05000286/LER-2009-005
| Indian Point 3 | |
| Event date: | |
|---|---|
| Report date: | |
| Reporting criterion: | 10 CFR 50.73(a)(2)(i)(B), Prohibited by Technical Specifications 10 CFR 50.73(a)(2)(v), Loss of Safety Function |
| 2862009005R00 - NRC Website | |
Note: The Energy Industry Identification System Codes are identified within —the-brackets
DESCRIPTION OF EVENT
On June 9, 2009, tfie inoperability during past operation of a 480 volt undervoltage/deg'raded grid (UV/DGV) relay {62} was identified during resolution of questions by NRC inspectors for a Problem Identification and Resolution (PI&R) inspection. NRC review of Corrective Action Program (CAP) condition reports (CRs) identified a CR that recorded an Agastat time delay relay (62-1/3A) for 480 volt bus 3A, {EDI. _ Therelay..failed its As-FOund acceptance tolerance including its -Tethnical Specification-(TS)'required-value during performance of surveillance test 3-PT-M62A.(480 Volt Undervoltage/Degraded Grid Protection System Bus 2A and 3a Functional) .on November 8, 2007. CR-IP3-2007-04210 recorded this condition. A review of priorests identified testing the previous month also showed it failed these test-criteria ..'Test failure of this relay on October 11, 2007, was recorded in CR-IP3-2.007.703869t A September 13, 2007 test was identified where the relay was-"outside the-As-F6Und calibration acceptance criteria but within the TS limit.
TS 3.3.5 (LosS of Power Diesel Generator Start Instrumentation) surveillance requirement SR 3.3.5.2 specifies for degraded voltage (480 volt bus) relay (Non Safety'Injection).an.allowable value of equal to or greater than 414 volts with a time delay of equal to or less than 45 seconds. The relay for each test was calibrated to within tolerance within the TS allowed outage time. In accordance -- with reporting-guideIines the discrepancy:is assumed to occur at the time of discovery. unless, there is. firm. evidence, based on review of relevant information such as equipment history and the cause of the failure, to indicate the discrepancy existed previoUsly.
A corrective.actiohCAY-was;initiated-for the: November_ 8, -2007 test failure, that included the Oetbber '11: 2007 test failure, which design engineering - instrumentation and control (DE I&C) evaluated for past operability to determine if the relay- went -out of calibration prior to the time of discovery. The DE I&C engineer assigned to enter component. drift data and evaluate failures was not specificallyknowledgeabia, in component. failure. analysis -nor drift analysis. The Drift MOnitoring .PrOgrah (DMPrhad transferred from Programs and Component engineering (P&CE).- to I&C. DE-in-early-2006: P&CE discontinued entering data and overseeing the. programat that time but DEJ&C,didnot assume responsibility :Until January 2007 WIiich'reSuitedin test data:not completely entered until August 2008. The lapse in the DMP adMinistration was recorded in CR-IP3-2007- 3426. There was no training on the DMP software nor expectations on evaluating the drift data:being entered into the program- except for guidance provided in the DMP procedure .0-PCE-ADOi (Drift.Moniori4-ig:PrOgram): DE I&C used the DMP guidance which stated :that- for; ""more than two of the last five test failures, then the failure is assumed to have.:occurred prior to the time of discovery.
Since.at the timeit:was not believed iela.y. 62-1/3A had more than two test failures'in the last fiVe tests, the DE I&C"engineer concluded that the relay did not demonstrate unacceptable performance._ However, the DE I&C engineer failed to recognize additionai-prOcedural_gUidance-whiCh Stated; ."The review will .
concentrate not only on the components PaSthiStory,- but the magnitude in which the component - waS-foundbut-of - tolerance.'1_ The DE I&C engineer incorrectly concluded the relay had been performing - .satisfactorily and-failed to recognize the significant magnitude of the failure which was well outside. the manufactUrer!s,repeat accuracy (drift). The significant drift and repeat failures would-indicate the relay was inoperable during past operation and exceeded the TS allowed outage time. The DMP software does not document TS values therefore, the October 2007 test failure to meet TS SR was not recognized and the lack of drift analysis knowledge led to not recognizing the large shift in As-Found test data as a degraded condition.
NRC FORM 366AU.S. NUCLEAR REGULATORY COMMISSION (9-2007) FACILITY NAME (1) DOCKET (2) LER NUMBER (6) PAGE (3) The relay was replaced in November 2007. The event was recorded in the Indian Point Energy Center corrective action program (CAP) as CR-IP3-2009-02664. The degraded relay is a time delay relay (62-1/3A) {62} Series E7000, Model E7014PD004 manufactured by Agastat {A109}.
The extent of condition applies to all As-Found surveillance test failures. Each time a component in a surveillance test fails its As-Found calibration acceptance criteria, the component history, direction of failure, and magnitude of failure, needs to be evaluated. A sample portion of surveillance test data from the time period of pre 2007 to mid,2008, was reviewed for large shifts in drift. The review did not identify any other issues.
- CAUSE, O EVENT- .
