ML20083D443

Special Rept SR-91-002-01:on 910318,emergency Diesel Generator a Experienced Crankcase Overpressurization.Caused by Stuck Piston Rings & Improper Temp Differential.Cylinder Liner,Piston Assemblies & Rod Bearings Replaced

ML20083D443
Person / Time
Site:	Waterford
Issue date:	09/23/1991
From:	Packer D ENTERGY OPERATIONS, INC.
To:	NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
References
SR-91-002-01, SR-91-2-1, W3B5-91-0268, W3B5-91-268, NUDOCS 9110010005
Download: ML20083D443 (7)

Text

_ . ._ _.. _

.g T Entergy 5 "' m on " a " '"<-

. Operations W3BS 91-0268 A4.05 QA September 23, 1991 U.S. 14uclear Regulatory Commission ATTEll110N: Document Control Desk Washington, D.C. 20555 SUBJEC'i : Waterford 3 SES Docket No. 50-382 License No. NPF-38 Reporting of Special Report Gentlemen:

Attached is Special Report Number SR-91002-01 for Waterford Steam Electric Station Unit 3. This Special Report is submitted per 10CFR50.36(c)(2) and Technical Specifications 4.8.1.1.3 and 6.9.2.

Very truly yours,

(~~  ;)

OMAG P D.F. Packer General Manager Plant Operations DfP/DDW/rk Attachment cc: Messrs. R.D. Martin G.L. Florreich J.T. Wheelock - INPO Records Center E.L. Blake N.S. Reynolds NRC Resident inspectors Office 0110010005 910 C --)

FDR ADOCK0500((. J t .

Special iteport 91 002-01 Valid inilure of Emergency Diesel Generat or ' A' due to overptessurization of the crankcase.

4 1H11ME11W On 14 arch 18, 1991, Wat eriot d Stenin I: lect ric Stationi Unit 3 was in cold shutdown when 1:roergency Diesel Generator (LDG) 'A' expericined a crankense ove :T r e t.s u r i z a t i c.n . The CDG was being run in accordance with the pret equisit es of the 18 suonth Emergency Diesel Engine Inspection proceduse, M 003 015, when the overpressurization occursed. The operat or obset ving the I:DG run depressed the emet ge-ncy stop but ton and pulled the overspeed t rip.

This, event has been classifled as a valid diesel failure and is being reported in accordance with Technical Specification 4.b...).3.

The root cause of this. event in stuck piston rings. A possible root cause, based on vendor bulletin informatlon, vos an in: proper temperature -

dif ferential between jacket cooling vator and lube oli teitperatures. The lube oil t o jacket water temperatut e differential han been eliminated as a root cause. A related possible contributing cause was an omission of relevant.

information in that the vendor did not. provide a service news bulletin which specifically identified the posulble adverse consequences of the previously-mentioned temperature differential. This possibic contributing cause is now considered a noncausal factor. -The piston and piston liner have been replaced and t he titproper temperature dif ferential will be corrected.

4

! NARRATIVE At 0457 hours0.00529 days <br />0.127 hours <br />7.556217e-4 weeks <br />1.738885e-4 months <br /> on March 18, 1991. Vaterford Steam Electric Station Unit 3 was in cold shutdown when Emergency Diesel Generator (EDC) (Ells Identifier EK ENG) 'A' experienced a crankcase (EIIS Identiller BLK) overpressurization.

Personnel operating the EDG depressed the emergency stop push button and pulled the overspeed trip after the overpressure condition occurred. This event has been classified as a valid diesel failure and is reported in accordance with technical Specification 4.8.1.1.3.

EDG 'A' was started at 0221 to connence the prerequisite 4 hour4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> continuous run of the 18 Month Emergency Diesel Engine Inspection procedure, XM 003-015. The purpose of the run was a pre maintenance heatup and engine analysis prior to placing the unit out of service for the EDG ' A' refueling outage inspection. EDG 'A' was loaded to 4400 kilowatts when the crankcase i overpressuriration occurred. All 10 cylinder relief assemblics lifted, filling the room with oil vapor. The room was exited and there were no personnel injuries. After the vapor cleared, the room was entered and a crankcase oil sample was drawn. This oil sample was found to be within specifications. An on site fuel dilution test also yielded satisfactory results.

An inspection of the EDG cylinders, pistons, rods (Ells identifier ROD),

and bearings was commenced as required by MM 003-015. The inspection revealed that the 5 left (SL) cylinder liner (Ells Identifier LNR) exhibited e intermittent 1 1/2 by 3 inch longitudinal rough spots or scuff marks perpendicular to the piston pin. Vertical scratch marks were noted throughout the length of the cylinder. Additionally, the metal on the underside of the piston had been discolored by heat. A borescope inspection of the other pistons indicated satisfactory results. Disassembly and inspection of four

I other pistons, SR, IL, IR and 8L, revealed normal wear.

Of particular interest was a missing piece of chromo irom the $L piston cylinder liner. 'the missing chrome flake is approximately 1/4 inch in diameter. The 5L cylinder liner was analyzed at the vendor's n.etallurgical lab to determine the cause and consequence of the missing chrome.

Cooper Bessemer Service News Bulletin ntunber 688, dated June 76, 1985, addressed the possibility of " cylinder scuffing, piston seizure and... crankcase explosion" if jacket cooling water temperature was not maintained 5 degrees Fahrenheit higher than piston luhe oil tempetature.

