ML20054L289
| ML20054L289 | |
| Person / Time | |
|---|---|
| Site: | Diablo Canyon  |
| Issue date: | 07/02/1982 |
| From: | Dick C, Jacobson M, Raymond W PACIFIC GAS & ELECTRIC CO. |
| To: | |
| Shared Package | |
| ML20054L277 | List: |
| References | |
| NUDOCS 8207070348 | |
| Download: ML20054L289 (75) | |
Text
,
i
).
1 UNITED STATES OF AMERICA 2 NUCLEAR REGULATORY COMMISSION 3
4 BEFORE THE ATOMIC SAFETY AND LICENSING APPEAL BOARD 5
6 7
)
8 In the Matter of ) Docket No. 50-275
)
9 PACIFIC GAS AND ELECTRIC )
COMPANY ) (Low Power Proceeding) 10 )
(Diablo Canyon Nuclear Power )
11 Plant, Unit No. 1 )
)
12 13 14 AFFIDAVIT OF WARREN A. RAYMOND, CHARLES W. DICK, AND MICHAEL J. JACOBSON 15 16 STATE OF CALIFORNIA )
17 ) ss.
CITY AND COUNTY OF SAN )
18 FRANCISCO )
19 20 21 22 The above being duly sworn, depose and say:
23 24 The purpose of this affidavit is not to establish 25 the absence of deficiencies in the implementation of the 26 PGandE Quality Assurance Program. Obviously deficiencies r207070393 820702
- DR ADOCK 05000275 PDR l
1.
1 did exist; however, those associated with the implementation 2 of QA/QC procedures for design activities are being 3 generically addressed by the Independent Design Verification 4 Program (IDVP) established by the NRC Order Suspending 5 License, CLI-81-30 (November 19, 1981). This affidavit will 6 establish that the construction work is not similarly 7 affected because of:
8 . the difference between construction and design work 9 process 10 . construction work history 11 . surveillance activities by regulatory agencies, 12 Quality Assurance / Quality Control groups, and "other 13 groups" 14 This affidavit will also demonstrate that Mr. Hubbard 15 mischaracterized and/or misstated many of the facts set 16 forth in his affidavit concerning alleged " breakdowns" in 17 the Diablo Canyon Quality Assurance Program. (The Hubbard 18 affidavit dated May 24, 1982 will be hereafter cited as 19 "Hubbard aff. at . ")
20 21 1. HISTORICAL DEVELOPMENT OF QUALITY ASSURANCE AT DIABLO CANYON.
22 23 In order to place the QA/QC issue in perspective, 24 it is useful to briefly describe the quality assurance 25 Program which has evolved over the past 16 years at DCPP.
26 ///
1 1 The Diablo Canyon project study started in 1964 2 and resulted in the issuance of a contract for the Nuclear 3 Steam Supply System INSSS) in November 1966 -- approximately 4 2-1/2 years before the proposed Appendix B to 10 C.F.R. Part 5 50 (" Appendix B") was issued for comment. The Unit 1 PSAR 6 was submitted in January 1967, prior to the identification 7 of formal quality assurance requirements, and thus the 8 requirements were not specifically addressed. Supplement 5 g to the Unit 1 PSAR was submitted in November 1967 and 10 described a quality program for the primary system. The 11 Construction Permit for Unit 1 was issued in April 1968 (one 12 year prior to publication of the proposed Appendix B), and 13 construction of the major buildings commenced in June 1969.
14 At this time the seismic design for Unit 1 was essentially 15 complete. Two months prior to the start of major 16 construction on Unit 1, the proposed Appendix B was 17 published for comment.
18 Formal quality assurance requirements were first 19 addressed in Appendix G to the Unit 2 PSAR which was 20 submitted in July 1969. That appendix addressed Appendix B 21 requirements as follows:
22 QUALITY ASSURANCE A comprehensive Quality Assurance Pro-23 gram for the design and construction of Unit 2 at Diablo Car., in will be estab- ;
24 lished by Pacific Gas and Electric Company in the discharge of its respon-25 sibility to build a safe and reliable nuclear power plant. The program will 26 be an extension of the quality assurance O
1 program developed for Unit 1, and will benefit from the experience gained from 2 the earlier program. . . .
3 This Appendix discusses the scope of the program, its organization and manage-4 ment, and the implementation of the technical and systems aspects of the 5 program. The organization of the material in this Appendix is based on 6 the divisions used in the proposed
" Appendix B - Quality Assurance Criteria 7 for Nuclear Power Plants", issued for comment by the Atomic Energy Commission 8 on April 17, 1969. Since there is substantial overlap in the material 9 covered in the various headings, the subject matter of an individual division 10 should not be considered separately but in relation to the entire Appendix.
11 12 PGandE established a Quality Engineering Department in 13 November 1969 to implement this commitment to a Quality 14 Assurance Program, and two months later the Company issued a 15 Quality Assurance Manual for Design and Construction. This 16 manual was developed utilizing the proposed criteria ( April 17 1969) since Appendix B was not formally published until June 18 1970. Since that time PGandE has continually reviewed and 19 modified, as necessary, its Quality Assurance Program and 20 organizational structure in order to be responsive to the 21 evolving regulatory guidance and industry standards.
22 Erection of major piping systems for Unit 1 23 started in November 1970, approximately six months after f
24 Appendix B was issued and four months before 10 C.F.R. Part 25 50, Appendix A was issued. The NSSS installation began in 26 September 1972, and Unit 1 construction was completed by 1 November 1975 and a " Hot Functional Test" was satisfactorily 2 conducted. One month later, Unit 1 was licensed to receive 3 and store fuel in preparation for fuel loading.
4 During this period, there were several concurrent 5 industry and NRC actions that had a significant impact on 6 the evolution of quality assurance requirements. Appendix B 7 was issued with a minimum of definition as to the detail 8 required for implementation. In 1970, industry began to 9 form technical groups to define Appendix B requirements.
10 This effort resulted in the development of the N45.2 series 11 of standards. Even though these standards did much to 12 interpret and pros.de guidance on the application of the 13 Appendix B requirements, the NRC found it necessary to 14 conduct a series of meetings in the mid-1970s and to issue l 15 clarifications (Rainbow Books) to provide guidance in the 16 establishment of industry quality assurance programs. An l 17 example of the difficulty encountered in developing a l
18 meaningful interpretation of Quality Assurance requirements 19 was illustrated by the development of ANSI N45.2.13. The 20 work on this standard began in September 1971 and it was not 21 completed until 1976 when the standard was issued --
a 22 Period of approximately five years. The NRC er. lorsed this 23 standard in July 1977 -- about six years after work on it 24 began.
25 In addition to the WASH documents (Rainbow Books),
26 the NRC has issued Regulatory Guides which affect the
O 1 interpretation of Appendix B. For example, the Standard 2 Review Plan references 17 Quality Assurance Programmatic 3 Regulatory Guides which are applicable to present day design 4 and construction of nuclear power plants. Many of these 5 Regulatory Guides were issued after Unit 1 construction was 6 complete but have had a significant effect on the current 7 interpretation of Appendix B.
8 The initial QA Manual including the section for g control of construction activities was issued in January, 10 1970 and has been reviewed and updated on a continuing basis 11 in order to be responsive to the evolving regulatory 12 guidelines and industry standards. In 1980, the QA Manual 13 for control of all departmental activities was incorporated 14 into a single corporate manual which controls all quality 15 related activities for PGandE. Detailed implementing 16 Procedures for control of construction activities were 17 issued as a separate manual. The DCPP site construction 18 departments (Mechanical, Electrical, Civil) started to icaue 19 formal, controlled procedures to implement the requirements 20 of the then proposed Appendix B requirements in October 21 1969, 8 months before Appendix B became a formal 22 requirement.
23 PGandE started formal audits of construction 24 activities in December, 1969. Audits of construction 25 activities have been conducted on a regular basis since that 26 time in accordance with NRC regulations as described in the
1 QA Manual. The corporate QA/ construction QC audits from 2 1969 through 1981 exceeded 1500. This total does not 3 include the hundreds of audits conducted by the NRC and 4 contractors in accordance with their QA programs.
5 6 2. INDEPENDENT DESIGN VERIFICATION PROGRAM 7 The IDVP was formally established in strict 8 accordance with the November 19, 1981 Commission order and 9 the letter from Mr. Harold Denton of the same date. The 10 IDVP is managed by Teledyne Engineering Services (TES) and 11 employs several consultants to review QA and the 12 safety-related design activities performed by PGandE and 13 consultants who contributed to the design effort.
14 It is important to recognize that QA programs are 15 only one means to the end of being sure that design 16 requirements are met. When complete, the IDVP and QA 17 Programs are not in themselves a total solution to all 18 Problems. The IDVP is designed to provide assurance that 19 design meets criteria and other requirements, and in doing 20 so will more than compensate for any past QA deficiencies.
21 This concept is well expressed in the TES Interim Technical 22 Report for DCPP, dated June 23, 1982, and is quoted below.
23 Although one purpose of the QA Audit and 24 Review is to provide certain information in direct response to the NRC Order, 25 another purpose is to obtain background information which impacts the review of 26 the design process. The QA reviews
1 1 indicate trends in activities such as design control, document control, and 2 test control as they existed during the design process. A12 hough these trends 3 may indicate the level of consistency in a given area of design, good QA cannot 4 guaranty good design nor does poor QA necessarily indicate bad design. There-5 fore, the QA audits and reviews are not in themselves an indicator of design 6 verification or the absence thereof.
Instead, the results of the QA audits 7 and reviews may affect the extent of verification of the design process. A 8 QA Finding reveals the potential for a low level of consistency and indicates 9 the possible need for additional verifi-cation.
10 11 The IDVP was specifically developed to provide 12 assurance that design activities were accomplished in a 13 manner consistent with the license application and does not 14 depend on the effective implementation of the original QA 15 Program. The quality assurance review associated with the 16 IDVP provided input to the independent review process to 17 identify those areas which might require additional review 18 over and above that originally anticipated.
19 The Commission Order also included a requirement 20 for an independent review of the QA Programs of PGandE and 21 its design consultants. R. F. Reedy, Inc. ("RFR") performed 22 this review for the IDVP and submitted a Report dated 23 March 8, 1982 (" Reedy Report"). The RFR effort had as an 24 objective the identification of design areas where the IDVP 25 could most effectively concentrate its efforts. (TES Inter-26 im Technical Report dated June 23, 1982.) To accomplish I this objective Reedy used present day interpretations and 2 understandings of Appendix B in place of the differing 3 interpretations and understandings applicable at the time 4 work was done. This approach established a low threshold 5 for the reporting of non-conformances in the area of Quality 6 Assurance.
7 The Hubbard Affidavit relies heavily on the 8 findings of the Reedy Report both in his general 9 condemnation of the PGandE Quality Assurance program and his 10 inference of " breakdowns" in the QA/QC program as applied to 11 construction work. The Reedy effort was directed at quality 12 assurance in the area of seismic design and did not involve 13 site or QA for construction activities. The Reedy Report 14 was thoroughly discussed at the NRC meeting of April 1, 1982 15 where Mr. Reedy himself stated that:
16 "I have to agree with you that the strict interpretation that we use[d] was 17 completely unfair. The order to me did not seem to be fair to begin with, and I 18 made a comment at the time that the evaluation should be done to the criter-19 ia that was in use at the time this pro-gram was accepted by the NRC and audited 20 by the NRC, but we did not go back and say we will accept what the NRC audited 21 and what they accepted. We will use the criteria in the order. Now I don't 22 think that is fair, but that is what we did." (Tr. pp. 32 & 33.)
23 24 Following that meeting PGandE responded to the Reedy Report 25 on April 15, 1982. That response is attached to this 26 affidavit as verified by Mr. Warren Raymond, Manager of
1 Quality Assurance for PGandE. It is incorporated herein by 2 this reference. That response evidences, contrary to the 3 conclusions drawn by Mr. Hubbard, thut PGandE did institute 4 QA/QC program control measures.
5 6 3. CONSTRUCTION QUALITY ASSURANCE 7 The Hubbard affidavit asserts that because of 8 quality assurance deficiencies in the area of design, one is 9 " led to believe" that deficiencies exist in the area of 10 construction. PGandE QA reviews and NRC inspections (Diablo 11 Canyon Reverification Ctudy, Transcript February 17, 1982, 12 p. 31 and Commission Briefing, Transcript November 9, 1981, 13 p. 23), confirm that there is no evidence of comparable 14 difficulties in construction activities.
