Text

Review & Evaluation of Tdi Diesel Engine Reliability & Operability - Shoreham Nuclear Power Station Unit 1
	ML20111A655
Person / Time
Site:	Shoreham File:Long Island Lighting Company icon.png
Issue date:	12/31/1984
From:	Laity W, Richmond W; Battelle Memorial Institute, PACIFIC NORTHWEST NATION
To:	; Office of Nuclear Reactor Regulation
Shared Package
ML20111A632	List: ML20111A629; ML20111A655;
References
	CON-FIN-B-2963 PNL-5342, NUDOCS 8501080060
	Download: ML20111A655 (121)
	v • d • e

--

1 PNL-5342 l l

Technical Evaluation Report Review and Evaluation of Transamerica Delaval, Inc.,

Diesel Engine Reliability and Operability-Shoreham

-Nuclear Power Station Unit 1 December 1984 Prepared for the U.S. Nuclear Regulatory Commission Division of Licensing Office of Nuclear Reactor Regulation under Contract DE-AC06-76RLO 1830 NRC FIN B2%3 Pacific Northwest Laboratory Operated for the U.S. Department of Energy l ;. by Battelle Memorial Institute 3 ,

'.y ?

s OBallelle B

!!828888R e 868882 PDR 5

- - ~ , , _ _ _ _ _ _ . . , . , _ _ _ . - . - - ~ _ _ . _ . . _ _ . _ _ _ _ _ _ . . _ . - _ _ . . _ _ _ _ . . . _ . _ _ _ _ _ - . _ _ _ _ _ _ _ _ . _ . _ _ _ . . - _

p:

O DISCLAIMER This report was prepared as an account (,f work sponsored by an agency of the .

United States Government. Neithat the United States Government nor any agency thereof, nor any of their employees, makes any warranty, express or imp!ied, or assumes any legal liability or responsibility for the accuracy, com-pieteness, or usefulness of any information, apparatus, product, or process disclosed,or represents that its use would not infringe privately owned rights.

Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise, does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof; PACIFIC NORTHWEST LABORATORY operated by BATTELLE for the UNITED STATES DEPARTMENT OF ENERGY under Contract DE-AC06-76RLO 1830 l

t i

. a-e=

' $g ,.

m 4

5 F

s_.

PNL-5342 t

Technical Evaluation Report REVIEW AND EVALUATION O' 0F TRANSAMERICA DELAVAL, INC.,

DIESEL ENGINE RELIABILITY AND CPERABILITY - SH0REHAM NUCLEAR POWER

' STATION UNIT 1 G

December 1984 Prepared for the U.S. Nuclear Regulatory Commission Division of Licensing Office of Nuclear Reactor Regulation under Contract DE-AC06-76RL0 1830 NRC FIN B2963 Project

Title:

Assessment of Diesel Engine Reliability / Operability NRC Lead Engineer: C. H. Berlinger o

.1

.o Pacific Northwest Laboratory Richland, Washington 99352

PACIFIC NORTHWEST LABORATORY PROJECT APPROVALS O-77 Date e c.. !

o

,i W. W. Laity, Project Manager Pacific Northwest Laboratory

,,-) , [ ' [ G : - _ _ -

.~ l gate 12- O '54 W. D. Richmond, Chairman Senior Review Panel Pacific Northwest Laboratory 0

-O iii

F FOREWORD This report is supplied as part of the Technical Assistance Project, Assessment of Diesel Engine Reliability / Operability, being conducted for the U.S. _ Nuclear Regulatory Commission, Office of Nuclear Reactor Regulation, Division of Licensing, by the Pacific Northwest Laboratory. The U.S. Nuclear Regulatory Commission funded this work under authorization B&R 20-29-40-42-1 FIN No. 82963.

v

.0 o-y c

CONTENTS PACIFIC NORTHWEST LABORATORY PROJECT APPROVALS ......................... iii FOREWORD ............................................................... v TABLES ................................................................. xiii 0

ABBREVIATIONS .......................................................... xv

1.0 INTRODUCTION

...................................................... 1.1 0

1.1 ORGANIZATION OF REPORT ....................................... 1.2 1.2 APPLICABILITY OF CONCLUSIONS ................................. 1.2 1.3 REPORT PREPARATION ........................................... 1.3

2.0 BACKGROUND

........................................................ 2.1 2.1 OWNERS' GROUP PROGRAM PLAN ................................... 2.1 2.2 SHOREHAM NUCLEAR POWER STATION ............................... 2.2 3.0 LILCO TESTS, INSPECTIONS, AND COMPONENT UPGRADES .................. 3.1 3.1 SHOP QUALIFICATION TESTS ..................................... 3.1 3.2 ONSITE PRE 0PERATIONAL AND OPERATIONAL TESTS .................. 3.2 3.3 CONFIRMATORY TESTS ........................................... 3.4 3.4 POST-INSPECTION TESTING ...................................... 3.5 3.5 REPORTED RESULTS AND CONCLUSIONS ............................. 3.6 3.6 PNL EVALUATION ............................................... 3.8 4.0 REQUALIFICATION OF COMPONENTS WITH KNOWN PROBLEMS ................. 4.1 4.1 ENGINE BASE AND BEARING CAPS ................................. 4.2 0

4.1.1 Component Function .................................... 4.2 4.1.2 Component Probl em Hi story . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2

+

o 4.1.3 Owners' Group Status .................................. 4.2 4.1.4 LILCO Status .......................................... 4.3 vii

4.1.5 PNL Evaluation and Conclusion ......................... 4.3 4.2 CYLINDER BLOCK ............................................... 4.5 4.2.1 Compo nent Fu nc ti o n . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.5 4.2.2 Component P robl em Hi s tory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.5 4.2.3 Owners ' Group Sta tu s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.8 .

4.2.4 LILCO Status .......................................... 4.11 4.2.5 PNL Eval uation and Concl usions . . . . . . . . . . . . . . . . . . . . . . . . 4.13 4.2.5.1 Camshaf t Gall ery Cracks . . . . . . . . . . . . . . . . . . . . . . 4.13 4.2.5.2 Circumferential Cracks in Liner Bore ......... 4.14 4.2.5.3 Ligament Cracks .............................. 4.15 4.2.5.4 S tu d- to- S tu d C rac ks . . . . . . . . . . . . . . . . . . . . . . . . . . 4.15 4.3 CRANKSHAFT ................................................... 4.17 4.3.1 Component Function .................................... 4.17 4.3.2 Component Problem History ............................. 4.18 4.3.3 Owners' Group Status .................................. 4.18 4.3.4 L I L CO S ta tu s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.20 4.3.5 PNL Evaluation and Conclusions ........................ 4.22 4.4 CONNECTING RODS .............................................. 4.25 4.4.1 Component Function .................................... 4.25 4.4.2 Component Problem History ............................. 4.25 4.4.3 Owners' Group Status .................................. 4.27 4.4.4 L I L C O S ta tu s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.30

4.4.5 PNL Evaluation and Conclusions ........................ 4.31 4.5 CONNECTING ROD BEARING SHELLS ................................ 4.32 4.5.1 Component Function .................................... .

4.32 4.5.2 Component Problem History ............................. 4.32 viii L_

4.5.3 Owners' Group Status .................................. 4.33 4.5.4 LILCO Status .......................................... 4.33 4.5.5 PNL Eval uation and Concl usion . . . . . . . . . . . . . . . . . . . . . . . . . 4.34 4.6 PISTON SKIRTS ................................................ 4.35 o 4.6.1 C omponen t Fu nc ti o n . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.35 4.6.2 Component P robl em Hi story . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.35 4.6.3 Owners' Group Status .................................. 4.36 4.6.4 L I L CO S ta tu s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.37 4.6.5 PNL Evaluation and Conclusions ........................ 4.38 4.7 CYLINDEP LINERS .............................................. 4.40 4.7.1 Componen t Fu nc ti on . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.40 4.7.2 Component Problem History ............................. 4.40 4.7.3 Owne rs ' G rou p Sta tu s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.41 4.7.4 LILCO Status .......................................... 4.41 4.7.5 PNL Evaluation and Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . 4.42 4.8 CYLINOER HEADS ............................................... 4.43 4.8.1 Compo nen t Fu nc ti o n . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.43 l 4.8.2 C omponent P robl em Hi s to ry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.44 4.8.3 Owners' Group Status .................................. 4.44 4.8.4 L I LCO S t a tu s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.45 4.8.5 PNL Eval uation and Conclusions . . . . . . . . . . . . . . . . . . . . . . . . 4.46 4.9 CYLINDER HEAD STUDS .......................................... 4.48 4.9.1 Component Function .................................... 4.48 4

4.9.2 C ompo ne nt P robl em H i s to ry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.48 4.9.3 Owners' Group Status .................................. 4.48 4.9.4 L I L C O S t a tu s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.49 ix

_ . - . . - . _ - _ - - - _ . . - - - - - ~

4.9.5 PNL Eval uation and Concl usions . . . . . . . . . . . . . . . . . . . . . . . . 4.49 4.10 P U SH ROO S . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.50 4.10.1 Compo ne n t Fu nc t i o n . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.50 4.10.2 Component P robl em Hi story . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.50 4.10.3 Owners' Group Status ................................. 4.50 .,

4 .10. 4 L I LCO S ta tu s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.51 4.10.5 PNL Eval uation and Concl u sion . . . . . . . . . . . . . . . . . . . . . . . . 4.51 4.11 ROCKER ARM CAPSCREWS ........................................ 4.52 4.11.1 Component Function ................................... 4.52 4.11.2 Component P robl em Hi story . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.52 4.11.3 Owners' Group Status ................................. 4.52 4.11.4 L I L CO S ta tu s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.52 4.11.5 PNL Evaluation and Concl usion . . . . . . . . . . . . . . . . . . . . . . . . 4.53 4.12 TURBOCHARGERS ............................................... 4.54 4.12.1 Component Function ................................... 4.54 4.12.2 Component P robl em Hi story . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.54 4.12.3 Own e rs ' G ro u p S ta tu s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.55 4.12.4 L I LCO S t a tu s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.56 4.12.5 PNL Evalu ation and Conclu sions . . . . . . . . . . . . . . . . . . . . . . . 4.57 4.13 J AC K ET W AT E R P U MP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.59 4.13.1 Component Function ................................... 4.59 4.13.2 Component Problem History ............................ 4.59 e 4.13.3 Owners' Group Status ................................. 4.59 4.59

4.13.4 LILCO Status .........................................

4.13.5 PNL Evaluation and Conclusion . . . . . . . . . . . . . . . . . . . . . . . . 4.60 4.14 HIGH-PRESSURE FUEL OIL TUBING ............................... 4.61 X

4.14.1 C omp o n en t Fu nc ti on . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.61 4.14.2 Component P robl em Hi story . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.61 4.14.3 Owners' Group Status ................................. 4.61 4.14.4 LILCO Status ......................................... 4.62 0

4.14.5 PNL Evaluation and Conclu sion . . . . . . . . . . . . . . . . . . . . . . . . 4.62 4.15 AIR STARTING VALVE CAPSCREWS ................................ 4.63-O 4.15.1 Component Function ................................... 4.63 4.15.2 Component Problem History ............................ 4.63 4.15.3 Owners' Group Status ................................. 4.63

. 4.15.4 L I LCO S t a tu s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.63 4.15.5 PNL Evaluation and Conclusions . . . . . . . . . . . . . . . . . . . . . . . 4.64 4.16 ENGINE-MOUNTED ELECTRICAL CABLE ............................. 4.65 4.16.1 Compo nen t Fu nc ti o n . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.65 4.16.2 Component Problem History ............................ 4.65 4.16.3 Owners' Group Status ................................. 4.65 4.16.4 L I L CO S t a tu s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.65 4.16.5 PNL Evaluation and Conclusion . . . . . . . . . . . . . . . . . . . . . . . . 4.66 5.0 PROPOSED MAINTENANCE add SURVEILLANCE PROGRAM . . . . . . . . . . . . . . . . . . . . . 5.1 5.1 MAJ OR MA INTEN ANCE ITEMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2 5.1.1 PNL Evaluation and Recommendations . . . . . . . . . . . . . . . . . . . . 5.2 5.1.1.1 Crankshaft ..................................... 5.8 e

5.1.1.2 C yl i n de r B l o c k . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.8 5.1.1.3 Connecting Rods ................................ 5.11 4

5.1.1.4 Connecting Rod Bearing Shell s . . . . . . . . . . . . . . . . . . 5.11 5.1.1.5 C yl i n de r Li n e r s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.12 5.1.1.6 C yl i n de r H e a d s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.12 xi L___

5.1.1.7 Head Studs, Air Start Yalve Capscrews, Roc ke r A rm B ol ts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.13 5.1.1.8 Turbochargers .................................. 5.14 5.1.1.9 Lube Oil Sampl ing and Anal ysi s . . . . . . . . . . . . . . . . . 5.14 5.2 ADDITIONAL MAINTENANCE ITEMS ................................. 5.15 a

5.2.1 R a ti o n a l e . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.15 5.2.2 P NL Ev a l u a ti o n . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.15 ,

5.3 OPERATIONAL SURVEILLANCE PLAN ................................ 5.19 5.3.1 R a ti o n a l e . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.19 5.3.2 P NL Rec ommenda ti o ns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.19 5.3.2.1 P re-Turbi ne Exhau st Temperatu re . . . . . . . . . . . . . . . . 5.19 5.3.2.2 Ai r Mani fol d Temperature . . . . . . . . . . . . . . . . . . . . . . . 5.21 5.3.2.3 Fuel Oil Transfer Pump Strainer Differential Pressure ....................................... 5.21 5.3.2.4 Starti ng Ai r P ressu re . . . . . . . . . . . . . . . . . . . . . . . . . . 5.21 5.3.2.5 Fuel 011 Da ywTa nk Level . . . . . . . . . . . . . . . . . . . . . . . . 5.21 5.4 STANOBY SURVE ILL ANCE PL AN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.21 5.4.1 Rationale ............................................. 5.21 5.4.2 P NL Ev al u a ti o n . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.22 6.0 OVERALL CONCLUSION ................................................ 6.1 6.1 GENERAL CONCLUSION ........................................... 6.1 6.2 BASIS FOR CONCLUSION ......................................... 6.1 e

'e xii

TABLES 3.1 Component Inspections Conducted by LILCO Following Confirmatory Testi ng of Emergency Diesel Generator 103. . . . . . . . . . . . . . . . . . . . . . . . 3.5 3.2 Significant Problems Encountered in Shoreham Emergency Diesel Generators Du ri ng Te sti ng. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.7 4.1 Loads and Engine Hours........................................... 4.21 5.1 Major Maintenance Items for Shoreham Nuclear Power Station....... 5.3 5.2 Recommended Additional Maintenar.ce Items for Shoreham Nuclear P owe r S t a t i on . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.16 5.3 Recommended Operational Surveillance Plan for Shoreham Nuclear P owe r S ta t i o n . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.20 5.4 Recommended Standby Surveillance Plan for Shoreham Nuclear P owe r S t a ti on . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.23 w

h xiii

ABBREVIATIONS ABS American Bureau of Shipping ASLB Atomic Safety and Licensing Board BMEP brake mean effective pressure

Code of Federal Regulations CFR DEMA Diesel Engine Manufacturers Association OR/QR design review / quality revalidation E80CR Engineering and Design Coordination Report EDG emergency diesel generator ET eddy-current testing FaAA Failure Analysis Associates FSAR Final Safety Analysis Report IACS International Association of Classification Societies LILCO Long Island Lighting Company LOCA loss of coolant accident LOOP loss of offsite power LP liquid penetrant M/S maintenance / surveillance NDE nondestructive examination NDT nondestructive testing NRC U.S. Nuclear Regulatory Commission OG Owners' Group; the TOI Diesel Generator Owners' Group OGPP Owners' Group Program Plan 0/R uperability and reliability PNL Pacific Northwest Laboratory t SNPS Shoreham Nuclear Power Station SWEC Stone & Webster Engineering Corporation TDI Transamerica Delaval, Inc.

TER technical evaluation report UT ultrasonic testing xv L.

l l

REVIEW AND EVALUATION OF TRANSAMERICA DELAVAL, INC.,

DIESEL ENGINE RELIABILITY AND OPERABILITY -

SHOREHAM NUCLEAR POWER STATION UNIT 1

1.0 INTRODUCTION

Long Island Lighting Company (LILCO) is seeking a full power operating license for the Shoreham Huclear Power Station (SNPS) Unit 1. One matter of concern to the U.S. Nuclear Regulatory Commission in considering the request is the operability and reliability (0/R) of the SNPS standby emergency diesel-engine generators manufactured by Transamerica Delaval, Inc. (TDI). The 0/R of these engines have been brought into question by a major crankshaft failure at SNPS in August 1983 as well as by other problems reported by owners of TDI diesels in nuclear and non-nuclear service.

Shoreham Nuclear Power Station Unit 1 is served by three TDI model DSR-48 diesel engines, designated emergency diesel generator (EDG) 101, 102, and 103.

These EDGs are inline 8-cylinder four-cycle, turbocharged, aftercooled engines.

Each is nameplate rated for 3500 kW, and operates at 450 rpm with a cylinder brake mean effective pressure (BMEP) of 225 psig. The latest information pro-vided by LILCO specifies the emergency load as a maximum of 3300 kW under design basis accident conditions coincident with a simulated loss of offsite power (LOOP).

In response to the NRC concerns about the SNPS TDI engines LILCO has undertaken a comprehensive analysis of all major engine components and com-pleted a number of component replacements and engine tests to ensure their 0/R. These LILCO actions are described in a number of documents and in testi-

mony before the Atomic and Safety Licensing Board (ASLB).

The Pacific Northwest Laboratory (PNL) has been requested by NRC to review

, and evaluate LILCO's overall efforts to ensure the engines' reliable perfor-mance. This technical evaluation report (TER) documents the PHL review and expresses the resulting conclusions and recommendations regarding the capability of the SNPS TDI diesel engines to serve their intended function.

1.1

1.1 ORGANIZATION OF REPORT

' This technical evaluation report is organiz2d as follows:

e Section 2.0 provides relevant background information on efforts by both LILCO and the TDI Diesel Generator Owners' Group (an ad hoc group of similar TDI engine owners) to resolve the TDI engine Concerns. ;

Section 3.0 presents PNL's review and evaluation of LILCO's tests, inspections, and component upgrades undertaken to prepare the engines ,

for nuclear service.

e Section 4.0 comprises a review and evaluation of LILCO's resolution of known problems in 16 engine components identified by the Owners' Group through a review of TOI engine operating history, e Section 5.0 provides PNL's review of LILCO's proposed maintenance and surveillance (M/S) program.

e Section 6.0 presents PNL's overall conclusions and recommendations regarding the suitability of the three TDI diesel engines to perform their intended function as emergency standby power sources for the SNPS.

1.2 APPLICABILITY OF CONCLUSIONS To derive the conclusions presented in this report, PNL reviewed the basic documents supplied by LILCO, participated in various meetings with LILCO and NRC, and observed components of the engines as disassembled for inspection following testing. Concurrently, PHL also reviewed various relevant Owners' Group documents and participated in their meetings with NRC, and drafted tech-nical evaluation reports on some elements of the Owners' Group Program Plan (OGPP). In addition, PNL and its consultants participated in the hearing on .

Docket No. 50-322-OL, extending from September through November 1984, before the Atomic Safety and Licensing Board. Information stemming from these reviews e and activities constitutes a major part of the data base from which PHL's con-clusions are drawn.

1.2

Immediately prior to the preparation of this TER, LILC0 submitted to the NRC a Final Safety Analysis Report (FSAR) revision that identifies 3300 kW as the " qualified load" considered necessary for each engine to support its designated share of the emergency power needs of SNPS. In recognition of that FSAR revision, this TER addresses the adequacy o; engine components relative to this load limit.

This TER precedes the completion of the final review by PNL and the NRC staff of the Owners' Group Program. Accordingly, the conclusions expressed in

~

this TER about the long-term suitability of the Shoreham engines for nuclear service are contingent upon final action by NRC on the following PNL recom-mendations. LILCO should commit to NRC to implement all applicable recommen-dations-and requirements identified in the NRC review of the Owners' Group Program. Completion of the ongoing Phase 1 and Shoreham Phase 2 reviews is anticipated early in 1985. In the opinion of PNL, the reviews of all Shoreham-related issues that require priority PNL/NRC attention have progressed suffi-ciently to consider these issues resolved, subject to the actions discussed in this TER. All recommendations and requirements identified in NRC's review of the Owners' Group Program should be implemented or be fully ready to implement by the end of the first reactor operating cycle. These actions will complete the resolution of the TOI engine issues at Shoreham.

1.3 REPORT PREPARATION This report is based in part on PNL's review of documents cited in Section 2 .0 . In addition, the PNL team visited the Shoreham Nuclear Power Station Unit 1 in March, May, and November 1984, for orientation and to observe the EDG 103 inspection following testing. PNL met with SNPS staff and management on these occasions, as well as in connection with TOI Owners' Group meetings

, and the Atomic Safety and Licensi'ng Board hearing during September through November 1984.

, The following PHL staff members and consultants were involved in this review and evaluation, and authored this report:

D. A. Dingee, PNL project staff e R. E. Oodge, PNL project staf f 1.3 L_

i l

I e J. F. Nesbitt, PNL project staff e F. R. Zaloudek, PNL project staff e A. J. Henriksen, diesel consultant to PNL

P. J. Louzecky, diesel consultant to PNL.

Others whose contributions were valuable in fomulating the conclusions presented herein include PNL Assessment of Diesel Engine Reliability /Operabil- '

ity Project team members J. M. Alzheimer, L. G. Van Fleet, and W. W Laity; and consultants S. H. Bush, B. J. Kirkwood, A. Sarsten, and J. V. Webber (repre-senting Ricardo Consulting Engineers). The report editor was A. J. Currie. -

l 1

e D

1.4

I 1

l l

2.0 BACKGROUND

This section presents background information on efforts undertaken by the TDI Diesel Generator Owners' Group and by Long Island Lighting Company to resolve the problems identified in the TDI diesel engines.

s 2.1 OWNERS' GROUP PROGRAM PLAN Thirteen nuclear utilities that own diesel generators manufactured by Transamerica Delaval, Inc. have established an Owners' Group to address ouestions. raised by the major failure in one TDI diesel engine at the Shoreham Nuclear Power Station in August 1983, and other problems in TDI diesels reported in the nuclear and non-nuclear industry. On March 2,1984, the Owners' Group submitted a plan to the U.S. Nuclear Regulatory Commission outlining a comprehensive program to requalify their diesel generator units as standby emergency power sources.

The Owners' Group Program Plan describes a two-phase approach for resolving the known and potential problems in TDI engines:

e Phase 1 addresses the evaluation and resolution of significant known problems in 16 components. These problems were identified by the Owners' Group through a review of the operating histories of TDI

. engines in nuclear and non-nuclear services, e Phase 2 entails a comprehensive design review and quality revalida-tion (DR/QR) to identify critical components of TDI engines in addition to the 16 referred to above, and to ensure that these components are also adequate for their intended service.

