ML16341B736
| ML16341B736 | |
| Person / Time | |
|---|---|
| Site: | Diablo Canyon  |
| Issue date: | 11/30/1981 |
| From: | Faulkenberry B, Andrea Johnson, Kirsch D, Morrill P NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION IV) |
| To: | |
| References | |
| 50-275-81-29, 50-323-81-18, NUREG-0862, NUREG-0862-I01, NUREG-8062, NUREG-862, NUREG-862-I1, NUDOCS 8112030289 | |
| Download: ML16341B736 (36) | |
Text
NUR EG-0862 Issue t
PON
~AC QSiL hl YIS CF'llKSpSCiCGoP9 lRSpOotrR of seismic related errors at Diablo Canyon Nuclear Power Plant, Units I and 2 Docket Nos. 50-275 and 50-323 License Nos. CPPR-39 and CPPR-69 Pacific Gas and Electric Company DEOX lg8~3 9
MsKs tume Ig)~g RQ@gg
~ /
(r~+
U.S. Nuclear Regulatory Commission Office of Inspection and Enforcement B. H. Faulkenberry, A; D. Johnson, D. F. Kirsch, P. J. Morrill, Region V
~ps Rfcy Wp0
~+a I
~J,,
E
+*+++
8112030289 811'130 PDR ADOCK 05000275 8
It' PDR
Available from GPO Sales Program
)ivision of Technical Information and Document Control U.S. Nuclear Regulatory Commission Washington, DC 20555 Printed copy'rice:
$2.75 and National Technical Information Service Springfield, VA 22161
NUREG-0862 Issue 1
Inspection Report of seismic related errors at Diablo Canyon Nuclear Power Plant, Units 1 and 2 Docket Nos. 50-275 and 50-323 License Nos. CPPR-39 and CPPR-69 Pacific Gas and Electric Company Inspection at: San Francisco, CA Inspection Conducted: October 14-23, 1981 Date Published: November 1981 pectors: B. H. Faulkenberry, A. D. Johnson, D. F. Kirsch, P. J. Morrill, Region V eport Nos. 50-275/81-29 and 50-323/81-18 Division'of Resident, Reactor Projects and Engineering Inspection Region V Office of Inspection and Enforcement U.S. Nuclear Regulatory Commission Washington, D.C.
20555
ABSTRACT Results of a special inspection by NRC Region V personnel to evaluate:
The equality Assurance Programs and other management control"systems in effect at Pacific Gas and Electric.and at URS/John A. Blume and Associates.
During the period from 1970 to 1981; the extent to which these equality Assurance Programs and management control systems were implemented as they relate to the development, transmittal, and use of safety related informa-tion, and, how the seismic related problems discovered at Diablo
- Canyon, during September-October,
- 1981, were caused and identi-fied.
0 I
SUMMARY
AND RESULTS
~Summa r Ins ection on October 14-23, 1981 (Re ort Nos.
50-275/81-29 and Areas Ins ected:
Special announced inspection by Region V
personne to evaluate:
the quality assurance programs and other management control systems in effect at PGEE and URS/John A. Blume and Associates during the period from 1970 to the present; the extent to which these quality assurance programs and management control systems were implemented as they relate to the develop-ment, transmittal, and use of safety related information; and, how-the recently identified seismic problems involving the con-tainment building annulus areas (Units 1
and 2), were caused and subsequently discovered.
The inspection involved 160 hours0.00185 days <br />0.0444 hours <br />2.645503e-4 weeks <br />6.088e-5 months <br /> by 4
NRC Region V personnel.
Results:
The results of this special inspection indicate there mmay mave been several instances, during the time period between 1970 and
- 1978, in which the requirements of 10CFR50, Appendix B,
were not 'properly implemented.
These apparent failures to pro-perly implement
- 10CFR50, Appendix B requirements will'e evaluated during subsequent NRC inspections to further determine their si.g-nificance, and follow corrective -actions necessary by the licen-see.
DETAILS 1.
Individuals Contacted a
~
Pacific Gas and Electric Com an b.
J. Schuyler, Vice'resident, Nuclear Generation J.
Hoch, Manager, Nuclear Projects
>I. Raymond, Manager, guality Assurance M. Tressler, Supervising Mechanical and Nuclear Engineer R. Bettinger, Chief Civil Engineer J.
- Rocca, Chief Mechanical and Nuclear Engineer R. Locke, Senior Counsel V. Ghio, Senior Civil Engineer J.
Horn, Engineer, Mechanical and Nuclear Engineering URS/John A. Blume and Associates (URS Blume c ~
D. Lang, Project Engineer N. Gleason, Manager, guality Assurance Dr. Robert Cloud and Associates R. Cloud d.
State of California Re resentatives R.
Hubbard
~G. Minor 2.
Recent Problems That Have Been Identified as Related to Containment Buildin Annulus Seismic Related Work Activities
'a ~
Incorrect A
lication of Containment Annulus "Dia rams'n September 27, 1981, the Senior Resident Inspector at Diablo Canyon was notified by the Plant Superintendent of a potential problem with the seismic analysis of the Residual Heat Removal System (RHR).
On the following day,,
September 28,
- 1981, the licensee again contacted the Senior Resident Inspector and stated that apparently Unit 2 piping drawings of the containment annulus area were erroneously used in the seismic analysis of Unit 1 piping.
The above reports were supplemented with letters from PG&E to the Director, Region V, dated September 28, 1981, a'nd September 30, 1981.
