ML20059H896
| ML20059H896 | |
| Person / Time | |
|---|---|
| Issue date: | 10/26/1993 |
| From: | Mcintyre R Office of Nuclear Reactor Regulation |
| To: | |
| Shared Package | |
| ML20059H858 | List: |
| References | |
| REF-QA-99900403 NUDOCS 9311100260 | |
| Download: ML20059H896 (20) | |
Text
,
ORGANIZATION:
CE Nuclear Energy San Jose, California r
REPORT NO.:
99900403/93-02 CORRESPONDENCE Mr. Patrick W. Marriott, Manager ADDRESS:
Licensing & Consulting Services GE Nuclear Energy 175 Curtner Avenue San Jose, California 95125 ORGANIZATIONAL Mr. Kenneth W. Brayman, Manager CONTACT:
Quality Assurance Systems (408) 925-6587 NUCLEAR INDUSTRY GE Nuclear Energy (GE-NE) is engaged in the supply of ACTIVITY:
advanced boiling water reactor designs to utilities.
GE-NE also furnishes engineering services, nuclear replacement parts, and dedication services for commercial grade electrical and mechanical equipment.
INSPECTION CONDUCTED:
September 7 through 10, 1993 SIGNED:
Aan~
c V
IC T3 Richard P. McIntyre, leam LeadeE Dats Reactive Inspection Section No. 1 Vendor Inspection Branch (VIB) 4/
LO % [q'3 APPROVED:
OrUldisPotapovs, Chief
/
Date Reactive Inspection Section No. 1 Vendor Inspection Branch (VIB)
INSPECTION BASES:
10 CFR Part 50, Appendix B and 10 CFR Part 21 INSPECTION SCOPE:
To determine if quality activities performed as part of the design of the Advanced Boiling Water Reactor (ABWR) project were conducted under the appropriate provisions of the GE-NE 10 CFR Part 50, Appendix B, l
quality assurance program, as implemented by the Quality Assurance Program Description (NEDO ll209-04A) that has been approved by the NRC.
PLANT SITE APPLICABILITY:
None l
9311100260 931028 PDR GA999 ENVGENE 99900403 PDR
1 INSPECTION
SUMMARY
1.1 Nonconformances 1.1.1 Contrary to Criterion III of Appendix B to 10 CFR Part 50 and Section 3.10 of the GE-NE QA Program Topical Report, NED0-Il209-04A, changes made to the computer codes REDYA, ODYNA, and SAFER were not being documented and verified in any formal manner. No official records of the changes are required to be kept and the changes themselves do not have to be reviewed independently.
(93-02-01) 1.1.2 Contrary to Criterion III of Appendix B to 10 CFR Part 50 and Section 4.4.1 of Engineering Operating Procedure (EOP) 40-3.00, " Engineering Computer Programs" (ECPs), the flow area of the internal recirculation pump used for the modeling of the SAFER code was based on an unverified hand drawn sketch with a reference to an individual who provided the information, instead of a reference to the applicable dimensioned design drawing.
(93-02-02) 1.1.3 Contrary to Criterion XVII of Appendix B to 10 CFR Part 50, GE-NE had not filed the quality assurance documents or the test log requiad by Test Plan and Procedure (TPP) TP-515.1078, "ABWR Full Integral Simulation Test (ABWR FIST)," Revision A, dated October 17, 1983, for the ABWR FIST tests in DRF E00-149 and subsequently could not produce the documents.
(93-02-03) 1.1.4 Contrary to Criterion XII of Appendix B to 10 CFR Part 50 and Sections 1.1 and 4.2.b of E0P 35-3.20, " Calibration Control," GE-NE purchased thermocouples used in the ABWR FIST tests from a commercial grade supplier, not on GE-NE's approved supplier list, and accepted and used the instruments as calibrated by the supplier without further verification of the quality or traceability of those calibrations.
(93-02-04) 1.1.5 Contrary to Chapter 17 of the ABWR Standard Safety Analysis Report (SSAR) which commits to ANSI /ASME NQA-1-1983, the calculation notebooks for the inputs to the SAFER, REDYA and ODYNA computer codes did not have a sufficient level of detail and in some cases were inadequately referenced.
In addition since changes to computer codes are design analyses they should also be documented in a sufficient level of detail.
(93-02-05) 1.1.6 Contrary to Criterion V of Appendix B to 10 CFR Part 50 and GE-NE SSAR for the ABWR, Chapter 17, Section 17.1.2, " Quality Assurance Program," GE-NE failed to perform an annual implementation review of Hitachi and Toshiba's QA program for the 1991 period.
This failure resulted in a 16 month interval between the audits performed in 1990 and the 1992 audits.
(93-02-06) 1.1.7 Contrary to Criteria V and VII of Appendix B to 10 CFR Part 50 and Section 7 of the GE-NE QA Program Topical Report, NED0-Il209-04A, " Control of Purchased Material, Equipment and Services," GE-NE failed to perform audits of Bechtel's ABWR QA Program Plan implementation for engineering services associated with GE-NE PO No. 190-ALWR-31387, and accepted safety-related services from Bechtel without them being listed on GE-NE's Approved Suppliers List for such services.
(93-02-07)
.)_
F 1.2 Unresolved Items 1.2.1 GE-NE has the following statement in Chapter 17 of the ABWR SSAR: "The lead responsibility to produce each specification and drawing is formally assigned to one design organization. However, the content of each document is reviewed and approved by GE-NE. While all comon engineering documents reflect the formal consensus of all parties, GE-NE is responsible for the 1
design and supporting calculations and record.; for the ABWR project."
It is not clear to the staff how GE-NE has met their SSAR commitment.
The ABWR system design record files (DRFs) have the Japanese plant (K6/K7) system design specification, process flow diagram (PFD), piping and instrument diagram (P&lD), and instrument block diagram for each system.
These received a formal GE-NE review via Engineering Review Memoranda (ERMs) and the resolution of comments is well documented. However, there is a scarcity of information on supporting calculations, particularly for those systems where the technical associates had the design lead. GE-NE had not documented a review of the supporting calculations for the reactor building cooling water system and the audit process of the technical associates did not examine the technical adequacy of the supporting calculations.
