IR 05000321/1998008
| ML17335A729 | |
| Person / Time | |
|---|---|
| Site: | Hatch  |
| Issue date: | 04/30/1998 |
| From: | NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION II) |
| To: | |
| Shared Package | |
| ML17335A728 | List: |
| References | |
| 50-321-98-08, 50-321-98-8, 50-366-98-08, 50-366-98-8, NUDOCS 9805110181 | |
| Download: ML17335A729 (36) | |
Text
U.S.
NUCLEAR REGULATORY COMNISSION REGION I I Docket Nos:
License Nos:
50-321 '0-366 DPR-57 and NPF-5 Report No:
50-321/98-08, 50-366/98-08 Licensee:
I'outhern Nuclear Operating Company.
Inc.
(SNC)
Facility:
E. I. Hatch Units 1 5 2 Location:
P. 0.
Box 2010 Baxley, Georgia 31515 Dates:
March 16 - 20, 1998 Inspectors:
- S. Alexander, Reactor Engineer, NRR. Division of Reactor Controls and Human Factors (ORCH).
Quality Assurance, Vendor Inspection and Maintenance Branch
- P.
Fi llion ~ Reactor Inspector, Region II, Team Leader F.
Gee.
Reactor Engineer, NRR
~
DRCH, Instrumentation and Controls Branch A. Pal, Electrical Engineer.
NRR, Division of Engineering, Electrical Engineering Branch Accompanying Inspector:
K. Naidu. Reactor Engineer.
NRR, DRCH, Quality Assurance, Vendor Inspection and Maintenance Branch (inspection observer)
.
Approved by:
K.
D. Landis. Chief Engineering Branch Division of Reactor Safety 9805iiOi8i 980430 PDR ADDCK 0500032i
PDR Enclosure
)
EXECUTIVE SUMMARY E. I. Hatch Nuclear Power Plant. Units 1 and
NRC Inspection Report 50-321/98-08 and 50-366/98-08 An inspection was conducted using the guidance of Temporary Instruction (TI)
2515/137, Inspection of Medium-Voltage and Low-Voltage Power Circuit Breakers.
issued March 1998.
The inspection team comprised Region II and Headquarters personnel.
On-site inspection time was four days plus an exit meeting.
The switchgear and surrounding areas were found in good material condition in a walkdown inspection with no deficiencies identified.
(Section M2. 1)
The team observed that certain problems with the 4160 V circuit breakers indicated the need for refurbishment.
However. the schedule for refurbishment of these breakers was not established.
The team noted that the 600 V breakers had a problem with the RMS-9 trip devices that was not resolved.
The team concluded that the circuit breakers were operable.
(Section M2.2)
The preventive maintenance procedures were of a type appropriate to the circumstances and included appropriate qualitative and quantitative acceptance criteria.
The team identified some manufacturer recommended steps and good industry practices which were not incorporated into the procedures.
(Section M3. 1)
The team confirmed through review of sample records that the licensee's stated preventive maintenance program was being implemented.
The team identified a
weakness in that the ci rcuit breakers did not have unique serial numbers or identifiers affixed to the breakers for recording on maintenance records.
(Section M3.2)
The team performed a detailed review of fai lures of medium-voltage ci rcuit breakers over the last three years.
and found that the licensee's root cause determinations and corrective actions met the requirements of Criterion XVI.
The team identified some weaknesses of detail and thoroughness in one of the licensee's Event Review Team reports covering ci rcuit breaker fai lures.
(Section M4. 1)
The team found that the licensee was complying with 10 CFR 50.65.
Maintenance Rule. with regard to treatment of power circuit breakers.
(Section M4.2)
With two exceptions.
the team concluded that the resolution of issues expressed in operating experience documents, Part 21 reports and vendor letters was good.
A weakness was identified with the Operating Experience Review procedure in that it did not specify review and resolution of vendor information letters (except for NSSS vendor).
(Section M5. 1)
The vendor interface program had weaknesses in that the licensee was not contacting the circuit breaker manufacturers which contributed to not receivirg all the relevant information.
(Section M5.2)
a P
The team concluded that while some overhauls were satisfactory.
other were not. but were apparently believed by the licensee. to have been equivalent.
Although most 600 V-AK breakers had been satisfactorily overhauled by GE, most of the AKs in Class 1E DC service had not. nor had most of the 4160 V
breakers.
Team further concluded that there were likely many DC AK breakers and 4160 V breakers that were overdue for overhaul.
The licensee stated that as a result of this information and its own analysis of recent failures't was undertaking a program to bring in a contractor to train its breaker technicians to perform overhauls on site and to develop better overhaul criteria and specifications.
