ML20135G971
| ML20135G971 | |
| Person / Time | |
|---|---|
| Site: | LaSalle  |
| Issue date: | 05/14/2020 |
| From: | Christopher Hunter NRC/RES/DRA/PRB |
| To: | |
| Littlejohn J (301) 415-0428 | |
| References | |
| LER 1992-012-00 | |
| Download: ML20135G971 (7) | |
Text
B-153 B.22 LER Number 374/92-012 Event
Description:
Reactor Trip With Degraded Reactor, Core Isolation Cooling Date of Event: August 27, 1992 Plant: LaSalle 2, B.22.1 Summary The reactor scrammed from 80% power because of a main turbine trip. The main turbine tripped due to a thrust bearing failure indication. The reactor core isolation cooling system (RCIC) auto-started, and the motor-driven feed pump (MDFP) was started in preparation for tripping the turbine-driven feed pumps (TDFPs). However, when the TDFPs failed to trip, the reactor water level rose, resulting in a trip of the MDFP and RCIC. In an attempt to prevent flooding of the steam lines, the outboard main steam isolation valves (MSIVs) were manually closed, resulting in a TDFP shutdown. Later, the operators experienced difficulty in starting RCIC for reactor pressure control. Water that had accumulated in the steam line passed through the pump turbine and into the exhaust header. Flashing of that water to steam prevented RCIC startup due to high exhaust pressure trip signals. The conditional probability of subsequent core damage estimated for the event is 6.1 x 10V The relative significance of the event compared to other postulated events at LaSalle 2 is shown in Fig. B.46.
LER 374/92-012 1E-7 1E-6 M5-5 IE4 lE&3 ME2 S
I U
ii + HPCS U
I precuuor cutoff -~
+ RcIC Fig. B.6. Relative event significance of LER 374/92-012 compared with other potential events at La Salle.
LER NO: 374/92-012
B-154 B.22.2 Event Description On August 27, 1992, while reactor power was being reduced to 80%, LaSalle 2 scrammed because of a spurious thrust-bearing wear detector turbine trip signal. The spurious signal was caused by a shift in the trip setpoint due to manufacturer error. Within seconds, RCIC auto-started on a spurious low reactor water level signal caused by pressure oscillations induced by closure of the turbine stop valves.
The MDFP was successfully started to control water level; however, the TDFPs then failed to trip because of oil contamination and blockage in the turbine oil system (both TDFPs failed to trip on high vessel level and after multiple attempts from the control room and locally at the pump). The increasing water level in the reactor eventually resulted in a trip of RCIC and the MDFP. The MISVs were manually closed 3 min into the event when the 73-in administrative limit was reached; to prevent flooding outboard of the MISVs. However, the RCIC steam line is inboard of the MSIVs and the transient water level rose to 130 in. which is 22 in. above the bottom of the main steam lines. Closure of the MSIVs resulted in a trip of the TDFPs and loss of the main condenser as a heat sink. The safety relief valves (SRVs) were then required for control of reactor pressure. Although the SRVs were used successfully for this function, corrosion-caused instrumentation failures prevented direct confirmation of closure of two SRVs.
Attempts were made to use RCIC for reactor pressure control. Two start-up attempts failed as a result of high-exhaust-pressure trips. The cause was water accumulation in the steam lines which passed into the exhaust header via the pump turbine. There, flashing of water to steam resulted in pressure peaks which triggered the RCIC trips. The RCIC steam line drains had operated as designed, but the time available for water drainage was insufficient. The third attempt to start the RCIC (approximately 5 min after the first trip) was successful.
B.22.3 Additional Event-Related Information LaSalle is equipped with high pressure core spray (HPCS) and RCIC, either of which can provide adequate reactor vessel makeup following a loss of feedwater (LOFW) or a loss of inventory from a stuck open relief valve. In addition, the MDFP can be used for reactor vessel makeup.
B.22.4 Modeling Assumptions The event was modeled as a LOFW with failed RCIC. Potential sequences associated with the event are described in Appendix A, section A.3.2, BWR Nonspecific Reactor Trip, and shown on the event tree included with this analysis documentation. The plant response observed during the event impacted the following branches on the event tree:
TRANSIENT (reactor trip occurs). The reactor tripped because of main turbine stop-valve closure.
Power conversion system provides core cooling. The MSIVs were manually closed during the event in an attempt to prevent flooding of the main steam lines. This resulted in unavailability of the PCS.
