Text

Update Rept on Tdi Standby Diesel Generators
	ML20084A175
Person / Time
Site:	Grand Gulf
Issue date:	04/30/1984
From:	; MISSISSIPPI POWER & LIGHT CO.
To:	;
Shared Package
ML20084A161	List: AECM-84-0240, Forwards Update Rept on Tdi Standby Diesel Generators. Info Provided in Support of Resolution of Tdi Diesel Generator Issues; ML20084A175;
References
	NUDOCS 8404240436
	Download: ML20084A175 (81)
	v • d • e

_ _ _ _ _

e.

CRAND GULF NUCLEAR STATION UNIT 1 UPDATE REPORT ON TDI STANDBY DIESEL GENERATORS April, 1984 8404240436 840420 gDRADOCK 05000416 PDR Z3rgl

e.

n.

i TABLE OF CONTENTS

fajL, ABSTRACT.

I

1.0 INTRODUCTION

2 2.0 PISTONS.

9 15 3.0 CYLINDER HEADS.

4.0 CONNECTING R0D BEARINGS.

19 5.0 PUSH RODS..

25 6.0 CRANKSHAFT.

29 7.0 L.P. FUEL LINE FAILURE.

36 8.0 H.P. FUEL LINE FAILURE.

38 9.0 CRANKCASE CAPSCREWS...

40 10.0 TDI PRODUCT IMPROVEMENTS..

42 11.0 QUALIFICATION / RELIABILITY DEMONSTRATION TESTING.

44 12.0

SUMMARY

49

13.0 CONCLUSION

S..

52 I.

14.0 REFERENCES

53 4

Y ATTACHMENT 1 - RESPONSES TO SIXTEEN PROBLEMS IDENTIFIED IN TDI OWNERS GROUP MEETING WITH THE NRC ON JANUARY 26, 1984...........

55 ATTACHMENT 2 - PISTON MANUFACTURING DETAILS.

75 TABLES TABLE 1 DELAVAL ENGINE SPECIFICATIONS..

4 TABLE 1-2

.CGNS D/G OPERATING DATA..

S' TABLE 1 DIVISION I AND II APPROXIMATE RUN HOURS UNDER LOAD SINCE PISTON SKIRT MODIFICATION IN NOVEMBER, 1981.........

6 TABLE 1-3A-RECENT SPECIAL TESTING...........

7 TABLE 1 D/G LOADINGS.....

8' i

Z3rg2

j TABLE OF CONTENTS (CONTINUED)

Page TABLE 2 RESULTS OF INITIAL INSPECTION OF GGNS MODIFIED AF PISTONS.

12 TABLE 2 RESULTS OF ADDITIONAL INSPECTION OF GGNS MODIFIED AF PISTONS.

13 TABLE 3 INSPECTION RESULTS OF DIVISION I CYLINDER HEADS.

17 TABLE 3 INSPECTION RESULTS OF DIVISION II CYLINDER HEADS.

18 TABLE 4 CHEMICAL SPECIFICATIONS LIMITS FOR ALC0A B850 22 6-T5 AL'JMINUM......-...................

TABLE 6 SHOREHAM AND GGNS CRANKSHAFT DATA.

32 TABLE 6 CRANKSHAFT STRESSES AS REPORTED BY VARIOUS ANALYSES...

33 TABLE 6 CRANKSHAFT LIQUID PENETRANT INSPECTION RESULTS..

34 TABLE 10 TDI PRODUCT IMPROVEMENT SIMS..

43 TABLE 11

SUMMARY

OF QUALIFICATION AND VALIDATION TESTING.

48 TABLE 15 ENGINE MOUNTED COMPONENTS PROBLEMS CAUSED BY TURBOCHARGER VIBRATION....

70 h

FIGURES FIGURE 2 LOCATION OF STUD BORE AREAS WITHIN PISTONS..

14 FIGURE 4 CONMECTING ROD BEARING DESIGN & NOMENCLATURE (SCHEMATIC) 23 FIGURE 4 COMPARISON OF CONNECTING ROD / BEARING CHAMFER ARRANGEMENTS.

24 FIGURE 5 WELDED BALL CONNECTOR ROD.

27 FIGURE 5 FRICTION WELDED PUSH R0D.

28 FIGURE 6 CRANKSHAFT COMPARISON..

35

..................=

FIGURE A12-1 ARTICULATED CONNECTING ROD ASSEMBLY.

64 FIGURE 15 ILLUSTRATION OF SHOREHAM TDI DIESEL GENERATOR LUBE OIL SYSTEM (REFERENCE TELECON SHOREHAM)...............

71 FIGURE 15 ILLUSTRATION OF GGNS TDI DIESEL GENERATOR LUBE OIL SYSTEM.

72 FIGURE 15 LEFT BANK TURBOCHARGER MOUNTING ARRANGEMENT 73 FIGURE A2-1 PISTON COMPARISONS.

'78 11 Z3rg3

ABSTRACT This report contains a detailed description of the program of preventive maintenance, replacement of components with improved quality products, engine testing, and engineering evaluations which have been undertaken by MP&L, GGNS Unit 1, to enhance reliability and to assure with a reasonable level of confidence that the Transamerica Delaval, Inc.

(TDI) diesel engines at Grand Gulf Nuclear Station (GCNS) will perform their required safety function.

This report is an update of the previous report submitted to the NRC in letter AECM-84/0103, dated February 20, 1984.

It addresses implementation of vendor recommendations, NRC directives, problems encountered on TDI engines at other locations, potentially significant items identified by the TDI D/G Owners Group, the results of several of the TDI D/G Owners Group technical reports, and updated operating experience information.

Z3rg4 1

1.0 INTRODUCTION

Grand Gulf Nuclear Station, Unit 1

is equipped with three diesel generators, two of which are supplied by Transamerica Delaval, Inc. (TDI).

These two diesel generators are sources of emergency power to the GGNS Division I and Division II ESF buses. The third D/G set, dedicated to the HPCS system, is supplied by the Electro-motive Division (EMD) of General Motors.

This report provides a detailed description of a program undertaken by MP&L to enhance the reliability and performance of the two TDI diesel generators at GGNS Unit 1.

The report contains a description of activities of preventive maintenance, replacement of various components with improved quality products, and testing of the two TDI diesel generators.

The improvement program includes specific actions which have been or are being taken to correct the problems experienced with TDI diesel generators during the start-up testing phase of GGNS Unit 1.

Potential problems identified to MP&L as a result of the experience with TDI diesel generators at other nuclear installations are also addressed.

The main emphasis of this report is to provide the results of an engineer-ing evaluation of the two TDI diesel generators at GGNS Unit 1 for their reliability and performance.

This evaluation is intended to provide reasonable assurance to the NRC that the TDI diesel generators will perform their required safety function. This report supplen.cnts earlier reports on the Division I and II diesel generators (Reference 12, 13, 14, 19, 20).

l Sections 2 thru 9 of the report contain descriptions of repairs or modifi-cations which have been performed. Section 10 concerns TDI's product im-provement program.

Section 11 focuses on the testing programs, both the testing done in the past and the testing performed af ter completion of the piston skirt changeout. A summary of the overall engineering evaluation is provided in Section 12. The conclusions reached from these evaluations are provided in Section 13. provides details of concerns raised at a meeting of the TDI diesel generator owner's group with the NRC on January 26 -1984, their applicability to Grand Gulf and their resolution. provides a summary of various piston skirt designs that have been or are in use in the GGNS TDI diesels.

Table 1-1 provides a list of the principal design specifications for GGNS Unit 1 TDI diesel generators.

Table 1-2 shows the total operating hours, starts, valid tests and valid failures for the GGNS TDI D/Gs.

Table 1-2 also shows that the ratio of valid failures (2) to valid starts (137) results in an excellent start reliability in excess of 98 percent.

Z3rg5 2

1.0 (Continued)

Table 1-3 shows the Division I and II approximate run hours under load since the originally furnished piston skirts were modified in November, 1981.

Table 1-3A shows additional testing that has recently been completed on the Unit 1 TDI diesel generators.

The 7 day runs, 24 hour2.777778e-4 days 0.00667 hours 3.968254e-5 weeks 9.132e-6 months runs and 100 hour0.00116 days 0.0278 hours 1.653439e-4 weeks 3.805e-5 months runs for the TDI D/Gs consist of 11 runs ranging between 60% to 110% load for a total of 695.7 hours8.101852e-5 days 0.00194 hours 1.157407e-5 weeks 2.6635e-6 months or an average of 63.2 hours2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months per run.

Table 1-4 shows the procurement specification estimated electrical loads and the present electrical loadings for the Division I and II diesel generator.

The significant work activities completed on the Division I and II cugines are:

o All piston skirts have been replaced with skirts of improved

design, i

o All 32 cylinder heads were inspected and eight cylinder hea'ds with rejectable indications have been replaced, o

All Division 1 and Division II connecting and main push rods have been replaced with components of improved design, o

All connecting rod bearings have been replaced, o

Inspection of both crankshafts has been completed, and o

Rework of turbocharger piping and components using ASME welding, procedures and materials has been completed, Realignment of the turbocharger has been completed.

o These work activities are intended to enhance engine performance and reliability.

They have insignificant impact on engine specifications, design criteria, subsystems or performance characteristics.

None of the work activities affect the design considerations listed in Table 1 of IEEE 387-1977.

As such, these work activities are considered minor design changes as defined by IEEE 387-1977.

Therefore, the post maintenance qualification and availability testing of these diesels was planned according to the guidelines established in IEEE 387-1977 for minor changes. Additional testing was also performed.

Z3rg6 3

i TABLE 1-1 DELAVAL ENGINE SPECIFICATIONS Model DSRV-16-4 Quantity 2

Engine Serial Numbr:r 74033-2624 & 74034-2625 Service Stationary generator for nuclear service Fuel Mode Diesel Configuration 45* "V" type No. of Cylinders 16 Bore (in.)

17 Stroke (in.)

21 Cycle Model 4 stroke Total Displacement (cu. in.)

76,266 Crankshaft Rotation CW (from Flywheel end)

Firing Order IL-8R-4L-5R-7L-2R-3L 6R-8L-1R-5L-4R-2L-7R 6L-3R Continuous Rating (kw) 7000 Overload Rating (kw) 7700 Crankshaft Diameter (in.)

13 Crank Pin Diameter (in.)

13 Z3rg7 4

L TABLE 1-2 GCNS D/C OPERATING DATA Total Run Hours Division I Division II Shop and Pre-Op Run Time (Hrs) 535 252 Since Date of OL Run Time (Hrs) 862 618 Total Run Time (Hrs)* )

1397 870 TOTAL NO. OF STARTS I

Delaval Shop Runs 310(2) 5 Pre-Operational Runs 60 60 Since Date of OL Runs 170 120 Total Starts ('}

540 185 NOTES:

1.

Source of Information - Delaval Technical Manuals.

2.

Division I engine had 300 prototype runs for reliability testing.

3.

Data as of April 4, 1984 4.

Valid Tests:

Div I -

84 Div II -

.53 137 Valid failures:

2 (1-Div I Control System Electrical Component) (1-Div.I - Unknown)

Reliability:

98.5%

Valid tests and failures'are as defined in Regulatory Guide

~1.108.

Z3rg8L 5

TABLE l-3 DIVISION I AND II APPROXIMATE RUN HOURS UNDER LOAD SINCE ORIGINALLY FURNISilED PISTON SKIRTS WERE MODIFIED IN NOVEMBER, 1981 TO APRIL 4, 1984 Load, + 5%

Division I llours Division II ilours

< 50 14 12 50 - 60 450 316 60 - 99 75 13 100 301 251 110 14 10 Z3rg9 6

i i

TABLE l-3A 1

RECENT SPECIAL TESTIN0 Pl!RPOSE Divinton 1 Divinton !!

7 Day Run 168 hr 0 60%

32 hr 0 60%

37 hr 0 60%

70 hr 0 60%

46 hr 0 60%

i After Platon Skirt Changeout Breakin Runn I hr 9 20%

1 hr 0 20%

1 hr 9 25%

2 hr 0 50%

2 hr 0 75%

1 hr 0 75%

2 hr 0 100%

2.5 hr fl 100%

2 hr 0 100%

1 hr 9 20%

4 hr 0 100%

.5 hr 0 100%

.8 hr 0 50%

.4 hr 0 100%

2 hr 0 110%

.2 hr 9 1101 23 hr 0 100%

1 hr 0 50%.

2 hr 9 110%

22.1 hr 9 100% '

Surveillnneen:

4 llour Rune 8 hr 0 100%

4 hr 9100%

(2 runs)

Rollability Runn i

100 lleur Rune 21 hr 0 100%

100.3 hr 0 100%

72.3 hr 0 100%

100 hr - 32 hr 0 100%, 68 hr 0 75%

4 1

l e

i t

N, s

23rg10 7

l I

l l

?

l l

i I

TABLF, 1-4 I

w D/G LOADTNCS

~

I 4

1 p

p --. :.

Division I Division 11 i

F Procurement Specification

$730 KW 6100 KW l

i

.s 4

7000 KW 7000 KW

Design DG Rating i

r Lo u of Offatto rouer Loada 3627 KW (31.8%)

4745 KW (67.8%)

f i

e

?

Post 1.0CA Loaje 4711 KW (67.3%)

3914 KW (55.9%)

[

a

~

l foc.it Conpacted EST Bus Load

$963 KW (85.2%)

6397 KW (91.4%)

i t

k

~

.-w I

1 t

t

[<

i

{

~

s, r,

j v,.

s

. +,,,

f y

.a i~

..l ;

ey.

+.

I w

i

- ~

23rstl O

8 s

l i

2.0 PISTONS

2.1 DESCRIPTION

MP&L received information from TDI that during a recent reassembly of the three TDI diesels at the Shoreham station, an inspection of the piston skirts revealed linear indications exceeding 1/16 inch in length in 23 of 24 modified "AF" piston skirts.

As a result of this finding, TDI generated a 10CFR21 report recommend-ing that GGNS and San Onofre inspect 25% of the modified "AF" piston skirts in each engine for linear indications. MP&L subsequently found rejectable indications in three of four modified "AF" piston skirts during the 25% inspection on the GGNS Division II engine.

All piston skirts on the Division II D/G were then inspected.

The results of these inspections are shown in Table 2-1 and 2-2.

The inspection criteria used for the inspection is described in Step 2.3 of this section.

2.2 ENGINEERING EVALUATION Failure Analysis Associates (FaAA) performed an inspection and analysis of the modified type "AF" piston skirts which were removed from the Shoreham diesels.

