ML19270F600
| ML19270F600 | |
| Person / Time | |
|---|---|
| Issue date: | 11/16/1978 |
| From: | Donna Anderson, Hale C, Sheron B NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION IV) |
| To: | |
| Shared Package | |
| ML19270F594 | List: |
| References | |
| REF-QA-99900081 NUDOCS 7902150490 | |
| Download: ML19270F600 (7) | |
Text
VENDOR INSPECTION REPORT U. S. IlVCLEAR REGULATORY COMMISSION OFFICE OF INSPECTION AND ErlFORCEMENT REGION IV Report flo. 99900081/78-03 Program flo. 44010 Company:
Exxon Nuclear Company 2101 Horn Rapids Road Richland, Washington 99352 Inspection Conducted:
October 30 - flovember 3, 1978 Inspectors: 77.M73/ac (b n/a/77 f,tD. G. Anderson CPrincipal Inspector, Vendor Date Inspection Branch Other Personnel: %/6B
[cu
__p/ fahr pB.W.Sheron,NgearEngineer, DSS /NRR Date
~77M75/ubb'd-w n/iubr pR. F. Audene, Nuclear Engineer, DSS /NRR
'Date Yl kr Y
y /a,/77 J. E. fiorian, Sectf6n Leader, DSS /NRR D&te f
Approved by:%/'((
u hA[7T
/ g. J. Hale, Chief (/ Vendor Program Evaluation Da'te Section, Vendor Inspection Branch Summary Special Inspection conducted on October 30 - November 3, 1976 (93900081/78-03)
Areas Inspected:
10 CFR 50, Appendix B, and Topical Report XN-NF-1 as applied to the establishment and implementation of procedures to control 7902150490 the development and revision of safety analysis computer codes.
The inspection involved thirty (30) inspector hours on site by one (1) USNRC Region IV inspector.
Results:
In the area inspected there were no unresolved items and one (1) deviation with two (2) examples identified as follows:
Deviation:
10 CFR 50, Appendix B, Criterion V, and Section 3.0 of Topical Report Xfl-NF-1-procedures do not exist for the evaluation and reporting of significant changes in the evaluation model as referenced in Criterion II.b.
(See flotice of Deviation, Enclosure, Example 1.)
Calculation Forms were not completed and Calculation Indexes were not prepared as required by procedure XN-flF-500, 002 (See Notice of Deviation, Enclosure, Example 2).
. DETAILS SECTION (Prepared by D. G. Anderson)
A.
Persons Contacted
- R. E. Collingham, Manager, System Model Development
- G. C. Cooke, Manager, Fuel Response Analysis
- M. S. Foster, Manager, EDP Programming
- K. P. Galbraith, Manager, Nuclear Safety Engineering
- S. E. Jensen, Manager, Systems Analysis
- D. C. Kolesar, Engineer
- G. F. Owsley, Manager, Reload Licensing
- D. K. Perry, Quality Assurance Engineer
- J. A. Perry, Manager, Quality Assurance
- Denotes those present at the exit meeting.
B.
Introduction This report covers a special inspection conducted to examine the establishment and implementation of procedure controlling safety analysis computer code development.
The objectives of this inspection are:
1.
To determine that _ adequate procedures to minimize the potential for analysis errors to go undetected have been established for control of the development and revision of these codes.
2.
To determine that these procedures were fully implemented during the development and revision of selected codes.
C.
Control of Safety Analysis Computer Codes 1.
Establishment of Procedures a.
Inspection (1) The Exxon Nuclear Company Topical Report, XN-NF-1, requires in Section 3.0, Design Control, that " Design of fuel assemblies and related components and the preparation of design documents are performed in accordance with approved procedure and techniques."
Section 3.5 states that "The adequacy of product designs may be verified in several ways, including in reactor experience of similar design performance
. -a.
\\
of design reviews, alternate calculations, or design testing." These commitments are implemented by the following Exxon Nuclear Company procedure:
XN-NF-500, 002, Fuel Design and Engineering Quality Assurance Procedures for Fuel Fabrication - Product Line and Related Components.
(2) An examination of XN-NF-500, 002, revealed that there are requirements for testing, documentation, verifica-tion, and control of computer programs.
The verifica-tion documentation is required to be maintained in a design file at the Control Files location.
Section 6.9.2 identifies those records to be included in the design file.
A Computer Code Council is responsible for the administrative and quality assurance aspects of Exxon Nuclear Computer programs and programming.
This procedure also details the requirements for development of ECCS Computer programs and the modifica-tion of existing computer programs are established along with the definition of responsibilities of individuals in the verification chain.
b.
Findings The procedure used to implement Exxon Topical Report, XN-NF-1 appears to meet the requirements of ANSI N45.2.11 as appropriate to verification, and documentation of this verification for Safety Analyses Computer Codes.
