ML20213A585
| ML20213A585 | |
| Person / Time | |
|---|---|
| Site: | Fermi  |
| Issue date: | 01/28/1987 |
| From: | Agosti F DETROIT EDISON CO. |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| Shared Package | |
| ML20213A587 | List: |
| References | |
| VP-NO-87-0015, VP-NO-87-15, NUDOCS 8702030370 | |
| Download: ML20213A585 (45) | |
Text
!
Frank E. Agosti Nuclear tions Detroit.
d r.m Edison IESEF" W m.
January 28, 1987 VP-!O-87-0015 U. S. Itaclear Regulatory Consission Attn: Document Control Desk Washington, D. C. 20555
References:
1)
Fermi 2 NRC Docket No. 50-341 NRC License No. NPF-43
- 2) Detroit Edicon to NBC Letter,
" Independent Investigation of EDG Bearing Defects," VP-86-0160, dated Noveit>er 21, 1986
- 3) Detroit Edison to NRC Letter, " Detroit Fdison's Response to Safety Evaluation of the Fermi 2 anergency Diesel Generators (EDGs)," VP-86-0145, dated l
October 14, 1986
- 4) Detroit Edison to NIC Letter, "Canpletion of the Fermi 2 anergency Diesel Generator Lubrication Analysis 18 l
Month Trending Program," VP-NO-87-9002, l
dated January 5, 1987 l
l
- 5) Detroit Edison to NBC Letter, " Request for Anendment of License Cordition on Bnergency Diesel Generator (EIG) Lube Oil Surveillance Program," VP-86-0073, dated Septenber 30, 1986
Subject:
Consolidation of Outstanding Issues Concerning the Dnergency Diesel Generators (EDG) Including Proposed Amendment to the Operating License for Fermi 2 Plant
'1his letter is in response to the Deceober le,1986 meeting in Bethesda between the NRC Staff and Detroit Edison personnel on Emergency Diesel Generators. As discussed in the. meeting, this letter provides all the information that is needed to support the proposed license amendment on the EDGs and for clarifying EDG bearing inspection requirements.
}0*l PDR ADOCK 05000341
\\\\1k 8702030370 870128 P
y January 28, 1987 VP-ND-87-0015 Page 2 n is letter also provides Detroit Edison's position on the conclusions and recommendations of < ne studies performed on the EDG bearings. % e Coln Industries report on the Air-Oil Boost Prelube Systen capacity and the Failure Analysis Associates report on the surface scoring of main engine bearings are attached. A discussion on the allowable period of time within which bearing inspections nost be performed is also included.
With this letter, pursuant to 10GR50.90, Detroit Edison Company hereby proposes to amend Operating License No.
NPF-43 for the Fermi 2 plant. W e proposed amendment is to change License Condition 2.C(10) such that previously referenced Detroit Edison letters to the NFC dated Parch 6, 14 and 15, 1985 are replaced by this letter. This will delete the comitment for the cycrterly inspection and subsequent spectrograpiic analysis of the EDG lube oil filter. See proposed wording on page 7 of.
Detroit Edison has evaluated this request against the criteria of 10GR50.59 and 10GR50.92 ard determined respectively that no unreviewed safety question and no significant hazards considerations are involved.
% e Fermi 2 Onsite Review Organization has approved the proposed changes. % e Nuclear Safety Review Group has concurred with the proposed changes based on their review of the original request for a change to the license condition in Reference (5).
% e application fee of one hundred fifty dollars
($150.00), required by 10CFR170.12(c), has been forwarded to the NBC with the preceding letter VP-86-0073, dated September 30, 1986.
In accordance with 10GR50.91, Detroit Edison has provided a copy of this letter to the State of Michigan.
If you have any questions, please contact Mr. Steven Frost at (313) 586-4210.
Sincerely, 0C
~
1
USNRC January 28, 1967 VP-ND-87-9015 Page 3 Attachments:
Sun m y (1)
Background / Safety Evaluation /Significant Hazards Consideration /Divironmental Inpoct/ Conclusion / Change Page from License / Table 1.
(2)
Detroit Edison Response to Battelle and Failure Analysis Associates Reports Conclusions and Reconnendations (3)
Detroit Edison Clarification of NRC Required EDG Bearing Inspections
-(4)
Evaluation of the EDG Air-Oll Boost
]
Prelube Systen (5)
Justification for the Elimination of the Requirement for Quarterly Lube Oil Filter Changes and Subsequent Spectrographic Analysis 4
(6)
Colt Industries Ehgineering Report,
" Analysis of Upper Bearing Lubricant Supply During Fast Starts" (7)
Failure Analysis Associates Report,
" Investigation of Surface Scoring of Main Bearings: Fairbanks-Morse 38'ID8-1/8 Diesels at Fermi 2 Power Plant" cc: Mr. E. Greenman Mr. W. G. Rogers Mr. J. J. Stefano Mr. G. C. Wright Supervisor, Advanced Planning and Review Section, Michigan Public Service Connission i
4 i
January 28, 1987 VP-!O-87-9915 Page 4 I, FRAE E. E OSTI, do hereby affirm that the foregoing statements are based on facts and circumstances which are true and accurate to the best of my knowledge and belief.
E
}
FRANK E. R h!TTI Vice President Nuclear Operations On this M day of d-*%, 1987, before me personally appeared Fr'ank E. Agosti, being first duly sworn and says that he executed the foregoing as his free act and deed.
%Of0J Notary Public MARCM CUCK Notary Public. Washtenaw County, MI My Commission Expires Dec.23,1987 hf [
J G~A
?L y
i
,, ~, -
SLEMARY
%e EDGs at Fermi 2 have experienced several engine bearing failures that occurred since early 1985. In response to these failures Detroit Edison initiated several studies to determine the causes of the failures and identify corrective actions. Detroit Edison also established additional surveillance to provide early detection of failures. n ese actions were taken to establish the root causes of the failures and make changes as necessary to the EDGs, and their operating procedures to avoid further bearing probl es.
Results of studies performed by Detroit Edison and its contractors indicates that the root causes of the bearing problems were inadequate lubrication in the upper crankline during starting, crankshaft misaligrment and improper bearing cap installation.
Measures have been taken to correct these probl es such that there is adequate assurance that the EDGs can reliably perform their intended design function.
Attachment (1) to this letter is the background discussion supporting the change to the License Condition, the Safety Evaluation, the Significant Hazards Consideration, Environmental Impact and Conclusion. A page frm the license showing the proposed change is included, plus a table of all the commitments in the three March 1985 letters.
Attachment (2) provides Detroit Edison's position with regard to the conclusions and recmmendations of the Failure Analysis Associates report and the Battelle report which was transmitted to the NRC by Reference (2).
Attachment (3) discusses the request for clarification of the required EDG bearing inspections, tranmitted by Reference (3),
that proposes allowable time periods within which to perform such inspections while being allowed to continue plant operations.
Attachment (4) presents Detroit Edison's position on the conclusions of the Colt report on the capacity of the Air-011 Boost Prelube systs. A copy of this report is attached.
Discussion of increasing the prelubrication by modifying the Air-011 Boost Prelube systs, or by other means, is included.
Attachment (5) provides justification for discontinuing the quarterly lube oil filter changes and spectrographic analysis of the filter media as requested in this letter. Trending analysis of the data resulting from these inspections and analyses was provided in Reference 4.
Detroit Edison, in Reference (5), requested elimination of the License Condition 2.C(10). 21s letter supersedes Reference (5) and proposes rmoving frm License Condition 2.C(10) the ccumitment for quarterly inspection and subsequent spectrographic analysis of the EDG lube oil filter.
Page 2
'1he License Condition currently references thrce letters from Detroit Edison to the NIC which made carmiitments regarding the Emergency Diesel Generators including the subject surveillance activities. Detroit Edison requests to replace the reference to these letters with thia letter. Table 1 to this letter discusses each of the comitments made in the three letters being replaced to ensure that these ccanitments have been or are being satisfied.
I i
i l
i
ATTACHMENT 1 Background for Modification to License Candition 2.C(IS)
In March 1985, Detroit Edison proposed several corrective actions intended to assure the reliability of the Fermi 2 EDGs. (See Table
- 1) %ese conmitments were incorporated by reference in License Condition (10), in Section 2.C of the Fermi 2 Operating License.
