IR 05000275/1982002
| ML20050N700 | |
| Person / Time | |
|---|---|
| Site: | Diablo Canyon  |
| Issue date: | 03/30/1982 |
| From: | Joukoff P, Morrill P, Thomas Young NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION V) |
| To: | |
| Shared Package | |
| ML20050N694 | List: |
| References | |
| 50-275-82-02, 50-275-82-2, 50-323-82-02, 50-323-82-2, NUDOCS 8204140426 | |
| Download: ML20050N700 (8) | |
Text
-
,
.
U. S. NUCLEAR REGULATORY COMMISSION REGION V.
50-275/82-02 Report No. 50-323/82-02 Docket No. 50-275, 50-323 (OPS)
License No. CPPR-39, CPPR-69 Safeguards Group Licensee:
Pacific Gas and Electric Company P. O. Box 7442 San Francisco, California 94106 Facility Name:
Diablo Canyon Units 1 & 2 (1) Pacific Gas and Electric Company, San Francisco, California (2) URS/Blume and Associates, San Francisco, California Inspection conducted:
(3) R. L. Cloud Associates Inc., Berkeley, California Inspection conducted:
November 13 and 30,1981, December 23, 29 and 31, 1981 and January 6-8, 1982 Inspectors:
JJ\\A, J
3 4 6/-8L
'
,
P'. Morrill, Rea tor nsp t Date Signed
[gY 3-XA-M P'.
Joukof f, Inve
~ tor Date Signed l
Approved by:
MM"
.
T' Young, Jr., ChSef, Reactor ffojects Section 2 Date Signed Reactor Operationk Project 1rMch Summary:
Inspection of November 13 and 30,1981, December 23, 29 and 31,1981 and January 6-8, 1982 (Report Nos. 50-275/82-02, 50-323/82-02)
Areas Inspected:
Routine inspection of activities associated with the licensee's Independent Evaluation of Seismic Service Related Contract work performed prior to June 1978 pursuant to NRC Order Suspending Licensee (CLI-81-30). This inspection effort involved 64 inspector-hours on-site by two NRC inspector / investigators.
Results: The findings contained in this report will be evaluated in conjunction with other ongoing reviews related to the seismic adecuacy of Diablo Canyon.
8204140426 820331 PDR ADCCK 0500027S O
.-.
.
.
DETAILS 1.
Persons Contacted Pacific Gas and Electric (PG&E) Company
- E. Kahler, Senior Quality Engineer
- V. Ghio, Senior Civil Engineer
- I. Wallak, Supervising Civil Engineer R. L. Cloud Associates Inc. (Cloud)
- R. Cloud, President
- P. Chen, Project Engineer URS/J. A. Blume and Associates (Blume)
D. Lang, Project Manager L. flalik, Manager of Structures Department D. Jhaveri, Vice President, Deputy Manager, Nuclear and Energy Division The inspectors also talked with and interviewed a number of other licensee and contractor employees.
- Denotes those attending the exit meeting on January 8, 1982.
- Denotes those attending the exit meeting on-December 31, 1981.
2.
Organization and Management The inspector examined the P,. L. Cloud and Associates (RLCA)
proposed seismic reverification program dated December 3,1981, and discussed the program implementation with Dr. Cloud and his employees.
The inspector observed that RLCA had six full-time and three subcontracted engineers working on the reverification program.
The inspector also observed that RLCA had contacted Stone and Webster for assistance on system design review in a letter dated December 23,
'
1981.
Dr. Cloud informed the inspector that the Ouality Assuran:.e (QA)
review to be done by Roger Reedy, Inc. (Reedy), was under way and that Reedy had set up his audit plans, was accumulating plans and manuals, had begun a preliminary review of the documents received to date, and was beainning to audit the various seismic subcontractors. The inspector also discussed RLCA reporting requirements with Dr. Cloud.
Dr. Cloud stated that on or about the time of the November 3,1981 meeting with the NRC his contract responsibility shifted from Mr. J. Rocca to Mr. G. Maneatis. As a practical matter, Mr. Rocca retained technical responsibility to assist RLCA in obtaining infor-mation while findings and identified problems were dealt with at Mr. Maneatis' level.