In these two letters, the licensee stated that the "diagrams" used to locate the vertical seismic floor response (VSFR) spectra in the Unit 1 containment annulus
- area, above the base slab through the operating deck at elevation 140 feet, were in error.
The "diagrams" used 'to locate the spectra
were applicable to Unit 2, but were identified for use in the selection of spectra for Unit' seismic design.
Also, in the letter of September 28, 1981, the licensee stated that fuel loading will be delayed until a resolution of all activities and modifications necessary for fuel loading are completed and concurrence is obtained from the NRC.
On October 14, 1981, representatives from the Region V office met with personnel from PG&E, at their offices in San Francisco, California.
The engineer who first discovered the inaccurate application of the containment annulus "diagrams" was present at this meeting.
Through discussions with these
- persons, the NRC determined the following:
(1)
The "diagrams," which are included as Attachments 1
through 4 were developed at PG5E and apparently given to URS/John A. Blume and Associates (URS/Blume) on March 8, 1977, for their use in the. development of vertical seismic response spectra for the Uni.t 1 and Unit 2 containment building annulus areas.
'2)
URS/Blume, when given the "diagrams",
knew the "diagrams" were applicable to Unit 2.
However, they were not aware that the Unit 1 and Unit 2 containment annulus areas are mirror images.
Therefore, during the development of the associated seismic response
- spectra, URS/Blume'ssumed that both Unit 1 and Unit 2.containment buildings were of the same configuration.
(3)
- PGEE, upon receipt of the seismic response spectra in May 1977 and July 1977, developed.by URS/Blume, assumed the spectra and associated containment annulus frame.
orientation "diagrams" were for,the Unit,.l. containment since it was identified as such by URS/Blume.
In actuality, the containment annulus frame orientation "di.agrams" represented the Unit '2 containment.
.PGSE, in turn, performed subsequent desi'gn calculations for Unit 1 using the Unit 2 containment annulus frame orientation "diagrams".
PGSE then, knowing that the Unit 2 c'ontainment is q mirror image of Unit 1 containment, "flipped" the "diagrams" for use in performance bf design calculations for Unit 2
- and, thus in turn, erroneously used Unit 1 containment annulus frame orientation "diagrams" for the development of Unit 2 design requirements.
(4)
On September 21,
- 1981, an engineer employed with PGSE in the hanger design group, while performing work which included Unit 2 IEB 79-14 related work. involving the use of the "diagrams",
became suspicious that the supposed Unit 2 "diagrams" did not accurately represent Unit 2 frame locations.
On September 21-22,
- 1981, he continued to investigate this apparent discrepancy and brought it to the attention-of-his-immediate superv'isor.
By
September 24, the responsible Senior Civil Engineer had been informed of the apparent discrepancy.
On September 25, second level PG&E management were. notified and they in turn contacted URS/Blume.
URS/Blume, confirmed that the wrong "diagrams" had been u'sed.
On September 26, PG&E management continued to evaluate the problem.
On September 27, the'lant Superintendent. notified the, NRC Senior Resident Inspector that a problem did indeed exist.
Upon confirmation that wrong "diagrams" were used in the development of Unit 1 design requirements, PG&E reanalyzed the design requirements using the apptropriate containment annulus frame orientation "diagrams"'nd, determined that, as a result of the error, modifications were required to be made on 31 Unit j.
pipe supports.
These modifications involved such actions as adding snubbers, changing the snubber size, adding braces, replacing structural
- members, and stiffening base plates.
As of October 30, 1981 none of the above modifications had been made at Diablo Canyon Unit 1.
The basic cause of this problem appears to be the informal manner in which certain. data were developed by PG&E and transmitted to URS/Blume and the lack of independent review of these data within PG&E prior to submittal to URS/Blume.
Inabilit to Verif Accurac of Hei hts Listed'on "Dia rams" and Used as an in ut or the Deve o ment o
Seismic es onse
~Sectra During the process of determing the extent of the design verification control implemented by PG&E on safety related information submitted by PG&E to.URS/Blume 'and other service type contractors,,Region V personnel requested PG&E officials to verify the accuracy of the weights that were listed on the "diagrams" and used as input into the development, by URS/Blume, of the seismic response spectra for the Unit 1 containment building annulus area.
Ilhen these weights could not be confirmed as being accurate, PG&E personnel took current as-built drawings and recalculated the weights of equipment and piping, in the Unit 1 containment, supported from the containment annulus structural steel.
These newly calculated weights varied from the weights listed on the "diagrams" that were furnished by PG&E to URS/Blume in 1977.
Upon determining that the new weights differed from those given to URS/Blume and used in the development of the seismic
response spectra in 1977, PG&E provided URS/Blume with the new weights and requested they develop new spectra.,
On October 26,
- 1981, Region V personnel were informed that, because of the new weight inputs, changes were being reflected in the newly developed'esponse spectra.
These changes involved a general shift to lower peak 'acceleration and lower frequencies in about one-half (ll out of 20) of the spectra.
In the other nine spectra, the peak acceleration was generally increased.
Sample calculatiqns by PG&E, using two of the "worst case" new spectra showed that piping loads may be significantly, increased in specific cases.
As a result of the above, PG&E officials on October 26; 1981, stated they intend to do a complete review and recalculation of response spectra and recalculation of seismic design requirements for the Unit.1 containment annulus area.
This will include a
review and verification of the model used by URS/Blume in development of the spectra, verification of the.accuracy of the input data provided to URS/Blume by PG&E, development of a new'set of vertical response
- spectra, and recalculation of the seismic design requirements.-
The basic cause of this problem appears to be the informal way in which certain data were developed by PG&E and transmitted to URS/Blume from PG&E, and the lack of independent review this data received within PG&E prior to submittal to URS/Blume.
c.