In response to the EDG bearing failures at Fermi 2 in January 1985, Detroit Edison proposed in the March 6, 1985 letter that the EDG lube oil filter be inspected and replaced quarterly, rather than annually. his camitment encompassed visual inspection of a sample of the filter discs to determine whether flakes of bearing material were present on the filter. If metal flakes were found, a bearing gap check would be required. In a ccupanion cmmitment, Detroit Edison proposed spectrcgra@ic analysis of the lube oil filter media. Detroit Edison initiated quarterly inspections of the filters in May 1985. Subsequently, in Novsber 1985 additional failures were experienced. As a result of these events, no significant number of metal flakes were found on the filter discs and nothing in the spectrographic analysis of the filter media would have predicted these bearing failures.
The inspection procedure for the EDG lube oil filter requires opening the filter housing, draining the oil in the filter housing, replacing the filter disc assemblies, refilling the filter housing and returning it to service. %is process entails the risk of introducing foreign material while the filter housing is open and the risk of making other errors while replacing the filter assenbly. It also requires that the diesels renain out of service for several hours.
Experience has shown that, as stated in the Region III Inspection Report No. 85046, this filter inspection and the spectrographic analysis of the lube oil filter media are ineffective in predicting incipient bearing failures. Therefore, Detroit Edison requests that the requironent for this inspection and analysis be deleted from the Fermi 2 license.
Cmmitments made in the Detroit Edison letters of March 6,14, and 15, 1985 were supplemented by cumitments made because of Detroit Edison's EDG Reliability Demonstration Test Program and response to the safety evaluation report related to the EDGs. Dece cm mitments are shown in Table 1.
to VP-NO-87-0015 Page 2 Safety Braluation ne proposed change to the Fermi 2 Operating License was reviewed against the criteria of 10CPR50.59 and found to involve no unreviewed safety question.
Is the probability of occurrence or the consequences of an accident or malfunction of amtr==t important to safety previously evaluated in the safety analysis report increased?
No. % e EDGs are the only equipment involved in the proposed chanse. The quarterly lube oil filter inspections were proposed to augment the EDG surveillance program until additional data were obtained. Detroit Edison hoped that increasing the frequency of filter inspections would provide additional information that might reveal an incipient bearing failure. Subsequent bearing failures indicate that the filter inspections and analysis do not provide this information. De elimination of this ccuraitment does not deprive the Fermi 2 operators from any useful data regarding potential diesel bearing failures. Furthermore, it decreases the probability of introducing foreign material into the lube oil systen while the inspection is performed and the probability of errors while replacirg the filter assembly. Also, by decreasing the time the EDGs are inoperable for filter surveillances it increases the time the EDGs are available.
W e proposed change to the license condition is considered strictly an administrative change and does not affect accidents or malfunctions evaluated in the Fermi 2 FSAR. %e mechanics of tracking and inplementing conunitments at Fermi 2 renains the same.
% erefore, the proposed changes do not increase the probability of occurrence or the consequences of an accident or malfunction of equipnent inportant to safety previously evaluated in the safety analysis report.
Is the possibility for an accident or malfunction of a different type that any evaluated previously in the safety analysis report created?
No. Experience has shown that the more frequent lube oil filter inspection and spectrographic analysis do not reveal incipient bearing failures. De EDG Reliability Dmonstration Test Program and other measures described in the March 18, 1986 (VP-86-0025) letter provide assurance that the Fermi 2 EDGs are a reliable source of alternate standby power. Elhninating the requirment to inspect and replace the lube oil filter to VP-!O-87-0015 Page 3 quarterly and to analyze the filter media, recognizes the low value of the more frequent filter inspection and analysis.
Furthermore, it does not affect plant operating conditions, the way in which any equipment is operated and does not create any new failure modes.
S e proposed change to the license condition is considered administrative and does not create unanalyzed accidents or malfunctions. Controls for implementing these canitments are in place and they are not affected by the change.
%erefore, the proposed changes 60 not create the possibility for an accident or malfunction of a different type than any evaluated previously is the safety analysis report.
Is the margin of safety as defined in the basis for any tecimical specification rart - d?
No. We lube oil filter inspections and the spectrographic analysis have been shown to be unreliable in detecting incipient bearing failures. W ey were proposed by Detroit Edison and accepted by the Staff to detect bearing failures and 4
improve engine availability. We deletion of the inspection and analysis has no effect on the margin of safety of the Fermi 2 EDGs as defined in the bases of the EDG-related Technical specifications.
We proposed change to the license condition has no effe::t on the margin of safety.
- ~,
~,. -
to VP-NO-87-0015 Page 4 Simifir ar* nawards Onnairbration ne proposed change to the Fermi 2 Operating License was reviewed against the criteria of 10TR50.92 and found to involve no significant hazards considerations.
h proposed change does not involve a significant increase in the probability or consequences of an accident previously evaluated.
%e quarterly lube oil filter inspections were proposed to augment the EDG surveillance progran until additional data were obtained. Detroit Edison hoped that increasing the frequency of filter inspections and spectrographic analyses from annually to quarterly would provide additional information that might reveal an incipient bearing failure. Subsequent bearing failures indicate that the filter inspection and analysis do not provide this information. The elimination of this ccmnitment reduces the probability of introducirg foreign material into the lube oil systen while the filter housing is open, and the probability of other errors while replacing the filter assenbly.
n e proposed change to the license condition is considered strictly an administrative change and does not affect the safety analyses presented in the Fermi 2 FSAR. The mechanism that Detroit Edison utilizes for tracking and inplementirg its ccanitments remains the same.
n erefore, the proposed changes do not involve a significant increase in the probability or consequences of an accident previously evaluate:1.
h proposed change does not create the possibility of a new or different kind of accident from any previously evaluated.
Experience has shown that the more frequent lube oil filter inspection and spectrograpiic analysis do not reveal incipient bearing failures. The EDG Reliability Denonstration Test Program and other measures described in the March 18, 1986 letter provide assurance that the Fermi 2 EDGs are a reliable source of alternate power. Eliminating the requirement to inspect and replace the lube oil filter quarterly and to analyze the filter media, recognizes the low value of the more frequent filter inspection and analysis.
to VP-NO-87-0015 Page 5
%erefore, the proposed changes do not create the possibility of a new or different kind of accident fra any previously evaluated.
'the proposed change does not involve a significant reduction in the margin of safety.
S e lube oil filter inspections and the spectrographic analysis have been shown to be unreliable in detecting incipient bearing failures. %us deleting the requirment for quarterly inspection and analysis has no effect on the nargin of safety in the operation, surveillance and naintenance of the Fermi 2 EDGs as defined in the bases to the EDG-related Technical Specifications. % e proposed deletion of the license condition does not alter inplementation of commitnents and necessary follow up. W e tracking and follow up mechanism for ccumitments remains the same. Thus, changing the license condition has no effect in the margin of safety.
to VP-NO-87-0015 Page 6 Ehvironmental Inglect Detroit Edison has reviewed the proposed Operating License changes against the criteria in 10CFR51.22 for environmental considerations. As shown above, the proposed changes concern maintenance procedures and instructions and elimination of tests.
'Ihey do not involve a significant hazards consideration, nor increase the types and amounts of effluents that may be released offsite, nor significantly increase individual or cumulative occupational radiation exposures. Based on the foregoing, Detroit Edison concludes that the proposed amendment meets the criteria given in 10CFR51.22(c)(9) for a categorical exclusion frcm the requirement for an Environmental Impact Statement.
GM2msHBI It is concluded, based on f.he considerations discussed above that by implementing the proposed changes no unreviewed safety question and no significant hazards considerations are involved. Fermi 2 will still continue to conply with all applicable regulations.
Furthermore, it is concluded that: (1) there is reasonable assurance that the health and safety of the public will not be endangered by operation in the proposed manner, and (2) such activities will be conducted in cmpliance with the Canission's regulations and the issuance of this amendment will not be inimical to the ccnsnon defense and security or to the health and safety of the public.
l l
l l
Attach ent 1 to VP-NO-87-0015 FERMI 2 OPERA' TING LICENSE Page 7 CHANGE PAGE 5
procedures and measures described in Section 9.S.1 and Appen-dix E of SSER #5 and #6 shall continue to be implemented, including removal of power from the Division Igooling tower bypass valve (No. E1150-F603A) and from either'the single series valve (No. E1150-F008) in the reactor heat removal (RHR) system or the two parallel RHR suction valves (hos.
E1150-F608 and E1150-F009) during normal plant operation until the independent alternate system is declared opera-tional.
(10) Emergency Diesel Generator Lube Oil Surveillance Program (5ection 9.5.7. 55ER 75)
DECO shall implement its comitments regarding the surveillance program for the lubricating oil system of the emergency diesel-generators as described in its letterg dated (P. arch 6 March IL and March'15,1985D Jan ary 27,1477.
