ML20117C188
| ML20117C188 | |
| Person / Time | |
|---|---|
| Site: | Waterford  |
| Issue date: | 04/30/1985 |
| From: | Barnes B EG&G, INC. |
| To: | NRC |
| References | |
| CON-FIN-A-6819 EGG-RST-6838, NUDOCS 8505090352 | |
| Download: ML20117C188 (128) | |
Text
{{#Wiki_filter:. ' * (Ok EGG-RST-6838 April 1985 WATERFORD III PROJECT 1 REPORT (4- 5 8 A Compiled by B. L. Barnes Idaho National Engineering Laboratory Informal Report Operated by the U.S. Department of Energy
;,:ys 'j?.W ;i+ W!%i'M,-G' jn_ ";. ~ f2; ,
Ms . ,* 1\ . Ifd . g gg **- -O schh$ h;t3.h.['? ~s Q w' W 5+Nr;pGry.rgy, Mg ; bkhlWC5=O
~..7 & J: _ a"s sM Y Ah.T_$__{MF W N E W -*r~r CWi .
m-mc - -; & W - + - m-
'orJunul W ~ ~ ~ fZ - [ 7 @ y (u-{a. M+ 'WRiM7 -
- g. =w .,
3r, .
.=:m . . -
! j
~ ' ^
'- ~%
% N N-
[ .Nsh-.-
'; -~-. - , .
W %t -n
,- -. 43 W ~Q h 1
Prepared for U. S. NilCLEAR REGULATORY COMMISSION Under DOE Contract No. DE-AC07-76ID01570 H N N . A6819 8505090352 850430 PDR ADOCK 05000382
) g g g g daho P PDR
EGG-RST-6838 WATERFORD III PROJECT 1 REPORT b l l Compiled by Bert L. Barnes l 4 l' 1 ! Published April 1985 l EG&G Idaho, Inc. Idaho Falls, Idaho 83415 l l-l- . l Prepared for the l-U.S. Nuclear Regulatory Commission l
- Under DOE Contract No. DE-AC07-761001570 t FIN No. A6819 t
l 1 l L_ _ -
ABSTRACT Eleven EG&G Idaho personnel participated as members of a 19 person NRC Quality Assurance Records Review Team at the Waterford III Nuclear Plant site. The team members reviewed 217 allegations pertaining to the overall quality assurance program at Waterford III, interviewed allegers to clarify these allegations, and made assessments of allegations. Quality assurance . .. related allegations are listed herein along with a description of the methods used to assess and resolve these allegations. 9 4 e 11
CONTENTS
- 1. INTRODUCTION .................................................... 1
- 2. WATERFORD III TASK FORCE DESCRIPTION ............................ 2
- 3. PROCEDURE EMPLOYED TO RESOLVE ALLEGATIONS ....................... 2
- 4. QUALITY ASSURANCE RECORDS REVIEW TEAM
SUMMARY
OF RESULTS ........ 4
- 5. RESULTS ......................................................... 8
- 6. QUALITY ASSURANCE ALLEGATION LISTING ............................ 9
- 7. ALLEGATIONS ..................................................... 12
- 8. CONCLUSIONS ..................................................... 127 s
W e m
WATERFORD III PROJECT 1 REPORT INTRODUCTION This report was prepared to document the primary activities of the eleven EG&G Idaho personnel listed below who participated as consultants to the NRC to help resolve quality assurance and quality control allegations-for the Waterford III Nuclear Power Plant owned by Louisiana Power and Light (LPL). EG&G Idaho Members of the Quality Assurance Records Review Team V. E. Berg D. L. Ross J. W. Farren, Jr. J. D. Stearns
. B. L. Freed D. L. Summers F. W. Ingram V. L. Wenczel
. T. L. Morgan J. E. Whittemore C. L. Nalezny It is important to note that the work contained herein was not all done by EG&G personnel; a great deal of this work was done by the NRC staff
] i and.other consultants. However, 11 of the 19 members of the Quality I Assurance Records Review Team were EG&G personnel. y Because of the controversial nature of all the Waterford allegations, it was agreed that Task Force Team members would not remove notes and related paper work from the plant site until which time the written words ] 6 had been edited, finalized, and signed off as acceptable by all Task Force j personnel. .For this reason this report was of necessity assembled after l .- the fact using the Safety Evaluation Report (SER), Supplement 7 (NUREG-0787). The majority of this report was directly taken from this SER )- since no other written record was-brought back to EG&G Idaho by the EG&G team members. 1 k
WATERFORD 3 TASK FORCE DESCRIPTION To achieve the objectives of the Program, a Waterford 3 Task Force was organized that was composed of about 40 technical specialists from NRC Headquarters and Regional Offices, and included several NRC consultants including EG&G Idaho personnel. The Task Force was divided into teams that looked into the areas of: (1) the civil and structural and mechanical and piping disciplines; (2) instrumentation and control; and (3) quality assurance and quality control. This report deals only with item (3), quality assurance (QA) and quality control (QC). A fourth group, the Inquiry Team, was included in the Task Force. The Inquiry Team had previously investigated concerns raised in May 1983 about LP&L's Quality Assurance (QA) Program. The Inquiry Team's report (July 14, 1983) identified specific QA concerns and recommended followup actions by Louisiana Power and Light (LP&L). LP&L responded in letters of September 29, 1983, and February 20, 1984, to the Director, NRC Office of Inspection and Enforcement (IE). The Inquiry Team worked onsite with the Task Force to evaluate the adequacy of LP&L's response. The Inquiry Team's findings are provided in Inspection Report 50-382/84-34 issued on July 20, 1984. PROCEDURE EMPLOYED TO RESOLVE ALLEGATIONS On April 2, 1984, the Task Force began a six-week onsite effort to determine the validity of the allegations, to evaluate their safety significance, and to assess their generic implications. The Task Force began by preparing action plan task sheets for each l allegation that characterized it and outlined an approach to resolve the
., allegation. In some cases, when a number of individual allegations were
! similar in subject, they were combined and evaluated together. In evaluating the allegations, the Task Force utilized various project 2
documents, such as specifications, drawings, procedures, and instructions.
- . This effort entailed the review of documents that were in place at the time of the allegations, as well as any subsequent revisions. Construction records for the various structural materials, including construction work packages, personnel qualifications records, inspection records, surveillance and audit reports, and other relevant documents, were also reviewed. Additionally, the Task Force gathered information from representatives of LP&L, from E8ASCO (the architect-engineering firm) and from their subcontractors. The Task Force also examined areas in the plant where only direct observation could provide needed information. And, as mentioned earlier, Task Force members met with some of the allegers to clarify allegations and discuss the status of evaluations. ,
Finally, the Task Force evaluated the facts gathered, the documentation examined, and the discussions held to determine the validity l ,, of each allegation. The safety significance was then determined, any generic implications were defined, and indications of potential management breakdown were noted. Where LP&L action was needed, the necessary action was requested. This report contains the safety evaluation of QA and QC related allegations or groups of allegations reviewed and resolved in part by EG&G Idaho personnel. Each evaluation includes: (1) the allegation number; (2) a brief statement in plain English describing the allegation; (3) a summary of the Task Force evaluation process; and (4) as appropriate, action required of LP&L. The evaluation section includes, as appropriate the safety significance of the findings with regard to the design, construction, operations, QA, and/or management controls; assessment of the generic implications of the findings; the root cause of the situation; and any management concerns identified. The Task Force pursued its evaluation of allegations and their safety significance to wherever the information led. In some cases, both EG&G personnel and the NRC staff looked into an allegation and found it to be of no safety significance, but discovered a new concern unrelated to the 3
. original allegations. These new concerns were also carefully assessed. In all, the Task Force identified 23 areas of potential safety concern needing resolution prior to the NRC staff's making its safety determination and licensing decision. In a letter to LP&L of June 13, 1984 LP&L was requested to propose a program and schedule for a detailed and thorough assessment of staff concerns identified by the Task Force. The program plan would include and address the root cause of each of the 23 potential concerns identified; the generic implications on other safety related systems, programs, or areas; and the collective significance of the deficiencies. The program plan should include the action proposed by LP&L to assure that such problems will be precluded from recurring in the future. Louisiana Power & Light responded to the staff's request by letters of June 28, 1984 and July 27, 1984, with a proposed program plan.
The NRC staff will complete its review of LP&L',s response to the
.. individual issues prior to making a licensing decision. The results of - that review will be published in a supplemental Safety Evaluation Report.
In most cases, the Task Force found no generic implications for-problems on safety-related systems. However, the NRC staff and other Task Force members recognized that some allegations had potential generic implications beyond the scope of the original allegation. As noted above, the Task Force identified 23 areas with potential safety significance, some of which had possible generic implications. The NRC requested, in a letter of June 13, 1984, that LP&L evaluate the generic implications, as well as the potential safety significance, for these areas. OVALITY ASSURANCE RECORDS REVIEW TEAM
SUMMARY
OF RESULTS The NRC Quality Assurance (QA) Records Review Team, including EG&G Idaho personnel, reviewed 217 allegations pertaining to the overall QA program and its implementation. After assessing each allegation on its own
-* merits and for its generic implications, the team grouped the allegations into 37 subject categories. These categories included allegations about unqualified QA inspectors, records reviewers, and welders; inadequate and 4
falsified QA documentation; improper disposition of nonconformances; ,, drawings that did not reflect as-built conditions; lack of material traceability; and improper changes to ASME Code dat'a reports and record packages. , f In evaluating the allegations, the team utilized both random and biased sampling techniques. Random samples were taken fromI QA reccrds, logs, inspector and welder listings, and other documentation, whlie biased samples came from information obtained from the allegers and from trends , found during the team review. , The QA Records Review Team was made up of 19 members, 1! of whom were from EG&G Idaho. This Team, collectively, represented approximately 300 years of experience in QA record reviews and audits, ASME Code interpretation, nuclear power plant inspection, testing and evaluation of materials, reactor construction and operation, instrumentation, and other pertinent disciplines. , The QA Records Team found that half of the allegations tley reviewed were substantiated. Of the 37 categories reviewed, the team found eight areas that indicate a breakdown of the overall QA program thit' have potential safety significance and generic implications. ;, The first of these areas involv'ed unqualified or incorrectly certified QA/QC inspection personnel. Approximately 35'/. of the inspectars for two site contractors were not qualified to be certified. The team found that background checks had not been made. The issue of a few potentially falsified resumes is under review by the NRC.Cffice of Investigation:-(0I) , (see Allegations A-01, A-02/208/274 A-28, and,A-57/314). l The team found the QA documentation inade'oujte. Problems included missing documentation for instrumentation installation, for in process and
. j
/
5
e I i final inspections, for welder identifications, and controls, and for base 1 and filler material identification and controls related to the reactor coolant system (see Allegation A-197 through 206/213/216). 1 The third area of potential safety significance and generic implications involved adequate review of QA recommendation. The team found 2 a violation of separation criteria for the installation of instrumentation tubing for the reactor coolant system (see Allegation A-279/280/282/284/ 288/27). The team found many cases of inadequately dispositioned and closed i nonconformance reports and discrepancy reports. Approximately 38% of EBASCO and Mercury nonconformance reports had problems, some of which included: missing documentation to support corrective action closure; lack of proper review for reportability in compliance with 10 CFR 50.55(e) for construction deficiencies; improper identification and resolution for an instrumentation tubing defect; no basis for voiding a nonconformance report; voided nonconformance reports missing from the QA files; the qualification of a welder was not based on the required welding test; a forged signature was discovered on an inspection report; a generic problem was detected for nonconformance reports and action to prevent recurrence was not addressed; and over 200 nonconformance reports were open indicating that corrective action is incomplete even though the plant is essentially
- complete. (For EBASCO nonconformance reports, see Allegations A-33/55/56/61/67/68/69/70/71/73/74/75/309/329/306v and w/221, A-18/164, and A-49/78/87/123.) (For Mercury nonconformance reports, see Allegations A-232/234/237/238/243/244/245/262/311/312/313/316/317/318/320/321/323/324/
325/326/327/328/331.) Also, there were problems with EBASCO discrepancy reports. Closure
. ,' references were incorrect or inappropriate; closure action was improper; documentation was inaccurate or incorrect; and disposition failed to address the discrepancy (see Allegation A-05).
6
Another potential problem area identified by the team involved inadequate welder qualifications and welding filler material control. Problems included welders not qualified to the correct welding procedure, and inability to locate qualification records. Also, welding filler material was not controlled as required by the ASME or AWS Codes for low hydrogen electrodes for the rebaking process. The allowable issuance time frame was exceeded, and the requirements for drying electrode coatings were not complied with (see Allegation A-215/175/240/239/306j). The team found corrective action documents that were not upgraded to nonconformance reports. Problems were found with the required corrective action system for upgrading certain documents to nonconformance reports. The documents not upgraded, and the percentages of them not upgraded, included: field change requests (55%); design change notices (29%); engineering discrepancy notices (60%); and deficiency notices (12%). Some documents had not received the proper EBASCO review, identification,
., corrective action, or closure, or had not received review for reportability ,, to 10 CFR 50.55(e) (see Allegation A-302/307/306b/306s/306x).
t , t Another area of potential safety significance and' generic implications involved vendor documentation. The conditional release system for issuance , of material, items, or components for installation was not complied with. l Some equipment was received at the site with missing or incorrect documentation, such as technical manuals, as-built drawings, test / calibration results, and certified material test reports. These problems were not properly documented by LP&L or EBASCO. Some of these problems have existed since 1976. The E8ASCO system in place for other conditional release items had identified and tracked various problems, but had failed to resolve the problems in a timely manner (see Allegation A-165/166/292/293/297/299/300). Finally, the team confirmed a QA program breakdown between LP&L, EBASCO and Mercury. In December 1982, the NRC took enforcement action against LP&L for a QA program breakdown and imp **ed I civil penalty. During this current QA review, the team fourJ I.bJt LFSL had failed to take 7
.. adequate corrective action. Additionally. LP&L, EBASCO, and Mercury failed to perform their audits as required and failed to take the required
, corrective action. Management audits performed by outside consultants identified problems and concerns for which LP&L also failed to take proper corrective action. The team found from this review that an overall QA program breakdown did exist, and that LP&L's failure to determine the root cause and their lack of comprehensive corrective action has allowed the breakdown to continue (see Allegation A-48/295),
for_the eight problem areas identified by the QA Records Review Team, LP&L shall submit a plan to the NRC identifying the safety significance of each issue together with an approach to resolution, including proposed corrective action and time frames for plan execution and completion. RESULTS
. The 73 quality assurance related allegations have been separated from
. . the 264 other allegations contained in NUREG-0787, Supplement 7. A
. numerically ordered listing of these allegations with a brief
" characterization" statement follows this section. Also following this section are the complete assessments for each quality assurance allegation. Related allegations with differing allegation numbers have been grouped together where the resolution of one is common to all.
i
\
9 4 8
QUALITY ASSURANCE ALLEGATION LISTING Allegation Num'er(s) b Characterization .-' A-01; A-02; A-03; Unqualified or incorrectly certified QA/QC A-28; A-57; A-304 inspection personnel. A-04; A-209 QC inspections were inadequate. A-05 EBASCO QA personnel used speed letters instead of deficiency notices (DNs) to identify problems on system turnover documentation packages. A-06; A-294; A-09 Untrained or unqualified QA records reviewers A-92; A289b. A-07; A-117; A-172 Falsification of QA documentation A-276; A-286 A-08 EBASCO QA records and documents were not stored in fire safe over night and some records were being stored in a warehouse in old file cabinets and boxes. A-13; A-19; A-340; Uncorrected electrical deficiencies A-341 A-18 Six EBASCO nonconformance reports (NCRs) were not included in the NRC system; i.e., inadequate nonconformance control. A-26 Nonconforming conditions may not have been properly identified and corrected by EBASCO. A-32 As-built drawings do not reflect actual plant configuration and EBASCO QA document reviewers were told to change redline drawings for T-8. A-33; A-59; A-227; NCRs improperly dispositioned and closed without A-232 substantiating indications. A-35; A-183a; A-197; QA documentation and records were inadequate A-223; A-230; A-308 A-36 Torque wrench was 38% under-calibrated. A-39 Nitrogen purge on piping system not performed.
. A-43 Improperly performed calibrations, including improper humidity and temperature control in a test lab (office trailer).
9
l 1 l Allecation Number (s) Characterization A-48 Complete breakdown in the QA program between EBASCO and the Mercury Construction Company. A-49; A-53; A-283 Personnel were not free to write NCRs; indications that the nonconformance program was nonexistent. A-54 Failure of E8ASCO and site subcontractors to implement their procedures. A-58 Many permanent plant items were not maintained by 4 subcontractors, EBASCO, or LP&L; items were rusting in place. A-63 Equipment not properly aligned or leveled when
-installed.
A-64 EBASCO site letters F-63724E, F-618958, and F-6147E imply that Mercury inspection activities may be unacceptable.
.. A-81 Changes made to N-5 documentation in packages for instrument lines without the knowledge of the Authorized Nuclear Inpsector (ANI) after his .. approval signature.
A-94; A-100; A-214 Lack of material traceability. A-143; A-186b; A-188 Review of QA documentation was inadequate A-279 A-165; A-253 Problems with vendor QA personnel and vendor QA documentation. l A-185 Written procedure does not exist for joint Mercury /EBASCO review of system turnover documentation. A-186a EBASCO QA surveillance did not monitor contractor work activity such as documentation. l A-187 Mercury instrumentation drawings were not correct because field changes were not incorporated into the drawings. A-196 An EBASCO Document Control Supervisor was not qualified for the position. i A-207 Mercury QA personnel were observed throwing away
. original documents after copies were made.
4 10
Allegation Number (s) Characterization i A-215; A-332 Inadequate welder qualifications and welding l problems. l A-229 An individual was directed to turn over l Operational Control Records (OCR) packages before ' system reviews were completed and the resulting documentation and hardware discrepancy lists submitted to LP&L documented inadequate reviews. A-249 EBASCO ONs were written and given to subcontractors who would process the DNs through their own engineering /QA personnel and would then either change the documentation or walkdown a system to eliminate the discrepancy, at which time the DNs would be discarded. A-263 A Mercury Construction Superintendent did not agree with Mercury audit findings and refused to correct problems identified. A-290 The disposition of corrective action documents
. was not adequate to correct the deficiency and changes were incorrectly made to QA records.. . A-302 Lower tier corrective action documents were not being upgraded to NCRs. Also, FCRs, DCNs, and EDNs were issued after the fact for nonconformances in lieu of NCRs.
A-306c Lack of EBASCO traceability documents on bolts measuring one inch or less. A-306t Noncertified document reviewers were making changes to QA documentation. A-306u T-B did not adequately measure the test equipment. s 11
e S 4 = 0
- e 9
6 ALLEGATIONS e D 9 4 12
Task: Allegation A-01 Characterization: It is alleged that the employment records for quality assurance (QA) personnel of the Mercury Construction Company have not been
- verified and that the implications of this alleged practice may apply to all Mercury employees.
Assessment of Allecation:' The implied significance of this allegation is that unqualified Mercury cuality control (QC) inspectors may have performed inspections on safety-related systems, making the validity of the inspections questionable and the quality of the work indeterminate. The NRC staff reviewed Mercury's procedure for qualification of inspection, examination and test personnel, inspector certification records, and personnel training and resume packages against Regulatory Guide 1.58, Revision 1, as suggested by NRC Generic Letter 81-01, ANSI N45.2.6-1973, and IE Circular 80-22. The NRC staff sampled 30% of the inspector certifications of QC personnel; the results revealed that no verification of past employment was documented. An additional staff examination of inspector qualification for Allegation A-02 has . indicated a generic problem of improperly certified inspectors. Mercury inspectors had performed inspections on safety-related systems. The
.- NRC staff believes this allegation has implied safety significance.
Actions Reouired: See Item No. 1 in the enclosure to the letter from D. Eisenhut to J. M. Cain (LP&L), June 13, 1984. k l 13
- n-- -
Task: Allegation A-03 Characterization: The allegation is that answers to Mercury Construction
- Company certification examinations for QA inspectors were given out prior to the tests.
Assessment of Allecation: The implied significance of this allegation is that Mercury personnel may have been certified incorrectly, based on fraudulent examinations, which could affect the acceptability of safety-related systems. The NRC staff reviewed a sample of examinations given to Mercury QC inspectors, the Mercury procedures for certification of inspectors, and previous reviews of this item by LP&L and the NRC Senior Resident Inspector. The examinations of the same inspectors used for the sample population of Allegation A-02 were reviewed for trends. No indications was found to support the allegation. The NRC staff found that although two examinations were graded incorrectly, in both cases the grade would have been passing anyway. The NRC staff was unable to interview any previous Mercury QC inspectors; however, the NRC Senior Resident Inspector interviewed six Mercury QC inspectors at the time of the allegation. They all stated that they had not
- been given the answers prior to the examination.
The NRC staff believes this allegation has neither safety significance nor generic implications. Actions Required: None. s i
\
14
l l I .". Task: Allegations A-02; A-208; A-274 Characterization: The allegation is that QC inspectors from the Mercury Construction Company were not properly certified for their positions. Assessment of Allecation: The implied significance of this allegation is that QC inspectors with the Mercury Construction Company may have been incorrectly certified because they lacked the required education and experience. The safety significance is that uncualified QC inspectors may have inspected safety-related systems, thereby rendering verification of the quality of these systems indetenninate. The original licensing commitment for QC inspector qualifications by LP&L as - stated in the PSAR was to the " green book;" ANSI N45.2.6,1973. The NRC task force staff determined that LP&L's comitments, as required in NRC Generic Letter 81-01, are inadequate in that they have committed to ANSI N45.2.6,1978, and Regulatory Guide 1.58, Revision 1, for Operations only. The commitment to these standards was never imposed by LP&L on the site contractors for the construction phase of Waterford Unit 3. It should be noted that the 1978 version of ANSI N45.2.6 is a less conservative standard, however, comitment to the 1978 standard and Regulatory Guide 1.58 would have made verification of resumes mandatory. Due to the findings related to Mercury Company activities the scope of the review was expanded to other contractors. This and GEO issue was Testing (ANST/TC-1A addressedQC inspectionby) reviews of theprocedures, qualification Mercury, Tompkins-Beck ANSI N45.2.6, 1973, and the certification and resume packages. NRC reviewed inspector certifications for 37 of the 100 Mercury QC inspectors, including certifications for all Level III personnel. Twelve inspector certifications were found questionable due to insufficient education or experience when compared to either the 1973 or 1978 versions of ANSI N45.2.6. The certification records of 38 T-B QC inspectors were selected at random and reviewed. Fourteen inspector certifications were found questionable due to insufficient education or experience. The T-B certification procedure was determined to be inadequate when compared to ANSI N45.2.6, 1973. The certification records of 40 GE0 Construction Testing NDE inspectors also were selected at random and reviewed to GEO Procedures and ASNT/TC-1A. All were found to be acceptable.
. This allegation has implied safety significance and is also The indicative full scope of theof a generic site problem for inspector certification.
problem cannot be determined until assurance is provided that the equipment
- and systems have been inspected by qualified inspectors.
Actions Reovired: See Item No. 1 in the enclosure to the D. Eisenhut letter of June 13, 1984, to J. M. Cain (LP&L). 15
Task: Allegation A-04 Characterization: The allegation is that some Mercury QC inspections were
, signed off as being done but were not.
Assessment of Alleaation: The implied significance of this aTlegation is that if inspections were not performed, installations would be questionable. The NRC staff reviewed 10 operational control record (OCR) packages of approximately 60 instrument lines with at least 10 welds in each line (600 welds). Additionally, the staff performed a walkdown.of 19 instrumentation installations and found no problems. The records reviewed by the NRC staff were complete and dates shown for performance of the inspections appeared feasible. Also see Allegations A-183a and A-184.. The NRC staff concludes that this allegation has neither safety significance' nor generic implications. Actions Required: None. 4 e o
=
D 16 i
l Task: Allegation A-05 Characterization: It is alleged that EBASCO quality assurance (QA) personnel used speed letters instead of deficiency notices (DNs) to identify problems on
- system turnover documentation packages.
Assessment of Allecation: The implied significance of this allegation is that, cecause speed letters are an unauthorized corrective action system, QA corrective action and nonconformance control systems may have been circumvented. Speed letters do not provide for accurate identification of deficiencies or discrepancies, and do not assure that corrective action has been taken
- regarding safety-related systems.-
The NRC staff revi~ewed EBASCO QA procedures for system turnover to determine if the procedurer provided for accurate records review to assure proper identiff-cation of n.onconformances. The staff also noted whether or not the procedures allowed the use of speed letters, and whether or not corrective action documents were generated based on the contents of the speed letters. The NRC staff reviewed EBASCO Quality Assurance Instruction (QAI)-9 and compared it to applicable ANSI 45.2 standards and 10 CFR 50 requirements. The staff found QAI-9 to be acceptable. QAI-9 provided guidelines for the collection, handling, and review of construction and installation QA records e and for transmittal to the EBASCO QA Records Supervisor for storage, handling, and maintenance. QAI-9 also provided instructions for describing the review status of construction and installation records. Quality assurance records
' reviewers were required to review packages for completeness, accuracy of content, proper fom, traceability, legibility, authenticity, and proper changes and supplements. Any deficiencies noted on the documentation were to be corrected, or, if unable to be resolved prior to submittal for turnover, they were to be documented on Form QAI-9.2. As a part of the review status, a separate reconenendation was to be made on the form to identify potential deficiencies affecting hardware.
The NRC staff found that EBASCO QA procedures did not address the use of speed letters as an alternative for identifying deficiencies or discrepancies. However, the staff did determine that speed letters were used to request infomation regarding engineering problems, to obtain engineering evaluations, and to question the disposition or closure of deficiencies, and that they w)ere also used by the EBASCO Quality Assurance Installation Review Group (QAIRG to identify problems noted on the authorized QAI-9.2 foms. The speed letters transmitting this infomation were then forwarded to lead coordinators who resolved or provided dispositioning instructions to the reviewer. The NRC staff reviewed the QAI-9.2 foms of 68 documentation packages and s detemined that the infomation noted in the EBASCO speed letters referenced previously identified problems stated on the QAI-9.2 forms. The staff found that corrective action documents, QAI-9.2 foms, NCRs, and DNs were generated
. as a result of these reviews.