Waterford 3 had not received this bulletin f rom the vendor, but a consultant who was familiar with bulletin 688 noted that Watertord 3 EDG logs showed that jacket cooling water temperature was being maintained 4 degrees lower than piston lube oil temperature. Maintaining the reconnended temperature dif ferential would reportedly prevent a crankcase explosion caused by frictional heat resulting from scuffing of the cylinder in the piston skirt area. The piston is cast steel with a thin outer coating of tin. The piston cylinder wall is coated with chrome. Although the reasons are not known, there is empirical industry evidence that indicates that an improper ten:perature differential causes the cast steel in the piston skirt area to become exposed. This cast steel surface then scuffs the chrome cylinder wall, eventually generating enough frictional heat to cause a crankcase explosion.

The fact that this temperature differential condition had existed at Waterford 3 on both EDGs for several years with no deleterious cffects and the fact that none of the other EDG 'A' cylinders or EDG 'li' cylinders had scuf f marks cast s doubt that the improper temperature differential is the root cause. Based on determinations made by the vendor and an outside consultant, the improper temperature differential has been eliminated as a root cause. A noncausal

4 factor is an omissien of relevant information in that Cooper Bessemer Service News Bulletin Number 688 was not received at Waterford 3 from the vendor.

This failure is the second failure during a valid test of EDG 'A' at Waterford 3 since the operating license was issued. There has also been one failure during a valid test of EDG 'B'. The jacket water and lube oil thermostatic valve thermo elements have been replaced and the teru1 ting temperature differential is zero. Establishing a lube oil temperature which is lower than jackat water temperature will require utilizing a different lube oli thermo element. A lower temperature lube oil thermo element has been requested from the vendor and the vendor will provide these thermo elements pending completion of their evaluation and procurement process. To address the noncausal factor, initial corrective a, tion consisted of coordinating with the Cooper Bessemer Nuclear Users Group to ensure that Waterford 3 now has all applicable Cooper Bessemer Service News Bulletins. To ensure all applicable Cooper-Bessemer Service News Bulletins are received in the future, the system engineer will receive Cooper Bessemer Service News Bulletins from the Cooper-Bessemer Nuclear Users Group. Additionally, the vendor has a designated point of contact for all nuclear users who will be used to closely track the bulletins.

A review of industry experience revealed that crankcase over.

+

pressurizations at other plants have been attributed to metal cuttings in piston. pin oil grooves,-improper pisten pin fit up, piston bushing failure, lube oil pump bearing failure, a loose piston pin bolt, and water leaks.

Additionally, four crankcase overpressurizations attributed to piston to-liner seizures have indeterminate root causes. All of these root causes have been systematically eliminated during the root cause investigation. Specifically, no metal cuttings were found in EDGA'. The vendor determined that although .

w. v---r, -r #

e * ,

two piston bushings needed to be reconditioned, the piston bushings could have been used in their previous condition. An improper piston pin fit.up would have caused a piston seizure soon after the initial EDG start. The EDG lube oil pump functioned satisfactorily. No loose piston pin bolts were noted and the crankcase oil sample was within specification.

EDG 'A' was declared operable at 1930 on April 12, 1991. The followind items were replaced during the refueling outage: 5L cylinder liner and piston assembly, SR piston assembly, IL piston assembly, and 5 and 1 rod bearings.

The IL and SR pistons and bushings were reconditioned.

Af ter further analysis of the cylinder SL components associated with the overpressurization incident, the vendor concluded that the overpressurization was caused by stuck piston rings. Deposits built up behind the rings until the rings were no longer free to move. Additional deposits forced the compression rings against the liner and significantly reduced lubrication.

The lack of lubrication resulted in ring chatter and increased surf ace temperatures on the piston and liner. As temperatures increased, the chrome plating on the liner cracked and began spalling off. A piece of chrome plate debris lodged between th niston and liner. The resilting friction caused the piston area around tr icbris to become incandescent, the incandescent area Ignited the crankcase oil vapor causing the crankcase overpressurization (REF:

Cooper Energy Services Materials Laboratory Report ID#1345).

Based on the inspection results of the other cylinders, the stuck ring condition on cylinder SL is considered to be an isolated incident. To ensure that no incipient failure mechanisms are active, however, EDC 'A' will be inspected during the next refueling outage. During this inspection, at least six pistons will be pulled from a sampling of cylinders. Liner chrome thickness and piston tin thi.kness will be measured to determine if abnormal l

1

wear is occurring. Piston rings will be checkad for residue buildup and debris. All cylinders from which the pistons are not removed will be inspected by borescope. Additional pistons will be pulled as necessary based on inspection results. Vaterford 3 intends to have vendor experts assist with the inspections of both EDO 'A' and EDG 'B' to help ensure a most thorough evaluation.

The current surveillance test interval for EDG 'B' is at least once per 31 days. The current surveillence test interval for EDG ' A' is once every seven days. These surveillance test interv.ds are in conformance with the schedule of Regulatory Position C.2.d. The reankcase relief assemblies functioned as designed and EDG 'B" remained operable theoughout the time (nat EDG 'A' was out of service therefore, this event did not present a hazard to the health and safety of the public.

Plant

Contact:

W.R. Brian, Plant Engineering Superintendent, 504/464 3127 l

l l

l

-