[
(11) Low-Pressure Turbine-Disc Inspection (Section 10.2.2. SER)
Deco shall perform an inspection of the low-pressure turbine-discs during the second refueling outage, including volumetric examination of the disc base using ultrasonic techniques. The frequency of subsequent inspections shall be in accordance with the turbine manufacturer's recomendations.
(12) Retention of Persons with BWR Operating Experience on Shift (Section 13.1, 55th #5)
~
At all times the plant is in an operating condition other than cold shutdown or refueling. DECO shall have a Ifcensed senior operator on each shif t who has had at least six months of hot operating experience on a similar type plant, including at least six weeks at power levels greater than 20 percent of full power.
and who has had start-up and shutdown experience. For those l
shifts where such an individual is not available on the plant staff. Deco shall provide an advisor who has had at least four l
years of power plant experience, including two years of nuclear
{
plant experience, and who has had at least one year of experience on shift as a licensed senior operator at a similar type facility.
Use of advisors who were licensed only at the reactor operator level or who otherwise do not fully meet the criteria for shift advisor, will be evaluated by the NRC staff on a case by-case basis, As a minimum, Deco shall train these advisors on the procedures. Technical Specifications and plant systems for the Femi-2 facility and Deco shall examine the advisors *.on these 1
topics at a level which will assure their familiarity with the i
plant. For each shift, the remainder of the shift crew shall i
l i
_ _ _ _ _ _ _ _ _ _ _ _ _ to VP-NO-87-0015 Page 8 Table 1 EDG-Related C<mnf %ts Reference Ccannirment Remarks NE-85-0460
- 1) Following prelubrication, prior Coupleted (March 6, 1985) to planned starts, the prelube punp will not be secured until the EDG reaches full speed (900 RPM). After the EDGs are modified to permit slow idle speeds securing the punp will be allowed when engine reaches idle speed.
- 2) After 20 starts have been Supplemented conducted without prelubrication by letter but not longer than 18 months VP-86-0099 the gap check will be performed.
(7/24/86)
- 3) Lube oil filters will be inspected Propose to and replaced quarterly. ne delete discovery of a visible acctanulation of bearing material will trigger a gap check.
- 4) Monthly oil sanples will be drawn Continuing from the sunp while the engine is consnitment for operating and will be analyzed monthly sanp-for metal concentrations. he ling, 18 month results will be trended for 18 trending com-months.
pleted; report issued in letter VP-NO-87-9002 5)
Detroit Edison will perform a Propoce to spectrographic analysis of lube oil delete filter media.
_ _ _ _ _ _ _ _ _ _ _ _ _ to VP-!O-87-0015 Page 9 NE-85-0459
- 6) A gap check is performed after 20 Clarifies (March 14,1985) dry starts but not longer than 18 Iten 2 above month intervals. If gap is greater than recmmended tolerance the bearing will be replaced, or manu-facturer consulted. If manufacturer reconmends actions the engine will not be started until actions are cm pleted.
- 7) One additional criterion for perform-Clarifies ing gap checks is the sighting of It s 3 above.
flakes of bearing material on the Propose to filter. 'Ite sanple will consist of delete a mininum of two discs from each of the seven filter stacks. Greater than an average of one fragment per disc for EDG 13 and 14 and two per disc for EDG 11 and 12 will initiate the inspection.
8)
Monthly oil samples will be taken Same as Itens while the engine is operating and 4 and 5 above.
the sanples will be analyzed for Propose to metal concentrations. In addition delete second a spectrographic analysis of sentence.
material deposits of the lube oil filter media will be performed.
- 9) Detroit Edison will establish a program A trending to evaluate and trend the data for use study of 18 in prediction and detection of nonths data bearing failure.
of lube oil has been com-p]eted and a report issued in letter VP-NO-87-0002
- 10) Detroit Edison will perform a study of Coupleted.
air boost prelube systen to ascertain See Attachnent if inprovments can be made to to this letter accomplish the systan's intended objectives.
NE-85-0462
- 11) Until the modification for slow Modifications (March 15, 1985) starting the diesels is inplenonted,
- cmpleted, the diesels will be accelerated therefore, not automatically within the 10 applicable to second time period to synchronous slow start speed and then the inachine will be mode.
loaded slowly per the manufacturer's recaamendations.
_ _ _ _ _, _. _ _ _ _ _,. ~ _ - _ _ _ _.. _ _ _
,y
,,,y
- = _ -. -
, _ -. to VP-ND-87-9915 Page 19
- 12) During surveillance testing due to Continuing loss of offsite power or loss of ccanitment diesel the other diesel will not be synchronized with the grid under conditions of grid instability to assure plant safety is not degraded.
I 1
i l
l 1
i l
l 1
Detroit Edison Response to Battelle And Failure Analysis Associates Report Conclusions and Recommendations A
I f
l I
4 L
i ts VP-ND-87-0015 Page 1 Detroit Edison Response to Battelle And Failure Analysis Associates Reports Conclusions and Recomendations he Battelle report was prepared in response to a comitment doctmented in a Detroit Edison letter to the NRC, Region III, VP-NO-86-0025, dated March 18, 1986. %e comitment was to have an independent laboratory analyze several bearings, with different conditions noted on each, to determine the cause of the surface conditions. % e bearings had been removed fr m the engines during the Novenber-Decenber,1985 inspections.
Failure Analysis Associates was retained by Detroit Edison to assist in evaluating problems with the EDG bearings. %e FAA report provides their findings and conclusions based on their analysis of the bearing problems. ne report is being transnitted by this letter to the NRC as a courtesy.
Detroit Edison's Positions on the Battelle Report
Reference:
Final Report on a Metallurgical Investigation of Various Defects on the Surfaces of Diesel Engine Alumin-Tin Bearings to Detroit Edison datal August 31, 1986 by R. D. Buchheit and W. A. Glaeser, Battelle Columbus Divisions, Columbus, Ohio.
Several bearing sanples and lube oil samples were renoved frm Detroit Edison's Fermi 2 EDGs and submitted to Battelle in Columbus, Ohio for metallurgical investigation. At the time when the bearing sanples were sutnitted, Battelle was also provided a general overview of the EDG application, but were not provided with engine design and maintenance procedure details. % is fact becomes evident in some of the report conclusions and recomendaticna which v re made to eliminate certain bearing surface conditiont Swestigated out which are not feasible due to engine design and EDG vendor recomended maintenance procedures. The specific points are identified below.
Following are Detroit Edison's positions on the Battelle Report conclusions:
1.
Battelle
Conclusion:
"% e condition of surface distress was apparently caused by inadequate lubrication and excessive frictional heating that led to adhesive wear metal-transfer damage to the bearing surface."
Epsition: Detroit Edison is in agreenent with the conclusion as stated. Detroit Edison came to the same conclusion during its' investigation of the surface distress condition. Failure Analysis Associates (FAA), a consultant to Detroit Edison on EDG bearing problems is also in agreement with the conclusion.
t
.- to VP-ND-87-9915 Page 2 2.
Battelle
Conclusion:
" Dull-rough and dull-smooth surface conditions most likely were caused by an excessive level of contamination by iron-rich particles and lapping conpound particles."
Position: Detroit Edison is in general agrement with this conclusion and believes that the majority of dull-rough / smooth surface conditions were caused by the lapping compound particles.
W e source of the iron-rich particles was residue frm construction of the engines. n e dull rough / smooth condition was observed predminately on EDGs 13 and 14 which were equipped with the original factory installed bearings.
%ere are two sources of lapping coupound particles that can cause the dull-rough / smooth condition. The first source is fra the crankshaft journal lapping procedure used by the manufacturer on new crankshafts or fr m re-lapped journals when replacment bearings are installed in the field. The lapping procedure requires cleaning the lapping compound frm the journal area, however sme abrasive particle residue is expected. % e second source whicn causes the bulk of the blemish, is from the manufacturer's procedure to apply a ccating of "bcaring conditioner" (which contains lapping ccupound abrasive particles) directly to new bearing surfaces for purposes of breaking-in new bearings. W e manufacturer has evaluated the lapping compound abrasive particles used in the bearing conditioner and has revised their specifications to use a lapping compound abrasive particle size more suitable for use on aluminum bearing surfaces.
Under normal EDG operating conditions no excessive level of contamination has existed in the EDGs as verified by lube oil analysis. Detroit Edison plans to continue monitoring lube oil on a routine basis to ensure contamination free EDG operation.
Furthermore the dull-rough / smooth bearing surface condition was not related to the bearing failures experienced by the EDGs.
3.
Mtt_e.13e conclusion:
"Einbedded iron rich particles were apparently cast-iron particles fr m the engine castings."