(93-02-08)
~
1.2.2 The GE-NE ABWR SSAR (Document 23A6100) contained inconsistent design information in that Figure 9.2-la, the PFD, did not depict the US ABWR reactor building cooling water (RCW) system configuration and Table 6.3-1 contained the incorrect main steam flow rate.
The team identified that the system P&lD, Figure 9.2-1, sheet I of 9, represented the ABWR configuration with 3 heat exchangers while the associated PFD, Figure 9.2-la, showed only 2 heat exchangers that is representative of the Japanese plant (K6/K7) design.
The system flow and pressure drop information on the PFD had not been re-calculated for the ABWR configuration as the analysis had been performed by the international associate. The team also pointed out that the main steam flow rate listed in Table 6.3-1 of the SSAR was inconsistent with the value used in the computer code SAFER 03 input.
(93-02-09) 1.2.3 The team reviewed the RCW system DRF, P21-00001, and determined that within the DRF there were several pages (sheets 554-560) of unchecked /
unverified calculations that evaluated the ABWR system differences from the K6/K7 design, including additional heat loads and the addition of a third heat exchanger. These calculations therefore would support the US ABWR certification.
The evaluation was very informally done and was not sufficiently detailed as required by ANSI N45.2.ll-1974, " Quality Assurance Requirements for the Design of Nuclear Power Plants," with respect to:
purpose, method, assumptions, design input, and references so that a technically qualified person could review and understand the analysis without recourse to the originator.
In addition, E0P 42-10.00 states that when a DRF is closed, the completed record shall be reviewed to ensure design verification requirements, where applicable, have been met.
i The RCW calculations that extrapolated the K6/K7 design to the certified ABWR design were performed in a manner not consistent with the GE-NE QA topical report (NED0-11209-04A) commitment to ANSI N45.2.11-1974.
(93-02-10) 2 STATUS OF PREVIOUS INSPECTION FINDINGS No previous inspections have been conducted in this area.
3 INSPECTION FINDINGS AND OTHER COMMENTS 3.1 Ouality Assurance Proaram The quality assurance (QA) program implemented for the ABWR program is described in the Advanced Light Water Reactor Program Quality Assurance Program Plan (QA Plan) that was prepared for the Department of Energy, Contract No. DE-AC03-86SF16563. This DOE QA Plan mandates the application, as appropriate to the contract scope, of the QA program described in Pevision 6, dated June 30, 1986, of the NED0-11209-04A, " Quality Assurance Program Description." This is the GE-NE topical report that has been reviewed and approved by the NRC and meets Appendix B of 10 CFR Part 50.
?
The DOE QA Plan contains a work element / implementing procedure matrix that contains 18 major subdivisions which correlate with the 18 criteria of Appendix B.
The 18 subdivisions are further broken down into 78 work elements committed to the QA Plan.
Four types of GE-NE procedures are described that implement the work elements of the QA Plan. These are Nuclear Energy Business Operations (NEBO) Policies and Procedures (P&Ps), BWR Engineering Operating Procedures (EOPs), Nuclear Systems and Technology Operation (NSTO) Policies and Instructions, and Nuclear Service Procedures (NSPs). During the 1986-1987 time frame, the NSPs were phased out and folded into the E0Ps.
The NEB 0 P&Ps are high level GE-NE policies that establish overall policies and responsibilities for NEB 0.
As a result of a reorganization, GE-NE nuclear activities are currently under the Vice President of GE Nuclear Energy (GE-NE) and NEB 0 no longer exists.
The E0Ps are a series of procedures that implement GE-NE policies and the QA program. NST0 Policies and Instructions deal with subjects such as cost schedules, budgeting, contract award, and business management and as such have no importance in implementing the QA Plan. The NSPs have been subsumed by the E0Ps.
The team reviewed the organizational hierarchy and found that several QA organizations are involved with performing verification activities for the ABWR design efforts. All of these organizations appeared to have the necessary independence to carry out their charter. Documentation reviews.and personnel interviews were conducted with selected staff from the various GE-NE organizations involved with implementation of the GE-NE QA program. This included Nuclear _ Quality A urance (NQA), Services Quality Assurance (SQA),
l Advanced Reactor Programs (ARP) Quality Assurance, and ARP engineering and management staff.
Selected aspects of the QA program elements were examined i
in further detail as described below. 4 f
I 3.1.1 Scope Of QA _ Program Implementation for Desige, f
The team was informed that all ABWR engineering work is performed in accordance with the guiding Engineering Operations Procedures (EOPs) and that none of the work is classified as non-safety-related. The generic application r
of E0Ps was observed during the course of the inspection.
3.1.2 Quality Council The Quality Council is comprised of QA representatives from a number of GE-NE organizations.
The purpose of the council is to assure uniformity in application of the QA program and to resolve QA issues.
The team reviewed the i
minutes from the following Quality Council meetings: 3/19/93, 12/14/92, 7/9/92, and 3/20/92. Only a cursory mention was made in the minutes of one issue related to the ABWR, that being a letter from a Japanese utility involving positive audit results of GE-NE. The meeting minutes did not reflect any other Quality Council review of ABWR issues. The NRC Safety Evaluation Report (SER) for the ABWR design certification refers to the Quality Council as an aid for NQA to fulfill its responsibilities.
It was not evident that the Council was making substantive contributions with respect to ABWR quality issues.
3.1.3 (ngineering Training The team reviewed E0P 70-30, " Personnel Proficiency in Quality Related Activities."
The E0P states that employees shall be trained on the quality system and that they shall read or be instructed in applicable procedures.
The team randomly selected two lead engineers associated with the ABWR project. Their self-study indoctrination records were reviewed.
Each document stated that the engineers had read and understood the pertinent quality procedures.
3.2 Instractions and Procedures 3.2.1 NQA Procedures The team reviewed the following NQA Practices and Procedures:
- 1.1 " Files and Records"
- 1.2 " Standard Distribution List"
- 2.1 " Conduct of Audits"
- 2.2 " Preparation of Corrective Action Requests"
- 2.3 " Audit Corrective Action Performance Report"
- 2.4 " Auditor and lead Auditor Qualification and Certification" The procedures were found consistent with the governing QA policies. The procedures provided working level directions for performance of NQA activities.
