05000271/LER-2014-001
| Vermont Yankee Nuclear Power Station | |
| Event date: | 09-29-2014 |
|---|---|
| Report date: | 11-21-2014 |
| Reporting criterion: | 10 CFR 50.73(a)(2)(i)(B), Prohibited by Technical Specifications 10 CFR 50.73(a)(2)(v)(D), Loss of Safety Function - Mitigate the Consequences of an Accident |
| 2712014001R00 - NRC Website | |
DESCRIPTION OF EVENT
On September 29, 2014, at approximately 1209, with reactor power operation at 96 percent, the 'A' train Emergency Diesel Generator (EDG) (EIIS=EK) did not complete the starting sequence during a scheduled Technical Specification monthly surveillance. When the under voltage relay auto test switch was moved to the test position, the air start solenoids energized but the engine did not rotate. Alarms were received in the Control Room indicating that the 'A' EDG had not started. The station had already entered a 7-day Limiting Condition for Operation (LCO) per Technical Specification 3.10.B.1 when the 'A' EDG was declared inoperable prior to performing the surveillance. Mechanics initially attempted to manually rotate (barring) the engine in the normal counter-clockwise direction and were unsuccessful, indicating mechanical or hydraulic lock. The engine was then successfully rotated in the clockwise direction, and at approximately 1722 an air roll of the `A' EDG was performed with a normal response.
On September 30, 2014, at approximately 0900, the 'B' train EDG was successfully started in order to demonstrate that the remaining EDG was operable within 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> as required by Technical Specifications.
On October 1, 2014, a cylinder air pressure test was performed on all twelve cylinders of the 'A' EDG and a leak was identified in the #6 cylinder. A boroscope inspection was performed on the #6 cylinder and identified a Jacket Coolant Water leak inside the combustion chamber at the location of the relief test cock valve adapter. During disassembly, the adapter was found at less than the expected torque, and visual inspection of the copper gasket for the adapter indicated that an apparent lack of gasket compression was the likely source of the leakage, estimated at 20 milliliters (mL) per minute. The adapter and copper gasket were replaced and properly torqued. A cylinder air pressure test was re- performed on the #6 cylinder and no further leakage was identified.
On October 2, 2014, post-maintenance testing and Technical Specification surveillance requirements were successfully completed. At approximately 0800, the `A' EDG was declared available for online risk.
At approximately 1506, the `A' EDG was declared Operable and the 7-day LCO was exited. The elapsed time from discovery of the failure until the EDG was returned to service was approximately 3 days.
There were no indications of a Jacket Coolant Water leak on the 'A' EDG prior to the event. No work had been performed on the `A' EDG since completion of the previous surveillance test run performed on August 25, 2014, that demonstrated its capability to meet its safety function.
No other systems, structures or components (SSCs) were inoperable at the start of the event and contributed to the event. No safety systems actuated or failed to actuate during the event.
The EDGs at Vermont Yankee are Fairbanks Morse model 3800TD8-1/8, each with a continuous rating of 2750 kilowatts (kW).
CAUSE OF EVENT
An investigation concluded that the direct cause of the event was leakage past the copper gasket for the relief test cock valve adapter in the #6 cylinder. The leak allowed jacket coolant water to fill the small space between the opposing pistons. In the as-found condition, as determined by the position of the air start distributor in relation to the cylinder position, the position of the opposing pistons in their compression stroke covered the exhaust ports and allowed the leak to fill the gap between the pistons with Jacket Coolant Water (incompressible fluid). This condition prevented the pistons from completing their stroke, and thus did not allow the engine to rotate or the starting process to progress.
Upon disassembly, the relief test cock valve adapter required considerably less torque than expected to remove. By visual inspection, the condition of the adapter copper gasket was not consistent with one that had been compressed to the vendor recommended torque value. Maintenance records for the 'A' EDG show that this copper gasket had been installed during the last 'A' EDG cylinder replacement in 1992.
Maintenance verifications had been performed every two years to check for indications of this type of leakage, and none were identified. The investigation was unable to determine a specific cause for why the copper gasket began leaking after it had provided adequate water tight sealing for 22 years of service.
