Text

Reliability Improvement of Medium Speed Diesels in Nuclear Standby Applications, Presented at 860512-14 Marine Propulsion Technology Conference in Washington,Dc
	ML20214E060
Person / Time
Issue date:	05/12/1986
From:	Berlinger C, Murphy E; NRC
To:	;
Shared Package
ML20214D670	List: ML20214D667; ML20214E060; ML20214E265; ML20214E423; ML20214E480; ML20214E485;
References
	FOIA-86-656 NUDOCS 8611240339
	Download: ML20214E060 (15)
	v • d • e

._

' Qg g TheEngineering Resource For M MnCfngMffl/ 400 COMMONWEALTH ORIVE WARRENDALE. PA 15096 SAE Technica Paper Series 860887 l

l Reliability improvement of Medium Speed Diesels in Nuclear Standby Applications Carl H. Berlinger and Emmett L. Murphy United States Nuclear Regulatory Commission Marine Propulalon Technology Conference Wachington. D.C.

May 1214,1986 g 12 g 9 861110 ELLI596-656, PDR a/3-

o The appearance of the code at the bottom of the first page of this paper indicates SAE's consent that copies of the paper may be rnade for persona! or internet use, or for the personal or internal use of specific Clients. This consent is given on the con-dition. however, that the copier pay the stated per article copy fee through the Copyright Clearance Center, Inc., Operations Center, P.O. Box 765. Schenectady,

N.Y.12301, for copying beyond that pertraitted by Sections 107 or 108 of the U S.

Copyright Law. This consert does not extend to other kinds of copying such as copying for general distribution, for advertising or promotional purposes, for creating new collective works. or for resate. ,

Papers published prior to 1978 may also be copied at a per paper fee of $2 50 under the above stated conditions SAE routinely stocks printed papers for a period of three years following date of publication. Direct your orders to SAE Order Department.

To obtain quantity reprint rates, permission to reprint a technical paper or per.

~ mission to use copyrighted SAE publications in other works, contact the SAE Publications Division.

o -- =_-

se . sat : M WA'5~

p3==x= h ;s T n 55fassar mmmewarwamomme

,a sa a se sees, n a.-

WWWW ses C,Lwai. isoasuTV Dai444sf I

No part of this publication may be reproduced in any form,in .

l an electronic retrieval system or otherwise, without the prior written permission of the publisher.

ISSN0148 7191 Copyright C1986Societyof AutomotiveEngineers,inc.

This paper is subject to revision. Statements and opmions ad- Persons wishes to submit papers to be considered for pre.

vanced in papers or discumon are the author's and are his sentation or pubbcation through SAL should send the manu-responsibdit), not S Al's.however the paper has been edited script or a 300 mord abstract of a proposed manuscript to by SAI for uniform st)hng and format Discussion u di be Secretar) . I ngineering Activity lloard, S AI .

pnnted with the paper if it is pubbshed in SAL Transactions.

I or permission to pubbsh this paper in fun or in part. contast Prmted in U.S. A.

the S AL Pubbeations Division.

860887 Reliability improvement of Medium Speed Diesels in Nuclear Standby Applications Carl H. Beilinger and Emmett L. Murphy United States Nuclear Regulatory Commission ABSTRACT In response to these problems, 11 (now 13)

U.S. nuclear utility owners formed a TDI Emergency standby diesel generators Diesel Generator Owners Group to address manufactured by Transamerica Delaval, Inc. operational and regulatory issues relative experienced a number of major problems to diesel generator sets used for standby during preoperational qualification testing emergency power. On March 2, 1984, the at several U.S. nuclear sites. Most notably Owners Group submitted a proposed program these have included a complete fracture of to the NRC which, through a combination of a crankshaft, an engine block failure, piston design reviews, quality revalidations, failures, and cracked and leaking cylinder engine tests and component inspections, was heads. These problems appear to stem from intended to provide an in-depth assessment deficiencies in design and manufacturing of the adequacy of the respective utilities' quality by the engine manufacturer. This TDI engines to perform their safety related paper discusses some of the more significant function [1]*.

problems experienced and actions taken by the The Owners Group Program Plan involved nuclear utility owners and the NRC to the following major elements:

reestablish confidence in the reliability of 1. Phase I: Resolution of known these engines and to qualify these engines generic problem areas intended for nuclear service. by th2 Owners Group to serve as a basis for the licensing of plants during the period prior to comple-j tion of Phase II of the Owners l

DURING THE 1970s, many utilities ordered Group Program.

diesel generators from Transamerica Delaval, 2. Phase II: A design review / quality Inc. (TDI) for installation at nuclear plants revalidation (DR/QR) of a large in the USA. The first of these engines to set of important engine components become operational at an operating plant were to assure that their design and at San Onofre Unit 1 in 1977. However, manufacture; including specifica-nuclear plant operating experience with TDI tions, quality control and quality Emergency Diesel Generators (EDGs) remained assurance and operational very limited until preoperational test surveillance and maintenance, are programs were commenced at Shoreham and adequate.

Grand Gulf Unit 1 in the early 1980's. . 3. Expanded engine tests and inspec-Concerns regarding the reliability of tions as needed to support Phases I large bore, medium speed diesel generators and II.

manufactured by TDI for application at Under Phase I, the Owners Group performed domestic nuclear plants were first prompted a comprehensive review of operating history by a crankshaft failure at Shoreham in of TDI Series DSR-4 engines in both nuclear August 1983. However, a broad pattern of and non-nuclear service for purposes of deficiencies in critical engine components identifying significant problem areas. The subsequently became evident at Shoreham and Owners Group evaluated the causes of these at other nuclear and non-nuclear facilities problems and issued recommendations to the employing TDI diesel generators. These individual owners concerning actions they deficiencies stem from inadequacies in design, manufacture and quality assurance /

Numbers in brackets designate references ,

quality control (QA/QC) by TDI. at end of paper.

2 860887 should take to resolve these problems contributed to the enhancement of the including needed component upgrades or reliability of these engines in nuclear modifications, component inspections, and standby applications.

engine tests.

