DCL-96-155, Revised Special Rept SR 95-05:on 950806,EDG 1-2 to Achieve Rated Output Frequency within TS Limits.Caused by Degraded Electronic Governor Performance.Replaced Electronic Governor Motor Operated Potentiometer & Electronic Governor
| ML20116B856 | |
| Person / Time | |
|---|---|
| Site: | Diablo Canyon  |
| Issue date: | 07/22/1996 |
| From: | PACIFIC GAS & ELECTRIC CO. |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| Shared Package | |
| ML20116B844 | List: |
| References | |
| DCL-96-155, SR-95-05, SR-95-5, NUDOCS 9607300315 | |
| Download: ML20116B856 (9) | |
Text
Enclosure 3 PG&E Letter DCL-96-155 REVISION TO SPECIAL REPORT 95-05, EMERGENCY DIESEL GENERATOR 1-2 FAILURES TO ACHIEVE RATED OUTPUT FREQUENCY WITHIN TECHNICAL SPECIFICATION TIME LIMITS DUE TO DEGRADED ELECTRONIC GOVERNOR PERFORMANCE in accordance with Regulatory Guide (RG) 1.108, Section C.3.b, the following information is included: i
- 1. Diesel aenerator involved: Emergency Diesel Generator (EDG) 1-2.
- 2. As of April 1.1996 the number of valid failures in the last 100 EDG i 1-2 valid tests: 3
- 3. Cause of failure: PG&E has determined that the most probable cause of these events was due to a defective electronic governor l
stability potentiometer on the electronic governor previously :
installed on July 19,1995. '
- 4. Corrective measures taken:
]
- a. Immediate corrective action: PG&E replaced the electronic governor motor operated potentiometer (MOP) and l electronic governor. PG&E then successfully tested EDG 1-2 in accordance with Surveillance Test Procedure (STP)
M-9A, " Diesel Engine Generator Routine Surveillance Test."
- b. To prevent recurrence: The dedication process for the EDG <
electronic governors has been revised. The dedication process requires that all electronic governors be functionally tested prior to being installed on a EDG. PG&E has replaced all the electronic governor units on EDG 1-1,1-3, 2-1, 2-2, and 2-3.
- 5. Time EDG was unavailable: EDG 1-2 was cleared to investigate the cause of the failure to achieve rated output frequency on August 6 and 7,1995. EDG 1-2 was declared operable after successful performance of STP M-9A on August 8,1995, at 1735 PDT. Thus, EDG 1-2 was unavailable for a total of 54 hours6.25e-4 days <br />0.015 hours <br />8.928571e-5 weeks <br />2.0547e-5 months <br /> and 41 minutes.
Enclosuro 3 PG&E Letter DCL-96-155 i
~
7.' Confirmation of proper test interval: As of April 1,1996, the total number !
of valid failures in the last 100 valid tests for EDG 1-2 is 3, and the total number of valid failures in the last 20 valid tests for EDG 1-2 is 0;
' therefore, the 31-day test interval is in compliance with the accelerated ;
test schedule of TS Table 4.8-1.
I. Plant Conditions Unit 1 was in Mode 1 (Power Operation) at 100 percent power, ll. Description of Event A. Summary In accordance with the requirements of Diablo Canyon Power Plant (DCPP) TS 6.9.2 and 4.8.1.1.4, and Revision 1 to NRC RG 1.108, PG&E is submitting this revision to Special Report 95-05 concerning two failures of EDG 1-2 on August 6 and 7,1995. l Both events involved the failure of EDG 1-2 to achieve rated output frequency within TS time limits due to degraded electronic govemor performance.
B. Background TS 4.8.1.1.2 requires that each EDG be demonstrated operable by verifying that the generator frequency be 60 i 1.2 Hz within 13 seconds after the start signal.
STP M-9A implements TS 4.8.1.1.2 and is normally performed on a 31-day testing frequency. STP M-9A starts and runs each EDG for
. a minimum of one hour. STP M-9A is performed to demonstrate - l proper startup to verify that the required voltage and frequency are automatically attained within acceptable limits and time. The EDG must reach an indicated frequency between 60 i 1.0 Hz within 13 seconds to meet the STP acceptance criteria. This test also verifies that the automatic startup components of the EDG unit are operable.
