ML20206D672
| ML20206D672 | |
| Person / Time | |
|---|---|
| Site: | North Anna  |
| Issue date: | 03/31/1987 |
| From: | TRIDENT ENGINEERS, INC. |
| To: | |
| Shared Package | |
| ML20206D632 | List: |
| References | |
| NUDOCS 8704130395 | |
| Download: ML20206D672 (124) | |
Text
.
LJ ENGINEERING ANALYSIS SERVICE AND MAINTENANCE NORTH ANNA POWER STATION FAIRBANKS-MORSE 38 TD8-1/8 DIESEL ENGINES PHASE II Prepared for VIRGINIA ELECTRIC AND POWER COMPANY Richmond, Virginia Prepared by TRIDENT ENGINEERING ASSOCIATES, INC.
Annapolire, Maryland l
j i
l ENGINEERING ANALYSIS
! SERVICE AND MAINTENANCE l NORTH ANNA POWER STATION FAIRBANKS-MORSE 38 TD8-1/8 DIESEL ENGINES PHASE II March 1987 8704130395 870331 PDR ADOCK 05000338 P PDR TRIDENT EN GIN E ERIN G AS SO CI ATES. IN C.
- k. .f EXECUTIVE
SUMMARY
In early 1985, Trident Engineering Associates, Inc., was retained by Virginia Electric and Power Company (Virginia Power) to assist in the reliability improvement program for the Emergency Diesel Generators (EDG) at the North Anna Power Station (NAPS) Unit 2. Trident participated in all the major overhauls and repairs for those EDG engines, as well as the majority of the maintenance activities to the present date.
This participation included inspection and evaluation of components, evaluation of procedures, technical specification review, and evaluation of engine operation in accordance l with technical specification requirements. In early 1986,
)
Trident was asked to prepare an engineering analysis of the Fairbanks Morse 38 TD8-1/8 turbo-charged Diesel engines at NAPS.
This engineering analysis focused on certain engine com-ponent loads, materials, lubrication, temperature effects and control systems. Materials and loads were analyzed as were predictive analysis programs, monitoring and control systems (see Virginia Power Letter Serial No.86-179 of 1
March 25, 1986.)
[
L As a part of the Reliability Improvement Program, Virginia
(
! Power has implemented certain actions over the past two years ii TRID ENT ENGINEERING AS S OCI ATES, IN C.
o resulting in improve nents in the operation and maintenance of the engines and subsequent availability improvement.
These actions include:
- 1. Revision of Technical Specifications to provide for slower starting and incremental loading during monthly surveillance testing;
- 3. Implementation of a periodic lube oil analysis program directed toward early detection of developing problems;
- 4. Assignment of a dedicated operator to monitor load level during routine testing to preclude engine overload;
- 5. Verification of fuel injection pump rack settings to ,
prevent excessive engine overload;
- 6. Formation of a reliability group consisting of Virginia Power, Trident, and Colt representatives;
- 7. Revision of preventive maintenance procedures and schedule for doing preventive maintenance per manu-facture's recommendations;
- 8. Revision of operating procedures to reflect recommen-dations in Fairbanks-Morse Colt Industrie's letter of April 17, 1985.
- 9. Performance of periodic vibration analysis;
- 10. Initiation of a performance trending program and utilizing a computer graphics package to display trended parameters; iii TRIDENT EN GIN E ERIN G ASSOCI ATES. INC.
bd 11. Replacement of cylinder liners with new high-temp seal liners;
- 12. Performance of periodic engine overhauls in accordance with manufacturer revised recommendations;
- 13. Replacement of piston assemblies with manufacturer rebuilt assemblies;
- 14. Cleaning the engine oil coolers;
- 15. Having the Manufacturer rebuild and calibrate all injec-tion pumps;
- 16. Setting of injection timing to obtain consistent firing pressures for all cylinders;
- 17. Factory training for maintenance engineers and mechanics;
- 18. Institution of semiannual piston pin floating bushing gap measurement;
- 19. Improvement of the overall engine material conditions by replacing bolts, gaskets, etc.;
- 20. Establishment of a requirement for inspection of blower and turbocharger running clearances during every refuel-ing outage.
The results of Trident's analyses are presented in detail in the following sections of this report. In summary, the con-clusions and findings of the analyses follow: (The numbers in parenthesis following each conclusion refer to the recom-mendations in this Summary that follow the conclusions.)
iv TRlDENT ENGINEERING AS SO CI ATES. IN C.
f:
4
~
W
(
ld 1.ffasUfailuresofthepistonpinfloatingbushings, piston skirts and cylinder liners are attributable to a combina-tio.n of factors resulting from inadequate lubrication g and engine overload. (1,2)
- 2. The Chevron Delo 6000 lube oil currently in use at NAPS in the EDGs is satisfactory for the intended engine service. (7,8)-
i
- 3. The assignment of a dedicated operator to monitor and l control engine overload, along with the current limits l
in engine load imposed by the new rack stop settings,'is satisfactory on an interim basis to prevent engine over-load; evaluation efforts to ine .1 improved alarms or controls should continue. (12,15)
- 4. Bearing load calculations indicate an operating region y ,
where piston pin bushing damage could occur due.to the
- use of a lubricating oil which does not provide an i
, adequate lubricating film. (1,2,7,8,11,12) y
- 5. The lube oil analysis program implemented at NAPS pro-vides assurance that the oil in service at any particu-lar time is adequate. (14) 6
% 6. The design of the main bearings is fully adequate.
Thrust face performance could be improved by enhancing bearing manufacturing tolerances and inspections. (3)
\* 7. There were no metallurgical indications of improper composition of material properties in the piston pin
)
bushings or main bearings. (4) v J < TRID EN T EN t.sl N EENIN G AS SU C f AT E S. I N C.
)
1 i
1 l
Ed 8. The lube oil booster system design is adequate for the service intended. (5)
- 9. Engine lubricant performance may be adversely affected by local high temperatures of the Calrod type oil heater. (6)
- 10. The roots-type blower in the scavenging air system is the only potential problem in the air system. Periodic l
monitoring of this item should continue. (9)
- 11. Significant improvements in the maintenance and monitor-ing of the EDGs, with regard to reliability improvement have been implemented. (10)
- 12. The data currently available on the condition of the engine should be expanded to establish trends that can be used in an enhanced preventive maintenance program.
Additional recording and monitoring instrumentation to provide this data should be considered. (10)
- 13. The analysis showed no inadequacies in the jacket water and inter-cooler systems, or in the water chemistry for those systems. (N.A.)
l 14. The fuel injection pumps and nozzle performance have been improved by upgrading preventive maintenance and
! procedures. The potential for air blockage of fuel delivery from the reserve tank should receive further review. (13) i l vi TRIDENT ENGINEERING AS S O CI ATE S, IN C.
l LJ 15. The starting air systems are considered to be satie-factory. (N.A.)
- 16. The design of the EDG foundation, the engine-generator alignment, and the generator spherical roller bearings were analyzed and determined to be acceptable. (N.A.)
- 17. There were no indications of cylinder liner or piston inadequacies. (N.A.)
Based on the results of this analysis, Trident recommends that:
[ 1. The gap monitoring program for the piston pin bushings be continued, although the inspection frequency may be revised based on gap trends.
- 2. Engine operation continue to be carefully monitored and limited to prevent overload conditions.
- 3. Virginia Power, in conjunction with the engine manufac-
, turer, consider quality control improvement for the main bearing manufacturing tolerances.
- 4. Virginia Power request that the engine manufacturer review the bushing data to assure that the bushings r
L being used are consistent with their specifications.
- 5. Virginia Power review additional test data on the upper l
lube oil booster system to ensure system performance.
I I
vii TRIDENT ENGINEERING AS SO CI ATES, IN C.
1 l
bd 6. Changes to the lube oil keep-warm system be considered to preclude the possibility of exposure of the lube oil to excessive temperatures.
- 7. Continue using Delo 6000 oil with an enhanced lube oil monitoring program and installed instrumentation.
- 8. No consideration be given to use of synthetic oils until adequate test data is available.
- 9. Consider the development of a program, to study blower failure and promote increased blower reliability.
- 10. Consider the installation of electronic monitoring of the engine during surveillance testing and adoption of a proven commercially available electronic surveillance system with improved instrumentation, monitoring, recording, and trending of EDG performance.
- 11. Continue the practices of setting the fuel rack stop at the one-half hour load limit for standby service, and consider setting the load limit on the hydraulic gover-nor at the 2,000 hour0 days <br />0 hours <br />0 weeks <br />0 months <br /> load limit during periodic testing.
- 12. Virginia Power continue to pursue with the engine manu-I facturer the development of a device or system that will automatically perform the load limiting function includ-ing alarms and controls.
viii TRID E NT ENGINEERING AS S O C I AT E S, I N C.
- 13. Consider a design review of the fuel transfer system to preclude the possibility of air blockage.
- 14. Continue the lube oil monitoring program including trend-ing. As a data base is established, some modification to the monitoring program may be desirable.
- 15. Continue the practice of assigning a dedicated operator to monitor and control engine overload.
l
(
l l
(
{
{
l
(
(
ix TRID ENT ENGINEERING ASS OCI ATES. IN C.
[
L bd TABLE OF CONTENTS Page
1.0 INTRODUCTION
.............................. 1 2.0 BEAR 1NGS.................................. 3 2.1 PISTON PIN FLOATING BUSHINGS (BEARINGS)... 3 2.1.1 Historical Perspective.................... 3 2.1.2 Design Maturity........................... 5 l 2.1.3 _ Failure Mode.............................. 5 2.1.4 Analytical Focus.......................... 6 2.1.5 Analytical Method......................... 7 2.1.6 Analytical Results........................ 9
{ 2.1.7 Conclusions............................... 10 2.1.8 Recommendations........................... 12
( 2.2 MAIN BEARINGS............................. 12 2.2.1 Bearing Design............................ 14 2.2.2 Lubrication............................... 16
( 2.2.3 2.2.4 Analysis.................................. 17 Conclusion................................ 19 2.2.5 Recommendation............................ 20 f
2.3 METALLURGICAL EXAMINATION OF BEARINGS..... 20 2.3.1 Examination Results....................... 20
( 2.3.2 Conclusion................................ 22 L 2.3.3 Recommendation............................ 24 3.0 LUBRICATION............................... 25 3.1 BEAR 1NG rA1tURES.......................... 25
[ 3.2 LUBE OIL SYSTEM ADEQUACY REVIEW........... 26 L
3.2.1 Lube Oil System Design.................... 27 3.2.2 Lube Oil Keep-Warm System................. 31 k 3.3 LUBRICANT SELECTION....................... 33 3.3.1 Conclusion................................ 35 3.1.2 Recommendations........................... 35
( 3.4 tUBR1 CANT MON 1 TOR 1NG...................... 35 3.4.1 Conclusions............................... 36 l 3.4.2 Recommendations........................... 36
{
{
X TRlDENT ENGINEERING ASSOCI ATES, INC.
l l
(
i I
( LJ \
TABLE OF CONTENTS (Continued) l l
Page 4.0 SCAVENGING AIR SYSTEM..................... 38 4 . '. SYSTEM DESIGN............................. 38 4.2 FAILURE ANALYSIS.......................... 41
4.3 CONCLUSION
S............................... 46 4.4 RECOMMENDATION............................ 46
~
5.0 SURVEILLANCE INSTRUMENTATION.............. 47 5.1 ENGINE ANALYSIS INSTRUMENTATION........... 47
{
5.2 cONCtUS10NS............................... 54 f 5.3 RECOMMENDATION............................ 54
[ 6.0 TESTING................................... 55 6.1 TESTING-PISTON PIN BUSHING END CLEARNACE.. 55 f
L 6.2 PROCEDURE................................. 56 6.2.1 Recommendation............................ 58
[ 6.3 SURVEILLANCE TESTING AND GOVERNOR OPERATION........................ 58
, 6.3.1 Conclusions............................... 59
[
6.3.2 Recommendations........................... 60 7.0 FUEL SYSTEM AND CONTROL................... 61
(
7.1 i ANALYSIS - INJECTION PUMP................. 63
( 7.2 ANALYSIS - INJECTION NOZZLE............... 66 7.3 ANALYSIS - FUEL TRANSFER SYSTEM........... 66
7.4 CONCLUSION
S............................... 68 7.5 RECOMMENDATION............................ 68
{
xi
( TRIDENT ENGINEERING ASSOCI ATES. INC.
l l
( TABLE OF CONTENTS (Continued)
Page 8.0 AUXILIARY SYSTEMS.........................
( 69 8.1 GENERATOR BEARING......................... 69 r 8.1.1 Analysis.................................. 69
( 8.1.2 Bearing Type Design Considerations........ 70 8.1.3 Lubricant and Lubrication................. 73
, 8.1.4 Bearing Mounting.......................... 73 l 8.1.5 Conclusion................................ 75 8.1.6 Recommendation............................ 75
( 8.2 JACKET WATER AND INTERCOOLER SYSTD.' . . . . . . . 75 L
8.2.1 System Design............................. 77 8.2.2 Evaluation - Cooling Effectiveness........ 78 r 8.2.3 Evaluation - Radiator Effectiveness
[ and Coolant Temperature Control........... 79 8.2.4 Evaluation - Jacket Coolant Pumps......... 80
, 8.2.5 Conclusion................................ 80 s
8.2.6 Recommendations........................... 80 8.3 PISTONS AND CYLINDER LINERS............... 80 r
8.3.1 Analysis - Cylinder Liners................ 82 t 8.3.2 Analysis - Pistons........................ 83 8.3.3 Conclusion................................ 84 r 8.3.4 Recommendation............................ 84 L
I s
e I
L r
L r
L I
xii TRIDENT ENGIN EERIN G ASSOCI ATES. INC.
l l
LJ
1.0 INTRODUCTION
l This report describes the result of Trident Engineering Asso-ciates Phase II engineering analysis of Virginia Power's, North Anna Power Station (NAPS) Emergency Diesel Generators (EDG).
