ML17300A787
| ML17300A787 | |
| Person / Time | |
|---|---|
| Site: | Palo Verde  |
| Issue date: | 04/28/1987 |
| From: | Office of Nuclear Reactor Regulation |
| To: | |
| Shared Package | |
| ML17300A786 | List: |
| References | |
| TAC-64398, NUDOCS 8705010148 | |
| Download: ML17300A787 (14) | |
Text
~R IltOI
~o A
Cl0 a
+~
gO
%y*y4 UNITEDSTATES NUCLEAR REGULATORYCOMMISSION WASHINGTON, D. C. 20555 ENCLOSURE 1
STAFF EVALUATION BY THE OFFICE OF NUCLEAR REACTOR REGULATION COVERING LICENSEE'S RECOVERY PROGRAM FOR EMERGENCY DIESEL GENERATOR-B AT PALO VERDE UNIT 3 TAC. NO. 64398
1.0 INTRODUCTION
'O HQ
'mlhQ
'AIOQ OO OO (Q&
',mM wD
~OQ 0
IA OQ.
l&Q
~ lmQ.Q 4
2.0 Arfzona Nuclear Power, Project (ANPP-the licensee) has submitted a program designed to demonstrate the acceptability'of repairs made to the damaged Emergency Diesel Generator (EDG)-B at Palo Verde Unit 3 by demonstrating operabil'ity of EDG-B.
The program was forwarded ender cover of ANPP letter, Serial 161-00103-JGH/RAB, dated March 24,'987.
The licensee's program consists primarily of running EDG-B in various modes for specified periods of time coupled with inspections of the EDG during and subsequent to operation.
This evaluation documents the staff's review of the licensee's recovery program with regard to whether or not the program includes (1) sufficient EDG operation at load for potential problems to surface, and (2) adequate inspection fn terms of frequency and depth to find indications of potential problems which may have surfaced.
In Section 3.0 of this evaluation, the staff lists ct fteria which must be met in order to declare EDG-B operable.
This evaluation fs concerned only with whether or not the licensee's recovery program, when implemented, is sufffcient to demonstrate these criteria.
No attempt fs made at this time to judge if the criteria in Section 3.0 will be met.
A determination of EDG operability will be made at the conclusion of testfng and inspection, and will be the subject of a separate evaluation.
BACKGROUND On December 23, 1986, EDG-'B sustained extensfve damage to engine interna'ls and structural components as a consequence of failure of the number 9
cylinder master connecting rod.
The damaged'nternals and components include the crankshaft at the number 9 crankpfn journal, crankshaft counterweights, pistons,, master and.articulated connecting rods, cylinder liners, center frame casting, right side cylinder block casting, and miscellaneous tubing and instrumentation.
Between December 23,- 1986 and January 20, 1987, the licensee conducted a
detailed investigation of the engine faflure fn order to determfne (1) the root cause of the failure and generic fmplfcations, ff any, and (2) what would be requ'fred in way of replacement or repairs to restore the
i>
li l
At a meeting in Bethesda oh January 20, 1987, the licensee presented the results of their investigation.
The root cause of the failure was identified as high cycle fatigue of the master connecting rod fn the area where the articulated rod connects to ft.
The high cycle fatigue initfated fn a I/16 inch thick iron materfal plated to the master rod bore during manufacture and propogated into the rod base material.
Generic implications were limited to master rods having the same iron plating.
The licensee also determined that it would be possible to repafr the existing engine without having to replace the crankshaft, the center frame casting, or the right cylinder block castfng.
The crankshaft would be machined in place to resotre the number nine crankpin Journal surface to an acceptable undersize condition.
The damaged engine castfngs would be repafred using a process called metal stitching.
Some detafls of this latter repair process were also provided.
A ffnal report on the failure of EDG-B was submftted by the licensee on February 9, 1987.
Thfs report fncluded details of the root cause analysis and findings, as well as details of the licensee's program for testing and requalffication of EOG-B subsequent to repairs.
The staff reviewed the licensee's submittal and concluded that the root cause fdentffied by the
'licensee was the most probable cause.