After comparing the GGNS Division II piston skirt inspection results with the Shoreham evaluation results (Reference 1), FaAA concluded that the GGNS Division I piston skirts could contain fatigue cracks of the same approximate depth as the Shoreham engines.

As a result of these early evaluations MP&L replaced all piston skirts in the two Unit 1 TDI engines with the improved "AE" style skirt provided by TDI.

(See Attachment 2 - for Details of Piston Designs).

MP&L worked with TDI in the final phases _of production and inspection of these piston skirts to assure that they are free of rejectable indications (as evidenced by fluorescent magnetic particle. examina-tion).

Results of an evaluation by FaAA (Reference 21) indicates that the.

"AE" type piston will exhibit substantially lower stresses than the replaced modified "AF" type under similar loadings.

FaAA has also indicated that these AE skirts have now been operated for over 300 hours0.00347 days 0.0833 hours 4.960317e-4 weeks 1.1415e-4 months in one of the Shoreham engines including 100 hours0.00116 days 0.0278 hours 1.653439e-4 weeks 3.805e-5 months of_ full power-operation.

Other AE skirts _ have accumulated over 6000 hours0.0694 days 1.667 hours 0.00992 weeks 0.00228 months in a stationery generating plant, and over 600 hours0.00694 days 0.167 hours 9.920635e-4 weeks 2.283e-4 months in an advanced development engine.. Inspection of these skirts (one after 6000 hours0.0694 days 1.667 hours 0.00992 weeks 0.00228 months , two after 600 hours0.00694 days 0.167 hours 9.920635e-4 weeks 2.283e-4 months ,-and four after 300 hours0.00347 days 0.0833 hours 4.960317e-4 weeks 1.1415e-4 months ) with a high-resolution eddy current procedure disclosed no cracking.

Experimental stress analysis of a type AE skirt was' conducted under hydrostatic loading of the-piston crown. The maximum stress measured was below the yield strength of the material and corresponds to cylic loading below the value required to produce a crack.

It was concluded that,l based on both analysis and test results, the.

"AE" type piston skirt-attachment'would not fail in fatique.

Z3rg12

.9

. s 2.3 PISTON INSPECTIONS t

(1) Division 11 modified "AF" piston skirts were nondestructively examined with liquid fluorescent dye penetrant and/or wet fluorescent magnetic particle processes.

The syecific area of concern was the filleted transition area between the skirt / crown stud hole bore and the skirt wall.

All critical filleted areas of each modified."AF" piston skirts were initially inspected with the liquid fluorescent dye penetrant process.

The results of this initial inspection are summarized in Table 2-1.

The following criteria were used in recording possible indications:

l 1.

all indications were to be evaluated, and 2.

a linear indication is defined as an indication in which its length is greater than three times its width.

Numerous indications were found, ranging from 1/32 to 9/16 inches in length.

The following additional inspections were performed to determine if the linear indications were superficial in nature.

Each linear indication was ground and/or sanded to a depth of approximately 0.062 inches. These indications were then re-inepected using the liquid fluorescent dye penetrant or wet fluorescent magnetic particle process.

Linear indications were found ranging from 1/32 to -1/2 inch in length.

The results of these additional inspections are summarized in Table 2-2.

To characterize these indications, a confirmatory metallurgical analysis will be performed.

The analysis will attempt to deter-mine the mode of cracking, characterize the crack propagation rate, and estimate the depth.

(2) All replacement "AE" piston skirts were nondestructively examined by TDI using the wet fluorescent magnetic particle process prior to installation in the engines. All TDI nondestructive examina-tion procedures were reviewed and approved by MP&L. The follow-ing criteria were established as levels of unacceptability:

1.

any linear indication greater than 3/16 inch long,

2. -

rounded indications with dimensions greater than 3/16.of~an

inch, 3.

four or more rounded indications in a line separated by 1/16 of an inch or less, edge to edge, and 4.

cracks and hot tears.

3 These acceptance criteria ' were ' derived ' from. ASTM Standard' E : 125-63, reapproved 1980.

s Z3rg13

' 10 ~

4

2.3 (Continued)

All piston skirt castings were accepted to the above criteria.

It

'should be noted that all acceptable indications that were found were documented by appropriate records.

2.4 MANUFACTURING DETAILS The manufacturing details for the "AF", modified"AF" and "AE" piston designs have been provided to MP&L by TDI.

The evolution of TDI's piston design is relevant to this report.

As such, the details of manufacture for each of these piston designs is presented in.

It is important. to note that the "AE" design utilizes a reinforced (lower stressed) casting and a half-stack. Belleville washer arrange-ment.

Also, "AE" skirts are heat treated to produce stress relieved nominal 100,000 psi tensile strength nodular iron.

The "AE" style skirt is interchangeable with existing R-4 piston crown and requires only minor hardware changes.

2.5 CONCLUSION

S As a result of the Division II modified "AF" piston skirt inspection, the piston skirts in the Division I and II D/Gs have been replaced with the type "AE" piston skirts.

l Based on results of the analytical work completed by FaAA and the operating experience and subsequent inspection of the piston skirts on the TDI Kodiak engine, R-5. test engine, and the Shoreham engine, MP&L has concluded that the "AE" design is. capable of - performing the required function at all running loads, and the " A E" piston skirt attachment would not fail in fatique.

4 J

1 1

.Z3rgl4.

.11

TABLE 2-1 RESULTS OF INITIAL INSPECTIONS OF GGNS MODIFIED AF PISTONS IN THE DIVISION 11 D/G Indication Length (Inches)(1)

Piston Stud Hole Bore Area (2)

Identification

1

2

3

4

1RB None 1/8 3/32 1/4,1/32,1/16

1LB None 1/32 None 1/16 1

2RB None 1/32 None 1/4,1/16

2LB 5/64,1/16 None 3/16 1/8,1/22

3RB 1/4 1/32,3/16 None 1/2

3LB 1/32 1/32 None 3/16

4RB None None 1/4 ~

None

4LB 1/16,1/8 None 3/32,1/8 1/16

5RB 1/3,2 1/32 1/4 1/4

5LB 3/32,3/32 1/8 9/16

'3/8

6RB 1/4 3/16 None 1/4 s

6LB 1/32 None 1/16-1/32

?#7RB None None None 3/32.

7LB None None

. None-1/32

8RB 3/32 None None None

8LB.

1/16,1/8 None.

1/16 1/4 General Notes:

i 1

(1) All. inspection performed using Liquid Fluorescent Penetrant Process.

-.(2)

See Figure 2-1 for location'of the-stud bore area within piston: skirt.

Z3rE15-

=12

.= -

TABLE 2-2 RESULTS OF ADDITIONAL INSPECTION OF GGNS MODIFIED AF PISTONS IN THE DIVISION II D/G Indication Length (Inches)(1)

Stud Hole Bore Area (2)

P ston Identification

1

2

3

4

1RB 1/8 (MT)

NAD (MT)

1LB NAD (MT)

NAD (MT)

2RB NAD (PT)

NAD (PT)

2LB NAD (PT)

NAD (PT)

3RB 1/4 (MT) 3/16 (MT)

NAD (MT)

3LB 1/8 (MT) 1/8 (MT) 3/16 (MT)

4RB 1/4 (MT)

4LB 1/8 (MT) 1/8 (MT) 1/16 (MT)

5RB NAD (MT)

NAD (MT) 1/4 (MT)

NAD (MT)

5LB NAD (MT)

NAD (MT) 1/2 (MT) 1/4 (MT)

6P.B NAD (MT)

NAD (MT) 1/4 (MT)

6LB NAD (MT)

NAD (MT)

7RB 1/32 (MT)

7LB NAD (MT)

8RB NAD (MT)

8LB 1/8 (MT)

NAD (MT) 1/4 (MT)

General Nutes:

(1) PT indicates Liquid Fluorescent Dye Penetrant Inspection.

(2) MT indicates Fluorescent Magnetic Particle Inspection.

(3) See Figure 2-1 for location of the stud bore area within piston skirt.

(4) -- Not performed. No discontinuities present during initial inspection.

(5) NAD No apparent defect.

Z3rg16 13

FIGURE 2-1:

LOCATION OF STUD BORE AREAS WITHIN PISTON

[

N x_-

i FIG 2-1 t

3.0 CYLINDER HEADS

3.1 DESCRIPTION

During disassembly of the Division II engine for piston inspections, red rust was reported in the area of the exhaust valve seats on the #5 right bank head.

Subsequent color contrast dye penetrant inspections showed cracks in the stellite exhaust valve seat overlays.

Because of the rusting, it is postulated that one of the cracks may extend into the water jacket.

3.2 INSPECTION AND ENGINEERING EVALUATION As a result of these cracks, an investigation has been initiated to determine the extent of the cracking and required corrective action.

The investigation has been divided into two parts; short and long term.

The short term investigation was initiated to determine if the extent of cracking is generic to all heads et GCNS.

All cylinder heads on both Division I & 11 D/Gs were removed and were nondestruc-tively examined in the area of the stellite seats with a color contrast solvent removable dye penetrant _ process.

The following criteria were established as levels of unacceptability:

1.

Any cracks or linear indications 2.

Four or more rounded indications in a line separated by 1/16 inch or less, edge to edge 3.

Any rounded indication with dimensions greater than 1/16 inch 4.

Linear indications are these indications in which the length is more than three times the width Two of the 16 heads on the Division II D/G and six of the 16 heads on the Division I D/G were determined to have Iej ectable indications.

Of these, only the Division II D/G #5 right bank cylinder head had an apparent through wall crack. No other visual evidence of cracking was found in the cylinder heads.

A description of'the indications found during these inspections are detailed for Division I in Table 3-1 and Division II in Table 3-2.

The heads which were_ rejected on the Unit 1 TDI engines were rejected for minor indications which were revealed during a liquid penetrant examination of the valve seat areas.

It is not known if-any of these indications would have propagated, but the decision was made. to install only clean heads with no indications. _ To address the long-term concern, a - failure investigation has been initiated.

A metal-lurgical evaluation will-be performed to determine the cause of crack-initiation, and the crack propagation mode.

.Z3rgl7 15

3.3 CORRECTIVE ACTION Based on the short term investigations, MP&L replaced the two heads on the Division II D/G and the six heads on the Division I D/G with heads that were examined and determined to have no rejectable valve seat indications. No further action is planned, pending the results of the long term investigation.

3.4 CONCLUSION

S Two heads on Division II and six heads on Division I were determined to have rejectable indications.

However, as demonstrated by the operability of the D/Gs prior to the replacement of these heads, the ability of the Unit 1 D/Gs to perform their safety function was not impaired.

The replacement of these eight heads with heads free of rejectable valve seat indications provides additional assurance that the potential for head cracking from this source is minimized.

To provide further assurance that any significant cracks in the heads will be detected, additional surveillance will be performed following D/G operation to detect the presence of water in the cylinders. These surveillances will be in addftion to the current surveillances which are designed to check for the presence of water - prior to manually initiated D/G starts.

r 1

J Z3rgl8

-16

TABLE 3-1 INSPECTION RESULTS OF DIVISION 1 CYLINDER HEADS Head Identification Inspection Results Number ILB No Apparent Defects 1RB No Apparent Defects 2LB Linear Indications on Fusion Zone Between Casting and Stellite Valve Seat 2RB No Apparent Defects 3LB No Apparent Defects 3RB No Apparent Defects 4LB No Apparent Defects 4RB Linear Indication in Stellite Valve Seat 5LB Linear Indicaticns in Stellite Valve Seat SRB No Apparent Defects 6LB Linear Indication in Stellite Valve Seat 6RB Linear Indications in Stellite Valve Seat 7LB No Apparent Defects 7RB No Apparent Defects 8LB No Apparent Defects 8RB Linear Indication in Stellite Valve Seat Z3rg19 17

i-I TABLE 3-2 INSPECTION RESULTS OF DIVISION II CYLINDER HEADS Head f, cation Inspection Results en f

ILB Incomplete fusion 5/16 inch long on intake valve seat 1RB No Apparent Defects 2LB No Apparent Defects l

2RB No Apparent Defects 3LB No Apparent Defects 3RB No Apparent Defects 4LB No Apparent Defects 4RB No Apparent Defects SLB Nr Apparent Defects SRB Twelve linear indications ranging' from 3/16 to 3/4 inch. All indications transverse to stellita overlay on two exhaust valve seats. All cracks' are contained

- within the valve seat except f or one, which cxtends:

from stellite into cast head material.'

6LB

_No Apparent Defects 6RB No Apparent Defects

'7LB No Apparent' Defects 7RB No Apparent Defects 8LB No Apparent Defects 8RB No Apparent Defects

~

P Z3rg20 18~

4.0 CONNECTING ROD BEARINGS

4.1 DESCRIPTION

Shoreham has experienced cracks in connecting rod bearings.

These cracks were discovered (See Reference 2), when LILCO disassembled the three diesel engines at Shoreham (TDI Model DSR-48) to investigate a crankshaft failure (See Section 6.0).

A complete inspection found that four of the forty-eight connecting rod bearing shells contained cracks.

Even though the Grand Gulf D/G design is significantly diff-erent (i.e., GGNS has articulated connecting rod design and reduced connecting rod bearing loads) an inspection and evaluation was per-formed to determine if this concern exists at Grand Gulf.

4.2 ENGINEERING EVALUATION, SHOREHAM BEARINGS A schematic of a cracked Shoreham bearing is shown in Figure 4-1.

FaAA performed an analysis (Reference 18) on one of the cracked Shoreham bearings.

The cracked bearing was checked for its chemical and physical properties.

A scanning electron microscopy (SEM) analysis of the fracture was also performed and dimensional checks were made for wear.

The chemical and physical properties met the current design specifications except for elongation.

The elongation was found to be below specification, however, the test specimen was not standard, and led to results that were inconclusive.

Reference chemical properties for B850.0-T5 are shown in Table 4-1.

A Shoreham SEM examination of the fracture face indicated that voids in the bearing shell may have been crack 1.iitiation locations.

In compret-

sien, voids in the " overhang" ar% would not pose a problem.

However, the bearing / rod arrangement on the Shoreham diesele did not support the end part of the bearings (Figure 4-2).

This unsupported end combined with the yawir.g of the crankshaf t would put the internal dianeter surface into tension.

The surfacc porosity acting as a stress intensifier may have contributed to crack initiation in the unsupported end

(" overhang" area).

