ML20141F833

From kanterella
Jump to navigation Jump to search
Insp Rept 99900403/85-01 on 850304-06.No Violation, Nonconformance or Unresolved Items Noted.Major Areas Inspected:Svc Info Ltrs,Status of Previous Insp Findings, Potentially Reportable Condition Files & Reactive Items
ML20141F833
Person / Time
Issue date: 12/30/1985
From: Jocelyn Craig, Robert Pettis
NRC OFFICE OF INSPECTION & ENFORCEMENT (IE)
To:
Shared Package
ML20141F819 List:
References
REF-QA-99900403 NUDOCS 8601090560
Download: ML20141F833 (12)


Text

ORGANIZATION: GENERAL ELECTRIC COMPANY

  • NUCLEAR ENERGY BUSINESS OPERATIONS SAN JOSE, CALIFORNIA REPORT INSPECTION INSPECTION N0.: 99900403/85-01 DATE(S): 3/4-6/85 ON-SITE HOURS: 25 CORRESPONDENCE ADDRESS: General Electric Company Nuclear Energy Business Operations ATTN: Mr. W. H. Bruggeman, Vice President and General Manager 175 Curtner Avenue San Jose, California 45125 ORGANIZATIONAL CONTACT: Mr. J. J. Fox, Senior Program Manager TELEPHONE NUMBER: (408) 925-6195 PRINCIPAL PRODUCT: Nuclear Steam System, Services and Fuel.

NUCLEAR INDUSTRY ACTIVITY: General Electric Company (GE) Nuclear Energy Business Operations (NEB 0), has a work force of approximately 4500 assigned to domestic power plant activity.

ASSIGNED INSPECTOR: N A 8%[d b R. t. Pettis, Special Projects Inspedtion

/4ho/s/

Date Section (SPIS)

OTHER INSPECTOR (S): P. Sears, SPIS W. Shier, BNL W. Banister, EG&G APPROVED BY: M bdNM John W. Craig, Chi # , SU S, Vendor 6 Program Branch

/2/3./gg Date INSPECTION BASES AND SCOPE:

A. BASES: GE Topical Report No. NED0-11209-04A and 10 CFR 21.

B. SCOPE: The inspection was conducted to review and obtain copies of selected GE Service Information Letters (SILs); review the status of previous inspection findings and Potentially Reportable Condition (PRC) files; and review various reactive items.

PLANT SITE APPLICABILITY: Limerick 1 (50-352); Fermi (50-341); Hope Creek (50-354); LaSalle 1 and 2 (50-373/374); Monticello (50-263); Oyster Creek (50-219); Perry 1 and 2 (50-440/441); Shoreham (50-322); Vogtle 1 and 2 (50-424/425).

8601090560 860107 PDR GA999 EMVCENE 99900403 PDR

ORGANIZATION:

GENERAL ELECTRIC COMPANY NUCLEAR ENERGY BUSINESS OPERATIONS SAN JOSE, CALIFORNIA REPORT INSPECTION N0.: 99900403/85-01 RESULTS: PAGE 2 of 12 A. VIOLATIONS:

None.

B. NONCONFORMANCES:

None.

C. UNRESOLVED ITEMS:

None.

D. STATUS OF PREVIOUS INSPECTION FINDINGS:

1. (Closed) Nonconformance (84-02): Contrary to Engineering Operating Procedure (EOP) 42-10.00, Section 4.2.d.4, concerning Design Record.

Files (DRFs), the DRFs that supported the verification of computer calculations for SAFER 02 computer code (DRFs No. A00-01249, A00-1320, and E00-137) did not identify the reviewer and date when performed.

The DRFs supporting the SAFER 02 verification calculations have been independently reviewed. In addition, two actions have been taken by GE to prevent recurrence of this type of nonconformance: The Manager, Core and Fuel Technology has issued a letter to all engineers responsible for Engineering Computer Programs reiterating the DRF requirements for verification of calculations; in addition, a Quality Assurance Newsletter (dated August 1984) has been issued to all engineers and managers that includes a "DRF Closecut Checklist" with reminders about signing and dating DRF entries.