The OGPP also describes a program element for .special or expanded engine tests

. and component inspections, as appropriate, to verify the adequacy of the engines and components to perfonn their intended functions.

.p' At NRC's request, PNL reviewed the OGPP. The results of that evaluation were reported to NRC in PNL-5161, Review and Evaluation of TDI Diesel Generator Owners' Group Program Plan (Pacific Northwest Laboratory June 1984).

2.1

Section 4 of PNL-5161 deals with considerations for licensing actions for nuclear stations prior to completion of the implemer.tation of the OGPP. Recom-mendations in that report relevant to LILCO's current request for licensing of the Shoreham Nuclear Power Station Unit 1 are:

1. The engines should be inspected per Section 2.3.2.1 of PNL-5161 to e ensure that the components are sound. #

2. Preoperational testing should be performed as discussed in 2

Section 2.3.2 of PNL-5161. .

3. The engines should receive enhanced surveillance and maintenance.

4. A " lead engine" as described in Section 2.3.2.2 of PNL-5161 should be tested to 107 cycles at " qualified" load to verify the design 4

adequacy of key engine components subject to fatigue stresses, if components of the same design have not already been operated that .

many cycles under the same or greater load.

+

l_ The first three recommendations are self-evident; namely, that the engines j

have sound parts, that appropriate preoperational tests have been satisfac-j torily completed, and that a suitable program of maintenance and surveillance

is established to ensure future performance. The fourth recommendation is '

! included to ensure that all components, including the pistons and crankshaft, have sufficient fatigue resistance to preclude fatigue fracture of these com-ponents with concomitant engine failure.

2.2 SHOREHAM NUCLEAR POWER STATION i

In its efforts to establish the operability and reliability of the SNPS TDI diesel engines, LILCO has performed engine modifications and has conducted tests and inspections. The utility has provided NRC with documents relevant to these activities. These documents and others that were used in the preparation

of this TER are listed below.

e a LILCO report dated August 23, 1983, Shoreham Nuclear Power Station .

Emergency Diesel Generator 102 Crankshaft Failure Analysis / Recovery -

Master' Plan - This report describes the steps LILCO will take in 2.2

investigating the failure of the EDG 102 crankshaft and in detemining the generic implications for EDGs 101 and 103.

e a LILCO report dated December 14, 1983, Shoreham Nuclear Power Station Emergency Diesel Generator Recovery Test Program - This report defines the diesel generator test program that will be

~

'+' implemented on the SNPS TDI engines to demonstrate 0/R following replacement of the engine crankshafts and reassembly of the engines.

. e a LILC0 report dated January 6,1984, Shoreham Diesel Generator Recovery Program Summary - Shoreham Nuclear Power Station - This document outlines all aspects of the SNPS diesel generator recovery program.

e a report entitled Delaval Diesel Generator Operation Experience (handout at TDI Owners' Group meeting, January 26, 1984) - This report outlines the experience of various owners of TDI diesels (including LILCO) with their engines to late 1983.

An OG report dated June 29, 1984, TDI Diesel Generator Design Review /

Ouality Revalidation Report - Shoreham Nuclear Power Station Unit 1 -

This 9-volume report documented the comprehensivo DR/QR effort perfomed on the SNPS TDI engines and the results of that effort.

LILCO Engineering and Design Coordination Report No. F-46505 and attachments dated July 17, 1984 - This report addresses the TDI diesel maintenance / surveillance program.

e an NRC report dated August 13, 1984, Safety Evaluation Report -

Transamerica Delaval, Inc. Diesel Generator Owners' Group Program Plan - This report presented HRC staff recommendations for TDI diesel generator test and inspection programs.

e a report by Stone & Webster Engineering Corporation dated August 1984, entitled Survey of Start Experiences and Causes of Unscheduled

+ Shutdowns of Transamerica Delatal Inc. Diesel Engines - This document summarizes data extracted from various diesel generators' logs (including the three engines at SHPS).

2.3

e a letter dated October 5,1984, from J. J. Range (LILCO) to A. R.

[ Dynner (Suffolk County) regarding duplication of photos for Suffolk County - This letter summarizes the status of the discovery made available the week of October 5,1984, by LILCO pursuant to Suffolk

. County's September 25 request,

j. .

a letter dated October 5,1984, from J. J. Range (LILCO) to R. J. t Goddard (NRC) transmitting LILCO Deficiency Report Number 1224 i including revision and several FaAA preliminary inspection reports.

l e ' a letter dated October 6,1984, from J. J. Range (LILCO) to A. R.

. Dynner (Suffolk County) transmitting receiving inspection reports, production routing sheets, etc., for LILC0/TDI engines.

~e a letter dated October 9,1984, from J. J. Range (LILCO) to A. R.

Dynner'(Suffolk County) transmitting photographs of the cam gallery, engine bed, and main bearing saddles at SNPS.

e a letter dated October 18, 1984, from J. D. Leonard, Jr. (LILCO) to H. R.-Denton (NRC), " Confirmatory Testing of TDI Diesel Generators at Shoreham Nuclear Power Station Unit 1, Docket No. 50-322" - This document provides NRC with LILCO's testing protocol for the 107 -cycle confirmatory tests.

e a letter dated October 22, 1984, from _J. Leonard (LILCO) to H. R.

Denton (NRC), " Submittal of FSAR Revision Qualified Load - TDI Diesel Generators at Shoreham Nuclear Power Station Unit 1, Docket No. 50-322" - In response to Item 1 of Section 4.6 of the NRC Safety Evaluation Report on TDI OGPP entitled " Interim Basis for Licensing",

LILCO has developed a " qualified load" by a combination of analysis and testing utilizing results of a recent preoperational test.

e a LILCO report dated December 3,1984, TDI Emergency Diesel

Generator 10310 7-Cycle Confirmatory Test / Inspection Report, Shoreham Nuclear Power Station Unit 1 - This report provides LILCO's tests and inspection results for the 107-cycle confirmatory test of the EDG 103.

2.4

e a letter dated December 6,1984, from B. Germano and M. Herlihy (LILCO) to D. Dingee (PNL), " Load and Engine Hour Telecopy dated 12/5/84 from D. Dingee to B. Germano" - This letter provides engine hours for all three SNPS EDGs following crankshaft replacement to November 4,1984. ,

1

. .* a letter dated December 7,1984, from B. Germano (LILCO) to D. Dingee b (PNL), "Telecopy dated 12/6/84 from D. Dingee to B. Germano Concern-ing Regulatory Tests Following DR/QR Inspections" - This letter

provides confirmation that LILCO has completed the required qualiff-cation tests following EDG 101 and 102 reassembly in the spring of 1984.

[ e a letter dated December 11, 1984, from B. Germano and M. Herlihy (LILCO) to D. Dingee (PNL), " Load and Engine Hour Data for DG101 and

DG102" - This letter summarizes approximate engine data for EDG 101 l and 102 from the time following the last crankshaft inspection to the

{ end of the EDG 103 long duration run on November 4,1984.

In addition to reviewing these documents, PNL visited the SNPS . site to

- observe engine inspections and review a~ sample of the LILCO procedures for i dispositioning component inspection findings. PNL and its consultants also

j. gained perspective on certain SNPS components (crankshaft, piston skirts, _

cylinder heads, and engine block) through participation in the Atomic Safety and Licensing Board hearing (Docket No. 50-322-OL) over 'the period extending from September 10.through November 16, 1984. The testimony and exhibits for this hearing are also reference material used in preparation of this TER.

i b-f

-0 2.5 4

3.0 LILCO TESTS, INSPECTIONS, AND COMPONENT UPGRADES The SNPS EDGs have been subjected to several testing / inspection programs and, as a result,- have undergone several component upgrades. These programs consist of 1)' shop qualification tests, 2) onsite preoperational tests, 3) con-firmatory tests, and 4) special post-inspection tests. The key inspections e

include those done on all engines in connection with the DR/QR activities as well as the EDG 103 inspection in November 1984 following the confirmatory

- tests. Other incidental inspections were also done.

A chronological discussion of these tests, inspections, and component upgrades is presented in Sections 3.1 through 3.4. The results and conclusions reported by LILC0 are documented in Section 3.5. PNL's evaluation of LILCO's program is presented in Section 3.6.

3.1 SHOP QUALIFICATION TESTS According to LILCO, the test program for the Shoreham EDGs began in the early 1970s with shop tests at the TDI manufacturing facilities in Oakland, California. These shop tests were performed to verify the operability of the EDG units, including the interrelated functional capability of engine com-ponents. The shop tests accomplished for all three engines included:

e load tests e air starting system tests

. e alarm and safety function tests.

LILCO noted that EDG 101 was used by TDI to qualify the R-48 series engines for nuclear service. This involved successfully completing 300 con-secutive starts and operating for 110 hours0.00127 days 0.0306 hours 1.818783e-4 weeks 4.1855e-5 months , mostly at loads above 50% of the

' engine's rated load.

LILCO reported that the shop tests required a minimum of 30 hours3.472222e-4 days 0.00833 hours 4.960317e-5 weeks 1.1415e-5 months of operation on each EDG; 10 of those hours were at loads equal to or greater than

, 100% load (3500 kW). In addition, each unit was required to start at least 10 times.

3.1

3.2 ONSITE PREOPERATIONAL AND OPERATIONAL TESTS Onsite preoperational tests were conducted to confirm EDG operability and to verify their functional capability to interface with various plant sys-tems. Tests pursuant to the SNPS FSAR, including NRC Regulatory Guide 1.108 tests, were also conducted.

LILCO has reported that the diesel generators were operated onsite for the a first time in October 1981 (EDG 102), March 1982 (EDG 103), and April 1982 (EDG 101). The preoperational test program, consisting of 12 preoperational ,

tests (four per diesel engine), was started in September 1982 and included 1) a mechanical test to check various mechanical trips on the diesel engine and the air starting capability of the air start system; 2) two electrical tests including various high-load rejection tests to demonstrate the electrical trips for the generator; and 3) the diesel generator qualification test, which demon-strated the capability of the diesel generators to successfully complete a total of 69 consecutive starts. On June 24,1983, LILC0 successfully completed the preoperational test program, including all mechanical, electrical, and qualification tests, for all three diesel generators.

In spring 1983, LILC0 commenced a cylinder head replacement program in response to minor cooling water leaks that were noted. In mid-August 1983, during testing in conjunction with this cylinder head ggrade program, the crankshaft of EDG 102 fractured. This engine had logged 671 total hours operation, including 254 hours0.00294 days 0.0706 hours 4.199735e-4 weeks 9.6647e-5 months at the rated load of 3500 kW and 19 hours2.199074e-4 days 0.00528 hours 3.141534e-5 weeks 7.2295e-6 months at 3900 kW. At that time, the engine hours logged on EDGs 101 and 103 were com-parable to those of EDG 102. An inspection of the EDG 101 and 103 crankshafts also revealed cracks.

Following the EDG 102 crankshaft failure, LILCO initiated an effort to investigate the failure and to assess any generic implications for EDG 101 and 103. This effort included inspections of the diesel engine components in addi-tion to the crankshafts. Several components were found to have problems.

LILC0 reports that four of the upper connecting rod bearings were cracked; all ,,

the AF piston skirts were found to have linear indications; and the governor of EDG 102 was found to be damaged as a result of the crankshaft failure and was returned to the manufacturer. The nature of these findings and their 3.2

disposition are provided in more detail in Section 4.0 for the 16 components with known problems identified by the OG.

In fall 1983, as part of the crankshaft failure recovery program, LILC0 made a number of engine modifications. New crankshafts with larger (12-inch diameter) crankpins were installed; all piston skirts were replaced with AE

skirts; all cylinder heads were replaced with the newer Group III models; and new connecting rods were installed with the 12-inch (rather than 11-inch) bear-ing diameter. New connecting rod bearings were installed and rod-eye bushings with relevant indications were replaced.

During the same time period, LILC0 became aware of TDI engines in non-nuclear service with block cracks (marine experiences with the ore-carrier MV Gott and the MV Columbia, a ship belonging to the Alaskan Marine Highway). In response to this and upon advice of their diesel consultants, LILCO conducted dye-penetrant inspections of the cylinder liner landing area. They reported no relevant indications in this area. However, numerous radial / vertical cracks were found in the ligaments between the bore and the stud holes on all SNPS EDGs.

LILC0 next proceeded with engine testing and inspections in support of their DR/QR activities. This included achieving a nominal 100-hour total operation on each of the engines at or above the rated load (3500 kW), followed by a comprehensive engine inspection. The inspection that the OG accomplished encompassed 168 components of the SNPS diesel generators. The TDI Owners' Group provided technical recomendations regarding special component inspec-tions. LILC0 was responsible for implementing these recommendations and for establishing acceptance criteria where none were established by the Owners' Group. Further details of the DR/0R inspections as applicable to the generic problem components are summarized in Section 4.0 of this report.

This period (fall 1983 through spring 1984) of testing was highlighted by a number of problems with turbochargers (failed nozzle ring capscrews, a lost

nozzle ring vane on EDG 103, and failed bearings on all SNPS EDGs). LILCO made suitable repairs and replaced the bearings.

3.3

^

1 J

l LILC0 reported that, during a test in spring 1984, the EDG 103 engine was found to have developed a crack down the front of the block. This block was subsequently replaced with TDI's new-design cylinder block. Microstructural analysis revealed the original block had an extensive degenerate microstructure that produced inferior mechanical properties. Cracks between the cylinder liner bore and cylinder head studs (ligament cracks) were reported in EDGs 101 and 102.

The DR/QR disassembly / inspection also resulted in a number of component replacements: one cylinder liner, several connecting rod bearings, the governor couplings, a number of rod-eye bushings, the jacket water pump, and both turbocharger thrust bearings.

3.3 CONFIRMATORY TESTS In addition to the above tests and inspections, LILCO undertook confirma-tory tests of EDG 103. These tests and the post-test inspections were per-formed in accordance with NRC staff recommendations described in Safety Evaluation Report - Transamerica Delaval, Inc. Diesel Generator Owners' Group Program Plan issued on August 13, 1984, and in subsequent discussions between NRC and LILCO, documented in LILC0's letter SNRC-1094 dated October 18, 1984.

~

1 The confirmatory testing on EDG 103 provided additional running hours (525 hours0.00608 days 0.146 hours 8.680556e-4 weeks 1.997625e-4 months ) to accumulate a total of in excess of.740 hours0.00856 days 0.206 hours 0.00122 weeks 2.8157e-4 months of operation at the load of 3300 kW on the replacement crankshaft. This corresponds to 107 stress cycles, generally considered adeouate and necessary to confirm long-term life at the tested load.

i Inspections following the confirmatory tests included those indicated for l each component listed in Table 3.1.

D 3.4

TABLE 3.1. Component Inspections Conducted by LILCO Following Confirmatory Testing of Emergency Diesel Generator 103 Component Inspection Perfonned Cylinder heads Ultrasonic inspection of firedeck thickness at six specified locations Liquid penetrant (LP) inspection of surfaces of intake and exhaust valve seats and the firedeck area between exhaust valves

. Visual inspection to determine if any heads had through-wall weld repairs of the firedeck where the repair was performed from one side only Engine block Fluorescent magnetic particle examination and LP exam-ination of the block top and in cam gallery locations where strain gauges had been placed during the con-firmatory testing Eddy-current (ET) examination in the stud hole region between cylinders No. 4 and 5 Connecting rod LP examination bearings Wrist pin and LP examination rod-eye bushings Turbocharger Visual inspection of the thrust bearings, the nozzle ring vanes and capscrews, and the turbocharger mounting f1ange bolts Bearing float evaluation Crankshaft LP and ET examinations as appropriate on all fillet areas and all crankshaft oil holes except the main bearings No. 1, 2, 10, and 11 Gears Visual inspection of accessible front-end gears and gear teeth

'~

Cylinder liners Visual examination for excessive scuffing Pistons LP examination of all piston skirts in the stud / boss

'+ region Visual inspection of crown-to-skirt contact surface 3.5 4 -r ,e --- -w w - - - , -w, , , - , --- --

w n , - -

LILC0 reported that all of the inspections listed in Table 3.1 showed acceptable results. Except for two connecting rod bearings (one that was damaged during disassembly and one that had an indication on the cuter diameter) and one cylinder head that was found to contain a plug weld, the components were released for reinstallation in the engine.

3.4 POST-INSPECTION TESTING LILCO has reported that, to demonstrate engine operability after reassembly, they have successfully completed SNPS FSAR testing including Regu-latory Guide 1.108 tests for EDGs 101 and 102. They have also completed the following tests on EDG 103 following reassembly after the confirmatory tests:

e ten modified starts to at least 1400 kW, but not to exceed 3300 kW e two fast starts to 3300 kW; run for a minimum of 4 hours4.62963e-5 days 0.00111 hours 6.613757e-6 weeks 1.522e-6 months after each fast start e one 16-hour test at load levels stepping up to and then down from 3300 kW - This includes a total of 4 hours4.62963e-5 days 0.00111 hours 6.613757e-6 weeks 1.522e-6 months at each of the following loads: 3300 kW, 2625 kW, 1750 kW, and 875 kW.

In addition to the immediate post-inspection operability tests, LILC0 will conduct a 3300-kW load test once every 18 months for 24 hours2.777778e-4 days 0.00667 hours 3.968254e-5 weeks 9.132e-6 months . This test replaces the previously required NRC 18-month load tests consisting of a 3500-kW load for 22 hours2.546296e-4 days 0.00611 hours 3.637566e-5 weeks 8.371e-6 months and a 3900-kW load for 2 hours2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months as specified in Regulatory Guide 1.108. This change will be incorporated in revisions to the plant technical specifications.

3.5 REPORTED RESULTS AND CONCLUSIONS Various problems occurred or were noted during the tests conducted on the SNPS engines. Those of significance are summarized in Table 3.2. LILCO ,

l j reports that all of these problems have been corrected or do not threaten engine reliability or operability.

The TDI Owners' Group has formally reported the results of their com-prehensive DR/QR effort in a nine-volume report entitled TDI Diesel Generator l Design Review / Quality Revalidation Report - Shoreham Nuclear Power Station l

! 3.6 l

TABLE 3.2. Significant Problems Encountered in Shoreham Emergency Diesel Generators During Testing Date Problem 3/81 Excessive turbocharger thrust bearing wear 12/81 Piston modifications to prevent crown separation 6 9/82 Engine jacket water pump modifications 6/82 Air starting valve capscrews too long for holes 9/82 Engine jacket water pump shaft failed by fatigue

~

Spring 1983 Cracks in engine cylinder heads Spring 1983 Cam gallery cracks in all three engines 3/83 Two fuel oil injection lines ruptured 3/83 Engine rocker arm shaft bolt failure 8/12/83 Broken crankshaft; cracks in other two crankshafts 9/83 Cracked connecting rod bearings 9/83 Cracked bedplates in area of main journal bearings 9/83 Unqualified instrument cable 10/83 Cracked AF piston skirts Spring 1984 Failed turbocharger nozzle ring capscrews and lost turbocharger nozzle ring vane on EDG 103 Spring 1984 Failed turbocharger thrust bearings, all three-EDGs Spring 1984 Cracks between cylinder liner bore and head studs (ligament cracks), EDGs 101, 102, and 103 Spring 1984 Stud hole to stud hole crack in EDG 103 I- Spring 1984 Deep crack in EDG 103 block from cylinder No.1 down

! front of block i

! Unit 1 dated June 29, 1984. The results of the confirmatory tests were l reported in a document published in early December 1984 entitled TDI Emergency Diesel Generator 10310 7-Cycle Confirmatory Test / Inspection Report, Shoreham Nuclear Power Station Unit 1 (undated). The details of LILCO's findings are discussed in Section 4.0 herein on a component basis. The results are o therefore not repeated here.

The conclusion drawn by LILC0 from both the Do./QR test / inspections and the j

confirmatory test / inspections is that all three SHPS EDGs are now suitable to l serve their function as standby emergency power sources.

l 3.7 i

i

., , .. , , - - . - . - -.-,..n.-- , --- . < ,,- -,

3.6 PNL EVALUATION In evaluating LILC0's engine tests, inspections, and component upgrades, PNL reviewed all available documentation of the tests, inspection results, engine operating history, and testimony / exhibits from the ASLB hearings on the SNPS TDI engines. Based on this review, PNL concludes that the testing and inspection program is adequate to uncover problems with engine components and e to confirm their ability to meet the load and service requirements. The com-ponent upgrades are viewed as responsive to the inspection findings and to the ,

recommendations of the OG. PNL notes that the tests conducted on the SNPS EDGs have subjected the engines to a number of starts comparable to that expected in actual service for the life of the plant. PNL finds that a sufficient number 7

. of hours (746 hours0.00863 days 0.207 hours 0.00123 weeks 2.83853e-4 months or 10 cycles) has been accumulated on EDG 103 to meet the criterion for proving the absence of high-cycle fatigue in the crankshaft.

l l

l i

l 1

3.8 l

4.0 REQUALIFICATION OF COMPONENTS WITH KNOWN PROBLEMS This section documents PNL's review of LILC0's actions to upgrade and/or requalify the 16 engine components known to have had significant problems (termed Phase 1 components). These components were previously identified by the Owners' Group through a review of the operating histories of TDI engines in nuclear and non-nuclear service.

Each Phase 1 component is discussed individually. The discussions are

~

i presented in a sequence reflective of component location within, on, or about the engine. The sequence generally progresses from bottom to top; that is, structural components, power train components, ancillary and auxiliary systems and components, on-engine and then off-engine.

Each component is described in terms of its function, operating history, and status as determined by the TDI Owners' Group and LILCO. This description is followed by PNL's evaluation and conclusion (s).

PNL's conclusions generally incorporate, without stating, the assumed commitment by LILCO to the modifications to their maintenance and surveillance i

program that are described in Section 5.0 of this TER, as well as the utility's commitment to appropriately implement the applicable recommendations and l

l requirements resulting from the NRC final review of the OGPP concerning these f- components. The conclusions also reflect PNL's finding, based on a sampling examination of LILCO's procedures for dispositioning component inspection l

findings, that these procedures are adequate with respect to both documentation I

and engineering considerations.

l 4

4.1 l

1 4.1 ENGINE BASE AND BEARING CAPS Part No. 03-305-A and 03-305-D Owners' Group Report FaAA-84-6-53 4.1.1 Component Function The' engine base itself supports the crankshaft and upper structures, and ,

carries the thrust of the cylinder combustion loads to the main bearings. The shaft is bedded in half-circle bearings set within " saddles" in the base. The bearing caps are structural members that hold the upper bearing shells in place over the shaft main journals while also absorbing the upward, reciprocating piston inertial loads. The studs and nuts hold the cap and therefore the shaft in place. A failure of base, cap, or bolting would allow shaft gyration or misalignment, potentially leading to shaft fracture and seizure, sudden engine stoppage, and possible ignition of crankcase vapors.

