ML20059F308
| ML20059F308 | |
| Person / Time | |
|---|---|
| Site: | Vogtle  |
| Issue date: | 10/26/1993 |
| From: | Office of Nuclear Reactor Regulation |
| To: | |
| Shared Package | |
| ML20059F298 | List: |
| References | |
| NUDOCS 9311040233 | |
| Download: ML20059F308 (4) | |
Text
[
[p na m
UNITED STATES y
- 4 S
NUCLEAR REGULATORY COMMISSION E*
WASHINGTON, D.C. 20555-0001 kv [
+**.*
SAFETY EVALUATION BY THE OFFICE OF NUCLEAR REACTOR REGULATION TEMPORARY EXEMPTION FROM APPENDIX J INTERVAL FOR LOCAL LEAK RATE TESTING OF CONTAINMENT PENETRATIONS V0GTLE ELECTRIC GENERATING PLANT. UNIT 1 DOCKET NO. 50-424
1.0 INTRODUCTION
By letter dated September 30, 1993, Georgia Power Company, et al. (the licensee), requested a license amendment to change the Vogtle Electric Generating Plant, Unit 1 (Vogtle), Technical Specification (TS) surveillance requirement 4.6.1.2d.
The requested change adds a footnote that extends the surveillance interval for the next required Type C leakage test of the auxiliary component cooling water (ACCW) supply and return containment isolation valves IHV-1974 (and associated check valve 1-1217-U4-113),
lHV-1975, lHV-1978, and 1HV-1979, to prior to entry into Mode 4 following the next scheduled refueling outage (or the next forced outage requiring entry into Mode 5), but no later than November 1, 1994.
The amendment provides a one-time only extension of the surveillance interval for the subject valves.
As presently written, TS 4.6.1.2d requires that 10 CFR Part 50, Appendix J, Section Ill.D.3, Type B and C tests for the subject valves be conducted at intervals no greater than 24 months.
In February 1992, the licensee prepared and implemented Licensing Document Change Request (LDCR) FS92-007 under the provisions of 10 CFR 50.59 and in accordance with Vogtle TS 6.4.1.6.
The LDCR revised Table 6.2.4-1 of the Vogtle Final Safety Analysis Report (FSAR), in part, with respect to the ACCW supply and return containment isolation valves.
Prior to the change, Table 6.2.4-1 stated that these valves were subject to 10 CFR Part 50, Appendix J,Section III.D.3, Type C leakage testing requirements, and that they were normally open during operation but closed under post-accident conditions.
However, as noted in footnote "g" to Table 6.2.4-1, ACCW flow should be l
maintained to the reactor coolant pumps (RCPs) under most post-accident i
conditions, if possible.
Therefore, the LDCR changed the leakage testing I
requirements from Type C to Type A and changed the post-accident position of the valves to "open."
In addition, the associated penetrations were added to FSAR Table 6.2.6-1 as penetrations that are not vented or drained during Type A testing. As a result of this LDCR, these valves were not Type C tested during the Vogtle Unit I spring 1993 refueling outage, although they had been tested during previous outages on both units.
1 The licensee's basis for the LDCR was that the subject valves do not receive a containment isolation signal (they are remote manually operated), and the associated penetrations are needed to maintain cooling water to the RCPs.
The 9311040233 931026 PDR ADOCK 05000424 P
pyg
licensee thought that the ACCW was a closed system because it does not communicate directly with the containment atmosphere or primary coolant.
Thus, when approving the LDCR, the licensee had concluded that Type A testing was sufficient for these penetrations.
However, during a recent document review, the licensee discovered that the safety evaluation for the LDCR was flawed.
The evaluation failed to consider that, while the ACCW system is seismic category 1 and the hard piping is fabricated of ASME Section III, Class 3 materials, the system had been installed in accordance with ANSI B31.1 and no N-stamp was affixed.
In addition, some components, such as motor coolers and flexible piping, are not composed of Class 3 materials. Therefore, the ACCW system does not meet the ANSI standard criteria for a closed system.
Consequently, the supply and-return isolation valves must be considered to perform an isolation function and should be subject to Type C testing.
2.0 EVALUATION The subject valves have been Type C tested during all previous refueling outages with the exception of the Unit 1 Spring 1993 outage.
The licensee reviewed the maintenance work order (MWO) history of the ACCW containment isolation valves.
This review found MW0s for seat leakage, packing leaks, flange leaks, preventive maintenance, and several inspections, but found no '
"as found" Type C local leak rate test (LLRT) failures after the initial entry -
into Mode 4 on either Vogtle unit.
-l The licensee also reviewed the LLRT history of the valves after initial Mode 4 entry and found this history to demonstrate the reliability and low leakage trends of these valves.
Listed below are the maximum values, taken from six refueling outages between the two units, for both the "as found" and "as left" l
LLRTs performed after initial Mode 4 entry. The below values indicate the i
" worst case" leakage.