However, during the NRC staff's review, several deficiencies were noted regarding item resolution or non-resolution. The staff believes that, due to the nature of the deficiencies, an NCR should have been written for: 17
o 02-CS-IC 9.2 Reviewed Item 63 - Inadequate documentation; should have been elevated to an NCR. o 02/3-FW/1C-8510 -9.2 Reviewed Items 18, 19, 20; 01-RC-1C-674 - 9.2 Reviewed Item 13. (1) Improper Closure Reference TBP-35; to be revised to correct deficiencies on February 15, 1983; latest revision of TBP-35 was June 18, 1982. An NCR should have been issued. (2) Removal of QC checkpoint was by improper authori.ty. o 02-SI-1C 9.2 Reviewed Item 17 - Incorrect / inadequate documentation; should have been elevated to an NCR. o QMC-APO-P47E - 9.2 Reviewed Item 26 - Closed OR with another DR, instead Penetrant test acceptance dates preceded the test request of an NCR. (prior to completion of the report). Both issues warranted an NCR. i o 80-IC-1143 - Traceability was required to warehouse only and not to the point of installaticn. Heat numbers were used interchangeably. Should have been upgraded to an NCR. o 'Q1-RC-LWS-RC Same as Q2/3-FW/1C-8510. o LW3-RC 9.2 Reviewed Item 11 - Flange retorqued but gasket installation was indeterminate. An NCR should have been issued. T
, o Q2-LW3-SI-10 F/E - 9.2 Reviewed Item 11 - Additional data added to a CMTR; procedural violation. An NCR should have been issued.
o CH-1C-342 - 9.2 Reviewed Items 19 and 25. Same as 02/3-FW/1C-8510. o CC-1C 9.2 Reviewed Item 1 - A DN was issued but did not relate to DR subject denoted. Flange was retorqued May 11, 1984 Potential Generic Issue - Use of 0-600 ft-lbs torque wrench for 90 ft-lbs when not calibrated at low range. (Identified as generic problem in DN T-B W-6531.) Resolution was "use as is" since the bolts are evenly torqued, but resolution did not address problem of torque wrench. Generic problem for all 9.2 reviews that closed out deficiencies referencing this DN.- Conflicting guidance with the above; FCR MP382 and IR 07012 (T-8) state , " torque to machine bolt specifications as opposed to evenly torqued." o AQMC-SI-P39-E - 9.2 Reviewed Item 10 - DN was written, but should have been an NCR. o OMC-HYPO-Pil-E - 9.2 Reviewed Items 43, 78, 81 - Inspection and documentation required by problem CIWA were not performed in accordance with procedures; an NCR should have been written, i The NRC staff found no indications that the use of speed letters did circumvent the corrective action and nonconformance control systems. EBASCO's QA records
- review procedures were in accordance with applicable standards of ANSI N45.2 18
and the requirements of 10 CFR 50. The staff also believes that QAI 9.2 forms (DRs) were acceptable to use for identifying potential problems with safety-related systems. 4 However, as noted previously, other discrepancies exist in the QA documentation packages reviewed by the NRC staff and, although reported on QAI 9.2 forms, these discrepancies required further corrective action by LP&L. Actions Reouired: See Item No. 6 in the enclosure to the D. Eisenhut letter of - June 13, 1984 to J. M. Cain (LP&L). e e G e e 19
Task: Allegation A-06 Characterization: The allegation is that EBASCO Quality Assurance (QA) reviewers have not completed on-the-job training (0JT) requirements. Assessment of Allecation: The implied significance is that without adequate QA indoctrination ana OJT, the record reviewers may not be properly performing their jobs. The NRC staff reviewed EBASCO training requirements, records, and work products and held discussions with the EBASCO training coordinator and QA personnel. The NRC staff found that records reviewers are required by EBASCO to have a minimum of from 40 to 80 hours of OJT depending on their previous experience, a high school level-education, classroom training, and QA reading assignments. There are no specific training qualification or certification requirements for document reviewers in NRC regulations, the ASME Code, or ANSI Standards. The NRC staff finds the EBASCO QA training program adequate. The NRC staff reviewed a sample of 40 training files of QA record reviewers and found them to be in compliance with EBASCO training requirements. The staff also reviewed a sample of work performed by QA records reviewers and determined it to be satisfactory.
- EBASCO had addressed the problem of incomplete OJT QA records for record reviewers with an EBASCO audit in October 1983. Subsequent EBASCO action updated and completed those records, and the subject audit was closed in - January 1984. The NRC staff has reviewed those audit files and concurs with the disposition, action, and closure.
The NRC staff noted that OJT requirements for QA record reviewers were not properly documented at the time of the allegation in August 1983; hcwever, corrective action has been taken by EBASCO. The NRC staff has verified that EBASCO QA records were properly corrected, and that work performed by QA record reviewers was satisfactory. Actions Required: None. s e 20
Task: Allegations A-07; A-29 Characterization: The allegation is that EBASCO quality assurance (QA) reviewers are " grooming" paperwork; e.g., having individuals sign documents
- years after the fact, and not dating to indicate a late entry, and that alterations in QA documentation are unauthorized, with no substantiating indications for changes.
Assessment of Allecation: The implied significance of this allegation is that records of safety-related systems may be inadequate. The NRC staff evaluated EBASCO's QA records review system and the qualifications of record reviewers. Also a sample of QA records was reviewed specifically for unauthorized changes, dates out of sequence, white-out, incorrect data, or appearance of falsification. The NRC staff found that EBASCO procedures provided guidelines for changes or corrections to documents. Documents were reviewed prior to submittal to record storage. EBASCO records reviewers were trained and oualified, and although problems in training had been encountered, those problems had been corrected. The NRC staff extensively reviewed documents submitted by various contractors and originally reviewed by EBASCO. One possible falsification of a signature on a nonconformance report (NCR) was observed. Otherwise, no, deliberate or obvious falsification of records was noted. The NRC staff considers the EBASCO records review system and qualifications of
- record reviewers adequate (see allegation A-06 and A-196). This allegation has neither safety significance nor generic implications.
Actions Required: None. e j 21
- Task: Allegations A-08; A-183b Characterization: The allegation is that quality assurance (QA) documents were not stored in fire safes overnight and that some records were being stored in a warehouse in old file cabinets and boxes.
Assessment of A11ecation: The implied significance of this allegation is that improperly stored records could be lost, misplaced, or destroyed due to fire, causing the required documentation to be irretrievable and the validity of installation and inspection to be indeterminate. The NRC staff reviewed EBASCO's QA records storage practices against the requirements of ANSI N45.2.9., and found them to be adequate. The records storea in the warehouse were found to be duplicates. Quality records and their protective storage are addressed in ANSI N45.2.9. Incompleted QA records do not fall under the definition of " valid records" as stated in ANSI N45.2.9. As defined in Section 1.4 of ANSI N45.2.9, a OA record exists when it has been completed; as defined by Section 3.2.1, " documents shall be considered valid records only if stamped, or initialed, or signed and dated by authorized personnel or otherwise authenticated. These records may be originals or reproduced copies."
^^
In conclusion, the NRC staff review did not find any indications of completed 0A records left out of the required storage areas. This allegation has neither safety significance nor generic implications. Actions Required: None. l l l k 9 22 L
Task: Allegation A-09
~
Characterization: The allegation is that contractor quality assurance people doing document reviews are being laid off and EBASCO engineers are being moved into quality assurance positions. Assessment of Allegation: The implied significance of the allegation is that without adequate quality assurance indoctrination and training the EBASCO engineers performing record reviews may not effectively perform their assignments. This issue was addressed by reviewing training requirements, records, and work products. In addition, interviews were conducted with the EBASCO training coordinator and other key personnel. The NRC staff found that record reviewers were required by an EBASCO procedure to have the minimum of a high school diploma, QA indoctrination consisting of lecture and reading assignments, and on-the-job training (0JT). For example, two electrical engineers originally from EBASCO's New York office, completed the QA indoctrination and reading phase of their training and one completed OJT on January 16, 1983. The other did not complete the OJT portion of the training and was transferred to the electrical engineering group onsite. He had never performed document reviews. The other individual performed e mechanical, instrumentation and control document reviews until he terminated his employment with EBASCO on March 15, 1984. Consequently, the NRC staff made the determination that the individual had not inspected any work for which he e had engineering responsibility. There are no specific training qualification or certification requirements for document reviewers in NRC regulations, the ASME Code, or ANSI Standards, and the EBASCO training program appears to be adeouate. NRC found no indications that GA document reviewers under contract were laid off for any reason, other than financial, in that EBASCO already had engineers like the two discussed on their payroll. Consequently, this allegation has no safety significance and is not indicative of any generic implications or management problems. Actions Required: None. p 4 e 23
Task: Allegation A-13, A-14, A-15, A-16, and A-17 Characterization: It is alleged that seismic supports for cable trays had documentation deficiencies identified by EBASCO Quality Assurance Installation Review Group (QAIRG) which were not being addressed. Assessment of Allegation: The implied significance of the five separate allegations in this area is that lack of corrective action on identified deficiencies could cause the quality of installation to be questionable. These five allegations involved specific problems with documentation of cable tray supports. The NRC staff evaluated each of the allegations for adequacy of documentation; and the staff inspected each installed support. LP&L had responded to each of the allegations in two submittals to the NRC Senior Resident Inspector (SRI). Their responses explained why they did not consider some of the allegations to be a problem, or if problems were evident, what corrective actions they were taking. The NRC staff reviewed LP&L's responses and considered them to be adequate. In assessing these allegations, the NRC staff verified that the documentation for each support matched each installed support inspected by the staff. The following is an itemized explanation of each allegation: Allegation A-13 involved defaced documents and inaccurate references to design
. documentation. The allegation was brought to the attention of the NRC staff on October 7,1983; however, on September 15, 1983, EBASCO NCR W3-6938 had been issued to document these concerns. EBASCO had reinspected the supports on October 12, 1983, and had closed the NCR on October 28, 1983. The NRC staff reviewed the documentation and inspected the supports, and determined that this nonconformance had been properly resolved.
Allegation A-14 was related to a support which contained three 3/8" concrete anchors instead of only two as indicated on the inspection documentation. LP&L agreed that a question existed, and they reinspected the hanger. Their reinspection revealed no deficiencies (Daily Surveillance Report W3-NY-12 TK-7, Feburary 28,1984). The NRC staff reviewed the reinspection records, performed a field inspection of the support, and determined that this problem had been properly corrected. Allegation A-15 involved a question of whether a support had been deleted, replaced, or remained, due to " confusing" information on inspection documentation. The NRC staff reviewed the subject documentation, performed field inspections, and determined that the support had been deleted and was replaced by a new support. The new support was inspected by the NRC staff and was found acceptable. Allegation A-16 involved a missing concrete anchor inspection form in the documentation package for support 32J13. EBASCO had issued NCR W3-7310 on October 19, 1983, to document this deficiency. The NRC staff reviewed the closure of this NCR and found that it had been properly dispositioned and
- closed.
1 i 24
Allegatica A-17 involved a question as to whether support 22E40 still existed in the field, due to a " confusing" hanger identification number in the documentation package (22E40/C100), and involved a concrete anchor inspection report which referenced this support, although the support contained no concrete anchors. The NRC staff review of this allegation revealed the folicwing: o The "C100" in the hanger identification number was the batch number of the paint used to coat the support. The staff determined that no conflict in hanger identification existed. o The anchor inspection report in question (303-87-87) had been revised on October 10, 1983, to indicate that the correct support identification number was 24E40, not 22E40. This correction of the report was considered appropriate. In addition, support 24E40 had since been deleted; therefore inspection report 303-87-87 no longer pertained to any existing installation and was removed from the records vault. The NRC staff considered this action to be appropriate. The NRC staff determined that EBASCO QAIRG adequately addressed and corrected seismic support documentation deficiencies. These allegations had neither safety significance nor generic implications.
.* Actions Recuired: None.
s 25 4
. Task: Allegations A-18 and A-164 Characterization: It is alleged that, due to EBASCO's inadequate
. nonconformance control, six nonconformance reports (NCRs) were found with
' ~
no numbers; e.g., not included in the NCR system. Assessment of Allecation: The implied significance is that improper use of the NCR system coulo lead to inadequate reporting, nonconforinance identification, timely corrective action, trending, reportability review, and correction of nonconforming conditions. This could render the quality of installed safety-related systems as being questionable. The NRC staff evaluated EBASCO's procedures for processing nonconformances to determine the control maintained and the methods used to generate NCRs. The
~
staff also reviewed'other, similar allegations and attempted to determine which unnumbered NCRs were not entered into the system. Regarding the unnumbered NCRs, the NRC staff attempted to determine if they had been entered into the system and if they had been properly dispositioned and closed. Also see Allegation A-53. The allegation referred to six missing EBASCO NCRs without numbers (and not in the NCR system): These NCRs were related to Fischbach & Moore tension test e problems related to expansion anchors. However, the NRC staff found seven unnumbered NCRs; the staff relayed this information to LP&L for their action, , and on February 15, 1984, LP&L provided a response, which stated that the original six nonconformance conditions had been addressed on EBASCO NCR W3-7533. The NRC staff review of LP&L's response disclosed that only four of the seven unnumbered NCRs were actually captured in W3-7533; the three missing NCRs were found to have been previously documented on EBASCO NCRs W3-7183, W3-7179, and W3-7561. LP&L also informed the NRC staff that additional problems with tension testing had been entered as EBASCO NCRs W3-7184, W3-7182, W3-7180, W3-7140, W3-7138, W3-7177, and W3-7139. The NRC staff reviewed the status of all of these NCRs and found them closed except for W3-7533, W3-7179, and W3-7140 (see Allegation A-33). Additionally, the NRC staff looked for other missing EBASCO NCRs and found that the following " voided" (missing) NCRs were not in the card index or in the QA vault: W3-27, W3-814, W3-859, W3-981, W3-1053, W3-1102, W3-1109, W3-1228, ' W3-1349, and W3-1438. The staff discovered that NCR W3-1215 was in the QA vault but was not in the NCR card index file. Overall the NRC staff's review disclosed that there were open NCRs with incomplete corrective action, which could result in unacceptable performance of safety-related systems. The staff could not review missing, " voided" NCRs; therefore, questions remain concerning acceptability of installed safety > systems, and the safety significance of this allegation could not be determined. The NRC staff believes this allegation has generic implications; as stated herein problems were identified with EBASCO and Fischbach & Moore. Actions Recuired: See Item No. 13 of the enclosure to the letter from D. Eisennut to J. M. Cain (LP&L), June 13, 1984. l l 26 i
. Task: Allegation A-19 Characterization: It is alleged that record deficiencies existed for 186
. separate Fischbach & Moore (F8M) documents.
Assessment of A11ecation: The implied significance of this allegation is that recoro ceficiencies may not have been properly identified and processed in the required nonconformance system and that they may have affected hardware installation. The 186 documents identified 192 separate " issues" requiring investigation. In assessing the allegation the NRC staff divided the " issues" into the following seven general categories: (1) 75 of the issues pertained to inspection checklists containing handwritten notes to the effect that the actual quantity of field-run supports attached to a specific engineered support needs to be verified. This concern was identified as a generic problem during the NRC construction appraisal team (CAT) inspection and will be tracked during CAT follow-up inspections. (2) 12 of the issues pertained to inspection checklists which had not been signed by the Quality Control (QC) Engineer in the space provided. A review of the procedures in effect at the time revealed that the QC Engineer's signature was not required provided that the OC Supervisor's signature appeared in the space provided for him. The procedure sections which clarified this are QCP-305, Revision 4; November 20, 1980, Section 6.4 and QAP-201W3, Revision 2, January 26, 1981, Section 5.3.1.C. The NRC staff review revealed the inspection checklists missing the signature of the QC Engineer were in fact signed by the QC Supervisor., (3) 29 of the issues pertained to documents with handwritten references to a
" Dead File" or a "Ouplicate File." In conversations with a former F&M records clerk, the NRC staff learned that if two or more duplicate records were generated for the same item or activity, only one was maintained as the official record. The others were placed in a file referred to as either a " Dead File," a " Historical File," or a " Duplicate File." This
" Dead File" is now located in F&M's Dallas office. The staff compared 13 i
of these " Dead File" records with the official records in LP&L's records facility and determined that the official records provide adequate documentation for the installation or activity involved. (4) 5 of the issues pertained to various documentation problems for electrical supports that were not adequately addressed. The NRC staff brought these
- items to LP&L's attention, whereupon they issued LP&L Site Surveillance
, Report W3S-84-16S to initiate appropriate corrective action. Items 1 through 4 on the site surveillance report encompassed these 5 issues. (5) One issue duplicated Allegation A-13 and is discussed in the assessment of that allegation.
. (6) 67 of the issues were listed for which the NRC staff either could not detemine the nature of the concern or determined that the concern had been adequately addressed during its normal review cycle.
. l i
1 27 l
(7) Three of the issues duplicated other issues. The NRC staff concludes that this allegation has neither safety significance nor generic implications. Actions Recuired: LP&L shall resolve the issues raised in LP&L Site Surveillance Report W35-84-16S. 0 1 4 6 28
Task: Allegation A-26 Characterization: The allegation is that interoffice correspondence, dated April 9,1983, from EBASCO to LP&L (File Reference CMH #83-601), implies that nonconforming conditions may not have been properly identified and corrected by
- - EBASCO, thus having hardware impact.
Assessment of Allecation: The implied significance of this allegation is that witnout proper toentification of nonconformances and proper corrective action, the quality of installation may be questionable. I EBASCO and LP&L issued a series of interoffice memoranca, including the subject correspondence, to discuss the status of various open items related to Start-Up
. System (SUS) SUS-48, Containment Vessel.
l' The NRC staff reviewed this interoffice memoranda and correspondence and found that the problem with the containment vessel documentation had been identified 1 as a result of an EBASCO quality assurance installation review group (QAIRG) review, and that the problem was discovered when Chicago Bridge & Iron (CB&I) turned the documentation over to EBASCO. The NRC staff found that EBASCO had properly identified and resolved the nonconformance problems on EBASCO QAI 9.2, Deficiency Reports (DRs). The requirements for upgrading of DRs to nonconformance reports (NCRs) were complied with when required (e.g., NCR W3-6231), and the interoffice memoranda and j correspondence were issued to discuss open item status, a S The NRC staff concluded that the subject correspondence was used only to document status and expedite closure of open items. All problems on the
, containment vessel were' properly identified and dispositioned on DRs and NCRs.
i Additionally, the NRC staff conducted an interview with the correspondence originator, who said that the status was obtained by reviewing the QAI 9.2 DRs and this listing was for information to further status and expedite open item closure. l The NRC staff concludes that this allegation has neither safety significance nor generic implications. Actions Required: None. I i 29
Task: Allegations A-32; A-50; A-83; A-84; A-85; A-86; A-121; A-128b; A-281; l 7p376 Characterization: The allegation is that as-built drawings may not reflect the actual plant configuration and that Ebasco 0A document reviewers were told to change red-line drawings for Tompkins-Beckwith (T-8). Assessment of Allegation: The implied significance of this allegation is that inaccurate as-ouilt orawings may not have received adequate engineering review and approval and the required inspections may not have been accomplished. The NRC staff addressed this allegation by reviewing the as-built program for Mercury and T-8. In addition to these reviews,19 Mercury-installed systems were walked dcwn and verified against the as-built drawings. Six T installed small-bore piping lines (2" or less) were also walked down. All of the drawings reviewed accurately reflected the actual plant configuration. These drawings were reviewed and found to have the required engineering review and approval. (Also see Allegations A-35; A-187; A-208; and A-230.) Allegations A-83, A-84, A-85, and A-86 address problems identified with the recording of heat numbers on as-built drawings. This requirement was not complied with as reported on Ebasco NCR W3-6786. 'The disposition and closure of this NCR is unacceptable to the NRC as stated in Allegation A-033.
. The NRC staff could find no indications of document reviewers changing red line drawings or " doctoring" the quality records.
. Based on the reviews conducted the NRC staff concludes that this allegation has neither safety significance nor generic implications.
Actions Required: None. 30
Task: Allegation A-33; A-55; A-56; A-61b; A-61c; A-61d; A-61e; A-67; A-68; T57; A-70; A-71; A-73; A-74; A-75; A-221; A-306 v&w; A-309; A-310; A-329
. Characterization: The allegation is that EBASCO nonconformance reports (NCR) were not entered into the NCR system, were improperly dispositioned and closed without substantiating indications, and that a questionable trending praalem was being utilized.
Assessment of Allegation: The implied safety significance is that' improperly - dispositioned NCRs could place the quality of installation in question, and a questionable trending program could cause uncorrected problems in'the QA program to continue undetected. The NRC staff conducted a. review of selected EBASCO quality assurance (QA) records and the NCR tracking system. The selected NCRs were reviewed for content, compliance to procedures, accuracy, completeness of the disposition and final closure. The staff disclosed that 34 of'89 NCRs (38.2 percent) reviewed were improperly dispositioned. The NRC staff evaluated EBASCO's NCR trending program. It was found to be adequate and was performing its intended functions. The issue of NCRs not being properly entered into the EBASCO NCR system was addressed in detail in Allegations A-18, A-53, and A-283.
. The following is a summary of the NRC staff review of EBASCO NCRs that contained questionable dispositions:
EBASCO W3 NCRs NCR 7139 - This NCR involved field inspection for the horizontal seismic supports for Radiation Monitors RE-HV 50285, RE-HV 50315, and RE-HV 0200.65. Only the data for the RE-HV 50315 support was the correct attachment (see Attachment No. 2 of NCR). LP&L reopened NCR and corrected seismic support identification problem (see Attachment No. 4 of NCR). The NRC staff identified this problem and LP&L took action which was acceptable to the staff.
'NCR-3912 - Involved nine 23J-2 type supports discovered during walkdown for which the fit-up inspection was by-passed. The original NCR disposition failed to address the actions required to prevent the reuse of the items. Attachment No.14 of this NCR identified this issue which was resolved by stating "it was not required for the disposition of this NCR..." No other NCR was reopened or referenced to resolve the issue. .
NCR-5565 - Identified that a QA inspector trainee signed off on the fuel handlingbuilding(FHB)cranequalityverification(QV) inspections. the NCR called for a complete reinspection. There are no documents attached to this . NCR showing a reinspection was performed. The NCR states "... corrective action per Attachment #4...." but there was no Attachment No. 4 to the NCR package. 31
.w -- - - - -.
. NCR-5563 - Identified that a QA inspector trainee dispositioned NCR No. W3-1728, for the Fuel Handling Building Crane, for J. A. Jones OV Department. This in violation of the requirements of ANSI N45.2.6. The inspections in question
. were signed off on August 27, August 28, and November 6,1979, and then by a co-signature on February 4, 1983, by a OA inspector who claimed to be present at the first inspection. This co-signature of the inspections in cuestion eliminated the requirement for a reinspection called for in the recomended disposition. The co-signature was applied 3 years and 5 months later. The NRC staff found this to be a questionable disposition of the NCR.
NCR-6159 - A sample inspection of Tubetrack welding identified that prior to July 1982 an unknown quantity of welding was performed using WPS "B" procedure without backing plates. This NCR raises the question of traceability problems that were not identified and addressed by the NCR. The NRC staff also found that the sample used when tensile testing the welds was questionable. The sample should have been representative of the weakest weld joint in lieu of the strongest (i.e., worst case example should have been used to conduct tests). 4 NCR-3919 - Was initiated due to a tubing crack discovered during a system hydrostatic test of instrument line PT-RC-0173, system 52A2, Reactor Coolant. This NCR also resulted in Significant Construction Deficiency (SCD) No. 61 being issued. It was detennined that the tubing failure was a result of a manufacturing defect (process, not metallurgical), and an attempt was made to ascertain that all tubing of this specific heat number was reinspected. Due
~
to the problem associated with traceability (being lost), the corrective action
.. was to reinspect all tubing installations to locate this heat of defective tubing. The reinspection reportly located all installation locations. The NRC staff review of this NCR revealed that operational control record (OCR) installation packages indicated that approximately 530 ft. more tubing was installed than was received on site. This was also verified by a review of warehouse issuance records. The " Requisition on Warehouse" fonn had been changed using liquid paper and a subsequent entry had been crossed out with ink.
SCO 61 states that the tubing in question, Heat No. 466023,1" tubing, was
- installed in 89 instrument locations. A sample of nine OCR packages for fuel handling building (FHB), reactor containment buf1 ding (RCB) and reactor auxiliary building (RAB) systems were reviewed to verify where the tubing was in fact installed. The nine packages (10 percent) were randomly selected from the list used to disposition the NCR. Only one OCR package actually reflected that the tubing in question was installed. This was also verified by an NRC staff walkdown. The walkdown on the eight remaining OCRs revealed different
' heat numbers installed; however, the OCRs were in agreement with these heat numbers. Therefore, the list used for dispositioning this NCR was invalid; this NCR and i SCD 61 were also invalid, as was the testing and disposition.
- These NCRs were closed out by referring to EBASCO letter F-61147E. The problem is that this letter did not close out these or other NCRs.
i 32
p -
.- ' NCR-6514 * - The problem of traceability for the weld being performed was still
- > in question; not addressed. The NCR also questioned use of some Bargen-Patterson designed supports installed by Mercury without traceability. This
. problem was also not addressed by referenced attachment.
flCR-3941-RI * - Identified that support number one fit-Qp inspection was by-passeo and the support had been completely welded out with only the welder's ID. NCR-5819 *. - Identified the problem of instrumentation supports being painted prior to final welded visual inspection. Disposition had been to inspect the welds through paint which was unacceptable. NCR-6221 - Identified that Weld Control records were signed off by an individual who was not a certified Level II inspector. Sign-off was based on Letter of Designation. The NCR disposition referred to the Tompkins-Beckwith (T-B). (April 1,1980) Quality Manual that was not in effect at the time the Letter of Designation was written (January 8,1979). Also, a reference given in the Letter of Designation did not allow designee sign-offs and was in effect as of March 15, 1983; the Letter of Designation also failed to meet the requirements of ANSI N45.2.6. NRC-6511 (Mercury-3336) - Stated that "during final inspection of installed I-Beam for support 1117-114, weld to existing beam 1A was rejected." The NCR only addressed the fact that the maximum gap was violated, but the weld was. - .- rejected for: (1) undersize, (2) lack of Fusion, (3) are strikes, and (4) undercut. Mercury NCR 3336 reconeended weld removal and rework. This
~
recommendation was crossed out and only the nonconforming fit-up gap was addressed. There were no records of rework or reinspection, and only copies of Mercury's NCR were attached to EBASCO's NCR. NCR-4219 (Mercury-614) - Identified a violation of QCP3110.4 paragraph 6. The nonconformance was that the sample system piping had been bent downward causing a low point in the piping. The piping was being forced down by support , - - SLRR-188. QCP-3110.4 stated that " tubing must be properly routed." This disposition stated that "... tubing was re-evaluated after support SLRR-188 and sample Ifne were installed, after completion of Penetration 29 work." There were no records for rework or reinspection to indicate satisfac' tory reinstallation of supports and sample lines. NCR-7432 - Identified a problem with concrete preplicament and post-placement documentation. The documentation could not be matched because the identiff-cation of the various placements were on different quality control (QC) forms. Also, this NCR was dispositioned by stating "...this problem was addressed on other NCRs and therefore voided..." f,o specific references were used; therefore, this disposition is unacceptable. Also, a QA engineer approved the recommended disposition and then voided the NCR. - NCR-4137 - Identified material and weid problems on support SLRR-238. This NCR has been closed out but failed to have 3 of*4 required welds on "M" gusset
" These McHs were closed out by referring to EBASCO letter F-61147E. The-problem is that this letter did not close out these or other NCRs. '
33
plates completed. This problem was identified during a QA review and a discrepancy notice (DN) was issued for the missing welds. This is an example of improper NCR closure and reopening, and the use of an incorrect reporting system (DN in lieu of NCR). NCR-4088 (Mercury-491) - This NCR identified numerous discrepancies found during a walkdown performed against drawing 160-T-035-A. There is no documentation that work was accomplished or completed. -The only document attached to this NCR was a Quality Control (QC) Report which listed numerous references but none were attached. EBASCO could not locate the referenced documents. NCR-5974 - Identified a problem with loss of heat number traceability for safety and non-safety grade related materials. This NCR was used to disposition approximately 150 to 200 DNs with "Q" prefix. NRC staff review indicated that in almost all cases loss of heat traceability was deemed acceptable. This NCR's disposition is questionable to the NRC staff. The problem still existed in that safety and non-safety grade material could have been mixed; i.e., there was questionable use of material with no pedigree or the material could have been the incorrect type and grade. NCR-5564 - Identified the problem of perfonning the final weld inspection after the weld has been painted. This NCR was closed out referencing an EBASCO
.- letter which stated" the inspection criteria will be without removing paint to inspect for undersize weld and lack of weld material where installation drawing calls for weld material to be deposited by the installer. This disposition is
.- unacceptable because visual inspections also include an examination for welding defects, not just dimensions.
