ML20080T132

From kanterella
Jump to navigation Jump to search
Testimony of RB Hubbard Re Design Qa.Verification Program Has Not Yet Demonstrated Plant Design Complies W/All Applicable NRC Regulatory Requirements & All Util Licensing Criteria.Related Correspondence
ML20080T132
Person / Time
Site: Diablo Canyon  Pacific Gas & Electric icon.png
Issue date: 10/16/1983
From: Hubbard R
CALIFORNIA, STATE OF, MHB TECHNICAL ASSOCIATES
To:
Shared Package
ML20080T133 List:
References
ISSUANCES-OL, NUDOCS 8310200178
Download: ML20080T132 (79)


Text

/~ ~ o J'i..

1 4

f #f 7 l $3 '

EE? D C0FT.E370'"CCE e

\_ .

i 1 UNITED STATES OF AMERICA UQ!

NUCLEAR REGULATORY COM. MISSION 2 ,

13 BCI 1921 :16 3 BEFORE THE ATOMIC SAFETY AND LICENSING APPEAL DOARD 0FFICE OF SEc[ET,t 4 CCClu.Ilui A sEs w:;

^"

5 }

In the Matter of )

J3 )

PACIFIC GAS AND ELECTRIC CO. ) Docket Nos. 50-275 O.L.

7 ) 50-323 0.L.

(Diablo Canyon Nuclear Power )

8 Plant, Unit Nos. 1 and 2 )

9 10 TESTIMONY OF RICHARD B. HUBBARD 11 REGARDING DESIGN QUALITY ASSURANCE 12 I.- INTRODUCTION 13 Q: Please state your name, address, and occupation.

14 A: My name is Richard B. Hubbard, and my business address is 15 1723 Hamilton Avenue, San Jose, California. I am 16 vice-president of MHB Technical Associates.

17 Q: Which of your qualifications and experience are relevant to 18 the design quality assurance (QA) matters you address in this 19 testimony?

20 A: I am a Professional Quality Engineer licensed by the State of 21 California (license number QU 805) . I hold a B.S. in 22 Electrical Engineering from the University of Arizona (1960) 23 and an M.B.A. from the University of Santa Clara (1969) . I 24 have nineteen years' experience in the design and manufacture

. 25 of systems and equipment for nuclear power generation 26 facilities, including eleven years' experience in responsible 27 elgineering and manufacturing managerial positions in the 1.

8310200178 831016 O PDR ADOCK 05000275 T PDR s

9' . 4 .

1 Nuclear Instrumentation Department (1965-1971), Atomic Power 2 Equipment Department (1971-1975), and Nuclear Energy Control 3 and Instrumentation Department (1975-1976) of the General 4 Electric Company (GE) . For the past seven years, I, along 5 with my co-founders of MHB Technical Associates, have 6 conducted numerous studies pertaining to the safety, quality, 7 reliability, and economic aspects of nuclear power facilities.

8 From November 1971 to February 1976, I was a Manager of 9 Quality Assurance for the manufacturing operations at the 10 San Jose, California, headquarters of GE's Nuclear Energy 11 Division. I was responsible for the development and 12 implementation of quality plans, programs, methods, and 13 equipment to assure that equipment for nuclear plants 14 designed, manufactured and procured by General Electric met 15 quality requirements as defined in NRC regulation 10 C.F.R. 16 Part 50, Appendix B (" Appendix B"); ASME Boiler and Pressure

17 Vessel Code; customer contracts; and GE corporate policies 18 and procedures. The product areas include radiation sensors, 19 reactor vessel internals, fuel handling and servicing tools, 20 nuclear plant control and protection instrumentation systems, l

21 and control room panels for the Nuclear Steam Supply System 2u (NSSS) and Balance of Plant (BOP). I was responsible for 23 approximately 45 exempt personnel, 22 non-exempt personnel, i

l 24 and 129 hourly personnel with a yearly expense budget of 25 nearly 4 million dollars and an equipment investment budget 26 of approximately 1.2 million dollars. While employed by

! 27 General Electric, I was responsible for developing a quality i

L 2.

F_, ,.

1 system which received NRC certification in 1975. The QA 2 system was also successfully surveyed for ASME "N" and "NPT" 3 symbol authorizations in 1972 and 1975, plus ASME "U" and "S" 4 symbol authorizations in 1975. I was also responsible for 5 the quality assurance program and its implementation at GE's J3 spare and renewal parts warehouse in San Jose. I am a member 7 of the IEEE Nuclear Power Engineering Standards Subcommittee 8 responsible for the preparation and revision of a number of 9 Quality Assurance standards for safety-related aspects of 10 nuclear power facilities.

11 Finally I have testified on safety-related aspects of 12 nuclear power facilities' quality assurance programs as an 13 expert witness before the NRC Licensing Boards; before 14 and at the request of the NRC's Advisory Committee on 15 Reactor Safeguards; before the Joint Committee on Atomic 16 Energy of the United States Congress; and before various 17 other federal and state legislative and administrative 18 bodies. A summary of r.y experience and professional

19 qualifications is set forth in the affidavit of 20 qualifications that is being filed with this testimony.

21 0: What sources of information have you relied upon in preparing 22 this testimony?

23 A: The facts and conclusions set forth in this testimony are l

24 generally based upon the information served on the parties in 25 the ongoing Diablo Canyon Nuclear Power Station, Units 1 and 26 2 (Diablo Canyon) , licensing proceedings that I have received 27 and reviewed between September 1981 and September 1983 in my 3.

pc -

.1 continuing role as technical consultant to counsel for the 2 Governor of California in these proceedings. I have reviewed 3 the semi-monthly status reports provided by Pacific Gas and 4 Electric Company (PG&E) and Teledyne Engineering Services 5 (TES) concerning the Independent Design Verification Program J (IDVP). Further, I have reviewed the Interim Technical 7 Reports (ITRs) issued by TES. In addition, I have reviewed 8 .the. IDVP Final Report and the Phase I and Phase II reports 9 released by the Diablo Canyon Project (DCP) resulting from 10 its Internal Technical Program (ITP). I have prepared and 11 submitted to the NRC detailed comments concerning 12 inadequacies in the proposed scope and methodology of the 13 Phase I and Phase II verification programs. I have discussed 14 these technical comments at the meetings between Mr. Denton 15 of the NRC Staff and the intervenors in the Diablo Canyon 16 proceedings _in San Francisco on February 17, 1982, and 17 September 9, 1982. In addition, I attended and made a 18 presentation on these matters to the NRC commissioners at a 19 meeting in Washington, D.C., on November 10, 1982. I 20 participated in a number of r.;eetings between the various 21 participants in the Diablo Canyon QA/QC investigations, 22 including on-site and off-site meetings with personnel from 23 PG& E , TES, NRC, Bechtel, Stone and Webster Engineering (S&W),

24 Robert L. Cloud Associates (RLCA), Roger Reedy Incorporated 25 (Reedy), and Brookhaven National Laboratory (BNL). Finally, 26 I am familiar with the PG&E license commitments set forth in 27 the Diablo Canyon Final Safety Analysis Report (FSAR) and the 4.

I 1 NRC staff reviews as documented in the Safety Evaluation 2 Report (SER) and its supplements, including Supplements 16 3 and 18 regarding the staff's most recent reviews of the plant 4 des ig n .

5 II. PURPOSE 6 Q: What is the purpose of your testimony?

7 A: The purpose of my testimony is to address the level of 8 assurance provided by the Diablo Canyon design verification 9 program conducted by the IDVP and ITP. Specifically, this 10 testimony addresses the matters set forth in Contentions 5, 11 6, 7, and 8.

12 Q: How is your testimony organized?

13 A: Part III of the testimony provides an overview of design 14 quality assurance. Terms utilized in the testimony are 15 defined and key assumptions are identified. My evaluation of 16 the matters encompassed by Contentions 5, 6, 7, and 8 are set 17 forth in Parts IV through VII of the testimony respectively.

18 First, my assessment of the effectiveness of the ITP's design 19 configuration control ef forto to assure that the as-built 20 Diablo Canyon plant conforms to the design documents is 21 delineated in Part IV. In Part V, the failure of the 22 verification program (i.e., the IDVP and the ITP) to reverify 23 a suitable sample of the design services subcontracted to 24 Westinghouse by PG&E is presented. The necessity for 25 identifying all the root causes which led to or provided the 26 basis for design errors discovered by the IDVP and ITP is 27 described in Part VI of the testimony, while in Part VII the 5.

I . .

1 adequacy of the ITP's quality assurance measures applied to 2 the design modifications developed since November 1,1981, is 3 reviewed. Finally, the conclusions resulting from my review 4 are summarized in Part VIII.

5 III. OVERVIEW OF QUALITY ASSURANCE / QUALITY CONTROL 6 Q: This testimony is about quality assurance. How is " quality 7 assurance" defined?

8 A: Appendix B uses the term " quality assurance" (QA) to comprise 9 "all those planned and systematic actions necessary to 10 provide adequate confidence that a structure, system, or 11 component will perform satisfactorily in service."

12 0: What does " quality control" (QC) mean?

13 A: Appendix B states that ' Quality assurance includes quality 14 control, which comprises those quality assurance actions 15 related to the physical characteristics of a material, 16 structure, component or system which provide a means to 17 control the quality of the materials, structure, component, 18 or system to predetermined requirements."

19 Q: What is " engineering assurance"?

20 A: Engineering assurance is a term often used to describe the 21 quality program measures applied by engineering to its design 22 control activities. For e? ample , the Chief, Engineering 23 Quality Control, develops and maintains the PG&E Engineering 24 Department quality control program.1/

25 /

26 27 1. Diablo Canyon FSAR, page 17.1-7.

6.

l

.1 Q: Does the preceding mean that not all QA program elements 2 required by the 18 criteria of Appendix B are necessarily 3 conducted by personnel of the QA organization?

4 A:- Yes. For instance, design verification is normally conducted 5 by members of the design organization, i.e. , the originating J engineer and the reviewing engineer. Other examples of QA 7 measures of ten not conducted by the QA organization include 8 procurement, special process specification and qualification, 9 document distribution'and records maintenance, control of 10 measuring and test equipment, and handling and storage of 11 material and equipment. In such cases, the QA organization 12 normally provides surveillance inspections and audits of the 13 organization performing the quality activity.

i 14 Q: What is an " audit"?

15 A: '" Audit" is defined in ANSI Standard N45.2.12 as 16 "A documented activity performed in accordance with written procedures or checklists to verify, by examination

_17 and evaluation of objective evidence, that applicable elements of the quality assurance program have been 18 developed, documented and effectively implemented in accordance with specified requirements. An audit should not 19 be confused with surveillance or inspection for the sole purpose of process control or product acceptance."

21 In general there are three major types of nuclear plant 22 design audits as follows:

l 23 (a) Program audit - an audit whose purpose is to compare 24 the design QA program to the applicable regulatory 25 requirements and Safety Analysis Report (SAR) 26 commitments.

27 /

7.

l

'2.._ _ _ . _.... __._.., _ ,_ . . . _ , _ . _ _ _ _ . _ , , , _ _ _ _ , _ _ . . _ , _ _ _ _ _ , _ _ _ , , _ _ _ _ _ _

l 1 (b) Process audit - an audit whose purpose is to ascertain 2 whether the existing design control procedures are 3 functioning properly and are being effectively 4 implemented.

5 (c) Product audit - an audit of the design documents whose

.@ purpose is to demonstrate that the design documents such 7 as specifications and drawings correctly reflect the 8 applicable regulatory requirements and the design basis 9 commitments set forth in the SAR.

10 Criterion 18 of Appendix B (" Audits") requires that a 11 comprehensive system of planned and periodic audits be 12 carried out to verify compliance with all aspects of the 13 quality assurance program and to determine the effectiveness 14 of the program. Thus, audits are intended to identify 15 conditions adverse to quality so that they can be corrected 16 and similar repetitive deficiencies precluded in the future.

17 0: What is an " error"?

18 A: In this testimony the term " error" will be utilized 19 consistent with the definitions adopted by the IDVP in 20 Appendix F of the Final Report as follows:

21 Error: A form of program resolution of an Open Item indicating an incorrect result that was verified as 22 such. It may have been due to a mathematical mistake, use of a wrong analytical method, omission of data or 23 use of inapplicable data. Each Error was classified as the most appropriate of the following:

(a) Error Class A: Design criteria or operating limits 25 of safety related equipment were exceeded and, as a result, physical modifications or changes in 26 operating procedures were required. . . .

27 /

8.

I

F. .

1 (b) Error Class B: An Error was considered Class B if design criteria or operating limits of safety 2 related equipment were exceeded, but were resolvable by means of more realistic calculations 3 or retesting. . . .

4 .(c) Error Class C: Incorrect engineering or installation of safety related equipment was found, 5 but no design criteria or operating limits were exceeded. No physical modifications were 4 required. . . .

7 Q: What is an Error Class AB?

8 A: On a number of occasions, the IDVP could not determine 9 whether resolution would or would not require physical 10 modifications, so the terminology Error Class A or B (ER/AB) 11 was used.

12 Q: Did the IDVP'also identify a category of discrepancies it 13 categorized as a " deviation"?

14 A: Yes. The IDVP defined a deviation as follows:

15 Deviation: A departure from standard procedure which is not a mistake in analysis, design, or construction. No 16 physical modifications are required. . . .