Phase II of the Owners Group Program 1.0 SOME SIGNIFICANT PROBLEM AREAS AND proceeded beyond known problem areas to THEIR RESOLUTION systematically consider all components (approximately 150 to 170 component types 1.1 CRANKSHAFTS FOR TDI MODEL DSR-48 per engine) important to the operability and ENGINES - TDI Model DSR-48 engines (used at reliability of the engines. Phase II was Shoreham and River Bend) are 8 cylinder intended primarily to ensure that significant inline engines with a 3500 kw nameplate new problem areas do not develop in the rating and a 3900 kw 2 hour2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months overload rating. '

future due to deficiencies in design or The Shoreham crankshaft failure in August quality of manufacture. The Owners Group 1983 occurred in the emergency diesel performed the Phase II design reviews and, generator (EDG)-102 engine during a two ,

as was the case for Phase I, recommended hour overload test at 3900 kw. At the time needed component upgrades and modifications of the failure, the affected engine had been and component inspections to validate quality run for a total of 671 hours0.00777 days 0.186 hours 0.00111 weeks 2.553155e-4 months , including 254 of manufacture and/or assembly. A major hours at loads 2 3500 kw and 19 hours2.199074e-4 days 0.00528 hours 3.141534e-5 weeks 7.2295e-6 months at element of the Phase II Program was the loads 2 3900 kw. Crankshafts in the preparation of a comprehensive engine Shoreham EDG-101 and EDG-103 engines were maintenance and surveillance program to be subsequently inspected and also found to

~

implemented by the individual owners. c ntain sev m cracks.

The staff concluded that the Owners Subsequent investigation by Failure Group Program Plan incorporated the essential Ana s s Ass ates, Inc. UaAA), a elements needed to resolve the outstanding consultant for the subject utility and also later for the Owners Group, revealed the concerns relating to the reliability of the failures to be fatigue related, caused by TDI EDGs for nuclear service [1]. The staff torsional vibration. Independent analyses has completed its final evaluation of the performed by FaAA established that the Owners Group findings and recommendations crankshaft had been overstressed relative stemming from this program and will be to the Diesel Engine Manufacturers publishing its final safety evaluation report Association (DEMA) standards [2].

shortly. In the interim period, prior to The original crankshafts at Shoreham completion of its overall review, the staff that had 11 inch diameter crankpins with 1/2 concluded that issues warranting priority inch fillets were replaced with new attention had been adequately resolved at crankshafts having 12 inch diameter crankpins several plants such that the TDI EDGs would with 3/4 inch fillets. Independent analyses provide reliable service through at least the performed by an expert consultant to Pacific first refueling outage (by which time the Northwest Laboratories (PNL) on behalf of the staff will have completed its overall review). NRC indicated that the crankshafts did not This finding permitted the staff to proceed meet DEMA at 3500 kw for combined orders. PNL with issuance of operating licenses for these and the NRC staff concluded that there was plants and has generally been based on (1) insufficient evidence to either approve or actions taken by the Owners Group and the disapprove the replacement crankshafts for individual owners to resolve known problem operation at engine loads at or above the areas, (2) implementation of an acceptable 3500 kw nameplate continuous rating. However, engine maintenance and surveillance program, PNL and the NRC staff concluded that unlimited and (3) incorporation of plant Technical fatigue life for the crankshafts could be Specification requirements and operating demonstrated by testing one of the Shoreham procedures which ensure that the engines will engines for 107 engine stress cycles (about not be operated in an overstressed condition. 750 hours0.00868 days 0.208 hours 0.00124 weeks 2.85375e-4 months ). That test would be conducted at Section 1 of this paper focuses on a load at or above the maximum emergency several of the known problem areas considered service Icad which could be placed on the .

under Phase I of the Owners Group Program engines during a design basis event. The and describes how these problems have been test load would be designated the " qualified resolved to the satisfaction of both the load" for the engine. Successful completion owners and the NRC staff. Section 2 of of such a test would be considered sufficient

this paper focuses on the role of periodic by PNL and the NRC staff to demonstrate that maintenance and surveillance in ensuring the " qualified load" is below the fatigue the continued reliability / operability of endurance limit [3].

the TDI engines for the life of the plant, In response to the NRC staff position, and also addresses certain testing and a 107 cycle test was completed for the operational considerations. Section 3 of Shoreham EDG-103 engine which established this paper focuses on several recent events 3300 kw as the qualified load level for the involving TDI diesel engines which have Shoreham engines. Subsequent NDE inspection of the crankshaft confirmed the absence of cracks at critical fillet and oil hole

3 860887 .

' locations, and provided the basis for inspected using NDE techniques sufficiently approving operation of these engines at sensitive to detect 10 mil cracks, then the loads up to 3300 kw. number of start stop sequences to propagate Although the River Bend engines were a crack from 10 to 18 alls should establish identical to the Shoreham engines, the the effective life of the crankshafts.

River Bend diesel generator set torsional Based on predicted crack growth rates, FaAA characteristics were found to be somewhat conservatively recommended that the different from those at Shoreham due to crankshafts should be inspected at intervals differences in their flywheels and of 50 start-stop sequences.

generators. Based on the 107 cycle tests The NRC staff and its PNL consultants conducted at Shoreham, the River Bend have completed their evaluation of the FaAA engines were approved for a qualified load findings and recommendations. In the interim, of 3130 kw; a load which produces comparable the staff authorized operation of San crankshaft stresses as those in the Shoreham Onofre 1 to its next refueling cycle [6]

engines operating at 3300 kw [4]. based on the fact that (1) all of the observed Plant Technical Specifications and cracks were removed during the previous engine operating procedures at Shoreham outage, (2) the crankshafts would be and River Bend have been revised to ensure reinspected at the next outage, and (3) the that the qualified load levels at the engines would not be operated above 4500 kw.

respective plants will not be exceeded The NRC staff and its PNL consultants have concluded that the San Onofre, Unit 1 in future service. crankshafts are adequate for continued

1. 2 CRANKSHAFTS FOR TDI MODEL DSRV-20 service subject to the following:

ENGINES - San Onofre 1 is the only U.S. plant with TDI DSRV-20 engines. This a) periodic inspections of main journal oil holes and fillets at each refueling model is a 20 cylinder engine in a " Vee" outage configuration with a continuous rating of a load limitation of 4500 kw 15% for 8800 kw. After over 1190 starts and 1275 b) t t g hours of running time, inspections performed yr n 9

as part of the Owners Group program revealed monthly surveillance tests linear indications in the vicinity of the oil d) routine monitoring of engine imbalance holes in various main bearing journals. The sensitive parameters should be conducted indications were subsequently removed by to minimize the possibility for either increasing the diameters of the oil sustained engine operation under those holes or by modifying the entry radius to conditions.

