ML20136B336
| ML20136B336 | |
| Person / Time | |
|---|---|
| Site: | Diablo Canyon, 05000000 |
| Issue date: | 04/15/1982 |
| From: | PACIFIC GAS & ELECTRIC CO. |
| To: | |
| Shared Package | |
| ML20136B092 | List:
|
| References | |
| FOIA-84-293 NUDOCS 8601020554 | |
| Download: ML20136B336 (22) | |
Text
P
^
~
4/15/82
'[
n 4
COMMENTS ON THE R. F. REEDY, INC., QUALITY ASSURANCE AUDIT REPORT ON SAFETY-RELATED ACTIVITIES PERFORMED BY PG&E PRIOR TO JUNE 1978 i
o 4
I.
INTRODUCTION At the April 1, 1982 meeting with the NRC, Pacific Gas and Electric Company (PG&E) committed to submit a formal respense 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 m development and refinement of the PG&E Quality Assurance Program, and to
) 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 Subcommittee on Interior and Insular Af fairs of the U.S. House of Representatives (Attach-ment 1).
In that statement, evidence was provided ar to the dedication of PG&E to formulating and implementing an effective quality assurance program.
We believe this presents a much more quality conscious image than that obtained "
from the Report.
PG&E did provide controls for the work performed, consistent with our commitments to regulatory requirements and their interpretations in esistence 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.
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 nov 16 years since the first major 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.
l l
Fo/A-NM3 l
f y LEIGHTOS4-293 PDR
.~
O
.q
_2-I i
j 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 and 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.
~
in mind in reviewing the following brief outline of significant milestones of
]
the design and construction of Diablo Canyon as referenced to the development j
of NRC Quality Assurance guidance.
We recognize the dilemma faced by R. F. Reedy, Inc., in attempting to go back and rememoer 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.
i (Tr. p 14)* As a result, it.is extremely difficult to achieve meaningful results from this type of audit, where the history is so extensive and. quality assurance requirements were evolving as project work proceeded.
j s
! ij i
i 1
l.
~
l
)
i 1
1 7
i i
?
- Refers to Teanscript of April 1,1982, NRC " Meeting With Pacific Cas and
~s' Electric Company to Discuss Seismic Design Review, Diablo Canyon Unit 1."
j f
[
2
_. ~. _. _... _. _,.. _ _
l '.,
's O,
g...
9 O 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 -
i approximately 2 1/2 years before the proposed Appendix B to 10CFR50 was issued for comment.
The Unit 1 PSAk was submitted in January 1967, prior to the identification of formal quality assurance requirements, and thus the require-cents were not specifically addressed.
Supplement 5 to the Unit 1 PSAR was submitted in November 1967 and described a quality program for the primary cystem. 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 j
construction on Unit 1, the proposed 10CFR50 Appendix B was published for Commeat.
_~
Formal quality assurance requirements were first addressed in Appendix G to the Unit 2 PSAR which was submitted in July 1969.
That appendix addressed _
y g
10CFR50 Appendix B requirements as follows:
QUALITY ASSURANCE A comprehensive Quali'ty Assurance Program for the design and 1
construction of Unit 2 at Diablo Canyon will be established h
by Pacific Gas and Electric Company in the discharge of its 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.
i This Appendix discusses the scope of the program, its organization and management, and the implementation of the 4
technical and sy' stems 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 i.
division should not be considered separately but in relation
}
to the entire Appendix.
i i
PG&E established a Quality Engineering Department in November 1969 to implement l
this connaitment to a Quality Assurance Program, and two months later the Company issued a Quality Assurance Manual for Design and Construction. This tanual 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 i
v Program and organizational structure in order to be responsive to the evolving regulatory guidance and industry standards.
I t
i
... - ~
(
n -
Erection of major piping systems for Uritt I 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.
i During this period,'there were several concurrent industry and NRC actions that had a significant impact on the evolution of quality assurance require-ments. Appendix 5 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 example of j
l
.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 was issued - a period of approximately five years.
i The NRC endorsed this standard in July 1977 - about six years after work on the standard was started.