9
[
3.2.2 Engineering Operating Procedures (EOPs)
The team reviewed a selection of E0Ps that govern the implementation of the QA program by both design / engineering and QA personnel. The following procedures were examined:
- E0P 55-2.00, " Engineering Change Control" E0P 60-3.10, " Engineering Records"
- E0P 65-2.00, " Product Safety Requirements"
- E0P 65-5.00, " Licensing Documentation"
- E0P 75-2.00, " Qualification and Certification of Personnel"
- E0P 75-3.00, " Corrective Action and Audits" E0P 65-2.00 specified that Product Safety Requirement (PSR) documents are to be prepared that define the safety and licensing requirements for standard plants. This E0P was found to be in the process of being revised to clearly reflect that the standard plant Safety Analysis Report (SAR) will serve as the PSR. This information was confirmed through discussions with engineering management and review of internal GE-NE memoranda. Thus for the ABWR, the SSAR is treated as a controlled design document from which safety and licensing requirements are translated into other design documentation.
3.2.3 Services Quality Assurance (SQA) Procedures The instructions utilized by the SQA QA group were reviewed.
In particular the following procedures were examined:
- AG-004, " Corrective Action"
- AG-008, " Processing Quality Records"
- AG-017, " Internal Audit Scheduling, Internal Audits and Auditor / Lead Auditor Qualification / Certification" As discussed in section 3.9 of this report, the team identified to GE-NE that while the procedural instructions for audits discuss the Corrective Action Requests and recommendations for handling audit findings, several audit reports were found to contain a variety of findings called: observations, unresolved items, and concerns.
3.3 Document Control Document control is prescribed by NEB 0 P&P 70-11, " Quality Systems Requirements" and numerous E0P's such as 15-2.00, "EOP Application;" 30-5.00,
" Engineering Records Documentation Supplied by External Sources;" 40-7.00,
" Design Reviews;" 42-5.00, " Engineering Requirements Document Release;"
42-6.00, " Independent Design Verification;" 42-8.00, " Document Issue and Application by ERM;" 55-2.00, " Engineering Change Control;" and 60-6.00,
" Drafting Manual Control."
The team found poor control of design and calculation documents related to the FIST, SSAR calculations, and computer code modeling.
Some calculation documents were kept in individuals' desks, and original design drawings were i
kept in an unlabelled shop drawing file.
Refer to Sections 3.5 - 3.8 of the report for details.
3.4 Quality Assurance Records Quality Assurance records are prescribed by NEB 0 P&P 70-11, Quality Systems Requirements, and numerous E0P's such as 35-3.00, " Engineering Tests;"
40-7.00, " Design Reviews;" 40-9.00, "ASME Code Design Verification;" 42-5.00,
" Independent Design Verification;" 42-10.00, " Design Record Files;" and 60-3.10, " Engineering Records Retention."
E0P 42-10.00 describes DRFs as formal, organized accumulations of information, which provide a controlled system for retention of documented engineering activities, necessary to substantiate significant design decisions. The DRF provides a mechanism for controlling and archiving important design records, such as design verification, studies and analyses.
It does not include documents, such as drawings and specifications, which are maintained under separate corporate design controls. The procedure also states that the DRF should provide for design notes, calculations, records and other supporting information, and cross-reference to related or supporting DRFs.
The team interviewed GE-NE configuration management staff in regards to practices for generating DRFs. DRFs are created by cognizant design engineers to include the necessary design documents. When the associated design activities are completed, the DRF is reviewed and forwarded for permanent retention on microfilm. Periodic reports are distributed to cognizant managers in the event DRFs are not being completed and microfilmed in-a timely fashion.
The reproduction area was examined and the process for handling incoming DRFs was reviewed.
The GE-NE records management personnel review the DRF for legibility prior to sending the DRF to the microfilming contractor.
The hard I
copy records are shipped to a vendor to be microfilmed and three microfilms are returned to GE-NE. One copy of the film is k.ept in the GE-NE library vault area, one is sent to the permanent repository, and one film along with the hard copy is given to the cognizant engineer. The team examined the library DRF files.
These were maintained in locked storage containers that are only accessible to personnel authorized by the appropriate DRF custodian.
Based on the DRF reviews, the team questioned GE-NE with respect to an SSAR
{
statement that GE-NE is responsible for common engineering documents that are 1
used for certification including the supporting calculations. These supporting calculations are not always included in the DRF. GE-NE currently has the following statement in SSAR section 17.1.1: "The lead responsibility to produce each specification and drawing is formally assigned to one design organization. However, the contents of each document is reviewed and approved by GE-NE. While common. engineering documents reflect the formal consensus of 1
all parties, GE-NE is responsible for the design and supporting calculations and records for the ABWR project."
The GE-NE DRF files do have the Japanese (K6/K7) plant design specification, process flow diagram (PFD), piping and instrument diagram (P&ID), and instrument block diagram for each system. These received a formal GE-NE review via Engineering Review Memoranda (ERMs) and the resolution of comments is well documented. However, there is a scarcity of information on supporting calculations, particularly for those systems where the international technical associates had the design lead.
The team reviewed the system calculations for the reactor building cooling water (RCW), a system where the technical associates had the design lead and identified that GE-NE has not documented a review of the supporting calculations and their QA audit process did not examine the technical adequacy of the supporting calculations.
GE-NE management indicated that GE-NE engineering reviews conducted of common engineering documents, participation in design meetings with the international associates, review of other design documents, and performance of QA audits of the associates fulfilled their responsibilities with respect to supporting calculations. As a result, Unresolved item (93-02-08) was identified during this part of the inspection.
3.5 Desian Control 3.5.1 Design Action List (DAL) i The team reviewed E0P 55-2.00 which identified that the Change Control Board (CCB) would maintain the DAL. The Chairman and secretary of the CCB explained the DAL process and provided a supplementary GE-NE administrative guideline that is used for the DAL process. GE-NE stated that the DAL items are those design issues which represent potential design changes for the US ABWR, such as changes required to meet the US regulatory requirements and US codes and standards. The DAL tracks the differences between the K6/K7 design and the certified US AEWR design. When a difference is identified between the designs, it is listed on the DAL and a decision is made whether to proceed with an Engineering Change Approval (ECA) for generic changes or an Engineering Change Notice (ECN) for singular changes. The DAL serves as a placeholder to track changes that need to be made at a later date affecting lower tier engineering documents. The First-of-a-Kind-Engineering (F0AKE) effort will translate the DAL items into the implementing design documents.
