ML20128N340
| ML20128N340 | |
| Person / Time | |
|---|---|
| Site: | Comanche Peak  |
| Issue date: | 05/29/1985 |
| From: | Wooldridge R WORSHAM, FORSYTHE, SAMPELS & WOOLRIDGE (FORMERLY |
| To: | Bloch P Atomic Safety and Licensing Board Panel |
| Shared Package | |
| ML20126J029 | List: |
| References | |
| NUDOCS 8506030240 | |
| Download: ML20128N340 (36) | |
Text
.
. WonsnAx,.FonsTTaz, SA.MPEI.s & WooI.DRIDGE twt tv. Piv..w.... 9 0 0,....a. to...
I>m.a. n== neo g......,,.,,......
..........s,..
- v. a.... a..,....
... '.,, "" *n...
':;r,t '.?'.".*,'t;;*'
"2";".* c.'!'.".*.
s.....,e.......
'!":. "'"'??.'"
- f.',".'.1 :01
- '.t"
- ':.:.*:"*,".......
- ~
3.I*Y',i:EI!~....-
May 29, Isas
~.
- A".'"?f.Z*;;
Peter B. Bloch, Chairman Administrative Judge Atomic Safety and..lcensing Board U.S. Nuclear Regulatory Commission Washington, D.C. 20555 Dr. Walter H. Jordan Administrative Judge 881 W. Outer Drive Oak Ridge, Tennessee 37830 Dr. Kenneth A. McCollom Administrative Judge Dean, Division of Engineering, Architecture and Technology Oklahoma State University Stillwater, Oklahoma 74078 1
Re: In the Matter of Texas Utilities Electric Company, et al (Comanche Peak Steam Electric Station, Units I and 2)
Docket Nos. 30-445-1 and 50-446-1 Supplementation of Applicants' Respons: to CASE's Request for Production
Dear Administrative Juages:
This is to nottfy you and all parties to the above dockett tnst Appilcants have identified a document which w believe to be within the scops of item 10 of CASE's Interrogatories and Requests to Produce dated July 7,1980, as clarified on August 4,1930. App!1 cants submitted their response to item 10 on Sept 6mber 3, 1930 and supp!.mented thsir re: pense by letters dated December 22,1930, March 3,1952, and Aprij 19, 1932. The document, a copy of which is enclosed with this letter, is a report prepared in May of 1978 by Management Analysis Company (MAC) following a mana6ement review and audit of the quality assurance program of the Comarche Peak Project.
Y 9co\\
ek
- e. t o
gg@9 W G2VO XA m
Administrative Judges May 29,1985 Page Two Recently, in gathering data for a prudence audit being performed for TUEC, a search was made by TUGCO personnel of inactive and closed corporate files located in TUGCO's Dallas office. In the course of such search, the enclosed report was found. A memorandum dated July 11, 1978, which de: ails TUGCO resolutions to the findings and recommendations made by MAC, is also enclosed.
TUGCO management is evaluating the failure to produce this document at ar).
earlier time and will advise the Board and parties of the results of this evaluation in the near future.
Respectfully submitted, Robert A. Wooldridge
/
Counsel for Applicants
=
z_
_. -RAW /klw Enclosures cc:
Service 1.ist G
6 e
9 e:
e T
e e
1 t
l
.. g-Management Annlysis Company
~
1110o RoseUe St, San Diego, CA 91121 714/4524391 May-T7,1978 MAC-JPJ-471 Mr. Perry Brittain President Texas Utilities Generating Company 200T Brym Tower Dallas,. TI 75201
Dear Mr. Brittain:
Enclosed is the report of the Management QuaTity Assurance Audit conducted for Texas Utilities Generating Company.
m The audit disclosed that, in general, the Quality Assurance activities were effective, that there is good team spirit between TUGCD/TUSI personnel and the Architect / Engineer and the Constructor. The audit resulted in the identification of some failures. to: comply with regulatory requirements, the Quality Assurance Pian or the.PSAR 4
These deficiencies are identified in an Audit Report as Appendix A.
The audit aTso identtfied areas. of potentiaTTy improved pract. ice. These are identified as.0bservations and Recomendations and Appendix B tc this letter. As you know,. MAC participated; irt an audit of the Comanche Peak site and significant improvement ts noted since that aucit.
Management Analysis Company received full cooperation from aTT personnel con-tacted during the audit, TUGC0/TUSI, Brown & Root, and Gibbs & HiTT. The6)
I general openness of personnel and their frank discussion not only enhanced the conduct of the audit, but exemplifies an attitude conducive to correction-of any deficiencies.
We appreciate the opportunity to be of service ts Texas. Utt11 ties Generating Company and Texas Utilities Services,. Inc. and hope to do so irr the future.
If there are any coments or questions regarding tMs wcrit, please contact Mr. J. ft. Norris or me at (714) 452-139T.
SincereTy, J
//
f y.
John-P'. Jac'kson Principa.1 Partner JPJ:bew Enclasures: Appendix A Appendix B 1
I
ey
~~
e e
e l
7 l
APPDIDIX A g
[
E U REPORT
(
b-e b
\\
o f
e 6t M
T 9
e B
e G
y
.7
AUDIT REPORT
Subject:
Audit of Texas UtfTities Generating company. DaiTas offfces and Comanche Peak Steant ETectric Station Construction Site Date of Audit-May 1-TZ,197E
~
Audit Scope:
A management audit was conducted of the Quality A surance Progrant of Texas Utilities Generating Company during th,e.
weeks of May I and May 8,.1978.
The purpose ' f the audit o
was ts.detenrin'ii the adequacy of the QuaTity Assurance Progrant as reTated to NucTear Regulatory Commission require-ments and the effectiveness of impTementatierr ta1neet program requirements and au.thority delegations. Activities were audited at both the TUGCO' offices frt DaTTas and at the
~
Comanche Peak constructforr site. Actfvitfes of the A. chitect/
~~
Engineer and Constructor were audited onTy at the construction site. The scope of the audtt incTuded conmitments made in the PSAR,, the Corporate QuaTity Assurance Manuai, the Comanche Peak Quality Assurance PTan,, the Prefect Procedures Manuai and the Brown & Root QuaTity Assurance Manuais and Procedures related ta the Comanche Peak stte.
Auditors:
Dallas office,May T-1,,T97E J. P. Jackson, MAC Audit Teant Leader J. M. Norrfs., MAC Auditor Comanche Peak Constructiott Site;,, May & & E., May 3-12,. 1978 J. P. Jackson, H4C Audit Teenr Leader J. M. Norris,, MAC Audfter J. A. Hendron,. MAC Auditor (May F-T2 onTy)
PersonneT NAME COMPANY TIT 1.E hhry D. N. Chapman TUGCU nager, *-1-2 R. G. Tolson TUGCQ b-.
Mgr, Site Surveillance, *-1-2 R. V. FTeck.
TUGCD/G&W Civ. Inspec. Supv.
- -1 J. V. Hawkins TUGCD/ Gal +
Prod. Assurance (QA), *-1
_---<,y-
i
[,.,
Personnel NAME COMPANY TITLE Co" 7ntery J. B. George TUSI Prof. General Mgr., *-i-2 J. T. Merrit TUSI Resident Manager, *-1 E. G. Eibson -
TUSI Project Engineer, *-1-2 r
B,. J. Murray TUSI Engineering Supy
- -1 l-J. J. Moorhead G&H Resident Engineer, *-1-2 y
B. C. Scott B&R Site QA Manager,. *-1 J. P Clarke B&R Site QC Manager, * ~
K. Mann B&R QA Records Coordinator. *
}
N
- 7. KirkTand B&R Prof. Generai Mgr,. *-T II. 7. DougTas B&R Project Manager. *-1 4
- 7. C. Frankum B&R Asst. Profect Mgr.,, *-1 g
P. Foscolo B&R Prof. Chief Engineer, *-T t Hancock B&R Mat'T Procurement, Con-struction Branch, *-1 A. Beren TUGCD Vendor Complianca
- A Vega TUGCC QA Central Staff Function C. Beggs TUGCQ Systems. Compifance,. *-T-Z R. Gary TUGCE V.F.
Operations, *-1 t.. Ffker TUSI V.P'.
Design & Procurement, *-1 P. Brittain TUGCO/TUSI President, 1 Interview I Pre-audit meeting-Z Post audit meeting Aud.it Method:
The aud.it was conducted through a series of interviews with responsible management end supervision and examinatiert of QuaTity Assurance manuals, procedures, records and work l
operations both at the Dallas headquarters of Texas UtiTities Generating Company and Texas Utilities Services, Incorporated anci at the Comanche Peak construction site.
Q Su::snary:
The audit disciosed that recent changes. in authority dele-gations had. been generalTy welt accepted. and that morale f
,-y
_..-.,,r.,y r
e.-
+-
l I.Q Sumary (6:nt'd): and taant spirit were good. However, the changes had n yet beerr fonnalized in revisions to the PSAR and the Conanche Peak Quality Assurance Plan. The audit also disclosed that present practicas in the control of design changes and of cartairr nonconformances do not provide the requisite level of review by the original designer.
In other instances it was evident that design changes were being used in Tien of nonconfonnance reports. Except f'or the areas noted. hereirt and below, there was generally good adherence to existing procedures.
Findings:
T. - The current activities of TUGCQ Quality Assurance per-sonnet are not cons'istent with the authority delegations to Browr: & Root and tar Gibbs & Hill as defined. in the PSAR and Comanche Peak Quat 4ty Assurance Plan.
Staf.TarTy, the QueTity Assurance Piarr and Procedures are not consistant with current and pTanned revisions in
~
autherfty delegations' to the Architect / Engineer and the Constructor, and fs not complete in addressing all eighteen criterte of 10CFR50 Appendix B.
,T,he lack of a weTT fdentified plan of reorganization and responsibility
~I causes uncertainty in carrying out some activities.
There needs to-be a plan for revising the Quality Assor-ance Program;. such a plan should include the establishment of art architecture of procedures to show-how other TUSI/
TUGCO: and contractor manuais inter-relate with the Quality Assurance Manuar. The TUGC0 QA Manager should establish-a schedule and assign responsibilities for completion of the necessary procedures. The scheduie should be supple-mented with a management effort to monitor adherence to the plarr and achievement of the schedule.
h Z.
The current site DC, DDA system of after the fact coordi-natierr of design changes with the original designer en
,,,__m
3, Findings:
provides a significant risk of design error and does (Cont'd) not meet the requirements of 10CFR50 Appendix B, nor of ANSI N45.2.IT, "QuaTity Assurance Requirements for the Design of Nuclear Power Plants".
A systen for expediting and documenting Gibbs'i Hill
~
home office approvals should be established using telephone, telecopier or telex as. a means of speedi,ng
__._I connunicatfort.
~
i 4
2.
De Comanche Peak Quality Assurance plan does. not provide for a Quality Assurance review of procurement documents and changes thereto prior to purchase order placemept, except for site originated procurements. Such a review m
is identified in 10CFR50 Appendix B,. Criterien IV and is a requirement of ANSI N45.2.13. It should be required y
ort aTT. safety related procurements.
E 4.
8 -
The current con 6fnatfort of Chapter 17 of the PSAR, the TUGCQ Corporate Quality Assurance ManuaT, the Comanche p
Peak QuaTity Arsurance Plan, Project Procedures and E
Browrr s Root Manuals and Procedures provides a complex array of procedures which is difficult to maintain l
current and consistent.
[-
5.
The current system of providing inspection ' instructions or checkTists t: in'spectors is too generic, placing an
?.,
undue burdert ort the inspector in attempting to determine applicable drawings and. specifications-and applicable revisions thereto. A review of records of concrete pours incidates that configuration reflecting the as-poured cond.ition is. not clearly defined. AppTicable DC DDAs are not 'noted in inspection documentation.
Configuration needs to be clearly identified to inspectors on e current basis, including all applicable l
. - - - - - - 1 o
Findingsi DC DDAs. and canpleted documentation must reflect the l
(Cont'd) status of the appTicabTe changes.
I 6.
Special, processing markings fer later in-service inspections.are carelessly appifed. The circia and f,
arrow used. for such marking is sometimes. inco6plete and not recognizable for its intended purpose.