Position: Detroit Edison is in agreement with the conclusion with the following clarification. W ere are few cast-iron parts (crankshaft, pistons, etc.) in the EDG engine, the basic construction is welded steel plate. W e iron-rich particles identified in the report were pralminately frm weld metal splatter rather than cast iron parts. Battelle is in agrement with this clarification.
(% is was a case of where Battelle was not knowledgeable of engine construction details.)
i to VP-ND-87-0015 Page 3 4.
Battelle Conclusionr. "%e bearings, examined in their present stage of bearing surface degradation, except that of " surface distress", were not considered to be significantly damaged and thus not to be in inninent danger of failure, but failures of the bearings would be expected to occur eventually if the apparent level of particulate contamination was allowed to persist."
1 Position: Detroit Edison is in agrement with the conclusion with the following clarification. Batte11e's statment "... but failures of the bearings would be expected to occur eventually if the apparent level of particulate contamination was allowed to persist" is based on their investigation of the
" dull-rough / smooth" bearing samples. This statment could lead to the false conclusion that the lube oil syst s is contaminated. n e dull-rough / smooth blemish condition is due primarily to the use of bearing conditioner on new and replacement bearings during installation. At this time the concentration of abrasive particles in new or replaced bearings is high during initial engine operation. However, the bearing conditioner is flushed into the oil as the lube oil passes through the bearings. n e total quantity of bearing conditioner used (less than an estimated pint for all bearings) is insignificant when the engine lube oil system capacity of approximately 550 gallons is considered and that the oil is filtered before it is returned to the bearings. Lube oil cleanliness has been confirmed by lube oil analysis for contaminates. Also, the Battelle Analysis of the used lube oil samples concluded that "No evidence of abrasive contaminates, such as silica or aluminum silicates were found in the oil sanples."
5.
Battelle
Conclusion:
"No metallurgical conditions of the aluminum-tin bearings that contributed to the damage of bearing t
surfaces were identified."
Position: Detroit Edison is in agrement with this conclusion.
Failure Analysis Associates also confirmed this conclusion during their evaluation of the January 1985 bearing failure.
6.
Batt.elle_
Conclusion:
"No unusual characteristics of the used lubricating oil sanples were identified."
Position Detroit Edison is in agreement with this conclusion.
It should be noted that subsequent problems were encountercd when the formulation of the EDG lube oil was changed by the oil supplier.
nese probles were corrected when Detroit Edison switched to Mcbil oil after conducting a survey of nuclear plants that have Fairbanks-Morse IDGs. Mobil oil was found to be the most popular. % 1s was discussed in a separate NRC
, to VP-NO 9915 Page 4 stbmittal and is unrelated to the bearing and oil sanples provided to Battelle for their investigation.
% e following paragraphs provide Detroit Riison's position on the three reccanmendations made by Battelle.
Battelle_feccannendation (1):
"Inprove the method used to renove lapping coupound from the journal surfaces and the general cleanup procedures."
Position: W is is a case where Battelle has unaware that lapping compound is a constituent of the bearing conditioner which the EDG manufacturer reccsumonds for use whenever new bearings are installed. We lapping conpound observed was most likely a residue frcan the bearing conditioner and not from lapping conpound used on the crankshaft journal. The tranufacturer's maintenance procedures call for cleaning lapping cc1 pound after use on crankshaft journal surfaces, but does not require renoval of all traces. This is because lapping compound would again be re-introduced when bearing conditioner is applied to new bearings. W e use of bearing conditioner as specified by the EDG manufacturer will be continued.
Battelle Recmunandation (2):
" Consider the use of steel-backed trimetal bearings with a babbit overlay to replace the solid aluminum alloy bearings".
Positiont %e primary purpose of this reconsnendation is to consider the use of a babbit overlay on the bearing surface.
Detroit R11 son's position is the same as stated later in this attachment for the Failure Analysis Associates conclusion concerning the use of a lead overlay bearing.
Battelle Recmsnandation (3):
" Maintain a continuous recirculating, pressurized lubrication systen to prevent corrosion in stagnant areas of the bearings and to assure adequate lubrication during rapid start ups of the diesel engines."
Eosition:
In 1982-1983 Detroit Riison installed modifications to each of four EDGs to provide continuous circulation of heated lube oil to all parts of the engine except the upper crankshaft and installed an air / oil booster systen to enhance Iubrication to the upper crankshaft bearings during autcznatic fant starts.
21s is another case where Batte11e's unfamiliarity of the detailed design of the orposed piston engine resulted in a nound roccanendation, but one that is not feasible. W e opposed piston design does not allow for continuous circulation of oil through the upper crankshaft bearings. Detroit E:dison's nonthly surveillance test runs of each EDG will circulate the lube oil sufficiently to minimize the potential for corrosion.
to VP-NO-87-0815 Page 5 As part of the corrective action from previous bearing problems Detroit Edison has revised all appropriate diesel operating procedures to require manual prelubrication before initiating any planned EDG start. This, in conjunction with the routine EDG surveillance test runs, meet the intent of Battelle's recormnendation for a continuous recirculation and pressurized lubrication systen.
__ to VP-ND-87-9915 Page 6 S e following are Detroit FAison's positions on the Failure Analysis Associates (FAA) Report reconnendations.
1.
FAA Rect==norris Elimiriatina the Use _oL Abranive Bearina Conditioner Detroit Ediann Position: ne diesel manufacturer has evaluated the use of abrasive particles, trade name Timesaver " green label", introduced with bearing conditioners used to break-in crankshaft bearings. Based on this evaluation the manufacturer has changed their formulation to provide use of a " yellow label" conpound in lieu of the " green label" previously used. %e
" yellow label" is reccamended for aluminum (the bearing naterial) whereas the green label is reconnended for nodular iron (the crankshafts material). W e use of green label causes a bearing surface condition referred to as " dull amooth" or
" dull rough". % is surface condition has not been linked as the cause to any bearing failures and is only a coanetic concern.
4 Detroit Edison will comply with the manufactures maintenance procedures which require the use of the revised " yellow label" bearing conditioner.
2.
FAA Rec-rids IncorDoration of an Electroplata] nahhit Bearina Overlay ontp_the_DeAring.
Retroit raisen Position
%e EDG manufacturer has previously experimented with this concept, and have a nunter of concerns.
One concern is in the area of appearance where plated bearings have the appearance of being bad as they wear to acconnodate slight differences in bearing / journal geometry. %is aptrarance condition could lead to prenature bearing relacements. ne second concern is that prchlemo have been experienced in developing a suitable method to al' ply a uniformly thin overlay with good adhesion characteristics.
ne bearing problems experienced by the EDGs resulted frce inadequate lubrication, crankshaft misalignant and inproper bearing cap installation. A lead overlay bearing would not ime I
prevented these types of problans, and therefore, Detroit Edison does not intend to pursue the developnent of an overlay bearing design.
3.
FNLBece== rx1s Decrenina the IHLYalue of the Shaft Microfinish.
Detroit rAlaon Position
%e EDG Manufacturer's curront design calls for a 19 microinch BMS finish on the crankchaft journals after polishing. After the polishing proccos has been completed, the crankshaft journals are then lapped to climinate the possibility of rucared material (i.e. nodular iron gra[hite
Attaclunent 2 to VP-NO-87-0015 Page 7 nodule caps) from breaking loose in operation. As a result of the lapping process, the final crankshaft journal finish is approximately 25-30 microinches, RMS. The FM recommendation is to develop a procedure to return the journal final surface finish to the 10 microinch RMS range.
Detroit Edison agrees, in principle, with the concept however, does not plan to pursue this for the following reasons.
a) The process, to our knowledge could not be done in the field and, therefore, only applies to ncw crankshaft requirments on a renewal parts basis. No further crankshaft replacements for the EDGs are expected during the useful life of the plant, b)
In-service crankshaft journals do achieve a high degree of final surface finish fran normal wear as the machines are operated.
c) The crankshaft journal surface finish was not a significant factor in the cause of the EDG bearing prob 1ms experienced in the past.
4.
FAA Recmends _a Possible Redesian_ of the Thrust __ Bearinc _t.o Decouple the Heavv_ Finnaes frce the Radial Portion of the Bearina.
Detroit Edison.lpsition: The EDG manufacturer has given considerable thought to this approach but has been unable to develop a design which can be acccrmodated by the existing engine configuration. A redesignal thrust bearing, incorporating this concept, thus would require considerable modifications to the current design, and would require an extensive testing prcgram to ensure reliability for nuclear plant applications. tkuertheless, Detroit Fdison would be very interested in incorporating a new thrust bearing design into the Fermi 2 EDGs if Fairbanks-Morse can bring it to fruition.
5.
