ML20205B957
| ML20205B957 | |
| Person / Time | |
|---|---|
| Site: | Hatch  |
| Issue date: | 08/13/1985 |
| From: | Ellen Brown NRC OFFICE FOR ANALYSIS & EVALUATION OF OPERATIONAL DATA (AEOD) |
| To: | |
| Shared Package | |
| ML20205B947 | List: |
| References | |
| TASK-AE, TASK-T507 AEOD-T507, NUDOCS 8509190241 | |
| Download: ML20205B957 (5) | |
Text
. . . - . . . . .. ,_
l AEOD TECHNICAL REVIEW REPORT *
. UNIT: E. I. Hatch, Unit 2 TR REPORT NO.: AE0D/T DOCKET N0.: 50-366 DATE:
LICENSEE:- Georgia Power Company EVALUATOR / CONTACT: Earl J. Brown t'SSS/AE: General Electric /Bechtel
SUBJECT:
STANDBY LIQUID CONTROL SYSTEM PRESSURE RELIEF VALVES LIFT AT A PRESSURE LOWER THAN REACTOR COOLANT PRESSURE EVENT DATE: May 5, 1984 (LER 366/84-005)
SUMMARY
Both loops of the standby liquid control system would have been unable to inject sodium pentaborate because, during a test, relief valves lifted at 600 and 700 psig rather than the set point range of 1350 psig 25 psig.
]
A review of available data-suggests that these events were isolated cases and were the result of degraded valves rather than an indication of a' generic problem. No additional AEOD effort seems necessary at this time.
j DISCUSSION It was discovered during bench tests that pressure relief valves in the standby liquid control (SLC) system at Edwin I. Hatch, Unit 2 lifted at a pressure lower than the set pressure of 1350 psig 25 psig. The event l
occurred on May 5, 1984 and is described in LER 366/84-005. The test pro-cedure-is performed at six month intervals. The "A" loop relief valve, 2C41-F029A, lifted at 600 psig and the "B" loop relief valve, 2C41-F029B, lifted at 700 psig. Since this' lift pressure was several hundred psig below the set pressure, it represents a situation quite different from normal set point drift.
A review of the P&ID drawing indicates that lifting of either of these relief r valves would result in recirculation of the pump discharge flow back to the intake side of the SLC pump. Hence, in the event of an injection signal, the as found condition of the relief valves.. indicates that the sodium pentaborate solution would just recirculate through the SLC pump rather than be injected into the reactor vessel when the reactor pressure is greater than 700 psig.
Investigation by the licensee revealed the stem on relief valve 2C41-F029A was broken at the cotter pin hole which permitted the valve spring to decompress and lose tension. Relief valve 2C41-F029B was found to have excessive wear on the valve stem and the valve stem guide.
I NE i
- This document supports ongoing AE00 and NRC activities and does not represent the position.or requirements of the responsible NRC program office.
The immediate corrective action by the licensee was to replace the damaged relief valves, Lonergan Model LCT-20, with a different model, LCT-30, supplied by the same manufacturer. Since this mode of failure has potential generic implicatiuns which resulted in complete inability to inject sodium pentaborate -
when needed, a search of the Sequence Coding Search System (SCSS) data base was conducted to review for similar events.
The data base search of SCSS was conducted for relief valves in standby liquid control systems and all events in standby liquid control systems. There were 43 events involving the SLC system, but only 11 events involved a relief valve in the SLC system. Table 1 is a list of the 11 events with LER number, plant name, event date, and event description with the cause of failure if it was stated in the LER. It is evident from the 11 events that low relief valve lift pressures of 600 and 700 psig only occurred for two valves at Hatch Unit 2 and one at Pilgrim. The valves at Hatch 2 were physically damaged with a broken stem on one valve and excessive wear on the other valve. However, one event at Pilgrim (item 3) identifies a low lift pressure at approximately 600 psig but the cause could not be determined. Subsequent testing of the valve at Pilgrim resulted in valve lift within set point specifications. Although the very low lift pressure of 600 psig was similar at Hatch and Pilgrim, the valves were supplied by different manufacturers. The valves at Hatch were manufactured by Lonergan and that at Pilgrim was manufactured by Crosby.
