05000397/LER-2009-004
| Docket Number | |
| Event date: | 08-05-2009 |
|---|---|
| Report date: | 05-02-2011 |
| Reporting criterion: | 10 CFR 50.73(a)(2)(v), Loss of Safety Function |
| 3972009004R01 - NRC Website | |
Plant Condition The plant was in Mode 1 operating at 100% when this event occurred.
Event Description
On August 5, 2009, at approximately 0750 the plant experienced an automatic actuation of the Reactor Protection System [JC] due to a main turbine trip. The turbine trip occurred immediately following a main generator [EL] differential lockout due to an electrical fault in the 6.9 kV electrical distribution system [EA].
The electrical fault was determined to have been caused by a failure in the non-segregated electrical bus [NSBU] which generated enough smoke in the turbine building to require declaration of an unusual event at 0812 due to toxic gases in amounts that could affect the health of plant personnel or safe plant operation.
The plant response to the disturbance in the electrical system and subsequent turbine trip and scram (RPS actuation) was as expected with the following exceptions:
1)The DEH system [JI] unexpectedly transferred to manual pressure control mode with the bypass valves remaining in the full open position. Consequently, reactor pressure dropped from a post turbine trip maximum of approximately 1081 psig to an approximate minimum of 396 psig in just over 3.5 minutes.
Plant Operators terminated the reactor pressure decrease by manually closing the inboard Main Steam Isolation Valves.
2) After the scram, the reactor coolant temperature dropped 106 degrees over a period of approximately six minutes, exceeding the Technical Specification (TS) allowable cool down rate of 100 degrees in an hour.
3)The feedwater [SJ] pumps tripped on low suction pressure within a few seconds of the scram. Water from the feed system was re-established when reactor pressure decreased to less than the shutoff head of the condensate booster pumps approximately two minutes after the scram.
There was no inoperable equipment at the start of the event that contributed to the event. This LER is submitted pursuant to 50.73(a)(2)(iv)(A) as an event or condition that resulted in automatic actuation of the Reactor Protection System.
Causes The most probable cause of the electrical bus failure was a loosening or relaxation of bolted connections on the center phase flexible link(s) due to repeated thermal cycles over time. The looseness resulted in increased heating of the joint, which degraded the joint. It is postulated that heating occurred in the joint to the point that insulation degraded, resulting in the formation of a high energy arc fault which shorted between phase conductors. The loose connections were not detected prior to the event due to a failure to perform PMs for bus torque checks as well as a lack of adequate temperature monitoring of the non segregated buses. Because the bus was destroyed, little forensic evidence exists to support these conclusions.
26158 R5 Subsequently, it was discovered that the bus had been improperly uprated in 1994 to support a modification that added the adjustable speed drive (ASD) loads to the bus. The increased bus loading resulted in reduced margin between the bus loading and the actual bus rating, leading to increased bus link temperatures. In addition, it was discovered that the ASD modification resulted in the addition of non-linear loads to the bus. This modification was installed in 1996. The non-linear loads caused an increase in total harmonic distortion (THD), which also results in increased heating on the bus. The impact of this increase in THD was not addressed as part of the design change. In both instances, less than adequate engineering rigor was used to analyze the impact of a plant design change on the electrical bus. As such, this lack of engineering rigor has been identified as a contributing cause to the electrical bus failure.
The unexpected transfer of the DEH system to manual pressure control was due to a design error introduced during the DEH system modification that was installed in 2007. The failure threshold in the DEH software code for the main steam throttle inlet pressure transmitters was set too low to accommodate the pressure spike that occurred when the turbine tripped from full power.
Design changes that were made to the feedpump suction trip setpoints and the digital feedwater reactor level control system in R18 (2007) contributed to the feedpumps tripping on low suction pressure.
Following the scram, the feedpump turbines ramped up speed to respond to the RPV low level signal, which caused an unanticipated drop in suction pressure below the trip limit. The change that staggered the feedpump suction trip setpoints did not achieve the desired result of preventing both feedpumps from tripping off in this event.
Corrective Actions Taken or Planned The following actions have been taken or are planned to address the issues identified above, prevent recurrence, and address the extent of condition/cause:
1) The damage to the 6.9 kV bus has been repaired, and windows have been installed in the bus duct covers to allow for more direct thermography monitoring of the links.
2) An appropriate frequency for thermography checks will be established in conjunction with performing torque checks for the other non-segregated buses during the next refueling outage to prevent recurrence of this type of event. The content and controls for PMs will be strengthened to ensure proper completion of critical steps, and appropriate levels of review and approval are applied to any changes.
3) The DEH system software has been modified to increase the failure threshold setpoint for the main steam throttle inlet pressure signal and to ensure the system will remain in automatic pressure control upon failure high of the main steam throttle inlet pressure signal.
4) Because the TS allowed cool down rate of the reactor pressure vessel (RPV) was exceeded, a fatigue analysis was performed on the RPV and concluded that this event is bounded by transients evaluated in the original RPV fatigue analysis.
26158 R5 5) The other non-segregated buses were visually inspected and had insulators torque checked.
Additional infrared thermography windows were installed in the bus housing covers at all flexible and rigid link locations with the exception of the two buses that are only normally energized during startups up to 23% power, in which only some of the thermography windows were installed. The remainder of the windows for the two startup buses are scheduled for installation during the next refueling outage in 2011 6) The logic for the Digital Feedwater level control system has been modified to limit feedwater pump turbine speed in response to a reactor scram. A staggered time delay for feedwater pump low suction pressure trips has also been added to allow suction pressure to recover before a second pump trips.
7) The non-segregated electrical bus is being upgraded to replace the existing busbar with a larger busbar with welded bus connections to increase the bus ampacity rating and regain margin between the bus loading and bus rating.
Assessment of Safety Consequences
For this event, all Emergency Core Cooling Systems (ECCS) were available to perform their intended safety functions. Both off-site power circuits were available and all three of the emergency diesel generators [EB and EK] were operable and available. This event did not involve an event or condition that could have prevented the fulfillment of any safety function described in 10 CFR 50.73(a)(2)(v). Therefore, this event posed no threat to the health and safety of the public or plant personnel and was of low safety significance.
Similar Events No previous similar bus failures have been reported by Columbia. A review of the Corrective Action Program condition report database found one other occurrence of an event with similar characteristics. In May 2007 evidence of electrical connection heating was discovered during performance of a PM on one of the non-segregated buses (not the bus that failed in this event). The connection was reworked and re taped. Visual inspections conducted to address the extent of condition on the other non-segregated buses identified and repaired instances of potential overheating on one bus (not the bus that failed in this event) and identified no major concerns for the bus that failed in this event.
Enemy Industry Identification System (EllS) Information EllS codes are bracketed [ ] where applicable in the narrative.
26158 R5