The Phase I report, issued on July 2, 1986, addressed certain service and maintenance aspects of Trident's engineering analysis. Phase I engineering evaluation efforts included reviewing Fairbanks-Morse (Colt Industries) Diesel generator mechanical failures at NAPS in great detail and at other nuclear power plants in lesser detail. Phase I activities centered around summarizing the EDG mechanical failures at NAPS, describing the associated failure mechanisms, analyzing and evaluating the effect of failures on reliability and recommending corrective actions to improve reliability.
L This Phase II engineering analysis focused on certain engine component loads, materials, lubrication, temperature effects, and control systems. Materials and loads were analyzed as were predictive analysis programs, monitoring, and control systems. (See Virginia Power letter Serial No.86-179 of i March 25, 1986.) The component load analyses performed were
- based on a variety of operating conditions including a range i
s 1 ,o c~ r c~ o,~ c ce,~ o .. oc,me. ,~ c.
l .
ld of engine power levels. Control systems were analyzed from the standpoint of their effect on engine reliability.
(- Specific EDG issues receiving detailed review in Phase II L
include: calculation of critical bearing loads, continued review of the lubrication system, metallurgical examination of failed bushings, review of overspeed protection and load control, evaluation of the engine scavenging air system, and a review of the EDG instrumentation. This report is organized to describe the detailed reviews, provide operating history perspectives, describe an analysis with results and present conclusions and recommendations.
{
L
[
r L
{
1
(
TRIDENT EN GIN EERIN G ASSOCI ATES. INC.
[
l tJ 2.0 BEARINGS Trident's evaluation of the engine bearings included analy-sis of the main and piston pin bearings (bushings). Analyti-cal efforts were focused on determining the loads experienced by the piston pin bushings since these bushings experienced failures. Prior to performing the analysis, firing pressure data war developed by a Trident expert. Since the data could not be obtained by measurements of the NAPS engines, it was derived from measurements of similar Fairbanks-Morse engines.
Trident believes, however, that the results of the analysis are within acceptable confidence levels for calculations of this type.
2.1 PISTON PIN FLOATING BUSHINGS (BEARINGS) 2.1.1 Historical Perspective The piston pin and its bushing (shown in Figure 2.1) have historically been a design problem in two-cycle Diesel engine development. The oscillatory motion of the connecting rod through a limited arc (20 to 25 degrees) does not allow the development of the classical oil film upon which journal bearings rely. In addition, the limited bearing size led to high loadings. The lubrication is essentially a " squeeze film" type in which the viscous oil is left in the load bear-ing arc for a limited load duration time. This lubrication mechanism depends on the load being relieved periodically TRIDENT ENGINEERING ASSOCI ATES, I N C.
/ \
\ \ INSERT BUSHINGS (FIXED)
// [\\\\\\\\ / / j PIS' ION ____
PIN - _ _ __ __ __
+ _
PISTON I\\\\\N\\\\V /
PIS' ION INTERNAL BOSSES
. /
\
FICATING j#p BUSHING N s N
SECTION THROUGH PISTON PIN BEARING FIGURE 2.1
/
_4_
TRID ENT EN GIN EERIN G ASSOCI ATE S. IN C.
I I
L3 to allow new oil to flow into the load bearing region. In the the case of high engine output there is the additional
( requirement of carrying away the considerable heat generated by the viscous shear of the thin oil film, both circumfer-
{
entially due to relative motion between the piston pin and
( bushing (bearing) and axially as the oil is squeezed out.
In contrast the similar part in the four cycle engine is
( relatively trouble free, due to the abundant opportunity for the oil film to be renewed. The heat is carried away during
(
, the idle cycle where the load is completely reversed over k half a revolution.
( 2.1.2 Design Maturity The Fairbanks-Morse 38 TD8-1/8 engine is of a very mature
(
design, having been in service for over 40 years. There is
( ample evidence that piston pin bushings of identical design to the NAPS units have a history of reliable service.
[ Fairbanks-Morse tests conducted at the 168 hour0.00194 days <br />0.0467 hours <br />2.777778e-4 weeks <br />6.3924e-5 months <br /> and 30 minute ratings also indicate satisfactory bearing performance.
{
( 2.1.3 Failure Mode Based on the visual observations of the bushings removed
[ from the NAPS engines, the failure appears to be one of lubrication film failure wherein the bushings first experi-
{
ence smooth burnishing, particularly near the bushing ends,
[
TRID ENT EN GIN EERIN G ABSOCI ATES. INC.
[
l l
L.f together with slight surface discoloration due to elevated temperatures. It is postulated that these temperatures ini-tially are the result of shear in the extremely thin lubri-cant films, without actual metal-to-metal rubbing. As the bulk temperature of the bushing increases, the hardness of the bronze begins to decrease and the lubricant film effec-tiveness decreases. The first effect appears to be a slight wiping of the bronze surface, which builds up sharp edges on the oil grooves, which further reduces the lubricant film and increases heat generation. This leads to a more rapid temperature rise and consequent plastic extrusion of the bushing. Throughout these failures the rubbing surfaces remains smooth and there is no heat discoloration of the piston pin or in the connecting rod eye, indicating that an adequate flow of lubricant is being supplied to those loca-tions.
2.1.4 Analytical Focus The piston pin bushing analysis focused on two questions.
First: Is there a demonstrable difference in upper and lower piston pin loading? This question arises from the observation that the bushing failures are primarily associ-ated with the upper crank shaft. Second: Does the opera-tional environment of the piston pin bushings represent an operating region where the increase in load (overload) and/or lubrication quality is particularly important?
TRIDENT ENGINEER 4NG AS S OCI ATE S. I N C.
l l
l LJ 2.1.5 Analytical Method The piston pin load at five or six degree increments was calculated in accordance with available pressure data. The calculated resultant load is the summation of the separate effects of the gas pressure, reciprocating inertial load of the piston mass, and the engine couple reaction to the mass moment of the connecting rod. Figure 2.2 illustrates the loading as a function of crank angle for the stated condition.
Upper and lower crankshafts and pistons are shown schemat-ically in the figure to aid the understanding of the varia-tion of load with crank angle. The gas pressure sets used were from sets of pressure versus rotation angle tabulations obtained from the engine builder (Fairbanks-Morse). These sets were not a coordinated set specific to the installation in question nor do they in general have adequate documenta-tion to identify possible differences regarding the NAPS units. They are nevertheless useful in establishing the general loading characteristics of the NAPS Fairbanks-Morse 38 TD8-1/8 engines. Differences in upper and lower piston pin load exposed to the same cylinder pressure are due to the difference in connecting rod length and to the 18 degree crank lead of the lower crankshaft. Appendix A contains more detailed information about the analytical methodology as well as additional figures. Figure A-1 shows a comparison of the upper and lower piston pin vector loads.
TRIDENT ENGINEERING ASSOCI ATES. I N C.
l1 -l1\lll,lll ,i J 1 o
a
- l
~
~
~ , E L
G
- \
- s' N A
\ Kk n
\\ No Ar
\
\
\ RC
\g Cr e
\ 2
.\
Sw .
/
f UoL 2 l
T* S-R E
- o o
/ f I o' o /
. O EL R U
f VC Y G
/ ' j*< D/ I F
- , AP I O
._ L2 6
_ R3 O
T C
_ E
- V t
2
- o
- < m o
s 0 0 m o o 5 4 2
- - - - - - ~
- ga9,E9g0W e
- 4I
- 490Mz< mz9znn2zo >mm8 ?"
t I'l'lll(l ,
7 s
r L
I I I
UA 2.1.6 Analytical Results Conclusions based on the results of the calculations using six different pressure data sets were:
- 1) Peak loadings are essentially the same for both upper and lower pistons. The peak loadings are directly related to peak firing pressures and fuel conditions.
Peak firing pressure is not a direct measure of the bushing distress, since it is only one component con-tributing to the bushing loading.
- 2) There is a significant difference between upper and lower pins at the time where they are subjected to their minimum loads. This is the time when the oil film between pin and bushing should be refreshed and heat removed. This load point occurs in the compression stroke, about 70 degrees before minimum volume (firing dead center) . Since this point is well before fuel injection occurs, the gas pressure is a function of the air receiver pressure and air density and not fuel com-bustion. Air receiver pressure is related to engine load, since the air is supplied by the exhaust driven I turbocharger. Specific air receiver pressures are a function of the engine auxiliary components (turbo-charger, compressor, etc.), adjustments, and ambient pressures and temperatures.
- 3) For the pressure sets analyzed, the upper piston pin has a minimum load of about 10 percent of maximum load.
TRIDENT ENGINEERING AS S O CI ATE S, IN C.
LJ This load is essentially constant over an arc tracing the connecting rod motion. The lower piston pin has minimum loads ranging from a high of about two percent of maximum loading to load reversal of about one percent over the arc of motion. For example, the direction of the load resultant for the lower piston pin bearing exceeds the connecting rod arc of approximately 20' as shown in Figure 2.3 for 300* and 310' lower crank angle.
- 4) The analysis is based on approximated rated load condi-tions and the analysis indicates it is probable that there is an operating regime where the piston pin bush-ing is particularly sensitive to either insufficient lubrication or increased loads due to high air receiver pressure.
- 5) Overload operation generates increased air receiver air pressures which are greater than the proportional increase in engine load.
2.1.7 Conclusions The bearing load calculations indicated an operating region where piston pin bushing damage could occur due to a marginal oil film caused by excessive engine loading. Inadequate lubricating oil, particularly oil which contained contami-nants or entrained air, could also causes overheating of the bushings and loss of bushing mechanical strength. Trident TRIDENT ENGINEERING ASSOCI ATES. INC.
I not 20
,\ !- " * -u- = _ ,< ,.
LOAD = LOAD = LOAD =
4350LBS 3771LBS 3404 b o ,
LBS fi's)'; . f 4'
- 4 f UPPER ,glll fllg .e '
kj f;f '
5 $ 5 t
' , '<; . k'"{l ',"
t'
'3 *' e+
g ..;ny p ,
4 .
g 8 ; ,
e' l E Lower Cronk Angle 290* 300* 3iO*
g i e i i
> 6*
N Y Av" g y j LOWER; g'fg g',
m -
,sa,wp 'ilj;j , $i -
[dk hk, 1
- 4. ,
mI ';'j$LOAD.
'g 3},/g,'1 wYi$,{h'p.
OAD= LdAD=
1151 LBS 573 LBS 399 l
nort
. i4.e.
,t (j
. LBS l MINIMUM PISTON PIN LOADS SET TWO-321 FP/CYL FIGURE 2.3 /
l Ub determined that the bushing failures experienced in the past were caused by a combination of these mechanisms.
2.1.8 Recommendations
- 1) Trident recommends that the gap monitoring program discussed in section 6.0 of this report be continued, with inspection frequency to be determined by trending i
the gap data. J:f any change in gap measurement is expe-rienced, Trident recommends that a more extensive inves-tigation be considered to ensure the adequacy of the l piston pin bushing lubrication. This further testing sould be performed to determine bushing loading, lubri-cation, and temperatures on the NAPS engines under a range of operating conditions.
- 2) Trident recommends that engine operation continue to be carefully monitored and limited to prevent overload conditions.
2.2 MAIN BEARINGS The main bearing shells are machined from cast aluminum.
l Each half shell has a central oil groove, with oil spreader grooves at the parting lines as shown in Figure 2.4. There is an oil entrance hole at the middle of each half, and a dowel and dowel hole for alignment with the mating half.
Location numbers and letters are stamped on the face towards the exhaust end of the engine, and on the side towards the
s s
b o
L L
SHARP EDGES
- I .. kfjffj
[ Q % -i di.%
) %
( e MAIN BEARINGS SHELL P
{
J%d SHARP EDGES $
j AND BURRS j"{
(
( I' MAIN BEARINGS FIGURE 2.4 1
- 1. _ ,
[ , i TRIDENT ENGINEERING ASSOCI ATES. INC.
h
( L3 The bearing shells are located axially and control side.
are rest. rained from rotating with the shaft by a dowel pin in the bearing cap.
Each crankshaft has a combination radial and thrust bearing immediately forward of the vertical drive bevel gear. These bearing shells, as shown in Figure 2.4 have a thrust face approximately 1.3 inches wide facing the gear and 1.1 inches facing away (forward) from the gear. These shells are solid
(
aluminum, also. The larger thrust face has four oil. spreader grooves in each half. The thrust face is flat, without any tapered lands or ramps to provide enhanced thrust load cap-abilities. Load capacity depended on the dynamic film gen-erated downstream of each oil groove.
2.2.1 Bearing Design Design of a main bearing for a Diesel engine must take into account the severe dynamic loading to be encountered, one ef fect of this dynamic loading is to causc. the halves of the
(
bearing shell to flex and vibrate in the bearing housing.
Any such flexing and vibration can result in the following failure modes:
- erosion of the back of the shell at points where the two surfaces repeatedly press tightly together and then spring apart;
" ' " " ' ' " '" " "'" ^"" ' ' S ' ' " C '
6 L
- fretting at the parting lines between the two halves of the shell;
[
- erosion of the face of the shells, particularly near the parting lines;
( ,
- cracking of the shells due to fatigue failure. 4 A prime factor in avoiding the above failure modes is ade-quate bearing shell thickness. This requirement is met for the aluminum bearings, in that at the time of their design
(
they were to be direct replacements for heavy, steel backed or bronze-backed babbitt bearings. A second factor, that of preventing movement of the bearing shells in their housing,
( is met by providing adequate " crush" on the shells. " Crush" refers to the lengthening of the shell halves by a few thou-
{
sandths of an inch, so that when the bearing cap is tight-ened to its saddle, the ends of the shell halves contact before the cap is drawn down fully. This tightening puts an elastic compression into the bearing shells. The adequacy of the design with respect to fretting, erosion and cracking
{
is demonstrated by the appearance of bearings which are in EDG and other service. Observations of the back surfaces show very little evidence of the onset of either fretting or erosion, and no cases of bearing shell cracking were reported.
TRIDENT EN GIN E ERIN G ASSOCI ATES. INC.
0
' LJ The crush imposed on the main bearing shells can be lost if a bearing experiences excessive temperatures. As the tem-perature increases, the aluminum shells expand much more
( rapidly than does the steel housing, and this expansion can result in exceeding the elastic limit of the aluminum in compression. When the engine is stopped and temperatures equalize, the shell, which has overheated, retracts and may have no crush or even may have a gap between the halves. If an inspection of the engine reveals a gap (using a 0.002" feeler gage) the bearing should be considered to have failed and must be replaced. If heating has simply eliminated the crush, then that shell may be subject to the effects of looseness in its housing, with resultant fretting and er o-sion. This has not been a problem at NAPS, and periodic exam-ination during each major outage provides adequate protec-tion for the engine.