In addition the license stated its plans to overhaul all 4160 V breakers beginning in the next outages with bus feeder and tie breakers (incoming) and then to place these breakers on a
6 to 9-year
'overhaul schedule.
(Section H8. 1)
The original calculation of ci rcuit breaker control voltage had weaknesses in terms of the design inputs and conclusions.
A safety assessment, performed during the inspection in response to the team's concerns about the voltage calculation.
was acceptable to support an interim operability determination.
The licensee indicated that a long term resolution would be developed to conservatively demonstrate on a continuing basis that the OC system could perform its intended f'unction. (Section H8.2)
r<,j'
li
REPORT DETAILS Summar of Plant Status Units 1 and 2 operated at power during the inspection period.
Introduction The focus was to verify the adequacy of licensee programs, procedures'raining.
equipment and supporting documentation for the maintenance of medium-voltage and low-voltage power ci rcuit breakers.
The overall scope of the inspection was defined in Temporary Instruction (TI) 2515/137, Inspection of Medium-Voltage and Low-Voltage Power Circuit Breakers.
issued March 1998.
The medium-voltage circuit breakers (4160 V system) 'were Type DHP manufactured by Westinghouse Electric Corporation, and the low-voltage ci rcuit breakers (600 V, 125 VDC and 250 VDC systems)
were Type AK manufactured by General Electric Company.
In 1989, the 600 V Type AK breakers were overhauled and retroiitted with General Electric Company Type RMS-9 digital based overcurrent trip devices.
E Maintenance H2 Maintenance and Haterial Condition of Facilities and Equipment M2. 1 Walkdown of Switch ear Areas Sho s and Trainin Faci lit a.
Ins ection Sco e
The team made a walkdown inspection of the switchgear areas.
The inspection included the 4160 V, 600 V and 125 VDC switchgear of both units'afety-related and non-safety-related.
At the team's request.
the licensee opened a few breaker compartment doors to allow viewing the interior compartment with the installed breaker.
The team inspected the maintenance shop to which breakers are brought when the requi red maintenance cannot be readily accomplished at the switchgear area.
The team inspected the electrical equipment maintenance training lab in the Skills Building to asses its suitability to support breaker maintenance training.
b.
Observations and Findin s
The switchgear and surrounding areas were found to be well maintained.
No broken of missing parts were observed.
and paint was in good condition.
No deficiency tags were observed.
Some low-voltage breakers were observed in the withdrawn position.
Upon request.
the licensee
=
provided a copy retrieved from the files of the evaluation which demonstrated that seismic qualification was maintained with breakers in the withdrawn positio L iI h
~ Removal of a breaker to the maintenance shop would-normally be necessary
'only for major maintenance involving substantial disassembly and repair of a breaker.
The maintenance shop was located in the plant such that it was accessible for all breakers.
except those that may not be able to be removed from thei r location (or with all components)
for radiological control reasons.
The maintenance shop was found to be clean, well organized, well lighted. well ventilated'nd well equipped, including ample work benches.
storage.
machine tools, and bas'ic test equipment.
There was no breaker maintenance in progress in the shop at the time of the inspection.
The team found that the electrical equipment maintenance-training lab was clean, well-lighted.
and well equipped.
In particular. the team noted that the lab had both 4160 V and 600 V training breakers and movable cubicle mockup stands to facilitate student access.
with readily accessible test equipment.
c.
Conclusions
'he switchgear and surrounding areas were found in good material condition in a walkdown inspection with no deficiencies identified.
Although the team did not directly observe br eaker maintenance in progress in the shop. the team was able to determine on the basis of observed conditions and equipment in the shop that it should provide a.
suitable environment for breaker maintenance in accordance with Hatch's current procedures, given their limitations discussed elsewhere in this report.
Similarly, the team w'as able to conclude that the electrical equipment maintenance training lab provided'an adequate environment to facilitate breaker maintenance training consistent with the knowledge and skill levels requi red by Hatch's current procedures and training documentation.
M2.2 Material Condition of Circuit Breakers Ins ection Sco e
The team evaluated the material condition of the circuit breakers.
b.
'Observations and Findin s The cause of failures of 4160 V circuit breakers indicated the need for refurbishment.
especially the problem of hardened grease (refer to Section M4. 1 for details).
The licensee planned to refurbish the 4160 V ci rcuit breakers in-house.
but a schedule for this work was not set.
This need for refurbishment caused the team to question the present operability of the 4160 V circuit breakers.
