ML17299A384

Special Rept Re 850509 Failure of Diesel Generator to Start. Caused by Failure of turbo-charger Manifold Temp Control Valve in Vent to Atmosphere Position & Major Starting Air Sys Leak.Control Air Supply Plugged to Temp Control Valves

ML17299A384
Person / Time
Site:	Palo Verde
Issue date:	06/10/1985
From:	Van Brunt E ARIZONA PUBLIC SERVICE CO. (FORMERLY ARIZONA NUCLEAR
To:	Martin J NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION V)
References
ANPP-32809-EEVB, NUDOCS 8506280079
Download: ML17299A384 (8)

Text

Arizona Nuclear Power Project P.o. BOX 52034 ~ PHOENIX, ARIZONA85072-2034 Mr. John B. Martin, Regional Administrator ANPP-32809-EEVB/GEC Office of Inspection and Enforcement June 10, 1985 U. S. Nuclear Regulatory Commission Region V 1450 Maria Lane, Suite 210 Walnut Creek, CA 94596-5368

Subject:

Palo Verde Nuclear Generating Station (PVNGS)

Unit 1 Special Report Diesel Generator Failures to Start Docket No. STN 50-528, License No. NPF-34 File: 85-056-026 G.l.01.10

Dear Mr. Martin:

Attached please find a Special Report prepared and submitted pursuant to Specifications 4.8.1.1,.3 and 6.9.2 of Appendix A (Technical Specifications) to the Palo Verde Nuclear Generating Station, Unit No. 1 Operating License. This report discusses diesel generator failures to start.

If you have any questions or concerns, please contact me.

Very truly yours, cu~

E. E. Van Brunt, Jr.

n ~

Executive Vice President Project Director EEVBJr/GEC/bg Attachments cc: R. P. Zimmerman A. L. Hon E. A. Licitra A. C. Gehr INPO Records Center 39lh NOI938 SO ai e Ci tlr SSI 850b280079 850b 28 PDR ADQCK 0500052 PDR 8

I

'":3? iJ 1 f spilt(,

sic ~ . s l3 p

~ ~ L h.'l <. l)l')

I t

PALO VERDE NUCLEAR GENERATING STATION UNIT 1 SPECIAL REPORT DIESEL GENERATOR FAILURES TO START Docket No. STN 50-528 On May 9, 1985,at 1418,with Unit 1 in Mode 3, Diesel Generator "B" was determined inoperable due to the failure to attain the required voltage, frequency, and speed within 10 seconds (actual time: 10.85 seconds, 13.25 seconds, and 12.0 seconds, respectively) as per Technical Specification 4.8.1.1;2.a.4(a) during the performance of Surveillance Test 41ST-1DG02. A subsequent start attempt of Diesel Generator "A" for Technical Specification 3.8.1.1 Action Statement "a" per 41ST-1ZZ02 at 1507 resulted in Diesel Generator "A" also being determined inoperable due to the engine failing to start within the required 10 seconds. With both Diesel Generators declared inoper-able, Unit 1 was required to comply with Technical Specification 3.8.1.1 Action State-ment "e" per 41ST-1ZZ02. These two start failures brought the total number of Diesel Generator start failures for Unit 1 to three (3) in thirty-two (32) valid tests, placing Unit 1 under the criteria of R.G.1.108.C.2.d.(3), which requires that the test interval should not be more than.7 days.

The Diesel Generator B start attempt was performed using a single air start receiver.'he onsite Diesel Generator (Cooper Energy Services) manufacturer representative was not familiar with the Technical Specification requirements and determined that our engines were not designed to meet the 10 second start time acceptance criteria with only one air start receiver operable. This was due to design pneumatic time delays and difference in the right and left bank starting air and control air piping. This hypothesis could not be confirmed, however, since the Cooper Energy Service (CES) design engineering group could not be contacted. Based on the information supplied by the onsite,CESrepresentative, surveillance test 41ST-1ZZ02 was performed at 1901 using both air'start receivers with all acceptance criteria being satisfied. Diesel Generator "B" was declared operable at 2023 on May 9, 1985 after having been inoperable for 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> and 5 minutes. This removed the requirements of Technical Specification 3.8.1.1 Action Statement "e" and was replaced by Action Statement "a".

The Diesel Generator "A" start failure was due to the engine not receiving fuel as indicated by analysis of a recorder strip chart from a Chart Recorder installed for surveillance testing. During the inspection following the start failure, a control air pressure regulator indicated an abnormally low system pressure which was not sufficient to actuate the engine fuel racks. If the (fuel racks) mechanical linkage that controls fuel supply to the diesel generator injectors are not actuated to the open position, the engine is deprived of fuel and cannot start. Troubleshooting of the low control air system pressure disclosed the following discrepancies: (1) a Turbo-charger Manifold Temperature Control Valve failed in the "Vent To Atmosphere" position, and (2) a major starting air system leak was found at the vibration shutdown switch that was caused by a cross-threaded Swagelock tubing fitting. The immediate corrective action was to plug the control air supply to the Temperature Control Valves with CES onsite representative concurrence under Temporary Modification Tag gl-85-DG187 and repair the vibration shutdown switch Swagelock fitting. On May ll, 1985 at 1702, Diesel Generator "A" started satisfactorily during a troubleshooting run and at 1724, Sur-veillance Test 41ST-1DG01 was performed with all acceptance criteria satisfied. Unit 1 Diesel Generator "A" was declared operable at 1942 after having been inoperable for 49 hours and 55 minutes. This no longer required that Technical Specification 3.8.1.1 Action Statement "a" be met since both Diesel Generators were operable.