Thfs sub)ect is dfscussed further in Section 3.0 of this evaluation.
The staff, however, did not agree with the licensee's progra'm -for testing and requalification.
The staff's position was that additfo'nal engine operation followed by non-destructive examination of engfne components was required.
The staff also identified an additional 14 items which should be addressed by the licensee to ensure complete root cause fdentiffcation as well as long term adequacy of engfne repafrs.
The staff posftion, fncludfng the above 14 ftems, were forwarded to the lfcensee by letter dated Natch 9, 1987.
The differences between the staff and lfcensee's positions fdentfffed fn the above staff letter were dfscussed at length during numerous conference calls between the staff and the lfcensee, but no resolution was reached.
Finally, on Harch 13, 1987, the staff visited the Palo Verde Unit 3 site to physically inspect repairs to EOG-B and to conduct fn-depth discussfons with the lfcensee regarding the above differences.
At the conclusion of the site visit, most of the differences between the staff and licensee posftfons had been resolved.
A few remaining items were resolved followfng the site visit:. and the lfcensee revfsed their program to ~ect these resolutions.
,The revised program for testing and requalfffcatfon of EDG-B prior to declaring ft operable was submitted on Narch 24, 1987.
The program submitted by the licensee is the subject of this staff evaluation.
3.0 STAFF EVALUATION
- 3. 1 Staff Concerns Faflure of the number 9 cylinder bank master rod on EOG-B imposed stresses for which the engfne was not specifically designed and which are not antfcfpated fn the normal operating 1~ of a diesel
.engine.
Although EOG-8 can be repaired, the staffwas concerned with the adequacy of this engine for nuclear service subsequent to repair because of fts being sub/ected to severe stresses.
This concern is
~I
'4l
)+
~
io
~
)
for stress related fatigue and vibration induced failure of engine component~
and repairs..
The staff is looking at 3 specific areas as follows: '
Crankshaft -
will the repaired crankshaft be acceptable for long term, reliable service considering the stresses that have been imposed on it and the fact that the number 9 crankpin journal is undersized by 0.175 inches.
Castings-Alignment-will the repaired engine castings (center frame and right side cylinder block) be able to with-stand the stresses of long term service without subsequent fatigue or vibration induced failure.
have the engine repairs been made in a manner that will ensure proper. alignment of critical engine components and preclude premature failure of the engine due to excessive wearing of mis-aligned parts.
The staff had additional concerns regarding the engine failure root cause determination.
The staff agrees with'he mechanics of the failure; i.e., the failure of the master rod initiated in the iron plating around the oil holes feeding -the articulated rod and propagated into the base metal.
However, the staff is not certain whether the failure was due to the iron plating, alone, or if there were other contributing factors.
Such factors could include a faulty master rod design or faulty workmanship.
The above staff concerns were forwarded to the licensee by letter dated March 9, 1987.
The staff letter included 14 specific recoaeen-dations which, if followed, would provide additional assurance regarding the adequacy of the repaired EOG for nuclear service and further define the root cause of the failure.
The March 9 staff letter also addressed the length of time EDG-B should be operated and physical inspections that should be conducted prior to declaring EOG-B operable.
The contents of the staff letter were discussed at length with the licensee in telephone conference calls and during a site visit to inspect EOG-B repairs on March 13, 1987.
Agreement on a recovery program for EDG-B was finally reached on March 17 and 18, 1987.
This agreement was documented in licensee letter, dated March 24,
- 1987, and enclosures thereto.
The licensee's response is discussed bel ow.
3.2 Licensee's Recove Pro ram The licensee has agreed to operate EDG-B at full load for at least 100 hours0.00116 days <br />0.0278 hours <br />1.653439e-4 weeks <br />3.805e-5 months <br /> as part of the recovery program.
The licensee will conduct inspections of the engine prior to operation, at various times during accumulation of the 100 hours0.00116 days <br />0.0278 hours <br />1.653439e-4 weeks <br />3.805e-5 months <br />, and subsequent to accumulating 100 hours0.00116 days <br />0.0278 hours <br />1.653439e-4 weeks <br />3.805e-5 months <br /> at full load.