By that time Diablo Canyon had completed construction, g
i sfl 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.
i For example, the Standard Review Plan references 17 Quality Assurance j
Programmatic Regulatory Guides which are applicable to present day design and construction of nuclear power plants.
Many of these Regulatory Guides j
were issued after Unit I 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, we are attaching a chronology of significant milestones of Unit i design and construction referenced to NRC regulations, industry standards, and Regulatory i
Guides.
(Attachment 2) i i
?
l 1
]
3:
I I
(
6
f O
q l
- q III. APPLICATION OF AUDIT -CRITERIA A principal concern which was discussed and acknowledged at-the April 1, 1982 eeeting 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 1
assurance program during the pre-June 1978 period.
In 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 an erroneous and misleading perspective regarding the relative effectiveness l~
.of the PG&E quality assurance program.
f i
In our meeting on April 1, 1982, Mr. Reedy stated that.the. Report did not I
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 j
the proper base for the development of audit criteria.
i I
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, j
that Mr. Reedy was not in agreement with that approach, but felt it was the k~
i caly 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
- strictnea.: of the criteria used to make the statements made in the Report, l
Mr. Reedy stated:
(
l "I have to agree with you that the strict interpretation that I
ve use was' completely unfair.
The order to se did riotTeem - -- - 7 '
'~;f" l
to be fair to begin with,'and I made a comment at the'feime~
that the evaluation should be done to the criteria th'T Vas
~"
','"W,.i
--=
't a
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
.i accept what the NRC audited and what they accepted.
We will i
use the criteria in the order.
Now I don't think that is fair, but that is what we did."
(Tr. pp 32 & 33) i 1
l 3
l When 10CFR50 Appendix B was issued in.1970, there insnediately arose many questions of interpretation.- The industry, through the American National.
l l
Standards Institute (ANSI) and the NRC (AEC) recognized the need for further
" guidance" on the interpretation of Appendix B.
Beginning~in August 1970 with
~f h, ostablishment of the ANSI N45 subcommittee, N45-3 and through the present time, s_/
l
e.
t A.
q
,, ~
f 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.
1 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 j
that John Hoch brought out earlier, that this is a. strict criteria when you use the end result to compare to th~e 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)
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 reviewed but that many of the documents ware not " controlled" in accordance with today's standards, and hence no credit was given.
In response to a staff question regarding the general. phi,losophy,given.to audit.,.
team members, Mr. Reedy replied:
.::t:
ag---
r "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 j
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 4
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) 4 d
l l
4
_,-.,,...--....m.-,,_-.-.._,_,._,_..%.-,,.-.,,,,._..,._...-..m..._
, _..,, _ ~,,
s.
Cs.
q
. O i
In our view, these observations should have been noted in the Report to ensure that credit was given for these quality assurance type controls that in fact existed.
This would enable any reviewer of the Report to obtain a l
more 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.
1 i
-t
_t>
.2:
- i :
O e
s-t l
Q q
).S 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 Qualiv; Assurance program for design work was not adequate in areas of policy, procedures and
- implementation.
The Quality Assurance organization had insufficient program responsibility."
{-
Comment:
I The Report appears to view the PG&E Quality Assurance Department 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 g#1 Assurance Department but also the Quality Control groups located in each department associated with nuclear. activities. The Quality Assurance Depart-4 cent and the Quality Control groups collectively monitor implementation of the Quality Assurance Program with the Quality Assurance Department performing cn 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 ' a Finding 1.
l' Conclusion 2:
"A general weakness existed in internal and external interface and document controls.
This I
questions whether appropriate design information i
was being exchanged and utilized by design groups l
and consultants. One concern is if the latest Hosgri seismic data was inputted for design
(
analysis."
(
-.---.,w, w
C'T
'~T s
()
' Comment:
On page 3 of the Report it is stated, "Where procedural coverage was not in place, the design documentation was used for determining if positive though informal controls were practiced." There is no discussion in the Report which
[
indicates' what conclusions were reached when design documentation was reviewed.
l l
The Report does not clarify that many documents were reviewed such as memos, l
implement and supplement Quality Assurance Manual procedures.