15 There are various reasons why program deficiencies 16 in the area of design QA should not be indicative of the 17 quality of construction or the effectiveness of a 18 construction QA/QC program.
19 QA programs for design and construction activities 20 function in significantly different ways. In construction, 21 one deals with a product (structures, systems, and 22 components) which permits hands-on type verification, such 23 as inspection and test. In design, one deals with a " paper" 24 product and verification must be accomplished by measures 25 such as checking, design reviews, approvals, and/or limited 26 testing. It is no accident that design quality has
I historically relied more heavily on professional judgment, 2 and less on QA/QC controls, than has construction. One 3 important reason for this has been the difficulty of 4 developing methods for design verification opposed to 5 construction verification.
6 For example, in applying the relevant criteria of 7 Appendix B there are differences between design and 8 construction in organizational structures and 9 responsibilities, verification approaches for assuring 10 design adequacy and product conformance, knowledge and 11 skills required for QA/QC personnel, appropriate remedial 12 and corrective actions, and, most important, the process of 13 verifying adequacy and detecting deficiencies. Contrary to 14 Hubbard's assertion (Hubbard aff, at 92), PGandE recognized 15 these differences in the organization of the original QA 16 Manual. It was written to specifically address the special 17 requirements of the various departments associated with 18 nuclear activities. For instance, one section of the QA 19 Manual prescribed engineering requirements (PRE); another 20 section related to construction work (PRC); and another 21 section related to procurement of materials (PRP). Other 22 job specific items were described in separate sections of 23 the manual. Thus, the PGandE QA manual incorporated a set 24 of separate requirements for engineering, procurement and 25 construction.
26 ///
l 1 With regard to controls and verification methods, 2 nuclear construction work has been subject to inspection, 3 testing, and other verification programs since the inception 4 of the nuclear power program. All construction work at DCPP 5 was performed by independent contractors with clearly 6 defined work scope and controls. The necessary interfaces 7 for construction work were specifically designated in 8 contract documents related to the product. Thus, 9 responsibilities were clearly defined and information flow 10 more easily controlled for construction work.
11 The historical time span covered by formal 12 construction verification programs for DCPP has been longer, 13 and the programs are more detailed than in design.
14 Construction controls were easier to develop and implement 15 than design controls. For example, it was not until 1971 16 that ANSI N45.2 was issued, and 1974 for ANSI N45.2.11 (the 17 basis for Regulatory Guide 1.64). These constituted the 18 first industry wide standards for design control QA 19 requirements; and Regulatory Guide 1.64 was the first on 20 design control issued by the NRC. As late as June, 1976, 21 Regulatory Guide 1.64, Revision 2, was issued, further 22 evolving the design control guidance issued by the NRC to 23 interpret Appendix B, Criteria III. Previously, other 24 requirements existed in codes and other standards, or drafts 25 thereof, but these were relatively specialized and were not 26 directed toward mangagement programs for quality.
1 In contrast, for construction activities at Diablo 2 Canyon, " quality assurance" requirements were an integral 3 part of every contract specification for construction, even 4 those issued prior to the time Appendix B was issued for 5 comment in 1969. (See Bain affidavit attached to PGandE 6 Response to Motion to Reopen.) These requirements covered 7 the areas of testing, material procurement, inspection, 8 reporting, personnel qualification, etc., either directly or 9 by reference to detailed codes and standards. These 10 requirements were further supplemented by the ASME 'N' Stamp 11 Program used by contractors working on the project. When 12 formal QA programs became a requirement for future nuclear 13 work, the changes for construction contractors at DCPP were 14 significantly less than the changes and additions required 15 for design activities. This was not unique to PGandE, but 16 was characteristic of the situation throughout the nuclear 17 industry.
18 For construction activities, the quality program 19 required multiple tiers of verification. Initial 20 verification of each contractor's work through first-line 21 inspection by contractor inspectors was performed. For the 22 Reactor Coolant System an additional inspection was 23 conducted by Westinghouse, the supplier of the Nuclear Steam 24 Supply System. Also verification was performed on a 25 Planned, routine basis, by PGandE's field engineering and 26 inspection personnel by continuous reinspection of 1 contractor work. In addition, quality control and quality 2 assurance included this activity in their scheduled audits 3 and continuous monitoring. This verification included not 4 only reinspections but witnessing of tests by contractors, 5 ongoing work of the contractors, and user tests of 6 materials, plus preoperational and startup testing of 7 systems. This verification was in addition to procedural 8 and other management controls.
9 Throughout the construction of DCPP well-defined 10 QA/QC controls were required of the construction 11 contractors. Strict requirements existed from the beginning 12 and were enforced. QA programs of the contractors were 13 routinely reviewed at the time of audit. PGandE also 14 Provided its own force of field engineers and inspectors to 15 assure compliance; and the AEC (now NRC) provided 16 inspectors. At the peak of construction the ratio of 17 inspectors (excluding NRC) to workers was approximately 1:8.
18 Further, other inspectors, (e.g., State of California, 19 Hartford Steam Boiler, and ASME) were much in evidence.
20 With regard to the general concept of audits, 21 PGandE required that routine (Program) audits be conducted 22 to assure the programmatic implementation of QA requirements 23 and activity audits be conducted to provide . reviews of 24 compliance to implementing procedures. These audits also 25 reviewed all records concerning:
26 ///
1 l
1 . Nonconformance reports.
2 . Minor variation reports.
3 . Design change notices.
4 . Equipment inspections.
5 . Inspector qualifications.
6 . Drawing control.
7 Some findings and deficiencies resulted from NRC a inspections and PGendE QA audits. This is to be expected 9 for any nucler.r power plant construction QA program. It is 10 also indicative that the programs were performing their 11 intended function. Despite a rigorous audit and inspection 12 Program, there have been no findings to indicate serious 13 deficiencies related to the construction QA/QC program at 14 DCPP.
15 The Hubbard affidavit cites only four discrepan-16 cies that could even remotely be considered to involve work 17 activities under the purview of the construction QA/QC i 18 program. Alleged " errors" IV.J. IV.Q, IV.R. and IV.T are
~
l 19 arguably associated with the construction QA/QC program, and l 20 these dwell heavily on the issue of problems with as-built i
21 drawings. (Hubbard aff. pp. 14-32.) In fact, IV.J and IV.T 22 are the same issue.
i 23 The assertion is made that discrepancies between 24 as-built drawings and as-built construction reflect poor l 25 construction. This argument is simply erroneous. The lc 26 appropriate question should be whether the as-built drawings i
i l
1 accurately reflect construction, not vice-versa. In point 2 of fact, as-built drawings are made after construction work 3 is complete. But the real test of construction quality is 4 whether the construction accurately reflects the approved 5 design including approved field changes.
6 7 4. OTHER MISCHARACTERIZATIONS AND/OR MISSTATEMENTS IN HUBBARD AFFIDAVIT.
8 9 In alleging that the PGandE QA Program was defi-10 cient, Mr. Hubbard has made a number of erroneous statements 11 and mischaracterizations in his affidavit. Some further 12 examples of these are described below.
13 Item 112 In referring to 10,000 anchor bolt 14 installations, reference is presumably made to 15 additional bolts installed as a result of the 16 Bulletin 02 program, and if so the statement is in 17 error. It is true a large number of additional anchor 18 bolts were installed but this was the result of a 19 cost-benefit decision which concluded it would be less 20 costly to add the larger number of bolts than to 21 perform reanalysis to justify existing installations.
22 If the reanalysis had been conducted, the number of 23 bolt modifications would have been minimal. This 24 problem was reflective of an industry-wide concern over 25 anchor bolt capacities.
26 ///
1 Item IV.V The affidavit concludes that, 2 based on the 24 design errors cited earlier in Section 3 IV and the additional discrepancies identified by 4 PGandE and the independent auditors in the semi-monthly 5 reports, a series of QA/QC " breakdowns" and associated 6 regulatory violations of NRC Regulations have occurred.
7 8 However, a substantial portion of the " errors" and 9 " discrepancies" referred to have no demonstrable 10 relationship to Hubbard's alleged QA/QC " breakdowns."
11 Many of the " additional discrepancies" are differences 12 in calculated results that exceed the 15% reporting 13 threshold established by the IDVP. These differences 14 can be due to many factors such as the use of different 15 assumptions or calculational methods. The mere 16 existence of a difference in results does not indicate 17 a violation of regulatory requirements or a QA/QC 18 " breakdown." Compliance of the original design work 19 with applicable QA/QC requirements and procedures can 20 not be called into question by differences in results 21 of this nature. This distinction becomes significant 22 when it is recognized that the affidavit repeatedly 23 characterizes the same items (i.e. those reported in 24 the semi-monthly reports ) as associated with "QA/QC 25 breakdowns" and uses clear differences in terminology 26 to describe them. Examples are:
1 a. pages 14 - " . . . one hundred and eighty-five 2 (185) additional design and construction 3 discrepancies have subsequently been disclosed in 4 biweekly status reports . . .
5 b. page 25 -
. . . the independent design 6 verificaton program has identified one hundred 7 fifty-six (156) potential errors in the Error and 8 OPen Item (EOI) reports.
9 c. page 82 - "
. . . the nearly 200 design and 10 construction deficiencies identifed to date, and 11 summarized herein in Section IV, had as a 12 significant root cause the failure of PGandE to 13 implement the required and committed to QA/QC 14 Program."
15 d. page 99 ". . . nearly 200 examples have been set 16 forth herein which document PGandE's failure to 17 Provide a QA/QC program . . .
18 Item V.A The Staff, PGandE, and ACRS did not 19 rely on the QA/QC program to allow reduced safety 20 margins. Mr. Hubbard cites his own earlier opinion in 21 Attachment V of his affidavit, but fails to cite any 22 factual evidence to support that conclusion. The 23 statements quoted by Mr. Hubbard simply do not support 24 his conclusion and to our knowledge neither does the 25 licensing record. In fact, to reduce margins because 26 of expectations of better quality performance is not 1 technical capability. Quality assurance is not a 2 factor.
3 Mr. Hubbard has misrepresented the use and 4 Purpose of the lists included in attachments G, H and 5 I. The list in attachment G is the Qualified Suppliers 6 List which was required until 1979 by QA Procedure 7 PRPS. Quality Assurance procedures prior to 1979 did 8 not require Contractors for Services (Consultants) to 9 be listed as a qualified supplier. Contracts for 10 Consultant work were processed and issued by the 11 Engineering Department in accordance with the 12 department's internal procedures. These procedures did 13 not require that a formal list be developed. PGandE 14 discontinued mandatory use of the Qualified Suppliers 15 List in 1979; however, PGandE still retains all the 16 requirements for establishing a qualified supplier.
17 The requirement for publishing a " List" was deleted 18 from the program due to the inordinate amount of work 19 required to maintain qualifications of the suppliers 20 listed. Infrequent use of particular suppliers 21 subjected them to unnecessary audits and more 22 importantly, wasted valuable QA resources which could 23 be used to better advantage in the surveillance of 24 other company safety-related activities.
25 Attachment H is a list developed for infor-26 mation used only in order- to perform a selection of
1 safety related contractors who would be subject to a 2 Quality Assurance review according to IDVP. It was not 3 and is not intended to represent a formal listing of 4 approved contractors, and therefore Consultants that 5 appear in attachment H would not necessarily appear in 6 Attachment G for the reason described in the preceding 7 paragraph.
8 The list included in Attachment I is a 9 Special list developed for environmental qualification 10 of equipment only. All suppliers listed or inferred 11 were qualified by PGandE Quality Assurance or were 12 subsuppliers to the supplier of the major component.
13 These subsuppliers were qualified under the provisions 14 of the procurement requirements required of the primary 15 supplier quality assurance program.