The team was informed that the majority of the DAL resulted from licensing review comments made by the NRC that necessitated changes from the K6/K7 design.
The team found no formal GE-NE procedure which ensured that each responsible engineer reviewed the international ABWR design to verify that it complied with the current set of applicable US requirements for each system.
GE-NE stated that the international ABWR was, in their opinion, licensable in the US. Therefore, the changes to the design that are captured in the DAL are those that were agreed to by GE-NE to resolve the NRC staff comments on the SSAR. According to GE-NE, the latest issue of the SSAR incorporates all DAL items issued to date. However, the actual implementation of the DAL will be addressed during the F0AKE activities.
_7_
8 3.5.2 Design Change Control f
-3.5.2.'1 SSAR Material l
The team.was informed that an international technical associate had the design t
lead.for the reactor building cooling water (RCW) system. The team identified that the system piping and instrument diagram (P&ID), Figure 9.2-1, sheet 1 of 9, represented the ABWR configuration with 3 heat exchangers while the associated process flow diagram (PFD), Figure 9.2-la, showed only 2 heat i
exchangers that is representative of the K6/K7 design. The system flow and t
pressure drop information on the PFD had not been re-calculated for the ABWR configuration as the analysis had been performed by the international associate. The team identified this inconsistency to GE-NE management. They stated that-they would either remove the PFD from the SSAR or revise the PFD information to be consistent with the US ABWR design.
The team pointed out that the main steam flow rate listed in Table 6.3-1 of the SSAR was inconsistent with the value used in the SAFER 03 input. GE-NE-agreed to correct this error in Amendment 32 to the SSAR. The team checked samples of other design input data against the SSAR and found them to be consistent.
The existence of inconsistent design information in the controlled SSAR is identified as Unresolved Item (93-01-09) as GE-NE was in the process of certifying the SSAR material to be submitted in Amendment 32 and the SSAR is a formal design document that is utilized by the staff to reach a safety i
judgement on the ABWR.
GE-NE does not have a procedure for controlling changes to the certified ABWR design. Currently GE-NE is exploring several ways of controlling the ABWR design after certification, and will adopt the approach recommended by the nuclear industry that is acceptable to the NRC.
i 3.6 Review of Safety Analyses and Desian Calculations The team selected for review examples of DRFs related to system design and analyses in support of SSAR Chapters 6 and 15. The analyses files selected were All-0009, A00-03024, A21-00001, and A21-00001-1. The review consisted of
(
verifying that input data and assumptions were properly documented and that independent review was performed.
The. requirements in E0P 42-10.00, " Design Record Files," were general and i
broad-based, and the DRFs met _the intent of this procedure.' For most of'the inputs the source references were listed. However, the team found that input
~!
information that was based on assumptions, engineering judgement, or previous j
GE-NE experience, was not identified as such. The documentation included in.
analyses files was lacking in' clear definition of the purpose, methodology and.
assumptions such that an independent reviewer who had not perfomed the analyses.would find it very difficult to review these files. The team could not confirm that the independent verification of the SAFER 03 analysis included checking of the data entry of inputs used in the computer. runs because the j
i l t
f
1 i
printout of the input data and evidence of verification were not found in the DRF.
An example of the poor documentation of input assumptions to the analyses was found in DRF All-00009. The flow area of the internal recirculation pump was based on a hand drawn sketch with a reference to an individual who provided the information, instead of a reference to the applicable design drawing.
GE-NE produced a vendor drawing from 1981 as the source of the flow area information. This drawing did not include any dimensions and indicated no scaling information. However, GE-NE stated the flow area was scaled from this drawing based on the referenced individual's recall from memory of the pump shaft diameter and based on the assumption that the drawing was to scale.
L This method of independent verification is not consistent with the GE-NE topical report referenced in the ABWR SSAR.
As a result, Nonconformance (93-02-02) was identified during this part of the inspection.
The above analyses were performed during the preliminary stages of the ABWR design, and GE-NE has not assessed the impact of the current ABWR design parameters on the conclusions reached in these analyses.
3.6.1 Residual Heat Removal System The team reviewed portions of the DRF associated with the residual heat removal (RHR) system, Ell-00052, which is still open.
Volume 3 of the DRF included Engineering Review Memorandum (ERM) DMH5432AY on the SSAR verification effort for the RHR system. The GE-NE verification appeared to be a comprehensive review of SSAR material (text, figures, and tables) for accuracy with respect to PAIDs, instrument block diagrams (IBDs), PFDs and selected Design Action List items (DALs).
Over 100 GE-NE verification review comments were generated.
The verification was completed on 6/25/93.
The team also reviewed DRF Ell-00032-1, Volumes I and 5 that included pertinent information on the RHR system. The DRF had been microfilmed on March 31, 1988. Design verification check sheets were included for GE-NE and the international associates for the common engineering documents. The design verifications provided for a comprehensive review by all three design organizations involved with the K6/K7 design and documented the reconciliations of the comments.
The DRF for the RHR system did not contain the original calculations in support of the design but provided references, scoping calculations by the responsible engineer to verify acceptability of the design parameters, agreements on design parameters reached between the parties responsible for the international design, comparisons with other BWRs, and engineering judgement. The system design specification along with the responsible engineer's justification of the design parameters was independently verified.
3.6.2 Reactor Building Cooling Water system The team reviewed the reactor building cooling water (RCW) system DRF, P21-00001. The RCW system DRF clearly documented the multi-party engineering reviews of common engineering documents and associated dispositions. However, ;
i i
within the DRF there were several pages (sheets 554-560) of unchecked /
unverified calculations that evaluated the ABWR system differences from the K6/K7 design, including additional heat loads and the addition of a third heat exchanger.
These calculations, therefore, would support the US ABWR certification.
The evaluation was very informally done and was not sufficiently detailed as required by ANSI N45.2.11-1974, " Quality Assurance Requirements for the Design of Nuclear Power Plants," with respect to:
purpose, method, assumptions, design input, and references so that a technically qualified person could review and understand the analysis without recourse to the originator.