(
2.2.2 Lubrication
( Both the main and connecting rod bearings sustained contin-uous complex loading with no indication of load reversal after examining the bearings' surfaces. However, the rota-tion of the journals, the shifting of the load vectors and the brief duration of peak loading permited the bearings to operate indefinitely with very little or no wear. In general, contact between the journal and its bearing shells was evi-denced by areas of light polishing, with minor scratching
'"* "" '""'" ^"" ''"
s
~
h
( Ub due to solid contaminants. It was assumed that most of this I contact resulted from twc operating modes:
initial run-in of new bearing shells; normal boundary lubrication conditions during start-up.
Each crankline has its own lubricating oil header. Oil is delivered from the engine-driven gear pump through a full flow filter, a heat exchanger and a strainer to the upper and lower headers in the engine. It then flows through jumpers to each main bearing cap and into the bearings.
Most of the oil continues on, via passageways in the crank-shaft, to the connecting rod bearings and on through the connecting rod to lubricate and cool the pistons. In the main bearings, lubricating oil flows into the load-carrying areas from the spreader grooves at the parting lines. Rota-tion of the crankshaft journal drags oil into the loaded zones to maintain unbroken hydrodynamic fluid films.
2.2.3 Analysis j Main bearing shells have not experienced any failures which were evident by inspection of the engine or which interfered with routine engine operation. However, the main bearing shells have experienced some scuffing which prompted their replacement.
TRlDENT EN GIN E ENIN G AS S O C I ATE S. INC.
~
LJ The scuffing found in the main bearings consisted of shallow removal of aluminum; a flaking or tearing mode of material removal was observed. Examination of the nature of the metal removal and the patterns observed led to the conclusion that the most likely cause was cavitation erosion due to oil film separation; however, fatigue failure and stray DC current were also considered to be possible failure initiating mechan-isms. The scuffing occurred with a lubricating oil which had displayed severe foaming properties and there have been no similar observations since changing to an improved lubri-cant. Electrical grounding of the engine is now carefully monitored. It is unlikely that any deficiency in bearing design or materials was a contributor. Even in cases where erosion covered extensive portions of a bearing, there was no indication that the bearing could not fulfill its load carrying function.
A feature of the main bearings and thrust bearings which might contribute to reduced oil films or oil film separation is the sharp intersections between oil spreader grooves and the bearing surfaces, as noted in Figure 2.4. Although the angles formed by these intersections are small, on the order of ten to 30 degrees, abrupt intersection of the two planes could act as a scraper and severely reduce the flow of oil into the very thin space between the bearing and the shaft.
TRIDENT ENGINEERING AS SOCI ATES, IN C.
s A
F Lj This effect of flow restriction and separation of the oil u film into individual streams has been demonstrated in various laboratory and field tests. Examination of replacement bear-ing shells showed that some of the oil groove edges were not only sharp but had burrs left from machining, which would have greatly enhanced the scraping effect. This condition was most serious where it occured on the thrust faces, and may have accounted for the numerous thrust bearing replace-ments. Correction of this problem involves providing smooth blends of the oil spreader grooves into the bearing surfaces.
All edges should have approximately one-quarter inch radii of curvature, and on the thrust faces the load capacity would be enhanced by a more gradual curvature. One example of the effectiveness of correct blending of the surfaces is in Colt Industries (Fairbanks-Morse) drawing 16 300 214, for the piston insert bushing, where it is stated that "all oil groove edges must be well broken."
2.2.4 Conclusion Trident's analysis of the main bearings revealed that the design of the bearings is adequate. However, the thrust face performance may be improved by the smoother blending during the manufacturing process of the edges of the oil groove in the thrust face of the bearing.
TRIDENT ENGINEERING ASSOCI ATES. INC.
l
/
I 2.2.5 Recommendation Trident recommends that Virginia Power in conjunction with the engine manufacturer investigate smoother blending of the edges of the oil grooves during the manufacturing process.
2.3 METALLURGICAL EXAMINATION OF BEARINGS Metallurgical examinations were conducted on an aluminum main bearing (upper crankshaft) and a bronze piston pin bear-ing (upper crankshaf t) . The examinations consisted of chem-ical analysis, hardness testing and metallographic examina-tion. These examinations were conducted to determine if the proper materials, with correct properties, were used in the bearings and whether the bearings themselves contributed or caused the observed bearing failures.
2.3.1 Examination Results 2.3.1.1 Bronze Ps:.'n*1 Pin Bushing Chemical analysis of this bushing revealed the following composition:
Zinc 0.10%
Tin 9.77%
Lead 7.79%
Nickel 0.30%
Copper 82.04%
l TRIDENT ENGINEERING AS SO CI ATE S. I N C.
I I
l L3 Fairbanks-Morse drawing 16 300 214, which was entitled " Bush-ing, Piston Insert", called for cast bronze SAE 64. According to the manufacturer, the composition of the bronze bushing is proprietary and therefore was not available for Trident's review.
Bushing hardness was measured at 6 points on the inner sur-face of the bushing and 6 points on the cross section. This testing indicated an average Brinnell hardness of 91 on the inner surface and 92 on the cross section. These hardness values are within expectations for materials of this type.
Metallographic examination of this bushing was performed on a mounted, polished, and etched specimen. A micrograph taken at 100X is shown in Photograph 1. The microstructure seen in this photograph is suggestive of centrifugally cast high lead tin bronze.
2.3.1.2 Aluminum Main Bearing Chemical analysis of this bearing revealed the following composition:
Copper 1.15%
Iron 0.44%
Silicon 0.23%
Manganese 0.05%
Chromium 0.05%
TRIDENT ENGINEERING AS S O CI AT E S. I N C.
/
f L
f ld Zinc 0.05%
Titanium 0.12%
l Nickel 1.12%
Magnesium 0.10%
Tin 6.00%
Aluminum 90.68%
The results of the analysis indicated that the composition of the aluminum main bearing were within the specification for alloy SAE 770. Hardness testing revealed a Brinnell hardness value of 57 when averaged over 6 measurements.
These hardness values are within expectations for materials of this type.
The area indicating damage on the inner surface of the bear-ing was examined at 50X as shown in Photograph 2. A section indicating damage was polished for examination. Photographs 3 and 4 (at different magnification and without etching) show subsurface pitting and consequent flaking possibly caused by stray direct current in the vicinity of the bearing.
2.3.2 Conclusion Trident found no metallurgical indication that the piston pin bushing failures or main bearing scuffing were due to improper composition or material properties.
TRID EN T EN GIN EERIN G ASSOCI ATES. I N C.
$?%
.y
< c. ..
t ' .
pl.
- &s 1 w ,
4'
- e
, 44 ,
eC' Ws
^
l
( s
^
PHOTOGRAPH 2.1
?% a-3 'E T. Micrograph of bronze bearing at 100X.
_,}h I
- 5. .%;-,,, g . -
- \
~
l , -+ '~
b J
,_ '* % a. ~3 s_ .y
=
.., .- ;?b .* '
iN) : .-..
, ._ 3, ,
J.A2. -,
~
l s
PHOTOGRAPH 2.2 i ' -[dk Bearing surface at 50X -
g A
,f 23 "#* R 5 %r !
sp.
.s . .-
LQ -
M
.&f' ;f :ah j , .g r } s ', ,y Y) W , ?! *
%t . k)eg, f ,'~shi' .
- f. 4
. l ,'
- N& * , 'f ,'- -
' 45* . k V y,- ,$f% i
. ' .h'fr..; f.I ?; .3 k f '4 ,' s.',j ? '
1, 1 -
.y n
' ~; ~ *y ;', ' , I
._ ,,. d 2 :!!
A1 ., ~ , L...
i PHOTOGRAPH 2.3 Micrograph of surface flaking 100X.
.o $ ' /, Ls 4 '
' ^
ni v se b. .
,. CT a *
? -n ' pk -
. . _b,,
s ?)- Y c #' j , p g i " i -
t/
ye .g .
',< h .
?
4 ggfS$&E.J.:
- 9 sl.,
" L,&+,.h f Q. g, L. '
PHOTOGRIsPH 2.4 Micrograph of surface flaking 1000X.
Ed 2.3.3 Recommendation Trident recommends that Virginia Power request that the engine manufacturer review the bushing data presented herein to assure that the bushings in the NAPS engines are in accor-dance with their specifications.
TRIDENT EN GIN E ERIN G ASSO CI ATES. IN C.
bd 3.0 LUBRICATION The lubrication systems of the EDG units at North Anna are important to ensure that every moving part receives an ade-quate supply of the proper grade of lube oil. Proper lubri-cation maintains a stable oil film which prevents metal-to-metal contact during all operational modes including start-up.
Correct lube oil system design and lubrication is, therefore, essential to satisfactory Diesel engine operation. Suitable lubricant selection and oil maintenance will help ensure engine reliability and extend engine life.
3.1 BEARING FAILURES Bearing failures in Diesel engines are frequently attributed to lack of proper lubrication. Although engine bearings are designed with a safety factor for maximum loads and speeds, abnormal overloading of the bearings may result in failure.
Heat, caused by the shearing of the oil in the bearing, must be carried away by the lubricating oil and by conduction through the bearing. At higher temperatures, lube oils may change chemically and attack bearing materials, especially alloys of cadmium and silver, copper and lead, or cadmium and nickel or silver, bronze, and brass alone. Bearing cor-rosion can be caused by organic acids produced by the products of combustion and oxidation of the lubricating oil when oil temperatures are high. Anti-oxidant and anti-corrosion oil TR ID ENT ENGINEERING AS S O CI AT E S, I N C.
Ld additives help keep bearing corrosion in check. Bearing corrosion may occur with an unsatisfactory lube oil selection. Bearing materials, therefore, have a direct influence in lubricating oil specifications.
Bearing failures which are oil related include those caused by:
abrasive particles brought to the bearing clearance space by oil inadequate oil supply improper bearing tolerance oil deterioration foaming of the oil improper oil viscosity impurities in the oil; for example, water, and/or coolant.
3.2 LUBE OIL SYSTEM ADEQUACY REVIEW To assess the overall adequacy of the lube oil system, the review was broken into two parts; a lubrication system design study and a lubricant selection analysis. The lubrication
! system design review covered the following areas:
upper lube oil accumulator system prelube procedures keep-warm system oil temperature Calrod immersion oil heater TRID ENT ENGINEER 6NG A f3 S O C I AT E S. I N C.
LJ
- temperature control control of impurities and emulsions filtration.
The lubricant selection review considered the following top-ics:
oil selection oil monitoring.
3.2.1 Lube Oil System Design The lube oil systems on the 38 TD8-1/8 engines at North Anna are expected to perform satisfactorily under a variety of conditions. Starting and operating from ambient, cold condi-tions are particularly demanding on engine lubrication. A single lubrication system design for a widely used Diesel engine, such as the basic Fairbanks-Morse 38 TD8-1/8 OP Diesel engine, may not have met all the requirements of the various engine applications.
There are four distinct sections of the engine requiring lubrication:
power cylinders, piston rings, and pistons bearing surfaces gear trains blowers.
TR ID E NT ENGINEERING AS S GCI ATE S. I N C.
LJ Piston pin bushing failures and scuffing of_some of the upper main bearings at NAPS have focused attention on the adequacy of the upper accumulator and associated piping to supply cufficient lubricating oil to the upper drive train during start-up.
3.2.1.1 Lube Oil Booster System There are two air activated booster prelube systems on each of the four Fairbanks-Morse engines at NAPS. One of these systems, the lower lube oil booster system, supplies lubri-cating oil to the drive end main bearing cn the lower crank-shaft, which support part of the generator weight. This system is supplied by the engine manufacturer with the engine.
The other system, the upper lube oil booster system, was installed in late 1982. Reliability concerns prompted the installation of the upper lube oil booster system to provide lubrication to moving parts in the upper section of the engine during start-up conditions. To obtain detailed information about the system, Trident examined the NAPS engineering review of the upper lube oil booster system design as found in Design Change DC-81-S05A, dated September 3, 1982.
TRIDENT EN GIN EERIN G AS S O C I ATE S. I N C.
E d 3.2.1.1.1 System Design The upper lube oil booster system is made up of an accumula-tor and associated piping. Figure 3.1 illustrates the system design. The accumulator discharges the oil when the accumu-lator piston is actuated by air supplied by a regulator in a starting air line. Piping routes the oil to a header for the upper crankshaft. As a result, oil is discharged to the i
bearings along the upper crankshaft at the same time engine parts began to move. The accumulator is refilled with oil from the main lubrication system af ter the engine is oper-ating and the starting air supply is shut off and vented.
Characteristically, when pressure lubricated opposed piston engines are shut down lube oil from the upper part of the engine drains down into the upper pistons and submerges the i l wrist pins and bushings. This residual oil that drains down into the piston provides adequate lubrication to the wrist pin bushings when the engine is restarted.
3.2.1.1.2 Conclusions Trident believes that the system design is adequate based on the information we have reviewed; however, further validation 1
TRlD E N T ENGINEERING AS S OCI ATES. IN C.
SUPPLY LIIEE Pitutt MEhDER b 10 SEARZIIG
,_ ,ER 3 ,
MkIII BEARIIBG L.O.
P,_
.E._
!I U*
- L.O. IREADER k
E I
..I
,Y k .O. SUPPLY LINE FDCM L.O. BOOSTER TO READER L.O. SUPPLY LINE PROM L.O. PUMP >
TO ALL UPPER BEARINGS
' I 9 ' ', l E
L.O. BOOSTER POR UPPER MAIN BEARINGS N
E STARTING AIR REDUCING l l n.h UPPER LUBE OIL BOOSTER FIGURE 3.1
\ r l
t T*'o=7 TRlDENT ENGINEERING AS SO CI ATES. IN C.
LJ of the adequacy of the accumulator lube oil volume should be confirmed by visual verification (during maintenance) of sufficient oil flows to each of the upper crankshaft bearings.
