ML20207P554

Ro:On 861213,emergency Diesel Generator 14 Failed to Start During Invalid Tests.Caused by High Coolant Pressure Low Alarm Setpoint & Low Coolant Pressure Switch Setpoint. Setpoints Revised

ML20207P554
Person / Time
Site:	Fermi
Issue date:	01/12/1987
From:	Lenart R DETROIT EDISON CO.
To:	NRC OFFICE OF ADMINISTRATION (ADM)
References
NP870016, NUDOCS 8701160239
Download: ML20207P554 (4)

Text

,--- - -

U. S. Nuclear Regulatory Commission January 12, 1987 NP870016 Page 2 ENCLOSURE 1 Emergency Diesel Generator Start Failure

1. Emeroency Diesel Generator Involved:

Emergency Diesel Generator (EDG) 14

2. Identification of Failure:

Failure Date: December 13, 1986, 1227 hours0.0142 days <br />0.341 hours <br />0.00203 weeks <br />4.668735e-4 months <br /> Description of Event:

This is the fourth start failure of EDG 14 that occurred during

" invalid tests" following the issuance of the Fermi 2 operating license. At the time of this start failure there have been 27 valid test starts since the operating license wcs issued. The December 13, 1986 start failure is considered a failure in an

" invalid test" under section C.2.e (2) of NRC Regulatory Guide 1.108 because the failure involved a " low jacket water pressure trip", which is bypassed during emergency starts. The number of start failures incurred during " valid tests" for EDG 14 subsequent to the issuance of the operating license remains unchanged at zero.

3. Cause of Failure:

On December 13, 1986, EDG 14 was being started to perform a

" slow" start surveillance test, which is required on a monthly basis. During a " slow start, the engine is started and speed is gradually increased to 940 RPM before loading. At approximately 800 RPM the engine unexpectantly tripped due to low jacket water pressure. A jacket coolant pressure low alarm was received.

This start failure on EDG 14 is similar to a start failure on EDG 12 that occurred November 14, 1986 and was reported in reference 2.

The low jacket coolant pressure alarm, was the only alarm received on December 13, 1986 when EDG 14 tripped. This indicates that the trip was caused by the Coolant Pressure Low (CPL) relay. The CPL relay trip is cypassed during emergency starts.

0701160239 DR 870112 ADOCK 05000341 PDR

r 4

U. S. Nuclear Regulatory Commission January 12, 1987 NP870016 Page 3 The CPL relay is a normally energized relay. During the start cycle, the normal speed relay keeps CPL energized until either the jacket coolant pump develops 24 PSIG head and closes the Coolant Pressure Switch (CPS) or the engine reaches 810 RPM.

Once the engine has developed approximately 10.5 psig jacket coolant pressure the CPLA pressure switch closes and keeps the CPL relay energized.

If CPLA does not close before either the tLchometer relay closes at 810 RPM, or CPS closen at 24 psig jacket coolant pump pressure, the engine will trip during non-emergency starts, and a low jacket coolant pressure alarm will be received. During emergency starts, only the alarm is received. .In this case a low jacket coolant pressure alarm was received and the engine tripped near 800 RPM, this indicates that CPLA did not close before the normal speed relay opened.

The root cause identified in Reference 2 was that the engine did not develop sufficient jacket water pressure to close CPLA before reaching 810 RPM. Additionally, it was determined that there was a potential for the Normal Speed Relay to open before 810 RPM.

This is because the CPS setpoint was too low. Either of the two possible causes identified for the EDG 12 trip may have caused the EDG _14 trip. Prior to the slow start modification installed on the engines in 1985, the low jacket water pressure trip / alarm was not allowed until 10.5 seconds af ter the engine reached 810 RPM.