Afro, the subsequent shell thickness measurements s;1 owed the bearing to be within the manufactur-ing tolerances, i.e., no appreciable wear.

FaAA has indicated (Reference 22) that two analyses were performed to determine the effect of the stress reduction on the fatique resistance of the new 12-inch bearing shells. A stress vs. number of cycles equation predicted that, based on the observed life of the 11-inch diameter bearing shells, the 12-and 13-inch shell fatigue life should be approximately 38,000 hours0 days 0 hours 0 weeks 0 months at full load, which is over ten times the usage expected over the 40-year service life of the nuclear stand-by diesel generators. Expected full load hours at GGNS are 75 full load hrs / year. An alternative analysis demonstrated that the decrease in the stress range is sufficient to prevent fatigue cracks, which indicates an infinite fatigue life for the bearing shells.

Z3rg21 19

Based on fracture mechanics analysis, an acceptance criterion for discontinuities in the aluminum was established. Voids up to 0.050 inch in diameter will not compromise the fatigue performance of the 12-inch and 13-inch connecting rod bearing shells in DSR-48 or DSRV-16-4 standby diesel generators. The maximum load required to be carried by the GGNS TDI D/Gs during an emergency event is only 68% of rating (Section 1-Table 1-4).

Further analysis is being performed by FaAA to determine fatigue life at lower loads assuming larger bearing voids in order to envelope less conservative conditions.

4.3 GGNS D/G INSPECTION The inspections delineated below were performed on Division II D/G components. New bearings were installed in both divisions to expedite the return to service of the diesel and to extend the replacement period of the bearings.

The integrity of these replacement bearings was based on an exact pare number exchange, visual inspection of the bearings before installation, and favorable results from the inspection / evaluation performed on the original bearings.

1.

All of the connecting rod bearings were dimensionally checked for wear and signs of unusual or abnormal wear patterns.

2.

Two (25% of total) of the connecting rod bearings were inspected by liquid penetrant (PT) and radiography. The radiography tech-nique utilized an x-ray tube radiation source and obtained a 2-2T film sensitivity yielding at 1 cast of 0.015 to 0.020 inches reso-lution. The PT inspections used a liquid fluorescent dye pene-trant process and met the requirements of ASTM Standard E163.

No rejectable indications were found.

Tests to check the chemical and physical properties of two (25% of total) original coanecting rod bearings are planned.

These tests will be performed in accordance with applicable ASTM 3

Standards. These tests are considered confirmatory in nature.

3.

The " overhang arrangement" of two bearings / connecting rod assem-blies was dimensionally checked for unsupported bearing material.

The chamfers on the connecting rods and rod bearings were dimen-sionally. checked to determine if the " overhang arrangement" exists and to verify that the = connecting rod configuration was of the correct design.

4.4 INSPECTION RESULTS The initial Division II D/G inspections indicated the following:

1.

Review of the radiographic film showed.that any bearing porosity was less than 0.030 inches, however, some linear type indications were present. All linear type. indications were directly trace-able to minor gouges and marring located on the bearing surfaces Z3rg22 20

i i

l 4.4 (Continued) which occurred during disassembly.

As indicated in Reference 22 porosity of less than 0.050 inches is predicted to be of little consequence to the satisfactory operation of the bearings.

2.

The results of the dimensional inspections confirmed that the bearings were within manufacturing tolerances. No signs of unusual or abnormal wear patterns were noted. This indicates that there was no misalignment between the connecting rod assemblies and the crankshaft.

3.

The results of the chamfer measuremer:ts indicate that there is no

" overhang" arrangement on the #7 connecting rod / bearing assembly and that the #2 connecting rod / bearing assembly has an " overhang" of approximately 0.016 inch (i.e., 0.016 inch of unsupported bearing material). This amount of " overhang" is insignificant compared to the 0.25 inch " overhang" that existed on the Shoreham bearings at the time of bearing cracking.

4.

The results of the Liquid Fluorescent dye penetrant examination indicated that no cracks were present.

4.5 CONCLUSION

S The differences in design between Grand Gulf and Shoreham (i.e.,

articulated connecting red design and reduced connecting rod bearing loads at Grand Gulf) preclude the types of problems that Shoreham has experienced.

However, inspections were performed to verify the ade-quacy of the hearings.

Inspections of the original Division II D/G connecting rod bearings showed that no appreciable wear or unusual wear patterns were present.

This confirms proper aligament of the connecting rod arsemblies to the crankshaft.

During the piston skirt changeout on the CGNS Unit 1 D/Gs the.

connecting rod bearings were inspected for _ unusual or abnormal wear pat te rns.

No signs of unusual or abnormal wear patterns were noted.

Two of the original Division II, D/G connecting rod bearings were inspected by radiography. The radiography _ technique = utilized an X-ray-tube radiation source and obtained a 2-2T film sensitivity yielding at

.least 0.015 to 0.020 inches resolution.. Review of radiographic film showed that bearing porosity was less than 0.030-inches which,is well below the preliminary acceptance criteria of 0.050 inches established by the TDI. D/G --Owners Group.

(. Reference - 22).-

Further. analysis is being performed by FaAA to determine fatigue. life at lower -loads assuming larger voids to envelope more realistic service conditions.

-Z3rg23 21

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

_,...m TABLE-4-1 l

CHEMICAL SPECIFICATION LIMITS FOR ALCOA B850.0-T5 ALUMINUM 4.

Element Composition %

Si 0.4 Max Fe 0.7 Max Cu 1.7 - 2.3 Mn 0.10 Max Mg 0.6 - 0.9 Ni 0.9 - 1.5 Sn 5.5 - 7.0-Ti 0.20 Other Flenents 0.30 Max Al Remainder a

s 4

3 i _

f

'22:

-23rg24-

FIGURE 4-1:

SHOREHAM CONNECTING R0D BEARING DESIGN AND NOMENCLATURE (SCHEMATIC) and N

Annular groove f

'N Cil he.

t Ocwel pin CRAcx

/

A furninum

/ tack i

Scresder ul

[G groove

[ ai, abbit Nb Parting line overlay on surfac 4

\\

l LENGT H orCRAC)r insid's

/

Longn diameter

/

6 FIG. 4-1

FIGURE 4-2:

COMPARISON OF CONNECTING ROD / BEARING CHAMFER ARRANGEMENTS

~

CONNECTING ROO 03 -540 - 0 3 - O C A x 43* CNMx \\

v.nyn e w,4

_,4$ e cgy, B MG. SHM::L a

0 3 * * ~ 0 3' # A

%A = 49.42 iMS

_-e

~

I t"DI A.

Q (ll" CRANKPIN)

SCHEMATIC 0F SHORE!!AM "0VERHANG ARRANGEMENT" IND_ICATING UNSUPPORTED BEARING MATERIAL CONNECTING ROO 02-340- s t - A J

,kz45*cHf.N3 x

x BRG. SHELL 0 #*

I6 O2-340 AG 3

--%A s 18.7.7 IN5 13~OlA.

gy (13* CRANKPIN)

SCHEMATIC 0F GGNS BEARING ARRANGEMENT WITH BEARING MATERIAL SUPPORTED FIG. 4-2

5.0 PUSH RODS

5.1 DESCRIPTION

On August 11, 1983, during the performance of unrelated maintenance, a rocker arm connector push rod was found to have a cracked weld. The push rod ball disengaged from the shaft as the push rod was removed from the Division I engine.

The defective connector push rod was replaced and the Division I engine was tested and returned to service with additional connector push rod inspection criteria specified.

During a subsequent inspection of the Division I engine, 14 of 16 connector push rods were discovered with cracked or separated welds.

This inspection revealed one of the connector push rod balls was cracked in addition to the weld cracks previously observed.

During the inspection of the Division II engine in December, 1983, 13 of 16 connector push rods were also found to have cracked tube-to-ball welds.

5.2 ENGINEERING EVALUATION AND CORRECTIVE ACTION There were two types of push rod designs used at CCNS. The main push rods had a tubular steel shaft fitted with hardened steel end pieces which were attached to the tube with four plug welds near the ends of the tube.

According to TDI, an estimated 2 percent of this design developed cracks in er adjacent to the plug welds on the rods.

The connector push rod consisted of a tubular steel body fillet-welded to carben stael tall bearings. This design is the t.ype thich erbth-it.ed defects at Grand Gulf and is shown in Figure 5-1.

A 1 1/?-inch high carbon steel ball bearit.g is fitted to 1 1/4-inch CD tubing with a 1/4-inch wall.

The inside edge of the tubing has a 45* chamfer which results in a 7/8-inch circular seating ring for the ball at the end of the tube.

The ball is attached to the tube with a centinuous 360* fillet weld. The materials of constructin are as follows:

Ball Material: AISI 52100 Tube Material: ASTM A519 Weld Material: UNIALOY 850 The first connector push rod found defective was subjected to metal-lurgical evaluation (Reference 3).

The initial weld defect resulted from lack of penetration of the fillet weld with the tubing. Destruc-tive examination of the ball and weld on the opposite end of the de-fective connector push rod revealed additional cracks in the heat-af fected-zone (HAZ) of the ball bearing. The velds exhibited lack of penetration and slag inclusions in the crevice area behind the weld.

The metallurgical evaluation concluded that the ball material is difficult to weld.

The possibility of finding underbead cracks all around the ball in the HAZ is very high.

Z3rg25 25

5.2 (Continued)

Previous operational experience did not indicate that the cracks would propagate out of the HAZ since the connector push rods are loaded in compression.

Furthermore, none of the MP&L defects or other reported defects were associated with underbead cracking.

Rather, all previous defects of this design were associated with insufficient weld penetra-tion.

Consequently, MP&L concluded that a push rod exhibiting these defects would not result in engine failure.

MP&L proceeded with an interim inspection program, until replacement connector push rods free of defects could be obtained.

The discovery of a cracked connector push rod ball in the Division I diesel, however, demonstrated that the underbead cracks could, in fact, propa-gate through the ball material.

A new replacement push rod design (Figure 5-2) had been developed by l

TDI.

This new design consists of a tubular steel shaft which is friction welded to cylinders of alloy steel on each end.

These ends are then machine finished and hardened.

The tube material is ASTM A-106 Grade B steel; the ends are AISI 8740/50 steel.

During December, 1983, MP&L engineers reviewed all aspects of the push rod fabrication and observed procedural qualification runs at Delaval's push rod fabricator in Los Angeles.

Samples of the qualification run were analyzed by MP&L and determined to be acceptable.

5.3 CONCLUSI0FS All intake, exhaust and connector push rods on both Unit 1 engines are the new friction welded push rods.

MPSL plans no further action on push rods; however, copies of metallurgical evaluations of the old and new push rods are being provided to the.TD1 D/G owners group.

FaAA has performed a cyclic wear test to 10 cycles en a sample friction welded push rod af ter which it was examined metallurgically.

No signs of abnormal wear or deterioration of the welded joint were observed.

A metallurgical evaluation of a connector push rod with 100 hours0.00116 days 0.0278 hours 1.653439e-4 weeks 3.805e-5 months of operation at 100% load is also planned.

MP&L considers this problem resolved with no further action required.

Z3rg26 26

FIGURE 5-1:

WELDED BALL CONNECTOR 03-390-04-AB F -0 05 - 074 FILLET WELD EACH END

/

1

_ _ __ _ _ i x x x s i sy ss s sxv_ _ _ _, _ _,

2 s

+

)

p N

f x m s. - - - h

\\ '<. - - - --

, x x s x s.

A l

4 I

FIG. 5-1

FIGURE 5-2:

FRICTION WELDED PUSH rod 1-07-AH O&39O AF O2-390-07-AF FRICTION WELD EACM END 4

__-_-_.m.

--. - =..

m.

_ j

}

9 e

1 l

e i

FIG 5-2

6.0 CRANKSHAFT

6.1 DESCRIPTION

The concern over the design adequacy of the crankshaft was prompted by a crankshaft failure that occurred at Shoreham with their TDI supplied standby diesel generators.

Investigation by FaAA revealed that the cause of the crankshaft failure was high cycle fatigue. This led the NRC to issue IE Information Notice No. 83-58 which identified Grand Gulf as having TDI supplied standby diesel generators with possible crankshaft design deficienc'.s.

6.2 ENGINEERING EVALUATION Investigations were immediately conducted by MP&L on the applicability of the failures at Shoreham to the Grand Gulf TDI diesel generators.

A physical comparison (Reference 19) of the DSR-48 (in-line eight cylinder) series engine crankshaf t that failed, with that employed in the DSRV-16-4 (Vee-16 cylinder)' series at Grand Gulf revealed some important differences.

Among the significant design improvements on the Grand Gulf engine are the larger web size and shape, larger crankpin

diameter, larger pin fillet radius and the use of counter-weights.

In addition, it was learned that TDI may have used potentially non-conservative (1st generation) design harmonic coefficients in 1974 when the Shoreham stress analysis was performed.

The original CGNS design stress calculations utilized later 1975 (3rd generation) harmonic coefficient values.

At the request of MP&L, TDI clarified the use of the CGNS 1975 Vs 1983 post-Shoreham 4th genera-tion coefficients and recalculated the GGNS D/G shaft torsional stresses using the latest coefficients (Reference 4).

The changes made in the newest harmonic coefficients were an analysis refinement that resulted from analytically generated results being compared to actual test results.

The changes were minor and did not substantially affect the analysis. The TDI results indicated that the CGNS crankshaft stresses are significantly less than the maximum allowable Diesel Engine Manufacturers Association (DEMA) standard and are also only ~~ 60% of the stresses in the failed crankshaft at Shoreham.

This confirmed that a substantial design margin existed in the GGNS crankshafts.

To verify the adequacy and results of the TDI crankshaft design analysis, MP&L requested Bechtel to evaluate TDI's analytical methods.

Bechtel concluded that the analytical methods used to predict crank-shaft stresses by TDI are in accordance with industry standard prac-tice and appear to be properly applied (Reference 5).

The results of the Bechtel analyses are summarized below:

o The shaft configuration lends itself to a simple dynamic model which adds assurance to the accuracy of the calcula-tion. The calculated first mode natural frequency has been confirmed by results of the torsiograph test, while the predicted single mode shaft stresses are within the DEMA allowables.

Z3rg27 29

1 l

6.2 (Continued) o The harmonic coefficients, cylinder firing sequence, and engine configuration are such that the response of the major orders of critical speed are minimized.

The harmonic coefficients have not been verified but it appears that a significantly detailed effort has been undertaken by TDI to provide accurate values, o

The TDI analysis did nor combine the response from the various harmonics of a given mode and of other modes to calculate total stress.