In one instance only a. portion of the circle resembling the Totter "C"' was. discernible. FaiTure t'e properly marie thesETocators now wiTT cause delay and possitrie error whers in-service inspections are made in highTy frradfated areas.
7.
Dispositiort of nonconforming items does, not always f
~
achieve the requisite myiew by appropriately qualified
~
desigre personnel. A procedure, Timited to. defects in
{
concate,. was. recentTy issued.which bypasses the estab-Iished nonconformance controT system. and,. thus,. violates
~
reguTatory requframents ire this regard.
In other fnstances., the DC DDA program: has beerr used to bypass y
the nonconformance reporting system. The nonconformance-I control systere should be the means for maintaining ~in-i
-f:
spector integrity, identifying problem areas and provide e driving. force for their correction.
i 8'.
The records storage facility does not currentTy have any means of internaI fire protectiert during hours it is
}
unmannedw although it is understood some method is pTanned. QueTity Assurance re' cords, such as personnel y
quaTifications,. are not maintained in the Records Canter, I
but are maintained in fireproof file cabinets in a trailer under the cognizance of Brown 1 Root training coordinator.
,Q 9.
Approximately twenty-four percent of Central Staff audits 7
have not been conducted as scheduTed. Comeining Central Staff audits, site audits and site surveillance activities
-t 4
w v
,.--w
- ---,.~~-
,,-- -,----.-,-----.--,w.,---mn,--..-
i---..,,---,_
_. TUGC0 AUDIT REPORT "O'
Findings:
(Cont'd) by TUGCD and by Brown & Root into a single, cohesive program would provide improved visibility to the overa.li audit and surve11Tance effort and permit evaluation and s
ajustment to the audit schedule to attainable and yet effective frequencies.
2~-
M
/
m.
S
=
6
\\
k l e
'O 9
4 g
6 O
6 e
APPENDIX B TUAE LITILITIES GENERATINE COMPANY OBSEVATIONS AND RECOM4ENDATIONS U
~
5 S
e S
b 6
T i
1 e
)b h
e
i TUGC3 AUDIT CBEVATICNS AND RSCDbCATIONS
.O 1.
c = m rtou
~
A.
General TUGCD Quality Assurance has undergene censideratie-reorganizatien in de i
past year. The general thrust of this effor: has been the assuraption of
\\
granter direct invcivemen: in the mnnagement and supervision cf the Cccanche Peak Quality Assurance Progra=. It is to be ncted that incr: ant shifts. in responsi5111ty were' being made at the time of SC's reviewJ ~
As e part of tnis assessment $ MAC evaluatati:ne reactions of key managers.
supervisers and inspec:crs :2 the overali changes that have taken piace to date.
+-
1: was. genereTTy observed that those interviewed thougn: that witn few
=
exceptions the changer. were for the better. There appeared ta be a team ef"cre orr the part of GA and Construction with excellen TUST executive management and project management supper cf the CA pregram. There was ne noticeabTe prebTem:witir organizationai prejudice becught about by the crgani:atienti intemixing c" iUSr Brewrr & Rcct or Gibbs. & hit 1 werk f rces and sucervision.
3.
Orcanizadon i
During the course of the audit MC discussed the value cf a revised crgani:ationai structure wi 'r the TUGC3 Qua11:7 Assurance.".anager and the Manager. Site Survei1Tance.
.It is reccamended that TUGC7 adopt arr organi:ational realignment of l
activities as set for:fr irr Exhibt: i wnereby Quatt:y Engineering and Inspection report to the Site QA Supervtser as two separate sub-j ceganizational entities. with responsibilities as defined in Exhibit T.
Such an omni:ation wf11 better supplement the existing Construction i
organization and will permit better organization for nandTing day-:c-day 0
site prebTem as well as iglementing reconmendatiens cf this re; r..
This is particularly sc in the area of inspection planning.
Q I
............~w~
a m ur.rznuM 10MS 2-
' C.
Quality' Surveillance Comnfttee ATT minutes of meetings of the QSC since its inception were reviewed.
/
It is noted that the QSC was established as a mechanism for providing top TUGCD management with a periodic update on such matters as " schedules and miTestones" or " audits and corrective actions".
r 1
ge#
/
It was noted that recent meetings dealt with ' tracking on the status of r_...
action ttems as set forth in the Outstanding Surevillance' Report Items 8"
or the Quality Assurance Items of Concern Report.
In such instances.it #
appeared the Quality Surveillance Connittee was. taking on the role of a task force or probTem solving group. The problem that exists if the QSC assumes. such a role is that probTems would tend te awaft the three month meeting cycTe before the necessary management attention is effected.
It is. recomended that TUGCQ re-evaluate the charter of the Q consideration given as. to its. value te the project recogni::ing that:
T.
AT.T action to. resolve problems. should be handlet on. a day-to-day b through the functioning organization, and 2.
The primary objective of maintaining; management awareness of Qu t J.
Assurance status might be accomplished more efficiently, effectively and on a more timely basis through a monthly Quality Assurance progress report distributed to the TUGCD/TUSI executives.
D.
QuaTificatierr of personnel MC reviewed the qualifications of all TUGCO/TUSI and Gibbs. & Hill Quality Assurance personnel and many of the Brown & Root personnel.-
It was observed that most of the TUGCD/TUSI Quality Assurance personnel have gained their Quality Assurance experience through Comanche Peak activities only.
Although the project has provided valuable experience it is ree-amended that any future assignments in Quality Assurance be filled with
{
quality engineers hired from outside the company with broad nuclear ex-perience,. preferably in construction. Such experience added to the existing: staff will serve TUGCD/TUSI well in accomplishin,g the important
..m=
m*
M e-v--
~
--.--,-,-----,.,-,my--,,--,-,y,-
- +. -. - - - -
i
.... -............ _. _.,.. ~
piping, electrical and startup activities ahead.
MAC had occasion throughout the audit to assess the quaTificatiens and I
expeHence of 20-30 inspectors throughout the construction site. These /[
observations are worth mentioning:
L 4
1.
The inspectors are generally young and inexperienced with many having h l
as little as six months experience in inspection.
,. g/
- 2. There was an obvious need for more seasened inspectors te work with the novice inspect:rs on a day-tc-day basis.
4 2.
Toc much responsibiTfty is piaced on the inspectors with respect to preparation of inspection planning, resolutiert of site problems and determinatfort o* the design configuration base for perfomance cf inspections.
E II.
QUAI.ITY ASSURANCE PROGRAl*
The Quality Assurance Progrant is deffned in three basic documents:
I h '-
The Corporate Quality Assurance Manual C.1 apter 17 of the PSAR The Comanche Peak QuaTity Assurance Plan These documents are not irt ::tal agrese.ent with one ancther. Since tne e is no other nuclear piant cur ently planned and since tne authority dele-k gations identified in the Corporate Manual are net in consen'ance with practices on Comanche Peak,71JGC3 should consider discentinuance of the
)
j -
Corpcrate Manual unTess there are other prtjects to wnich it is to be applied. If a Corporata ManuaT is required at a later date, a new one c uld be prepared based ort Comanche Peak 'experient.e and the requirements of any new projects to which 1: wouTd be appifed.
E The Comanene Peak Quality Assurance Plan addresses caly the fellowing e
criteria of the signteen identified in 10CFR50 Appendix B.
.....,,..wn
. annraumuuns Q
Organization Design Control Procumment Administration Inspectiorf Nonconfomance Contrei Document Control Records Audits 4f ttr the expanded responsibil.ities of the TUGCU 'QuaTity ssurance Depart-ment, the pTan needs ta be expanded to address aTI eighteen criterie to reflect the creation and functions of the Procurement Department and to be consistent wittr the authority deTagations and functions stili resting
!I with Gibbs & Hi1T and witte Browrr & Root.
m
~
There needs te be e pian for procedurat identiffcation and development and a schedule and assigned responsibilities for their completion, including-a compTete architecture of' Quality Assurance procedures,. pro,1ect procedures and interfacing procedufes of the Architect / Engineer and Constructor.
The effort. shouTd be te minimize the nusber of procedures mquired and to eliminate dupTicating or overlapping procedures through consolidation of detafT arrd joint approvais of the' organizations involved.
It is recomended that the QuaTf ty Assurance Manager use his organization as. the driving l
force to achieve required procedural coverage arr schedule.
~
It was noted that TUGCC is planning ort obtaining its own Code manual.
The stated reason for this was the fear that Brower & Root would not a Code acceptance. The auditors feel that the Brown i Root manual would be acceptable ta the Code Survey Team and that its weTd practices as exempli-find in. the Weld Shop are vary &cceptable. The auditors are of the opinion that obtaining a Code Stamp wiTT be difficult where all the werk of fapiamenting the progrant is perfomed by others.
III.
DESIGN CONTROL The present system of expediting field changes by referring design changes t:r.the originaT design organization for approval after the fact does not t
l
,____,...,m.,
- 2..
. lutico DBSERVATIONS & RECOMMENDATIONS
-I-L p
f
'Q meet the intent of 10C'R50 Appendix a nor of ANSI N45.Z.11, which requ that field changes be subject to design controls comensu exercised on the original design.
T1JGC0 audits have already disclosed that the Architect / Engineer has not been reviewing field originated c on a concurrent basis, thus the design engineer's coments may be feceived
~
Y after the specific construction work is complete 'resulting in possible less of design integHty, undue pressure on the designer to justify what has been done, loss of designer responsibiTity or possible*exte'ns.ive ~
repairs.
It.is recomended that a system for expediting review and approva! b'y the origineT designer be estabitshed on al'1 safety related changes using teTephone,. teTecopier or teTex as necessary to coordinate and document change approvais.
IV.
PROCUREMENT DOCLHENT CONTR0i Except for site procurements, the Comanche Peak Quality Assurance pian
, does. net provide for a review cf procurement documents and their cha prior to placing a purchese order. Thris fr. contrary to requirements cf 10CFR50 Appendix B',. Criterfort IV and ANSI N45.Z.T3, " Quality Assurance Requirements for ContrcT of Procurement.....". There is a review of pro-curement documents by Quality Assurance during Design-Review, but is was ascertained that this was a review of the drawings and specifications and not the purchase order or contract.
Procurement document review by Quality Assurance should assure that a nacessary requirements for access. to the supplier's facilities are provided and that necessary contreis and. documentation have been specified. and that the appropriata configuration has. been defined.
The review shouTd.
alse assure that requirements imposed are appropriate to the procureme and that there are no excessive requirements for quality program develeo-ment or for the delivery of unnecessary documentation.
Same of the pro-curecent packages reviewed appeared to. have both blanket requirements for Quality Assurance programs and excessive requirements for docume~ntat O
Y.
INSTRUCTIONS A.
Insoection plannino I
The curmnt system of providing inspection instructions or checklists to the inspectors is too generic in nature.
In the case of concrete inspection planning. the inspector filis out at simple pour card with an attached Concrete Placement Checklist, a Reinforcing Steel, Electrical Nehanical and Embedded Item Placement Checklist and a Stainless Steel Liner Checklist,.
[
the combination of which:
1.
Provides no. information with respect to-unique,. embedments or penetrations to be incorporated. in the pour.
k; I
R; 2'.
Piaces an undue'burderr on the inspector in attempting. to detemine' applicable drawings, specifications,, applicable revisions and m
y applicable DC DIM's. Muc!r of this input should be prbvided by g
eiericai support under the direction and subsequent approvai of g
()
{,
- a. quality engineer,
~
Inspectors estimated that 45-70%. of their time is spent on docu-mentatfort rather thart physical inspectiort activity. Welt thought out planning could do much to alleviate this situation.
Traceability It was observed that Comanche Peak has established a program of
~
unnecessary material traceability which,hased on one estimate, consumes fe at Teast a three-marr Tevet of effort and perhaps as ! sigh as a six man level of effort if one considers aTT the support functions required to-
{_
implement the program. ATT anchor bolts and B series cadweids are fully traceable to heat numbers such that through an elaborate and exter:sive system of mapping aTT instaliations,. the capability exists of identifyina each embedded anchor bo115,. B series cadweids and other standard embeds to its heat number. There exists no such NRC or industry requirement for this degree of traceability.