In addition, E0P 42-10.00 states that when a DRF is closed, the completed record shall be reviewed to ensure design verification requirements, where applicable, have been met.
DRF P21-00001 was identified as closed on a DRF status run dated September 9, 1993. No formal process appeared to be in-place to ensure that while the DRF had been microfilmed, an outstanding activity had to be accomplished with respect to design verification of the calculations.
The failure to ensure that the RCW calculations were design verified prior to closeout of the associated DRF appears to be inconsistent with E0P 42-10.00 requirements.
i The team questioned the lead system engineer as to how the RCW surge tank capacity had been sized as the ABWR ITAAC includes a verification of 16 cubic meter volume for the tank. The RCW design specification includes a statement that the tank is sized so that it can function for 30 days without makeup following a seismic disturbance that could cause a failure in some portion of the system piping. No supporting calculations existed in the GE-NE DRF that t
provided the details for the tank sizing, and the team was informed that such calculations had not been reviewed during GE-NE interaction with the international technical associates. At the end of the inspection, the team was informed that forthcoming SSAR amendment 32 will contain additional information about the surge tank capacity.
The RCW calculations that extrapolated the K6/K7 design to the certified ABWR design were performed in a manner not consistent with the GE-NE QA topical report (NEDO-11209-04A) commitment to ANSI N45.2.11-1974 and with GE-NE E0P 42-1.00 requirements. As a result, Unresolved Item (93-02-10) was identified during this part of the inspection.
3.6.3 Containment Pressure and Temperature Calculations The ABWR containment pressure and temperature calculations in DRF T11-0008 were approved and issued during the inspection. This file generally complied with the format and content requirements for calculations specified in procedure E0P 42-1.00, " Design Process," which was issued in December 1992.
The short-term and long-term accident response analyses for the U.S. ABWR containment were performed using the international ABWR data except for decay l
heat data, wetwell temperature, and ultimate heat sink temperature.
In response to the team's query regarding the inconsistency between SSAR Table 6.2-1 and the DRF, GE-NE stated that the SSAR table would be revised and submitted along with Amendment 32. !
t
3.7 ABWR Full Intearal Simulation Tests (FIST) Test Control 3.7.1 ABWR FIST Background The teams reviewed documentation for the ABWR Full Integral Simulation Tests (FIST), performed October 18, 1983, through January 18, 1984, to verify implementation of GE-NE's QA program. The tests were performed at the GE-NE FIST facility site in San Jose, California, which was used for other FIST tests in addition to the ABWR FIST tests.
Test Plan and Procedure (TPP) TP-515.1078, ABWR Full Integral Simulation Test (ABWR FIST), Revision A, dated October 17, 1983, provided information on test objectives, quality assurance requirements, procedures to be followed when performing the test, and the instrument list.
In addition, the TPP referenced other documents which were applicable to the tests such as facility drawings and information related to the FIST tests performed prior to and following the ABWR FIST tests.
1 The TPP stated that the objective of the tests was to obtain and evaluate basic thermal-hydraulic data from the test system configuration which had calculated performance characteristics similar to an ABWR with 8x8 fuel bundles during hypothetical loss-of-inventory and limited operational transients.
The FIST facility was developed to closely simulate the ABWR with one full-size fuel bundle of electrically-heated rods producing full bundle heat output. Other components included an external recirculation pump to provide specified core flow, a scaled steam separator, a heated feedwater supply system, and three emergency core cooling systems. The facility was run at realistic pressures, temperatures, bundle power, and coolant flow rates.
Approximately 500 instruments were connected to the system and the information was supplied to a highspeed data acquisition system whici, monitored pressures, temperatures, water levels, and flows throughout the facility.
3.7.2 FIST Test Control Section 2.0 of the TPP provided the QA requirements for the ABWR FIST tests.
Paragraph 2.1.4 of the TPP, "Q/A Forms," listed the following set of forms to be completed and filed in DRF E00-149 for each test:
(1) FIST Facility Configuration Confirmation, (2) Quality Surveillance Check Sheet, (3) FIST Pre-test Check List, (4) FIST Run Log, (5) FIST Test Procedure, and (6) Test Instrument List. These forms were to provide the quality assurance basis for each test performed. The team reviewed the applicable portions of DRF E00-149 and was unable to locate these forms for the tests that had been performed and GE-NE personnel were unable to provide these forms from a source other than the DRF.
Section 4.3.i of E0P 35.300, " Engineering Test," dated February 4, 1982, stated that the responsible test engineer should assure that test logs were established and maintained. The team was unable to locate the test log in ORF E00-149 and GE-NE personnel were unable to produce the document. The team concluded, based on the review of DRF E00-149, that GE-NE had not filed the required QA documents for each test or the test log in the DRF and subsequently could not produce the documents.
GE-NE had not maintained l
l sufficient records, identifiable and retrievable, to furnish evidence of i
activities affecting quality such as the results of reviews, inspections, tests as required by Criterion XVII, " Quality Assurance Records," of Appendix B to 10 CFR Part 50. As a result, Nonconformance (93-02-03).was identified q
during this part of the inspection.
The following records and documentation, which are required to'be contained or referenced in the FIST DRF, were missing: (1) as-built facility drawings, I
(2) reference to original data tapes from the tests, (3) records of disposition of all test anomalies, (4) test log and a complete set QA forms 1
for each test, and (5) documentation of analytical' or experimental verification of engineering calculations.. In addition,- the Joint Study Final Test Report, NEDC-30622, contained obsolete design drawings. Considerable effort was needed to determine what drawings contained the as-built elevations of key components of the test facility. Additionally, there were no final-approved revisions for the drawings reviewed.
t The team reviewed the six summary reports for the ABWR FIST tests:
(1) NEDC-22185, " Internal Pump Plant Blowdown Test," (4/1/82-9/30/82), dated September 1982; (2) NEDC-30031, (10/1/82-3/31/83) dated March 1983; (3 & 4) NEDC-30214 (two reports) (4/1/83-9/30/83), dated September 1983; and (5 & 6) NEDC-30516 (two reports) (10/1/83-3/31/84), dated March 31, 1984; and the final report, o
NEDC-30622, " Internal Pum Plant Blowdown Test - Final Report," dated June 1984.