However, because of the expected random phasing, the reduced effects of higher modes, the first mode stress margins compared to DEMA, and the torsiograph results, the total stress remains acceptable, A comparison of the Grand Gulf and Shoreham crankshafts has o

been provided in Table 6-1.

The improvement in the web area, fillet radius, properly applied counterweights, and shot-peened fillet radius surface finish provide for a significant reduction in stress concentrations, o

The torsiograph results provide verification of front end angle of twist and an indication of shaft stress even though it is not a direct stress mersurement.

One important piece of information suggested by the torsiograph tests is that the first mode dominates the response of the crankshaft.

This would tend to confirm TDI's use of first mode response to predict crankshaft stresses.

o To address the total stress in the circular portion of the shaft, Bechtel performed an independent dynamic analysis using the normal mode method and applying modal superpost-tion (Reference 16).

Five sets of harmonic coefficients were considered in the analysis with the most important-being the actual measured gas pressure values obtained from.

an engine of the same configuration and BMEP. The harmonic-coefficients used by TDI are in good agreement with those derived from the measured gas pressure values.

The results of-the single order and total stress calculations are tabu-lated in Table 2 along with other crankshaft stress results for comparison.

o It should be noted that TDI's, analytical crankshaf t stress determination is based on individual. harmonics within a given mode. - TDI did not determine the stress for a specific harmonic due to the response of all modes, or -- sum the

=

effects. of ' all harmonics and - stresses from.. experimental shaft deflection measurements to which a theoretical - de-i flection/ stress relationship was applied. The theoretical deflection / stress relationship is based solely on the characteristics of the first mode, whereas, the measured-deflection includes the response of all modes.1 Z3rg28-30 -

__m

~

i 1

I 6.2 (Continued)

I o

The value.of overall stress. reported by FaAA for the l

Shoreham crankshaft represents the average stress taken over i

the peak stress excursion.

To provide a meaningful com-l parison a similar average stress was computed by Bechtel for the GGNS crankshaft.. An average stress is a useful value i.

for comparison with DEMA standards since the measure of stress reversal is directly. related to fatigue life.

The peak stress calculated by Bechtel for GGNS crankshaft is 6034 psi.

Both the peak stress and the averaged reversed stress for the GGNS crankshaft are within, the limits for allowable stress published by DEMA.

More importantly it should be noted that the total GGNS crankshaft stress is 4

lower than the FaAA calculated stress.for the new Shoreham crankshaft, even though the rated output of the CGNS diesel is twice that of the Shoreham diesel.

As a further verification of crankshaf t - adequacy, during December, 1983, and January, 1984, when' the Division I and II engines were dis-assembled for maintenance and replacement of existing piston skirts with improved piston skirts, the Division I and II crankshafts were i

inspected using accepted NDE methods.

No rejectable indications. were i

discoverad.

1 All the rod bearing journalc.were examined using a' liquid fluorescent dye penecrant process. The entire journal surface was inspected wi*.h.

particuler. a*:tentf or to the journal fillets.

All licear _ indications were evaluated with respect to integrity. The results of the examina-cion are showa in Table 6-3.

t

6.3 CONCLUSION

S i

The method of analysis used by TDI has been reviewed and is in acenrd--

~

ance with industry standard practices.'. Additionally the total stress i

values not addressed in TDI analysis have been calculated based upon_-

measured gas pressure _ input and are shown.to be within. the ~ DEMA

~

limits. Liquid penetrant' examination has shown no defects to exist.on the journal fillets.

I~

The total stress analysis results'are lowerl than DEMA' recommendations and when combined with acceptable liquid' penetrant' examinations alle-

.viates the concerns over the design adequacy of-the.GGNS crankshafts.

~

1 t

Z3rg29-

[31

.. 2

. - - - ~

-. ~

i*

TABLE 6-1 SHOREHAM AND CGNS CRANKSHAFT DATA i

!~

Shoreham R Series Grand Gulf RV Series

. Web Width 21 in.

25 in.

4 i

Web Thickness 4 1/2 in.

5 1/8 in.

Web Shape Flat Sided Round Crank Pin Dia.

11 in.

13 in.

T Fillet Radius 1/2 in.

3/4 in.

Fillet Finish Not Shot. Peened Shot Peened v

f s

s i

i 1

l l

.i 1'

i

.Z3r 30.-

. 32 -;

~

.. ~

TABLE 6-2 CRANKSHAFT STRESSES-AS REPORTED BY VARIOUS ANALYSES Single Order Stress (psi)

Total Average Stress (psi)

Crankshaft Bechtel FaAA TDI Bechtel FaAA TDI Shoreham (11" pin) 5790(15) 4570(2) 8910(15) 5314( }

Shoreham (12" pin) 3300( )

2990( )

5640(I )

4208( )

GCNS (13" pin) 2389(16) 1967(4) 5084( 6) 3507( ' )

General Comments:

(A) DEMA limit for single order stress is 5000 psi and 7000 psi'for the total stress.

(B) Referer, es to the source of information are identified in parentheses.

(C) The differences between the TDI and Bechtel calculations are primarily'due to Bechtel's summation of stresses from all modes (modal superpositions)-and TDI's method of including only harmonics within a single mode. For additional information see discussion in Section 6.2.

Z3rg31.

33

1 TABLE 6-3 CRANKSHAFT LIQUID PENETRANT INSPECTION RESULTS Rod Learing.

Journal Number Inspection Results Div I All Journals Indications were present - evaluated as wear surface marks and marring caused by micrometer measurements.

No apparent defects.

Div II

1, #2, #3 Indications were present - evaluated as wear surface

4, #5, #6 marks and marring caused by micrometer measurements.

and #8 No apparent defects.

Div II 9

7 No indications pr. tent - No apparent defects 1

-r

~

'Z3rg32 341

FIGURE 6-1:

Crankshaft Comparisons 13"CRAFEPIN

. FRACTURE PLANE p

\\

l3" SHAFT l

i 2

-)l - e -

p-COUNTER WElGitT

~

g 1

[j jM

_ _ _ + _ _ j_

j Q)

N SHOREHAM - REPORTED SECTION "AA NO COUNTERWElWTS a

ROTATED 90*

GGNS CONFIGURATION EHOREHAM CONFIGURATION w2 i

'l 4.4 i

t;5 e

c 36 j

>A b

f" (

h' V

"~

(' },

l 4n C

buli, d

'i =

I 4

4 16 CNN ER a'

si g,

l

_____.I__

i:

M-563G 8 M-6530

_ rl d-l

-f 1 - ' - -

-- P

- - =

l I

l

,:ss....J L.. J_..ss.l, ENNNv3 e

i----.

r

[

n,

\\

->-A Fig 6-1

f 4 >.

.,s

,,v..'

7.0-L.P. FdEL LINE FAILURE

7.1 DESCRIPTION

4

~

On September 4, 1983,the Division I D/G was started for maintenance operation. The engine was manually stopped and the outside fresh air fans secured when a ftre was reported at the engine. The fire was caused by a break in a 1-inch fuel oil supply header. The break sprayed fuel. oil'onto the exhaust gas piping to the left bank turbo-charger. Closer examination revealed that the tubing cracked circum-ferentially'along a line between the two ferrules of the Swagelok connector which connects the 1-inch tubing to the cross connect pipe

'ikcueen the the right and left bank fuel oil supply lines. The fire r(quired extensive rework and replacement of various components.

s.2 ENG'NEERING EVALUATION T)'ree possible causes of the tubing failure identified in the analysis by Middle South Services (Reference 6) are as follows: (1) an improper tubing ma.;erial. (2) inproper fitup and assembly of the tubing con-nector, and (3) vibrdtion loading.

t

11) 'The strength of a Swagelok type connection depends on controlled deformation of ' the tubing between the body of the connector and the front and back ferrules.

Consequently, the tubing must be ductile enough to deform significantly without cracking.

Swagelok recommends an ASTM A179 material.

Metallurgical analysis revealed that the tubing composition, hardness and ductility were all within the specified ranges for ASTM A179 material and that the tubing was acceptable for the application.

The 0.049-inch tube wall was the minimum recommended by Swagelok, but more than adequate for the operating pressures.

The tubing was replaced with Delaval standard spares.

(2) A Swagelok representative from the Oakland Valve and Fitting Company, inspected the failed tubing and the associated Swagelok conne'etor which had been sectioned for analysis.

The Swagelok representative stated that the tubing had been properly deformed and that fitup or assembly problems would have been very un-likely. The Swagelok fitting was replaced with Delaval standard spare n.aterial.

1 (3) Vibration and fatigue were the most likely causes of the failure.

The assembly of the Swagelok fitting forms a small ledge which acts as a stress concentrator.

There were no supports on this section of tubing, although Delaval drawing 02-450-13 shows a clamp or support as item number 7.

W e-Si

~

Z3rg33 36

(3)

(Continued)

Analysis revealed no evidence of cracking at the other end of the tubing at the fuel oil filter Swagelok connection. The crack in the failed section of tubing initiated at the root of the ledge.

The crack was initiated and propagated by high cycle fatigue mechanisms. This particular section of tubing had been subjected to unusual vibration loading by a defective left bank turbo-charger.

The root cause of the f ailure was determined to be the unusual vibra-tion loads imposed by the defective turbocharger combined with the absence of any supports to isolate the Swagelok connector from vibra-tion loading.

The defective, out-of-balance turbocharger (combined with a period of operation after the turbocharger mounting bolts were discovered to be loose) is suspected as the initiating source of vibration.

The fuel oil header, to which the failed tubing section connects, is mounted to the turbocharger mounting pedestal. This turbocharger, however, was replaced prior to the ultimate failure of the tubing which resulted in the fire.

7.3 CORRECTIVE ACTIONS MP&L designed and installed a tubing support for this section of tubing on both standby diesels.

In addition, following completion of all rework related to the fire, the engine was subjected to a mainte-nance run to verify that all components were functioning properly.

During the maintenance run, the engine was instrumented for vibration analysis.

The results of the vibratory analysis revealed that the engine exhibited vibration levels which were well within the limits which could be expected from this type of machinery.

These actions were described to the NRC in Reference 12.

MP&L contracted Technology for Energy Corporation (TEC) to perform vibration testing on both Unit 1 engines following the September, 1983 fire rebuild effort.

During the course of this testing. all piping systems in the area of the turbochargers were inspected along with most other major engine components.

No areas of abnormal or suspect vibrations were reported.

In conclusion, TEC stated that both engines had normal vibration levels and that no further vibration problems should be anticipated in the operation of these engines.

MP&L has further committed to develop and implement a vibration monitoring program to routinely inspect both Unit 1 engines.

7.4 CONCLUSION

S The root cause of the low pressure fuel line failure was attributed to unusual vibration loading and' the absence of any supports to isolate the Swagelok connector from this loading.

The correct;Lve actions taken alleviates the loads imposed on these lines. Therefore, further failures of these lines are not expected.

Z3rg34 37-

o.

8.0 H.P. FUEL LINE FAILURE

8.1 DESCRIPTION

Shoreham experienced a failure of a fuel inj ection line during pre-operational testing.

TDI filed a 10CFR21 notification on July 20, 1983 to alert the NRC to a deficiency involving a possible draw seam on the ID of the high pressure fuel injection lines supplied on TDI diesel generators. The tubing failures at Shoreham were attributed to the draw seam which acted as a stress riser and failed when subjected to repeated operating cycles (about one million cycles).

At approximately the same time of the notification, a high pressure fuel injection line on the GGNS, Unit 1, Division I diesel generator failed.

An analysis of the failed tubing attributed the failure to the tubing manufacturing flaw.

8.2 ENGINEERING EVALUATION AND CORRECTIVE ACTION All of the GGNS D/G fuel lines were original equipment, except one on each division, and had been subjected to more than ten million operat-ing cycles.

Therefore, they were considered free of defects of this type.

The two lines that were not original equipment had been re-placed during startup testing because of leakage around the fittings.

One of these two replacement lines subsequently failed, as stated above, at approximately one million cycles.

Based on the results of an analysis performed by Middle South Ser-vices, (Reference 7), the failed tubing exhibited a crack which initiated from a manufacturing flaw on the inside surface of the tube.

The flaw, which ran the entire length of the failed tubing section, was formed by a defective mandrel during the initial extrusion phase of the forming process. Additional rolling operations lapped over the flaw, which was about 6-8 mils deep.

The fuel inj ection line operating pressure, which cycles between atmospheric pressure and about 5000 psi, provided the fatigue loading which produced cracks along the stress riser provided by the manufacturing defect.

The preexisting flaw acting with the fatigue stresses gererated by the cyclic operating pressures produced the failure.

These evaluations and actions were described to the NRC in Reference 12.

8.3 CONCLUSION

S The TDI 10CFR21 notification indicates that the failures occur at approximately one million operating cycles and that fuel lines that have in excess of ten million operating cycles without failure are satisfactory.

All of the original lines on the Division I ' and -II diesels were, therefore, considered free of internal flaws of ' this type because they have in excess of ten million operating cycles and have not failed.

One line on the Division I diesel, the one that failed and was replaced, and one line on the Division II diesel were not original lines and were considered suspect.

Replacement lines were ordered and installed in place of these two lines.

Z3rg35 38

8.3 '(Continued)

TDI has inspected the fuel line material used for the new MP&L lines using a sampling technique where a 1-inch long portion is cut from each end of each 17-20 foot long stock tube. These short sections are split axially by saw cut, and the bore surfaces checked for draw deficiencies.

The basic assumption of the inspection is that any deficiencies in the tubing would exist throughout the entire length.

If there are no imperfections found in the end pieces, then there are none in between, and the tube is considered acceptable.- The new MP&L lines successfully passed the TDI' inspection.

This problem is, therefore, considered resolved for the CGNS Unit 1, TDI diesels.

H

.I Z3rg36 39 1

?.

4 9.0 CRANKCASE CAPSCREWS

9.1 DESCRIPTION

During the performance of a 24 hour2.777778e-4 days 0.00667 hours 3.968254e-5 weeks 9.132e-6 months run test on March 15, 1982, the Division II D/G tripped on a " Generator Differential" which was accom--

panied by an observed electrical arcing flash inside the generator.

I In a subsequent inspection of the generator it was found that the stator insulation had been damaged and that a 15/16 inch capscrew head

-from a 5/8 UNC X 1-3/4 inch long capscrew had imbedded in the stator and damaged the generator.