17 Q: Will you use the term " deviation" in your testimony?

18 A: No. In my judgment the term " deviation" cannot be precisely 19 differentiated from an " Error C." Consequently, contrary to f

20 the IDVP definition, it appears to me that a departure from a 21 standard procedure is in fact a mistake and should be 22 categorized as an Error. Further, my judgment appears to be 23 consistent with the OA measures of criterion 5 of Appendix B, 24 which requires in part that activities affecting quality 25 shall be prescribed by documented instructions, procedures, l

26 or drawings and shall be accomplished in accordance with I

! 27 these instructions, procedures, or drawings.

l 9.

l t

1 0: Does each design error disclosed by the IDVP or NRC audits 2 represent a multiple failure of the design control system?

3 A: Yes. Each error represents at least two failures; Lne 4 original failure or error itself, and the accompanying 5 failure in the QA program or its implementation which allowed 6 the original failure to remain undetected. An undetected 7 error may also represent a failure of more than two " gates"

~8 in the QA program. For example, an error may represent a 9 failure by the originating design engineer, a failure by the 10 verifying engineer to detect the error, a failure by the 11 engineering assurance organization to detect the error in its 12 surveillance activities, and a failure by the QA organization 13 to detect the error during its audits. In the preceding, the 14 term " gate" means a control measure in the design process at 15 which a design attribute is checked or verified for 16 conformance. Thus, the effect of the multiple gates is to 17 provide several opportunities to detect a design 18 nonconformance.

19 Q: Will a good QA program assure that design failures will be 20 totally eliminated?

21 A: No. Rather, in designing any complex facility, errors by 22 the originating engineer are inevitable because people are 23 not infallible. QA programs recognize human imperfections 24 and thus impose a management control system to detect 25 these inevitable errors and, therefore, to ensure that the 26 facility is, in fact, designed to the requisite licensing 27 criteria. ]

10.

l  !

)

\

1 Q: Will a good QA program detect all design discrepancies?

2 A: No. Clearly while the goal of a QA program for a nuclear 3 plant is to assure zero defects, it is equally clear that 4 some defects will escape detection. In my experience, the 5 gates in the QA program for safety-related items for a

,6 nuclear plant are designed to assure that all critical errors 7 (Classes A, B, and AB) will be detected. Further, it is 8 expected that the vast majority of major errors (Class C) 9 will be detected. A lesser detection rate is acceptable for 10 minor design discrepancies. Such a systematic concept for 11 classifying characteristics of error provides a rational 12 basis for designing the gates in a QA program. Thus, design 13 features with critical characteristics require special 14 emphasis in the QA program. Correspondingly, such a system 15 provides a reasonable basis for applying a lesser degree of 16 QA controls to design features which might result in a minor 17 error.

18 Q: Have you made any significant assumptions in preparing 1

19 this testimony?

20 A: Yes. Based on my review of the Board's August 16, 1983 21 order, and my attendance at the pre-hearing conference on 22 August'23 and 24, it is my understanding that the reopened 23 Diablo Canyon design quality assurance proceeding will 24 focus on whether the verification program (the IDVP and 25 ITP) has demonstrated that the design of Diablo Canyon is 26 now in compliance with the applicable NRC regulatory 27 requirements and the PG&E licensing criteria. The i

11.

i

F. .

1 licensing criteria (or commitments) are contained in the 2 PSAR, FSAR, Hosgri Report, SER, SER Supplements, and other 3 licensing documents including letters from PG&E to the NRC.

4 Consequently, my testimony will not address the question of 5 whether PG&E and its design subcontractors complied with the 6 Commission's design quality assurance regulations and 7 associated SAR commitments in the period prior to November 1, 8 1981. Rather, it is now assumed that the design QA prccess 9 and its implementation prior to November 1,1981, cannot be 10 relied upon to assure the adequacy of design.

11 0: Does the preceding assumption identify a fundamental question 12 you will *3 dress in the testimony?

13 A: Yes. In my judgment the fundamental question regarding the 14 Diablo Canyon design, which encompasses all the matters set 15 forth in the contentions, has as its essence one basic 16 question: Whether the IDVP and the ITP "after the fact" 17 verification efforts provide an equivalent level of assurance 18 regarding the design of Diablo Canyon as would have been 19 obtained by a QA/QC program conducted in a timely fashion in 20 compliance with the regulatory requirements of Appendix B.

21 IV. FAILURE TO ASSURE AS-BUILT PLANT CONFORMS TO 22 TO DESIGN DOCUMENTS 23 Q: What contention addresses the verification program efforts to 24 assure that the as-built plant conforms to the design 25 documents?

26 A: Contention 5. This contention, as set forth in the Board's 27 August 26, 1983 Order, reads as follows:

12.

,.7

[

l aThe verification program has not verified that Diablo Canyon Units 1 and 2 ' as built' conform to the design 2

~

drawings and analyses."

3 Q: .W hat was the initial concern regarding the matters 4 encompassed by tihis contention?

5 A: All reviews of PG&E's design control practices for its design 6 activities conducted prior to November 1, 1981 disclosed 7 numerous examples where the as-built Diablo Canyon plant 8 failed to conform to *he design documents. This pattern of 9 configuration control noncompliance was identified by PG&E in 10 its review in response to NRC Bulletin 79-14. A similar 11 pattern of differences between the as-built plant and design 12 documents was disclosed by the IDVP in the Phase I reviews 13 and by Brookhaven National Laboratory-(BNL) in its 14 independent analysis of the vertical response of the 15 containment annulus structure.

16 Q: Did other reviewers of the pre-November,1981 design 17 activities determine that design and modification; problems 18 indicate the need for improved engineering support?

19 A: Yes. The Institute of Nuclear-' Power Operations (INPO) 20 visited the Diablo Canyon site during the week of January 25, 21 1982. In its report dated February 12, 1982, INPO recommended 22 changes to the design change control practices as follows:

23 " Improve the existing moSification program to ensure that changes to the plant are controlled and performed 24 in a timely manner. Fc. ? example:

25 a. Complete revisions to affected documentation before modified systems,are returned to service.

b. Issue final as-built documentation and update 27 procedores as soon as possible.

13.

m_.-___.____ -.__-

1 c. Assign review and approval responsibilities for non-critical modifications to on-site technical support 2 department engineers." (Governor's Exhibit (" Gov.

Exh.") 11.)

4 Q: Did the DCP initiste changes in its design configuration 5 control practices for.the post-November 1981 design

.6 activities?

7 A: Yes. The engineering review of plant modifications resulting 8 from IDVP identified errors has been performed in accordance 9 with Engineering Manual Procedure 3.60N; PG&E Engineering 10 Manual Procedure No. 3.7, Rev. 5; PG&E Engineering Manual 11 Procedure No. 3.7, Rev. TR-9 ; Diablo Canyon Project 12 Engineer's Instruction No.13, Rev. 0; and Diablo Canyon 13 Project Engineer's Instruction No. 13. These procedures 14 require that Engineering review the result of construction 15 activities which differ frcm the Design Change Notice (DCN) .

16 Q: Did PG&P initiate other reviews of the as-built condition of 17 the plant?

18 A: Yes. A description of the design and construction 19 configuration control process as well as the proposed 20 as-built walkdown activities by the DCP was provided to 21 the NRC in a June 24, 1983, letter from Schuyler of PG&S to 22 Eisenhut.

l 23 Q: Have the preceding corrective measures fully resolved the 24 as-built configuration control problem?

25 A: No. The IDVP's recent review of the sample of design 26 documents resulting from the DCP's corrective action program 27 /

14, l

r

.o .

1 identified a number of' instances where the as-built plant 2 differed from the design documents.

3 Q: Were EOIs issued for these conditions?

4 A: Yes. The most significant EOIs resulting from the IDVP 5 corrective action review are briefly summarized in Table 8-1 6 which is appended to the testimony. Configuration control i

7 errors identified by the IDVP included the following:

8 (a) Differences disclosed between "as analyzed" and 9 "as-built" bolt sizes (EOIs 1120, 1121).

10 (b) Differences disclosed between "as-built" and i

11 "as-analyzed" instrument tubing support (EOI 1123).

12 (c) Design analysis finite element model of the control room 13 . slab used to generate Hosgri spectra not agreeing with i

14 the field verified location of the supporting wall 15 (EOI 1124).

, 16 (d) Incorrect valve modeling in DCP seismic reanalyses 17 (EOI 1133, 1135, 1137).

18 Q: Did the IDVP also identify other discrepancies which were not 19 the subject to EOI's?

20 A: Yes, the IDVP review of the DCP corrective action measures, 21 as summarized in the ITRs, identified a large number of 22 configuration control discrepancies. The discrepancies t

-23 disclosed in the ITRs are summarized in Table 5-1 which is 24 appended to this testimony. For reasons not documented by I

25 the IDVP, the majority of the discrepancies documented in 26 Table 5-1 were not the subject of EOIs. In my judgment, the 27 failure of the IDVP to initiate EOIs for these matters is a 15.

1 serious omission. Further, the failure to systematically 2 initiate EOIs appears to be contrary to the IDVP's procedures 3 for identifying and evaluating potential errors.2/

4 Q: Can you provide examples of configuration control 5 discrepancies identified by the IDVP which were not

.6 documented in EOIs?

7 A: Yes. Configuration control discrepancies between the 8 as-built plant conditions and the design documents identified 9 by the IDVP during its verification of post-November 1981 10 design activities are denoted with an asterisk in Table 5-1.

11 Some examples of such configuration control discrepancies in 12 design documents are the following:

13 (a) Pipe weight: A 2000 pound flow element (weig ht equivalent to a pipe length of about 2.7 times 14 pipe diameter) was not included in the DCP model (ITR 59).

5 (b) Piping geometry: The DCP coded one portion of 16 20-inch diameter pipe 3 feet shorter and another portion of 12-inch diameter pipe 4 feet longer, 17 than indicated by IDVP field verification (ITR 59).

18 (c) Support modeling: Support 55S/64R was modeled as a rigid + Y-directional support, whereas the IDVP 19 field verification found it to be a gravity support (ITR 59).

(d) Valve Modeling: The weights for valves LCV-113 and 21 -115 were modeled 45% low for valve bodies and 8%

low for valve operators. In addition, minor 22 differences in the DCP eccentricity calculations were noted (ITR 59).

24 2. The IDVP in its Phase II Program Plan stated that 25 "Open Item Reports are prepared for the purpose of reporting an IDVP response to a QA and Design Control 26 Practices deficiency, a violation of the verfication criteria or an apparent inconsistency identified in the 27 performance of the work."

16.

~

l' (e) Support locations: A 3-foot difference in location of one of the supports was noted (ITR 59).

(f) valve Modelings. The DCP analyzed Valve 8805B with 3 the operator in the vertical position, but an IDVP field verification found this operator to be in the 4 horizontal _ position. Also, differences were found for valve center of gravity locations for valves 5 8805A and 8805B and IDVP [ sic) and for operator support locations (ITR 59).

(g) The IDVP field varification noted a weld across the 7 top of a member attached to the process pipe.

The DCP drawing did not show this weld (ITR 60) .

8 (h) The IDVP field verification noted that one of the 9 four restraints comprising support #98/83 was a small box ' frame bilateral rather than a tee-shoe 10 and clamp assembly as shown on the DCP support drawing (ITR 60) .

11 4

(1) Unintentional restraints, as shown on the DCP 12 walkdown isometric and by IDVP field verification, were not explicitly addressed in the analysis 13 (ITR 61).

14 (j) DCP sketches and as-built data did not correlate with the support analysis. In addition, DCNs for 15 modifications were omitted from the documentation

!. package (ITR 63).

(k) Two of the bolts in the four bolt plate joint 17 between two column members had been cut out to prevent pipe movement interference. The impact of 18 the reduced section was not evaluated (ITR 65).

19 Q: Based on the foregoing, what do you conclude?

CO A: The IDVP's reviews to date of a sample of the product (the r

-21 design documents) relating from the QA/QC process for the 22 DCP's corrective action measures demonstrate to me that 23! configuration control deficiencies continue to exist at 24 Diablo Canyon. Such configuration differences between the 25 as-built plant and the design documents are, in my judgment, 26 contrary to the design control and document control 27 requirements of Criteria 3 and 6 of Appendix B. The 17.

1

,_ .. . _ _ _ - - , . _ - , - . - _ , , , _ _ . . . - , - . . - - , , , , . , . , , . - - . . . , , - . , , , . . . _ , , , . , - . ._.._,.,.,-,,.___,.,-.m.,-_.

1 configuration control deficiencies also indicate a failure 2 to comply with the requirement of Criteria 10 and 11 that 3 inspections and tests be conducted to verify conformance with ,

4 drawings in that the proper conduct of such tests and )

5 inspections would not result in differences between the 6 as-built plant and design documents remaining undetected.

7 Finally, the continued existence of discrepancies between the 8 "as-built" and "as-designed" configuration of the plant ,

9 indicate that, contrary to Criterion 16 of Appendix B, the 10 corrective actions by the DCP have not been adequate to 11 assure that all conditions adverse to quality are identified 12 and corrected, and that the cause of the discrepancy is 13 determined and action taken to preclude repetition of similar 14 discrepancies.

