ML20161A165
| ML20161A165 | |
| Person / Time | |
|---|---|
| Site: | Turkey Point  |
| Issue date: | 06/09/2020 |
| From: | Christopher Hunter NRC/RES/DRA/PRB |
| To: | |
| Littlejohn J (301) 415-0428 | |
| References | |
| LER 1988-011-00 | |
| Download: ML20161A165 (3) | |
Text
B-4 4 Accident Sequence Precursor Program Event Analysis LER No: 250/88-011 Ri Event
Description:
Diesels unavailable after 16 h Date of Event: May 29, 1988 Plant: Turkey Point Unit 3 Summnary Turkey Point 3 was at 100% power when operators conducting a test of the emergency diesel generator fuel oil transfer pumps discovered that a diesel oil storage tank isolation valve was locked closed instead of locked open, as required. This could result in diesel failure after 16 h of operation. The event has been modeled as a potential LOOP of duration greater than 16 h. The conditional probability estimated for the event is 9.9 x 10-6. This estimate is believed conservative since the impact of core cooling during the previous 16 h was not taken into account. The relative significance of the event compared with other events at Turkey Point is shown below.
LER 250/881-011 RI IE-8 I1E-7 I E-6 I-5 1IE-4 I1E-3 IE-2 TrIip Trip w/I AFW L LO OP L EP Unavail 30hs Train Unavail AFW Unavail (360 hrs)
Event Description On May 31, 1988, at 1445, with both units at 100% power, operations personnel attempted to perform procedure OP 4304.4, Diesel Oil Transfer System Periodic Test of Pumps. Proper flow through the "A" emergency diesel generator fuel oil transfer pump could not be developed. An investigation determined that valve 70-003, a normally locked-open diesel oil storage tank isolation valve, was locked closed. The valve was opened and verified to be locked open. With valve 70-003 closed, fuel oil was available for approximately 16 h of DC operation.
B-4 5 An investigation conducted subsequent to the event determined that valve 70-003 had been isolated on May 29, 1988, at approximately 1700, by a chemistry technician while obtaining a sample from the EDO fuel storage tank. The technician erroneously believed that the valve was locked closed, and therefore opened the valve to allow a second technician to obtain a sample from the downstream sample valve (70-004). After the sample was obtained, the first technician closed and locked valve 70-003, believing that this was the original position.
Event-Related Plant Design Information Turkey Point Units 3 and 4 have two diesel generators that are shared between the two plants. EDG K4A provides backup power to the 3A and 4A 4160-ky buses, and EDG K4B provides backup power to the 3B and 4B 4160-ky buses. Buses 3A and 3B are the vital 4160 buses for Unit 3, and 4A and 4B are the vital 4160 buses for Unit 4.
The fuel oil transfer system at Turkey Point consists of a single fuel oil storage tank that normally contains a minimum of 40 000 gal of diesel fuel oil. One EDG day tank is provided for each of two ED~s.
Separate fuel oil transfer pumps are provided to transfer fuel oil from the EDO storage tank to the EDO day tanks. However, both fuel oil transfer pumps take suction from a common line containing valve 70-003. Each EDO day tank gravity feeds to an EDO skid mounted tank.
At the time of the event, the skid tank for the "A" EDG contained 210 gal of fuel oil and the day tank contained 3300 gal of fuel oil. The skid tank for the "B" EDO tank contained 205 gal of fuel oil and the day tank contained 3400 gal of fuel oil.
ASP Modeling Assumptions and Approach The event has been analyzed assuming the fuel oil supply to the D~s will not be recovered and that the D~s will fail at 16 h. At that point, a median diesel generator recovery time of 4 h was assumed.
The frequency of LOOP > 16 h at Turkey Point was developed based on information provided in NUREO-1032. This frequency is estimated to be 2.6 x 10-3 /yr. This results in a probability of blackout for the 46-h unavailability period of 2.0 x 10-5.
This value was used with the seal LOCA models used in the ASP Program to estimate the conditional probability for two sequences: blackout with subsequent seal LOCA and failure to recover AC power prior to core uncovery, and blackout without subsequent seal LOCA and failure to recover AC power prior to battery depletion. The analysis assumed a battery lifetime of 2 h (as specified in the FSAR) and the use of old 0-ring design RCP seals.
13-4 6 Analysis Results A conservative estimate of the conditional probability for this event is 9.9 . 10-6. The analysis is conservative in that it did not address increased grace time prior to core damage from the 16 h of core cooling prior to DG failure, nor did it address the potential for the plant to be on RHR at the time, which would have reduced the likelihood of seal failure. In addition, Turkey Point has nonsafety-grade black-start D~s that can be used to supply AC power.
The dominant sequence following the postulated blackout involves an RCP seal failure (p = 0.7) and failure to recover AC power prior to core uncovery (p = 0.68). Because of the approach used to model this event, neither an event tree with highlighted sequence nor a conditional prob-ability calculation sheet is provided.