It'was determined that the capscrew head was from a capscrew on the diesel's rear crankcase cover that had sheared off and entered the generator through the air gap on the end 4

of the generator. The generator was replaced with a generator from Unit 2 and all rear crankcase cover capscrews on the Unit 1,, Division I and II diesels, were replaced with new replacement capscrews.

An independent lab performed an analysis (Reference 9) of the 42-capscrews removed from the Unit 1, Division I and II diesel genera-tors. A review of the analysis produced the conclusion that the failure mode was due to a low-stress fatigue front expanding from an initial small crack. It was also noted that the failed capscrews had a decarburized skin which may have contributed to the failure.

On October 4, 1982, the rear crankcase cover capscrews were checked for the correct tightness (60 ft-lbs). Three of'the capscrews on the Division II diesel generator were found'to be less than 60;ft-lbs'(20, 23 and 35 ft-lbs).

Any. capscrew not within 2 f t-lbs of the. 60 f t-lbs was to be torqued -to - within the acceptable range.

When the capscrew that was: found at '20. f t-lbs 'was tightened,, it.. sheared ~ of f -

approximately one inch from the bottom side of the head before; reach-ing 60 ft-lbs.

4

.9.2 INSPECTION'AND TESTING The Division II D/G was instrumented by Nutech in January of 1983 and data was obtained during an operational test run.--The test' data

indicated that the highest vibration amplitude occurred during the startup and shutdown of the diesel, with.capscrew stresses at 6000 p s i.' The vibration amplitude was much less during steadyistate' opera-tion at 450 RPM, vith the capscrew stresses at 3000 psi..However, the test-results.were inconclusive as~to the root causes of the vibration source. The present information indicates that.the capscrews failed

~

by a combination of metallurgical and transientLvibration factors 'and-that the failures'are' unique to'the Division-II D/G.

lZ3rg37 140

9.3 CORRECTIVE ACTIONS The main thrust of the corrective action taken was the design and installation of protective screens for the generator air gaps. The failure of the rear crankcase cover capscrew, by itself, would not prevent the diesel from performing its safety function.

On the other hand, the entry of foreign material into the generator could cause failure, therefore, the screens were installed to protect against a similar mode of failure.

At the same time fatigue resis-tant, high strength capscrews and tab washers were installed to extend the life of these capscrews.

One of these capscrews was pulled from each division and subjected to destructive analysis.

While there was no sign of crack initiation there were signs of frett-ing on the threads of the capscrew removed from the Division 11 D/G.

The expected life of these capscrews has not been confirmed.

After the metallurgical report is evaluated, the schedule for removing another bolt from the Division II D/G for analysis will be determined.

9.4 CONCLUSION

S Although MP6L is continuing to inspect the crankcase cover capscrews and isolate the source of cyclic loading, the possibility of failure of one of these capscrews no longer poses a threat to diesel generator operability due to the installation of protective screens on the gene-rator air gaps.

Z3rg38 41

10.0 TDI PRODUCT IMPROVEMENTS TDI has a product improvement program which addresses both changes that are required to ensure diesel generator operability / reliability and changes that are developed to extend component life, allow easier maintenance operations, or use improved manufacturing techniques. The TDI program classifies changes as follows:

(1) changes required to correct 10CFR21 deficiencies, (2) changes developed to improve diesel generator performance or reliability (not as a result of a potential defect) and issued to custo-mers under TDI's Service Information Memo (SIM) program, and (3) changes developed by TDI that are determined by TDI to be relatively insignificant to diesel generator operation and therefore do not necessitate immediate customer notification.

The TDI program of product improvement has included applicability reviews for the diesel generators installed at Grand Gulf and the applicable changes have been identified to MP&L (Reference 2).

The TDI Nuclear Check List for SIMS identifies those that are applicable to TDI diesels at nuclear stations.

The thirty-three SIMs identified by the list were reviewed by MP&L to determine which SIMs could be considered product im-provements.

Four categories; product improvement, instructions, informa-tion and guidelines were utilized for the review.

Eight of the thirty-three SIMs reviewed were considered to be product improvements, nine SIMs as recommended instructions, ten SIMs as informational and six SIMs as guidelines.

A listing of the eight SIMs considered product improvements is provided in Table 10-1.

Review of the vendor manual for the TDI diesels and other documents indicates that the eight product improvement SIMs have been incorporated on the Unit 1, Division I and 11 diesel generators.

A continuing review will be performed for TDI SIMs as they are received to determine their applicability to the CGNS TDI diesels and appropriate actions taken, as deemed necessary.

Z3rg39 42

i TABLE 10-1 TDI PRODUCT IMPROVEMENT SIMS SIM NO.

SUBJECT 64 1.

Increase link rod torque - 735 to 1050 ft/lbs 2.

Increase rod bolt torque - 1 1/2 in bolt 1200 to 1700 ft/lbs 1 7/8 in bolt 1800 to 2600 ft/lbs 3.

Product improvement designed to increase reliability 4.

Deletes SIM 270 5.

Use in conjunction with SIM 332 6.

Incorporated during "AF" piston skirt modification in November, 1981 307 1.

Change in ring end gaps on new piston rings in 4 valve R & RV engines 2.

Incorporated 313 1.

Information on removing intake manifold supports on 4 valve RV engines to reduce oil leakage at the camshaft covers 2.

Incorporated 324 1.

Modification of type "AF" piston skirt 2.

Incorporated on Unit 1. Unit 2 "AF" piston skirts have not been modified 3.

The modified "AF" piston skirts have been replaced with the "AE" style piscon skirts on the Unit 1 Division I and II D/Gs 324A 1.

Information for reuse of piston crown studs 2.

Incorporated 332 1.

Never harder washers on connecting rod bolts RV enginec 2.

Incorporated 360 1.

Information on possible problem of air start ~ valve capscrews being too long 2.

Incorporated on Unit 1, Div I and II, tracking document issued for Unit 2 361 1.

Information on potential problem with commercial grade cable in certain engines and panels 2.

Incorporated on Unit 1 Division I'and II engines - Cable replaced with Class 1E qualified cable, tracking document issued for Unit 2 Z3rg40 43

O

11.0 QUALIFICATION / RELIABILITY DEMONSTRATION TESTING 11.1 HISTORY All the GGNS Unit I diesel generators have been tested and qualified in accordance with the requirements of Regulatory Guides 1.9 and 1.108 and IEEE Std. 387-1977.

The Division I and Division II engines were shop tested by TDI, including a 300 prototype test run on the Division I engine as required by IEEE 387-1977.

On-site testing was done by Bechtel and MP&L before fuel loading in June, 1982.

Since then the engines have been tested in accordance with the plant surveillance test procedures, as described in the plant technical specifications.

Augmented testing such as a 7-day performance run was performed on both of the TDI engines under a directive of MP&L management (Reference 10) before the present maintenance and parts replacement work was started in December, 1983.

To verify the operability and reliability of the Division I D/G following the D/G rework af ter the fire, the 18 month functional test was repeated for the Division I D/G.

This additional 18 month func-tional test included the following:

1.

Starting air receiver capacity test 2.

Testing of D/G trips and response to ECCS actuation signals 3.

100% load rejection 4.

Simulated loss of offsite power followed by the loss of and restart of the D/G 5.

Simulated loss of offsite power in conjunction with ECCS actuation signals 6.

24 hour2.777778e-4 days 0.00667 hours 3.968254e-5 weeks 9.132e-6 months load test 7.

LOP /LOCA test This additional 18 month functional test was completed satisfactorily.

The following sections outline the recent tests that were performed on the Division I and II diesels following completion of maintenance work, before each of the two engines was returned to service.

11.2 REQUALIFICATION TESTING REQUIREMENTS Testing requirements for modifications to a previously qualified diesel generator unit are set-forth in IEEE Std. 387-1977. The recent maintenance and parts replacement work on the two TDI diesels had no significant impact on engine specifications and design criteria, related subsystems, or engine performance characteristics.

Nor, did these work activities involve changes in plant load characteristics Z3rg41 44

11.2 (Continued) for the two TDI engines. No modification of the generator or related electric ar instrumentation circuitry was performed. Therefore, none of the design considerations listed in Table-1 of IEEE Std. 387-1977 were modified or altered. As such, the various tasks performed during the current maintenance activities were considered minor design changes as defined by IEEE 387-1977 criteria. Appropriate testing was conducted to verify satisfactory operability of the engines.

11.3 REQUALIFICATION/ DEMONSTRATION TESTING FOLLOWING PISTON SK~RT REPLACEMENT The requalification testing is described in the following section.

11.3.1 To perform TDI's recommended breakin run, following the installation of the "AE" piston skirts, the engines were started and run at 300 rpm and no load for about 15 minutes.

During this run the D/Gs were inspected to ensure that the rocker arms, valves, push rods, fuel injection pumps, nozzle holders, high pressure fuel injection lines and drip return headers were secure, functioning properly and that there were no fuel leaks.

The engines were then stopped, the crankcase side door covers removed and various internal components checked for indication of excessive heat.

The covers were replaced and the engines run at 20% load for about one hour. After this run the engines were inspected as above.

The engines were then run at levels varying between 25% to 100% load for approximately 8 hours9.259259e-5 days 0.00222 hours 1.322751e-5 weeks 3.044e-6 months .

After this run hot crankshaft web deflection checks were per-formed. The engines were then allowed to cool and another inspection as above was performed.

11.3.2 The load rejection tests were accomplished by performing Test #3 of Surveillance Procedure Nos. 06-OP-1P75-R-0003 and 06-0P-1P75-R-0004 " Standby Diesel Generator (SDG) 11 (12) 18 Month Functional Test".

These tests demonstrated the capability to reject a full load (7000 kw) without exceeding speeds or voltages which could cause tripping, mechanical damage, or harmful overstresses.

11.3.3 In addition to the required testing, 24 hour2.777778e-4 days 0.00667 hours 3.968254e-5 weeks 9.132e-6 months run tests were performed; 2 hours2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months at 110% load followed by 22 hours2.546296e-4 days 0.00611 hours 3.637566e-5 weeks 8.371e-6 months at 100%

load. These tests demonstrated the capability of the D/G to carry the rated load for an extended period.

11.3.4 The starting, load acceptance and design load tests were accomplished by performing Surveillance Procedure Nos.

06-0P-1P75-M-0001 and 06-0P-1P75-M-0002, " Standby Diesel Generator (SDG) 11 (12) Functional Test".

These tests demonstrated the ability of the D/G to start and reach rated frequency and voltage within 10 seconds after the start signal, the capability to be loaded to at least 100% load within 60 seconds and to operate for at least one hour at full load.

Z3rg42 45

i 11.4 TESTING NOT REQUIRED FOR REQUALIFICATION The main consideration in developing the requalification test program described in Section 11.3 above was that any engine component or subsystem that was replaced, modified or reworked would be adequately tested, followed by an integrated testing of the total diesel genera-tor system.

Accordingly, an engine component or subsystem that was not af fected by the maintenance activities and was previously quali-fied, was net tested individually or in conjunction with engine test-iag.

11.5 D/G RELIABILITY ENHANCEMENT TESTING MP&L has developed comprehensive maintenance programs and established operating practices to assure a high level of diesel generator reli-ability.

This program was developed using vendor recommendations as well as good engineering practice and operating experience.

This program covers the diesel generator as well as its supportive equip-ment.

Critical diesel generator parameters such as jacket water temperature, lube oil temperature, jacket water standpipe level, generator bearing oil level, turbocharger lube oil flow, starting air pressure, heater operation, and alarm checks are performed once per 8 hours9.259259e-5 days 0.00222 hours 1.322751e-5 weeks 3.044e-6 months ; while other various supportive equipment is checked once per day by the Operations Department.

These checks will assure that the diesel generators are in a satisf actory state, and that potential problems are identified.

During the monthly surveillance test, operating parameters are checked to verify that the diesel generator is operating as required.

The generator operating parameters monitored are voltage, amperes, fre-quency, VARS, DC volts-field, DC Amps-field, RPM and watts.

The engine operating parameters monitored are lube oil temperature and pressure, jacket water temperature and pressure, turbocharger lube oil pressure, lube oil filter differential pressure, fuel oil pressure, fuel oil filter differential pressure, combustion air pressure, crank-case vccuum, RPM cylinder temperatures, and exhaust stack tempera-tures. The monitoring of these parameters aids in detecting any prob-lems which would affect engine operation and reliability.

11.6 ADDITIONAL DEMONSTRATION TESTS Since the discovery of the failed crankshafts at Shoreham, additional testing / monitoring of the D/Gs at Grand Gulf has been implemented (Table '

).

This includes the completion of a 7-day equivalent test run on cuch D/G units, 24 hour2.777778e-4 days 0.00667 hours 3.968254e-5 weeks 9.132e-6 months run tests (22 hours2.546296e-4 days 0.00611 hours 3.637566e-5 weeks 8.371e-6 months @ 100 percent and 2 hours2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months @ 110 percent power),

monitoring of vibration levels by Technology for Energy Corporation (Reference 8), increased emphasis on pre-action planning sessions for persons involved in planned operational and maintenance activities and an improvement in the working relationship with TDI (Reference 10 and 11).

Z3rg43 46

1 11.6 (Continue'd)

Additional testing also includes the completion of a 100 hour0.00116 days 0.0278 hours 1.653439e-4 weeks 3.805e-5 months run on the GGNS Unit 1 D/Gs. The Division 11 D/G performed well during a 100 hour0.00116 days 0.0278 hours 1.653439e-4 weeks 3.805e-5 months 100% load run.

As a precautionary measure, the Division 1 D/G was shutdown 21 hours2.430556e-4 days 0.00583 hours 3.472222e-5 weeks 7.9905e-6 months into the 100 hour0.00116 days 0.0278 hours 1.653439e-4 weeks 3.805e-5 months 100% load run when a maintenance inspection revealed that two bolts were missing from the left bank turbocharger. The bolts were replaced and the 100 hour0.00116 days 0.0278 hours 1.653439e-4 weeks 3.805e-5 months run restarted.

Seventy-two (72) hours into the run, three bolts were again discovered broken or missing from the left bank turbocharger and as a precautionary measure the engine was shutdown.

Following the second incident, an extensive maintenance effort was undertaken to ensure that the left bank turbocharger was aligned correctly.

Following this effort the Division I D/G performed well during another 100 run with 32 hours3.703704e-4 days 0.00889 hours 5.291005e-5 weeks 1.2176e-5 months at 100% load and 68 hours7.87037e-4 days 0.0189 hours 1.124339e-4 weeks 2.5874e-5 months at 75% - load.