Penetration 28 is the ACCW supply line and penetration
- l 29 is the ACCW return line.
j PENETRATION 28 PENETRATION 29 MAXIMUM LEAKAGES MAXIMUM LEAKAGES IHV-1978 - 20.5 sccm 1HV-1974 = 152 sccm*
IHV-1979 - 40.4 sccm IHV-1975 - 62.0 sccm 2HV-1978 = 49.2 sccm 2HV-1974 - 99.6 sccm*
2HV-1979 = 90.6 sccm 2HV-1975 = 136.3 sccm
- Includes leakage through associated check valve 1-1217-U4-113 The Vogtle Inservice Inspection Program currently specifies a maximum allowable leakage of 1000 sccm for each butterfly valve and 1500 sccm for the check valve.
The leakage limit for the combination of valve IHV-1974 and-check valve 1-1217-04-113 would be 2500 sccm.
These limits were not based on Appendix J requirements, but were established based on the low leakage history of these valves and define the point at which repair would be required. The.
Appendix J leakage limit for all penetrations subject to Type B and C testing (0.6L at Vogtle is 228,273 secm.
The current total for Type B and C test leaka,)e at Vogtle, as of September g
10, 1993, is 14,398.8 secm. As of the last 9
. LLRT, the leakage for each of these four valves was as follows: 1HV-1974 - 152 i
sccm (this includes leakage past check valve 1-1217-U4-ll3 in parallel with j
1HV-1974); IHV-1975 - 11.6 sccm; IHV-1978 - 9.3 sccm; and 1HV-1979 -
11.4 sccm.
The test pressure, P was 45 psig at the tine these numbers were obtained.
The test pressure has,,since been reduced to 37 psig in accordance with previous license Amendments 63 (Unit 1) and 42 (Unit 2), and the leakage would be less at this lower pressure.
]
During the last outage for Unit 1, the licensee performed maintenance on 1HV-1979 that could have affected its leakage, but performed no LLRT since it was not required by the FSAR at the time.
The maintenance involved removal of the motor and gearbox and altering the limit switch settings, but no work was done that would have affected the valve seat. The standard work practice for setting limit switches on this type of soft-seated butterfly valve following this type of maintenance is as follows:
first, the valve is manually closed using the hand wheel until 0 (fully closed) is reached, and the limit switch is set.
Then, the limit switch is tested by manually operating the valve again.
Finally, the valve is stroked using the motor until the limit switch actuates.
At this point, the hand wheel is used to ensure that the valve is seated properly after the limit switch actuates. As a reference point, in the Spring of 1992 this type of work was performed on Unit 2 valve 2HV-1978 and ;
pre-maintenance and post-maintenance LLRTs were performed.
The pre-and post-maintenance leakage was well within the leakage limits for this valve.
The probability of containment isolation failure following a core damage accident is modeled in tne Vogtle individual plant examination (IPE). The IPE was submitted by letter dated December 23, 1993.
In order to model a more conservative scenario of containment isolation failure than was considered in the base case Vogtle IPE, the licensee assumed that the occurrence of any core damage scenario would cause a break in the ACCW flow path and that the operator would be required to isolate the ACCW system for successful containmeni, isolation.
Based on a Type C test interval of 2 years, the frequency of cora danege with containment isolation failure was found by the licensee to be on the order of 10-7 per reactor year. The licensee has stated that extending the requirad Type C test interval for these valves beyond the Appendix J 2-year period has a negligible impact on that probability. Thus, the probability of an event that leads to core damage and a failure of the ACCW piping inside containment with a failure to isolate containment is not considered to be credible by the licensee.
The staff concurs that the additional operation period, between expiration of the current leak tests to prior to entry into Mode 4 following the next scheduled refueling outage (or the next forced outage requiring entry into Mode 5), but no later than November 1,1994, is not expected to significantly decrease the margin between expected as-found leak rate and L,.
The ACCW system is seismic category 1, and the hard piping is fabricated of ASME Section III, Class 3 materials.
Some components, such as motor coolers and flexible piping, are not fabricated of Class 3 materials. The licensee concluded that, even though the ACCW does not meet the ANSI standard criteria for a closed system, it can be considered to be highly reliable and that there is reasonable assurance that for most events its integrity would be maintained.
The staff concurs with this conclusion.
9
l 1 The NRC staff also finds that the 2-year interval requirement for Type 0.and C components is sufficient for timely detection of significant deterioration while, at the same time permitting the tests to be performed during plant outages.
Leak rate testing of the penetrations during shutdowns is preferable because of the lower radiation exposure to plant personnel.
Some penetrations can not be tested at power.
For those penetrations that can not be tested during power operation or for which testing at power is inadvisable, the increase in confidence of containment leaktight. integrity following a successful test is slight and does not justify a plant shutdown specifically j
to perform the tests within the 2-year time period, considering the factors discussed above.
3.0 CONCLUSION
Based on the above evaluation, the NRC staff finds the requested one-time only exemption to TS surveillance requirement 4.6.1.2d, is acceptable.
As provided in the footnote, the surveillance interval for the next required Type C leakage test of the ACCW supply and return containment isolation valves 1HV-1974 (and associated check valve 1-1217-04-113), 1HV-1975, 1HV-1978, and IHV-1979, is extended for Vogtle Unit I to " prior to entry into Mode 4 following the next scheduled refueling outage (or the next forced outage requiring entry into Mode 5), but no later than November 1, 1994."
1 Principal Contributor:
C. E. Carpenter, Jr.
Date:
October 26, 1993
)
l
)
W-