The copper gasket is subject to normal cyclic fatigue and cyclic changes including thermal expansion and contraction and pressure cycles, and it is likely that these factors in combination with the as-found condition of the copper gasket resulted in the Jacket Coolant Water leaking into the cylinder.
The root cause was that, during installation in 1992, the adapter was not torqued to the vendor recommended value. With the limited amount of historical information available, the investigation was unable to entirely eliminate some possible causes such as a problem with the torqueing equipment or incomplete information provided in the work package instructions. The investigation concluded that the most likely possible cause was the individual installing the adapter not using accepted human performance practices to ensure the work instructions were completed in accordance with vendor recommendations.
Review of the maintenance records and work instructions did not reveal any specific underlying reason for this cause, since the work instructions as they should have been provided to maintenance personnel were adequate and were capable of being understood by the trained and qualified technicians who performed the work, which included both site and vendor personnel. Installation of the adapter was performed as part of a cylinder liner replacement under a Waiver of Compliance (see LER 92-017-00 and LER 92-017-01), therefore situational pressures were present.
ANALYSIS OF EVENT
The safety objective of the Station Auxiliary Power System, described in the Updated Final Safety Analysis Report, Section 8.4, is to provide a reliable power supply for the starting and operation of engineered safety feature loads during and following any design basis accident. Normal power is supplied to the buses through the unit auxiliary transformer which is supplied by the main station generator, and one of the two startup transformers. An alternate supply is provided via the startup transformers from the 345/115 kV transmission system. There are two standby EDGs, each connectable to a safety bus. Each EDG has sufficient capacity to supply each of the redundant emergency loads required under postulated design basis accident conditions. Backfeed through the main step-up transformer, the Vernon Hydroelectric Station and the Station Blackout Diesel Generator are available as reliable power sources that can be manually aligned to supply adequate capacity to support safe shut down of the plant. Analysis demonstrates that the plant can cope with a complete loss of normal and back up EDG power for up to two hours.
Based on available information, the hydraulic lock condition probably developed shortly after the `A' EDG was placed in a standby condition following completion of the previous monthly surveillance on August 25, 2014. During the time period that the `A' EDG was potentially inoperable, a monthly surveillance run of the redundant 'B' train EDG was performed which could have prevented fulfillment of a safety function to mitigate the consequences of an accident during two brief windows (approximately 15 minutes each) prior to and immediately following the surveillance run, but was otherwise available. With both EDGs inoperable, Technical Specifications allow 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> for the reactor to be placed in cold shutdown.
Since the 'A' EDG was potentially inoperable since completion of the previous monthly surveillance run, this event is reported in accordance with 10 CFR 50.73(a)(2)(i)(B) as an operation or condition prohibited by Technical Specifications. Since during this time period, a scheduled monthly surveillance was performed on the 'B' EDG as described above, this event is also reported in accordance with 10 CFR 50.73(a)(2)(v)(D) as an event or condition that could have prevented fulfillment of a safety function needed to mitigate the consequences of an accident.
As noted in the event description, the `A' EDG was able to be manually rotated to remove the hydraulic lock condition, so that it could be recovered within a few hours. Based on the recovery actions for the `A' EDG, and the availability of the redundant `B' train EDG, normal and alternate off-site power sources including the Vernon Hydroelectric Station, and the Station Blackout Diesel throughout the time period when the `A' EDG was potentially inoperable, the event did not pose a threat to public health and safety.
CORRECTIVE ACTIONS
The subject adapter and copper gasket were replaced, and a cylinder air test of the 'A' EDG was satisfactorily performed on the #6 cylinder on October 1, 2014. A cylinder air test of the 'B' EDG to confirm no similar condition existed on the redundant EDG was satisfactorily performed on October 14, 2014. A review of work order instructions based on vendor recommendations determined that these instructions are in compliance with the planning procedure and identified no nonconformities.
ADDITIONAL INFORMATION
A scheduled monthly surveillance run of the `A' EDG was satisfactorily completed on October 27, 2014.
The copper gasket and adapter that were replaced were sent to the manufacturer Fairbanks Morse for analysis. Although not expected, if this analysis reveals information that results in a significant change to information provided in this report, a supplement will be submitted.
PREVIOUS SIMILAR EVENTS
No similar event has occurred at Vermont Yankee within the past five years.