C. Event Description ,
On July 19,1995, PG&E replaced the EDG 1-2 electronic governor.
2-
)
. Enclosure 3 PG&E Letter DCL-96-155 1
Between July 27 and August 4,1995, five STP M-9As were performed satisfactorily, j On August 5,1995, during performance of STP M-9A a single load spike of 300 kW was observed when a control room operator used the electronic governor MOP to adjust load. The 300 kW load spike would translate to a possible 0.35 Hz frequency change on the 4 kV bus. An engineering evaluation was performed and I confirmed by the electronic governor vendor, Woodward Governor, j that the spike would not be observed on the 4 kV bus due to slow l response time of the hydraulic actuator and high inertia of the !
EDG.
)
i Using the guidance of RG 1.108, Sections B and C.2.e, PG&E .
determined that this event is not a valid failure because EDG 1-2 j successfully loaded within the time required by TS 4.8.1.1.2, and '
all equipment required to be operating during emergency ;
conditions -was capable of performing their intended safety ;
function. 1 On August 6,1995, at approximately 1116 PDT, during )
performance of STP M-9A for EDG 1-2, the frequency went up to
)
62.5 Hz and was difficult to adjust. Subsequently, while EDG 1-2 .
was idling, the frequency drifted between 59 to 60.5 Hz. At this I time, EDG 1-2 was declared inoperable and TS 3.8.1.1, Action i Statement b, was entered. PG&E replaced the MOP based on the following indications that the MOP may be the cause of the problem: (1) a 300 kW load spike occurred during a previous test on August 5,1995, when the operator was using the MOP to adjust load, (2) EDG 1-2 frequency came up to 62.5 Hz and was difficult to adjust and, (3) even though difficulty in adjusting the frequency could be indicative of a degraded electronic governor, PG&E did not replace the electronic governor at this time because the electronic governor was recently replaced on July 19,1995. l On August 7,1995, at 1445 PDT, during surveillance testing activities to verify correction of the problem identified on August 6, .l 1995, a frequency oscillation occurred and EDG 1-2 was shutdown i for electronic governor adjustments. EDG 1-2 was restarted and l frequency came up to 62.5 Hz and could not be adjusted with the !
electronic governor MOP. PG&E replaced the electronic governor with a newly refurbished electronic governor since the electronic governor MOP had already been replaced, and the inability to adjust frequency was indicative of a degraded electronic governor.-
.. - -~ ._. .- - . - - . - . . - . - - - . . - - - . . - - - . - . .
Enclosura 3 PG&E Letter DCL-96-155 On August 8,1995, at 1735 PDT, EDG 1-2 was declared operable after the successful performance of STPs M-9A, M-9D1, and M-9D2, Both events are considered to be valid failures in accordance with the guidance of RG 1.108, Sections B and C.2.e.' Both' events were due to the failure of EDG 1-2 to achieve rated output -
frequency within TS time limits due to degraded electronic govemor performance.
On August 13,1995, during performance of STP M-9A, a load drop of approximately 200-300kW was observed on the output of EDG 1-2.
On August 15,1995, after running EDG 1-2 at greater than 2560- )
kW for approximately 2.5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br />, a load drop of 250 kW occurred. l At this time, EDG 1-2 was declared inoperable and subsequent l troubleshooting / data gathering run for approximately 20 hours2.314815e-4 days <br />0.00556 hours <br />3.306878e-5 weeks <br />7.61e-6 months <br /> did :
not reveal any load swings or drops. Subsequently, PG&E and !
L Woodward Governor Co. determined that the mechanical governor i needed to be replaced. The cause of the degraded mechanical l governor has been determined to be wear of critical mechanical i
- governor components.
- On August 17,1995, the mechanical governor and actuator was replaced. In addition, PG&E has replaced the original mechanical governor units with refurbished units on EDG 1 -1,1 -3, 2-1, 2-2, and 2-3.
