05000219/FIN-2018001-02
From kanterella
Jump to navigation
Jump to search
Finding | |
|---|---|
| Title | Enforcement Action (EA)-18-007: No. 2 Emergency Diesel Generator Ring Lug Failure |
| Description | On October 9, 2017, during a routine surveillance load test, the No. 2 emergency diesel generator failed approximately 5 minutes into the run due to a broken ring lug on a current transformer. Laboratory analysis of the broken ring lug determined that the ring lug failed due to fatigue cracking that was initiated due to stresses caused by bending and twisting of the electrical lug. Exelon last conducted a load surveillance on the No. 2 emergency diesel generator on September 25, 2017. Corrective Actions: Corrective actions included replacement on the broken ring lug on the No. 2 emergency diesel generator, extent of condition inspections on the No. 1 and No. 2 emergency diesel generators for additional bent or twisted ring lug connectors, and revision to the electrical ring lug installation and emergency diesel generator procedures to include inspection for bent or twisted ring lugs. Corrective Action Reference(s): Issue report 4060815 Enforcement:Violation: Oyster Creek Technical Specification 3.7.C.2.b states, in part, that if one diesel generator becomes inoperable during power operation, the reactor may remain in operation for a period not to exceed 7 days. Contrary to the above, on October 9, 2017, it was recognized that one diesel generator was inoperable for greater than the technical specification allowed outage time of 7 days, and Oyster Creek continued power operation. Specifically, on October 9, 2017, No. 2 emergency diesel generator failed to run during a routine surveillance test due to a broken ring lug on a current transformer, which resulted in a total inoperability time of 6.5 months.Severity/Significance: For violations warranting enforcement discretion, Inspection Manual Chapter 0612 does not require a detailed risk evaluation, however, safety significance characterization is appropriate. A Region I Senior Reactor Analyst (SRA) performed a best estimate analysis of the safety significance using the Oyster Creek Standardized Plant Analysis Risk (SPAR) model, Version 8.50 and Systems Analysis Programs for Hands-On Integrated Reliability Evaluations (SAPHIRE). The evaluation estimated the total (internal and external events risk) increase in core damage frequency (CDF) to be in the mid to high E-6/yr range, or a low to moderate safety significance. The SRA evaluated the internal events risk contribution due to the inoperability of the No. 2 emergency diesel generator for an approximate 6.5 month exposure time. The exposure time relative to when the No. 2 emergency diesel generator was no longer capable of meeting its 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> mission time is uncertain due to the effect of vibration induced fatigue, and therefore the method prescribed within the RASP handbook guidance was used.
9 The analyst used the guidance in Section 2.5 of the Handbook, Revision 2.0, to estimate the exposure time of 6.5 months based on the cumulative 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> summation of the No. 2 emergency diesel generator surveillance test proven run time. This approach is appropriate for periodically operated components that degrade during operation (i.e. vibration induced fatigue only occurs while the emergency diesel generator is in-service/operating). Given this approach, the dominant internal events, loss of offsite power were evaluated for the estimated internal risk increase. This contribution was estimated at 2E-6/yr increase in CDF. The dominant sequences involved loss of offsite power events with a concurrent failure of the No. 1 emergency diesel generator, failure of the combustion turbines, and failure to recover offsite power or recover an emergency diesel generator prior to core damage.The SRA performed various modeling changes after a review of revised calculations for DC battery life:Analysis noted that Oyster Creek Generating Station recirculation pump seals are similar in design to those tested in reports generated for Nine Mile Point Unit 1 with the use of CAN2A seals. Therefore, the failure probability of the seals in the station blackout sequence wasadjusted from 0.1 to 5E-2 similar to Nine Mile Point Unit 1 SPAR model 8.50.The failure to load shed action (DCP-XHE-XM-LSHED) in the model was calculated using the SPAR-H method and revised to 1.2E-2 versus being assumed to always fail (TRUE).Failure probabilities for 1, 2, or 3 stuck open electromatic relief valves were revised to be consistent with the previous model version 8.22 because of the isolation condenser design at Oyster Creek Generating Station which limits cycling and significantly reduces the probability of a failed open electromatic relief valve due to isolation condensers controlling pressure.The depressurization function using electromatic relief valves, if required, was calculated through SPAR-H to be 1E-2 for sequences where total seal failure is assumed (DEPSEALFAIL) (conservatively assumed limited time available).The diesel driven firewater pumps are both available and were set to calculated fault tree failure probabilities instead of always failed in the previous model. These are 2,000 gallons per minute pumps with a large supply of water and relatively simple operator actions to inject to the reactor pressure vessel. The firewater was assumed to fail at 0.1 when a total recirculation seal failure occurs due to assumed time constraints.The offsite power and the emergency diesel generator required recovery time events were increased to 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> for events where DC load shedding was successful, without seal failures and isolation condenser success along with diesel driven firewater success.The SRA noted the No. 2 emergency diesel generator was recoverable. In fact, the diagnosis of the failed condition was performed in a nominal 8-10 hours from the failure. Therefore, a probability of failure to recover event for the conditional case was developed. The SRA used SPAR-H as simple guidance, which conservatively supported a reasonable