- The apparent cause was personnel error due to inadequate knowledge of the drift monitoring program and component drift performance. The DE assigned to the DMP had no specialized training or knowledge in component drift and calibration analysis. The inadequate knowledge allowed the large drift that exceeded TS SR criteria in October 2007 to be treated as a normal As-Found failure instead of prompting a more rigorous evaluation.. Contributing causes (CC): CC1: TS values are not specified in surveillance tests providing an error trap allowing test reviewers to believe that if an As-Found failure is calibrated back within tolerance the component was functioning acceptably; CC2: Failure to initiate a CR on the As-Found failure for the surveillance test of relay 62-1/3A performed on September 13, 2007. This failure to record in the CAP allowed the second failure to be considered a single test outlier instead of a second consecutive As-Found failure.
CORRECTIVE ACTIONS
.� _ The:following corrective actions have been performed under. Entergy's Corrective Action Program to addresScause_and_prevent recurrence:
- I&C SuperVisOrS-were'coachedtb reinforce Management's expectation for initiating CRs for' as'-.found failures., -
- Expectations- and guidance, will be developed for evaluation and disposition of aS-foUnd surveillance-failures. Scheduled completion is August 31, 2009.
- The functionalfailure determination.for prior relay failures will be re evaluated: Scheddled- cOmpletionis'Aligust 31, 2009.
EVENT ANALYSIS' _ The event is. reportable under 10CFR50,73(a)(2)(i)(B). The licensee shall report any operation'or cOndition -whichwath.pz-ohibited by the plants Technical Specifications. Ori-November 8,"2007, the surveillance test of Agastat relay 62- 1/3A ..(480. volt bus 2A/3A) As-.Found did not meet the TS calibration acceptance criteria: TheNovember-test'faiIure was_the second test failure of the TS value.
- These two failures along. With a preVious- As-Found failure indicated the relay was a degraded component and exhibited abnormal drift. In accordance with reporting guidelines of NUREG-1022, the discrepancy is assumed to occur at the time of discovery unless there is firm evidence, based on review of relevant information ,such as equipment history.andth'catise of the failure, to indicate the discrepancy existed previously.. , The-test results provided evidence that the relay drifted outside its acceptance criteria and therefore was inoperable during past bperation:- * The condition of inoperability exceeded the TS allowed outage time.
The inoperable Agastat relay did not result in the loss of any safety function.
Agastat relay 62-1/3A is a time delay relay whose timer setting drifted outside its acceptance criteria which would have caused it to actuate later than required.
However, the relay would have still actuated upon demand. Relay actuation later than specified could result in the motors running on the 480 volt bus to trip on overcurrent if a degraded grid condition continued before the relay transferred bus loads to the emergency diesels. If the degraded voltage was accompanied by a safety injection (SI) signal, the 10 second relay timer would have transferred the bus loads to the emergency diesel generators (EDG). Engineering judgment concluded that the safeguards motors could have operated without damage. Therefore, there was no safety system functional failure reportable under 10CFR50.73(a)(2)(v).
PAST SIMILAR EVENTS
A review-was performed.of Licen-see Event Reports (LERs) for the past three years for any events reporting TS prohibited conditions due to relays out of calibration specification resulting in -inoperability. No applicable LERs were identified.
SAFETY SIGNIFICANCE
This event-had no significant - effect on the health and safety of the public.
There were no actual safety consequences for the event because there were no UV or DGV conditions that challenged-the bus or loads during past operation when the relay-may have been out of specification.
The DGV protection on each 480 volt safety bus consists of two DGV relays arranged in:a two-out-of-two- logic.- Functional actuation includes a time delay of 10 seconds if- a coincident safety injection (SI) signal indicates accident conditions_exist and& time.delay of 45 seconds-if no SI signal is generated (i.e., non-accident condition). The DGV relay reported in this LER was for DGV protection for the non-accident condition. When the two-out-of-two logic was made up, the _function would not have been performed within 45 seconds. The 10 second time was .operable sq-if a_two-out-of-two logic was made up coincident with an SI signal:. the function would actuate as designed. The drift of the DGV relay -timer aetting'_wouldhave caused it to actuate later than specified but it would have actuated on demand and- prbvide 480 volt bus load transfer to the onsite emergency AC poWei- source_AEDGs). The actuation of the DGV relay later than specified. could result in the - motors running on the 480 volt bus to trip on overcurrent if a degraded grid condition continued before the DGV relay actuated and transferred bus loads.to the.EDGs. Engineering judgment indicates this would not have occurred. In tlie.aafety evaluation for TS Amendment No. 54, the original proposed.time delay for DGV relays was less than or equal to 210 seconds for- all- conditiona-.Hrhe DGVfor.'n-on-accident conditions was not proposed for equipment. Protection -but was selected to allow sufficient time for the offsite power transformer:attomati0..tap-_.changer .to . attempt to restore offsite power voltage:thereby,PreVenting',disCOnnection fi-om the preferred power source. The 210 second'bGV:time delay. waa:Oonaidered unacceptable for accident conditions and was reviaed-tb reas -than-or'eqtal. to 10 seconds. The safety evaluation report
- noted the 210 second time'delaywas reasonable for non-accident conditions when minimum safety equipment is operating, when accepting the proposed revision to 45 sedonds, The magnitude of the-DGV time delay, although exceeding the 45 second :TS:liMit;- Was. well below'the210 seconds.