NRC-6786 - Identified that many Mercury NCRs were irsued concerning the lack of neat numbers. These NCRs were closed by referencinq a generic series of EBASCO NCRs. The EBASCO disposition stated that the possible heat numbers will be documented on the Mercury as-built drawings. This data is not recorded on the as-built drawings, however, the Mercury Company NCRs have been closed. The disposition of this NCR does not address where the required heat numbers were recorded or how traceability was maintained. NCR-7177 - It was noted that Fischbach and Moore (F&M) violated Procedure QCP-309, 6.3.2.4, that is, they failed to test three additional expansion anchors for every anchor that failed. It was also identified that an uncalibrated pressure gauge was used on the tension tester and tension testers serial numbers were not recorded. The NCR disposition stated that "QCP-309 did I not require recording of serial numbers;" this violates ANSI N45.2, Section 13, that requires the traceability of measuring and test equipment to point of usage. F&M should have written an NCR. Inspection Report (IR) 311-06-70 and IR 310-36-43 identified bolt failure due to excessive slippage. Dispositions prescribed by these irs were in violation of QCP 309, Section 6.3.2.2(d) and 6.4.3. NCR-7182, NCR-7180, NCR7181, NCR-7184, NCR-6723 - These NCRs also involve a violation of ANSI N45.2, Section 13 requirements in that QCP 309 did not
>* require the tension testing equipment's serial number, calibration date, and pressure gauge number to be recorded.
34 .
P CR-7547 - Noted discrepancies against OCR-1830 and Mercury NCR-0806. The
!' disposition of this NCR is unsatisfactory due to the system passing a hydrostatic test is used as the basis for acceptability of fit-up discrepancy between the 3
union and tubing. This does not take into account the effects of service conditions such as vibration and cyclic loads, and an engineering evaluation was not performed. NCR-1650 - Ungraded DN No. MC 2128, which identified that the pressure gauge on the anchor bolt tension tester, was found to be out o.f tolerance reading (+450 psi higher than actual). The NCR disposition was to retest all anchor bolts installed prior to the date the tension test guage was determined to be out of calibration. The question is how was it determined which bolt (s) to retest when QCP 309 did not require the recording of tester serial number, on 3 . the previous tests? NCR 1803 also referred to gauges being out of calibration. Also see NCRs 7177, 7182, 7180, 7181, 7184, and 6723. NCR-6623 - Identified a heat number and signature being falsified. The tubing in question was removed and replaced in accordance with Mercury NCR 3696. The question was what was done to assure that no other heat numbers were falsified by Mercury, and who forged the signature and entered the heat number on QC Material Verification Mechanical Inspection Report? The NCR's disposition did not address the falsification question and the reference to SCD-57 deals only
- with traceability.
NCR-5586 - Weld Testing Laboratory was not surveyed (audited) and placed on the Approved Vendors List by Mercury prior to welder performance qualification i
. taking place. This item was not addressed in the NCR disposition. Also, the
- statement provided by the test lab that "a Mercury inspector reviewed all tests,"
l 1s not adequate. Mercury should have had the person (s) who perfomed this l surveillance document this activity. l l NCR-6165 - States "... welder R-1 is not qualified to this procedure..." The l disposition states, "... Measures taken to preclude recurrence is required..." No indications of this action could be located. NCR-7099 - Identified improper weld on cabinets 48A and 488. Numerous problems were noted with this NCR disposition. FCR-IC-P-416 Revision 1, Sk-1, called for a fillet weld where a flare bevel weld was required. There is no documentation to adequately support the NCR disposition. Also, the evaluation of disposition by EBASCO states, " Evaluation indicates that the stresses are low." There is no documentation indicating what stresses were being referred to. In addition, the recommended disposition "that ESSE (EBASCO Site Support Engineering) evaluate the cabinet base metal cracks" was not addressed. Proper weld length could not be achieved because the incorrect sized embed
- plates were installed and there were bolt holes that had been filled with grout in the areas required to be. welded. Weld size and length were not adequately addressed.
NCR-4593 - Disposition inadequate. See-A-232 for details. The NRC staff also discovered that the EBASCO requirement for having the NCR closed within a 20-day period or having an extension granted was consistently violated. The following table is an example of NCRs found to exceed the closure time requirements of ASP-III-7, Section 6.1.3.a: 35
TABLE OF NCRs NOT MEETING THE 20-DAY ACTION LIMIT Extension Letter Attached
. . Verification of (See Notes for NCR No. Issue Date Discosition Date Exolanation)
W3-6597 07/06/83 11/28/83 No W3-6950 07/03/83 10/18/83 No W3-6615 07/27/83 10/25/83 1 W3-6032 04/06/83 10/21/83 2 W3-5713 02/17/83 11/04/83 3 W3-4504/R3 07/08/83 12/03/83 No W3-5124 11/06/83 09/13/83 4 W3-5063 10/22/82 10/19/83 4 W3-5071 10/22/82 10/19/83 4 W3-5070 10/22/82 02/09/84 4 W3-5074 11/04/82 10/19/83 4 W3-5232 11/08/82 10/13/83 4 W3-5231 11/08/82 03/19/84 4 W3-5398 12/13/82 02/01/84 '4 Notes
. 1. Extension requested on October 6,1983. Extension granted to October 20, 1983.
- 2. Extension requested on October 11, 1983, again on October 21, 1983.
Extension granted to October 24, 1983.
- 3. Extension requested September 2,1983, granted to September 30, 1983.
4 Extension requested on September 2, 1983, granted to September 30, 1983, but NCRs have no indication of this extension or any others. This 20-day working requirement was removed from the ASP-III-7 procedure with the issuance of change issue "K." Issue "K," Section 6.3.1.a. now states
" Monitor the status and required corrective action to assure timely completion."
The definition of timely completion needs to be addressed. The NRC staff also found that the 10 CFR 50.55(e) for evaluation for reportability did not include the proper engineering involvement. The evaluation was generally accomplished by a QA engineer only. ESSE apparently was not involved in the required engineering input.
- W3-6530 - Identified a specific problem with EBASCO Letter No. W3-QAIRG-569 dated June 24, 1983, supplied by the alleger. This letter identified a T-8 welding processing violation of EBASCO Specification 884-75, General Welding 7
Requirements, Section II 6.04, which requires that the first weld pass in nuclear piping service shall be a gas tungsten arc weld (GTAW) process. Contrary to this requirement, T-8 performed socket welds in carbon steel nuclear piping with a shielded metal arch weld (SMAW) process. NCR No. W3-6530 was issued on July 12, 1983, and was subsequently accepted by EBASCO 36
engineering based upon the welds passing hydrostatic test and liquid penetrant .'. or magnetic particle testing. Based upon this action and the fact that the ASME Code was not violated, the NRC staff concluded that the disposition was satisfactory. W3-7241 and W3-7218 - The NRC staff reviewed these ESASCO NCRs and found them to ce acequately aispositioned ar.d properly closed. In conclusion, the NRC staff believes NCRs were improperly dispositioned and closed out without corrective action being completed. This allegation has safety significance and generic implications and is indicative improper management controls. Action Reouired: See Item No. 6 in the enclosure to D. Eisenhut letter of June 13, 1984, to J. M. Cain (LP&L). O e 9 og 9 m
~~
37
Tasli: Allegation A-35
~ " ~
Characterization: The allegation is that because of inadequate documentation, LP&L and EBA5CO cannot verify that piping systems installed and inspected by Mercury, NISCO, T-8, and EBASCO, were completed in accordance with the ASME code. Assessment of Allecation: Adequate documentation may not be available, to verify tnat systems were built in accordance with the ASME Code even though parts of systems may have been reviewed and accepted. Mercury, NISCO, and T-B (the contractors) installed safety-related piping and control systems, and EBASCO and LP&L reviewed and accepted the documentation of the work. The implied safety significance is that without adequate documentation, the accept-ability of the work is questionable and the quality of the installation is 4 indetenninate. Documentation control procedures of the contractors, EBASCO, and LP&L were compared to the requirements of ASME Code,10 CFR 50, and applicable ANSI N45.2 requirements. These procedures were found to be adequate. Implementation was verified by reviewing objective indications to substantiate documentation adequacy. Start up system (SUS) 52A, Reactor Coolant (RC) piping, was used as an example to verify the documentation relationship among NISCO, T-8, and Mercury. The record turnover package for SUS 52A-2 was used to identify large bore, small-bore, and instrumentation interface with corresponding contractor identity e and applicable Waterford Start Up Group -(WSG) drawings. From the WSG drawings, NISCO's installation /weldment documentation turnover packages were obtained by referencing the RC ISO listing. Using EBASCO's piping component list, and sorting line number by ISO, small bore weldment packages of T-B were retrieved. Mercury's Master Instrumentation List cross-indexed the instrumentation called out on the WSG drawing to Mercury instrumentation tubing documentation packages. Documentation packages were reviewed by the NRC staff to determine if:
- 1. Adequate / detailed quality records were maintained.
- 2. Records were reviewed for completeness prior to filing as historical records.
- 3. Inspection / test results were documented and traceable to the material.
- 4. Records were retrievable when required.
The NRC staff found that the contractors, EBASCO, and LP&L have implemented the
- program requirements. Therefore, the allegation has no safety significance nor generic implications.
'.. Actions Recuired: None.
38
Task: Allegation A-36 Characterization: A torque wrench utilized by Gulf Engineering for work on the Emergency Diesel Generators was 38 percent under calibration. Initial Assessment of Sionificance: An improperly calibrated torque wrench may have invalidateo tne work performed while in an uncalibrated condition. The NRC Office of Investigation (0I) has investigated this allegation because of charges of falsification and intimidation. They, in turn, returned two technical issues to LP&L for response. LP&L responded to these issues and took corrective action. The NRC staff addressed this item by a review and evaluation of the LP&L response and subsequent corrective action. Technical Item No. 1 - The alleger reported that the exhaust pipes to both generators A & B do not have any . type of cover over them. Subsequently, during normal rainfall the pipes fill up with rain water and leak into the engine compartment. The alleger reported this problem because during work on diesel generator "B," to remove water and to repair damage in 1980, the torque wrench was found out-of-calibration. e Discrepancy Report No. 047 was written on September 24, 1980, to document the out-of-calibration torque wrench; however, the alleger stated that he was forced to accept it as is and the re-torque was never performed. The corrective action taken to prevent further damage by rain water was to finish the roof over the diesels and to install elbows in the exhaust system. There has been no recurrence of this problem. Technical Item No. 2 - Because of the uncertainty of the torque values achieved during work on the fuel nozzles, LP&L presented a recent (Spring 1984) work package where the nozzles had been torqued as objective indications of corrective action. The NRC staff review of the work package indicated that the work was accomplished in accordance with established procedures and was acceptable. In conclusion, the NRC staff found that although problems existed in 1980, the problems have been adequately corrected. Therefore, this allegation has neither safety significance or generic implications. Actions Reauired: None. e 4 39
l f l
. Task: Allegation A-39, A-42 Characterization: Multipleallegationshavebeenmadethatthenitrogen . purge on piping systems was not performed prior to welding, specifically for diesel generator piping systems.
Assessment of Allecation: The implied significance of these allegations is that tne ASME Code anc contractor welding procedures may not have been complied with. The result could affect the acceptability of the welds. The NRC staff addressed the issue by reviewing contractor (Tompkins-Beckwith) welding procedures and the ASME Code, and by physical inspecting of portions of the diesel generator piping systems. Specifically, the NRC staff selected for evaluation the lubrication cooling water system and the lubrication piping system. In both cases, the systems'were found to be welded in compliance with the ASME Code requirements for purging. Since both systems were carbon steel, the omission of the purge was permitted by the Code and T-B Procedure 1.8. Additionally, the physical inspection of welds revealed no deficiencies. It should also be noted that nitrogen is not normally a welding purge gas, the gas normally used at nuclear construction sites is argon. Based on this review the NRC staff concludes that this allegation, has neither safety significance nor generic implications. Actions Recuired: None. m 9, 40
Task: Allegation A-43
'~
Characterization: Improperly performed calibrations including improper humidi and temperature control in Magna Flux Company's test lab (office
, trailer .
Initial Assessment of Sionificance: Improperly calibrated instruments could invalicate tne acceptacility of safety-related systems. This issue was addressed by a review of the response to this item provided by LP&L. LP&L stated that no purchase orders had been issued to the Magna Flux Company for work on the Waterford site. The NRC staff review determined that calibrations were correctl} performed in accordance with approved procedure. When procedures were violated, nonconformances were documented. The OI investigation of allegations in this area has indicated that the alleger, when interviewed, stated that "I intentially overstated my concerns to get NRC attention." Since the scope of the work of this allegation (Diesel Generator Installation) did not require the calibrations to be accomplished in a controlled atmosphere laboratory,
.the calibrations were deemed acceptable.
- In conclusion, this allegation has neither safety significance nor generic implications.
- Action Required: None.
6 6 9 e 41
l Task: Allegations A-48 and A-295 Characterization: The allegation is that there was a complete breakdown in the
. QA program between EBASCO and the Mercury Construction Company.
Assessment of Allecation: The implied significance of this allegatfor is that a "ccmpiete" QA breakccwn could cause the quality of all Mercury cons ruction work to be questioned and cculd imply that all work at the Waterford site might be questionable. NRC Inspection Report No. 50-382/82-14, dated December 6, 1982, identified a partial QA program breakdown and imposed a civil penalty against LP&L. LP&L acknowledged a QA program breakdown had occurred between EBASCO, Mercury, and Tompkins-Beckwith. Therefore, the allegation, as stated, is true. A QA program treakdown did occur between EBASCO and Mercury construction company. Corrective action is discussed in LP&L's response to the civil penalty, SCD-57, and in LP&L's responses to preliminary questions asked by the NRC inquiry team. The NRC staff performed an additional review in this area to determine whether LP&L identified and corrected the root cause of the OA program breakdown. The NRC staff interviewed ' site workers, reviewed Mercury programs, procedures audits, and related documents with a goal of attempting to understand how the QA program breakdown occurred and how it could have been avoided. Interviews Through interviews with Mercury employees, the NRC staff learned that from mid-1978 through mid-1983, the Mercury Company had six different project superintendents, six different project engineers, and six different QA site supervisors. Site QA/QC personnel interviewed felt that they received little or no support from Mercury corporate QA management, that EBASCO and Mercury management were "in bed together," ran " roughshod" over QC, and that QA/QC decisions were constantly overridden by cost and schedule. This was said to have resulted in a high turnover of Mercury QC inspectors. Several QC inspectors were said to have written letters of resignation describing these poor conditions and were threatened with dismissal if these letters were not withdrawn. Mercury Corocrate Annual Review Mercury Internal Audit No. 1-9-80 (January 8, 1980) documented that Mercury management had not reviewed the OA program since September 1978. The audit log shcwed this finding to be closed on the same date as the audit report. However, EBASCO Audit CBB/AEZ-83-2-3 determined that the
~
audit was not signed off as closed and that only one annual review, for July 1981 to July 1982, had been made. This annual review was reviewed by the NRC staff. The staff concluded that the review did not have the appropriate scope or depth to adequately evaluate the Mercury site QA program. 42
Percury Internal Audits Mercury Internal Audit No. I'-11-80 (danuary 11,1980) documented that an
. internal audit on QAM Sections 10,11,14,15, and 17 was not performed in 1979, despite the requirement that internal audits are to be completed at least once a year. Mercury Internal Audit No. 5-11-82 (May 11, 1982) documented that QAM Sections 2, 7, 13, 14, and 15 were not audited during 1981. No Corrective Action Request (CAR) was initiated as a result of the 1980 audit, but CAR No. 44 was issued May 11, 1982, more than 2 years after the same finding in Internal Audit No.1-11-80.
The NRC staff learned that Mercury had not audited QAM Section 5, Procurement, from 1978 through 1983 because, as they stated, they had purchased no materials. However, Mercury Letters WQ-1216 and 1225 (December 15, 1982, and January 4, 1983) discuss Mercury's issuance of a , purchase order to Welders Testing Laboratory for welder qualification services in late 1978 that remained in effect as of January 1983. LP&L acknowledged this finding and stated that Mercury had also purchased calibration services from another firm. The NRC staff reviewed Mercury files and determined that QAM Sections 5, 12, 17, and 18 were not audited in 1980. The failure to audit Section 12, " Nonconforming Items and Corrective Action," is significant because of Mercury's apparent failure
,. to take corrective action regarding their failure to perform management reviews and internal audits, and because Section 12 was not audited in 1981. The staff also discovered that QAM Sections 5, 14, and 16 were not audited in 1981, and that QAM Sections 5, 11, 12, 13, 14, 15, and 16 were not audited in 1983. Although the staff found response letters to audit findings describing corrective action in the file for calendar year (CY) 1979, no responses were found for CY 1980-1983.
This partial QA program breakdown has never specifically been reported to the NRC even though Mercury identified the deficiency in January 1980 and EBASCO identified the problem in June 1982. Mercury Audits of Implementing Procedures The NRC staff review of Mercury QA, construction, and spe~cial process procedures showed that Mercury had not audited the following Mercury Company procedures during the life of the project: MCP-2140, 2170, 2175; SP-650, 651, 652, 653, 654, 655, 656, 657, 658, 661, 662, 663, 668, 670, 672; WPS-B, P, G; BP-1; and WPS-WE-4. EBASCO Audit No. SW-82-6-1 (June 7, 1982) documented the finding that Mercury implementing procedures had not been audited. Many subsequent documentation and hardware deficiencies identified were related to those procedures not having been audited by Mercury. In addition, on those areas that were audited, it appeared that special or
. supplemental audits should have been performed in addition to annual audits of these areas because of trends indicated by nonconformances, deficiencies, and audit findings.
43
Mercury Corrective Action Reports (CARS) The NRC staff reviewed 143 CARS issued by Mercury Company from December 19, 1978, through August 9,1983, including all CARS relating - to audits. This review revealed that EBASCO should have been aware of adverse trends concerning audits at least as early as December 19, 1978. Other trends which should have been recognized and investigated by EBASCO concerned: (1) red line "as-built" drawing control and installation problems (for example, CARS 019, 024, and 025);-(2) craft personnel that were not following procedures (for example, CARS 020, 021, and 022); and (3) failure to inspect welds because craft personnel were ignoring hold points and because of QC failure to inspect weld joint fit ups (for example, CARS 027,034,and061). EBASCO Audit Program Observations The ESASCO audit and corrective action systems did not function properly in that EBASCO audits NB-79-9-4, NB-80-8-3, and NB-81-5-1, performed in 1979, 1980, and 1981, stated that Mercury audited each section of the QA manual, despite the fact that the Mercury CAR log and two or more audits stated tha't internal audits were not conducted annoally on each section of the Mercury QA Manual . EBASCO Audit SW-82-6-1 (June 7, 1982) finally identified Mercury's failure to audit their QA program. Other additional problems were identified in the above EBASCO Audits which should have alerted management to problems with Mercury as early as September 1979. It is not apparent that any action was taken as a result
- of the issues found during the audits.
LP&L Management Audits of the Site CA Program The NRC staff reviewed these audits to determine if LP&L had regularly reviewed the status and adecuacy of the program on site. The audits performed at the site between 1972 and August 1979 were not very comprehensive. They were essentially document reviews. Audits after this were more comprehensive, but were not thorough enough to cover all elements of the QA Manual and implementing procedures. Site Audits Scheduled Versus Audits Performed The NRC staff reviewed LP&L audits scheduled and performed from 1979 through 1984. The following table lists the audits scheduled versus the audits performed: Year. LP&L Site Auditors Scheduled Performed 1979 4 193 80 1980 4 181 60 1981 4 178 53 1982 5 203 91
. 1983 5 168 42 TOTALS 923 326 44
In justifying this discrepancy, LP&L representatives said that audits were not performed because LP&L QA Engineers and technicians were routinel SystemSUS) (y diverted reviews,to support unannounced Construction NRC inspections, Deficiency Reporting, and other Start-Up unscheduled work. The existing manpower was insufficient to meet both routine auditing and special task assignments. An LP&L representative stated that there was no requirement for LP&L to conduct these routine audits. The NRC staff found that this issue was identified as an audit system deficiency in 1979 during a management audit for LP&L by Middle South Services (MSS) (audit reports E79-6 (AFR E-88) (August 27 to August 31, 1979), and E80-15 (AFR E-122) (December 8 to December 16,1980). Corrective action was apparently not taken and it appears that no further action would have been taken to complete the scheduled audits since MSS Audit Report E81-12 reported the conduct of QA Audits and QA Construction site audits as " satisfactory" when only 53 of 178 of those scheduled were perfomed. Similarly, Audit Report E82-15 (December 6 to December 22, 1982) considered it normal when 91 of 203 scheduled audits were conducted. (Those perfomed may be fewer than those scheduled since these figures include reaudits and may also include QA surveillances, which are not to be considered os audits per ANSI N45.2.12.) This condition was discussed with the LP&L Site QA Manager, who did not consider these findings to be a problem. LP&L Audit of EBASCO OA Program The NRC staff reviewed LP&L audits of EBASCO site activities performed in 1981, 1983, and 1984. Nine audits were conducted in 1981, and seven in 1983, none of which were sufficiently comprehensive to meet the requirements of 10 CFR 50, Appendix B and LP&L QP 18.3, Revision 3. No
- audits were conducted during the first several months of 1984, which according to LP&L representatives was because of their need to support the NRC CAT inspection and due to NRC Waterford Task Force effort.
Conclusion The NRC staff reviewed the-Mercury audit program and determined that it failed to audit some key quality control and construction procedures , where deficiencies occurred. Proper auditing of these procedures from 1978 through 1983 may have prevented the deficiencies identified against the activities of the Mercury Company.
- In comparing procedures which had been audited with those that had not, there appeared to be a correlation between documentation problems and hardware deficiencies at Mercury Company, in addition to several significant breakdowns in the Mercury QA program, specifically those for the audit, nonconformance and corrective action systems. The NRC staff requested LP&L management to furnish existing documents which would demonstrate what contractor procedures had been audited by LP&L EBASCO, and other site contractors. the LP&L site QA manager and QA construction 45
engineer stated that they did not know what procedures had been audited
. during plant construction. LP&L representatives stated that they did not
. manage the audit function in this manner, but simply monitored auditing activities in a broad sense, and that there was no requirement to know
. what specific procedures were audited.
The NRC staff considers this failure to audit Mercury implementing procedures to be an issue that should have been identified in LP&L's corrective action response to the proposed civil penalty identified in hRC Inspection Report No. 50-382/82-14, issued December 6,1982. LP&L's corrective action was inadequate in that LP&L did not thoroughly evaluate, determine all the causes, and identify the entire extent of the QA breakdown at Mercury Company. As indicated above, LP&L, through EBASCO, had several opportunities.to identify and correct the problems associated with Mercury. The long held policy of LP&L has been to conduct their business related to engineering, construction, and operation of power plants by utilizing a very ' lean' in-house organization with almost total reliance on the Architect-Engineer (AE) constructor for engineering and construction. This policy of ' lean-ness' and almost total reliance on the AE/ constructor is one of the root causes of the problems associated with Waterford 3.
. This problem was previously identified by outside consultants but was not acted on. LP&L staff was not sufficiently involved in the planning, monitoring, and control for site activities and site organizations were lacking in numbers and in comercial nuclear power plant experience necessary to effectively monitor and control the Waterford 3 project.
In conclusion, the NRC staff found that there was a breakdown in the QA programs at EBASCO and Mercury. Also, a breakdown of the LP&L QA program was indicationsd in that: (1) LP&L did not thoroughly evaluate, determine the root cause, and take effective corrective action to preclude recurrence of the identified problems; and (2) LP&L did not take action to implement the recommendations of its consultants and the NRC to increase its manpower and invcivement with the Waterford 3 Project. LP&L's failure to effectively implement their QA Program has potential safety significance and the inadequate management controls which led to this QA breakdown, has generic implications on the question of management's ability to safely operate the Waterford 3 facility. Other NRC Task force findings identified in this SSER are further indications of the QA program breakdown between EBASCO and Mercury and are indicative of a breakdown of the LP&L QA program. LP&L has been requested to address this issue. Action Recuired: See Item No. 23 in the enclosure to the D. Eisenhut letter of June 13, 1984, to J. M. Cain (LP&L). 46
Task: Allegations A-49; A-78; A-87; A-123 Characterization: It is alleged that individuals were prevented from writing NCRs or forced to rewrite specific NCRs and that subsequent quality assurance (QA) documentation may be falsified. Assessment of Allecation: The NRC staff assessed the technical aspects of fcur specific allegations. Ine implied significance of these allegations is that identification and correction of nonconformances may have been curtailed and may have impacted installed systems. Also, inadequate hydrostatic testing may have been performed on safety systems, which might not have been verified if QA documentation was falsified. Allegation A-49 alleged that individuals were not free to write NCRs. The NRC staff assessed this allegation in conjunction with Allegation A-283 by reviewing procedures utilized by Mercury, Tompkins-Beckwith (T-8), and E8ASCO concerning the reporting, disposition, and correction of discrepancies, deficiencies, and nonconforming items. The NRC staff found that adequate procedures existed to allow employees to report apparent abnonnalities in facility construction. (See Allegations A-53; A-93.) The procedures specified that supervisors and management were e responsible for specifying the processing and disposition of items, and that the emphasis was that other personnel were also responsible for reporting items. The NRC staff found no indications, other than the allegers testimony, to indicate that contractor management exerted pressure on individuals or otherwise suppressed the writing of NCRs. The staff review revealed that the n'onconforming conditions have been subsequently entered into the NCR system and the issues resolved. This allegation has neither safety significance nor generic implications. Allegation A-78 alleged that NCRs written by Mercury quality control (QC) inspectors on welding problems were not processed by QA/QC supervisors or management, and that documentation was subsequently falsified to reflect the absence of nonconfonning items. The NRC staff evaluated this allegation in conjunction with Allegations A-81 and A-82 by reviewing data packages for 28 different Mercury Operational Control Records (OCRs) on various instrumentation lines. The staff also performed system walkdowns on 19 of these installations. Both the data packages and the systems walked down were found to be correct. It should be noted that procedures on the disposition of welding discrepancies generally stated that weld defects that are in the process of inspection or repair were not required to be reported as nonconfonning items, discrepancies, or deficiencies and that only the final inspection was required to be
.. documented. The NRC staff agrees that this has been acceptable industry practice. Nonconformance reporting procedures also stated that reporting was a supervisory / management decisien. The NRC staff found these reporting procedures to be adequate.
47
The NRC staff found the installations to be correct and the procedural requirements to be adequate. This allegation has neither safety significance nor generic implications.
. . Allecation A-87 alleged that a QA engineer assigned to Mercury, but employed by ESAsco,.was forced by EBASCO QA management to rewrite an NCR.
The NRC staff reviewed this allegation by evaluating NCRs W3-6719, W3-6719/R1, and an unnumbered NCR. It was noted that the unnumbered NCR was dated August 17, 1983. The original NCR W3-6719 was dated August 17, 1983, and closed out on August 22, 1983, with an EBASCO note stat.ing that the QA engineer's concerns were invalid due to the generalities voiced. This NCR was , subsequently " voided" and reissued as Revision 1. NCR W3-6719/R1 was issued on August 23, 1983. The staff could not determine if the individual had been
" forced" to rewrite the NCR. The NRC staff reviewed NCR W3-6719/R1, which was closed out on October 13, 1983. The above hydrostatic test conditions were assumed by EBASCO to be the " worst case" and therefore that "all" other hydrostatic tests performed by Mercury were deemed satisfactory. This was not the case, since only one test was reviewed by EBASCO.
Due to the items.noted above, the NRC staff determined that EBASCO QA management had improperly dispositioned and closed NCR-6719/RI. Allegation A-123 alleged that ESASCO QA records reviewers working on Mercury
- records were not allowed to look in the field because they.were finding too many problems, and that reviewers working with T-B encountered similar problems.