Failure to Use Latest (Jul 1977 Revision of Vertical Res onse S ectra sn Des>
n of on ust and Cab e Tra Su orts On October 15, 1981, representatives from HRC Region V. met, with personnel from PG&E at their offices in San Francisco, California.
The Senior Civil Engineer who had been designated as the principal PG&E interface person with URS/Blume was present at this meeting.
Through discussions with these
- persons, the NRC determined the following as related to activities conducted under the PG&E URS/Blume contract:
(1)
PG&E designated a Senior Civil Engineer as the principal point of contact or principal interface person for URS/Blume contract activities.
(2)
Under the above designated Senior Civil Engineer, other PG&E engineers were designated as responsible for activities conducted within each building," i.e.,'ontainment. building, auxiliary building, turbine building, etc.
(3)
In addition to the above building engineers, additional engineers were responsible for special type activities that transcended building boundaries.
One of these special type activities was the development of the design (4) of seismic support details for electrical conduit and cable trays.
On August 3,
- 1977, URS/Blume submitted to PG&E, via a
- letter, two copies of "Diablo Canyon Containment Interior Structure - Annulus Vertical Floor Response
- Spectra,
- 7. 5m Hosgri Earthquake".
This set of spectra was a revision to the.previous set of spectra that had been issued in May 1977.
This revision was made because it had been determined that the connection between the crane wall and annulus slab at elevation 140 feet,was best modeled as a
pin type connection rather than a fixed (solid) type connection.
(5)
(6)
The PG&E Senior Civil Engineer, upon receipt 'of the August 3 letter and new set of response
- spectra, added a
handwritten note to the letter and sent the letter and spectra to the assigned containment building engineer.
The handwritten note requested the containment building engineer to get copies to involved parties.
Contrary to the note type request, the letter and revised spectra were not provided by the containment building engineer to the engineer responsible for the design of seismic support details for electrical conduit and cable trays.
As a result, the final design requirements for electrical conduit and cable trays in the Unit 1 and
'nit 2 containment annulus areas. were not based
'upon the final set of response spectra.
This problem was discovered on September 28, 1981.
It was discovered as a result of the review conducted by PG&E to determine the extent of the problem resulting from the improper application of the "diagrams" (see Section 2.a of this report);
On September 28, when the design engineer who was responsible for developing seismic support design requirements for electrical conduit and cable trays was contacted, PG&E officials determined his final designs were based upon tray 1977 vertical seismic response spectra and he had no knowledge that the July 1977 spectra had been issued.
For work directly related to the PG&E-URS/Blume Contract, Region V representatives could find no evidence that, during the 197Q-1977 time period:
A written procedure or document ever existed that specified how and to whom within PG&E specific 'information and data received from URS/Blume should be channeled after receipt.
by PG&E, (2)
PG&E supervisors or managers routinely or otherwise checked with (audited) their design engineers to assure
. they had in their possession and were using correctly the latest design input data, or (3)
A formalized control system ever existed that would routinely or otherwise identify for applicable design engineers the latest design input information that had been received from URS/Blume.
The basic cause of this error appears to be the failure by PG8E to implement adequate control systems,to assure proper distribution within PG&E of data dnd information recei,ved from URS/Blume, to assure all appropriate PG&E personnel received such data and information, and to periodically verify that all appropriate persons were properly using the. data and information.
3.
PG&E Oualit Assurance Program NRC Region V representatives examined the PG8E quality.assurance program and associated procedures as related to,the design and construction of safety related structures,
- systems, and components at Diablo Canyon Units 1 and 2.
This examination was primarily directed toward an evaluation of those portions of the program addressing design control and review, independent verifications, procurement of outside services, and documentation.
Reoortedly, the licensee has implemented the quality assurance program for the design and construction of the Diablo Canyon Unit No; j. nuclear power plant as described in the licensee's anplication for'he construction permit of the No.
2 Unit (see Compliance Report 50.-
275/69-6).
Prior to 1970, the PG&E quality assurance program consisted of standard practices and procedures that had developed over the years.
During 1969 and early 1970 PG&E developed a quality assurance manual (Red Book) which consisted of the basic ouality assurance policies and procedures necessary,to address Appendix B criteria.
During late 1969 and early 1970, the Division of Comoliance of the AEC, Region V inspectors, performed= an inspection of the PG&E quality assurance program.
The results of that i,nspection concluded that the program met the requirements of 10CFR50 Appendix B (effective July 27, 1970).
In addition, the Region V inspectors found that PG&E's activities, r'elated to procurement and construction of Diablo Canyon Unit 1, had been consistent with--the. quality assurance program outlined in the PSAR for Diablo Canyon Unit 2, and that this program met,the requirements of the AEC regulations.
Concerning design activites, the Region V inspectors, during 1969-
- 1970, found that written procedures to guide design reviews were not developed and fully implemented until early 1970.
For example, certain calculations examined by the inspectors in December
- 1969, relating to the design of the component cooling water system, had not been checked by another engineer even though it was reportedly standard practice within the licensee organization to have calculations checked, by another engineer.
Subsequent to the initial development and publication of their quality assurance manual and associated procedures in 1969-1970, PG&E has, throughout the years, periodically revised and upgraded their program.
Revisions to the program procedures were made in
- 1972, 1974,
- 1975, 1977,
- 1978, and.1979.