It is. inter-
{
' f esting to note that rebar does not require traceability on Comanche Peak (and shouTdn't) MAC knows. of no other project that imposes this require-
[
~ ~ ~
~~
~
-/-.
ment and could not identify a Comanche Peak specificatierr or procedure requiring it. Accordingly, it is recomended that this practice be dropped imediately. Such a move wouTd enhance inspector morale as those involved are aware that the practice serves no useful purpose.
k C.
Procedure Simolification Newly established procedural systems are such that Construction and
. Quality Assurance issue procedures on similar subject ma,tter jointlyr
~
for example,. the recentTy issued procedure on shop travelers was.
(
jointTy prepared by Construi: tion and Quality Assuran'ce.
It is recom-mended that important procedures such as those related te concrete be
{
revised and fssued as a single 'procedum approved by Constructiert and Quality Assurance.
Similarly,, those procedures related to piping and eTectricai should: be revised and jointTy issued as a single Comanche 2
Peak procedure.
D.
Procedures. Endeoendent of~ Houstarr '
The present systent of obtaining; Browrt & Root., Hcuston office approvai on construction procedures should be modified. Guidelines should be worked out wfth the Houston office whereby they appmve only top levei procedures permitting the site fuTT flexibility in revising detailed p
site procedures.
Perhaps the Brown & Root, Houston office could retain approval authority on those top Tevel documents that establish Brown &
Root policy, controT the necessary type of forms etc.
e However,.
detailed operating procedures should be changed with site approval onTy.
Perhaps the Houston office wouTd agree ta a retroactive review procedure n
E..
Conffeuration ControT
.N A review of records. for completed concrete pours indicates that the V' configuration refTecting the as-poured condition is not property p
defined.
It was noted that the inspectors record the particular drawing e
number and revision letter, however, all applicable DC 00As are not
-O noted anywhere in the-inspection supporting documentatiort.
g e
e
-c-
- s
{
F.
Preol'anning of Construction Work In-discussions with construction management personnel it was indicated i
that a new scheme of construction planning is being developed. This new scheme provides for a d'etailed material takeoff on all Gibbs & Hill b'
drawings which provider detailed instructions to the crafts as to-?he i
civil, mechanical and electricaT items to be included in each segment of work. This femalized approach of taking material takeoffs in the
[
office and providing this infomation to the field forces on an approved
~
materiaT takeoff Tist will daemuch to. improve the quaTity of the work.
Since the materiet takeoff is a femai process accomplished by construction engineers welt fn-advance of the work, it provides a significant measure j
of preplanning, including the process.ing of necessary design changes tn accomplish the work. Such art effort will do much to minimize field errors with respect to. Taft out embedments or inability to complete work as a result. of design errors.
It is reconsnended, however that this effort bed 0 femalized inta e Comanche Peaje site procedure. As such, it will be recognized' as part of the system-and will de much to assure that Gibbs, &
Hill drawings are forwarded,ts.the site on a timeTy basis to accomplish this preplanning effort.
VI.
DOCUMENT CONTROL While there appeared ts be some problems with bringing the Automatic Records Management System on Tine, the manual system backing it up appeared.ta be functioning satisfactoriTy. The auditors found no
' deficiencies in doc:snent control.
VII.
CCNTROL OF PURCHASED MATERIAL, EQUIPMENT AND SERVICES The QuaTity Assurance PTan is not up to. date in regard to TUGCO's responsibfifties for procurement, source evaluation-and source sur-veillance. TUGCC has developed. e program for rating supplier perfam-ance and shows evidence of action: when reatings are unsatisfactory.
h The list of suppliers mquiring evaluation and source surveillance is not kept up to date by the Architect / Engineer. The-list in use is over g..
f I_-
. -.. = w m m a
- four months old, but is maintained manually by the TUGC0 Supervisor of Supplier Compliance.
There does not appear to lie any method. of assuring that the latest con-figuration is supplied to the personneT performing source inspection prior to shipment of procured. items. Thesourceinspectorappeakto be at the mercy of the supplier irr determining W1at changes have been j
identified and incorporated. Thus it is conceivable that items wi1T,be-shipped to. the site that det not meet the desired configuration evert though requirements of the purchasing document have been met.
Such recafpts can cause delays. and unwarranted costs in meeting the proper configuration.
,I It is reconenended that. a practice be estabitshed of identifying and confirming required configuration prior to procurement and prior to shipment of purchased materials and components.
VIII.
IDENTIFICATION AN0' CONTROI:. OF MATERIALS. PARTS AND CO k,
No deficiency noted.
MateriaT reviewed irt the warehouse,. in open storage and in the weld. shop ap'peared to be adequately identified.
l!
ki IX.
SpECIAL PROCESSES l
A.
Radiocrachy
,h Iridium 192 is being used as the radiation source for all radiography at the site.
This isotope has its optimum capability at about 1.5 inch thickness of steeT and is not reconrnended by the Code below.75 inches
{_
It is pemitted for lesser tMcknesses when the use of other radfation sources is not practical and when resolution-of the outline and 4T hoTe size of the penetrameter can be demonstrated. The energy levels of iridium isotopes are higher than optimunr for materiais.375 inches or I
thinner, resulting in a flat image and Tack of contrast.
Because exposure time relates to distance, the isotope is nomally placed against the pipe opposite the film.. With a.100. inch' source size, this causes blurring
{
of the image.
Lack of contrast and a blurred image makes it unlikely f
'~~
\\
10-
\\
\\
that hairline cracks will be seen and difficult to accurately define i
large indications.
The use of f ridiunr 1g2 meets the minimun requirements cf the Code, but by not providing optimum identification of observed anomolies it does three undesirable things. First it causes unnecessary removal and rupair of indications that can be seen but not properly k
identified; secondly, it masks narrow cracks, tight lack of weld pene-tration and non-fusion which can be detrimental to service life; thirdly, l
it does not provide art adequate base line for in-service. inspections-performed after the plant has, gone-inte operatforr.
FaiTure to have cTear identification of the originaT' indications at that point can cause delays t
the cost of which greatTy exceeds the cost of providing. better identiff-I.
cation and necessary repair of defects found in the construction-phase.
f' Recossoendation - It is recommended that TUGC0 require x-ray for sh
~
and consider its use where practical for construction weTds.
X-ray machines in the range of tttirty pounds of weight are avafiahTe and are f
nearly as. portabTe as the.' isotope. Because of its smeTier focal spot size andi variable voltage,. x-ray can gtve superior radiography.
The feedback of informatforr to.weiders can improve the quaTity of welds and i
minimize the potential' for defects. The abiTity to discriminate between indications having roundness or sharpness at the ends can eTiminate re The ability to positively identify in the construction phase those indi-cations which have a potential for growth and fa1Ture can permit econom repair without radiation hazards that are inherent if found later in the operating phase.
R.
Weiding No causes for concern and na procedurai noncompitances. were found. irr review of the weld shop. There seemed-to be a general opiniorr that after radiography repatrS are being required that are acceptable witnin the Code.
A review of a smell quantity of rejected film-indicates this.
generatization may be valid.
It was disclosed during the audit that radiographs may be reviewed by as many as five individuals.
7 cessive review leads to supercritical evaluation of film and to excessive[
V Such ex-repair.
As. previously stated., better radfography permits better
__..-._.7,_-___i
ivwu Uc x MVATIONS & RECOMMENDATIONS
~
-7 7-u identification. of conditions acceptable within the Code.
Unnecessary repairs increase cost and reduce pipe reifability.-
)
R Recomendations - Have radiographs which have been rejected for defects reviewed by TUGC0 Level III radiographer. If a reasonable statisticai l-sample shows that excessive repair of weids has been required, establish the policy that Code acceptable indications shalT remain untouched, but
}
shall be recorded on the reports.
As ax economy, consider re5scing the number of persons. performing sequentiat review of radfographs.
C.
NDE Qualifications.
The site NDE Level III situation; is. uncTear.
~
Only LeveT II certification by Brown & Root was avaiTable for the NDE Supervisor; however, it is understood that TUGCC has. issued. a Tetter identifying him 'as Levet III.
1 f-Recomendation - Clarify the authority and responsibility of the NCE supervisor in administering-tests and evaluating and certifying per-sonnei.
This is very important as reTated to Code work since,t,he Level III will be working under the authority of the holder of the Code stamp.
1 o e X.
INSPECTION-
[ '--
There were no deficiencies. noted relative to inspeccion; howev~er, it war noted that a large number of inspection. personnel are receiving
~
their first nuclear construction experience on the Conanche Peak site As a consequence, it is necessary to improve the quality cf inspection planning and to increase the-TeveT of supervision and quality en Inspection planning should identify the required configuration support.
including applicable DC DDAs,. the features to be verified, the~ inspect method and acceptance criteria in order te minimize possible confusion h
and error, i
E,
TUGC0 OBSERVATIONS & RECOMMENDATIONS '
~
XI.
TEST CONTROL
'j A review was made of the TUGCa startup administrative procedures, with the following observatione.
1.
The procedures. appear to be written around the old organi:ation;
(-
that is, in several instances they refer to the Brown & Root QA/QC input required in the preparation of "startup work requests".
Z.
It was noted that an unique system is being established to handle I
noncanformances during the startup. phase. It is rsecamended that wherever possible exis' ting schemes utilized in construction be used during the startup process. This fs important since most personnel involved in dispositioning such items as nonconformances and design f_
changes wf1T be the same persons. involved in construction.
=-
XII.
CONTROL OF MEASURING & TEST EQUIPMENT t
' [
.e.,.
It was observed that out'of 24 instruments sampied which are utili:ed in civil,, structurat, mechanical and. eTectrical work, approximately 50 percent had not been withdrawer from the caThratierr laboratory since 1:s Tast caTibration date. This is. particularTy significant when it is recognized that the present systant is such that if a calibration date becomes due, the instr:anent is recalibrated whether or not it has been issued for use.
It is reconmended that consideration be given ts simply changing the calibration date rather than going througn a calibration cycle if the toot has not been used.
It was noted that many constmetion toofs are calibrated.
It is.
important to note that calibration of constructi~on toeTs fs not necessary with respect to.10CFRSIT Appendix E.
Although calibration and maintenance is extremeTy important on construction tools, it may be that frequencies may be relaxed.
XIII.
INSDECTION, TEST AND OPERATING STATUS No deficiencies were noted in this area. Material and equipment observed in receiving inspection, in the warehouse and outside storage 4
.a
,,n-,
l.,,.
-u.
s area appeared to: be adequately identified. No tests were observed.
XIV.
{
HANDLING, S~pRAG? AND SHIPPING
~
Exterior storage pactices should be reviewed. The protective coverings of many items are damagedi some reported on monthly surveillance-reports have not been corrected.
Large temporary structures, such as -those over the emergency diesel engines, require wind bracing to prevent further damage Because of soil chemistry, rain and humidity, the -current pi ac- ~4 tice of eTTowing Targe stainless steel piping to remain uncovered sho be reviewed.
Sensitized. stainless. is extremely sensitive to chloride,.
fluoride and suTphide contamination dich with water as a couplant can 1
cause intergranu. Tar corrosion and premature fafTure.
XV.
CONTR0t. OF NONCONFORMANCE;S '
i =._
There appears to be an effort to reduce the number of documented non-conformnnces.
It was noted that DC 00As, w:re being utiTized for nonconformance reports.
ATthougtt this was observed on a smalI percentage of DC DDAs issue the month of Apr1T, it is reconmended that this practice be stopped imediately.
The TUGC0 system is. correctTy established whereby non- -
confonnances are written after the fact and DC DDAs are reserved design changes before the fact. It is important that this practice be enforced.since DC DDAs. prepared after the fact necessitate that workers be directed verbally to.vioTate the drawing since the deviation will be handTed after the fact with DC DDAs. This is t poor Quality Assurance practice.
Procedure CPQT-AB,. Rev. O., dated 5-5-78 was issued for the purpose of providing expedient disposition of concrete discrepancies.
The procedure infers that discrepancies of 72'F versus 70*F or 6.2% air content versus 6.Ct. maximum is perfectly acceptable when it is signed off by the field t
engineer.
Such a system shortcuts the established nonconforming material control system as defined in Brown & Root and TUGC0 procedures and should Y
--._-,-,,--_,--e----
,,.-n, _,._ --
m
,,w,_.,_
e,--
. olo
.J. n....,
]
o me.mr.mm i ovu
_m
,.r,,.