In addition to the above two possible causes, a third potenial cause for the trip has been identified. After EDG 14 had been started and synchronized a functional check of the new tachometer signal generator that had just been installed to improve tachometer signal generator reliability, indicated that the tachometer was varying i 50 RPM from 900 RPM (actual engine speed). It is possible that the signal generator provided a spurious signal of 810 RPM before the engine had actually reached that speed. This may have caused the engine to trip, due to insufficient jacket coolant pressure to close CPLA.

Based on the above, the root cause of the EDG 14 trip was one of the following:

A) CPLA setpoint is too high. The engine was unable to de elop sufficient jacket coolant pressure to close CPLA at the point in the start cycle the trip is actuated.

B) CPS setpoint is too low. This would have actuated the jacket coolant pressure trip before the engine reached 810 RPM.

z -- _

..o

'U. S. Nuclear Regulatory Commission January 12, 1987 NP870016

- Page 4 C) The new signal generator installed created a spurious 810 RPM signal before the engine developed sufficient jacket coolant pressure.

None of the above root causes would affect the engine's operation during an emergency start, as the jacket coolant pressure low trip is a non-essential trip which is bypassed during emergency starts.

After the start failure on December 13, 1986, EDG 14 CPLA setpoint was changed from 10 psig decreasing to 8 psig decreasing and CPS setpoint was revised from 24 psig increasing to 25 psig increasing.

At the completion of the setpoint changes EDG 14 was restarted and surveillance testing was successfully completed on December 14, 1986.

4. Corrective Measures Taken:

As a result of the EDG 12 Start Failure of November 14, 1986 Detroit Edison has decided to revise the CPLA set point on all four Fermi 2 EDGs from 10 psig decreasing to 8 psig decreasing, and revise the CPS setpoint on all four Fermi 2 EDGs frova 24 psig increasing to 25 psig increasing. The above corrective actions had not yet Laen implemented on EDG 14 prior to its start failure. Detroit Edison is in process of completing these actions on the remaining diesel (EDG 13). Detroit Edison will also investigate the problem with the EDG 14 tachometer, and take appropriate corrective action. These measures will ensure the CPLA switch closes prior to the engine reaching 810 RPM, and should preclude recurrence of a similar event. The new satpoints will not affect the Emergency Diesel Generators' ability to perform their safety function.

5. Lenoth of Yime FDG Unit Unavailable:

EDG 14 was out of service 14 hours1.62037e-4 days <br />0.00389 hours <br />2.314815e-5 weeks <br />5.327e-6 months <br /> to determine the cause of the start failure and to make the setpoint changes to CPLA and CPS.

6. Current Surveillance Test Interval:

The current surveillance test interval is once per 31 days.

7. Verification of Conformance of Test Interval:

Because this start failure is considered an " invalid test" start failure as defined in Regulatory Guide 1.108 the EDG surveillance test interval is not affected. The current surveillance test interval is in conformance with Regulatory Guide 1.108 and with Fermi 2 Technical Specification Table 4.8.1.1.2-2.

~

t Oi,;ay" Rsg. Guida'1.108 Fermi-2 6400 North D6xie Highway

$Uj'e"c;"fygan4aiss January 12, 1987 Nuclear s

NP870016 "'""*

U. S. Nuclear Regulatory Commission Attention: Document Control Desk Washington, D. C. 20555

References:

(1) Fermi 2 NRC Docket No. 50-341 Facility Operating License No. NPF-43 (2) Emergency Diesel Generator Start Failure Report, NP860650, Dated December 12, 1986 =

(3) Emergency Diesel Generator Start Failure Report, NP860636, Dated November 26, (

1986

Subject:

Emeroency Diesel Generator Start Failure Please find attached our report on one start failure of Emergency Diesel Generator (EDG) 14 which occurred on December 13, 1986. This report is submitted to you in accordance with Fermi 2 Technical Specifications 4.8.1.1.3 and 6.9.2.

If you have any questions regarding this report, please contact Mr. Lewis Bregni at (313) 586-5313.

Sincerely, t

R. S. Lenart Plant Manager cc: B. A. Davis W. G. Rogers J. J. Stefano ge\

t g

.