The NRC staff reviewed EBASCO QA procedures which defined document review responsibilities. The staff's review revealed that those procedures adequately l addressed the responsibilities of QA records reviewers, and that the procedures 1 provided detailed instructions for the collection, handling, status, and review of QA records for construction and installation. The procedural requirement for field verification was established in EBASCO procedure QAI-9 which stated,
" Concerns regarding discrepancies in plant configuration shall be brought to the attention of the .EBASCO QA Surveillance Group for action." This statement was not a requirement making it necessary for a QA records reviewer to go into the field.
EBASCO QA/QC review and verification groups were available to the records reviewers for any field work. The NRC staff investigated this allegation l in conjunction with Allegation A-283. This allegation has neither safety significance nor generic implications. i. Actions Required: With regard to Allegation A-87 (NCR W3-6719/RI) see Item No. 6 in the enclosure to the D. Eisenhut letter to J. M. Cain
- - dated June 13, 1984 ,
l- .. 3 9 48
Task: Allegations A-53 and A-93 Characterization: It is alleged that identifying and clearing problems ~was difficult cue to an "almost non-existent" program, and the Document Reviewers
- had no way of reporting problems.
Assessment of Allecation: The implied significance of this allegation is that nonconforming items may not have been identified or corrected. The NRC staff reviewed the EBASCO and sub-contractor's procedures for nonconformance identification and found that an adequate system did exist. The allegation, however, alludes to the existence of a system " road block" that prevented the processin (See Allegations A-18 and A-164.)g of nonconformances The NRC staff did find indications and discrepancies. of NCRs that were " Voided" or "Could Not Be Located" (see Allegations A-18 and A-232). A total of 33 NCRs were voided, and 12 could not be located in the EBASCO card index. Ten other NCRs listed on the card index could not be located in the EBASCO QA Vault. One NCR in the OA vault was not on the card index. The ten
" voided" and " missing" NCRs are the only indications of problems with NCR processing that the NRC staff detected during this review. The basis for these allegations may have existed at the time the allegation was made, but the staff could find no indications that the problem existed at the time of this NRC review.
- This allegation has neither safety significance nor generic implications.
Actions Required: None. e 49 b= - --- __ - ,_ . _ _ , .
, -.,y
Task: Allegation A-54 Characterization: It is alleged that there was a failure by EBASCO and site suocontractors to implement their procedures. Assessment of Allegation: The implied significance of this allegation is that the lacx of proceoural implementation could cause installations and inspections to be invalid and thereby render the quality of installation as indeterminate. The NRC staff assessed this allegation to determine if failure to implement procedures was generic (site wide). The staff: (1) examined EBASCO's program for controlling the preparation, approval, and revision of site procedures; (2) walked through EBASCO and Mercury procedural interface; (3) verified that
~
NISCO, Tompkins-Beckwith (T-8) and Mercury reviewed turnover quality records in accordance with procedural requirements; and (4) examined the corrective action mechanism for enforcing procedural implementation. EBASCO Program for Controlling Site Procedures EBASCO's program for procedure review and approval includes procedures generated by EBASCO engineering and all site contractors. The site quality assurance engineering department and/or other EBASCO disciplines, as required, reviewed procedures affecting quality prior to implementation. Regarding EBASCO's e review and approval of site procedures, see Allegations A-188, A-190, A-191, and A-193.
< Mercury /EBASCO Procedural Interface The NRC staff obtained a copy of EBASCO's review, approval, and coments for the initial and current revision of Mercury's QA Records and Control Procedure (QCP-3010). The objective was to verify that EBASCO had reviewed and approved Mercury's procedure prior to implementation. These documents contained recorded coments with resolution, and approval granted to Mercury for procedure issuance and implementation.
l A number of Mercury turnover QA documentation packages for the Reactor Coolant instrument lines were reviewed by the NRC staff. These packages had been l reviewed and approved by Mercury's QA Document. Review personnel using procedure QCP-3010. EBASCO QAIRG also reviewed and approved these packages using proce-dures QAI 9 and supplement QAIRG No. 15. The results of package and other Mercury documentation reviews are discussed in Allegations A-308, A-183, A-184, and A-197. Mercury, NISCO and T-B Turnover 0A Records Review The NRC staff reviewed startup system 52A, Reactor Coolant, to verify generic procedural implementation for turnover QA records. Mercury, NISCO and T-B performed QA records reviews in accordance with approved procedures. EBASCO
.. QAIRG conducted a 100 percent review of contractor documentation packages.
LP&L used a 10 percent sampling plan, based upon Mil-STD-105D, to review and audit turnover packages. 50 l i
. ~ . - - - - - - - - . - .
NRC staff observed that:
- 1. Quality record reviews by contractors, EBASCO, and LP&L were adequately documented (stamped, signed, and dated).
- 2. Structured checklists were used to designate the status of records and nonconformance actions and to accept records.
- 3. Reviews were conducted for technical adequacy, completeness, proper form, legibility and authenticity.
- 4. Inspection / test results were documented in detail and traceable to installed hardware.
Fo'r additional detail of Mercury, NISCO, T-B, EBASCO, and LP&L review of SUS turnover QA records, see writeups for Allegations A-35, A-143, A-150, A-162, A-163, and A-308.
. Corrective Action:
When, in the judgement of contractor QA/QC personnel, a nonconformance required corrective action, or a recurring condition existed, a request for Corrective Action Report (CAR) was initiated. Corrective action included:
- 1. Determination of cause.
- 2. Recommended course to prevent recurrence.
- 3. Implementation. -
- 4. Follow-up to verify effectiveness.
EBASCO performed inprocess audits (surveillance) of construction activities. The NRC staff reviewed a number of surveillance report results. Included in these audits were the evaluation of contractor nonconformance reporting and corrective action to prevent recurrence.- NRC noted that corrective action was used to enforce procedural implementation. Allegation A-186(a) details NRC staff review of EBASCO surveillance reports. Additionally, during the NRC review of all allegations, numerous procedures were reviewed and found to be adequate. In conclusion, based upon information reviewed, the NRC staff found that site procedural implementation was verified as adequate. Therefore, this allegation has neither safety significance nor generic implications. Actions Required: None. i . i 51 i
- Task: Allegations A-57 and A-314 Characterization: It was alleged that Mercury Construction Company Level III inspection personnel had no certifications to inspect at the Waterford site and that two Level III in pectors, the Corporate QA Manager and the training officer, were not qualified to be certified due to insufficient educaticn and . previous experience.
Assessment of Allecation: This allegation implies that the Level III inspection personnel were not certified and that two Level III inspectors were not qualified due to insufficient education and experience. If true, this could affect the certification of Level I and Level II inspection personnel performing safety-related inspections, thereby rendering the verification of the quality of safety-related systems indeterminate. This item was addressed by NRC staff review of Level III inspection personnel packages, certifications, and resumes. The NRC staff found that 13 Level III inspection personnel at Waterford had certification records on file. The certifications did not mention the Waterford site specifically, neither did the certifications for all levels of Mercury inspectors. The Level III inspectors were found to have been certified. The NRC staff reviewed the certifications and supporting documentation for all site Level III inspection personnel, including the two in question. The certifications of three Level III personnel reveal that they lack sufficient prior experience, although the certifications for the two inspectors in question were found to be acceptable. This item has potential safety significance. The questionable certifications will be tracked with Allegation A-02. Actions Recuired: See Item No. 1 in the enclosure to the D. Eisenhut letter of June 13, 1984, to J. M. Cain (LP&L). i l l l l
~
1 l l 52 l
Task: Allegation A-58 Characterization: It"is alleged that many permanent plant items were not being properly maintained by sub-contractors, EBASCO, or LP&L and that the items are
. - rusting in place, even though such maintenance was the responsibility of LP&L and its contractors.
Assessment of Allecation: The irrolied safety significance of this allegation is that permanent plant equipment may have been damaged due to lack of preventive maintenance. During the review of this allegation, the NRC staff noted that LP&L, EBASCO, and other sub-contractors had written care and maintenance procedures. These procedures-appeared to be in accordance with ANSI N45.2.2, to which the licensee was committed. The allegation stated that tube track, hangers, supports etc., were not maintained by Mercury or EBASCO. The NRC staff's detailed review of LP&L and EBASCO audits indicated that a basic maintenance program was utilized during construction. LP&L and EBASCO audit reports further indicated that Mercury had some deficiencies in their care and maintenance program. These deficiencies were cited in the audits and corrective action was initiated. These were no nonconforming conditions and the deficiencies did not require generation of an NCR. Galvanized tube tracks, sway struts, and structural items were classified as storage Level D items. Level D items could be stored outdoors in a designated area which was well drained and marked for storage. These items were to be stored on cribbing.or equivalent means to allow air circulation and do not require a special care and maintenance procedure. They were inspected on an area basis, with the inspection documented by Mercury on Form 239, Material Handling and Storage Report. , Mechanical and hydraulic snubbers were classified as Level C storage items. Level C items require indoor or equivalent storage. These supports did have a special care and maintenance procedure. A sample review of EBASCO's main warehouse QC care and maintenance reports indicated that from February 1977 to September 1983 monthly reports were completed for care and maintenance of those items. To address the problem of rusting in the field, memo #F-55764E, dated October 14, 1981, was issued requiring all onsite contractors to notify EBASCO Services of any permanent materials requiring painting prior to installation in the reactor containment building. j i NCR W3-6940 documented a good example of the concerns expressed in this allegation. The NCR stated that " Indications are the items identified in the cc,mponent description above (Bergen-Patersr.n Hangers issued to Mercury) were degraded by the Mercury Company as a result of storage and handling practices and techniques." An NRC staff review of the above statement indicated that the supports in question were either sway struts or mechanical snubbers. Mechanical snubbers were Level C items and were stored in a warehouse. They were only
. requested from the warehouse prior to installation, so that there was little 1
chance for their degradation. An EBASCO representative also stated that Mercury ' 53
~
- did not install any mechanical snubbers. Rigid supports such'as sway struts were installed by Mercury. These were Level D items and required very little maintenance, except to ensure that the grease preservatives were still effective. - This was done during care and maintenance inspections performed by EBASCO and Mercu ry.
NCR W3-6940 also stated that " Inasmuch as there is no documentation (Form (s) 239) available, it is indeterminate as to whether the Mercury Company performed any final location inspection of items identified in the description above." A detailed examination of the issues raised in this statement by the NRC staff indicated that the review of the EBASCO records review group was done after Mercury left the site. Mercury in fact did not leave records (Form (s) 239) for permanent plant files. However, ANSI N45.2.2, paragraph 6.6, " Storage Records," does not require care and maintenance inspection results to be kept as permanent plant records. Contract specification W3-NY-15 did not require these records to be kept as permanent plant records. Additionally, the NRC staff reviewed preventative maintenance performed on safety-related rotating shaft equipment, tanks, and instruments. EBASCO prepared Construction Maintenance Instructions for this equipment and issued these instructions to the responsible contractors.
< Responsibility for preventative maintenance was transferred to LP&L at the time of equipment turnover. The NRC review of contractor maintenance records revealed an adequate history of preventative maintenance, from receipt through e installation to equipment turnover. LP&L maintenance records were also reviewed and demonstrated a continuation of adequate preventative maintenance following equipment turnover. The review for contractor-controlled equipment included:
low pressure safety injection and high pressure safety injection pumps and motors, emergency diesel generators, emergency feedwater pumps and motors, and shutdown heat exchangers. The review for equipment currently controlled under the LP&L system included: low pressure safety injection pumps and motors, various pressure indicators, and transmitters. The safety significance of this issue would be minimal because even without inplace inspection of supports and hangers, existing onsite procedures ASP-IV-121 and 138 were being implemented as final checks to assure that supports and restraints for the piping systems were properly installed and would perform their intended function. In conclusion, LP&L and EBASCO surveillance audits indicated that Mercury did perform care and maintenance inspections. The NRC inspection of plant safety-related items, whether in storage or installed revealed no discrepancies. This allegation had no safety significance and no generic implications. Actions Required: None. 9 54
i i Task: Allegation A-59
~ '
Characterization: The allegation is that there was little control over nonconforming conditions, in that no Hold Tags were placed on items in the
- - field by EBASCO.
Assessment of Allecation: The implied significance of this allegation is that nonconforming materials, parts, or components may not have been properly identified to prevent their inadvertent use or installation. The individual who wrote the letter raising this concern is unknown to the NRC staff. On December 27, 1983, the Senior Resident Inspector at Waterford brought this concern to the attention of LP&L QA for action in accordance with NRC policy for handling technical allegations at that time. LP&L responded that EBASCO diJ not replace Mercury Construction Company Hold Tags that were hung in the field when EBASCO assumed the responsibility for Mercury's contract. EBASCO did renumber approximately 80 percent of the Mercury Hold Tags with the EBASCO Noncompliance Report (NCR) numbers. The remaining 20 percent of the hold. tags were not required by EBASCO procedure. The NRC staff noted that not all NCRs required the use of Hold Tags. The EBASCO procedure for Processing of Nonconformances, Number ASP-III-7, Issue X, stated that when a Discrepancy Notice (DN) was upgraded to an NCR, the NCP, number was added to the DN tags. If an NCR was generated on a discrepant hardware item not previously documented, a Hold Tag was to be prepared. Hold Tags were not issued for NCRs involving documentation deficiencies. The NRC staff found that there is no Hold Tag Index, but that the Hold Tags are indicated on the open NCR status card file. The NRC staff obtained this information from a review of the EBASCO NCR procedure, and interviews with LP8L and EBASCO QA/QC personnel. This allegation has neither safety significance nor generic implication. Actions Required: None. l I I \ i i 55 l l
Task: Allegation A-63 j Cha racterization: The allegation is that equipment was not properly aligned or i levelled when it was installed. j Assessment of Allecation: The implied safety significance of this allegation is that sarety-relatea rotating eouipment with improper alignment will not continue to function as designed. In assessing this allegation, the NRC staff reviewed EBASCO Specification MC-1, General Specification Covering Installation of Mechanical Equipment, dated April,26, 1976, and Gulf Engineering Company Procedure PR 10.0, Traveler Sys. tem,' dated December 10, 1982. The NRC staff found that these installation requirements and controls were adequate. Additionally, the NRC staff reviewed the installation records (travelers) for the following equipment: Emergency diesel generators High pressure safety injection pumps Low pressure safety injection pumps Charging pumps Emergency feedwater pumps Reactor coolant pumps The installation records were found to be complete and acceptable, including QC
- inspection results.
The NRC staff found that EBASCO Specification MC-1 was followed unless the manufacturer's requirements were more stringent or contained special require-ments; in that case, those requirements were complied with. For example, the reactor coolant pumps were installed in accordance with the requirements established by the Nuclear Steam System Supplier, Combustion Engineering, and the pump manufacturer, Byron Jackson. Additionally, the staff observed a low pressure safety injection pump in operation. The pump and motor were operating with virtually no vibration, indicative of proper setting and alignment. Accordingly, this item has no safety significance and is not indicative of any generic implications or management problems. Actions Reouired: None. 1 e e l 56
Task: Allegation A-64; A-65; A-66; A-322
, '. Characterization: The allegation is that EBASCO site letters F-63724E, F-61895E, and F-61147E imply that Mercury inspection activities may be unacceptable.
Assessment of Allegation: The implied significance of this allegation is that unacceptacle inspections by Mercury could have permitted invalid installations to have gone undetected. The NRC staff reviewed the letters identified in this allegation. .The staff also reviewed Mercury letter WA-2065, February 25, 1983; E8ASCO letter F-61120E, January 13, 1983; and Mercury letter WA-2014, December 21, 1983. The numerous solutions called for in these documents were implemented, and resulted in the issuance of approximately 100 nonconformance reports (NCR) and discrepancy notices (DN). As a result of the reinspections required by the above letters, and the issuance of two Significant Construction Deficiencies (SCDs), SCD-57 and SCD-84. The staff concluded that sufficient inspection was performed. The NRC staff concluded that the problems identified in the EBASCO letters were subsequently identified in the EBASCO nonconformance system. EBASCO has performed the necessary additional work and inspections to bring the Mercury installation work up to an acceptable level. The NRC staff found this action to be acceptable. This allegation therefore has neither safety significance nor generic implications. Actions Recuired: None. 6 e e d 57
Task: Allegations A-81 and A-82 ~~ Characteriza' tion: The allegation is that changes were made to N-5 accumentation packages for instrument lines without the knowledge of the .. authorized nuclear inspector (ANI) after his approval signature had been placed on the package. Assessment of Allecation: The implied significance is that the requirements of the ASME Code, Section III, NA-8000, were not complied with because the completed Code Data Reports were. incorrect. The NRC staff reviewed N-5 data packages for 28 different instrument lines. There was no indications of changes to N-5 data after the ANI's approval. The NRC did not interview Mercury's ANIS, as they were not available onsite. The staff did, however, interview the EBASCO ANI who was not aware of any changes made to N-5 packages after the Mercury ANI had approved them. Three instrument lines were found to have conflicts between the drawings, weld records, and N-5 data. DPT-RC-120/121 (low pressure) N-5 data showed field welds FW-1C, 10, 51, 61, and 31 which were not on the drawings or final weld records. Field welds FW-47, 56, 57, 72, and 73 were on the drawings and weld records but were not on the N-5 Data Report. DPT-RC-120/121 (high pressure) had FW-30 and 31 on drawings and final weld records but were not on the N-5 Data Report. DPT-RC-124Y (high pressure) showed FW-15 on the N-5 data form but not on the drawing or the final weld records. The NRC staff conducted a walkdown of the three lines. The
- walkdown revealed that the drawings accurately reflected the installed system. ~.
LP&L initiated a Site Surveillance Report to document and follow-up on these discrepancies. The NRC staff found no indications of changes to N-5 data having been made after the ANI's approval, and the staff's walkdown of the three lines that had deficiencies in the N-5 data revealed that the hardware ~ installed was accurately reflected on the drawings and supporting QA records. The NRC staff concludes that this allegation has neither safety significance nor generic implications. Actions Reouired: None. eg 6 58
l
. Task: Allegation A-92 Characterization: The allegation is that EBASCO record clerks closing out
. document deficiency reports (DRs) were not qualified to do so.
Assessment of Allecation: The implied significance of the allegation is that unqualified personnel may have closed ors for quality assurance (QA) records of safety-related systems which could be adversely affected. The NRC staff reviewed EBASCO's records review system, training requirements . for records reviewers, training files of records reviewers, and a sampling of closed out DRs. The NRC staff found that EBASCO procedures required records reviewers to have a high school education, classroom training, on-the-job training (0JT), and reading assignments. The EBASCO system required document ors to be initiated and closed out by records reviewers. The NRC staff selected a random sample of work packages (QA records), and reviewed the records of 27 individuals who signed off document DRs in those work packages. Nine of the 27 individuals (33 percent) signed off document DRs without having completed the records reviewer training requi-ed by EBASCO. The problem of inccmplete training records for records reviewers, in particular OJT completion, was addressed by an EBASCO audit in October 1983. Subsequent e action updated and completed those records or addressed open items, and the subject audit was closed in January 1984 The NRC staff reviewed that audit and concurred with the disposition action and closure. The NRC staff concluded that there were no specific training qualifications or certification requirements for document reviewers in NRC regulations, ASME Code, or ANSI Standards. The EBASCO training program appeared to be adequate. The NRC staff found that document DRs were closed by individuals who had not completed EBASCO records reviewer training requirements. However, corrective action by EBASCO satisfactorily addressed the problem, and work performed by records reviewers was considered satisfactory by the NRC staff. This allegation has neither safety significance nor generic implications. Actions Required: None. Gg l 59 ,
1
' - Task: Allegations A-94 through A-99 4
Characterization: There are six miscellaneous allegations relating to an i
- unidentified nonconformance report (NCR) dated September 9,1983, which j
makes reference to various problems with stainless steel piping and tubing. l Assessment of Allegation: The implied significance of this allegation is that ' i improper reporting and corrective action of NCRs, missing quality assurance 1 (QA) records, lack of access to necessary documentation and key personnel, and the lack of procedures could cause the QA process to be compromised and place [ the quality of construction in question. e
! The NRC staff reviewed these six allegations as follows:
! 1. Allegation A It was alleged that an individual had on fi.le an* . 2 unnumbered NCR and complete documentation on how EBASCO would not allow people to report problems. A copy of the NCR was obtained from LP&L and
- was identified as EBASCO NCR No. W3-6943. This NCR therefore had been
- properly entered into the EBASCO system.
- 2. Allegation A It was alleged that Attachment 1 of the " unidentified" NCR (W3-6943) states that two heat numbers had been used for three l .- different schedules of pipe tubing. The NCR disposition was to:
1 * (1) "... provide a listing of suspect pipe runs..."; (2) "...QC to take UT 1 measurements on suspected runs..."; and (3) "... evaluate any deficiencies
- .- reported by QC..." The purpose of this inspection was to detemine if the proper schedule of material had been installed. The results of the UT inspection revealed that schedule 160 piping had been installed in the appropriate systems. This corrective action was acceptable to the NRC
- staff.
- 3. Allegation A It was alleged that Mercury Construction Company took a
. trailer full of documents to their home office. The removal of the j documents was contrary to LP&L's commitment to ANSI N 45.2.9; however, the NRC staff determined that these records had been returned to the site.
LP&L provided all Mercury records requested by the NRC staff during this i review, even though the records system was cumbersorhe and retrieval was l not always timely. 4 Allegation A The allegation is that EBASCO reviewers were not allowed l access to heat number records and could not research the extent of the. problems. Six EBASCO document reviewers were interviewed and all of them stated they had no problem obtaining access to Mercury records. The site QC' supervisor of the EBASCO Verification Group stated emphatically that inspectors working for him physically checked tubing for heat numbers upon
- proper request from the document reviewers. There was no indications of EBASCO personnel being denied access to Mercury records. Although this -
, . allegation may have existed prior to and at the time of the allegation, .
- access was not a current problem and all QA records requested by the NRC
- staff were available. !
p. i i 60
- ~ - - - -
-6 o
- 5. Allecation A The allegation is that Mercury Authorized Nuclear Inspectors (ANIS) were off limits to reviewers. The six EBASCO document reviewers interviewed, who were on site at the time of this allegation, agreed they had ready access to Mercury ANIS. There were no indications that the reviewers were denied access to the ANIS, although this situation may have existed prior to or during the allegation. Mercury ANIS were not available on site to be interviewed by the NRC staff.
- 6. Allegation A The allegation is that there are no procedures for review of occuments. The NRC staff reviewed the following procedures and found them to be adequate for review of documents: EBASCO Procedure QAI No. 9, " Review and Handling of Construction-Installation Records,"
original issue dated October 31, 1979, current revision issued April 20, 1983; EBASCO Procedure QAI No. 9A, " Documentation.Statusing Review Instruction," dated December 13, 1982; Tompkins-Beckwith Procedure T8P-20, "QA Records Turnover," dated February 7,1983; and Mercury Company Procedure QPC-3010, " Quality Assurance Records Control," original issue dated September 7, 1978, current revision issued November 1, 1978. Although the use of supplemental procedures related to a particular contractor could be a improvement over the base procedure, the base procedure (QAI-9) was deemed adequate. Additionally, contractors' documentation packages were reviewed by the NRC staff and generally found acceptable; see Allegations A-143, A-150, A-162,
- and A-163.
The NRC staff has determined that these six allegations had neither safety
- significance nor generic implications.
Actions Reouired: None. m 9 61
Task: Allegation A-100, A-101,.A-102,'A-103, A-127, A-128a, A-128c, A-285a Cha racterization': The allegation is that the Mercury Construction Company had
. the following material traceability problems: (1) in one case a Mercury QA Supervisor directed clerical personnel to improperly add valid heat numbers to records when missing heat number were discovered; and (2) numerous cases where the original heat numbers were not transferred to the tubing when it was cut.
Assessment of Allegation: The implied significance of this' allegation is that the records for material may not accurately reflect what is actually installed in the field, a condition:which could affect the acceptability of safety-related < systems. The NRC staf'f addressed this issue by reviewing Mercury work packages for indications of added or changed hea't numbers. After examining a random sample of heat numbers for tubing, fittings, valves, and filler material from the j packages, the staff found that the records were adequate to show the acceptability of the material for its intended use. The Mercury weld records were difficult to follow because of crossouts, weld deletions, additions, and a poor system of recording applicable notes. The records consisted of photocopied material and original pen-and-ink data. Falsification of heat numbers could not be determined from the staff review conducted. The NRC staff also walked down 19 installations to verify heat number markings and found the numbers to be correct. Mercury QC and document review personnel involved in material traceability problems are no longer on site and were not interviewed. Numerous NCRs and DRs were issued on material traceability problems. The following are typical examples. NCR-W3-4593 - This NCR identified Mercury installed tubing which did not have heat number traceability. To resolve this NCR, five samples of installed tubing were cut and sent to a testing laboratory for analysis. The analysis l i revealed that the tubing was SA-213 Type 316, which was acceptable. The NCR was then closed -and the NRC staff considers this action acceptable. l l p NCR-W3-6943 and W3-7538 - These NCRs identified the loss of heat number l traceability by Mercury, which caused a concern as to whether the thin wall tubing had been installed in the system requiring heavy wall tubing. To resolve this NCP, the tubing was ultrasonically tested (UT). The UT identified
,that heavy wall tubing had been installed in the appropriate system and this NCR was closed. The NRC staff considers this action acceptable.
~
NCR W3-3919 - This NCR was initiated because of a tubing crack discovered during a Reactor Coolant hydrostatic pressure test. This NCR resulted in
, .SCD #61 being issued. This NCR and SCD should be reopened because the testing and disposition were invalid. For details See Allegation A-33.
Discrepancy Report - An EBASCO quality control review turned up discrepancies in heat numoers for tubing unions between Mercury inspection records and system isometric drawings and issued a discrepancy report. The discrepancy report initiated Field Verification Inspection Requests to verffy heat number l r 62
- discrepancies and to correct Mercury inspection records to substantiate that
. ~~
- installed hardware is accurately depicted on system isometric drawings. EBASCO QC made over 4800 such comparisons in correcting these discrepancies. The NRC staff review found this action acceptable.
In conclusion, the NRC staff found that Mercury's heat number traceability to installed material was adequate. Therefore, this' allegation has neither safety significance nor generic implications. However, Allegation A-33 did identify some problems with heat number traceability. Actions Required: None. 6 4 4 6 63
. Task: Allegation A-117, A-120
. Characterization: The allegations are that Mercury and Tompkins-Beckwith (T-8)
. records were falsified "to whitewash what may be serious construction defects" and " doctored to give them the appearance of compliance with federal safety regulations."
Assessment of Allecation: -The implied significance of the allegations is that if records of safety-related systems have been falsified the quality of those systems are indeterminate. This issue was addressed by an NRC staff review of the T-B and Mercury document review system and the qualification of record reviewers'. The staff also >- reviewed a sample of documents specifically for unauthorized changes, dates out of sequence, incorrect data, or the appearance of falsification. The NRC staff found that T-B and Mercury procedures provide guidelines for changes or corrections to-documents. Documents were reviewed by subcontractors including T-B and Mercury quality assurance (QA), turned over to EBASCO for an additional review by QAIRG, and then submitted to EBASCO for records storage. The NRC staff did an extensive review of documents submitted by T-B and Mercury as well as other subcontractors. One possible falsification of a Mercury employee's signature on a nonconformance report (NCR) was observed and was referred to the NRC Office of Investigation (0I) for follow up. No other
.- ' apparent falsification of records was noted.