16 Item VII.F The assertion is made that PGandE 17 failed to conduct adequate design reviews in a timely 18 fashion and that audits of the design review process 19 did not result in appropriate follow-up action to cor-20 rect conditions adverse to quality. This assertion is 21 based on the results of three PGandE audits performed at various times. However, Hubbard's assertion is 22 23 based on mischaracterizations of the audit results he 24 presents. Some examples of the mischaracterizations 25 are as follows:
26 ///
l 1 a. Paragraph 80, concerning an Audit performed in 2 April 1977, states that PGandE auditors once again 3 concluded that compliance with design review 4 measures was inadequate. Review of the report 5 itself (Attachment N to Hubbard's affidavit) 6 yields the opposite conclusion. Section 2.0 of 7 Audit No. 77016 states in part:
8 The designs of all the selected structures, systems and components 9 have been reviewed. These reviews comply with the requirements of 10 Procedure PRE-6 except that (1) some apparent design deficiencies 11 identified by the reviews have not been resolved and (2) several de-12 sign review reports lack required approvals.
13 14 Section 3.0 of the same report states in part:
15 The activities reported as incom-plete and the resolutions consider-16 ed open to question could not be considered as infractions or defi-17 ciencies until after the deadline for completion of the design 18 reviews.
19 It is clear from the foregoing that the auditors 20 recognized and identified an item which would have 21 to be completed. It is not unusual for this 22 condition to exist during the conduct of an audit, 23 and it provides proof that there was an active, 24 functioning design review program in place.
25 ///
26 ///
1 The second quotation (regarding ANSI N45.2.11) by 2 Hubbard from this audit does not reflect any 3 deficiency. The statement is a recommendation 4 from the audit team.
5 6 There is additional information on record 7 indicating the presence of an adequate design 8 review program during the 1977 period. See 9 Attachment 1, p. 13 (response to the Reedy Report) 10 which states in part:
11 At this same time, the NRC reviewed PGandE's procedure PRE-6, "Compre-12 hensive Design Reviews," and found that it provided adquate guidance 13 for the conduct and approval of comprehensive design reviews (NRC 14 Report 50-275/77-12). Subsequent NRC inspections and PGandE quality 15 Assurance audits of these compre-hensive design reviews confirmed 16 that the reviews were completed in accordance with the requirements of 17 this procedure.
18 b. Paragraph 81, concerning an Audit in February 19 1979, states that PGandE auditors identified 20 significant deficiencies in the PGandE design 21 verification. Again, review of the report yields 22 a different picture. The cover letter 23 characterizes the audit results as follows:
24 The attached audit report concludes that completed design verification 25 conforms to requirements but that much of thE work is still in prog-26 ress.
1 The cover letter specifically characterizes the 2 four. discrepancies noted as less significant than 3 "nonconformances" which, in the PGandE program, 4 would be the classification for significant 5 conditions adverse to quality. As before, this 6 audit report indicates the presence of an active, 7 functioning design review program.
8 ///
9 ///
10 ///
11 12 13 14 15 16 17 18 19 20 21 22 I
23 24 25 26 c-.. -n., . , -, --
1 2 Since the audits cited above do not, when taken as 3 a whole, indicate failure to conduct design 4 reviews in a timely manner, the basis for claiming 5 a " breakdown" in th QA/QC program is not apparent.
6 .
7 Dated: July 2, 1982 8
9 10 N A -
Warren A. Raym Charles W. Dic V 11 12 M a
- b**
Michael J. Ja' cobs'Us 13 14 15 Subscribed and sworn to before 16 me this 2nd day of July, 1982 17 18 19tincy J. Lemaster, SEA]L
~
19 Notary Public in and for the City and County of San Francisco, 20 State of California.
My commission expires April 14, 1986.
21 22 23 24 25 26 1 PROFESSIONAL QUALIFICATIONS OF 2 WARREN A. RAYMOND 3
4 5 My name is Warren A. Raymond. I am the manager of 6 Quality Assurance for Pacific Gas and Electric Company and 7 have the responsibility of implementing a Corporate Quality 8 Assurance Program in the areas of engineering, construction, 9 procurement, and operation.
10 My educational background is as follows:
11 University of Minnesota - BS in Electrical Engineering, 12 1950.
13 Atomic Power Equipment Department (GE) - Dresden Technology 14 Course, 1957.
15 University of California -
Radioactivity and Nuclear 16 Structure, 1958.
17 University of Idaho - Public Utility Executive Course, 1964.
! 18 I joined PGandE in 1950 as an engineer in power 19 plant operation with responsibility for equipment l 20 performance and instrument maintenance. In 1957 I was 1
21 assigned the position of Plant Superintendent at the 22 Vallecitos Boiling Water Reactor Facility in Pleasanton, i
23 California. From that position, in 1961 I assumed the 24 position of Start-up Coordinator for Humboldt Bay Power 25 Plant Unit 3 (nuclear unit) and in 1971 became Plant 26 ///
1 Superintendent. I was a licensed Reactor Operator from 1958 2 to 1963 and a Senior Reactor Operator from 1963 to 1979.
3 In 1979 I transferred to my present position of 4 Manager, Quality Assurance (previously Director), with 5 overall management responsibility for the implementation of 6 the Corporate Quality Assurance Program. I have 29 years of 7 power plant experience involving all aspects of operation.
8 I have been a member of the ANS Committee for the 9 Development of Simulators for Use for Reactor Operator 10 Training, and am presently a member of the ASME Steering 11 Committee for Owners Certification.
12 -
13 14 15 16 17 18 19 20 21 22 23 24 25 26 f
i
- i I
J 1 PROFESSIONAL QUALIFICATIONS OF 1 1
2 CHARLES W. DICK 3
4 5 My name is Charles W. Dick. I am a project 6 manager and a member of the project management team of the 7 Diablo Canyon Project consisting of the integrated 8 organization of Bechtcl Power Corporation and Pacific Gas 9 and Electric Company, and with responsibilities which 10 include quality assurance. I am a Licensed Professional 11 Engineer in the states of California, New York, and 12 Pennsylvania.
13 My education background is as follows:
14 California Institute of Technology - BS in Electrical 15 Engineering, 1946, 16 Stanford University - MS in Electrical Engineering, 1948.
17 I have also had additional training through 18 Advanced Engineering Programs, in Business Administration, 19 and from various technical and business courses.
20 Prior to my recent assignment to the Diablo Canyon i 21 Project, I was been Manager of Division Quality Assurance l
l 22 Division at the Bechtel Power Corporation from 1980-1982.
23 My responsibilities included formulating the QA programs for i
l 24 implementing such programs, and training QA personnel for
, 25 some 14 nuclear projects.
26 ///
i L
I joined Bechtel Power Corporation in 1965 and w rked as a project engineer on various nuclear and 2
f ssil-fuel projects. I was responsible for project 3
engineering work for a number of different types of power 4
projects and studies, for nuclear standards development and 5
f r licensing. Beginning in 1973, I became an Engineering 6
Manager and subsequently Manager of Engineering, with 7
verall management responsibility for the project 8
engineering work on more than 20 power plant projects.
g Prior to nty employment at Bechtel, I was engaged 10 as an engineer with the General Electric Company beginning in 1948. During that time, I was involved in marketing and application engineering related to nuclear power facilities.
Prior to that I was assigned as an electric utility applications engineer and provided consultation services 15 involving heavy electrical apparatus.
I am a senior member of IEEE and a member of the 17 American Society for Quality Control. I was also a member g
of the industry working group for development of ANSI Standard N45.2.ll (Quality Assurance Standards for Design of Nuclear Power Plants).
22 23 24 25 26
i 1 PROFESSIONAL QUALIFICATIONS OF 2 MICHAEL J. JACOBSON 3
4 5
6 My name is Michael J. Jacobson. I am the Project 7 Quality Assurance (QA) Engineer for the Diablo Canyon 8 Project consisting of the integrated organization of Bechtel 9 Power Corporation and Pacific Gas and Electric Company. I 10 am a Registered Professional Quality Engineer in the State 11 of California.
12 My educational background is as follows:
13 Sacramento State College - BS in Civil Engineering, 1970.
14 Golden Gate University - Business Management Certificate in 15 Management, 1979.
16 I joined Bechtel Power Corporation in 1970 as a 17 quality assurance engineer responsible for various aspects is of design phase quality assurance on a nuclear power plant is project. I was subsequently responsible for performing 20 structural design and seismic analysis activities on the 21 project. Later, I was assigned as Project Quality Assurance 22 Engineer responsible for supervising project QA activities, 23 including direction of quality audits of construction 24 activities.
25 Subsequently, I was assigned Project QA Engineer 26 on various other nuclear power plants, where I was i
1 responsible for directing project QA programs. I was 2 responsible for ensuring that project construction and site 3 activities, as well as quality control aspects, met 4 applicable QA and regulatory requirements.
5 I was assigned to the Diablo Canyon Project in 6 1982 to direct and control the QA program for this project.
7 8
9 10 11 12 13 14 15 16 17 18 19 20 21 22 1
23 24 I 25 26
4/15/82 O
COMMENTS ON THE R. F. REEDY, INC., QUALITY ASSURANCE AUDIT REPORT ON SAFETY-RELATED ACTIVITIES PERFORMED BY PG&E PRIOR TO JUNE 1978 l 4 3o.
I. INTRODUCTION At the April 1, 1982 meeting with the NRC, Pacific Gas and Electric Company (PG&E) committed to submit a formal response to the R. F. Reedy, Inc., Quality Assurance Audit Report (Report) within two weeks. At this meeting it was confirmed the audit was based on a 1982 interpretation of 10CFR50 Appendix B, as well as guidance from ANSI N45.2.11 (Design Control), and applied to the work performed prior to June 1978. For this reason we believe that co=ments on the Report are appropriate.to place the findings in proper perspective, to correct certain inaccuracie's or omissions, to describe the historical development and refinement of the PG&E Quality Assurance Program, and to O present some different conclusions.
Important factors regarding our quality programs were presented in the November 19, 1981 statement of our Mr. G. A. Maneatis to the Subcoc=ittee on Interior and Insular Af fairs of the U.S. House of Representatives (Attach-ment 1). In that statement, evidence was provided as to the dedication of PG&E to formulating and implementing an effective quality assurance program.
We believe this presents a much more quality-conscious i= age than that obtained from the Report. PG&E did provide controls for the work performed, consistent with our co=mitments to regulatory requirements and their interpretations in existence at the time the work was accomplished; however, our engineering activities were not documented by 1982 Quality Assurance standards. It is
= also important that the past work is being further confirmed by comprehensive design verification activities being performed by independent consultants,
- by PG&E, and by certain design contractors.
g The key to understanding our position is to recognize that Diablo Canyon is a project mostly designed and built to quality standards of the late 1960s and early 1970s. It is now 16 years since the first =ajor project commitments were made. During this time, the NRC (AEC) regulations requiring quality assurance programs were first issued, interpretative standards were developed, and enforcement practices evolved.
O
l t
O PG&E's Quality Assurance effort evolved from the preliminary proposed statement of the 18 Criteria in early 1969 to the comprehensive regulatory scheme of 1982. During this period, the NRC inspection effort has also evolved from an inspector's individual inspection plan in the early 1970s to a sophisticated computer-controlled inspection program in 1982. The Report spans almost this entire period; however, no acknowledgment is made of the many interpretations cnd changes implemented by Regulatory Guides, Appendices to existing regula-tions, or experience gained as a result of evaluation of the construction and operation of nuclear power plants. This evolutionary process should be kept i ^
in mind in reviewing the following brief outline of significant milestones of
, the design and construction of Diablo Canyon as referenced to the development of NRC Quality Assurance guidance.
We recognize the dilemma faced by R. F. Reedy, Inc. , in attempting to go back and remember what the interpretation of Appendix B was in the 1972 era before issuance of the subsequent interpretative ANSI Standards and Regulatory Guides.
We agree with Mr. Reedy that to do so would be impractical, if not impossible.
(Tr. p 14)* As a result, it is extremely difficult to achieve meaningful results f rom this type of audit, where the history is so extensive and quality assurance requirements were evolving as project work proceeded.
