ML20126F333
| ML20126F333 | |
| Person / Time | |
|---|---|
| Site: | Vogtle  |
| Issue date: | 12/23/1992 |
| From: | Office of Nuclear Reactor Regulation |
| To: | |
| Shared Package | |
| ML20126F331 | List: |
| References | |
| GL-89-10, NUDOCS 9212300237 | |
| Download: ML20126F333 (7) | |
Text
. _ - _ _ ._ - --__ - _ _ .
.;k*R ~
3 '
n NUCLEAR REGULATORY COMMISSION-k Ek : WASHINGTON. D. C. 20555
- u / ,
....+ .
SAFETY EVALUATION BY THE OFFICE OF NUCLEAR REACTOR REGULATION REGARDING SCHEDULE EXTENSION-TOL COMPLETE GENERIC LETTER 89-10 PROGRAM V0GTLE ELECTRIC GENERATING PLANT. UNITS 1 AND 2 DOCKET NOS. 50-424 AND 50-425 Introduction "
On December 27, 1989, Georgia Power Company (the licensee) submitted its response to Generic Letter (GL) 89-10, " Safety-Related Motor _0perated Valve.
Testing and Surveillance," for the Vogtle Electric Generating Plant (Vogtle),
Units 1 and 2. In that submittal, the licensee stated that each motor-operated valve-(MOV) in its GL 89-10 program would be set up l statically,-but :
that it planned no differential pressure testing of MOVs beyond"that already performed in conjunction with NRC Bulletin 85-03. The licensee questioned the applicability of the results of-NRC-sponsored tests to typical MOVs in nuclear plants, Nevertheless, the licensee stated that any. additional-data and-guidance would be factored into its M0V program to ensure that the program remained current.
In a reply on September 25, 1990, the NRC staff stated that research results-and operating events had revealed that, in many instances; static tests of MOVs did not demonstrate capability under design-basis conditions.-
Consequently, the staff recommended full differential pressure testing where practicable. If an alternative could not be justified, the NRC staff recommended the two-stage approach as described in GL 89-10 and its supplements. The NRC staff stated that- the licensee's justification for alternatives to differential pressure testing would.be evaluated during inspections.
From January 27 to 31,1992, the_ staff inspected the Vogtle: program. developed by the licensee in response to GL 89-10. NRC Region II-led the--inspection-with the participation of an NRR contractor from-the Idaho National Engineering Laboratory (INEL).- The inspectors found that-the' licensee had.
identified a total of 280 M0Vs to be included in the GL 89-10 program at:
Vogtle Units 1 and 2. ._The licensee had completed its determination of _ worst-case differential pressure for MOVs within the program. The-licensee had:also
. performed' sizing and switch setting calculations for MOVs within its program..
For Westinghouse valves, the licensee ~used-the methodology prescribed by Westinghouse for- evaluating M0V size and switch settings. - For valves' that were not supplied by. Westinghouse, the licensee attempted to use a valve factor of'0.5 in the Limitorque' standard equation, but'if the MOV was_not adequate for this valve factor, then 0.3 was substituted. 'The' licensee assumed a stem friction coefficient of 0.15,.but had not justified this _
< assumption for its 36-month stem lubrication frequency. The licensee had not 9212300237 921223 DR ADOCK O g4 y
t
..z .
~
formalized.the' consideration of diagnostic equipment inaccuracies, torque 1 switch repeatability, or load: sensitive behavior-(rate of loading). The licensee's MOV Program
Description:
Document!did not indicate any testing ~of MOVs- under differential pressure and-flow conditions to' be performed in -
response to GL 89-10. Under NRC. Bulletin 85-03, the-licen'see:had_-tested 45 MOVs, but the inspectors found:that the testing was conducted below design-basis differential pressure conditions and the inspectors-had. concerns-with:
diagnostic equipment accuracy, test conditions, and fluid parameters.- The inspectors reported that the licensee also had concerns with 'the use of NRC Bulletin 85-03 test data. The licensee's MOV Program Description Document stated that the two-stage approach of GL 89-10 would be followed. However, the inspectors found that the licensee was not intending to follow the GL 89-10 two-stage approach because only static testing was plannSd without testing under maximum achievable. conditions.
The inspectors found that the progress and near-term plans for implementing the_ licensee's MOV program were limited.- The inspectors were concerned that the schedule specified in GL 89-10 for resolving the MOV issue would not-be 3 met at Vogtle. Among the concerns identified by the inspectors were:-
(1) The licensee's plan for demonstrating MOV operability at design-basis conditions had not started; (2) Program development for in-situ test guidelines and procedures and the prioritization of MOVs to be tested had not started;-
(3) Procurement of additional diagnostic equipment was being
. considered; and, (4) Modification of approximately 54 MOVs planned for 1993 refueling outages for Units 1 and 2 would impact the remaining refueling outages in.which testing could be conducted.
