 incorporc":ed, with oversights and proper corrective measures as 5- required to bring the plant up to the level of confidence required to 6 insure the public health and safety.
7 8
9 10-11 12
- 13 14 15 16
.17' 18 ,
19 20
^ 21 .
22 23 24 25 79 ,
, ,4 , ,e The preceding CASE's Motion for Summary Disposition was prepared under my personal direction, CASE Witness Jack Doyle. I can be contacted through
~
CASE President, Mrs. Juanita Ellis,1426 S. Polk, Dallas, Texas 75224,214/
946-9446.
My qualifications and background are already a part of the record in these proceedings. (See CASE Exhibit 842, Revision to Resume of Jack Doyle, accepted into evidence at Tr. 7042; see also Board's 12/28/83 Memorandum and Order (Quality Assurance for Design), pages 14-16.)
I have read the statements herein, and they are true and correct to the best of my knowledge and belief.
0 ao (Sig 2 #
) Jack loyle '[
Date:
W
/ /98Y STATE OF N\ c.. ,u c_ ,
COUNTY OF Ti C t l'. -
On this, the i ,d day of S,dr < , 1984, persona 11'y appeared Jack Doyle, known to me to be the person whose name is subscribed to the foregoing instrument, and acknowledged to me that he executed the same for the purposes therein expressed.
Subscribed and sworn before me on the \ d day of OO c- i ,
1984.
ME s ,,4 b c Notary Publjp in and for the,,7 State of u ,. ..~,3o -
My Commission Expires:
l t
- - . ~ . ..-,,..-4 -,,_.-,,----,_.-...----4, _ , , . . - - - - , -
-. c
,. w
.l o .
l UNITED STATES OF AMERICA
, NUCLEAR REGULATORY COMMISSION .
BEFORE THE ATOMIC SAFETY AND LICENSING BOARD In the Matter of }(
}{
TEIAS UTILITIES ELECTRIC }{ Docket Nos. 50-445-1 COMPANY, et al.- }{ and 50-446-1
'(Comanche PTak Steam Electric }{ !
Station, Units 1 and 2) }{
CERTIFICATE OF SERVICE By my signature below, I hereby certify that true and correct copies of CASE's Motion and Offer of Proof; and CASE's First Motion for Sumary Disposition Regarding Certais Aspects of the Implementation of Applicants' Design and QA/QC for Design -
have been sent to the naass listed below this '6th day of October ,193],
by:- Express Mail where indicated by
- and First Class Mail elsewhere.
- Administrative Judge Peter B. Bloch
- Nicholas S. Reynolds, Esq.
U. S. Nuclear Regulatory Commission Bishop, Liberman, Cook, Purcell 4350 East / West Highway, 4th Floor & Reynolds Bethesda, Maryland 20814 1200 - 17th St., N. W.
Washington, D.C. 20036
- Ms. Ellen Ginsberg, Law Clerk U. S. Nuclear Regulatory Commission
- Geary S. Mizuno, Esq.
4350 East / West Highway, 4th Floor Office of Executive Legal Bethesda, Maryland 20814 Director U. S. Nuclear Regulatory
- Dr. Emaneth A. McCollos, Dean Commission Mary 1 sad National Bank Bldg.
Division of Engineering, Architecture and Technology - Room 10105 Oklahoma State University 7735 Old Georgetown Road
. Stillwater, Oklahoma 74074 Bethesda, Maryland 20814 i *-Dr. Walter N. Jordan Chairman, Atoale Safety and Licensing 881 W. Outer Drive Board Panel Oak Ridge, Tennessee 37830 _U. S. Nuclear Regulatory Commission Washington, D. C. 20555 l
l 3
l l 1 k
. as Chairman Renes Hicks, Esq.
Atomic Safety and Licensing Appeal Assistant Attorney General
^
Board Panel Environmental Protection Division U. S. Nuclear Regulatory Commission Supreme. Court Building Washington, D. C. 20555 Austin, Texas 78711 John Colline Regional Administrator, Region IV U. S. Nuclear Regulatory Commission 611 Ryan Plaza Dr., Suite 1000 Arlington, Texas 76011 Lanny A. Sinkin 114 W. 7th, Suite 220 Austin, Texas 78701 Dr. David H. Bolts 2012'S. Polk Dallas, Texas 75224 Michael D. Spence, President Texas Utilities Generating Company Skyway Tower .
400 North Olive St., L.B. 81 Dallas, Texas 75201 Docketing and Service Section (3 copies)
Office of the Secretary U. S. Nuclear Regulatory Commission Washington, D. C. 20555
/
(jfris.) Juanita Ellis, President CASE (Citizens Association for Sound Energy) 1426 S. Polk Dallas, Texas 75224 214/946-9446 i
e 2