In addition, the staff has recommended a by FaAA, and torstograph that the licensee proceed with an evaluation testing established that the San Onofre f various options to modify the engines to crankshaft stresses were well within DEMA alleviate the crankshaft torsional stresses l allowables at rated load. However, transient that have led to crack initiation under

! torslograph testing and subsequent analysis startup and coastdown transients. The NRC revealed that, under startup and coastdown staff may require that such modifications, conditions, stresses can be developed which as necessary, be implemented if linear exceed the endurance limit and which could uture therefore lead to crack initiation. The chinsreappearduring level of stress was determined by FaAA to 1. 3 CONNECTING ROD BEARING SHELLS -

be dependent on the type of startup (rapid Connecting rod bearings in TDI Series DSR-4 starts produce the maximum stress) and the engines consist of two half-shells assembled angular position of the crankshaft prior into each connecting rod. The half-shells l

to a rapid start [5]. are fabricated from aluminum-6% tin alloy i FaAA analyses and PNL analyses performed which are electroplated on the inner surface on behalf of the NRC staff have indicated with a lead based babbit.

three closely spaced criticals occurring Inspections performed subsequent to the at 217, 240, and 264 RPM, respectively, crankshaft failure in the Shoreham EDG-102 l

which provide a possible explanation for engine revealed that one upper bearing shell the sensitivity of stress during engine from the EDG-102 engine and three upper starts and coastdowns. In this situation shells from the EDG-103 engine were cracked the vibrations initiated by the first through the thickness of the shells. One of critical could still be " ringing" when the the cracked shells from engine EDG-103 had shaft hits the next critical, and be once actually fractured into two pieces although again augmented by the third critical leading it had not affected the operability of the to large torsional vibration amplitudes. engine up to the time it was discovered.

A fracture mechanics analysis by FaAA Analyses of the failed bearing shells determined that crack depths up to 18 mils indicated that they were of the proper deep could be tolerated before the cracks composition and ultimate strength, would be subject to rapid propagation under Metallurgical and analytic evaluation

, steady state cyclic stress conditions. FaAA suggested that three factors contributed concluded that if the oil hole regions are l

4

860887 to the observed cracking: 1) the geometry engine was shutdown and subsequent inspection of the connecting rod and bearing shell was revealed additional stud to stud cracks. The such that a small unsupported length of original stud to stud crack first observed in bearing shell occurred at its extreme end; March 1984 was determined to have grown to a

2) the calculated peak oil pressure was depth of 3.9 inches. The owner elected to 29,700 psi, which exceeds the 26,000 psi replace the block for the EDG-103 engine.

commonly used in normal industrial practice; Subsequent metallurgical tests and and 3) edge loading of the bearings resulted photomicrographs established that whereas the in the concentration of the operating loads block material for EDGs 101 and 102 at on the unsupported bearing ends. In addition, Shoreham exhibited the appearance and ultimate scanning electron microscopy of the fracture tensile strength of normal gray cast iron, surface of one of the cracked bearings Class 40, the material of the original EDG-103 -

revealed voids approximately 0.020 to 0.030 block was found to be of a degenerate inches in diameter that appeared to be the (Widmanstaetten) graphite composition with initiation site for the cracks [7]. an ultimate tensile strength much inferior to '

New 12 inch diameter bearings were that of typical gray cast iron, Class 40.

installed in the Shoreham engines consistent At the Owners Group recommendation, with the 12 inch diameter crackpki journals other utility owners have also checked their of the replacement crankshafts. The new blocks for similar degenerate graphite 12-inch bearing eliminated the unsupported microstructure. To date, one engine block length of the bearing shell. Although at Washington Nuclear 1 has been found with the edge loading condition was not changed this microstructure and is being replaced. '

in the new design, the Owners Group analysis Another block at San Onofre I was also found showed that the larger 12-inch diameter to have degenerate graphite microstructure, but journal reduced the maximum tensile stress it will not be replaced because the degree to 50% of the value in the original ll-inch of contamination was much less severe and design. Stress distributions in the the engine load is restricted to 4500 kw.

13-inch bearings used in TDI Model DSRV-16-4 Based on the results of strain gage and OSRV-20-4 engines are approximated by tests and calculations using two dimensional those calculated by the Owners Group for the analytical models, FaAA has reported [8] that 12-inch bearings used in the DSR-48 engines. for material exhibiting minimum acceptable Acceptance criteria were developed by the tensile strength, initiation of " ligament Owners Croup, based on fracture mechanics cracks" is predicted to occur after analyses, concerning maximum allowable void experiencing operating hours at high load sizes in the aluminum bearings which could be and/or engine starts to high load. Ligament tolerated without degrading their fatigue cracks are not a significant concern in-and-performance. The Owners Group has recommended of themselves; however, such cracks do result that each owner perfrom a radiographic in increased stress and thus increase the inspection of all connecting rod bearings to potential for crack initiation between the ensure compliance with these criteria. stud holes of adjacent cylinders. Such " stud Application of these criterla have led to the to stud" cracks are considered to be more replacement of numerous bearing shells at a serious than ligament cracks since they can number of plants. potentially degrade the overall mechanical 1.4 ENGINE BLOCK - Cracks have been integrity of the block and its ability to reported in cylinder blocks of both TDI DSR-4 withstand cylinder firing pressures.

(in-line) and DSRV-4 (" Vee") engines in An FaAA cumulative damage analysis has nuclear and non-nuclear applications. indicated that given the existence of ligament Numerous " ligament cracks", which are vertical cracks and the absence of stud to stud cracks cracks extending between the cylinder prior to a loss of off-site power / loss of counterbore and an adjacent cylinder head coolant accident (LOOP /LOCA) event, even if stud hole, had been observed on the top a stud to stud crack were to initiate during surfaces of all three Shoreham engine blocks such an event, the crack would not propagate prior to March, 1984. In March, 1984, a " stud sufficiently during the event to impair the to stud" crack was initially observed in operability of the engine. FaAA has .

engine EDG-103 which extended vertically recommended that the absence of " stud to (from the block top) between adjacent stud stud" cracks be confirmed by eddy current holes of c.djacent cylinders to a depth of inspection subsequent to any period of

1.50 inches. operation above 50% of rated load. The In April 1984, engine EDG-103 experienced NRC staff has required that these an abnormal lord excursion while being recommendations be incorporated into the operated at full load (3500 kw). Subsequent engo maintenance and surveillance to the load excursion, engine EDG-103 was prs ; for each plant [3],[4].

sutjected to a 3900 kw overload test. At .5 PISTON SKIRTS - Piston skirts in a point less than 2 hours2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months into the test, a the two piece piston design for the TDI R-4 crack was observed to extend from a stud series engines have been undergoing an hole at the top of the block to approximately evolution since their original introduction 5 inches down the front of the block. The in 1970. This evolution has been largely ,

- __ x

8$887" 5

([1responsetoproblemsidentifiedduring b) periodic inspection of stud boss e service experience with nuclear and regions to confirm continued non-nuclear applications. adequacy.