{
be disc'ontinued.
If tolerances are unrealistic such that the 72*F is acceptable, then the design specification should be changed to se indicate.
It is recomended that good inspection-planning be provided inspectors, identifying the characteristics to be inspected, the method. of inspection and acceptanca criteria and that inspectors identify nonconformances~to such criteria. Th,is will maintain the integrity of inspectors and provides identification of problern areas and provides a means for the,ir correcti.on.
i It is reasonable te assume thaN orr a project as Targe.as. Comanche Peak, there will be severai thousand nonconfomance reports. The number does not reflect adversely on the quaTity of construction, but the failure to identify nonconfonnances does refTect adverseTy on the integrity of i.
inspectors and Teaves unknown the quaTity of the plant.
=---
L XVI.
CORRECTIVE ACTION There were no deficiencies noted reTative te corrective action on hard-ware. The SuppIfer CompTfance Supervisor has established a method of tracking vendor performance and ' hows positive results. frorr actions taken s
t: correct supplier quality probTems. A review of reports of site sur-veillance conducted by TUGC0 showed corrective action responses. were
~
being promptly received. A review of reports of survei1Tance actions by Brown & Root showed generaT1y adequate response and resolution of corrective action acept for e period of four months wher; surveiTlance personnel were assigned to other tasks.
~
Irr general,. corrective action appears ts, be adequate and timely on vender and site related problems but some deffefencies identified. in audits of malcr contractors. stilt persistu Tome of the changes in authority dele-gation to major contractors appears to be actiert taken to correct inadequats-or untimely response by those organizations;. however, other actions. taken, such as handling of field changes and nonconfonnances, appear t6 be those l
of circumventing the problem rather tharr correcting it.
l W
.__,,.,,_.,_,,_,._.,__,,___,,__.,,,,,.,,,__,-m_..._._,
r l....
XVII.
RECDR E Except for lack of internal fire protection, the quality rec:rds area is considered to be satisfact:ry. Some Quaitty Assurance re::r:s, such as personneT qualffications, are ner s : red in the records canter but are mintained separately by the Brown & R :: training cocrdinat:r.
There is not curren:Ty a catalog er Tisting of required ree:rds air.ncugn,
it is being prepared. A review of a' selection of Quality Assurance records showed the documents, in them s have been preperly cac:pleted..
)-
and frt :ne c:rrec: crder.
l
~
Recoemendatiert - The installa:ferr of arr fra r: gas fire extinguishing
\\'
systeer er the identfficatfert of geograstricaTiy separate duplicate rec rds 1
shcuid be expedfred. TUGC3 shouTd review the fire prete:: fen capabfif ties cf storage facfTf tes in the training supervtscr's trat1er and consider
_-. _ a dupTicate set c" suca rec:rds t:r be maintained in the-ree:rds cantar.
i XVIII.
AUDITS There are-several audit anci surveiTTance progras:s in ef'e::. Audits by the Quai 1:y Assurance Departner:: Central Staff are per*crued en site a :fvities,. mujer==ntractors and suppTiers. Site surveillance ac:icas are per'erned under the direction of the TUGCU QA Site Superviser. Si=ilar'
'i 66 surveillance activities are carried cut under the direction cf :he Brown &
Rect Site Quality Assurance Manager. Whfie calied surveilTance a:.iens, the surveillance pr: grams are forst1Ty planned and seneduled, utili:e l
[
cneckTists to guide the a :fvity and record resuits,. and issue rescr:s cf deficiencies and require correctiert. Exce::: for fernal and doc:ane.v.ed pre-audit and post-audit metings, aTT the elements of an audit pr: gram are irr place. It was repcrted dat the reason for ca!iing the activity "surveiTTance" was := avcid cutside audit:rs finding the prog a= defiefent because it did nc: incTude the dec:anented pre-and pes:-audf: meetings, l
t yet the auditors fcund that such meetings wre conducted, but en ar.
infereal basis.
Recamendation - The auditors cens.ider :na present pr:gra: := te an effe::ive ::ci sich c=uld be further i:x:r:ved. TUGC0 shculd ::nsider t
..e,
---.y w
~- - --- -- ' - - -- -'~
combining the audit and survefTiance activities into a single, cohesive effort. Such air integrated effort could cover required areas. more' efficientTy, without duplicattorr and. at a frequency that can-be mein-tained. Such art audit progrant should be described in written procedures and include e descriptierr of both the formai audit and the continuour audit plan (survei1Tance) and the method of' conducting pre-and post-audit meetings should be described to preclude later criticisms by outside organizations. The resulting audit program-should be a superior tool for management assessment of progrant impTementation and effectiveness.
=
e r
ee -
- 6 k
.g e
e O
~, - -
r
.r,,
r--,,,,n-e.-
--,,w-,-
rE...-
s 1-us 5
e 8
m e
m I
ECE 8" E E
M
=mE v
U mg-am-a
<1 EdCE*C
.r mm
= ~
m>=
- s
..=<=E<-.- 8
=
1 n
am 1
5 1
=.
E. e.-
c.
w E
E. -
COE m.
u m-a
=
CE 2
M m
- k."
- k. *
-w:
5 25=
vE 5"
[
=
4*
E w
g 5 =.
w t"aWEW 8.
-=
-m
-w We u
~g=~#e~t 5 EWtERS g-
-wo-E m.
m a
-=
m WM Q
W=RE Qm==6 W
I O
b e,
as w. oggummmy e MM e 6 age
- "'N~
9 y.
..- r.
._..,.y
8 ',yy.,,
'bTECv UTILITIES GENERATING COA.;iNY
,-. s.
1, k
l O FFICE M E M O R-AND IT bt A
- s ro,c p.m c. 9m.
w r.. uyn,an J.} f 9 Subjec.
Nnaement Ouality Assurane* Audit RECETVRh JUL 12197-i P. G. BRITTAIN Attached are our resolutions to the findings and; recanenendations. made-
,)
by Management Analysis Company as a result of their audit in May.
Our analysis of the audit results has been discussed in general tems with John Jackson, and we see no need to respond fennally to this audit.
en og a 1
a.p. cary h
9 f
L. F.Hkar -
e h
e
/
,e
/
{
l l
4 1
FINDINGS (APPENDIX A) 1.
Findine Sumary: Our QA Plan and Procedures. do not reflect current autnority delegations to B&R and to G&H.
R'esconse: At the time of, the audit we were operating under properly approved deviations. These deviations were incorporated into a permanent QA Plan Manual revision on July 1,1978.
2 Finding Sumary: The current practice of after-the-fact design cnange review provides significant risk of error and is. in
~
noncompliance with 10 CFR 50 Appendix B.
Response: We disagree; orr both-counts. The G4H Resident Engineer has. exercised extremely good judgment in implementing the authority delegated to him. Of the approximately 2000. changes /devia-tions/clartfications issued under the system, we are aware of none that hrave provided exposure to. a significant risk of error. To.
provide greater visibility of the design change functiott, a system was, implemented on May 25, 1978 that provides an analysts of all changes and permits, continuing evaluation of the field efforts.
Thts syster is current for ongoing activities at the present time and wtT1 be completed for past activities on or before July 15, 1973.
Our belief that we are in fact in compliance with 10 CFR 50, Appendix B' is, supported by internal audits by TUGC0 QA personnel and independent audits. by twar separate NRC Inspectors. We propose to Taave the design change system as is.
I'.
Findine Sumary: TUGC0 QA does. not review-all procurement doctments and enanges thereto prior to release.
Response: We disagree that' this is, a requirement. A separate QA requirement section,. approved by TUGC0 QA, is included with each purchase order and is. applicable to all supplements. Changes to these recuirements are authorized only by Quality Assurance.
4.
Finding Sumary:. The current array of CA manuals. and procedures is complex ana a1fficult to maintain, Response: We agree. The new. Plan manuaT was issued July 1,1978.
The corporate QA Program. Manual. is currently under s*wdy with the goal of streamlining it.
5.
Finding Sumary: Records do not reflect the as-poured configuration clearly.
Response
Configuration has always been made visible to inspectors, i
l
..m
,.,_,.m.
and steps have,been taken to improve the visibility for the record.
6.
Findine Sunnary: Markings for in-service NDE inspections were not always cistinct.
Response: We agree that_such markings should be Tegible. QC will inspect special process ISI markings prior to turnover.
7.
Finding Summary: We are using the DC/DDA (design change) program ta bypass the nonconformance system.
Resoonse: This is not true. If construction identifieiand' corrects. a. defect or obtains an approved engineering change prior to:
QC. inspection,. no NCE is. required.
~
8.
Findine Sunnan: The records, storage faciTity does not have internal fire protectfort during. off-duty hours.
Response: An inert gas fire protectiorr system is ort order by TUSI.
Target date for installatf ort is. August 1,1978.
9 Finding Sunnary: ApproximateTy 24% of audits scheduled by DalTas staff nave not been conducted. Audits by TUGCG and.B&R shouTd be comotned tre one overali effort.
Response: Our audit scheduie is constantly being revised to reflect.
cnanging manufacturing status and ta aTion us to, use audits. to investigate probTem areas of the most innediate concern as they arise.
We believe we cart defend our audit program, and are leaving -
it as it is. j t
9 e
e 9
u.
,m
..-k.
OBSERVATIONS AND RECOMMENDATIONS. (APPENDIE B)
O
~
I Organization Recommendation: Separate inspection from Quality Engineering (planning).
Response: We had begun work on this before the audit. Our peditiet assurance group has taken responsibility for Quality Engineering.
Recomendation: Re-evaluate the charter of the Quality Surveillance'
~~
connittee.
Responsac We have decided te discontinue the Quality Surveillance Committee. Instead TUGCC % Manager wili issue a report quarterly to keep top management apprised of the status of QA matters,.
including quality trends. The first suctr. report mili be issued by August 15, 1978.
m Reconnendation: Hire more seasoned inspectors. Hire any future engineers from outside the company wittr experience in nuclear plant constructiorr.
Response-We are ort the Tookout for weTI qual.ffied personnel and continue ta review applicants frour outside the company. The reTatively young inspectors wil1 be strengthened best and quickest by taking from thes the responsibility for inspection pTanning and providing them wittr adequate, but concise instructions and checklists. This has. beert done and is effective.
i II Qualtty Assurance Program Recomendation: Revamp our present QA manual system.
Response: We agree. TUGC0 % issued a revised QA Manual on. July I, 1978.
Opfnfon: Obtaining: our ASME Code stamp. wiTT be difficuit if all the wort is done by others.-
Response: We know, it will be dtfficult, but it can be done. The report does not accurately record our stated reasort for obtaining an H-stamp.
l III Desten Control
~
Recommendation:
Abandon our present system of expediting field
..... ~..
changes.-
O R oon : we w4ii 1 ve <= => te <s.
~
1 l
IV Procumment Document Control Recoussendation: All procurement doctanents should be reviewed by QA.
Response: We disagree for reasons. stated on Page 1.
Y Instructions Reconsnendation: Streamline-inspect!on planning and checklist preparat1on Resconse: We agree,. and have been active it: this, effort since Januart.
Reconnendation: Discontinue mapping individual standard imbeds.
m Response: Mapping has been discontinued, but we have retained a:
~~
reasonaole degree of traceabtTity on embedded itens.
Reconnendation:- Combine conhtfon and-QA procedures.
Response: Sfnce January we have been doing this for new procedures anc wnerr revising old procedures.
Reconsnendtion: Discontinue requiring BAR Houston approval of procecures.
Response: All procedures except those invoTving ASME Code work are now approved at the site.
Observation:
AT1 appTicable DC/DDA's. are not included in supporting doctmentatton.
t Response: Refer to Page 1, Appendix A,. Iteur F. -
Reconsnendation:
Establish a formal site procedure for planning-construction woric.
Response: This planning is. being done. We don't intend to create a.
tormal procedure for it.
O l
l
VI Document Control Observation: The ARMS system should be backed up by a: manual system tstatee versally in exit management interview).
Response: The auditors' lack of confidence in ARMS was the result of insufficient familiarization. with the systen arr their part.
VIL Control of Purchased Material, Eauipment, and Services.
Observation: The QA Plan is not up to date in this. area..
