05000458/LER-2014-006
| River Bend Station - Unit I | |
| Event date: | 12-25-2014 |
|---|---|
| Report date: | 02-19-2015 |
| Reporting criterion: | 10 CFR 50.73(a)(2)(iv)(A), System Actuation |
| 4582014006R00 - NRC Website | |
REPORTED CONDITION
On December 25, 2014. at 0836 CST, a reactor scram occurred while the plant was operating at approximately 85 percent power. This event resulted from the loss of power on the Division 2 reactor protection system (RPS) (**JD**) bus, in conjunction with a pre- existing half-scram on Division 1. The loss of Division 2 RPS power also resulted in a Division 2 containment isolation signal.
Approximately four minutes after the scram, reactor water level increased to the Level 8 setpoint, causing the running main feedwater pump (**5J**) to trip. As reactor water level decreased back through the normal operating range, operators attempted to re-start main feedwater pump "C," but its supply breaker failed to close. Main feedwater pump "A" was subsequently returned to service. As reactor water level decreased to the point at which the startup feedwater regulating valve (FRV) should have opened to establish automatic control, the valve failed to open. Attempts to open it with a manual input signal were unsuccessful, and the "C" main FRV was put back into service. By that time, reactor water level had decreased slightly below the Level 3 RPS actuation setpoint, resulting in a second scram signal.
This event is being reported in accordance with 10 CFR 50.73(a)(2)(iv)(A) as an automatic actuation of the RPS system and the primary containment isolation logic (**JM**).
INVESTIGATION and IMMEDIATE ACTIONS The Division 1 half-scram had been inserted two days prior to the event, in compliance with Technical Specifications, following the failure of an instrumentation channel on the no. 2 main turbine control valve. Teams were formed to investigate the separate significant aspects of the event, as follows:
Loss of Division 2 RPS Bus Power was lost when the output breaker on the RPS motor-generator (MG) in the Division 2 subsystem tripped. The mostly likely cause of the output breaker trip was an intermittent failure of the MG field flash card due to a degraded capacitor. The capacitor was replaced, and the MG was tested and returned to a standby condition as a backup power supply. The alternate power supply will remain in service carrying the bus until completion of a modification to eliminate the field flash card as a potential source of recurrence of this problem.
Misoperation of the Startup FRV The operation of the startup FRV was investigated to determine why it was unresponsive to either the automatic controller or the operator's manual input. Maintenance technicians discovered a failed circuit card in the "manual" side of the valve controller. The "automatic" function of the controller had operated correctly in the post-scram environment once reactor water level had returned to normal, and this was confirmed again during the troubleshooting.
Additionally, Engineering personnel determined that the performance of the valve was consistent with its design criteria. The valve and its control logic are designed for flow control, and not simply for isolation. When the valve receives a gradual "open" signal from the closed position, there is no specific time response requirement. This design feature can allow a delayed response that, in the scram recovery scenario, may be too slow to arrest a significant downward trend in reactor water level.
The procedural guidance to isolate the main FRVs and use the startup FRV in a scram recovery was developed in 2010 to mitigate the effects of any seat leakage through the main FRVs. At that time, Operations developed the procedure change and validated its feasibility in the training simulator. However, no cross-discipline review by Engineering was requested. The investigation of December 25 event found that the simulator model does not accurately reflect the actual operating characteristics of the startup FRV, in that the simulated performance of the valve is more responsive than the actual valve. An Engineering review could have brought the actual operating characteristics to light. This factor, in conjunction with the simulator performance, reinforced the concept that the startup FRV was capable of controlling reactor water level in the post-scram transient condition.
Malfunction of Main Feedwater Pump "C" The failure of main feedwater pump "C" was found to have been was caused by an "over-racked" condition of its supply breaker (that is, the breaker racking mechanism had slightly over-travelled the last time the breaker was returned to service). This caused the limit switches that detect the position of the breaker mechanism within the cabinet to give the control logic circuit a false indication that the breaker was not connected to the bus. Interim instructions have been implemented to have electricians verify the condition of all similar breakers each time they are racked in.
CORRECTIVE ACTION TO PREVENT RECURRENCE
The following corrective actions are planned, and will be tracked and documented in the corrective action program:
Loss of Division 2 RPS Bus A design modification will be installed to eliminate the potential for a failure in the field flash card to affect the operation of the MG set once it is loaded.
Misoperation of the Startup FRV Operating procedures have been revised to direct that, in the scram recovery environment, one of the main FRVs will remain in service. The startup FRV is not to be used in that condition.
The main FRVs are on the schedule of work items for the upcoming refueling outage. Valve seat leakage is one of the deficiencies to be addressed.
Malfunction of "C" Main Feedwater Pump Guidance on racking the main feedwater pump breakers (and those of similar design) will be incorporated into the pertinent operating procedures.
Additionally, the deficiency that necessitated the insertion of the Division 1 half-scam on December 23 was corrected prior to plant startup.
PREVIOUS OCCURRENCE EVALUATION
River Bend Station previously reported an event that occurred on December 6, 2014, in which the Division 2 RPS bus de-energized unexpectedly. That condition resulted in a Division 2 containment isolation and a half-scram. The investigation did not conclusively determine the cause of the trip of the Division 2 RPS motor-generator (MG) output breaker, which initiated the event. Repairs were made to the MG control circuits to address the most likely causes of the trip, and the MG was load tested for approximately 30 hours prior to being placed back into service. . It was postulated that another intermittent failure was occurring in the field flash circuit card, and a design modification is being developed to mitigate that failure, but that remains a future activity. As described above, the source of the MG output breaker trip on December 25 has not been conclusively identified, so it is not certain whether the corrective actions for the December 6 event should have precluded the December 25 event.
SAFETY SIGNIFICANCE
No plant parameters requiring the actuation of any standby diesel generators or the emergency core cooling systems were exceeded.
No reactor safety-relief valves actuated. All reactor control rods properly inserted at the scram signal. This event was, thus, of minimal significance to the health and safety of the public.
(NOTE: Energy Industry Component Identification codes are annotated as (**XX**).)