The overall maintenance pzogram and surveillance testing gave reasonable assurance that the
(
lI
breakers were operable.
The failures were precursors indicating that refurbishment should be initiated on an aggressive schedule.
The 600 V breakers had been refurbished.
That refurbishment replaced the obsolete trip devices with new type trip devices.
The 600 V ci rcuit breakers had a generic problem with the RMS-9 digital based overcurrent trip device, where the device would cause sympathetic tripping of breakers when a ground fault occurred on the system.
This appeared to be an inherent design problem with the device. at least for applications on certain type systems.
The licensee was expeditiously working on this problem.
but it was difficult to "prove" any resolution as ground faults are a rare occurrence.
The breakers were considered operable due to the low probability of the specific type of ground fault that may result in sympathetic tripping occurring.
Conclusions The team concluded that the ci rcuit breakers were operable.
However.
root cause evaluations for recent failures indicated that refurbishment of the 4160 V breakers should be performed on an aggressive schedule.
The 600 V circuit breakers had a generic problem with the RMS-9 overcurrent trip device.
Maintenance Procedures and Documentation Maintenance Procedures for Power Circuit Breakers Ins ection Sco e
The team reviewed the preventive maintenance procedures for the medium-voltage and low-voltage power circuit breakers.
The procedures were compared to the maintenance section of the manufacturer's instruction manuals.
The procedures were compared to the recommendations in Electric Power Research Institute (EPRI) publication NP-7410. Circuit Breaker Maintenance.
The procedures were reviewed in light of the team's knowledge of good industry practice for breaker maintenance.
as well as specific guidance in Tj 2515/137.
These documents and concepts provided a framework for assessing the quality of the maintenance procedures.
The ultimate acceptance criterion applied by the team was that the procedures were of a type appropriate to the circumstances and included appropriate qualitative and.quantitative acceptance criteria, as stated in 10 CFR 50. Appendix B. Criterion V, Instructions.
Procedures and Drawings.
Observations and Findin s The team observed that the maintenance procedures were clears detailed.
incorporated sign-offs on individual steps by craft persons and overall
[V
i II
by supervisors.
and included Quality Control hold points.
Data sheets provided a good record of the results of all measurements made and breaker condition at time of maintenance.
The team found that insulation tests and contact resistance test specified correct values.
Steps that specified examining a particular part, such as arc chutes'ere general.
The training department lesson plan covering breaker maintenance also did not contain specifics on examining parts.
The team noted that there was ao training specifically on the 600 V AKF (field breaker) ci rcuit breakers which are substantially different than the AK breakers.
The procedure for medium-voltage circuit breakers was 52PM-R22-001-0S, 4160 Volt AC Switchgear and Associated Electrical Components Preventive Maintenance.
The team reviewed in detail revision 14. dated February 24, 1998.
Since several key steps were revised recently, the team also reviewed portions of certain previous revisions to obtain 'a clear picture of the program as implemented in the recent past.
The team made the following observations on this procedure:
The current revision of the procedure included a step to lubricate the main and arcing contacts with a thin layer of grease.
This step was consistent with the manufacturer's instruction book.
However, previous revisions of the procedure did not have this step.
This meant that until recently the procedure directed craft personnel to clean the main and arcing contacts, but did not direct them to replace the grease.
To ascertain to what extent the possible lack of grease affected the condition of the contacts, the team examined two circuit breakers which had been in service for a long time.
The team found that the contacts on these two ci rcuit breakers were in good condition and there was no excessive contact friction.
The team also observed. that a
corrective maintenance procedure was not revised to specify applying grease to the main and arcing contacts.
The manufacturer's instruction manual (IB 32-253-48, page 40)
recommended to check clearance between the tripping trigger and the trip latch roller, and similar for the closing trigger.
The team observed that this step was not incorporated in the licensee's procedure.
.The team observed that the procedure did not include any type of breaker operating time check.
The team considered timing checks an accepted industry practice to confirm proper operation of the mechanism and contact.
Cognizant engineers stated that breaker timing checks were performed by a separat procedure on incoming (supply) breakers to confirm proper operation of the fast bus trans er.
This was not verified by the tea The team observed that the procedure did not include any functional test at reduced control voltage.
All functional checks were performed at battery float voltage'.
The team considered the reduced control voltage functional test a good industry practice predictive tool.
It could also be requi red in cases where the calculated control voltage is a lower value than the rated control voltage.
There is further discussion on this topic in Section M8.2.
The procedure for low-voltage circuit breakers was 52PM-MEL-012-0S, Low-Voltage Switchgear Preventive Maintenance.