RLCA employees felt that this arrangement had signi,ficantly speeded the flow of information from PG&E to RLCA.
.
- e
.
-2-3.
Review of Cloud Progress Reports The inspector examined the RLCA reports, listed below, and the RLCA files, and discussed items identified in the progress reports with RLCA personnel to verify adherence to the NRC's order (CLI-81-30) of November 22, 1981.
a.
Progress Report of Seismic Service-Related Contracts Prior to June 1978.
Progress Report No. I 11/2 - 11/10/81 Progress Report No. 2 11/11 - 11/23/81
-
Progress Report No. 3 11/24 - 12/8/81 Progress Report No. 4 12/9 - 12/21/81 b.
Design Verification Program - Seismic Service Related Contracts prior to June 1978 dated December 3,1981 (Program).
The inspector observed that the program had evolved through the use of the progress reports, in that RLCA personnel had deleted Westinghouse and General Electric from further consideration on the basis that they were primarily equipment vendors and whatever services they offered were in support of licensing. The inspector stated that this interpretation
of the NRC's order should be documented and requested that RLCA
employees document and justify their position.
Subsequently, the documentation effort requested was included in RLCA's Progress Report No. 4 dated December 21, 1981, on pages 1 and 4.
The inspector also observed that RLCA employees had identified several items requiring follow-up and/or more information.
During discussions with RLCA employees the inspector also determined that
,
there was confusion regarding the applicable seismic spectra for equipment in the auxiliary building (see Paragraph 6 of this report).
'
As a consequence of these observations, the inspector asked RLCA employees how they were keeping track of errors and items requiring
,
follow-up.
RLCA employees stated that they would use an " error /
'
l open item" identification system and would include copies of the
" error and open item list" in subsequent progress reports.
Subsequently,
'
the inspector verified that this commitment was met in RLCA Progress Report No. 5 dated January 6, 1982.
4.
Review of R. L. Clouds' Quality Assurance (0A) Program Implementatien The inspector examined the RCLA QA Manual, Revision 5, dated November 1981, and the RCLA QA Supplement, Revision I, dated December 1981, for the Seismic Reverification Program. The inspector also examined the records used to indicate which personnel had read and understood the manual and its supplement. The inspector observed that two engineers and one secretary assigned to the reverification program had apparently not read or signed the records.
RLCA employees committed to have all appropriate project personnel read and sign this documentation.
During the inspection period the inspector verified that this commitment was complete.
'
.
-3-5.
Review of Work in Progress a.
The inspector examined the seismic reverification program generic sample work in progress for the following items to verify accuracy and completeness.
i)
Component cooling water piping seismic model li) Auxiliary building seismic model iii) Component cooling water heat exchanger seismic model l
iv) Charging System pump suction piping seismic model The inspector discussed this work with the responsible engineers and verified that their work was based on PG&E drawings which had been verified at the plant by RLCA employees. The inspector determined that in one case, the auxiliary building, the RLCA employees were verifying the location of concrete block walls (i.e.:
PG&E Drawing 438431, Revision 12, shows added block walls around the 480 volt switchgear) and other changes made after the Hosgri report was prepared, but were not verifying embedded items or all structural steel details.
b.
The inspector examined the RLCA files (logs) related to the seismic reverification program to verify selected
,
!
conclusions stated in the RLCA progress reports.
The following specific items were examined.
i)
Electrical raceway supports:
seven of 20 examined by RLCA appeared in error.
PG&E is following up on this item.
,
l ii) Heating, ventilating, and air conditioning equipment:
four equipment items were found by RLCA to have
'
'
deficiencies in their qualifications.
PG&E is
[
following up this item.
l l
iii) Electrical conduit supports:
RLCA questioned the method l
of qualifying the conduit. PGAE personnel were doing additional calculations to qualify these supports.
RLCA intends to followup this PG&E work when it is completed, iv) Differences between field walk down data and PG&E drawings:
PG&E resolving these on a case-by-case basis.
.
g-eem
-,w--
-.-eep
+
,-m-y
---wy
,se e-qn-,
,
n-+--
m
,,,, -g
,a sy,
.
wv n.m y,,
4ym p-m
,em.
y
,
p...