During the 1977-1978 time
- frame, a significant modi'fication to the prooram was made when PG&E developed and published a set of engineering department procedures which supplemented the procedures contained within the quali.ty assurance manual.
.These new procedures provided more definition in such areas as comprehensive design reView, correspondence.
- control, design criteria, design calculations, design verification, drawing preparation and approval, design documents prepared by A/Es and consultants, and contracts for engineering services.
pp and 1978 when various aspects of the PG&E quality assurance.
program and associated procedures were not properly implemented.
4.
PG&E-URS/Blume Contract Activities Based upon the results of the current (October 14-. 23, 1981) review performed by NRC Region V on PG&E quality assurance program (manu'als and procedures) historical documents, and a review of the. results of inspections (NRC Inspection Reports) conducted from 1969 through, 1981, it appears that the PG&E quality assurance, manual and assoc)ated procedures, as-initially developed in. 1970 and periodically revised and upgraded throughout the years, have been adequate to provide the necessary guidance, if properly implemented, to meet the requi.rements as described in 10CFR50, Appendix B.
- However, as described in Sections 2.a, Z.b, 2.c, 4.a(1), 4.a(5), 4.a(6),
and 5 of this report, there a
ears to have been numerous instances between 1970 a.
During the period of October 14-23, 1981, NRC Region V repre-sentatives conducted a review of past PG&E-URS/Blume contract activities as they related to the development, transmittal, and use of safety related design information and data.
This review was conducted at both the PG&E and the URS/Blume offices in San Francisco.
PG&E first contracted with URS/Blume, for performance of work related to the Diablo Canyon projects, in 1966.
This initial work involved consulting services for investioations on earthguake exposure and design criteria for submission to the Atomic Energy Commission.
The contract
- document, issued by PG&E on October 24,
- 1966, was a short letter which authorized URS/Bluje to perform services for PG&F., as set forth in the URS/Blume proposal submitted to PG&F. on October 20, 1966.
From 1966 on, a series of letters were iss'ued by PG&E to URS/Blume, each of which extended the dollar amount that could be used by URS/Blume for services to be provided to PG&E.
On May 21,
- 1976, PG&E issued a letter to URS/Blume extending the dollar amount of the contract. to do Hosgri Fault evaluation work.
On November 2, 1976, a letter from URS/Blume to PG&E stated that work to be performed within the next several months would involve the containment structure.
On Jul/ 12, 1978, a letter was sent to URS/Blume from PG&E.
This letter confirmed that URS/Blume would continue as
.a consultant for PG&E for.Hosgri seismic evaluation work for Diablo Canyon Units 1 and 2.
This letter also stated that all further work under this contract shall be conducted in accordance, with the URS/Blume Ouality Assurance Manual dated December 16, 1977.
It also stated that during the life of this contract, no changes shall be made to the manual without prior approval by PG&E.
The above described letter appears to be the first written document that PG&E submitted to, URS/Blume committi,ng URS/Blume to a quality assurance program.
However.,
URS/Blume submitted a letter to PG&E. on August 30, 1977, in which they described the quality assurance procedures followed by URS/Blume in the course of their wo'rk on the Diablo Canyon Project.
This letter is included as Attachment 5, and states in part:
"Since our work on the DDE analysis of Diablo Canyon (pre-Hosgri) predated formal QA requirements, we followed good engineering analysis practices including checking o$ all calculations, computer programs, and reports.
Our work on the Hosgri evaluation. of'he Diablo Canyon structures employed the same thorough engineering practices and in addition included many of the formal quality assurance requrements specified.in
Our present and future work on the structural modifications to the Turbine Building will employ complete quality. assurance requirements in accordance with the URS/Blume Quality Assurance Manual."
NRC Region V representatives reviewed the way in which PG&E personnel developed, transmitted, and used information and data related to.the PG&E-URS/Blume contract.
This, review involved discussions with PG&E personnel and included a review of pertinent documents at their offices in San Francisco.
From this review, the NRC determined the following:
(1)
Contract documents prior to July 12,
- 1978, were rather informal documents, of the letter type, and did not really address quality assurance control requirements
'ntil July 12, 1978.
Personnel at PG&E indicate this
(2)
(3) resulted because of.the early time period (1966) in whicl.
the contract was first initiated and because when it was first initiated it was for a relatively small dollar amount.
PGSE did establish a'Senior Civil Engineer as the principal point of contact or principal interface oerson with.
URS/Blume (see Section'.c.
of this report).
,However, engineers under the Senior Civi.l Engineer f'requently communicated information to URS/Blume without going through the Senior Civil Engineer.
Various types of information were communicated to URS/Blume during the conduct of meetings, via telephone communication, and by written transmittals.
The written transmittals were submitted by various means including:
typewritten and signed letters, handwritten transmittal
- forms, and by being handed directly to URS/Blume personnel without the use of a covering transmi.ttal document.
(5)
(6)
Certain documents and data, given to URS/Blume, specifically engineering drawings and supporting data that were processed through PGEE's Design Engineering Department, were well controlled.
They received proper design reviews and approval by supervision which appeared tq fully meet the intent of Appendix B reauirements.
Other documents and data, s'pecifi.cally the "di,agrams" and associated weight data (see Attachments 1 through 4) were not well controlled and appear to have received no design review and verification prior to submittal by PGSE to URS/Blume.
Documents reviewed at PGEE as well as discussions with PG8E personnel confirm that work activities.
on the PGSE-URS/Blume contract, as related to the development, transmittal, and use of information, and as performed by PG&E personnel, were performed in an "informal" manner.