assumption of a 0.10 conditional probability of failure to recover the emergency diesel generator within 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. The base case utilized a calculation within SPAR of 0.33 failure to recover probability for 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> sequences. To estimate the external risk contribution, the SRA identified that the most significant external risk contribution was from fire events. Seismic, external flooding, and high wind events were not significant contributors for the issue. From discussions with Oyster Creek Fire probabilistic risk analysts and a review of this failure condition, the increase in CDF due to the failed No. 2 emergency diesel generator for the assumed 6.5 month exposure time was estimated at 4.5E-6/yr [(8.5E-5/yr-4.5E-5/yr) x (6.5/12 months) x 0.2].The DC safety-related battery life would be at least a nominal 14 hours1.62037e-4 days <br />0.00389 hours <br />2.314815e-5 weeks <br />5.327e-6 months <br /> and longer if DC bus stripping occurred, this allows for extended isolation condenser or electromatic relief valve function, with injection from diesel driven firewater. Given the time considerations and characteristics of the failure, an assumed recovery at a failure probability of 0.2 (slightly higher than internal due to less time) was applied for the No. 2 emergency diesel generator, which was a best estimate determined through SPAR-H insights. The dominant fire sequence was a fire affecting the A and B 4kV switchgear rooms, where combustion turbine support would be lost, with failure of the No. 1 emergency diesel generator breaker to close, and failure of locally operating the isolation condenser due to eventual loss of power. The SRA noted that FLEX credit was not quantified and would result in a lower risk estimation likely in the low E-6/yr range. Combining internal and external risk contributions, the total increase in CDF was 6.5E-6/yr, or low to moderate safety significance. The SRA determined that Exelon uses a Large Early Release Frequency (LERF) factor value of 8E-2. This value takes into consideration operator action for those relevant high pressure vessel breach scenarios (fuel-coolant interaction, liner-melt-through, and direct containment heating). This also credits procedure strategies where other mitigating actions are taken such as flooding the drywell. The SRA review of the dominant sequences and time to core damage affirmed that LERF did not increase the risk over that determined from the increase in CDF.Basis for Discretion: The inspectors determined that the ring lug failure was not within Exelons ability to foresee and prevent. As a result, no performance deficiency was identified. The inspectors assessment considered:1. Exelons review of emergency diesel maintenance performed in 2015 checked allconnections of the current transformer for tightness. The inspectors did not identify any gaps or deficiencies in the 2015 inspections. Inspectors also reviewed completed biennial inspections of the connection dating back to 1991 and did not identify any gaps.2. At the time of the failure, the current transformer connections did not have a time directed replacement frequency recommended by the Emergency Diesel Generator Owners Group. The inspectors did not identify any additional vendor or industry recommendations specific to the failed component or considerations specific to the failed component that existed prior to the failure.3. Industry operating experience information available to Exelon did not identify the potential for the fatigue cracking of the bent wire ring lug that was experienced.4. The bent ring lug failure was not the result of a failure on the part of Exelon staff; no standards existed on bending of the lug during installation and is considered skill of the craft.The NRC determined that it was not reasonable for Exelon to have been able to foresee and prevent this violation of NRC requirements, and as such, no performance deficiency existed. Therefore, the NRC has decided to exercise enforcement discretion in accordance with Sections 2.2.4 and 3.10 of the NRC Enforcement Policy and refrain from issuing enforcement action for the violation of technical specifications (EA-18-007). Further, because Exelons actions did not contribute to this violation, it will not be considered in the assessment process or the NRC Action Matrix. Exelons equipment corrective action program evaluation report (ECAPE) determined that the ring lug failed on the No. 2 emergency diesel generator as a result of fatigue cracking, which was initiated due to excessive stress caused by bending and twisting of the ring lug beyond limits specified in industry guidelines. The inspectors noted that the ECAPE did not provide supporting information regarding how the ring lug was bent and twisted beyond industry guidelines. Specifically, industry guidance states that ring lugs can be bent up to 90 degrees. The broken ring lug found in the No. 2 emergency diesel generator was bent at approximately 45-55 degrees per the ECAPE, which was within industry guidelines. Additionally, the ECAPE did not include specific guidance on twisting allowances for ring lugs. Exelon documented the inspectors observation in Issue Report 4089829. As a result of the inspectors observation, Exelon revised the ECAPE to say the ring lug failed on the No. 2 emergency diesel generator as a result of fatigue cracking, which was initiated due to excessive stress caused by bending and twisting of the ring lug. |
| Site: | Oyster Creek |
|---|---|
| Report | IR 05000219/2018001 Section 4OA2 |
| Date counted | Mar 31, 2018 (2018Q1) |
| Type: | Violation: Severity level Enforcement Discretion |
| cornerstone | Mitigating Systems |
| Identified by: | NRC identified |
| Inspection Procedure: | IP 71152 |
| Inspectors (proximate) | A Patel E Andrews F Arner J Furia J Kulp J Schoppy D Pelton |
| Violation of: | Technical Specification - Procedures |
| INPO aspect | |
| ' | |
Finding - Oyster Creek - IR 05000219/2018001 | |||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Finding List (Oyster Creek) @ 2018Q1
Self-Identified List (Oyster Creek)
| |||||||||||||||||||||||||||