4.1.2 Component Problem History Four incidents of cracking have occurred in the engine base saddles of inline DSR-4 engines, causing this component to be evaluated as a generic issue:

e SNPS EDG 102, reported following an inspection in September 1983 e SNPS EDG 103, reported following an inspection in September 1983 e U.S. Coast Guard cutter Westwind (a TDI DSR-46 engine) e ' U.S. Coast Guard cutter Northwind (a TDI DSR-46 engine).

4.1.3 Owners' Group Status Failure Analysis Associates (FaAA), a consultant to the Owners' Group, analyzed the base, bearing saddles, bearing caps, nut pockets, and bolting /

nuts. FaAA conducted a finite element analysis 'to determine stresses acting on -

critical sections of the bearing saddle under lateral loading from the crank-shaft. The loads were determined from a journal orbit analysis. The bearing ,

cap, through-bolts, bearing studs, and nuts were similarly analyzed. The studs and bolts were tested for hardness.

4.2

l l

l FaAA concluded that the base assembly components have the strength necessary to operate at full rated load for indefinite periods, provided that all components meet manufacturer's specifications, that they have not been damaged, that mating surfaces are clean, and that proper bolt preloads are maintained.

o The Owners' Group concluded that the cracks in the engine base saddle of SNPS EDG 102 were due to the crankshaft failure. The cracks in EDG 103 resulted from improper engine disassembly procedures. Cracks in both U.S.

Coast Guard cutters' engine base saddles were the result of undertorquing.

4.1.4 LILC0 Status Cracks were found in the main bearing saddles of EDG 102 and EDG 103 during an inspection (September 1983). FaAA evaluated these cracks and concluded that the EDG 102 cracks resulted from the crankshaft failure and those in EDG 103 were caused by improper engine disassembly procedures. In September 1983 the engine base saddle for EDG 101 was also inspected and no cracks were noted. However, because no inspection requirement or instruction existed at that time, no records were made. LILCO concluded, based on the analytical testing and inspection cited in Section 4.1.3 above, that adequate margins of safety for ultimate and fatigue loading exist for the main bearing saddles. Based on the finite element analysis, also cited in Section 4.1.3, they further concluded that the existing cracks in EDG 102 and 103 bearing saddles will not propagate. They also concluded that adequate safety margins exist against failure of the through-bolt and bearing cap bolt nut pockets.

Periodic . inspection (at alternate refueling cycles) via fluorescent-dye penetrant of the EDG 101,102, and 103 bearing saddles is to be done to verify that cracks will not initiate and that existing cracks will not propagate.

4.1.5 PNL Evaluation and Conclusion PNL notes that no cracks were found in EDG 101. PNL believes that the origin of the cracks observed in EDGs 102 and 103 was properly diagnosed and

'that the analysis conducted is appropriate to conclude that those cracks will not propagate in service. PNL also concurs with the periodic inspections planned to verify that the cracks will not grow.

4.3 L

On the basis of the inspections, diagnostics, and actions taken by LILCO,

- PNL concludes that the engine base and bearing caps in EDGs 101, 102, and 103 are acceptable for their intended service, subject to a confirmatory inspection to be performed according to the 0GPP recommendations noted in Section 5.0 of this TER.

3 1

d 9

4.4

4.2. CYLINDER BLOCK Part No. 03-315-A Owners' Group Report FaAA-84-5-4 4.2.1 Component Function

. The cylinder block, which is bolted to the engine base, provides structural support for the cylinder liners, cylinder heads, camshaft and valve assemblies, and other miscellaneous components. It also serves as the outer

~

boundary for the engine coolant. The block is subjected -to both mechanical and thermal stresses resulting from the combustion processes. Structural failure of the block could lead to inadequate support of components that confine combustion pressures, and thereby result in a sudden engine shutdown.

4.2.2 Component Problem History Cracks have been reported in cylinder blocks of both DSR-4 (inline) and DSRV-4 ("V") engines in nuclear and non-nuclear applications. Several types of cracks have occurred in cylinder block tops. Cracks have also occurred in the camshaft galleries of inline engines, in the vertical wall just above the camshaft bearing supports.- The following is a summary of the types of cracks and the engines in which they have been found.

1. Ligament cracks - A ligament crack is oriented vertically and extends between the counterbore for the cylinder lirier landing and a cylinder head stud hole. Numerous cracks of this type have been identified in the top surfaces of the Shorehan EDG 101, EDG 102, and original EDG 103 engine blocks. Crack maps for the three blocks are presented in FaAA-84-5-4, Design Review of TDI R-4 and RV-4 Series Emergency Diesel Generator Cylinder Blocks and Liners.

Ligament cracks have also been reported by FaAA in the marine and stationary installations listed below. These engines have operated with such cracks from 6,000 to 28,000 hours0 days 0 hours 0 weeks 0 months .-

s T

4.5 L_ u

T TDI Engine Series Installation DSR-4 Copper Valley Electric Corporation DSR-4 MV Trader DSR-4 MV Traveler DSRV-20-4 Homestead, Florida DSRV-16-4 MV Gott .

DSRV-16-4 MV Columbia 3

2. Stud-to-stud cracks - A stud-to-stud crack is also oriented vertically, and extends between two cylinder head 'itud holes of adjacent cylinders. In nuclear applications, stud-to-stud cracks have been identified only in the original block for the Shoreham EDG 103 engine. Following replacement of the crankshaft in that engine and an engine test of 100 hours0.00116 days 0.0278 hours 1.653439e-4 weeks 3.805e-5 months at or above the nameplate rating of 3500 kW, a crack was discovered that extended between two adjacent studs on the exhaust side of cylinders No. 4 and 5. Later, after EDG 103 had experienced an abnormal load excursion while being operated at full load, and had then been operated for a brief period (less than two hours) at 3900 kW, reexamination of the engine block revealed additional between-stud cracks. Furthermore, the original stud-to-stud crack between cylinders No. 4 and 5 had grown, as documented in the FaAA report referenced above. (The original EDG 103 block was replaced, as discussed later in this section.)

3. Circumferentia1 cracks - Cracks of this type are found in the corner formed by the cylinder liner landing and the cylinder liner counter-bore. They may extend circumferential1y around the landing and down-ward into the block. Such cracks were discovered in the original EDG 103 block through destructive metallurgical examinations, which ,

revealed a maximum crack depth of approximately 3/8 inch. Because of the relatively sharp corner where these cracks occur, they are dif-ficult to identify through nondestructive tests. PML anticipates that similar cracks may occur in the EDG 101 and 102 blocks, because 4.6

of the relatively high stress concentration associated with the geometry of the cylinder liner landing.

4. Cam gallery cracks - This type of crack appears as a horizontal indication in the upper radius of a camshaft bearing support, and extends in essentially a horizontal plane toward the engine jacket

, cooling water system. Cracks of this type have been discovered in the cam galleries of the EDG 101, EDG 102, original EDG 103, and replacement EDG 103 cylinder blocks. Weld repairs that are essentially cosmetic in nature were performed on the cam gallery cracks in the first three blocks. These repairs did not involve complete removal of the crack; furthermore, additional cracking occurred between the weld " nuggets" and the base material in all three blocks. The cara gallery cracks in the replacement EDG 103 block are much shallower than those in the other blocks.

Another crack of a type that differed from those described above appeared in the original EDG 103 block after the follcaing sequence of events. During a test at full load, EDG 103 experienced an abnormal load excursion. The engine slowed to 390 rpm, at which time a breaker tripped, removing the electrical load. The engine continued to operate at rated rpm (450) for about 10 minutes, and was then shut down. After the engine was restarted and loaded to 3900 kW, a crack was observed extending down the front of the block from cylinder No. 1, and the engine was again snut down. Reexamination of the block revealed addi-tional stud-to-stud cracks discussed earlier in this section. LILCO decided to replace the block.

Metallurgical examinations of the original EDG 103 block by. FaAA revealed an extensive degenerate graphite microstructure that produced markedly inferior mechanical properties. FaAA concluded from metallurgical examinations of the

- EDG 101 and 102 blocks that they did not exhibit similar degenerate microstructures.

. Several indications were discovered in the DSRV-16-4 engines at Comanche Peak that also differ from the types of cracks described above. These indica-tions are oriented vertically and extend radially into the block from the cylinder liner landing and cylinder liner counterbore. Through metallurgical 4.7

I 1

examinations, FaAA identified these cracks as interdendritic shrinkage or porosity resulting from the casting process. They have not been found in any other TDI engines in nuclear service.

4.2.3 Owners' Group Status Because no cracks other than those found in the Comanche Peak engines have been reported for any other TDI engines in nuclear service, all efforts have -

been directed toward determining the significance of the various cracks in the SNPS engine blocks. ,

To this end, FaAA on behalf of the OG conducted an investigation that consisted of 1) an analysis of loads on the block that influence fatigue and fracture and 2) a stress analysis to estimate the levels of stresses caused by ,

these loads, as input to their fracture and fatigue life evaluation.

The load analysis considered the combined effects of 1) the preload on the cylinder head studs, 2) the load distribution between the head and the block, f 3) the load between the head and liner, and 4) the thermal and pressure loads

! - between the liner and the block. These loads were used as input to the stress analysis to provide estimates of the stress levels in the block.

l The stress analysis included strain-gauge testing on EDG 103 at various loads and types of starts, as well as two- and three-dimensional finite element analyses of the top of the block. The finite element analyses were used to i

1) analyze the stresses in the ligament resulting from firing pressure, l 2) obtain the ratio of stresses in the ligament resulting from thermal expan-f sion, 3) detemine the radial stress distribution on the inside surface of the block resulting from a unifom pressure on the inside surface of the liner for i

both the cracked and uncracked ligament, and 4) determine the effect of varying

! the liner-to-block radial clearance. The results of the finite element analy- ;

ses were used to gain insight on the distribution of stresses and to determine .

! scaling factors to relate stresses at gauge locations to those at the crack i initiation sites. <

In addition, sections of the original EDG 103 block were cut out and sub-jected to full metallurgical tests of materials, including fractography and metallography, and visual inspection of cracks in counterbore to stud hole, 4.8

. . - - .,,,,..m._.,,,z.,,....m_= . , _ , . , _ . - , . , , , _ _ , . .._,,,m._.--,m,-,,,__,.--...m, .-.,, , -..,,,,vm,%,y,

- - .. - - . _ - - - . - - . - _ - _ _ _ _ - _ - . - ~

stud hole to stud hole, and counterbore radii and camshaft gallery areas.

Metallurgical tests were also conducted on samples from EDG 101 and 102 blocks.

FaAA findings are summarized as follows:

. e Initiation of cracks in the ligament between stud hole and liner counterbore is predicted to occur after accumulated operating hours

at high load and/or engine starts to high load. These cracks are f benign because the cracked section is fully contained between the

, liner and the region of the block top outside the stud hole circle.

Field experience is consistent with both the prediction of ligament

, cracking and the lack of immediate consequences. These cracks are not expected to extend below the cylinder liner counterbore landing

, (aoproximately 1.5 inch deep.)

1 ,

o The presence of ligament cracks between stud holes and liner counter-l bore increases the stress and the probability of cracking between the stud holes of adjacent cylinders such that stud-to-stud cracks are predicted to initiate after additional operating hours at high load and/or engine starts to high load. The deepest measured crack in

! this region was originally estimated to be approximately 5.5 inches

deep, but later, when a cutout section was available for measurement, i

determined to be 3.9 inches deep. This did not degrade _ engine

[.

operation or result in stud loosening.

e The apparent rate of propagation of cracks between stud holes in the

'origiral EDG 103 block at SNPS, when compared with LOOP /LOCA require-ments, indicates that blocks with ligament cracks are predicted to f withstand a LO0P/LOCA event with sufficient margin, provided that 4 1) inspection shov: no stud-to-stud cracks prior to the event and

2) the specific block material of EDG 103 is shown to be sufficiently 1ess resistant to fatigue than typical gray cast iron, Class 40.

I Metallurgical tests and photomicrographs demonstrated that EDG 101

. and EDG 102 block material had the appearance and ultimate tensile

(- strength of typical gray cast iron, Class 40. However, the material of the EDG 103 original block was found to be of a degenerate ;

4.9 1

grr.phite composition with ultimate tensile strength much inferior to that of typical gray cast iron, Class 40.

e The block tops of engines that have operated at or above rated load should be inspected for ligament cracks. Engines such as those at Catawba and Grand Gulf that are found to be without ligament cracks can be operated without additional inspection for combinations of ,

load, time, and number of starts that produce less expected damage than the cumulative damage prior to the latest inspection. The allowable engine usage without repeated inspection can be determined from cumulative damage analysis.

The blocks of engines that have been operated without subsequent inspection of the block top should conservatively be assumed to have ligament cracks for the purpose of defining inspection intervals.

e For blocks with known or assumed ligament cracks, the absence of detectable cracks between stud holes of adjacent cylinders should be established by eddy-current inspection before the engine is returned to emergency standby service after any period of operation at or above 50% of rated load. If crack indications are found, removal of the adjacent heads and detailed inspection of the block top are necessary. In addition, it is necessary to ensure that the mic.rostructure of the block top does not indicate inferior mechanical properties.

Engines that operate at lower maximum pressure and temperature than those in the SNPS engines may have increased margins against block cracking that could allow relaxation of block top inspection recuire-ments. Modifications to other parameters such as increased liner-to-block radial clearance and reduced liner protrusion above the block (proudness) will reduce stresses, and site-specific analyses of such modifications could also permit relaxation of inspection requirements. ,

e The cracks in the cam gallery of the EDG 101 and 102 blocks and the EDG 103 replacement block are shrinkage cracks that originated during 4.10

the cooling-down period after the blocks were cast, while they were still in the mold. During operation the areas in question are under continuous compressive stress and, thus, pose no problems due to crack growth.

4.2.4 LILCO Status As part of the DR/QR program, LILCO assembled and reviewed the component documentation including the Owners' Group evaluation of the component described

. in the previous section. They also performed a series of dimensional checks and NDT examination of the block that are summarized as follows.

Engine 101 e A liquid penetrant test was performed on the cylinder liner landing along the top landing surface, fillet radius, and vertical face adjacent to the surface on cylinders No. I through 8. Indications of landing cracks were reported and reviewed by FaAA. Based on operating experience, FaAA judged these indications to be normal and to have no impact on the safe operation of the engine.

e Liquid penetrant and ultrasonic tests were performed in the area of the cylinder block stud holes. These tests revealed linear indica-tions in the landing area of the stud hole counterbore for cylinders No. 1, 2, 3, 4, and 6. The indications were reviewed by FaAA and, on the basis of experience, judged to not compromise safe operation of the engine.

e Visual, liquid penetrant, and ultrasonic tests of cylinder block liner landings performed in conjunction with a 100-hour test reyealed indications on cylinders No. 3, 4, 5, 7, and 8. Indications wEre reviewed, and judged to be normal and to have no impact on the

+ function of the seating surface.

Eddy-current tests were performed on the cylinder block in the area

. between the cylinder head studs. No relevant indications were reported.

4.11

e Measurements were taken to establish the as-built dimensions of all cylinder block liner landings, and recorded.

Engine 102 e Liquid penetrant tests were performed on the cylinder liner landing along the top landing surface, fillet radius, and vertical surface adjacent to cylinders No.1 through 8. Linear indications were -

found, which, after review, were judged to not compromise the safe operation of the engine. ,

e Visual and eddy-current tests were performed prior to and after the 100-start test. The locations of all indications were reported to FaAA, who judged them to be not detrimental .

e Liquid penetrant tests were performed around the head stud bolt circle on top of the block. Ligament cracks were reported to FaAA, who judged them to be not detrimental.

Eddy-current tests were performed on the cylinder block in the stud-to-stud areas between the cylinders, and no relevant indications were revealed.

e Liquid penetrant tests were performed on cam gallery saddles No. 3 and 5, revealing linear indications. The indications were reported to FaAA, who judged them to be not detrimental.

The dimensions of the cylinder block liner landings were verified.

Engine 103 Replacement Cylinder Block (following the 746-hour confirmatory test) e Fluorescent magnetic particle examination of the block top surface revealed no recordable indications. Eddy-current examination of the four adjacent stud holes between cylinders No. 4 and 5 revealed no ,

recordable indications.

Prior to the endurance run, liquid penetrant and magnetic particle ,

examinations of cam saddles No. 2 and 8 and the areas adjacent to the bolts were performed and indications mapped. A surface resistance probe was used to measure the depth of the indications. These 4.12

I l

examinations were repeated following the confirmatory testing.

Comparisons of the results of these tests established that these indications did not grow as a result of the testing.

e Prior to the endurance run, strain gauges were placed in critical areas of the camshaft gallery. Adjacent tie rods were loosened and the strain gauges were adjusted to zero. Next, the tie rods were retorqued to their proper values, and the strain gauges registered compressive stress. The strain gauges were monitored and readings

~

were recorded at varying load conditions. At no time during operation did the readings indicate tensile stresses.

The replacement block for EDG 103 is a typical gray cast iron, Class 45.

Ultimate tensile strength has been checked and proven to be normal. Class 45 is superior to Class 40 gray cast iron. Therefore, the replacement block for EDG 103 is of a superior strength material to that of EDG 101 or 102.

On the basis of the review of the design review documentation and the results of the testing and inspections summarized above, LILCO concluded the cylinder blocks for EDGs 101,102, and 103 are acceptable for their intended function at SNPS.

4.2.5 PNL Evaluation and Conclusions PNL's review of the SNPS EDGs included consideration of 1) the FaAA design review of the cylinder b1'ocks, 2) inspection reports for the SNPS engines, and

3) the testimony exhibits of the applicant, intervenor, FaAA, and NRC given at the ASLB hearing. The implications of the observed camshaft gallery, ligament, circumferential, and inter-stud hole cracking observed in both nuclear and non-nuclear applications were considered.

4.2.5.1 Camshaft Gallery Cracks

- Evidence available from recent tests and metallurgical investigations strongly suggests that the known camshaft gallery cracks originated during the casting and subsequent cooldown of the cylinder blocks, and that the cracks have not grown since that time. Strain-gage measurements taken by FaAA on EDG 103 demonstrate that the areas where the camshaf t gallery cracks occur are subject to compressive stresses during engine startup, operation, and shutdown.

i 4.13 I

_ . . . _ _ _ _ _ _ _ . . _ _ _ , . _ . ~ . , _ _ _ _ _ _ _ _ _ _ _ _ _ _ . . . _ _ . - . _ . . . ,. _ __._ . _ _ _ _ _ _._. . .

Although PNL concurs that compressive loads introduced during engine assembly should prevent growth of the cam gallery cracks, PNL is less certain of the level of residual stresses in the vicinity of the cracks and the consequences of those stresses when compressive loads are reduced or removed. The residual stresses could conceivably lead to crack " pop in" when a block is unbolted from its base. It is also conceivable (although admittedly unlikely) that the unknown residual stresses, combined with reduced compressive stresses during -

engine operation, could exceed the imposed compressive stresses at the crack tip and lead to crack growth during operation. Therefore, PNL is of the ,

opinion that monitoring of crack behavior is indicated for the camshaft galleries of EDG 101 and 102. PNL's recommendations, and reasons for not recommending that EDG 103 also be monitored, are presented in Section 5.1.1.2.

4.2.5.2 Circumferential Cracks in Liner Bore Circumferential cracks in the liner counterbore and counterbore landing were observed in the Shoreham engines and in other engines in non-nuclear applications. These cracks were not analyzed in the FaAA original design review; however, they were later dealt with by both visual examination of cracks in the cutout section of the original EDG 103 block. PNL believes that the FaAA analysis of the origin of cracks, namely stresses induced by cylinder liner proudness, is correct.

Further, FaAA's finite element analysis of the area reveals that the ,

above-described region of high tensile stresses is immediately surrounded by a region of high compressive stresses resulting from the bolt-up of the cylinder head to the block. Therefore, it is PNL's judgment that any cracks formed in ,

the cylinder liner counterbore and landing would be rapidly arrested as they move into the region of compressive stress, and will not represent any hazard to engine reliability. This judgment was supported by the results of section-ing of the circumferential crack that had propagated only 1/8 to 3/8 inch into ,

the' block even though this block had degraded mechanical properties. Further confirmation that such cracking is benign is furnished by operating experience; there are no records of any nuclear or non-nuclear engine failing because of cracks of this type.

4.14

4.2.5.3 Ligament Cracks PNL concludes that, provided the ligament cracks in cylinder blocks for EDG 101 and EDG 102 are properly monitored as indicated in Section 5.0 of this TER, they will not impair service during an eventual LOOP /LOCA event. This conclusion is based on review of FaAA analyses, including three-dimensional finite element analysis, LILCO's DR/QR, and the fact that, although numerous reports on cylinder block ligament cracks exist from TDI DSR-4 and DSRV-4 engines in operation, there are no reports on these cracks rendering an engine

~

nonfunctional.

4.2.5.4 Stud-to-Stud Cracks Stud-to-stud cracks are considered more serious than ligament cracks because they degrade the overall mechanical integrity of the block and its ability to withstand firing pressures and piston side thrust. The analysis performed by FaAA indicated that, once ligament cracks occur, the stresses in the stud-to-stud region increase, providing a greater potential for cracking in this region. From cumulative damage analyses, FaAA determined that approxi-mately the same amount of accumulated damage would be required to form stud-to-stud cracks following the formation of ligament cracks as would be needed to originally cause the ligament cracks themselves. Furthermore, the amount of damage that would be caused by operation during a LOOP /LOCA accident would be much less than that required to produce a stud-to-stud crack greater than 4 inches deep. Therefore, FaAA concluded that a block was able to meet its intended function if tests showed the absence of stud-to-stud cracks.

Based on the FaAA analysis of the cracks present in the SNPS blocks and on the LILCO inspection results showing the absence of cracks between studs of l adjacent cylinders, PNL concluded that the cylinder blocks currently installed

! in EDGs 101,102, and 103 are suitable for continued use. This conclusion is l

subject to verification that no cracks have developed between stud holes of l

adjacent cylinders in EDGs 101 and 102 following each operation of the engine at 50% of qualified load or above. If cracks are found, further analysis should be made to determine the suitability of the block for continued service. Because the absence of ligament cracks in the block of EDG 103 was l confirmed following the completion of the confirmatory tests, inspection for i

! 4.15

inter-stud cracks 'is not necessary. However, it is recommended that the block of EDG 103 be reinspected for ligament cracks at intervals based on the formula described in the report FaAA-85-5-4.

In consideration of the above cited analyses and inspections and PNL's examinations of the blocks, PNL concludes that the blocks installed on EDGs 101,102, and 103 are acceptable for the intended service, subject to ,

monitoring of cracks as noted above.