The engine vendor has run tests to determine the adequacy of the upper lube oil booster system. The data associated with this testing was not available for Trident's review.
3.2.1.1.3 Recommendation Trident recommends that Virginia Power request the engine vendor's test data and review the test data to verify proper system design.
3.2.2 Lube Oil Keep-Warm System Surveillance testing and operating requirements for the EDG units at NAPS requires the use of starting support systems.
Lubricant performance during starting is very important.
Since oil viscosity increases with decreased temperature, startup with low lube oil temperatures increases the torque required from the starting air system. Increased torque requirements result in lower cranking speeds and increased starting times. The Diesel generator sets are supplied with a lube oil keep warm system consisting of a heater tank and a small circulating pump. This keep warm system maintains oil supplied to the engines at 125-135 degrees F.
TRIDENT ENGINEERING AS S O CI ATE S. IN C.
LJ 3.2.2.1 Immersion Heater Evaluation A thermostatically controlled direct contact immersion heater, which is located in the engine lube oil system, heats the engine oil. Heat is transferred to the lube oil by convec-tion from the heater.
The lubricant and associated additive package should be pro-tected from contacting hot metal surfaces in excess of 200'F within the keep-warm system. Diesel engine lubricant condi-tions may be adversely affected by the local high surface temperatures which can occur with the Calrod type heater.
In addition, thermal control with this type of heater is somewhat limited. Trident found that in one isolated instance oil blackening occurred in one oil sample due to excessively high heater temperatures when the controller failed to properly cycle.
An analysis of surface temperature characteristics for the Calrod heater was performed and is found in Appendix C. This analysis shows that the local oil film temperatures and heat fluxes for improper use or impaired Calrod heating elements may exceed the values which are acceptable to maintain oil integrity.
TRID E N T ENGINEERING AS S OCI ATE S. INC.
- k. .f 3.2.2.2 Conclusions High local temperature on the existing Calrod heaters could cause a reduction in oil lubricating properties.
3.2.2.3 Recommendations Although the present Calrod lube oil heating system is ade-quate when properly maintained, Trident recommends that Virginia Power review the present Calrod heater design to determine whether modifications, such as encapsulation of i
heater elements, or redesign, such as an' indirect Calrod steam heater system are warranted. Such a study would con-sider safety, manufacturers' warranties, and site require-ments as well as the desire to further protect the oil from hot spots within the keep-warm system. Measurement of the l
pressure drop across the heater and/or oil flow rates would further ensure proper lubricant flow and oil protection with the present system.
l 3.3 LUBRICANT SELECTION Trident reviewed reports which had discussed lubrication issues associated with the piston pin bushing failures . A letter dated April 22, 1986, from Chevron Research Company, signed by R. E. Crocker, indicated the bushing damage was associated with water contamination and oil film rupture.
Another letter dated May 28, 1986, from Fairbanks-Morse Engine Division, signed by Ed D. Greene, indicated that the TRf DENT ENGINEERING AS SOCI ATES. IN C.
U d bushing failures were caused by lack of lubrication result-ing from an inadequate anti-foaming package used in the Gulf XHD-40 oil. Examination of the failed parts by experienced personnel can allow postulation of the cause of failures.
The lubricant selection process should involve the recogni-tion of the empirical nature of lubrication art. Although certain technical considerations are useful in specifying an acceptable oil for a certain application, the major consider-ation should be the performance of the oil in a similar appli-cation.
During the course of the previous two years, considerable effort was directed to the question of lubricant selection.
Because of the recognition of the empirical nature of lubri-cant selection, good past performance in these and similar engines was the paramount criteria. In August 1985, after selection of Chevron Delo 6000, Fairbanks-Morse confirmed that this oil had performed well in their engines. During subsequent EDG overhauls, all oil was changed to Delo 6000.
Experience to date, including monthly analysis of the oil, have confirmed the satisfactory performance of this oil in the 38 TD8-1/8 engines.
TRIDENT E N GIN E E R IN G AS S O CI AT'_S. IN C.
ld 3.3.1 Conclusion Trident concludes that the Chevron Delo 6000 oil is satisfac-tory for use in the NAPS Fairbanks-Morse 38 TD8-1/8 engine based on the favorable results seen over approximately one year of service at NAPS. During this time period, no failure of mechanical components was attributable to the lubricating oil.
3.3.2 Recommendations
- 1) Continue to use Chevron Delo 6000 lubricating oil. Con-tinue and expand, as discussed herein, the lube oil moni-toring program to identify early any developing problems.
- 2) At this time, Trident cannot recommend the use of syn-thetic oils until adequate engine durability data is available.
3.4 LUBRICANT MONITORING To assess completely the quality of the oil in an engine as it is being operated is impractical and unnecessary.
Certain outstanding and characteristic properties can be assessed quickly and in a practical manner to determine the continued useability of the oil. These are:
ash (sulphated) ASTM method D-874 metals - atomic absorption spectrometry fuel oil-dilution - calculated from viscosity values for the fuel oil and lubricating oil TRID ENT ENGINEERING AS SOCI ATE S. IN C.
l l
l bd
- water - ASTM method D-95 or D-1744
- solide (insoluble in pentane and benzene) - ASTM J method D-893 l 1
l viscosity - ASTM method D-445 '
total acid number and total base number - ASTM i 1
method D-664.
(See Appendix B for detailed discussion.)
During engine operation, samples are periodically taken and analyzed to ensure the oil continues to meet its specifica-tions and requirements. A summary of the items analyzed are
.found in Appendix B.
3.4.1 Conclusions A lube oil monitoring and analysis program, that is adequate with the addition of the recommendations below, has been established at NAPS to monitor the performance of the oil and its additive package.
3.4.2 Recommendations
- 1) Continue to use the lube oil monitoring programs to identify early any developing problems.
- 2) Visual observations should be made when samples are taken to notr abnormal physical conditions, such as oil foam (as entrainment) or oil-water emulsions.
TRID E NT EN GIN E ERIN G ASS OCI ATES. IN C.
Ed The contact of Diesel engine lubrication with water will usually result in some separation of the several chemi-cals which have been added to the oil to improve its performance. Any indication of water contact, either in storage or while in the engine, should be followed up by l
an analysis to determine if the oils composition and performance characteristics have been affected.
- 3) Use engine oil samples taken during operation to provide data regarding water, wear metals, total base number, oxidation, total solids, viscosity, and insolubles for use in trending the lubricant performance.
- 4) Consider collecting additional operating engine oil information, such as oil flow rate, pressure drop across the keep-warm system, and on-line ferrography.
- 5) The condition of the oil should be controlled from the time of its arrival, while it is in storage, and con-l tinually during its use in the engine. Sufficient phys-ical and chemical tests should be made of each lot intro-l duced into the engine to ensure that the oil has the characteristics and properties specified.
TRIDENT EN GIN EERIN G AS SO CI ATES. IN C.
4.0 SCAVENGING AIR SYSTEM A two-cycle Diesel engine must have positive means for forcing an air charge into each cylinder at the end of each power stroke to scavenge out the exhaust gases and to provide air, for combustion. During start-up, idling and low-load opera-tion an exhaust gas-driven turbocharger can not fulfill this function because of low exhaust gas flow rates. Therefore, the Fairbanks-Morse 38 TD8-1/8 turbocharged Diesel engine incorporates a gear-driven, Roots-type positive displacement blower for starting and for low load operation. The blower and two turbochargers are arranged for series-parallel ope-ration, as shown in Figure 4.1.
4.1 SYSTEM DESIGN Room air enters the Roots blower through a set of inlet air filters, and is delivered by large diameter piping to the suction side of the turbochargers. The air from the two turbochargers is combined and then passes through two inter-coolers and into the engine air box. A second air inlet, shown in Figure 4-1, delivers air directly to the turbochar-gers. This inlet is, closed at low and medium power outputs, and opens automatically near full rated power. The two intercoolers are cooled by a separate loop of the engine jacket water cooling system, interconnected to the main sys-tem but having their own circulating pump, heat exchanger and temperature regulating valves.
N TRIDENT ENGINEERING A S S O C I ATE S. I N C.
~--- - -- - -__ _ ____________ _ ___
INLET l AIR FILTER COMBUSTION AIR u
TO ENGINE h h 4
- ( ___q__4
>0 _ ~ _ _ _ _
- - CHECK Z
n f
VALVE z
m 7 E
E i
C ) N
,: s, _
.: 0 M TURBOCHARGERS 9
4 J ROOTS u
'm INLET A BLOWER
, - AIR FILTER
~ ' r SCAVENGING AIR SYSTEM SCHEMATIC t FIGURE 4.1 m
LJ At engine start-up, the Roots blower immediately begins to force air through the turbochargers, through the intercoolers and into the engine. As engine speed increases to 900 rpm at zero load, the pressure of the scavenging air in the engine air box rises to approximately 7 psi. There is not sufficient energy in the exhaust gas for the turbochargers to provide any significant boost in pressure during startup.
At 900 rpm, as load is applied, the turbochargers gradually increase in speed and began pulling air away from the Roots blower and boosting the scavenging air pressure in the engine airbox. Therefore, as the load increases on the Diesel engine the airbox pressure increases and the pressure at the discharge side of the Roots blower drops. At approximately 85-90 percent load the pressure at the turbocharger inlets becomes low enough for the check valve to open, allowing the turbochargers to pull in outside air in addition to the air supplied by the blower. With these conditions, the pressure decreases at the blower discharge and blower horsepower and temperature differential are at a minimum. At the zero load, 900 rpm condition the blower horsepower and air discharge temperature are at their maximum. Air temperatures at the inlet to the intercoolers ranges, typically, from 200 to 265*F, and discharge temperatures from 90 to 110 F, with the higher temperatures at the full load condition. Temperature differentials across the Roots blower range from 135 F at 900 rpm and zero load, to 70'F at full-rated load.
TRIDENT ENGINEERING AS S O CI AT E S. I N C.
LJ 4.2 FAILURE ANALYSIS Failures of Roots blowers have typically been one of three types:
' rotor-to-housing and rotor-to-rotor clearances changed to an out of specified limits condition; rotor-to-rotor contact was made, usually near the mid-length of the rotors; rotor-to-end plate contact was made, usually at the engine end of the lower rotor.
In 1985, the Roots blower internals on engine number lJ were found to be scratched and the rotor-to-end plate clearances were out of the manufacturer's specifications. The damaged blower was replaced with a new unit and sent to Fairbanks-Morse for rebuild.
A November, 1984, Fairbanks-Morse Engines Division Service Information Letter discussed Roots type blower internal heatup and rubbing degradation and attributed the problems to operation at zero load and low engine load conditions.
The Service Information Letter advised that zero load full speed operation be limited to periods of not more than five minutes. In addition, Fairbanks-Morse recommended that the rotor-to-housing clearances be increased from .025" + .004 to .031 .006. The NAPS blowers are presently being system-atically returned to Fairbanks-Morse for clearance increases.
TR ID ENT ENGINEERING AS S O C I ATE S. IN C.
f
'L3 Observation of engines during run-in after overhaul showed that as load was applied to the Diesel engine the turbocharger increased its speed and output, the airbox pressure increased and the blower discharge temperature dropped. However, Trident's observations indicated that a considerable load (in excess of 1000 KW) was required to reduce the air temper-ature differential to less than 100 degrees F. This pre-sents a concern if the engines are started by an Engineered Safety System Actuation and are required to run at zero load for an indefinite period of time, while normal electrical power is available.
Data available to date indicates that there may be movement of the aluminum rotor on the steel shaft (see Figure 4-2).
Possible contributing factors to this phenomena are:
- 1. Design of the attachment of the rotor to the shaft;
- 2. Inlet air restrictions contributing to greater than allowable temperature increases;
- 3. Flow restrictions between the blower and engine resulting in greater than allowable temperature dif-forentials and blower delivery pressures;
- 4. Dirty or non-free-running turbochargers resulting in higher blower pressures and temperature differen-tials at zero load or low load operation.
TR f D ENT ENGINEERING AS SOCI ATES. IN C.
s s s s
%s s s s s s \
N s N \
s s s N
\
.s N \
dB Ns s s s \
w s N
% \___ \
s s s N
\ \
8$ N s N s
.n s s s a s
\
88 '
s N g
\
s s q -
-s
!a- , s s s s s s a s s em s s \
!E s N N N s
?W s s \
\ e g_ 's
\
hk s .-
s s.-- s gm, N ss \ SU l '
s s s N' '? mg m i N N 'd $co 4 I i 43 g l EE o 5
$ h E gm Pn f H l 5 I
~
j m
M 2 d s
TRIDENT ENGINEERING ASSOCI ATEF. I N C.
LA A program which could be used to examine the blower reliabil-ity issue is presented in the following discussion. This l
program includes the performance of an examination of a rotor taken out of service that has shown a significant I decrease in rotor-to-housing or rotor-to-rotor clearances.
This examination would include:
Review as-installed clearances data, rotor-to-rotor and rotor-to-end plate; Measurement of rotor-to-rotor and rotor-to-end plate clearances after being in service; Inspection and measurement of bearings, housings, lock nuts, and fit-up as the unit is being disas-sembled; Detailed mapping of any rotor distortions; Sectioning of a rotor and shaft and examination of the rotor to shaft fit-up, in particular looking for relative movement.
If this examination shows that there is movement of the rotor relative to the shaft, then a redesign should be considered. If no relative movement is found, then a program should be initiated to determine what unfavorable operating conditions are contributing to blower degradation.
Elements of such a program would be:
TRIDENT EN GIN EERIN G AS S O C I AT E S. I N C.
bd
- Very accurate pressure and temperature measurements throughout the scavenging air system taken over the I
full range of operating conditions.
- A thorough examination of the system, with special emphasis on the inlet air section and the section between the blower and engine air box, to identify any elements that may restrict flow or set up pul-sations or standing waves which could have adverse flow effects.
Trident believes that there should be a resolution of the concern that the engine may be operated at no load conditions following engineered safety features actuation with no loss of off-site power, and the engine manufacturer's service letter stating that the engine should not be operated for greater than five minutes at the no-load condition. From the standpoint of protecting the engine, limiting no-load operation is an adequate interim measure but further engi-neering review and analysis is recommended.