For example:
the development and transmittal of certain data and information were not well controlled, certain data and information were not prooerly reviewed and approved prior to transmittal to URS/Blume, certain design information received from URS/Blume was not properly channeled to all pertinent engineers, and sunervisors apparently did not periodically check (audit) their design engineers to assure they had received and were using the latest design data.
(7)
Because of the "informal" nanner in which activities were performed by PGEE on the URS/Blume Contract, there is little assurance that other errors of the type identified in Section 2 of this report do not exist in work developed under the PGSE-URS/Blume Contract during the 1966-1978 time period.
On October 22,
- 1981, NRC Region V representatives conducted a
review, at the URS/Blume offices in San.Francisco, of activities performed under their contract with PGSE. as related to the development, transmittal, and use of informati.on.
Thi.s review involved the review of documents and discussions with URS/Blume personnel.
As a result of this review, the NRC determined the following:
(a)
URS/Blume engineers were aware that 'the "diagrams" (see Attachments 1 through 4) were a representation of the.
Unit 2 containment annulus;
- however, they assumed. that both the Unit 1 and Unit 2 containments were the
- same, i.e., were not mirror images..
(b)
In general, guality Assurance controls appear to have been followed, to.-the extent as described in the URS/Blume letter dated August 30, 1977 (se'e Attachment
- 5) by URS/Blume engineers in the performance of their work for PGSE.
- However, NRC Region V representatives examined about 40 URS/Blume calculation sheets related to Diablo Canyon containment building work.
Calculation sheets 11 through 14, titled "Containment Interior-Vertical Stiffness-Elevation 114 ft. to 140 ft.", performed on March, l5,,1977,.did not have any documented evidence to show that an independent check had been performed on these calculations.
The URS/Blume. Project Manager could not provide evidence that these calculations had been
- checked, but he thought they had been and the checker failed to sign off on the check sheets.
(c)
A Project Engineer was designated by URS/Blume as responsible for control of all work done on this contract.
All input data to PGSE from Blume appears to have been transmitted by this Project Engineer via standard typewritten transmittal forms.
(d)
The URS/Glume Project Engineer stated that information provided to URS/Blume by PGSE, did in'ome cases, come from personnel
'other than the PG8E Senior Civil Engineer and was given',to various people within URS/Blume other than the URS/Blume Project Engineer.
(e)
The Project Engineer stated that design reviews were performed on'RS/Blume developed information submitted to PG&E; however; these design reviews did.not include associated information provided by PG&E.
(f)
The degree of control and amount of independent review performed at URS/Blume by URS/Blume appears to have been significantly better than that performed at PG&E by PG&E.
5.
'ther Service Related Contracts; NRC Region V representatives conducted a review of several other service type contracts in effect during the 1970.-j.980 time period.
This review concentrated on the development, transmittal, and use of data and information for work performed under these contracts and was in'ddition to the review performed on the URS/Blume Contract (see Section 4 of this report).
a ~
Contract Documents Contract documents for the followi'ng identified contracts were reviewed to determine the degree of defi.nition and specificity included to describe the work to be performed, how it would be performed and reviewed, and the quality assurance controls that would be applied.
(1)
Nuclear Services Corporation (NSC), Contract Nos. 5-'7-73 and 5-5-75 These contracts were issued in 1973 and 1975 respectively, and involved the performance of analyses of postulated pipe failures outside containment (PBOC)..
They were issued on a standard PG&E contract form, instead of a
- letter, and included a description of the work to be performed.
No reference was made in these contract documents regarding the quality assurance controls that would be applied to the work to be performed.
In a letter from.flSC to PG&E dated December 2, 1974, NSC states that because of AEC emphasis on QA, additional QA procedures have had to be applied.
However, it was not until September 28, 1977, that PG&E defined the quality assurance controls that should be applied to these contracts.
(2)
Earthnuake Enaineerin
- Services, Incor orated, Contract No. 5 7 This contract was issued on t1arch 9,
- 1977, and involved a seismic reanalys~s of piping systems.
It was issued on a standard contract form and contained a very detailed description of the work to be performed, how it would be performed, and who would check calculations.,
Procedures were included for checking calculations, and a good definition, of the qual,ity assurance controls that were to be applied to the work were included.
(3)
EDS Nuclear, Incorporated, Contract No. 5-21-78, Issued 41a 31, 1978 This was in general a well defined contract on a standard PGSE contract form. It contained a paragraph which states the contract work will be performed in accordance with Revision 11 of contractor's gA manual dated August 1, 1977.
It also states that no changes will be made in the manual without prior approval of PG&E.
- b. ~iR NRC Region V inspectors examined the PGSE contract files, civil engineering files, and mechanical engineering files.
Yarious contractor's records were selected to be examined as described below:
(1)
Nuclear Services Corporation (NSC) Contract No. 5-17-73 (Unit 1) and Contract No. 5-5-75 (Unit 2) to evaluate the consequences of a Pipe Break Outside Containment (PBOC) event and provide design of components to mitigate the effects of a PBOC event.
(2)
Engineering Data Systems - Nuclear (EDS) Contract No. 5-21-78 to develop a valve qualification program to demonstrate valve operability following or during a Hosgri seismic event.
The examination included letters,
- reports, meeting minutes, records of telephone calls, contract revisions, memoranda, and transmittals from the contractors.
After the contract was underway, the inspector observed that PGSE and NSC used the PGSE drawing control system including the drawing change notices, to control and distribute design drawings.
The NSC design work appears to have been coordinated with the PGSE construction organization, the steel fabricator and PGSE Engineering.
