ML20032D907
| ML20032D907 | |
| Person / Time | |
|---|---|
| Site: | Diablo Canyon  |
| Issue date: | 11/07/1981 |
| From: | Georgiou B CALIFORNIA, STATE OF |
| To: | Palladino N NRC COMMISSION (OCM) |
| Shared Package | |
| ML20032D898 | List: |
| References | |
| NUDOCS 8111180280 | |
| Download: ML20032D907 (8) | |
Text
.
.__4m Stair af California i
~
l W N!
GovsnNon s ovrsce A -
ocn-wress u 916/445-1915 ECMUNO G. BROWN Jm.
saa==os November 7, 1981 Nunzio J. Palladino Chairman U.S. Nuclear Regulatory Commission Washington, D.C.
20555 RE:
Pacific Gas and Electric Co.
(Diablo Canyon Nuclear Power Plant, Units 1 and 2)
Docket Nos. 50-275, 50-323
Dear Chairman Palladino:
This latter is in furtherance of Governor Brown's letter to you dated October 30, 1981, in which the Governor asked the Commission tc order a cruly independent audit of the safety features of the Diablo Canyon Plant before the Commission permits any operation of t
the plant.
Attached is a list of thirteen separate seismic design and con-struction errors at Diablo Canyon that have been discovered since September 21, 1981.
These errors, which involve a large number of systems, components, and equipment critical to safe operation of the plant, demonstrate a serious and widespread breakdown of the quality assurance program at Diablo Canyon.
The errors are particularly significant because they were overlooked by PG&E and NRC inspectors for four years, during which time both PG&E and the NRC repeatedly gave assurances that.the seismic design and safety features of the plant were being analyzed with the most careful and detailed attention.
The NRC's regulations, specifically the Quality Assurance require-ments of 10 C.F.R. Part 50, Appendix B, are designed to ensure compli-anc.e with the NRC's technical requirements and, thereby, to detect t
the very types of errors that were overlooked by PG&E.*/
PG&E did not detect these errors because PG&E did not comply with Appendix 3.
In short, PG&E viol.ated the NRC's regulations.
- /
The NRC Staff pointed out to PG&E on October 9 that errors of the type discovered at Diablo Canyon should have been detected if Appendix B had been properly implemented by PG&E.
See October 9 transcript, p.
87.
Also, PG&E's Mr. Maneatis, Senior Vice President, stated en November 3 that PG&E's error was "the result of failure to follow established practice and represented a clear violation of our quality assurance program."
Transcript,
- p. 131.
Finally, on October 9, Dr. Denton stated that had the Staff known of PG&E's errors, the Staff would not have recommended issuance of the low power license.
8111180280 W11109 PDR ADOCK OS M 75 0
?:)R t
Chairman Palladino November 7, 1981 Substantial uncertainty now exists concerning the actual degree of quality achieved at Diablo Canyon.
The Staff has recognized, to some degree, this uncertainty.
As a result, on November 3, the Staff directed PG&E to conduct an expanded audit of Diablo Canyon safety systems.
However, the Staff's directive does not reach far enough.
It does not establish the basis for a truly independent audit of the existing and potential errors at Diablo Canyon by outside experts who have no real or apparent interest in the results of their audit.
At the November 3 meeting with PG&E, the Staff did not direct an audit of PG&E's errors by independent experts in accordance with the Governor's request.
Instead, PG&E was permitted to perform an audit by a consultant selected exclusively and unilaterally by PGGE.
The Governor pointed out in his October 30 letter that an audit h PG&E of the very errors which PG&E itself committed and overlooked for four years simply would not be credible.
If the Commission authori:es the kind of audit permitted on November 3 by the Staff, the credibility of the audit itself and the credibility of the Commissien will be undercut.
We remind the Commis-sion that any hope for public confidence in the Commission's Diablo Canyon determinations has been shattered by the recent post-licensing disclosures of errors at Diablo Canyon.
We submit that a truly inde-pendent audit ordered by the Commission is the only means by which the NRC can recapture any degree of credibility.
The importance of quality assurance at nuclear power plants vas recently emphasized by the Commission in the NRC's 1980 Annual Report.