The i

loads carried by the D/Gs during these runs are substantially larger than those that could be experienced during emergency engine service.

During a loss of power event the Division I loads are 52% of the D/G full load rating and Division II 68% of the D/G full load rating.

The successful completion of these runs following the correction of 4

the alignment ~of the Division I left bank turbecharger demonstrates that the GGNS TDI D/Gs will reliably carry loads substantially larger than those experienced during emergency engine service.

i t

~Z3rg44 47

~.

I 4

}

{

TABLE 11-1

SUMMARY

OF QUALIFICATION AND VALIDATION TESTING f

ADDITIONAL

. TESTING PRIOR TO INSTALLATION OF "AE"-

DEMONSTRATION I.

PISTON SKIRTS REQ SURVEILLANCES Qualification-Testing (

X 4

Preop Testing ( )

X Tech Spec Testing X

18 Month Functional Test, Division I D/G X

' Day Equivalent Test X

Vibration Test Runs X

"AE" PISTON SKIRT INSTALLATION j

Piston Inspection X

Crankshaft Inspection X

Rod Bearing Inspection X

Cylinder Head Inspection X

TESTING FOLLOWING INSTALLATION OF "AE" I

PISTON SKIRTS-5 Break-In Run X-Load Rejection X-24-Hour Run - 2 Hr.@ 110%, 22 Hr @ 100%

X V

Monthly Surveillance

'X-Additional 100% Power Runs

'X-(Div I.101 Hrs, Div II: 100 Hrs)-

l

^

~

(1)- Qualification Testing includes 300 Start Prototype Tests Performed by TDI.

(2). includes Starting, Load, Acceptance, Overload, Design Load - Rejection, i

Reliability, Electrical'and Subsystem Tests.

- (3)' Includes Monthly Surveillance and 18 Month Punctional Tests.'

Z3rg45' 48

~

w

12.0

SUMMARY

Specific actions have been taken to correct problems identified during testing of the Division I and II TDI diesel generators and to also evaluate and resolve problems identified to MP&L as a result of experience with TDI diesel generators at other nuclear installations.

Significant actions that have been completed, or are planned, are as follows.

The suspect modified type "AF" piston skirts in the Division I o

and 11 D/C have been replaced with new type "AE" pistons.

The new type "AE" piston skirts were inspected prior to installation to assure they were free of the type of rejectable indications found on the type "AF" piston skirts and to establish documented baseline data for the new skirt. The FaAA report (Reference 21) for the TDI D/G Owners Group on "AF" and "AE" piston skirts concluded that, based on both analysis and test results, the type "AE" skirt attachment would not fail in fatigue.

During removal of cylinder heads on the Division 11 D/G the o

stellite overlays on the exhaust valve seats on the #5 right bank cylinder were discovered to have cracks.

There was also incom-plete fusion on the intake valve seat of the #1 left bank head.

Inspection of the Division I heads found six with rejectable in-dications.

The eight heads with rejectabic indications were replaced with heads that had no rejectable indications.

To address a long term concern, a failure investigation has been initiated to determine the cause of the crack initiation and the crack propagation mode, o

As a result of the connecting rod bearing failure identified at Shoreham, MP&L initiated an inspection of the connecting rod bearings and connecting rods during the scheduled piston replace-ment on the Division 11 D/G.

The inspection results indicate that the integrity of the bearings is good and not affected by previous service. A final analysis for chemical and physical properties is planned.

The FaAA report (Reference 22) for the TDI D/G Owners Group conservatively concluded that the 12 and 13 inch bearing shell fatigue life should be approximately 38,000 hours0 days 0 hours 0 weeks 0 months at full load.

This is over 10 times the useage expected over the 40 year service life of the nuclear standby diesel generators.

Numerous weld failures between the D/C connector push rod ball o

and tube have been discoveted.

MP&L concluded that it was not likely that a failed push rod would result in engine failure.

However, during a recent inspection one of the push rod balls was cracked in addition to the weld cracks.

At this point a new replacement design was pursued.

The new design has been determined to be acceptable by MP&L and replacement connecting and main push rods have been installed in the Division I and II diesels.

Z3rg46 49

12.0 (Continued) o Due to the crankshaft failure at Shoreham an engineering evalua-tion of differences in design between the Shoreham and GGNS TDI diesel crankshafts was performed. This evaluation shows that the potential for the type of failure experienced at Shoreham does not exist at GGNS.

During the piston skirt replacement the Division I and II crankshaf ts were inspected.

These inspections did not indicate defects of the type found at Shoreham.

o A fire in the Division I D/G room on September 4, 1983 was determined to be caused by the break of a low pressure fuel oil line.

Analysis indicated that the line break was caused by a combination of unusual vibration loads imposed by a turbocharger that had been replaced several weeks before the fire, and the absence of any supports to isolate the Swagelok connector from vibration loads.

Tubing supports were designed by MP&L and installed on the Division I and II diesel-generators.

o A 10CFR21 notification to the NRC by TDI dated July 20, 1983 identified a possible draw seam on the ID of high pressure fuel oil lines supplied on the Division I and II D/Gs.

A high pressure fuel oil line on the Division I D/G also similarily failed.

An analysis of the failed tubing attributed the failure to a manufacturing flaw (draw seam) in the tubing.

The TDI letter of July 20, 1983 indicated that the failure occurred at approximately one million operating cycles and that fuel lines that have in excess of ten million operating cycles without failure are acceptable.

Using this rationale, all of the original lines on the Division I and 11 D/Gs were considered to be free of flaws of this type, however, the replacement for the failed line and one line on the Division II diesel were not original lines and were considered suspect. Replacement lines were ordered and installed in place of the two suspect lines.

o The generator on the Division II D/G was damaged and replaced in mid-year of 1982 when a head from a capscrew on the rear crank-case cover sheared off and entered the generator via the genera-tor air gap.

Protective screens have been installed on the Division I and 11 generator air gaps to prevent recurrence of damage to the generator from an incident of this type.

Subse-quent testing indicated that the problem of the capscrew shearing was unique to the Division II D/G and that the failure was due to low stress high cycle fatigue, however, test results were incon-clusive as to the root causes of the vibration sources.

High strength capscrews and tab washers were installed to extend the life of the capscrews.

Periodically a capscrew will be removed from the crankcase covers and subjected to destructive analysis in an attempt to obtain further information for identifying the root cause.

Z3rg47 50

12.0 (Continued)

Following piston skirt replacement, qualification / reliability o

testing in accordance with IEEE Std. 387-1977 was performed on the Division I and II diesel generators.

Testing of the D/Gs prior to this maintenance included the satisfactory completion of a 7-day equivalent test run on both D/Gs.

Post maintenance testing included breakin runs, twenty-four hour runs, load rejection tests and surveillance tests.

Additional post maintenance demonstration testing has resulted in approximately 158 hours0.00183 days 0.0439 hours 2.612434e-4 weeks 6.0119e-5 months at 100% load on the Division I diesel generator and 131 hours0.00152 days 0.0364 hours 2.166005e-4 weeks 4.98455e-5 months at 100% load on the Division 11 diesel generator.

1 i

t Z3rg48 51

o.

O

13.0 CONCLUSION

In conclusion, the specific corrective actions, engineering evaluations and testing that have been completed, enhance the reliability of the D/Gs and provide assurance, with a reasonable level of confidence, that the GGNS TDI engines will adequately perform their required safety function.

I

)

i i

i I

F d

Z3rg49

$2

14.0 REFERENCES

1.

FaAA Preliminacy Report on GCNS Modified AF Piston Skirts.

2.

TDI Letter Dated 12-15-83 " Nuclear Power Plant Standby Diesel Generator User's Group Minutes of November 30, 1983, Meeting".

3.

Metallurgical Evaluation of Diesel Engine Push Rod Weld From Grand Gulf Nuclear Station-Unit 1 Emergency Diesel Generator (Division 1),

prepared by Middle South Services.

4.

TDI Response to MP&L for NRC Request of Additional Information on TDI D/Gs, Dated November 2, 1984.

5.

Preliminary Standby Diesel Generator Crankshaft Design Analysis Review Grand Gulf Nuc1 car Station, prepared by Bechtel Power Corporation.

6.

Metallurgical Evaluation of Diesel Engine Fuel Oil Line Failure from Emergency Diesel Generator - Division 1, Grand Gulf Nuclear Station -

Unit 1, prepared by Middle South Services.

7.

Metallurgical Evaluation of Diesel Engine Fuel Injection Tube f rom Unit 1 Emergency Diesel Generator Grand Gulf Nuclear Station Prepared by Middle South Services.

8.

Test Evaluation Report on the Grand Gulf Nuclear Station Division I and Division II Diesel Generators (TEC Report No. R-83-033), prepared by Technology for Energy Corporation.

9.

Engineering Investigation of the Failure of Rear Crankcase Cover Capscrews for the Delaval Standby Diesel Generators at MP&L, GGNS, LETCO Job No. G-8847, Dated August 17, 1982, by Law Engineering Testing Company.

10.

PMI 83/12569 J. P. McGaughy to J. B. Richard Letter on D/G Enhancement.

11.

PMI 84/0210, J. E. Cross to J. F. Pinto Letter on Plant Staff Response to NRC D/G Questions.

12.

AECM-83/0689 - GGNS Dierel Generator Reliability Report, October 26, 1983.

13.

AECM-83/0724, CGNS Diesel Generator - NRC Request for Additional Information, November 15, 1983.

14.

AECM-84/0030, GGNS Diesel Generator - NRC Request for Additional Information, January 18, 1984.

15.

FaAA-83-10-2 PA07396 - Emergency D/G Crankshaft Total Stress Analysis Summary, February 2, 1984.

'Z3r 50 53

r l

J 4

i t

14.0 (Continued) 16.

Bechtel Standby Diesel Generator Crankshaft Total Stress Analysis Summary, February 2,1984.

I i'

17.

FaAA-83-10-02 PA07396, Analysis of the Replacement Crankshafts for l

i Emergency Diesel Generators, Shoreham Nuclear Power Station, October 31, 1983.

18.

FaAA-83-10-16, PA07396, Emergency Diesel Generator Connecting Rod l

Bearing Failure Investigation Shoreham Nuclear Power Station, October 31, 1983.

19.

AECM-83/0653, Applicability of Shoreham Diesel Generator Crankshaft j

Failure to GGNS, October 14, 1983.

[

i 20.

AECM-84/0103, CGNS Standby Diesel Generator, Comprehensive Reliability i

Report and Status, February 20, 1983.

i 21.

FaAA-84-2-14 FME-R-6/7389, Investigation of Type AF and AE Piston l

Skirts, February 27, 1983.

I 22.

FaAA-84-3-1, PA0 7389/LAS-M&T-3A, Design Review of Connecting Rod Bearing Shells For Transamerica DeLaval Enterprise Engines March i

12, 1984.

>A I

d I

I P

i i

23rg51 54

ATTACHMENT 1 TO THE UPDATED REPORT ON GGNS DIVISION I AND II TDI DIESEL GENERATORS RESPONSES TO SIXTEEN POTENTIALLY SIGNIFICANT PROBLEMS IDENTIFIED IN TDI OWNERS GROUP MEETING WITH THE NRC ON JANUARY 26, 1984 April 1984 Z3rg52 1

1.0 INTRODUCTION

A meeting of the Transamerica Delaval, Inc. (TDI) diesel generator (D/G) owners group with the NRC Staff was held on January 26, 1984.

During the meeting the owners group presented a slide summarizing significant potential problems with TDI diesels.

These potential problem areas are detailed below:

o Crankshaft o

Connecting Rod Bearings o

Pistons o

Cylinder Heads o

Cylinder Liners o

Cylinder Block o

Engine Base o

Head Studs o

Push Rods o

Rocker Arm Capscrews o

Connecting Rods o

Electrical Cable o

Fuel Injection Lines o

Turbocharger o

Jacket Water Pumps o

Air Start Valve Capscrews Further details of these concerns, their applicability to Grand Gulf, and their resolution are described in the following sections.

23rg53 56

2.0 CRANKSHAFT A summary of the concern and its resolution on Grand Gulf is provided in Section 6.0 of the Final Report.

3.0 CONNECTING ROD BEARINGS A summary of the concern and its resolution on Grand Gulf is provided in Section 4.0 of the Final Report.

4.0 PISTONS A summary of the concern and its resolution on Grand Gulf is provided in Section 2.0 of the Final Report.

5.0 CYLINDER HEADS A summary of the concern and its resolution on Grand Gulf is provided in Section 3.0 of the Final Report.

6.0 CYLINDER LINERS

6.1 DESCRIPTION

A concern has been raised regarding cylinder liner damage in TDI D/Gs.

One incident was listed for GGNS, AECM-82/157, dated April 15, 1982, which transmitted the final report on PRD-81/45 dealing with the separation of piston crown from the piston skirt during testing of the Division II D/G.

An additional deficiency noted in this report was damage to a cylinder liner on the Division I D/G.

The damaged i

cylinder liner was discovered during disassembly of the Division I D/G for corrective action for the piston skirt / crown separation.

The damaged Division I cylinder liner was found to be grooved in three places. These grooves were approximately 10 inches long and 1/16 inch deep. As indicated in the PRD final report, the grooving was probably caused by debris that entered the cylinder during assembly or initial startup.

6.2 ENGINEERING EVALUATION AND CORRECTIVE ACTION The grooved cylinder liner was replaced with a new liner.

The Division I lube oil was flushed and replaced and the lube oil sump was cleaned.

At a meeting between MP&L, LILCO and TDI on February 2, 1984, TDI indicated that the only case of a cylinder liner failure occurring without some other initiating event causing it, occurred on the ship Z3rg54 57

6.2 (Continued)

Columbia. This damage was attributed to the high vanadium content of the light-heavy fuel oil and the high ash content of the lube oil (heavy oil).

Despite the cracking of the liner which resulted from the use of these oils, the engine continued to perform its function.

The GGNS TDI diesels use light fuel oil with a lower vanadium content and utilize light lube oil.

During the recent piston skirt changeout on the GGNS Unit i TDI engines, the cylinder liners were subjected to a close visual inspec-tion before and after honing the liners to receive the new rings. No obvious damage was discovered during these inspections.

6.3 CONCLUSION

S Based on lube oil cleanup efforts, recurrences of the subject problem is considered to be resolved.- To date, no known cylinder liner fail-ure has been the root cause of a TDI engine failure.

Neither liner material, manufacturing process nor design are con-sidered to be the root cause of the damage on the Division I GGNS engine.