Revision
ED I, dated September 11, 1997.
The team reviewed this procedure in detail, and made the following observations:
~
The above observations on breaker operating time check and reduced control voltage functional test applied to low-voltage circuit breakers as well.
~
The procedure did not include a check of the anti-pump circuit.
The team considered this check a good industry practice.
In general.
the team observed that the licensee was receptive to the above comments on the maintenance procedures.
The team was aware that the licensee was participating in circuit breaker owners groups which are working toward recommendations on breaker maintenance.
c.
Conclusions The preventive maintenance procedures were of-a type appropriate to the circumstances and included appropriate qualitative and quantitative acceptance criteria.
The team identified some manufacturer recommended steps and good industry practices which were not incorporated into the procedures.
The procedures for the breakers were recently revised.
and probably would be revised in the near future to add additional steps or possibly revise some steps.
M3.2 Review of Documentation to Confirm Im lementation of the Maintenance
~Pro ram a.
Ins ection Sco e
The team selected a sample of ci rcuit breaker s by load and function.
and requested preventive maintenance data sheets for the last two times the preventive maintenance was performed.
The records were reviewed to determine whether the licensee's program in terms of the interval between maintenance and the scope of maintenance was implemente I fl
b.
Observations and Findin s Review of the selected sample of data sheets confirmed that the interval between preventive maintenance wa's equal to or less than the licensee's stated program.
The maintenance interval was consistent with general industry practice.
The team identified a weakness in record keeping for maintenance on the ci rcuit breakers.
The licensee's practice was to track the maintenance program and work orders by use of the associated switchgear cubicle number.
Individual ci rcuit breakers did not have a serial number or.
unique identifier affixed to the circuit breaker.
The licensee's practice was keep each circuit breaker with its associated cubicle, with very little if any switching of circuit breakers among cubicles.
This practice should ensure that the basic program is implemented, and provides documentation of the basic program.
Nevertheless, lack of a unique identifier and recording of that unique identifier on all maintenance records means a certain ambiguity exists with regard to work performed on individual ci rcuit breakers.
Unique identifiers would provide additional confi rmation of which records belong with which circuit breakers.,
C.
Conclusions The team confirmed through review of sample records that the licensee's stated preventive maintenance program was being implemented.
The team identified a weakness in that the circuit breakers did not have unique serial numbers or identifiers affixed to the breakers for recording on maintenance records.
Naintenance Staff Knowledge and Performance Root Cause Evaluation and Corrective Action for Circuit Breaker Failures Ins ection Sco e
At the team's requests the licensee provided a summary of corrective maintenance performed on 4160 V ci rcuit breakers since the beginning of 1995.
The licensee also provided a copy of an Event Review Team Report which addressed failures of 4160 V circuit breakers.
The team reviewed these reports and discussed failures and corrective actions with the cognizant engineers.
The basic requirement applying to the scope of inspection was
CFR 50. Appendix B, Criterion XVI Corrective Action.
Observations and Findin s There were nine failures of 4160 V ci rcuit breakers since 1995.
All these failures were failure to close on demand and all occurred on
safety-related buses.
Four of the failures were due to a worn levering-in device.
A worn levering-in device results in the safety floor tripper maintaining a trip signal while at the same time the device turns free giving the operator the feedback that the breaker is fully inserted.
Corrective actions for this problem included checking the air gap between the tripping trigger and the mechanical floor tripper
.
cycling each inserted breaker before declaring it operable, replacing worn levering-.in devices (when identified) with a new device of improved design.
The inspector confirmed that operating procedures included the air gap check.
The inspector examined an in-service breaker and verified that an instruction sign was affixed to the breaker prompting the operator to check the air gap.
The inspector also observed that the parts at the air gap had been painted red to facilitate seeing the gap.
These corrective actions should prevent recurrence of the above described problem.
However, the NRC team was aware that the procedure requi ring the air gap check had been in effect since October 1994.
Apparently there were instances when the procedure was not implemented.
Three of the failures of 4160 V breakers were due to problems with
either the latch check switch or the motor cut off switch.
In the short term. the corrective action for this problem was to require the cycling of each breaker before declaring it operable (this is only partially effective).
In the long term, the licensee was considering modifying the control ci rcuits to provide an indicating lamp which will give positive indication of continuity in the close circuit.
Two of the failures were due to hardened grease.
A short term corrective action taken was to time test all the main incoming breakers using an already existing test procedure.
The proposed corrective action for this problem was to begin a program of refurbishment of breakers.
The team's understanding was that the main incoming breakers would be refurbished at the next refueling outage for each unit respectively.
