ML20043B101
| ML20043B101 | |
| Person / Time | |
|---|---|
| Site: | Pilgrim |
| Issue date: | 05/10/1990 |
| From: | Recasha Mitchell GENERAL ELECTRIC CO. |
| To: | Grimes B Office of Nuclear Reactor Regulation |
| References | |
| MFN-040-90, MFN-40-90, RCM-9017, NUDOCS 9005240190 | |
| Download: ML20043B101 (5) | |
Text
bklk C
,q.
tL jg3
.v
' ~
GE Nuclear Energy
_.,~._. _,
m csw un. sr u cm:s May 10,--1990 MFU 040-90 RCM-9017 i
l e
Mr. Brian K. Grimes, Director Division of Reactor Inspection and Safeguards Office of Nuclear Reactor Regulation l
US Nuclear Regulatory Commission Washington, DC 20555 4
l
Dear Mr. Grimes:
This letter is in response to your reference letter of April 20, 1990. The GE Nuclear Energy (GE NE) evaluation has discovered no generic implication for the GE manufacture or refurbishment of circuit breakers because of the GE Model AK-2A-50 circuit breaker failure to automatically trip at Pilgrim Nuclear Power Station. Attached you will find a detailed.
'1 report of the areas that you questioned.
In summary, our breaker overhaul and refurbishment activities-were found to be fully adequate,- including the 1987 Pilgrim maintenance activity.
We evaluated the failed AK breaker, with the missing snap ring.
It was again refurbished by GE.
Then, GE tested it without the
{
snap ring and demonstrated it could operate only for a very few cycles j
(approximately 5) before mechanical failure would occur. Since 1987 this AK breaker was tested and operated a minimum - of. 40 times after 1
maintenance.
Therefore, the recent failure to operate is concluded to be il a failure of the snap ring itself.
Failur,(es of this type are the reason for.the recommendation of periodic yearly) routine inspection, maintenance and testing.
1 Our recent reinspection at Pilgrim revealed no circuit breaker, as left after maintenance by GE in 1987, was in danger of failure to operate 1
L due to variations in hardware configuration.
The listing of all the missing co_mponents, to which your letter refers, was prepared by GE at
-i the utility's specific request, without regard to the significance of the findings.
GE evaluated each of the missing components and has concluded that most of them were inconsequential and none affected breaker operability.
t
'Our manufacturing process is also believed to be fully adediate.
L The reports of " missing parts" for the most part,.in fact, simply represent the way the breakers are configured at the factory to meet only the specific functions (breaker options) requested in the customer order.
l This comparison was to the " standard" configuration of the respective breakers, which was the customer selected benchmark for preparation of the list. Additionally, GE did not initially supply approximately 25% of bbk$06hdOp3 t\\\\
h.
1 s..
s Mr. Brian' K. Grimes Page 2 May 10, 1990
'the Pilgrim breakers.
The breakers have been maintained under the utility's program from the original 1968/69 purchase date.
GE -first
~
- serviced some of the Pilgrim. breakers in 1984.
~
To Jaddress a concern expressed in the NRC audit exit meeting 'of -
l April 26, _1990, GE: NE field maintenance procedures are continuously.
revised to _ reflect the _ latest information we learn through our maintenance. activities.
Specifically, we do now note the "as found" configuration of breakers when we do maintenance.
As' implemented by GE NE, the breaker maintenance program is prudent and reliable.
The minor variations observed' in the Pilgrim installed-breakers are - the result of original variations from- " standard" configuration due to customer selected options; changes in manufacturing practices over time; and non-0EM suppliers. -It is GE NE's position that these variations do not adversely impact operability or reliability of
-the Pilgrim breakers.
In summary, equipment which has been maintained under the GE-NE program is suitable for ' service and is reliable.
This is further evidenced by the entire investigation at Pilgrim Station.
We will advise you of any further findings which reflect on this matter.
Please call if you have any questions or comments.
' Sincerely,
- h. C.