Inspections of the Division I and II D/G cylinder liners during the recent piston skirt changeout in December, 1983 did not reveal any indication of liner damage.

Based on the above conclusions and root cause, cylinder liner failure is not expected to occur at GGNS.

7.0 CYLINDER BLOCK

7.1 DESCRIPTION

The non-nuclear industry has reported cracks occurring in the area around the cylinder liner landing. Cracks may also propagate from the head stud / stud bore to the jacket cooling water passage.

MP&L has also been recently advised of the discovery of cracks in the area of engine head studs on'a cylinder bicek used in a nuclear application.

7.2 ENGINEERING EVALUATION If this cracking were -to occur and propagate into the jacket water passage it would be possible for an extremely low flow of jacket cooling water to come into contact with the head studs and cylinder head. This flow would be prevented from entering the cylinder by two spiral wound head gaskets.

It is unlikely that jacket cooling water would enter the firing chamber (cylinder) and only a very slow loss of jacket cooling water to the outside of the engine would be evident.

To prevent this cracking, TDI has indicated that proper torque must be placed on the cylinder head studs.

Z3rg55 58

~

- =_

7.3 CORRECTIVE ACTION AND CONCLUSIONS MP&L considers that no corrective action is required for these condi-tions, since the postulated condition would not interfere with the operation of the engine and because the proper torque of 3600 foot-pounds, as recommended by TDI, has been applied to the head studs of the GGNS Unit i TDI D/Gs.

Even if cracks were to occur they would be expected to propagate very slowly because of the large mass of metal in the cylinder block.

However, MP&L will continue to closely follow the investigation findings of the TDI D/G owners group.

l 3

d J

l I

i Z3rg56 59

8.0 ENGINE BASE

8.1 DESCRIPTION

Linear indications have been found on the bearing base journal of several marine diesel engines.

These indications were apparently caused by improper torquing of the bearing holddown studs during assembly of the engine.

8.2 CORRECTIVE ACTION TDI issued SIM #286 to correct this problem.

This fix resulted in an increased preload being placed on the holddown studs.

8.3 CONCLUSION

S Grand Gulf's TDI D/Gs were assembled after SIM #286 was issued.

GCNS installation of main bearing bolt nuts, as witnessed by GCNS Plant Quality, indicate that correct preload values were verified during recent engine disassembly at the site on all main bearing studs. This problem, therefore, is not expected to occur at Grand Gulf since no defects have been reported to have occurred in engines using the proper torque.

9.0 HEAD STUDS This concern is related to the cylinder block concern described in Section 7.0 of this Attachment. Refer to this section for further details.

10.0 PUSH RODS A summary of the concern and its resolution on Grand Gulf is provided in Section 5.0 of the Final Re ort.

11.0 ROCKER ARM CAPSCREWS

11.1 DESCRIPTION

Shoreham has experienced problems recently with fatigue failure of a rocker arm capscrew.

11.2 ENGINEERING EVALUATION AND CORRECTIVE ACTION The failure at Shoreham was apparently caused by undertorqued poor quality capscrews.

New capscrews made of ASTM A-193 material were installed and torqued to specified torque values to correct the problem at Shoreham.

Z3rg57 60 4

11.3 CONCLUSION

S GGNS rocker arm capscrpws have not experienced this type of failure after greater than 10 cycles of operation.

The " Emergency Diesel Generator Rocker Arm Capscrew Stress Analysis" report, dated March, 1984 prepared for the TDI Owners Group by Stone and Webster Engineering Corporation, concluded that both the original and modified rocker arm capscrews are adequately designed for the given service conditions. The CGNS rocker arm capscrews are original components of the diesels and have been properly torqued to 365 ft-lbs.

MP&L considers this issue resolved with no further action required.

A

?

a i

f 1

1 l

Z3rg58

.61 i

12.0 CONNECTING RODS

12.1 DESCRIPTION

TDI has informed MP&L of several incidences of connecting rod failure.

At a meeting between MP&L, LILCO and TDI, on February 2, 1984, TDI defined the historical problem with connecting rods.

Cracking of the connecting rod link assembly in a master rod-longitudinal plane through the bottom of upper bolt holes (See Figure A12-1) has occurred on several non-nuclear applied diesel engines built by TDI.

12.2 ENGINEERING EVALUATION AND CORRECTIVE ACTION TDI Vee-type engines of a comparable size to GCNS Division I and II utilize either 1-1/2 or 1-7/8 inch connecting rod bolts. The original design of the GGNS engines (TDI's earlier design) uses the larger of the two bolt sizes. TDI originally specified that these bolts should be torqued to 1800 foot-pounds.

TDI initiated an evaluation of the problem based on the operating history of the engines with failed or cracked connecting rods.

For example, several instances of connecting rod cracking were reported to have occurred in a marine diesel on the ship Columbia.

The average hours of operation between occurrence was approximately 10,000 hours0 days 0 hours 0 weeks 0 months .

Evidence of fretting in the " rack-teeth" almost always accompanied connecting rod failure or cracking.

The first design change to remedy the situation was a decrease in the connecting rod bolt diameter to 1-1/2 inches. Decreasing the connect-ing rod bolt diameter effectively increased the amount of base metal where cracking was occurring.

Since the cause of the cracking was thought to be relative motion between the rod parts and flexure of connecting rod parts, an increase in the base metal adjacent to the crack initiation site should increase stiffness and hence decrease incidence of cracking or failure.

A decrease in cracking frequency was noted.

However, connecting rods using both 1-1/2 and 1-7/8 inch bolts were still reported exhibiting cracking.

It was then thought that fretting of the " rack-teeth" was due to lack of clamping force between the connecting rod link and the master rod and box assembly.

TDI issued Service Information Memo (SIM) 64 to rectify the suspected clamping force problem.

SIM 64 effectively increases the required torque on 1-1/2 and 1-7/8 inch connecting rod bolts from 1200 to 1700 foot-pounds and from 1800 to 2600 foot-pounds, respectively.

This design change greatly reduced the reported cases of connecting rod cracking.

The GGNS Division I and 11 engines were originally assembled at the vendor's shop using the pre-SIM 64 torque values. Therefore, the GGNS engines have been run part of the present total sum times with the connecting rods torqued to pre-SIM 64 torque values and the balance at post SIM 64 torque values. The table below indicates the approximate run times on the Division I and 11 engines before and af ter SIM 64 was implemented:

Z3rg59.

62

12.2 (Continued)

Division I Division II At Assembly 0

0 Before SIM 64 332 44 After SIM 64 1065 826 Present Run Times 1397 870 At a recent TDI D/G owners group component selection committee meeting, the owners group diesel generator specialists agreed that the type of cracks reported by TDI would propagate very slowly.

The cracking of connecting rod parts on non-nuclear diesel engines were reported to have occurred at relatively large run times (greater than 10,000 hours0 days 0 hours 0 weeks 0 months ).

12.3 CONCLUSION

S To date, all engines using the 1-7/8 inch connecting rod bolts exhi-biting failures or cracking have been suspected of being under-torqued.

Further, no known failures have occurred on connecting rods using 1-7/8 inch bolts that were properly torqued.

All torques used on the subject bolts at GCNS have been verified to be in accordance with SIM 64.

Based on low probable propagation rate of incipient cracks, relatively low run hours on Division I and 11 at pre-SIM 64 torques, and the expected low future run times, (estimated 200 hours0.00231 days 0.0556 hours 3.306878e-4 weeks 7.61e-5 months / year) deleterious cracking of the CGNS connecting rods is not expected.

J t

Z3rg60 63

FIGURE A12-1:

Articulated Connecting Rod Assembly i

a l

D 84pM i

a 5

l

[f~

I w)),y.

. ese i'

?>

G l7

?b'\\

0 r.

\\

f A

j ri s'5 %

o';

t

\\

EMP8'"

w'/

(s

' ',~,

e:

~

t ro g

o*

p l

O s

\\

.uan.

l*

l

.r,Ni

'M

?

,/

F.,

j h,

W'

-..,,):

~

O

s 6

{

aam'ei

[*

myg)l

'\\"R'Jll#"""

7 Fig. A12-1

13.0 ELECTRICAL CABLE

13.1 DESCRIPTION

Memo (SIM) No. 361 concerning certain Class 1E cable which failed the IEEE 383-1974 insulation flame test was issued by TDI. The content of this SIM is detailed in Table 10-1.

This SIM identified the affected cable as being the shielded cable from the terminal block to the Air-pax tachometer relay in the engine control panel, the shielded cable from the Airpax magnetic pickups to the junction boxes on the side of the engine and the multiconductor cable from the engine side mounted junction box to the Woodward governor actuator.

Another notification from Delaval received by MP&L on October 20, 1983 (API-83/0974), idicated that the manufacturer's temperature rating for the cable insulation may be exceeded during operation of the diesel generator. Delaval recommended that these cables be replaced with 90*

rated cable.

13.2 ENGINEERING EVALUATION AND CORRECTIVE ACTION It was determined by Nuclear Plant Engineering that this potential deficiency could create a substantial safety hazard.

A Design Change Package (DCP-82/3196) was implemented for Unit 1 in which the in-stalled commercial grade shielded cable on the Division I and II D/Gs was replaced with Class 1E IEEE 383-1974 qualified cable.

Further investigations into the problem subsequently revealed that Bechtel Design Specification M-018.0, Section 6.8.2.6, calls for com-pliance with Design Specification Appendix N which requires compliance with IPCEA Publication No. S-19-81, Section 6.

In responding to API-83/0974 it was determined that the af fected cable had previously been replaced on the Unit 1, Division I and II D/Gs.

Therefore, no further action was initiated for Unit 1.

Bechtel has issued NCR 6762 to track this concern for the Unit 2 D/Cs.

13.3 CONCLUSION

S This issue is considered closed for the Unit 1 D/Gs. The replacecent electrical cable meets the appropriate requirements of IEEE 383-1974 and TDI's recommended temperature rating.

14.0 FUEL INJECTION LINES A summary of the concern and its resolution on Grand Gulf is provided in Section 8.0 of the Final Report.

Z3rg61 65

15.0 TURBOCHARGER

15.1 DESCRIPTION

Original problems with supports and components adj acent to the.

Division I D/G left bank turbocharger have been attributed to a turbocharger that had exhibited signs of unusual vibration.

This turbocharger was replaced with a spare turbocharger in August of 1983.

Both Division I D/G turbochargers were replaced following the D/G fire in September of 1983 because they were located in the fire area and their ability to carry out their design function was in question.

Recurring problems with alignment on the left bank turbocharger were experienced following the replacement in August of 1983.

This was corrected in February of 1984 when an extensive maintenance effort was undertaken to correct the alignment problem.

Following this effort the Division I D/G completed a 100 hour0.00116 days 0.0278 hours 1.653439e-4 weeks 3.805e-5 months run (32 hours3.703704e-4 days 0.00889 hours 5.291005e-5 weeks 1.2176e-5 months @ 100%, 68 hours7.87037e-4 days 0.0189 hours 1.124339e-4 weeks 2.5874e-5 months

@ 75%) with no further problems.

Turbocharger abnormalities resulted in broken mounting bolts.

When the turbocharger is not anchored correctly, the intercooler and the jacket water piping are forced to partially support the turbocharger and thereby absorb a larger amount of fatigue stress.

This stress would normally be absorbed by the turbocharger mount.

Since neither of these two auxiliaries was designed to support the turbocharger, they both developed cracks and broken welds.

Table 15-1 presents a summary of past problems, causes, and corrective action taken. Further details are provided below:

15.1.1 CRACKED WELDS AND BASE METAL ON INTERCOOLERS Cracks developed in the base metal on the top of the inter-cooler along an extruded seam.

This seam has air.ce been redesigned by Delaval and a piece of flat bar stock welded over the top of the extruded vee shape to stiffen it.

The stay rods extend from one side of the intercooler to the other through a heavier block of steel on the outside. The rod is then welded-to this heavier block, thio is the weld which broke on the right bank intercooler.

Several other-stay rods were observed to have deficient welds and were also cut out and revelded.

15.1.2 CRACKED WELDS ON JACKET WATER PIPING There were several cracked welds which developed on flanges and fittings where the jacket water system ties into the turbochargers. Since more than one repair was necessary the header was refabricated using standard pipe, fittings, and ASME Section III Welding & NDE Criteria in order to work with codes with which MP&L maintenance and engineering personnel were acquainted.

Z3rg62 66

1^

i 15.1.3 LOW PRESSURE FUEL OIL HEADER FAILURE l

[

On September 4,.1983, the main fuel oil. line feeding the Division I engine headers failed due to fatigue.

The oil sprayed onto the turbocharger exhaust gas header transfor-mation piece and ignited.

All affected components were repaired or replaced. The failed tube and Swagelok fitting were subjected to a metallurgical evaluation, and the cause i

of the failure was identified as high cycle fatigue com-pounded by the absence of tubing supports.

Further dis-j' cussion is provided in Section 7.0 of this finsi report.

15.1.4 TURBOCHARGER MOUNTING BOLT FAILURES There haveJbeen several instances of turbocharger mounting bolt. failures on the GGNS Division I D/G left. bank turbo-charger.

15.1.5 INDUSTRY EXPERIENCE Turbocharger problems at other nuclear plants have also been experienced. Recently, Shoreham has experienced a i

failure of turbocharger thrust bearing in two of their engines.-

3 15.2 ENGINEERING EVALUATION AND CORRECTIVE ACTION The turbochargers on the Division I D/G left bank have exhibited signs of unusual vibration and misalignment in~.the past'.. Improper.

turbocharger alignment and running of the engine ~ with broken / missing turbocharger mounting bolts, has produced conditions conducive. to.

j fatigue crack initiation and propagation in adjacent - supports and, components.

During the rework I of the ' Division I engine after the fire, the replacement turbochargers were removed and re-seated - twice before proper fitup was considered attained.- The result was an engine that.

had no noticeable areas.' of high vibrations, as - attested to by Technology for Energy Corporation' when_ they instrumented the. Division

~

I and Division Il engines.after the fire rework was completed.-

-f Thrust bearing failures similar to.those at Shoreham have not been'

^

identified at GGNS and'are not expected because of;the differences in;

^

design of the lubrication systems. : The failures ~ of the" turbocharger -

-thrust bearings-at Shoreham have been attributed to probable. lack of lubrication during' manual engine. starting.