,
i
-
.
'
.
-4-6.
Outdated Seismic Spectra for Auxiliary Building a.
While examing the RLCA work which identified problems described in the RLCA progress reports the inspector observed that the RLCA seismic analysis of piping located in the auxiliary building was on hold pending resolution of the seismic spectral inputs (spectra) RLCA employees showed the inspector that the docketed "Hosgri Report" (Section 4) dealing with the auxiliary building was different and in some cases less conservative than the URS/Blume final Seismic Report for the auxiliary building, dated October 1979.
RLCA had requested resolution by PG&E as to which were the controlling seismic spectra in a letter, Cloud to Rocca, data November 5,1981.
In a subsequent telephone call with URS/Blume (Cloud file P105-4-593-015),
an RLCA employee was told that the north-south spectra for the auxiliary building in the May 1977. preliminary report were incorrect in that the soil spring (north-south) was left out of the model, that the October 1979 report corrected this error and that this (later)
spectra is in all cases lower-RLCA employees were also sent an
.
internal PG&E memorandum (DCM C-15 dated November 19,1981), which indicated that figures 17 through 34 of the October 1979 report superseded figures 4-110 through 4-127 in the Hosgri report.
When questioned as to why RLCA did not bring this item to the NRC's attention, Dr. Cloud stated that at the time thi!, issue came up he was interested in obtaining a set of spectra to conduct independent calculations to complete his work promptly and well.
Prior to the NRC inspector discussing the issue, Dr. Cloud didn't know the status of the October 1979 report versuses the Hosgri report but stated that the time issue would have to be clarified before his review could be completed.
He felt that the RLCA work was still valid since his work could check the PG&E calculation and analysis methods using the Hosgri spectra and that wouldn't change the reivew of methodology.
Dr. Cloud felt it was PG&E's responsibility
,
to clarify the design basis of the plant and in the interim he intended to use the Hosgri (as defined by the ground spectra) report since it was based on close scrutiny by the NRC and a host of others.
He also stated that any significant problem would have come out of the RLCA Auxiliary Building' review in any case.
Based on this information, the inspection staff concluded that it was appropriate to continue the inspection efforts at URS/Blums'
(Blumes') and PG&Es' offices in San Francisco in order to determine the significance of the different spectra and the cause of the apparent delay for reporting this matter to the NRC. The inspectors met with Blume personnel on January 8, 1982 and with PG&E personnel on January 7 and 8,1982 in San Francisco.
-..
__
-
. -
-
-
__
.--
,
- -.
- -.
.
.
.
,
-5-Blume personnel sta'.ed that the auxiliary building report prepared b.
in May 1977 was pre 7iminary and marked as such. They went on to explain that their design review (dated October 27, 1978) had discovered the inad <ertent omission of the soil springs at elevation 100 feet in the north-south direction and that this omission was corrected in the final October 1979 Report.
Since this omission led to conservative results for the auxiliary building structural analysis, the Blume design reviewer concluded that the preliminary analysis was conservative. The spectra developed from the building responses were different and Blume expected PG&E to take appropriate action.
Since most of the spectra peaks were lower it was thought.oy Blume personnel that loads would generally decrease. The inspector observed that generally translational spectra were reduced, torsional spectra increased, and frequency of peak responses increased. The inspector asked Blume personnel why they had told RLCA personnel on November 17, 1981 that the October 1979 report's spectra were
"in all cases lower" RLCA telecon record P105-4-593-015). Blume personnel responded that they had intended to communicate that
" responses were lower" and that the October 1979 Auxiliary Building Report contained the spectra which should be used by RLCA.
In response to the inspector's questions Blume personnel stated that PG&E had requested them to analyzeithe impact of the changes in spectra on equipment, piping and components.
They went on to state that only equipment in the Auxiliary Building, east side, above elevation 100 feet appeared to require some seismic qualif.ication checks since other spectra for the Building were conservative. At the time of the inspection no adverse effect on equipment qualification had been identified.
Based on.an examination of the 1977-78 design reviews, Blume personnel felt that-this_ problem (omission of soil springs in analysis) was unique.