________.___-._______I"!""*'"'"""'"^""**''"*
4.3 CONCLUSION
S Trident concludes that the Roots blower is the only potential problem in the scavening air system that might interfere with the EDGs performing their intended service. It is Trident's opinion that the design of the attachment of the rotor to the shaft may allow for the movement of the rotor on the shaft and the subsequent closing of the rotor-to-end plate clearances to the point where rubbing and possible catastrophic failure can occur.
4.4 RECOMMENDATION Trident recommends that Virginia Power consider the develop-ment and implementation of a program, to study blower failure and promote increased blower reliability.
TRIDENT EN GIN EERIN G AS S OCI ATE S. IN C.
5.0 SURVEILLANCE INSTRUMENTATION The Fairbanks-Morse 38 TD8-1/8 Diesel engine is used in a variety of applications. Use of these two-stroke engines to provide reliable standby emergency power within the nuclear industry requires that their design be satisfactory to meet the service requirements. Reliability of the engines, will improve with good maintenance and testing practices since:
Engine component wear patterns are generally the
{l
. result of specific causes which can be determined 5 and addressed.
W Diesel engine monitoring and maintenance is labor intensive and computerized data handling can improve productivity.
- EDG maintenance records can be categorized for refer-ence, analysis, and trending.
An improved awareness of the key components and procedures essential to the EDG operation at NAPS has been observed by Trident personnel on site over the past year. Additional action in this area is in progress.
5.1 ENGINE ANALYSIS INSTRUMENTATION A variety of routine Diesel maintenance activities can now be effectively coordinated using PC computer operated main-tenance programs. These programs provide the maintenance TRID ENT ENGINEERING ASS OCI ATE S, I N C.
l l
l 1
L3 engineer with the tools necessary to easily plan, schedule, track, coordinate, and record all maintenance activities.
Many industries have already moved into computerized control and reporting maintenance programs, including:
offshore drilling industry manufacturing industry chemical processing industry marine industry aviation industry power generation industry Computer-based EDG monitoring programs can accrue several benefits which can contribute to the improved performance of the EDG sets at NAPS including:
on-line and readily available historical data records for key operations; development of an information base that can be exchanged between sites having similar PCs; an ability to easily accumulate, create, and review data and performance for trending.
Preventive maintenance programs are being expanded and estab-lished at NAPS to ensure the future reliability of their Diesels.
TRID ENT ENGINEERING AS SO CI ATE S, I N C.
L3 Sophisticated diagnostic instrumentation is now commercially available that can monitor, collect, and record crucial operational data that is pertinent to the mechanical condition of any operating Diesel engine. Monthly surveillance testing and semiannual fast-start requirements for the EDG units at NAPS require variable or cyclical operation for the major portion of their total accrued run time. This new technology can be used to accumulate valuable performance data during transient and emergency operation of these engines.
Trend data obtained by unobtrusive Diesel diagnostic instru-mentation can be the basis for further upgrading EDG mainte-nance at NAPS. A review of the performance data gathered can be useful in assessing the mechanical performance of the machinery prior to preventive maintenance averhaul and what components, if any, need attention during scheduled mainte-nance. This approach, different from disassembly, review, repair, and reassembly maintenance scheduling programs, could help minimize unscheduled machinery downtime and thereby increase the Diesel reliability. Figure 5.1 illustrates the manner in which the mechanical and thermodynamic events which are occuring in the engine can be monitored and provided to the engineering staff.
TRIDENT ENUINEERING AS S O CI ATE S. INC.
l f
PROCESSOR
[
d 2 PRDITER CONNECTI(31 SOE I
INJECTICM PRESSURE ANPLIFIER MULTIPLEKOR '
TBAllSDUCER COMBUSTICEI l ABI2)
PRESSURE TRANSDUCERS l Q sCAVD GE AIR Maca: ETIC CTL I CYL 2 CYL 3 CTL 4 CYL 5 CYL 6 CTL 7 CTL 8 CYL 9 lTL 10 ::YL 113rL 32 ] PICK-UP DIE SEL ENI INI . G PIS108 POSITICE TYPICAL ELECTRONIC ENGINE MONITORING dYSTEM FIGURE 5.1
\ r T" "
TRf DENT ENGINEERING ASSOCI ATES, INC, ,
- - _ _ _ . - _ _ _ _ _ _ _ _ _ _ _ - _ _ _ l
bd Diesel performance monitoring instrumentation can be inte-grated into existing PC maintenance programs to provide plant maintenance engineers with additional diagnostic data for analyzing and eliminating root causes associated with specific component failures. Specific illustrations of information readily available with the PC computerized Diesel monitoring technology include:
vendor specifications and performance envelopes for EDG comparison with actual operation; pe formance data gathered by appropriate instru-mentation during assurance testing or emergency operation; actual engineering data accumulated for use in evaluating generic Diesel problems;
- accumulated failure history trend data for critical components; projected equipment performance trends from opere-tion history.
I f
TRIDENT ENGINEERINC AS S O CI ATE S. I N C.
L3 Table 5.1 lists the performance parameters for the EDG which should, in general, be obtained during operation to provide a complete description of the engine performance. Trident recommends the following data be monitored:
Pressures
- 2. Cylinder cooling water pump discharge at cooling water header.
- 3. Air cooler water pump discharge and at cooler discharge.
- 4. Scavenging air pressure
- 5. Crankcase vacuum.
- 6. Maximum cylinder pressure.
Temperatures
- 1. Lube oil before and after cooler.
- 2. Cylir. der jacket water temperature after the cooler and at discharge from the engine.
- 3. Air cooler water temperature after the water cooler and air cooler discharge.
- 4. Pyrometer temperature. These temperatures are all accurate if thermocouples are kept clean and wire connections tight.
TRlDENT ENGINEERING AS SOCI ATES. IN C.
L.f TABLE 5.1 PROSPECTIVE ON-LINE MEASUREABLE EDG PERFORMANCE PARAMETERS Engine Component Monitored Parameter Bearing Vibration level Blower Pressure, temperature, vibration level Coolant Pressure, temperature, dissolved oxygen Cylinder Peak pressure, temper-ature, indicated Mean Ef fective Pressu::e (MEP)
Engine Speed, torque, brake MEP, critical vibration levels Fuel Injection rate, injection timing Generator Voltage, amperage, fre-
) quency, wattage, power factor, reactive KVA, vibration Lubricant Pressures, temperatures TRtDENT ENGINEERING ASSOCI ATE S. IN C.
LJ
5.2 CONCLUSION
S Trident concludes that the EDG maintenance personnel at NAPS were aware of the critical nature of their engines. All data recorded in the engine room during a run is manually entered into a computer in the maintenance engineer's office. This data, however, should go directly to a computer electronically from the engine room, thus eliminating the potential for human error. Peak firing pressure is obtained using a mechanical indicator and if the correct procedure is not followed, pressure data could have been inaccurate.
Computer technology for recording and presenting many months of data at NAPS would allow more meaningful decisions concerning service activities and their scheduling.
Complete performance data for the NAPS engines during their operation is not available.
5.3 RECOMMENDATION Trident recommends that Virginia Power consider electroaic monitoring of the emergency Diesels by use of a proven com-mercially available system.
TRlDENT ENGINEERING AS S O CI ATES. IN C.
Ld 6.0 TESTING Trident's engineering analysis of the NAPS Diesel engines revealed three areas where testing issues need to be dis-cussed. These issues are: (1) the long-term monitoring of piston pin bearing performance (2) fast starts and loading during semiannual Technical Specification surveillance test-ing and (3) engine go"ernor effects regarding the potential for engine overloading during surveillance testing. Issues 2 and 3'are discussed jointly in a section because they ere related to surveillance testing.
6.1 TESTING - PISTON PIN BUSHING END CLEARANCE At the time of overhaul of the EDG's at NAPS in November 1985 and March 1986, inspections had identified several fac-tors which had caused or contributed to piston pin bushing failures. Several steps were taken to correct or eliminate these factors, and two circumstances dictated the need for close surveillance of bushing performance:
' various materials examinations had not been completed;
- the possibility that some transient events may have
) resulted in momentary excessive loading or other circumstance which would have contributed to failures.
Consequently, a program to measure the gap between the piston pin floating bushing and the cheek of the piston insert boss TRIDENT ENGINEERING AS S O C I ATE S. INC.
Ld was initiated, so that piston pin bushing elongation could be detected. This program monitors the upper pistons only, since they were involved in almost all of the bushing fail-ures.
6.2 PROCEDURE Gap clearance measurements require that the top cover of the engine be removed, and that lubricating oil be removed from the pistons to expose the upper part of the connecting rod eye as in Figure 6.1. Feeler gauges are then used to measure the maximum gap between the rod eye and piston pin bushing and the insert bushing, as shown in the illustration.
The feeler gauges are brought down to contact with the pis-ton pin to ensure proper measurement. This gap measurement is taken on each side of the connecting rod eye to ensure that burrs or dirt did,not cause erroneous readings to be taken.
As shown in Figure 6.1, the bronze piston pin bushing might be a few thousandths shorter than the steel rod eye bushing.
Consequently, the bronze bushing could begin to elongate slightly without causing any change in gap measurements.
However, observations of bushing failures have indicated that once elongation begins, it continues to the point that the elongation can be measured.
TRID ENT EN GIN E ERIN G AS SO CI ATES. IN C.
FEELER GAGE CCENECTING W '/ / ROD ISTON J I a - PISTON INSERT 3 I I l
l PISTON PIN BOSS l
^ ////W///; P 1
3 hIhk / C 1
/
i ] l l C l J E PISTON PIN BEARING GAP MEASUREMENT ,
7 FIGURE 6.1 7..
g TRIDENT EN GIN E ERIN G AS S O CI ATE S. I N C.
l l
l bd Therefore, the gap measurements do identify the onset of bushing extrusion prior to failure.
6.2.1 Recommendation Trident recommends that the gap measurement procedure be performed once a year or after 40 hours4.62963e-4 days <br />0.0111 hours <br />6.613757e-5 weeks <br />1.522e-5 months <br /> of engine operation, or whenever the engine has possibly been overloaded. The scheduled surveillance interval can be altered based on gap trends.
6.3 SURVEILLANCE TESTING AND GOVERNOR OPERATION The Diesel engines at NAPS are equipped with Woodward gover-nors. These governors are of a hybrid design and consist of a mechanical control system coupled to an electrical control system. During engine start-up the hydraulic control system senses engine speed by means of fly weights and modulates fuel flow. When the generator field is flashed, the electri-cal control system in the governor starts assuming control of fuel flow. The electrical control system senses the fre-quency and output voltage of the generator. During surveil-lance testing, governor operation is particularly important because of the potential for engine overloading.
The NAPS Technical Specifications requires periodic surveil-lance testing of the Diesel generators. Testing consists mainly of two types; tests with " fast starts" and tests with
,s TRIDENT ENGINEERING AS S O C I AT E S. I N C.
bd " slow or soft starts". Fast starts are conducted on 6-month and refueling time intervals and require that the engine be brought up to 900 rpm and loaded in a very short period of time. Slow starts are performed monthly on an alternating or staggered basis and do not require rapid start-ups and loadings. Time is allowed for temperature equalization in the piston crowns and liners so that cracking of these parts is less likely and so that cooling water seals are not sim-ilarly thermally stressed. Generator load is manually increased during the monthly testing to simulate electrical loading of safety equipment. In the past, during the monthly testing, the potential existed for overloading the engines.
Presently, with a dedicated operator observing the generator output, overloading is considered unlikely. In addition, a load limit is established on the hydraulic governor during surveillance testing to prevent engine overloading.
6.3.1 Conclusions Trident concludes that the assignment of a dedicated operator to monitor generator output during testing and the establish-ment of a conservative load limit on the governor are adequate interim measures to prevent engine overloading. Trident also recognizes that Virginia Power has obtained a technical specification change to minimize the deliterious effects of
" cold, fast starts".
TRIDENT EN GIN EERIN G ASSOCI ATES. INC.
Ed 6.3.2 Recommendations
- 1) Trident recommends that, for the long term, the fuel rack stop continue to be set at the 1/2 hour load limit.
For monthly testing, Trident recommends that the load limit on the hydraulic governor be set at the 2000 hour0.0231 days <br />0.556 hours <br />0.00331 weeks <br />7.61e-4 months <br /> load limit of 3000 kw.
- 2) Trident recommends that additional monitoring instrumen-tation be considered which would. annunciate or provide an alarm for overload conditions.
TRIDENT ENGINEERING AS SO C I ATES, I N C.
bd 7.0 FUEL SYSTEM AND CONTROL Each cylinder of the Fairbanks-Morse 38 TD8-1/8 engine has two individual fuel injection pumps and nozzles; one on the control side and one on the opposite side. Figure 7.1 shows an injection pump. The pumps are mounted in a position inverted from the upper crankcase, and are actuated by one camshaft on each side of the upper crankshaft. The injection l
nozzles are located opposite each other in the combustion chamber area of the cylinder liner. A short, heavy steel delivery tube connects each injection pump to its nozzle.
The injection pump assembly consists of a cast steel pump housing with a cam follower assembly at the upper end. Lubri-cation for the assembly is provided by oil splashed by the upper crankshaft. The pump barrel and plunger, and the deliv-
! ery valve, are at the lower end. The pump has variable timing for the start of injection and constant timing for pressure release. Rotation of the plunger, to change the period of fuel injection is effected by a toothed rack in the side of the pump housing. A pointer attached to the pump body and a i
scale on the rack indicates the rack position relative to its zero delivery point.
All of the pump racks are moved by the governor simultane-ously, by means of the control (reach) rods running alongside all of the pumps on each side of the engine. These control TRIDENT ENGINEERING AS S OCI ATES. IN C. 1
- )
RACK POINTER / \
FUEL
- RACK \ _
g @
]ma o i
~
d L
~
PUMP BODY o o i I I I f YOKE lJliin J :(Wj
.g ,
DELIVERY VALVE FUEL INJECTION PUMP FIGURE 7.1 T-.=
i TRIDENT ENGINEERING AS S O CI ATE S. IN C.
u ld rods are connected together by a transverse yoke at the exhaust end of the engine. The connection between a control rod and the 12 injection pumps on a side of the engine is such that when moving in the direction to increase fuel delivery the connection is solid metal-to-metal, and the racks of all 12 pumps move together. However, when moving in the direction to decrease fuel delivery, movement is assisted by a spring. This is a safety feature so that if an injection pump rack or plunger jams while the engine is in operation, the remaining 23 pumps are not prevented from reducing to idle position or zero delivery position. Each pump rack is readily disconnected from the reach rod by means of a spring-loaded knob and snap connection.
he injection nozzles are of the spring-loaded pintle type.
