05000443/LER-2014-002
| Seabrook Station | |
| Event date: | 04-06-2014 |
|---|---|
| Report date: | 08-07-2014 |
| Reporting criterion: | 10 CFR 50.73(a)(2)(i)(B), Prohibited by Technical Specifications 10 CFR 50.73(a)(2)(vii)(A), Common Cause Inoperability 10 CFR 50.73(a)(2)(v)(A), Loss of Safety Function - Shutdown the Reactor |
| 4432014002R01 - NRC Website | |
Description of Event
On April 6, 2014, while the plant was in refueling outage sixteen (OR16), it was determined during surveillance testing that three of four reactor coolant pump (RCP) [AB, P] undervoltage (UV) reactor trip channels exceeded the Technical Specification (TS) channel response time acceptance criteria of 1.5 seconds for the RCP UV reactor trip function. The channel response time was exceeded because the RCP UV time delay relays (62V) [AB 27 ] exceeded their maximum allowable response time. The condition for the RCP UV time delay relays exceeding tolerance was experienced on all four channels; however, only three of the four channels did not meet their TS required response times. Since this condition involved multiple similar components, there is evidence indicating that this condition may have arisen over time and three channels of RCP UV were inoperable. This resulted in the plant operating in a condition prohibited by the TS from September 20, 2012 (last calibration date for 1-RC-P-1-C relay) to April 1, 2014.
Cause of Event
A root cause analysis was conducted for this event. The root cause of the event was determined as:
Revisions of the Design Control Manual in 1991 and earlier did not require a failure modes and effects analysis as part of the design change packages that installed the E7022PA relays. The RCP UV time delay setting was not identified as a critical attribute and the manufacturer's recommendations for applications requiring very precise time delay settings were overlooked and not incorporated into station procedures.
The Root Cause team identified that the RCP undervoltage channel calibration procedures do not contain adequate guidance to assure performance of the relays. Specifically, the procedures do not follow the manufacturer's recommendations for warm-up and periodic monitoring of response time.
The calibration procedures were based on the 7022PJ relay datasheet, which did not provide any guidance for very precise time delay settings. However, when 90MMOD508 installed an E7022PA relay in one channel, the E7022PA relay data sheet included recommendations for very precise delay application. This datasheet was included in the 1990 design change, but was not specifically recognized in the design change as a necessary calibration methodology change. Nor was this identified by Maintenance design reviewers as a required change to the existing procedure.
The recommendations for very precise time delay settings should have been identified and considered during the review of the 1990 design change. Although the datasheet was specific to the E7022PA relays, the time delay setting should have been identified as a critical attribute that required precise control, resulting in discussion with the manufacturer regarding this same application — very precise delay setting — on the 7022PJ relays.
The contributing cause for this event is:
Monitoring and trending of the RCP UV 62V Relays failed to elevate the significance of the condition of the time delay relays not meeting their acceptance criteria. This was demonstrated by failure to communicate results to station leaders, determine cause, and initiate corrective actions.
Analysis of the Event
An out of tolerance condition was experienced on all four RCP 62V relays during OR16. The out of tolerance condition was identified during surveillance testing and resulted in three of four RCP UV trip channels exceeding their allowable response time.
Surveillance requirement (SR) 4.3.1.2 requires verification of the response time of each reactor trip function listed in TS Table 3.3-1, Reactor Trip System Instrumentation, every 18 months. The response time for the RCP UV reactor trip channel is 5 1.5 seconds. During the 18 month surveillance testing of the RCP 62V relays, three of the four RCP UV relays exceeded their maximum allowable response time, resulting in the undervoltage reactor trip channels exceeding the limit of 5 1.5 seconds.
TS 3.3.1, Reactor Trip System Instrumentation, requires four channels of RCP UV instrumentation in Mode 1 and provides an action for the case that one RCP channel is inoperable. For this condition, however, three channels of RCP UV were inoperable due to failure to meet the time response requirement of SR 4.3.1.2. Because no action addresses more than one channel inoperable, TS 3.0.3, which requires a timely plant shutdown, would be applicable to this condition. Plant operation with three RCP UV channels inoperable exceeded the shutdown times specified in TS 3.0.3; therefore, the condition resulted in an operation or condition prohibited by the TS. This condition existed from the time the relays were last calibrated in refueling outage fifteen (September 20, 2012) until the plant entered refueling outage sixteen (April 1, 2014).
This condition is being reported as a Safety System Functional Failure under 10 CFR 50.73(a)(2)(v)(A) as the condition involved the failure of a reactor trip function credited for shutdown of the reactor and mitigation of consequences of an accident.
This condition is also being reported under 10 CFR 50.73(a)(2)(vii)(A) as it involved the failure of an RCP UV reactor trip function credited for shutdown of the reactor and mitigation of consequences of an accident and caused multiple channels to become inoperable. During OR16 surveillance testing, all four RCP UV time delay relays exceeded their response time acceptance criteria. As a result three of the four channels exceeded their TS channel response time for the RCP UV trip function.
Although determined to be a Safety System Functional Failure, the millisecond delays in reactor trip signal on RCP UV is not risk significant because the RCP UV trip signal is one of several redundant and diverse reactor trip signals in the reactor protection system. The RCP UV trip, along with the RCP underfrequency trip, serve as a backup to the low reactor coolant system (RCS) [AB] flow trip, which provides protection for a partial loss of RCS flow. The three trips also provide protection for a complete loss of RCS flow. The observed behavior in each UV trip circuit included a slight extension (0.0067, 0.0289 and 0.1165 second) in the total circuit actuation time, not a failure of the circuit to actuate. Although the actuation time exceeded the TS response time limit, the additional actuation time is very small and would have negligible impact on plant risk. In addition, the RCP underfrequency trip, the RCS low flow trip, and manual trip provided redundant reactor trip capability during this event. The Station intends to not count this event as an SSFF in the 2nd Quarter 2014 indicator as the condition would not have caused the reactor to fail to trip.
Seabrook Station Infocollects.Resource@nrc.gov , and to the Desk Officer, Office of Information and Regulatory Affairs,
Corrective Actions
The following corrective actions were completed 1-RC-P-1-B undervoltage relay was replaced with a new relay and adjusted to the required response time.
1-RC-P-1-C and 1-RC-P-1-D undervoltage relays were adjusted to the required response times.
The Root Cause Team identified the following corrective actions:
1. Revise LX0563.02, Reactor Coolant Pump Undervoltage Channel Calibration and Relay P.M., to incorporate the following requirements:
a. Energize relay for at least 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> before performing dropout testing.
b. Perform a recheck of time delay dropout time after being energized for at least 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br />. Recheck is to be performed during the outage.
c. Perform additional rechecks if adjustments were made.
2. Revise Agastat maintenance calibration procedure LS0563.12, Testing Of Agastat 125 VDC (7000 Series) TDDO Timing Relays, to ensure the 62V and 62F relays are energized for at least 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> before de-energizing for calibration.
3. Develop and present a case study on lessons learned from the RCE. Emphasize the importance of not accepting long-standing performance issues and margin management.
Similar Events testing had not verified the response time of emergency feedwater relays (LER 2012-001) and reactor trip breaker shunt trip circuits (LER 2012-002). These events resulted from ineffective methods utilized in the mid 1980's to verify that surveillance test procedures ensured compliance with the TS.
Additional Information
The Energy Industry Identification System (EIIS) codes are included in this LER in the following format: [EIIS system identifier, EIIS component identifier].
Seabrook Station