Eliminati_rgMutUppsprarguy Circymferential 011 Groove irtthe
'1hrustJcAr.ing.
l l
Detroit nihen Position: The intent of this proposal is to raluce the thrust bearing running surface distortion resulting from thernal expansion and the preload placed on the bearing.
'Ihe EDG nunufacturer has designed a thrust bearing eliminating the unnecessary oil groove in the loadtd thrust bearing half r; hell. The manufacturer has placed an order for a test quantity of these bearings and plans to run these in a laboratory engine for testing. Following the laboratory engine tests the bearing design will be released for testing in a field application.
Upon successful ccepletion of the testing program and when the rctiesigncd thrust bearing becomes available for nuclear plant applications, Detroit Fdison will install the ncw design bearing as a rcncwal part on an as-necded basis.
Detroit Edison Clarification of NBC Required EDG Bearing Inspections
- _ - t@
g VP-PO-87-0015 Page 1 Detroit Edison Clarification of NRC Required EDG Bearing Inspections s
%e NBC's letter to Detroit Edison entitled " Safety Evaluation of the Fermi 2 anergency Diesel Generators (EDG)," dated July 16, 1986 stated that the EDG Denonstration Test Progra had provided reasonable assurance that the EDGs will function as a reliable onsite power source. It also required that a gap check of the crankshaft main bearings be performed after every unprelubricated start.
In addition, a special visual inspection of the #13 upper thrust bearing of EDG
- 13 was specified to be performed after reaching the envelope of slow or fast starts as defined by the EDG Demonstration Test Program.
~
,n Detroit Edison requested a clarification of the allowable period of time within which to perform the inspections. % e requests were:
(1) seven days to perform the bearing gap check after an unprelubricated EDG start occurs, and (2) thirty days to perform the special visual inspection of the EDG 13 upper thrust bearing after the envelope nunber of engine starts is reached. A further clarification requested was that the EDGs would be considered operable until they are removed 3'
from service to perform the actual inspection.
% ese requests were made merely to allow the Fermi 2 operators the flexibility to plan for the inspections, conduct the in an orderly manner, at the best time possible and taking into consideration the status of the plant.
Allowing seven days to perform a gap check would prevent imediate plant shutdown if one or both EDGs per division experienced an auto (dry) start. Past i
experience indicates that numerous dry starts can be sustained before bearing failure will occur. Allowing the EDGs to be considered operable up to the performance of the gap check would, therefore, not conprmise the safety of the plant since two independent divisions of onsite emergency power would be maintained.
Similarly, allowing 30 days to perform the EDG #13 upper thrust bearing disassably and inspection would prevent imediate shutdown of the plant when the end of the envelope, established by the EDG Denonstration Test Progrm, has reached the limit of 10 fast starts or 30 slow starts. Detroit Edison has never experienced a failed thrust bearing on any EDG. Allowing 30 days to perform the inspection, while considering EDG #13 operable, would not comprmise plant safety as two independent divisions of onsite mergency power would be maintained until the inspection. W e 30 days would also allow time for the NBC to be present at the inspection.
No significant objections to these requests were voiced by the NBC at the s
Dec mber 10, 1986 meeting with Detroit Edison. %e NRC stated tiet Detroit Edison could consider the requests granted until formally notified.
5 r
-mr.
--- -~---- - - - - - -, - -
9 9
W N
^-
A httachnent 4
- s Emergency Diesel Generator (EDG) Air-Oil Boost Prelube Systen Study 4
4
,4 K
a Y
4 m
(
l I
h"
'e a
e
_ to VP -M)-87-9915 Page 1 nrergency Diesel Gerierator (EDG) Air-Oll Boost Prelube Svsten Study In SER Supplement No. 5 the NIC Staff reccamended that Detroit Edison study whether is should increase the reservoir tank volume of the EDG oil booster /accurtulator systen to sane appropriate volume larger than 1.25 gallons.
Colt Industries, the engine manufacturer, performed a study of the capacity of the systen as presently installed. A copy is included with this letter. We results of the analysis are as follows:
1.
During the first one second time interval fran the start signal, the booster cysten provides oil to the bearings, but it is at essentially zero pressure. ne booster does initially fill the power assenbly passageways with oil; 2.
It takes approximately three seconds for the engine driven main lube oil punp to fill the remaining enpty systen passageways and to develop pressures at the main bearings.
%e analysis indicates that the booster systen initially fills the enpty passages with oil at essentially zero pressure. % e only potential benefit of increasing the capacity would be if it were able to pressurize the oil. %e ctudy also shows that increasing the volume of the booster systen would simply continue the same flow rates for a longer time and would not necessarily pressurize the lube oil systen.
While the air boost prelube systen does not, independently, prevent dry starting of the EDG upper crankline, the systen does contribute to engine prelubrication by initially distributing oil to all the oil passages. Detroit Edison has a thorough program of maintenance and inspection to assure the continued reliable operation of the EDGs. %e air boost pre-lube systen is not an essential ccmponent for continued reliable EDG operation but does provide some positive benefits.
Detroit Edison has requested that Colt analyze two possible options available to enhance the existing 1.25 gallon air booster design. ne two options are (i) to install an auxiliary pre-lube ptanp which would begin ptmping prior to engine starting, and (ii) to increase the size of the air oil booster volume to three g llons.
For option (i) Colt Industries nodekd the affects of a 200 gpn and 100 gpn cuxiliary air or dc pre-lube punp. neir analysis shows that a 200 gpn punp would produce a minimum of 8.3 psi pressure to the urper crank 11ne bearings and fill the voids within 2.3 seconds. A 100 gpn ptmp would produce a minimum of 3.6 psi oil pressure and fill the voids within 4.7 seconds. Additional tine is necessary for the punp to start, reach rated speed and pressurize the bearings.
W e total delay to lubricate the bearings would be 4.8 seconds for the 200 gpn punp and 8.7 seconds for the 100 gpn punp. ne ergine driven main lube oil punp to VP-NO-87-0015 Page 2 pressurizes the bearings at three seconds. D ere appears to be no gain from an auxiliary pre-lube punp unless the engine start is delayed. Se present 10 second starting time is established as one of the conditions in the design basis accident (10GR50 Appendix K) analysis. W erefore, Detroit Edison does not plan to pursue this option.
ne model for the 1.25 gallon booster shows the oil is delivered to the upper crank line bearings for about one second of the 3 second time interval before the main lube oil pung takes over. For option (ii) Colt has calculated the effects of the three gallon booster. During the first second the booster is filling the voids. During the second and third seconds a slight pressurization is indicated at the bearings, however, it is not a significant inprovement.
n e air / oil booster is a design enhancement, but is incapable of making a significant contribution to inproved lubrication during auto starts. This is because no significant oil pressure can be produced by the cysten regardless of the quantity of oil available. A majority of the nuclear plants using the same model EDGs as Fermi 2 do not incorporate any air / oil boost systen. Detroit E%lison will not close the door to future inprovements in the ait/ oil booster system, however, should they become available frctn the EDG vendor.
i
it 4
[
Justification for Elimination of the Requirement for Quarterly Lube Oil Filter Ganges and Subsequent Spectrographic Analysis 4
4 f
-r--
-.n-----..-.
- -,,_ -, --=
to VP-NO-87-0015 Page 1 Justification for Elimination of the Requirment for Quarterly Lube Oil Filter Changes and Subsequent Spectrographic Analysis License Condition 2.C (10) presently contains a requirement that the Emergency Diesel Generator (EDG) lube oil filters will be removed, disassembled for inspection and replaced quarterly. In addition sanples of the filter media will have a spectrographic analysis performed. %e intent of the filter replacement and analysis was to attempt prediction of EDG bearing problems. Detroit Edison has requested the license condition be revised based on (1) the ineffectiveness of the analysis to predict bearing problems, (2) other methods in place to monitor bearing condition and (3) the potential for introducing debris into the engine lube oil systen during the filter changes.
It was stated in the NRC Special Inspection Report (No. 50-341/854046(DRS))
dated July 3, 1986 that the filter inspections had faile0 to predict the inpending bearing failure.
At the December 10, 1986 meeting with the tmc in Bethesda, Detroit Edison reviewed the request for elimination of lube oil filter changes and subsequent spectrographic analysis. The basis presented at the meeting was that the filter inspections had not been a reliable indication of incipient bearing failure.
Additionally, oil analysis is being performed on a nonthly basis and the frequency of gap checks has been increased. Also, Fairbanks Morse reccanends quarterly filter inspections without changing the filters. 'Ihe NBC requested the data be provided that indicated the filter inspections had not been a reliable indicator.