-Shoreham's TDI D/GsJ1ube' oil; systems utilize two. pumps, one an engine driven pump ;and the other an electric driven heater pump -(See Figure -

15-1). ' The GGNS D/Gs olobe oil' system utilizes three pumps,Eone.an lengine driven pump. -one -an electric driven heater pump and ' the other an electric - driven ' auxiliary pump J (See 1 Figure ; 15-2).

[

Z3rg63-67l

~

v.

~

w

i j

15.2 (Continued)

Presently, on a manual start Shoreham does not have the means of i-supplying oil to the turbocharger thrust bearing other than through a turbocharger lube oil drip system.

The turbocharger lube oil drip system is also used at GGNS and is essentially the same as Shorehams.

However, prior to a manual start of the D/Gs at GGNS the engine is prelubed for two minutes or less with the auxiliary lube oil pump which pressurizes the turbocharger thrust bearing with lube oil. This precludes the type of failures reported at Shoreham.

- dditional occasions when the Recently, there have been several a

Division I D/G 1 eft bank turbocharger mounting bolts. failed. The main reason for these failures was again considered misalignment of the turbocharger with -its associated piping. and components.

An engineering evaluation of - the turbocharger mounting arrangement is.

being performed and procedures designed to preclude misalignment have i

been implemented.

The alignment problems with the turbochargers are partially a function of ~ the custom fit arrangement of the turbocharger with the flanged connections on the mating piping and apparatus.

Each engine j

supplied by TDI has slightly different piping due to-the fact that it is hand built. The specific problems encountered with the lef t bank turbocharger on Div. I was the. misalignment of the turbocharger and intercooler mating flanges. -These flanges were ' misaligned such that if you tightened down the flanged connection, the_ turbocharger was cocked on the mounting pedestal approximately 30 mils. (Figure 15-3.)

i This problem was attacked several times before. a satisfactory resolution was achieved. TD1 recommended the following:

1)

Intercooler flange be cut.

2)

The turbocharger bolted down 3)

The intercooler flange be bolted to the turbocharger 4)

,The intercooler flange tack welded to the intercooler t

5)

The turbocharger removed; 6)~

The intercooler flange rewelded j

7)

The turbocharger remounted

~

8)

The turbocharger bolted down

9) 'The intercoolercflange bolted up-t 10)L The turbocharger. mounting bolts removed-

11) The mounting plate checked.for clearances

~~

i

12) _ The' turbocharger remounted. ~

l Z3rg64' 68-

]

R

,o yr

,g.'

ym

-+e y

9

,- p -

15.2 (Continued)

13) Perform a maintenance run

14) The turbocharger mounting bolts removed 15)

The mounting plate rechecked for clearances

16) The mounting bolts installed At the end of this process the engine was tested for 100 hours0.00116 days 0.0278 hours 1.653439e-4 weeks 3.805e-5 months and the mounting clearances rechecked.

The final conclusion was that the problem had effectively been solved.

The final clearances were 0, 0, 0, and 3 mils. op each of the four (4) corners of the mounting plate.

These values are acceptable and the successful completion of the 100 hour0.00116 days 0.0278 hours 1.653439e-4 weeks 3.805e-5 months run demonstrated that the problem has been solved.

15.3 CONCLUSION

S The susceptible areas in the piping and components around the turbo-chargers have been identified by past failures. The integrity of these areas has been enhanced by the use of approved ASME code

welding, procedures, and materials during rework.

Since these enhancements, the weld and component failures have not raoccurred.

The original problems with the left bank Division I turbocharger were due to a turbocharger that exhibited signs of abnormal vibration.

Since this turbocharger was replaced there have been alignment problems with the left bank turbocharger resulting in broken mounting bolts.

In February of 1984, the turbocharger was removed and carefully refitted with all attached piping and mating surfaces and correctly installed.

A 100 hour0.00116 days 0.0278 hours 1.653439e-4 weeks 3.805e-5 months performance run was performed af ter which the turbocharger was inspected and no problems were discovered.

There have been no further problems experienced since this was done.

MP&L feels confident that this problem has been successfully corrected.

As a part of the effort by the TDI Diesel Generator Owner's Group, MP&L will fund a thorough study of turbocharger mounting arrangement and will take any corrective actions deemed necessary.

Z3rg65 69

m

_ _ _ ~..

n TABLE 15-1 e

ENGINE MOUNTED COMPONENTS PROBLEMS ATTRIBUTED TO TURBOCHARGER VIBRATION d

ITEM DESCRIPTION OF PROBLEM-CAUSE CORRECTIVE ACTION-11 Cracked welds and base metal Fatigue compounded by high Repaired welds and base metal cracks on intercoolers.

vibration from turbocharger, cracks. Reseated turbocharger to eliminate undue stresses caused by misalignment.

2' Cracked welds on jacket water Fatigue compounded by high Repaired welds. Refabricated flanges.and piping headers.

vibration from turbocharger, header to ASME III Class 3 reseated turbocharger.

3

-Low pressure fuel oil header Fatigue compounded by high Replaced fuel line and fittings failure.resulting in Div I vibration from turbocharger.

reseated turbocharger.

fire.

~

15 W

~

.Z3rg66

FIGURE 15-1:

Illustration of Shoreham TDI Diesel Generator Lube 011 System (Reference Telecon Shoreham)

Filter Y

Orifice 1

4 g,

Sight Glass e

Drip Lubrication j

Supply Turbo a

Pressurized Oil Supply __

Prelube Flow Shoreham

~

f 1

Main L.O. Header n

Z Z

z z

a Inlet Inlet Str Str t

Full a

Flow Fltr

.L.

TCV d

(

Cooler

)

~*

be Filter a

CO CO Engine L.O.

Driven Heater

' Pump Pump Fig. 15-1

-.>....u=-..-~..

.~..:.

-=:...

l 1

l FIGURE 15-2:

Illustration of GGNS TDI Diesel l

Generator Lube Oil System Filter W

1 1

e Orifice % _

18 Drip Lubrication Supply Turbo s Turbo Pressurized Oil Supply i

t i

h Prelube Flow p

GGNS -

(

Main Main

(

L.O.

Hdr.

4 L.O.

A fEdr.

f2 Z

+

e Z

hZ n

Inlet Inlet Str.

Str.

-e-

-s-h Full Flow Fltr.

~

t

/.

Km Cooler

)

Pre Lube Filter i..

CO CO CO Engine Electric L.O.

Driven Aux Pump Heater Pump Pump Fig; 15-2

i

....-~..--...u.-

2 w.u.,a a _.. a >

a

.,1 e.....z.

1.a a.-

i

/

o a

o O

O O

N/

M Condition of Turbocharger-s,'-

fj' a

N Intercooler Flanges with Turbocharger Base Bolted

[

," {

Down

/ /j

\\

O O O v

I l

g,-

I c

r i

i m

N.030 O

O O

m

..g.

L

)

~

Condition of Turbocharger Base When Turbocharger-Intercooler Flanges Bolted Together e

LEFT BANK TURBOCHARGER MOUNTING ARRANGDIENT FIGURE 15-3

16.0 JACKET WATER PUMPS

16.1 DESCRIPTION

Shoreham has experienced a jacket water pump shaft failure.

16.2 ENGINEERING EVALUATION This problem apparently affects only the in-line engines. No jacket water pump shaf t failures have been reported on Vee-type engines to date.

16.3 CONCLUSION

S As of this time, MP&L has been unable to obtain evidence of generic jacket water pump shaf t failures on DSRV-16-4 engines.

Since this problem appears to be unique to in-line engines, it-has not and is not expected to occur on the GGNS diesel engines.

17.0 AIR START VALVE CAPSCREWS

17.1 DESCRIPTION

On May 13, 1982, TDI reported a potential defect concerning the-cap-screws that are used to retain the air start valves in the cylinders heads to the.NRC under the provisions of 10CFR21.. The 3/4-10x3 inch long capscrews were' suspected of bottoming out in the tapped holes in the cylinder heads. JThis c'ould result in insufficient or unequal clamping forces being applied to the air start valve.

17.2 ENGINEERING. EVALUATION AND CORRECTIVE ACTIONS TDI recommends replacement with 2 3/4 inch long.capscrews or machin-ing 1/4 inch off the existing 3 inch capscrews. A design change was issued by MP&L to implement corrective actions. The air start valve capscrews on the Unit 1, Division I and II D/Gs were modified _by-machining 1/4 inch off the length.

17.3 CONCLUSION

S Corrective action is considered complete in regards torthe air start-valve capscrew problem on the Unit 1 D/Gs.

Z3rg67-74

ATTACHMENT 2 TO THE UPDATED REPORT ON GGNS

. DIVISION I AND II TDI DIESEL GENERATORS i

h 4

1 4 _

I i

. PISTON MANUFACTURING DETAILS 1

i J

i t

1 1

t

)

APRIL, 1984 i

i i

Z3rg68.

4 9

+1-?

= 34

v4 m rtee sF=ma

td" W

T T7 9

'i-9

)

b k

't

?*p--

f g f

e ATTACHMENT 2 PISTON MANUFACTURING DETAILS 1.0 GENERAL As reported by TDI (Reference 2), all 450 RPM rated " Enterprise" R-4 series engines have been furnished with two-piece pistons which incorporate a cast steel piston crown attached to a cast modular iron piston skirt by means of four studs. This piston design has evolved since its inception in 1969 to incorporate design improvements for high reliability and less costly manu-facture. As horsepower ratings of engines increased in the mid-1960's, Transamerica Delaval and other medium speed diesel engine manufacturers abandoned the older style single piece piston design. The two piece piston is inherently better equipped to deal with the higher thermal inputs of high Brake Mean Effective Pressure-(BMEP) engines, because it allows thermal growth of the crown without causing excessive bending stresses in the skirt. The two piece piston design is also better equipped to handle the higher pressure and inertia loadings of the increased horsepower engines. The modular iron skirt has passed through several design changes.

Five different designs have been.used and are identified by TDI terminol-ogy as "AF", modified "AF", "AN Old Style", "AN New Style", and "AE".

2.0 GGNS BACKGROUND Only the "AF", modified "AF" and "AE" piston skirts have been used at GGNS.

The "AF" piston skirts were originally supplied by TDI on the Division I and_II GGNS engines. When problems were encountered with material quality of the washers (piston crown / skirt bolt) and CGNS experienced a piston crown / skirt separation, MP&L responded by upgrading hardware in accordance with SIM 324 to the modified "AF" piston skirt decign. The cracking dis-covered later on Shoreham modified "AF" piston skirts and rejectable indi-cations found at inspection prompted MP&L to change out all piston skirts at GGNS to the latest "AE" design piston skirts.

3.0 PISTON TYPES 3.1 "AF" AND MODIFIED "AF" PISTON SKIRTS "AF" piston skirts use spherical washers on the four studs which attach the crown to the skirt. These spherical washers provide l

fastener flexibility. These commercially supplied' washers proved to have inconsistent quality and large variatione in heat treatment and manufacturing tolerances. As a result,.a small number of-the washers failed in service, resulting in piston, skirt / crown separation. One such separation occurred on the Division II D/G during field-testing.

To solve the spherical washer problem, the design was modified to incorporate a " full stack" Belleville washer arrangement resulting in a modified "AF" piston skirt.

Z3rg69 76

e N:

3.1 (Continued)

The "AF" style piston skirt casting received.-the following heat treatment:

o Heat to 1750 degrees F. (near the upper critical tempera-ture) for 3 hours3.472222e-5 days 8.333333e-4 hours 4.960317e-6 weeks 1.1415e-6 months . Normalize (air cooled) in still air.

This results in a pearlitic structure with 100,000 psi tensile strength.

Re-heat to 1050 degrees (slightly below the lower critical o

temperature).for 3 hours3.472222e-5 days 8.333333e-4 hours 4.960317e-6 weeks 1.1415e-6 months and cool in still air. This tempering process produces the desired ductility in the nodular iron.

3.2 "AE" Pistons The "AE" piston, the latest R-4 piston skirt design, just concluding research development testing, incorporates the field experience on the l

R-4 series engine and the R-5 series engine.

The '.'AE" design utilizes a " half stack" Belleville washer arrangement.

All "AE" skirts are heat treated to produce strecs relieved 100,000 psi tensile strength nodular iron. All piston skirts in the TDI units at GGNS Unit I have been replaced with this design. The "AE" style

[

skirt is interchangeable with existing R-4 piston crowns and requires i

only minor hardware changes.

i 4

d i

i i

)

4 1

- Z3rg70

77 tr e

T

l

,:7..- - -....

.. - -. ~

. _ =

FIGURE A2-1:

PISTON COMPARISONS N

xys\\

s t

3.4

/

f

///

\\

E

'\\

</ss/

c.

0 - 0 4.= A F smanteAL wetMGM

~

sin.4 4

"AF" SKIRT J:.*

..m-

., p.

.u7.-+..~,,,,,,,,...

p.

xw rw.=

M v

q

/>

FULL $7ACK

$XlRT BELL VILLS STYLE MODIFIED "AF" SKIRT i

r.4

2. 4 d

4

.N NNN1 y

3,. r., =e

.>:a...,

.4 r

(

,,\\

I i

~

c_

sxw f-

'J l

>=

b

\\

4

z

nex sxre r setu'u n srvem_

os.go ca.,,

ngga ggggy Fig. A2-1

---

ML20084A175

Text

1.0 INTRODUCTION

SUMMARY

13.0 CONCLUSION

14.0 REFERENCES

SUMMARY

1.0 INTRODUCTION

2.1 DESCRIPTION

2.5 CONCLUSION

3.1 DESCRIPTION

3.4 CONCLUSION

4.1 DESCRIPTION

4.5 CONCLUSION

5.1 DESCRIPTION

6.1 DESCRIPTION

6.3 CONCLUSION

7.1 DESCRIPTION

7.4 CONCLUSION

8.1 DESCRIPTION

8.3 CONCLUSION

9.1 DESCRIPTION

9.4 CONCLUSION

SUMMARY

SUMMARY

13.0 CONCLUSION

14.0 REFERENCES

1.0 INTRODUCTION

6.1 DESCRIPTION

6.3 CONCLUSION

7.1 DESCRIPTION

8.1 DESCRIPTION

8.3 CONCLUSION

11.1 DESCRIPTION

11.3 CONCLUSION

12.1 DESCRIPTION

12.3 CONCLUSION

13.1 DESCRIPTION

13.3 CONCLUSION

15.1 DESCRIPTION

15.3 CONCLUSION

16.1 DESCRIPTION

16.3 CONCLUSION

17.1 DESCRIPTION

17.3 CONCLUSION

Navigation menu

Search