During the examination of this implementing procedure and the documentation generated as a result of following the procedure, the following items were identified and Exxon Management indicated that they would consider further action, as appropriate.
Any action taken by Exxon will be examined during a future inspection.
(1) There are no formal procedures for the documentation, timely assessment, and resolution of potential safety concerns, nor are there any procedures for reporting ECCS model changes per the requirements of 10 CFR 50, Appendix K.
(2) There are no procedures considered by Exxon to be applicable to code development.
Existing procedures are only considered applicable to code use.
(3) Verification information should be documented in the project files.
Exxon also stated that Management signatures on topical report constituted verification by design review.
We do not believe this to be an acceptable method of verification.
(4) When changes are made to the code, letters to code users informing them of the changes should be issued and included in the code file.
When code versions are approved, the approval letters should be kept in code file.
(5)
In some instances, signatures indicating that independ-ent checks were performed on final calculations were not found.
(6)
It was found in some cases that input data calculations were dated after the ECCS analyses were completed and the topical report approved by Exxon Management and submitted to NRC.
Independent checks were dated as much as 3 months after Management approval.
This item was identified by Exxon in an Internal Audit.
Correc-tive Action has been initiated.
(7) There is no procedure which requires input data, as it is put into the code, to be independently checked.
(8) Exxon should have procedures specific to how codes developed by Exxon contractors are to be checked out and verified by Exxon.
These procedures should be more specific than those that presently exist in the area of verification requirements.
(9)
Procedures do not exist which describe the detailed preparation, development, review, approval, and control of computer codes.
(10) Procedures should address the details of preparation, completion, checking, signing, identification, indexing, etc., for calculation forms which are used in code development.
(11)
Procedures do not address how errors are identified, corrected, or flagged in master codes.
(12) The categorization of codes into 3 classes; master, developmental, and special, does not appear to be sufficient to assure protection of NRC-approved codes.
. More specific categorization which provides better protection of HRC-approved codes against unauthorized modification is recommended.
(13) There is no evidence that code design and project files are indexed, nor is there any requirenent that they be indexed.
Without indexing, there is no record of_ what documents constitute a code design or project files.
(14)
Definitions in procedure XN-NF-500, 002 should be
~
more extensive, encompassing such items as verification, quali fication, etc...
(15)
In some instances, calculational files are only signed by the checker on the first page. This signature is intended to indicate that all pages of the calculation were checked.
Better confidence that all pages were checked would be obtained if (1). all pages were signed, and (2) the checker documented in a brief statement the extent and scope of the check.
2.
Implementation of Procedures a.
Inspection The development and revision of the following computer programs was evaluated for meeting the requirements of procedure XN-NF-500, 002 related to verification and control:
WREM:
The base for the Exxon Nuclear Company PWR generic water reactor evaluation model.
The PWR evaluation model includes RELAP-4 EM, RELAP-4 FLOOD and T00DEE 2.
RELAP-4 EM computes the space and time variation of the thermal-hydraulic conditions of the priccry and secondary systems during the decompression of the primary system following a LOCA.
RELAP-4 FLOOD is used to calculate the reflooding rates starting at the beginning of core recovery. T00DEE 2 calculates the hot fuel rod temperature analysis from bottom of core recovery through the reflood period until core quench.
PTSBWR 2:
A digital computer program, written in Fortran language, which simulates the behavior of non-jet pump BWR's under irregular operating conditions,
O
. such as steamline isolations, coolant recircula-tion e lfunctions and feedwater malfunctions.
Calculctes fluid conditions, flow rates, heat flux, reactor power and reactivity as a function of time.
The inspector examined the documentation contained in the design file on these two (2) computer programs which consisted of eight (8) memoranda transmitted between Exxon and Service Computer organizations, sixteen (16) internal memoranda related to computer code changes, seven (7) computer code Topical Reports contained in Exxon documents, twenty-nine (29) calcuation forms related to verification of computer programs, seven (7) meeting minutes of the Computer Code Council, one (1) internal audit of the Computer Code Council, twenty (20) Standard Sample Problem Verification Forms, fif teen (15) verification indexes, five (5) Master Computer Code Approval Forms, one (1)
Comput.er Code Masterization Procedure Checklist, two (2)
Master Code listings, two (2) telephone memoranda related to questions rrom Service Computer organizations about computer code use, and nine (9) computer code runs related to code verification.
b.
Findi_ngs In this area of the inspection, no unresolved items were identified.
One (1) deviation from commitment with two (2) examples was identified (See Notice of Deviation, Enclosure).
D.
Exit Meeting An exit meeting was conducted with management representatives at the conclusion of the inspection on November 3, 1978. Those individuals indicated by an asterisk in Section A of the Details Section of this report were in attendance.
The inspector discussed the special nature of this inspection and the scope and finding indentified during this inspection.
Management representatives of Exxon acknowledged the statments by the inspector with respect to the one (1) deviation presented.