W O

4.16

1 l

4.3 CRANKSHAFT Part No. 03-310-A Owners' Group Report FaAA-84-3-16 4.3.1 Component Function

. The crankshaft receives the reciprocating power strokes from the cylinders (via the pistons and connecting rods),. converts them to rotary motion, and transfers the shaft power to the generator. It also drives the gear train that operates the camshaft, which, in turn, operates the cylinder-head valves, fuel' injection pumps, governor, etc. The crankshaft is supported by journal bear-ings mounted in the engine base. The crankshaft begins as a forged steel' billet, which is subsequently formed into the crankshaft configuration by a further process of forging and twisting, after which it is machined. By means of holes drilled throughout the crankshaft, pressurized oil is picked up from the main journal bearing supply points and transmitted to connecting rod bearings, wrist pins, undersides of the pistons, and other parts.

The crankshaft is subject to a variety of very complex stress fields.

These include direct and torsional shear stresses and bending stresses due to the piston thrusts; inertial effects of reciprocating masses; torsional, axial.

and flexural vibration stresses; bending stresses due' to overhung flywheel; bending stresses due to wear-down in main journal bearings; and variation in external support alignments. These nominal - stress combinations are augmented in local stress fields due to the stress-raising influence of oil holes and crankweb/ journal transition zones. Residual stresses due'to forging and heat

~

treating procedures, operating conditions, and operating accidents also affect the final stress spectrum. The machined surfaces of the crankshaft journals-

.and crankpins are subject to damage from oil impurities, bearing deterioration, and excessive heat. Therefore, crankshaft failures may occur. At worst, a crankshaft may actually fracture (through fatigue) and separate,- leading to -

' immediate engine shutdown and probable significant conjunctive damage to other 1

4.17 t

components. Precursory damage leading to failure (such as cracking) can some-times be prevented via surveillance and maintenance (e.g., periodic crankshaft deflection checks).

4.3.2 Component Problem History In August 1983, the SNPS EDG 102 crankshaft fractured during plant pre-operational tests. This fracture occurred at the crankpin journal of cylinder

No. 7, separating the crankshaft into two pieces. The fracture involved the web connecting the No. 7 crankpin journal to the adjacent No. 9 main bearing ,

j ournal . Inspection revealed severe cracking in the crankshafts of the other two SNPS engines. Independent studies performed by FaAA and the Franklin Research Center subsequently determined these failures to be due to torsional vibrations. No other torsional failures of DSR-48 crankshafts have been reported.

The original crankshafts that had 11-inch diameter crankpins with 1/2-inch fillets were subsequently replaced with new crankshafts having 12-inch diameter crankpins with 3/4-inch fillets.

4.3.3 Owners' Group Status The OG initiated an extensive investigation of the causes of the SNPS crankshaft failure. FaAA and SWEC were retained by LILC0 to carry out inten-sive inspections, and analytical and experimental investigations. The NRC reauested that the Franklin Research Center provide an independent review. The conclusion of these investigations was that the crankshaft failed from tor-sional vibration stresses resulting from operation too near a critical speed.

The Owners' Group next evaluated the adequacy of the replacement Shoreham crankshafts. This was performed by FaAA and consisted of 1) reviewing TDI calculations of stresses fron single torsional vibration modes and SWEC torsio-graph tests on both the old and new crankshafts to verify that the new crank- '

i shafts did meet Diesel Engine Manufacturers Association (DEMA) standards and 2) performing a fatigue analysis of the crankshaft to determine the factor of ,

safety against fatigue. In addition, TDI obtained certification from the American Bureau of Shipping ( ABS) for sizing of the crankpins, journals and webs.

4.18

The analysis of the factor of safety against fatigue failure consisted of

1) a torsional dynamic analysis to compute the nominal stresses at each crank throw, 2) a three-dimensional finite element analysis to determine local stresses in the crankpin fillet, 3) stress measurements at the points of maxi-mum stress indicated by the finite element analysis, and 4) a determination of the factor of safety by comparing the measured stresses with the endurance limit for the failed Shoreham crankshaft.

FaAA reached the following conclusions (which are documented in

~

FaAA-84-3-16):

e The TDI calculations of stresses using single orders are appropriate and show that the stresses in the replacement crankshafts are below DEMA recommendations for single orders of torsional vibration.

e The SWEC torsiograph tests show that the stresses in the replacement crankshafts are below DEMA-recommended limits for both single and combined orders of torsional vibration at 3500 kW (100% load) and at 3800 kW. A linear extrapolation to 3900 kW also shows compliance.

Calculations of torsional stresses over the range within five percent above and below rated speed (450 rpm) at 3500 kW show compliance with DEMA within the accuracy of the analysis. These stresses were cal-culated by FaAA using the modal superposition method together with harmonic data obtained by SWEC at 3500 kW and 450 rpm.

e On the basis of an endurance limit established for the failed crank-shafts and scaled to account for the higher ultimate tensile strength of the replacement crankshafts, together with stress levels computed from strain gauge data, the factor of safety against fatigue failure of the replacement crankshafts is 1.48 for operation at 3500 kW.

This factor of safety does not account for the beneficial effects of shotpeening, and is even greater if the shotpoening of the Shoreham crankshafts is considered.

The replacement crankshafts are suitable for unlimited operation in the emergency diesel generators at SHPS at the nameplate engine rating of 3500 kW and at the two-hour-per-24 hour rating of 3900 kW.

4.19 L

Other evaluations of the adequacy of the replacement crankshafts were per-fomed for LILC0 by Dr. Franz F. Pischinger, president of FEV (Research Society for Energy, Technology and Internal Combustion Engines) and a professor at the University of Aachen in West Germany; and by Dr. Simon K. Chen, owner and president of Power and Energy International, Inc., a private consulting firm in Beloit, Wisconsin. Dr. Pischinger independently reviewed the work performed by '

FaAA on the crankshafts, and he compared the design of the crankshafts against the Kritzer-Stahl design criteria. He concluded that the crankshafts should have unlimited life for operation at 3500 kW, and that the crankshafts should -

be able to operate at 3900 kW for a minimum of 600 hours0.00694 days 0.167 hours 9.920635e-4 weeks 2.283e-4 months . Using 12 orders of vibration and harmonic coefficients based on data from Lloyd's Registry of Shipping standards (" Guidance Notes on Torsional Vibration Characteristics of Main and Auxiliary Oil Engines," 1976), and the TORYAP computer program, Dr.

Chen concluded that the replacement crankshafts comply with DEMA standard practices at 3500 kW and 3900 kW.

4.3.4 LILCO Status The replacement crankshafts installed by LILC0 were manufactured by the West Geman finn of Krupp Stahl, A.G. using a forged slab, hot-twist fabri-cation process. Nondestructive examinations performed by Krupp included ultra-sonic testing (to detect subsurface flaws) and magnetic particle inspection (to detect surface and near-surface flaws). Krupp's inspections revealed no relevant indications.

The fillet areas of two of the three replacement crankshafts were shot-peened by TDI before these crankshafts were shipped to LILCO. They were repeened for LILCO by Metal Improvements Company, Inc., when receipt inspection revealed that the original shotpeening did not meet LILC0 requirements. Before the repeening was performed, the crankshafts were subjected to magnetic par-ticle testing and liquid penetrant testing of the fillets. These inspections ,

showed that no relevant indications were introduced in the first shotpeening.

The third crankshaft was shipped directly from Krupp to LILCO, and shotpeened ,

by Metal Improvements Company.

As part of LILC0's DR/QR program for the engines, the three crankshafts were reinspected in the areas of highest torsional stress after each crankshaft 4.20

d had been operated for approximately 300 hours0.00347 days 0.0833 hours 4.960317e-4 weeks 1.1415e-4 months , including approximately 100 hours0.00116 days 0.0278 hours 1.653439e-4 weeks 3.805e-5 months at 3500 kW and above. The reinspection of each crankshaft involved high-resolution eddy' current testing and liquid penetrant testing of the crankpin journal fillets of cylinders No. 5, 6, 7, and 8. No rejectable indications were found.

- On October 22,1984, LILC0 submitted a Final Safety Analysis Report (FSAR) amendment to the NRC staff for a " qualified" engine load of 3300 kW (i.e., the maximum emergency load that would be imposed on any of the three engines under design-basis accident conditions). In a letter to the NRC staff dated October 18,1984, LILC0 described the protocol for a 740-hour confirmatory test ,

of the EDG 103 engine at the qualified load of 3300 kW. The test was completed early in November 1984.

Inspections performed by LILCO following the test referred to above included 1) liquid penetrant testing of all crankshaft fillet areas and external radii of oil holes except for fillets and oil holes at main bearings No.1, 2,10, and 11; 2) eddy' current inspection for evaluation of all record-able indications; and 3) eddy' current inspection of oil holes to three inches

from the journal surface except oil holes in main bearing journals No.1, 2, 10, and 11. (The latter journals, which are not the most highly loaded, were not accessible for inspection because the cylinder block remained installed on the engine base.) All recordable liquid penetrant indications were evaluated by LILCO and found to be nonrelevant.

l Summarized in Table 4.1 are the loads and corresponding hours accumulated on all three engines following the installation of the new crankshafts. This information was provided by LILCO.

i 6

i 4.21

TABLE 4.1. Loads and Engine Hours Engine Number

. 101 102 103 Total hours (all loads) 611 557 1323 Approx. hours at 3300 kW 0 0 525 Approx. hours at 3500 kW 147 117 119 Approx. hours at load greater than 3500 kW 99 68 101 Approx.' hours at 3900 kW 6.5 7.5 7 Number of starts 216 250 319 -

4.3.5 PNL Evaluation and Conclusion PNL reviewed the post-test inspection 0F EDG 103 following the endurance test described above. The purpose of the review was to determine, through an independent audit of the condition of key engine components, whether or not they exhibited any evidence of abnormal behavior under the conditions imposed during the test. This audit was performed by consultants under contract to PNL who have extensive experience in diesel engine technology, and by a PNL specialist in nondestructive testing. PNL's findings on the crankshaft are documented in a report dated December 3,1984, for the Atomic Safety and Licensing Board. In summary, PNL's consultants found nothing in their visual inspections of the crankshaft journals and the corresponding bearing shells that would be indicative of crankshaft deficiencies. Furthermore, no reject-able indications were found in the nondestructive examinations witnessed by PNL's NDT specialist.

1

[

PNL's consultants also performed independent reviews of the adequacy of the replacement crankshafts relative to DEMA standard practices and to the rules established by several classification societies for marine engines.

Although TDI was not obligated to follow rules of marine classification ,

societies in the design of the Shoreham engines, such rules provide a con-servative basis for an independent evaluation. The results of PNL's reviews are summarized as follows:

l 4.22 1

[

e DEMA Prof. Sarsten, a PNL consultant from the Norwegian Institute of Technology at Trondheim, Norway, used a computer program called COMHOL to calculate torsional stresses for single orders and for the sum of 24 orders of vibration. His analysis employed the same harmonic data used by FaAA. His results predict that the crankshaft stresses meet DEMA standards for single orders, but exceed DEMA standards for the sum of orders at 3500 kW. At 3300 kW, his results predict that the crankshaft torsional stresses are below DEMA limits for the sum of orders over the speed range of 5% below rated speed through 5% above rated speed, except for that portion of the speed range above 466 rpm. His results predict that stresses exceed the DEMA limit of 7000 psi by a maximum of approximately 250 psi at 473 rpm (at 3300 kW) .

e American Bureau of Shipping PNL consultants confirmed that crankshaft web dimensions satisfy ABS rules. However, torsional stresses predicted by PNL consultant A.

Sarsten for single and combined orders at 3500 kW (approximately 3600 psi and approximately 7100 psi, respectively) exceed limits calculated by TDI (3357- psi and 5035 psi, respectively) that would be allowed under the 1984 ABS rules.

e International Association of Classification Societies Ricardo Consulting Engineers of England calculated the factor of safety of the Shoreham replacement crankshafts for PNL according to the proposed rules of the IACS. For 3200 kW, 450 rpm, and a maximum cylinder pressure of 1650 psi, the calculated factor of safety was 0.926 in comparison to the IACS-proposed minimum of 1.1.

e Det Norske Veritas Allowing for the stationary application of the Shoreham engines and for an estimated influence of shotpeening, this classification society in Oslo, Norway, concluded that the crankshaft safety margin would not be adequate for loads exceeding 3200 kW.

4.23

PHL has the following comments on the various analyses of the replacement crankshafts and the recently-completed endurance test of EDG 103:

e In light of the conflicting results of the analyses performed for LILCO and the analyses performed independently for PHL, the analy-tical evidence alone does not provide a sufficient basis for con-cluding that the crankshafts are adequate for the qualified load of ,

3300 kW.

e The crankshafts for the Shoreham engines do not have to meet any or ,

all of the requirements of the various marine classification societies. Even if a crankshaft does not meet such rules, it may still perfom adequately. The rules of the classification societies contain inherent conservatisms that reflect the rigors and uncer-tainties of marine service. Furthemore, the rules are often subject to interpretation and discussion with the classification society, and l approval does not necessarily depend on strict compliance with the rules.

e The results of the EDG 103 test completed in November 1984 provide important and, in PNL's view, definitive information regarding the l

fatigue resistance of the Shoreham crankshafts for service at l

i 3300 kW. The test duration of 746 hours0.00863 days 0.207 hours 0.00123 weeks 2.83853e-4 months at the rated engine speed of 7

450 rpm corresponds to just above 10 crankshaft stress cycles at or above the cualified load in a 4-cycle engine. This number of cycles is generally accepted as sufficient to demonstrate high-cycle fatigue resistance in metal structures, provided that no cracks develop under the conditions imposed during the test. The post-test examinations of the EDG 103 crankshaft demonstrated that it had completed the test with no indications of cracking.

On the basis of the following considerat. ions, and subject to the recom-mendations for surveillance discussed later in this section, PNL concludes that the replacement crankshafts for EDG 101, EDG 102, and EDG 103 are acceptable ,

for their intended service, provided that they are not operated during engine 4.24 3 < >

tests at loads in excess of the qualified load of 3300 kW(a) . The primary considerations on which this conclusion is based are as follows:

e The torsiograph tests performed for LILCO by SWEC as discussed earlier in this section provide experimental evidence that the crankshaft torsional stresses are essentially in compliance with DEMA standards at 3500 kW and the rated engine speed of 450 rpm. Al though torsiograph tests were not conducted for underspeed and overspeed conditions at that power level, the results at rated speed provide a level of assurance that actual torsional stresses at 3300 kW are likely to be essentially in compliance with DEMA standards over the limited frequency range and associated speed range to which the EDGs are controlled at Shoreham, e Ultrasonic tests of the crankshafts during manufacture revealed no significant subsurface defects. Magnetic particle, eddy-current, and liquid penetrant examinations performed on several occasions prior to installation of the crankshafts in the engines and following opera-tion of all three crankshafts at load levels to 3500 kW have revealed no rejectable indications.

e The 746-hour endurance test of the EDG 103 engine at or above the qualified load of 3300 kW and the absence of any rejectable indica-tions on critical crankshaft surfaces following that test provide definitive evidence of the fatigue resistance of the design under the conditions imposed during the test.

(a) In a report filed with the ASLB on December 3,1984, the NRC staff identi-fied several transient conditions under which the. SNPS diesel generators could be loaded for brief periods (a few seconds to a few minutes) above the qualified load of 3300 kW during an energency. It is PNL's understand-ing that these transient loads would not exceed 3900 kW. Recognizing that the engines have operated for many hours at loads above 3300 kW (see Table 4.1), that inspections following such operation have revealed no defects in the crankshafts, and that the duration of any transient loads above 3300 kW will be very brief relative to the hours already accumulated, PHL concludes that the transients will not jeopardize the operability of the engines.

4.25

e The three engines have comparable engine load histories above 3300 kW as shown in Table 4.1.

In light of the results of the analyses performed for PNL by Ricardo -

Consulting Engineers and by Det Norske Veritas as summarized earlier in this section, PNL has concluded that it would be prudent to examine certain high-stress areas of all three crankshafts periodically to confirm that no cracks ,

develop in service. These examinations should include the. nondestructive tests listed below. If these examinations reveal nothing of significance, LILCO may wish to propose a change in the examinations to NRC.

During the first refueling outage, the fillets of the three crankpin journals (Nos. 5, 6, and 7) subject to the highest stresses should be examined with liquid penetrant and, as necessary, eddy current in the crankshafts of both the EDG 101 and 102 engines. The fillets in the two main journals between these three crankpins should also be examined in this manner. In addition, the oil holes in these crankpin and main bearing journals should be examined in the manner used in the most recent examination of _the EDG 103 crankshaft. These inspections are not considered necessary for the EDG 103 crankshaft at the first refueling outage because of the inspection performed on this crankshaft in November 1984.

e In subsequent refueling outages, two of the three most heavily loaded crankpin journals in each of the three crankshafts should be examined as noted above. The main bearing journal between them should also be examined in this manner.

0 4.26

i 4.4 CONNECTING RODS d

Part No. 03-340-A Owners' Group Report FaAA-84-3-13 4

4.4.1 Component Function The primary function of the connecting rod is to transmit the engine cylinder firing force from the pistons and piston pin through the rod to the e crankshaft such that the reciprocating motion of the pistons induces rotation and output torque of the crankshaft. The connecting rod must have sufficient column buckling strength and fatigue resistance to withstand the cylinder

. firing forces and ir.ertial loads. The wrist pin bushing (or rod-eye bushing) and the crankpin bearings are contained by the connecting rod. The flexure of

, , the rod must be such that the bearings are not unacceptably distorted. The passages within the rod must remain unblocked to provide cooling and lubrication to the bearings and pistons. Sufficient clamping force must be maintained by the bolts on the connecting rod cap to prevent relative motion of the components The rod cap bolts must support the necessary preload without yielding, fracture, or unacceptable thread distortion. The wrist pin bushing must support the cylinder firing forces and inertial forces.

4.4.2 Component Problem History i

Only one inservice failure of connecting rods in TOI DSR-48 series engines has been reported. This . failure consisted of a longitudinal split through the j oil hole in a DSR-46 engine at Glenna11en, Alaska (Copper Valley Electric Corporation) . Reportedly, this crack was initiated from fatigue. The failure report supplied by TDI' did not identify the origin of the crack; however, no material abnormalities were reported. This engine had operated for over 8000 hours0.0926 days 2.222 hours 0.0132 weeks 0.00304 months and, for part of that time, at noch higher peak firing pressures (1975 psi) than those measured for the Shoreham engines (1680 psi).

4.4.3 Owners' Group Status The adequacy of the TDI inline connecting rods was addressed by FaAA for the Owners' Group. .The objectives of their efforts were to assess the f . structural integrity of connecting rods in TDI model DSR-48 engines in standby 4.27 I ,. - .-

emergency diesel generator sets at Shoreham, River Bend, and Rancho Seco nuclear power stations, and to determine the connecting rods' suitability to perform their required function.

The Owners' Group evaluation considered four major parts of the inline connecting rod assembly: ttie rod-e3e bushing, the rod eye, the connecting rod bearing housing and cap, and the connecting rod itself. The rod-eye bushing, ,

which is of the same design as those in the V-engines, was analyzed because

! linear indications have been found in the bronze bushings during field inspec-s tions. Journal orbit analyses, metallurgical evaluations, and stress and

{ fracture mechanics analyses were performed. The rod-eye end of the connecting rods was evaluated by stress and fracture mechanics analyses, which included assumed surface flaws. The connecting rod bearing housing and cap were evaluated by stress and fatigue analyses. The connecting rod itself was analyzed for buckling stability.

The connecting rod is attached to the crankpin bearing cap with four bolts extending entirely through the connecting rod. Prestressing of these bolts I

creates compressive stresses in the connecting rod itself and tensile stresses in the bolts. The two extreme loading conditions, firing stroke and exhaust f stroke, were considered. The stresses in the bolts and connecting rods were determined for the two load cases, and the fatigue crack propagation in the bolts was investigated because they were the most critically stressed compo-nent. A critical crack deptt of 0.133 inch was determined at the thread root.

l While cracks in the root of the bolt threads are not permitted, the analysis showed that a crack as large as the critical crack could be tolerated and would not propagate. Fatigue was determined not to be a problem.

The buckling stability of the connecting rod was assessed under the maxi-mum cylinder firing pressure. The margin factors of 6.28 for yielding and of 5.72 against lateral buckling of the connecting rod were determined. ,

Wrist pin bearing performance was analyzed using a journal orbit analysis computer program. The oil pressure profiles imposed on the rod-eye bushing ,

under piston firing and inertial loads were determined. A peak oil film pressure of 97,400 psi was predicted to occur at the bottom of the bushing due 4.28

4 to power stroke. A peak oil film pressure of 5000 pst(a) was also predicted by '

7 FaAA to occur at the top of the bushing due to the inertial effects of the exhaust stroke. These two cases provided input to a rod-eye bushing stress anal ysis.

The calculated circumferential stresses and the oil film pressures were

used as input to a fracture mechanics analysis. This fracture mechanics model indicated that bushing defects would not propagate if they originate on the

, outside diameter. The model also indicated that bushing defects on the inside diameter will not propagate unless they originate within +15 degrees from the bottom center. Even if inside diameter (ID) defects are within +15 degrees of

the bottom center, they are predicted not to propagate unless the crack faces are exposed to the full range of oil film pressure. Because of the compressive hoop stress in the bushing, it was considered unlikely that the crack faces would separate and allow oil pressure to be exerted.

! In conjunction with the rod-eye bushing stress analysis, the rod eye 1 itself was analyzed with the same finite element and curved beam models and for the same load cases. The stress range calculated was below the fatigue l initiation stress range for the rod material. Because of the possibility of

pre-existing defects, as in the case of the Glenna11en failure, the threshold crack size for fatigue was estimated by a fracture mechanics analysis using conservative values for the threshold range of stress intensity factor. A 0.043-inch deep flaw was determined to be the critical crack depth for the

! maximum tensile stress range (calculated) for load case 1. For load case 2, ,

the maximum critical crack depth of 0.04 inch at the rod eye was determined.

The Owners' Group could find no explanation for the one reported rod eye fatigue failure. However, fracture mechanics analyses indicate that fatigue cracks could propagate from a 0.04-inch deep surface discontinuity at the intersection of the oil hole with the bore of the rod eye. Such discontinui-l ties on the smoothly polished surfaces were felt to be readily apparent on

.. visual examination.

. .(a) The FaAA value reported in FaAA-84-3-13, page 2-4, was 500 psi. This was corrected by G. Derbalin (LILCO) in a telephone conversation with D. Dingee (PNL) on December 9, 1984.

4.29 .

i I

. - . , _ _ _ . . _ _ _ . _ . _ . - . _ . . _ _ . _ . _ . . _ . . . _ _ . _ _ _ . ~ . . _ _ , _ . _ . - . . _ _ . _ . _ . _ _ . _ , , _ ,

t t

1 n

Based on their evaluations the OG concluded that the inline DSR-48 con-necting rod is adequate for its intended purpose, provided there are no bushing defects in the region within 15 degrees on either side of the bottom dead center of the bushing.