These documents j
instructions, and interdiscipline letters.which were issued to further i
provide evidence that PG&E did.indeed exercise controls for internal design documents. This is discussed further under Programmatic Finding 3.
l l
~
~
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."
[
1
' Comment:
Contrary te the conclusion, PG&E's design verification program was formalized in Quality Assurance Procedurs PRE-6 issued in 1970.
The Report concludes that l
there vs:s major gaps in the design verification based on a "three element" j
[. ;.
definition which we believe goes beyond even present standards in some respects.
We submit that this definition is not appropriate criteria for this audit.
i 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 i
program review.
The application of these criteria was inappropriate and i
apparently resulted froc a misinterpretation of Attachment 1 to the November 19, 1981 order, paragraph 1(a)(4) which intended that the criteria i
for the desian verification program consider the relevant guidance in ANSI f
N45.2.11.
This is discussed further under Programmatic Finding 4
,I i
=
j I
l I
i t
,,,_,._.-,,.,--.m.._,
T e- }
?
V.
FINDINGS 4
i Programmatic l
Finding 1:
" Quality Assurance as defined in the QA Manual was essentially an audit role.
The Quality Assurance j
group was not assigned a primary role in determining I'
QA requirements."
i Comment:
The Quality Assurance Department, including its predecessor, Quality j
Engineering Department, has been responsible for th'e development and issuance l
of corporate policies and procedurec since the first issuance of the Quality Assurance Manual in 1970.
Implementation authority and responsibility for eSe 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 coquents received from other affected departments.
The Checklist on page 37 states:
"Until March 1972 Quality Engineering,
()whichhadresponsibilityfortotalqualityprogram..
This statement 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 ar.angement was not unusual, but was probably typical and in conformity with accepted practice in the industry.
- :; = a ent:..t..
L....: :.2:
Programmatic Finding 2:
"PG&E had no procedure for assuring the completeness
[
of the QA program to address the requirements of j
Comment:
i 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:
4 fr
- No.
There was no procedure and I don't know that a l
'.[j)p procedure is really esquired."
l i
4
le l
~
_11_
b The finding implies that PG&E's Quality Assurance Program did not I
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 l
is based on the divisions used in the proposed j
" 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 l
various headings, the subject matter of an individual
' division-should not be considered separately but_in relation to the entire Appendix."
1 This commitment is further emphasized in the Letter of Authorization for the Quality Assurance Manual signed by J. F. Bonner in 1970:
i
"... the PG&E... Diablo Canyon Site, will be 4
j designed and constructed in full compliance with the Quality Assurance Program developed from the plan described in Appehdix G of the Preliminary Safety.
4
( 7.
Analysis Report.
- v l
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 i
from the NRC (AEC) at the time; e.g., the t. hen draf t version of 10CFR50, Appendix B.
The Policy Section of the Quality Assurance Manual in a
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."
i This Policy Section was issued in 1970.
)
l f
Programmatic
~
Finding 3:
"There were no provisions for document control of l
correspondence and design documents."
Comment:
There were programmatic requirement provisions in the Quality Assurance Manual i
far document control of correspondence and design' documents.
There were '
centrols which described internal' design interfaces and responsibilities for, doctament control; however, these controls were contained in the work procedures
. cnd the implementation assigned to the Responsible Engineer.
(Examples are
{
'A specific procedure concerning the;ae items only was
! 'y-( given in Attachment' 3.)
"j not issued at that time. The Checklist, page 27, refers to Quality Assurance i
Manual, Revision 0, Section 3.2.3, Information Control,.which states:
't t
+.
\\-
y
G q
f}
" 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 problems found with consultants had not been found to exist with equipment suppliers.
This finding on document control is discussed again in the Report as Finding 5 under " Implementation Deficiencies."
Implementation Finding 5:
"There was no effective document control syste=
established."
Comment:
These controls were implemer ted 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, but the memoranda were not part of a formally controlled system.