[~ ' *Ref ers to Transcript of April 1, 1982, NRC " Meeting With Pacific Gas and
\s Electric Company to Discuss Seismic Design Review, Diablo Canyon Unit 1."
y ,
,()II. HISTORY The Diablo Canyon project study started in 1964 and resulted in the issuance of a contract for the Nuclear Steam Supply System (NSSS) in November 1966 -
cpproximately 2 1/2 years before the proposed Appendix B to 10CFR30 was issued for comment. The Unit i PSAR was submitted in January 1967, prior to the identification of formal quality assurance requirements, and thus the require-nents were not specifically addressed. Supplement 5 to the Unit 1 PSAR was submitted in November 1967 and described a quality program for the primary system. The Construction Permit for Unit I was issued in April 1968 (one year
- prior to publication of the proposed Appendix B to 10CFR50), and construction of the major buildings commenced in June 1969. At this time the seismic design for Unit I was essentially complete. Two months prior to the start of major
. construction on Unit 1, the proposed 10CFR50 Appendix B was published for Comment.
Formal quality assurance requirements were first addressed in Appendix G to the Unit 2 PSAR which was submitted in July 1969. That appendix addressed 10CFR50 Appendix B requirements as follows:
QUALITY ASSURANCE A comprehensive Qualit.y Assurance Program for the design and construction of Unit 2 at Diablo Canyon will be established by Pacific Gas and Electric Company in the discharge of its O- responsibility to build a safe and reliable nuclear power plant. The program will be an extension of the quality assurance program developed for Unit 1, and will benefit from the experience gained from the earlier program. . . .
This Appendix discusses the scope of the program, its organization and management, and the implementation of the technical and systems aspects of the program. The organi-zation of the material in this Appendix is based on the divisions used in the proposed " Appendix B - Quality Assurance Criteria for Nuclear Power Plants", issued for
. comment by the Atomic Energy Commission on April 17, 1969.
Since there is substantial overlap in the material covered in the various headings, the subject matter of an individual division should not be considered separately but in relation to the entire Appendix.
PG&E established a Quality Engineering Department in November 1969 to i=plement this commitment to a Quality Assurance Program, and two months later the Company issued a Quality Assurance Manual for Design and Construction. This manual was developed utilizing the proposed criteria (April 1969) since 10CFR50 Appendix B was not formally published until June 1970. Since that time PG&E
. has continually reviewed and modified, as necessary, its Quality Assurance s ) Program and organizational structure in order to be responsive to the evolving regulatory guidance and industry standards.
-- Erection of major piping systems for Unit 1 started in November 1970, approximately six months after Appendix B was issued and four months before 10CFR50 Appendix A was issued. It has been estimated that design at that time was about half complete. The NSSS installation began in September 1972, and Unit I construction was completed by November 1975 and a " Hot Functional Test" was conducted. One month later, Unit I was licensed to receive and store fuel in preparation for fuel loading.
During this period, 'there were several concurrent industry and NRC actions e t hat had a significant impact on the evolution of quality assurance require-ments. Appendix B to 10CFR50 was issued with a minimum of definition as to the detail required for implementation. In 1970, industry began to form
. technical groups to define Appendix B requirements. This effort resulted in the development of the N45.2 series of standards. Even though these standards did much to interpret and provide guidance on the application of the Appendix B requirements, the NRC found it necessary to conduct a series of meetings in the mid-1970s and to issue clarifications (Rainbow Books) to provide guidance in the establishment of industry quality assurance programs. An exa=ple of the difficulty encountered in developing a meaningful interpretation of Quality Assurance requirements was illustrated by the development of ANSI N45.2.13.
The work on this standard began in September 1971 and it was not completed until 1976 when the standard yas issued - a period of approximately five years.
The NRC endorsed this standard in July 1977 - about six years after work on the l
standard was started. By that time Diablo Canyon had completed construction, hot functional testing, and was ready to load fuel. Since that time, PG&E has been involved in the Hosgri investigation, redesign for the postulated Hosgri event, and the TMI-related backfits.
In addition to the WASH documents (Rainbow Books), the NRC has issued Regulatory Guides which affect the interpretation of 10CFR50 Appendix B.
For example, the Standard Review Plan references 17 Quality Assurance Programmatic Regulatory Guides which are applicable to present day design and construction of nuclear power plants. Many of these Regulatory Guides were issued af ter Unit 1 construction was complete but have had a significant effect on the current interpretation of 10CFR50 Appendix B.
. In order to place the significant events involving Diablo Canyon in perspective, l
we are attaching a chronology of significant milestones of Unit I design and l
construction referenced to NRC regulations, industry standards, and Regulatory i, Guides. (Attachment 2) l l
l l '-n
- I III. APPLICATION OF AUDIT CRITERIA A principal concern which was discussed and acknowledged at the April 1, 1982 neeting is that the Report was compiled utilizing "today's" understanding of the 10CFR50 Appendix B criteria for evaluating the adequacy of the PG&E quality cssurance program during the pre-June 1978 period. I'n our. view, the use of current interpretations of requirements as criteria for a historical audit of a quality assurance program over the past four to twelve years has created on erroneous and misleading perspective regarding the relative effectiveness
- of the PG&E quality assurance program.
In our meeting on April 1, 1982, Mr. Reedy stated that the Report did not consider the PG&E quality assurance program as committed to in the PSAR.
(Tr. p 12) We strongly believe that the PSAR information would have been the proper base for the development of audit criteria.
The fact that the Report used current interpretations of 10CFR50 Appendix B criteria was made quite evident several times during said meeting.
(Tr. pp 13 & 14)
It is important to note that 'the use of "today's" interpretation of Appendix B arose from the R. F. Reedy, Inc., interpretation of the Order, and further, that Mr. Reedy was not in agreement with that approach, but felt it was the O only avenue available. (Tr. p 14)
Even more important, Mr. Reedy apparently agrees with our position when in response to a question from the NRC Staff regarding his feeling about the strictness of the criteria used to make the statemeats made in the Report, Mr. Reedy stated:
"I have to agree with you that the strict interpretation that we use was completely unfair. The order to me did not seem to be fair to begin with, and I made a comment at the time
. that the evaluation should be done to the criteria that was in use at the time this program was accepted by the NRC and audited by the NRC, but we did not go back and say we will
. accept what the NRC audited and what they accepted. We will use the criteria in the order. Now I don' t think that is fair, but that is what we did." (Tr. pp 32 & 33)
When 10CFR50 Appendix B was issued in 1970, there i= mediately arose many questions of interpretation. The industry, through the American National Standards Institute (ANSI) and the NRC (AEC) recognized the need for further
" guidance" on the interpretation of Appendix B. Beginning in August 1970 with establishment of the ANSI N45 subcommittee, N45-3 and through the present ti=e,
O there has been an evolving set of Standards and Regulatory Guides that continue to refine, define, and in some cases redefine, the requirements of Appendix B.
Mr. Reedy succinctly covered this issue when he stated that:
"I think it is important to understand that the QA as written in the original 1970 Appendix B has had a lot of inter-pretations that were put out in safety guides and then in regulatory guides and then in adapting or adopting various -
- the rainbow books, for example, the N45.2 series as an educational progression of changing standards in QA, and this occurred starting about 1970 and went on to the present time
- so you have a knowledge increase as to what is required in QA as interpreted by the NRC, but we were not taking that into account because the criteria established in the order was take 10 CFR 50, Appendix B, and I think that is the comment that John Hoch brought out earlier, that this is a strict criteria when you use the end result to compare to the learning curves that goes on over a period of twelve years."
(Tr. pp 12-13)
The NRC Staff appeared to be in general agreement with this opinion.
(Tr. p 41) 7~
U l An example of the impact of the use of today's strict interpretation of Appendix B is evidenced by the general issue of documentation. The Report implies that procedures did not exist or reviews were not held. The Report goes on to admit that many procedures and documents were in fact revit:wed but that =any of the documents were not " controlled" in accordance with today's standards, and hence no credit was given.
In response to a staff question regarding the general philosophy given to audit team members, Mr. Reedy replied:
i=
l "The primary consideration was, we will consider the document if it's a controlled document. In other words, we
- felt that our work could be audited by anybody and the only way you could audit what we did and conclusions that we drew was on the basis of controlled documented evidence. A piece of paper pulled out of someone's file that shows a certain bit of information to me is not a controlled document, and we based our conclusions and things in the report on those documents that were officially controlled."
(Tr. pp 29 & 30) l .
V
_7_
O In our view, these observations should have been noted in the Report to ensure that credit was given for these quality assurance .ype controls that in fact existed. This would enable any reviewer of the Report to obtain a nore comprehensive understanding of the overall quality of the product rather than the more restrictive view given in the Report.
In summary, it is significant to note that the Staff expressed appreciation for the interpretive insight into the Report received at the meeting, because
. it altered the impressions received from initial reading of the Report.
(Tr. pp 36, 43, 102) We certainly agree with the Staff's response that when you take these observations into account you certainly get a different
, impression regarding the PG&E Quality Assurance Program during the pre-1978 period.
With the foregoing thoughts in mind, we now turn to our specific comments on the Report's conclusions.
e O
O
O
-- IV. REVIEW OF REPORT CONCLUSIONS The Report draws three conclusions, presumably developed from a review of the findings, to which we have the following response:
Conclusion 1: "The PG&E Quality Assurance program for design work was not adequate in areas of policy, procedures and implementation. The Quality Assurance organization had insufficient program responsibility."
Comment:
The Report appears to view the PG&E Quality Assurance Departeent as synonymous to the Quality Assurance Program. This viewpoint disregards the wording in 10CFR50 Appendix B that states: " Quality Assurance includes quality control, which comprises those quality assurance actions related to the physical characteristics of a material, structure, component or system which provides a means to control the quality of the material, structure, component or system to predetermined requirements."
PG&E's present Quality Assurance Program includes not only the Quality
% Assurance Department but also the Quality Control groups located in each x, department associated with nuclear activities. The Quality Assurance Depart-ment and the Quality Control groups collectively monitor implementation of the Quality Assurance Program with the Quality Assurance Department perf o rming an overview function.
While the PG&E Quality Assurance Program is structured so that the audit function of the Quality Assurance Department is highly visible, the Quality Assurance Department has always had, and executed, the responsibility for Quality Assurance Program definition and for assuring that an appropriate program is established. This is discussed further under Programmatic Finding 1.
Conclusion 2: "A general weakness existed in internal and
. external interface and document controls. This questions whether appropriate design informatiu, was being exchanged and utilized by design groups and consultants. One concern is if the latest Hosgri seismic data was inputted for design analysis."
Comment:
l On page 3 of the Report it is stated, "Where procedural coverage was not in l place, the design documentation was used for determining if positive though l informal controls were practiced." There is no discussion in the Report which indicates
- what conclusions were reached when design documentation was reviewed.
The Report does not clarify that many documents were reviewed such as memos, instructions, and interdiscipline letters which were issued to further implement and supplement Quality Assurance Manual procedures. These documents provide evidence that PG&E did indeed exercise controls for internal design documents. This is discussed further under Programmatic Finding 3.
Conclusion 3: "The design verification program was not formalized and was inconsistently implemented and documented.
This included major gaps in design overviews of the design approach for mechanical and other equipment."
Co= ment:
Contrary to the conclusion, PG&E's design verification program was formalized in Quality Assurance Procedure. PRE-6 issued in 1970. The Report concludes that there were major gaps in the design verification based on a "three element" definition which we believe goes beyond even present standards in some respects.
O,Wesubmit L that this definition is not appropriate criteria for this audit.
The Report on page 2 clearly states under Evaluation Criteria that selected parts of ANSI N45.2.11 were considered for guidance in the quality assurance program review. The application of these criteria was inappropriate and apparently resulted from a misinterpretation of Attachment 1 to the November 19, 1981 order, paragraph 1(a)(4) which intended that the criteria for the design verification program consider the relevant guidance in ASSI N45.2.ll. This is discussed further under Programmatic Finding 4.
D O
l l
)
V. FINDINGS Programmatic Finding 1: " Quality Assurance as defined in the QA Manual was essentially an audit role. The Quality Assurance
. group was not assigned a primary role in determining QA requirements."
Co::: ment :
. The Quality Assurance Department, including its predecessor, Quality Engineering Department, has been responsible for the development and issuance of corporate policies and procedures since the first issuance of the Quality Assurance Manual in 1970. Implementation authority and responsibility for the Quality Assurance Program is assigned to the Manager, Quality Assurance (previously Director, Quality Assurance). Revisions to the Policy Sections of the Quality Assurance Manual have always been reviewed and approved by senior management of PG&E and by the Manager, Quality Assurance. These revisions have always been developed by the Quality Assurance Department with due consideration to comments received from other affected departments.