4.4.4 LILC0 Status Rod-eye bushings in the replacement connecting rods for all three SNPS EDGs were inspected prior to the startup testing. Linear indications were found by liquid penetrant testing on all rod-eye bushings and were determined -

to be casting defects. These indications were found on both the inside and outside diameter of the bushings. Similar indications were also found in new, unused bushings. No fatigue growth of these cracks was noted after 100 hours0.00116 days 0.0278 hours 1.653439e-4 weeks 3.805e-5 months ' operation at full load. The linear indications were determined to be the i

result of interdendritic shrinkage or porosity. No service-induced fatigue extension of the casting defects was observed.

Metallurgical evaluations were performed on several bushings. A chemical analysis performed on one of the bushings removed from a SNPS engine showed the bushing composition to be within the range for the specified bronze alloys, C93200 or SAE 660.

As a result of the DR/QR inspection, all the following wrist pin bushings, which had indications of cracks within +15 degrees of the bushing bottom, were replaced:

o EDG 101 - bushings in connecting rods 1, 2, 3, 4, 6, and 8 1

e EDG 102 - bushings in connecting rods 1, 3, and 8 l

EDG 103 - bushings in connecting rods 1, 3, and 5.

Following the 746-hour confirmatory test on EDG 103, all wrist pin bushings were again inspected with liquid penetrant. None of the bushings i showed any porosity. They did show slight wear patterns and some light ,

scratches.

In addition to the bushings, LILC0 also confirmed the condition of the ,

rod-eye, the rod bolts and the connecting rod itself. The connecting rods were inspected at TDI with liquid penetrant with a LILCO inspector present. This

inspection considered the whole connecting rod including the rod eye. The 4.30

connecting rod bolts were inspected at TDI and subsequently inspected visually at SNPS prior to rod assembly. LILCO confirmed that the connecting rods cur-rently installed in the Shoreham EDGs did not contain any cracks or discon-tinuities that could lead to fatigue failure.

As a result of these inspections of the connecting rod and wrist pin bushing, LILCO believes that these components are acceptable for their intended design function.

4.4.5 PNL Evaluation and Conclusions The PNL reviewers evaluated the Owners' Group report and supplementary infonnation on inline connecting rods. They found that the Owners' Group examined the appropriate significant failure modes (namely, the cracks in the rod-eye bushing; fatigue in the rod eye itself; fatigue and possible pre-tension loss in the connecting rod bolts; stiffness and buckling of the con-necting rod; and size of the oil cooling holes and path). The bounding load cases of exhaust stroke inertial loads and firing pressure loads were correctly used in the analyses. The analytical methods used by the Owners' Group were judged to be appropriate.

Both known and postulated cracks in components have been included in the Owners' Group analyses. PNL concurs with the Owners' Group position that linear indications are acceptable in the rod-eye bushing so long as they do not occur within +15 degrees of the bottom center, because the indications are in compression. PNL also concurs that cracks larger than 0.046 inch deep in the rod eye or 0.133 inch at the root of the bolt threads are not acceptable.

PNL also reviewed the inspections performed by LILCO: 1) inspection of the connecting rods and bushings for signs of distress; 2) liquid penetrant test on all wrist pin bushings; 3) determination of the connecting rod and cap

- material; 4) determination of the hardness of the connecting rod and caps; and

5) the TDI/LILCO inspection procedures for the rod eyes and bolts. Based on these evaluations and reviews, PNL concludes that the connecting rods and O

bushings installed in the Shoreham engines are acceptable for the intended service.

4.31

4.5 CONNECTING R00 BEARING SHELLS Part No. 03-340-B Owners' Group Report FaAA-84-3-1 4.5.1 Component Function The connecting rod bearings interface the connecting rods with the crank- ,

shaft. They are of cast aluminum alloy with a thin babbitt overlay, and are furnished in two identical halves. They are lubricated under pressure, and a substantial flow of oil proceeds through machined channels in the shells from the drilled crankshaft oil holes to the passageways within the connecting rods and on to the pistons and intervening bearing surfaces. The upper bearing half is subject to the piston firing loads and is therefore more susceptible to failure.

Failure can occur through inadequate oil flow or pressure, excessive or unplanned loadings, structural anomalies (from design or manufacture), or fatigue and erosion of the ba5bitt lapr in crucial areas. Bearings are also subject to particle, chemical. or water contamination of the oil, or improper oil selection for the duty, either of which can lead to degradation and failure. The failure mechanism usually is gradual, and its onset generally can be detected by prudent surveillance of oil and filter conditions. However, a substantial structural problem, excessive cylinder loads, or heavy water contamination can lead to rapid failure. This can affect the crankshaft journals, sometimes with irreparable results.

In light of the severe conditions affecting bearings, the need for replacement is not uncommon. However, in customary service, bearing life generally is measured in multiples of 10 4hours, given reasonable service conditions.

4.5.2 Component Problem History Five incidents of cracking in the SNPS EDG connecting rod bearing :; hells have been reported. All but one occurred during operation with the original 11-inch crankshafts and were discovered during disassembly after the crankshaft failure on EDG 102. A number of bearings, other than the cracked ones, have 4.32

- . ..- - .-- - - - _ ~ - -- . - . - - - - .- . .

F i also been replaced because of inservice conditions or nonconformance with the Owners' Group criterion for subsurface voids. No other connecting rod bearing shell incidents have been reported on any DSR-4 engines. .

j 4.5.3 Owners' Group Status Failure Analysis Associates analyzed the connecting rod bearing shells for

the Owners' Group. The analyses, which encompassed both 11-inch and 12-inch l

diameter shells, included:

e journal orbit analysis to determine the pressure distribution in the hydrodynamic film o finite element analysis to determine the stress distribution in the connecting rod bearing shell e fracture mechanics analysis to determine the resistance to fatigue cracking e computation of acceptance criteria using radiographic NDE e evaluation of babbitt adhesion.

Based on their analyses, FaAA concluded that the cracking of the four -

11-inch diameter bearing shells was due to bearing shell overhang causing undue bending stresses. They attributed the crack in the 12-inch bearing shell to '

excessive voids in the subsurface of the bearing shell in the area of the

> crack. The overall conclusion was that, provided they conform to the manu-

! facturer's specifications and meet the criterion for subsurface voids developed '

by FaAA, the bearings are suitable for the intended service.

l 4.5.4 LILCO Status

! Following recommendations and instructions issued by FaAA and approved by the Owners' Group, LILCO performed radiographic and liquid penetrant examina-

! tions on all 16 bearing shells in each engine.

'~

The crankpin bearing shells for EDGs 101, 102, and 103 were inspected after the nominal 100-hour test, which was run to support the DR/QR activities.

The crankpin bearings for EDG 103 were again inspected after the 746-hour endurance test.

4 4.33 2

7 l

l The results of these inspections are as follows: ;

I e EDG 101 - All the crankpin bearings were found to be satisfactory and reusable except the upper shells for cylinders No. 7 and 3, which were replaced because they did not meet the OG criterion for subsur-face voids. The lower shells for cylinders No. 4 and 6 were aporoved ;

for use as lower shells for 300 hours0.00347 days 0.0833 hours 4.960317e-4 weeks 1.1415e-4 months of running. The shells for ,

cylinder No. 5 were interchanged, the lower shell being approved by i FaAA for 300 hours0.00347 days 0.0833 hours 4.960317e-4 weeks 1.1415e-4 months of use as a lower bearing shell. Analysis relevant to the use of lower shells No. 4, 5, and 6 for 300 hours0.00347 days 0.0833 hours 4.960317e-4 weeks 1.1415e-4 months of - ;

operation was provided by FaAA in a letter dated March 3,1984, from l C. H. Wells (FaAA) to P. Martin (LILCO). !

e EDG 102 - The crankpin bearing shells for this engine were found in satisfactory condition except the upper shells for cylinders No. 2, 5, and 8 and the lower shells for cylinders No. 5 and 8, which were replaced because they did not meet the OG criterion for subsurface voids. ,

1 e EDG 103 - All except two crankpin bearing shells for this engine were found to be satisfactory. The lower shell for cylinder No. 2 was !

damaged in handling; the upper shell for cylinder No. 6 had a small surface inclusion on its outside surface. Both were replaced. .

- 4.5.5 PNL Evaluation and Conclusion Based on review of the FaAA analyses and LILCO inspection reports, and on a number of visual inspections conducted by PNL consultants, PNL concludes that the connecting nHi bearing shells are acceptable for the service intended. The ,

above conclusion is based on the cor.dition that if the bottom bearing shells 1 No. 4, 5, or 6 for EDG 101 should exceed the allowable 300 hours0.00347 days 0.0833 hours 4.960317e-4 weeks 1.1415e-4 months of operation during the first refueling cycle, they must be replaced prior to starting the ,

cycle.

9 4.34 i

4.6 PIST0h SKIRTS Part No. 03-341-A Owners' Group Report FaAA-84-2-14 4.6.1 Component Function '

4

. The piston (an assembly that includes the piston crown, piston skirt, rings, piston pin, etc.) receives the thrust of inertia and combustion and transfers it to the connecting rod. The cast steel crown is subject to the direct combustion pressure and thermal conditions. The skirt, made of ductile

, iron, actually transfers the load to the piston pin / connecting rod and guides the reciprocating motion of the piston within the cylinder. Such a two-piece piston structure is relatively common to large, modern, high-output engines.

In general, failure is most apt to result from excessive pressure and thermal stresses of both high-cycle and low-cyle character. Durability is i

affected by material selection, fabrication quality, and design characteris-tics. A crown separation will require immediate shutdown; it is likely to lead quickly to serious cylinder, head, and rod damage, and to piston seizure, with adverse impact on the crankshaft and possible crankcase explosion. Hence, adequate attachment of crown to skirt is necessary.

4.6.2 Component Problem History TDI has utilized several skirt designs, including types AH, AN, AE, and modified type AF, in their R-4 series engine. Most early engines for nuclear service were furnished with type AF and AH skirts, although one plant had AN skirts. The SNPS engines were originally furnished with 23 modified type AF piston skirts and one type AN skirt.

The modification to the type AF skirt, perfonned by TDI in 1981, consisted

, of spot-facing each of the four bosses through which the studs extend to secure the piston crown and replacing the originally supplied spherical washer set with two stacks of Belleville washers. This spot-finishing reduced the height of the stud attachment bosses from 2 inches to approximately 0.25 inch.

During an early inspection of the SNPS piston skirts, all 23 of the type AF piston skirts were found to contain linear indications in one or more of the 4.35

skirt-to-crown attachment bosses. The single type AN piston did not exhibit these indications. Subsequent metallurgical examinations of these indications revealed that they were fatigue cracks. Similar cracks were observed in the type AF piston skirts at Mississippi Power & Light (MP&L) Company's Grand Gulf Nuclear Station. LILCO subsequently replaced all 24 piston skirts in the Shoreham EDGs with type AE skirts of the latest design. This type AE design restores half the original height of the attachment bosses and incorporates one stack of Belleville washers instead of two. In addition, the piston bosses are wider and more smoothly blended into the skirt wall. -

Prior to their use at Shoreham, one of the msjor sources of experience wf th the type AE piston skirt was the experimental TDI R-5 engine. In this engine, the type AE piston skirts were observed to contain no cracks, even after 622 hours0.0072 days 0.173 hours 0.00103 weeks 2.36671e-4 months at a peak firing pressure of approximately 2000 psi.

4.6.3 Owners' Group Status The TDI Owners' Group experimentally and analytically evaluated both the type AF and type AE piston skirts. The 0G first evaluated the cracked type AF skirts to assess the :1ature of the problem. This evaluation revealed that the observed cracking was the result of fatigue. Subsequently, both skirt types were experimentally tested for stress in a static hydraulic test, and these stresses were evaluated by finite element analysis of the skirt only. Then, the thermal stresses in the piston crown were evaluated by finite element analysis, and their effect on the stresses in the skirt determined. Finally, a fatigue and fracture analysis was performed.

It was concluded that the type AF skirts would crack in service at TDI nameplate rating, but the cracks would not grow once they move out of the highly streced region near the boss. For type AE skirts, the analysis indicated tb . cracks may initiate at high loads but will not grow. On these bases, the OG concluded that the modified type AF skirts are adequate for service, provided that they are 100% inspected for cracks in the stud boss area prior to use and that they are inspected periodically. Recommendations for ,

operating load levels and inspection intervals were to be made on a plant-by-plant basis. Furthermore, the OG concluded that the type AE piston skirts as currently installed in the SNPS EDGs were adequate for unlimited life.

4.36

4.6.4 LILC0 Status As part of the component revalidation process, LILCO assembled and reviewed the component documentation including 1) the Owners' Group evaluation of the component described in the previous section and 2) NDT evaluations of the skirts performed prior to placing them into service. The utility also

- performed a series of NDT tests and inspections; these are summarized below.

EDG 101, in February / March 1984 - A liquid penetrant examination of

- the piston attachment bosses was performed on piston skirts No. 5, 7, and 8. Piston skirts No. 5 and 7 were found to be satisfactory. A nonrelevant indication was found on piston skirt No. 8. An eddy-current test was performed on the piston skirt attchment bosses on piston skirts No. 5, 7, and 8, and all three skirts were found to be satisfactory. Dimensions of the piston groove height, rir.g height, and piston pin bore diameter were confirmed for cylinders No. 5, 7, and 8. Piston skirts No. 5, 7, and 8 were visually inspected and signs of scuffing were observed. The pistons were cleaned and reinstalled.

e EDG 102, in February / March 1984 - A liquid penetrant test was per-formed on the piston skirt attachment bosses for cylinders No. 5, 6, 7, and 8. Satisfactory results were obtained. Eddy current tests were performed on the skirts of pistons No. 5, 6, 7, and 8 in the same area. The results were reported as satisfactory. Visual inspections were performed of the outside diameter of the skirts on pistons No. 5, 6, 7, and 8. Unsatisfactory conditions were corrected and the pistons were reinstalled.

e EDG 103, in March 1984 - A liquid penetrant examination was performed at the piston skirt attachment bosses for bolt attachment to the crown on pistons No. 5, 7, and 8. The test revealed that all areas examined were satisfactory. An eddy-current examination was per-

formed on pistons No. 5, 7, and 8 in the same area. The regions inspected were found satisfactory. Dimensions of the piston groove height, ring height, and piston pin bore diameter were confirmed for 4.37

l 1

i pistons No. 5, 7, and 8. A visual inspection of the skirt outside l diameter was performed and the piston skirt was returned to service.

I y e EDG 103, in May/ June 1984 - Pistons No.1 through 8 were visually inspected. A light carbon deposit was found on the outside edge of

the crown and down to the top ring on all pistons. No " unusual" i scuffing or scratching was noted on the outboard portions of the ,

pistons and piston skirts.

i e EDG 103, in November 1984 (following completion of the 746-hour l

endurance test at 3300 kW) - All eight cylinder liners were

inspected; no scuffing was found. Breakaway torque for crown-l to-skirt attachment bolts revealed no degradation of original torque i

values. Liquid penetrant tests of the piston skirt at the crown-

! to-skirt attachment bosses revealed no recordable indications. No j eddy current evaluations were required because no indications were found. Visual inspection of crown-to-skirt contact areas for exces-i sive and abnormal fretting revealed only minor, normal operational

fretting in several small areas. Specific activities performed by PNL representatives in conjunction with this inspection and the conclusions reached by these representatives are addressed in a report to the ASLB dated December 14, 1984.

Based on their review of component documentation and test results, LILCO concluded that no adverse indications exist relevant to the integrity of the AE f piston skirts currently installed in the SNPS EDGs; hence, these piston skirts j are acceptable for their intended design function.

! 4.6.5 PNL Evaluation and Conclusions

PNL's evaluation of the SNPS EDG piston skirts is limited to the type AE pistons, because this is the piston skirt type currently installed in the ,

engines.

The primary conclusion of the Owners' Group analysis of the type AE piston skirts was that cracks may initiate but will not grow. PNL reviewed this analysis and found the stress field in the region of the stud bosses so complex that is was difficult to conclude with any degree of certainty whether cracks 4.38

would initiate or not, and, if they did initiate, whether they would grow or not. However, available operating experience appears to support the conclusion that this piston type is suitable for its intended function.

This operating experience was obtained from both the TDI R-5 test engine and from the SNPS EDG 103 confirmatory test. In the R-5 engine, two type AE piston skirts were installed and the engines tested for 622 hours0.0072 days 0.173 hours 0.00103 weeks 2.36671e-4 months at 514 rpm and a peak firing pressure of 2000 psi, about 20% higher than that expected at Shoreham. The type AE piston skirts used in this test were not quite identical to the same type AE skirts used at Shoreham. However, they were sufficiently comparable to conservatively extrapolate the results to the Shoreham engines.

The 622 hours0.0072 days 0.173 hours 0.00103 weeks 2.36671e-4 months of operating time in the R-5 engine were equivalent to 9.6 x 106 stress cycles in the type AE skirts. This number of cycles very closely approaches the fatigue limit for long-term operability of a mechanical design. Therefore, this R-5 test engine experience gives considerable confidence that the type AE skirt design is adequate. The other experience was obtained in EDG 103 during the 746-hour endurance test at 3300 kW. This test subjected the piston skirts to in excess of 10 7stress cycles; subsequent nondestructive testing revealed no apparent crack initiation. The successful completion of this test without occurrence of apparent fatigue of the piston skirts provides considerable confidence in the suitability of the skirt design for the intended function.

PNL also visually inspected all the piston skirts identified in Section 4.6.4 above. Based on 1) the LILCO procedure for handling inspection findings, 2) the PNL' examination of many of the piston skirts, 3) the suitability of the design as indicated by the above-described experience,

4) the current serviceability of the piston skirts now installed in all three engines as confirmed by the component revalidation tests for all three EDGs, and 5) the NOT inspection of EDG 103 following the confirmatory test, PNL concluded that the type AE pistons in the SNPS EDGs are acceptable for the intended service.

4.39

4.7 CYLINDER LINERS-Part No. 03-315-C Owners' Group Report FaAA-84-5-4 4.7.1 Component Function Engines of this size and character are designed with individual, removable .

cylinder liners, which fit inside the cylinder block. The liners contain the pistons and are capped at the upper end by the cylinder head. Thus, they act ,

as containment for the firing forces, subject to the stress and heat thereof, and the reciprocating travel of the pistons. The outer surfaces are cooled by jacket water circulating within the block. The lower end is sealed against an opening in the block with 0-rings. The upper end has an external, circumferen-tial ledge, which seats on the block's " liner landing." The head is gasketed and bolted in compression against the upper liner annulus, to seal in the high-pressure combustion gases. The liner is of nodular iron, selected for its

~

strength, castability, and durability against the rubbing action of the pistons and rings.

Liners generally do not fail, but they can be adversely affected by inade-

, quate or inappropriate lubrication, the forces and heat of the combustion pro-cesses, the character of the pistons and rings, and the quality of fuels and oils. Failure most often is in the form of scoring by broken rings or carbon deposits, or " scuffing" by the action of the piston on the cylinder walls, due to one or more of the factors mentioned. If such conditions are severe enough, a piston will seize and cause significant damage to liner, head, and connecting i rod, and even to the crankshaft. A crankcase explosion can result.

4.7.2 Component Problem History Only one incident of cylinder liner " failure" in nuclear service is '

known. This failure occurred in 1982 at Grand Gulf when a piston crowr.

separated from the skirt during testing of the Division II engine and marred the liner. +

4.40

4.7.3 Owners' Group Status ,

I The OG included considerations of liners in their study of cylinder bloc ks. Two concerns were uncovered:

2 e The TDI design calls for the liner to protrude slightly above the top deck of the block, to ensure a tight, compressive fit against the head and gasket. However, this produces bending moments in the head and substantial shear stresses on the cast iron liner landing of the 4 -

block. Both aspects are suspect in some of the real or incipient failures in those components. TDI has approved remachining to reduce the protrusion, termed " proudness".

e The design also calls for a tight fit between the outer ring of the liner ledge and the matching counterbore of the block. There is some concern by the Owners' Group that this could increase hoop stresses in the block, which might lead to block cracks. TDI has approved reducing this fit in the cylinder block.

4.7.4 LILCO Status

^

The bores of all cylinder liners were inspected by the Owners' Group for dimensions, signs of interior, wear, scoring, scuffing, or cracking.

Cylinder liner No. 7 on EDG 101 showed a crack existing from the top edge of the liner down 2-5/32 inches on the inside surface. A metallurgical exami-nation suggested that the crack was not related to service loading. This liner was replaced.

Cylinder liner No. 5 on EDG 102 showed evidence of scuffing. It was brush deglazed and reused. Liner No. 7, however, was found pitted and was replaced.

EDG 103, in which both the liner landing height and outside diameter were

, remachined to reduce bending and hoop stresses, was dismantled for inspection following the recent 107 -cycle endurance test. The cylinder liners were not removed from the block. Evidence of liner surface spot glazing was found.

Following the inspection, all liners were deglazed by honing per TDI instruc-tions. The deglazing was witnessed by a TDI representative.

4.41

LILCO concludes that the liners in EDGs 101,102, and 103 are suitable for nuclear standby service.

4.7.5 PNL Evaluation and Conclusion PNL representatives viewed the liners from EDG 101 in March 1984 and those from EDG 103 in June and November 1984. The liners were glazed and showed some hard rubbing spots. However, their appearance was typical of liners that had been in service. The liners did not appear to have any scuffed surfaces or other defects that could not be removed by deglazing. -

PNL concludes that the liners in all SNPS EDGs are acceptable for their intended service. This conclusion is based upon:

e a review of LILCO's actions for all EDGs with respect to inspection, remachining, and replacement (as needed) e PNL's examination of the liners from EDGs 101 and 103 e the good service record for these liners.

r l'

V 4.42

4.8 CYLINDER HEADS !

Part No. 03-306-06-0F

' Owners' Group Report FaAA-84-15-12

. 4.8.1 Component Function

. The cylinder heads cap the cylinders and, with the cylinder liners, pro-vide the enclosure needed to direct the combustion forces against the pistons.

In the TDI engine design, each cylinder uses a separate cylinder head assembly.

The bottom surface of the cylinder head, facing the piston, is called the fire-deck. There is also a top deck to enclose the internal water cooling passages and an intermediate deck that provides structural . rigidity to the assembly.

The cylinder head assembly contains two inlet valves, two exhaust valves, a fuel injector, air starting valve, and a test cock.

Each head is bolted to the cylinder block by means of eight studs extending through the head from the block. On top of the cylinder heads are two more components: the subcover or rocker box, which supports the valve actuating mechanism, and a light top cover.

The TDI DSR-4 heads are cast from an alloy steel. The casting cores that produce the complex system of internal water, air and exhaust gas passages are large and are difficult to hold in place during the casting process. They can shift during manufacture, causing uneven and/or incomplete sections, and can lead to a variety of flaws or indications, some of which can be repaired during subsequent manufacturing processes.