In its letter of March 4, 1992, forwarding the_ inspection report to the licensee, the NRC staff requested the licensee to respond to_three specific concerns:
-(l) Identification of' the methodology for implementing the.. .
two-stage. approach. described in GL 89-10-and its
, supplements; s
(2) Performance of.-in-situ testing at design-basis conditions for each applicable-MOV, where practicable; and (3) Compliance-with the GL 89-10 schedule completion date of June 28, 1994.
In its response -to'the inspection-report on July'9,1992, the licensee stated that its'GL 89-10 program w'as being implemented in accordance with'the two--
stage approach of the Generic Letter..;The licensee stated that it was-setting:
MOVs statically using diagnostic-equipment under_ Stage 1, and that_it would L verify.its methodology using a variety of techniques, including design-basis -
l
, s.#-, =., ,.,-c . - -- _ , , . __: . r =
differential pressure testing, under Stage 2. In reviewing the licensee's response, the NRC staff found that the licensee was not following the recommendations of GL 89-10 for design-basis testing and the two-stage approach, in GL 89-10, the NRC staff requests licensees to test MOVs under design-basis conditions where practicable. Where such testing is not practicable, the NRC staff describes Stage 1 in Supplement 1 to GL 89-10 as testing the M0V under maximum achievable conditions and verifying the size and setting of the MOV using the best data available. GL 89-10 states that, under Stage 2, licensees should verify MOV sizing and settings using applicable data.
In its July 9, 1992, response, the licensee noted that it had not committed to perform any additional testing under GL 89-10 in its December 1989 response.
However, the licensee stated that it had reassessed its position on differential pressure testing in response to a recent concern for the accuracy of its MOV diagnostic equipment and additional concerns raised relative to anomalous valve behavior under high differential pressure conditions. The licensee provided a plan for testing a total of about 75 MOVs under maximum achievable conditions at Vogtle Units 1 and 2. The licensee described its method of selecting MOVs for testing based on design-basis dif ferential pressure conditions, analytical margin, and achievable test conditions.
In its response of July 9, 1992, the licensee stated that its static testing of MOVs would be completed within the five-year (June 28, 1994), or three-refueling-outages schedule of GL 89-10. The licensee noted that 54 MOVs required modification based on its GL 89-10 design review and sizing calculations. The licensee intended to complete those modifications by June 28, 1994 (or the GL 89-10 three-refueling-outages schedule), but stated that it might delay completion until June 28, 1995, if problems occurred. The licensee then requested an extension until June 28, 1995, to complete differential pressure testing of about 75 MOVs and to justify the capability of MOVs not tested. The licensee based its request for the schedule extension on:
(1) Fif ty-four MOVs requiring modification; (2) Problems identified with ITI-MOVATS diagnostic equipment in the summer of 1991 which limited the amount of static testing in subsequent outages; and (3) The Electric Power Research Institute (EPRI) MOV Performance Prediction Program scheduled for completion in mid-1994.
Based on its review of the July 9 submittal, the NRC staff did not consider the licensee to have provided adequate justification for an extension of the GL 89-10 schedule. On September 22, 1992, the NRC staff met with the licensee's representatives to discuss its M0V program and to determine the extent to which the licensee could demonstrate the capability of its safety-related h0Vs within the recommended schedule of GL 89-10. Following-that meeting, the NRC staff requested additional information on the licensee's plan l and schedule for its GL 89-10 program. The NRC staff also stated that it I
t
. 4 would audit the licensee's plans and schedule upon receiving the additional information. The licensee provided the information by letter on October 9, 1992.
A11dit On November 12 and 13, 1992, the NRC staff and its INEL contrac'or audited MOV activities for Vogtle at the licensee's corporate offices in Birmingham, Alabama. During the audit, the staff discussed the licensee's plans and schedule for completing its program in response to GL 89-10.
As part of its response to GL 89-10, the licensee is verifying that the MOVs in its GL 89-10 program (except Westinghouse valves) are sized and set to accommodate the thrust requirements associated with a 0.5 valve factor and inner valve seat diameter (or orifice area). For valves supplied by Westinghouse, the licensee is verifying that the motor actuators can deliver sufficient thrust to operate the valves using the standard Limitorque equation and the Westinghouse-provided valve friction coefficients (which are similar '
to 0.5 valve factors). The licensee had previously relied on stall thrust data provided by Westinghouse which has been revealed to not constitute an adequate basis for determining MOV operability (see NRC Information Notice 92-70). The licensee is planning to modify 54 MOVs that cannot accommodate the thrust requirements with these valve factors and friction coefficients such that all MOVs within its GL 89-10 program will be able to accommodate thoss thrust requirements by the end of 1993.