The T-B and Mercury record review systems were considered adequate. Personnel performing record reviews are considered qualified. Previous instances of falsified documents were properly corrected by EBASCO (See Allegation A-07). The one possible falsified signature discussed above appeared to be an isolated case. .In conclusion, this allegation has neither safety significance nor generic implications. Actions Required: None. I e e 4 64-
. \
. - . , - . - . - . - , , - . _ - . . , ..- - _ - - ,,n_ _ . . . . n._ - . _ _ . _ _ - , . _ _ _ _ _ _ _ _ - , - . - , . - . , . . .,
Task: Allegation A-143, A-150, A-162, A-163 Characterization: The allegation is that LP&L, EBASCO, and contractors did not perform adequate reviews of QA records. Assessment of Allecation: Inadequate review of documentation to assure
* ^
completeness ano tecnnical adequacy of QA records could cause the quality of installation work to be in question. The procedures delineating guidelines for the collection, handling, and review of construction-installation quality assurance (QA) records for NISCO, Mercury, Tompkins-Bechwith.(T-B), EBASCO, and LP&L have been evaluated by the NRC staff for compliance with ANSI N45.2.9 and ASME requirements, and found acceptable. To verify procedural implementation, the NRC staff reviewed a random sample of work packages for each company. The startup systems (SUS) sampled included 36-Component Cooling, 52-Reactor Cooling, and 60-Safety Injection. The hardware reviewed included large and small-bore piping, and instrumentation and control lines. The NRC staff reviewed the documentation packages to verify completeness, accuracy of content, proper form, traceability, legibility, authenticity, and pertinent changes. The Staff also conducted system walkdowns. The NRC staff made the following observations:
' (1) QA records review, acceptance and approval was adequately documented (stamping, signature, and dating).
(2) Technical adequacy was acceptable. (3) Records were complete. (4) Corrections were performed in accordance with prescribed procedures. (5) Record retrievability was timely. (6) Auditability was adequate. For additional information regarding 0A record review by contractors, EBASCO and LP&L, see Allegations A-35 and A-308. Technical records review versus records accountability is addressed in Allegation A-186(b). The NRC staff concludes that the contractors, EBASCO, and LP&L have implemented adequate documentation requirements, including reviews. Therefore, Allegations A-143, A-150, A-162, and A-163 have no safety significance, nor generic implications. Actions Required: None. 9 l e 9 65 4
.'. Task: Allegations A-165, A-166, A-292, A-293, A-297, A-299, A-300 Cha racterization ~ . The allegation is that the activities of EBASCO vendor . quality assurance (QA) personnel, and the vendor QA records were not adequate.
Assessment of Allecation: The implied significance of this allegation is that tne ESASCO QA personnel dealing with the vendors were not performing their jcb adequately and that problems *dith QA records generated by the vendor could exist which could affect the acceptability of safety-related material and equipment. In assessing this allegation,-the NRC staff reviewed the following types of information: (1) EBASCO's Nuclear Quality Assurance Program, particularly the sections dealing with surveillance, (2) the letter that the allegation was based on, (3) letters on file dealing with EBASCO's attempts to clear deficien-cies with vendor documentation, (4) interviews / conversations with EBASCO personnel, (5) the EBASCO deficiency record file, (6) the EBASCO master list of deficiencies dealing with documentation, (7) the EBASCO master list of deficien-cies dealing with radiograph files, (8) two audits reports, (9) receipt, receipt inspection, and conditional release system, and (10) selected vendor documentation quality records. This review revealed that EBASCO was aware of the issues identified in the
- allegation and that it was based largely on an EBASCO assessment of problems with vendor / subcontractor QA records that had been uncovered by EBASCO during a QA records review prior to turning those records over to Louisiana Power and Light (LP&L). The NRC staff investigation also disclosed that EBASCO conducted a thorough review of the vendor records. EBASCO detected some problems with those records that were identified on NCRs and DRs. The NCR and DR dispositions included:
(1) EBASCO engineering specialists, who were cognizant of the work, closed out nonconformance reports (NCRs) and deficiency reports (DRs) which were based on incomplete applicant inspection reports. (2) Reinspecting all the work done by American Bridge (AB), redoing faulty work by AB, reinspecting the rework, and generating complete inspection records for all the AB work. (3) Obtaining (material certifications for material "D" supplied materialby by Chicago CB&I. Brid and Iron ("D" not did material was used require material in non-ASME/non-certification) pressure boundary situations and
- The above allegation, which has been resolved was largely based on a situation that had existed prior to the allegation, and was being resolved at the time the allegation was made.
The NRC staff also reviewed a selection of Dravo and Southwest Engineering documentation packages and found them to be complete, technically adequate, and appropriately reviewed and approved by EBASCO. In addition, the staff reviewed selected documentation for the following EBASCO-procured stock material: pipe, 66
_ _ . - . _ _ _ _ _ _ _ .. . . . _ _ _ _ . _ ~_. _ __ ___ tubing, valves, weiding filler materials, and bolting. This material was
" ~
installed by EBASCO, Tompkins-Beckwith (T-8), and Mercury. The staff review found that this material meets the requirements. Various traceability issues
- were addressed in NCRs, and the results of the NRC review for those is denoted under Allegations A-33,' A-55, A-56, A-61, A-67 through A-77, A-84, and A-329.
The staff has concluded that the specific allegation has little safety significance. However, during the NRC staff review, a list of deficiencies associated with conditional certification of equipment (C of E) was found for equipment supplied by CE. For example, one conditional C of E for the reactor vessel and internals was issued because as-built drawings, material certifications, and the fabrication plans (as-built drawings) had not been forwarded when the equipment was delivered to LP&L in 1976. The missing documents'were reportedly sent to EBASCO sometime in 1978, according to the a EBASCO quality records supervisor, but were apparently lost prior to being placed in the EBASCO document control system. The conditional C of E was found when a check of all files was~ made in April or May 14, 1984. The missing documents have been requested from CE, and a deficiency report was issued and placed on a master deficiency list. This problem has existed since July 20, 1976. An important aspect of the reactor vessel documentation deficiency was that the e deficiency was not identified by either the EBASCO or LP&L QA program and was not on a master tracking list, it was found only as a result of a check of the files. The use of _a formal document tracking system was not initiated until the problem was identified by the NRC staff. The absence of a formal tracking system may also mean that not all CE conditional releases have been identified. Additionally, the requirements for proper identification for nonconformances, corrective actions, and the use of conditional releases was not complied with by EBASCO or LP&L, indicating a partial breakdown of the QA program. It should also be noted that the plant is now constructed, inspected, tested, and ready to load fuel. This oversight was not identified to LP&L Startup and Testing as a turnover exception, nor were they aware it existed prior to NRC detection. Deficiencies in EBASCO's identification of conditional C of E's, and in the tracking system have been corrected. Reportedly, vendor quality records have been reviewed for conditional C of E's. EBASCO has included them in their computer tracking system 'as a means to alert EBASCO management of problems, and strong efforts have been made to complete the record deficiencies. In par-ticular, the conditional C of E and associated documents for the reactor vessel and hardware were provided to LP&L in May 1984 The NRC staff during this review also discovered that the EBASCO system for conditional releases (CRs) was not adequately implemented in that the major overriding factor, " schedule," caused the system to be watered-down. Addition-
. ally, the EBASCO procedure ASP-IV-86, Conditional Release Control, did not address the time frame for the resolution and closure of conditional releases.
The time frame for 15 conditional releases had been revised from " Prior to System Turnover" to " Prior to Fuel Load" to " Prior to Commercial Operation." Two of the 15 CRs were for pressure boundary parts, flanges, and end caps. These CRs were resolved and closed during this NRC inspection. 67
Other CRs affected Safety Injection and Component Cooling Water Valves, which are primarily equipment Qualification problems. It should, however, be noted that all the CRs have been identified and tracked as system turnover exceptions. Actions Recuired: See item 5 of the enclosure to the D. Eisenhut letter to J. M. Cain cated June 13, 1984 O D6 e i 68
Task: Allegations A-172, A-174 Characterization: The allegation is that Tompkins-Beckwith (T-B) heat number records may have been falsified. Assessment of Allecation: The implied significance of this allegation is that heat numoers may nave oeen falsified and that the quality assurance (QA) documentation may not reflect the actual hardware in the plant. The NRC staff investigated the technical aspects of this allegation by reviewing LP&L, EBASCO, T-8, and Mercury Company document system procedures and comparing them with the ASME Code and the applicable ANSI N45.2 requirements. The procedures address authorized changes to QA documents and were found to be adequate. (Also see Allegations A-35, A-308.) The NRC staff reviewed a sample of T-8, Mercury, EBASCO Construction (Force Account), and NISCO turnover documentation packages and found them adequate. ' The results are detailed in Allegation A-308. The NRC staff selected a random sample of heat numbers from these documentation packages, where along with heat number data presented by T-B and EBASCO satisfied the NRC staff that the heat numbers were valid. The NRC staff found no indications that heat numbers had been falsified. In conclusion, based on the NRC staff review, no indications of falsification was detected. This allegation has neither safety significance nor generic implications. Actions Reouired: None. 44 69
Task: Allegation A-183(a), A-184 ~~ Characterization: It is alleged that Mercury documentation in OCR record packages contains generic deficiencies which indicate a partial breakdown of .- the QA program. Assessment of Allecation: The implied significance of this allegation is that generic oeficiencies in the Mercury QA OCR record packages could cause the quality of construction to be questionable. The NRC staff reviewed ten Mercury operational control record (OCR) packages. The ten packages documented approximately 60 instrument lines with at least ten welds in each line. Although the weld records were very difficult to trace, they were found to be complete and acceptable. Durin staff noted that numerous records (approximately 80 percent)g the review the were photocopies, with varying percentages of original entries. The remaining records (approx-imately 20 percent) were all originals. All record copies were properly completed, including initials for cross-outs. The NRC staff also interviewed key personnel from EBASCO and LP&L that were cognizant of QA documentation turnover package preparation, review and filing. Mercury Company personnel were not available to be interviewed.
~
The allegation pertaining to a QA program breakdown is addressed in Allegation A-48. The NRC staff concluded that this allegation has neither safety significance nor generic implications. It is also noted that there is no requirement in ANSI fl45.2.9 for records to be originals. Actions Required: None 70
. Task: Allegation A-185 Characterization: The allegation is that a written procedure does not exist
; for joint Mercury /EBASCO review of system turnover documentation.
Assessment of Allecation: The lack of a written procedure may cause an inadequate review of quality-related documentation which could have affected
. verification of the quality of safety-related systems.
Based upon Significant Construction Deficiency (SCD) No. 57, EBASCO instituted a joint Mercury /EBASCO review of contractor turnover packages. SCD 57 was
' issued to identify problems with Startup System's ! SUS)
( Nos. 59, 60A, B, and C. Because Mercury QA records contained inadequate installation and turnover documentation, a joint Mercury /EBASCO review of Mercury's turnover documentation was performed to eliminate excessive Mercury documentation deficiencies; prior to turnover to EBASCO. Mercury performed turnover documentation review at their facility in accordance with Mercury QC Procedure 3010 (N49720 Suppl) - Quality Assurance Records Control. The review compared the turnover package QA records against project specifications and identified deficiencies and corre-sponding corrective actions to close these deficiencies. At this point, an EBASCO reviewer would evaluate Mercury's package review for adequacy. EBASCO would either agree, disagree, or add additional deficient-action items. Each
- action Mercury completed was documented on a Mercury checklist.
EBASCO procedure QAI-9 (original Revision 0, dated October 29,1979), Review
- and Handling of Construction Installation Records, was in place and was being originally used for the review process. A specific, detailed EBASCO procedure was not written for joint Mercury /EBASCO turnover package reviews. However, EBASCO procedure QAI 9A, Revision 0, dated September 21, 1982, Record Statusing to Support Startup System Turnover, was in effect during joint EBASC0/ Mercury reviews. This procedure was generic in nature and provided instruction to delineate guidelines for in-process and status-checking of construction-installation QA records. The procedure included guidelines for:
responsibilities; review of records completeness; accuracy of content; proper form; traceability; legibility; and authenticity. It also provided forms to document record review with the status (accept / reject) of individual documents within the package, and noted that deficiencies detected during review were to be documented on Form QAI 9.2. On October 8, 1982, EBASCO issued a supplement to QAI 9A (QAIRG No.15) for EBASCO QAIRG reviewers to evaluate Mercury's turnover packages for Contract W3-NY-15. This procedure delineated the specific documentation, by form number, that Mercury's package should include. 4 Even though no specific EBASC0/ Mercury joint review procedure was in existence, by adhering to EBASCO procedure QAI 9, QAI 9A, and Mercury QC Procedure QCP-3010, i joint Mercury /EBASCO reviewers had adequate procedural guidance to perform and document review activities. j . l l - L 71
It should also be noted that the EBASCO joint review with other site contractors .. was performed using the same procedural mechanism as previously described for Mercury. . '. The NRC staff concludes that this allegation has no safety significance and is not indicative of any adverse generic or management implications. Actions Required: None. F e 72
Task: Allegation A-186(a) Characterization: It is alleged that EBASCO quality assurance (QA) surveillance did not monitor contractor work activity, e.g., documentation.
' ' ~
Assessment of A11ecation: The implied significance is that without adequate QA surveillances, tSASCO and LP&L could not assure that the contractor's QA programs were being properly implemented. Therefore, the quality of the installation and inspection programs could be questionable. The NRC staff reviewed 462 surveillance reports covering a 5-year period of construction activity. The reports were reviewed for type of inspection performed, type of documents reviewed, validity of corrective action, and follow-up and response time for corrective action. The staff reviewed trend analysis-reports to determine transmittal of surveillance report information, corrective action response, and closure. QA surveillance personnel qualifications were also reviewed. The NRC staff found that corrective action was not always timely but was followed up, and that reports were closed out in all cases reviewed. In some cases, corrective action was not accepted by EBASCO AQ , and surveillance reports were returned to the contractor for satisfactory resolution. Surveillance reports indicated that all contractors were monitored at various times during their work and documentation activities. The NRC staff determined that EBASCO personnel performing the surveillance activities were certified and qualified to perform their surveillance duties. Based upon its review, the NRC staff concluded that the EBASCO surveillance program meets the intent of the EBASCO Quality Assurance program. Therefore, this allegation has neither safety significance nor generic implications. Actions Required: None. J i 9 O 73
Task: AllegationA-186(b) ~ Characterization: It is alleged that an EBASCO Records Reviewer stamped off 1,600 civil records (concrete) in an 8-hour work day in September 1982. '~ Assessment of Allecation: The implied significance of this allegation is tnat inccmplete recoros, and' improper review of quality records for technical adequacy, could cause the quality of construction and the results of inspec-tions to be questionable. The NRC staff interviewed the EBASCO records reviewer and the imediate supervisor, and compared the individual's qualifications to the job prerequisites. The staff examined EBASCO's review of contractors' civil records stored in EBASCO's vault, and also reviewed the basis for the excessive amount of documents stamped off in a short time frame. The NRC staff found during the interview of the individual and his immediate supervisor that the individual's role in documentation review of civil concrete quality records was as follows:
- 1. His actual job description was Records Coordinator, not Records Reviewer as stated by the alleger. The Coordinator's responsibility was to assemble, into one task file, various quality records from contractors participating in concrete placement (i.e., concrete pour package). It
. appeared that the alleger was unaware of the difference in the job descriptions between a Records Reviewer and a Records Coordinator. . 2. The types of records assembled by this indivi. dual into the files that were transmitted to EBASCO's vault for EBASCO QAIRG review included (a) top tier concrete drawings (DWGs); and (b) individual pour packages associated with those DWGs, including batch (concrete placement pour) slips, concrete material testing reports, gradation sheets from the batch plant for sand and gravel, etc.; and (c) quality related activity and inspection records for preplacement (rebar, inplacements, and forms), placement of concrete, curing of concrete, and postplacement (form removal and checking for voids).
Contractors involved with preplacement, placement, postplacement activities were J. A. Jones, American Bridge, Allied, Gulf Engineering, and Peabody Testing (GEO). Document review and collation by the Records Coordinator was for sorting and accumulating concrete documentation by task or activity. This was not a rev.iew for technical adequacy. This review included completeness, proper form, legibility, and authenticity of documentation. Upon completion of
- records collation and review, the Records Coordinator grouped the records by type and task. Prior to submission to the EBASCO vault for QAIRG review, the Coordinator stamped the records to signify accountability and review action complete.
Transmittal packages to the vault could contain up to 250 records. In a day, between 500 to 600 records were stamped and submitted to the vault. The governing procedure for this activity was EBASCO generic procedure QAI-9.
;74
The NRC staff compared the individual's qualifications to his job assignment. It was determined that his qualifications complied with the job performance prerequisites. It was also noted that he did not receive the QAIRG review training; however, since no review for technical adequacy, accuracy of content, or traceability was required, his past experience, training and education provided adequate qualifications. The allegation identified that during an EBASCO QAIRG technical review of 70 of the 1,200 civil concrete pour package quality records numerous problems were identified. Based upon . deficiencies noted during this review, several discrepancy reports (DRs) were initiated, and nonconformance reports (NCRs) were then issued identifying generic discrepancies. To disposition the NCRs, the QAIRG initiated a 100% re-review of the concrete packages. The NRC staff interviewed the alleger, who stated he was satisfied with EBASCO's corrective action. For further information regarding the alleged individual's stamping (" bean counting") and document review, see Allegations A-143 and A-289. In conclusion, the individual's stamping was not to document technical review, but to account for and collate records. This allegation has neither safety significance nor generic implications. Actions Reouired: None. e S I e 9 4 75
Task: Allegation A-187
- Characterization: The allegation is that Mercury instrumentation drawings were not correct because field changes were not incorporated into the drawings, many drawings contained red lined changes, and two or three drawings for the same installation were marked differently.
Assessment of A11ecation: The implied significance of this allegation is that inadequate control of field changes could result in as-built drawings that do not reflect the actual plant configuration. Multiple red lines on the same drawing may.cause confusion that could lead to inadequate inspection of the affected systens. In assessing this allegation, the NRC staff reviewed the Mercury Company's procedures for red lining drawings and for document control, examined Mercury drawings, and conducted a walkdown of a sample of completed systems, using the as-built drawings. Out of 19 randomly selected instrument lines which were field checked by the NRC, one process tubing deviation from the as-built drawings was identified. This deviation was for an LP&L-installed modification that had not been completed. LP&L had documentation on file reflecting the change and indicating the incomplete status of the field work. As a result of this review, the NRC staff determined that the Mercury system for red lining drawings was cumbersome. It allowed controlled copies of a
- drawing and its revision to show different changes. This occurred if nonpressure boundary field changes were made after the pressure boundary portion of the drawing had been completed.
In conclusion, the as-built drawings reflected the actual condition of the installed hardware, and although accurate, this allegation had neither safety significance nor generic implications. Actions Required: None. Se 0 76
Task: Allegations A-188, A-190, A-191, A-193 Characterization: The allegations are that Mercury Construction Company's procedures for review of quality assurance (QA) records or documentation '~ packages were vague, loose, had not been properly reviewed by EBASCO and LP&L Engineers, and did not meet the requirements of ANSI Standards or the ASME Code requirements.
- Assessment of Allegation: The implied significance of this allegation is that if the Mercury records review procedures were inadequate, the records review of safety-related systems may not have assured the acceptability of the systems installed by Mercury.
The NRC staff investigated the allegation by evaluating: (1) Mercury's " Review and Handling of Construction-Installation Records," procedures for compliance with ANSI and ASME requirements; (2) EBASCO's program for reviewing and approving contractor procedures prior to issuance; and (3) task reports regarding the NRC staff review of Mercury's turnover documentation packages for procedural implementation. Mercury Company Review The NRC staff compared Mercury's QA record control procedures with the ASME e Code, 10 CFR 50, Appendix B, and the applicable ANSI N45.2 requirements and found them acceptable. The procedures provided guidelines and established minimum requirements for the collection, control, review, filing, storage,
' maintenance, and disposition of quality assurance records. Their " System Turnover Document Package" included: test records; drawings; equipment lists;
. and other records supporting construction activities. The types of quality records which were to be included in the System Document Turnover Packages were:
- 1. Documentation requirements index, which identified required qual.ity records and included package review completion and acceptance signatures dated by Mercury and the Authorized Nuclear Inspector (ANI).
- 2. Equipment Installation Form 277.
- 3. Expansion Anchor Installation Form 277A.
- 4. Piping Tubing Inspection Report Form 276-1.
- 5. Tube Tray Inspection Report Form 262-A-1.
- 6. Hanger and Support Inspection Form 262-1.
. 7. Weld Data Reports Forms 197-1, -2 or -3.
- 8. Material Verification Reports Form 198.1-1 (as applicable).
- 9. Failed Anchor Reports Form 211A.
77
- 10. .QC Report (general) Form 211 (as applicable).
- 11. Process Control Travele'r Form 208.
.. '. ~ 12. Operation Control Report Form 110.
- 13. Reference to Code Data Reports or Copies of Code Data Records, a
[ 14. -QC Pressure Test Reports Form 216. l 15.. Pressure Testing Requirements.
- 16. NDE Reports.
, ,17. Mercury Drawings.
EBASCO Review EBASCO's program for procedural review and approval included procedures generated by EBASCO Engineering and all site construction contractors. The ! Site Quality Assurance Engineering Department and/or other EBASCO disciplines, as required, reviewed procedures affecting quality prior to implementation. Two procedures governed procedural review and approval. Procedure ASP-III, Preparation of Site Procedures, delineated guidelines for preparation, scope of l - procedures, instructions for procedural content, procedure, approval, issuance, and the revision mechanism. Procedure QAI-2, QA Review of Site Generated
*. Procedures of Activities Affecting Quality, details the methods for procedural
. . review and documenting comments applicable to the review; the resolution and acceptance of procedural comments; and final acceptance of the procedure. The NRC staff obtained a copy of EBASCO's review / comments and approval for the initial and current revision of Mercury's QA Records Control Procedure QCP-3010. The objective was to verify that EBASCO had reviewed and approved Mercury's procedures prior to implementation. These documents contained recorded comments with accepted resolution, and approval granted to Mercury for issuance and implementation. The staff also reviewed a number of Mercury turnover QA documentation packages for the Reactor Coolant instrument lines. These packages had been reviewed and approved by Mercury's QA personnel, using QCP-3010, documentation review a procedure. EBASCO QAIRG also reviewed and approved these packages. Contained in the packages were the following types of quality documents: A. EBASCO Review Check List (dated and signed by EBASCO QAIRG reviewer) B. Mercury Documentation Requirements Form /RPT-Form 209 (Documentation Index) C. Operations Control Report (OCR) cover sheet D. Process Control Traveler I E. Pipe and Tube Inspection Report 78
F. Material Verification-Heat Numbers Traceable, CMTRs and C of C.
- ~
G. Weld Data Sheets
', H. NDE Reports I. Quality Control Reports-acceptance of work J. NCRs K. CIWAs ,
L. As-Built / Red-Lined Drawings (latest revisions) M. Hydrostatic / Pneumatic Test (as applicable)
- 1. Hydrostatic / Pneumatic Test Instructions
- 2. Hydrostatic / Pneumatic Data Sheet
- 3. Valve Line-Up Sheet i 4. Hydrostatic / Pneumatic Test Discrepancy List
- 5. Weld Data Sheet - information copy to verify weld number (not used for B31.1 test)
. 6.. Calibration Sheets
- 7. Piping and Instrumentation Drawing For further information regarding the staff's review of Mercury's documentation system, see Allegations A-308, A-183, A-184, A-197.
In conclusion, Mercury's QA records control procedures for the review of QA 1 turnover documentation packages were found to be acceptable. Their procedures were compared with the appropriate ANSI and ASME requirements and found to be acceptable. EBASCO reviewed and approved Mercury's procedures prior to implementation. The NRC staff concludes that these allegations have neither safety significance nor generic implications. 2 Actions Required: None. 4 t O 9 l t s 79
Task: Allegation A-196 Characterization: The allegation is that the ESASCO Document Control Supervisor was unqualified for the position. Assessment of Allegation:-The implied significance of this allegation is that a accument control supervisor without proper quality assurance (QA) training and background could adversely affect document review and records management. The NRC staff reviewed training requirements and records, and held discussions with the EBASCO training coordinator and the Document Control Supervisor in examining this allegation. The NRC staff.found that EBASCO record reviewers were required to have, as a minimum, a high school education, and QA indoctrination, including lectures, reading assignments, and on-the-job training (0JT). The Document Control Supervisor held a BS degree in Economics, and had been employed by EBASCO in records management since 1976 as QA Records Coordinator (1976-1977), QA Records Representative (1977-1978), and QA Records Supervisor (1978-1984). The Document Control Supervisor had been an administrative manager over another supervisor, lead reviewer and record reviewers, and has supervised up to 150 EBASCO personnel after completing all the EBASCO requirements for records
,. review supervisors.
There are no specific training, qualification or certification requirements for ' records reviewers in NRC regulations, the ASME Code, or ANSI standards. The EBASCO training program for records reviewers appeared to be adequate. The NRC
' staff concluded that the QA Records Supervisor and Documentation Manager is qualified. This allegation has neither safety significance nor generic
' implications. Actions Required: None. Y a. I v } 80
Task: Allegations A-197 through A-206, A-213, and A-216 Characterization: It is alleged that Mercury's corrective actions for problems relating to quality control (QC) weld record data sheets were not documented .. adequately, nor were the reasons for the corrective actions denoted. Assessment of Allegation: The implied significance is that the lack of adequate review of documentation for technical adequacy anc completeness could cause the quality of installation to be in question.. Without proper. corrective action to determine the root cause and to prevent recurrence, the problem may be corrected, but the system could continue to be deficient. The NRC staff reviewed 10 operational control record (OCR) packages [ refer to Allegations A-183(a) and A-184] and did not substantiate this allegation. However, the records reviewed by the NRC staff were at.a later date and consequently, the record deficiencies identified by the alleger could not be verified. The NRC staff discovered the following problem pertaining to missing QC records. Communications between LP&L and EBASCO had prompted an EBASCO revision to an LP&L drawing to clarify the " class break" for N1 instrument lines. The revision imposed ASME Class requirements for all installations between the process piping and the instruments for instrument lines installed after April 7, 1982. Prior to the revision, a class break was defined to show the
- location where ASME Class requirements did not apply and where ANSI B31.1 ,. guidelines applied.
Although ANSI B31.1 guidelines do not address records retention, 10 CFR 50 Appendix B does require installation and inspection records. Therefore, for locally mounted N1 instruments, even though they were installed prior to April 7, 1982, the QC records could not be located. The NRC staff found several deficiencies in N1 instrument records of installation and inspection in the zones classified under ANSI B31.1, including weld reports, welder identification, weld fillbr material, base material, and weld inspection results. The NRC staff concludes that the lack of QC records for instrumentation installation to ANSI B31.1 is in violation of the requirements of 10 CFR 50
~
Appendix B, and related QA/QC program elements. Actions Reouired: See Item 2 of the Enclosure to the letter from D. Eisenhut to J. M. Cain (LP&L) dated June 13, 1984. p 81
. Task: Allegation A-207 l Characterization: The allegation is that on many occasions, Mercury quality
. assurance (QA) personnel were observed throwing away original documents after copies were made.
Assessment of Allecation: The implied significance of this allegation is that improper maintenance of QA records associated with the construction of safety-related systems can affect the quality of those systems. The NRC staff reviewed this allegation against the standard QA good work practices outlined in ANSI N45.2.9, Section 3.2.1, which reads "QA records submitted for. retention shall be legible, completely filled out and adequately identifiable to the item involved" and "These records may be either the original or a reproduced copy." , The Mercury procedure allowed the original records to be destroyed after a revision or change was accomplished. The final records did reflect the as-built system and were found to be adequate. The staff found no evidence of any " completed" records missing. There is no requirement to maintain ' inprocess records because of the definition in ANSI N45.2.9:
" Quality Assurance Records - Those records which furnish documentary evidence of the quality of items and of activities affecting quality.