15 Q: What do you now recommend?

16 A: As a minimum, the verification program should take the 17 following steps:

18 (a) Examine the numerous past and current examples of 19 known discrepancies between physical configuration 20 and design documents, determine the root causes for 21 those discrepancies, and make all changes in design 22 documents and physical installations required by 23 that analysis.

24 (b) Once it is believed that all root causes have been 25 identified and all resulting discrepancies located 26 and corrected, the conclusion should be confirmed 27 /

l l

18.

g e ,- ----

I 1 by examining a random sample of installations and 2 verifying that they conform to the design 3 documents.

4 (c) Modify the procedures for the design-construction 5 configuration control interface to insure that all 6 deviations from design documents made by 7 construction are promptly examined, approved and 8 documented by engineering in compliance with 9 regulatory requirements.

10 V. FAILURE TO REVERIFY WESTINGHOUSE DESIGN SERVICES 11 Q: What contention addresses the verification program efforts to 12 reverify the design services subcontracted by PG&E to 13 Westinghouse?

14 A: Contention 6. This contention, as draf ted by the Board in 15 its August 26, 1983 order, reads as follows:

16 "The verification program failed to verify that the design of safety related equipment supplied to PG&E by 17 Westinghouse met licensing criteria."

18 0: To what extent did the ITP reverify the safety-related design 19 activities performed by Westinghouse for Diablo Canyon?

20 A: The ITP, with assistance from Westinghouse, conducted a 21 limited verification of the design of Westinghouse supplied 22 equipment. The ITP review of Westinghouse was characterized 23 by PG&E as follows:3/

24 /

25 l 26

3. PG&E's Answer to Governor Deukmejian's Third Set of 27 Interrogatories, September 19, 1983, page 53.

19.

l -.- .. ..

1 INTERROGATORY NO. 51:

2 With respect to contention 6, do you deny that the verification program failed to verify that the design of 3 safety related equipment supplied to PGandE by Westinghouse met licensing criteria?

5 RESPONSE TO INTERROGATORY NO. 51:

4

~6 No. The seismic design of all safety related equipment furnished by Westinghouse was not reanalyzed.

7 Whenever findings of the verification program altered the input to specific pieces of safety-related 8 equipment, that equipment was requalified by Westinghouse and reviewed by the DCP.

10 Q: Was a similar interrogatory addressed to the IDVP?

11 A: Yes.

12 Q: What was its response?

13 A: The IDVP responded to Governor's interrogatory as follows:d/

14 ANSWER TO INTERROGATORY NO. 51 15 Since the IDVP did not review any verification that the ITP may have performed of the design of 16 safety-related equipment supplied to PG&E by Westinghouse, the IDVP neither admits nor denies the 17 portion of contention 6 that relates to any such activities by'the ITP. With respect to activities 18 by the IDVP, although the IDVP verified the Westinghouse-PG&E interfaces within the scope of the 19 IDVP programs, it did not verify the design of safety-related equipment supplied to PG&E by 20 Westinghouse."

21 Q: How would you characterize the extent to which the IDVP and 22 the ITP reverified the safety-related design activities 23 performed by Westinghouse for Diablo Canyon?

24 /

25 /

26

4. IDVP's Answers to Governor George Deukmejian's Third Set

. 27 of Interrogatories, September 21, 1983, page 48.

20.

i

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

.- ',c 1 A The preceding statements by PG&E and the IDVP indicate there 2 is general agreement that neither the IDVP nor the ITP 3 conducted a systematic verification of the design of 4 -

Westinghouse-supplied safety-related Nuclear Steam Supply 5 Equipment.E/ In addition, the IDVP did not perform a 6 verification of the Westinghouse design services summarized 7

in Appendix A of ITR-9. Rather, the IDVP conducted only a 8 limited review of the Westinghouse-PG&E design interface.

9 For example, with respect to seismic design, when the IDVP 10 examined the transmittal of Hosgri spectra it only verified 11 on a sampling basis that the applicable spectra were actually 12 used for equipment qualification. Similarly, the IDVP review 13 of the non-seismic safety aspects of the Auxiliary Feedwater

-14 System design, as well as the Reedy Phase II QA audit, failed 15 to involve anything more than an examination of the design 16 interface between PG&E and Westinghouse.5/

17 Q: Did the limited IDVP review disclose potential design errors 18 .in the Westinghouse activities for Diablo Canyon?

19 A: Yes, there is evidence that design errors have remained 20 undetected by the Westinghouse QA program. For example, the 21 vertical spectra used by Westinghouse for qualifications of 22 the accumulators is in error. For the vertical direction, 1

23 Westinghouse used two-thirds of the tau filtered spectra, 24 rather.than two-thirds of the unfiltered spectra as committed 25 26 5. Also, see IDVP Final Report, pages 4.1.4-3.

L 27 6. IDVP . Final Report', Section 4.1.3 and ITR-ll.

21.

_ , _7.

1 to at page 4-3 of the Hosgri Report. Further, in the BNL 2 review of ITR-11, BNL reviewers noted that errors were 3 d.isclosed in 30% of the Westinghouse samples examined by the 4 IDVP. Therefore, BNL questioned, as I question, the adequacy 5 of the IDVP's verification of Westinghouse seismic design

,,6 activities as follows:1/

7 "Further, the large percentage of exceptions (30%),

where Westinghouse qualification spectra did not 8 completely envelope the Hosgri spectra, would warrant additional samples if a complete check of the spectra 9 criteria was intended."

10 However, there is no evidence that TES implemented the BNL 11 suggestion to conduct additional sampling.

12 0: Was the IDVP review of Westinghouse design activities further 13 limited?

14 A: Yes. For example, the verification of system design 15 pressures and temperatures for safety-related systems, 16 including its use in equipment specifications, resulting from 17 a generic concern were not included in the IDVP's additional 18 verification program for items within the Westinghouse design 19 scope, but rather were limited by the IDVP to PG&E design 20 scope systems 8/

21 0: Did the ITP's limited review also reveal design errors in the 22 Westinghouse activities?

23 A: Yes. The seismic review of the main control boards (MCB) 24 conducted by Westinghouse in response to new spectra for the 25

7. Summary and Evaluation Report, ITR-21, TR-5511-2, Rev. O, 26 Brookhaven National Laboratory, November 2, 1982. (Gov. Exh. 12.)

27 8. SER Supplement 18, page C.4-25.

22.

1 auxiliary building developed by the ITP identified an error 2 in the original seismic qualification analysis. The MCB was 3 procured by Westinghouse from Reliance, and Reliance used a 4 private consultant to seismically qualify the MCB by 5 analysis. The original analysis in the early 1970's 6 predicted the lowest natural frequency of the MCB to be above 7 70 Hz based on the analytical model used. In the current 8 evaluation process, the MCB was modeled using field 9 measurements and results of in-situ tests. The in-situ tests 10 pointed out the existence of natural frequencies between 15 11 to 28 Hz which is much below the 71 Hz calculated originally.

12 Decause of this error, and because of the severity of the 13 current Hosgri spectra at the base of the MCB in the lE to 33 14 Hz range, Westinghouse has provided modifications to the MCB.E/

15 Q: Based on the foregoing, what do you now conclude?

16 A: It is evident that the conclusions resulting from the IDVP 17 and ITP reviews on samples of other design service 18 contractors cannot be extended to provide meaningful 19 conclusions as to the adequacy of Westinghouse-supplied NSSS 20 equipment or of the adequacy of Westinghouse design services.

21 Further, Westinghouse was the responsible design organization 22 for over 70% of the Diablo Canyon safety-related systems. As 23 NSSS contractor, Westinghouse had responsibility to develop 24 and implement the majority of the non-structural Diablo l~

25 26 9. SER Supplement 18, Section 3.5.3. Also, cee transcript of Westinghouse /PG&E/NRC meeting on May 20, 1983, regarding 27 seismic qualification of the MCB.

23.

l l

1 Canyon safety features committed to by PG&E in the FSAR and 2 other licensing commitments provided in response to the NRC 3 regulations. In particular, Westinghouse supplied the Diablo 4 Canyon designs provided to assure compliance with a 5 significant number of the General Design Criteria set forth 6 in Appendix A to 10 C.F.R. Part 50 of the NRC regulations.

7 Q. What do you recommend?

8 A. What is needed is a systematic verification of Westinghouse 9 design activities. The verification should include a 10 suitable sample of Westinghouse design documents and 11 activities sufficient to assure that all Diablo Canyon 12 licensing criteria are met, to assure the efficacy of the 13 Westinghouse QA process, and to assure that all basic causes 14 and generic implications of any errors detected have been 15 thoroughly assessed.

16 VI. FAILURE TO IDENTIFY ROOT CAUSES 17 Q: What contention addresses the verification program efforts to 18 identify the root causes of the design errors detected by the 19 verification program?

20 A: Contention 7. This contention, as set forth by the Board's 21 Order, reads as follows:

22 "The verification program failed to identify the root causes for the failures in the PG&E design quality 23 assurance program and failed to determine if such failures raise generic concerns."

25 Q: How do you define " root cause"?

26 A: " Root cause" is defined as the underlying basis that precedes 27 and usually induces an effect or result.

24.

I l

L , . _ _ . ~ - - _ - - _ . . . _ _ . _ . , _ . _ __ _ _ _ .

1 Q: Does the term " root cause" mean the same as " basic cause" as 2 used by the IDVP?

3 A: Yes. The IDVP in Section 6.3 of its Final Report defined 4 basic cause as "the underlying problem or concern which led 5 to or provided the basis for an identifiable error of

,p commission or omission," which is equivalent to the preceding 7 definition of root cause.

8 Q: What does the term " generic concern" mean?

9 A: " Generic concern" refers to the potential of each error to 10 exist in a similar manner in other, unreviewed parts of the 11 plant. Thus, the corrective action verification for a 12 potential generic concern should be conducted to the depth 13 and extent required to ascertain whether the specific error 14 is one of a number of similar errors in other, unreviewed ,

15 items of the Diablo Canyon plant.

16 Q: Were generic concerns identified by the IDVP?

17 A: Yes. Generic concerns were purportedly identified by the 18 IDVP "where one or more specific errors had been identified 19 or because the IDVP believed that a generic concern could 20 exist even though a specific concern was satisfactorily 21 resolved."AS/ Thus, the term generic concern, when used by 22 IDVP, is intended to indicate that the error (Class A or B) 23 is "potentially applicable to structures, systems, or 24 /

25 /

26 27 10. IDVP Final Report, Section 3.5.5, page 3.5-4.

25.

1 components in addition to that for which it was first 2

identified.11f 3 Q: Was an assessment of the basic cause of the identified errors 4 required of the IDVP?

5 A: Yes. The Commission order and the November 19, 1981 Staff 6 Letter both require that the IDVP conduct an assessment of a

7 the basic cause of all the design errors identified by the 8 IDVP. The Staff and the Commission required that the IDVP 9 provide the following three part assessment for each 10 identified design error:

11 "A technical report that fully assesses the basic cause of all design errors identified by this program, the 12 significance of design errors found, and their impact on plant design." See Commission Order, Attachment 1, at 13 part 1(a)(5)(b) and Staff Letter at parts 1(b), 2(b),

and 3(b).

14 15 C: Where is the IDVP assessment documented?

16 A: The IDVP Final Report in Section 6.0 contains its evaluation 17 with the basic cause of design errors documented in a very 18 general manner in Section 6.3. The significance is set forth 19 in Section 6.4, while the impact is briefly discussed in 20 Section 6.5.

21 Q: Did the ITP also provide a general statement describing its 22 determination of the basic causes of the identified errors?

23 A: Yes. The ITF Mocumentation of basic causes is provided in 24 Section 1.8 of the Phase I Final Report and Section 3.0 of 25 the Phase II Final Report. In no case,however, did the ITP, 26 27 11. IDVP Final Raport, Section 5.1, page 5.1-2.

26.

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

1 or the IDVP, correlc'e the basic causes cited to the 2 identified errors.

3 Q: Should the IDVP and ITP have made such a correlation?

4 A: Yes. The IDVP's and the ITP's failure to make this correla-5 tion is contrary to the corrective action requirements of 6 Criterion 16 of Appendix B.12_/ -

7 Q: Why is this so?

r 8 A: Criterion 16 requires that a QA audit ascertain the causes of 9 QA program failings so that an appropriate corrective action 10 program can be devised. Part of any prcper corrective action i

11 program is a determination as to whether the observed failure 12 has generic implications. Fundamental to any investigation 13 of the generic implication of any QA failure is a 14 determination of the root cause of that failure. It is only 15 when the root cause of a failure is identified that the 16 question of its generic implications can be addressed.

17 Instead of analyzing the root cause of each design error it 18 uncovered as a mechanism toward assessing the generic 19 implications of that error, the IDVP and the ITP provided no 20 more than global conclusions regarding basic cause with no 21 specific reference to any of the identified errors. Further, 22 the global basic causes identified by the IDVP and the ITP 23 primarily relate to the seismic errors. Thus, the multiple 24 failures (basic causes) which resulted in the non-seismic 25

12. The failure to develop the correlation is also contrary 26 to the requirements for the assessment of basic causes set forth in the Commission Order and the Staff's November 19, 1981, 27 Letter. l 27.

er -

7=-- - -.4.- --

..--_,s_.,---

s., p e- , 4 , ,_..,-,.,e,- -,,-,.,.--,-,,,----,,-m .--..--~.e .------,,-----,---.--,-m- - . - - - - - - , . - - . - - -

1 design errors were not systematically addressed. This is a 2 serious omission.