Figure G-2d, sheet 4, of the Final Report had one signature block filled (" drawn"), three left blank (" checked," "DRTG," and "ENGRG"), and the issue date block left blank. Figure G-1 of the Final Report, " FIST Piping and Instrumentation Drawing," had one signature block ("ENGRG") left blank, the revision block left blank, and the issue date block left blank although NEDC 30031, dated March 31,-1983, an earlier document, contained the drawing
.j with all signature blocks signed, the revision block filled, and an issue date of March 25, 1983.
j I
The team also examined the ABWR FIST testing and SAFER code qualification i
based on ABWR FIST.
GE-NE stated that the FIST test was a " licensing" test and had appropriate QA measures in place to assure the integrity and accuracy of the data acquired and that only programs that related directly to reactors require a safety-related classification. The DRF indicated that the testing was conducted as a non-safety-related activity.
The NRC asserts that the FIST test program comprised a safety-related
\\
activity, and that the program's purpose was, in part, to obtain design basis data for the ABWR.
Therefore, the use of the data falls under type B2 of E0P 35-3.00, and also constitutes a safety-related activity. Appropriate QA procedures should therefore have been in place, commensurate with the requirements for safety-related tests, per the requirements of Appendix B to j
10 CFR'Part 50 and GE-NE's QA program description.
i 3.7.3 Instruments, Calibration, and Procurement Sections 1.1 and 4.2.b of E0P 35-3.20, " Calibration Control," dated January 2, 1981, stated that maintenance and test equipment calibrations were to be l
performed using controls which assured traceability to certified equipment j
i 1 lb
_ _ ~ ~. _,
having known valid relationships to nationally recognized standards.
Section 2.2 of the TPP, " Instrumentation," discussed the instrumentation used to gather data during performance of the tests and indicated that the differential pressure transducers and other instruments were calibrated by 1
GE-NE personnel, that manufacturer's calibrations were to be used for i
thermocouples, and that the calibration histories were to be filed in DRF EC9-149.
The thermocouples used for the ABWR FIST tests were purchased by GE-NE from Claude S. Gordon Co. as commercial grade items without further verification of the adequacy of the calibration or performance characteristics. GE-NE had not i
audited or performed surveys of Claude S. Gordon Co. and had not placed them on the GE-NE approved suppliers list. The purchase orders did not specify that any quality assurance program was to be in place, or that the criteria of 10 CFR Part 50 Appendix B or 10 CFR Part 21 applied. Therefore, GE-NE did not have assurance that the performance of the thermocouples was as specified by the manufacturer or traceable to certified equipment having known valid relationships to nationally recognized standards.
GE-NE had not verified that the thermocouples, instruments used in an activity affecting quality, were properly controlled, calibrated, and adjusted to maintain accuracy with necessary limits as required by Criterion X11. " Control of Measuring and Test Equipment," of Appendix B to 10 CFR Part 50. As a result, Nonconformance (93-02-04) was identified during this part of the inspection.
3.7.4 Documentation of Test Anomalies and Deficiencies The teams reviewed DRF E00-149 and the summary and final test reports for disposition of anomalies. The DRF contained a " Questionable - Channel List" which listed changes made to instrumentation such as changed orifice constants, the addition of instruments, and failure of instruments. The items typically listed the channel identification and a brief description of the reason for the entry.
The form also provided an area for a " decision or action" in which disposition of the item could be documented. A number of items entered, which described failed thermocouples, did not have any disposition entered.
GE-NE personnel indicated that the instrumentation system had been designed to be redundant to account for failures of tnermocouples and that a disposition was not required.
l The TPP did not provide for a method to document test anomalies or deficiencies other than the Questionable Channel List which was specific to i
instrumentation channels. The team noted that one test was required to be re-performed due to an inadequately sized blowdown orifice.
The repeat of this test was documented on an Engineering Work Authorization sheet and included in DRF E00-149.
3.8 ABWR Computer Code Modelina The team reviewed the code qualification and computer modeling for the GE-NE ABWR thermal-hydraulics calculations included in Chapters 6 and 15 of the SSAR. The three computer codes and associated modeling examined were REDYA, ODYNA and SAFER.
REDYA is a point kinetics transient analysis code that is only used for slow transients. ODYNA is a one dimensional (10) kinetics i
e
t transient analysis code that is used for fast transients includ' ing all pressurization transients. SAFER is GE-NE's LOCA evaluation computer code. A review of the DRF for each computer code revealed that no official
- documentation of the implementation, testing, and independent verification of computer code changes exists in the DRF.
In some cases, the developer keeps
-i his own personal records of this implementation and testing.
In one case the code developer stated that code changes were not even documented through internal code comment statements. As a result, Nonconformance (93-02-01) was identified during'this part of the inspection.
GE-NE's method of independent verification is called a Design Review. During i
the design review, the results of code qualification calculations are I
presented to the design review team. This "high level" independent verification seems to assume that low level verification of the implementation of code changes and modeling has' already been done.
Since this is not required, the high level review can allow errors to slip through as previously identified in the August 1993 inspection of the SWBR and TRACG for the GIST test.
In the case of REDYA and ODYNA, GE-NE qualified the codes for modeling internal pump plants without comparison to experimental data.. An unreferenced and later report compared the codes to internal pump data from two European internal pump plants. GE-NE also has not been able to obtain internal pump experimental data from the technical associates. The technical associates will orly supply GE-NE with the information needed for code inputs and not the data they were obtained'from.
The ABWR calculation notebooks reviewed were found to be sloppily kept and not I
self-contained enough to review without the analyst present. The calculation notebooks and analysis DRFs also do not seem to meet the GE-NE topical. report which requires stand alone documentation with all assumptions clearly stated and that a technically qualified person is able to review it without any outside help. As a result, Nonconformance (93-02-05) was identified during j
this part of the inspection.
1 3.9 Ouality Assurance Audits l
3.9.1 Nuclear Quality Assurance (NQA) i The team reviewed the NQA audit plans that had been prepared for 1990 through 1993. Two audits, Q9008 and Q9306, involved ABWR activities.