The licensee stated that 12 of the 54 MOVs to be modified had been tested under dynamic conditions as part of its response to Bulletin 85-03. During the audit, the staff reviewed licensee's plans to modify the remaining 42 MOVs and did not identify any immediate operability concerns. However, in-anticipation of the NRC audit, the licensee had determined that two of those MOVs had their torque switches set too low to ensure their operability. The licensee declared the two MOVs inoperable on November 9,1992, and set the switches to an adequate level. The licensee had previously determined that nine MOVs without any automatic actuation signals had inadequate capability to isolate flow in the event of a postulated passive line break. The licensee.
established administrative controls in the form of standing orders on-November 11, 1992, to direct reactor operators to reduce the differential pressure before attempting to operate the MOVs from the control room. NRC Region 11 will conduct a follow-up inspection at Vogtle to review the promptness of the licensee's response to the determination of inadequate capability of those 11 MOVs and the adequacy of standing orders to provide directions to reactor operators for reducing differential pressure.
The licensee is assuming a stem friccion coefficient of 0.15 for all of its MOVs for both static and dynamic operating conditions. During the audit, the NRC staff reviewed preliminary test data provided by the licensee that revealed that the stem friction coefficient for some MOVs was near 0.15 under static conditions. For those MOVs, the. stem friction coefficient may be greater than 0.15 under cyn&mic conditinnt. The licensee plans to verify its j
- ~ . _ . , _ . . . _ _ - _ - . _ _ - ---- _ ..... _ . _ _ .._ .. . - _ _ , _ ,,-,,,m. , , , , _
5 stem friction coefficient assumption as part of its GL 89-10 program. The 14RC staff will review the licensee's verification of its stem friction coefficient assumption during future inspections.
The licensee stated that all of the MOVs within its GL 89-10 program will be statically tested by June 1994. In its letter of July 9, 1992, the licensee described its method of separating the MOVs with active safety functions within its GL 89-10 program into priority classes for the purpose of determining the MOVs that will be dynamically tested. In the July 9 letter, the licensee stated that, by June 1995, it planned to have differential-pressure-tested all MOVs in its two highest priorities (Priority Classes 5 and
- 6) where such testing was practicable and a differential aressure greater than 50% of the design basis differential pressure could be aciieved during the test. The licensee estimated that approximately 75 of the 99 HOVs in Priority Classes 5 and 6 would be dynamically tested by June 1995 under its criteria.
During the f1RC's November 12-13 audit, the licensee stated that, by June 1994, it would incorporate the information from its dynamic testing performed by that date into its methodology for sizing and switch setting calculations for all Priority 5 and Priority 6 MOVs and applicable MOVs in the other priority classes.
During the November 12-13 audit, the NRC staff reviewed the MOVs within the GL 89-10 program that the licensee had grouped by type, size and manufacturer.
The NRC staff noted that the licensee's test plan did not include several groups of MOVs. For example, the test plan excluded all butterfly valves and all 6 , 8 , 10 , and 14-inch gate valves, further, the staff noted that the licensee was planning a significant amount of testing for the last outages at each unit before June 1995, which might not allow sufficient time to resolve any problems that develop before the June 1995 completion date.
The NRC staff requested that the licensee respond to the audit by letter to confirm the schedule.for completing the MOV modifications, the static testing, and the dynamic testing. The licensee agreed to provide the written response and stated that it would reconsider its test plan. .
In a letter on December 2, 1992, the licensee made the following commitments:
(1) To implement, in the 1993 refueling outages, all modifications currently identified for GL 89-10 MOVs.
l (2) To ensure, upon completion of the 1993 modifications, based on an
! analytical evaluation, that all Anchor-Darling gate valves will be capable of operating assuming a 0.5 valve factor, and that all
, Westinghouse EMD gate valves will be capable of operating using the I
Westinghouse supplied disc coefficients.
(3) To perform, by June 28, 1994,- the initial static testing-of all GL 89-10 MOVs using diagnostic equipment.
I h
(4) To dynamically test, by June 28, 1995, all valves in its revised Vogtle Differential Pressure Test Plan. (The licensee indicated that its test plan has been revised to add several butterfly, 6-inch gate, and 8-inch gate valves, while removing several globe valves from the plan.)
(5) To apply, by June 28, 1994, all Vogtle differential pressure test data that are available following the 1993 refueling outages to similar valves, as applicable.
in the list of priority classes for MOVs within the GL 89-10 program at Vogtle provided in the licensee's July 9 lotter, the NRC staff found that the priority class assigned to MOVs 1HV8471 A and B appeared to be incorrect based on the licensee's method of assigning priority classes. The licensee agreed and stated that it would review the list of priority classes for its MOVs. In its December 2 letter, the licensee provided an updated differential pressure-test priority matrix.