Meeting minutes and documented telephone calls appea'r to have been used extensively to document data, transmitted and agreements reached.
Although not required by PGSE until 1977, NSC did have a
gA program as early as April 1975.
The information requested by NSC appears to be documented in their letters to PG&E as do the drawings and reports forwarded'y NSC to PGSE.
In mid 1975, NSC requested (and obtained) confirmation in writing from PG8E, regarding the accuracy of the major assumptions and source terms for NSC's work.
The inspectors were not able to find any specific documentation of a design revi ew; however, the extensive exchange of telephone calls, documentation of meetings, and drawing review process used for this work appear to demonstrate continuing and.
detailed'esign review by the.engineers and their superviso'rs..
The inspectors observed that NSC had utilized some structural seismic analysis data developed by J.
Blume and Associates (Blume).
The inspectors subsequently asked licensee representatives if this Pipe Break Outside Containment work had been reexamined for validity following the Hosgri seismic reevaluations by Blume.
Licensee civil and mechanical engineering supervisors responded that it had, and that no additional changes were required.
In the case'f the EDS contract, the contractor asked for and was apparently given "Carte Blanche" to exami,ne and copy the PG&E files, for the puroose of compiling a '"master file" to qualify required valves for the Hosgri seismic event.
The inspector observed that in some cases handwritten data was sent to EDS via letter and memo from PG&E, while the EDS work appears to have been sent by letter to PG&E.
As in the NSC
- contract, a great deal of information appears to have been transmitted and documented in meeting minutes and telephone calls.
The FDS contract required EDS to do all their work in accordance with their gA manual which had been reviewed and aporoved by PG&E.
This gA manual reportedly'overed criteria 1, 2, 3, 5, 6, 16, 17, and 18 of the 10CFR50 Apoendix B criteria.
The inspector was unable to find any record of OA audits or identified design reviews of the EDS work.
As with NSC, the EDS work appears to have been reviewed and exymined by the engineers and their supervisors while it was in progress.
The end product of the EDS work appears to be verification of valve operability for the valves listed in the Hosgri. Report (tables 7-7, 7-7A, and 7-8) based on PG&E's and Westinghouse's, liqts of valves required for safe shutdown as well as seismic acceleration.
furnished by Blume and Earthquake Engineering Services vi.a PG&E.
Transmittal and Use of Information Based on the NRC Region V inspectors'xamination of PG&E files for the NSC contracts (5-17-73 and 5-21-78) as well as EDS contract (5-21-78), it appeared that PG&E. kept no records of drawing transmittals to their contractors.
When questioned, PG&E representatives'stated that PG&F., did not retain the drawing transmittal slips after the contract work was completed.
The inspector also observed that design drawings and reports from both NSC and EDS were apparently transmitted by letter or transmittal sheet.
The numerous
- meetings, memos, and telephone calls which were documented, generally have distriution lists including PG&E as well as contractor personnel.
The HSC output consisted of:
(1) design drawings used for fabrication and installation of pipe restraints, jet impingement barriers, and steam venting pathways; (2) recoranendations to change the location of equipment to mitigate the effects of Pipe Break Outside Containment; and (3) reports/analyses of stress in piping systems and for submission to the NRC to document PG&E's Pipe Break Outside Containment analyses.
The EDS work apparently resulted in "master files" on each valve to demonstrate its operability as reflected in tables 7'-
7, 7-7A, and 7-8 of the. Hosgri Report.
PG&E forwarded these; tables to EDS by letter, for review prior to their use in the report.
It appears tha't the seismic accelerations for valves were developed by Blume', and by Earthquake Engineering Systems.
Some of these data were'ransmitted to EDS via a handwritten-note with attached sheets.
In one case, the inspector observed a rough sketch of valve orientation transmitted by letter.
In reviewing the licen'see's files, the inspectors concluded that based on the PG&E documentation examined it did not appear possible to verify all the sources of information used by the contractors.
In the review of contract documents from 1966 through, 1980, it appears that there was a continual evolution of quality assurance control as regards the definition of work, how the work would be performed, and the quality assurance controls that would be. applied.
In general, service type contracts were not brought under the quality assurance umbrella until late 1977 and early 1978.
During this time period, a requirement was established at PG&E to have all service contracts channeled through the guality Assurance Department for review and definition of qu'ality assurance controls.
- Also, during this time frame the guality Assurance Department reviewed all service type contracts, still in existence that were, issued prior to 1978.
They evaluated these contractors'uality assurance programs and issued revisions to the contracts which specified the quality assurance controls that would be in effect for the work performed.
All service type contracts issued after about mid 1977 had good definition of,quality assurance control requirements.
and contain a good definition of how work will be performed.
The degree of "informality" in the review, transmittal and use of information appears to differ ~with each service type contractor.
However, it anpears that the greatest degree of "informality" existed with the URS/Blume Contract (initiated in 1966) with lesser degrees of "informality" existing with other service contractors that performed work at least up through mid 1977 to early 1978.
It is important to note that URS/Blume data were inputed into the NSC contract work and possibly other service contractors'ork.
To date it has not been determined if the data that were inputed needs to be revised as a result of the errors identified to date and described in Section 2 of this report.
6.
Neetin Held with PGSE Officials on October 23, 1981 On October 23,
- 1981, a representative from the NRC Region V Office met with Senior Level Nanagement officials. at their offices in San Francisco.
At this meeting, the Region V representative conveyed the following information:
a ~
b.
