05000220/LER-2001-001
| Event date: | 08-22-2001 |
|---|---|
| Report date: | 10-22-2001 |
| Reporting criterion: | 10 CFR 50.73(a)(2)(iv)(A), System Actuation |
| 2202001001R00 - NRC Website | |
I. Description of Event
On August 22, 2001 at 0508 hours0.00588 days <br />0.141 hours <br />8.399471e-4 weeks <br />1.93294e-4 months <br />, Nine Mile Point Unit 1 experienced an unplanned scram, from approximately 100 percent power, duo to a generator trip. The reactor was operating at steady state conditions prior to the scram. The immediate cause for the generator trip was actuation of the Negative Phase Sequence Current Relay generator protection logic in response to a grid perturbation, All control rods fully inserted. In response to the generator trip, reactor pressure rose to 1115 pounds per square inch gage (psig). The increase in pressure resulted in all six Electromatic Relief Valves (ERVs) opening, as designed. The ERVs closed within four seconds of the scram.
Feedwater control transferred to the High Pressure Coolant Injection mode upon receipt of the turbine trip signal as designed. Approximately 37 seconds post scram, the Main Steam Isolation Valves (MSIVs) automatically closed.
The MSIVs receive a close signal when the mode switch is in RUN and reactor pressure lowers to 850 psig. During the event, operators were performing post scram reporting and did not reposition the reactor mode switch from RUN to SHUTDOWN prior to reactor pressure reaching 850 psig. The scram reporting involves communicating the status of the reactor parameters, including control rod position, to the Control Room Supervisor. Minimum pressure reached was approximately 815 psig. The maximum pressure reached while the MSIVs were closed was approximately 920 psig. Approximately six minutes after closing, the MSIVs were re-opened, following an evaluation of plant conditions.
Post scram, reactor water level reached a minimum level of 30 inches. The feedwater flow control valves functioned as designed and recovered reactor water level. Reactor water level reached 90 inches approximately one minute after the scram. At 90 inches all feedwater flow control valves were closed. Heatup of the water injected post scram resulted in reactor water level increasing to approximately 106 Inches. Feedwater pumps 11 and 12 tripped as designed when reactor water level exceeded 95 inches. The high reactor water level resulted in water level partially covering tho Emergency Condenser (EC) nozzles. Water entry into the EC lines had been previously analyzod and the ECs remained operable. Water level remained below the level at which water entry into the Main Steam Lines would occur.
An evaluation post scram concluded that the cause of the trip was a generator load rejection caused by the actuation of the Negative Phase Sequence Current Relay. A review of fault recordings indicated that a fault occurred on tho 345 kilovolt (KV) grid outside the station, which was detected by the Negative Phase Sequence Current Relay. The fault event lasted only four cycles, approximately 64 milliseconds, However, the Negative Phase Sequence Current Relay has a time dolay which should have prevented the relay from actuating.
The Negative Phase Sequence Current Relay is a General Electric (GE) 12SGC21A1A relay. The relay is designed to protect the generator against possible damage from unbalanced currents resulting from prolonged faults or unbalanced-load conditions. When actuated, the relay causes a generator load rejection that would be followed by a turbine trip.
A review of operating experience identified a similar event at Palo Verde in 1992 and identified GE Service Advisory Lollor (SAL) 189.1 as applicable to the Nine Mile Point Unit 1 event. GE SAL 189.1, Issued in 1987, documented a design flaw in the time delay circuitry of relay 12SGC21A1A that could occasionally result in an almost instantaneous actuation. A review of records indicates that Nine Mile Point Unit 1 had not previously evaluated this GE SAL or the Palo Verde event.
Nn 5701-1M1GGA (1-2001) .FACIUTY NAME (1) DOCKET (2) JAN-11-2002 FR1 02:41 PM � FAX NO. 3 � P. 05 II, Cause of Event The reactor scram was caused by a generator load reject. The faulty design of the Negative Phase Sequence Current Rolay caused the generator load reject. The relay has a time delay that should have prevented actuation, given the length of time the grid perturbation was present. However, a design flaw in the relay can result in the time delay becoming nearly instantaneous. This design flaw is documented in GE SAL 189.1.