Page 24 of the Checklist notes that drawing control procedures were acceptable for PG&E drawings. Page 66 of the Checklist states that " evidence exists to verify. that correspondence was reviewed but no formal procedures were used."
These statements all support our comment that controls did exist.
- : :11:i: -.
~.
The Report implies that the overall drawing control system was inadequate.
This conclusion is apparently based upon reviewing five or six pipe support drawings (see page 87 of the Checklist).
It is not appropriate to attempt a conclusion based on only. -eview of the pipe support drawings.
' 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 written to control support drawings and changes.
=
O q
.-)
~.!
Programmatic Finding 4:
"During Phase I, there were no controlled procedures i
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."
Comment:
There were existing controlled procedures in the Quality Assurance Manual covering these program elements.
The Checklist, particularly on yages 17 throurgh 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:
J PRE 2 - Design Development PRE 3 - Drawing Preparation, Review and Approval 1
PRE 4 - Specifications
]
PRE 6 - Comprehensive Design Reviews lin. addition to these, there were " uncontrolled" implementing procedures (Attachment 3) as noted by Mr', Reedy.
(Tr. p 54) l 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 stated 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.
e This finding on design controls is directly related to Finding 4 discussed in the Report under " Implementation Deficiencies."
~
Implementation
. Finding 4:
9PG&E design verification on in-house activities and i
suppliers was unstructured and applied inconsistently.
We consider that design verification consists of the following three elements:
"1) Design overview for design approach, methods, design input selection, and assumptions.
ry
j "2) Detailed checking of design steps and completed design documents.
D m
--r
-w,---s'
--e-
'rv
N q
j t
/]
l "3 ) Verification of approved "As Built" condition against approved design.
I i
" Activities for element 3 were not initiated l
until 1979 and are to be reviewed during Phase II.
l Documentation showed detailed checks to be. performed i
on PG&E work with design overviews being performed l
i 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:
1 i
Elements 1 and 2 can be acceptable methods for design verification even when used individually.
They are not consistent with the Commission order a
l
,Q dated November 19. 1981 even for applying today's*
D/
criteria.
- Element 3 is not required for. design verification by applicable quality assurance requirements.
3 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 time frame ~,-- ' "
j element 2 was generally accepted industrywide (until N45'.2'.M 'snd R'egirlator'y"
~
= 'Ar j
Guide 1.64 were issued for additional supplemental guidance).
Regarding element 3, it clearly has merit for some situations, but it is
~
4 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-builtin6, and has never considered as-builting as an aspect
- i of design verification.
Simply stated, a design requirement is considered l
"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.
v 1
.. - -. =... -,
1
,3
- 'l 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 assentially 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.
Thisfindingonqualityassurancerequirementsfordesignconsultants[,isalso -
discussed in the Report as Finding 3 under " Implementation Deficiencies."
Implementation Finding 3:
" Design consultants were not required to implement Quality Assurance Programs."
Comment:
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 startic
- in mid-to-late 1977 and were revie' ed-and qualified by mid-1978.. As indicated on' page 46 of the Checklist,
w two consultants implemented a Quality Assurance Program prior to mid-1978
~
cne in 1974 and the other in 1977.
....2-In addition, PG&E utilized a Responsible Engineer in re,si.dence,f.or, seismic testing.
(Tc. p 89)
Programmatic Finding 6:
" Corrective action provisions were not addressed except with respect to audit deficiencies and deficiencies at the rite."
Comment:
The Transcript clarifies that the Report only applies to safety-related seisnic design activities. Looking at this finding with that viewpoint, this finding l is correct. However, the Reporc does not clarify that this programmatic N/
deficiency was corrected in late 1975.
x
O m
~ c..,,,
r]
9, The Quality Assurance Manual for Operating Nuclear Power Plants issued for i
trial use in March 1975 contained a preliminary version of Quality Assurance Procedurs (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 cont'rol 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:
j "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 to evaluate our implement'ation of the program.
By the early part of November we will need to demonstrate to the NRC inspectors a fully Laplemented 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 theTeffectiveness- -
of the program."