1:xcessive, undelivered fuel drains through jumper lines to a common fuel drain.
7.1 ANALYSIS - INJECTION PUMP The injection pumps supplied for engines in nuclear service are essentially the same pumps which have had a long record of service in naval and commercial engines of the same type.
There is one modification introduced in recent years, and that was the installation of an anti-erosion sleeve in way of the pump barrel port. The bore of the pump casing was TRf D ENT ENGINEERING AS SO CI ATES. IN C.
bd experiencing erosion in the area where the backrush of fuel, at the end of the injection stroke, was impinging on it.
This had not been a problem in steel pump casings in engines rated at lower power levels, but was observed in nuclear l
service. There have been no reported failures due to this i
j erosion, but hardened steel sleeves were installed to pre-L l clude a:.y future problems. All injection pumps at NAPS were rebuilt during the 1985-86 overhauls and have the hardened I sleeves installed. They have also calibrated by the manu-facturer.
Fuel injection pump concerns which have occurred in the opposed piston engines at NAPS and at other locations have been generally in the following categories:
fuel leaks at the seal between the delivery valve and the pump casing; cracking between the stud holes and the bore of the pump casing, at the delivery valve end of the pump; jamming of the pump plunger; fuel leaks at the threaded fittings.
l l
These proolems were attributed primarily to maintenance actions and procedures. The first two, leaks and cracking at the delivery valve area, are frequently related. When the nuts holding the delivery valve were run up on their studs it was very easy to cock the valve body and find, later, TRlDENT ENGINEERING AS SO CI ATES, INC.
Ed that the joint was leaking fuel. This discovery usually was made when the engine was running and the first corrective action taken might have been to try tightening the nuts.
This tightening might not have corrected the original mis-aligned condition and, even worse, the torque might then have exceeded the specified value and caused cracking between the stud holes and the bore of the pump casing. Avoiding over-torquing but achieving a tight joint without leakage required proper training, procedures, care and judgment on the part of the personnel doing the assembly work. Each pump must have been properly shimmed at time of installation or field repair. The importance of this work was stressed to all the station employees involved, and staff engineers monitored performance of the work and equipment.
I L
Jamming of the pump plunger in its barrel could have been
! caused by mechanical distortion of the pump casing, by extremely fine dirt particles, or by corrosion. Most prob-lems of this nature occured with newly installed pumps or after a long period of shutdown. Periodic operation of the l '
engines, carrying 75 percent load or greater for a minimum of two hours, was cignificant in reducing corrosion effects and reducing pump failures. This type of failure has not been a problem in the engines at NAPS.
TRIDENT EN GIN EERING ASSOCI ATES. IN C.
bd 7.2 ANALYSIS - INJECTION N0ZZLE The pintle-type injection nozzles are identical to nozzles in commercial service which provided excellent service over many thousands of hours of operation. However, in EDG ser-vice there were indications of fuel impingement on piston crowns and during engine overhauls, a high percentage of nozzles did not meet the pop test requirements and had to be replaced. It appears that short periods of operation and long periods of standby made these nozzles quite sensitive to water and other contaminants in the fuel oil, causing interference with the free movement of the pintle, corrosion, and carbon buildup on the nozzle tip. This has not been a problem which threatens the load carrying capability of the engines.
7.3 ANALYSIS - FUEL TRANSFER SYSTEM Fuel is pulled from the standby fuel day tank in the engine room and is pumped through the fuel filters and into the injection pumps by a gear-type transfer pump driven from the same shaft which drives the hydraulic governor. A spring-loaded relief valve maintains 10- 5 psi pressure and per-mits sufficient excess flow through the pump so excessive pump temperatures are avoided. The fuel filters are of the disposable element type which removes solids greater than five micron in diameter.
TRIDENT ENGINEERING AS SO C I ATES. IN C.
bd The fuel transfer system has been in use on many Fairbanks-Morse opposed piston engines for over 40 years, and it seldom requires any parts replacements or repairs. Power for the transfer pump is taken from a gear-driven shaft, and all gears and shafting are supported by oil lubricated ball bear-ings. Fuel pump packing, of the braided type, has proven durable and has seldom required any adjustment to the gland nuts. An independent DC electric motor driven transfer pump is provided also, to ensure that the system is primed after being drained for maintenance work or loss of fuel oil pres-sure.
One characteristic which appears in fuel transfer systems where the reserve tank level is below the level of the fuel filters, is that a leak in a filter seal could allow air to enter the fuel lines. If the fuel filters are drained and the elements replaced, the personnel conducting the main-tenance should pump fuel into the filters to ensure engine starting. In installations where the reserve tank is above all of the fuel system elements, priming of all components in the system does not require these special efforts.
TRIDENT ENGINEERING A3 SO CI ATE S. IN C.
- e
7.4 CONCLUSION
The design of the fuel transfer system with the reserve tank level being below the fuel filters allows for the possibility of air blockage of gravity feed of fuel to the engine.
7.5 RECOMMENDATION
- 1) Trident recommends that a design review of the system be performed to determine what, if anything, needs to be done to preclude the possibility of air blockage of gravity fuel feed.
- 2) Trident recommends that normal maintenance practices which emphasize cleanliness, be continued on the fuel injection pumps, nozzles, and transfer system.
- 3) Trident recommends that the practice of making uniform rack settings be continued.
TRf DENT EN GIN E ERIN G ASSOCI ATES. INC.
Ld 8.0 AUXILIARY SYSTEMS During Trident's engineering evaluation, certain auxiliary equipment was reviewed because of the equipment's influence on overall engine performance and reliability. The results of these engineering evaluations are presented below.
8.1 GENERATOR BEARING All four EDG sets at the North Anna Power Station have iden-tical single-bearing generators. These EDG sets have the engine end of the generator rotor supported by a solid cou-pling to the engine crankshaft. The outboard end of the rotor is supported by an oil-lubricated double-row spherical roller bearing. Lubricating oil is contained in a sump, with a sight glass to indicate the oil level. When the sump is filled to the mark on the gage glass the roller train is submerged in oil to a depth of approximately one-half inch.
8.1.1 Analysis An important feature of the bearing installation in the gen-erator is that it is insulated from the generator endbell.
The purpose of this insulation is to provide a break in the rotor-to-engine-to-rotor loop, minimizing any tendency for stray electric currents in the shaft and through the bearing.
Even very small currents through the bearing could cause TRIDENT ENGINEERING AS SO CI ATES. IN C.
Ld surface pitting and eventual failure of the rollers and the raceways.
In evaluating the adequacy of the generator bearing for its
[
role in supporting approximately two-thirds of the rotor weight, the following areas were addressed:
- bearing type, size, and load rating, and
- the lubricant and lubrication system design of the bearing mounting.
These areas are discussed in the following sub-paragraphs.
8.1.2 Bearing Type Design Considerations The advantages of a double-row spherical roller bearing are:
internally self aligning
- high load capacity
- ease of lubrication
- availability of replacement bearings.
Disadvantages are:
lack of axial movement susceptible to corrosion
- easily damaged by dirt.
The necessity for sore self-aligning capability in the gener-ator bearing is quite evident, in that the generator frame could be aligned for the optimum air gap at the field poles without regard to bearing alignment. Where soft metal, split TRIDENT EN GIN EERIN G ASSOCI ATES, !N C.
L d half journal bearings are used, it is necessary to provide a spherically seated bearing shell, and that form of self-aligning system is burdened by frictional forces, greater size, and greater complexity.
Concerning the design of the roller bearing, the dynamic load rating for the 22240 bearing, under Anti Friction Bear-ings Manufacturers Association (AFBMA) standards, is listed 1
as 274,000 pounds. Using the AFBMA load / life formula for roller bearings: ' t 3.13 L = (Cyp where L is life in millions of revolutions to the 10%
failure point I C is dynamic load rating P is load (assume 20,000 pounds, for example)
At 900 rpm, the 10% failure point exceeded 80,000 hours0 days <br />0 hours <br />0 weeks <br />0 months <br />.
Lubrication is achieved by the roller train dipping into a passive oil reservoir, without the need for a circulation system. Further, there is not sufficient heat generation to require the lubricant to serve as a heat transfer medium.
As an added benefit, the lack of an oil circulating system reduces the complexity of the shaft seal inside the endbell.
Grease lubrication could be used, but with somewhat less assurance than with oil. AFBMA rates lubrication severity, TRIDENT ENGINEERING AS SO CI ATES, IN C. ,
l
bd as size and rotational speed increases, using a DN number where D is the diameter of the bearing bore in milimeters, and N is the rpm. A DN of 200,000 is about the upper limit of practical grease lubrication. The bearing in question, at 900 rpm, has a DN number of 180,000, so it is considered to be close to the limit.
Replacement bearings are readily available since numerous manufacturers produce similar bearings in a range of industry specified grades of precision and material quality.
The lack of internal axial clearance is a distinct disadvan-tage. The generator roller bearing and the thrust bearing in the Diesel engine might set up some conflicting forces if the generator frame is not positioned extremely precisely.
A bearing with free end-float, such as a cylindrical roller bearing or a conventional journal bearing, would have been advantageous. However, visual examination shows no evidence of a problem, and the small amount of axial positioning allowed in the bearing housing provides adequate accommoda-tion for the positioning of the generator frame and for ther-mal expansion of the shaft and rotor. Therefore, no problem with lack of internal axial clearance exists at NAPS.
Considering the susceptibility of rolling contact bearings to' damage from dirt and corrosion, EDG service provides a TRIDENT ENGINEERING ASSOCI ATES. INC.
Ed relatively protected environment with very little likelihood of significant amounts of contamination in the bearing. Fur-ther, roller bearings can sustain a substantial amount of pitting and scratching without rapid failure.
8.1.3 Lubricant and Lubrication The lubricating oil used in the NAPS EDG generator bearings is an SAE 40-weight non-detergent oil, which is commonly used in medium and high-speed gears and bearings. This type of lubricant is considered suitable for achieving the load /
life rating inherent in the bearing design, in that it incor-porates a satisfactory combination of viscosity, extreme-pressure additive, and corrosion inhibitor.
The action of the roller train dipping into the lube oil provides sufficient flow to satisfy lubrication requirements and a minor heat transfer function.
8.1.4 Bearing Mounting The bearing mounting is conventional in that the inner ring is an interference fit on the shaft, and is held in align-ment by a shaft shoulder. The outer ring is a clearance fit in its housing. The bearing housing consists of a rela-tively thin, cylindrical steel ring with an insulating ring of a fibrous material between it and the bore of the gener-ator end bell. The end caps and cap screws have insulating TRIDENT EN GtN E ERIN G AS SOCI ATES. INC.
ld spacers and sleeves, so that the outboard end of the rotor is fully insulated from the frame. This type of construction is used frequently in motor and generator designs, and greatly reduces stray currents in the generator-engine shaft assembly.
Two problems can arise from the incorporation of insulating materials into the bearing housing. One problem stems from the low elastic modulus of the insulation, which could permit the steel ring and consequently the outer ring of the bearing to become out-of-round. This out-of-roundness can cause excessive vibration, and in a machine which operates for thousands of hours each year, can contribute to bearing dam-age. In the case of EDG service with its low number of hours operating, this is unlikely. In addition, vibration monitor-ing of these components is conducted monthly. The second problem is ensuring that the integrity of the insulation is maintained.
Since the shaft-to-frame path through the engine represents a short circuit, at the time of Trident's review, there was no positive way to measure insulation resistance except when the generator was uncoupled from the engine, or through other paths. No evidence of any electrical or corrosion damage was found by visual inspection of the assemblies and the electrical grounding is periodically monitored.
THIDENT ENGINEERING AS S O CI AT E S, I N C.
Ed 8.1.5 Conclusion Evaluation of the generator bearing, lubrication, and mount-ing, together with the experience from the large numbers of Diesel generator sets in a wide variety of services, leads to the conclusions that the spherical roller bearing instal-lation has a high degree of reliability and design adequacy.
8.1.6 Recommendation Trident recommends that normal maintenance and surveillance practices, flushing of the bearings and visual inspection at each refueling outage be continued.
8.2 JACKET WATER AND INTERCOOLER SYSTEM There are two interconnected jacket water cooling systems for each of the EDG engines. One system provides cooling water flow through the exhaust belts, through the cylinder jackets, through the turbochargers, and then through a liquid-to-air radiator as shown in Figure 8.1. The second system provides cooling water flow through the scavenging air inter-coolers and then to a separate radiator. Each system has its own engine-driven circulating pump and temperature regulating valves. The jacket water system has a keep-warm feature, consisting of a small motor-driven pump and thermostatically controlled immersion heater to maintain a water temperature of 130-135 degrees F while the engine is not operating.
TRID E N T ENGINEERING ASSOCI ATE S, INC.
ll l' I
.I E
TV O L . gO$@ 1 MA AV ,I
) lX RX AO X C EB G
(
O) 2O$@ ,
M E
l AAV G
l' la N
I 1 L .
L O 8 s u o B O C E I = ~
R R U N E G O - T I I A F SK W NN "
- K MN N C OI S \
' RG FN
%GIN A J
R E E E E l G L O
N E R I O h
V A AC l C
S n /
h /
I l
8N8 hO z* ZS35 >$O ; :u. - 9
b 8.2.1 System Design The radiators are built into the sides of a large closed compartment located at the exhaust end of the engine. A ducted fan, with adjustable pitch on the blades, pulls cool-ing air from the engine room in through the radiators and exhausts it up through a vent above the roof of the building.
This fan is driven by an extension of the lower crankshaft of the engine, driving through an angle gear box in the radi-ator compartment and through a vertical shaft to the fan.
Air enters the engine room through large louvers, passes around a missile-shield wall and then is drawn into the radi-ator compartment or into the engine scavenging air system.
