ML20116L685

Special Rept 92-008:on 921014,emergency Diesel Generator 11 Failed Due to Water in Cylinder.Caused by Crack in Cylinder Head.Head Replaced & EDG 11 Retested

ML20116L685
Person / Time
Site:	Grand Gulf
Issue date:	11/13/1992
From:	Cottle W ENTERGY OPERATIONS, INC.
To:	NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
References
92-008, 92-8, GNRO-92-00139, GNRO-92-139, NUDOCS 9211190071
Download: ML20116L685 (3)

Text

r-M N =# ENTERGY E"#~"'""*'

Rvt G tsort MS 391f0 Id M1437 600 I

November 13, 1992

(("*

we, h 3f'ai b Y !i,({~W ! - 7 U.S. Nuclear Regulatory Commission Mail Station F1-137 Washington, D.C. 20555 Attention: Document Control Desk

Subject:

Grand Gulf Nuclear Station Unit 1 Docket No. 50-416 License No. NPF-29 Special Report 92-008 Valid Failure of Emergency Diesel Generator 11 Due to Water in Cylinder GNRO-92/00139 Gentlemen:

In Decer5er 1986, the Commission issued Amendment 26 to the GGNS Operating Ljcense. This amendment adtu maintenance and surveillance requirements for the TDI EDGs. The surveillance requirement specified for the cylinder heads was that an air roll be performed at specified times to ensure water leakage into a-cylinder would be detected and the' appropriate action taken to ensure diesel reliability. This requirement is implemented through approved plant procedures.

On October 14, 1992, Operations personnel performed a venting procedure (air roll) on Emergency Diesel Generator (EDG) 11 in accordance with Section 4.2.2.a of system Operating Instruction 04-1-01-P75-1.

During the air roll of EDG 11, plant perV7nel observed _ water coming from the Number 2 Right Bank cylinder stopcock. The air roll I was_ performed twice more with the same res.ults.

A work order was generated to replace the head and retest EDG 11.

The cracked head was replaced and' the diesel was retested.

satisfactorily.

Another work order was generated to perform non-destructive examination inspections on the cylinder head to determine if the head was actually cracked. The results of the examination revealed a crack - (approximately three inches in length) in the top left exhaust port of the Number 2 Right Bank cylinder head.

., 4 o -

921.190071 921113 4 -.

PDR ADOCK-05000416 g f4 l D; S PDR ,

g . . . _ , - - -- -- _ ._

• . j November 13. 1992

.GNRO*-9 ?/ 0013 9 Page 2 of 4 The cause of the crack is believed to be manufacturing defects in the head. Based on operating experience of cylinder heads cast before October 1978 for the Transamerica Delaval, Inc. Enterprise (TDI) inline and V-type R-4 series engines, plant personnel were aware that heads of this type, presently installed on EDG 11 and 12, were not stress-relieved and are susceptible to fatigue crack growth in thin sections and/or from fabrication-induced defects.

This information was provided in a report by Pacific Northwest Laboratory in December 1905.

The cylinder head that was cracked on EDG 11 was manufactured January 1975; therefore it is categorized as a Group I head. Based on failure of heads for the TDI R-4 engines, TDI made changes in manufacturing procedures and in quality control, as well as minor design changes. Those cast prior to October 1978 are Group I; those cast between October 1978 and September 1930 are Group II; and those cast after September 1980 are Group III.

Failure history of TDI R-4 series engines indicates that Groups I and II heads were - subject to core shift, inadequate control of Palidification, and inadequate control of the stellite valve seat weld deposition process. The Group III heads are less prone to manufacturing defects. GGNS presently has some Group III heads installed, however Group I heads are still in use.

During the air roll, a condition (e.g., water in a cylinder) s is identified that could have resulted in the failure of the diesel generator unit during test or during response to a bona ' fide-signal. The amount of water discovs ed in the cylindar would not have caused a failure. However, if tne leak had gone undetected, a sufficient amount of water could have accumulated in the cylinder to cause failure of the EDG, Therefore this occurrence is considered a valid failure of the EDG unit. This was the first valid failure in the last 20 valid tests and the fifth valid failure in the last 100 valid tests. In accordance with Technical Specification Table 4.8.1.1.2-1, the surveillance frequency for EDG 11 was increased to once per 7 days.

Yours truly, w r emic l

WTC/RR/

l cc: (See Next Page; l

l

I

>-,.. .r 13, 199L

.GitRG -92/ 0013 9 Pago 3 of 4 cci Mr. R. 11 . Bornhard Mr. D. C. Ilintz Mr. R. B. McGoheo Mr. H. S. Reynolds Mr. !!. L. Thomas Mr. Stewart D. Ebneter Regional Adminir,trator U.S. liuclear Regulatory Commission Rogion II 101 Mariotta St. , 11.51. , Suito 2900 Atlanta, Georgin 30' 23 Mr. P. W. O'Connor (w/2)

Of fico 7f flucl ear Reactor Regulation U.S. Iluclear Rog'11atory Commtsalon Mail Stop 13113 Washington, D.C. 20555

_ - - _ - _ _ - _ _ _ - - _ _ - _ - -