I The Checklist on page 37 states: "Until March 1972 Quality Engineering, which had responsibility for total quality program . . . ." This statement v supports our comment that the Quality Assurance Department is responsible for developing, issuing, and auditing implementation of the program.
If we interpret these Checklist comments as " deficiencies" against Appendix B requirements for independence of Quality Assurance, we are confident in asserting that for the 1970-72 period such an organizational arrangement was not unusual, but was probably typical and in conformity with accepted practice
- in the industry.
l Programmatic
- l. Finding 2: "PG&E had no procedure for assuring the completeness I of the QA program to address the requirements of 10CFR50, Appendix B."
Comment:
l We disagree that a procedure is required to demonstrate a one-to-one correlation of the Quality Assurance Program with the elements of Appendix B.
On page 47 of the transcript, Mr. Reedy stated when asked if PG&E had any procedure which was intended to accomplish this activity:
"No. There was no procedure and I don't know that a O procedure is really required."
l
O The finding implies that PG&E's Quality Assurance Program did not completely address NRC requirements. PG&E committed to develop a program consistent with the proposed requirements of 10CFR50 Appendix B in the Unit 2 PSAR Appendix G in 1970:
"The organization of the material in this Appendix
. is based on the divisions used in the proposed
" Appendix B - Quality Assurance Criteria for Nuclear Power Plants", issued for comment by the Atomic Energy Commission on April 17, 1969. Since there is substantial overlap in the material covered in the various headings, the subject matter of an individual division should not be considered separately but in relation to the entire Appendix."
This commitment is further emphasized in the Letter of Authorization for the Quality Assurance Manual signed by J. F. Bonner in 1970:
". . . the PG&E . . . Diablo Canyon Site, will be designed and constructed in full compliance with the Quality Assurance Program developed from the plan described in Appehdix G of the Preliminary Safety Analysis Report . . . ."
From the foregoing and a review of the manual, it can be seen that the Revision 0 Quality Assurance Manual was based on the best available guidance from the NRC ( AEC) at the time; e.g. , the then draf t version of 10CFR50, Appendix B. The Policy Section of the Quality Assurance Manual in Section 2.1.2 states: "The Director, Quality Assurance . .. . His responsibility is to review continually the Quality Assurance Program and to report on its adequacy and the extent to which it is being carried out."
This Policy Section was issued in 1970.
Programmatic Finding 3: "There were no provisions for document control of correspondence and design documents."
Comment:
There were programmatic requirement provisions in the Quality Assurance Manual for document control of correspondence and design documents. There were controls which described internal design interfaces and responsibilities for document control; however, these controls were contained in the work procedures and the implementation assigned to the Responsible Engineer. (Examples are given in Attachment 3.) A specific procedure concerning these items only was not issued at that time. The Checklist, page 27, refers to Quality Assurance s Manual, Revision 0, Section 3.2.3, Infor=ation Control, which states:
l l
i l
l
-- " Written procedures are required to control ,
correspondence, drawings, specifications, manufacturers' _.
information, etc. , in order to assure that the engineers
~
and designers involved in the design (1) receive the necessary information to develop the design, and '
(2) are kept informed of the design being developed by other interfacing design organizations. These procedures provide for the review, approval, release, distribution, and revision of design documents."
~
Also, during the exit interview it was stated that probless found with
- consultants had not been found to.crist with equipment suppliers.
m This finding on document control is discussed again in the Report as Finding 3 under " Implementation Deficiencies."
Implementation Finding 5: "There was no effective document control syrtes established." .
Comment:
These controls were implemented in project instructions, Engineering discipline' memos, and departmental procedures and were not part of a formal system of procedure control such as we have today. ,
Page 39 of the Checklist states that coordination activities were controlled by memoranda, tut the memoranda we're not part of a formally controlled' system.
Page 24 of the Checklist notes that drawing control procedur,es were acceptable for PG&E drawings. Page 66 of the Checklist statet tharf "evi,dence exists to verify that correspondence was reviewed but no formal procedures were used."
These statements all support our comment that controls did exist.
- The Report implies that the overall drawing control system was inadequate.
This conclusion is apparently-based upon reviewing five or six pipe support i drawings (see page 87 of the Checklist). It is not appropriate to attempt
. a conclusion based on only a review of the pipe suppor,t drawings. PG&E recognized the control of pipe support drawings involved problems due to the number of supports and the number of revisions required. This was documented
- in a 1978 Nonconformance Report in which special procedures were writt,en to -
l control support drawings and changes.
i -
t s' s
,/ e r: av l' ^
,- ~;
m -
i f.
[ ' ] *l**'. y '
t l
f~,. p
- ~ ,
~
. ne' w /*
./ . i ? ', ~
w t g- -
a,
~
-13_
/
O Programmatic Finding 4: "During Phase I, there were no controlled procedures for design control, design interfaces and design responsibilities. PRE-9 and PRE-10 on these subjects were released in 1979 and are to be audited during Phase II."
l <.'
Comment:
The're were existing controlled procedures in the Quality Assurance Manual covering these program elements. The Checklist, particularly on pages 17 E through 24, indicates these were reviewed and in many cases found to be acceptable. These requirements were addressed in Section III of the Quality Assurance Manual Policy and the following procedures:
PRE 2 - Design Development
- PRE 3 - Drawing Preparation, Review and Approval PRE 4 - Specifications PRE 6 - Comprehensive Design Reviews In addition to these, there were " uncontrolled" implementing procedures (Attachment 3) as noted by Mr',. Reedy. (Tr. p 54)
It should be noted that during this period, PG&E's use of consultants for engineering services was limited since a very significant amount of Phase I design work was done in-house and was completed prior to June 1970 when 10CFR50 Appendix B was issued.
As we statec at the November 3,1981 meeting, the NRC and PG&E, in 1977, recognized that comprehensive design reviews needed to be completed on all licensee designated Class I structures, systems, and components prior to fuel loading. PG&E agreed to perform the design reviews. Also in report No. 50-275/77-12 issued in 1977, the NRC stated that PG&E Procedure PRE-6 provided adequate guidance for the conduct and approval of comprehensive design reviews.
- This finding on design controls is directly related to Finding 4 discussed I in the Report under " Implementation Deficiencies."
Implementation Finding 4: "PG&E design verification on in-house activities att suppliers was unstructured and applied inconsistently.
'Je consider that design verification consists of the following three elements:
i "1) Design overview for design approach, methods, design input selection, and assumptions.
{ "2) Detailed checking of design steps and completed design documents.
e f
O "3) Verification of approved "As Built" condition against approved design.
" Activities for element 3 were not initiated ,
until 1979 and are to be reviewed during Phase II.
Documentation showed detailed checks to be. perf ormed on PG&E work with design overviews being performed on a selective basis. Most of what PG&E refers to as Design Reviews consists of element 2."
Comment:
As stated above, PG&E's design verification process was structured in Quality Assurance Procedure PRE-6 and was reviewed and found acceptable by the NRC in report No. 50-275/77-12.
Further, the three elements together are not appropriate because:
- Elements 1 and 2 can be acceptable methods for design verification even when used individually.
They are not consistent with the Commission order dated November 19, 1981 even for applying today's O criteria.
Element 3 is not required for design verification by applicable quality assurance requirements.
The Report draws the conclusion that most of PG&E's design reviews consisted of element 2 which in itself is consistent with N45.2.11. While listed as an implementation deficiency, it appears to draw the conclusion that this is a programmatic deficiency also. PG&E submits that in the 1970-74 ti=e frame, element 2 was generally accepted industrywide (until N45.2.11 and Regulatory Guide 1.64 were issued for additional supplemental guidance).
Regarding element 3, it clearly has merit for some situations, but it is neither required nor implied by current Regulatory Guides, ANSI Standards, or 10CFR50. PG&E has always maintained and has always had procedures and requirements for as-builting, and has never considered as-builting as an aspect of design verification. Simply stated, a design requirement is considered "as-built" when the construction or installation is verified as being within the tolerance, configuration, etc., of the design document. PG&E has always maintained a construction inspection engineering staff to assure that the plant is built as designed; and when deviations are encountered, they are as-built and returned to the Engineering Department for approval.
O O
\
Programmatic Finding 5: "PG&E did not require design consultants to implement Quality Assurance requirements."
Comment:
PG&E concedes that during the Phase I period, the program did not formally
'
- require design consultants to implement Quality Assurance program requirements.
Guidance on control of procurement of items and services was widely interpreted, cnd we submit that it was neither uniformly understood nor typical industry practice to impose such requirements specifically on design consultants during ossentially all of the Phase I period. Evolution of the PG&E Quality Assurance program as previously discussed in Section II was consistent with the rest of the industry.
This finding on quality assurance requirements for design consultants is also discussed in the Report as Finding 3 under " Implementation Deficiencies."
Implementation Finding 3: " Design consultants were not required to implement Quality Assurance Programs."
O Com=ent:
This finding does not clarify that PG&E began reviewing the quality assurance programs of seismic consultants in 1977 and that all consultants were required to submit a Quality Assurance Program starting in mid-to-late 1977 and were reviewed and qualified by mid-1978. As indicated on page 46 of the Checklist, two consultants implemented a Quality Assurance Program prior to mid-1978 -
one in 1974 and the other in 1977.
In addition, PG&E utilized a Responsible Engineer in residence for seismic testing. (Tr. p 89)
Program =atic
- Finding 6: " Corrective action provisions were not addressed except with respect to audit deficiencies and deficiencies at the site."
Comment:
The Transcript clarifies that the Report only applies to safety-related seis=ic design activities. Looking at this finding with that viewpoint, this finding
/N is correct. However, the Report does not clarify that this programmatic
\s- deficiency was corrected in late 1975.
O The Quality Assurance Manual for Operating Nuclear Power Plants issued for trial use in March 1975 contained a preliminary version of Quality Assurance Procedure (QAP) No. 8.1, "Nonconformances." Page 52 of the Checklist states under Audit Instruction, " Prior to issuance of PRM-3 (2/1/78), what measures were established to identify and control conditions adverse tct quality and obtain corrective action to prevent recurrence?" The implication here is that PRM-3 was a satisf actory procedure. It should be noted here that PRM-3 was developed for the Design and Construction Quality Assurance Manual from QAP 8.1. The Report did not identify the existence of QAP 8.1, presumably
- because it was contained in the Quality Assurance Manual for Operating Nuclear Power Plants.
Although contained in the Operating Quality Assurance Manual, QAP 8.1 was applicable to Diablo Canyon Power Plant work at the time it was issued as evidenced by the following statements from the September 26, 1975 cover letter issuing the manual:
"With this letter the revised Manual is being distributed.
The new version has been approved and signed by the President and the Senior Vice President. It is important that any aspects of the quality assurance program which may not yet be fully in effect be implemented without delay."
() " Inspectors from Region V, Nuclear Regulatory Commission, have been reviewing the Manual and have begun inspections By the to evaluate our implementation of the program.
early part of November we will need to demonstrate to the NRC inspectors a fully implemented quality assurance program at Diablo Canyon or run a risk of delaying the operating license."
"The Quality Assurance Department will be conducting audits in the near future to evaluate the effectiveness of the program."
QAP 8.1 contained the following paragraph under Section 2.0, APPLICATION:
- "2.1 The Plant Superintendint, Manager, Station Construction, and Project Engineer are responsible for seeing that non-conformances occurring in work under their control are promptly disposed of in accordance with this Procedure."
This paragraph clearly describes the Project Engineer's responsibility to disposition deficiencies found in design activities in accordance with QAP 8.1.
QAP 8.1 was developed during the review of the corporate program which took
'T place for the Operating Quality Assurance Manual in anticipation of Unit 1 going into operation in late 1975 or early 1976.
l p
This finding on corrective action provisions for design deficiencies is directly related to Finding 2 discussed in the Report under " Implementation Deficiencies."
t Implementation Finding 2: "The PG&E audit system and corrective action system were not effective."
Comment:
It is inaccurate to imply that the overall audit and corrective action systems were not effective. All of the substatements made for this finding apply to the pre-1976 period. As discussed above, QAP 8.1 was issued in September 1975 to correct this programmatic deficiency.