This results in a rapid flow of air through the engine room whenever the engine is in operation. This does not affect the EDG although it can be uncomfortable for workers in the area.
The two cooling systems are interconnected by a jumper pipe between the pump suctions and by connection to a common expan-sion tank. The cooling water is treated with Calgon CS dur-ing the warm months and with Calgon CS plus antifreeze in the winter months. Coolant temperatures coming from the engine and coming from the intercoolers are controlled by THIDENT ENGINEERING AU SO CI AT E D. IN C.
bd two Amot thermostatic valve assemblies which bypass the radi-ators until the set temperatures are reached.
8.2.2 Evaluation - Cooling Effectiveness Flow of coolant through the exhaust belts and through the cylinder jackets readily meet all requirements for cooling at the elevated horsepower ratings for nuclear service. The temperature rise of the coolant is far from excessive, and detailed inspections of numerous exhaust belts and cylinder liners did not show any evidence of hot spots. Examination of the water passages showed them to be relatively clean and free of corrosion and deposits.
Flow of coolant through the intercoolers is adequate. At full load with a water inlet temperature of 80 degrees F, scavenging air enters the intercoolers at 245 degrees F and enters the engine airbox at 90-95 degrees F. This relatively low airbox temperature assures good volumetric efficiency of the engines and consequently good load carrying capacity.
It depends heavily on the cleanliness of the air-side sur-faces in the intercoolers to effect the potential heat exchange. Inspection of these intercoolers at NAPS showed that they remain reasonably clean.
Coolant flow through the turbocharger water jackets results in a reasonable temperature rise and adequate cooling in the TNIDENT ENGINCENING AS SOCI ATES. INC.
l Ed turbos. No definitive temperature and flow measurements are available but evaluation was been made on the basis of hands-on temperature estimates and on the inspection of several turbo rotor assemblies.
8.2.3 Evalutaion - Radiator Effectiveness and Coolant Temperature Control The radiators are deemed to be adequate in that the temper-ature control valves continue to bypass some water flow at maximum engine output and high summertime air temperatures.
The radiators performance depends on maintaining reasonable cleanliness on the air-side surfaces. However, as noted for the case of the intercoolers, air contaminants are a minor problem at NAPS and the dust which did stick in the radiators was easily removed.
Coolant temperature control by means of the Amot thermostatic valves is not a problem for the NAPS engines. Each valve assembly contained six individual thermostatic plugs, so one malfunction has little effect of fluid temperatures. One problem which arose at overhcul time was that recalibration showed that a number of elements had responses slightly out-side of the specified temperatures, and required replace-ment. This did not detract from their capability to perform satisfactorily to meet all service demands.
THIDENT ENGINEERING AHHUCI ATES. (NC.
7 Ld 8.2.4 Evaluation - Jacket Coolant Pumps The pumps are engine-mounted, gear-driven units with pres-sure lubricated ball bearings and mechanical seals. Their service history, in nuclear plants and in a wide variety of other engine installations, show them to be very reliable and durable.
8.2.5 Conclusion Trident found that the performance and design of the jacket water and intercooler system was satisfactory.
8.2.6 Recommendations Trident recommends that normal maintenance practices be con-tinued on the jacket water and intercooler system.
8.3 PISTONS AND CYLINDER LINERS The cylinder liner assembly consists of a cast iron liner with scavenging air ports near the upper end and exhaust ports near the lower end, and with a combustion chamber sec-tion at the mid-point, as shown in Figure 8.2. The bore of the liner has a porous chromium plating throughout its length, for improved wear life. There are four threaded ports in the combustion chamber; two for the injection nozzle adapters, one for the starting air adapter and one for the indicator cock adapter. A water jacket fits over the mid-length of the liner, from just below the scavenge ports to just above
'"'"""'"""'""""'""^""C'^' ' N C '
urran connesessna <w
" b .
l sm- - ,
)
russar --
a
- J_ - l ll w s rut.rr ata roars % -
,J h crunoam s.z:n-.-
f)>
k
{
l e
a L unras ancarr-e: : :
E : : l umersa mora_e .
- 0, l l i
zXMAUST r0 ,61,
', si,
-s -s M m s, nne- '
axxAust asL: A _
-s
)
s fa 0 3 aman erstm . y s
l l
l l
\ L
- m m -is ,
s -
' t m2rs=. ; . _
-_a ;
- C' l
sman cownscttmo : :
aco CYLINDER LINER, TWO PISTONS, AND EXHAUST BELT FIGURE 8.2 . j' To.a.u,,
i T HID E N T E N GIN C E NIN G Atit1D CI AT E S. IN C.
Ld the exhaust ports. O-rings at each end provide the jacket-to-liner water seal.
The pistons have removable inserts which carry the piston pin, as opposed to the type where the piston pin was carried in heavy bosses in the piston skirt. This means that all loading is transmitted from the piston crown directly to the pin and to the connecting rod, without stressing the piston t skirt. This avoids the necessity for having the thick, stress concentrating bosses cast into the piston skirt. The pistons have chromium plating on the crown and a lead-tin flashing on the skirt. Just below the crown are three ccmpression rings and at the lower end of the skirt are grooves for three oil rings.
8.3.1 Analysia - Cylinder Liners Problems which have occurred at NAPS in this type of cylinder liner are:
- 1. loss of the chromium plating and exessive wear *
- 2. chromium chipping or flaking *
- 3. cracking at the port bridges *
- 4. scuffing and scoring *
- 5. leakage at the jacket 0-rings *
- 6. erosion of the adapter threads
- As a result of piston pin bushing failure, j
'"' ""' """'""""'"" ^"" " ' ' ' " " ' ' " " '
L3 This liner design has a long history of operation in many types of engine service. Although the 375 horsepower per cylinder rating for nuclear service is greater than for non-nuclear engine applications, its performance as a reli-able and relatively long-life component is an indication of its durability.
Liner failures in nuclear service, such as cracking of port bridges, scuffing, and scoring have been attributed largely to catastrophic failures of pistons. Disintegration of the chromium in the combustion chamber has apparently been the result of fuel nozzle malfunctions. There have been isolated cases of cylinder liner scuffing and minor cracking at the adapter holes, but in general their performance has been satisfactory.
8.3.2 Analysis - Pistons The two modes of piston degradation at NAPS have been:
cracking of the chrome plating on the crown and crown erosion, and
- scoring and breakup of the piston skirts (after bushing failure).
The crown cracking and erosion was due to fuel impingement on the crown and/or water leakage. Ilowever, the engine /
cylinder hydrostatic testing performed each refueling oatage TNID E N T EN ulN L E MIN G Af180CI AT E E lN C.
LJ coupled with the injection pump and nozzle testing should reduce the probability of these two failure modes. The scor-ing was due to piston pin bushing and insert bushing distor-tion, which forced the piston skirt out of round and went through scuffing to total piston failure. Analysis of bush-ings failures was discussed in a section 2.1.3.
8.3.3 Conclusion There is no indication of any cylinder liner or piston defect which could threaten the ability of the engine to respond to t'e need for emergency power. Consequently, there does not appear to be a need for any additional in-depth analysis of the design or the materials.
8.3.4 Recommendation Trident recommends that normal maintenance be continued on the pistons and cylinder liners.
T H IO 6' N T t: N D I N G. t; H I N G A 91 t10 C I A f t. tl. INC.
LJ APPENDIX A PISTON PIN DUS!!ING ANALYSIS fp410 t N T E N L11 N t. E N I N t3 A t11113 C I A f 8' ll. I N C .
Ld APPENDIX A PISTON PIN BUSHING ANALYSIS This Appendix contains information which supplements the description of the piston pin bushing analysis presented in Section 2.1 of the report. This additional information is presented to explain in more detail the elements of the calcu-lational model including diagrams and pertinent mathematical relationships.
Dynamic Model The load on the piston pin bushing changes in direction and magnitude with respect to time. Therefore, it is necessary to determine the load as it varies over a complete loading cycle.
Three force components act on the piston pin bushing; the gas force, the reciprocating inertia force of the piston assembly, and the reaction force to the rod inertia couple not considered by the two maos substitution model.
Al
'"'""'"""'""""'""^"""""N
T' L3 1. Calculation of Gas Force, Fg A = Piston Area
~
= Pressure of gas in cylinder at crank (Pg)g angle f
Fg = -A (Pg)g l
l
- 2. Calculation of Piston inertial force, P. y L = Connecting rod length R = Crank radius a = Crank angle from firing center (top of stroke)
X = Piston displacement from X=0, top of stroke W = Weight of piston and pin
= W daX Fy g dta Top of l Stroke
. I d ! ! !
l l I l
c c l . .
R COS o I.
~ '
X
/ L -R 2 SIN c4 2
A2 T it lD E N T 8. N O I N . F 84 8 e ll A ft fl d C I A T E ft. I N C .
Ed let X = R_
L X = L + R - R cos a - La-Ra sina a .
t . y
= R (1 - cos a + f 1- 1-Ka sin'a [
dx dt
= R U [sina +
dt k sin 2 a )
l l 2 1-K3 sin 3 a A h = Constant d8X #da) a f K cos 2a K8 sin 8 2a i dta
= R y \cosa 1-Ka sin a a 4(1-K sin a '/2[
a
- 3. Calculation of Rod Couple l We l Port of l Rod Moss l
P T T -
0
' ' ~ ~* 'N
/ P Wp + W
__pis,on a Port 0 A of Rod Mose W = weight of connecting Roo er h,= distance from rod CM to contorlino of crank pin h, a distance from rod CM to contarline of piston pin L= longth of connecting rod (hg + ha )
W h, Ws =
L w,= Wh 3 L
11 = radius of gyration of connecting rod about its contor of mass W = weight of pin p
rniorst cunesd3:su annociartu, eso.
d LJ Consider that the connecting rod is replaced by a two-mass substitution, W 1
and W 2 such that W 1 is at the crankpin centerline, and W is at the piston pin centerline. Th'e 2
motion of W is pure rotation along with the crankpin. The 3
motion of W 2 is pure reciprocating translation along with the piston. Neither W 1 nor W 2 influences the load on the piston pin.
l
. The inertia couple for the completo rod ist wha dag g dta The radius of gyration for the two masses ist Wt hs a + W a h a* = hht W
Thorofore the inertial couple of the two masses ist j
Whhg t da g t The remaining couple not considered by two mass sutstitution is obtained by subtraction:
I W Ha da W h hri dag wr t da6 C =
9 g 9 371
='
9 hh,-Ha)dta 3
The couple "C" may be replaced by the two couples P.A6 (see diagram) and - TR whero T = P con a Thus C = P.K5 - TR
= P (L con 6 + R cona ) - PR con a and P = C/L con 6 The couplo P.X5 acts on the engine framos thorofore, the force "P" must be transmitted through the piston pin.
A4 THIO E N T E N UIN E E MIN G AISMUCI AT E M. INC. l
Ed Therefore:
i da g P.EO =
E 9 [h, t
h - Ha j [( L Lcos cos 8 + R j cosdt8a \ -
The variable 8 may be eliminated and the couple written in terms of crank rotation p,g W ' sin a
,_e(hh-Haf'K(1-Ka)
}- 1 2 , (1-Ka sin 8 a)8/ a K2 (1-K2) sin 2 a fdala 2 (1-Kasin2a) a, (3t[
The analysis shows that engine air receiver pressure has a major offect on the ability of the piston pin bushing to continuously roostablish lubricant films on its lands during the low pressure portion of each compression stroke. At low '
gas pressure, there is a point during the compression stroko where the inertia force of the piston exceed the force gon-erated by the air receiver pressure and the lubricant film is roostablished readily. As load increanos, air rocciver procsure rises and therefore the minimum gas force on the bushing increanos, making lubrication more difficult and the load carrying oil film thinnor. Approaching rated load, the air roccivor prousure is increased at an increasing rato, no that at rated load the savority of the lubrication problem in the piston pin bushing is very sensitive to changos in loading.
AS TNIDENT E N tilN E E NIN tl Af8 tiU CI AT E L INC.
bd The results of the piston pin bearing calculations are presented in Figure A-1 through A-5.
Figuro A-1 shows the piston-pin vector loadings for both the upper and lower pistons throughout an entire cycle (one revo-lution). As can be noted, the engine was being operated at a much higher load than required of the North Anna engines l (415 IIP /cyl-4980 llP total). The curve shows that during firing the portion of the cycle where maximum bushing loads occur, the loadings on the upper and lower bushings are ascentially the samo (64,000 pounds) . The curven do, how-over, show a significant dif ference betwoon upper and lower minimum loadings which occur about 300 degrees after innor contor.' This in the time when the bushing must replenish its lubricant for its next power otroke.
Piguro A-2 shown in general the offect of engine load (dif-forent power outputs; 321, 362, and 415 BilP/cyl) on uppor piston-pin bushing loadings. As previonaly described the data was obtained from existing vendor filos and the test hardware combinations woro not noconsarily the como for each of tho three different power runo. The data in, however, considered unablo in developing an understanding of the basic problemn annociated with lubricating two-cycle pinton pin l A6 THODENT E N talN E E NING A88ttOCI A T E N. IN C.
Fairba n ks -
Morse 8 1/8 x 10 0.P. Diesel Piston Pin Bushing Vector Loads -
Pounds 415 BHP /Cyl -
900 RPM LEGEND L Platon M 60000-
_ U Pistor o.
1 5000o-E g 4 moo-h '
f.r m oo-q 1 ta \
gsoooo- \
\
'***' %-- _ - _ /
0 00 180 270 Lower Cran'< - Dogroes Af ter Inner Conter FIGURE A-1 A7 T NIO L N T E N UIN E E NIN t1 Ata t10 CI A T E M. IN U.
\
Fairbanks -
Morse 8 1/8 x 10 0.P. Diesel Upper Piston Pin Vector Loads -
Pounds 321, 362 & 415 BHP /Cyl -
900 RPM LEGEND
,,,, 321 BHP /C
, _ _ 362 BHP /C
'E 8 socoo. - . 415 BHP /C
, l soooo- ,
b fsoooo-e 8 soooo- , ,
E I to000- ,_
N_ ,
^ ^
0
- O 90 100 270 Lower Crank - Dogroos Af ter Inner Conter FIGURE A-2 A8 TNIDENT E N til N E E N I N () A14 f4 () C I A f t l4, I N C.
l l3 bushings. First, the peak bushing loads are NOT signifi-cantly different for the three power levels. Peak bushing loads increase 6% for a 29% increase in power output.
