05000397/LER-2007-005
| Columbia Generating Station | |
| Event date: | 12-10-2007 |
|---|---|
| Report date: | 02-07-2008 |
| Reporting criterion: | 10 CFR 50.73(a)(2)(i)(B), Prohibited by Technical Specifications |
| 3972007005R00 - NRC Website | |
Plant Condition The plant was operating in Mode 1 at 100 percent power, with the remaining DG and offsite power systems operable at the time these conditions were discovered. There were no structures, systems, or components that were both a) inoperable at the time of the events, and b) contributed to the events.
Event Description
On June 5, 2005 the Division 3 Emergency Diesel Generator [DG] was started for monthly operability surveillance testing under procedure OSP-ELEC-M703. This procedure starts the DG at idle speed for a warm up, then increases speed and places the electronic governor in control via taking the unit mode selector switch from MAINTENANCE to AUTO. When the electronic governor [65] is placed in service, the exciter field flashes and generator terminal voltage is developed. When the DG was started there was no indication of voltage or frequency after the unit mode selector switch was taken to AUTO at either the remote or local control panels. Operations declared the DG inoperable due to the lack of indications of voltage and frequency. Investigation revealed that the fuses [FU] on the primary side of the metering and relaying potential transformers [XPT] had cleared. The potential transformers provide power to the electronic governor as well as the local and remote indications for voltage and frequency of the DG. The clearing of the fuses at that time was determined to be age related. The fuses were replaced, and the DG was declared operable on June 7, 2005 following satisfactory performance of the required surveillances.
On November 8, 2007 the Division 3 DG was started for monthly operability surveillance testing under procedure OSP-ELEC-M703. Similarly to the incident discussed above, Operations declared the DG inoperable due to lack of indications for voltage and frequency. Engineering performed a walk down of the DG and control panels and observed normal exciter field current and field voltage which indicated the field had flashed and the terminal voltage had developed. DG speed was approximately 925 RPM which indicated the speed was being controlled by the hydraulic governor vice the expected 900 RPM when controlled by the electronic governor.
Troubleshooting revealed that the fuses on the primary side of the metering and relaying potential transformers had cleared. An investigation comparing the two fuse failure incidents resulted in a discovery on December 10, 2007 that the operating procedures that were performed prior to discovery of the two cleared fuse events shut down the Division 3 DG by taking the engine control switch to STOP while the unit mode selector switch was still in AUTO. Since both of these events occurred during shut down of the DG, no indication was available to alert Operations personnel of the cleared fuses condition until the next attempted performance of the monthly surveillance.
The May 3, 2005 fuse clearing incident was caused by performance of procedure TSP-DG-E501, "Simultaneous Start of All Three Diesel Generators." This procedure is used to satisfy SR 3.8.1.20 which is performed once every ten years. This procedure placed the engine control switch to STOP while the unit mode selector switch was still in the AUTO position to shut down the DG.
26158 R3 control switch to STOP without first placing the unit mode selector switch to MAINTENANCE. On October 19, 2007, OSP-ELEC-C703 was run using the revised shut down method for the first time.
This procedure is performed when there is a potential common cause failure that requires a start/run of the other diesels.
Review of the Division 3 DG control circuitry indicated that by placing the engine control switch to STOP without first placing the unit mode selector switch to MAINTENANCE shuts the DG down without collapsing the excitation field. In this sequence, the voltage regulator will attempt to maintain voltage as the engine speed and frequency drop. It was concluded that this results in an over excitation condition and excessive volts per hertz operation causing the fuses to clear.
It was concluded that from the performance of TSP-DG-E501 on May 3, 2005 and OSP-ELEC-C703, on October 19, 2007, the Division 3 DG fuses were cleared and the Division 3 DG was inoperable until the fuses were replaced and the DG restored to service on June 7, 2005 and November 10, 2007 respectively. The time period that the fuses were cleared on the Division 3 DG exceeded the Technical Specifications allowed completion time for the required actions of LCO 3.8.1 Condition B for both incidents. In addition during the 2007 incident, the Division 2 DG was inoperable for maintenance for just over seven of the 22 days that Division 3 DG was inoperable with the cleared fuses. Seven days with two DGs inoperable exceeds the Technical Specifications allowed completion time for the required actions of LCO 3.8.1 Condition E.
During the time the Division 3 DG fuses were cleared, the electronic governor would have been out of service as well as the local and remote indications of voltage and frequency for the DG. Because of the lack of indications and reliance on the hydraulic governor, Operations could not have satisfactorily completed all surveillance requirements rendering the DG inoperable.
Immediate Corrective Action In both incidents, the fuses were replaced and the diesel operability surveillance was satisfactorily completed prior to declaring the DG operable. OSP-ELEC-C703 was deactivated to preclude use of this procedure pending revision to correct the inadequate shut down method. TSP-DG-E501 was not deactivated; however it is not scheduled to be performed until 2015 and as discussed below will be revised prior to its next use.
26158 R3 Columbia Generating Station
Cause
A combination of change and barrier analyses was used to evaluate these events. The root cause team determined that the procedure revision process was adequate and that the actions of the licensee personnel were logical given their level of knowledge.
The root cause for this event was caused by a lack of knowledge of the Division 3 DG shut down logic by Operations and Engineering personnel. This lack of knowledge led to the direct cause of introducing inadequate procedural guidance that created an over excitation condition which cleared the fuses.
Further Corrective Action The procedures described above will be revised to ensure that the correct method for shut down of the Division 3 DG is utilized. The requisite lesson plans for Operations and Engineering training will be updated and training provided to ensure the lessons learned from this event address the knowledge shortfall.
Assessment of Safety Consequences
The Division 3 DG is used to supply power to the High Pressure Core Spray (HPCS) [BG] system in the absence of the normal/startup power sources. During the times the Division 3 DG fuses were cleared, the electronic governor would have been out of service as well as the local and remote indications of voltage and frequency for the DG. If an actual load demand had occurred during the time frames the fuses were cleared due to the loss of normal/startup power sources, Energy Northwest has concluded that there is reasonable assurance that the hydraulic governor would have controlled the DG with the High Pressure Core Spray system loads to meet safety function requirements.
For a demand start of the Division 3 DG that involves HPCS system load acceptance, the Division 3 DG would pick up the load upon DG output breaker closure. The 4.16 kV critical bus with the Division 3 DG power supply and HPCS loads running would reach a steady state frequency according to the Division 3 DG output frequency versus load droop control characteristic set into the hydraulic governor actuator. According to Original Equipment Manufacturer guidance, the hydraulic governor speed droop control knob on the front of the governor actuator is set for 3% droop when the electronic governor is controlling the unit. This means at full load on the Division 3 DG with no electronic speed control, the critical bus frequency is expected to be in the range of 61.2 to 61.0 Hz (where 61.0 Hz would represent a minimum recommended separation between the electronic and mechanical governor operation to avoid interference). This frequency would meet Technical Specification requirements for loaded operation with a concurrent loss of offsite power. Hence, the Division 3 DG and associated High Pressure Core Spray system would have fulfilled their safety functions during these conditions.
26158 R3 Columbia Generating Station In addition, the normal/startup power sources remained available to the affected bus during the entire time frames that the Division 3 DG was inoperable.
Similar Events No other events involving clearing of these fuses occurred in the previous ten years.
EIIS information denoted as [XX].
26158 R3