2. (Closed) Noncenformance (84-02): Contrary to E0P 42-1.00, Section 3.3.2, regarding design control, no documentation was available for the analyses described in GE Topical Report HEDE 23785-1-P, Vol. 11, and NEDE 24984. These topical reports were submitted to the Office of Nuclear Reactor Regulation for review.

The two topical reports referenced above describe the analytical basis for two safety-related computer codes (SAFER 02), NEDE 23785-1-P and (0DYN04) NEDE 24984 A review of the extensive verification programs for these computer codes has indicated that sufficient testing and comparison of code calculations with other analytical and experimental results was performed to preclude the need for additional DRFs.

ORGANIZATION:

~

GENERAL ELECTRIC COMPANY NUCLEAR ENERGY BUSINESS OPERATIONS SAN JOSE, CALIFORNIA REPORT INSPECTION N0.: 99900403/85-01 RESULTS: PAGE 3 of 12

3. (Closed) Nonconformance (84-04): Contrary to GE Quality Assurance Topical Report NED0-11209, Rev. 4, Section 3.12, " Design Change Control," Engineering Operating Procedures (E0P) 40-3.00 " Engineering Computer Programs" (ECPs), does not require that Control Components (responsible engineers for ECPs) define other design documents affected by computer code changes or errors, or coordinate these changes with other responsible engineers whose documents are affected. Further, Section 4.1 of the same procedure (E0P 40-3.00) does not require that the Control Component interface with responsible engineers affected by a computer code error, and assess effects of computer code errors on designs, past and present.

E0P 40-3.00 has been revised (Change Notice A, December 19, 1984) to require that responsible engineers for ECPs document all errors in approved Level 2, 2R and 3 computer programs. In addition, these errors will be classified according to their potential impact on previous analyses. This documentation is then reported to all User Design and Development Ccmponent Managers for evaluation. These managers also acknowledge receipt to the responsible engineer.

The inspector reviewed an example of the implementation of the procedure. This included the ECP error description and potential impact evaluation, the distribution to the component managers, and the acknowledgement of receipt returned to the ECP responsible engineer.

4. _( Closed) Nonconformance (84-04): Contrary to E0P 40-3.00,

" Engineering Computer Programs," the Design Record File (DRF) for the CRNC-04 computer code (No. A00-01619) did not include all of the code testing specified in the Software System Specification.

The Software System Specification for CRNC-04 has been revised to indicate that the code verification testing will include a comparison of results with resul.ts from previous versions of the code. This is considered sufficient since the current version of CRNC-04 does not include any significant analytical model modifications or editions.

5. (Closed) Nonconformance (84-04): Contrary to E0P 42-6.00,

" Independent Design Verification," the verification of calculations described in GE Topical Report NEDE 25518 was not completed until after issuance of the report.

I

ORGANIZATION: GENERAL ELECTRIC COMPANY NUCLEAR ENERGY BUSINESS OPERATIONS SAN JOSE, CALIFORNIA REPORT INSPECTION N0.: 99900403/85-01 RESULTS: PAGE 4 of 12 No action regarding the design record file for NEDE-25518 was required. However, as part of the action taken to prevent recurrence of this type of nonconformance, a memorandum was written emphasizing the requirement for completion of the independent review and verification prior to issuance of reports.

6. (0 pen) Nonconformance (84-04): Contrary to E0P 42-10.00, " Design Records Files," the DRF for the PANACEA Core Design System (No.

670-0005) did n>t always identify the originator, reviewer, or date performed.

GE will review DRF entries to assure that originators, reviewers and dates of entries are adequately identified. A record of this review will be incorporated into DRF 670-0005. In addition, the Manager, Core and Fuel Technology will issue a letter to all engineers responsible for ECPs reminding them of their responsi-bility under the referenced E0P. These requirements will also be emphasized in QA training course documents related to DRFs.

This item will be reviewed during a future inspection.

7. (0 pen) Nonconformance (84-04): Contrary to Section 3.10 of the QA topical report NED0-21109-04A, application of the SAP 4G07 code was not fully verified in the following areas:
a. Two options of the beam element (fixed end forces and shear deformation analysis) and one option of the pipe element (the ASME code analysis) had no verification provided,
b. One nodal point option (slaved degrees of freedom) and one option of the beam element (released degrees of freedom) had verification for the latest version only. However, an earlier version of the SAP 4G07 code (which is a Level 3 program), is still available for use on safety-related designs.