Cylinder head deficiencies that have been experienced have tended to be mostly superficial linear indications with inconsequential results. However, some deficiencies have led to warpage or cracks. The latter, if through the jacket water passages, can result in the leakage of water into the affected

' cylinder when the engine is inoperative, and the introduction of combustion gases into the cooling jackets during operation. If an attempt is made to

, start an engine with water present in one or more cylinders, severe structural damage can result.

l t

l 4.43 i-t

.~.-_._.m._..~. . - _ _ _ _ , . - , _ . , _ _ _ _ . - . . . . , _ . _ _ . _ . , _ . , _ _ - , _ _ . , , _ . , - - - . . -

4.8.2 Component Problem History Numerous failures of TDI cast steel cylinder heads have been reported in both nuclear and non-nuclear applications. For identification, TDI cylinder heads have been classified by the Owners' Group as belonging to one of three groups. Group I heads include all those cast prior to October 1978. Group II heads were cast between October 1978 and September 1980. Group III heads were .

cast after September 1980. The distinction among groups involves both design changes to ' facilitate better casting control and improvements in quality con- ,

trol. Most instances of cracked heads have involved Group I heads. Only five instances of cracks resulting in water leaks have been reported in heads of Groups II and III, and these have all been in marine applications. Most of these cracks were observed to have originated at the stellite faced valve seats.

The most recently reported head failure of a TDI nuclear EDG occurred at Mississippi Power & Light (MP&L) Company's Grand Gulf Nuclear Station. A 2-inch through-wall crack occurred in the right exhaust port casting surface between the valve seat area and the exhaust valve guide in their Division I diesel engine. This crack allowed water from the cooling jacket to enter a cylinder; the presence of this water was detected during the "barring-over" of the engine with the cylinder cocks open. The specific head group classifica-

tion of this head was not reported. However, the affected head was supplied with the engine and had undergone 1500 hours0.0174 days 0.417 hours 0.00248 weeks 5.7075e-4 months of operation, including 335 hours0.00388 days 0.0931 hours 5.539021e-4 weeks 1.274675e-4 months at 100% load (7000 kW, 225 BMEP) and 31 hours3.587963e-4 days 0.00861 hours 5.125661e-5 weeks 1.17955e-5 months at 110% load. MP&L believes that this was a unique, isolated event.

4.8.3 Owners' Group Status l

l Failure Analysis Associates performed mechanical and thermal stress I calculations for the Owners' Group to determine if these heads are suitable for the intended service. The results indicated that heads from all three groups would be suitable. However, FaAA recomended that Group I and II heads be inspected for cracks using liquid penetrant and magnetic particle testing. .

They also recomended that the firedeck thickness be determined by ultrasonic testing. For Group III heads, sample inspection as described for Groups I and II was recomended. For all three groups, FaAA recomended that the engine be 4.44 l

1

i rolled over before manual start with the cylinder cocks open to assure that no water was leaked into the cylinders.

4.8.4 LILCO Status The SNPS EDGs originally were supplied with Group I heads. During early operation, leaks developed in three heads. These were attributed to casting 1 defects resulting from the coring and mold design or lack of stress relief.

l All heads were subsequently replaced with Group III heads. The replacement

. heads were operated for at least 300 hours0.00347 days 0.0833 hours 4.960317e-4 weeks 1.1415e-4 months , including 100 hours0.00116 days 0.0278 hours 1.653439e-4 weeks 3.805e-5 months at or above i

3500 kW load. Many were inspected by nondestructive testing in the DR/QR program as summarized below.

e EDG 101 - A liquid penetrant test was performed on the exhaust and intake valve seats and firedeck area between exhaust valves on cylin-

[

ders No. 5, 7, and 8 following about 100 hours0.00116 days 0.0278 hours 1.653439e-4 weeks 3.805e-5 months of full power opera-tion. The surface integrity of the inspected area was found to be -

satisfactory.

I

e EDG 102 - A liquid penetrant test was performed on the exhaust and i

intake valve seats and the firedeck area between the exhaust valves I on cylinder heads No. 5, 7, and 8 following about 100 hours0.00116 days 0.0278 hours 1.653439e-4 weeks 3.805e-5 months of full y al set n for s cracki w per d th v1 seat area with the valves in place on all four valves of all f eight cylinder heads. No evidence of cracking was noted.

f e ENG 103 - Liquid penetrant tests were performed on cylinders No. 5, 7, and 8 following about 100 hours0.00116 days 0.0278 hours 1.653439e-4 weeks 3.805e-5 months of full power operation. Head .

' No. 7 was found to be satisfactory. Indications on heads No. 5 and 8

' were dispositioned for engineering action. A visual examination was perfonned of the firedeck on cylinder heads No. 1, 2, 3, 4, 6, 7, and i

8 for indications of surface damage. Heads No. I and 6 were found to be satisfactory. Observations regarding cylinder heads L

4.45 L

l' . - - - - - . . _ _ _ . _- _

No. 2, 3, and 4 were resolved by engineering action. Superficial surface abrasions observed on heads No. 7 and 8 also were resolved by engineering.

The thickness of the firedeck of the heads of all three engines was verified by ultrasonic techniques.

Following the 746-hour confirmatory test of EDG 103 at the qualified load -

of 3300 kW, additional inspections were performed on the cylinder heads. These inspections included 1) ultrasonic inspection of the firedeck at six locations ,

to verify that minimum thickness requirement of 0.400 inch was met, 2) surface inspection (either liquid penetrant or magnetic particle) of intake and exhaust valves to verify their freedom from unacceptable surface defects, and 3) a determination if any heads had through-wall weld repairs of the firedeck where the repair was performed from one side only. Based on the results of these inspections, LILCO concluded:

With one exception the cylinder heads passed the confirmatory tests. The one exception was head No. 4, which, although having performed satisfactorily and having passed both ultrasonic and liquid penetrant inspections, was replaced due to having been plug-welded in the fuel injector area.

As a result of the design review of the cylinder heads performed by FaAA for the Owners' Group, a review of existing documentation on the cylinder

heads, and the results of the above tests, LILCO concluded that the cylinder l heads are acceptable for their intended function at SNPS, provided that the engine barring-over procedure is conducted at " appropriate intervals" after shutdown and before manual starting.

4.8.5 PNL Evaluation and Conclusions PNL reviewed the FaAA mechanical and stress analyses of the TOI cylinder ,

l l

heads, the service history of the Group III heads currently installed on all three SNPS EDGs, and the results of the nondestructive tests performed as part of the component revalidation program and following the 746-hour confirmatory l tests of EDG 103. PNL concluded that the cylinder heads currently installed on all three SNPS engines are acceptable for the intended service, provided that l

l 4.46

the engine is air-rolled at appropriate intervals with open cylinder cocks after and before planned operation to verify the absence of cracks that may allow water leakage into the cylinder. It is recommended that this procedure be performed 4 to 8 hours9.259259e-5 days 0.00222 hours 1.322751e-5 weeks 3.044e-6 months , and again 24 hours2.777778e-4 days 0.00667 hours 3.968254e-5 weeks 9.132e-6 months , after any operation and, thereafter, prior to any planned start. If leakage is indicated by the ejection of water or steam from any of the open cylinder cocks during air-rolling, the affected head should be removed, inspected, and replaced, if defective.

4 l

4.47

4.9 CYLINDER HEAD STUDS Part No. 03-315-E Owners' Group Report Emergency Diesel Generator Cylinder Head Stud Stress Analysis (SWEC March 1984).

-4.9.1 Component Function Eight studs per cylinder are used to bolt the heads to the cylinder block. Together they transmit the power load from the head to the block and ensure a required preload on the cylinder head gasket.

Head bolts are not normally found to yield or break; however, these occurrences are possible, due to faulty design, materials, or fabrication, or excessive firing pressure. Fatigue failure is a greater concern, given reasonable operating conditions. This will occur if preload is insufficient and the bolts go through many cycles of loading. Once a bolt yields or breaks, its neighbors must carry increased burden, and the head is unevenly stressed.

This generally results in escaping combustion gases, with the attending hazards of heat and fire, as well as physical and metallurgical damage to head and block.

4.9.2 Component Problem History To date, no cylinder head stud failure has been reported in the nuclear industry. However, some isolated failures have been reported in the non-nuclear field. The cause has not been established.

TDI has employed two basic stud designs recently. One is of straight shank diameter, and there has been concern that its tight fit within the block stud opening, coupled with inadequate preload, could put side thrusts on the block and contribute to block fractures. A second design uses a necked-down shank. This design not only avoids any possible stud-to-bore contact, but ,

reduces the preload needed to maintain positive stresses during the firing cycle.

4.9.3 Owners' Group Status Stone & Webster Engineering Corporation (SWEC) has analyzed both the old design studs and new necked-down studs developed by TDI to minimize potential 4.48 i

W i

cylinder block cracking. SWEC has concluded that both stud designs are adequate for the service intended, provided proper stud preload is applied.

4.9.4 LILCO Status The SNPS engines are all equipped with studs of the necked-down design.

No failures have been noted and no studs have been replaced. In connection with the DR/QR inspection effort the bolts were checked on a sampling basis on all engines for 1) visible signs of distress, 2) dimensional conformance to specifications, and 3) material hardness and compaction. In addition, on EDG 102, the breakaway torque was measured following 100 starts. All sampling tests were completed without any rejectable indication.

4.9.5 PNL Evaluation and Conclusion PNL concludes that the studs now installed on the SNPS EDGs will be acceptable for the intended service. This conclusion is based on the following findings resulting from PNL's evaluation:

e The SWEC analysis has satisfactorily demonstrated the stud design is adequate.

e No failures of cylinder head studs have occurred in TDI engines in nuclear service to date.

e LILCO's action of inspecting and torquing the studs is deemed acceptable.

e LILCO has confirmed to NRC that these bolts were installed with proper preloading.

4 4.49

4.10 PUSH RODS Part Nos. 03-390-C & D and 03-390-04-AB Owners' Group Report FaAA-84-3-17 4.10.1 Component Function Push rods transmit the cam action from the camshaft on the engine side to ,.

the' intake and exhaust valves in the head. One main rod extends from the cam-shaft to the subcover where it acts directly on the intake valve rocker lever. ,

' The second main rod transfers cam action to an intermediate rocker in the sub-cover and on through an' intermediate (connector) push rod to the exhaust valve rocker arms. They are subject to high-acceleration compressive forces and cylinder pressures on the valves as they respond to the cams. Fundamentally, these are steel tubes with rounded ends, to fit the various mating sockets.

A failure would, at the least, reduce valve action and, thus, cylinder perfomance. Total inoperability of a cylinder could result, but would not necessarily lead to immediate engine shutdown. Because these components are always in compression, failure modes are limited, assuming reasonably good design.

4.10.2 Component Problem History TDI push rods originally had tubular steel bodies fitted with forged and hardened steel end pieces, attached by plug welds. An estimated 2% reportedly developed cracks in or around the plug welds. A " ball-end" push rod design introduced later consisted of a tubular steel- body with a high-carbon steel ball fillet-welded to each end. This design proved to be prone to cracking at' the weld. A third design, consisting of a tubular steel body friction-welded on each end to a forged plug having a machined, hemispherical shape, was then introduced. This third configuration is referred to.as the friction-welded design.

4.10.3 Owners' Group Status Because industry (both nuclear and non-nuclear) had expressed concern about the continued integrity of TDI push rods, the TOI Owners' Group included the component in the known generic problem category for specific study and 4.50

resolution. Failure Analysis Associates performed stress analyses as well as stress tests to 10 ~ 7 cycles on samples of both the plug-welded and the friction-welded push rods, at conditions simulating full engine nameplate loading. No sign of abnormal wear or deterioration of the welded joints or ends was observed. Other nuclear owners have run these versions in actual service beyond 10 7 cycles with no adverse results. The 746-hour test on SNPS EDG 103 was completed successfully without any observed push rod failures.

FaAA concluded from their analyses and tests that both the plug-welded and friction-welded designs are adequate. They provided stipulations for inspec-tion and action, including destructive examination of a random sample.

4.10.4 LILCO Status 4

Originally the SNPS EDGs contained push rods of the fillet-welded ball-end design. Subsequently, the main push rods were replaced with those of the forged-head plug-welded design and the connector rods with those of the friction-welded design.

The component revalidation included visual inspection, LP inspection, and metallurgical analysis and hardness test of the rods on the three engines.

From the results, LILCO concluded that the installed push rods are acceptable

+

for their intended use. LILCO's conclusion is supported by the findings of the FaAA analysis.

4.10.5 PNL Evaluation and Conclusion PNL reviewed and concurred with the FaAA report. PNL also reviewed docu-mentation of LILCO's actions and noted the favorable record of push rods in extended service elsewhere. On these bases, PNL concludes that push rods of both the plug-welded and the friction-welded designs are acceptable for their ,

intended service.

O i

4.51

4 t ,

4.11 ROCKER ARM CAPSCREWS Part No. 03-390-G 1 <

Owners' Group Reports Emergency Diesel Generator Rocker Arm Capscrew Stress Analysis (SWEC March 1984, July 1984).

4.11.1 Component Function ,

The rocker arm capscrews bolt in place the rocker arm shaft in the sub-cover assemblies. They transmit camshaft rolling loads, valve spring loads, ,

and residual cylinder pressure forces from the rocker ann shaft to the cylinder heads. They are made from fairly standard bolting materials. A failure would weaken or cancel the restraints on a rocker shaft and cause malfunction of intake o exhaust valves. Reduced engine output would result.

4.11.2 Component Problem History Rocker arm capscrew failures due to improper bolt preload have been reported at SNPS. There have been no reports of similar failures elsewhere.

4.11.3 Owners' Group Status Stone & Webster Engineering Corporation performed stress analyses of both th'c original capscrew design with a straight shank (the type that failed at SNPS) and a newer design incorporating a necked-down shank. SWEC has concluded that both designs are adequate for the service intended. SWEC attributed the failure at SNPS to insufficient preload.

4.11.4 LILCO Status The DR/QR report states that, prior to April 1984, all rocker arm cap-

~

screws in the.SNPS engines were esplaced with the newer, necked-down design. A sample of these capscrews was inspected in connection with the Shoreham DR/QR and found'to be of the correct design material and hardness. The bolts were ,

installed with the specified 365-ft/lb preloads. LILC0 concluded on the basis of the analysis and inspections that the capscrews are acceptable for their intended service.

u 4.52

'k 4

4.11.5 PNL Evaluation and Conclusion PNL concludes that the rocker arm capscrews in all SNPS engines are acceptable for the intended service. This conclusion is based on 1) a review of the OG analysis, 2) LILCO's actions to replace all bolts, 3) the favorable checks of materials and design as-installed, 4) the confirmation of installa-tion preloading, and 5) LILCO's commitment to perform periodic preload checks as described in Section 5.0 of this TER.

4 6

4.53

4.12 TURBOCHARGERS Part No. MP 017 (Model BC0-90G)

Owners' Group Report FaAA-84-6-56 4.12.1 Component Function The turbochargers on the LILC0 TDI DSR-48 engines are Model 90G units .

manufactured by the Elliott Company. One turbocharger per engine provides pressurized air to the cylinders for combustion of more fuel than would be ,

possible with a "normally aspirated" engine. The turbochargers consist principally of a turbine, driven by engine exhaust gases, directly driving an air compressor wheel or impeller. The associated housing ducts the air and exhaust to and from the rotors; the exhaust inlet guide vanes direct the exhaust gases toward the turbine wheel blades. Turbine speed changes with engine load (i.e., gas volume, pressure, and temperature), with maximum speed depending on specific turbine selection and design parameters.

Because close tolerances and high rotating speeds are necessary for effi-ciency, and because temperature levels can approach 1200*F at the exhaust inlet, all components are sensitive to temperature, pressure, structural loads (vibrations), and contaminants or particles in the gas and air streams. The radial and thrust bearings require particular care and lubrication.

Vanes and blades are sometimes lost due to heat and vibration, or frac-tured by impact of particles, such as bolt heads, fractured vanes, or valves.

Undue stresses or vibration from connected exhaust piping or inappropriate supports can cause rotor wear at stator interface. Inadequate bearing lubrication (and the cooling the oil provides) can lead to bearing failure.

Depending on the severity of the situation, diesel engine shutdown can come quickly, but usually is not imediate.

4.12.2 Component Problem History Various problems have occurred in the turbochargers on TDI DSR-4 engines in nuclear service. The principal problem has been the rapid deterioration of the combination turbine thrust / radial bearing, which has occurred at the Shoreham, Comanche Peak, Catawba, and San Onofre nuclear plants. There also 4.54

have been problems regarding missing exhaust inlet vanes, missing or broken i capscrews joining the vane disc to the turbocharger at the inlet, and broken capscrews and welds in the support mounts. At Shoreham, capscrew failures have been reported in EDG 101, and lost nozzle vanes have occurred in EDG 103.

Y Because nuclear EDGs have, to date, had unusual quick-start requirements--

. and are tested extensively to assure reliability for such duty--the owners and TDI investigated the failure parameters early in the history of such service.

It was recognized that the bearing and bearing lubrication systems inherant in the 90G design were not adequate to provide lubrication of the bearing thrust pads and rotor thrust collars under fast startup conditions to high loads. TDI I initiated two steps of modifications in an attempt to address this problem; one instituted and modified the oil drip system and the second provided for manual prelubrication prior to planned starts.

4.12.3 Owners' Group Status On behalf of the Owners' Group, FaAA undertook an extensive study of causes of reported failures in nuclear service. The net result was an affirmation of inadequate startup lubrication. Briefly, the resulting recomendations were:

i e Retain and use a " drip system" that directs a small flow of oil toward the bearings at all times in standby, but increases the flow of oil to 0.35 gph. (Higher flows are apt to flood past the bearing into the exhaust manifolds and create fire risk a't startup.)

e Provide and use an auxiliary prelubrication pump to direct l substantial flow to the bearings immediately prior to planned

! 'startups.

e Maintain oil filtration at 10 microns or better and utilize

= spectrochemical and ferrographic oil analysis regularly.

Enhance bearing inspection programs. .At least one bearing should be e inspected at a station following every 100 starts of any nature'.

-Inspection should also be done following 40 starts without manual prelube.

4.55 l

In a separate study, FaAA also considered the various nozzle ring compo-nent failures that have been reported in Elliott 90G turbochargers. They concluded that, on the basis of operating experience, these types of failures do not affect the operation of the turbocharger and, therefore, do not com-promise the ability of the EDGs to perform their intended function. They did, however, recommend that the engine operation be monitored to ensure that ,

exhaust gas temperatures do not exceed maximums specified by Elliott.

4.12.4 LILCO Status As a result of LILC0's engine start tests, the turbocharger thrust bearings were wiped and had to be replaced. LILC0 subsequently instituted the drip system and later the prelube bearing oiling system.

In early 1984 after the new crankshafts were installed in the engines, the turbochargers were disassembled for inspection subsequent to the 100-hour engine qualification test. All three units were thoroughly inspected and the thrust bearing in each turbocharger was again replaced.

In addition to replacing the thrust bearings, LILCO made the following repairs to the turbochargers:

EDG 101 - replaced oil seal and snap ring, and staked inlet casting plugs, as welds were found cracked e EDG 102 - staked inlet casting plugs, as welds were found cracked e EDG 103 - replaced oil seal and snap ring, and staked inlet casting plugs, as welds were found cracked.

In addition to these efforts, LILC0 has committed to the OG plan calling for inspection of the turbocharger thrust bearings if any engine experiences 40 fast starts (starts without manual prelubrication of the bearings) or 100 total starts. ,

The turbocharger on EDG 103 was disassembled for inspection following the 746-hour crankshaft test. The turbocharger was found in good condition. The ,

thrust bearing did show a few slight circumferential dirt scratches across its 4.56

face, which were judged to be of no consequence by LILCO and PNL consultants.

In addition to the 10 -cycle 7 test, EDG 103 had made 319 starts prior to the turbocharger inspection.

Based on their having made the cited changes and implementing the OG/FaAA recommendations listed in Section 4.12.3, as well as the recent inspections and

,, the fact that the exhaust temperature is measured, LILCO concludes these turbochargers now installed will adequately perform their intended function.

. 4.12.5 PNL Evaluation and Conclusions

! PNL has reviewed the FaAA report referenced above, the results of the Owners' Group meeting with representatives of FaAA, the Owners' Group, NRC, and PNL, and the inspection data presented by LILCO. PNL also has examined the prelube system at other, similar plants. Based on these reviews and on the recent inspection of the EDG 103 turbocharger, PNL concludes that a similar prelube system now installed on th'e diesels at SNPS will provide sufficient

, additional lubrication to augment the protection of the turbocharger bearings during planned fast starts. Further, in PNL's view, the few unplanned fast l

starts that may occur without prelube during a given operating cycle will not l 1ead to bearing failure prior to scheduled maintenance of the bearing (see l Section 5.1.1.9). According to Failure Analysis Associates, as confirmed in a telephone conversation between PNL (W. Laity) and FaAA (T. Thomas) on July 20, 1984, the shortest known time-to-failure of a turbocharger thrust bearing l

subjected to " dry" starts (for which no forced bearing prelubrication was-provided) occurred at SNPS. That bearing experienced at least 62 " dry" starts -

before failure. The new operating procedure instituted at SNPS suggests that each engine is likely to experience very few, if any, " dry" starts in a given operating cycle.

PNL also notes that LILCO has established a planned program of relevant maintenance and surveillance and, at the next refueling outage, has agreed to implement the OG recommendations for inspections. LILC0 has also committed to

. comply with OG recommendations regarding capscrews, vanes, and mounting and r

supports that may result from the Shoreham DR/QR effort.

i l

! 4.57 l

PNL notes that the engines at SNPS will be run at a BMEP of about 213 psi. This engine rating is slightly below the TDI full load BMEP of 225 psi. This reduces the pre-turbine exhaust temperature, which is beneficial to the turbocharger.

On the bases of the above considerations and the recent inspection of the EDG 103 turbocharger, including the PNL consultant examination, PNL concludes ,

that the turbochargers on SNPS EDGs 101,102, and 103 are suitable for their intended service.

4 4.58 L

i 4.13 JACKET WATER PUMP Part No. 03-425-A Owners' Group Report Supplement to Emergency Diesel Generator Engine Driven Jacket Water Pump Design Review (SWEC July 1984).

,. 4.13.1 Component Function The engine driven jacket water pump furnishes water to the engine jackets (i.e., the cylinder block surrounding the liners, the heads, the coolers, and the exhaust manifold). Water is also circulated through the turbocharger water jackets. The pump is a typical centrifugal pump, driven from the front-end gear.

Without the water pump (or an emergency backup), the engine would quickly shut down due to excessive temperatures. Such pumps generally are trouble-free, but occasionally develop problems of shaft seals, bearings, and drive mechanisms.

l 4.13.2 Component Problem History The jacket water pumps at Shoreham have encountered one significant problem: a pump shaft failure. This led to redesign of the method of attach-ing the impeller to the shaft. There is no history of other jacket water pump failures.