3 Q: What do you mean by multiple basic causes?

4 A: In general each error identified by the IOVP was the result 5 of multiple causes. For example, as discussed in Part III of 6 this testimony, each design error detected by the IDVP or ITP 7 represents at least two failures: the failure itself, 8 and the accompanying failure or failures in the QA gating 7

9 program or its implementation which allowed the original 10 failure to remain undetected. Thus, an adequate evaluation 11 of the basic cause must address the cause underlying each j 12 failure.

i 13 Q. Did the ITP or IDVP assess the potential generic concern 14 which can result due to the . failure to establish or implement 15 the QA/QC measures required by Appendix B?

16 A. No. This is a serious deficiency in the. verification 17 program's assessment of basic causes.

18 Q: Is it possible to go back in time to ascertain the cause of a l

19" design error?

20 A Yes. Indeed, Criterion 16 of Appendix B requires that the 21 causes of conditions adverse to quality be identified and 22 corrected.

23 Q:' Can you provide an example of how to conduct the two part 24 assessment of basic causes for a particular error?

25 A: Yes. The original mirror image design error provides such an 26 example. The error was that the diagram used to locate l

27 Vertical Seismic Floor Response (VSFR) for the Unit 1 ,

t

28. j t

i

~ . . - e ,e. - +

- -w F - c r.e--.. ,.t,-,-.ws,r ..,e....., ,,,-,,r, ,4 ,~.,,-y. , ,m, .,...-.-.c%-.--, -r,

, - --r_-w, i, -- - , . - , - - - ,rw-.

o .

1 containment annulus was applicable to Unit 2 but was 2 identified as being that of Unit 1. Since the units are 3 opposite hand, this resulted in an incorrect orientation of 4 VSFR spectra for Unit 1 component and system design. The 5 origin of the error was in the PG&E submittal to its 6 principal seismic design subcontractor, John A. Blume and 7 Associates (Blume), of an unverified, unlabeled, handwritten 8 shetch of the Unit 2 opposite hand geometry in place of the 9 Unit 1 geometry.12/ Blume personnel further compounded the 10 " sketch" error by one of their own. Blume assumed that the 11 layout of the annulus areas of Units 1 and 2 were identical 12 when, in fact, they were mirror images.11/ In my judgment 13 the underlying cause of the initial failure was the breakdown 14 of design interface control in that unverified and 15 uncontrolled design data were provided to Blume contrary to 16 Criteria 3, 4, and 5 of Appendix B regarding design control, 17 procurement document control, and procedural control.

18 The initial failure was not detected because the design 19 control and document control measure failed to assure, as 20 required by Criteria 3 and 6, that documents were reviewed 21 for adequacy and approved for release by authorized personnel 22 and were distributed to and used at the location where the 23 prescribed activity was performed. In addition, Blume was 24 not contracturally obligated by PG&E to a QA program until 25

13. LER 81-002/01T-0, October 12, 1981. (Gov. Exh. 13.)

26

14. NRC Inspection Report 50-275/81-29, page 2. (Gov.

27 Exh. 14.)

29.

1 1978, eight years after Appendix B was adopted and twelve 2 years after Blume's first engineering services on Diablo 3 Canyon, contrary to the requirements of Criterion 2 that a OA 4 program be established "at the earliest practicable time."

r 5 Also, PG&E's qualification and evaluation of service 6 contractors as required by Criterion 7 did not occur until

~

7 after completion of the subject engineering. Finally, audits 8 of suppliers, such as Blume, were not carried out in a timely 9 fashion to verify compliance with the QA program requirements 10 and to determine the effectiveness of the program as required 11 by Criterion 18. Thus, the appropriate corrective action 12 measures as set forth in Criterion 16 were not initiated by 13 PG&E or its subcontractor.

14 Q: Did the IDVP or ITP conduct such a two part evaluation as you 15 have suggested?

16 A: No. In no design error did the IDVP or ITP specifically 17 identify and document the second failure or failures in 18 the quality assurance program or its implementation 19 which allcwed the initial failure to remain undetected.

20 Further, in most cases, neither the IDVP nor the ITP 21 provided documentation identifying the underlying cause, 22 ,

or series of causes, leading to the initial failure.

23 Q: Do you have any examples to illustrate the weaknesses you 24 described in the IDVP and ITP treatment of basic cause?

25 A: Yes. The resolutions of EOI's No. 7002, 8010, 8017, 8022, 26 8023, and 8060 all demonstrate a failure to completely 27 address the basic causes of the identified errors.

30.

m 4 w f y - , q - ,p- , - - , , - . ,-,-,--s w- m . ,, ,,m.- -_,_m-- - , , - , .n -, y-- . ,

1 Q: How does the resolution of EOI 7002 exemplify a failure to 2 address the basic cause?

3 A: EOI 7002 resulted from an R. F. Reedy finding that no 4 objective evidence was available to demonstrate that the 5 effects of jet impingement on components inside containment 6 had been considered.dE/ In order to resolve EOI 7002, the 7 DCP developed the analysis of jet impingement inside 8 containment as committed to in FSAR Section 3.6, paragraph 9 3.6.15/

10 The above resolution, while it may address the specific 11 error which led to the establishment of EOI File No. 7002, 12 does not address the question of why no documentation of the 13 jet impingement review was available when the FSAR stated 14 that a review had been performed.

15 PG&E, in its Final Report on Phase II of the ITP, 16 discussed the cause of EOI 7002 as follows:11/

17 As was typical for plants of this vintage, formalized analyses for jet impingement were not done for the 18 original plant design, which was based on inherent separation of safety systems through an appropriate 19 arrangement of equipment, piping, walls, and other structures. The current NRC guidelines for formal jet 20 impingement evaluation defined in Section 3.6 of the Standard Review Plans had not been issued at that time.

21 Jet impingement effects were, however, taken into account. The plant arrangement was designed so that a 23

15. Potential Program Resolution Report, File No. 7002, 24 revision 0, IDVP, October 11, 1982. (Gov. Exh. 15.)

25 16. Open Item Report, File No. 7002, revisic1 5, IDVP, July 26, 1983. (Gov. Exh. 16.)

17. " Phase II Final Report, Design Verification Program",

27 Pacific Gas & Electric Company, June 1983, pages 3-17.

31.

- - , , ,4 -- -

,. , ..,e , , , , - , , - , - - - ,--

1 catastrophic break of a high-energy line or other unexpected phenomenon would not affect a redundant 2 safety system, and the original design included appropriate consideration of break locations and 3 calculations for jet impingement forces on major structures.

4 As nuclear plant design progressed and requirements 5 became more formalized and were upgraded, more rigorous documentation of the adequacy of jet impingement design 6 was required by the NRC. The DCP has upgraded the

- original design basis by the performance of a 7 formalized, rigorous investigation of jet impingement utilizing the current requirements and techniques."

8 One can readily understand that when Diablo Canyon was 9

first designed, there was no requirement for a formal and 10 rigorous jet impingement analysis. However, between the 11

. original design of Diablo Canyon and October 11, 1982 (when 12 Reedy finding was dated), PG&E made a commitment in the R. F.

13 their FSAR that the effects of jet impingement inside 14 l containment would be addressed. Therefore, the PG&E 15 l explanation does not identify the the reason why PG&E's QA 16 program did not identify this failure to perform the jet 17 impingement analysis earlier?

18 Q: How does the resolution of EOI 8010 exemplify a failure to

- 19 address the basic cause?

20 A: EOI 8010 resulted from an IDVP concern that low pressure 21 piping and components would be overpressurized by a variety of operational occurrences.18,/ In order to resolve EOI 8010, 23

/

24 l

1 /

~

25 23

( 18. Opca Item Report, File No. 8010, revision 0, IDVP, 27 September 13, 1982. ( Gov . Exh. 17. )

l 32.

l i

y . w yew . g ** . = =+

y---e, yw .wD, , , , . , , e- - _g_ ,_._y .y,,,.., ,,,,v..,_ym,,_,gn%,.9% . p -.,w,3., y,,_y.%,-c. ,. ._

-w_ ,__ ,,.

1 PG&E implemented modifications to assure the protection of 2 low pressure components.1EI Further, to resolve the generic 3 implications raised by EOI's 8009, 8010 and 8062, the DCP 4 provided additional verification of the selection of design 5 temperature, pressure, and differential pressure across power 6 operated valves.

7' The IDVP reviewed part of the DCP reanalysis and 8 reported its findings in ITR 46. On page 2-2 of ITR 46, the 9 IDVP stated, " concerns similar to those originally found in 10 the pressure / temperature review of the AFW system were found 11 by PG&E also to exist in the Main Steam System and portions 12 of the Component Cooling Water System." However, the IDVP 13 documentation of the resolution of EOI 8010 and of the 14 additional verification of design temperature / pressure does 15 not identify the reason behind the original error, or the 16 subsequently-identified errors in other systems.

17 PG&E, in its Final Report on Phase II of the ITP, 18 discussed the cause of EOI 8010 as follows:20/

19 "...EOI 8010 arose as a result of a later system modification to improve start-up flow to the coolers.

20 In implementing the flow improvement, the designers failed to recognize the small increase in pressure 21 resulting from the design change."

22 While the above explanation does identify a reason why 23 the error (EOI 8010) occurred, it raises further questions 24 26 19. Error Report, File No. 8010, revision 8, IDVP, March 4, 1983 (Gov. Exh. 18); Program Resolution Report, File No. 8010, 26 revision 11, IUVP, June 1, 1983 (Gov. Exh. 19).

27 20. Ibid . 17, pages 3-12.

33.

I which should be answered before the basic cause may be 1

2 considered to have been addressed. For instance, does a 3 procedure exist which requires the designer to evaluate 4 possible pressure changer whenever a modification is made to  :

5 a fluid system? Also, what caused the additional errors  :

6 which were reported in ITR 46 but given no EOI designation?

7 For the above reasons, the resolution of EOI 8010 does 8 not address the basic cause leading to the specific error, 9 nor does it identify the QA crror which allowed the original

, 10 error to remain undete :ted.

11 Q: How does the resolution of EOI 8017 exemplify a failure to 12 address the basic cause?

13 A: EOI 8017 resulted from an IDVP concern that separation 14 criteria were violated by an electrical control transfer 15 switch where control power from two redundant safety-related 16 sources was brought together.21/ In order to resolve the 17 specific concern of EOI 8017, the DCP modified the transfer 18 switch to provide separation of the power sources.22/

19 Further, in order to resolve the generic implications of 20 EOIs 8017 and 8057, the DCP provided additional verification 21 of electrical separation which the IDVP reviewed and reported 22 in ITR 49. On page 2-1 of ITR 49, the IDVP states, "The PG&E

! 23 review resulted in the identification of separation and 24 /

25 l 21. Open Item Report, File No. 8017, revision 0, IDVP, 26 October 4, 1982. (Gov. Exh. 20.)

27 22. Ibid. 17, pages 3-19.

34.

I single failure concerns similar to those addressed in the 2 initial sample."

3 The IDVP documentation did not address the reason for 4 the original error, c the QA error which allowed the 5 original error to remain undetected. PG&E, in their Final 6 Report on Phase II of the ITP, discussed the cause of EOI 7 8017 as follows:23/

8 "In the early design of the plant, it was not intended 4 that the separation criteria be literally applied all 9 the way to the redundant devices and terminal blocks .

where the circuits came together within panels. Within  :

10 that context, it was a standard practice to weigh competing considerations for low probability failure 11 events. The value of operational flexibility gained by cross-tying redundant trains, for example, would be 12 weighed against the unlikely occurrence of failure at the point of the cross-tie. Often the gain in 13 operational flexibility outweighs the risk of the local failure at the point of cross-tie." 3 14 15 As in the case of EOI 7002, it is readily understood

-16 that certain requirements were not effective at the time of a 17 time of Diablo Canyon's initial design. However, the error 18 which led to the establishment of EOI file 8017, and the 19 errors which were identified in the DCP's additional review i 20 of separation represent violations of separation criteria i

21 which were committed to in the FSAR. PG&E's explanation does 22 not address the question of how these errors occurred or how 23 they remained undetected by the PG&E QA program.

24 Q: How does the resolution of EOI 8022 exemplify a failure to 25 address the basic cause?

i 26 j 27 23. Ibid. 17, pages 3-20 and 3-21.

35.

, -  % . p --- gy,_,s* * . ~

,,_,m, ,,___m.,, , , . _ , , . , , _ , . ,

. . - - - , .,e...--,.y.,ty- .,,.-v grr,,_ , , . _, , _ , ,.,g , , , _ _ _ , , _ , . _ , _ _, _ _

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

o .

i 1 A. EOI 8022 resulted from an IDVP concern that certain circuit 2 breakers in the 4160V safety-related electrical distribution 3 system had interrupting ratings less than the short circuit 4 interrupting currents in certain operating conditions.2f/ In 5 order to resolve EOI 8022, the DCP obtained a letter from GE 6 documenting the actual short circuit interrupting capacities t 7 of the circuit breakers as significantly higher than the 4

8 rated capacity, and also higher than the calculated 9 duties.21/

10 Although the GE letter may have resolved the specific  !

$ 11 concern of the under-capacity of circuit breakers, neither 12 the IDVP nor the DCP documentation has addressed the question 13 of how the design error originally came about, and why the 14 error was not detected by the PG&E QA program.