As a result of the reviews conducted by NRC Region V
r epr esentatives during the period of October 14-23,
- 1981, NRC has developed
.concerns regarding the general lack of quality assurance controls applied to work performed by service type contractors prior to mid 1977 and early 1978.
PGSE was requested, at the next scheduled meeting with the NRC to be prepared to discuss why the NRC should not require the reverification study to be broadened to include non-seismic safety related service type contract work.
0
. Oy Attachment
.1
Oo Attachment 2
/
lL~ 'j gg.9 Zips
Attachment 3
Attachment 4
URQ I' <';l ie<<4IV<'A< I'<'<e<< e<a',
<<ee e
~
I URS/JOHN A. BLUME8 ASSOCIATES, ENGINEERS 130 JESSIE STREET (AT NE1V MONTGOMERY)
SAN FRANCISCO. CALIFORNIA94105 TEL: (415) 397.2525 CABLE: OLUMENGRS August 30, 1977
<'I:e'el"e I <iY
<'I<<<<'
I<i'e
! I'i'II<1.0 hlr. J.
O. Schuyler
- hlanager, Nuclear Projects Pacific Gas 6 Electric Company 77 Beale Street, Room 2645B San Francisco, CA 94106 FF<'!I Rc,CiIVEi)
D~P IIUCLEAR PROJECTS DES JBH DEPART MF I<IT EI'EP 21977 FILE
SUBJECT:
Quality Assurance for URS/Blume IIIork for the Diablo Canyon Project
REFERENCE:
URS/Blume Quality Assurance hlanual
Dear hlr. Schuyler:
This letter summarizes the quality assurance procedures followed in the course of our work on the Diablo Canyon proj ect.
Due to the fact that the present 10"FRSO, Appendix B quality assurance requirements came into existence well after our work,on Diablo Canyon was begun, it is necessary to separate our work into three parts.
Since our work on the DDE analysis of Diablo Canyon (pre-Hosgri) predated formal QA requirements, we followed good engineering analysis practices
.including checking of all calculations, computer programs, and reports.e Our work on the Hosgri evaluation of the Diablo Canyon structures employed the same thorough engineering practices and in addition included many of the formal quality assurance requirements specified in 10CFRSO, Appendix B.
Our present and future work on the structural modifications to the Turbine Building will employ complete quality assurance requirements in accordance with the URS/Blume Quality Assurance htanual.
In general, the URS/Blume quality assurance standards require three primary areas of action during the course of a project:
design control, document
- control, and audits.
Design control requires conformance to regulations in "the following eleven categories:
1.
Criteria, or Basis for Design a ccords.n 2.
Calculations Attachment 5
Page 1 of 5 a.
hlay be hand or computer.
a.
Hill be clear and complete.
b.
URS/Blume Project hlaITager is responsible for development.
c.
URS/Blume Division. Director is responsible for approv;Il a>>d client agreement.
d.
Cannot be changed informally.
All changes will be made g to Quality Assurance hlanual section on revisions.
hfr. J. 0. Schuyler 2
URb AN Iti!FllfihlH)lihlI'>~ '
F. >&444
! h1'l,F 5
', ': ll
>4~
August 30, 1977 b.
c ~
d.
e ~
Index for all calculations will be included.
i<ill show all applicablo criteria, assumptions, and refer-ences.
i<ill show enough so that any qualified engineer not familiar with the project can understand what is being done.
Verification oE results:
sign oEf by originator and checker.
URS/Blune Project hfanager signature indicates completion.
Revisions according to Quality Assurance Manual.
3.
Com uter Pro! rams a ~
b.
c ~d-,
e.
f.
All programs used on quality assurance projects are verified programs.
(In the event an unverified program is used, a
second program will be run as a check.)
All certified programs will be logged with name, version, principal use, and personnel available for consultation.
In-house programs will be certified by our computer group.
Certi'fication involves docunentation of everything that went into making up the program (flow charts, listings, user
- manuals, and verification of output).
Outside programs:
verification calculations will be available.
Ãhero verification is done outside the
- company, documentation vill be on file as part of the project documents.
hfodifications:
all modifications to programs will be approved by the URS/Blume Division Director.
hfodifications will be made only to copies of the program, never to original versions.
All modifications will be documented.
hfodified programs will be logged with proper version of identification after they have been certified.
0 4.
~Dravin 4
a ~
b.
C ~
d.
All drawings will be checked for compliance to calculations, project criteria.
Final <<pproval by tho URS/Dlumo Project hfanager.
Index of all drawings will be maintained by the chief draftsman.
Revisions to drawings will be documented and follow the same checking procedures as originals.
5.
Field Investi !ation a ~
b.
All field work will be documented.
Equipmcnt and instrumentation used in the field will bc cali-
- brated, and calibration records will be filed with tho project documents.
6.
Interface Control a ~
URS/Dlumc Project Manager is responsible Eor establishing the procedures between URS/Blume and other organizations (architects or other engineers).
Attachment 5
Page 2 of 5 0
URG Mr. J. 0. Schuylcr August 30, 1971 b.
UlhS/Blume Quality Assurance Manager reviews all procedures to assure compliance with quality assurance procedures.
c.
All procedures are documented.
d.
Objective of interface control is to keep all involved parties informed and prevent unnecessary re>>ork.
7.
~Re orts a.
URS/Blume Division Director has overall responsibility for information and recommendations.
b.
URS/Blume Project Manager responsible for preparation.
c.
Revisions documented and reviewed; one copy to project files.
a.
URS/Blume Project Manager has major responsibility.
b.
All specifications shall be checked for agreement to project drawings and other project requirements.
c.