Page 53 of the Checklist shows aspects of the PG&E audit program that were found acceptable. Page 31 of the Checklist shows the portions of the Quality Assurance Program covering corrective action for suppliers, contractors.
on-site discrepancies, and audit results, all of which were found acceptable, s_s/ In the audits conducted in the Engineering Department between 1969 and 1978, deficiencies were identified and corrective actions were specified. All audit reports were distributed to senior management.
We would also note that Mr. Reedy clarified that he did not intend to imply that corrective action was not taken; rather, he did not see a corrective action document that he was accustomed to seeing. (Tr. p 84)
The third subparagraph of this finding, which states: " Corrective action verification was by re-audit only" apparently finds verification by reaudit to be a deficiency. This practice was an acceptable technique recognized by the NRC (AEC) guidance document WASH 1283 (Revision 1, May 24, 1974), which contains Draf t 3, Revision 4, of ANS1 N45.2.12. Section 4.5.2 thereof states (for the auditing organization):
" Followup action may be accomplished through written communication, reaudit, or othe,r appropriate means."
(Underline added.]
O
O Program =atic Finding 7: " Indoctrination and training were not addressed in the QA Manual or procedures."
Comment:
, The 1970 Quality Assurance Manual, Revision 0, contains Appendices which are bound into and made a part of the manual. Appendix B thereto, page B-6 states:
" Personnel involved in the Quality Assurance Program will be trained and qualified in their respective fields. Personnel performing non-destructive examina-tions must meet the qualification requirements of appropriate Codes and Standards. Proficiencies will be monitored and training activity taken when necessary."
This topic is discussed in the Transcript on pages 47 and 48:
fe's MR. REEDY: ". . . There was a statement in the program that there would be training and indoctrination but nothing as to how it would be done. That you will do this. Each manager will do that, et cetera. We put down this as a programmatic deficiency in the respect that no detail -- just addressing it as an overall thing -- it wasn't provided, but other people could make a different interpretation of that. The fact that it did address it could be interpreted differently."
MR. BISHOP: "So you concluded there was no indoctrination or training?"
MR. REEDY: "It wasn't formalized and documented to the extent that you had controlled documents showing who received what training when, on what subject."
O
m Programmatic .
Finding 8: "The QA Manual contained no provisions for PG&E management review of the QA program for status and adequacy."
Comment:
The requirement in Appendix B is:
"The applicant shall regularly review the status and adequacy of the quality assurance program. Management of other organizations participating in the quality assurance program shall regularly review the status and adequacy of that part of the quality assurance program which they are executing."
We believe the process described below constitutes a review by managenent and satisfies the intent of the Appendix B requirement.
As noted in our comments on Programmatic Finding 1, all program policy changes have been reviewed and approved by senior management for authorization.
This review compared the revisions with Appendix B requirements. We have always considered that a revision to a procedure constitutes a review of that procedure. Between 1970 and 1978, there were at least five revisions made to each individual section in the Policy See. tion of the Quality Assurance Manual.
Further, audit reports have been distributed to senior management as verified on page 35 of the Checklist. NRC I&E Reports received the same distribution as PG&E Quality Assurance audit reports.
This topic is discussed on page 47 of the Transcript:
MR. REEDY: "For example, there has been an awful lot of discussion about Item 8 which is management review of the QA program; and there was management review of all che changes at the program. There was management review of all the audits, but the verification of that was extremely difficult to find. It wasn't documented that we could pull out as a controlled thing that yes, this occurred. It occurred because you might see a signature on someone's revision to the program but . . ."
7- ~
\~ A
( ) Further, on page 65:
MR. REEDY: "Now, there was QA or review on the part of management of audits. There was certainly review of program changes, so there was a management involvement in that, but we were trying to hone in specifically on the activity to verify the effectiveness of the program, -
, and to me a review of audits doesn't tell you'how effective the auditors are doing their work. It tells you what they found but not were they really effective auditors, and that is the intent of my statement."
MR. BISHOP: "What do you expect to see to have a
- positive finding on that type?"
MR. REEDY: "I would expect that someone would go in on a periodic basis to see if they are arriving at similar conclusions to what the auditors would be finding, the internal PG&E auditors, finding out if they are effec-tive. In other words, do you concur with the type of findings that they are coming up with?"
MR. BISHOP: "The third party?"
MR. REEDY: " Third party, accompanying people, but just O doing the same function again."
We observe that only in recent experience has the NRC encouraged use of third party auditors, or has the industry regarded them as useful.
In addition, page 68:
MR. HAASS: "Was there implementation deficiencies tied in with any use of stricter QA requirements than one would read in Appendix B?"
MR. REEDY: "I think it would be easy for people going back to 1970 to say our management involvement in assessing our program and having discussions with QA managers and looking at audit reports and talking to people was adequate to fulfill that requirement in 1970.
That type of activity in my mind was the type of activity that was carried on in 1970. It was certainly carried on there, but now when you are looking for how is that documented as to what occurred? Where are the minutes? They weren't there. Again this is a very strict interpretation of today's requirement. What you would look for today. You would look for some kind of minutes of the meeting of what occurred. So I'm not
() saying management was not involved. That is not the intent of that state =ent."
O This finding on management review is discussed again in the Report as Finding 1 under " Implementation Deficiencies."
Implementation Finding 1: "PG&E management did not review and assess.the effectiveness of the Quality Assurance Program."
Com=ent:
PG&E's senior management has always received and reviewed copies of audit reports. By this process the overall Quality Assurance Program has been continually assessed using both Quality Assurance Department and NRC reports es input. As discussed above, changes to the Policy Section of the Quality Assurance Manual required the review and approval of senior management.
Further, oral reports regarding the status of the Quality Assurance Program were given by the Quality Assurance Director to the Executive Vice President.
The Report mentions a program review by Energy Inc. (December 22, 1975).
We submit that the Energy Inc. review is an important example of where senior management review and assessment were functioning. The Energy Inc. review was commissioned by PG&E management and results acted upon. Indeed, major changes
() in the PG&E -Quality Assurance / Quality Control Program occurred following the
(_) Energy Inc. report. For example, the Engineering Quality Control Department was established in 1976 and charged with the responsibility of developing and issuing a manual controlling Engineering Department activities. Additionally, a new Director of Quality Assurance was hired from outside the Company to provide aggressive and experienced leadership in the Quality Assurance Program.
This documented evidence refutes a finding that there was no management review and assessment of the program. While they may not have been documented, they, in fact, did take place and along with the other evidence noted above support a conclusion that management was indeed assessing the effectiveness of the Quality Assurance Program activities.
\
d
lO VI. CONCLUSION In conclusion PG&E submits that:
- Standing alone, the Report gives an erroneous impression of the pre-1978 PG&E Quality Assurance Program.
- The use of 1982 interpretation of 10CFRIO Appendix B for
, a historical audit of pre-1978 activities is inappropriate; however, PG6E recognizes the impracticality, if not the impossibility, of attempting to reconstruct acceptable interpretations of Appendix B during that time frame.
The April 1, 1982 meeting did much to clarify these issues of concern to PG&E regarding the Report.
- The foregoing comments provide further clarification of our Concerns.
- Finally, we believe that the Report, supplemented by these comments and the Transcript, provides an adequate basis for Teledyne Engineering Services to assess adequacy of the work performed by R. F. Reedy, Inc. , and the implications of his work as they may impact upon the remainder of the Independent O' Verification Program.
O
- ATTACHMENT 1 1 STATEMENT 2 by 3 George A. Maneatis 4 Senior Vice President - Facilities Development 5 Pacific Gas and Electric Company 6 before the
, 7 Subcommittee on Energy and Environment 8 Committee on Interior and Insular Affairs 9 U.S. House of, Representatives 10 November 19, 1981 11 12 13 Good morning. My name is George A. Maneatis and I am the 14 Senior Vice President - Facilities Development for Pacific Gas
() 15 and Electric Company. I am here today to discuss our Quality 16 Assurance Program in the context of the recently identified 17 errors in the seismic design for our Diablo Canyon Nuclear 18 Power Plant and to give our views on the implications of 19 these errors relative to quality assurance for reactors 20 under construction.
21 22 First, let me state that PGandE is committed to construct and 23 operate the Diablo Canyon Nuclear Power Plant in a manner that 24 will assure the public health and safety. Paramount to this 25 commitment is the development and implementation of a Quality ,
26 Assurance Program. The purpose of PGandE's Quality Assurance 27 Program is to provide policies, procedures, controls, and
}
1
I mechanisms to meet the requirements of 10 CFR 50, Appendix B,
(} 2 3
" Quality Assurance Criteria for Nuclear Power Plants and Fuel Reprocessing Plants." Such a program in addition to other 4 practices provides confidence that the designed and constructed 5 Diablo Canyon Nuclear. Power Plant will operate safely and 6 perform satisfactorily. Quality Assurance is a management 7 '
. and administrative tool which prescribes controls to detect 8 and correct errors in the design, construction, and operation 9 of nuclear f acilities and, hen,ce, should minimize the proba-10 bility of undetected errors. However, Quality Assurance .
Il cannot, in all cases, prevent errors from.being made~nor can 12 it guarantee that all errors will be detected. It should be 13 emphasized that Quality Assurance is only one of many factors "
14 which contribute t'o the overall safety of a nuclear power plant. '
() 15 Conservative design and the redundancy and diversity of equip-l 16 ment and systems important to safety are other factors which 17 contribute to safety.
18 19 20 PGandE QUALITY ASSURANCE 21 22 To put Quality Assurance at PGandE in historical perspective, l 23 a construction permit for the Diablo Canyon Nuclear Unit I was 24 issued by the Atomic Energy Commission (AEC) on April 23, 1963.
25 During the initial design and construction stages of the plant, ,,
26 Quality Assurance criteria were being formulated by the AEC
{ } 27 and were formally adopted and issued as Appendix B to 10 CFR 50 2 .
l
1 in June 1970. Thes,e Quality Assurance criteria became
, 2 requirements for both Diablo Canyon units.
3 4 In anticipation of these requirements, PGandE organized a 5 Quality Engineering Section (now called Quality Assurance 6 Department) in November 1969 and assigned it responsibilities for assuring that the design and construction of nuclear power i
, 7 8 plants complied with applicable policies and procedures. In 9 January 1970, PGandE developed and issued a Quality Assurance 10 Manual to govern audits of construction and engineering 11 activities. Since that time, PGandE has continually reviewed 12 and modified, as necessary, its Quality Assurance Program 13 and organizational structure in order to be responsive to .
14 evolving regulato y guidance and industry standards.
() 15 16 When the Quality Engineering Section was first established 17 it consisted of three engineers. Currently we have a total l 18 of 47 Quality Assurance / Quality Control engineers to assure 19 implementation of the Qual'ity Assurance Program. At the peak -
20 of construction activity at the Diablo Canyon Power Plant, 21 the Quality Assurance / Quality Control personnel involved on 22 the project (including contractor personnel) numbered 298 or 23 about one Quality Assurance / Quality Control person for every 24 ter, craftsmen.
25 26 3
1 A further step in the evolut' ion and improvement of PGandE's
() 2 Quality Assurance Program was the consolidation and updating 3 of the Engineering Department procedures into an Engineering 4 Manual.
5 6' Since 1969, over 1400 internal Quality Assurance audits have 7 been conducted of which about 230 were of Engineering Department ,
8 activities. Deficiencies identified during these audits were 9 reported to senior management and corrective actions were 10 specified. These actions have either been or will be imple-11 mented and verified in accord'ance with Quality Assurance 12 requirements. Additionally, throughout plant design and 13 construction, the NRC has continually made reviews of the 14 audits conducted by' aur Quality Assurance Department and O 15 performed independent inspections of various design, 16 construction, and operating activities.
17 18 As an example of the NRC review process, the NRC and PGandE 19 determined in June 1977 th&t the existing design control 20 activities needed stronger controls. At that time, PGandE 21 agreed to complete comprehensive design reviews on licensee-
. 22 designed Class I components, structures, and systems to verify 23 correctness of design. At this same time, the NRC reviewed 24 PGandE's Procedure PRE-6, " Comprehensive Design Reviews,"
25 and found that it provided adequate [d. dance for the conduct ,
26 and approval of comprehensive design reviews (NRC Report
( ) 27 50-275/77-12). Subsequent NRC inspections and PGandE Quality 4
1 Assurance audits of these comprehensive design reviews confirmed that the reviews were completed in accordance with the require-() 2 ments of this procedure. The NRC affirmed the overall adequacy 3
4 of the PGandE program in testimony before the ASLB in October 5 1977.