4.13.3 Owners' Group Status Stone & Webster has investigated the jacket water pumps as installed on

i. the TDI inline and vee engines. They reviewed these jacket water pumps from the standpoints of mechanical design, material suitability, and hydraulic f

performance. SWEC found the pump as modified at SNPS is acceptable,- with the recomendation that the proper torque be used for holding both the gear and impeller on the shaft taper.

4.13.4 LILCO Status A water pump at SNPS sustained a broken shaft. The failure was analyzed l

to be caused by a stress concentration at a keyway. SNPS obtained a new jacket water pump in which the keyway was eliminated; the impeller in this new pump is 4.59

now held to the tapered shaft with a preloaded lock nut. The modified design was reviewed by SWEC, who stated it was satisfactory.

In connection with the Shoreham DR/QR effort, the jacket water pump for EDG 102 was subjected to:

e visual inspection for signs of distress or excessive wear or pitting of both the drive and driven gears -

e verification of bearing surface contact between impeller and shaft

~

e liquid penetrant examination on the contact surface of the impeller and gear to shaft, shaft taper, impeller, and gear bore e verification of material and hardness of shaft and impeller o visual inspection of the clearance or wear rings for evidence of galling or wear e dimensional checks of the shaft.

LILCO has reported all results to be satisfactory. No comparable inspections or tests were conducted for EDGs 101 or 103.

During the inspection of EDG 103 after 746 hours0.00863 days 0.207 hours 0.00123 weeks 2.83853e-4 months of operation at a load of 3300 kW, the water pump was examined in place in the engine. The water pump seal showed no leaks, and the water pump drive and drive gear showed no wear or pitting pattern.

Based on the OG design review studies and the inspections listed above, LILCO has concluded that the jacket water pumps are acceptable for their intended function at SNPS.

4.13.5 PNL Evaluation and Conclusion On the basis of the redesign of the impeller /shaf t attachment, and the

- fact that all SNPS EDGs are equipped with this new design, PNL concludes that .

these pumps are acceptable for their intended service.

4 4.60

4.14 HIGH-PRESSURE FUEL OIL TUBING Part No. 03-365-C Owners' Group Report Emergency Diesel Generator Fuel Oil Injection Tubing (SWEC April 1984).

4.14.1 Component Function The high-pressure fuel oil tubing carries the fuel oil from the cam-driven injection pumps on the engine sides to the injector nozzles (spray nozzles) in

' ^

the h'e ads. This oil is under pulsating and quite high pressure (~500 psi to i

15,000 psi once each cy:le); hence, any flaws in the steel tubing or fittings used, or any breaks caused by vibration or other factors, will release an oil spray in high-pressure bursts, with consequential personnel and fire risks.

! 4.14.2 Component Problem History High-pressure fuel tubing leaks have developed during preoperational l cngine testing on SNPS and Grand Gulf engines. No other failures in nuclear applications have been reported.

4.14.3 Owners' Group Status l SWEC analyzed the failed high-pressure fuel tubing and concluded that the l failures originated in ine.er surface flaws that were introduced-during fabri-cation. If, through eddy-current inspection, the inner surface condition of l new tubing is found to be within the manufacturer's specification, SWEC has concluded the high-pressure tubing is suitable for the service intended. I,t was also recommended, however, that all future replacement lines be of a superior material and be " shrouded" with a sock to protect against open oil sprays in the event of future leakages.

The OG also has reviewed the compression fittings and concluded that they a are adequate, provided that the injection lines are properly installed. The OG p recommends that inspections for fuel leaks near the compression fittings be performed while the engine is running.

4.61 u

4.14.4 LILC0 Status In March 1984, following a series of tubing failures, all of the fuel oil injection tubing was replaced with new, high-pressure, shrouded tubing. The ends of this tubing were eddy' current inspected to detect flaws greater than 0.003 inch, which is 1 css than the flaw size determined by the OG to be critical for propagation.

4.14.5 PNL Evaluation and Conclusion PNL concurs with the OG analysis of the critical flaw size. On the basis ,

of 1) the actions taken by LILCO to replace all tubing, 2) the successful ET, and 3) LILC0's commitment to check tube fittings monthly for leakage. .PNL con-cludes that the high-pressure tubing on all SNPS engines is acceptable for the intended service. PNL recommends that checks for oil leaks be done only while the engine is not running.

g O

e 4.62

4.15 AIR STARTING YALVE CAPSCREWS Part No.03-359 Owners Group Report Emergency Diesel Generator Air Start Valve Capscrew Dimmension and Stress Analysis (SWECo March 1984).

4.15.1 Component Function 9

These capscrews serve to hold the air start valves in place in the cylinder head. A failure, or an inappropriately long capscrew, will prevent the air starting valve assembly from seating securely in the head. The con-sequences of being incorrectly secured are the loss of power in one cylinder due to escaping combustion gases.

4.15.2 Component Problem History No actual failures of these capscrews have been reported. However, on May 13,1982, TDI reported a potential defect dt:e to the possibility of the 3/4-10 x 3-inch capscreris " bottoming out" in the holes in the cylinder heads,

-resulting in insufficient clamping of the air start valves.

4.15.3 Owners' Group Status SWEC and TDI both have recommended that the 3-inch capscrews be either shortened by 1/4 inch or replaced with 2-3/4-inch long capscrews.

4.15.4 LILCO Status Two TDI upgrades associated with the air start valves have been imple-mented at Shoreham. They are 1) installation of new 2-3/4-inch capscrews and

2) installation of new gaskets.

In conjunction with the SNPS DR/QR, a number of inspections were p?rformed to verify that the capscrews for all three engines are of correct material and length, bolt holes are clean and lubricated, the locking pin is in the valve ann lock nut, the 0-ring and grooves are in good condition, and the gasket seal to the cylinder head is of the proper material. Material testing of the cap-screws was also performed. Other visual inspections were conducted for corro-sion and deterioration. LILC0 confirmed that all capscrews were torqued to OG reconsnendations.

4.63

On the basis of the modifications, inspections, and maintenance plans, LILCO concludes that the air start valves, capscrews, and gaskets for all engines are suitable to serve their intended function.

4.15.5 PNL Evaluation and Conclusions The inspections and actions taken by LILC0 to eliminate potential problems are judged to be adequate to prevent failures. PNL therefore concludes that, '

with the continued use of LILCO's installation procedures to control torque of I

bolts, studs, and screws to specified ranges, and LILCO's recommended mainten- ,

ance procedures, these components will not present future problems on the SNPS engines. Thus, PNL concludes these components on EDGs 101, 102, and 103 are acceptable for their intended service.

i I

l l

o l

! 4.64 l

[

4.16 ENGINE-MOUNTED ELECTRICAL CABLE Part No. 03-688-8 Owners' Group Report SWEC Report of April 1984 4.16.1 Component Function These cables serve the Woodward governor / actuator and the Air-Pax magnetic pick-up, both mounted on the engines. Inappropriate cable materials, not able r . to withstand the temperature or service environment, could lead to short cir-cuits, with adverse impact on the component functions and possible risk to

, personnel.

4.16.2 Component Problem History Two defective cables were recorded by TDI in a 10 CFR 21 report. Also, a TDI Service Information Memo warned of potentially defective engine-mounted cables.

4.16.3 Owners' Group Status Analyses of the subject wiring, and of the recommended replacements, were conducted by SWEC, both generically and specifically for SNPS. The replacement cable and teminations were deemed serviceable for this duty.

4.16.4 LILCO Status LILCO has evaluated the wiring and terminations used on all three SNPS engines. The terminal components and blocks were found to be within specifi-cations. The insulated wires associated with the crankcase vacuum fans and the i standby air supply solenoid valves were replaced to meet operating temperature requirements. All wiring and cables now installed are of acceptable flame-l retardant construction and meet specified current and thermal requirements. ,

& On the basis of the evaluations and changes implemented, LILCO concludes that the engine-mounted electrical cable and terminations used at SNPS are suitable to serve their intended function.

~

I a

4.65 .

t

'l L

4.16.5 PNL Evaluation and Conclusion Based on the review of the actions taken by LILCO, PNL concludes that the subject terminations and cables on EDGs 101,102, and 103 are acceptable for their intended service at SNPS.

O J

t l

l

? A

~

i

\

- 4.66 i--

. . _-. - - _ . . - . . - . - . = - - . . . . _ _ _ . - . - . . . - - - - _ . . . _ _ _ _ .

5.0 PROPOSED MAINTENANCE AND SURVEILLANCE PROGRAM In PNL's review of the TDI Diesel Engine Owners' Group Program Plan t (PNL-5161, June 1984), maintenance and surveillance (M/S) is identified as "a key aspect of the overall effort for establishing TDI engine operability and '

, reliability". NRC also recognizes the importance of a comprehensive M/S pro-gram and has provided guidelines for the development of such a program in the NRC staff SER dated August 13, 1984.

A Long Island Lighting Company has developed an M/S plan for the Shoreham Nuclear Power Station. The plan is presented in the Stone & Webster Engineer-l ing and Design Coordination Report (E8DCR) No. F-46505, which is to be incor-

[ porated into the TDI Instruction Manual whereby it will be implemented into the fl Shore'1am M/S Program. E8DCR F-46505 includes, as an attachment, the Suggested Maintenance Schedule from the TDI Operation, Maintenance and Instruction Manual, Model DSR-48 Diesel Engine. Subsequent references to the LILCO M/S plan in this TER will, therefore, also encompass TDI's recommended maintenance

! and surveillance procedures. LILCO's M/S plan is also partially reproduced in Appendix II'of the Shoreham TDI Diesel Generator Design Review / Quality Revali-l- dation Report (June 29, 1984), although LILCO has indicated that this appendix

is subordinate to E&DCR F-46505. ,

l i E80CR F-46505 lists many more maintenance items than will be discussed in j this report. Those that are not itemized here are judged to be-beyond the scope of this effort, which is focused on components and systems critical to l engine operation and/or with failure histories. However, PNL has added a few items where special attention is deemed appropriate. Specifically, recommendations for M/S are provided when, in the opinion of PNL consultants, I their inclusion in an' overall M/S plan is important to ensuring operability and

. reliability. This report is not intended to supplant LILCO's M/S plan, but 4

, rather to augment and clarify it.

l . , ,

This section documents PNL's review and evaluation of LILCO's M/S plan in i light of the opinions and recommendations of recognized experts in diesel .

engine technology. Comments on four aspects of a comprehensive M/S program are l

presented in the subsections that follow. The four aspects are: ,

5.1 l

, -- . _ _ _ _ . , , , _ _ . _ . . _ . _ . . . , _ , , . . _ _ _ . _ _ _ _ - - - - _ . _ ~ _

e major maintenance items e proposed additional maintenance items e operational surveillance e standby surveillance.

5.1 MAJOR MAINTENANCE ITEMS ,

Components classified as major maintenance items include engine structural and moving parts. Parts with a failure history, even if they are static and/or ,

nonstructural, also are included. LILC0's proposed M/S plan for these compo-nents is summarized in Table 5.1.

5.1.1 PNL Evaluation and Recommendations PNL found that LILCO's proposed M/S plan does not identify several impor-tant components and systems requiring periodic maintenance. In addition, PNL consultants recommended that LILCO's maintenance procedures for several items be revised.

PNL recommends that LILC0 add procedures for the following components to its M/S plan:

o main bearing shells e studs / bolts.

PNL also recommends modifications to the LILC0 M/S plan for the following commponents:

e crankshafts e - cylinder blocks e connecting rod bearing shells e cylinder liners e cylinder heads e turbochargers. e PNL recommendations for M/S of these components are presented in Table 5.1 for comparison with LILCO's M/S plan. Where the two recommendations diverge ,

substantively, explanation for the difference is provided in the subsections that follow.

5.2

.- > S.

TABLE 5.1. Major Maintenance Items .for Shoreham Nuclear Power Station (a)-

Item LILCO Proposal PNL Recommendation Crankshaft Check and record crankshaft deflections. Concur with LILCO. Measure web deflec-tion, hot and cold,- once each refueling.

EDG 101 and EDG 102: Inspect crankpin journals No. 5, 6, and 7 and the main bearing journals between them at the first refueling outage, using LP and ET as appropriate. Also inspect the oil holes in these journals.

EDG 101, 102 and 103: During the second and subsequent refueling outage, inspect two of the three crankpin journals sub-ject to the highest stresses (No. 5, 6, and 7) and their oil holes, using LP and ET as appropriate. Perfonn this inspec-tion or each engine.

Cylinder Perfonn a visual and eddy-current Concur with LILCO. In addition, visually Blocks inspection of the cylinder block, paying inspect daily during any period of con-close attention [to the areas] between tinuous operation and inspect under -

stud holes of adjacent cylinders prior intense light while operating monthly.

to returning the engines to emergency standby service after any period of At each refueling outage, inspect top operation greater than 50% load. surface of block exposed by removal of two heads. LP recommended with UT as appropriate.

Monitor behavior of several representa-tive cam gallery cracks in EDG 101 and 102. Alternative approaches and fre-quency discussed in Section 5.1.1.2.

(a) Refer to Section 4.0 for additional details on Owners' Group designated generic issues.

TABLE 5.1'. (contd)

Item .LILCO Proposal PNL Recommendation-Cylinder ~ Perform LP and/or UT. inspection of Bl ocks .' -

cylinder. liner landing at any time a liner is removed..

Connecting Rods None provided' -Check preload on connecting rod bolts at-each refueling outage.

Main Bearing- Check at alternate refueling outages. Concur with LILC0 Shells (method unspecified).

Connecting Inspect and measure connecting rod Concur with LILC0 Rod Bearing bearing shells every 5 years.

Shells Perform an x-ray examination on new Concur with LILC0

- bearing shells to' acceptance criteria

' developed by Owners' Group prior to installation.

Bump test for bearing wear at each Concur with LILCO.

refueling outage.

Replace lower bearing shells No. 4, 5,and 6 on EDG 1 at first refueling,. or sooner per FaAA recommendations (300 operating hours).

Pistons Inspect and measure skirt and piston Concur with LILCO pin every 5 years (TDI inspe t maintenance fonn No. 5-D-10)yajon and Cylinder Perform a boroscopic inspection of Concur with LILC0 Liners liners for potential progressive wear at each refueling outage.

Measure / record dimensions at each dis, assembly.

(a) Included in Shoreham DR/QR report, but not in E&DCR F-46505.

E .a

O 1

TABLE 5.1. :(contd)

Item LILC0 Proposal PNL Recommendation Cylinder Visually inspect cylinder heads (all Concur with .LILCO Heads eight cylinders) at each refueling outage.

Inspect firedeck between exhaust valve seats and all valve seats for two adjacent heads with LP at each refueling -

outage. Select heads such that all heads are inspected every four refueling i

outages.

l Cylinder Air-roll before planned starts, and Heads 4 to 8 and 24 hours2.777778e-4 days 0.00667 hours 3.968254e-5 weeks 9.132e-6 months after each shutdown.

Valve Gear / Visually inspect camshafts, tappets, Concur with LILC0 Li fters rollers, rocker arms, push rods, and i valve springs, and adjust lifter, at each refueling outage.

a, b Head Studs, Air None provided(a) Check preload on 25% of head studs, Start Valve Cap- 100% of air start valve capscrews i

, screws, Rocker and 25% of rocker arm bolts at each l Ann Bolts refueling.

}

Gear Train Visually inspect cam, idler, and crank- Concur with LILCO

! shaft to jacket water pump gear for i chipped or broken ' teeth, excessive wear,

, pitting, or other abnormal conditions

] during alternate refueling outages.

! Measure gear backlash at each refueling Concur with LILC0 i outage.

i

, (a) LILCO does propose to remove and inspect the air start valves at each refueling outage and would i therefore retorque all air start valve capscrews.

i i

TABLE 5.1. (contd)

Item LILC0 Proposal PNL Recommendation Turbochargers Measure rotor endplay monthly until Concur with LILCO, but also record confidence is established, measurements and monitor for trends.

Inspect thrust bearing after Concur with LILCO, except inspect after 40 nonprelubed starts. 40 nonprelubed starts or after 100 starts (inclusive), whichever occurs first.

Perform spectrochemical ferrographic Concur with LILCO engine oil analysis quarterly or after 20 starts. Inspect nozzle ring and check rotor. float at each refueling outage.

"Y" Strainers Inspect monthly. Concur with LILC0 in Starting Air System Replace filter element at each Concur with LILC0 refueling outage, f' Lube Oil Check lubricating oil with a viscosi- Concur with LILC0; sample from lube Sampling and meter for fuel oil dilution. Send a oil filter inlet monthly or every Analysis sample of oil to laboratory for analy- 24 hours2.777778e-4 days 0.00667 hours 3.968254e-5 weeks 9.132e-6 months running.

sis monthly.

Lube Oil Drain lubrication oil system and clean Concur with LILCO.

Sampling and sump tank at every refueling outage.

Analysis Depending on the results of lube oil analysis, refill with new oil .

Sample from bottom of sump (check for water) monthly.

Air Start Remove, clean, and inspect at each Concur with LILC0 Valves refueling outage.

Inspect the piston cap guide housing Concur with LILCO sliding surfaces at each refueling outage.

Ensure that the refrigerant dryer is Concur with LILC0 working properly daily. .

. .. , e I- _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

- '.- > c TABLE 5.1. (contd)

Item LILC0 Proposal PNL Recommendation Base Assembly Perform a fluorescent dye liquid pene- Concur with LILC0 trant inspection of the linear indica-

.~tionspresentonthebearingsaddlega) during alternate refueling outages.

Fuel Injection Check tubing for leaks at fittings Concur with LILCO. PNL recommends that Tubing monthly. leakage inspection and corrections be made with the engine stopped.

(a) The Shoreham DR/QR report calls for magnetic particle examination. PNL concurs with E&DCR F-46505, which specifies liquid penetrant examination of the base assembly.

5.1.1.1 Crankshaft LILCO propose _ to measure crankshaft deflections at each refueling outage. PNL concurs and recomumends in addition certain WE examinations of the crankshaft at refueling outages.

In light of the analyses performed for PNL by Ricardo Consulting Engineers and by Det Norske Veritas, PNL concludes that it would be prudent to examine -

certain high-stress areas of the crankshaft at each refueling outage. The areas to be examined and the examination methods and frequency are provided in ,

Section 4.3.5.

PNL Recommendations PNL concurs that LILC0 take crankshaf t deflection readings at every refueling outage. LILC0's M/S plan does not prescribe hot and cold deflection measurements or the timing of such measurements. PNL recommends the following instructions be added to LILCO's M/S plan. The hot deflection measurements should be taken immediately after the 24-hour preoperational testing, so as to reflect representative operational foundation temperatures. The hot checks should be initiated within 15 to 20 minutes after shutdown, and completed as rapidly as possible, preferably within 1/2 hour, starting with the last throw of the engine (generator end). Such a schedule, although strenuous, is deemed achievable. If the crankshaft deflection readings are outside the acceptable range, the foundation bolts should be checked for proper preload.

PNL also recomends that crankshaft. journals be LP inspected whenever corresponding bearings are being inspected.

5.1.1.2 Cylinder Block Following any period of engine operation at greater than 50% load, LILCO proposes to perforia visual and eddy current inspections of those portions of the block top that are accessible between cylinder heads. The purpose of these

inspections is to verify the continued absence of detectable cracks between stud holes of adjacent cylinder. PNL concurs with this proposal (with the ,

clarification that 50% load is 50% of the " qualified" load), but recomumends that additional surveillance and maintenance procedures also be perforised.

5.8

. . -_. - . . _ = . . - - . - . - - - . _ _ _ . _ _ - - . - -- -. - -.- - _-

4 Cracks known to exist in the cylinder blocks of the three Shoreham EDGs q are described in Section 4.2.2. These include ligament cracks in EDG 101 and i EDG 102, cam gallery cracks with weld overlays in those same engines, and cam

gallery cracks in EDG 103.

PNL Recommendations

, PNL recommends that the visual examinations proposed by LILCO be augmented with daily visual examinations when the engine is in continuous operation, and '

with a visual examination under intense light when the engine is operated monthl y.

[ PNL recommends the following nondestructive examinations for each cylinder block in addition to the eddy-current inspections proposed by LILCO:

o Block top - During each refueling outage, the portion of the block top exposed by removal of two of the cylinder heads (for inspections i discussed in Section 5.1.1.6) should be examined for any new cracks that may develop into stud-to-stud cracks. This inspection should also identify any new ligament cracks in the exposed areas of the top j surface, and any changes detectable through surface examination of known ligament cracks. For any new cracks and/or changes in known l

cracks that extend from stud holes, the studs should be removed to l gain better access to the holes, and the depth of the cracks along i the counterbores should be measured. PNL recommends that LP be used h to perform these examinations, and that UT be used as appropriate to better define the extent of any new cracks or changes in known cracks.

e Cylinder liner landing - An LP and/or UT inspection of the cylinder liner landing should be performed at any time a liner is removed from any of the three engines, to determine if circumferential cracks have >

developed. PNL recognizes that liners are likely to be removed only

. . infrequently, and does not recommend removal of a liner for the sole a purpose of this inspection. If a circumferential indication is found, an attempt should be made to characterize the depth and length through appropriate nondestructive tests. PNL also recognizes that i

5.9 ,

f

r, , - ~ , , -,,,-,-<---,,-,-----n.~.,,-__,-n.~,~,-c,--,,n.,,,-.- -e,,,.,----,,,----,- -~n.--

t these cracks are difficult to identify unequivocally through nondestructive tests due to the relatively sharp corner where they may occur.

e Cam gallery - Several representative cracks 111 the camshaft bearing saddles of EDG 101 and EDG 102 should be monitored to deteriiine if long-term changes are observed that would be indicative of crack , ;

propagation under transient conditions of engine startup and shutdown, and during engine operation. PNL recommends that the ,

second camshaft bearing saddle inboard of each end of the engine be selected for monitoring. PNL does not recommend monitoring the cam gallery cracks in EDG 103, because the known cracks in the replacement block are much shallower than those in EDG 101 and EDG 102 and have not been repair welded.

4 The monitoring should focus on crack behavior, rather than on mea-surement of :; tresses in the vicinity of a crack. (Representative stresses in the area of interest were measured in the replacement EDG 103 block by FaAA in October 1984.) This could be accomplished i

through any of several alternative approaches. One approach would be to install strain gages to monitor changes in crack opening and/or in crack length. The gages could be monitored during monthly engine

f. tests. If no changes indicative of crack growth were observed in such tests over a period of 6 months, the frequency of testing could be reconsidered. The monitoring could be discontinued if no changes were observed over one refueling cycle.

An alternative approach would be to monitor crack depth periodically with an appropriate surface probe. Depth readings measured in this manner lack accuracy, but would probably be sufficient to show any significant change in crack depth. Because access covers would have

to be removed to expose the cracks, the engine would not be imme-diately available for emergency service. To obtain the desired ,

information with minimal disruption of engine availability, it would be sufficient in the opinion of PNL to take the depth measurements approximately every 3 months through one refueling c3cle. The a 5.10 I

J

monitoring could be discontinued at the end of that period if no changes indicative of crack growth were observed.