In addition, audit Q9107 had been planned then was cancelled as it was determined by GE-NE to be a redundant verification with a Quality System Review effort-The team reviewed the associated ABWR audit documentation including: audit checklists, auditor qualifications, audit plans, audit findings, completed audit checklist, summary audit report, NQA audit report, and associated Corrective Action Requests (CARS). Audit Q9008 was performed between September 24, 1990, and October 5, 1990, by two audit personnel.- The audit reviewed the ABWR Reactor Pressure Vessel (RPV) design specification. The audit uncovered the fact that some DRFs had not been microfilmed -in a timely manner. -Appropriate corrective action and preventive action was specified for
.(
the Corrective Action Request. Audit Q9306 had been partially completed. The
' j
?
.)
audit checklist included attributes such as DRF technical completeness which had not yet been audited.
3.9.2 Services Quality Assurance (SQA)
The team reviewed the SQA audit planning for 1993. No audits were planned in the ABWR area as the manager of Advanced Reactor Programs (ARP) QA had not requested any audit assistance from SQA. However, SQA (formerly PQA) has audited the ABWR project in the past.
The team reviewed a sample of those audits and associated documentation as discussed below:
PQA Audit 91-2: This audit principally covered Japanese work on the K6/7 project, thare were no findings of import for the ABWR.
PQA Audit 92-1: Two auditors participated on this audit. Aspects of the interface between GE-NE and Japanese partners was audited.
PQA ABQ 91-1: This audit related to the control and issuing of the ABWR Standard Safety Analysis Report (SSAR). The audit identified a concern that some diagrams had been submitted in the SSAR that had not been verified.
3.9.3 Advanced Reactor Programs (ARP) QA The 1993 audit plan for the Advance Reactor Program (ARP) QA group for 1993 was reviewed and it was found that no audits were planned for the ABWR.
However, the team was informed that a scheduled audit on the First-of-a-Kind-Engineering (F0AKE) would actually cover the ABWR follow-on engineering work.
That audit is planned for the fourth quarter of the year.
ARP QA had performed one ABWR audit in 1992, Q9203. The team reviewed ARP audit Q9203 that was conducted on both the ABWR and SBWR.
The audit report was issued on January 25, 1993. The following aspects were documented as having been examined for the ABWR: review and status of design verification items for the SSAR, review and status of Design Action List (DAL) items and associated engineering change documentation, review and status of SSAR preparation, and follow-up to a previously identified SQA audit issue. The audit checklist and audit report did not contain sufficient information to document that the audit scope had covered the breadth of QA aspects claimed to have been performed (Appendix B Criterion 1, 2, 3, 4, 5, 6, 7, 16, 17, and 18). During discussions with the lead auditor, the team was informed that the report had been structured in a brief format to conform to management expectations. The team expressed a concern that audit reports should contain sufficient information regarding the scope of audit activities to allow the evaluation of the audit at a later date.
The team was informed that ARP QA surveillances are also performed by quality personnel to supplement the formal audit process.
The team reviewed ARP QA surveillances for reviews on several DRFs. The QA staff had compiled a checklist of several attributes to check with respect to DRF administrative content. These surveillances had been performed for the Control Rod Drive (CRD) restraint, reactor pressure vessel (RPV) and fuel transfer DRFs. The QA i
review found some unsatisfactory aspects that the team was informed has been rectified. The performance of supplemental surveillances is a good practice to augment the more formal and infrequent ARP QA audits.
1 The team questioned QA management about the technical composition of the audit teams.
GE-NE management stated that over the last 5 years, none of the ARP QA audit teams has been supplemented with technical personnel to perform a deep review of the integrity of the design process.
3.9.4 Audits Performed By External Groups In June of 1992, a Nuclear Procurement Issues Committee (NUPIC) audit was performed by auditors from Florida Power and Light, Entergy, Wolf Creek, Nebraska Public Power District (NPPD), and Illinois Power. Audit finding SA92-05-02 was generated because GE-NE engineering calculations in two DRFs (involving NPPD work) were found to not always conform to the requirements of ANSI N45.2.ll that had been contractually invoked.
GE-NE implemented corrective action that included generating a new E0P, 42-1.00, " Design Process," that implements the ANSI N45.2.ll requirements and GE-NE rectified the two DRFs as needed. GE-NE had not evaluated the adequacy of the pre-existing ABWR calculations with respect to the new E0P requirements. GE-NE management stated that prior ABWR work was suitably controlled.
See Section 3.4 of this report regarding reviews of DRFs and associated calculational files.
3.9.5 GE-NE Audits of Hitachi Limited and Toshiba Corporation The NRC inspectors reviewed several GE-NE audits performed of Hitachi Limited and Toshiba Corporation (Japanese technical associates) which are approved for engineering services related to the design of safety-related systems and components for the Japanese ABWR plant K6/K7. GE-NE is required to perform an annual review of Hitachi and Toshiba's QA program implementation in accordance with GE-NE's Joint Venture Agreement (JVA) and GE-NE's QA Plan. The audits reviewed during the inspection included the 1988, 1989, 1990, and 1992 audits performed by GE-NE at the Hitachi Works, located in Hitachi City, Japan, and Toshiba's Isogo Nuclear Energy Center, located in Yokohama, Japan. The audits were performed by one member from GE-NE QA, located in San Jose, California, and one member from the General Electric Technical Service Company (GETSCO) office, located in Japan.
Each audit was usually one to two days in duration.
The audits of the Japanese technical associates were performed to review the implementation of the Joint Venture QA Basic Plan and focused on various aspects of their overall QA program which included document and design
+
control, engineering computer codes, design reviews, quality training, QA program changes, document maintenance, and internal QA audits. The QA programs are based on " Quality Assurance Guidelines for Nuclear Power Plants" (JEAG 4101), which was established on the basis of Appendix B to 10 CFR Part 50.
The first edition of this guideline was issued in 1972 and was later modified on the basis of International Atomic Energy Agency (IAEA) Code of Practice 50-C-QA and issued as a second edition (JEAG 4101-1981). Reflecting IAEA Safety Guides Series 50-SG-QA, the guidelines were modified once again and issued as a third edition (JEAG 4101-1985).
In addition, a series of l
detailed guidelines (JEAG 4102-4109) have been issued as supplementary material for JEAG 4101-1985. These were introduced during the period 1985 to 1988, however they thereafter came to be reviewed on the basis of IAEA safety standard 50-C-QA (Revision 1) and were eventually issued in a comprehensive version (JEAG 4101-1990) which remains effective today.