The valves identified for all events in Table 1 were manufactured by either Crosby or Lonergan. The valves at Pilgrim and Vermont Yankee are Crosby valves and those at Hatch and Brunswick are Lonergan valves. Hence, only four plants have reported problems with relief valves lifting outside of specifica-tion requirements. Furthermore, at least two sites have identified that relief valve lift problems appear related to the test procedure. Discussions with licensee staff at Brunswick (items 7, 8, and 9) revealed that the test procedure was the primary cause of relief valve lifting outside of specifica-tion limits. The reduced time interval between tests was apparently not effective (18 months to 6 months) in reducing improper lift pressure. However, the test procedure was changed from using the SLC pump pressure and audio recognition of valve lift to bench testing with water and a portable positive displacement pump. Since that procedure was implemented, there have been no further reports involving SLC relief valves.
A search of NPRDS was conducted in an attempt to further assess potential generic implications. There were no additional BWR plants identified in that search that use the Model LCT-20 relief valves. However, the search did identify use of the same model valve at a few PWR plants in the service water system, component cooling water system, feedwater system, and auxiliary feed-water system. However, no unusual events were identified.
FINDINGS AND CONCLUSIONS A review of the available operating data appears to suggest that the two relief valves lifting at 600 to 700 psig at Hatch 2 were isolated events.
Furthermore, those valves were degraded in a manner that has not been reported by other licensees. It also appears that certain test methods used (like at Brunswick) to establish the pressure at which these valves lift may be the primary cause of reported lift pressures outside the acceptable set pressure range.
Based on available information, it is concluded that these events do not represent'a generic issue for BWR plants relative to inability to inject sodium pentaborate into the reactor vessel when needed. There does not appear to be a need for additional NRC action at this time.
k
.i e
4
~ -
_4_
TABLE 1 RELIEF VALVE EVENTS IN STANDBY LIQUID CONTROL SYSTEM Event LER Number / Plant Date Event Description
- 1. 271/81-031 11-9-81 Relief Valve SR-39-B was found Vermont Yankee to have set point of 1540 psig.
Set point range was 1400-1490 psig. Suspected cause of drift was crystallization of sodium pentaborate.
- 2. 271/84-013 7-21-84 Relief valves SR-11-38 A&B were Vermont Yankee found to have a set point below the required range. Most probable cause of low setpoint was the testing technique.
- 3. 293/85-001 1-1-85 SLC system was inoperable due to Pilgrim 1 debris in SLC system storage tank and test tank. The "A" relief valve lifted at about 600 psig vs.a set point of 1450 psig. The "B" relief valve, which lifted below setpoint but within range, was found with pieces of rubber gloves between the blowdown adjus-ting orifice and plunger. Cause of "A" lifting low is unknown. It lifted at the required setpoint on subsequent bench. tests.
- 4. 321/81-006 1-19-81 SLC loop "A" relief valve, IC41-Hatch 1 F029A, lifted at 1200 psig vs. a setpoint of 1325 75 psig. Cause was attributed to setpoint drift.
- 5. 321/81-020 3-18-81 SLC loop "B" relief valve, Hatch 1 1C41-F029B, lifted at 1180 psig vs.a setpoint of 1325 1 75 psig.
Cause was attributed to setpoint dri f t.
'- . 5 l
. 1 TABLE 1 (Continued)
- 6. 321/83-102 10-11-83 'SLC relief valve IC41-029A would Hatch 1 not lift in the range 1325 75 psig. Valve was disassembled, and cleaned and bench tested successfully.
- 7. 324/82-041 3-14-82 SLC relief valve 2-C41-F029A Brunswick 2 lifted at 1465 psig vs setpoint of 1400 50 psig. Further evalua-tion is planned. Test frequency was changed to 6 month intervals from 18 month intervals.
- 8. 325/80-057 7-3-80 SLC relief valve 1-C41-F029B did Brunswick 1 not lift in the setpoint range of 1400 50 psig. Cause was attributed to setpoint drift.
Test interval increased to every 6 months.
! 9. 325/83-019' 4-25-83 Both units had SLC relief valves Brunswick 1 outside the setpoint range of 1400 50 psig. Unit I relief valves 1-C41-F029A and F029B lifted i
at 1321 psig and 1529 psig respec-tively. Unit 2 relief valves 2-C41-F029A and F0298 lifted at 1403 psig (acceptable) and 1323 psig respectively. Cause appears to have been the test method which utilized SLC discharge pressure and an audible report of valve opening.
- 10. 366/81-014 3-18-81 SLC relief valve failed to lift Hatch 2 at 1400 psig vs. set pressure of 1350 50 psig. Cause was set-point drift.
- 11. 366/84-005 5-25-84 SLC relief valves in loops "A" Hatch 2 -
and "B" (2C41-F029A and 2C41-F029B) lifted at 600 psig and 700 psig respectively. Would have been unable to inject sodium pentaborate into reactor vessel.
Cause was damaged and worn valve components.
.