6
. 7 .
8 IDENTIFICATION AND REPORT OF ERROR 9
10 On September 27, 1981, we determined that an error had occurred 11 which potentially affected the plant's seismic design. As 12 a result, we immediately suspended scheduled fuel loading 13 operations and informed the NRC. -
~
14 15 We immediately assembled a team of engineers to analyze the 16 matter thoroughly and hired an independent consultant to conduct 17 a separate and independent review and analysis. The Quality 18 Assurance Department also immediately initiated an investigation 19 of the error to determine its cause and extent and cooperated -
20 fully in the NRC's independent investigation. We presented our 21 findings to the NRC Staff in formal public br'iefings held on 22 October 9, 1981 and November 3, 1981. Additionally we met with 23 the staff and intervenors in lengthy public work sessions during 24 the week of October 12, 1981. The staff has proposed, and 25 PGandE has agreed in principle, to a f ar-ranging reverification ,
26 program to review and confirm the adequacy and accuracy of work 4
performed for PGandE by its service-related contractors.
( ) 27 5
1 Our findings indicate that although PGandE's Quality Assurance 2 Program has been in overall compliance with NRC requirements, 3 a deficiency in its implementation occurred in 1977 which 4 involved the informal exchange of design information between 5 PGandE and one of our seismic consultants during the seismic 6' reanalysis for the Hosgri event. In this particular case, 7 inaccurate weights for equipment and piping in a limited area ,
8 of the containment structure were prepared and furnished the 9 consultant on an incorrectly or,iented diagram. This informal 10 information exchange did not conform to Company practices 11 and Quality Assurance procedures which required the checking 12 of design calculations. Had these procedures bee.n followed, 13 we believe that the errors would have been detected at that .
~ ~
14 time.
15 16 As it turned out, the error was not identified until a piping 17 design engineer in PGandE's Department of Mechanical and Nuclear 18 Engineering who was reviewing Unit 2 piping systems raised 19 questions about the correctness of the diagram. '
20' 21 During our investigation we also identified a failure of 22 Company personnel to use revised 1977 seismic data for check-23 ing the adequacy of the design of conduit and cable tray 24 supports.
25 l
26 !
[
6
- .~.
1 This error involved a failure to follow established Quality
( ) 2 Assurance procedures which require formal distribution of 3 design data. Again, had these procedures been followed, 4 we believe that this error would also have been detected 5 at that time.
6
- 7 To avoid such errors in the future and to assure that past .
8 errors have not af fected plant design margins, the Company is 9 embarking on a two-phase progra,m. The first phase involves 10 establishment of a retraining program f or engineering personnel 11 which emphasizes the adherence to quality procedures. The 12 second phase involves the retention of a consultant to carry 13 out a reverification of the existing design and to report 14 any deficiencies without restriction.
15 16 Notwithstanding the occurrences of these errors, as well as 17 several other minor errors found during the investigation, we 18 continue to believe that because of the original conservative 19 design and the rather large' margins of safety associated -
20 with the affected systems, th'ere would have been no safety 21 consequences even if the errors had not been detected and 22 the plant had gone into operation.
23 24 25 26 7
. a 1 IMPLICATIONS FOR PLANTS UNDER CONSTRUCTION I
3 The errors which have been identified at Diablo Canyon resulted 4 f rom a f ailure to follow the requirements of our Quality 5 Assure Program. However, a review of the numerous PGandE 6' Qua~ surance and NRC audits shows that while deficiencies 7 occur, .aey are generally detected and resolved in a timely 8 manner. This incident vividly illustrates that no matter how 9 embarrassing or costly an error, may be, it will be reported and 10 promptly corrected. We believe that existing NRC and industry 11 quality assurance requirements when properl,y implemented are 12 adequate to safeguard the public health and safety. Also, 13 the industry record of identifying, reporting, and correcting 14 problems exemplifies a strong commitment to adhere to Quality 15 Assurance requirements.
16 17 We are confident that the errors identified at Diablo Canyon 18 do not represent a generic weakness in the Quality Assurance 19 regulations or in industry's responsibility to design plants 20 that will operate safely. We believe the present program 21 of inspections by industry and the NRC is adequate to assure
. 22 a high degree of reliability in the detection and correction 23 of deficiencies.
24 25 26 27 8
. s ,
1- CLOSING REMARKS .
2 CE) 3 In closing let me emphasize that:
4 5
- PGandE voluntarily suspended fuel loading when 6- we discovered the error and we immediately took
. 7 steps to notify the NRC and to correct the ,
8 error.
9 ,
10 - PGandE is firmly committed to design, 11 construct, and operate Diablo Canyon in a safe 12 manner. In carrying out this commitment, we 13 have: -
14 - Assignsd competent, experienced, and 15 dedicated people to perform these 16 functions; 17 -
Used and continue to use extreme con-18 servatism in the design and construction 19 of Diablo Canyo'n; 20 - Contracted with o'ne of the most qualified 21 consultants in the country to carry out an 22 independent reverification of the existing 23 design and to report any deficiencies 24 without restrictions.
25 - Agreed to make any further changes that ,
26 are identified as a result of the reverification program; and
( ) 27 9
1 - Conducted ourselves and reported the
(} 2 3
results professionally and forthrightly even though it has been embarrassing 4 to the Company, to our engineers, and 5 to the nuclear industry.
6-
. 7 -
I represent the highest level of PGandE .
8 management in assuring this Committee, the NRC, 9 and the public that we a,re totally committed 10 to the continued implementation of a sound and 11 effective Quality Assurance Program,and to 12 building nuclear power plants that are safe in 13 every respect.
14 15 16 17 18 19 20 21 22 23 24 25 26 10 lo e imii
ATTACHMENT 2 DESIGN AND CONSTRUCTION MILESTONES o
DATE DESIGN AND CONSTRUCTION STATUS COMMENTS NOVEMBER 1966 NSSS CONTRACT AWARDED JANUARY 1967 UNIT 1 PSAR SUBMITTED JULY 1967 PROPOSED GENERAL DESIGN CRITERIA ISSUED (10CFR50 APPENDIX A)
APRIL 1968 UNIT 1 CONSTRUCTION PERMIT AUGUST 1968 STARTED SITE PREP JUNE 1968 UNIT 2 PSAR SUBMITTED APRIL 1969 PROPOSED QA CRITERIA ISSUED FOR COMMENT (18 CRITERIA)
JUNE 1969 START CONSTRUCTION MAJOR SEISMIC DESIGN FOR BUILDINGS BUILDINGS COMPLETE JULY 1969 UNIT 2 PSAR APPENDIX G ADDED QUALITY ASSURANCE PROGRM
" PROGRAM WILL BE AN EXTENSION DESCRIBED USING PROPOSED 10CFR50 Of THE QAP DEVELOPED FOR APPENDIX B UNIT 1 . . ."
NOVEMBER 1969 PG&E FORMED QUALITY ENGINEERING DEPARTMEh7 JANUARY 1970 PG&E ISSUED RED QA MANUAL FOR DESIGN AND CONSTRUCTION JUNE 1970 10CFR50, APPENDIX B ISSUED NOVEMBER 1970 START ERECTION OF MAJOR PIPING DESIGN ALMOST HALF COMJLETE DECEMBER 1970 UNIT 2 CONSTRUCTION PERMIT ISSUED
l
\
DATE DESIGN AND CONSTRUCTION STATUS COMMENTS
) FEBRUARY 1971 10CFR50 APPENDIX A ISSUED OCTOBER 1971 N45.3.11 SUBCOMMITTEE l ESTABLISHED l
(LATER RENAMID N45.2.11)
MARCH 1972 PG&E FORMED QUALITY ASSURANCE i DEPARTMENT JUNE 1972 SAFETY GUIDE 23 ISSUED (QAP DESIGN AND CONSTRUCTION)
ENDORSES ANSI N45.2-1971 SEPTEMBER 1972 START INSTALLATION NSSS NOVEMBER 1972 SAFETY GUIDE 33 ISSUED (OPERATING QAP)
ENDORSES ANSI N45.2-1971 JUNE 1973 GRAY BOOK ISSUED MAY 1974 GREEN BOOK ISSUED JUNE 1974 . N45.2.11 ISSUED (QUALITY f'"x ASSURANCE REQUIREMENTS FOR THE
(,,) DESIGN OF NUCLEAR POWER PLANTS)
FEBRUARY 1975 R.G. 1.'4 ENDORSES ANSI N45.2.11-1974 AUGUST 1975 CONTAINMENT LEAK TEST SEPTEMBER 1975 SER #3 APPROVES FSAR NOVEMBER 1975 HOT FUNCTIONAL TEST DESIGN AND CONSTRUCTION COMPLETE FOR FUEL LOADING DECEMBER 1975 NRC ISSUES LICENSE TO RECEIVE AND STORE FUEL FEBRUARY 1976 N45.2.13 ISSUED (QUALITY ASSURANCE REQUIREMENTS FOR CONTROL OF PROCUREMENT . . .)
JULY 1977 R.G. 1.123 ENDORSES ANSI N45.2.13-1976 SINCE DECEMBER 1975, DIABLO CANYON WORK HAS BEEN CONCERNED WITH HOSGRI p)
'\, , NOTE:
, INVESTIGATIONS, HOSGRI REDESIGN, TMI BACKFIT, AND REVERIFICATION OF DESIGN.
,,3 l . U .71.* ='i L ATTACHMENT 3 e j g,4D _
Nov:;mbar 25, 1969 l bchanical Syste: Design Responsibilities For Class I Syste=s and Equipment Units 1 & 2 - Diablo Canven Site h' .ORANDUM TO FILE:
Th2 purpose of this comorandum is to confirm, update and clarify the mechanical system design responsibilities.
i ,
Ecch Engineer (System' Engineer) who has been assigned a system is responsible for l tho total system design including establishment of des,ign criteria, overall design, piping size and instru=entation.
The Lead Engineer is responsible to see that the Engineer (s) assigned to the system carries out the foregoing function.
Generally, either the Lead Engineer or the Engineer is responsible for equipment procurement for this assigned system. Where the System Engineer does not procure the equipment, the equipment procurement responsibility is shown below. The System Engineer, however, is still responsible for the equipment design criteria.
In addition to system responsibility as given above, Mr. G. A. Abbett is responsi-ble for the design of the instrument systems based on the performance criteria established by the System Engin'eer. Mr. Abbott is also responsible for all con-y trol valves.
In addition to system responsibility as given above, Mr. A. G. Walther is responsi-ble for the design of the piping for the various systems (except drainage piping) including hanging, valves and insulation.
The System Engineer is siso responsible for furnishing the Mechanical Design Draf ting Section, Electrical Design Section and Civil Design Section the necessary criteria for the design of items done by the foregoing sections.
Organization With Prime PG&I Orranization
- System or Maior Ecuipment Responsibiliev Lead Entineer Engineer Main Steam and Feedvater PG&E J3Gegan AGWalther j ,
" RPFawcett Steam Gen. Safety Valves Westinghouse JPFinney JPFinney Reactor Coolant JPFinney HJGormly l Chemical Vole =e & Control Liquid Holdup Tanks ' PG&E JPDuffy Westinghouse' JP?inney J?Finney Safety Injection Rosidual Heat itanoval
" JPFinney NLZicmek O Sa=pling Westinghouse GAAbbott GAAbbott
ATTACKHENT 3 b 2-r~' Organization k ,)s With Prime PG&E Organi:ation Systee or Maior Ecuirment Resconsibility Lead Engineer
- Engineer Containment Spray Westinghouse / JPFinney/
PG&E WJLindblad RWWood Component Cooling PGLE JPFinney NLZio=ek
" J3Gegan DOBrand Auxiliary Sal: Water Fire Protection
" WJLindblad RPFawcet:
" JPFinney NLZiomek Diesel Fuel Oil Auxiliary Building Ventilation WJLindblad RWWood***
E:ergency Air
" GAAbbott JJCarpenella Instrumentation & Con:rol Westinghouse / GAAbbott RWMiner PG&E n JVRocca Control Boards & Panels Containment Ventilation PGLE WJLindblad RWWood***
" WJLindblad Containment Isolation Valves
" AGWalther Main Steam & Feedwater Containment Piping Penetration. " WJLindblad JPDuffy Diesel Generator Units (Mech.)