Handled similarly to reports.
9.
Subcontractors a.
Must have quality assurance program or be subject to pro-cedures of URS/Blume Quality Assurance Manual.
b.
URS/Blume Project Manager responsible for selecting subcontrac-tors.
c.
URS/Blume Quality Assurance Manager responsible for subcontrac-tors'ompliance to a quality assurance program.
Must approve subcontractors before selection.
a.,There will be at least one review per project, depending upon project size.
b.
A review will be a check of selected portions of a project to assure conformance to the project criteria and project require-ments as well as a spot check of calculations.
c.
Whenever possible, the quality assurance review will be performed by qualified persons not working on the project.
d.
All discrepancies will be identified and documented.
e.
A design review report will summarize all discrepancies and corrective action taken.
11.
Revisions a.
No revisions >>ill be made without proper documentation.
b.
Revisions and all necessary documentation apply only to approved documents.
The control of.qual-ity-relate'd docunents require that file registers or indexes be maintained at all locations where quality assurance required documentation is stored.
The retention of all documents willJe as directed Attachment 5
Page 3 of 5
Mr. J.
O.. Schuyler URS AN <4n fellAf/>tlAIHE0cf,; i'ii, ~.el i ~r.~ r~AGA:liZA'VH>
by PG6E.
An index of all material in the Document Control Center will be maintained by URS/Blume.
Internal audits will be performed within URS/Blumc to vcr-'fy that quality assurance standards have been properly implemented and are functioning effectively.
Audit teams will be made up of members of the firm not in-volved in the project.
Pre-and post-audit meetings will be held to prepare for the audit and to discuss any discrepancies found.
All audits will be completely documented.
A compilation of the quality assurance requirements which have been or are being met for the Turbine Building modifications and the recent Hosgri evaluation work are shown in Attachment A.
The measures described herein represent a summary of the quality assurance requirements specified in the URS/Blume Quality Assurance hlanual.
1<e will be happy to provide you with a more detailed description should you require.
Very truly yours, URS/JOHN A.
BLUME 6 ASSOCIATES, ENGINEERS Davi A.'ang Project hlanager DAL:bern Attachment cc:
Mr. Vince Ghio/PG6E hlr. Darrell Policy/PG6E Attachment 5
-Page4of 5
ATTACllhlENT A URS/BLUihlE UALITY ASSURANCE DIABLO CANYON NUCLEAR P01<ER PLANT DESIGiN CONTROL TURBINE BUILDING MODIFICATIOiNS RECENT HOSGRI REEVALUATION 1"ORK 1.
CRITERIA 2.
CALCULATIONS 3.
COMPUTER PROGRA11S 4.
DIVNINGS S.
FIELD INVESTIGATIONS 6.
INTERFACE CONTROL
-7.
REPORTS 8.
SPECIFICATIONS 9.
SUBCONTRACTORS 10.
DESIGN REVIEl(S N/A ll.
REVISIONS DOCUhlL'NT CONTROL AUDITS
~
REQUIREhlENT hlET OR BEING llET o
RL'QUIREhlENT NOT hlET N/A NOT APPLICABLE attachment 5
Page 5 of 5
NRC FORM 335 I7 77)
U.S. NUCLEAR REGULATORY COMMISSION BIBLIOGRAPHICDATA SHEET
- 1. REPORT NUMBER JAssiynedby DDCJ NUREG-0862
- 4. TITLE AND SUBTITLE JAdd Volume No., ifaPpropriateJ Special Inspection of Seismic Related Errors at Diablo Canyon Units 1
and 2
- 2. (Leave blank/
- 3. RECIPIENT'S ACCESSION No.
B. H. Faulkenberry, A. D. Johnson, D. F. Kirsch, P. J. Morrill
- 6. DATE REPORT COMPLETED YEAR MONTH
- 9. PERFORMING ORGANIZATIONNAME AND MAILINGADDRESS (Include Zip Codei Inspection and Enforcement, Region V
Reactor Construction Projects Branch Division of Resident, Reactor Projects and Engineer-ing Inspection, U.S. Nuclear Regulatory Commission DATE REPORT ISSUED MONTH November
- 6. (Leave blankl B. (Leave blank)
YEAR 1981
- 12. SPONSORING ORGANIZATION NAME AND MAILINGADDRESS (include Zip Codel
- 10. PROJECT/TASK/WORK UNIT NO.
- 11. CONTRACT NO.
- 13. TYPE OF REPORT IE, Region V Inspection Report PERIoD covEREo (Inclusive datesl October 14-23, 1981
- 15. SUPPLEMENTARY NOTES
- 14. JLeatre blankJ
- 16. ABSTRACT i200 words or lessl Results of a special inspection by NRC Region V personnel to evaluate:
The Quality Assurance Programs and other management control systems in effect at pacific Gas and Electric and at URS/John A. Blume and Associates.
During the period from 1970 to 1981; the extent to which these Quality Assurance Programs and management control systems were implemented as they relate to the development, transmittal, and use of safety related informa-tion, and, how the seismic related problems discovered at Diablo Canyon, during September-October, 1981, were caused and identi-fied
- 17. KEY WORDS AND DOCUMENT ANALYSIS 17a. DESCRIPTORS 17b. IDENTIFIERSJOPEN ENDED TERMS
- 16. AVAILABILITYSTATEMENT Unl imited NRC FORM 335 17 77I
- 19. SECURITY CLASS IThis reporti Unclassified 20.I~Q Ptfppgthis pagel
- 21. NO. OF PAGES
- 22. PRICE S
0