GE stated, in their May 22, 1985 written response to the NRC, that the SAP 4G07 code is a fully verified Level 2 computer program, meaning it satisfies GE's design review process requirements for independent verification with results comparable to those from either experi-mental data or their alternate solution techniques, as per E0P-40-3.00. The response also stated this documented design review fully conforms to the verification requirements of Regulatory Guide 1.64 and NEB 0 commitments outlined in NED0-11209. This item will be reviewed during a future inspection.

ORGANIZATION: GENERAL ELECTRIC COMPANY NUCLEAR ENERGY BUSINESS OPERATIONS SAN JOSE, CALIFORNIA REPORT INSPECTION N0.: 99900403/85-01 RESULTS: PAGE 5 of 12

8. (0 pen) Nonconformance (84-04):

Contrary to E0P 42-6.00, the method from which analytical results were obtained in the SAP 4G07 computer program verification problems 4.1, 4.2, 5.1, 8.1, and 14 was not referenced, nor were any hand calculations included.

As stated in GE's written response to the NRC regarding Nonconformance 7 above, the SAP 4G07 computer code has been fully and independently verified by a NEB 0 design review team and judged to be adequate for its intended purpose. This item will be reviewed during the next inspection.

9. (Closed) Nonconformance (84-04): Contrary to E0P 40-3.00,

" Engineering Computer Programs," users had been reporting potential computer code errors verbally to the responsible engineer without the required documentation. GE personnel stated that no potential computer code errors had been discovered since the orevious NRC inspection and that proper procedures will be followed in the event of future code errors.

10. (Closed) Unresolved Item (84-04): This item concerned errors that were discovered in the RVRIZO2 computer code that is used by GE in containment system and piping design calculations. The ccmputer code has also been distributed to utilities for their own use.

During this inspection it was determined that:

a. The various utilities who obtained the computer code have been notified of the error and advised of potential consequences;
b. A survey of GE users of RVRIZ02 determined that no additional safety-related code applications had been performed;
c. The computer code has been' removed from the approved Level 2 status (i.e., not approved for safety-related analyses).

As a result of the corrective action taken by GE, this item is considered closed.

l OR.GANIZATION: GENERAL ELECTRIC COMPANY NUCLEAR ENERGY BUSINESS OPERATIONS SAN JOSE, CALIFORNIA REPORT INSPECTION N0.: 99900403/85-01 RESULTS: PAGE 6 of 12 E. OTHER FINDINGS OR COMMENTS:

1. Sodium Pentaborate Curve Error A potential deficiency in the Standby Liquid Control System (SLCS) at the Fermi 2 plant was the subject of a GE Field Deviation Disposition Request (FDDR) on January 9, 1985. This deficiency was related to an error that was discovered in the sodium pentaborate concentration data supplied in SLCS system specifications for FERMI.

This item was reviewed with the GE cognizant engineer, for the SLCS, who determined that the error was within the margin included in the system design and did not jeopardize the plant's ability to achieve safe shutdown.

2. Control Rod Drive Filters Movable inner filters for the control rod drive mechanisms (CRDMs) at the Monticello Nuclear Plant were supplied by GE as spare parts with incorrect mesh size, 2 mil instead of 10 mil. This situation occurred once in 1974 and twice in 1984. The new 10 mil replacement design is " stationary" in contrast to the former which was " movable,'-

i.e., moved along with the index tube during a reactor scram.

The old design may have resulted in screen clogging; the new 10 mil design allows for the passage of larger particles, thus reducing the possibility of clogging. GE stated that this new filter is easily recognized by the fact its screen material appears on the outside of the filter casing rather than the inside as in the case of the earlier 2 mil design.