The licensee's intention was to refurbish all safety-related 4160 V breakers.
Four of the above mentioned failures were addressed by an in-house Event Review Team.
Their report was issued on December 10.
1997.
The NRC team found the following weaknesses with this report:
~
The report gave a cause as inadequate lubrication of crank shaft bearings when the actual problem was with the pole shaft bearings.
The report states that the failure which occurred on October 31.
1997.
was due to hardened grease and that this breaker had never been refurbished.
The failed breaker was examined by the team and the team observed that this breaker had a refurbishment sticker, which had been affixed by an outside company.
At that point
~ the licensee believed the Event Review Team report was in error and-
It 8'
tI II
'I
that the breaker had been refurbished in 1990.
There is further discussion on this matter in Section H8. 1.
As a result of the NRC.
team's observation, corrective actions beyond those stated in the licensee's report may be requi red.
~
The report did not discuss when the last preventive maintenance was performed on the failed breakers and whether anything about the preventive maintenance procedure or the frequency of past maintenance played a role in each of the failures.
The report did recommend future refurbishment.
The report did discuss the difficulty in adjusting the latch'check switch with the method in the procedure.
\\
Host of the failures discussed above are age related failures, which indicates the need for refurbishment.
Various documents made different statements and various personnel made different statements concerning the schedule for future refurbishment.
The team evaluated the nine fai lures of 4160 V breakers, and considered the date of the failures and the timing of corrective actions.
The team found that the root cause determinations and subsequent corrective actions met the requirements of Criterion XVI.
c.
Conclusions The team performed a detailed review of fai lures of medium-voltage circuit breakers over the last three years.
and found that the licensee's root cause determinations and corrective actions met the requirements of Criterion XVI.
The team identified some weaknesses of detai l and thoroughness in one of the licensee's Event Review Team reports covering circuit breaker failures.
H4.2 Maintenance Rule Issues a.
Ins ection Sco e
The team reviewed the licensee's compliance with 10 CFR 50.65.
Requirements for Monitoring the Effectiveness of Maintenance at Nuclear Power Plants. with regard to power circuit breakers.
b.
Observations and Findin s The licensee defined an Electrical Distribution System for maintenance rule purposes.
which included incoming and tie circuit breakers.
Individual load ci rcuit breakers were considered part of the associated mechanical system.
However.
due to fai lures." circuit breakers were being monitored as a separate class of components under
CFR 5Q.65(a)(1).
The 4160 V breakers were classified in March 1998.
and the
~
~
t
600 V breakers were classified in February 1998.
The team verified that circuit breaker failures had resulted in the associated mechanical system being classified
CFR 50.65(a)(1).
The team found that the licensee had criteria and goals for ci rcuit breakers that were commensurate with safety.
The team observed that the licensee established a "4 kV Breaker Reliability Enhancement Plan" to
..help ensure that those breakers could be returned to (a)(2) status.
c'.
Conclusions The team found that the licensee was complying with 10 CFR 50.65.
Maintenance Rule. with regard to treatment of power circuit breakers..
M5 Maintenance Organization and Administration M5. 1 Review nd Evaluation of 0 eratin Ex erience Documents Part
Re orts and Vendor Letters a.
Ins ection Sco e
The team reviewed the licensee's resolution of all the NRC Information Notices and Bulletins and vendor letters listed in TI 2515/137 applicable to the type of switchgear at Hatch.
The team reviewed a
sample of 10 CFR 21 Reports for resolution.
The team reviewed the licensee's Operating Experience Review procedure and procedure adopted pursuant to 10 CFR 21.
b.
Observations and Findin s As described in Section M5.2. the licensee had not received all the applicable vendor letters.
With regard to resolution of letters that were received, the team found that. with two exceptions.
the licensee resolved the issue discussed in the letter.
The exceptions identified were letters from General Electric Co. applicable to safety-related low-voltage breakers which had been erroneously declared not applicable based on the title and superficial review of the contents.
The two letters were:
SAL 175-9.20.
Maintenance and Upgrade of AK-25 Circuit Breakers with Undervoltage Trip Devices Used as Reactor Trip Devices.
dated March 21 '984.
SAL 175-9.21, AK-25 Trip Breakers
- Shunt Trip Paddle, dated February 7.
198 Resolution of NRC generic communications and Part 2l reports on the subject of circuit breakers was good.
The Operating Experience Review procedure did not specify that the non-NSSS-scope-of-supply information letters be reviewed and resolved, although the team concluded this activity was in fact taking place.