For the purposes of this standard a document is considered a quality assurance record when the document has been ccmpleted." The NRC staff reviewed Mercury QA records and found them legible, complete, and retrievable although many records were reproduced copies. The NRC staff concludes, although the maintenance of reproduced copies in lieu of original records is somewhat unusual, that this allegation has no safety significance nor generic implications (pertained to Mercury Company only).- Actions Required: None. i 82
Task: Allegation A-209 Characterization: It w'as alleged that review of Mercury weld records identified some welds that had not been fit-up or final inspected. Assessment of Allegation: The implied safety significance of this allegation is that missed inspections could cause the quality of installation to be in question. l The NRC staff reviewed 10 operational control record (OCR) quality control (QC) packages for 60 instrumentation lines averaging 10 welds per line for a total of 600 welds. The weld records reviewed were found complete. However, the records review by the NRC staff took place at a later date and consequently the record deficiencies identified on earlier records by the alleger could not be verified. The NRC staff review did reveal that an EBASCO nonconformance report (NCR) had been written and dispositioned for a weld FW-5 in OCR-1830 for the condition of no fit-up inspection (see Allegation A-232). The staff review provided no indications to show lack of fit-up or final inspections of welds. Where inspections had been missed, they were identified in the NCR and corrective action had been taken by Mercury. ! # The NRC staff concludes that this allegation has neither safety significance
. nor generic implications.
Actions Required: None. D l l l l l f r O ee 9 83
Task: Allegation A-214 Characterization: The allegation is that quality control (QC) weld data reports from Mercury Construction Company during 1981 contained incorrect
- filler material certification because QC personnel would obtain a heat number for filler material and continue to record that number whether or not it was correct.
Assessment of Allegation: The implied significance of this allegation is that traceability of materials may have been lost, thus causing the validity of the installed material containing the required certification to be questionable. InassessingthetechnicalIspectsofthisallegationtheNRCstaffreviewed Mercury weld data reports and took random samples of filler material heat numbers for verification of their traceability to Certified Material Test Reports (CMTRs). Procurement Purchase Orders and CMTRs were also reviewed for heat numbers sampled and were found to be acceptable. In addition, the staff reviewed six operational control record (OCR) packages which contained weld records for 28 different instrument lines. No indications of adding or falsifying heat numbers was found. However, many of the first issue (pink sheets) Mercury weld records were a combination of photocopy and original (pen and ink) data. Thus the original documents in most cases were not avail-- able for review. It is noted that there is no requirement for the maintenance of original records in lieu of photocopies. (See Allegation A-207 for NRC staff finding pertaining to the descruction of historical records.) In conclusion, this allegation has neither safety significance nor generic implications. Actions Required: None. Se 84
Task: Allegation A-223 Characterization: The allegation is that there were documentation discrepancies between the field construction and QC packages, that Operational ,', Control Records (OCRs) 492 and 903 had Penetrant Test (PT) reports and signatures missing and that repair data and documentation was confusing. Assessment of Allegation: The implied significance of this allegation is that inccmplete, missing, and conflicting documentation raise a question regarding the quality of the installation and the inspection results. The NRC staff investigated this allegation by reviewing nine OCR packages, including OCRs 492 and 903. This was done to determine if the problems identified were generic. Systems reviewed included Reactor Coolant, Safety Injection, Feedwater, Charging Water, and Main Steam. The review of OCRs 492 and 903 revealed no missing PT reports nor missing signatures. EBASCO originally reviewed all Mercury packages in accordance with QAI-9. Deficiency Reports (DR) were issued and tracked for all discrepancies noted during the EBASCO review. Deficiencies were corrected prior to system turnover. - During the NRC staff review of OCRs 1782 and 1924, one weld from each package I was found to have missing PT records. Since the records could not be located in a timely manner, LP&L decided to perform additional penetrant testing to determine weld acceptability. LP&L and EBASCO had pts performed on these two
# welds and both welds were acceptable.
The staff also found that field construction records (green sheets) were destroyed and that the only remaining record of work performed was the QA copy (pink sheets). These QA records were generally in poor physical condition and because of the many crossouts and explanatory notes, were confusing and difficult to follow. However, after an extensive review, the NRC staff was able to determine through actual inspection and review that the records, although in a less than desirable condition, were acceptable and adequately documented the actual as-built configuration. Later revisions of weld record QA sheets showed significant improvement in the quality of the records and had original data recorded on the original record form. The NRC staff reviewed the QA documentation for approximately 600 welds and concluded the following: the missing PT records were isolated cases; the records were poorly maintained; weld history was difficult to follow; the filing system was extremely cumbersome, making retrievability difficult; and records were not always original copies (but originals are not a requirement). Even though the noted problems are a NRC concern, all requested records were
- available and were found to be acceptable. This allegation has neither safety significance nor generic implications.
Actions Required:
- None.
85
1 Task: Allegation A-227
' ~
Characterization: It is alleged that a recommended system for identifying and. cispositioning nonconformances was written by an EBASCO employee for seismic
. . Category I supports but was rejected by ESASCO management. ,
Assessment of Allegation: The implied significance of this allegation is.that EBA5CO's nonconformance system for seismic supports may have been inadequate and that nonconformances may not have been properly identified and dispositioned which render the quality of the supports as indeterminate.
> The NRC staff evaluated Mercury Construction Company procedures for controlling nonconformances to determine the adequacy of the system, and to determine if 1 there was a compatible system in place such as that recommended by the former
~ employee. The staff also interviewed EBASCO personnel who were involved in the -
review, audit, inspection, disposition, and closure of Mercury nonconformance 4
- reports (NCRs).
In its review of Mercury procedures, the NRC staff found that the former EBASCO employee's recommended procedure'had not.been specifically incorporated, and ' that no subsequent changes were made to the nonconformance system. ' Mercury Procedure SP664 required the writing of NCRs when necessary.
# The NRC staff also interviewed personnel who were involved in. Mercury
. nonconformance activities. The staff found that Mercury's system did not include any type of nonconformance document (such as a discrepancy notice (DH)), other than an NCR, for a suspected nonconformance. Only NCRs were
. entered into the system; this caused many'more NCRs to be issued than were actually necessary. This use of NCRs, rather than DNs or other types of reports, was supported by the statements in procedure SP664 which stated, "The reporting of nonconformances shall not be limited to QA/QC department. Whenever a suspected nonconformance is discovered, it shall be the responsibility of the individual or department to initiate a nonconformance report." (Emphasis added.]
In the interviews, the NRC staff also found that, to assure proper disposition, a 100 percent review based on EBASCO NCR W3-7317 was performed on Mercury NCRs that had been dispositioned "use-as-is." During system turnover, NCRs were also reviewed as part of the turnover packages. The NRC staff believes the responsibility and decision to accept or reject an , employee suggestion on a procedure was the option of management. EBASCO and Mercury management made the decision not to accept the EBASCO employee's ; suggestion for a new procedure. The NRC staff found the nonconformance system used by EBASCO and Mercury acceptable. This allegation has neither safety significance nor generic implications.
" The NRC staff also reviewed other allegations related to Mercury NCRs. See j the staff assessment of Allegation A-232.
f* Actions Required: None. I 86 1
- - - - -+...,_-.._,-,y .-,,, ,,w.--e..e,,--__-.,,,,-m, .-,_,_,-vypg-m-. . ._._.w.,g,-., __,,,---,,,,-._m--,,e.v..
Task: Allegations A-229, A-306g e '~ Characterization: It is alle operational control recordsOCR) (gedpackages that an individual wasquality [ system startup directed to turn over assurance ,, (QA) records] before system reviews were completed, and that the resulting documentation and hardware discrepancy lists submitted to LP&L documented inadequate reviews. Assessment of Allecation: The implied significance of this allegation is that if inadequate reviews of turnover documentation packages were performed, the acceptability of safety-related systems may have been affected. The NRC staff examined an EBASCO interoffice memorandum which provided package turnover status. The staff interviewed EBASCO and LP&L personnel involved with the Mercury turnover documentation p'rogram, and examined the discrepancy list identified in the EBASCO memorandum to determine if referenced corrective action was documented. Also, the NRC staff reviewed a sample of Mercury SUS turnover packages to verify that deficiencies noted by EBASCO Quality Assurance Installation Review Group (QAIRG) had been adequately resolved. The EBASCO interoffice memorandum provided the status of five SUS documentation c packages prior to turnover the LP&L. The topics included EBASCO and Mercury integrated SUS turnover package review for approved checklists; training of
! Mercury and EBASCO personnel for parallel review; and a notation.that the five . SUS packages reviewed were missing preliminary as-built (PAB) drawings (PABs represented the as-built configuration). Also, two SUS turnover packages that were given to the EBASCO startup group were identified as not being reviewed . against the PAB, and that the EBASCO startup group had rejected the two packages and caused all five SUS packages to be rereviewed using the PABs as baseline documents.
The NRC staff interviewed LP&L and EBASCO personnel, including the addressee of the EBASCO interoffice memorandum, involved with Mercury and EBASCO reviews of turnover packages. The staff found that, between December 1982 and February 1983, there was a change in the Mercury and EBASCO interface as to parallel and individual company documentation review. There was some confusion between Mercury, EBASCO, and LP&L reviews, and, as noted in the EBASCO memorandum, two SUS packages were submitted to EBASCO startup group without being evaluated against PABs. The EBASCO startup group rejected the packages, which caused re-review by Mercury and EBASCO QAIRG. The NRC staff believed this to be a good example of quality assurance system checks and balances working effectively to maintain system effectiveness. The NRC staff could not locate the discrepancy lists for the two systems alleged to have been reviewed without PABs. However, LP&L did provide the
. staff with original turnover deficiency and status documents received by LP&L from EBASCO QAIRG. These records identified that Mercury and EBASCO reviews were performed using PABs as the baseline documents. After reviewing these
- current records, the NRC staff found no procedural violation.
Ten Mercury generated SUS turnover packages were reviewed by the NRC staff. The objective was to verify that Mercury and EBASCO documentation review did cover technical adequacy, completeness, and authenticity. The staff concluded that Mercury and EBASCO QAIRG reviews were adequately accomplished and 87
documented. In addition, the NRC staff performed walkdcwns for 19 instrumenta-tion installations and found no problems. For further details, see the NRC staff assessments of Allegations A-188/A-190/A-191/A-193, A-35, and A-308. The NRC staff fcund that fiercury, EBASCO, and LP&L had implemented adequate control of SUS turnover package reviews. This allegation has neither safety significance nor generic implications. Actions Required: None. O 5 0 . e e m. 88
. _. .- - .= - .. .. . Task: Allegation A-230 Characterization: The allegation is that an EBASCO review of Startup System . (sus) 525, Reactor Coolant System , found many Mercury documents that were incomplete and red lined on drawings that did not match the as-built plant configuration. This allegation also claimed that there were generic deficiencies in Mercury turnover packages.
Assessment of Allegation: The implied significance'of this allegation is that missing or inadequate documentation and incorrect red lined drawings could place the quality of installation in question. The NRC staff investigated this allegation by reviewing four of the six safety-related Operation Control Reports (OCRs) associated with SUS 52B and five OCRs for other systems. EBASCO reviewed 100% of the OCRs prior to Mercury turnover of the packages and LP&L performed a 10% sample review. The NRC staff found that Mercury, EBASCO, and LP&L reviews were documented on completed checklists and various other forms in the turnover packages. QA documentation deficiencies were identified on EBASCO DRs and resolved. (See Allegation A-05). The issue involving problems with red lined drawing: was addressed by Allegation A-187, for which the NRC review identified no problems. The NRC staff concludes that this allegation has neither safety significance nor generic implications. Actions Required: None. E O e 89
. _ . . ._ __- . ._ ~ . ._
Task: Allegations A-232, A-234, A-237, A-238, A-243, A-244, A-245, A-262, A:3T1, A-312, A-313, A-316, A-317, A-318, A-320, A-321, A-323, A-324, A-325, A-326, A-327, A-328, and A-331 Characterization: It is alleged that the Mercury, and to a lesser extent the ESA5CO, nonconformance systems did not: (1) properly identify nonconforming components; (2) prevent the installation of ronconforming materials, parts, and components; (3) provide for the proper disposition of nonconformances; (4) give Quality Assurance (QA) personnel the freedom to write nonconformances; and (5) assure that corrective actions were adequate. Assessment of Allegation: The implied significance of this allegation is that the installation of safety-related systems could be questionable. Although this allegation appeared to be generic, the large number of specific examples provided to the NRC staff indicated that these problems were related only to certain activities. Some allegations were also related to charges that discrepancy notices (DNs), deficiency reports (DRs), field change requests (FCRs), design change notices (DCNs), and speed letters were used to circumvent the requirement for nonconformance reports (NCRs). In evaluating this alleg2 tion, the NRC staff: (1) Reviewed Mercury and EBASCO procedures for processing NCRs, and examined the NCRs identified by the allegers to determine if they were properly dispositioned and to determine if corrective action was taken; and Interviewed Mercury, EBASCO, and LP&L quality engineers involved in the (2) NCR system. The NRC staff reviewed each NCR identified by the allegers to determine how adequately it was reported, reviewed, and dispositioned. These NCRs were further examined for generic problems which could affect other safety-related systems. The NRC staff.found the following with respect to each specific allegation:
- 1. Allegation A-232 - It was alleged that EBASCO NCR W3-4352 was improperly dispositioned in order to meet ASME Code requirements and ANSI standards and dispositioned "use-as-is." The NRC staff reviewed the disposition of the NCR and concluded that the corrective action taken for closure was appropriate.
- 2. Allegation A-237 - Examples of some NCRs were provided which were alleged to be examples of " things not properly handled." Additionally, it was alleged that most Mercury NCRs were not properly addressed or closed.
The NRC staff reviewed between 125 and 150 liercury NCRs. Of these, the staff found that 19 NCRs had questionable or improper dispositions; they were as follows: 90
Mercury NCRs
~
313, 322, 337 Identified seven i" stainless steel lines for P2 instruments
- that were damaged by weld spatter. The NCR stated that the lines were replaced and documented as such in operational l
control record (OCR) 995 and OCR 1020, but it could not be ascertained frem these rework packages that the repair and reinspection was either started or completed. There was no l documentation with these NCRs to prove that corrective action was completed. 363 Indicated a problem with fitup of emergency diesel generator fuel oil tank "A." This was a safety-related system; therefore, an authorized nuclear inspector (ANI) review should have beeri performed, but was not. l 554 Noted numerous problems with supports during a walkdcwn. l There was no proof of work being performed to correct these problems other than a memo (Form 211) stating that work was performed. , 658 Identified problems with OCR 1671 seismic Category I support, B-430-x23-J-42. The NCR stated "the disposition has been completed, all rework documented." There was no other - i documentation in the package other than the NCR W3-7317
.- acceptance letter.
572 Noted that the weld on support location #26 was undersized. I The NCR stated that the weld was reworked and weld metal added to bring weld to sufficient size. There was no reference as to what OCR was issued to perform this rework or traceability of weld metal used in the performance of this job. Also, there were no i~nspection reports identified or contained in the package. 673-678 These NCRs were closed out by the statement:
" Administratively closed B31.1 to be tracked and resolved by Mercury Engineering Department." This resolution was unacceptable as the requirements of 10 CFR 50 Appendix B apply to safety-related installations as committed to by LP&L. (Also, all of these NCRs were reviewed by EBASCO under NCR W3-7317 and accepted "as-is.")
673 Identified problems with instrument tubing installed by OCR #723. 674 Identified problems with the electromagnetic control panel worked by OCR #1246. 675 Identified problems with instrument tubing installed
- by OCR #720.
91
~- . - - . . . - . . . - . . - - - .
---. - ~ . . . .
676 Identified problems with instrument tubing installed by OCR #720. 677 Identified problems with instrument tubing installed by OCR !1332. 678 Identified problems with instrument tubing installed by OCR #723. 888 Indicated problems with personnel qualifications; e.g.,
"Several QC [ quality control] type personnel have been certified level II without documented indications of qualification requirements per QCP 3110 paragraph 1.4 and ANSI N45.2.6." Recomended disposition was marked "N/A" yet the recomended disposition completed stated "This NCR not processed: 1) Initiator not a Mercury employee at time of writing; 2) QCP 3110 . . . does not apply to W3;
- 3) ANSI N45.2.6 previously incorporated by QCP 3050 is approved. All M Co. QC techs are trained and tested per QCP 3050 prior to performing inspection or tests."
889 Indicated problems dealing with piping supports installed by Mercury in that the installed hangers were different than those noted in Mercury's QC support installation documentation. As with NCR 888, the recommended disposition
- was marked "N/A" and the recommended disposition was completed by saying "This NCR not processed: 1) Initiator not a M_ Co. Employee at time of writing; 2) . . . "
2234 Stated that no heat numbers could be verified between FW13 and FW13R. This is for OCR 666, System 528. The recommended disposition was per attachment #4 of NCR W3-4593. 3149 Indicated that there was no documented indications that welder N-343 was qualified to welding procedure specification D (WPS-D). Disposition of this problem was by use of a weld test coupon subsequently found on April 27, 1983, but no longer available. No documenta-tion existed on the qualification of this welder or on his retest. Thus, all welds made by this welder were l suspect. Generic Problems with NCRs o The validity of several dispositions were questionable because the referenced letters used to close these NCRs did not adequately address them. o Several Mercury NCRs identified that discrepancies existed between drawings and documentation. The solution to this problem was to modify the documentation so that it " agreed with what was installed in the field." The l l 92
adequacy of design was also questionable because of a lack of indications that a design review was performed by Mercury Engineering.
'. o Some NCR dispositions and QA reviews perforr:ed by Mercury did not have sufficient documentation to justify those actions and reviews.
- 3. Allegation A-238 - It was alleged that Mercury Corrective Action Report (CAR) No. 129 of December 13, 1982, was not handled correctly because the actual problem raised was evaded. The NRC staff reviewed this issue and determined that CAR-129 was reviewed by EBASCO and subsequently upgraded to NCRs W3-5669 and W3-5671. These NCRs were dispositioned and closed on February 18, 1983. The staff concluded that this CAR was dispositioned properly and that this allegation has neither safety significance nor generic implications.
- 4. Allegation A-234 - It was alleged that many Mercury NCRs were improperly dispositioned and never received EBASCO NCR numbers. While conducting its general review of Mercury NCRs and the NCRs identified in the allegations under A-237, the NRC staff sampled between 125 and 150 of these NCRs to determine if there were Mercury NCRs which should have been elevated to EBASCO NCRs and were not. In conducting its review, the staff found that EBASCO had issued NCR W3-7317 (October 26, 1983) to disposition the
. violation of Mercury Procedure SP-664 in that when a Mercury NCR was dispositioned "use as is," the Mercury QA Supervisor elevated the applicable NCR to EBASCO. The NRC staff found that this NCR was closed on December 1, 1983, after a 100 percent review of Mercury NCRs had been conducted.
During this review, EBASCO found that 437 Mercury NCRs had been disposi-tioned "use as is" and had not been reviewed by EBASCO. An EBASCO team composed of QA and construction engineering personnel reviewed these and either concurred in or rejected the dispositions. The disposition of 36 NCRs was rejected by EBASCO and these were subsequently elevated to EBASCO NCRs. The NRC staff reviewed them and a portion of other Mercury NCRs to determine if there were any marked "use as is" which were not captured by EBASCO's review team and to determine for those dispositioned "use as is" whether they had been elevated to EBASCO NCRs. The staff's findings were that the review conducted by ESASCO appeared to have been adequate in that there were no NCRs out of the sample examined dispositioned "use as is" which had not been reviewed by EBASCO and that none were found which were dispositioned "use as is," which should have been elevated to EBASCO NCRs and had not been. This concern has neither safety significance nor generic implications.
- 5. Allegation A-243 - The allegation concerned a phone conversation on
- November 15, 1982, between a Mercury QA Document Reviewer and an employee of the Magnaflux Company about the proper method of obtaining successful magnetic particle test results. The NRC staff reviewed this concern and determined that the events discussed during the conversation violated neither test codes nor procedures. Therefore, the disposition of the Mercury NCRs addressing this problem was proper. This concern has neither safety significance nor generic implications.
93
-n -
. ~ . - . - -. .-. _ ,.
- 6. Allegations A-244-245 - It was alleged that multiple nonconformance reports
- were not dispositioned correctly. The NRC staff reviewed Mercury NCRs 996, 399, 854, 867, 922, 950, 889, 888, 952, 960, 990, 995, 1025, 1042, and
- 1027 concurrently with Allegation A-237.
- 7. Allegations A-311-312-313 - These allegations were that ESASCO NCR W3-4593 (Mercury 881) was not properly dispositioned in that LP&L was not transfer-ring heat numbers to as-built drawings; that LP&L was not transferring all possible heat numbers when they did make transfers; and that heat number charts used for tracking heat numbers were difficult to interpret and were incorrect. The NRC staff investigated this allegation during its general review of EBASCO NCRs. EBASCO NCR W3-4593 was reopened February 16, 1984, with the recognition that approximately 25 percent of tubing installed could not be directly traced to certified material test reports (CMTRs) or certif-icates of compliance-(C of Cs). This review was also conducted because additional heat numbers for instrument tubing which may have been a part of the Mercury installation were identified. The final disposition (on March 23, '984) of this NCR was that direct heat number traceability was not required for Mercury tubing installation. The disposition also stated that Mercury did not have a materials control program meeting the require-ments of ASME Code Section III or 10 CFR 50, Appendix B, Criterion.VIII.
This NCR was further dispositioned by requiring Mercury to install
- materials required by design based on general site controls. However, the NRC staff, in its review, concluded that because of the lack of heat number traceability, section(s) of non-safety tubing issued simultaneously
/ with safety-related tubing of similar size could have been installed as safety-related tubing. Therefore, the staff is not confident that what is supposed to be installed as safety-related tubing is in fact safety-related tubing. Further, both Mercury's lack of a materials control program, and their having traceability to only the warehouse, do not meet ASME Code requirements or 10 CFR 50, Appendix B, Criterion VIII.
- 8. Allegation A-316 - The allegation was that a draft NCR improperly dispositioned a problem with the thickness of stainless steel tubing.
The NRC staff reviewed EBASCO NCR W3-7538 and found it to be properly dispositioned. The tubing in question was ultrasonically tested to determine tubing wall thickness. All tubing was found to have the appropriate wall thickness. Thus, all safety concerns were resolved.
- 9. Allegation A-317 - The allegation was that draft Mercury NCR 1830/NCR 806 was improperly closed. The NRC staff review of EBASCO NCR W3-7547 revealed that there were disposition problems. For specific details, see Allegations A-33 and A-55.
. _10. Allegation A-320 - The allegation concerned a draft NCR on Mercury instrumentation supports. The NRC staff review of EBASCO NCRs W3-6514, W3-3941/R1, and W3-5819 revealed that the concern addressed by this allegation was discussed in these EBASCO NCRs. (See Allegation A-33.)
- 11. Allegation A-321 - The concern over a draft NCR on Mercury Procedure SP-664 was addressed in EBASCO NCR W3-7317. The NRC staff determined that NCR W3-7317 was dispositioned properly and that it resolved the problems addressed.
94
= _. -. _ -=:- . . . . . . ~- , . . . - - - -
l 1
- 12. Allegation A-323 - The allegation was that hold tags for Mercury NCRs 2663
*- and 2665 were removed illegally or prematurely. This allegation was reviewed by the NRC staff who notM that these NCRs were upgraded to
- EBASCO NCRs W3-5879 and W3-5881, respectively. The documentation indicated
'* that the field welds were increased to the required dimensions and success-ful NDE results were obtained. The staff could find no indications that hold tags were removed illegally or prematurely.
- 13. Allegation A-324 - The allegation was that EBASCO NCR W3-3894 might have to be reopened to address weld data report deficiencies. The NRC staff I reviewed this NCR and concluded that it was dispositioned properly and that all safety concerns were resolved.
- 14. Allegation A-325 - The allegation was that Mercury NCR 3557 was improperly dispositioned. This allegatiori concerned over-pressurization of instru-mentation lines during hydrostatic testing. The NRC staff reviewed this NCR and discovered that Mercury NCR 3557, along with NCR 3438, were upgraded i to EBASCO NCR W3-6440. The EBASCO NCR was dispositioned by reviewing stress limits of components within the test boundaries. No stress limits were exceeded. The staff concluded that the disposition of the item was proper and that it had neither safety s,ignificance nor generic implications.
- 15. Allegation A-326 - This allegation concerned proposed stop work orders and was addressed by the NRC staff during its review of Allegations A-311, A-312, A-313, and A-315.
- 16. Allegation A-327 - The concern of NCR W3-6159 was addressed in the evaluation for Allegations A-33 and A-55.
- 17. Allegation A-328 - This allegation concerned the EBASCO Site Support Engineering (ESSE) review of hardware-related NCRs and was reviewed by the NRC staff. This concern was evaluated when the NRC staff reviewed Allegations A-33 and A-55 when it was discovered that EBASCO NCR W3-7317 was properly dispositioned and closed out.
- 18. Allegation A-331 - This allegation concerned the improper closure of NCRs and has been evaluated in the NRC staff review of Allegations A-33 and A-55.
+
- 19. Allegation A-318 - The concern over Mercury hold tag installation and retrieval, as identified in interoffice correspondence W3-QA-26547, was reviewed by the NRC staff. The initial audit described in correspondence W3-QA-26547 revealed five of the ten NCRs audited had missing hold tags.
In response to this problem, LP&L reported they had located eight of the ten NCR hold tags; two were assumed to be lost. ! 20.- Allegation A-262 - This allegation was that welder qualifications were falsified, and it has been evaluated in the NRC staff review of Allegation A-215. ,- 21. Nonconformances and Corrective Actions - QAM Section 12 and SPP-664, Revision 4, requires that an NCR be written "When the suspected nonconform-ing material, items, and services are determined by the QA Manager to be i 95 ;
nonconforming, . . . " The NRC staff determined that draft nonconformance
,'. reports (NCRs) submitted to the QA manager could be rejected or voided by him as allowed by the procedure. It was, however, found that draft or voided NCRs were not being retained as QA records. Therefore, the disposi-tion or justification for rejection was not auditable because the draft or voided NCRs were destroyed. The NRC staff interviewed former Mercury and other site personnel who had direct knowledge of the NCR system. They acknowledged that such a system may have impeded project personnel from processing NCRs. An LP&L QA engineer stated that LP&L had addressed this problem when Mercury management was supplemented by EBASCO in 1982.
Mercury personnel were then encouraged to write NCRs and as a result the number increased from several hundred to more than 3000. ThisHowever, happened the when about 80 percent of Mercury's work had been completed. NRC staff commented that this increased NCR activity would not necessarily solve the problem of failure to process or document rejection or voiding of draft NCRs that were submitted by personnel who left the site prior to LP&L's initiation of this policy. In con"" n, the NRC staff found that the inadequate disposition of Mercury NCRs ai roper voiding of NCRs is a safety significant issue and has potent ..neric implications for the Mercury NCR system. This issue will require action on the part of LP&L. Acticns Recuired: See Item No. 6 of the enclosure to the D. Eisenhut letter
- to J. M. Cain dated June 13, 1984, and the D. M. Crutchfield letter to J. M.
Cain dated September 19, 1984. l l l-i 96 L
i Task: Allegation A-249
~~
Characterization: 'It is alleged that EBASCO Discrepancy Notices (DNs) were written and given to sub-contractors who would process the DNs through their
.. own Engineering / Quality Assurance (QA) personnel and would then either change the documentation or walk dcwn a system to eliminate the discrepancy at which time the DNs would be " trash canned".