In the event the block of either EDG 101 or EDG 102 is unbolted from its base during the time the cam gallery cracks are being monitored, measurements should be taken to determine whether or not crack 0 " pop-in" occurred as a result of the relaxation of compressive stresses in the cam gallery.

A 5.1.1.3 Connecting Rods LILCO has not addressed the inspection of the connecting rod bolt pre-load. PNL recommends that the preload on all connecting rod bolts be checked at each refueling outage.

PNL believes that it is good practice to inspect the preload on the con-necting rod bolts periodically. Checking the bolt preload during regularly scheduled outages is a simple procedure and is easily justifiable on the basis of the potential damage to the engine that could result from the loss of these bol ts.

PNL Recommendation PNL recomends checking all connecting rod bolts' preload at each refueling outage.

5.1.1.4 Connecting Rod Bearing Shells LILCO proposes to inspect and measure the connecting rod bearing shells every 5 3 ears and to " bump" test for bearing wear at each refueling outage.

PNL concurs and reconsends replacement of the lower bearing shells No. 4, 5, and 6 in EDG 101 before they accumulate a total service of 300 hours0.00347 days 0.0833 hours 4.960317e-4 weeks 1.1415e-4 months .

l l The Owners' Group design review report (FaAA-84-3-1) concluded that the bearings were adequate at site loads for the lifetime of expected usage. Based on these findings, LILCO has proposed inspection of these bearings every a 5 years and to bump test all connecting rod bearings at each refueling outage.

l l LILCO has also identified three connecting rod bearing shells that have l minor defects. These bearing shells are installed in the lower position of crank journals 4, 5, and 6 of EDG 101. PNL believes that these bearing shells I.

l l 5.11 1

i

are acceptable for operation through the first refueling outage, provided that the total operating hours accumulated on any of these bearings do not exceed 300 hours0.00347 days 0.0833 hours 4.960317e-4 weeks 1.1415e-4 months , as recommended by FaAA (C. H: Wells) in a memo to LILC0 (P. Martin) dated March 24, 1984.

PNL Recommendations PNL concurs with LILCO's plan. PNL also recommends replacing lower bear- 4 l ing shells No. 4, 5, and 6 of EDG 101 at the first refueling outage, subject to the 300-hour limitation discussed in the preceding paragraph. ,

5.1.1.5 Cylinder Liners LILCO proposes to perfone a boroscopic inspection of the liners for poten-tial progressive wear at each refueling outage. PNL concurs and recosumends a l recorded dimensional check at every disassembly.

l l Because liner wear provides an important indication of engine reliability

! and operability, it should be monitored whenever reasonably possible.

! PNL Recommendation l PNL concurs that all liners should be visually inspected by boroscope at I each refueling outage. In addition, two of the liners from each engine should be measured for wear at every disassembly, and the dimensions recorded for trend analysis.

5.1.1.6 Cylinder Heads LILCO proposes to visually inspect all eight cylinder heads at each refueling outage. PNL concurs, but recosumends further that two heads from adjacent cylinders be LP inspected at valve seats and firedeck at each refueling outage. In addition, PNL recomumends that the engines be air-rolled before all planned starts, 4 to 8 hours9.259259e-5 days 0.00222 hours 1.322751e-5 weeks 3.044e-6 months after each shutdown, and 24 hours2.777778e-4 days 0.00667 hours 3.968254e-5 weeks 9.132e-6 months after each shutdown. 3 Air-rolling the engine will expel any accumulation of water in the cylinder, which would most likely be the result of a cracked cylinder head or ,

I liner. Substantial water accumulation in a cylinder could cause severe damage i

5.12 i.

to head, piston, crankshaft, or bearings on engine startup. Detection and expulsion of water in the cylinder liners is essential to ensuring engine operability.

PNL Recommendations PNL concurs with LILCO's plan and, in addition, recomends a schedule for 8

air-rolling as follows:

e an initial air-roll at least 4 hours4.62963e-5 days 0.00111 hours 6.613757e-6 weeks 1.522e-6 months (but not over 8 hours9.259259e-5 days 0.00222 hours 1.322751e-5 weeks 3.044e-6 months ) after engine shutdown e a second air-roll approximately 24 hours2.777778e-4 days 0.00667 hours 3.968254e-5 weeks 9.132e-6 months after shutdown e thereafter, an air-roll immediately prior to any planned engine operation.

In view of the potential for crack initiation noted in Section 4.8, PNL also recomends removal of two heads and visual and LP inspection of the firedeck at each refueling outage. The firedeck should be inspected between exhaust valves. The heads to be inspected should be selected such that all heads are inspected every four refueling outages.

5.1.1.7 Head Studs, Air Start Valve Capscrews, Rocker Arm Bolts LILC0 has not addressed head studs, air start valve capscrews, and rocker am bolts in their M/S plan. PNL recemends that these items be inspected for proper preload at each refueling outage as specified below. '

i

! Loss of preload on cylinder head studs, rocker arm capscrews, and air start valve capscrews can adversely affect engine operability if it goes j unnoticed. Because of their operational history, these items are included on l the Owners' Group list of components with significant known problems. Thus, I these components warrant regular maintenance and surveillance.

PNL Recommendations PNL recommends that the preload be checked on a sample of 257, of the head

^

l studs and rocker am capscrews at each reactor refueling outage. However, l

5.13 i

s n

> because the air start valve capscrews are more susceptible to relaxation (due

^' to the associated soft metal gaskets),,PNL recommends that they all be checked at each refueling outage. 4 5.1.1.8 Turbochargers LILCO proposes to measure turbocharger rotor endplay monthly until con-fidence is established, to visually inspect the thrust bearing af ter 40 1 nonprelubed starts, and to perform a spectrochemical engine oil analysis quarterly or after 20 starts. PE concurs, with certain qualifications, and i also recommends that rotor float be measured and stationary nozzle rings including vanes and capscrews be inspected at each refueling outage.

A' recurring problem in the turbochargers on TDI engines at Shoreham and at other TDI installations has been thrust bearing wear. A modification to the lubrication system to provide minimal lube oil to the thrust bearing during engine ' standby proved to be inadequate. Subsequent modifications to the system have increased bearing prelubrication, which has substantially mitigated the thrust bearing wear.

~

The turbochargers on the Shoreham EDGs have also experienced a failed nozzle ring capscrew and a lost nozzle ring vane.

PNL Recommendations PNL recommends that LILC0's M/S plan be modified to include visual thrust bearing inspection after 40 nonprelubed starts or after 100 starts (inclusive),

and to include rotor float measurement and stationary nozzle ring including vanes and capscrews at each refueling outage.

4 l 5.1.1.9 Lube Oil Sampling and Analysis LILCO proposes to check the lube oil with a viscosimeter and send a sample of oil to the laboratory for analysis month 1y. LILCO also proposes to drain e

and clean the sump tank at each refueling outage. PE concurs with LILCO's plan and recommends, in addition, a monthly inspection for water.

Proper maintenance and surveillance ;of the lubrication oil is desirable not only to ensure proper lubricationiof moving engine components, but also to obtain diagnostic information from which engine wear can be inferred. The oil 5.14 s

~ , - - . - - ~ . -

samples to be analyzed should, therefore, be taken from the lube oil filter i nl et. Because water will sink in oil, a sample should also be taken from the bottom of the sump tank to detect the presence of water.

PNL Recommendations g PNL concurs with LILCO's plan to take the lube oil samples from the filter inlet with the engine running. PNL further recommends that monthly samples be taken from the bottom of the sump tank to check for the presence of water.

A 5.2 ADDITIONAL MAINTENANCE ITEMS In addition to the major maintenance items discussed in Section 5.1, cer-tain other engine components and maintenance procedures warrant inclusion in a comprehensive M/S plan. These include:

e fuel injection pump e lube oil keep warm filter e fuel injection nozzle e lube oil filter

= governor oil change e heat exchangers e air intake filter e jacket water system flush e fuel oil . drip tank e engine balance e fuel oil filter and duplex e governor drive coupling strainer 5.2.1 Rationale These additional maintenance items also are important to EDG reliability

and' operability although they have not been related directly to any TDI engine failure to date. In PNL's opinion, their inclusion in an M/S plan is consis-l tent with good practice. The items are presented here as recommendations /

suggestions.

5.2.2 PNL Evaluation PNL reviewed the available LILC0 documentation to determine if the i 13 items noted above were included in the utility's proposed M/S plan. The 1 x l first two columns of Table 5.2 summarize PNL's findings.

PNL's recommendations are summarized in the third column of Table 5.2.

Explanatory notes are included where necessary to augment these l 5.15 l

l I

l TABLE 5.2. Recommended Additional Maintenance Items for Shoreham Nuclear Power Station Item LILC0 Plan PNL Recommendation Note Fuel injection Not included Verify calibration and opera- (1)

. pumps tion at every third refueling.

Fuel injection Not included Check " popping" pressure and (2) nozzles spray pattern at every ,

3 refueling outage. J l Governor Drain, flush, refill, and Concur with LILC0 l oil change vent actuator oil system 2 '

with oil from a clean l container, ensuring that appropriate cleanliness l

, procedures are followed, I at each refueling outage. l l

i Air intake Inspect air intake filter Concur with LILC0 l f filter every 3 to 6 months.

Replace as required. l Fuel oil Not included Check monthly, clean as l drip tank required.

Fuel oil Change . filter elements Concur with LILC0; also change

filter and at each refueling outage. filters if there is a 20-psi l duplex pressure drop across filter.

j~ strainers l Record strainer differen- Concur with LILC0 i tial pressure monthly; clean or replace if pres-sure drop is greater than l 5 psi.

-Lube oil Change filter elements at Concur with LILCO keepwarm each refueling outage or filter when filter differential pressure reaches 10 psid.

Lube oil Manually cycle lube filter Also change with 20-psi filter and strainer monthly. pressure drop across filter. ,

Jacket water Flush daily. Concur with LILCO

[ -heat exchanger l Inspect tubes, tube pro- Concur with LILCO

l tectors and tube sheets l for corrosion and fouling i

at each refueling outage.

5.16 l

[

l I.

t TABLE 5.2. (contd)

Item LILCO Plan PNL Recomendation Note

~

I Inspect and clean lantern Concur with LILC0 ring at each refueling outage.

r Replace packing ring at Concur with LILC0 each refueling outage.

Lube oil heat Inspect for leakage at Concur with LILC0

'- exchanger neoprene packing daily.

Inspect tube, tube pro- Concur with LILCO tectors and tubc sheets for corrosion and fouling at each refueling outage.

Replace packing rings at Concur with LILCO each refueling outage or when packing becomes hard or when leakage at the packing is noted and cannot be stopped by tightening.

Jacket water Not included Flush system at alternate system flush refuelings.

Engine Not included Record firing peak pressure balance and exhaust temperatures and adjust per TDI manual at each refueling.

Governor Check that coupling is Concur with LILCO

, drive tight on shaft at each coupling refueling outage.

Replace the present Concur with LILCO neoprene elastomeric inserts in Koppers y coupling.

A 5.17

-___ , ~ .- .. _ _ _ . .._ . .._ ._ _ ___ --

. .- ._. . - . - . . - - =

4 NOTES (1) Fuel injection pumps on the LILC0 engines have not been a source of pro-bl ems. Due to the precision and close-tolerance nature of the fuel injec-tion pumps, they can be damaged easily during a disassembly, thus requiring raplacement of parts when otherwise unnecessary. Fuel injection pumps can be checked for proper operation and calibration at any reliable diesel ser-vice center; faulty or questionable pumps can then be put aside for dis-assembl y. It is important to note that the same test should be performed on all pumps if they have been disassembled. The pumps should also be '

recalibrated after disassembly.

(2) Fuel injection nozzles are similar to injection pumps, in that very close 2 tolerances are encountered; thus, they are also susceptible to damage dur-ing maintenance inspection. Proper testing of the nozzles for leakage,

. " popping" pressure, and spray pattern would give a complete indication of

, the status of each nozzle. Then, only nozzles giving questionable results would need to be disassembled. The same tests should still be performed on

)

all nozzles after reassembly, should they be disassembled.

PNL recommends checking " popping" and closing pressure and spray pattern of all fuel injection nozzles at every refueling outage. Should operating surveillance (i.e., cylinder exhaust temperature) indicate a potential fuel injection nozzle problem, the suspect nozzle should be tested and, as necessary, disassembled.

J 1

A.

5.18

recommendations. In PNL's opinion, the recommendations presented in Table 5.2 should be considered in finalizing LILCO's M/S plan for the Shoreham plant.

5.3 OPERATIONAL SURVEILLANCE PLAN A third aspect of M/S is operational surveillance, which refers to the r parameters to be monitored and/or recorded during engine operation. This

~

typically includes temperatures and pressures at key locations in and about the engine, as well as cumulative parameters such as engine hours and tachometer readings..

5.3.1 Rationale Operational surveillance is necessary to ensure safe and efficient opera-tion of the diesel engine. By monitoring and recording. key engine parameters, trends in degradation can be detected, allowing timely preventive maintenance.

Trend monitoring may also prevent major engine damage by providing early warning to allow engine shutdown prior to damage.

5.3.2 PNL Recommendations j

PNL' has not received a detailed operational' surveillance plan from LILCO.

Recognizing the importance of operational surveillance as indicated above,

[

PNL's recommended surveillance plan is provided in Table 5.3. Justification l for several of the included parameters when the recommendations are not obvious is provided in the subsections. that follow.

5.3.2.1 Pre-Turbine Exhaust Temperature Pre-turbine exhaust temperature is valuable because: -

e The individual cylinder exhaust pyrometer reports only a time average of highly variable function.

2 e The turbine inlet temperature may be higher than any cylinder exhaust because of more hot puffs per time, and also because of possible exo-y thermal reactions in the exhaust manifold.

e Blades and nozzle rings may be damaged by temperatures above the

manufacturer's limit, which Elliott states is 1200'F.

t 5.19 l ..

l l

I

TABLE 5.3. Recommended Operational Surveillance Plan for Shoreham Nuclear Power Station Item PNL Recommendations Lube oil inlet pressure Log hourly Lube oil to turbocharger pressure Fuel oil- to engine pressure

~

1 Fuel oil filter differential pressure r.

Air manifold pressure tube oil filter differential pressure Jacket water pressure (inlet and outlet)

Crankcase vacuum All cylinders exhaust temperature Exhaust temperature at turbine inlet >

Lube oil;. temperature (inlet and outlet)

Jacket water temperature (inlet and outlet)

Tachometer Hour meter

, Generator load Air manifold temperature U l Fuel oil transfer pump strainer Log hourly unless strainer is

differential pressure - auto / duplexed and alarmed Starting air pressure Check hourly L

L Fuel oil day-tank level Check hourly Air manifold Drain condensate every 4 hours4.62963e-5 days 0.00111 hours 6.613757e-6 weeks 1.522e-6 months of engine operation A

(

Visual inspection for leaks Monthly or after 24 hours2.777778e-4 days 0.00667 hours 3.968254e-5 weeks 9.132e-6 months ' operation L

i-5.20

~

t - --_ - .-

'E 5.3.2.2 Air Manifold Temperature Air manifold temperature indicates the-effectiveness of the turbocharger aftercooler, and its efficiency is dependent on water flow and temperature and on fouling. The effects of elevated air manifold temperatures are reduced 2

maximum load and less efficient combustion. Although the potential for such I y problems is less at the low engine loads, it is considered prudent to monitor and . trend the air manifold temperature.

?

, t 5.3.2.3 Fuel Oil Transfer Pump Strainer Differential Pressure

, This pressure should be monitored continuously and recorded hourly unless

the pump is automatically valved duplex with alarm to protect fuel feed.

5.3.2.4 Starting Air Pressure This pressure must be monitored to ensure sufficient pressure is available

[ for restart at all times.

5.3.2.5 Fuel Oil Day' Tank Level This level must be monitored to ensure fuel availability.

5.4 STANDBY SURVEILLANCE PLAN In Appendix 11 of the Shoreham DR/QR report, LILCO has provided a list of I specific. items to be maintained on a daily basis. In addition, the attachment to E&DCR F-46505 contains several pages from the TDI Instruction Manual that specify maintenance procedures to be followed.

5.4.1 Rationale Standby surveillance is important to ensuring the operability of the

diesel engines. The parameters monitored on an engine in standby status are

-intended to indicate the engine's preparedness to start rapidly and accept

~#

1oad. The two factors that contribute most to this are engine temperature and lubrication. By keeping the engine warm and all oil passages pressurized, the EA time lag associated with load acceptance is minimized. In addition, a ready supply of' quality compressed air is required for starting the engines.

5.21

-5.4.2 - PNL Evaluation PNL reviewed LILCO's proposed standby surveillance plan and found that the proposal covers a number of items considered important to monitoring engine condition while on standby status. However, several important items were not included'in LILCO's proposal. For other items, PNL consultants' )

recommendations differ from those of LILCO. In general, justification for 9

[ these recommendations is based on engineering judgment.

Recommended items to be included in a standby M/S plan are presented in /

Table 5.4. The information in this table is not intended to supplant any '

maintenance procedure recommended by the manufacturer, but rather to provide l additional perspective in developing an integrated M/S plan.

l Two points regarding the keepwarm lube oil filter are important:

i

1. Entrained water or bacteria (in the absence of bactericide use) will

tend to plug some filter media (or weak' en others), and so would grad-f vally increase pressure drops.

. 2. The continuous keepwarm flow through the filters will (purposefully)

!' continuously filter the oil, with gradual buildup of contaminants in ,

j the media; the material scavenged out thereby itself helps filter even finer particles as time continues.

Thus, it remains valid to monitor oil filter pressure drop during standby. The l

changes are slow enough that a weekly check is (eemed sufficient.

In conclusion, it appears that the LILCO standby surveillance plan needs to be formalized as a separate entity within the overall M/S program for Shoreham. PNL recommends that the plan include the items and time intervals listed in Table 5.4.

1 i

d I

i l 5.22

TABLE 5.4. Recommended Standby Surveillance Plan for Shoreham Nuclear Power Station

' Item LILCO Proposal PNL Recommendation Starting air pressure None provided Visual checks every 8 hours9.259259e-5 days 0.00222 hours 1.322751e-5 weeks 3.044e-6 months .

Log every 24 hours2.777778e-4 days 0.00667 hours 3.968254e-5 weeks 9.132e-6 months .

Lube oil temperature in/out Daily (a)

P

' Jacket water temperature Daily (a) in/out

-(

Lube oil sump level Daily Fuel oil day-tank level None provided Room temperature "

Annunciator test Test every 8 hours9.259259e-5 days 0.00222 hours 1.322751e-5 weeks 3.044e-6 months .

Log every 24 hours2.777778e-4 days 0.00667 hours 3.968254e-5 weeks 9.132e-6 months .

Check alann clear Daily

. Check compressor air trap Daily operation F Fuel rack and linkage Daily Concur and lube monthly.

operation

,. ' Governor oil level Daily Concur with LILCO Inspect for leaks Daily, with detailed Concur with LILC0

. inspection monthly

! Check freedom of-air butter- Monthly Concur with LILCO fly valve and cylinder

' . Keepwarm oil filter None provided Check weekly differential pressure

-Test jacket water for_ pH, Monthly Concur with LILCO

, conductivity, corrosion i .2 inhibitor Air start distributor filter Monthly Concur with LILCO h e Air. start adnission valve None provided Check quarterly strainer (a) Both in and out temperatures are to be measured.

l 5.23

6.0 OVERALL CONCLUSION This section presents PNL's overall conclusion regarding the capability of the Shoreham EDGs to perfonn their intended function as emergency standby power sources.

C 6.1 GENERAL ~ CONCLUSION

( PNL and its consultants conclude that EDGs 101, 102, and 103 at the Shoreham Nuclear Power Station are suitable for nuclear standby service at the

" qualified" load of 3300 kW, subject to the following comments. !

e PNL recommends that the NRC staff obtain a commitment from LILCO to implement the maintenance and surveillance recommendations documented in Section 5.0 of this report. In the revised surveillance and main- j tenance plan, LILCO should retain all of the provisions currently .

proposed by the utility that are not addressed in Section 5.0.

LILCO's commitment should be a prerequisite for a license from NRC.

Recognizing that this report precedes the final review by NRC of the Owners' Group Program Plan, PNL recommends that the NRC staff obtain a commitment from LILCO to implement all relevant recommendations and requirements identified in the NRC review. . This commitment should be obtained as a prerequisite for licensing, and all relevant actions should be completed by LILCO before or during the first refueling otstage, insofar as this is practicable.

~

, 6.2 BASIS FOR CONCLUSION l

PNL based this conclusion on reviews of 1) the TOI Owners' Group Program Plan, 2) LILCO's actions to resolve generic and plant-specific problems and to y

upgrade the Shoreham EDGs, 3) the results of engine testing (including requalification and confirmatory testing), and 4) the results of associated D engine inspections.

p I

i 6.1 I~

k- _.__ _ ,-_ _., _ _ __ _ ._ _ _ ,_..,_._ _._ _

PNL-5342 DISTRIBUTION 14o. of- No. of Copies Copies OFFSITE P. Lang, NE-14 U.S. Department of Energy

.17 Division of Licensing Office of Nuclear Energy p Washington, DC 20555 Office of Nuclear Reactor Regulation U.S. Nuclear Regulatory ONSITE

$D Commission Washington, DC' 20555 DOE Richland Operations Office ATTN: C. Berlinger (10)

M. Carrington (2) H. Ransom /M. Plahuta D. Corley D. Eisenhut Pacific Northwest Laboratory A. Marinos F. Miraglia 6 Consul tants M. Williams H. Hardy

-12 NRC Plant Project Managers A. Henriksen Division of~ Licensing B. Kirkwood U.S. Nuclear Regulatory P. Louzecky Commission A. Sarsten Washington, DC 20555 T. Spaetgens ATTN: B. Buckley S. Burwell 5 Senior Review Panel D. Hood D. Houston F. Albaugh K. Jabbour S. Bush' T. Kenyon C. Hill E. McKenna W. Richmond M. Miller L. Williams S. Miner C. Stahle 22 Project Team

-J. Stefano J. Alzheimer E. Weinkam A. Currie 2 U.S. Nuclear Regulatory D. Dingee Commission R. Dodge

'11 -

Public Document Room W. Gintner Division of Technical W. Laity (15)

Information and Document J. Nesbitt Control F. Zaloudek b' Washington, DC 20555 Technical Information (5)

Publishing Coordination (2)

Distr-1 '

i:

I i d

ML20111A655

Contents

Text

Title:

1.0 INTRODUCTION

2.0 BACKGROUND

1.0 INTRODUCTION

2.0 BACKGROUND

Navigation menu

ML20111A655

Text

Title:

1.0 INTRODUCTION

2.0 BACKGROUND

1.0 INTRODUCTION

2.0 BACKGROUND

Navigation menu

Search