The licensee's practice was to treat vendor information letters as Part
reports.
but the Part 21 report procedure did not address this.
Conclusions M5.2 With two exceptions, the team concluded that the resolution of issues expressed in operating experience documents, Part 21 reports and vendor letters was good.
A weakness was identified with the Operating Experience Review procedure in that it did not specify review and resolution of vendor information letters (except for NSSS vendor).
Vendor Interface Pro ram Ins ection Sc e
The team reviewed the licensee's vendor interface program.
'The team reviewed the licensee's response to NRC Generic Letter 90-03, Relaxation of Staff Position in Generic Letter 83-28, Item 2.2 Part 2 "Vendor Interface for Safety-Related Components
- and it
- Supplement 1. both of which promulgated the NRC Staff's position on vendor interface programs.
GL 90-03 required at least annual contact with vendors of key safety-related equipment.
Supplement 1 mentioned Electrical Distribution system equipment.
b. Observations and Findin s The licensee's master list of vendors used to make the -annual contact to verify receipt of all pertinent technical information omitted General Electric-Electrical Distribution 8 Control and Westinghouse Nuclear Services Division, both of which supplied the licensee's switchgear.
The team also noted that Coltec Co., supplier of the emergency diesel generators, was omitted.
Apparently. the licensee had never contacted the sources of technical information on ci rcuit breakers to establish themselves as customers who should be on the distribution list.
The
.
licensee did not receive all the Service Advisory Letters (SALs)
pertaining to the Type AK circuit breakers from General Electric Co.
The licensee did not receive one Technical Bulletin applicable to Type DHP circuit breaker s from Westinghouse Corporation.
The team concluded that omissions in the vendor master list contributed to failure to receive the vendor information.
The team also noted that the vendors themselves bear some responsibi lity for this problem.
For example.
the
.licensee had a contract with General Electric Nuclear Energy for its
Vendor manual Subscription Service which should have delivered the SALs in question.
Westinghouse Corporation sent the Technical Bulletin in question only to its Ty'pe DS.circuit breaker customer list.
The team found that failure to receive technical information through the primary channel did not result in any actual hardware problem, due to the fact that the information was received through other channels such as NRC Information.Notices.
Nevertheless this weakness in the program, if left uncorrected, could lead to problems in the future.
At present, the licensee is actively participating in ci rcuit breaker owners groups which should correct or overcome the administrative problems noted above.
Conclusions The vendor interface program had weaknesses, in that the licensee was not contacting the ci rcuit breaker manufacturers which contributed to not receiving all the relevant information.
Miscellaneous Haintenance Issues Review of Purchase Order s Ins ection Sco e
As discussed elsewhere in this report.
breakers at Hatch were not on a
=
regular (periodic) overhaul or refurbishment schedule.
Rather
~ they were typically sent to various subcontractors for repair, modification.
and/or refurbishment, on an as-needed basis.
To assess the scope and depth of breaker overhauls, the team reviewed selected files of overhaul records.
The principal contents of these were typically the GPC purchase orders (POs) to repai r facilities for the work and the reports by the repair facilities documenting what work was accomplished, parts replaced, test results.
etc.
Observations and Findin s The licensee issued PO 94-6018474, dated October 6, 1994. to Westinghouse Electric Co.. for'ork on Westinghouse 4160 V type DHP circuit breakers at the Eaton-Cutler Hammer (ECH) Greenwood.
South Carolina. facility. during the Hatch Unit 1 1994 outage.
Note that this was around the time of the'ale of the DHP product line to ECH.
The PO required an appropriate QA program and stated the applicability of
CFR Part 21.
The work originally specified was repai r and refurbishment'ut without further specificity, except that the PO stated that the charging ratchet mechanism needed to be checked and
, repaired as it was not allowing the breaker to be charged, as well as
~
L
other necessary repairs.
However.
a change order required the breaker to 'be completely "remanufactured,"
bringing it up to "like new" condition. including complete inspection, cleaning.
and replacement of pole units," contacts, puffer tubes, pawl kits, and relay.
Renewal of lubrication was not specifically mentioned.
It was also not clear how the breaker was identified except for its style number and another designation,
"STL g8."
The certificate of conformance returned with the breaker was issued by Westinghouse Nuclear Services Division (WNSD), not ECH.
It stated that the remanufactured breaker had "Serial Number" 8-940.521-1.
The attached repair report stated that upon receipt the breaker received visual inspection, mechanical operation, electrical operation, and a
primary conductance test.
Deficiencies noted were damaged pawls (in the spring charging ratchet mechanism)
-damaged contacts.