Assessment of Allegation: The implied safety significance of this allegation is that improperly processed EBASCO DNs and lack of adequate corrective action could cause the quality of installation to be questionable. The NRC staff reviewed EBASCO DN control procedures for adequacy and indications of implementation, and reviewed EBASCO DN logs and files for completeness. The staff interviewed personnel involved with the processing of DNs, and selected and reviewed approximately 20 DNs related specifically to system numbers 36 (Component Cooling Water), 52a (Reactor Coolant), and 60 (Safety Injection). Some were for a time period in 1981. The NRC staff review revealed the-following:
- 1. EBASCO procedures addressed the initiation, processing, logging, resolution, reviewing, reporting, and filing of Discrepancy Notices (DNs). These procedures were adequate for the control of DNs.
- 2. The DN logs and files in both the Quality Material Control, Construction
~
Quality Control areas, and the EBASCO QA vault were found to be in order
- with the DN numbers all in sequence. There were no missing or lined out DNs that were unaccounted for. The files also had no DNs that were missing.
- 3. The personnel that were procedurally responsible for the implementation of the DN system, were interviewed by the NRC staff and found to be knowledgeable and conscientious about their responsibilities.
- 4. The random selection of approximately 20 DNs were reviewed by the NRC staff and found to be properly closed, by a specific action or by reference to an NCR number, with appropriate signatures.
Additionally, DNs were being entered into the EBASCO tracking system to assure an adequate status was maintained. In conclusion the NRC staff determined that EBASCO's procedural content and implementation of the DN system was adequate, that all DNs were accounted for, and that adequate corrective action was taken. This allegation had neither safety significance nor generic implications. Actions Required: None. 97
~ , . _ , y ,.m__ - - _ y - _ - , . . , , - _ . , . -
- e. .
Task: Allegations A-253, A-254, A-257
~'
Characterization: It is alleged that Vendor Records Reviewers were not qualifieo to assure requirements were met, and that some vendor records were
.. bad, particularly those for Bergen-Patterson, a subcontractor of Dravo.
Assessment of Allegation: The implied significance of the allegation was that QA records reviewea ano approved by unqualified reviewers could place the records in question. Inadequate QA records for Bergen-Patterson components could place the acceptability of those items in question. The NRC staff's methodology for reviewing the allegation was as folicws: (1) the quality records supervisor was interviewed; (2) the appropriate quality assurance procedures were reviewed; (3) the applicable purchase orders and specifications were reviewed; (4) applicable quality records packages for pipe supports were reviewed for compliance with the purchase order and specification; (5) EBASCO and NRC Audits of Bergen-Patterson Pipe Support Co. were reviewed; and (6) the EBASCO-approved Vendor List was reviewed. For the details of allegations pertaining to Dravo and CB&I see Allegation A-292. The NRC staff assessment of this allegation revealed that it was probably based on a lack of understanding of the responsibility of the quality, records section. By procedure, the responsibilities of the quality records section was to check each quality records packages to assure that it was legible and complete. If a package was not complete or was illegible, a deficiency report was filled cut, logged in the master tracking system, and issued for corrective action. The primary responsibility for reviewing vendor records rested with vendor quality assurance representatives and the Project Quality Assurance Engineers, who were assigned to particular engineered components. In addition, when necessary, documents were sent to the engineering disciplines for review. For example, seismic analyses were reviewed by the appropriate seismic engineering discipline. An NRC staff sample review of quality records packages revealed that EBASCO Vendor Quality Assurance representatives were reviewing all documents, performing all required quality checks, and completing all reports legibly and completely. This was confirmed by the Quality Records review group. The results of an April 1982, inspection of Bergen-Patterson, by NRC was reviewed. The inspection covered the engineering and design functions of the Bergen-Patterson offices in Pittsburgh, Pennsylvania and reveales that Bergen-Patterson had an acceptable QA program covering design and engineering. The NRC inspection was indications that Bergen-Patterson's C of Cs were valid for engineering and design of pipe supports. An NRC review of the EBASCO Approved Vendor List revealed that audits of all Bergen-Patterson facilities had been performed including the manufacturing facility at Laconia, New Hampshire. In addition, the approved vendor list showed that the Laconia facility possessed an ASME Code stamp (NPT)N-1217, which expires September 8, 1984. An audit of the Laconia facility dated-
- January 1983 was also reviewed. The Audit covered the Bergen-Patterson QA program compliance with the purchase specification. The results of the audit demonstrated that Bergen-Patterson had a quality assurance program that meets 98
all requirements of the purchase specification. Sufficient indications was
. provided to verify that the manufacturing of pipe supports complied with the
. - C of C.
- In conclusion, the staff's review revealed that the allegation has neither
- - safety significance nor generic implications.
Actions Recuired: None. 0
. b .
a 9 6 ee 9 99
i Task: Allegation A-263 Characterization: It is alleged that a Mercury Construction Superintendent
, did not agree with fiercury audit findings and refused to correct prcblems . - identified.
Assessment of Allecation: The allegation arose from an audit finding in Mercury Internal Auoit No. 12-1-82 (December 12, 1982), which documented that tubing did not have end caps. It was alleged that the Construction Superintendent refused to take corrective action. The NRC staff reviewed the Mercury audit files and noted that'the audit recommendation block had the Construction Superintendent's comment, "Do not agree with your recommendation." The Mercury auditor had also referenced Corrective Action Report (CAR) 124-125, closed out February 27, 1983. The NRC staff learned that a Mercury Construction Supervisor assigned to the area where end caps were found missing had immediately taken corrective action. 2 His superintendent received the audit finding and noted an additional recommendation made by the Mercury auditor, that supervisors and lead personnel do not adequately instruct and train subordinates in correct interpretation of Mercury procedures and that a lack of enforcement exists. The superintendent
,. disagreed with this additional recommendation, not with the missing end cap problem for which corrective action had already been taken.
The NRC staff determined that the Mercury auditor's finding did not contain sufficient indications to support the auditor's additional recommendation; even though the auditor thought the situation to be true, he did not document his finding. The CAR which the auditor referenced, and which he had included in the audit report, stated that disciplinary action was not appropriately applied or given to personnel. The auditor expected disciplinary action to be taken against certain Mercury construction craft personnel. The NRC staff agreed with the Construction Superintendent's conclusion that this type of recommendation was not appropriate to include in an audit report. However, the review of the Mercury audit program identified other problems as described in the NRC staff's assessment of Allegation A-48. This specific allegation regarding the Construction Superintendent's refusal to
. take corrective action has neither safety significance nor generic implications.
Actions Required: None. 4 e 5 l J 100
Task: Allegation A-276 Characterization: The allegation is that in one case the QA Manager for Mercury construction Company directed the Mercury Training Officer to change .. the resume of'one inspector to reflect more experience than the individual actually had. Assessment of. Allegation: The implied significance of this allegation is that if the inspector in question was incorrectly certified to a higher level of performance than his previous experience would allow under ANSI N45.2.6 and the Mercury procedure for qualification of QC inspectors, inspections performed by this inspector on safety-related systems could render the quality of those systems as indeterminent. l The NRC staff, working in conjunction with the NRC Office of Investigation (OI), has determined that the resume of the inspector in question was falsified and that he performed safety-related inspections while incorrectly certified. This allegation has safety significance and is indicative of a generic problem for inspector certification. Actions Reouired: See Item No. 1 in the Enclosure to the D. Eisenhut letter of June 13, 1984 to J. M. Cain (LP&L). 9 S. 101
Task: Allegations A-279, A-280, A-282, A-284, A-288, A-27 Characterization: The allegation is that inadequate documentation reviews were performed by Tompkins-Beckwith (T-B), EBASCO, and Mercury in that open
.'. deficiencies were found in packages which had been reviewed and accepted.
Assessment of Allegation: The implied significance of this allegation is that inadequate review of quality documentation could cause installation to be questionable and inspection results to be invalid because of a lack of assurance of technical content and completeness. The NRC staff reviewed the turnover packages (system QA documents) for T-B, EBASCO Construction (Force Account), and Mercury. The staff also reviewed the following procedures: Mercury Co., QPC 3010; T-B, TBP-20 and TBP-35; and EBASCO, QAI No. 1, QAI No. 9, QAI No. 9A. Mercury, T-B, and EBASCO Quality Assurance Installation Review Group (QAIRG) had reviewed 100% of the QA documents. Tompkins-Beckwith (T-B) In its QA documentation review, T-B found the turnover packages to be acceptable. A subsequent review by EBASCO QAIRG revealed problems such as I incomplete / incorrect documentation. T-B took actions and corrected the
. deficiencies, as confirmed by a later EBASCO QAIRG review. The NRC staff reviewed selected documentation for technical adequacy, completeness, and indications of proper contractor and EBASCO review and concluded that QA
. documentation was adequate.
Mercury Company A review was conducted by the NRC staff that included six Mercury Operational Control Records (OCR). No open Nonconformance Reports (NCRs) were found in the Mercury documentation reviewed. OCR Package 1782, drawing 172-L-012-C, Revision 4, had a handwritten note on it identifying two lines, DPT-RC-9116 SMB (HP) and DPT-RC-9116 SMA (HP), where the separation criteria had been violated. Subsequent to the problem identification by the NRC, LP&L initiated an Engine'ering Discrepancy Notice (EDN) to document this problem and subsequently generated an NCR (W3-7702). This' problem was not identified during the review and closecut of the Mercury OCR packages and as such was an open item. In addition to the NCR, LP&L was looking at all OCRs in Startup System (SUS) 52A for any other separation criteria violations. The NCR analysis showed that the expansion loops, which were the cause of the violation, can be removed (reworked) thus eliminating this problem. The safety significance and generic implications will depend on the outcome of LP&L's review of the rest of SUS-52A. In OCR package 1020 for line DPT-RL-120/121 (LP), duplicate weld record sheets were found for field welds FW-1, 1C, 10, 51 and 52. One weld record showed these welds as deleted and the other sheet showed them as being acceptable. A walkdcwn of the system revealed that these welds are not a part of the
, ins.tallation and that the drawing accurately reflects what is installed. LP&L has initiated the required actions to remove the inaccurate record from the OCR package. The NRC staff considered this is an isolated occurrence and found no safety significance or generic implications.
102
. EBASCO Construction (Force Account)
The NRC staff review of EBASCO Force Account quality documentation turnover
. packages disclosed that they were, in general, adequately reviewed. The exception was work accomplish,d e as a result of the disposition of nonconformance reports (NCRs). NCR work packages were closed out and submitted to the . vault without an independent review by EBASCO QA or QAIRG - Upon detection of this problem by the NRC staff, EBASCO instituted the folicwing corrective action: (1) the applicable procedures were revised to. assure
- independent review by EC-QA and QAIRG; (2) affected QA/QC personnel were instructed, in formalized training classes, of the procedural changes; and (3) the number of affected packages (twelve) were identified and submitted to EC-QA and QAIRG for review. The NRC staff re-r.eviewed five of the twelve work packages and found them to be acceptable.
Basedupontheabovereviews,theNRCstafffoundthatEBASCOhadan'ade4uhte system for identifying and resolving deficient items in work packages. ESASCO also informed LP&L System Startup about system status via a cover letter noting exceptions as " System Turn-Over Exceptions." Indoctrination and training of personnel performing document reviews was reviewed by the staff and is documented in Allegations A-109 and A-294. See
~
, also Allegations A-143, A-150, A-162, A-163, A-188, A-190, A-191 and A-193 for more details in this area.
In conclusion, the reviews of contractor and EBASCO QA documentation by the NRC staff revealed the procedural controls and actual documentation to be technically adequate and complete, and that documentation deficiencies were identified and corrected in most cases prior to system turnover. The items with corrective action remaining open were identified as system turnover exceptions. The issue on the instrumentation separation criteria violation was significant and may have generic implications; the other issues have neither safety significance nor generic implications. Actions Reouired: (See Item 3 in the Enclosure to the D. Eisenhut letter to J. M. Gain, June 13, 1984.) h i 9 103 r,-g-- s- o- e,- - ,-m , , - - - - , , .
- +w w - w-,- < - 1 + - - - -n-e - --
, Task: Allegations A-283; A-181 -
Characterization: The allegation is that the process for initiating NCRs was
. discouraging.
-Assessment of Allecation: The implied significance of this allegation is that tne lack of processing cue to the impediments in the NCR process may have prevented adequate identification and timely corrective action on non-conformances and that consequently the installations may be questionable.
The NRC staff reviewed this allegation in conjunction with Allegations A-49 and A-123 using the following methodology: 1) Evaluated EBASCO procedures which were used to control nonconformances; and 2) Evaluated related correspondence and LP&L responses to that correspondence. EBASCO Procedures The NRC staff reviewed the EBASCO procedures that address NCRs and found that both the 1978 and subsequent 1984 revisions provided complex instructions for the initiation and processing of nonconformances and discrepancies. These instructions, by the nature of their complexity, may discourage some individuals from going through the process of initiating nonconformances or discrepancies. The procedure although complex was found by the NRC staff to be an acceptable one.
- Related Correspondence Selected correspondence on this subject was reviewed but conclusive substantiation of this allegation could not be made.
Based on this assessment and other reviews on this subject, the NRC staff concludes that this allegation has neither safety significance nor generic implications. However, it should be noted that the NRC did identify several problems with the NCR system and its implementation (see Allegations A-18, A-33, A-302, and A-232). Additionally, LP&L initiated an exit . interview process to identify employee , concerns in January 1984. Although this system was identified by the NRC Task Force to have some deficiencies (Eisenhut letter to Cain, dated June 13, 1984) the system, when properly implemented, should help resolve this type of an employee concern.
- Actions Required
- None.
es e 104
Task: Allegation A-286; A-289a Characterization: The allegation is that Tompkins-Beckwith (T-B) Records Reviewers falsified heat numbers to demonstrate adequate traceability. ,.- Examples of affected startup systems (SUS) included SUS-36, Component Cooling Water and SUS-60, Safety Injection. Assessment of Allegation: The implied significance of this allegation is that heat numoers may have been falsified and that the quality assurance (QA) documentation may not reflect the actual hardware in the plant, placing the quality of installation in question. The NRC staff reviewed this allegation by evaluating T-B document system procedures and comparing them to the ASME Code and applicable ANSI N45.2
! requirements. The NRC staff found T-B's procedures to be adequate (also see. Allegations A-35, A-172, and A-308).
The NRC staff reviewed a sample of T-B turnover packages and found them to be adequate. This review is discussed in detail in Allegation A-308. Included in this sample were work packages from SUS-36 and SUS-60. No obvious falsification of heat number entries was observed (see Alleg~ations A-35 and A-308). Heat numbers were selected from the T-B work packages by the NRC staff, ad
~
documentation was presented to the satisfaction of the staff that the heat numbers were valid, including back-up material test reports and heat number 4 identification marking on the installed items (see Allegations A-100 and A-172). The staff's review of T-B turnover packages could not substantiate the allegation. The NRC staff interviewed six EBASCO document reviewers who were onsite at the time of the allegation. They all stated they had no problem obtaining access to contractor records. The EBASCO Site Supervisor of the Quality Control (QC) Verification Group stated that "when requested by the EBASCO Document Reviewers, a questionable or missing heat number would be verified by the.QC inspector, who would physically check material for heat number marking" (see Allegations A-97 and A-172). The QC Verification Group was available to any reviewer concerned with falsification of records or any problem with heat numbers. Issues could have been resolved during the review of the packages. Two specific individuals were alleged to have made heat nember changes: one was the T-B Project Engineer, and the other was a T-B QA Coordinator. The NRC staff did not find any indications that the Project Engineer made any changes T-B to QA records. The QA coordinator was authorized to change QA records. Inspector certifications are discussed in Allegations A-28, A-304, and A-285b. The T-B system provided for document changes by authorized personnel. The NRC staff found no indications that any unauthorized changes were made.
. The NRC staff found no indications of falsified heat numbers. This allegation has neither safety significance nor generic implications.
. Actions Required: None.
105
~w -y y,.-
= _ . _ - - - - - . --~ . ~ . - , - . ~ . . - . - . - . - . _ . - - . . - - -
)
l i Task: Allegation A-289(b) l Characterization: The allegation is that quality assurance (QA) documentation of system 60 (safety injection) did not have an adequate review because the EBASCO reviewers were not qualified. Assessment of Allegation: The implied significance of this allegation is that tne review of records for System 60 may have been ineffective because it was conducted by unqualified personnel, and the quality of the installation of this system may be indeterminate. The NRC addressed this allegation by reviewing EBASCO record requirements, record reviewer qualifications, and a sample of records from System 60. The NRC staff found that ESASCO procedures require records reviewers to have a high school education, classroom training, on-the-job training, and to have completed assigned reading lists. The NRC staff selected a sample of turnover work packages from System 60, and reviewed the qualifications of records reviewers who reviewed and approved these packages. Although formal training of records reviewers was not complete at the time of System 60 turnover, this system was reviewed by QA personnel (inspectors) that were qualified to review records. Records reviewer quali-fications are also discussed in Allegations A-06, A-92, and A-294. Work performed by records reviewers is discussed in Allegation A-35.. A problem concerning inccmplete training records for records reviewers was addressed by EBASCO during an audit in October 1983. Subsequent action updated
- and completed those records or addressed open items, and the audit was closed in January 1984. The NRC staff reviewed the audit and concurs with the disposition, action, and closure.
The EBASCO training program was considered adequate by the NRC staff. There are no specific training, qualification, or certification requirements for
- document reviewers in NRC regulations, the ASME Code, or ANSI. standards.
The NRC staff found no indications of records reviews by unqualified personnel. The EBASCO system of records reviewer qualification was adequate and problems I encountered in formal training were properly addressed. This allegation has l neither safety significance nor generic implications. Actions Reouired: None. l l i 106 i
Task: A11egatiens A-290 and A-301 Characterization: The allegation is that the disposition of corrective action i documents was not adequate to correct the deficiency and changes were incorrectly made to QA records. Assessment of A11ecation: The implied significance of this allegation is that if it is true, icentified deficiencies could still be uncorrected, making the installation and inspection of safety-related systems questicnable. In assessing this allegation, the NRC staff (1) reviewed EBASCO and Tompkins-Beckwith (T-B) procedures for corrective actions; (2) sampled T-B QA documents that had deficiencies corrected by either an NCR or by a speed letter, and (3) evaluated selected NCRs, DRs, and DNs for the proper disposition and corrective action. The NRC staff also compared T-B and EBASCO procedures for corrective actions and for processing nonconformances (NCRs), DNs, and DRs, with the applicable requirements of the ASME code, ANSI 45.2, and 10 CFR Part 50, and found them to be acceptable. The use of speed letters and memos was not addressed in these procedures and the NRC staff concluded, after several interviews, that speed letters and memos were only used to provide information or to request information between EBASCO and its subcontractors. The NRC staff assessed the following specific allegations. . 1 Allegation A-290 alleged that after a Deficiency Report was transmitted to Tompkins-Beckwith, it was questionable as to how corrective actions, including walk downs, were accomplished. (The NRC staff reviewed this allegation concurrently with Allegation A-33.) The staff determined that Quality Assurance Instruction (QAI) 9 provided adequate instructions for forwarding the deficiencies discovered during records review to the responsible organization for resolution. It also included actions to be taken by the reviewer should a deficient item require resolution through the implementation of an NCR or DN. However, the staff found that the procedures lacked instructions for followup action or resolution of these deficiencies to assure proper closure. Additionally, the NRC staff found that T-B's procedures did not address how these types of deficiencies would be resolved or closed. Allegation A-301 alleged that about 40% of the changes made to records were not done correctly for both programmatic and hardware deficiencies. Most of these deficiencies were identified at least on a speed letter or on an E3ASCO Deficiency Report. In many cases, the change made was not adequate to correct the deficiency. (See Allegation A-05 for examples of questionable disposition and inadequate or incomplete closure.) Based on a sample review of the
. packages and deficiency reports relating to the package, the NRC staff believes that the~re are still open deficiency reports.
.. The NRC staff also reviewed T-B DNs in Allegation A-302.
-Actions Required: See Item 6 in the Enclosure to the D. Eisenhut letter of June 13, .1984 to J. M. Cain (LP&L).
107
-- . ._~ --- .. .- _. - _ . .
't $. Task: Allegation A-294 Characterization: The allegation is that the personnel us.ed by ESASCO to
- perform documentation reviews were not qualified to adequately perform record
--- reviews in the short period they were given.
Assessment of Allecation: The implied safety significance is that if the E5A5CO recoro reviewers were not qualified to perform the record reviews in the short period they were.given, the quality of records review for the applicable safety-related systems may be indeterminate. The NRC staff determined that the EBASCO procedure for review of records required record reviewers to have previous experience in quality documentation
- and to complete.on-the-job training (0JT). The alleger gave a specific person to contact and the specific time period for this allegation was April to July
, of 1983.
Following a discussion with the person specified, the NRC staff selected resumes of 24 records reviewers who had completed OJT in the time span specified. ! A review of their resumes indicated that all had previous experience in quality assurance, engineering, or design. The staff also reviewed EBASCO indoctrination and training records for the reviewers and found that all had completed the
~
. required . training and OJT required by EBASCO procedures. Previous reviews of allegations regarding documentation have not indicated any problems caused by unqualified record reviewers (See Allegation A-35).
There are no specific training, qualification, or certification requirements in NRC Regulations, the ASME Code, or , ANSI Standards for document reviewers. Regarding the adequacy of reviews "in a short period" of time, the NRC's review of QA documentation revealed these documents had been reviewed adequately and were acceptable. -Also see Allegations A-35, A-183(a), A-197, A-223, and A-230. The NRC staff finds no indications supporting the allegation of inadequate documentation reviews by unqualified personnel, and believes the EBASCO training program was adequate. This allegation has no safety significance or generic implications and is not indicative of any management problems.
- Actions Required: None.
9 GS 9 108
f Task: Allegations A-302; A-307; A-303b; A-306s; A-306x Characterization: The allegation is that lower tier corrective action documents were not being upgraded to NCRs. Also, FCRs, DCNs, and EDNs were issued after-the-fact for nonconformances in lieu of NCRs. Assessment of Allegation: The implied significance of this allegation is that 4 without proper identification of nonconforlances, the proper disposition, timely corrective action, determination of root causes and actions to prevent recurrences cannot be taken; that quality trending cannot be accomplished; and that the requirement of 10 CFR 50 for the reporting of Construction Deficiencies (50.55(e)), cannot occur without the identification provided in an NCR. ! This allegation was addressed by an NRC staff review of selected Field Change Requests (FCRs), Design Change Notic'es (DCHs), and Engineering Deficiency Notices (EDNs). The methodology used in this evaluation included an evaluation of random FCRs, DCNs, and EDNs for the Reactor Coolant, Safety Injection, and Component Cooling Water Systems. Additionally, FCRs, DCNs, and EDNs were selected at random and evaluated from the various document issuance logs. The selected design change documents were then reviewed for content and for the basis of issuance; that is, for whether they were issued "before-the-fact," as-a design change or "after-the-fact," as a nonconformance report. Finally, the staff conducted a system walkdown to verify proper identification and change. control completion. The NRC staff also reviewed Tompkins-Beckwith Discrepancy Notices (DNs) for proper review and upgrading to EBASCO Nonconformance Reports (NCRs). This review was conducted by selecting DRs from the DR log and the QA records vault. .The NRC staff also reviewed Request For Information (RFI)
- i. records and responses to those requests. (See Allegation A-187.) The following is a summary of the NRC review.
Field Change Requests (FCRs) The NRC staff reviewed 63 FCRs and 21 revisi.ons to those FCRs. Of the 63 FCRs reviewed, 35 should have been NCRs (55%). An additional four may have reflected conditions that warranted an NCR, for a total of 39 (61%). EBASCO procedure-ASP-I-4, Design Control dated June 7, 1983, states in paragraph 6.1.4, "FCRs shall not be generated in place of nonconformance reports." The practice of issuing FCRs in lieu of NCRs not only occurred in the past but continued during the staff review, as in'the following examples:
- 1. F-MP-1818 (May 25, 1984). Miscellaneous piping; incorrect weld rod was used; the weld was removed and replaced with correct material. Contractor
- T-8.
- 2. F-AS-3698 (January 6, 1984). Reactor Coolant (RC) Pipe Seismic Support; bolts and embed documentation changed to reflect as-built. Contractor -
NISCO.
.. 3. F-AS-3648 (December 2, 1983). Miscellaneous pipe supports; changed drawings to reflect as-built embed plate size and weld size. Contractor not indicated.
109
, er-v - -. --
- 4. F-AS-2338 (February 2, 1982). RC supports; unacceptable weld gap between beam and embed, because embed was cut too short; riisposition was to change
"~ design; seismic review may be required. Contractor - NISCO. .'. 5. F-MP-1434 (February 8, 1981). RC spool piece installed backwards in pipe chase; disposition was to leave as installed. Contractor not indicated.
- 6. F-AS-1631 (December 20, 1979), and Revision 1 (October 10, 1980). Reactor Coolant pipe support. "Despite many attempts at repair and rewelding, these welds continue to crack." Design changed. Contractor - NISCO.
~7. F-E-3089 (June 21, 1983). Combustion Engineering (CE) supplied enclosures on Reactor Coolant Pumps speed sensor pulse amplifiers. The CE supplied enclosures were replaced with a different enclosure. The replacement was done without CE review or approval. Additionally, no environmental or seismic qualification review had been done prior to change.
Subsequent review by CE identified the enclosures as acceptable, however the gaskets on the new enclosures had high chlorine content and were unacceptable. An FCR was generated to replace the gaskets with " qualified" gaskets. Contractor - NISCO.
- 8. F-MP-2138 (September 26, 1982). Replaced broken 1-inch valves (cracked
; seats) in the RC system. Contractor - T-B.
- 9. F-MP-2151 (October 1, 1982). Replaced a valve because it would not hold pressure for cold hydrostatic test in the RC system. Contractor T-B.
- 10. F-E-2288 (August 14, 1981). Five cables were pulled through wrong conduit.
The corrective action was to change documentation; that is, to reflect cable routing as-built, in lieu of correct routing. Contractor - Fischbach
& Moore.
Design Change Notices (DCNs) The NRC staff reviewed 14 DCNs and 5 revisions to those DCNs; of the 14 DCNs reviewed, 4 should have been NCRs (29%). The misuse of DCN, although not as prevalent as the FCR misuse, is a serious problem due to the fact that the NCR system was circumvented and adequate corrective action was not taken. The DCN system allows the option for the QA review to be determined by Engineering, consequently most of the DCNs were not reviewed by QA. A proper review by QA could have turned DCNs into NCRs, as in the following examples:
- 1. DCN-703 (September 24, 1982), and Revision 1 (September 27, 1982). Called for replacing four 3/4-inch valves with 1-inch spare valves in the RC System because the 3/4-inch valves had cracked seats. Revision 1 identified an additional seven valves with the same problem. This clearly should have been reported under 10 CFR 50.55e, or 10 CFR 21. This issue appears to be related to FCR-MP-2138.
- 2. DCN-IC-478 (March 30, 1981). Unidentified or misidentified valves in the warehouse were retagged via this DCN. This should have been an NCR.
Contractor - EBASCO. 110
~
- 3. DCN-ME-30 (January 6, 1983). " Replacement of unqualified solenoid valves
- with Class IE environmentally qualified valves..." The subject system was
*
- the RC System. The reason for change, to meet environmental qualification requirements..." (IEEE 323 is dated 1974; this is not a new requirement.)
i DCN-ME-30, Revision 1 (May 5, 1983). Stated " Revise sol.enoid model to document As-Built." The NRC could not determine if a properly qualified correct model solenoid was actually installed.