Early TOI DSR-4 engines employed type AF TDI engines at a number of other nuclear piston skirts. In response to problems with plants were initially supplied with type AH type AF skirts relating to the use of pistons skirts. This skirt was manufactured spherical washers in the stud boss attachment from type AF casting patterns which were region, T0! introduced " modified" type AF modified to accommodate in the "as cast" skirts and type AH skirts. The modified skirt the aforementioned machining type AF skirts incorporated machining modifications to convert type AF skirts to modifications (primarily as a field retrofit) " modified" type AF skirts. Owners Group to the stud boss attachment region to permit analyses [11] indicate that type AH skirts use of a double stack of Belleville washers, may initiate cracks in the stud boss region Pistons of this design were provided by TDI under transient thermal conditions associated to a number of nuclear plants. with engine start ups prior to reaching During an early inspection at Shoreham, steady state conditions. As in the case of all " modified" type AF skirts were observed the modified type AF pistons, however, the to contain linear indications in the skirt- Owners Group predicts that any cracks will to-crown attachment bosses which were later not propagate beyond the stud boss region.

confirmed by metallurgical examination to be Nonetheless, type AH skirts at plants fatigue cracks. Similar indications were seeking near term operating licenses (Comanche later found in " modified" AF skirts at Peak Unit I and Perry Unit 1) have been Grand Gulf Unit 1. Experimental and replaced with the improved type AE skirts analytical evaluations by the Owners Group as a conservative measure.

~

indicated that although fatigue cracks may Another piston design, type AN, has initiate if the engines are operated near been found by the Owners Group to be unsuited full rated load, the cracks will not continue for nuclear standby service [11]. Although to grow after they have moved out of the geometrically similar to type AH and modified highly stressed region near the boss [9]. Af pistons, many type AN pistons have Tr.e Owners Group concluded that the modified experienced relatively high levels of AF piston skirts are adequate for service residual stress due to differences in provided that they are inspected for cracks thermal treatment received by these pistons.

prior to use, and periodically thereafter. Although many AN pistons have reportedly The Owners Group findings notwithstanding, been operated satisfactorily for extended the modified AF piston skirts at Shoreham, periods, there have been numerous reports of Grand Gulf Unit 1, and at other plants have cracks including instances of actual breaking been replaced with an improved piston skirt of the piston skirt into numerous pieces with design, type AE, discussed below [3],[10]. catastrophic consequences to the engine (non-To date, San Onofre Unit I has been the only nuclear). Accordingly, AN piston skirts at plant to seek NRC approval to operate with Catawba Unit I have been replaced with type

" modified" AF pistons installed. NRC AE skirts [12].

initially approved use of modified AF skirts The AE piston skirt design was at San Onofre for one refueling cycle introduced by TOI in 1982 to alleviate based upon a number of considerations problems with the AN design. It incorporates including (1) that a 25% sample inspection an increased stud boss thickness (relative revealed no evidence of cracks, (2) that the to " modified" AF, AH, and AN piston skirts)

San Onofre engines will not be operated and a stress relief to relieve residual above 4500 kw, which corresponds to a cylinder stresses believed to have been responsible firing pressure of about 50% of the firing for the observed cracking in AN skirts.

pressures at normal rated load c6nditions Owners Group analyses indicate stress levels for TDI OSR-4 engines, and (3) that similar to be substantially reduced over earlier piston skirts at a non-nuclear facility skirt designs. Furthermore, operating

~

in Homestead, Florida have operated for experience provides considerable confidence more than 107 cycles (750 hours0.00868 days 0.208 hours 0.00124 weeks 2.85375e-4 months ) at loads that this design will provide adequate comparable to those at San Onofre with no service. Two type AE pistons were run evidence of cracks during subsequent in a TDI test engine for 622 hours0.0072 days 0.173 hours 0.00103 weeks 2.36671e-4 months at inspections (6). The NRC staff and its 514 RPM and at a peak firing pressure PNL consultants have completed their review 20% higher than in TDI engines in nuclear of the modified type AF piston skirts and service. The 622 hours0.0072 days 0.173 hours 0.00103 weeks 2.36671e-4 months of operating time have concluded that they are adequate for corresponds to 9.6 x 108 stress cycles.

continued service beyond the next refueling Subsequent inspections revealed no cracks.

outage for the San Onofre 1 engines subject In addition, type AE pistons were installed to the following requirements: in the Shoreham EDG-103 engine during the a) a load restriction of 4500 kw 15%, 746 hour0.00863 days 0.207 hours 0.00123 weeks 2.83853e-4 months endurance test (107 stress cycles) and, at 3300 kw discussed earlier in Section 1.1.

Again, subsequent inspection revealed no evidence of crack initiation.

i o

6 - 86%a7 o 1.6 CYLINDER HEADS - Numerous instances auxiliary prelubrication pump was provided by of cracks and leaks in TDI cast steel TDI to direct a substantial oil flow to the cylinder heads have been reported in both bearings immediately prior to all planned nuclear and non-nuclear application. From starts. _

an operability standpoint, the major The Owners Group recommended that the O,'

concern is that cracks in the jacket owners maintain oil filtration at 10 microns '%

water passages can result in the leakage or better and utilize spectrochemical and --

of water into the affected cylinder when ferrographic oil analyses regularly as part the engine is in a standby mode. If an of the preventive maintenance programs at attempt is made to start an engine with their plants. The Owners Group has also water present in one or more cylinders, recommended that one bearing be inspected severe structural damage can result. at a plant following an initial 100 starts TDI cylinder heads have been classified of any nature. Furthermore, any bearing by the Owners Group as belonging to one of experiencing 40 automatic starts without manual prelube should be inspected.

three groups [13]. Group I heads include all Finally, the Owners Group has concurred those cast prior to October 1978. Group II with TDI recommendations for monitoring heads include those cast between October 1978 turbocharger rotor axial clearances.

and September 1980. Group III heads include The Owners Group has emphasized the need those cast after September 1980. The not only to confirm that the clearance distinction among groups involves both is within TDI/Elliot specifications but, design changes to facilitate better casting also to trend any increase in clearance control and improvements in heat treatment which may be indicative of thrust bearing and quality control. Most instances of degradation [14].

cracked heads have involved Group I heads.

Only five instances of cracks resulting in 2 MAINTENANCE AND SURVEILLANCE TESTING water leaks have been reported in heads of AND OPERATIONAL CONSIDERATIONS Groups II and III, and these have all been in marine applications. Most of these cracks pg g g were observed to have originated at the engine surveillance to be performed in stellite faced valve seats. conjunction with periodic engine tests will To minimize the potential for leaks, provide the primary means for monitoring the the individual utilities have inspected effectiveness of the Owners Group program in the cylinder head fire decks and valve seats resolving known problem areas and in for cracks pursuant to recommendations by validating the design and manufacturing the Owners Group. In addition the fire adequacy of key engine components. In decks have been checked for proper thickness

addition to a confirmatory role, it is clear To further verify the absence of cracks from the preceding discussion of known which may allow water leakage into the problem areas that periodic inspections and cylinder, the staff has required that the surveillance practices as recommended by the surveillance program for TDI engines include Owners Group or as required by the NRC are an provisions for air rolling of the engine integral element of the technical resolution at appropriate intervals with open cylinder of these issues. In some cases (e.g.,

cocks before and after each planned operation. DSRV-20-4 crankshafts, engine blocks, cylinder The staff has concluded that such air rolls heads, turbocharger thrust bearings), the should be performed 4 to 8 hours9.259259e-5 days 0.00222 hours 1.322751e-5 weeks 3.044e-6 months and again initiation of cracks or abnormal wear during 24 hour2.777778e-4 days 0.00667 hours 3.968254e-5 weeks 9.132e-6 months following any engine operation and, future service cannot be precluded on the thereafter, prior to any planned start [2]. basis of operating experience and/or analysis.