The report gave a
list of replaced parts and stated that the breaker was cleaned, moving parts relubricated, and tested satisfactorily including timing and conductance tests.
The teary found this work to be reasonably complete.
A similar PO, 94-6018258, specified
"Clean, test, inspect and replace parts in accordance with Westinghouse's standard remanufactured breaker program."
Similar attached documents indicated similar results.
The licensee issued PO P-02757 on March 1.
1989, to General Electric Nuclear Energy (GE NE) at its Atlanta, Georgia, switchgear service shop.
Part 21 and Appendix 8 were applicable.
The PO was apparently one of
.
several that covered multiple AK type 600-volt breakers.
The PO originally specified "service and install RMS 9 overcurrent trip systems
~ but a change order required refurbishment in addition to the RNS 9 upgrade.
The vendor's report listed electrical and mechanical test data.
parts replaced, and provided a product quality certificate and a certificate of seismic qualification.
The work appeared to be comprehensive and reported test results were satisfactory.
The licensee's PO C12347, issued October 3.
1988, to Power Distribution Technologies (PDT), Inc.. (with Appendix B. and Part 21 applicable)
simply indicated that the safety-grade breaker was to be refurbished.
The PO stated that the vendor should ensure that the breaker's make.
model and serial number should be on all documents.
but what the vendor thought was the serial number.
was the shop order number which is not unique to that breaker.
and in fact would be common to several breakers of a similar type for the same Hatch unit.
The vendor's report stated that the breaker was overhauled and tested.
It listed the conditions found.
and action taken.
including the replacement of several worn or broken parts.
However. there was no evidence that the vendor performed major disassembly.
cleaning and relubrication. particularly of the pole shaft and crank shaft beariags and other moving mechanism parts.
This document indicated an unsatisfactory overhaul.
This report was consistent with the fact that this was the breaker that had recently
failed in service. its failure.attributed to hardened grease in the crank shaft (although the licensee acknowledged that the report used incorrect terminology and was meant to say pole shaft bearings).
Conclusions M8.2 The team concluded that while some overhauls were satisfactory, other were not.. but were apparently believed by the licensee to have been equivalent.
Although most 600 V AK breakers had been satisfactorily overhauled by GE, most of the AKs in Class 1E DC service had not, nor had most of the 4160 V breakers.
Team further concluded that there were likely many DC AK breakers and 4160 V breakers that were overdue for
'verhaul.
The licensee stated that=as a result of this information and its own analysis of recent fai lures, it was undertaking a program to bring in a contractor to train its breaker technicians to perform overhauls on site and to develop better overhaul criteria and specifications.
In-addition the license stated its plans to overhaul all 4160 V breakers beginning in the next outages with bus feeder and tie breakers (incoming) and then to place these breakers on a
6 to 9-year overhaul schedule.
Breaker Control Power Issue Ins ection Sco e
The team performed inspection to establish whether breaker operation was assured at minimum operating voltage as specified in the vendor's manual (rated voltage)
or minimum calculated voltage.
whichever is the lowest as stated in the TI.
b.
Observation and Findin s The team found that all periodic functional testing within the preventive maintenance procedure was carried out at battery float voltage.
With regard to the 4160 V breakers, minimum calculated voltage was below the rated voltage.
This fact led to further inspection effort by. the team which =-is described below.
Calculation No.,88 ELECT.
DC Control Circuit Voltage Drop in Diesel Battery Systems'
"Rev. 3. dated March 19.
1998 'ocumented the worst case minimum available dc control circuit voltage 'at the closing. coil of 4160 V breakers.
The design basis of the calculation was LOOP/LOCA with battery charger failure for a two-hour duty cycle.
The first 12 seconds of the load profile was comprised of momentary loads of, breaker operations and field flashing plus a steady state load dominated by emergency lighting.
The two hour steady state load was emergency lighting.
The team found this load profile was consistent with the Final Safety Analysis Report.
The worst case calculated voltage was
~
i II
74.5 V at the close coil for Emergency Diesel Generator 1B.
The published minimum operating voltage for the closing coil, as shown on manufacturer's information attached to the calculation, was 105 V.
In-house testing of three breakers in 1989 showed that the close coils operated at from 55 V to 65 V.
Based on this test.
the calculation assumed the minimum acceptable voltage at the closing coils of 4160 V
breakers was 65 V.
Although not mentioned in the calculations the
. licensee, per Technical Specification surveillance requirements was performing a loss-of-normal AC-power surveillance test at each refueling outage.
This surveillance test essentially duplicated the design conditions of the first 12 seconds of the calculation load profile.