The application of disciplined engineering practices and thorough management and programmatic controls to the design, fabrication, construction, and cptracit: of nuclear power plants is essential to the protection of public health and safety and of the environment.
Quality Assurance (QA) provides this necessary discipline and control.
Through a QA program.that meets NRC requirements, all organizations performing work that is ultimately related to the safety of plant operation are required to conduct that work in a preplanned and documented manner; to independently verify the adequacy of completed work; to provide records that will confir.n the acceptability of work and manufactured items; and to assure that all individuals involved with the work are properly trained and qualified to carry out their responsibilities. (p. 79)
These words have been put to a critical test by the multiple CA errors at Diablo Canyon.
If the public is to believe that the Commission is genuinely serious about QA, then Diablo Canyon must not be permit-ted to operate until a truly independent audit is completed and full compliance with Appendix B is demonstrated.
Chairman Palladino November 7, 1981 Accordingly,_in furtherance of tne Governor's October 30 letter, we hereby ask that the Ccmmission:
1.
Order that an audit of the errors at Diablo Canyon be per-formed by outside experts who are independent of PG&E.
At a minimum, these cutside experts should be persons who have not worked for PG&E or on the Diablo Canyon project.
(PG&E's current auditor, Dr. Cloud -- who was approved by the Staff on November 3 --has previously worked on Diablo Canyon.) Moreover, the outside experts should not be selected unilaterally by PG&E; they should be acceptable to all parties in the Diablo Canyon proceeding.
The final selection of the independent auditor should be approved by the Commis-sieners, following the Staff's consultations with all parties.
We have already submitted to the-Staff a suggested list c2 nine possible indeperdent auditors.
Surely, independent firms which are acceptable to all parties can be found.
2.
Order that Diablo Canyon shall not be permitted to cperate until the entire audit is completed.
On November 3, the staff indicated it would permit low power operation before completion of the audit.
However, there is no technical or legal basis for quality assurance differentials between operation at low power or greater power. Put briefly, Diablo Canyon should not operate at any power level unless it complies with the NRC's technical requirements and regulations.
3.
Order that the staff convene a working session which leads to selection of outside experts, acceptable to all parties, who will perform a truly independent audit.
At such a working session, any and all parts of our October 30 pro-posal could be discussed, and appropriate modifications to that proposal could be evaluated by all of the parties in a ccoperative atmosphere.
We note, en the basis of discussions with the Staff, that it appears that our proposal for a Steering Committee has been misunder-stood.
We understand that the Staff apparently regards our proposal cs unacceptable because it appears to intrude upon the NRC's regula-tory authority.
We are sensitive to the Staff's concerns and thus wish to clarify that our proposal was intended primarly as a point of departure for discussion.
We believe that this is a subject that surely can be resolved to the satisfaction of all parties at the working session.
Our intentien so far has been to approach the independent audit as being an issue above and beyond the on-going adversarial proceed-
- ings, We cannot conceive how the Governor's proposal for a truly independent audit could adversely affect the interests of any party to this proceeding.
In our view, PG&E should welecme such an
Chairman Palladino November 7, 1981 audit, and the Ccmmission should sie:e it as a means to shore-co the damage suffered by the NRC in mistakenly licensing Diablo Canyon as a facility safe to operate.
Fe bring to the Ccmmissioners' attention the fact that the follow-ing governmental bcdies have already called for an independen: audit:
City Council of San Luis Obispo; City Council of Santa Barbara; City Council of Pismo Beach; City Council of Tnousand Oaks; Board of Super-visors of Ventura County; Board of Supervisors of Santa Barbara County.
In addition, several major California newspapers have editorialized in support of the Governor's proposal for an independent audit.
We believe that it would be a profound mistake for the Commission to permit an audit of PG&E's errors under the terms directed by the Staff on November 3.
In actuality, such an audit amounts to a busi-ness-as-usual approach that in no way befits the extraordinary recent disclosures at Diablo Canyon.
We reiterate our interest and availabil-ity to work cooperatively with the NRC in establishing the framework for an independent audit of the quality of Diablo Canyon.
We ask that the Commission now take action that results in a truly independent audit worthy of belief by the affected citizens of California.
Sincerely,
~
.t.,p~\\.