1.7 TURBOCHARGER THRUST BEARINGS - TDI Periodic inspections are, therefore, critical diesel generators in nuclear service employ from the standpoint of assuring that any turbochargers manufactured by the Elliot problems are identified and corrective actions Company. Elliot Model 90G are used for TDI tiken on a timely basis.

Models DSR-48 and DSRV-16 and Elliott Model The Owners Group has prepared a 65G for TDI Model DSRV-20. Turbochargers comprehensive set of maintenance and at several nuclear plants have experienced surveillance recommendations as part of the rapid detsrioration and failures of the Design Review / Quality Revalidation Report combination thrust / radial bearings. It was prepared for each plant. These recognized that bearing and bearing lubrication recommendations reflect Owners Group findings systems inherent in the turbocharger designs stemming from both its Phase I and Phase II

- were not adequate to provide lubrication of efforts and also reflect review by the Owners the bearing thrust pads and rotor thrust Group of TDI Instruction Manuals, TDI Service collars under fast startup conditions to high Information Memos, and TDI correspondence on loads. In response to this problem, the oil specific components. The staff believes that drip system was modified to provide for these recommendations should be followed by increased flow toward the bearings at all each owner in developing its plant-specific times during engine standby. In addition, an maintenance / surveillance program. In addition, each owner should implement an I

7

'860887 -

operational surveillance program to monitor 3.1 GRAND GULF - During an engine test and record key temperatures and pressures at on November 6, 1985, immediately following an key locations in and about the engine. By extensive preventive maintenance outage, one monitoring and recording key engine of the TDI diesel engines at the Grand Gulf parameters, trends in degradation can be nuclear power plant underwent an engine detected, allowing timely preventive overspeed event which caused physical damage maintenance. The staff has required that to bearings and other critical components.

each owner comit to implementation of an The engine startup test being conducted at acceptable maintenance and surveillance the time was intended to verify that the program prior to issuance of an operating preventive maintenance operations just ifcense. completed had been performed correctly.

In Generic Letter 84-15,[15] the NRC Such a test is routinely performed after has encouraged utilities to propose changes maintenance of this type is completed, to the Technical Specifications to address but prior to placing the diesel engine into staff concerns regarding the effects of an operable status. During the recent frequent fast start tests on engine wear and maintenance outage considerable work had tear for TDI and non-TDI engines alike. As been performed on the Woodward governor one example, frequent fast start, fast load which required that the governor lube oil /

tests during preoperational testing was an hydraulic fluid be totally drained from the aggravating factor contributing to the rapid speed control unit. However, during refill deterioration of the turbocharger thrust of the governor lube oil system, the system bearings in several TDI engines as a result was not properly vented and refilled.

of inadequate prelubrication. Technical During the preoperational engine startup

- Specifications currently in place at River test the governor, which controls engine speed Bend specify that each surveillance test be (rpm), did not function because of the preceded by an engine prelube period. inadequate lube oil level and the engine Further, all surveillance tests, with the accelerated to a speed well in excess of its exception of once per 184 days, may also be design / operating speed. As a consequence preceded by warm-up procedures and may als bearings, bushings, and other critical engine include gradual loading (> 150 seconds) as components were damaged. In addition to the recommended by the manufacturer [4].

It has also been customary for plant governor speed contr.ol malfunction, the.

Technical Specifications to require that engine overspeed trip mechanism, which is monthly surveillance testing be performed designed to rapidly shutdown the engine at the nameplate engine rating specified during an overspeed, event before the engine by the manufacturer, with a 2 hour2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months overload rpm reaches a damaging level, failed to trip test every 18 months. For the TDI engines, the engine quickly enoug'. e prevent engine however, the staff is concerned that such damage. Following this e r t the engine was testing could overstress certain components disassembled and inspected by the utility.

(such as DSR-48 crankshafts, for example) The governor speed control unit was and thus, increase the potential for a removed without repair and was shipped to premature failure during a loss of offsite the manufacturer, Woodward, for failure power event. Therefore the staff has assessment and testing.

required that surveillance testing not Based on the information provided by exceed the " qualified" load of such the licensee, Woodward governor and TDI, the components as established on the basis cause of the governor malfunction, which led of appropriate analysis, testing, and/or to the overspeed event and subsequent engine operating experience. However, survelliance damage, was clearly identified as inadequate testing must meet or exceed the maximum lube oil level. New procedures were developed emergency load requirements (as specified to properly refill the governor after future in the plant FSARs) for a design basis maintenance operations. This information was LOOP or LOOP /LOCA event. In addition, promptly forwarded to owners of similar the utility must have adequate operating engines. Subsequent to the assessment of the procedures and operator training to ensure overspeed event, a potential design deficiency that operators have proper guidance and in the overspeed protection trip system was instruction against overloading the diesels identified and a design modification was above the qualified load [3],[4]. developed by Woodward and TDI to reduce the trip response time and enhance the required 3.0 RECENT PROBLEMS engine overspeed protection.

3.2 CATAWBA - Two recent failures of the Several recent events which have number 7 main bearing on a Catawba TDI EDG involved TDI emergency diesel engines are occurred on November 20 and December 5, 1985.

worthy of note because the assessment and The first failure occurred after 2 180 hours0.00208 days 0.05 hours 2.97619e-4 weeks 6.849e-5 months resolution of these problems have of operation. The second failure occurred contributed to the enhancement of reliability after less than 90 seconds. The available of these engines in nuclear standby evidence leaves considerable uncertainty l applications- regarding the cause(s) of the failures.

l

8 .. 860887 e However, the licensee and the NRC staff REFERENCES believe that several factors may have contributed to the failures: [1] United States Nuclear Regulatory a) misalignment of bearing shells Commission (U.S. NRC), " Safety during assembly. Evaluation Report, Transamerica b) a dimensional or other physical DeLaval, Inc., Diesel Generator anomoly in the engine base, Owners Group Program Plan,"

bearing cap or crankshaft. Washington, D.C., August 13, 1984.

c) excessive wear and scoring of the [2] Failure Analysis Associates (FaAA),

bearings and crankshaft journals " Emergency Diesel Generator Crankshaft resulting from operation of the Failure Investigation, Shoreham engine with contaminated lube oil. Nuclear Power Station, "FaAA Report The NRC staff does not believe that No. FaAA-83-10.2.1, Palo Alto, CA, improper installation was a common causal October, 1983.

factor since the first failed bearing was [3] U.S. NRC, " Supplemental Safety 1 installed at the TLI factory and the other Evaluation Report, Shoreham Nuclear l was installed by licensee personnel at the Power Station, Docket 50-322," l reactor site. In addition, no other failures Washington, D.C., December, 1984. j cf this type have been reported for TDI [4] U.S. NRC, " Safety Evaluation Report engines in nuclear serv P_. Bearing failures related to the Operation of River due to improper installat'on or misalignment Bend Station, Docket 50-458,"

characteristically occur, es did the second NUREG-0989, Supplement 3. Washington, failure, at Catawba within 4. few minutes. D.C., August 1985.