The latter inspection will cover 19 individual items, including non-destructive examination (NOE) of the number 9 crankpin
41
.and visual inspection of the number 9'crankpin bearing shells.
Details of the licensee.'s engine inspections are included as Enclosure 6 to the licensee's Harch 24, 1987, letter.
In the process of accumulating 100 hours0.00116 days <br />0.0278 hours <br />1.653439e-4 weeks <br />3.805e-5 months <br /> at full load, EOG-B will be subjected to a series of tests designed to demonstrate the accepta-bility of the diesel generator instrumentation and controls, safety
- systems, and electrical generating capability.
These tests will include 35 starts in. order to requality EOG-. B in accordance with Regulatory Guide l. 108;.
Details of these tests and the total hours at various loads for each of them are included as Enclosures 4 and 5, respectively, to the licensee's Harch 24, 1987, letter.
The licensee also responded to the 14 staff recommendations.
These responses, included as Enclosure 1 to the licensee's Harch 24, 1987 letter are discussed below.
~
Recommendation No. 1:
Recommendation No. 2:
Recommendation No. 3:
Recommendation No. 4:
Recommendation No. 5:
Recommendation No.
6:.
The licensee provided calculations to show that the torsional stresses on the number 9 crankpin will increase by only'pproximately 4.7X due to. the reduced crankpin
.diameter.
These calculations are supported by calcula-tions of torsional stresses in a typical Cooper KSV-20-T crankshaft and actual torsiograph values of a Palo Verde KSV-20-T engine.
The calculations and torsiogt aph values were included as Attachments 1-3 to Enclosure l.
The licensee has provided a finite element analysis of the master connect-ing rod.
The staff is currently review-ing this submittal.
The licensee has made provisions for a complete inspection of critical engine.
parts in the recovery. program for EDG-B.
Crankshaft hot and cold deflection tests are=included as part of the recovery program for EOG-B..
The licensee has provided an analysis to show that operation for a sustained period of time (50 min) at off-design speed (296RPH) has not damaged the crankshaft.
The licensee will visually inspect all crankshaft journals for damage (cracks).
In addition, NOE will be performed on the Nos.
8 and 10 journals, and twice on the No.
9 crankpin journal.
4i I;
'I 1
Recommendation No.
Recommendation No.
Recommendation No.
7 8:
9
~
The vibration damper will be inspected in accordance with vendor recoamenda-tions prior to'eclaring operability.
The jacket water system of EOG-B will be hydrostatically tested and inspected for leaks prior to operating the engine.
The engine will be inspected again for leaks following break in, and the lube oil will be analyzed for water content as weIl as for dirt, viscosity, and wear metals.
The number 9 crankpin jourMI will be inspected subsequent to EOG-B testing and prior to declaring it operable.
Recommendation No.
10:
Recommendation No. 11:
Recommendation No. '12:
Recommendation No. 13:
Recommendation No. 14:
3.3 Evaluation All irqn plated connecting rods have been removed from service at Palo Verde.
The licensee has verified that no con-necting rods are in use which have a
nickel spray applied in high stress areas.
(This is a subset of the master connecting rod finite element analysis.)
Connecting rod high stress areas have been identified.
(This is a subset of the master connecting rod finite element analysis.)
The licensee will have a sample of lube oil from the damaged engine analyzed for wear metals, dirt, viscosity, and acidity.
The licensee has provided data to demonstrate that the fire extinguishing chemicals introduced into EDG-B at the time of the casualty have not had an adverse effect on engine components.
As stated
- above, the staff wascohcerned about stress related fatigue and vibration induced failure of engine components.,
It is the staff's position that the adequacy of EOG-8 for nuclear service can best be demonstrated by operating it at full load for a specified period of time.
The operating time is determined primarily by the mechanical properties of the materials in the engine, including materials used for the repairs.
These properties can be represented by a S-N (stress vs number of cyclic loadings) curve.
For the types of materials used in EOG-B, including repairs, the S-N curves show
0 4l t
that operation in excess of 10 cycles at stress levels commensurate with full-load operation without failure of any kind is adequate to demonstrate that the materials can withsta'nd an infinite number of cyc1es at those same stress levejs.