F 0

t

." Special Report-Diesel Gen tor Failure to Start

":Page 2 of 3 During the troubleshooting of Diesel Generator "A" control air system low pressure, the Air Manifold Temperature Control Valves on both Diesel Generators were found to be venting control air to atmosphere for abnormally long periods of time, varying from approximately 20 minutes to continuously. The air leakage caused by the Temperature Control Valves reduced the fuel rack actuating portion of the control air system by 5 psi. This ranged from approximately 30 psi to 25 psi and significantly increased start times. The pressure decrease could also cause a non-start. due to the failure of the rack to open because the contzo3 air pressure could not attain the required actuation pressure of the pneumatic 'valves required for fuel rack initiation. The low control air system pressure caused the apparent failure of Diesel Generator "A" and, in conjunction, with the single air receiver start, caused the slow start of Diesel Generator "B". A non-start was simulated on Diesel Generator "A" by observing that the fuel rack failed to actuate with the Temperature Control Valve venting to atmosphere but actuated immediately when the valve vent hole was plugged. The difference between the non-start of Diesel Generator "A" and the slow start of Diesel Generator "B" was due to the difference in the 30 psi regulator and fuel rack pneumatic valve setpoints.

Considering the regulator and valve setpoint tolerances, the difference between a Diesel Generator start and a non-start could be as small as 3 psi in the fuel rack actuation portion of the control air system.

The Temperature Control Valve problem was probably a contributing factor and possibly the major cause of earlier reported Unit 1 Diesel Generator slow starts and non-starts (LER 85-018, ;dated 4-4-85.&: 6-10-85; LER 85-019, dated. 5-.25-85', LER 85-'026; dated 4-29-85)

The intermittent nature of the start failures, which could not be repeated after the engine had been run, can now be explained. as follows: The control air leakage is due to slow actuation of the Manifold Temperature Control Valves which leaves the valve in an intermediate position, venting control air to atmosphere'. When the engine has been in standby for an extended period of time, such as the interval between surveillance tests, the Temperature Control Valves actuate more frequently to maintain the Turbocharger Manifold temperature at 105'F by regulating the 140'F Jacket Water flow thru the Intercooler box. Due to the slow actuation, the Temperature Control Valves are left venting control air a large portion of the time. In addition, when the Spray Pond Pump is started in preparation for a surveillance test start, the Inter-cooler temperature drops, since it is cooled by the Spray Pond water, causing the Temperature Control Valves to actuate. The valves take 20 minutes or longer to actuate and the Diesel Generator is normally started within 10 minutes after the Spray Pond Pump is started, therefore, the probability is high that the Temperature Control Valves are in mid-actuation, venting control air to atmosphere, when the DG start signal is initiated. Start attempts following a slow start usually appear normal because the Turbocharger Intercooler box is in excess of 105'F and no Temperature Control Valve actuation is required as well as the engine being in a preferred start condition after recently being exercised.

Cooper Energy Services (CES) was contacted on May 13, 1985, regarding the single air start criteria. The CES representative verified that the Diesel Generators 'eceiver were designed to start on one air receiver in 10 seconds. The Diesel Generator B slow start was earlier attributed to the single air receiver start by the CES onsite repre-sentative and declared operable on that premise. As a result of the new information obtained from CES and the observed Temperature Control Valve slow actuation which reduced control air pressure by approximately 5 psi, the conclusion reached by Operations Engineering was that the single receiver start in conjunction with the control air leak through the Temperature Control Valves caused the slow start. This is a justified evaluation since either the single air receiver start or the Temperature Control Valve air leakage can be shown to cause slower start times and with both conditions occur'ring simultaneously, the start time could easily exceed the 10 second criteria.

<l jf

~

.'.<,Special .

" Diesel Ge tor Failure to Start Page 3 of -.

Both Diesel enerator start failures occurred as a result of the same mechanical valve failure which appears to be a generic design problem since Unit 2 has experienced the same valve failures during Start-'up testing. These two start failures were essentially due to the same cause and should be counted as a single failure on a per nuclear unit basis, thus allowing Unit 1 to return to a l4 day test interval per R.G.1.108.C.2.d.(2 ).

This complies with NRC recommendations to reduce the number of "Fast Starts" (Ref.

IE Notice 84-15) since a 7 day test interval would serve no useful purpose in this case, and would potentially reduce the Diesel Generator reliability by increasing the probability of unrelated component failures.

The following corrective action will be implemented:

l. Engineering will evaluate the 30 psi pressure setpoint for J-DGA-PCJ281 and J-DGB-PCV282, to the control air system reliability, considering the fuel rack (Mechanical Linkage that controls fuel supply to the diesel generator injectors) pneumatic valve setpoints and tolerances.

2., Pressure Indicators J-DGA-PI273, J-DGB-'PI274",' J-DGB-PI276 'will'e added to the Operator's Log for surveillance during Standby operation of the Diesel Gener-ators to ensure adequate control air pressure for the fuel rack actuation.

3. Cooper Energy Services was contacted for recommendations to increase the reliability of the Manifold Temperature Control Valves and/or decrease their effect on the control air system pressure.

~

$P

~ ~

I