Response: Manual revision was. congTeted orr July 1.
Statement:. Source inspectors. (TUGC0 QA) appear to be at the mercy of the supplier irt deterisining what changes. have beerr incorporated.
Response: This is not true. We (TUGC0 04) prepare our own checklists after searching the appropriate files. We agree with the m
auditors' (verbal) consents that this. method places an additional burden on ourselves-however, the original (and. Togical) approach of
- c depending upon Galt tar prepare the checklists did not work.
VIII Tdentificatf ort and Control' o'f Materiais, Parts, and Comoonents No observations IX Soecial Processes Recommendation: The use of Iridium 192 should be replaced by x-ray for all shop welds and for field welds where practicaT.
Response: Browrr ir Root is studying this recommendation,. and is cosmitted to have e report for TUGCO/TUSI by July 15.
Recommendation: Reduce the number of individuais reviewing.
ractograpns and establish the policy that Code-acceptable indications. be recorded. but not repaired.
Response: This had been accomplished prior to the audit, but apparently the auditor talked with someone who wasn't aware of it.
f Reconmendation: Clarify the responsibility of the NDE supervisor relattye to code work.
O an.
m
_,,7m--
,,., y
Resoins7: Th2 auditors were confustd in this area. Our.only Laval 111 at the site is in civil work, not ASME Sec. III Div. I work.
X Insoection Included here was a re-statement of observations covered elsewhere in the report.
- ~
XI Test Control Recomendation: Existing nonconfomance control systems should be..
useo, during startup.
~
~
Response: The new QA ManuaT,. with fnput fros '11GC0 Operations, will T
accress. nonconfomance controi systans.
XII Contret of Measuring and Test Ecufoment.
Reconsnendation: Equipment should not be recaTibrated on the due i
date if it has not beert used.
~
Response: The reconsnandation has. been adopted for items whose caltoratiorr is not subject te change while not frr use.
Implementatfort date was July 1 XIII Inspectiert, Test and Operating Status No observations XIV HandTine Storace and Shinoine ti l
Recomendation: Exterior storage & protectiotr practices should be reviewed.
-Response: We had previously reviewed the storage practices and have-no: reason to beTieve that a problem exists. Various NRC fnspectors have also inspected this activf ty. However,. Westinghouse is rewfewing this, and their matallurgist will report by July 15.
XV ControT of Nonconfomances-Observation: This is a restatement of concern over nonconferinance control.
Response: We reject the inference t' hat problems are ciretanve'nted ratner than corrected at CPSES.
Mfic findings.
XVII Reconsnen system. dation:
This sectiott restat Resjo,n,se,:
The inert gas ystemed the need for fire pro XYIIIA s
tection M
will be installed by A Recommenda ugust 1, 1978.
ad~ EGtiT4E" tion:
sniiuld be countne(d.The audit D Res sAurveti1ance.e intend to leave ou nd surv 11Tance (sit nse:
W e
\\
audit progam. separate f m
m.m
~
rom ite s
N...
S*
e e
e m-I
~
7<
S h
9 6
f h.
y
f'
{4N f_t s
/
In Reply Refer To:
h c, {'
~
Dockets:
50-445/84-32 FEB 15 HE S0-446/84-11 yd 1
,ltI g1>
Texas Utilities Electric Company ATTN:
M. D. Spence, President, TUGC0 Skyway Tower 400 North Olive Street
/p
. / C *g
[rvV Lock Box 81
- 0. W 4 '"
Dallas, Texas 75201 p } .
9 A
Gentlemen:
P
'W This refers to the inspection conducted under the Resident Inspection Program by Mr. H. S. Phillips of this office and NRC contract personnel during the period August 20, 1984, through September 20, 1984, of activities authorized by NRC Construction Permits CPPR-126 and CPPR-127 for the Comanche Peak facility, Units 1 and 2, and to the discussion of our findings with Mr. D. Chapman and other members of your staff at the conclusion of the inspection.
Areas examined during the inspection included a review and evaluation of how effectively Texas Utilities Electric Company management has implemented the corporate quality assurance (QA) program for design, procurement, and construction activities.
Special emphasis was placed on evaluating the management of the audit program; management's action to regularly review the status and adequacy of the QA program; and followup on findings pertinent to program management identified by previous NRC and consultant inspection teams.
Within these areas, the inspection consisted of selective examination of procedures and representative records, interviews with personnel, and observations by the inspectors.
These findings are documented in the enclosed inspection report.
During this inspection, it was found that certain of your activities were in i
violation of NRC requirements.
Consequently, you are required to respond tc this violation, in writing, in accordance with the provisions of Section 2.201 of the NRC's Rules of Practice," Part 2, Title 10, Code of. Federal Regulations.
Your response should be based on the specifics contained in the Notice of Violation enclosed with this letter.
These violations may be related to findings identified by the NRC Technical Review Team (TRT).
If the issues are considered to be similar, you may respond to the items separately or as part of the Comanche Peak Response Team Action Plan.
l 1
h h
\\
RRI St gv.
TL/TF DMS&S C/RP82 D/ ORS &P NRR 48 1
SPhillips/It OHunn cutt R5hngart 0 Hunter R0enise Noonan
% /)'. /8f p /p/8f
' / N]84 l /p/8g &/84 I* nh/84
,\\
7cMO x
Texas Utilities Electric Company.
Should you have any questions concerning this inspection, we will be pleased to discuss them with you.
4 Sincerely, "Crfginal Slgned byf a
D M. HU N N;ctT.T**
D. R. Hunter, Chief Reactor Project Branch 2
Enclosure:
1.
Appendix A - Notice of Violation 2.
Appendix B - NRC Inspection Report 50-445/84-32 50-446/84-11 cc w/ enclosure:
Texas Utilities Electric Company Texas Utilities Electric Company ATTN:
B. R. Clements, Vice ATTN:
J. W. Beck, Manager i
President, Nuclear Nuclear Services Skyway Tower Skyway Tower 400 North Olive Street 400 North Olive Street Lock Box 81 Lock Box 81 Dallas, Texas 75201 Dallas, Texas 75201 bec to DMB,(IE01) bec distrib. by RIV:
RPB1 RRI-OPS TX State Dept. Health RPB2.
RRI-CONST.
Juanita Ellis EP&RPB R. Bangart Renea Hicks R. Martin, RA J. Gagliardo Billie Pirner Garde C. Wisner, PA0 D. Hunnicutt S. Phillips R. Denise, DRSP TRT (CPSES) (2)
RIV File S. Treby, ELD MIS System
- Eisenhut, NRR t
4
E 1
I APPENDIX A NOTICE OF VIOLATION l
Texas Utilities Electric Company Dockets:
50-445/84-32 Comanche Peak Steam Electric Station, Units 1 and 2 50-446/84-11 Construction Permits:
CPPR-126 CPPR-127 4
Based on the results of an NRC inspection conducted during the period af i
August 20, 1984, through September 20, 1984, and in accordance with t'ie NRC Enforcement Policy (10 CFR Part 2, Appendix C), 49 FR 8583, dated March 8, 1984, the following violations were identified:
1.
Failure to Recularly Review the Status and Adequacy of the'QA Program Criterion II of Appendix 8 to 10 CFR 50, as implemented by the Preliminary Safety Analysis Report (PSAR) and the Final Safety Analysis Report (FSAR),
Section 17.1, " Quality Assurance Program," and ANSI N45.2-1971, requires i
that the quality assurance program shall provide for the regular review by the management participating in the program, of the status and adequacy of
~
the part of the quality assurance program for which they have designated responsibility.
4 Contrary to the above, the applicant did not establish quality assurance j
procedures to regularly review the status and adequacy of the construction l
quality assurance program; nor did the applicant appear to have reviewed j
the status and adequacy of the construction quality assurance program.
This is a Severi'.y Level IV Violation.
(Supplement II) (445/8432-02; j
446/8411-02) l 2.
Failure to Establish and Implement a Comprehensive System of Planned and
)
Per1 odic Audits l
Criterion XVIII of Appendix B to 10 CFR 50, states, in part, "A comprehensive system of planned and periodic audits shall be carried out to verify compliance with all aspects of the quality assurance program and to determine the effectiveness of the program." The requirements are addressed in the PSAR and FSAR, Section 17.1, " Quality Assurance Program,"
j 1
which references Regulatory Guide 1.28 (ANSI N45.2) and ANSI N45.2.12 i
(Draft 3. Revision 4).
Those commitments require that a comprehensive system of planned audits be performed on an annual frequency.
4 j
Contrary to the above, the following examples were identified which demonstrate the failure to establish and implement a comprehensive system of planned and periodic audits of safety-rplated activities as required, as noted below:
~
I Notice of Violation a.
Annual audits were not adequately addressed by the audit implementation procedures.
TUGC0 Procedure DQP-CS-4, Revision 0, dated August 9, 1978, only required two audits of vendors fabricating reactor coolant pressure boundary components, parts, and equipment; one audit of vendors fabricating engineered safeguards components, parts, and equipment; and audits of balance of plant (safety-related) as required by the quality assurance manager.
TUGC0 Procedure DQP-CS-A, Revision 2, dated April 16, 1981, required only that organizations will be audited on a regularly scheduled basis.
TUGC0 Procedure DQP-CS-4, Revisions 2 and 10, did not specify auditing frequencies for design, procurement, construction, and operations activities.
TUGC0 Procedure DQP-CS-4, Revision 10, based audit requirements on Regulatory Guide 1.33, Revision 2, February 1978. This 4
commitment did not fully address the requirements of the construction quality assurance program.
The above procedure and subsequent revisions failed to describe and require annual audits in accordance with commitments and requirements.
Earlier audit procedures were not available to determine if they met requirements, b.
Planning and staffing to perform 1983 audits was inadequate to assure that a comprehensive system of audits was established and implemented to verify compliance with all aspects of the quality assurance program, in that, of 656 sa7ety-related procedures (which control safety-related activities) the NRC review revealed that the applicant
^
sampled only 165, or 25 percent, during the 1983 audit program.
Consequently, significant aspects of the safety-related activities were not adequately audited.
c.
The Westinghouse site organization, established in 1977 to perform Nuclear Steam System Supply (NSSS) engineering services, was not audited by TUGC0 during the years of 1977, 1978, 1979, 1980, and i
1981.
d.
Audits of vendors that manufacture or fabricate parts, components, j
and equipment for reactor coolant pressure boundary and engineered i
safeguards systems have not been conducted annually dating back to August 9, 1978.
This is a Severity Level IV Violation.
(Supplement II) (445/8432-03; l
446/8411-03) t
l Notice of Violation 3.
Failure to Properly Certify a Vendor Compliance Inspector Criterion V of 10 CFR 50, Appendix B, states, in part, " Activities affecting quality shall be prescribed by documented instructions, procedures, or drawings, of a type appropriate to the circumstances and shall be accomplished in accordance with these instructions, procedures, or drawings."
TUGC0 Procedure DQP-VC-4, Revision 6, dated January 5, 1984, requires that Level II inspectors (Corporate QA) shall attend and satisfactorily complete nondestructive testing courses including eddy current testing.
Contrary to the above, one of six inspector's files had no documentation 1
to show that the inspector had attended and completed an eddy current testing course. Subsequent, discussions revealed that he had been certified without meeting this requirement.
The vendor compliance supervisor stated that this inspection skill is not needed since there is no present vendor work activity which would require this skill; therefore, this procedure was revised and the requirement omitted during this inspection.
This is a Severity Level V Violation.
(Supplement II) (445/8432-05; 446/8411-05)
Pursuant to the provisions of 10 CFR 2.201, Texas Utilities Electric Company is hereby required to submit to this office, within 30 days of the date of this Notice, a written statement or explanation in reply, including:
(1) the corrective steps which have been taken and the results achieved; (2) corrective steps which will be taken to avoid further violations; and (3) the date when full compliance will be achieved.
Consideration may be given to extending your response time for good cause shown.