15 Q: How does the resolution of EOI 8023 exemplify a failure to 16 address the basic cause?

17

[ A: EOI 8023 resulted from an IDVP concern that, under certain 18 accident conditions, the voltages on the Engineering 19 Safeguards 480V system buses may be insufficient for 20 continuous operation.26/ In order to resolve EOI 8023, PG&E 21 /

. 22 23 24. Open Item Report, File No. 8022, revision 0, IDVP, October 12, 1982. (Gov. Exh. 21.)

24

25. Error Report, File No. 8022, revision 5, IDVP, March 10, 25 1983; Interim Technical Report 24, ravision 1, IDVP, May 4,1983.

4 (Gov. Exh. 22.)

26

26. Open Item Report, File No. 8023, revision 0, IDVP, 27 October 12, 1982. (Gov. Exh. 23.)
36. .

4

- . , , , .,._n_, , - , . - , , - - , - -

w y-. --.., -.-,- . , , , ,y,, ., ,y e.,-, ,..[-.,,, e_m.,,%,,r_,,_ wy,_m.- -, ,,-.,,-.__--_rmy,

o .

1 adjusted the transformer tap settings on the 230 kV start-up 2 transformers and on Buses IF, IG and lH.22/

3 The IDVP documentation does not address the question of 4 why the potential for undervoltage existed in the design, nor 5 why the condition was not detected by PG&E's QA program. The 6 basic cause of EOI 8023 is not discussed in PG&E's Final 7 Report on Phase II of the ITP. Thus, neither the IDVP nor 8 PG&E have addressed the basic cause of EOI 8023.

9 Q: How does the resolution of EOI 8060 exemplify a failure to 10 address the basic cause?

11 A EOI 8060 resulted from an IDVP concern that an interaction 12 involving the' runout control system could limit Auxiliary 13 Feedwater (AFW) flow to less than minimum values in certain 14 operating conditions.21/ In order to resolve EOI 8060, PG&E 15, calculated new runout control system setpoints and 16 implemented the changes in the field.22/

17 The IDVP documentation does not address the question of 18 why the design included the potential for adverse interaction 19 between the runout control system and AFW flow, nor does it 20 identify.the QA error which allowed the design error to 12 1 remain undetected. The basic cause of EOI 8060 is not 22 /

23

27. Program Resolution Report, File No. 8023, revision 5, 24 IDVP, March 11, 1983. ( Gov. Exh. 24. )

25 28. Open Item Repert, File No. 8060, . revision 0, IDVP, j October 29, 1982. (Gov. Exh. 25.) i 26

29. Error Report, File No 8060, revision 5, IDVP, March 15, 27 1983. (Gov. Exh. 26.) .

l 37.

l 1

l

1 riiscussed in PG&E's Final Report on Phase II of the ITP.

2 Therefore, neither the IDVP nor PG&E has addressed the basic 3 cause of EOI 8060.

4 Q: Are the preceding six EOI resolutions the only examples of 5 failures by the IDVP and ITP to address the basic causes of 6 design errors?

7 A: No, they are only indicative of the methodology used by the 8 verification program to identify basic causes for errors.

9 Further examples have not been presented in my testimony in 10 the interest of brevity.

11 Q: In summary, what is your conclusion regarding Contention 77 12 A: The corrective action measures initiated by the ITP in 13 response to the design errors identified by the IDVP do not 14 in a number of cases fully respond to the root cause.

15 Therefore, potential generic concerns were not raised, or 16 were only partially addressed by the verification program.

17 /

18 /

19 /

20 /

21 /

l 22 /

23 /

24 /

25 /

26 /

27 /

38.

i

, '~_ _

. . . , . - . [ ._ _ . _ . _ _ . , . _ , . . . _. . _ . _ __ ___ _.__,_

1 VII. INADEQUACIES IN THE ITP QA MEASURES I

2 INCLUDING ITS CORRECTIVE ACTION PROGRAM 3 Q: What contention addresses the adequacy of the ITP's quality 4 assurance measures applied to the design modifications

! 5 developed since November 1,1981 as well as the sufficiency .

6 of the ITP's Corrective Action Program?

7 A Contention 8. This contention, as set forth by the Board, 8 reads as follows:

9 The ITP failed to develop and implement in a timely manner a design quality assurance program in accordance 10 with 10 CFR Part 50, Appendix B to assure the quality of the recent design modifications to the Diablo Canyon 11 facility and the IDVP failed to ensure that the corrective and preventative action programs implemented 12 by the ITP are sufficient to assure that the Diablo Canyon facilities will meet licensing criteria.

. 13 14 Q: What sources of information measure the adequacy of the ITP's 15 quality assurance program including the Corrective Action 16 Program activities?

4 17 A: There are three measures of the post-November 1, 1981 QA 18 program and its implementation: design product verifications; 19 design process audits; and regulatory compliance reviews.

20 Q: What do you mean by design product verification?

21 A: A design product verification provides a comparison of the 22 Diablo Canyon design documents with the applicable design 23 criteria. Such a review constituted the major focus of the 24 IDVP's verification efforts. The IDVP corrective action 25 review addressed a sample of the seismic and non-seismic 26 safety-related design activities conducted since November 1, 27 1981. The design product verification is also very important 39.

t w e.; -- p se e .- ~~ w

-n -

,w-, N -. >---n ,,,,---6 ,L e--~~-,---a.a -- m,_ e,--e-, ,---a n---- a-e, r, ,,e---,. .ma,, +,,nv-s-,--v,,.,,ne r., wnmm.m--.-,o---,,rar-s,--

1 since it provides a direct measure of the adequacy of the 4

2 as-released for construction design documents.

3 Q: What were the results of the IDVP review?

4 A: The errors identified in the IDVP's design product review of 5 design modifications since November 1,1981 are summarized in 6 Table 8-1 which is appended to this testimony. As set forth 7 in Table 8-1, the review to date has identified approximately 8 five (5) errors of Classes A and B, twelve (12) Class C 9 errors, and one (1) so-called deviation.

10 Q: Did the IDVP in its review of the ITP Corrective Action

, 11 Program also identify other discrepancies which were not the 12 subject of EOIs?

13 A: A design product verification provides a comparison of the

14 Diablo Canyon design documents with the applicable design 1

15 criteria. Such a review constituted the major focus of the 16 IDVP's verification efforts. The IDVP corrective action 17 review addressed a sample of the seismic and non-seismic 18 safety-related design activities conducted since November 1, 19 1981. The design product verification is also very important 20 since.it provides a direct measure of the adequacy of the  !

21 as-released for construction design documents.

22 Q: What were the results of the IDVP review?

23 A: The errors identified in the IDVP's design product review of 24 design modifications since November 1,1981 are summarized in 25 Table 8-1 which is appended to this testimony. As set forth 26 in Table 8-1, the review to date has identified approximately 27 /

40.

y.._ . . _ . _ _ . . . . _ . . .

e 9: r- ,$ye -* *y. -y 4,- , ,, F-,y ,+ . - . . , ,,.,,,,,--,.~,-,--,v--,+--- ,s.- s - -.-re,-.,-,- -,,,.o, w,, ---,,, , ,,,y, -,..--.--.,,w.,--,,,, -

i 1 five (5) errors of Classes A and B, twelve-(12) Class C 2 errors, and one (1) so-called deviation.

3 Q: Did the IDVP in its review of the ITP Corrective Action 4 Program also identify other discrepancies which were not the

~

5 subject of EOIs?

6 A: Yes, as previously disc'ussed in Part IV of this testimony, 7 the IDVP's corrective action review identified numerous 8 design discrepancies which could have, but did not result in 9 the issuance of an EOI. A list of such discrepencies is 10 provided in Table 5-1 which is appended to this testimony.

11 Q: Should the IDVP have issued EOIs for the discrepancies 12 tabulated in Table 5-1?

13 A: ,Yes. In my judgment, the discrepancies in general met at 14 least-one of the IDVP's criteria for issuing an EOI in that 15 they represented:'

16 (a) A deficiency in a QA and Design Control 17 Practice; . _

18 (b) A violation of the verification criteria; or 19 (c) An apparent inconsistency identified in the 20 performance of the work.

21 Q: What is the significance of the IDVP's failure to issue EOIs 22 'for identified discrepancies?

23 A: The EOIs identified by the IDVP in its regiew of the DCP 24 Corrective Action Program unders?>tc. the< nature and extent of 25 the discrepancies actually d.6svy9 ..J by the IDVP.

26 A: Can you provide an example of how this understatement is 27 reflected in the ,IDVP's Final Report?

- 41.

  • M 7

g

1 As In its Final Report, the IDVP concluded that "only one r.ddi-2 tional analysis was found to use an improper value (stress 3 intensification factor), as reported in EOI 1138"?3 p/

4 Contrary to the IDVP's conclusion, Table 5-1 of this testimony 5 sets forth over 10 examples drawn from ITR 59 where improper 6 values of stress intensification factors (SIFs) were disclosed 7 in the ICVP's verification of the ITP's post-November 1981 8 design activities. Moreover, further examples of misapplied 9 SIFs, as' document in ITR 60, are also presented in Table 5-1.

10 Q: What do you mean by design process audits?

11 A: A design process audit provides information concerning whether 12 the DCP effectively implemented its quality assurance program.

13 Q: What are the key documents which constitute the DCP design 14 quality assurance program?

15 A: The following documents form the DCp design quality assurance 16 program and implementing procedures:

17 (a) The Bechtel Topical Report BQ-TOP-1, Rev. 3A, October 18 1980 (Gov. Exh. 27), describes the quality assurance 19 program to be implemented by the DCP. Additional 20 information and clarification of the program was 21 provided to the NRC in a letter dated August 13, 1982, 22 which modified the scope of the DCP QA Program; 23 (b) The Bechtel Nuclear Quality Assurance Manual (NOAM) 24 (Gov. Exh. 28), as amended to correctly identify the 25 Diablo Canyon Project quality assurance policies; 26 27 30. IDVP Final Report, Section 5.6, page 5.6-2.

42.

3 - J- 9- ~ . _ . _ . ~ y,,- -s,._ , _ _ . . . .

1 (c) The Bechtel Quality Assurance Department Procedures 2 Manual (QADP) (Gov. Exh. 29) as amended for DCP provides 3 procedures for Quality Assurance Department personnel 4 utilized on the project; 5 (d) The PG&E Engineering Department Engineering Manual 6 (Gov. Exh. 30) establishes the controls and procedures 7 to be used by the DCP for design activities; 8 (e) The Diablo Canyon Project Engineering Instructions (PEI) 9 (Gov. Exh. 31) provide supplemental instruction for the 10 application of the Engineering Manual to the Diablo 11 Canyon Project.

129 Q:  ?::: a design process audit of the DCP quality assurance 13 program conducted by the IDVP?

14 A: Yes, the results of a series of design control audits 15 conducted by Reedy between November 11, 1982 and December 7, 16 1982 are set forth in ITR 41. In addition, Reedy performed a 17 follow-up audit on March 17, 1983.

18 Q: What were the results of Reedy's design process audit?

19 A: The IDVP audit basically took a snapshot of the DCP design QA 20 program at one point in time. In its limited review, the 21 IDVP design control audit disclosed 24 deficiencies in the 22 DCP quality assurance program development and implementation 23 including incomplete records documentation, lack of 24 procedures, procedures not being followed, inadequate 25 training, failure to implement commitments in a timely E

28 manner, inadequate document control, deviations in design 27 control activities, and failure to control procurement 43.

l

-+v ye.- by *- .--

=+ g -4w9 --- a.-- ,.

w w r-w-y--- ---w ..-------.--gg---

1 activities. However, these conditions were determined by the 2 IDVP "to be due to incomplete documentation, because this 3 audit was performed in the early stages of the DCP QA Program 4 implementation."31,/ While it is true that the DCP QA program 5 was only established on August 20, 1982, it should also be 6 remembered that the Reedy audit took place nearly one year i af ter the issuance of the Order suspending the Diablo Canyon 8 license. Thus, in my judgment one would have expected more 9 progress in QA program implementation.

10 Q: Do you agree that none of the audit results identified by 11 Reedy could have either a potential or real impact on the 12 quality of design activities? ~

13 A: No. The Diablo Canyon QA/QC measures presumably were drawn 14 up such that (a) the QA/QC measures were designed to achieve 15 a necessary objective; and (b) if implemented properly, the 16 QA/QC measures would have achieved the objective. In fact, 17 however, the necessary implementation was not achieved.

18 Instead, over a number of years,_there were recurring 19 observations of lack of necessary QA/QC attention.

20 The failure to implement the design control measures 21 represents a serious concern primarily because it reflects a 22 lack of discipline in and management attention to the QA/QC 23 prog ram. The QA/QC management program requires specified 24 standards be reliably and repeatedly achieved, and that 25 program objective was not obtained. In QA/QC, such lack of 26 27 31. ITR 41, page 2.

44.

