05000446/LER-2004-002
| Event date: | |
|---|---|
| Report date: | |
| Reporting criterion: | 10 CFR 50.73(a)(2)(iv)(A), System Actuation |
| 4462004002R01 - NRC Website | |
I. DESCRIPTION OF REPORTABLE EVENT
A. REPORTABLE EVENT CLASSIFICATION
An event or condition that resulted in valid actuation of any system listed in 10CFR50.73 (a)(2)(iv)(A). Specifically, an EDG automatically started and the Turbine Driven Auxiliary Feedwater Pump started.
B. PLANT OPERATING CONDITIONS PRIOR TO THE EVENT
On October 19, 2004, prior to the event, Comanche Peak Steam Electric Station (CPSES) Unit 2 was in Mode 1 and at 100% Power Operations.
C. STATUS OF STRUCTURES, SYSTEMS, OR COMPONENTS THAT
WERE INOPERABLE AT THE START OF THE EVENT AND THAT
CONTRIBUTED TO THE EVENT
There were no inoperable structures, systems, or components that contributed to the event other than the DeCordova feeder line to the CPSES 138 kV switchyard was out of service due to maintenance at the start of the event.
D. NARRATIVE SUMMARY OF THE EVENT, INCLUDING DATES AND
APPROXIMATE TIMES
On October 19, 2004 and before 0520, the DeCordova feeder line to the CPSES 138 kV switchyard was out of service due to maintenance on the CPSES switchyard circuit breaker [EIIS: (FD)(52)] and the associated DeCordova switchyard circuit breaker relays. The Stephenville feeder line was in service.
The two feeder lines to the CPSES 138 kV switchyard are from DeCordova and Stephenville switchyards. All five feeder lines to the CPSES 345 kV switchyard were in service.
At approximately 0520 on the same day, a circuit breaker in the Stephenville 138 kV switchyard tripped and failed to reclose causing the loss of power to the CPSES 138 kV switchyard. Loss of power to the 138 kV switchyard caused the loss of offsite power to the Unit 2 138/6.9 kV Startup Transformer XSTI.
As expected, both Unit 2 6.9 kV safeguards electrical trains, Trains A and B, momentarily lost power due to the loss of the 138 kV switchyard to the Unit 2 Startup Transformer. A loss of normal power supply to the 6.9 kV AC safeguards buses results in a transfer to the alternate power supply, which also includes load shed of the bus and initiation of the blackout sequencer in order to reload the bus.
Additionally, initiation of the blackout sequencer results in an automatic start of the Turbine Driven Auxiliary Feedwater Pump [EIIS: (BA)(P)]. If the transfer to the 6.9 kV Train A and B bus alternate power supply is successful, the respective EDGs will not start. In this event, a slow transfer to the alternate 6.9 kV Class IE source was initiated and both blackout sequencers started, and the Turbine Driven Auxiliary Feedwater Pump started as required.
In response to the start of Turbine Driven Auxiliary Feedwater Pump and delivery of the relatively cold feedwater, the reactor operators (utility, Licensed) lowered power by 20 MW from 1190 MWe to 1170 MWe. Additionally, the Unit 2 Train B EDG [EIIS: (BK)(DG)] started unexpectedly, but was not required to load because the safeguards buses were re-energized following the slow transfer from the CPSES 345 kV switchyard. The bus undervoltage relays should have caused both Train A 2EA1-1 and Train B 2EA2-1 circuit breakers to open in less than or equal to one-half second. 2EA1-1 operated correctly, but 2EA2-1 did not open for approximately 30 seconds due to the failed Agastat relay. The DG output circuit breaker was prevented from closing due to the 2EA2-1 breaker still being shut, thereby preventing the alternate source breaker 2EA2-2 from closing first.
The Unit 2 Train A equipment preformed as required.
The Turbine Driven Auxiliary Feedwater Pump was promptly secured per site operating procedures.
The Stephenville line was returned to service at approximately 0606 on the same day, restoring power to the CPSES switchyard and XST1. Investigation revealed that the unexpected start of the EDG was due to the erratic behavior of the Preferred Offsite Source Bus Undervoltage Agastat relay. Once the relay was replaced and the Stephenville 138 kV line was verified for continuous operation, the preferred 138/6.9 kV Class lE source was returned to service at 2324 feeding both safeguards buses.
E. THE METHOD OF DISCOVERY OF EACH COMPONENT OR SYSTEM
FAILURE, OR PROCEDURAL OR PERSONNEL ERROR
Control board indicators and alarms alerted the Reactor Operator (Utility, Licensed) of a loss of power to transformer XST1, the Unit 2 Train B EDG auto start, and the Turbine Driven Auxiliary Feedwater Pump start.
II. COMPONENT OR SYSTEM FAILURES
A. FAILURE MODE, MECHANISM, AND EFFECTS OF EACH FAILED
COMPONENT
The automatic start of the Unit 2 Train B EDG, in response to the bus transfer, was unexpected. Troubleshooting determined that the time delay relay 27BX 1/ST1 was operating erratically.
B. CAUSE OF EACH COMPONENT OR SYSTEM FAILURE
Performance of multiple tests on the time delay relay 27BX-1/ST1 determined that the time delay relay degraded to a point where repeatability was lost.