P Dated:
i i
t I
l l
o APPENDIX B U. S. NUCLEAR REGULATORY COMMISSION REGION IV NRC Inspection Report: 50-445/84-32 Construction Permit: CPPR-126 50-446/84-11 CPPR-127 Dockets:.50-445 Category: A2 50-446 Licensee: Texas Utilities Electric Company Skyway Tower 400 North Olive Street Lock Box 81 Dallas, Texas 75201 Facility Name: Comanche Peak Steam Electric Station (CPSES), Units 1 and 2 Inspection At: Dallas Corporate Office, Dallas, Texas Inspection Conducted: August 20, 1984 through September 20, 1984
////!f8 Inspector:
H. S. Phi' lips, Senior Resident Reactor Date Inspector Construction i
NR: Contract Personnel:
B. Freed, Senior Project Engineer, EG&G Idaho, Inc.
G. Thomas, Quality Engineer, EG&G Idaho, Inc.
Approved:
M'
.wh@
M!43/f9
- 0. M. Hunnicutt, Team Leader Date
/
Region IV Task Force Inspection Summarv Inspection Conducted August 20 through September 20, 1984 (Report 50-445/84-32; 50-416/84-11)
N g2b[
U 3U fb
I Areas Inspected:
Routine, announced inspection to determine how effectively corporate management has implemented the QA program for controlling design, procurement, and construction activities; and to determine how site management interfaces with corporate management.
The inspection involved 74 inspector-hours by one NRC inspector and 176 inspector-hours by two NRC contract personnel at the corporate office and the site.
Results: Within the two areas inspected, three violations were identified (failure to regularly review the status and adequacy of the QA program -
paragraph 2b.; failure to establish / implement a comprehensive system of planned and periodic audits paragraphs 2c.(1) and 2d.(3)(a); and failure to properly certify a Level II vendor compliance inspector, paragraph 2d.(3)(f).
t
DETAILS 1.
Persons Contacted W. Clements, Vice President Nuclear Operations, Texas Utilities j
Generating Company (TUGCO)-
- D. M. Chapman, Manager, Quality Assurance (QA), TUGC0
- R. G. Spangler, Supervisor, QA Services TUGC0
- D. L. Anderson, Supervisor, QA Audits, TUGC0 A. H. Boren, Supervisor, Vendor Compliance, TUGC0 "S. L. Spencer, QA Auditor, TUGC0 l
D. Z. Hathcock, QA Auditor, TUGC0 H. R. Napper, QA Auditor, TUGC0 A. Vega, Site QA Manager, TUGC0 L. M. Bielfeldt, Supervisor, Quality Engineering, TUGC0 l
C. Welch, Supervisor, QA, TUGC0 J. H. Roberts, Supervisor, Construction /Startup, TUGC0 J. T. Merritt, Assistant Manager, Engineering and Construction, TUGC0 j
R. Gentry, Manager, Project Support Services, TUGC0 l
F. Peyton, Supervisor, Purchasing, TUGC0 M. Strange, Supervisor, Engineering Support, TUGr 3 R. Baker, Staff Engineer, TUGC0 l
H. Harrison, Supervisor, Technical Services, TUGC0 G. Krishnan. Supervisor Stress Analysis Group, TUGC0 j
R. Williams, Drafting Supervisor, TUGC0 i
G. Purdy, Site QA Manager, Brown & Root Inc. (B&R) i R. L. Moller, Site Manager, Westinghouse
- Denotes those attending one or more exit interviews.
2.
Texas utilities Manaaement of QA Activities d
a.
Introduction The objective of this inspection was to determine the status of the
+
construction QA program and the effectiveness of implementation of the corporate QA' program for ongoing design, procurement, and
~
construction activities.
The NRC inspectors reviewed the QA commitments described in Section 17.1, " Quality Assurance During Design and Construction."
Texas utilities Electric Company (TUEC), as the applicant, has delegated to Texas Utilities Generating Company (TUGCO) the responsibility and authority for engineering, design, procurement, construction, operation, and QA activities at Comanche Peak Steam Electrical Station (CPSES). Gibbs & Hill Inc. (G&H), is the Architect-Engineer (AE) and provides TUGC0 with design, engineering, and procurement services as requested. Westinghouse (W) is the Nuclear Steam Supply System (NSSS) supplier and provides TUGC0 with the design, engineering, procurement and fabrication services for the NSSS and the initial supply of nuclear fuel.
Brown and Root, Inc.
o.
)
)
_4 T
l (B&R) is the Construction Manager / Constructor and provides construction services at the site, including the QA program for ASME Division 1 Code work.
4 b.
Organization i
The TUGC0 corporate management structure and responsibilities were described in the Final Safety Analysis Report (FSAR); and the various TUGC0 QA manuals and procedures described how FSAR requirements were implemented to control design, procurement, and construction activities.
Recent organizational changes pertaining to the QA program were described in FSAR figures 17.1-1, 17.1-2, 17.1-3, 17.1-4, and 17.1-5 which were included in Amendment 50 dated July 13, 1984.
f Recently, there have been three important QA personnel changes.
A new site QA manager reported in March 1984, a new site quality engineering supervisor reported in August 1984, and a new vendor l
compliance supervisor was recently selected. These organizational j
changes were made to replace individuals who were reassigned or j
promoted to other positions, and these changes were reported to the NRC. The independence and effectiveness of the QA effort do not appear to be adversely affected by these changes.
i j
The assistant project general (APG) manager reports to both the VP of l
engineering and construction and to the TUGC0 Executive VP of l
operations. Discussions with the APG manager confirmed this and that he was supervised by both. This management practice is questionable.
The CPSES QA Plan Section 1.2, paragraph 1.2.1, does not describe the APG manager's interface with or the responsibility to the VP nuclear operations.
Subsequent discussions with TUGC0 QA personnel revealed that this position was discussed in the startup QA manual. This item is considered unresolved pending clarification of the QA plan and further review during a subsequent inspection.
(445/8432-01; i
446/8411-01) i c.
QA Proaram TUGC0 QA Program Plan and subtier procedures for design, i
construction, engineering, and procurement described the control of l
all related project and quality activities.
A sample of these l
procedures were reviewed and documented in NRC Inspection Report No.
50-445/84-22; 50-446/84-07.
l The Quality Assurance Program (described in the FSAR) provided the l
delegation cf design, engineering, construction, and procurement j
functions to prime contractors, subcoptractors, and vendors.
It stated that the TUGC0 audit program assured that these organizations i
had adequate QA programs and verified implementation of the overall QA program within TUGCO.
1 1
i
_y.._y_,,,,.
---w-
- ~,, - ~.
..,..,,,---_..--e.-.y-.-,...
..._ _____-- r._,-w
,,,.,__...y..,,
,-r,,.--,_,,_.,,m, e.r.
r,
l
. The inspectors reviewed the QA program procedures and any objective evidence to determine if the applicant regularly reviewed the status and adequacy of the QA program as required by Criterion II of Appendix B to 10 CFR 50, the PSAR and FSAR, and ANSI N45.2-1971.
Reviews and discussions revealed no documented requirements or evidence that the QA program status and adequacy had been reviewed by the applicant.
In order to determine if the QA program had been assessed, the inspectors reviewed additional information.
In late 1981 and 1982 audits were performed by a consultant (Fred Lobbin), by Sargent and Lundy (using INPO criteria), and by TUGC0 (using INPO criteria).
Each of these audits evaluated limited aspects of the QA program.
In 1983 Cygna evaluated the design program.
The Lobbin Report (February 4, 1982) R-82-01, contained four major findings:
level of experience within the TUGC0 QA organization is low; i.e., commercial nuclear plant design and construction QA experience; staffing for the audit and surveillance functions is inadequate; the number and scope of design and construction audits conducted by TUGC0 QA to date has been limited; and QA management has not defined clearly the objectives for the surveillance. program resulting in a program which, in the author's (Lebbin) opinion "is presently ineffective."
The TUGC0 QA manager responded to these findings in an office memorandum (QSC-18), dated February 23, 1982. This response basically concurred with these findings.
The respense committed to recruit nuclear experienced individuals, to increase the number and scope of site audits, and to scre effectively use the surveillance program. Two program reports (QBC-25 and 29) regarding these matters were issued from the QA manager to the VP nuclear operations on May 21 and August 31, 1982, respectively.
Following the Lobbin Report, the NRC performed a CAT inspection (IR 445/83-18; 446/83-12 dated April 11,1983) and included a review of the TUGC0 audit program at the corporate offices. The inspection included a review of 18 audits (conducted between 1978 and early 1983), auditor qualifications, audit planning and scheduling, audit reporting and followup, and audit program ef fectiveness. The report concluded that weaknesses existed jn the established QA audit program and included the scheduling and frequency of audits, the lack of effective monitoring of the construction program, and the lack of effective resolution of certain audit findings. The inspection also indicated that the QA program should have been more effective.
o.
I Based on the findings in the Lobbin report, and the findings in the NRC CAT report, the QA program continues to exhibit weaknesses.
The continuing weaknesses in the QA program over a significant period of time reinforce the need for the applicant to routinely assess the status and adequacy of the QA program routinely to ensure that the areas are identified and adequate and timely corrective action is taken to correct the QA program weaknesses.
The failure to regularly review the status and adequacy of the QA program as required is a violation of Criterion II of Appendix B to 10 CFR 50.
(445/8432-02; 446/8411-02) d.
Management of the TUGC0 Audit Program (1) Program Requirements FSAR Subsections 17.1.2, "QA Program," and 17.1.18, " Audits," require internal audits of (TUGC0 corporate and site activities) and external audits (prime contractors, subcontractors and vendors) to evaluate the effectiveness of the QA program by verifying conformance with design requirements; compliance with established requirements, methods and procedures; and implementation of corrective action.
These commitments require the establishment and implementation of a comprehensive system of planned and periodic audits of all aspects of the QA program.
The TUGC0 audit program consisted of internal and external audits of design, construction, engineering, and procurement activities.
TUGC0 also retained responsibility for the external audits that were usually delegated to the AE and NSSS organizations; i.e., audit of vendors.
In addition to construction and vendor audits, the TUGC0 audit group was also responsible for performing preoperational/startup and plant operation audits.
TUGC0 committed to the audit requirements of ANSI N45.2.12-1973, Draft 3, Revision 0, Section 3, " Audit System," and these program management objectives are:
to determine that a QA program has been developed and documented in accordance with applicable requirements; to verify that the program has been implemented, to assess program effectiveness; 4
to identify program nonconformance; and to verify program correction where appropriate.
?.
. This section al'so stated that to achieve these ANSI standard objectives full management backing, manpower, funding, and facilities shall be available to implement the system of audits.
(2) NRC Evaluation of Plannina/ Implementation of Program The NRC inspector reviewed and evaluated the applicant's plans, procedures, and number of audits performed (see paragraph 2e below) and determined that planning was inadequate. This audit effort was too large for the four available TUGC0 auditors in 1981, even though additional specialists were utilized to assist with the audit activities.
(a) The inspector' reviewed and evaluated planning documents (formal and informal) used by the TUGC0 QA manager, supervisor QA services, and supervisor QA audits. The review and discussions with these individuals revealed that annual audit plans were based on the audit of organizations rather than activities.
TUGC0 Audit Procedure DQP-CS-4, Revision 0, dated August 9, 1978 required:
semiannual internal audits, semiannual construction audits, annual AE audits, annual NSSS audits, and annual plant operation audits.
However, for vendor audits the procedure required:
first audit at 15 percent; and second audit at 60 percent
" item completion" by reactor coolant pressure boundary vendors; one audit of engineered safeguards vendors at 25 percent ites completion; and audit of balance of plant (other safety-related) vendors as determined by the manager QA.
This does not meet the requirements of paragraphs 3.4.1 and 3.4.2 " Scheduling," of ANSI N45.2.12 which requires, " Auditing be initiated as early in the life of the activity as practicable... applicable elopents of the QA program shall be audited at least annually or at least once within the life of the activity whichever is shorter."
l I
i
. Furthermore, Audit Procedure DQP-CS-4, Revision 2, April 16, 1981, and Revision 10, June 4, 1984, have further reduced the (scheduling) frequency of audits.
Revision 10 now states, in part, "3.2.1, The following organizations will be audited on a regularly scheduled basis but in accordance with Regulatory Guide (RG) 1.33, Revision 2, January 1978, Regulatory Position 4:
- a. AE; b. NSSS; c. constructor; d. TUGC0 Internal;
- e. Preoperational/Startup; f. Plant Operations;
- g. Subcontractor...