Detroit Edison's letter VP-MI-87-8002, dated January 5,1987, provides the results of the first 18 months of lube oil and lube oil filter analysis for the EDGs. Tables 2-1 thru 2-4 of this transnittal are attached. As shown in table 2-3 the level of Aluminum found in the filter on Novenber 5,1985, just before the Novenber 13,1986 EDG 13 bearing failure was very low. In addition, table 2-1 shows that the analysis did not indicate on August 14, 1985 and Decenber 18, 1985 any increasing trend in aluminum although the EDG-11 bearings were found to be damaged on Decenber 3,1985. However, in both instances numerous metallic fragments were observed prior to renoving the filters from their vessels. The disassenbly of the filter media and subsequent spectrographic analysis does not appear to be any more effective then visual inspection of the exterior of the filter media.
to VP-NO-87-0015 Page 2 Fairbanks Morse has revised their reccamended maintenance inspections to include a quarterly lube oil filter inspection. Detroit Edison plans to follow this reccanendation as part of the preventive traintenance program. ne action criteria in our procedures, to perform gap checks of the main bearings if excessive particles are found to be deposited on the media, will be maintained.
Also, the filter elements will be changed whenever excessive differential pressure exists or during the planned refueling traintenance on the engines (ncminal 18 month intervals) whichever comes first. No further spectrographic analysis of filter media will be performed.
ne lube oil analysis program is also continuing as outlined in VP-NO-87-0002 and is tracking Aluminum concentration in the engine lube oil. If given thresholds are exceeded, bearing gap checks are requiral. nis progra requires sampling at every monthly surveillance test run period.
When the original comitment for lube oil filter changes was made, gap checks of bearings were performed at 18 month intervals. If bearing particle thresholds were found to be exceeded during filter inspections, the corrective action required was a gap check. Since we have increased the frequency of gap checks to every 6 months, filter inspections are to some extent a repetitive activity.
A final point is that during the process of changing the lube oil filters a potential exists to contaminate the " clean" lube oil side of the filter and piping with unfiltered oil or debris. his potential is minimized if only a visual inspection of the filter media exterior is performed.
In summary, the justification for the deletion of the filter change requirements are:
1)
We media internal inspections and analysis has not proved to be any more effective in monitoring bearing condition then visual inspection of the filter media exterior.
2) ne filter housing will be opened and the media exterior will be examined on a routine basis.
3) ne frequency of the bearing gap inspection has been increased from 18 months to six months.
4) he lube oil monitoring progra is continuing on a monthly basis.
5)
Eliminating unnecessary filter changes will reduce the potential for introducing contamination in the lube oil systen.
to VP-NO-87-0015 Page 3 Table 2-1 EDG 11 Lube Oil Filter Sample Analysis Data Sample Deposited Metals in micrograms (by CTC)
Date Fe Cr Sn Al Si Zn Notes 5/8/85 19 0
0 59 10 5
8/14/85 11 3
6 81 8
9 12/18/85 22 6
21 74 16 1
1 1/27/86 57 11 67 291 43 7
2 4/5/86 31 7
33 18 20 6
9/11/86 7
0 0
6 4
4 Notes: (1) Some bearing distress observed in late December 1985 (2) Metallic particles were observed inside filter vessel.
l Table 2-2: EDG 12 Lube Oil Filter Sample Analysis Data f
Sample Deposited Metals in micrograms (by CTC)
Date Fe Cr Sn Al Si Zn 5/8/85 18 2
4 71 12 0
8/14/85 15 1
1 123 13 0
12/13/85 24 2
6 80 26 3
9/18/86 4
1 0
5 2
2
~
_ to VP-NO-87-0015 Page 4 Table 2-3: EDG 13 Lube 011 Filter Sample Analysis Data Sample Deposited Metals in micrograms (by CTC)
Date Fe Cr Sn Al Si Zn Notes 6/7/85 17 4
1 109 20 1
11/5/85 16 2
3 13 44 8
1 1/22/86 33 4
72 248 18 9
2 9/25/86 94 33 146 25 126 60 Notes:
(1) Upper connecting rod bearing (#3) fracture November 13, 1985 (2) Metallic particles were observed inside filter vessel.
l Table 2-4: EDG 14 Lube 011 Filter Sample Analysis Data j
l l
Sample Deposited Metals in micrograms (by CTC)
Date Fe Cr Sn Al 31 Zn 7/28/85 29 7
7 18 90 54
.11/7/85 28 8
8 14 100 35 4/17/86 139 56 28 47 63 1
/
9/4/86 150 22 54 73 72 8
i, -
l
MTPGMENT 6 OIT D00STRIES ElGINEERIlE REPGtr l
1 w-w-m wm--v w
w'=
" ' - =
"vs
. _.. ~. _. ~. -
~
,c, ENGINEERING REPORT con rit.E NUMBER R-5.08-501-001 EMERGENCY DIESEL GENERATOR, DETROIT EDISON C0.,
ATE 8/25/86, Rev. I sus 1EcT FERMI 2 NUCLEAR GENERATING STATION PREPARED sY W. A. Brill ANALYSIS OF UPPER BEARING LUBRICANT SUPPLY "fr#Y' DURING FAST STARTS t
APP VED Introduction The subject engines have shown some history of trouble with their upper main bearings.
This has been associated with the requirement for rapid start-load sequences without prelubing the engine.
As part of their continuing program to improve this situation, Detroit Edison has contracted with FMED to conduct the investigation and analysis reported herein. (DE Contract NS-318586; FMED Sales Order 071918).
The goal of this investigation has been to detennine the time development of oil flows and pressures at the upper main bearings commencing with the engine starting signal.
An analytical flow model of the engine and skid was prepared.
This includes all important flow paths, and pumps, filters, etc.
Also included is the auxiliary pneumatic booster system.
Calculations of the behavior of this model ~were conducted by utilizing a commercial, general purpose computer program called FAAST (Fluid Analysis And Simulation Technique).
Conclusions 1.
The derived computer model of the engine, skid, and booster lube oil system accurately represents the real system, based on a comparison of computed results to test data.
2.
During the first one second time interval, the booster provides oil to the bearings, but it is at essentially zero pressure.
The booster does initially fill the power assembly passageways with oil.
3.
It takes approximately three seconds to fill the remaining empty system passageways and to develop pressures at the upper main bearings.
~
ENGINEERING REPORT PA*ZE *
~,
not NUMBER R-5.08-501-001 EMERGENCY DIESEL GENERATOR, DETROIT EDISON CO.,DATE 8/25/86, Rev. 1 s m tcT FERMI 2 NUCLEAR GENERATING STATION ParPARES. A. Brill sy REPORT uTLE DURING FAST STARTS APPROVED BY Discussion The FAAST computer program is a general purpose hydraulic analysis progra which is commercially available through the Control Data Corp. "Cybernet" time-sharing service.
and flows to match specified boundarThis program iteratively balances the system pre a network of _ elements (pipes, pumps,y conditions.
The system is specified as Fluid properties, element properties, boundary values and job co
_ nodes.
statements must be input to the program.
The program calculates quasi-steady state results.
This is believed to accurately represent the flow developments in the engine, as these occur over the relatively long time 'of a few seconds.
The pressure wave velocity is estimated to be 900 ft/sec, and longest acoustic path perhaps 50 ft out and back.
Since such waves are expected to be pretty well decayed after 4 or 5 reflections, such transients would only last on the order of 1/4 second; thus, validating our quasi-steady assumption.
The model of the engine-plus-skid system is sketched on page 8.
sketch of the auxiliary booster system is on page 9.
A separate numbers, and a few labels are included in these sketches. Node numbers, element include the following:
Features of the model 1.
The main engine-driven pump is a gear-type pump (flow is primarily a function of speed), for which measured data was inputted.
2.
Flow out of the system occurs mainly as coolant at the pistons.
Other outflows occur at upper and lower main bearings and conn rod bearings, and at the several " spitter" type cooling and lubricating lines to various gears and accessory bearings.
3.
The rotating crankshaft acts as a centrifugal pump on oil flowing through it to each connecting rod.
4.
Elevations are specified only for nodes at which output flows and/or pressures are to be calculated.
It is noted that calculated heads /
pressures are not severely affected by elevation; 1 psi = 31 inches of oil column, while low point to high point of the system is only 100 inches overall.
r'assfLMenMS morse
~
PATE ENGINEERING REPORT c.nio n.in..
C RILE NUMBER R.5.08-501-001 i
EMERGENCY DIESEL GENERATOR, DETROIT EDIS0N CO.,
8/25/86, Rev. 1 sarra FERMI 2 NUCLEAR GENERATING STATION PREPARED av ANALYSIS OF UPPER BEARING LUBRICANT SUPPLY W. A. Brill REPORT TITLE DURING FAST STARTS APP ED 5.