C. SYSTEMS OR SECONDARY FUNCTIONS THAT WERE AFFECTED BY
FAILURE OF COMPONENTS WITH MULTIPLE FUNCTIONS
Not applicable -- No failures of components with multiple functions have been identified.
D. FAILED COMPONENT INFORMATION
Manufacturer: Tyco Electronics Model No. E7012PA002 Agastat relay
III. ANALYSIS OF THE EVENT
A. SAFETY SYSTEM RESPONSES THAT OCCURRED
Both Unit 2 Trains A and B blackout sequencers started, the Turbine Driven Auxiliary Feedwater Pump started, and a slow transfer to the alternate 345/6.9 kV Class lE power supply XST2 was initiated. The Unit 2 Train B EDG started, but did not load. The bus undervoltage relays should have caused both Train A 2EA1-1 and Train B 2EA2-1 circuit breakers to open in less than or equal to one half second. 2EA1-1 operated correctly, but 2EA2-1 did not open for approximately 30 seconds due to the failed Agastat relay. The Unit 2 Train A equipment preformed as required.
B. DURATION OF SAFETY SYSTEM TRAIN INOPERABILITY
Not applicable -- However, the offsite 138 kV switchyard was inoperable from 0520 to 0606.
C. SAFETY CONSEQUENCES AND IMPLICATIONS
Both Unit 2 Trains A and B blackout sequencers started which function to reload the associated 6.9 kV safeguards bus in a pre-established sequence following the undervoltage condition on the bus and re-energization from the alternate power source. When the time delay relay 27BX-1/ST1 was slow to time out at approximately 30 seconds instead of less than or equal to one-half second, as designed, the preferred 6.9 kV Train B feeder breaker did not open as expected and the Unit 2 Train B EDG started, but was not required to load because alternate 6.9 kV power was supplied from the CPSES 345 kV switchyard via the Startup Transformer XST2. As a result of the slow transfer to the alternate 6.9 kV source, the Unit 2 Train B safety bus lost power for approximately 30 seconds. The unanticipated start of the EDG is attributed to the delay in the tripping of the Train B 6.9 kV switchgear preferred feeder breaker during slow transfer of the Safeguards buses to their alternate source.
Because the Train A safeguards bus was unaffected by this event, this condition did not result in a safety system functional failure, under 10 CFR 50.73 (a)(2)(v).
Furthermore, there was no disruption in the non-safeguards loads required for normal plant operations because normal plant operations are unaffected by the loss of the 138 kV switchyard: Therefore, this event did not adversely affect the safe operation of CPSES Unit 2 or the health and safety of the public.
The Turbine Driven Auxiliary Feedwater Pump start is bounded within the existing accident analysis.
IV. CAUSE OF THE EVENT
The bus undervoltage relays should have caused both Train A 2EA1-1 and Train B 2EA2-1 circuit breakers to open in less than or equal to one-half second. 2EA1-1 Train B EDG started due to the bus undervoltage, but was not allowed to close its output breaker due to 2EA2-1 breaker remaining closed. TXU Power believes that the 27BX 1/ST1 undervoltage relay for 2EA2-1 was operating erratically causing the start of the Train B EDG.
The Turbine Driven Auxiliary Feedwater Pump started as designed.
V. CORRECTIVE ACTIONS
The relay 27BX-1/ST was replaced with a new relay. The new relay was tested and within design specifications and subsequently placed in service. The defective relay was sent off site for a more detailed failure analysis to provide further understanding of the failure mechanism and potential extent of conditions for other similar Agastat relays on site.
After performing a failure analysis using Southwest Research Institute (SRI) resources, some particles considered as foreign materials (FM) were found in the "clean" area behind the diaphragm of Agastat relay, 27BX1/ST1. The conclusion of the report is that particles were transported to the exit port and disk pathway used to bleed-off the air compressed by the diaphragm when the relay coil is energized. Restricting the exit path would cause the relay to time long. FM may also cause the relay to time erratically since these particles may move in the exit pathway as air passes them. SRI also determined that the foreign materials were introduced into the clean air cavities of the relay at the manufacturer's assembly area. However, based on the reliable operation of the relay (past calibrations 1995-2001 have been within set point criteria) the migration of the particles is random and not predictable and as such, the failure of relay 27BX1/ST1 is considered a random failure.
The extent of condition applies to relays of the same manufacturer and design that are used on Unit 2 Train A and both trains in Unit 1; however, CPSES considers this failure to be a random failure with low risk.
VI. PREVIOUS SIMILAR EVENTS
There was a previous event involving an Agastat relay on October 7, 2002. Specifically, unexpected auto start of the Unit 1 Train B EDG. Corrective actions from the previous 2002 event included development of a schedule to replace 212 of the safety related relays which perform a safety-related function and approximately 64% of these relays have been replaced. The schedule was based on availability and on risk of failure of these relays based on past calibration data and failure history.
This specific relay was scheduled to be tested and replaced, if required, on October 25, 2004 as a result of the previous event. The current completion date to replace the remaining safety related relays is December 31, 2005.