3.2.1 In lieu of regularly scheduled audits of vendors TUGC0 QA will perform the following:
- a. Monitor the individual vendor ratings which are based on vendor performance... b. for those vendors who cannot be evaluated based on vendor ratings... regularly scheduled audits will be performed based on level of activity." The NRC inspector discussed with TUGC0 management the fact that RG 1.33 is for operations and does not fully address the requirements of the construction QA program.
4 i
This failure to develop audit program procedures which adequately address and describe QA program requirements and commitments is a violation of Appendix B, 10 CFR Part 50, Criterion XVIII (445/8432-03a; 446/8411-03a).
(b) In addition to evaluating to determine if annual audits were planned, the NRC inspector requested objective evidence which would demonstrate that planning for audits for calendar years 1983 and 1984 included a method to verify compliance with all aspects of the QA program and to determine the effectiveness of the QA program. The review of the objective evidence revealed that the planning was not adequate, particularly regarding the audit. basis, status, and tracking.
The only objective evidence available consisted of a listing of planned audits of internal organizations and contractors each year and a summary of 1983 audit results and criteria audited; however, this data in many cases did not list the criteria audited and while reviewing older audits it was noted that an "after the fact" review resulted in identifying the applicable criteria covered for various organizations.
The inspector requested a listing of selected site procedures which were in effect in 1983 that were representative of site
' safety-related activities and subject to audit by TUGC0 corporate QA. The review of the listings provided and the 1983 i
audits revealed the following information:
4 l
l r
m.
. Audits of Total Procedures
% Audited Procedures Procedures Audited / Referenced in 1983 TUGC0 Quality Documents Index (December 20, 1983) 295 71 24 TUSI Engineering Instruction Index (December 2, 1983) 65 16 25 TUSI Nuclear Engineering Procedures / Instructions Index (September 26, 1983) 26 18 69 TUSI Engineering Procedures Index (November 4, 1983) 30 12 40 B&R Quality Document Index (November 22, 1983) 51 20 39 B&R Construction Procedures Index (June 20, 1983) 189 28 15 Total 656 165 25 Only 25 percent of the procedures (specific safety-related activities) were audited in 1983. Although audits on a sampling basis are acceptable, there was no evidence that all safety-related areas were audited.
The audits did not encompass all aspects of the QA program in order to determine effectiveness.
The failure to properly plan or produce evidence of adequate planning for a comprehensive audit program to verify compliance with all aspects of the QA program resulted in the failure to audit significant parts of the QA program is a violation of Criterion XVIII of Appendix B to 10 CFR 50 (445/8432-03b; 446/8411-03b).
The NRC inspector contacted the Westinghous {W1 Site manager to review the procedure listing for safety-related activities which TUGC0 had audited. As indicated below, no audits of NSSS site activities were performed in 1983. Discussions with the QW site manager revealed that no audits had been performed by IUGC0 QA in 1977, 1978, 1979, 1980, or 1981.
This was discussed with the TUGC0 audit staff and QA manager who did not disagree with the stated audit frequency.
1 i QW Site Organization hternal Total Procedures
% Audited Procedures Procedures Audited / Referenced in 1983 Westinghouse (W) Site Applicable Procedure, QA Manual, May 1983 18 PPD Procedures 14 Installation Procedures 29 1 The failure to audit LW), procedures (safety-related activities) l annually as required by ANSI N45.2.12, Oraft 3, Revision 0, of the QA program is a violation of Criterion XVIII of Appendix B to 10 CFR Part 50, (445/8432-03c; 446/8411-03c).
(c) The NRC inspector discussed The staffing of the Audit Program with TUGC0 QA management the findings of the Lobbin Report and the NRC CAT Team Report regarding the staffing of the audit functions. The discussions revealed that the TUGC0 audit staff had been increased from 4 to the present number of 12 between 4
1982 and 1984, and TUGC0 management has been looking for 3 or i
4 additional nuclear experienced auditors to further increase the j
audit staff. However, it was also revealed that management had not determined the total audits required nor the manpower needed 4
I to accomplish the audits.
j This matter is an unresolved ites pending the determination of the number of audits and auditors that will be needed to effectively implement the audit program (445/8432-04; 446/8411-04).
{
(d) The NRC inspector determined through review of charts and lj procedures that current organization provided organizational freedom from cost and schedule.
l.
(e) The NRC inspector evaluated audit personnel qualifications by j
reviewing 14 personnel files of lead auditors and auditors.
This included presently employed and formerly employed auditors.
i l
These personnel were qualified as required by TUGC0 Procedure DQI-QA-2.1, Revision 7, and ANSI N45.2.23-1978,
" Qualification of Quality Program Audit Personnel for Nuclear l
Power Plants."
l (f) The NRC inspectors reviewed TUGC0 Audit Procedures DQP-CS-4, Revision 10(June 4,1984),andpQI-CS-4.6, Revision 7 i
l (April 13,1984). As previously discussed in paragraph 2.C(1),
DQP-CS-4 does not include adequate commitments to perform annual l
audits and failed to address both design and construction and plant operations audit requirements.
I f
i 4
1 e.
Implementation of the TUGC0 Audit Program The NRC inspectors selected three areas of the audit program to review and evaluate implementation.
Results of this evaluation are discussed in the following paragraphs.
(1) Internal Audits of Site Activities - The NRC inspector reviewed the index which showed all site audits and found that Audits TCP-1 through TCP-112 had been performed between March 1978 and August 1984. The number per year are:
(1) 4 in 1978; (2) 3 in 1979; (3) 10 in 1980; (4) 11 in 1981; (5) 30 in 1982; (6) 29 in 1983; and (7) 22 during the first 8 months of 1984.
After the audit program was found inadequate in the consultant's report (Lobbin), the number of audits increased from less than 1.0 per month in 1982 to 2.5 per month in 1982. After the NRC CAT inspection report in 1983 this 4
number increased to 2.7 per month for the first 8 months of 1984.. This indicates that positive action concerning these reported weaknesses was taken; however, as previously discussed objective evidence was not available that the required number of audits and auditors has'been identified.
This item was previously identified above as unresolved.
The 1983 and 1984 audit schedule included each_ audit scheduled, cancelled, and any additional audits planned or performed.
Where audits were cancelled, they were rescheduled and other audits were added and performed.
This effort was well documented.
In 1983 the TUGC0 audit group performed 158 audits.
Sixty-five internal audits of site activities are as follows:
construction /QC/ engineering - 33 audits; startup - 5 audits; and oparations - 27 audits.
The NRC inspector selected and reviewed 31 TCP 1983 audits of site activities. The audit files included notification to the organization audited, an audit plan, checklists, an audit report, audit response, and evaluation / closeout of findings.
Audit reports reflected good preparation and execution.
Substantial findings generally resulted and were resolved.
Several lead auditors were interviewed concerning the management of the TUGC0 audit program. They stated that the audit program had weaknesses or deficiencies in 1978 but they had witnessed dramatic improvements and were confident that the audit program was currently working well.
t
. (2) Assurance of Design Control - TUGC0 management verified that design was controlled in accordance with the QA program requirements and procedures through administering an effective audit program. The design control functions were delegated to the AE ahd IWl; however, TUGC0 was designated the engineering organization responsibility for plant design.
The NRC inspector reviewed and evaluated the results documented in 15 TUGC0 internal and external audit reports which specifically relate to Criterion III of 10 CFR Part 50, Appendix B, design and applicable procedures. These represent all audits design and consisted of 8 audits of TUGCO, 3 of (W),
and 4 of G&H, engineering organizations. All audit findings, concerns, and deficiencies were closed through correspondence and were later verified through subsequent audits.
Management involvement was evident as the VP nuclear operations was on concurrence and was furnished status reports by the QA manager.
In October 1982, TUGC0 initiated a special audit effort to review design using the Institute of Nuclear Power Operations (INPO) performance objectives and criteria.
Sargent & Lundy personnel were used to perform this audit. This audit identified 13 findings and TUGC0 audit No. TNO-2, dated June 1983, verified corrective action.
(3) Assurance Control of Procurement Activities - TUGC0 management elected to retain procurement responsibilities except for certain functions delegated to the AE and NSSS. The NRC inspector selected several functions retained by TUGC0 to determine if their audit program effectively monitored or verified that procurement activities were accomplished in accordance with the QA program and applicable procurement procedures. Management involvement with procurement documents, bid / source evaluation, and specific QA inputs were reviewed by the inspector. The vendor audits and evaluation of vendors were a large work effort. The following are the results of this review and evaluation.
The NRC Comanche Peak Special Review Team Report dated July 13, 1984, at the site identified a potential violation, i.e.,
failure to perform annual audits of vendors. The report documented an inspection of the procurement effort at site and part of this inspection included determining the frequen y of vendor audits. As a result of the special inspection, the TUGC0 QA manager approved an FSAR change request, dated August 3, 1984, which asked that TUGC0 be allowed to adopt NRC RG 1.144 audit requirements in lieu of ANSI N45.2.12. Oraft 3, Revision 0, for construction and ANSI N45.2.12, Oraft 4, Revision 2 for operations. This requested change would not change the requirement to perfore internal audits annually but i
8 t
t 13-would reduce the requirement to perform annual audits of suppliers. Considering this requested QA program change which had not been approved by the NRC, the following are the l
inspection results:
(a) The NRC inspector reviewed the TUGC0 vendor audit program for 1983 to determine compliance with commitments (FSAR Section 17, paragraph 17.1.18), ANSI N45.2.12 and TUGC0 l
1 procedures DQP-CS-4 and DQI-CS-4.5.
l I
i The annual audit schedule revealed that 60 vendor audits l
were scheduled during 1983. Audit TCLC-2 was cancelled (lack of activity with Purchase Order CPC-307) and i
audit TBS-3 was rescheduled (delayed by 1 week) as a result of NRC CAT Team inspection findings.
The NRC inspector l
selected 3 vendor audit files, TVO-1, T)M-3, and TBF-2, for review to determine the extent of the audits as applicable to the audit plan checklist, noted deficiencies, concerns, j
and comments. Also included in this review were the i
corrective actions and/or preventive action documented in j
writing by the vendor in response to the applicable audit i
findings. Documents in file closed the audit findings and indicated that followup on corrective action would be j
verified during the next audit.
The NRC inspector reviewed the vendor audit frequency to determine if TUGC0 established a schedule to annually audit vendors. The licensee commitment to ANSI N45.2.12, Draft 3, Revision 0, requires annual audits or at least once within the life of the activity. Neither procedural requirements were established nor were vendors audited i
annually.
i i
The failure to establish procedural requirements and to l
perform annual vendor audits is a violation of I
Criterion XVIII of 10 CFR Part 50, Appendix B and ANSI N45.2.12, Draft 3, Revision 0 (445/8432-03d; j
446/8411-03d).
(b) The NRC inspector reviewed the approved vendors list (AVL) program for 1983 to verify that methods used by TUGC0 to i
qualify vendors to supply safety-related materials, parts, i
j and services were consistent with the QA plan, procedural l
requirements, and commitments described in ANSI N45.2.13-1976. A review of supplemental menos and preaward survey files and revisions 9 through 12 of the AVL verified that the AVL was current.
This review showed i
33 additions, 40 status changes, and 1 deletion to the AVL for the period January 24, 1983, through December 20, 1983.
The preaward survey files reviewed were consistent with i
i
~
i
- I
{
procedures DQP-CS-4, Revision 10, and DQI-CS-4.2, A
i Revision 3, December 1, 1982.
During the review of l
preaward survey files, the inspector confirmed that formal l
l identification letters, the survey date, and the scope of the survey (checklist) were consistent with the vendor QA
]l program. Also, the corrective action responses by the suppifer concerning noted deficiencies, concerns, and l
comments were reviewed, and followup action verified in a subsequent audit.
i (c) The NRC inspector reviewed the vendor performance j
evaluation (VPE) system to determine compliance with commitment and procedural requirements. TUGC0 Procedure DQP-CS-4.3, paragraph 1.1 stated that the purpose of the l
evaluation was to establish a comprehensive method of identifying system weaknesses in vendor QA programs through 1
acceptable / unacceptable hardware information generated as a i
result of vendor release inspections. The VPE files included release inspection trip report cover sheets, vendor rating sheets, releases, and the inspection checklists as required by TUGC0 Procedure DQI-CS-4.3, Revision 4, paragraph 3.1.