Although the cause of the second failure is [5] FaAA, " Evaluation of Transient believed to be from unidentified installation Conditions on Emergency Diesel Generator problems, the first failure cannot be attributed Crankshafts at San Onofre Nuclear to a similar cause. No physical or dimensional Generating Station Unit 1,"

anomolies were found during recent inspections FaAA-84-12-14, Palo Alto, CA, April 1985.

which could cause repeated alignment problems. [6] U.S. NRC, " Safety Evaluation Report, Physical examinations of engine San Onofre Nuclear Generating Station 1, components prior to and following the failures Reliability of TDI Diesel Generators, did indicate that the degree of wear and Docket No. 50-206," Washington, D.C.,

sciring of bearings and bearing journals November 19, 1984.

had progressed. In addition, solid metallic [7] FaAA, " Design Review of Connecting Rod particles found imbedded in the aluminum Bearing Shells for Transamerica Delaval be: ring shells and in the lube oil system Enterprise Engines, "FaAA-84-3-1, Palo were analyzed and were found to be identical Alto, CA, March, 1984.

to the .020 to .040 inch diameter shot [8] FaAA, " Design Review of TDI R-4 Series bitsting material used at the site during Emergency Diesel Generator Cylinder fiald piping modifications and repairs of the Blocks and Liners," FaAA-84-5-4, Palo lube oil system. It is believed that this Alto, CA, June 1984.

gritty material had contributed to the [9] FaAA, " Investigation of Types AF and AE scoring and accelerated wear leading up to Piston Skirts," FaAA-84-2-14, Palo Alto, the first failure. What has not been CA, May, 1984.

cxplained is why the 10 micron lube oil [10] U.S. NRC, " Safety Evaluation Report filters and strainers failed to remove this related to be Operation of Grand Gulf debris. Nuclear Station, Units 1 and 2, Docket The NRC staff concluded that there was Nos. 50-416 and 50-417," NUREG-0831, l considerable uncertainty as to the cause(s) Supplement No. 6, Washington, D.C.,

cf the bearing failures, particularly the August 1984.

first failure. The recommendation was made [11-] FaAA, " Investigation of Type AN and AH to carefully install and confirm proper Piston Skirts," FaAA-84-10-30, Palo clignment of the lower bearing shells by Alto, CA, November 1984.

I lifting / jacking the crankshaft during [12] U.S. NRC, " Safety Evaluation Report cssembly. In addition, extensive flushing related to the Operation of Catawba and cleaning of the lube oil system was Nuclear Station, Units 1 and 2, performed prior to placing the engine back Docket Nos. 50-413 and 50-414,"

into operation. The effectiveness of these NUREG-0954, Supplement 4, Washington, I measures will be determined by a 100 hour0.00116 days 0.0278 hours 1.653439e-4 weeks 3.805e-5 months D.C., December 1984.

confirmatory endurance test during which [13] FaAA, " Evaluation of Cylinder Heads o' the number 7 bearing would be disassembled Transamerica Delaval Inc. Series R-4 and carefully inspected after 1 hour1.157407e-5 days 2.777778e-4 hours 1.653439e-6 weeks 3.805e-7 months and Diesel Engines," FaAA-84-5-12, Palo again at the completion of the 100 hours0.00116 days 0.0278 hours 1.653439e-4 weeks 3.805e-5 months Alto, CA, August 1984.

cf testing.

l l

l J

9 860887 e -

[14] FaAA, " Design Review of Elliot Model [15] U.S. NRC, " Proposed Staff Actions to 90G Turbocharger used on Transamerica improve and Maintain Diesel Generator Delaval DSR-48 and DSRV-16 Emergency Reliability (Generic Letter 84-15),

Diesel Generators Sets," FaAA-84-5-7, " Washington, D.C., July 2, 1984.

Palo Alto, CA, July 1984.

I 9

i f

.. w

4 i

)

t This paper is subject to revision. Statements and opinions ad- Persons wishing to submit papers to be considered for pre-venced in papers or discussion are the author's and are his sentation or publication thsough SAE should send the manu-responsibdity, not SAE's;however, the paper has been edited script or a 300 word abstract of a proposed manuscript to:

by SAE for unsform styhns and format. Discussion udt be Secretary, Engineering Activity Board,SAE.

printed a t:ta the paptr ifit is pubbshed in SAE Transactions.

I or permission to publish this paper in fuu or in part, contact Printed in U.S.A.

the SAE Pubbcations Division.

s -

.* ~

mergenc oowersucoles supply to the diesel was immediately measurable gain to the oserall opera- clear power plant need this special kind shut off and the day's supply tank tion of the plant. remedial action is of att.'ntion from the risk assessment drained back into the main fuel tank. called for. engineer? By nature the diesch. or their But there is a problem in hasing these it is not scry reawuring if the plant equivalent counterparts, are a typical additional safety measures coming into safety engineer expands on probability comentional installation. made up of play too early. while there is stdl a fair calculations, which would ultimately equipment designed and produced for hope that the diesel operation can be prose that the emergency diesel instal- general industrial applications continued satisfactoril). Cuttmg off the lations are not really needed at all. The lioweser. it is their role in connection fuel supply should be part of a control- fact that these installations hase been with the specific requirements of mam-led action by the operator in the control prosided and that they are part of the taimng and supportmg nuclear safety in room. and not automated It should nuclear safety related equipment should emergency conditions that places estra aho be noted that the better the risk be the starting point of this kind of demands on their a*.ailabiht) under the separation between redundant Jiesch discuwion. and ifit ean be shown that a most adserse conditions. These con-or within indnidual diesel compart- protection desice is prosided that can ditions should be imestigated in detail memt the lew necessary it will be to cut jeopardue the operation of the diesel for each installation. On no account the diesel operation in case of fire. that might be needed to cope with a 'should additional difficulties be created A ducted combustion air supply ob- nuclear safety problem, it should be by ill-balanced protection measures.

siously prosides additional flesibility in proef enough that this protection desice in this connection a nsk assessment dealing with a fire in the diesel room. should be remosed. check list has been deseloped. The application of thi, check list to an Asoiding common mode failure. The Risk awcwment (huk tht. Why do the emergency diesel installation h de-degree of redondancy can be further emergency diesel mstallations in a nu- scribed in the panel on p50.