The team took the position that since calculated voltage was significantly below rated voltage (i.e. 74.5 V vs 105 V) the calculation should have recommended to perform periodic testing to demonstrate on a continuing basis that the breakers can perform their function.
The team questioned whether the surveillance test in fact provided this periodic testing.
The team concluded that it did not, because the surveillance test was only a go-no-go functional test.
Specific criteria were needed in terms of voltage which would demonstrate that sufficient margin existed to ensure good operation until the next test, given that battery capacity will drop with age.
In response to the team's question the licensee performed a safety assessment during the inspection, which the team reviewed.
The safety assessment focused on the breaker operations in the first 12 seconds (Time from LOOP to EDG output breaker closure which restores power to the battery charger).
This was the critical consideration due to the large voltage drop in the control circuits.
There was no concern with the abi 1,ity to carry emergency lighting for two hours.
The calculation within the safety assessment removed the aging factor and design margin included in the No.88 ELECT calculation.
This was justified due to having new batteries and recent capacity tests showing at least 100 percent capacity., This new calculation also adjusted the cable resistance values to more realistic values.
Apparently. the 105 V
rating in the No.88 ELECT calculation was not correct.
Westinghouse Electric Corporation supplied information indicated that an enhanced design coil was furnished with the Hatch shop order.
Westinghouse stated that the new design had a published rating of 90 V. and there was considerable margin below this rating. which could be confirmed by on-site testing.
As stated above.
the licensee had performed some on-site testing.
This supplementary information gave more credibility to the use of 65 V in the original calculation.
but did not mean that
V was an actual rating.
The safety assessment calculation indicated
V at the EDG 18 close coil as compared to 90 V coil rating.
Margin was demonstrated by the 1989 test results mentioned abov l ll'
When the aging factor and the design margin are put back into the calculation as inevitably they must be, the calculated worst case voltage will be less than 90 V.
This concept means that a future long term resolution to the issue raised during this inspection must be developed.
The team realized there were several options to resolving the issue.
To ensure NRC review of the final resolution of this issue and other issues and weaknesses described in this report Inspector Follow Up Item 98008-01.
Review of Lo'ng Term Resolution of Circuit Breaker Maintenance Weaknesses and Issues'as established.
c.
Conclusions The original calculation of circuit breaker control voltage lacked the rigor normally seen in safety-related calculations in terms of the design inputs and conclusions.
The team concluded that this was a
weakness but did not.
by itself, rise to the level of a violation of the requirements for design control.
A safety assessment, performed during the inspection in response to the team's concerns about the voltage calculation.
incorporated supplementary information and removed the aging factor and the design margin.
The safety assessment was acceptable to support an interim operability determination.
The licensee indicated that a long term resolution would be developed to conservatively demonstrate on a continuing basis that the DC system could perform its intended function.
H8.3 Closed IFI 97010-08
Review of 4160 Volt Breaker Failure Analysis and Preventive Maintenance Program.
The team inspection performed according to TI 2515/137 completed the review intended by this IFI.
Mana ement Heetin s
Exit Heeting Summary The team presented the inspection results to licensee management on March 20.
1998.
The licensee management acknowledged the results as presented.
The team handled some proprietary information during the inspection, but no proprietary information is included in this report.
PARTIAL LIST OF PERSONS CONTACTED Licensee
~J.
Betsi ll, Assistant General Manager,
.Plant Operations M. Googe.
Maintenance Manager
- J. Lewis, Manager.
Plant Training
- T. Moore. Assistant General Manager Plant Support
- P. Roberts.
Manager
~ Outages and Planning
- S. Tipps, Manager.
Nuclear Safety and Compliance
- P. Wells. General Manager Nuclear Plant
'
NRC and Other Or anizations
- R. Carrion. Senior Resident Inspector (Acting)
- B. Crosby.
Component Engineer, Commonwealth Edison.
Representing Westinghouse Electric Corporation Circuit Breaker Owners Group
- M. Thomas.
Chief (Acting) Engineering Branch.
Region II
"Attended exit meeting
.
INSPECTION PROCEDURES USED TI 2515/137 Inspection of Medium-Voltage and Low-Voltage Power Circuit Breakers ITEMS OPENED, CLOSED AND DISCUSSED
~oened 50-321.
366/98008-01 IFI Review long term resolution of circuit breaker maintenance weaknesses and issues (Section M8.2).
Closed 50-321.
366/97010-08 IFI Review of 4160 Volt breaker failure analysis and preventive maintenance program (Section M8.3).
J'f 0'
I