- 4. DCN-E-790 (February 8, 1982). Revised a cable routing list to reflect as-built conditions.
Engineering Discrepancy Notices (EDNs) The NRC staff reviewed 76 EDNs for proper identification and control. An additional 35 were identified as being improperly " voided" and I was noted open with no action ever taken. Of the 76 reviewed, 51 should have been NCRs. Of the total, 5 were turned into NCRs, the balance of 46 (60 percent) should have been NCRs. Additionally, 3 EDNs were identified non-safety related that should have been safety-related. The 46 EDNs, which should have been NCRs, plus the 14 safety-related EDNs of- those incorrectly voided, yield a total of 66 percent
' which should have been NCRs, or which were improperly processed. The 35 EDNs
" voided" were the result of actions by a clerk via a " speedy-memo" because
- these 35 EDNs could not be located, they were in effect lost or,were never processed. In accordance with Section 6.3.2 of EBASCO procedure ASP-IV-70, Handling of Engineering Discrepancy Notices, "EDNs that are safety-related shall be forwarded to the Quality Assurance Supervisor for his concurrence . . .
. and upgrading to an NCR if required . . . " Since the site engineering group determined what EDNs were routed to QA for this review process, and many were not forwarded to QA, part of the system breakdown appears to have occurred in
< this area. Also there were no indications that the QA Supervisor reviewed all safety-related EDNs forwarded to him, which accounted for the remaining part of the breakdown.. There is also no method in the procedure to void EDNs. The following examples are indications of misuse of EDNs in lieu of NCRs: l 1. EDN-EC-1476 (September 6, 1983). Stated " Weld No. W101 on Whip Restraint l R-BD-2-R23, for System 19-16, that the MT or PT on the weld root pass was bypassed." The disposition was "use-as-is" based on acceptable UT.
- 2. EDN-E-1548 (November 19, 1983). Stated " Safety related cable was damaged."
The disposition was to repair. The approval of the corrective action and reinspection was signed by the same individual. The signatures were later crossed-out and replaced with initials. The approval and reinspection are both indeterminate. Also the signatures were made by an engineer, not by a qualified inspector. l* l -
- 3. EDN-EC-1502 (October 5, 1983). Stated " Control room and control panel conduit was not installed per the drawing requirements." The disposition j
was to issue an FCR to change the drawings. i 111 l t
I
- 4. EDN-EC-1479 (September 8, 1983). Stated "A snubber was installed, RCSR-4167 l (Reactor Coolant), as a seismic Category 1 and safety classification 1.
. . The snubber was procurred on a purchase order as non-safety related (total number ordered on this purchase order was 4)." The disposition was
. " Install the snubber as received. QA records to review documentation to
- verify snubbers are acceptable as safety-related items. If documentation does not-support safety related requirements, ESASCO Purchasing Department is to obtain documentation from the vendor." This EDN was later voided by a speed memo with no explanation provided. Documents attached to this EDN included Bergen Patterson C of C dated May 4, 1982, certifying hangers "7738-02 and 7838-03 Random Short Form" were manufactured in accordance with B31.1. Also attached was the Bergen Patterson Bill of Lading stating
" Inspection Not Required, Standard Travel Stops, CMTR not Required, Project - Nuclear Plant." Status of these snubbers was indeterminate.
- 5. Thirty-Five (35) voided EDNs - The voiding was accomplished.via a speed-memo; M of the voided EDNs were identified as safety related. The following chronology outlines this issue:
o September 20, 1983 - The EBASCO Construction Administrative Engineer issued a memorandum distributed to 28 individuals stating that "the attached list identifies EDN numbers assigned to you for which we have not received the original EDN for distribution. This is an indication that EDNs have not been written'or have been voided." He further states that unless his department received information regarding the EDNs-by October 1, 1983, they would be ' voided. o October 27, 1983 - Later the Administrative Engineer issued an additional memorandum stating all EDNs with the exception of those on the page 2 of the attacted list (containing 36 EDNs) had been clarified. o December 12, 1983 - An engineering clerk, issued a speed-memo.to QA Records (Supervisor) stating 35 EDNs-were
" voided." The clerk had previously received a memorandum from the Administrative Engineer
, " voiding" the EDNs. 1 The Administrative Engineer was interviewed by the NRC during the week of April 21, 1984, and he indicated that no one made any effort to review the EDNs for content or safety implication of required corrective action prior to Examples of the voided EDNs were as follows: voiding them. s
. 1. EDN-EC-0630 (October 21, 1982). Stated " Inadequate drainage at -35' elevation floor in the RAB - Reactor Auxiliary Building." .
- 2. EDN-EC-ll75 and EDN-EC-1176 (March 18, 1983). Stated " Material on Hold" and "QC Volume AGW QC.1"; the specifics of these EDNs are unknown.
4 112
It should be noted that the EDN logbook indicated these EDNs were
,' initiated by the EBASCQ General Material Administrator who was interviewed and indicated that he had no knowledge of opening any EDNs. Both EDNs were marked safety related.
- 3. EDN-EC-1140 (March 2, 1983). Stated " Operators for valves 3FW-V-607A and 6CD-V343 are installed on the opposite valves." Although the EDN was closed, the actual installation was not verified or was incorrectly verified. The NRC inspection of these valves revealed the folicwing:
i o Valve 6CD-V343 has the operator tagged and identified on the name plate as 3FW-V605B.
^
I o Valve 3FW-V607A has the operator tagged and identified on the name plate as 6CD-V348. An additional tag was also noted attached to the operator ' identifying it as 6CD-V348, a Although the EDN only identifies problems with two valves, four valves were actually involved. Tompkins-Beckwith (T-B) Discrepancy Notices (DNs) The NRC staff reviewed procedure TBP-12, Nonconformances and Discrepancies, that states in section 6.2 "DNs are required to be upgraded to.EBASCO NCRs when
. the following criteria applies..." (as defined in section 4.1).
Nonconformance - A deficiency in characteristic, documentation or procedure
, - which renders the quality of an item or service unacceptable or indeterminate.
Examples of a Nonconformance include: physical defects, test failures, incorrect or inadequate documentation; or deviation from prescribed inspection or test procedures, drawings, Code and Contract requirements. !' The NRC review revealed that T-B failed to upgrade DNs as required. The following DNs are examples that fall into this category: W-6519 The DN was written identifying that a torque required by the flange control record was outside the range of the torque wrench. The disposition stated that an adapter was used but this was not required. Also, it states " accept-as-is" because of the successful hydrostatic test. This does not answer the question of the torque values. The L problem is that the bolts were over torqued. This DN should be upgraded to an NCR because of the incorrect documentation and because the quality of the flange is indeterminate. Service conditions were not reviewed for impact. W-6183 These DNs were written against a torque wrench being used W-63zz outside of its calibrated range. The disposition stated W-6519 -that the torque wrench was not required because the Code
- only requires all bolts to be tightened equally. This does not resolve the problem of the torque wrench use. The bolts used to make up the flange still have torque requirements and the question of under torquing has not been resolved. An NCR covering the torque value of all flanges that have been improperly torqued should have been issued.
113
W-3656 Identified a problem with the incorrect heat number being used for Weld FWII RWIR-1. The disposition of this DN appeared to be invalid because not all of the disposition required items were addressed;
- i.e., the attachments did not show corrected rod slip or the ,
l QC-accepted heat number change on the weld record, W-5755 This DN identified a problem with the heat number for filler material.
.This DN was dispositioned without a justification for the actions taken. The action taken was changing the heat number on a rod slip because it was stated to be a clerical error.
W-742 1.oss of power to ovens for an unknown length of time. The disposition was to bring the ovens back to the hold temperature for eight hours prior to issuance of weld rod. This was a generic problem and a Code violation because the rod or wire was not rebaked es required (see Allegation A-215). W-5917 The DN was issued on a heat number problem for filler material. (See concerns for W-3656 and W-5755 for disposition.) Also, there was no indications that the weld record was corrected or that the QC inspector's failure to note the filler heat number problem, if in fact it was incorrect. W-381 Identified the problem that welds were being painted. prior to the final visual inspection. The recomended disposition was to comply with EBASCO letter F-33795-E. There was no indications that the reinspec-tion was performed utilizing the initial inspection criteria. An
- inspection through paint was unacceptable under ASME and AWS Code requirements. Closure appeared to be invalid.
The following additional DNs should have been upgraded to NCRs: W-1876, W-5824, W-4112, W-5047, W-5692, W-5416, W-6243, W-5916, W-381, W-6349, W-2105, and W-728. Note: There appeared to be a generic problem with heat numbers being entered incorrectly or clerical errors being made on rod slips. Examples are W-5824, W-3656, W-4648, W-4968 and W-4969. A DN, NCR, or CAR was not issued to prevent recurrence. Request For Information (RFI) or Information Requests (irs) i The NRC staff also reviewed the various forms of Request for Information and how those requests were resolved; that is, via a clarification, a Nonconformance Report (NCR), a design change, a Field Change Request (FCR), or a Design Change Notice (DCN). The staff conducted the following review by contractor: l l NISCO i i The NRC staff reviewed the actions taken by EBASCO engineering to resolve 38 f selected RFIs. This review was to determine if the response to the RFI was a clarification, an NCR, or an FCR. The NRC staff also reviewed the supporting documentation for work, repair, use-as-is, or the design change. The RFI 114 l
document was utilized as a tool to correct prcblems, but it was not issued as a method to resolve design problems. The RFI did make reference to the document used to resolve the problem. The staff review revealed that the actions taken were correct and that the quality records were complete and technically adequate. Temokins-Beckwith (T-B) The NRC staff reviewed the actions taken by EBASCO Engineering to resolve Information Requests (irs). A sample of 20 irs were selected from the EBASCO Engineering files. The review revealed that EBASCO responded to the request by clarification (referring to appropriate drawing revision or specification / procedural requirement), NCR, DN, FCR, or DCN. This provided direction to T-B on how to proceed. Design information was not provided on this document; rather to the referenced document containing the required information. In conclusion, the NRC staff found that the QA program requirements for nonconformance identification, control, and proper corrective action was not ccmplied with. The design change control system was incorrectly substituted to report "after-the-fact" nonconformances and T-B DNs were not properly upgraded to NCRs. Additionally, the QA program was, in effect, circumvented, and the required review for 10 CFR 50.55(e) reportability was not accomplished. Actions Required: See Item No. 4 of the enclosure to the D. Eisenhut letter to J. M. Cain dated June 13, 1984. 1 115
J f Task: Allegation A-304, A-305, A-285b characterization: The allegation is that changes t'o quality assurance (QA)
; documentation were made by a Tompkins-Beckwith (T-B) records reviewer who was not a QA inspector.
Assessment of Alleoation: The implied safety significance of this allegation is tnat an unqualified person made changes to QA documentation which could affect the acceptability of safety-related piping systems. The NRC staff reviewed T-B procedures for documentation review of piping systems and for certification of personnel reviewing piping documentation.-The staff also reviewed the certification and resume of the records reviewer who made changes to the QA documentation. The NRC staff found that the reviewer was correctly certified as a records review specialist in accordance with T-B procedure and the procedure met the requirements of ANSI N45.2.6, 1973. The staff noted that T-B procedures did not require changes to be made only by QA inspectors, and that records review personnel with the required indoctrination and training could make changes to QA documentation. ! ihis allegation has neither safety significance nor generic implications and is not indicative of any management problems.
^
Actions Required: None. e I e 116
Task: Allegation A-306c
- - Characterization: The allegation is that there was a lack of EBASCO' documents for traceaoility.of bolts one inch or less.
Assessment of Allecation: The implied significance of this allegation is that if the neat numoer traceability requirements of the ASME Code and 10 CFR 50, Appendix B, Criterion VIII were not' met, the acceptability of safety-related systems could be indeterminate. The requirements for the traceability of, bolting material of one-inch diameter or less during receipt, storage, and installation are acdressed in ASME Code ' Section III, NB-2150, NC-2150, and NA-3766.4, and Criterion VIII of 10 CFR 50, Appendix B. Heat code traceability requirements from the material supplier was documented by a Certificate of Compliance (C of C) or a Certified Material Test Report (CMTR). The NRC staff reviewed samples of these C of Cs and CMTRs and found them acceptable. The type of certification required foi heat number traceability documentation was dependent upon the requirements of the ASME Code, EBASCO specifications, and the EBASCO Purchase Order. The NRC staff found that the Bill of Material and the issuance records documented the heat or code number for installation and denoted the correct material code. For installation traceability, the NRC staff reviewed a selected sample of
- records which revealed that the correct bolting material type (i.e., SA-307, SA-193, SA-194) was actually installed. The NRC staff found the EBASCO and contractors' system for procurement, acceptance and installation of bolting materials and the method of control for traceability and identification, to be acceptable.
The NRC staff has concluded that this allegation has neither safety significance ~ nor generic implications. Actions Recuired: None. 117 r .-. - - - -
Task: - Allegation A-306t
~~
Characterization: The allegation is that noncertified document reviewers were making changes to QA documentation,
. Assessment of Allecation: The implied significance of this allegation is that changes to QA recorcs of safety-related systems made by unqualified persons could affect the acceptability of those systems.
Allegation A-306t concerns the qualification of a specific document reviewer and changes.to a nonconformance report (NCR). This issue was addressed by reviewing qualifications of the subject reviewer, as well as the procedure requirements for records review and quality record reviewer qualifications. The NRC staff found that the person specified was correctly certified as a QA Document Controller. Review of Tompkins-Beckwith (T-B) procedures allcwed changes or corrections to be made to records by records review personnel. There are no specific training, qualification or certification requirements for record reviewers in NRC regulations, the ASME Code, or the ANSI standards. The T-B record review system is considered to be adequate. In conclusion, the person cited in Allegation A-306t was qualified to perform record reviews and the T-B record review system is considered adequate. This allegation has neither safety significance nor generic implications. . Actions Required: None. O O
. l s
118
Task: Allegations A-306u and A-306z Characterization: The allegation is that Tompkins-Beckwith (T-B) did not adequately control measuring and test equipment (M&TE) and that this could cause test and measurements to be invalid. Assessment of Allecation: The implied signif'cance i of this allegation is that inadequate control of M&TE used in activities affecting quality could invalidate the results of test and inspections and render the quality of installations as indeterminate. The NRC staff addressed this allegation by reviewing the T-B procedures and quality manual section associated with M&TE calibration and control. The procedures establish a sound M&TE control system. A review of work packages that contained hydrostatic testing records was conducted to verify M&TE calibration and control. Two gauges were used during the hydro tests and both were calibrated prior to and after each test. The hydro test results were not accepted until the post-test calibration was complete and acceptable. The acceptance criteria specified that one gauge must exhibit proper post-test calibration. This system was.also observed in the Mercury Company packages. All of the calibration records reviewed had a quality control. signature for observation of the calibration and review of the data by quality engineers. The instruments used for the test were documented in the hydro test data. The pre- and post-test calibration sheets were also included in the hydro
. test data. -
The NRC staff generically looked at NCR/DN dispositions for calibration as a part of its assessment of Allegation A-33. Based on the review conducted, the NCR staff believes this allegation has neither safety significance nor generic implications. Actions Required: None. l 0 l l 119
Task: Allegations A-308; A-306j, A-306r Characterization: The allegation is that the piping system documentation of various contractors and LP&L is inadequate and that it cannot be verified that piping systems were installed and inspected in accordance with the ASME Code. Assessment of Alleoation: The implied significance of this allegation is that inadequate cocumentation reviewed and accepted by Tompkins-Beckwith (T-B), Mercury, EBASCO, and LP&L could render the acceptability of installation and inspection of various piping systems as indeterminate. The NRC staff reviewed documentation procedures of various sub-contractors, EBASCO, and LP&L to verify that they were in accordance with the ASME Code, 10 CFR 50, and the applicable ANSI N45.2 requirements. The Staff verified implementation of the procedures by reviewing Mercury and T-B turriover packages for technical adequacy and content. The T-B packages included both large and small bore piping for safety-related j systems. The selected startup systems (SUS) sampled included Ccmponent Cooling Water, Reactor Coolant, and Safety Injection. Packages withdrawn from the vault were sealed and each package had been reviewed by both T-B and EBASCO's Quality Assurance Installation Records Group (QAIRG). Each review was documented on a checklist noting accept / reject status with the reviewer's signature, initials, and date. LP&L had sampled 10 percent of the turnover packages,
. and their review was also documented with accept / reject status.
l Included in the T-B documentation packages were the following quality documents /
- records: Construction Installation Records Review Check Lists signed / dated by QAIRG (1 each for T-B and EBASCO); Traveler Index's; Bill of Material; Weld Control Record (s); Travel Erection Sheet (s); Weld Repair and NCR Index's; Cleanliness Reports (as applicable); Code Data Reports, CMTRs and C of C's; NCRs; Flange Control Documents; NDE Reports; Piping Modification Travelers; Discrepancy Notices; CIWAs; and ISO-Drawings.
l The NRC staff observed that T-B and EBASCO maintained adequate, detailed ! quality records. The accept / reject status was noted by appropriate inspector / l reviewers by signature and date. The NRC staff found that records had been reviewed for accuracy and completeness prior to being filed as historical records, that inspection / test results were documented and traceable to the material, and that records were readily retrievable. l The NRC staff found that the T-B and EBASCO documentation system was adequate. l This portion of the allegation had nc safety significance and was not indicative of any adverse generic implications. The Mercury packages reviewed were for the Reactor Coolant instrument lines. l The packages were stored in the Mercury and EBASCO vaults and retrieved by LP&L for review purposes. Each package reviewed had been previously reviewed
, by Mercury QA and EBASCO QAIRG. The Mercury review was denoted on each 120
document, and overall package acceptance was delineated on the Mercury Documents Requirements Form. EBASCO QAIRG review was documented on the Construction
'. Installation Records Review and Status Check List.
Included in the Mercury packages were the following documents: (1) EBASCO Review Check List (dated and signed by Ebasco QAIRG reviewer); (2) Mercury Dccumentation Requirements Form /RPT-Form 209 (Documentation Index); (3) Operations Control Report (OCR) ccver sheet; , (4) Process Control Traveler; (5) Pipe and Tube Inspection Report; (6) Material Verification-Heat Nos. Traceable, CMTRs, and C of C;
, (7) Weld Data Sheets; (8) NDE Reports;
. (9) Quality Control Reports-acceptance of work; (10) NCRs; (11) CIWAs; (12) As-built / Red-Lined drawings - (latest revisions); and (13) Hydrostatic / Pneumatic Test (as applicable), including (a) Hydrostatic / Pneumatic Test Instructions, (b) Hydrostatic / Pneumatic Data Sheets, (c) Valve Line-up Sheets, (d) Hydro / Pneumatic Test Discrepancy List, (e) Weld Data Sheets (info copy to verify weld no.; not used for B31.1 tests),
(f) Calibration Sheets, and (g) Pip,ing and Instrumentation Drawing (P&ID). 121
The NRC staff reviewed Mercury's quality records and verified the following:
- 1. Documentation of Mercury QA and EBASCO QAIRG review.
- 2. Documentation of Authorized Nuclear Inspector (ANI) review and acceptance of documentation package.
- 3. That-EBASCO QAIRG reviewed 100 percent of the packages and LP&L reviewed packages on a 10 percent sampling basis (similar to Mil-Std 105D sampling techniques).
- 4. That inspection / test results were documented and traceable to installed hardware.
- 5. That inspection weld records, drawings, and other quality records-generically appeared to be complete. However, they were awkward and cumbersome, and it was difficult to follow the traceability of individual records.
- 6. That record retrievability was not timely.
- 7. That weld records were not always original documentation (see Allegations A-197 to A-206).
- 8. That system walkdown disclosed that the installation (as-built) matches final drawings. Two N-5 installation Code Data Reports were found to be
~
- in nonconformance with the drawings.
The NRC staff concludes that this allegation has neither safety significance nor adverse generic implications. See Allegations A-183, 184, 197 to 206, 223, and 230 for additional-information. Actions Reouired: None, e 9m D 122
Task: Allegation A-332, A-333, and A-334 Characterization: The allegation is that (1) DCN-MP-920 is ina'dequate, (2) welder certification is questionable for welding on piping between the spent
~
fuel pool and the low pressure safety injection pump and the suction line from the condensate storage tank to the emergency feedwater puma, and (3) N stamp and Code classification on lines as signed by inspector 307783215 are questionable. Assessment of Allegation: The implied significance of this allegation is that improper design change, unqualified welders, and lack of or improper Code classification of installations could cause the quality of construction to be questionable. The NRC staff reviewed the three allegations as follows: A-332 The NRC staff reviewed DCN-MP-920, dated September 12, 1983, for technical content and adequacy. The intent of the DCN was to designate system code classification boundaries on piping from tanks, pools, canals, and connecting nozzles. The system boundaries were clearly defined and the DCN was considered appropriate for the intent. A-333 The staff reviewed welder certification records for personnel
. performing welds on the low pressure safety injection pump suction piping from the fuel storage pool, and the suction line to the emergency feedwater pump from the condensate storage tank. Field . weld records were compared with welder artification records and no violations were noted.
A-334 The staff found Code classification of the lines for the tanks, pools, and spent fuel transfer piping were correctly designated on DCN-MP-920 and were fabricated to the requirements of ASME Code Section VIII using the material requirements of ASME Section III. The subject piping was classified as required to be non-safety or safety-related with the applicable seismic category or 10 CFR 50, Appendix B requirements assigned. This type of classification, l using the various codes for one system, was common practice in the nuclear industry. l In conclusion, the NRC staff found the design change, the welder qualifica- ! tions, and the code classifications to be acceptable. These allegations have neither safety significance nor generic implications. Actions Required: None. 66 123
4 Task: Allegation A-340 Characterization: It is alleged that cable trays were not cleaned prior to installation of cable tray covers. Assessment of Allecation: The implied significance of this allegation is that deoris left in cable trays could cause or propagate a fire. Cable tray covers were being installed by EBASCO in accordance with Administrative Site Procedure ASP-IV-48, Revision N, " Permanent Plant Equipment Removal and Reinstallation Procedure". 4 In assessing.this allegation, the NRC staff examined Equipment Service Form 84-3-21, Sheets 3 and 4 of 39, which was an example of the detailed procedural
- requirements for the installation, inspection, and documentation of cable tray cover installation activities. Item 5 on sheet 3 of 39 states " Verify cleanliness immediately prior to cover installation." A review of the installation and inspection documentation (sheets 5 thru 39 of 39) revealed the following:
- 2 installations were inspected for cleanliness the same day as covers were installed.
- 6 installations were inspected for cleanliness 6 days before covers were installed.
- - 8 installations were inspected for cleanliness up to two weeks before i ,
covers were installed.
- 7 installations were inspected for cleanliness up to three weeks before covers were installed.
- 10 installations were inspected for cleanliness over one month before covers were installed.
Based on this review, the NRC staff concluded that the requirement for QC verification of tray cleanliness "immediately prior to cover installation" was not, in many cases, complied with. Because of this discrepancy in procedural conformance, the staff requested that LP&L remove installed tray covers in various plant areas to determine if the trays were, in fact, clean. Covers were removed by LP&L and inspected by NRC in the following areas:
- Reactor Auxiliary Building, Elevation +21, Switchgear Room A (3 Sections);
- Reactor Auxiliary Building, Elevation +21, Switchgear Room B (3 Sections);
- Reactor Containment Building, Elevation +21 (1 Section below floor grating);
- Reactor Containment Building, Elevation +35 (1 Section below floor i grating).
4 124
No debris was found in any of the trays inspected.
- In summary, although many tray sections were inspected for cleanliness well in advance of actual cover installation, the basic requirement that trays be . " clean" at the time of cover installation, was acceptable. This allegation has neither safety significance nor generic implications.
Actions Required: None. e
/
m e
~
125
Task: Allegation A-341 Characterization: The allegation is that deformed cable trays were " cinched up witn a come-a-long" in order to install the cable tray covers. Assessment of Allegation: The implied significance of this allegation is that the metnoc of using a mechanical device to compress cable trays to install the covers may have caused damage to cables installed in the trays. Since the alleger did not give specific locations where such instances occurred, the NRC staff inspected 1800 feet of cable tray as follows: o Reactor Auxiliary Building: +21, -4, and -35, elevations, o Reactor Containment Building: +21, +35, and +46 elevations. The NRC staff's inspection revealed no instances of damaged or deformed cable trays that would have indicated damaged cables. This allegation has neither safety significance nor generic implications. Actions Required: None. 9 9 O d 126
CONCLUSIONS Based on its review of 217 allegations and related information received to date of poor construction practices at the Waterford 3 site, the QA Records Review Team made the following observations and conclusions.
- 1. Quality Assurance activities during most of the construction period were principally delegated to the major contractor, EBASCO, by the utility. The lack of a fully staffed and effective utility QA program, along with EBASCO's failure to fully carry out the QA responsibilities delegated to them, led to quality problems during construction.
- 2. The QA Records Review Team verified the allegations that the project document requirements were not fully met. Failure to meet these requirements contributed to the quality problems at Waterford.
- 3. The number and diversity of the QA problems found is indicative of a program that is not meeting its objectives and requirements. The majority of the problems evaluated to date, in and of themselves, do not have an adverse impact on safety.
- 4. Potential violations of NRC regulations identified by the Task Force are being forwarded to the NRC Office of Inspection and Enforcement and to the Administrator, NRC Region IV Office for appropriate action.
e 37404 127 l
- i u Oa<Nua u,e:~ riOC. ve,N..,#-,
u s ,.uct A. a1vi.ArO., e g,,0===
%',"2#2- BIBUOGRAPHIC DATA SHEET EGG-RST-6838 sit iN3YmuCTsON$ ON fue mEvtast J LEAvt SLANK
& TITLE ANO sustiTLE WATERFORD III PROJECT 1 REPORT 4 OATE mEPOmiCOM*Liff 0 MONTH TEAR
, . Au r ,.O. ,,, Apri1 1985
. O Ari anaOar issuno Compiled by B. L. Barnes .ONr. veA.
j April 1985 S PROJECTtTA5E/womE UNIT NuwSER 7 sEmsomWiNG ORGami2Af TON NAME ANo uasLING ADORE 55 tsachele Cees Pro.iect 1
* *iN Oa GaANT Nuana EG&G Idaho, Inc.
Idaho Falls, ID 83415 A6819 Ita TYPt 08 REPORT
- 10. 5PONSOReNG ORGAmi2 ATION NAME AND MasLsNG ADORE 55 ffachele Ceses Division of Licensing Office of Nuclear Reactor Regulation Regulatory U.S. Nuclear Regulatory Commission * ** a'a mv 'a'o "- ~~
Washington, D.C. 20555 u sue u NrAav NO'*5 This report was compiled from NUREG 0787, Supplement 7, to document the primaryactivities of EG&G Idaho personnel who supported the NRC to resolve the Waterford III auality assurance related allegations. 13 A$$TR ACT fJOE? oores er 'esss Eleven EG&G Idaho personnel participated as members of a nineteen person NRC Quality Assurance Records Review Team at the Waterford III Nuclear Plant site. The team members reviewed 217 allegations pertaining to the overall quality assurance program at Waterford III, interviewed allegers to clarify these allegations, and made assess-ments of allegations. Quality assurance related allegations are listed herein along witn a description of the methods used to assess and resolve these allegations. 19 Av A46A44 Ls r v 14 DOCuwt NT ANALvlis - e a tinomO5 OtSCM*PTOMS "^**'a Pressurized Water Reactors Unrestricted l
~
Construction Distribution j Quality Assurance 16 SECumiTV CLAs$18tCATION tr. ur,# o iO NTisitas,'OrtN eNoeO fsaus Unclassified Allegations Waterford Unit No. 3 ~ Unclassified
"***""*5 Quality Assurance
.....Ct r}}