05000364/LER-2005-001
| Joseph M. Farley Nuclear Plant - Unit 2 | |
| Event date: | 12-04-2005 |
|---|---|
| Report date: | 01-31-2006 |
| Reporting criterion: | 10 CFR 50.73(a)(2)(i)(B), Prohibited by Technical Specifications |
| 3642005001R00 - NRC Website | |
Westinghouse -- Pressurized Water Reactor Energy Industry Identification Codes are identified in the text as [XX]
Description of Event
On December 4, 2005, at 0358, with Unit 2 at 29 percent power, while shifting on-service trains of High Head Safety Injection (HHSI) during power ascension following a refueling outage, the 2A (A Train) HHSI Pump [BQ] was found to have been inoperable, due to gas accumulation in its suction line. Although the actual time of inoperability is not known, it is apparent that the pump became inoperable some time prior to Mode 3 entry on November 30, 2005.
The unit was being returned to service following a refueling outage. The Chemical and Volume Control System (CVCS) work, and all associated testing procedures, had been completed. On November 24, 2005, at 1521, the A Train HHSI Pump was secured, and the B Train HHSI Pump was started, in accordance with plant procedures. On November 25, 2005, at approximately 1900, the decision was made to cool down and de-pressurize the primary system to repair a leaking seal on a Reactor Coolant Pump (RCP). During the Reactor Coolant System (RCS) cooldown, Volume Control Tank (VCT) level was allowed to vary between 20% and 65%, resulting in pressure variations between 25 psig and 55 psig. The root cause investigation determined that this large pressure variation can cause gas to come out of solution and migrate to the idle HHSI pump suction header. On November 26, 2005, a second significant VCT pressure change occurred as a result of gas leakby through two normally closed hydrogen isolation valves. VCT pressure increased from 20 psig to 46 psig before the leakby was terminated, at which time the VCT was manually vented to reduce pressure to 25 psig. The rapid venting of the VCT back to the normal operating pressure resulted in additional gas coming out of solution in the HHSI pump suction header. After the RCP seal repairs were completed, the reactor was taken critical on December 1, 2005.
During the on-service train shifting on December 4, 2005, at 0358, the 2A (A Train) HHSI Pump was started, and the 2B HHSI Pump (swing pump, then aligned to B Train) was secured. At 0359, the control room operator received indication that charging flow and seal injection flow were both zero. This was shortly followed by oscillations in both flow indicators. The pump was run in this condition for approximately 37 seconds. At 0359, the 2B HEST Pump (swing pump) was re-started (and verified operating normally), and the 2A HHSI Pump was secured. Indications returned to expected values.
TS Limiting Condition for Operation (LCO) 3.5.2 Required Action Statement (RAS) for inoperable A Train Emergency Core Cooling System (ECCS) was entered. The 2B HHSI Pump was aligned to A Train (and verified operating normally), and TS LCO 3.5.2 was exited on December 4, 2005, at 0635. With the 2B HHSI Pump aligned to A Train, the 2A HHSI Pump suction line was then vented. Following satisfactory completion of surveillance testing, the 2A HHSI Pump was returned to service on December 4, 2005, at 1636.
As a result of this event, venting frequency was increased to every 48 hours5.555556e-4 days <br />0.0133 hours <br />7.936508e-5 weeks <br />1.8264e-5 months <br /> for all three HHSI pumps pending root cause investigation. No gas was observed during venting of any of the three HHSI pumps through December 21, 2005. Based on these results, venting frequency was reduced to every 7 days. However, a similar event occurred on December 25, 2005, when a significant VCT pressure transient occurred during demineralizer flushing operations. Gas was detected on December 27, 2005, in the 2A HMI pump suction line during increased frequency venting evolutions resulting in declaration of inoperability. This confirmed the cause and identified additional routine operational activities that can result in VCT pressure transients.
HHSI pump suction line venting was placed on increased frequency during the investigation period.
All safeguards equipment, except the 2A HHSI Pump, functioned as designed. TS Action requirements were met following this occurrence. The A Train HHSI Pump functioned normally when tested following venting of its suction line.
Cause of Event
Root cause investigation determined the cause of the event to be gas bubble formation in the A Train HHSI Pump suction line due to pressure transients on a system saturated with dissolved gas. These transients occurred due to insufficient procedure guidance to maintain relatively constant VCT pressure during VCT level variations.
Safety Assessment Considering component/system operational and venting activities at the time of the occurrence, only one train was affected; therefore, this event did not constitute a safety system functional failure. The B Train pump and swing pump suction lines, which are at a lower elevation than the A Train pump, were verified to be unaffected by this event. During the period in question, two operable HHSI pumps were available at all times. Prior to the December 4 event, the B Train pump was operable and the swing pump was operable and aligned to B Train. During the December 25 event, the swing pump was aligned to the A Train and would have started automatically in the event of an A Train pump trip. The swing pump could be manually started from the Main Control Board throughout both events.
Only one train of HHSI was affected, and a safety injection was not required; therefore, the safety and health of the public were unaffected by this event.
Additionally, the system alignment at the time of the occurrence differs significantly from the alignment assumed in response to an accident. Therefore, firm conclusions, regarding the expected performance of the 2A pump in the post-accident alignment, based solely on the observed response at the time of this occurrence is not prudent. The post-accident alignment transfers the pump suction source to a higher pressure source (Refueling Water Storage Tank (RWST)) and significantly reduces the discharge-side system resistance of the pump(s).
The 2A HHSI pump would have been expected to purge the gas from the suction side in a reasonably short period of time without any near-term mechanical damage and would have then been expected to return to near pre-occurrence hydraulic performance. This is supported by the indications observed during the December 4 event, where flow initially dropped to zero but then began to recover as evidenced by flow oscillations.
While the exact time required to purge the gas through the pump is not known, the greater the suction flow, the shorter the expected purging duration.
In order to assess the impact of a delay in ECCS flow delivery associated with purging of a gas void, Westinghouse evaluated limiting accidents based on the specific flow characteristics of the 2A pump and the most recent as-left flow balance data with the 2A pump operating. These evaluations concluded that even with a 3-minute delay in ECCS flow delivery, use of the realistic ECCS flows demonstrate that the events acceptance criteria continue to be met. Based on this assessment, it can be concluded that if pre-event charging pump performance can be obtained in less than 3 minutes there would be no negative impacts with regard to limiting accidents.
On this basis, it can be concluded that although not desired, some gas accumulation can be tolerated without impacting the pump's ability to perform its safety function. Although during the event on December 4 flow appeared to be recovering, SNC conservatively considered the event to be outside the bounds of acceptability.
Based on the smaller volume of gas detected, engineering judgment was used to conclude that the 2A charging pump was not rendered incapable of performing its intended function during the second event on December 25.
Corrective Action The A Train HHSI Pump was vented, surveillance testing performed satisfactorily, and the system returned to service.
Applicable Operations procedures have been revised to clearly define the normal VCT pressure control band to minimize gas formation. In addition, criteria to vent the HHSI pumps, if the normal control band is exceeded, has been added.
Operations Plant Staff have been trained on this event.
Additional Information
The Westinghouse Owners Group (WOG) has an approved project to assess the significance of gas accumulation on the mechanical and hydraulic performance of pumps similar to the HHSI pump design. As a participating member of the WOG, Farley Nuclear Plant will remain cognizant of the project and will consider, as appropriate, all conclusions and/or recommendations.
The following LER's have been submitted in the last two years due to TS violations.
Failure Troubleshooting Water Inoperable Inoperable