GE personnel stated that this improved design was a response to slow scram times experienced at Oyster Creek in late 1971. This was accomplished by the issuance of Product Service Information Letter 71-21, dated December 29, 1971, advising customers to convert to the new 10 mil filter design. In reviewing another incident involving excessive scram times, at Monticello in December,1984, the major cause was attributed to the paper purge dam material (DESOLV0) and corrosion / cleaning byproducts in the system, and not the incorrect size filters furnished by GE.

Oyster Creek, the only other plant using 2 mil filters, is presently converting its remaining 12 CRDs to the 10 mil filters. At present, all 2 and 10 mil filters are in storage at GE and have been quarantined pending disposition instructions.

L

1 ORGANIZATION: GENERAL ELECTRIC COMPANY NUCLEAR ENERGY BUSINESS OPERATIONS SAN JOSE, CALIFORNIA REPORT INSPECTION NO.: 99900403/85-01 RESULTS: PAGE 7 of 12 GE's Potentially Reportable File (PRC 84-62) classified the CRD filters as a non-safety-related component which would not impact scram time performance even if totally plugged.

3. Defective Circuit Breakers at Vogtle Approximately 239 defective GE circuit breakers (models AKR30 and AKR50) were identifiad at Vogtle. The defective breakers were originally identified to GE by the A/E, Bechtel, on February 6,1984.

The breakers were in the Plainville, Connecticut, warehouse scheduled to be shipped to the Hope Creek Generatino Station. They had apparently been reworked per GE Service Advices 175-9.6, 175-9.7, and 175-9.11 by factory personnel who had not been previously involved in rework programs and consequently was accomplished without adequate retesting and reinspection. Per letter by GE-Contractor Equipment Business Operation (CEB0) dated February 23, 1984, all subsequent rework and generic reinspection were accomplished. GE-CEB0 notified NRC on February 24, 1984, of a potential safety hazard. Since GE-CEB0 is a subcontractor to the A/E, GE-NEB 0 did not have responsibility for this problem. This item will be reviewed with GE on a future inspection of the Plainville facility.

4. Neutron Monitor Power Supply Failures on Limerick 1 As a result of a -20 Vdc power supply failure for the Intermediate Range Neutron Monitors (IRM) during which the reactor failed to trip, GE-NEB 0 was requested to investigate. GE concluded that the system worked as designed and met all system specifications. However, Philadelphia Electric Company requested a change in the hardware design to provide information to the operator if the -20 Vdc supply should fail. This new design was documented in Field Deviation Disposition Request (FDDR) HH1-4460, Rev. 0, dated December 6,1984.
5. Dikkers Safety Relief Valve Equipment Qualification Test Failure A Potentially Reportable Condition file (PPC 83-12) was initiated by GE on April 19, 1983, concerning a test failure of actuator solenoids which initiate valve operation for the Automatic Depressurizer System. This item was subseouently determined not reportable to the NRC pursuant to 10 CFR Part 21. This conclusion was based on the performance capability of the equipment in accordance with the original GE specifications and requirements in effect prior

ORGANIZATION:

~

GENERAL ELECTRIC COMPANY NUCLEAR ENERGY BUSINESS OPERATIONS SAN JOSE, CALIFORNIA REPORT INSPECTION NO.: 99900403/85-01 RESULTS: PAGE 8 of 12 to the TMI eve..t. Subsequently, the failure of the actuator solenoids was reported to the NRC by GE's customers, Niagara Mohawk Power Corporation and Cleveland Electric Illuminating Company, June 1, 1983 (Region I) and July 29, 1983 (Region III), respectively.

These interim reports were issued pursuant to 10 CFR 50.55(e).

A separate file, PRC 84-44, on the same basic problem was reviewed for its contents concerning the Dikker Safety Relief Valves. There was no cross-reference between the two PRC files despite their almost identical problem. PRC 84-44 was opened on June 26, 1984, and reopened on November 11, 1984, and again determined not reportable under 10 CFR Part 21, but was found to be a condition germane to safety. The NRC was notified of this conclusion on November 28, 1984.

6. Perry Feedwater System Pipe Rupture Analysis Gilbert / Commonwealth (G/C), the A/E for Perry, made a 10 CFR 21 report which indicated design forces and other data originally calculated from G/C's Feedwater System (FW) pipe rupture analysis may possess unconservative assumptions, in light of a recent reanalysis performed by the NSSS supplier, GE, at the request of Cleveland Electric & Illuminating Company (CEI).