The input by the booster tank is modeled as a constant pressure, which agrees with actual tests.*
6.
For purposes of this investigation, the booster system and the engine-skid system models were not actually connected.
The separate models cause no real problem since the two effects act essentially sequentially, rather than simultaneously.
The models were left separate to reduce the solution bandwidth and computing time.
In retrospect, this was unnecessary in view of the modest 2 to 3 minute computing times actually incurred.
The accuracy of the models was checked by comparing computed results with two sets of experimental data; -(a) booster flow to main bearings on a six-cylinder engine,* and (b) steady state, full engines during manufacturer's tests., speed (900 rpm) operation of the Fenni 2 SAE 40 oil was used in the tests, and corresponding viscosity values were used in the calculations.
Tests of the booster system were conducted on a six cylinder engine of the same model as Fermi 2.
The tests were with room temperature oil.
The booster tank pressure reportedly " jumps to over 90 psi immediately after opening the air starts.
It remains at that pressure for 1.5 seconds and then after emptying, it drops to zero."
Total volumes of oil were also measured at each main bearing.
The computer program was run with 55*C (131*F) oil and 100 psi booster pressure, which are reflective of the keep-warm conditions of Fermi 2. This resulted in a flow rate in element 2 of 250.99 cu.in/sec, so the time to empty the 257.5 cu.in tank is computed to be, t = 257.5 + 250.99 = 1.03 sec.
In view of the slightly higher pressure and lower viscosity, this is in reason-able agreement with the measured 1.5 sec.
Volumes at each bearing are shown in the following table.
(The model was trimed to 6 cylinder by closing a valve at element
- 110, and bearings are renumbered in the table.)
Agreement between measured and calculated values is within 8 to 15%.
- Weekly Report R5.08-000, 8/9/82 Weekly Report R5.08.000, 8/16/82 t Contract test logs dated 5/16/75, 6/12/75, 6/20/75, and 7/23/75
N=*s Mme
~
PA:E ENGINEERING REPORT
"* 4 con,
RILE NUMBER R-5.08-501-001 EMERGENCY DIESEL GENERATOR, DETROIT EDISON C0.,
8/25/86, Rev. I sus:Ea FERMI 2 NUCLEAR GENERATING STATION PREPARED BY W. A. Brill REPORT ANALYSIS OF UPPER BEARING LUBRICANT SUPPLY TITLE DURING FAST STARTS APPROVED BY Six Cylinder Element Measured Calculated
- Bearing No.
Number Volume (cu.in)
Volume (cu. in.)
1 106 26.9 31.2 2
96 27.3 31.3 3
86 29.0 31.5 4
76 29.5 31.7 5
66 29.5 32.1 6
56 30.3 32.5 7
46 30.3 33.2 8
36 37.1 33.9
- For example at element 36, Y"
Fo e
ostdrVolume of booster =
x 257.5 = 33.9 Averaged data for the Fenni 2 engines running at 900 rpm at steady state were taken from the engine test logs. This is compared to the computer model in the following tabulation.
The oil temperature for tests and calculations was the same (77 C=170*F), and test pressures are averages of ranges actually measured.
Again good agreement is observed.
System Location Model Location Measured Calculated Main pump flow (cu.in/sec)
Element 208 2074.
2077.5
)
Pump discharge pressure (psi)
Node 410 74.
77.095 Pressure to filter (psi)
Node 424 72.5 75.501 Pressure from filter (psi)
Node 426 70.
72.207 Pressure to strainer (psi)
Node 456 53.5 57.219 Pressure to engine (psi)
Node 464 48.
51.117 Pressure to upper header (psi) Node 700 33.5 36.726
Fairbanks Morse
~~
PA2E ENGINEERING REPORT
"" 8 c ~..
g rn.E NUMBER R-5.08-501-001 ATE EMERGENCY DIESEL GENERATOR, DETROIT EDISON CO.,
8/25/86, Rev. 1 sumiEc7 FERMI 2 NUCLEAR GENERATING STATION PREPARED BY W. A. Brill REPORT ANALYSIS OF UPPER BEARING LUBRICANT SUPPLY TITLE DURING FAST STARTS APPROVED BY The computer models are next 'used to analyze cases for which no test data exists.
As in the earlier calculations, viscosity values corresponding to SAE 40 oil are used.
The first such case is that of the booster system acting on a 12 cylinder (14 upper bearing) engine, with oil at keep-wann temperature (55*C = 131*F), and with 100 psi booster pressure.
These parameters are reflective of the Fermi 2 engines in the keep-warm state. The computer results are sumarized as follows:
Booster flow rate (node 98) = 271.03 cu.in/sec Time to empty booster tank = f = h 3= 0.95 sec.
Twelve Cyl Element Flow Rate Total Volume Bearing No.
Number (cu.in/sec.)
(cu.in) 1 166 17.570 16.7 2
156 17.616 16.7 3
146 17.709 16.8 4
136 17.849 17.0 5
126 18.036 17.1 6
116 18.270 17.4 7
106 18.553 17.6 8
96 18.885 17.9 9
86 19.267 18.3 10 76 19.775 18.8 11 66 20.446 19.4 12 56 21.311 20.2 i
13 46 22.413 21.3 14 36 23.326 22.2 i
l
(
MO h
~
PA!E '
ENGINEERING REPORT c.n sawm, g
rn.t Numara R-5.08-501-001 EMERGENCY DIESEL GENERATOR, DETROIT EDIS0N C0.,
ATE 8/25/86, Rev. 1 sustrcT FERMI 2 NUCLEAR GENERATING STATION PREPAntS. A. Brill av nEPORT Tm.t DURING FAST STARTS APP VED At the time of start signal, t = 0, the upper end of the jumpers, the fitting to the main bearing cap, the cap itself, and the bearing groove are all empty.
These have a calculated volume of 7.3 cu.in. per bearing.
If one adds the empty passages in the crank, rod bearing, and conn rod, the total empty volume is 22.2 cu.in.
Noting this figure equals or exceeds the oil supplied per bearing, it is concluded that this initial oil is available to the bearing, but at essentially zero pressure.
The booster does help to initially fill empty passages that otherwise would have to be filled with flow from the main pump.
continue the same flow rates for a longer time. Increasing the volume of the booste The second case for the computer model is actually a series of cases at one second intervals during the time (8 sec.) the engine is accelerating to full speed.
Data from such a fast start test gives the following correspondence:
time (sec.)
0 1
2 3
4 5
6 6
8 speed (rpm) 0 45 145 275 405 630 655 775 900 Calculated pressures and flows at the upper bearings are sununarized on the graph on page 10.
All of these cases are run with the oil at keep-warm temperature (55 C = 131*F).
Main pump and " crank centrifugal pump" parameters were adjusted to reflect the various engine speeds.
In interpreting the results of these cases, it is important to remember the following. The top of the filter (volume = 125 cu.in), the supply pipe from the lower transverse header to the upper header (volume = 350. cu.in.), and the upper parts of the upper main jumpers (4.75 cu. in. each) are initially empty of oil.
The program presumes they are full, so the pressures calculated are not in fact produced until the flow has filled the void.
For example, during the first 3 seconds, the total flow is calculated
- to be:
V = f x 141.47 x 1 + f (141.47 + 416.57) x 1 + f (416.57 + 75
= 936, cu.in.
- The flow rates from the main pump at t = 0,1, 2, and 3 seconds are 0, 141.47, 416.57, and 756.54 cu. in./sec. respectively. The flow ra tes are presumed to be linear functions of time between these points.
~
PA"E '
ENGINEERING REPORT c.= w n.*
g rn.E NUMBER R-5.08-501-001 EMERGENCY DIESEL GENERATOR, DETROIT EDIS0N CO.,
ATE 8/25/86, Rev. I suarEcT FERMI 2 NUCLEAR GENERATING STATION PREPARED my W. A. Brill ANALYSIS OF UPPER BEARING LUBRICANT SUPPLY E#dE DURING FAST STARTS APP VED Thus, considering flows throDgh the lower rod assemblies, it is concluded that close to 3 seconds will have elapsed before the system is filled and the calculated pressures developed at the upper bearings.
Detroit Edison has measured modest pressures during this time interval, but those results followed a two minute prelube of the engine which would fill the voids.
There is no time for such a prelube cycle during a normal fast start.
While it was not part of the scope of this investigation, a brief discussion is made here of the conditions required for successful bearing operation.
There must be some minimum acceptable pressure which is a decreasing function of the length of time during which it acts.
to determine, " adequate" is not.
While the " minimum" is difficult and full speed shows those conditionsSuccessful operation at normal steady state to be adequate.
Similarly, prelubing the engine is known to create conditions for gcod bearing life.