,, improved if care is taken that the filtered combustion air intake cannot be exposed to a common mode type of failure.

Where the intake filters of otherwise redur. dant emergency diesel instal-Resolym.g the Transamerica

'"fi?"'Me""u""&s'TinTi ife t=

s Delaval diesel generator issue direction or towards a common focal point then a single eyent. for example. a By Carl Berlinger and Emmett L. Murphy burning ty re producing a dense cloud of -

smoke, could in fact reduce the degree of redundancy to nil. In the early 1980s major reliability problems were experienced The modern approach is therefore to at American nuclear power plants with standby emergency spread the redundant dicscis over the

site which eliminates esen that some- d.iesel generators manufactured by Transamerica Delaval. h..mce what far-fetched powibihty. But if then, however, the utility owners group and the NRC has e taken sabotage presention is taken seriously it measures MI:ich allow these engines to be qualified for nuclear pays to consider such es ents.

While dncussing problems of this se,.vice and which have helped to re-establish confidence.

nature probabihstic considerations are often thrown in to defend a chosen The failure of a crankshaft in an emerg- nuclear and non-nuclear facdities. The design solution. It should be remem- ency diesel generator during oserload deficienci es appeared to stem from in-bered that at one time nuclear safety testing at the Shoreham nuclear power adequacies in design. manufacture and iwues were totally dominated by deter- plant in August 19S3 was the first quahty awurance quality control.

ministic considerations. Whates er could indication of reliabihty problems with The units installed at Shoreham (and happen. how ever unlikely . was assumed Transamerica Delas al Inc (TDI) aho at Riser Bend) are TDI Model to occur and hence if not acceptable had standby diesel generators. DSR-48 units. These are eight-cylinder to be ehminated by design. In those During the 197th pround 60 TDI in-hne engines - hasing a 3500kW days probabilistic considerations w ere diesel generators had been ordered for nameplate rating and a 3900kW two-resers ed for cases of conflicting require- use at 16 reactors in the United S:ates hour os erload rating.

ments which could not be resobed by belonging to 13 utihties. The first of The Shoreham crankshaf t failure in the deterministic approach alone. Thus these engines became operational in August 1983 occurred during a two-the cnoice between two unasoidable 1977, at San Onofre 1. But it was not hour oserload test at 3900kW. At the i esils had to be based on probabilistic until the carly 19Nh when preoperation time of the failure the affected engine considerations. In those da>% howeser, test programmes were commenced at had been run for a total of 671 hours0.00777 days 0.186 hours 0.00111 weeks 2.553155e-4 months this was the exception, not the rule. It Shoreham and Grand Gulf 1. that a includmg 24 hours2.777778e-4 days 0.00667 hours 3.968254e-5 weeks 9.132e-6 months at loads of 3500kW seems that more of that kind of thinking significant body of operating esperience or more and 19 hours2.199074e-4 days 0.00528 hours 3.141534e-5 weeks 7.2295e-6 months at loads of is needed again. with the TDI standby diesch started to 3900kW o, more. Crankshaf ts in the A disadsantage, esen of low probabi- be budt up. It was at this point that a two other Shoreham standby diesch lity. should not be negotiable against a broad pattern of deficiencies in critical were inspected and aho found to con- l marginal advantage either in cost or in engine compoaents began to become tain sescre cracks The problem turned I operational amenities. As soon as it can esident, at Shoreham and at other out to be fatigue related - caused by be shown that a nuclear safety sositise cad Herhnger and I mmett L Wrptn are mth the part of the reactor installatirn can be tis 'suac.u n<getoon commmon. w sngnin Subsequently . further detects came to handicapped in its operabihty with no Dc :os55. Unneo su A light at other plants and in ether modeh Septen der 1986 5t

2mergenc 00wersuooies e'

q e

v

/

4 g r

. ,i

~

', -n

. s o D.,W.

i i

, u c

.r . .

+.

e ssa q s

~ *

y. h

~ :. < . p I

i a f e j

j- I (*

3 ._

f jl.

8 :n. .; *j

' ?

v -

+

?&

e

r, '

>V ;- J

\(

b 4 .-

g 5 i i 1 ,

J7 } -

7

b, .

of I!)] thesel generators m addition to Innional fatigue fadure of the tranlo. haft in ations to the mJimiual owners conec tn-the Shi i reham t.nlure l hese included, a t ransamerita inviasal inc illsli LWR-M ine actions thes shoult! take to resohe f or esample. t rankshaf t t rat ks. cracks diesel engine at 5hort ham nutirar power these problems including neetted com-m tonnettine ri%l bearine shelk. eneme plant. l his failurc, whnh naturred during ponent usyrades or modiheations. com-blot k (ut ks. piston skut crxks. c(hn- usenperd testing in \ugust pmA was one of

'nt impettions. and engme tests

, de r heait t rat k s and leak s. and turbo- ,in Phase !! of the ow ners croup pro-t et t uti r la ts a i t

)

thareer bearing probkms uhnh Ird to a widespread indusin SRC gramme has proteeded beyond known -

\ number i>t other. less signitn ant- (tfort to rrstore ws.lident e.

problem areas to a systematic con-pri M m .acas h o m.' potentialh gen- sideration of all components 'around i era impin aioas were aSo esentuah. problem areas to proude a basis for the 140 to l'O component types per eneme) n!c nti ied f hcensmg of plants durine the period im por t a n t to the opt raNhts and re-prior to completion of Phase 11 of the habihty ot the engines Phase !! is t hine rs urnup. lo dea! with the prob- ow nen group programme intendeil pnmanly to ensure that ueni-im 11 inow lip i5 nuJear utihty 9 Phase 11 design resieu quality re- ticant new problem areas do not de-w nt is ti imed a i!)I ihesci cencrator i

sahdation of a larce set of kes eneme selop m the future due to dehaencies in ow ners cri'up to adihess the operational components to ensure that their desien de,.icn or quahty of manutatture. The anJ reculamn iuut s it f arme to shesel and manutatture speuhcahons quahty ow ners enoup pertornvd the Phase !!

ee r 'e . s c.ed ter tandhs s mtig- tontini and quahis awu: .mt e . and deuen reuews and, as was the cee for ena pow er In \larth IM the ownen operational sun cilia nce and main- Phee 1. recommended needed som-