On August 18,1995, EDG 1-2 was declared operable after the successful performance of STPs M-9A, " Diesel Generator Routine Surveillance Test; " M-9D1, " Diesel Generator Full Load Rejection l
Testing;" and M-9D2, " Diesel Generator Partial Load Rejection Testing."
- The load drop events on August 13 and 15,1995, were determined i not to be valid failures in accordance with the guidance of RG 1.108, Sections B and C.2.e, and because EDG 1-2 successfully started and loaded in accordance with TS requirements.
On August 19,1995, during performance of STP M-9A, EDG 1-2 was paralleled at approximately 30 degrees out of synchronization .
- with the offsite power source. Subsequently, VAR oscillations 1
i-l~ . . -
' ~
Enclosuro 3 PG&E Letter DCL-96-155 occurred on EDG 1-2. The load on EDG 1-2 was reduced and the voltage regulator MOP was adjusted to various settings exercise the MOP. The EDG 1-2 load was subsequently raised to 2600 kW and no further VAR oscillations were observed.
The immediate cause of the initial VAR swings was due to paralleling EDG 1-2 out of phase with the offsite power source.
VAR swings that occurred after the EDG.was paralleled were due to the degraded voltage regulator MOP. PG&E believes that the most probable cause of the degraded voltage regulator MOP was due to a transient caused by paralleling EDG 1-2 out of phase with the offsite power source that caused arcing or carbon buildup on ,
the MOP. j l
Using the guidance of RG 1.108, Sections B and C.2.e, PG&E <
determined that this event was not a valid failure because EDG 1-2 successfully loaded within the time required by TS 4.8.1.1.2 and all equipment required to be operating during emergency conditions.
were capable of performing their intended safety function.
An evaluation was performed to determine if the VAR swings of this event were related to the VAR swing experienced on June 21, 1995 (reference DCL-95-158). A comparison of the two events -
was made and it was determined that the EDG 1-2 symptoms and responses were similar; however, firm evidence does not exist to determine conclusively if the two events were related.
On September 7,1995, during performance of STP M-9A, slight VAR swings were observed when the voltage regulator MOP was set at 500 VAR out. At this time the voltage regulator motor operated autotransformer (MOA) and voltage regulator MOP were replaced.
This event was determined not to be a valid failure in accordance with the guidance of RG 1.108, Sections B and C.2.e. PG&E has determined that this event is not a valid failure because EDG 1-2 was successfully started and loaded in accordance with TS requirements. VAR swings were only present when the voltage regulator MOP was adjusted. In an emergency condition the EDG is operated in the isochronous mode and the voltage regulator MOP and MOA are not adjusted. Therefore, the EDG 1-2 components needed for operation in the emergency mode would have been available during an accident.
Enclosuro 3 ,
PG&E Letter DCL-96-155 - !
i l
, On September 16,1995, STP M-9A was performed satisfactorily.
During the surveillance test, slight VAR oscillations were observed. j Accelerated testing in accordance with TS Table 4.8-1 was :
stopped and EDG 1-2 test frequency was returned to once every 31 days.
This event was determined not to be a valid failure in accordance with the guidance of RG 1.108, Sections B and C.2.e. PG&E has determined that this event is not a valid failure because EDG 1-2 was successfully started and loaded in accordance with TS j requirements. VAR swings were only present when the voltage l regulator MOP was adjusted. In an emergency condition, the EDG l is operated in the isochronous mode and the voltage regulator I MOP is not adjusted. Therefore, the EDG 1-2 components needed )
for operation in the emergency mode would have been available I during an accident. l 1
Between August 8,1995 and April 1,1996,33 STP M-9As were performed. In addition, during the Unit 1 seventh refueling outage a loss of offsite power event occurred on Unit 1 and all three EDGs automatically started and loaded on their associated 4 kV buses.
The EDGs supplied power to the Unit 14 kV vital busses for approximately 18 hours2.083333e-4 days <br />0.005 hours <br />2.97619e-5 weeks <br />6.849e-6 months <br />. Review of the strip-chart recorders in the control room revealed that no voltage oscillation (VARs in droop mode)'or frequency oscillations (Load in Droop mode) occurred on EDG 1-2. This operational data confirmed previous conclusions -
that with the EDG operating in the isochronous mode a degraded voltage regulator MOP would have no impact on EDG 1-2 operation.