" JPFinney NLZio=ek lhAuxiliaryFeedwater JBGegan RPFawcett Makeup Water
" J3Gegan HJGormly**
R:dioactive Waste Disposal
" HJGormly HJGormly Reactor Coolant Drain Tank
" JPDuffy
" JPDuffy Equipmen: Drain Receiver Tanks
" JPDuffy
.- Gas Decay Tanks Piping Syste=s
" AGWalther JPDuffy WC Ha=
. .$ c w 4 J. O. SCHUY' ,
JGS: pac See attached distribution.
(~ ?* Civil Engineering responsible for de: ail design of tanks.
's /*w* Civil Engineering responsible for de: ail design of ventila: ion systems.
)
ATTACHMENT 3 c DISTRIBUTION
- Engineering Services A. T. Lomas
- R. H. Spalding Elec. Distributien Engineering
- J. W. Colwell Civil Engineering R. V. Bettinger E. J . Ross ,,
Mechanical Engineering D. V. Kelly G. A. Abbott D. O. 3 rand J. 'J. Carpenella J. P. Duffy j
t R. P. Fawcett J. P. Finney f
l J. B. Gegan H. J. Gormly l
W. C. Ham W. J. Lindblad
- R. W. Miner J. V. Rocca A. G. Walther
~~
R. W. Wood N. L. Zier-ek e
ee
+m--.,w- . w, -.w -- - , - - - - - - . - - . ,- - . --- -
w-- -- r-- ,---y 7-,-y -------w,,--- g - - - - --,ew-- - -- m -. - _ _-- -- -_ _ __ _ _ _ -------
~
t"ACit- li t.z Ab A N U t:.L t:.L i t< l L LU
- ATTACHMENT 3 d HEC 1!ANICAL & INCII).R C;CIlic.EPlNG
- 1.23 Drawing Questions Units 1 and 2- Diablo Canyon Site 9
January 7, 1975 MESSES . D. NIELSD!
E. P. WOLI).K Please have Diablo Canyon ECO's or drawings requiring coordination sent to the following engineers:
i
- 1. Turbine Duilding John Sale
- 2. Auxiliary Buildins Richard Bacher
- 3. Centsignant Jesse Ante l 4. Intake Structure Panos Antioch'os They will be responsible for our routing and l returning them to you.
- H. J. COPJG.Y JWSale/ja .
cc: JAAnte
- PGAntiochos RE3acher /
DLPolley/
JBGegan RMLaverty WJStracke AC'a'a lthe r L
m, ...n 2.OCO ren mtu--counny own ATTACHMENT 3 o Wikc. RECElVED Jay PRolECT ENGlHE?R omsim en EPW DIActo r'ayon Sh,N owenmcur ELECTRIC DISTnIbtrrIon ENG1::EERI;;c 7
<" c No.
LETTCn CF 18.25 }{J3 JAN 151974 e d
't0 w er Diablo Canyon Sito Units 1 and 2 DN Preparrtion of Design Reviews For Class I Electric Systems -
f, f
& M~ 2 f January 11, 1974 ME 'OCAMDIT.!:
This memorandum defines the Class I cicetrical systems which will have design revicu and also describes the general requirements for these revicu.t. The specific design revicus covered are thccc that are in the ccope responsibility of the Electric Distribution Department.
Systems and equipment by 1.'estinghcuse on the MSSS Contract are not included, and neither are those in the responsibility of the !!cchanical, Civil and Architectural Departments.
The design revicus shall be prepared es deceribed in Qac);tj Accurance Procedure l'R.E-6 for Diablo Canyon Site and also shc11 ec c. ply with Part IIt, " Des gn Control", of 10CTRSO, Appendix E. Procedure
[ ( '; PRE-b generally is limited to systems in our case.
6 6, L ./
Design revict's shall be made for the following:
1_. Each of.these Class I electric power syctems:
- a. 4160 volt, 60 hert::, Buscs F, G, and H.
- b. 480 volt, 60 hert::, Euscs F, G, and H.
- c. 115 volt, 60 hertz, instrument c.c. power syste...
- d. 125 volt d.c. electric power sys. tem.
- c. 125 volt d.c. emergency lighting systcm.
- 2. Class I clectrical conductors for equipment inter-
- connections, including electrical pencer:tions of the containment.
- 3. C1cas I cicetrical raceasys.
/4 Clacs I heat tracing system.
- 5. Class I ventilation control system.
- 6. Scttinsn of electr'ic.a1 protectivo devices.
- 7. Miner:11 anew:;, as directed by the Supervicing Enginwr.
L-- (n._ )
\ .
2 }. , ,, 4 ;. .
, i ATTAC10iENT 3 f a v s
.N Meinorandum A
- s , , - January 1.1, 19 W -
p
, , - x ,.. ,,
Thedesigncriteriaselected,fu'tbccledtrisi'syetems #"
d shall include those giver dn the TSAR *and ?;f,is, and. the t.EC " General Design Criteria" listed .in Appopdix A of,1CCRF50'to the extent pf cur concittment in the PS/.R and TjdRs, alco'{ included are' the Westinghcurt criteria, and those industry , standards and Company practices in ef fect as the design progressed. Cere should/'.m; taken to snot incl uM thos e, criteria or standards'which have appearid.since'thb design was con.mitted and were not, adopted for the project. , c l %~ ,, ,, ,,_
~
Design reviey means'the' critical review of the' design 19
. order to provide,fu'rther assurance that the actions leading to,the /
design output such as drauings, calculations, analysis, and specifi-cations have been satisf actorily performed and the information
. included in the design output is correct. Established procedures for design revieu shc11 be followed, results of the revicu doccacnted, and measures taken to ensure that the findings arc impicmented.
Whether the review is conducted by one individual or an organization there are a number of basic questions that thould be addressed where applicable:
(1) Were the inputs correctly selected and incorporated ints design? 6 ' ,
. l ',,
(2) If assumption ( were ricccstacy to perform thb design
activity are th'c basis of tbc as,sumptions adequately
' describt.d and reasonabic'? ' , -/ f e"
p (3) Are the cirtvet quality requirc6ents Epucified? '
- >. ; . u%
(4 ) Have the appliEinble codes;'standeirds and c regulatory, requirement ]been met? ,.,e j < %
~,/ s. .
(5) llave the design inter" face , requirements been i satisfied?, [., - h ,, C i
,m, (6)
~
Was an appropristc dUsign y ?]9 F used?
.' "i g?
(7) Is the output reasonable csmpared to inputs?
f l (8) Are the specified parts, equi'p'Eent, and,procenses suitable for the required app (1 cation?
t (9). Are the specified materials et=patibic with each other and the design environmpntal conditions to
' which the material uill be expcsea?" - - '.
- / :.D ,
(10) lias adequate aceccsibility been provided to perfer~i "
needed maintenance and repair?
l l
3 t,
- _ _ ,, y2e- - _- - . . _ .
h / '" (.., (
N. ATTACHMENT 3 g
- ~ :
a W N %, , ,
January 11, 1974 s - y. f. Memorandum . :.', ~ff ~,
[ ',/ (11) llave adequate maintenance requirements been specified?
r &
.'/ (12) Have acceptance criteria been delineated in the design
, I documents, such as drawings, instructions or other
'- supporting documents which are sufficient to assure that adequate standards are maintained and that the activitics prescribed by the design documents have been satisfactorily accomplished?
~
.~
The design review documentation shall be organized and completed to meet the requirceents of PRE-6.
The review shall contain the following:
*# 1. A description of the system covered.
o'~.
- A statement giving the objective and scope of the
- - . 2.
d '3 review. ,
n >'
- 3. A summary giving the results and conclusions reached, 5 ~
and any design deficiencies formed not in accordance
,], q- > -
with the criteria or standards.
Any explanations, comments and recommendations.
V- 4
- /
, .o J ,, / 5. A list of the design criteria and the standards used in the revieu.
- 6. The signature of the review and the Supervising
. Engineer and the date of completion.
.u A set of attachments containing the material supporting 7.
/ * /y,/,
the details developed by the results and conclusions of the revieu. Material should include the analyses and
- data developed by the revicuer and copics of documents pertinent to the review.
The co=plete review shall be assembled in a folder or set of
- folders, labelled with title of the review and then placed in the files
. of the design section.
The status of the design review at this time are as follows:
- 1. The electric power systeus are partially complete, with much of the system analyses completed. The design revice of the Class I motors has been completed.
l PO l
l
- ATTACIDiENT 3 h *l Mtmorandum January 11, 1974 7'
. tA)
- 2. Design review of the electrical conductors for the contaitrnent fan coolers have been ecmpleted. The remainder has not been begun.
3., 4., 5. E1cetrical raceways, heat tracing, venti-lation control, and relay settings have not been design revicued.
- 6. A design review has been completed for the hydrogen releases in the battery rooms.
Persons assigned to perform and complete these design reviews will have access to any records that are availabic, and are requested to make the review as independent as practical of the original design.
7 / D*Y' DONALD NIELSEN DN:YG e
(
o m
9 e
9 O
e e
i
- v
,, ,-r .,, , ydTTACIDiENT 3 i ;. .-
- ../. : p C '
/,/ /[,/
/ ,.e /
- July 2S, 1972
- t Design Chane;er to cnd Approvc1
'~5 fi[ s of Piping Sche.; tics File No. 146.10
!?E"OP.a tm" 7 0 '.i L__T__/.!-1 i c r m' 'ECR'a'TC.% r.::nt:',: :ts :
1.s of neu, escentir.lly all piping schematics are issued "/.pproved for Consrrr tion." 1his remorencum will revise the centrol over changes to piping a schau:t;cz and super:cdc the previous memo of July 29, 1972, on this subject.
Any sys tccc, responsible, instrument or piping enginaer may initiate i a derint. chanye or revision to piping schematics. Changes sheuld be limited to caly tho:e absolutely necessary. This will be done by an intra-company memorant'un form with or uithout sketches. The memorandum shall be addressed to I. F. Hall and shall be routed through:
~
- 1. Responsibic Engineer
- 2. Piping Engineer
- 3. Instrument Engineer
- 4. J. V. Rocca Each will initial a'nd date his approval of the proposed change.
( '( ) Copics will be kept as follows :
yellow - originator pink - mechanical file whit e - I . F. Ha ll Anyone wishing copics for their own file will use the Xerox.
Another area that requires a more formalized control is the "Line Designation Table" (LDT). The LDT will be issued " Approved for Construction" a f ter being reviewed and signed off by oech system engineer. Af ter that, any changes in line nu=ber, flow, pressure or temperature must be processed through
, A. G. Walther/J. R. del Mazo on the attached format addressed to I. F. Hall.
Again, this should be routed through:
- 1. Responsible Engineer
- 2. Piping Engineer
- 3. A. G. Walther/J. R. del bb:o Each will initial and date his approval of the proposed change.
Copies will be the responsibility of each individual.
m f
- n-h' . I -M
[ / H. J. Gormly JYPocca/ja Attachment cc: IFHa ll
ATTACHMENT 3 j
,u3.s
. .. en... . >. . . ..1 u ... . . .. - p ..
a v
Line ';u. .
Date Chang.c Dc.ncr *p don i
9 S igned by:
System Engineer .
e:.
b; Counterr.fr,ned by:
O I have read the foregoing 22 page document entitled
" Comments on the R. F. Reedy, Inc., Quality Assurance Audit Report on Safety-Related Activities Performed by PGandE .
Prior To June 1978" including Attachments 1 through 3j.
The contents of said document and attachments are true and correct to the best of my knowledge, information and belief. .
June 24, 1982.
% I Warren A. Raym M; z -
Manager of Qua ity Assurance Pacific Gas d Electric Company i
Subscribed and sworn to before me this 24th day of June, 1982.
'7 *
)
SEAL
Nancy J. Eemaster, Notary Public '
l in and for the City and County of San Francisco, State of California.
My Commission expires April 14, 1986.
r 7
V
- - - - - - - _ - - -