. r:

-,--g , ww... - - - - , - - - -

, ,-y -e- -- - -- -

.~ -

1

. . 1 l

l l

1 attention to prescribed measures cannot be tolerated. Each l 2 QA/QC measure, once issued by responsible management, must be 3 assumed to be important . The fact that the IDVP now in l

4 hindsight apparently finds the instances of non-compliance to 5 be acceptable (or at least not a significant concern) 6 represents a lack of attention to the necessary discipline 7 and detail which constitutes a basic ingredient of a 8 successful QA/QC. program. Further, Mr. Reedy acknowledged 9 during his deposition that he had not reviewed the recent 10 EOI's 1120 to 1144 resulting from the IDVP's review of 11 corrective action in order to determine the significance of 12 the errors in terms of the QA program.32/ In my opinion, 13 this is a significant omission in Mr. Reedy's review.

14 Q: Were other design process audits of the DCP quality assurance 15 program conducted?

16 A: Yes. Both Region V of the NRC in its routine inspections and 17 the DCP in its audit program evaluated the implementation of 18 the design process. Similar to Mr. Reedy, Mr. Morrill of 19 Region V acknowledged at his deposition that he had not 20 evaluated the QA/QC significance of EOIs 1120 to 1144 in 21 forming his opinion on the adequacy of the DCP QA program 22 implementation.33/ Further, in spite of the past history of 23 design process discr;epancies at Diablo Canyon, Region V 24

32. Deposition of Roger F. Reedy, September 22, 1983, 25 pages 41, 42, 64, 65.

26 33. Deposition of Philip John Morrill, September 28, 1983, page 99. In addition, SER Supplement 18 does not address the 27 potential QA/QC significance of these EOIs.

45.

.- , ~. _ _ . _

'Y$ $ ~

t'- -n m -

--p-, _ , , _ . , ,_, _ , , _ , , , _ , , _,

o' .

1 personnel did not significantly increase their design 2 inspection activities at the DCP home offices during the past 3 two years. Rather, in my judgment, a business-as-usual 4 approach generally prevailed for design inspections.

5 Q: What do you mean by regulatory compliance review?

6 As A regulatory compliance review provides an assessment of the 7 quality program measures as compared to the regulatory 8 requirements of Appendix B.

9 Q: Did the IDVP conduct a regulatory compliance review?

10 A: No. Rather, the IDVP relied upon the NRC's review of the DCP b

11 QA program as assuring that the program adequately addresses 12 the requirements of Appendix B. -

13 Q: Based on your assessment of the foregoing three measures, 14 what is your conclusion?

15 .A: The results of the IDVP design product verification, as well 16 as its design process audits demonstrate that the ITP failed 17 to satisfactorily execute a design quality assurance program 18 for the design modifications developed since November 1, i 19 1981. The results further indicate that, contrary to the 20 requirements of Criteria 1 and 2 of Appendix B, the DCP 21 failed to establish and execute a design quality assurance 23 program. Further, contrary to Criterion 3, the DCP's design 23 control measures failed to assure that the Diablo Canyon l 24 design criteria were correctly translated into design 25 documents. ' Indeed, thd errors and potential errors set forth 26 .in Tables 5-1 and 8-1 demonstrate that the DCP QA program

[ 27 failed to adequately implement the required audits and

4C.

l l

g r,.__ ,- ,__c., g ,- - ~ . , ,,_; ,.e- , r.,., --.p.. . - , - - , , , , , . . . . , . - ,.,r,- , - , - , , ,.,_.y.,.,_--..-wu ..,-,,.,-,,,m_,. ..----,---w-, - . . -.,-- ,

1 corrective action measures contrary to Criteria 18 and 16 of 2 Appendix B. This conclusion inevitably follows as a result 3 of the identified errors since Criterion 18 requires that 4 PG&E and its design contractors perform planned and scheduled 5 audits to verify compliance with all aspects of the quality 6 assurance program and to determine its effectiveness.

7 Follow-up action is intended to be initiated to address the 8 identified discrepancies. Guidance for such follow-up action 9 is provided by Criterion 16 which requires that appropriate 10 corrective action be initiated to correct the identified and 11 any similar discrepancy, to determine the cause of the 12 discrepancy, and to preclude recurrence of further similar 13 discrepancies.

14 Q: What do you recommend?

15 A: Given the demonstrated number and nature of the errors 16 disclosed in the IDVP's review of a sample of the Corrective 17 Action Program, in my judgment it is reasonable to conclude 18 that further critical errors (Class A or B) exist in the 19 design of the plant which can only be uncovered by a rigorous

l. 20 and thorough verification program. Further, the cumulative 21 impact of the major errors (Class C) when coupled with the 22 critical errors indicate the necessity for further l

l 23 verification of the post-November 1, 1981 design activities.

l

[

24 /

i 25 /

26 /

27 /

l 47.

, p .%gm--w -

u.m8-s m- -,.o n -u--.-- - -,

,.,--w -,-. ,.ye -, , p , , , , , -y o--

1 VIII. CONCLUSIONS 2 Q: Based on the foregoing, what have you concluded?

3 A: The IDVP and the ITP have detected and corrected a number of 4 critical errors in both the Diablo Canyon design product, and 5 the quality assurance program measures provided to control 6 the design process. However, in my judgment the verification 7 program has not yet demonstrated that the design of Diablo 8 Canyon is now in compliance with all the applicable NRC 9 regulatory requirements and all the PG&E licensing criteria 10 in that:

11 (a) The verification program failed to verify a 12 suitable sample of safety-related design activities 13 for the design services subcontracted to 14 Westinghouse by PG&E.

15 (b) The verification program failed to systematically 16 identify the basic causes for either the initial 17 failure or the failure of the PG&E design quality 18 assurance process which allowed the errors 19 disclosed by the IDVP and ITP to occur and remain 20 undetected. As a result, potential generic 21 concerns were not raised, or were only partially J2 addressed by the verification program.

23 (c) The IDVP's field verifications of a sample of the 24 design documents resulting from the ITP's 25 Corrective Action Program demonstrated that all 26 configuration control errors have not been detected 27 and corrected. Rather, the continued existence of 48.

I discrepancies between the "as-built" and 2 "as-designed" configuration of the plant indicate 3 that the corrective action measures initiated by 4 the DCP have not been adequate to ensure that all 5 conditions adverse to quality have been detected 6 and corrected, and that the cause of such 7 discrepancies is determined and action taken to 8 preclude repetition of similar discrepancies.

9 (d) The results of IDVP's review of a sample of the 10 post-November 1981 design activities conducted by 11 the ITP's Corrective Action Program establish that 12 the DCP failed to satisfactorily execute a design 13 quality assurance program for its activities.

14 Given the demonstrated number and nature of the 15 design errors identified in the IDVP's review of 16 the Corrective Action Program, in my judgment it is 17 reasonable to conclude that further critical errors 18 exist in the design of Diablo Canyon.

19 For the preceding reasons, the IDVP and ITP design verification 20 efforts conducted to date have, in my judgment, failed to provide 21 an equivalent level of assurance regarding the design of Diablo 22 Canyon as would have been obtained by a QA/OC program executed in 23 a timely fashion in compliance with the regulatory requirements 24 of Appendix B.

25 /

26 /

27 /

49.

l .. . .. -

4 4 9

i TABLE 5-1

. . -~~. .

TABLE 5-1 DISCREPANCIES IDENTIFIED IN CORRECTIVE ACTION REVIEW DCP Analysis EOI Condition Noted by IDVP Issued A. ITR #49'(Rev. 0)

  • The PG&E review resulted in the No identification of separation and single failure concerns similar to those addressed in the initial sample.

Modifications were subsequently made to provide consistency with FSAR separation commitments.

B. ITR #54 (Rev. 1)

In the verification of "as-built" No

  • conditions, the IDVP found two minor instances where the as-built condition did not match the design drawings. The first case concerned the clearance of the crane wheels and guide struts versus the crane rail. Plates were modified to allow the proper clearance, and these modifications had no impact on the structural integrity. Secondly, a nonstructural plate was welded to the outer box instead of the inner box of the guide strut. This discrepancy will be resolved by the DCP for the final resolution of the operational capability of the polar crane.

C. ITR #55 (Rev 1)

Calc 52.15.6.1.0 The DCP used a ZPA of 1.6g for the No control room slab. The revised analysis of the control room slab showed various nodes with a maximum ZPA of 2.0g.

In certain cases, the DCP considered No

  • openings which were only temporary blockouts, or neglected a minor penetration opening in a wall.

The condition reflects a discrepancy between as-built conditions and the design documents.

DCP Analysis EOI Condition Noted by IDVP Issued D. ITR #59 (Rev. 1) 4-100, rev. 0 1. Valve modeling: The DCP neglected the No

  • weight of flange bolts which accounted for 7% of the total valve wcight.
2. Stress computation: The DCP used a No preliminary design pressure which was 240 psi (59%) below the value given in the subsequent DCP DCM.
3. Piping geometry: A difference of 1.42 No
  • feet was noted for the length of a pipe segment.

9-108, rev. 0 1. SIF applications: The DCP used 1.0 Yes for a taper transition at the heat exchanger nozzle. The IDVP calculated the SIF to be 1.9.

2. Interference: The DCP did not address No the interference between the regenerative heat exchanger and the pipe (including interference due to -

insulation).

1-119, rev. O Pipe weight: A 2000 pound flow element No *

' (weight equivalent to a pipe length of about 2.7 times pipe diameter) was not included in the DCP model.

2-111, rev. 0 1. Loading inputs: The DCP applied

' No auxiliary building SAM displacements at Support 11-93SL. This support was field verified by the IDVP to be attached to the containment exterior building. Pressure effects, and thermal anchor motions at two supports were not considered in the DCP analysis.

2. Piping geometry: The IDVP found that No
  • in the DCP analysis the location of
Support 11-93SL differed from the as-built condition by 3 feet 2-3/16 inches.

EOI DCP Analysis Condition Noted by IDVP Issued D. ITR 959(Rev. 1) (Contd)

The straight pipe between the elbow at nodes 15 and 23 was analyzed 4 feet 3 inches shorter than the dimension field verified by the IDVP.

98110, rev. 0 1. Piping geometry: A 3-foot difference No

  • was noted for elevation of Support 10-17SL between information provided by the support drawing and that used in the DCP analysis.

A 1-foot difference between IDVP field information and DCP input for one pipe segment was also noted.

2. Mass point spacing: A mass point No between two horizontal supports was not modeled in the DCP analysis.
3. Loading inputs: Two thermal load No cases were not considered by the DCP.

4-102, rev. 1 Piping geometry: The DCP coded one No

  • portion of 20-inch diameter pipe 3 feet shorter and another portion of 12-inch diameter pipe 4 feet longer, than indicated by IDVP field verification.

4A-100, rev. 0 1. valve modeling: The DCP neglected the No

  • weight of the flanges attached to Relief Valve RV-52. This resulted in a 9% difference in the total valve weight.

EOI '

DCP Analysis Condition Noted by IDVP Issued D. ITR #59(Rev. 1) (Contd)

2. SIF application: An SIF of 1.8 was No not applied at a butt weld on straight pipe near Support 57N/72R. A SIF of 2.1 was also not applied for the

, socket weld at a socket.

3. Loading inputs: DCP pressure and No temperature inputs (based on preliminary design data) were up to 59% lower than values provided in the subsequent DCM.
4. Interference: Pipe deflection at No Support 57N/61R (Y restraint) exceeded support clearance in the Z direction.

8-106, rev. 1 1. SIF application: An SIF of 1.0 was No applied at the pipe flued head interface, where an SIF of 1.9 was appropriate for a taper transition joint. SIFs of 2.1 were not specified for all applicable socket welding locations.

2. Interference: The DCP did not address No
  • the unintentiona] restraint found by the IDVP on the vertical portion of the 3/4 inch vent line.

4-113, rev. 0 1. Support modeling: Support 55S/64R was No

  • modeled as a rigid + Y-directional support, whereas the IDVP field verification found it to be a gravity support.
2. Valve qualification: The gravity No effect (1g) was not included in the acceleration qualification for Valve FCV-356, which has a horizontally mounted valve stem.

EOI DCP Analysis Condition Noted by IDVP Issued D. ITR 459(Rev. 1) (Contd) 4-113, rev. 0 3. SIP applications: SIFs applied at a No 20-inch X 2-inch connection, and at a butt weld on straight pipe were up to 50% lower than SIFs calculated by the

, IDVP.

4. Loading inputs: The design pressure No (preliminary) in one of the lines was 57% lower than the subsequent DCM value.

2-105, rev. 0 1. SIF applications: fae DCP applied an No SIF of 1.0 for three branch connections. The required SIFs were determined by the IDVP to be 1.628, and 1.07.

The DCP did not apply a taper transition SIF (1.9 maximum) to a flange / pipe interface.  ;

SIFs for butt welds on straight pipe <

were also not considered by the DCP.

2. Loading inputs: The DCP vertical No Hosgri spectra (based on preliminary data) was found to be lower than that determined alternately by the IDVP using subsequent DCM values.

2-120, rev. 3 Valve modeling: The weights for Valves No

  • LCV-113 and -115 were modeled 45% low for valve bodies and 8% low for valve operators. In addition, minor differences in the DCP eccentricity calculations were noted, i

l j

EOI DCP Analysis Condition Noted by IDVP Issued D. ITR 459(Rev. 1) (Contd) 6-101, rev. 2 1. Interference: The pipe deflection No during a thermal accident mode exceeded clearance available at Supports 40/21R and 40/22R.

2. Support modeling: Support 10/44SL was No
  • modeled with an eccentricity from process pipe OD of 4.25 feet instead of 4.5 inches.
3. Load summary: The coordinate system No was incorrectly shown on the flued head load summary.