- . ~ym a
Byron S. Georgiou Legal Affairs Secretary cc:
Commissioners Service List
QUALITY ASSURANCE ERRORS REVEALED.
SINCE SEPTEMBER 21., IEE1
~
l In the 'six weeks sin'ce September.~21 19E1,.' 15 seriour c.rrors in seisatc-design have been.discovere,d at Diablo Canyon.
These errors. have primarily invoIved problems !in tire development,.
I distribution, and use of design dats >y FG8E and its engineering -
services-subcontractors These errors were-discussed at Commission and Stuff.maetings in Washington. B.C.. and Bethesda.,
Maryland and at a series of meetings during October,.1981, with 1
j PGEE in Sart Francisco, California.-
As a. result of these discussions and inves.tigatiorrs, it is nor clear that each error involved a. failure of PG6E tv implement properly
- number of-the. ?S. quality ass.urance cri.teria of 20 C.F R. Part 50, Appendix.
-B.
The errars are described.below:
'a.
Error I C OppositiIrand Besi.gzr.
i On Septenber 28,1981c PEGE reported that a diagram ^
~
error had~ been found, Unit I NucIsar Fower Plant (ICPP-1).
in a portion af the seismic qualification of the Diable Canyon Tnis error resul.ted irr an incorrect applicatium of the seismic fIcar response spectra in the crane vall-containment ~.shell annulus of. the Unit' 3 Containment. Building' The error was. thst.
the diagram-used tv 3ecate Yestica'E Seismic Floor Response (VSFR) spec.tra for~ the Unit I contvimacnt annuIns was :erreaeour. - The diagram was applicable to. Unit 2 but was identified 'as. being.
tha.r of Unit'1.
Since the Uni.ts. are opposite hand,. this resultyd y
in an incorrect orientation of VSFR spectra for Unit 1~ cosqurnent and.systen desigrr The origin of the error was in the PGEE i
transnittal,, to a su' contracta- (John 'A BIune and Associates),
af an un.verif2nd., handwritt&r sketch of the Unit Z hand geoznetry in place of tho Unit I geometry. // (opposite Alse see E'rrar 3).
b.
Error 2 -JDscutent Distribution l
At the October 9 raeting-between the KRC Staff and PGGE, PG4E disclosed that the Scist:ic Category I electrical cable trays and. conduit supperts had been qualified to design i
response spectra which hel seen superseded The error was spectra to the rt-sponsible engineer. gthe 1atese revised caused-hy PE6E's failure to distribut
~
l I/
LER SI-M2/01T-&,. October 12,. IsSl
{
2/
Wher 9' meeting transcript, pp. 105-E7.
-L-l
-..-w.-,-
,--e-+
c...-re.m-,_,-m.--,,,,,,,,w y--r.,-,wc-,v.4~,,..ew,+.,,,.,
,,w-.---,,,..r--%,eer*v*'-*--*"-w**'wP-~4--v'ws**-*-e-**e' r
- P*
- -sr'T--wTT78-'-'**'--"-W'Tf e Wu'
1 Errorl3 - Incorrect Weights,
.c.
On Oct6her 22, 1981 ins.pectors from the NRC's Office of Inspection and. Enforcement determined that, in addi. tion to the improper application of the diagran as reported to the NPI by PGE on September 28, 1981, the' weights listed on the disgran and;used.' as an impat to John A.. Eltme and Associates for their development of response spectra, could. not be Vsverified sa being accurate PGW representatives recaI.culated.
the weight:;, using current as-Su_Lt. drawings, and determined i
the new weights to b-e different.F PGCE concluded that the " substantial" weight vzriations resu3ted fram three principal. causes:
(i)
'Hze large bore piping equi e weights were.not associated v.ith correct frames because the -Usit I piping orientation was used in conjunction witft the-Unit 2. frame orieutKtion..
(ii)
PGT,Es current calcula,tions, include additional can.tributors to the total: weight:
e.g.,
' which were considered.ys and. stecI grating, conduit and cable-tra to be insign'.licant in' the 1977 analysis.
(iii}
A n:rre detailed calculation of large bore piping weights, piping s,upport weights and equipment weights.
d.