.' r i tu p sul*mitNJ a pri pised piin teriance. are adequate ponent upgrades aint nuidifications and er.mune to the N Rt ~ whah. throuch a 9 lapanded engme tests and mspec temponent mspettions to uhdate qual-t ombmahon of design resiews. quahty tion as needed to support Phases I and ity of manuf atture and or auembly A s c uhda hons eneme tt sts and tom- 11 maior element of the Phee 11 pro-pon e n t m punons w as miended to Under Phec 1. the ow ners croup gramme we the preparation of a com-proside an m depth esessment of the performed a temprehenme reuew of prehensne eneme m.nnten.ince and

.hlcquat s ill t he r espet tis e hil!!tles' l l)1 the niperatine histor) t il !!)1 diesel su r s eilla nce procramme til be imple-enemes to perform their safety related eenerators m both nutlear and non- mented hs the mJiudual ow ners tuutmo nudcar scruce ter purposes of idenri-t he ow nen croup prociamme plan tung uemheant problem art s lhe NRC euluation. t he aJequacy of the un ok ed the f ollow me maior elemenn ow ners croup he enduated the tauses ow nen croup procramme he now been O Phase I reso!ution et know n genene of the problems and moed ictomna nd- culuated by the NRC statt and their NUCLE AR ENG NEE A NG 'NTE AN ATICNAL t

r r

.~

e . .-

~

umergency oowersuooies contractor Pacific Northwest Laborato- The bearing was replaced but the At the Forsmark. Oskarshamn and ries. The NRC staff conclusions. as set replacement alvi failed - after less than Barseback nw ns these are automatically out in a recently pubh.hed Safety 90 seconds of operation. While im- started on certain reactor trip functions Evaluation Report, are in essence, that proper installation of the new bearing and also connected to the unit on low if the owners group recommendations may have contributed to this second voltage. l in such areas as quahty resahdation. failure, the reaums for the first were At Ringhals, where one unit is a sw a I inspection, component modifications difficult to determine. One contributing with steam drisen auxiliary feed water and replacement load restrictions and factor may have been metalhc particles and core injection pumps, the gas operating precautions are followed. this in the lubricant, which for some un- turbines are started by remote control if will estabhsh the adequacy of TDI known reason were not picked up by needed and some manual action is diesel generators for nuclear standby the oil filters. needed to connect this power to the service. In this way a technical resolu- Assessment and resolution of these appropriate unit.

tion is asailable to address the so-called recent esents has further contnbuted to For emergency pow er supplies.

"TDI diesel generator nsue" the rehabihty of TDI engmes for nu- on-site diesel generators are also in-clear standby applica' ions They have. stalled.

Recent esents. In the meantime. prob- also sersed to emphasize the impor-lems with TDI emergency die.els re- tance of training and the development Testing diesel generators. The Ringhals cent!) encountered at two American of procedures for proper maintenance. units hase two main turbines each.

plants are worthy of note. Each mam turbine generator supplies in Nosember 1985 one of the TDI one main danion of each unit's elec-diesels at Grand Gulf suffered physical trical sptem. Each disision is broken damage to bearings and oth:r critical Further readmg down into two subdisisions - making

- comp,nents as a result of an oserspeed C. H. Berhnger and E. L. sturphy. Lmerrearv four subdisisions in all. Each reactor esent during a routme preoperational d'"'""""'""'""""'""""##" I'""'""""'" has four emergency diesels, one for f,. [",',"[ #'[s ),$["mNn"l$n""p,"ll"' each of those subdisisions. The site, test that was being conducted following required presentise maintenance, espenence raemg io onw ticonc po.cr mur. which has three rwus and one nwn.

Inadequate oil lesel in the engine's m I endon. E ngland. Wimtober IM therefore has 16 fixed diesels in all.

C' "' """"" '*d E' **"rP hy . Reliabd"' The AC power distribution is at gosernor was found to be the cause and "

the esent led to the deselopment of new ["[,[f [$ "',7.Nt$$mI[$,d'y",' 6.6kV and 0.5LV. The 6.6kV system is maintenance procedures. xemt divided into two parts, the ordinary net Also in Nosember 1985 a main bear- N R C- 5"'< adI""""" "P"n ocuabdur of and the net containing the emergency ing failure occurred at one of the '/[l,l'"1,"ll$l[,jf "

),""1"j s"'","j"[lj,,g'"uesel generators, with two intercon-Catawba diesel generators after 180 Commwon. w ashmpton DC. l'nated states. June nection breakers m series. Both break-hours of operation. tvsn ers open on the black-out signal from

he diesel sequenci..g equipment, black-out being defined as a loss of normal Improving the reliability of the e e d "

"#lres;iger;;","l;,emnnec,ed to the 6.6kV diesel net. one for each of Rmghals diesel generators the feu, sebdmisi. ,. ^, the cennected loads are the same for each pair of diesels, there is an interconnectmg tie By Lars Fredlund which could be used for handhng cross-

. wise faults during loss of offsite power.

As well as having access to the gr.d i for electrical supph.es, all Also. the DC and mserter fed AC Swedish nuclear power plants have dedicated gas turbines systems are in four trains. with the nearby. In addition, emergency diesel generators are provided exception of some ssstems which only

' 'he tu bin "

on site. Efforts have been made to increase the reliability of these Thc e le,g n e motors in-on-site supplies by introducing a " soft start" testing technique. stalled on un'its l'and 2 (SAChl 15m A mobile diesel generating set has also been acquired to rpm. 2.9N1W) differ from those on units 3 and 4 (N hab. Imo rpm 2.7N!W).

supplement the 16 diesels already installed at the four-unit But the same design basis applies to all R.mghals site. four un ts.

The diesels are water-cooled with The operational safety of a nuclear made as reliable as possible. compressed air starting systems. They power plant depends to a large extent Each of the four units at the Ringhals hase no shared auxiliary equipment.

on the constant availability of the elec- nuclear power station, situated on the Control and supervision is from the trical power supplies, both AC and DC. w est coast of Sweden, has its own central control room, where a sincie Thus the nort-al power supply, pro- 400kV transmission line connecting it to summary diesel alarm is prosided - o'ne vided by main turbine generators or the national grid. During grid disturb- for tripping and the other for non-from off-site sources, should be de- ances internal electrical systems can be tripping faults.

signed in such a way that the risk of its supplied from the turbine generators. The technical specifications for a being lost is minimized. In addition, the if both the 400LV system and the nuclear power plant call for regular tests emergency power supplies must also be turbine generators fail, off-site power is to demonstrate the capability of the asailable from the grid at 130kV. If this diesel sets to start and to maintain speed Board."R nghat also fails, al' Swedish plants have dedi- and voltage, and to serify loaded con.

e P wer P ant.

Varobacka. Smeden_ cated gas turbines at or near the plant. ditions.

Septemt>er 1986 53

ML20214E060

Text

Navigation menu

Search