GE's reanalysis postulated a pipe break per NUREG-800 and calculated jet impingement loadings for selected target locations, including the effects of fluid thermodynamics and state, pipe friction, and the reactor vessel's contribution to such jet forces. The erroneous assumptions made by G/C considered only the FW pump side of the rupture, while ignoring the contribution of the reactor vessel to the total force. This assumption produced jet forces and shape profiles which underestimated the total pipe rupture effect.

1 1

ORGANIZATION: GENERAL ELECTRTC COMPANY NUCLEAR ENERGY BUSINESS OPERATIONS SAN JOSE, CALIFORNIA REPORT INSPECTION NO.: 99900403/85-01 RESULTS: PAGE 9 of 12 GE's final report, DRF B21-00306 dated November 14, 1984, indicated a total jet force 25% larger than that calculated by the G/C analysis.

In addition, the jet shape resulting from such a postulated rupture yielded an entirely different configuration than the G/C analysis, which may result in the failure of protection devices for essential equipment. A further review by G/C indicated the nonconservate assumptions existed only for the FW system.

CEI's final report to the NRC, dated February 14, 1985, indicated that the incident was reportable to the NRC pursuant to 10 CFR 50.55(e).

Their evaluation of the safety implications revealed increased jet forces affected four target locations thus requiring additional or modified equipment shielding. However, only two of the four targets have safe shutdown functions since they affect the control rod drive tubes at the bioshield wall interface. This report further states that following a design basis pipe rupture, on the reactor side of the FW pump, a loss of control rod insertion capability coupled with an inoperable Standby Liquid Control System, may impair the ability to achieve safe shutdown.

Corrective action is to include modification of all four target locations, which for Unit I have already been performed.

Modifications for Unit 2 will be completed consistent with its construction schedule.

7. GE Supplied Steam Leak Detection System AnintegratedelectricaltestwasconductedatShorehaminwhicb offsite power was cut-off to initiate the test. When the diesel generators picked up the load, it was discovered that both the Reactor Water Cleanup (RWCU) and High Pressure Coolant Injection (HPCI) systems had been isolated by temperature instruments in the Steam Leak Detection (SLD) system furnished by GE. The cause of the isolation was attributed to incorrect settings of the time-delay relays contained inside the circuitry of the Riley, Model 86, Temperature Switch Modules, which measure ambient and/or differential temperatures in the Emergency Core Cooling (ECC) systems equipment areas. ,,.

When power is first applied to these relays, they receive a temperature trip signal before achieving a steady-state condition.

This same condition was previously discovered at Limerick and several

ORGAN 12ATION: GENERAL ELECTRfC COMPANY NUCLEAR ENERGY BUSINESS OPERATIONS SAN JOSE, CALIFORNIA REPORT INSPECTION N0.: 99900403/85-01 RESULTS: PAGE 10 of 12 other BWRs. Corrective actions taken involved either a GE modi-fication to the internal wiring of the module or a relay replacement in the isolation circuits for systems containing Agastat time-delay relays.

The NRC inspector reviewed GE's evaluation of this incident at Shorehani, PRC File 84-47, which concluded that the condition was not reportable to the NRC, pursuant to 10 CFR Part 21, since the plant could be brought to a safe shutdown without the availability of the HPCI or RCIC systems, therefore, no significant safety hazard would exist. GE based their conclusion on a five plant FSAR review which indicated the availability of at least two ECCS pumps following any single failure in addition to isolation of the HPCI system. In addition, GE's review stated it could be shown that with one pump available, sufficient make-up flow would exist to provide adequate core cooling.

On January 15, 1985, GE issued Service Information Letter (SIL)

No. 416, "Riley Temperature Switches" to all BWR/4 operating plants.

GE's recommendation was for owners of BWP/4 plants and LaSalle (BWR/5) to review their temperature switch designs based on information contained in the notice. FSAR licensing calculations may have to be updated to reflect the modified set of available systems, should affected plants refrain from modifying SLD circuitry.