In response to questions by the inspectors, PG&E personnel stated c.
that they believed that the October 1979 Report on the Auxiliary Building was given to PG&E by Blume to fulfill the Blume/PG&E contract.
Four copies of the report were received, looked at briefly, and filed. None were sent to the NRC or used to amend the Hosgri Report as PG&E was unaware that the spectra changed from the preliminary (May 1977) report.
The inspector pointed out that the original Blume Report on Design Review dated October 27, 1978 had been reviewed by PG&E personnel on December 1, 1978 and accepted by the responsible supervisor on December 13, 1978 (PG&E Civil Engineering Department,
.
.
-_.__.
_
-
_
,,
-
. _..-
-
.
.
-6--
-
file 40.31). PG&E personnel stated that they did not do a
" design review" of Blumes' work but only verified that Blume had asked and addressed the appropriate questions in their (Blumes')
internal design review.
PG&E personnel stated that, had they been aware of the different spectra between the preliminary and final reports, they would have done something.
Reportedly when Cloud personnel raised the issue in late Ocrober 1981, PG&E personnel prepared overlays to look at the differences and concluded that the differencies "didn't look too significant".
PG&E personnel did not report this item to the f4RC because they were not sure if it was a problem and consequently tasked Blume to evaluate which areas of the Auxiliary Building were affected and what that would do to equipment qualifications.
PG&E personnel stated that there were insufficient resources to devote to each " thing" that appeared to be of low impact from a technical viewpoint and that they were directing their resources to major items first.
Reportedly the building was unaffected and only a limited number of equipment seismic qualifications might be unconservative. As a consequence, PG&E personnel cancelled the internal memorandum, DCM C-15 of November 19, 1981 (directing the use of the October 1979 report)
with another memo dated November 25, 1981 until Blume's review is completed.
PG&E personnel also informed the inspector that they had initiated a nonconformance report (NCR) on this problem and that Engineering Quality Control (EQC) would followup resolution of the problem. The NRC, DC0-81-QA-N00S, indicated that PG&E personnel had reviewed reportability under 10 CFR 50.36 and 50.55(e) and had concluded that reporting cannot be made until the PG&E civil engineering review Was completed.
The inspector examined the following documents to verify the licensee's and contractor's statements.
Hosgri Report - flRC Docket Files May,1977 and October 1979 " Auxiliary. Building Dynamic Seismic Analyses for the 7.5M Hostri Criteria" Reports May 11, 1977 Transmittal Sheet - Transmitting the May 1977 Report November 17, 1981 Dension/Long Telecon Cloud File P105-4-593-015 November 19 and 25, 1981 Letters (DCM C-15) Wollak to PG&E Engineering Staff
.-,-
.
-7--
October 27, 1978 Blume Report on Design Review of the Diablo Axuiliary Building December 13, 1978 PG&E - Hosgri Design Verification of The Auxiliary Building November 20, 1981 - List of URS/Blume Hosgri Reports (total of ten)
December 30, 1981 - NCR, DC0-81-0A-N005 Auxiliary Building Undated Overlays of Seismic Response Curves for the Auxiliary Building The inspector observed that the entire matter had been handled in an informal manner in that (1) Blume was conducting the review based on verbal instructions, (2) the matter had not been identified to the NRC until the inspectors discovered it, (3) the items being identified by Cloud, PG&E and others were not being systematically identified and tracked in any one place or manner, and (4) the NCR written by PG&E and shown to the inspector was missing several signatures and dates. The inspector also observed that he had not detected any apparent deception or with holding of information and that the persons contacted had been very cooperative with the inspectors.
PG&E personnel expressed concern that the problem of the Auxiliary Building spectra may not be a safety problem at all and that to report it is a problem prematurely would make it an issue regardless of its safety implications. They also pointed out that reporting this type of item was outside regulatory requirements.
The inspector stated that these points were correct but added that there was extreme sensitivity to anything related to the seismic analysis of Diablo Canyon and that it might be in the licensee's best interest to report items of concern before someone else did.
7.
The inspectors met with RLCA representatives on December 31, 1981 and with the licensee representatives (denoted in paragraph 1) at the conclusion of the inspection on January 8, 1982. The inspectors summarized the scope of the inspection and the findings as described above.