Twenty eight (28) hours of operation at full load equals 10 cycles for the Cooper KSV-20-t engines.
The licensee has agreed to operate EOG-B for 100 hours0.00116 days <br />0.0278 hours <br />1.653439e-4 weeks <br />3.805e-5 months <br /> at full load, which is equal Co 3.6 x 10 cycles.
Assuming the NOE of the number 9 crankpin journal does not show any damage, the crankshaft will have been shown to be adequate for nuclear service with a safety margin of approximately 3.6.
The same concept is applicable to the engine repairs; i.e., if there is no evidence of failure after 100 hours0.00116 days <br />0.0278 hours <br />1.653439e-4 weeks <br />3.805e-5 months <br /> of operation, there is adequate assurance that the repairs will retain their integrity for an infinite period of time.
Finally, if the visual inspection of the number nine crankpin journal and bearings does not show any excess wear, it can be concluded that the alignment of critical engine components was correct at the completion of repairs and was not altered as a consequence of operating at full load for 100 hours0.00116 days <br />0.0278 hours <br />1.653439e-4 weeks <br />3.805e-5 months <br />.
Based on the above, the staff concludes that the licensee's recovery program for EOG-B adequately addresses the staff concerns regarding the crankshaft, engine castings, and alignment and is, therefore, acceptable.
The finite element analysis of the master connecting rod provided by the licensee is'till under review by the staff.
Me will report our findings in a separate evaluation at a future date, but prior to Palo Verde Unit 3 entering Node 4.
The absence of staff findings regarding this analysis does not affect the licensee's recovery program for EOG-B, however.
The design of the master connecting rod ss a generic issue which, if found deficient, would affect all d>esel generators made by Cooper Industries.
This design, however, is in no way impac-ted by the repairs to EOG-B, and staff review of the design analysis can proceed independent of the recovery program for EOG-B.
The licensee's responses to each of the staff's 14 recommendations have been reviewed and evaluated as follows.
For recoaeendations 1,
4 and 5, the licensee has provided analytical data regarding the adequacy of the repaired crankshaft.
The data indicates that the repairs to the crankshaft wi 11 not significantly change the stress levels in the repaired (undersized) area and, therefore, will not affect its suitability for use in nuclear service.
The staffhas reviewed the data provided regarding crankshaft adequacy andfinds it'cceptable.
Recommendation 2 deals with the master connecting finite element analysis and is discussed above.
Recommendations 3, 6, 7 and 14 relate to inspecting engine components other than those in the affected number 9 cylinder bank.
The purpose of these recommendations is to provide assurance that engine components outside of the affected areas were not damaged, or.that any damage was identified and compon-ents repaired or replaced.
The licensee has agreed to include these recommendations in the recovery program, thereby providing the assur-ance sought.
The staff finds this acceptable.
Recommendations 8
and 9 pertain to inspections to determine the adequacy of specific
I
engine repairs.
The licensee has also agreed to include these recom-mendations in the recovery program.
The staff finds this to be accep-table, also.
Recommendations 10, 11, and 12 are subsets of recommen-dation 2 and will be discussed along with it 'in a future evaluation.
Th'e licensee has provided all the information necessary to comply with recommendations 2, 10, ll, and 12.
Finally, recommendation 13, when implemented, will provide 'additional information regarding the accur-acy of the root cause determination.
The licensee has agreed to in-clude this recommendation in the recovery program for EDG-B, and the staff finds this acceptable.
- 4. 0 CONCLUSIONS Based on our review of the data provided by the licensee, the staff con-cludes that the scope of the licensee's recovery program for EOG-B is such, that, when implemented, will provide all the information required to determine the adequacy of EOG-B for nuclear service.
The scope of the program includes detailed design information on critical engine components as well as complete test and inspection procedures.
The combination of design analyses and engine test/inspections addresses all of the staff's concerns.
The staff concludes that the licensee's recovery program for EOG-B, as detailed in the licensee's March 24, l987, letter and enclosures
- thereto, is acceptable.
4t P
~ g~