ML20198S101

Special Rept:On 860505,diesel Generator 2A Declared Incapable Due to Contaminated Fuel & Broken Rocker Arms on Seven Left Cylinder.Caused by Bent Retaining Rings & Insufficient Generator Crosshead Procedure.Parts Replaced

ML20198S101
Person / Time
Site:	Byron
Issue date:	05/29/1986
From:	Querio R COMMONWEALTH EDISON CO.
To:	James Keppler NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION III)
References
BYRON-86-0541, BYRON-86-541, NUDOCS 8606100328
Download: ML20198S101 (4)

Text

.

^ 8:]/18-rN Commonwealth Edison Byr:n Nuciar St: tion

/ 4450 North GIrrnIn Church Rold FMCMTY EfJilNG

( 1 Byron, Illinois 61010 f

%f/ ./ __.L_J .L_

O v 98)f7L n i ql s

O(' a May 29,1986 PA FILS LTR: BYRON 86-0541 FILE: 2.7.400 Mr. James G. Kcppler Regional Administrator '

Region Ill U. S. Nuclear Regulatory Commission Glen Ellyn, Illinois 60137

Dear Sir,

This Special Report is submitted in accordance with Byron litdtion License ,

NPF-37 Appendix A, Technical Specification 4.8.1.1.3, reporting of Diesel Generator Failures.

This report addresses one valid test and failure experienced on the 2A Diesel Generator. The criteria used to determine valid tests and failures is in eccordance .

with Section C.2.e of Regulatory Guide 1.108.

t During an extended run of the 2A Diesel Generator to support Unit 2 System Auxiliary Transformer switching, an increase in crankcase oil level was observed.

Sampling indicated that diesel fuel oil had contaminated the crankcase oil.

All other operating parameters!were normal. The Unit 2A Diesel Generator was declared incapable at 0415 on May 5,1986, and Limiting Condition for Operation Action Requirement (LCOAR) 8.1.3-la was initiated. Investigation into the fuel oil leak identified a loose fitting on the fuel oil return line connection on the eight right cylinder. In addition, on May 7,1986 broken rocker arms were discovered on the seven left cylinder. It is believed that the rocker arm failure occurred during the May 5th run.

On February 12,1986 and February 17,1986 similar failures were experienced on the five left and seven left cylinders respectively. Cooper Bessemer, the diesel engine vendor, concluded that the root cause for these failures was a misalignment of the cylinder head and the cylinder liner. This misalignment in turn caused abnormal side loading and deformation of the crosshead guide which eventually lead to the rocker arm failure. Allindications from the May 5th failure show that the root cause for this failure is not the same as the root cause for the February 12th and 17th failures. Following the repairs for the February 12th and 17th failures, and prior to the May 5th failure, the 2A Diesel Generator ran for 67 hours7.75463e-4 days <br />0.0186 hours <br />1.107804e-4 weeks <br />2.54935e-5 months <br />.

8606100328 860529 4 DR ADOCK 0500 g g 5 0

j(0 \

e .

BYRON 86-0541 Page 2 Cooper Bessemer was brought on site to assist in the investigation of the current failure. The following actions were taken.

1. The seven left cylinder head was removed and inspected for physical deformation. Special emphasis was placed on correct clearances between the crosshead (camfollower) and its guide due to the previous failures of this kind. In addition the cylinder liner was checked for proper alignment. Three abnormal conditions were detected.

1) One lifter on the exhaust rocker arm was found collapsed.

2) Indentations were found on the connection of the intake crosshead and its pushrod.

3) Bent retaining rings were found on both the intake and exhaust rocker arms.

2. The five left and ten right cylinder heads were removed due to similar failures experienced in February,1986. A similar inspection to that of the seven left cylinder was excecuted. Again, bent retaining rings were discovered on the intake and exhaust rocker arms.

3. Cam shaft covers for all twenty cylinders were removed to verify proper clearance between all crossheads and their respective guides.

In addition, all rocker arm assembly covers were removed to inspect the rocker arm assemblies. Bent retaining rings were discovered on 14 of 20 rocker arm assemblies.

4. A Commonwealth Edison Systems Material Analysis Department (SMAD) representative was brought on site to investigate a possible material defect. Initialindications revealed that the exhaust rocker arm, manufactured by JG Corp. had a lower Brinell hardness than the intake rocker arm, manufactured by Cooper Bessemer. This condition was also noted in the two previous rocker arm failures dated 2/12/86 and 2/17/86.

Byron Station, in conjunction with Cooper Bessemer, identified four possible root causes for the failure root cause. The four possible root causes include;

1) collapsed lifter

2) crosshead indentations

3) weak rocker arms, and

4) bent retaining rings.