During the audit, the NRC staff reviewed the licensee's determination of MOVs that met its criteria for dynamic testing. The NRC staff found that MOVs HV8804A and B for Vogtle Units 1 and 2 appeared to meet the licensee's criteria for being dynamically tested, but were not identified to be tested.
The licensee agreed and stated that it would review its identification of MOVs that will be dynamically tested, in its December 2 letter, the licensee provided a revised list of the higher priority MOVs that it considers testable.
The licensee has completed its evaluation of the motor-operated block valves HV8000A and B for the power-operated relief valves (PORVs) at Vogtle Units 1 and 2 in response to item II.D.) of NVREG-0737. However, the licensee stated ,
that its assumptions in predicting the thrust required to close the block ,
valves would be verified through industry testing being conducted at Comanche Peak Unit 2. The NRC staff noted that recent testing performed at Comanche Peak revealed higher thrust requirements to close the PORV block valves than predicted at that facility. The NRC staff requested-that the licensee determine the applicability of the data obtained from Comanche Peak for its PORV block valves and provide its plans to the NRC staff. In its December 2 letter, the licensee stated that, based on discussions with Westinghouse,'the PORV block valves at Vogtle are different from those at Comanche Peak.
Therefore, the licensee does not believe'that-the Comanche Peak test data are directly applicable to the PORV block valves at Vogtle. However, the licensee stated that EPRI is evaluating the test data from-Comanche Peak and data from testing a similar Westinghouse valve at a test facility. In its December 2 letter, the licensee committed to evaluate the EPRI analysis for applicability to the Vogtle MOVs when that analysis is made available to licensees.
During the November 12-13 audit, the NRC staff noted that, contrary to the recommendation of GL 89-10 and its supplements, the licensee is.not planning to test MOVs at maximum achievable conditions as part of the first stage of the two-stage approach. The licensee stated that only a few MOVs could be tested at less than 50% differential pressure with any significant differential pressure. Testing at partial differential pressure helps validate valve factor and stem friction coefficient assumptions and provides
1 7
information to help justify grouping of MOVs. Further, the Vogtle Final Safety Analysis Report states, in Section 17.2.3, that adequacy of design changes shall be verified by the performance of design reviews, alternate calculations, or qualification testing. During future ins)ections, the ,
t licensee will .be expected to demonstrate design-basis capa)ility of safety-related MOVs and the NRC staff will consider the basis for that demonstration, including, where applicable, the lack of dynamic test data, f in reviewing the licensee's operability evaluations for MOVs that are to be i modified, the NRC staff found the evaluations for MOVs IHV-88040 and 2LV-0122E to be unclear. The licensee agreed that, although an operability concern did not exist, theta evaluations were not clear. The licensee stated that the evaluations would be clarified.
The NRC staff discussed the licensee's use of varying differential pressure requirements for specific MOVs for different purposes, for example, the '
licensee had used a higher differential pressure for u) grading certain MOVs than when determining whether the MOVs could perform t1eir safety functions.
The NRC staff stated that the licensee should document more fully its justification for the use of varying differential pressure requirements.
In its letter of October 9, 1992, the licensee stated that it was developing a procedure for the evaluation of data from tests of MOVs under differential-pressure and flow conditions. During the November 12-13 audit, the NRC staff discussed the licensee's plans for developing that procedure. The licensee will need to ensure that thrust, torque, load sensitive behavior, eculpment inaccuracy, and torque switch repeatability observed under test concitions are addressed for design-basis conditions before each tested MOV is returned to service. ,
Conclusions from the inspection of the GL 89-10 program at the plant site and the audit at the licensee's corporate offices, the NRC staff considers the licensee's test plan to be capable of demonstrating, by June 1994, the adequacy of its overall methodology for verifying the proper sizing and setting of MOVs at Vogtle. in demonstrating the adequacy of the methodology, the NRC staff will expect the licensee to verify its valve factor and stem friction coefficient assumptions.
Further, although the licensee has expanded the scope of its test program, the licensee will need to obtain test data from other sources because its plan '
does not include any 10 or 14-inch gate--valves at Vogtle, Based on the commitments in the licensee's letter of December 2,1992, (previously identified in this Safety Evaluation), the NRC staff concludes- that the licensee's schedule for completing its MOV testing program-in response to GL 89-10 at Vogtle is acceptable. The NRC staff will evaluate the implementation of the licensee's program in response to GL 89-10, including its December 2, 1994, commitments, during future inspections.
Date: December 23, 1992