4A-133, rev. 1 SIF modeling: The SIFs for intermediate No butt welds on straight pipe were not considered.

4-101, rev. 1 1. Valve modeling: The DCP used a 29% No

  • lower valve weight for FCV-365 than that shown by the valve drawing. The DCP did not model valve eccentricities for 9 remote-operated valves at 11 locations. Accelerations for FCV-361 were also extracted from an incorrect location by the DCP.
2. SIF applications: The DCP ME-101 No piping analysis program did not apply the taper transition SIF of 1.9 to the elbow side of the valve / elbow interface.

2-114, rev. 1 1. SIF application: The DCP did not No apply an SIF of 1.8 to account for the butt welds on straight pipe.

The DCP applied an SIF of 3.268 at an

! unreinforced fabricated tee where the l IDVP determined that the SIF for this tee should be 4.95.

O o

~

EOI DCP Analysis Condition Noted by IDVP Issued D. ITR 459(Rev. 1) (Contd)

The DCP also applied an SIF of 2.0 at a branch connection where the IDVP determined that the SIF should be

2. 58. .

The DCP also applied an SIF of 2.0 at a branch connection where the IDVP determinea that the'SIF should be 2.58.

2. Valve.modeling: The DCP modeled 75 No
  • pounds for the weight for the 4-inch check valve. The valve drawing shows the weight to be 102 pounds.

7-103, rev. O SIF application: The DCP did not apply No socket weld SIFs consistently at socket weld elbows. Some were treated as short or long radius elbows having lower SIFs than a socket weld. The DCP also did not use a socket weld SIF of 2.1 for a swage fitting socket weld.

8-116, rev. 1 1. SIF applications: The DCP used an SIF No of 1.5 at one 3/4-inch branch at node

32. However, the IDVP found that an SIF of 3.34 should have been used.
2. Support locations: A 3-foot No
  • difference in location of one of the j supports noted, l

8-117, rev. 2 1. Valve modeling: The DCP lumped only Yes two-thirds of the total valve weight i

at the overall center of gravity for l valve 9003A. The weight of the valve I contents (30 pounds) was neglected.

l EOI DCP Analysis Condition Noted by IDVP Issued D.- ITR 459(Rev. 1) (Contd)

2. SIF application: The IDVP did not No apply SIFs at various locations including the flued head, branch connection and butt weld along straight pipe. The IDVP determined SIFs ranging from 1.9 to 1.5 should be used.
3. Loading input: A 63% difference (as No compared to the subsequent DCM value) in design pressure was noted for small portions of piping.

4A-111, rev. 0 1. Loading inputs: The vertical Hosgri No spectra used by the DCP (based on preliminary input) was lower than that given in the DCM.

One of the thermal modes shown in the DCM was omitted from the DCP analysis (based on preliminary input).

2. SIF application: The DCP applied an No SIF of 1.9 at a socket weld connection instead of 2.1.

8-102, rev. 2 1. Valve modeling: The DCP analyzed No

  • Valve 8805B with the operator in the vertical position, but an IDVP field verification found this operator to be in the horizontal position. Also, differences were found for valve center of gravity locations for valves 8805A and 8805B and IDVP.{ sic] and for operator support locations.
2. SIF application: The DCP used a SIF No of 2.44 for the 3/4 inch branch on the 8 inch elbow and a SIF of 3.658 was determined by the IDVP. SIFs for two branches (3/4 inch and 2 inch), field verified by the IDVP to be attached to

< the 8 inch tee, were not included in the DCP analysis.

EOI DCP Analysis Condition Noted by IDVP Issued D. ITR 459(Rev. 1) (Contd)12-101, rev. 0 .1. SIF application: The DCP did not No apply an SIF of 1.8 { sic] butt weld on straight pipe locations.

2. Loading input: The DCP Hosgri spectra No input (based on preliminary design data) was lower than the IDVP alternately enveloped spectra (based on subsequent DCM values).

E. ITR 460 (Rev. ?)

Calculation 4 A-123 (Rev. 6) 1. In the natural frequency analysis, the No tributary weight for the piping on only one side of the anchor was used.

2. The minimum fillet weld size for the No stanchion-to-floor plate weld was incorrectly calculated.
3. The pad-to-process pipe weld was not No analyzed.
4. Reinforced pad stresses were not No examined.

S-1087 (Rev. 1) The stresses in the tee-shoe and its weld No to the pipe clamp were not addressed for the friction force loading.

A-35 (Rev. 6) 1. The torsional stiffness of one member No in the computer model was three times the correct value.

2. The weld between the stanchion and the No process pipe was not analyzed.
3. Reinforced pad stresses were not No examined.

- .+ __ - -_ __

EOI DCP Analysis Condition Noted by IDVP Issued E. ITR #60 (Rev. 1) (Contd)

Calculation #

A-79 (rev. 1) 1. An analysis revision sheet was not No included.

2. The analysis load sheet did not note No that the tributary pipe weight is in local coordinates whereas the remaining loads on the load sheet are in global coordinates.
3. The load case 4 weld stress, which was No a controlling condition, was not analyzed.
4. The support frequency in the No restrained direction was not addressed for the Revision 1 loads.
5. The stress at the cutout in a plate No member not addressed.
6. Stresses in several welds were not No addressed.
7. The analysis did not consider the No stress in the support immediately adjacent to the process pipe.

A-103 (rev. 5) 1. The 10 inch diameter sleeve, on which No

the support is mounted, was not i

considered in the natural frequency calculation. The IDVP considered the sleeve to be the most flexible member of the support by engineering judgement.

, 2. The 1/4 inch groove weld between the No existing lug and the process pipe was incorrectly analyzed.

l l

l 1

EOI l DCP Analysis Condition Noted by IDVP Issued E. ITR #60 (Rev. 1) (Contd)

Calculation #

3. The 1/4 inch fillet weld between a Yes member and the existing lug was (1129) incorrectly analyzed. The IDVP determined that this weld stress exceeded the allowable by factoring by correct inputs.
4. The Revision 4 sheet was not included. No
5. The IDVP field verification noted a No
  • weld across the top of a member attached to the process pipe. The DCP drawing did not show this weld.

A-148 (rev. 4) The DCP did not directly address the No support natural frequency. However, by engineering j udgement, the IDVP found that the natural frequencies were greater than the required minimum 20 Hz.

H-1279 (rev. 3) 1. The stress in the weld between two No members [ sic).

2. The computer analysis does not use the No maximum loads.

H-1040 (rev. 2) 1. The stresses in the shear lugs and the Yes attachment welds were not addressed. (1131)

2. A list of references was not included. No
3. The DCP analysis included a support No
  • drawing which showed a weld on two sides of a beam at its attachment to the existing insert. The IDVP field verification indicated an all-around weld on four sides of the beam (i.e.

the DCP analysis was conservative).

EOI DCP Analysis Condition Noted by IDVP Issued E. ITR #60 (Rev. 1) (Contd)

Calculation i A-22 (rev. 5) 1. The maximum combination of loads was No not analyzed.

~

2. The stress in the pipe attachment No adjacent to the process pipe and in the attachment-to-process weld were not analyzed.

H-1052 (rev. 3) The stresses in the shear lugs and No lug-to-process-pipe weld were not specifically addressed.

H-1054 (rev. 3) Member 3 (3x3x3/8 angle) was not analyzed No for stress.

H-359 (rev. 4) 1. The length-to-thickness ratio No (buckling) of a 2x2x3/8 angle member was not addressed.

2. The stresses in the welded attachment Yes and its weld to the process pipe were (1131) not addressed.

S-274 (rev. 3) 1. An analysis revision sheet was not No included.

2. An incorrect load was input into the No computer analysis. ,
3. Several differences were noted between No
  • the drawings and as-built conditions.
4. The analysis did not consider the No cutout in a member (a built-up section of WF and angle members). Affected are the stress in the member and the frequency of the support in the X direction.

EOI DCP Analysis Condition Noted by IDVP Issued E. ITR #60 (Rev. 1) (Contd)

Calculation #

5. The stress in the computer model No members 116 to 123, 126, 128 to 133, 141, 142 and 151 to 156 were not explicitly evaluated to AISC criteria.
6. The drawing does not reference another No
  • support which is mounted on Support 85N/31R.
7. The local reduced cross section of a No member was not considered in the frequency of the suppcrt and in the stress analysis of the member.
8. The weld having the highest stress was No not analyzed.
9. The local bending of a plate was not No analyzed.
10. The pipe welded attachments and their No welds were not addressed.

H-32 (rev. 5) 1. The support frequency in the No unrestrained direction was not addressed.

2. The support frequency in the No restrained direction did not consider the flexibility of the stanchion.
3. The bending and axial stresses were No interchanged in the stress interaction equation.
4. The Civil Verification Transmittal No sheet (notifies the civil group of support loads > 500 pounds) was not included.

EOI DCP Analysis Condition Noted by IDVP Issued E. ITR #60 (Rev. 1) (Contd)

Calculation #

5. The revision sheet was not included. No
6. The IDVP field verification noted a No
  • 1/4 inch plate between the baseplate and Member 9 which was not shown on the support drawing (no impact).

S-245 (rev. 5) 1. There was insufficient documentation No to verify the displacement value on which the frequency was based.

2. The frequency analysis did not No consider all the flexible members in the support.
3. Incorrect acceleration values were No used in the analysis.
4. The pipe clamp was not stress No analyzed.
5. The analysis of a weld did not No consider one moment.
6. The stresses in the shear lugs and No their welds were not addressed.

MP-398 (rev. 2) 1. The stress of Member B (1/4 inch plate) No was not addressed.

2. The torsional stress calculation of No Member 7 (3x3x3/8 angle) did not use the member thickness.
3. The analysis incorrectly computed a No shear area for a clevis.
4. The analysis did not consistently No incorporate the evaluation of increased design loads.

EOI DCP Analysis Condition Noted by IDVP Issued E. ITR 960 (Rev. 1) (Contd)

Calculation i MP-155 (rev. 3) The maximum stress was reported as No approximately half the maximum stress found by the IDVP in the STRUDL analysis.

MP-306 (rev. 2) A moment load was omitted in the analysis No of the critical weld.

MP-249 (rev. 2) 1. The weld stress analysis used No incorrect values for the MX load and for the section modulus.

2. Certain member and weld stresses for No Lead Cases 1 and 2 were evaluated against incorrect allowables (i.e.,

based on Load Cases 3 and 4) .

3. The IDVP field verification noted that No
  • one of the four restraints comprising this support was a small box frame bilateral rather than a tee-shoe and clamp assembly as shown on the DCP support drawing.

MP-983 (rev. 1) 1. Loads provided by the applicable No piping analyses did not appear, in all cases, to be those used in the support analyses.

2. Determination of loads at overlap No supports was not addressed.
3. There were some discrepanciec in the No member properties used in the hand calculations (e.g., torsion constant for Member 52).
4. Discrepancies were noted in some weld No section properties and in some member properties (e.g., for the majority of the frame, the properties used were for M4x13 instead of W4X13).

- - -- I

J DCP Analysis EOI Condition Noted by IDVP Issued E. ITR 460 (Rev. 1) (Contd)

Calculation #

M-178 (rev. 2) 1. The analysis used a shear area of the No full channel in calculating the stress in the channel, rather than using just the web area consistent with each direction.

2. Welded attachments (lugs) and their No welds to the process pipe were not evaluated.
3. The analysis erroneously calculated Yes and compared the frequency in the (1139) restrained direction to an incorrect value.

7-301 (rev. 0) 1 Preliminary thermal operating modes No

-were used by DCP in Revision O.

Revision of this analysis used values from the subsequent DCM.

2. An SIF of 1.3 (instead of 2.1) was No '

t applied at two socket weld locations without appropriate DCP field verification documentation. Also, at one location, an SIF = 1.0 (instead of 2.1) was applied for a half coupling.

3. One piping geometry modeling No
  • difference (location of data point 40A) between the computer model and the walkdown isometric was identified.

8-305 (rev. 1) From the IDVP field verification, No

  • unintentional restraints were not shown on the PGandE isometric.

**^'7~*. -m o = *'

o ,

EOI DCP Analysis Condition Noted by IDVP Issued E. ITR #60 (Rev. 1) (Contd)

Calculation #

8-306 (rev. 3) 1. In several cases, pipe geometry No

  • modeling differences exceeded DCP l' tolerances.
2. The weight of one valve was input as No
  • 1 25.7 lb instead of 50.7 lb. 1 l

8-310 (rev. 2) 1. One pipe length, as shown on the DCP No

  • l walkdown isometric was modeled i exceeding the DCP tolerance.
2. Unintentional restraints, as shown on NC
  • the DCP walkdown isometric and by IDVP field verification, were not explicitly addressed in the analysis.
3. Preliminary thermal operating modes No were used by DCP in Revision 2. The DCP compared the preliminary data against values in the subsequent DCM and judged these acceptable.

8-311 (rev. 4) 1. Differences were noted between the No documentation package (e.g. stress j isometric) and the computer analysis. l These documentation differences were resolved in Revision 5 of this i analysis. l l

2. One three-way support is modeled as a No
  • two-way. The analysis did not address the modification.

9-304 (rev. 1) Thermal operating modes and large bore No piping displacements (SAM / TAM) used by the DCP in the analysis were listed as preliminary. Revision of this analysis used values from the subsequent DCM.