Finally, the Failure Analysis
- Assoc, report
- calculates that an adequate oil film thickness of 100 microinenes will be maintained with a supply pressure of 3 psi at 300 rpm and approximately 2 psi at 100 rpm.
These values are calculated with 130*F = 55*C oil.
" Investigation of Engine Bearing Distress in Fairbanks-Morse Emergency Diesel Generators at Enrico Fenni II Power Plant", April 1985, pages 5 and 11.
roupennS Morse ENGINEERING REPORT PASE *
[]
FILE NUMsER R-5.08-501-001 EMERGENCY DIESEL GENERATOR, DETROIT EDISON CO.,
DATE sus EcT 8/25/86, Rev. 1 FERMI 2 NUCLEAR GENERATING STATION PREPARED ANALYSIS OF UPPER BEARING LUBRICANT SUPPLY REPORT
'av W. A. Brill DURING FAST STARTS TrrLE APPROVED t
av
. ENGINE - SKID
}
Ls/SE D/L Flow MODEL C
.3 Est **dsv r A*ro Mo bt nWMef4 *NG v
y bp l.
f Y
I 4
N,";, :.
?
"",$'I y
t a
i i
le i 3', $
s i
e v, 8 g
3 3
g'i (
l l
I i
i i
,.I 3 M
(
i
=*-
.W..,.. i;.. u,gE i E!E E353 1
3 a
R 3
E
'- =
- u,4,m, arcs 7 sse
- <~ :
3*
re sI i.i I EII I!EF 1
K s,.
- in
. o s, ss.
n "F
f,,,
5 :s
,s= g. es. en.
.e,-
- s<n
~
f r 4
tra leer
,a g sos
( l g ***
.t s.. 3.1,,,3,! ! !
$252 5
g 5
.<s s,sl
- h I
~
a
-se i
- 4,c
m*
5 u, a... m
.u m.- - " "
c,e
,,e 5
I I
I
,,Em e..I I k b!E
,es ss.,
2 E
h rs, u,,,,n m....E I
I m
- ~,~
'4 im
..u n ~s 5
E F
E E
Elf
,cil
,,11 g *=
2 m
III3 i
y E
i u...,
m g.a au
.u c
- m'~
}
test _
7f a a.<
ou am
,*u -- * ' '
~~
a a
n A
sb a
, # e, re m
.. g. =
.u-Ey R
g' m.,,, a m a,.s e t e
e I I ;w5 E
m
'= ~ w j
Im 55=
0**
.u.u. u
.u.g m > > <
7,ar ss e
3 aEr!
ss!1 a4 3
I a
2 o,
.,,, u, q s.<, m, m
.o-
< > =
- U
.,,. 4, q m a5 det d
=
4 3
I I
EI$
b!3 5 d
S m.,,, )m y$
99E see M
- f,
,,,q,.,,
.. q,,a,aq m
- e. = w A
14,7
%fes
=
,a
~
3 9
E E
E EII kIII I
E E
- a..
ses l
=c w:,,,
s,.
u,
,,, m g au
,, uc
- ~, ~
24sr m
~
44
,a
~
,., 5.....,, g 5 iti EliE l
E E
?
1 ges res9
{
- =
et,
,.c
- , = ' ~
5tes t
isat
. =,. a'
?
~
oot i
E E
E E
EIE IT3-3I I
I I
3
,.s sss l
V"
"< w..u et.
n n
.sc
< <. ~, ~ '
n am
.="
y1 a'
I E
E I
E EEE t * ~,
e I$
rs. an u,
m m a eu as-
- * * * =
s.,. a q.$,I,,, I I
I
=
a r,.
48 8 s24 El s
Gr9
- d9 es sos tas1 f
a 4,,,
"'\\ v i,,a a r
t' 12e it, u:;
3 O=
y s
e
., as m
Esos g
"a =
I* !
se,
,i,'"
en I
n
.L lQ or, u.
es, e*e en
)g r 0,,
k '"
3 e4, 3
/'
t I
I y
Fdtm*s Mwse
~
ENGINEERING REPORT
,4 g '
c g
rn.t wuxaER R-5.08-501-001 EMERGENCY DIESEL GENERATOR, DETROIT EDIS0N CO.,
8/25/86, R n. 1 j SUBJECT FERMI 2 NUCLEAR GENERATING STATION PREPARED ANALYSIS OF UPPER BEARING LUBRICANT SUPPLY 3Y W. A. Brill
$1rzY' DURING FAST STARTS APPQED I
4%
- r it t
E5 Y
BOOSTER q
k 9
e H
y Q
Q R
y y
1 lpWW ^ sso s si v.or Eve Z 91 est 'IK LUBE OrL, FL OW MODEL E.LEMtow!' MD ACDE NUMBERsNG A
4 5
m N
9 E
4 r, ces nyt 1st oss E.si 2si tsi - str h
e w
e e
r# sw
- 994 sir jpt.
,pt ~jpt Zyf
$pt
~
g m
~
s Urt 9Et Aft pet EEt ZE!
15' 9
t
?.
A SA
~ 9Ze S EJ pZt GZi Z Zi 4Z4 '
~
o I
e
'o *"
,n en " zu
'9*
su
,a su e
d k
R 2
R A*
g* *N n
e, eas Sos ~ p s ses Zoi so, 14 7 g
a 8e j
g 4
3 4
et
- w ss '.s ' s s in u
u,.
1
'96 h
8 k
k E
U so +n
,s ",, " n"w
=
osz u
wi i
+
a ss ",1
",a
',c U#
or u
st
,s o*z
+0i E
M N
E E
to,w",,",,",",,",,",,.
4 ors soi e
I O
3 2
0 2
0
',, g 2r ewe " v ",, g,, ",, ',, ",, -
! WI tu t
i
$4r 1
t e
2 see Y,, "e e",, "e e
tra l
g, Its Err 0
'e w
arr u
,, - rei E.
8 E
I N
N b
~
~
'#a U8
(
9s sc et is ZG
- E
~@8
ENGINEERING REPORT PATE' c,
h mz i
NUMBEn R-5.08-501-001 EMERGENCY DIESEL GENERATOR, DETROIT EDISON CO.,
8/25/86, Rev. 1 FERMI 2 NUCLEAR GENERATING STATION PREPARED W
ANALYSIS OF UPPER BEARING LUBRICANT SUPPLY W. A. Brill
)
mEPony TITLE DURING FAST STARTS AP ED
)
I'I----&
l }
I[*'l*
l i
f t
l*
i t
i 1
l t t
I I
l P
I PP'
' ' ' ' CdMedi"@
~~ ; -
-~
i cuwa M n!rwmm:..
n.
4
$..V
.4.-
m---.m h.
..g--
... 4
+e d.
.m y,
.is.y
..gm.i gm
. g a,
/
/y
/,,
..J g
' ' ' '~~~'.'. ' '
. ~.. - ~' '. ' _
~.. '..,
.. '-.~~--
d
... ~.~.~'
' ~~~ ~-~~
~.
g
/
p
. )k.
g----
~
- 2:222.~..' 2.~
~~'
~~~'
' ~ "M'
~~'
W}
$C2M Ef'M f/Tl
~
.gt
,.Qg.h-
. j f
.2- :
i L
//.
. U..
.,. y,.
y m'
g i
O
.-.A.
- p.
-_/
-g._
7
~
y.
y T.~. ' ~ ~1'~
~
~ 'O g/.
- /
. M MF WM.'
y
.=-
f
-/
..4...
I
....)
j/
/
g q
.../V.
s
/.
/,.
.../
...1..
7
'4'--
....y/
y---
.. W
- ,.".M a
,sy
'U N
. g W
.,Y
.7
-~
r g
I e
/
J W
..y
~5
/
/
/
'. C_~Z '-~
2-7
.g_
s.
/.c.
_- );-n 7
/
s-f f
\\
T hr-
..~
<w l
l f
.r
,, /
. (..
- e m. -p.WI
~
/
'-un e r..ww.--
g...
/
.-p 3 p
. p
/
y..
-,-y y
....--y a
/.M.
j
<.i
.l.p
. j
/
p--
c,.
. }
..p s
s
./
.,,,y.
/
/
s.
[
/
s l
s y
}
a
/ ~--..
g,-p
~...
(
.iI k.
. d
. g%
g ik
.l g,,
I em bb.
esF.
.nW.
~..
d i
d I
Id
.d i
W.. j q.
i
- q-
.d.s_a.i
.. I., s,,
a W
< -. u Jd j.i JW j.J w.J. I.].j
.I a
, 4,
,-~,,j..
4
,3 J. <
1,j i
j(
4 4
,.j j
jj i
.