8. Ground Break Relay Deficiency in Class 1E Units at Hope Creek Eight Model TGSR-12 ground break relays manufactured by GE were found to have a defective component. The relays were supplied by one of GE's non-nuclear operations under a subcontract to the A/E, Bechtel. GE-NEB 0 has not received GE Service Advice Letter 175-9.2, which describes this deficiency to affected non-nuclear customers, therefore, they assume it is not in the NSSS area of responsibility. Bechtel has replaced the defective relays with acceptable units provided by GE and documented the deficiencies and corrective action.

No nonconformances or violations were identified during this part of the inspection.

ORGANIZATION: GENERAL ELECTRIC COMPANY NUCLEAR ENERGY BUSINESS OPERATIONS SAN JOSE,-CALIFORNIA REPORT INSPECTION N0.: 99900403/85-01 RESULTS: PAGE 11 of 12 E. PERSONS CONTACTED:

  • M. Blich
  • B. Smith N. Barclay
  • J. Fox
  • J. Case
  • G. Stramback R. Hill
  • E. Giambalvo R. Waldman B. Simon E. Chu
  • F. Hopkins
  • J. Wood D. Saxena R. Valencia R. Gridley H. Hwang J. Atwell N. Barker A. Amiri R. Bloomstrand T. Herczeg R. Siemer' C. Canham
  • Attended exit meeting.

F. DOCUMENTS EXAMINED:

1. Procedure, dated December 19, 1984, Engineering Computer Programs Change Notice A for E0P 40-3.00.
2. Internal memo, dated November 9, 1984, J. Fox to distribution, memo concerning Engineering Computer Program (ECP) Error Control.
3. Report, document no. DRF-B21-00306, Rev. O, dated November 13, 1984, Feedwater Line Postulated Break Analysis, Perry Unit 1.
4. Report, document no. PRC 84-47, Shoreham Leak Detection System.
5. Report, document no. PRC 84-62, CRD Inner Filter (Monticello).

ORGANIZATION: GENERAL ELECTRIC COMPANY NUCLEAR-ENERGY BUSINESS OPERATIONS SAN JOSE, CALIFORNIA REPORT INSPECTION NO.: 99900403/85-01 RESULTS: PAGE 12 of 12

6. Letter, dated February 23, 1984, from GE Contractor Equipment Business Operations (GE-CEB0), Phillip Piqueir, to Claude Turnbow, Bechtel Power Corp., Hancock, NJ.
7. Letter, do:.ument no. B015, dated February 24, 1984, from GE-CEBO, David Dixon, Manager QA, to NRC, Richard C. DeYoung.
8. Specification, document no. FDDR HH1-4460, Rev. O, dated December 6, 1984, SURNMS Elementary Diagram.
9. Letter, dated November 21, 1984, Licensee Event Report - Failed 20 Volt SRM/IRM Supply Preventing RPS Actuation.
10. File, document no. PRC 83-12, dated March 6, 1985, GE file on Dikkers Safety Relief Valve (SRV).
11. Letter, Revision 0, dated April 19, 1983, from J. Jacobsen to G. G. Sherwood, Potentially Reportable Condition of the Electro -

Pneumatic Actuation Assembly on the Dikkers Main Steam Safety Valve (SRV) to perform its Class IE Function under NUREG-0588, Category 1 Oualification Requirements.

12. Letter, Revision 0, dated May 13, 1983, from G. G. Sherwood to J. Jacobsen - PAC 83-12, Electro-Pneumatic Assembly on the DikkersSafety/ReliefValve(SRV).
13. Letter, Revision 0, dated February 24, 1984, from G. G. Sherwood to J. Jacobsen - same subject.
14. File, document no. PRC 84-44, dated March 6, 1985, Dikkers Solenoid Valve Failure During EQ Testing. ,
15. Letter, document no. HE 84-32, dated June 26, 1984, from H. Ehson to G. Sherwood, Potentially Reportable Condition Failure of the Dikkers Safety Relief Valve Sole'noid during the NUREG-0588 Category I EQ Testing.
16. Internal memo, dated November 26, 1984, PRC 84-44, Dikkers Solenoid Valve Failure During Environmental Qualification Testing.

d