EOI DCP Analysis Condition Noted by IDVP Issued E. ITR 660 (Rev. 1) (Contd)

Calculation #

9-307 (rcv. 1) The weight of a valve and a branch in the No

  • overlapped region were not modeled in the analysis.10-301 (rev. :2) 1. Thermal operating modes, large bore No displacements (SAM / TAM) and contain-ment dilation movements were based on preliminary data. Revision of this analysis used values from subsequent DCMs and other controlled documents.
2. The analysis did not address valve No operator support requirements.

Revision of this analysis was performed to address this subsequent DCP procedural requirement.

3. At four locations, an SIF = 1.0 No (instead of 2.1) was specified for couplings.
4. The weight of insulation for a portion No
  • of the pipe was not considered in the analysis.

3-303H (rev. 3) 1. It was noted that the walkdown No

  • isometric and analysis considered the line to be insulated, whereas the IDVP 3 found it to be uninsulated.  !
2. A short segment of pipe up to the code No l break valve was 431 degrees Fahren- j heit, which exceeded the recommended  ;

span rule temperature range. The DCP I analysis adequately considered these temperature effects.

3. The effects of X, Y, and Z displace- No ments were considered separately rather than considering the resultant displaceruent perpendicular ' to the pipe l span for pipe flexibility evaluation. l t

1 1

. n_.- .

7- ~ - . , , , ,_, __ _ _ , , . _ _ _ . _ _ , , __ ___ __ ___

,\

r i

/ l .

1 .

I,

~

EOI DCP Analysis Condition _.Noted by IDVP Issued E. ITR 460 (Rev. 1) (Contd)

Calculation 4

4. Components such as valves and elbows No were considered to have the same flex-ibility as the pipe for the thermal flexibility evaluation of a short run of pipe.
5. The response spectral acceleration of No auxiliary building flexible slabs was not considered in the analysis.
6. The seismic stress acceptability was No

- not specifically, documented in the analysis, s s

6-301H (rev. 3) 1. A short segment of pipe up to the code No break valve was 365 degrees Fahren-heit, which exceeded the recommended span rule temperature range. The DCP analysis adequately considered these temperature effects.

2. The valve weight (one case) shown on No
  • the valve drawing was greater than that used in the analysis.
3. The equivalent weight used to No

- ' determine a seismic span containing concentreted weights was incorrectly

, calculated.

4. Support loads due to SAM / TAM for one No support were not summarized on the small bore hanger review sheet.
5. The effects of an anchor in the Yes non-seismic portion of the piping were not evaluated for pipe flexibility consider &tions. (See EOI 1142 below.)

9-327H (rev. 2) The analysis underestimated the design No

  • loads and effective weight at one support.

4

. _19_

4 N

r

-._.-#. -2 _

e e EOI DCP Analysis Condition Noted by IDVP Issued E. ITR 860 (Rev. 1) (Contd)

Calculatiop 4 19-307H (rev. 2) 1. A portion of the pipe was 280 degrees No Fahrenheit which exceeded the recommended span rule temperature

= range. However, the temperature used '

in the DCP analysis for this piping was 120 degrees Fahrenheit.

2. Documentation was not provided for the No active valve acceleration qualification. ,
3. Support loads due to SAM / TAM for one No support were not summarized on the small bore hanger review sheet.

F. ITR 863 (Rev. 1)

Analysis 4 HV-59 The IDVP field verified that only three No

  • of the five analyzed supports existed in the field (-05, -07 and -13). Support 07 was analyzed as the worst case support with a duct tributary length of 15.75 feet.

HV-88 The IDVP review of HV-88 resulted in the Yes issuance of EOI 1143 citing the misapplication of seismic coefficients.

HV-104 Ceiling connection detail for support No

  • 59357-38 shows a gap under a ceiling mounting plate. The methodology used to analyze the plate did not adequately account for the gap.

HV-116 DCP sketches and as-built data did not No

  • correlate with the support analysis. In addition, DCNs for modifications were omitted from the documentation package.

i

\, O EOI DCP Analysis Condition Noted by IDVP Issued F. ITR #63 (Rev. 1) (Contd)

Analysis #

HV-119 Analysis did not include weight of No insulation in determining duct frequency.

H7-81 EOI 1134 was issued as a result of the Yes HV-86 IDVP review of the DCP Corrective Action HV-87 Program. The DCP used an approximate HV-96 procedure to determine a response frequency based on the Rayleigh-Ritz method as performed by the ICES STRUDL II computer code. This procedure was used for the seismic analysis of HVAC ducts and supports.

The EOI was issued because, in some cases, frequencies reported by the DCP were significantly different from those alternately calculated by the IDVP. This difference occurred because the DCP frequencies did not always correspond to the first mode natural frequency.

S-80B 1. The design analysis did not consider No the support daadweight.

2. The design analysis did not explicitly No evaluate column stability.

S-262 One of the as-built supports had a member No

  • length slightly longer than the length used.

S-314 The dead analysis did not consider the No support deadweight.

S-356 1. The generic calculation was not based No on maximum generic conduit weight as required by DCM C-15, Revision 3.

2. Stresses at key support connections No were not evaluated.
3. The analysis did not account for the NO manufacturer's allowables in the analysis of Unistrut.

\

  • l EOI DCP Analysis Condition Noted by IDVP Issued i F. ITR 663 (Rev. 1) (Contd)

Analysis 4 S-562 Analysis neglected deadweight of the No ,

support structure and did not apply peak  !

accelerations. However, the DCP analysis used twice the approximate weight of the attached box.

S-623 The DCP computer model did not fully No r account for proper boundary conditions  !

and for all restraint reactions.

(Field The IDVP field verification determined No

  • Verification that a conduit clamp modification for of Corrective support CSR-127-6-471 had been Action) incorrectly implemented. DCN DCT-EC-3604, Revision 0, specifies changing a conduit clamp on conduit K7218 based on Analysis ACSR-127-T1. The IDVP f.ield documentation (Reference 8A) shows the new clamp to be on conduit K8375, i which is adjacent to conduit K7218.

ITS-2 (rev. 1) The IDVP reviewed ITS-2 using a written No checklist. The results of the review indicate that ITS-2 does not follow the established DCP analysis methodology.

ITS-4 (rev. 1) 1. Incomplete support weight. No i 2. Unreferenced seismic acceleration No coefficients are lower than the latest i spectra acceleration values.

3. Analytical methods which may have No provided unconservative results.

ITS-5 (rev. 1) Of the seven support analyses, one Yes *

, support used an assumed member size and (1123) section property which did not agree with

! the as-built member size. .

m- -1u,__._. , ,_ m_= .u .__,__n__._.__. , _ , . _ . . . . . _ . _ . , . , _ _ _

i

  • DCP Analysis EOI Condition Noted by IDVP Issued F. ITR 463 (Rev. 1) (Contd)

Analysis #

ITS-6 (rev. 1) 1. One support type did conform to DCP No

  • as-built information.
2. Two of the supports analyzed no longer No exist.
3. Four of the supports carried loads No
  • greater than those used in the DCP analysis.
4. One support could not be located in No
  • the field.

(Field Four of cixteen supports were found to No

  • Verification) carry loads greater than those documented by the DCP.

U-131 The DCP effective length (13.1 inches) No differs from the IDVP alternate calculation (28 inches).

U-192 The analysis assumed that the closed loop No double U-bolt was fixed at midlength.

This resulted in an effective gap which led to a pipe rupture load that exceeded the established NSC allowable.

U-313 The DCP rod effective length (48 inches) No differs from the IDVP alternate calculation (91 inches).

U-355 The DCP U-bolt effective length (103 No inches) differs from the IDVP alternate calculation (87 inches).

-23.-

, e i

EOI DCP Analysis Condition Noted by IDVP Issued F. ITR 963 (Rev. 1) (Contd)

Analysis #

(DCP test The DCP test program is responsive to the No program for IDVP QA finding noted in ITR 942. Based U-bolt on negative test results, several design connectors) modifications have been planned and implemented. In particular, the nuts and couplers are being replaced with split wedge designs. During the connector testing, the DCP noted a concern with the ductility of the U-bolts and rod beams.

DCP Open Item 42 has been issued to track this ongoing work.

S-20 1. The adequacy of the baseplate was not No explicitly addressed in the analysis.

2. A dynamic impact factor of 1.8 was No not applied.

S-30 1. The stresses in one key member were No not evaluated.

2. The adequacy of the base plate No connections was not evaluated for plastic moment transmitted through coped flange joints.

S-130 The analysis did not explicitly apply a No 1.8 dynamic impact factor or evaluate the rock bolts.

S-150 1. The concrete shear cone area was No underestimated by 50% in the design analysis.

2. The adequacy of the base plate No
connections was not evaluated for plastic moment transmitted through coped flange joints.

4 o EOI DCP Analysis Condition Noted by IDVP Issued F. ITR 963 (Rev. 1) (Contd)

Analysis #

S-240 1. The design analysis incorrectly No calculated the allowable values for

, the bearing plate.

2. The DCP incorrectly calculated the No fundamental frequency in the plane normal to the frame.
3. The capacities of the through bolts No were not evaluated.

S-260 1. The design analysis incorrectly No analyzes the weld stress in a column

. plate.

2. Rock bolts were evaluated using a No dynamic impact factor of 1.25 rather than 1.8.
3. The coupled U-bolt / substructure No analysis results were incorrectly evaluated.

S-329 1. The endmost line loads were modeled No incorrectly in the analysis.

2. The design analysis does not evaluate No the anchor under rupture loads from each of the 5 pipe lines and all 6 load components of a single pipe rupture.

S-331 1. Two of the bolts in the four bolt No

  • plate joint between two column members had been cut out to prevent pipe movement interference. The impact of the reduced section was not evaluated.
2. Bolts connecting beam base plate to embedded plate were not evaluated for shear stresses developed between the plates.

Cs N

e TABLE B-1 <

l I

I

- "- r - - . _ _ _ , _

me 4 TABLE 8-1 ERRORS IDENTIFIED IN DESIGN PRODUCT REVIEW (Design Modifications Since November 1, 1981)

  • Appendix B EOI Error Description of Error Criteria
  • Reference Class
1. Differences disclosed between 3,6,10,16 EOI 1120 Error C "as analyzed" and "as-built" EOI 1121 bolt sizes.
2. Differences disclosed between 3,6,10,16 EOI 1123 Error C "as-built" and "as-analyzed" instrument tubing support.
3. Design analysis finite 3,6,10,16 EOI 1124 Error B elenent model of the control room slab used to generate Hosgri spectra does not agree with the field verified location of the supporting wal] .
4. Revision 1 of the HVAC 3,6,16 EOI 1125 Error C compressor seismic calculation used incorrect and unconservative spectra.
5. DCP used improper stress 3,6,16 'EOI 1126, Combine w/

intensification factors EOI 1138 EOI 1098 (SI?). (Error A/B)

6. Deficiencies in DCP 3,6,10,16 EOI 1128 Error C reanalyses of station battery racks regarding bolt diameter and resolved shear force.
  • The designation of Appendix B criteria relevant to the identified deficiencies is intended to highlight the major QA/QC criteria violated and, as such, is not intanded to be an exhaustive list.

The preceding limitation is necessary also becauce of the high degree of interrelationship between a number of the criteria of Appendix B.

l

sp. e l

Appendix B EOI Error Description of Error Criteria

  • Reference Class
7. Errors in the design 3,6,16 EOI 1129 Error C reanalyses for large bore pipe support 565/3A.
8. Reanalyses of large bore pipe 3,6,16 EOI 1131 Deviation sGpports not evaluated as required by the DCP procedure.
9. Failure to perform an 3,6,16 EOI 1132 evaluation of Auxiliary Combine w/

EOI 1097 Building slabs for in-plane loadings contrary to the PG&E (Error A/B)

Final Report dated May 18, 1983.

10. Incorrect valve modeling in 3,6,16 EOI 1133, Combine w/

DCP seismic reanalyses. EOI 1135, EOI 1098 EOI 1137 (Error A/B)

11. Use of incorrect bolt 3,6,16 EOI li36 Error C allowable stress in the DCP reanalysis.
12. Error in the design analysis 3,6,16 EOI 1139 Error C calculation of frequency of a small bore pipe support.
13. DCP analysis failed to 3,6,10,16 EOI 1140 Error C examine the discharge nozzle flanged joint. As-built configuration does not conform to PG&E piping specification.
14. DCP failed to identify all 3,6,16 EOI 1141 high energy lines inside and Combined w/

EOI 1098 outside containment. (Error A/B)

15. Pipe support loads due to the 3,6,16 EOI 1142 Error C effects of various loading combinations not considered in the design analysis contrary to the DCP design criteria procedure.

as .*

Appendix B EOI Error Description of Error Criteria

  • Reference Class
16. DCP analysis does not 3,6,16 EOI 1143 Error C

. correctly consider the effect of the revised vertical and

-horizontal Hosgri spectra.

17. Design analyses performed to 3,6,16 EOI 1144 Error C generically qualify vents and drains may not be conservative.
18. Hosgri design response 3,6,16 EOI 3009 Error C spectra for the containment interior structure developed by DCP does not envelope raw spectra developed by the IDVP.

I l

l

- 9' 4 5 9 9 F * '%9 9Y p- -y-+-g + g y ** * *-- -9 - -


w- - *ye ~-m-~-=+