LER NO: 374/92-012
B-155 Feedwater provides reactor pressure vessel (RPV) makeup. The turbine-driven feedwater pumps shut down when the MSIVs were closed. The motor-driven feedwater pump tripped on high RPV water level. The motor-driven pump was assumed to be recoverable with a non-recovery probability of 0.12 (ASP non-recovery class R3, see Appendix A, sect. A. 1). This value was chosen because the tripped pump appeared recoverable in the required period from the control room, but, because of the main steam line flooding and the problems with the turbine-driven feedwater and RCIC pumps, recovery was considered to be non-routine and burdened.
RCIC provides reactor pressure vessel makeup. RCIC tripped twice on high exhaust pressure because of water accumulation in the steam lines. RCIC was. assumed to be recoverable with a non-recovery probability of 0.12 (ASP non-recovery class R3), for the same reasons as FW. This non-recovery probabilty for RCIC may be conservative, since the steam line drain valves operated as intended and the third RCIC startup attempt was successful.
The current ASP event trees for LaSalle do not model the potential use of RCIC to provide RPV makeup in the event of a single stuck-open SRV. The use of RCIC for this purpose was included in the NUREG-1150 PRAs and utility-sponsored IPEs. To address this, the conditional probabilities for applicable sequences (sequences 25, 26 and 28) were reduced by the probability of failing to successfully use RCIC for this purpose. This is the probability that either RCIC fails, two or more SRVs fail to close given one or more fail to close, or long-term core cooling fails given RCIC is successful and only one SRV fails open. Since long-term core cooling is reliable, this probability can be approximated by p(RCIC) + p(2 or more valves fail open I 1 or more valves fail open).
The failure probability for RCIC during this event was estimated above as 0.12. A value of 0.027 was estimated for p(2 or more valves fail open I I or more valves fail open), based on an estimated probability for two or more SRVs stuck open of 0.0015 (see NUREG/CR-4550, Vol. 1, Rev. 1, Analysis of Core Damage Frequency: Internal events Methodology, january 1990, p. 6-10) and an estimated probability of one or more SRVs stuck open of 0.056 (developed as described in Appendix A, sect. A.4).
The probability multiplier used to adjust sequences 25, 26 and 28 to account for the potential use of RCIC to mitigate the effects of a single stuck-open SRV is therefore 0.12 [p(RCIC)] + 0.027 [p(2 or more SRVs fail open I 1 or more SRVs fail open)] = 0.15. The conditional probability for sequence 28 (the only dominant sequence of the three sequences - 25, 26, and 28) was manually revised from 5.8 x 10' to 8.7 x 10-l to reflect this. This reduces the core damage probability estimated for the event from 6.6 x 10-1 indicated on the calculational sheets to 6.1 X 10-1.
B.22.5 Analysis Results The estimated conditional core damage probability associated with the event is 6.1 X 10-. This probability was calculated by reducing the core damage probability shown in the calculations (6.6 X 10-')
by the change. in sequence 28 (from 5.8 x 10-7 to 8.7 x 10-1, a factor of 0. 15) as discussed in the last paragraph in the modeling assumptions section. This has been reduced from the value shown on the calculational sheets to reflect the potential use of RCIC to mitigate a single stuck open relief valve, as LER NO: 374/92-012
B-156 discussed in the modeling assumptions section. The dominant core-damage sequence, highlighted on the event tree in Fig. B.47, involves an effective LOFW with successful reactor vessel makeup and failure to remove decay heat in the long term. Note that failure of RCIC does not contribute to the dominant sequences associated with the event.
This analysis addressed the potential loss of core cooling caused by failures of systems associated with transient mitigation. If the MISVs had not been closed, failure of the main steam line could have resulted. The potential for core damage from this sequence was not addressed in this analysis due to the difficulty in estimating the required steam line failure probabilities.
Additional information concerning this event is included in Augmented Inspection Team report 50-374/92020(DRS).
LER NO: 374/92-012
B-157 PM I"INCI=
I JVCj INg& I I ISM D NOSD OK 120M Mc (w
10 M(
OK 14 0D SC Is c 15 Sc 17 0D Sc 200C 100 20K Sc 24 C0 250C 260C L7 0D 290C OK Sc 31 0o OK wc
=E E =20 Sc Sc 330o Sc Sc Ca0 Sc 376W A C0
_______________________________________________________" AMW Fig. B.47. Dominant core damage sequences for LER 374/92-012 LER NO: 374/92-012
B-158 LER NO: 374/92-012
B-159 LER NO: 374/92-012