Error 4 - Containcent Spray Svster Pipe Supports On Septeuber 18, 1931, the NRC's 'rcsident inspector was notified by/ elephone of a deficiency reportable under 10 CFR t
50 55(e)
.T The repxrt from PGE,E addressed deficiencies in.
the design of the containnen; sprey systen pipe supports located within containnent.
The following four deficic::cies. were identified:
(il An incorrect thernal. analysis. was used for hauger Joeds.
3/
PND-Y-al-59 Preliminary NRC' Notification of Event or Unusua.1 Occurre:nce,. October 26,1?S'l.
4/
Letter from Crane to Engelken, October 19,1081.
Inexplicab1) this error was not brought to the NRC's 'C6:_tissioners" atten.t:
Septenher 21, 1981.
A
~
(fil This analysis was. perfcreed. with cne saubbar codeled cs o rigid menbar.,
(iii)
The variable spring settings. for the pipe supports were improgerly set, ~ based,
on a deadland analysis wtich assumed, incorrectIy,. that the ~ pipes contained water.
Civ)
In designing a. pipe anchor, the: loads from only one :n.de. were used.
The root cause of the preceditrg series of errors has not yet been explained by PG4E. '
e Error 5 - Trong Spectra During the period of Oc.tober 14 through16,1981, representatives of the NPI staff and. their. consultants from Broolhaven National. Laboratories net with the PGEE staff in.
San Francisco.
During the necting,. piping probler (PG5E f6-11) was reviewed PGEE' initially asserted that this problem did not require reevaluation as a. result of the opposzte hand error.
Kowever, it.was subsequently deterz ined that the.
eriginal PGEB calculation used erronesus spectra input and hence required reanaIysis with the appropriate spectra 5_/
The tause of the error has not yet been identified by PGRE f.
Errors 6 to 10 - Additional Design Errors At the Nevenber 5 neeting between FGEE and the h%C, PG68 disci & sed. that during Its. internel review undertaken es. a result of the diagram frme-orientation error, it has identi.fied five additional design errors resguiring plant modifications from causes. net related to the diagran error.
{
Thcsc design deficiencias airc:
h K_
(i}
In. a single case, parallel piping lines which were qual'ified and designed. from a single analysis ac.tusIIy require two analyses to properIy mdel both config-urations.
~
5/
NRC Meeting Sursary for October-14-15,. 1981, Discnssions and Preliminary Audit of Seiscic Isralysis for Equipment and Components in Diablo Canyon, Unit 1 Containment lamulus, p. 4.
5
h.
' Error 11 - Misapplications of m,sgri Spectra.
Error 11 involved misapplication of the Hasgri spectra.
Electrical raceway and condui^t suppurts are unistrut type supports, a're all Class I equipment, and are all Iaturally braced.
The PG6E seismic analysis is based. on an enveloping procedure using static analysis.
In this analysisi which contains a large number of configurations,, the Isrgest weight that a partietslar'conilgu-a.tiert is considered ta be able to have appIied to it is determined arid the highest acceleration the support can experience owing to its. location in the buNing, is. alsa determined.
Then, with thase tuo inputs, the first mode frequency of the supperts. is calcuInted, and the corres-ponding acceleration level is. taken fron.the response spectra,
and the stress analysis is condtteted De misapplicatiom errors were basically of two kinds.
First, tire analystjselected the wrong number off the 'resgense spectra curve; and siennd, in sone cases the engineer apparently used one of the Hesgri spectra from a. different location irt the building.
As hefer the cause of the errors has not yet been. deternined by FGEE i.- Error 15 - Further Spectra. Misannlication For the heating and ventiIating systen components,-
Dr. Cloud reviewed the seisni.e input for the fans and daepers.
We found one instance where the Hesgri spectra were misapplied.
Once againt the menner. in whi.ch this analysis was conducted is very simiIsr to'that for the conduit supports (Errar IZl.
The engineer confirmed tha.t the equipment was rigid, and then went to the zero portion of the ~ response spectra curve-and selected the wrong value for the acceleration level.
In' this case, PG4E believes that the engineer used a spectra -froz e differerst location of the building.-
The cause of the error has not been determined by PG6E 10/
1/
November 3 nereting transcript, pp. ZG4-ZG5, 10/ Kovecher 5 caeting transcript, p. ZG6.
5
,