A contributing cause is a failure to use operating experience. In 1987, GE issued SAL 189.1 documenting a design flaw in the time delay circuitry of relay 12SGC21A1A. In 1992 a similar event occurred at Palo Verde. A review of records Indicates that Nine Mile Point Unit 1 had not previously evaluated this GE SAL or the Palo Verde event.
The MSIV closure was caused by unclear standards regarding when to reposition the mode switch to SHUTDOWN post scram, Interviews with operating crews identified that repositioning the mode switch post scram was not consistent among the crews. Some crews reposition the mode switch to SHUTDOWN prior to providing a scram report to the Control Room Supervisor while other crews provided the scram report prior to repositioning the mocio switch.
III. Analysis of Event
The reactor scram and automatic closure of the MSIVs are reportable in accordance with 10 CFR 50.73(a)(2)(iv)(A), "Any event or condition that resulted in manual or automatic actuation of any of the systems listed in paragraph (a)(2)(iv)(0).,.." Paragraph (a)(2)(iv)(B) lists actuation of the reactor protection system including reactor scram and containment isolation signals affecting multiple main steam isolation valves.
During this event, all plant safety systems operated as designed. The ERVs opened consistent with the analysis of a load rojection with a scram as described in the Final Safety Analysis Report (Updated). Additionally, no Emergency Core Cooling Systems actuated or should have actuated during the event.
Although the MSIVs closed during the event they were re-opened within six minutes. During the time that the MSIVs were shut reactor pressure reached a maximum of 920 psig. No ERV set points were reached, and the condenser remained available.
Post scram the foedwater system functioned as designed. Heatup of water Injected post scram contributed to high reactor water level which resulted in water level partially covering the EC nozzles and the tripping of foedwator pumps 11 and 12. Water entry into the EC lines had been previously analyzed and the ECs remained operable. Reactor water level remained below the level at which water entry into the Main Steam Lines would occur. Feedwater pumps 11 and 12 were available for restart upon resetting the high level trip and would have automatically restarted if reactor water level had reached 53 inches.
A Probabilistic Risk Assessment (PRA) evaluation of the event concluded that the estimated Conditional Core Damage Probability (CCDP) was 6.68 E-7 and therefore the event was not risk significant.
Based on the above, the event did not pose a threat to the health and safety of plant personnel or the public.
DOCKET (2) LER NUMBER (6) NUMER1R FACILITY NAME (1) PAGE (3) 0 JAN-11-2002 FRI 02:41 PM FAX NO, 3 P. 06
IV. Corrective Actions
1. The Negative Phase Sequence Current Relay was replaced with a GE relay of a now design.
2. Nine Mile Point will conduct an operating experience review of GE SALs not previously evaluated, 3. The Operations Manual was modified such that the scram report is provided after the mode switch Is placed in SHUTDOWN.
4. Operations crews were briefed that the management expectation Is that the first immediate action post scram is to place the mode switch In SHUTDOWN.
V Additional Information
A. Failed Components
General Electric (GE) 12SGC21A1A relay
B. Previous similar events:
A review of Nine Mile Point Unit 1 Licensee Event Reports (LER) from 1995 to the present did not reveal events with a similar cause, faulty relay design. LER 95-02, "Reactor Scram Caused by Failure of Generator Protective Relay" and LER 96-11, "Reactor Scram Caused by the Main Generator Lockout Relay Trip" describe similar events, i.e. generator initiated scrams. The cause identified in LER 96-11 is failure of a flexible link in the generator exciter. The cause identified in LER 95-02 is failure of an oil filled capacitor in the foss of excitation relay, The corrective actions from the previous LERs would not have prevented this event.
C. Identification of components referred to in this Licensee Event Report Components IEEE 805 System ID IEEE 603A Function Emergency Condenser BL N/A Feociwater System
SJ N/A
High Pressure Coolant Injection
BJ N/A
SB N/A
Main Generator TB N/A Main Generator Output Power System EL N/A Main Turbine TA N/A Control Rod Drive System AA N/A Valve SB ISV Pump SJ P Control Rod AA ROD Relay EL, TB 46, 40 Nozzle BL NZL Capacitor EL CAP Exciter
TB EXC
Link TB N/A