The NRC inspector reviewed 3 VPE packages to determine that the quality assurance services (QAS) group's review was l
consistent with procedural requirements. One vendor file (Paul Monroe Hydraulic) was still active pending engineering review and evaluation on the 0-ring discrepancy identified during release inspection at Remo Hydraulics (Purchase Order CPF-11436-5 issued to Paul Monroe l
Hydraulics) for 20 hydraulic snubber assemblies. As l
required by DQP-VC-3, one vendor package (Meddco Metals) l was being held on a yellow flag sheet to alert TUGC0 auditors of next request for release so that TUGC0 auditors could accompany the TUGC0 vendor compliance inspector to l
resurvey the vendor. One other vendor (Volumetrics) i performance evaluation record was reviewed and it showed a vendor rating of greater than 90. The NRC inspector j
i interviewed the QA audit supervisor to determine what j
objective evidence (as required by referenced TUGC0 Procedure DQI-CS-4.3, paragraph 3.2) was used to perform-j the vendor evaluation and support vendor ratings. Preaward surveys, previous audits, and receiving inspection reports i
J were used as objective evidence to give the rating.
i The NRC inspector reviewed the receiving inspection activity for previous relea,se inspection shipments relative to the aforementioned vendors.
Receipt inspection consisted of shipping damage inspection, receipt of 3
documentation, identification, and quality assurance release.
I l
....c...
,_m
_ _ _ _ _ _ _ _, _ _. _,.. _. - _. ~,. _.,... _, _ _,,., _. _ _.__
.._-,.,_.,_...g_..
~
e, 1
I 1' (d) The NRC inspector reviewed the method by which the licensee i
performed source selection to determine that procedural j
requirements were met. QA plan Section 4.0, Revision 4, 3
July 31,,1984, required that a purchase order for i
safety-related items not be issued to a vendor unless TUGC0 l
QA had reviewed and accepted the purchase order; i.e., QA i
determines whether QA provisions are adequate and determines that a preaward evaluation recommends selection i
of the vendor, i
When procurement solicited bids outside the AVL, TUGC0 QA i
requested that an uncontrolled copy of the vendors quality
)
assurance manual be sent with the bid response. In.the i
event of a positive bid response from the unapproved supplier, the TUGC0 procurement group fomards the QAM and I
a request for QA program evaluation, Form QA-VE, to the TUGC0 QA audit group supervisor to initiate a preaward survey per QA Procedure DQT-CS-4.4, paragraph 3.1.
However, until the preaward survey is completed and a supplemental memo has been issued by the audit group supervisor, no further procurement action was taken.
j The NRC inspector reviewed the actions taken when an l
acceptable bidder takes exceptions to the purchase order or j
subcontract. Upon receipt of the exception, procurement i
filled out an expediting request, assigns a procurement log i
number, and forwarded this request to the field requisition originator for engineering review and evaluation.
Should 4
l the engineering group allow the exception, the necessary actions; i.e., design changes, were initiated.
The a
expediting request was returned to procurement accoepanied by a field requisition documenting the change with the j
approval signatures of engineering and QA.
j' (e) The NRC inspector reviewed the method by which TUGC0 performed vendor item acceptance of safety-related i
}
materials, parts, and components.
TUGC0 l
Procedure OQP-VC-1, Revision 8, June 4, 1984, l
l paragraph 1.1, specified that the purpose was to establish,
guidelines for performing final inspection and release of' TUGC0 purchased equipment and applies to both i
safety-related and nonsafety-related equipment. This f procedure allowed for a waiver, in which case the.
l inspection checklist applicable to the procurement specification became the responsibility of CPSES receiving 3
1 inspection as described in B&R CPSES Procedure CP-QAP-8.1, j
Revision 8. June 11, 1984, paragraph 3.4.1 (f) The NRC inspector reviewed six vendor compliance inspector's files to determine if training / certification i
e
i 1
l 1
i ]
j records met the requirements of ANSI N45.2.6-1978 and TUGC0 Procedure DQP-VC-4, " Guidelines for Certifying Vendor Compliance Personnel." Section 3.2.2 states that a i
Level II inspector shall attend and satisfactorily complete the nondestructive examination (NDE) courses. One '
inspector had not completed all of the NDE courses but had
{
been certified.
This finding was discussed with the vendor J
compliance supervisor who stated that there is no real need j
for certification in eddy current testing since inspectors i
do not utilize this NOE technique and the requirements
}
would therefore be deleted from the procedure. The NRC r
inspector verified the deletion of this requirement and j
procedural revision during this inspection.
The failure to certify the inspector in accordance with the i
procedure is a violation of Criterion V of Appendix 8 to i
10 CFR 50 (445/8432-05; 446/8411-05).
[
No other violations or deviations were identified.
1 i
3.
TUGC0 Corporate QA - Site QA Activities Interface l
i j
Appendix 8 to 10 CFR Part 50 requires TUGC0 to establish proper i
organizational and management interfaces, and procedures must describe how various organizations coordinate and comusunicate design, procurement, engineering, construction, and QA/ control activities and information.
The
{
following paragraphs describe inspection of this requirement.
a.
Site Oraanization i
i
.TUGC0 Procedure CP-QP-3.0, Revision 15, July 30, 1984, described the site QA organization for design and construction. This organization consisted of a site QA manager, QA supervisor, and a QC supervisor.
The site group performed no audit function, however, they did perform t
QA surveillances. The site group consisted of 13 QA/QC managers and more than 150 lead /QC inspectors and quality engineers. These personnel inspected non-ASME work.
\\
j B&R QA manual and ' implementing Procedure CP-QAP-03.01, Revision 6, described their responsibilities for QA/QC and construction i
l activities pertaining to ASME work. This organization consisted of a j-QA manager. QE supervisor, and a QC supervisor. The total QA/QC work
(
force involved with design / construction activities was approximately i
100.
t
{
Several other site subcontractors such as Bahnson, Brand Industrial j
Services, Inc., and Chicago Bridge angl Iron, have small QA groups on l
}
site and, as is the case with B&R, these organizations were audited j
by their respective corporate offices.
i 1
.I
Ob
- t
. 1 The NRC inspector interviewed the TUGC0 site QA manager to determine how the site QA group interfaced with the corporate QA office.
He stated that daily conversations occur between managers of these organizations, however, he did not make written summary reports.
Quarterly trending reports which analyze reported nonconformances and deficiencies are sent to the corporate QA manager.
b.
Site Surveillances The NRC inspector noted that surveillances were briefly mentioned in TUGC0 Procedure DQP-CS-4, Revision 10; however, there was no mention of how or if the surveillances would be used to complement the audit program. During discussions with 'the QA manager and other personnel, it was revealed that procedures were not tracked to assure that all were audited. The present audit staff could not audit all site procedures annually. The NRC inspector pointed out that the surveillance function may complement and be used to (1) check that all procedures,are implemented; (2) identify nonconforming trends; and (3) to feed potentially deficient or weak areas to the audit group which could, in turn, factor this information into the audit program. Audit priorities could then be established and the audit personnel could be more effectively used.
TUGC0 Surveillance Procedures CP-QP-11.2, 19.3, 19.4, 19.5, 19.6, 19.7, 20.0, and 27.0 described the surveillances of specific activities; however, no general procedure which describes the overall surveillance program was provided.
The present program did not appear to have sufficient purpose, direction, coordination, and feedback in relationship with the overall QA program.
Furthermore, the inspection revealed that the surveillance staff had been reduced from a supervisor and eight technical personnel to four technical personnel.
Considering the Lobbin Report this reduction of surveillance effort may not be a prudent action.
As noted in the findings in the'Lobbin Report; i.e., QA management had not clearly defined the objectives and scope of the surveillance i
program, it appeared that TUGC0 needed to strengthen the surveillance program. The TUGC0 management decision to commit to a surveillance l
program was a strength, but this lack of purpose and direction and support was a program weakness.
Additionally, the surveillance group was no longer observing work in Unit 1 but will now place most of their effort on Unit 2 construction activities.
~ This matter is considered unresolved pending clarification of the audit and surveillance program efforteand further review during a subsequent inspection (445/8432-06; 446/8411-06).
1
- ~,, _,
..-.,-.,y,<m-4-,~..,,..-e,,,--.,,.,-----w~
s l The NRC inspector randomly selected and reviewed 28 surveillances performed in 1982, 1983, and 1984.
Findings and resolutions of these findings were reviewed and in each case, written responses and corrective action were adequate.
c.
Site Desian Activities The NRC inspector reviewed and evaluated selected site activities pertaining to design verifications, design changes, design inputs, and control of vendor drawings as follows:
(1) Design Verification - The NRC inspector interviewed the TUGC0 supervisor of engineering, support, and other engineering personnel to determine how design verifications were performed, and examined the related procedures, logs, and design verification packages. Authorized design verifiers were maintained on lists and an automated tracking system was in place to assure that all design changes, i.e., design change authorizations / component modification cards (DCA/ CMC) were verified. Three design verification reports were reviewed to assure that the design verifier was on the authorized list.
Design verifiers were not to be involved in the original design review to assure an independence.
It was noted that each DCA/ CMC was being reviewed for verification.
If there was no authorized signoff, then the design was verified.
Audit TGH-23, conducted during August 1984, concentrated on Unit 1 quality related activities for which onsite G&H design review team had responsibility.
The audit involved evaluation of the program established and implemented for site review and processing of changes (CMA and DCC) associated calculations and 287 design review packages were reviewed.
No major technical problems were identified during this audit.'
(2) Design Changes - The NRC inspector interviewed engineers and draftsmen in TUGC0 engineering to determine how design changes were processed and examined the related procedures, files, reports, and tracking systems.
A master list was maintained identifying those individuals who were authorized to approve design changes and G&H updates this list by memo.
The NRC review of three design review files verified that the reviewers were on the authorized list.
The NRC inspector also reviewed the method used to incorporate field changes (OCA/ CMC) into related drawings and the subsequent review, approval, and incorporation of changes into as-built drawings. One observation requi, red additional discussions. The drafting supervisor's (piping support) authority to incorporate a change into a drawing was transmitted and signed by a clerk.
This was clarified as being acceptable by management because it
1 19-i was in accordance with established procedure (CP-EI 4.6-8,
)
paragraph 3.3) and also, as a final control, the as-built drawing was reviewed and approved by an authorized project engineer prior to release.
The NRC inspector examined how the TUGC0 administrative services group handled NRC IE Bulletins, Circulars, and Information Notices. These documents were coordinated by the operations support department and were distributed to the appropriate TUGC0 engineering group for action.
Design changes resulting from these inputs were processed in accordance with established design control procedures.
Responses from personnel receiving these reports were reviewed to verify that the reports were adequately addressed.
Summary reports and log sheets'are used to keep management current as to the status of the responses..
An INPO audit of the operating experience review program in 1982 noted the following good practice, "The procedures for handling industry experience are excellent and are expected to provide a firm base for developing an effective industry experience program."
TUGC0 QA audit Report TUG-41 was conducted in December 1983 to review implementation of the operations support program for evaluating and responding to NRC IE Bulletins, IE Notices, IE Circulars, and generic letters.
The auditors found the program in compliance with procedural requirements and the overall effectiveness of the program appeared to be adequate.
(3) Design Document Control - Two packages were reviewed and these contained evidence of vendor data checklists, indexes, approval letters, and the vendor stamp on drawings was observed.
d.
Site Precurement' Activities The NRC inspector determined that the TUGC0 procurement function was delegated to the TUGC0 site organization. The major procurements occurred several years ago; however, present procurement activities associated with items procured offsite for installation were performed by TUGC0 or were contracted to G&H, {Wl, or B&R who were evaluated and qualified by TUGC0 QA.
Procurement documents were reviewed, approved, and controlled; and receipt inspection of safety-related items on site was performed in accordance with written procedures and checklists.
The NRC inspector selected two procurement actions for review:
P.O. CPF-1233-S issued to Combustion-Engineering for the 1
procurement of a heated junction thermocouple system.
CPF-10469-S issued to Paul Monroe Hydraulics to refurbish four Rockwell International actuators.
[
i
-.n..
y n.
-,., -,. ~.. - - - -
..,,, -,