D. Inoperable Structures, Components, or Systems that Contributed to !
the Event None.
E. Dates and Approximate Times for Major Occurrences
- 1. August 6,1995, at 1116 PDT: Event / discovery date for event 1, EDG 1-2 was l shutdown when required frequency was not achieved. EDG 1-2 was declared inoperable.
Enclosuro 3 PG&E Letter DCL-96-155 l
- 2. August 7,1995, at 1445 PDT: Event / discovery date for ;
event 2. EDG 1-2 did not l reach the required output frequency during surveillance testing.
- 3. August 8,1995, at 1735 PDT: EDG 1-2 was declared l operable after successful I performance of STPs M-9A, M-9D1, and M-9D2.
- 4. September 16,1995: Accelerated testing in accordance with TS Table 4.8-1 was exited and EDG 1-2 test frequency was returned to once every 31 days.
F. Other Systems or Secondary Functions Affected None.
G. Method of Discovery Plant personnel identified the problem during the performance of STP M-9A and additional testing.
H. Operator Actions None.
- l. Safety System Responses None.
Ill. Cause of the Event .
A. Immediate Cause The immediate cause of both events was degraded performance of l
. EDG 1-2 equipment.
Enclosuro 3 PG&E Letter DCL-96-155 B. Root Cause -
The most probable root cause of this event has been determined to be a defective electronic governor stability potentiometer.
IV. Analysis of the Event Safety-related (Class 1E) electrical loads are supplied from three 4160 V vital buses (F, G and H) for each unit. Each vital bus can be supplied from the 500 kV switchyard, the 230 kV switchyard, the main generator, or the EDGs. The EDGs can provide power for engineered safeguards (Class 1 E) motors and loads used for emergency core cooling, reactor shutdown, and other vital safety functions when the main generator and offsite power sources are not available in the event of a loss-of-offsite power or other design-basis event.
l During a design-basis event, with EDG 1-2 in a degraded condition, one bus of vital 4160 V power may not have been available for Unit 1 support.
However, EDGs 1-1 and 1-3 were available to provide power to the other two Unit i vital 4160 V buses during the time that EDG 1-2 was inoperable. Subsequent to this event, EDGs 1-1 and 1-3 successfully completed STP M-9X, " Operability Verification."
[ Since only two vital buses are necessary to support DCPP accident analysis, the health and safety of the public were not adversely affected l by this event.
V. Corrective Actions
- A. Immediate Corrective Actions l
l 1. The EDG 1-2 electronic governor MOP and electronic l governor unit were replaced and functionally tested.
! 3. PG&E performed accelerated testing between August 8 and i
September 16,1995, at least once every 7 days to determine and validate the cause of the degraded EDG i equipment.
! 4. The electronic governor was sent offsite for failure analysis i and refurbishment.
8-
.. Enclosuro 3 PG&E Letter DCL-96-155
- 5. Additional temporary load monitoring equipment was installed on EDG 1-2.
B. Corrective Actions to Prevent Recurrence
- 1. The dedication process for the EDG electronic governors has been revised. The dedication process requires that all electronic governors be functionally tested prior to being installed on a EDG.
- 2. PG&E has replaced all EDG electronic governor units. l VI. Additional information A. Failed Components Component: Potentiometer,0-150 kOhms Manufacturer: A&B Model Number: 202760 Type: J B. Previous Similar Events On March 29,1994, at 2144 PDT, with Unit 1 defueled during performance of STP M-9A, EDG 1-1 failed to meet acceptance criteria for achieving the required frequency within the TS time limits during a simulated undervoltage start. The most probable cause for the failure to achieve rated frequency within TS time requirements has been determined to be due to electrical governor performance. At the time, the electrical governor was adjusted, and DG 1-1 was successfully tested in accordance with STP M-9A.
4 I
i i
1 i 4
.