The NRC inspector's review of these audits identified weaknesses in that the audit file documentation did not reflect the necessary Jetail to support an effective implementation audit of the QA program arcas reviewed. The eight audits reviewed by the NRC inspectors did not identify any findings or weaknesses which required corrective action, however two audits of technical associates identified that the quality system was effective in assuring the quality of common engineering and the JVA activities, with several exceptions.
One exception, documented in a 1988 audit report of a technical associate, identified that the associate's method of verification, attesting to the completion of individual design reviews performed for design verification, was not always documented per Section 5.3 of JEAG 4101-1985.
This exception appeared to have the potential of a nonconformance.
It was also noted that GE-NE failed to perform an annual audit of the technical associates' QA program as required by the JVA commitments and GE-NE procedures.
As a result, Nonconformance (93-02-06) was identified during this part of the inspection.
3.9.6 GE-NE Audits of Bechtel North American Power Corporation The NRC inspection team reviewed several purchase orders (P0s) between GE-NE and Bechtel North American Power Corporation (BECHTEL), San Francisco, California, for engineering services associated with the ABWR contract.
BECHTEL, acting as a subcontractor to the U.S. Department of Energy (DOE), is providing engineering services to GE-NE under DOE Contract DE-AC03-86 SF16563,
" Technology Programs in Support of Advanced Light Water Reactor Plants," dated August 27, 1986. The contract includes work scopes for both the Simplified Boiling Water Reactor (SBWR) and the ABWR program development and requires all contractors to establish, implement, and maintain a QA Program Plan which meets the requirements of Appendix B to 10 CFR Part 50 (Appendix B), and ANSI N45.2-1977. Work scopes for the ABWR include structural reanalysis and design verification of the reactor building and other facilities (Task 110.1), and dynamic analysis (Task 130.1) used to support Chapter 3, " Design of t
Structures, Components, Equipment and Systems," of the ABWR SSAR, which is associated with ABWR licensing certification.
The NRC inspection team reviewed GE-NE P0 No. 190-ALWR-31387, issued to BECHTEL on April 22, 1987, which included Tasks 110.1 and 130.1.
The P0.-
invoked no QA requirements with respect to the manner in which the work was to be processed and referenced -the " General Provisions" section of the DOE contract.
The work scope section of the GE-NE P0 stated, in part, that "Bechtel would furnish engineering services in support of GE-NE's contract with DOE to provide a licensing submittal to the NRC in support of Chapter 3 of the ABWR SSAR. This task is for safety-related systems (emphasis added)."
An audit of BECHTEL was performed by GE-NE in August 1991 (QE 9104) which verified satisfactory implementation of Bechtel's QA program used to support i
the SBWR.
However, no implementation audit of BECHTEL was ever performed in support of the ABWR.
In addition, BECHTEL appears on GE-NE's Approved f
i
Suppliers List (ASL) only for the Advanced Liquid Metal Reactor project, and i
not for the ABWR.
Since 1987, over 21 amendments to the PO have been processed by GE-NE for additional work scope associated with the ABWR certification program without the benefit of an audit of the Bechtel QA Program Plan (latest version is Revision 4, dated November 5,1992). As a result, Nonconformance (93-02-07) was identified during this part of the inspection.
During the NRC inspection, BECHTEL provided a letter to GE-NE (BLG-0100, dated September 10, 1993) which confirmed that although the GE-NE P0 did not reference or invoke quality program requirements, all work supporting safety-related activities under the P0 was processed in accordance with Bechtel's Nuclear QA Program Plan and Nuclear QA Manual, which complies with Appendix B.
3.9.7 Observations of QA Audit Activities Based upon a review of audit activities performed by several GE-NE QA organizations, the team had the following observations:
Audit reports had a variety of ways of categorizing the resultant findings, such as CARS, concerns, observations, recommendations, and unresolved items. The procedural controls only identify that CARS and recommendations result from audits. The team was informed that a draft procedure was under preparation for unresolved items. The other findings, while a reasonable explanation was given by GE-NE regarding their use, had not been explicitly described to ensure common use among the auditors and recipient organizations regarding corrective and preventive action requirements.
A weakness in the GE-NE audit approach was that audit teams were not supplemented with technical engineering experts to perform more intensive design reviews to supplement the QA programmatic audits.
The ARP QA follow-up of previously identified areas of concern is a good practice to ensure that corrective actions have been effectively implemented.
Another identified weaknesses in the GE-NE audit approach is that the audit file documentation for the technical associates did not reflect the necessary detail to support an effective implementation audit of the QA program areas reviewed.
4 PERSONNEL CONTACTED t
GE Nuclear Enerov:
Bob Berglund, General Manager, Advanced Reactor Programs (ARP) i J.F. Quirk, Program Manager, ABWR Certification P.E. Novak, Quality Assurance Manager, ARP 1
Joe Case, Manager, Nuclear Quality Assurance (NQA)
Ken Brayman, Manager, QA Systems, NQA
a Forrest Hatch, Manager, Services & Projects Quality
^
Craig Sawyer, ARP Chandler Eason, NQA Nil Patel, ABWR Certification Jcy Murray, QA Audits Manager, NQA Frink Paradiso, ABWR Engineer Bob Mitchell, Safety Evaluation Programs Paul Billig, SBWR Test Programs N.E. Barclay, Audit Programs Manager Elias Delmurd, Auditor / Engineer C.V. Nguyen, QA Engineer / Auditor R.W. Schrum, Core and Safety Methods Gary Dix, Manager, EQA & Automation H.T. Kim, ABWR Bruce Matzner, Core and Fuel Advanced Design Nuclear Reculatcry Commission:
Richard P. McIntyre, Team Leader, Vendor Inspection Branch (VIB)
Leif J. Norrholm, Chief,. VIB, George Thomas, Nuclear Engineer, Reactor Systems Branch Robert Pettis, Senior Reactor Engineer, VIB Billy Rogers, Reactor Engineer, VIB Robert Gramm, Section Chief, Performance and Quality Evaluation Branch S.K. Malur, Senior Operations Engineer, Special Inspection Branch Joseph Staudenmeier, Reactor Engineer, Analytical Support Group H.S. Cheng, Physicist, Brookhaven National Laboratory b
i i
1 6
i,
-