Cooper Bessemer indicated that one collapsed lifter could not have independently caused the rocker arms to fail. The indentations at the union of the pushrod and the crosshead were proven to have taken place after the failure of the rocker arms. Additionally these indentations had no effect on the clearance between the crosshead and its guide. Although SMAD did identify a lower Brinell hardness in the exhaust rocker arm, the minimum safety factor for the weaker rocker arms was in excess of 10.

BYRON 86-0541 Page 3 The intermediate root cause for this event is the bent retaining rings. The function of a rocker arm retaining ring is to hold the tappet inside the rocker arm during disassembly of the rocker arms. The tappet is the intermediate piece between the valve stem and its respective rocker arm. The lifter, which provides a dampening effect during valve operation, is housed by the tappet.

Although a normal retaining ring has no operational function, the bent rings obstructed the pathway for the tappet during valve adjustment. This in turn created a gap between the valve stem and the rocker arm. In addition, the collapsed lifter on the seven left cylinder increased the gap between the valve stem and the rocker arm. This gap caused the failure through either constant impact loading, or by allowing a pushrod to travel out of its normal position.

With the pushrod out of position, the rocker arm can over extend, creating high stresses within the rocker arm.

The root cause for this event is the lack of detailed procedure for the inspection of the Diesel Generator crossheads. Following the corrective action for the February 12 and 17 failures, the crossheads were inspected. During the inspection the valve adjusting screws, which position the tappet in the correct position, were backed out of their taps. The crankshaft was then rolled using the turning gear. This was done to check the crossheads for full freedom of movement.

Ilowever, by using this method, the tappets were forced against the retaining rings causing deformation of the rings. This deformation, as explained in the previous paragraphs, led to the failure of the seven left cylinder rocker arms.

It must be noted that failure of one rocker arm, on a given head, willinevitably cause the failure of the other rocker arm. Therefore, finding two failed rocker arms on a given head is expected.

The root cause for the contaminated lube oilis the broken rocker arms. When the rocker arms break, combustion no longer takes place in the cylinder. Ilowever, fuel oil is stillinjected into the cylinder. When the piston strokes, the fuel oil is compressed between the cylinder head and the piston. The fuel oil eventually leaks by the piston rings and enters the crankcase. Although a smallleak was found on a fuel drain line, this leak was minimal in comparision to the fuel leak by the piston rings.

The following corrective actions were taken to return the 2A Diesel Generator to capable status. A Cooper Bessemer Representative assisted in all reassembly procedures.

1) The loose fitting on the fuel oil return line was tightened to prevent lube oil contamination. In addition all 20 fuel oil injectors were replaced to isolate any undetected leakage.

2) The retaining rings on all 20 cylinders were removed and replaced.

3) The broken rocker arms and the collapsed lifter on the seven left cylinder were removed and replaced.

4) At 0302 on May lith,1986, the 2A Diesel Generator was started for an operational surveillance. At 0445, after successful completion ,

of the one hour run, LCOAR 8.1.3-la was exited.

. (

BYRON 86-0541 Paga 4

5) At 1013 on May lith, the 2A Diesel Generator began a 99 hour0.00115 days <br />0.0275 hours <br />1.636905e-4 weeks <br />3.76695e-5 months <br /> fullload run. At 1300 on May 14th the 99 hour0.00115 days <br />0.0275 hours <br />1.636905e-4 weeks <br />3.76695e-5 months <br /> run was successfully completed. All previous rocker arm failures Imve occurred in less than 68 hours7.87037e-4 days <br />0.0189 hours <br />1.124339e-4 weeks <br />2.5874e-5 months <br />, therefore this test verified rocker arm integrity.

6) The Diesel Generator Operational Procedures have been temporarily changed to alert the operator on how to detect rocker arm failure, and the proper actions to take if failure occurs.

7) Cooper Bessemer is currently working on detailed procedures for crosshead inspection. Crosshead Inspections will not be performed until the procedure is received.

The 2A Diesel Generator was unavailable for 6 days following the May 5th failure. This failure will have no effect on the testing frequency of the 2A Diesel Generator. Currently Technical Specification 3/4.1.8.3 applies to the 2A Diesel Generator. This specification has no requirement for altering testing frequencies due to valid tests and failures, therefore Section C.2.d of Regulatory Guide 1.108 is not applicable. This failure will be recorded for future use in determining the testing frequency upon acquisition of the Unit 2 Operating 6 License, blVS k R. E. Querio Station Manager Byron Nuc! car Power Station REQ /MMR/bf ec: J. Ilinds, NRC Resident inspector U. S. NRC Document Control Desk, Washington, D.C.

INPO Record Center CECO Distribution List M. Robert J. Cook

1. 03/059

-. _ - _ - _ .