Mapping the track of the lomi vector onto the perimeter of a circle which represents the piston pin results in Figure A-3, showing the pin loading for a lower piston during the 362 BilP test, which very clearly shows a load reversal per-mitting lubricant replenishment. It also shows the load scans approximately 100 degrees of the pin's surface per-mitting introduction of fresh oil. Figure A-4 shows the upper pin loadings at the 415 BLIP /cyl power level. The load doas not reverse on the upper pin and it scans only about 7 degrees of the pins surf ace as compared to 100 degrees on the low 0r. It is considered that these differences are the key factora explaining the sensitivity of the upper bushings to lube oil quality and power output icvels of the engine.
Figure A-5, showing lower pin loadings at the 415 BilP/cyl power icvel, is very interesting because it shows that increasing the power to this high level reduces appreciably the number of degrees the load scans the lower pin. At 362 DitP/cyl the load scanned approximately 100 degrees of the pin's surface. At the 415 CllP/cyl power level, it is reduced to approximately 7 degrees.
A9 TNieENT ENUINEENING AtlH(IC I A T E H. IN C.
PISTOI PIB 300 310 l
l I
l l
'l J20 i
J90 270 r gg 150' IErfATIces 0
340 ,
- i 00
,80 70 10.000 lbe/la
, 60 350 ,,
i 50 0*i ,
640 i
PEAR 14AD 64,903 0 20' 10 >
< ' 30 l l PISTON PIN LOAD - LOWER SimFT SET FOUR 362 IIP /CYL - 900 RPM FIRING PRESSURE 1434 PSI FIGURE A-3 a10 THIDE N T EN GIN E E RIN G Ata EOCI ATEll. $ NC.
e PISTou PIM g
1 l %
i i
l' I
,r 300 p
- l *H 50 100 340t lohD VECTOR - ,,60 0' 30rATIcer =-
10,000 1he/in 350<> ,g, a
PEAK toAL l 0<. 64,511 4 20'
< >40 l
4, I
4 i
10<
( /
1 1
PISTON PIN LOAD - UPPER SHAFT SET ONE i
415 HP/CYL - 900 RPM l FIRING PRESSURE 1454 PSI FIGURE A-4 All T N,O E N T ENGINCENING ASSOCI ATES. INC.
e PIST01 FIN g .
I I
I l
I
.I i
1 300 315 i 130 220 m 340 >
e 90
,s0 350 ,
a 50 LOAD VEC'!CR AT 40 ',20rATIW 04 10,000 lbs/in 40 e
sl -
I >35 10 FFAK tahD o 30 s4,511 e 20+
PISTON PIN LOAD - LOWER SHAFT SET ONE 415 BHP /CYL - 900 RPM FIRING PRESSURE 1454 lbs/in2 FIGURE A-5 A12 THIDENT EN QlN EERIN G ASSOCI ATES. IN C.
e - -- ,-,-- -, y-- .m-- - , , _ . . - - , . , , , , , - - , , , , , , _ , , , - - ,- ,+wn
m i
L.i APPENDIX B IN SITU OIL ANALYSIS TRIDENT ENGINEERING ASSOCI ATES. IN C.
1 1
I ld APPENDIX B IN SITU OIL ANALYSIS
(
This Appendix discusses in detail the elements of an in situ oil analysis program. Each additive or contaminant to be monitored is listed along with a discussion of the perfor-mance of the analysis.
Sulphated ash: Additive type, heavy duty Diesel engine oils contain metallo-organic compounds such that the quantity of sulphated ash contained in the new oil is directly related to the amount of such compounds contained in the oil. A reduction in such ash content during use of the oil is indi-cative of removal of such additives. Identification of and the contents of specific metals can be determined by ASTM methods 810 and 811. Once the principal additive metals are l
identified in the new oil, any loss of such metals during engine operation means a reduction in oil quality. The oil should be changed if the principal metallo-organic additive content is reduced by one-third from the content in the new oil.
Water: Contamination of oil with water can be most danger-ous. The circulation of free water through the lubricating Bl TRIDENT ENGINEERING AS SOCI ATE S, IN C.
Ud oil system can cause immediate failure of bearings. However, oil and water do not mix well; vater is also more dense than oil, and it evaporates readily from hot oil sumps. Conse-quently, an engine catastrophy can often be prevented by immediately securing the engine if any considerable quantity of free water is inadvertently admitted to the system. Small quantities of water contaminating the oil in the form of an emulsion require securing the engine, emptying, cleaning and flushing the oil system, and replenishment with a complete fill of new oil. An emulsion of water and oil will appear as a light grey to milky colored mixture easily detectable visually. Small quantities of water may unite with some of the oil additives present and separate by gravity to the bottom of oil reservoirs or in filters. Usually, the pres-ence of such combined water can be identified in the ASTM D-95 test of oil samples. If such tests indicate water present above a trace amount, a search should be made prior l to further engine operation to locate and remove the water.
If other oil tests indicate deteriorated oil quality then the oil ahould be replaced with new oil.
Solids: Incomplete combustion of fuel oil, oxidation of lube oil located in high temperature area, such as cylinder walls and piston crowns, result in formation of carbon-rich compounds, such as tars, gums, and resins. Those compounds l
B2 TRf DENT ENGINEERING AS SOCI ATES. IN C.
f bd insoluble in pentane and benzene will tend to plug filters, settle out in oil lines, obstruct oil passages in bearings, coat heat transfer surfaces, cause piston ring sticking and foul the tips of injectors. Additives in the oil tend to mitigate such harmful effects by keeping the carbonaceous material in a fine suspension in the oil sufficiently fine so that much of it will pass through normal oil filters caus-ing the oil to appear quite black almost as soon as it is added to the engine system. Solids contained in the oil that are insoluble in benzene include materials that are almost all carbon in composition and various inorganic sub-stances, such as wear, erosion and abrasion products from bearings, cylinders, piston rings, journals, etc. Micro-scopic examination of the benzene insolubles sometimes gives a good indication of poorly fitting parts which require main-tenance. Such material will normally be removed from the oil by the filter so that an examination of the filtered I solids also will aid in detecting degradation within the 1
engine.
l Viscosity: The two primary functions of an oil are to carry t
away heat and to reduce friction between moving parts, and oil viscosity is of prime importance in performing the latter l function. Viscosity is a measure of the oil's ability to form a liquid film between the moving parts. The higher the I B3 TRlDENT ENGINEERING ASSO CI ATES. IN C.
bd viscosity, the thicker the film, the lower the friction energy loss, and the lower the friction wear of the engine parts. When oil oxidizes in the engine due to high tempera-ture and an oxidizing atmosphere, higher viscosity compounds are usually formed and the oil is said to " thicken". Thus, an increase in viscosity during operation denotes oil degra-dation. Under some conditions, oil degradation is accompa-nied by a reduction in viscosity, but this phenomenon is not an important factor in internal combustion engines. The chief reason for the reduction of viscosity in Diesel engine oil has been the contamination and dilution of the lube oil by the engine fuel.
Dilution and contamination should be closely monitored since these engines make an inordinate large number of starts for the few hours they'are operated per year. During a start the cylinders are receiving maximum quantities of fuel and they have very little air for combustion.
An oil preventive maintenance analysis program has begun at NAPS with Precision Mechanical Analysis, Inc., P. O. Box 735, Branden Florida. Three ounce samples of lube oil are taken after the monthly two hour runs, approximately 5-15 minutes after shutdown for analysis. Samples include oil removed from the crankcase through the dipstick port, l
B4 TRIDENT ENGINEERING AS S O CI ATES. IN C.
LJ strainer, and after the five micron filter. The test program provides the following data to the utility:
- 11 wear metal tests - (particle count)
- non metals count as % by volume
% fuel
% water
% solids pH, total acid number, total base number
- total acid, total base viscosity @ 40*C and 100*C IR analysi,s for water, hydrocarbons, oxidant, nitrogen, and glycol some spectrographic analysis DR ferrography.
l Reports of the oil analyses from PMA have been reviewed by Trident as a part of the ongoing EDG Reliability Improvement l Program.
l l
l l
1 i
l B5 TODENT ENGINEEONG AS S O C I ATE S. I N C.
(
LJ APPENDIX C IMMERSION HEATER ANALYSIS I
l l
l I
TRIDENT ENGINEERING AS S OCI ATES. I N C.
bd APPENDIX C IMMERSION HEATER ANALYSIS This Appendix contains information which supplements the description of the direct contact keep-warm heater system presented in Section 3.2.2.1 of this report. The following material is provided to support the engineering concern about the potential for excess surface temperatures occurring with the Calrod heaters in the lubrication keep-warm system of the engine.
Reliable starting of the NAP's EDGs depends on many factors including the viscosity of the lubricating oil at start-up.
Since oil viscosity increases with decreasing oil temperature, Colt recommends that a keep-warm system be used to ensure a proper oil temperature to maintain proper lubricating oil flow at start-up. Too low a temperature, with its increased oil viscosity, for example, would result in lower cranking speeds and lower oil circulation rates; whereas, too high a temperature would produce too great an oil flow and insuffi-cient lubrication. Therefore, the lubricant is maintained by a keep-warm system to allow rapid oil circulation, to enable the engine to start immediately, and to accept the prevailing load rapidly without detrimental effects en the engine.
Cl TRIDENT EN GIN E ERIN G ASSOCI ATES. INC.
b3 Thermostatically controlled immersion heaters which are submerged in the keep-warm system lubricant maintain a minimum oil temperature during shut-down and out of service periods for these engines as well as provide an adequate oil condition for cold starts. Excessive accumulation of solids around heater elements can cause hot spots to exist which, in turn, can short circuit the heater element. It is also important to prevent the oil from coming in contact with such hot spots which can cause degradation of the oil and/or additives package. Any localized boiling and/or oil coking, produced by unacceptable immersion heater conditions or improper heater operation, can be probable causes of thermal oxidative deterioration of the oil. Since the lubricant system is so critical to engine performance, it is important to minimize any system effect which could adversely affect its performance.
Thermal Analysis Ideally, the rate at which thermal energy is transferred to the oil is fixed by the rate of energy generatio.i within the keep-warm immersion heater element. For steady state opera-tion, an energy balance on a clean cylindrical Calrod heater element would require that the joule heating produced within the heater element be equal to the convective heat transfer to the oil from the heater surface. For a cylindrical heater element, this would equal:
C2 TRtDENT EN GIN E ERIN G ASSOCI ATES. I N C.
b q' = I2Re'L = h(HDg L) (T 3 -
T ,) (1) where q' = heat transfer I = current Re' = resistivity at the heater element L = element length h = convective heat transfer coefficient Dg = Calrod diameter Tg -
Calrod surface temperature T, = oil keep-warm temperature It is apparent from equation (1) that for the basic heater the surface temperature of the clean heater equals T
s
= T, + I gD R' (2)
Any layer of carbonaceous oil residue will act as an insula-tor, and the heat transfer process will, in this instance, involve both contact and convective resistances to heat trans-fer from the heater element. In this instance, the heat transfer rate, determined by the heat generation, requires that the heat transfer equals T
s - T, 9
Rt' ,'c 1
+
HD L 7
hHDg L (3)
C3 TRIDENT EN GIN E ERIN G ASSOCI ATE S, INC.
L3 where Rf'c
= thermal contact resistance per unit area The fouled Calrod surface temperature, T s f, is then equal to l~
f=T. +IHDRe' T _p ,t, c
, g) s, E, f
and the ratio of the fouled surface temperature to the clean Calrod surface temperature would equal I2R
e T
- ta c T HD s, f 1 I2r
- + 1 (5)
T .
T, + HD.
1 Equation (5) can yield values of fouled heater surface temper-atures which can be as high as twice the initial value for ,
the bare heater element.
In addition, the effect of the added coking insulation is to increase the local heat transfer to the oil until an outer insulation radius, r g, is equal to the critical radius of insulation, r , defined as c
k r
c E (6) where k = thermal conductivity of carboneceous insulator layer h = convective heat transfer coefficient C4 TRIDENT EN GIN EERIN G AS SO CI ATE S. (N C.
LJ The radiation heat transfer nature of the carbonaceous soot can also further contribute to an increased rate of heat transfer into the oil from a fouled immersion heater element.
These above-mentioned factors can contribute to the following operational effects on the oil with the use of a direct con-tact heater:
impair the performance of the heater elements and potentially cause the heater element to overheat and foul; loss of control and uniform heating of the oil; oil breakdown by coking and oxidation with potential adverse effects on the additive package.
Direct contact heaters are often used in systems in which heavy fuel oil must be preheated, a situation which is dif-ferent than this critical prelubricant preparation. The use of a Calrod heater to supply heat to the lubricating oil is via a constant wall heat flux mechanism. Dependent upon the flow regimes and heat exchanger performance, the local heater wall temperature and oil film temperature will vary.
Plugged filters or strainers or any impediment to proper lubricant flow can also cause the oil to see temperatures and heat fluxes at the heater surface that may be far in excess of their ideal keep-warm condition.
C5 TRIDENT ENGINEERING AS S O CI ATES, I N C.
(J Proper engine lubrication in the O.P. engines at NAPS requires that the oils be prevented from contacting hot metals in excess of 200*F in the keep-warm system. A further design review to determine whether alternate configurations or modi-fications which could provide a constant wall temperature versus the present constant wall heat flux heater is war-ranted. Alternative configurations could use jacketed heater elements or an indirect Calrod heater-steam system approach to protect the oil. A thorough review of the proposed system by the engine manufacturer should be pursued.
l l
l l
l t
l l
s C6 TRIDENT ENGINEERING ASSOCI ATES. INC.
(
l t
i j
i
?
l i
I t l
l l
l l
TRIDENT ENGINEERING ASSOCIATES. INC.
Forty-Eight Maryland Avenue Annapolis, Maryland 21401 Telephone 301267-8128
__ __. ..___ __._. __ __ _ _ _ _ _ _ . _ . _ _ _ _ _ _ . . . . _ . . _ . - . _ . . . _ . . . . ._.