1 2 ' :;--
'::-- ' t;";
- r...
n.
QAP 8.1 contained the following paragraph under Section 2.0, APPLICATION:
"2 1 The Plant Superintendent, 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.
l
,-(
' QAP 8.1 was developed during the review of the corporate program which took l
- ) ' place for the Operating Quality Assurance Manual in anticipation of Unit 1 i
going into operation in late 1975 or early 1976.
e.
~
e- - -
y-,-~,,e-
~n
..,m.-we-,s-a
.-p-w--
ww
O q
_17_
r~'\\
This finding on corrective action provisions for design deficiencies is directly related to Finding 2 discussed in the Report under " Implementation Deficiencies."
Implementation Finding 2:
"The PG&E audit system and corrective action system were not effective."
Comment:
U
- ~~ -
a 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&5 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.
r In the audits conducted in the Engineering Department between 1969 and 1978, s
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 stater:' '"Corre'c't1v'e' ac tio'n ; ; -
~; - --
verification was by re-audit only" apparently finds 'v' rification' by re' audit e
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 ANSI N45.2.12.
Section 4.5.2 thereof states (for the auditing organization):
" Followup action m' y be accomplished through written a
communication, reaudit, or othe,r appropriate means."
f (Underline added.]
\\
\\
_l
A.
m 8 -
m Programmatic 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:
I
~ 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 appropriat'e Codes and Standards.
Proficiencies will be monitored and training activity taken when necessary."
i j
This topic is discussed in the Transcript on pages 47 and 48:
I es MR. REEDY:
"... There was a statement in the program
,J;)
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
(('i.
h--$--_ '
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."
-j F
,,_m.
,,,_,,,__,__...r,-,.,
.,-..-._,..,n
, _, - - +
.-m.
_..y.--..-_,-,,,y..,
e
'7 q
19-0 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:
CE "The applicant shall' regularly review the status and adequacy of the quality assurance program. Management of other organizations participating in the quality assurance progran 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 management and satisfies the intent of the Appendix B requirement.
(Q i
e 9" As noted in our comments on Programmatic Finding 1, all program policy l
changes have been reviewed and approved by senior management for authorization.
j 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 i
each individual section in the Policy Section of the Quality Assurance Manual.
Further, audit reports have been distributed to senior management as verified.
i on page 35 of the Checklist. NRC I&E Reports received the same distribution
[
as PG&E Quality Assurance audit reports.
i I
_ -... r : :.:.- -.
This topic ~ is discussed on page 47 of the Transcript:-
F
' b MR. REEDY: '"For example, there has been an awful lot f
of discussion about Item 8 which is management review of~the QA program; and there was management review of l
all the 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 i
-4 but M
_, _ -. - -.. - ~
_m...
t
, r
O f-A 1 Further, on pagea65:
HR. 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 t ha t, 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 ef fective 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?"
,g MR. REEDY:
" Third party, accompanying people, but just
(. ;)
doing the same function again."
y ke 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 implemeccation 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 goin'g!'
"'i
- ~
2' 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 treren' 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
,e minutes of the meeting of what occurred.
So I'm not
.'.JO saying management was not involved.
That is not the
-~#
intent of that statement."
l l
O q
g.- '
A r
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."
Comment:
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 as 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 resuler 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.
o e
i i
t l
O
,.- )
- O 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 10CFR50 Appendix B for a historical audit of pre-1978 activities is inappropriate; however, PG&E recognizes the impracticality, if not the l
impossibility, of attempting to reconstruct acceptable interpretations of Appendix B during that time frame.
i 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 I
our concerns.
Finally, we believe that the Report, supplemented by these comments and the Transcript, provides en adequate basis for Teledyne Engineering Services to assess adequacy of the work performed by R. F. Reedy, Inc., and the implications of his l
work as they may impact upon the remainder of the Independent
,c_
V.~-
Verification Program.
4 l
~
t r
i
~
.i i
l l
[
l 4
l l
,3 M
i i
_-