R. C. Mitchell, Manager Nuclear Products Licensing M/C 187, Phone (408) 925-2755 rmw Attachment l
Reference:
Letter from Brian K. Grimes (NRC) to Robert C. Mitchell (GE), dated April 20, 1990.
t i
cc:
E. W. Brach (NRC Rockville) l L. S. Gifford (GE Rockville)
P. W. Marriott (GE San Jose) f
,,, e i
y f
- q s-Mr. Brian K. Grimes.
1 Page 3
- May 10 1990 a
3 ATTACHMENT 1.
AK-2A-50 Circuit Breaker (S/N 224All26-312AE)
The report references the 3/20/90 failure of this circuit breaker I
and attributes the failure to a " missing" snap ring.
On 3/24/90 this breaker was serviced by GE in its Philadelphia-Service Center.
This work was performed under the GE NE QA Program (QAM-002) and-part' of the work scope included determination of the A
root cause of the failure.
It was determined (as stated in the Reference NRC letter) that the
~
breaker failed to operate.
This failure was due to one of the ball bearing assemblies which support the latch prop.
The bearing assembly had fallen. out of the frame causing the prop to be-displaced and jamming the breaker mechanism.
The bearing assemblies are held in place by snap rings.
It was reported to us by BEco personnel that they had not been able to find the snap ring for the bearing and-that they suspected it might have been omitted during GE's maintenance activities in 1987..
In an effort to determine the validity of' this assertion GE' simulated the postulated failure in the service center on 3/24/90.
First, the breaker was reassembled, adjusted and operated per normal procedures.
Then the new bearing snap ring was removed. and the breaker was-cycled.
After three operations the bearing had moved laterally to the point at which it was decided that it would fall out within one or two more operations.
Not wanting to risk damaging the breaker mechanism, the test was discontinued at this point.
i This test was witnessed by both the GE NE Field Engineer and the QC Supervisor.
After the circuit breaker was maintained and serviced by GENE (3/87).
it was operated 25 to 30 times before-being turned over the BECo.
The BECo Post Maintenance Test Program called for the breaker to be cycled in the cubicle before being placed into service.
The breaker was then operated successfully numerous times (estimated by BEco to be a minimum of 20 times) over a three-year period before the failure.
Based on the shop test and years of switchgear experience it is our opinion that the snap ring was definitely in place when the breaker was returned to service in 1987, and that the breaker failure was
l' g:
J.
?
c.
i
,Y Mr. Brian K. Grimes Page 4 May 10, 1990.
ATTACHMENT (Cont')
x therefore the ' result of a snap ring failure.
Neither BEco nor GE personnel were able to locate the snap ring or any broken pieces.
o GE offers to re perform this test for NRC or BEco personnel.
2.
Missing and Misaligned Parts As part of the maintenance and testing of the Pilgrim 480 Volt AK circuit breakers in 1990 a check of the fasteners on each breaker i
was conducted..This check included nuts, bolts, screws, snap rings and related hardware items. The "In-Process Control Sheets (IPCS)"
l for this inspection were prepared on site and were based on station breakers available for inspection.
Deviations from the IPCS were documented on BEco corrective action documents.
Of the 56 plus breakers (and hundreds of fasteners)- inspected only one significant item was identified.
An over-sized snap ring was found on the arcing contact pivot pin of one phase of an AK-50 breaker.- The breaker functioned properly for three years and would
' have continued to do so as the contact configuration did not place a -
lateral stress on.the ring.
Several of the breakers required routine adjustment and/or the replacement of worn and damaged components; however, we found l-nothing that was unusual in this area. The breakers require regular maintenance.
The manufacturer recommends annual' maintenance, at a-L' minimum.
Recognizing the operational cycles of the nuclear power l
plants, GE NE believes that a prudent' maintenance program contains a major overhaul and baseline testing every five years, followed by minor. maintenance and testing every outage (18 months) in. the l
interim.
l In summary, we found no Boston Edison circuit breaker, maintained by z
GE NE in 1987, that was not fully functional (except for normal l
. wear).
The majority of the missing washers were flat washers that are not required for operation.
In fact, most are not installed on current 1;
production equipment.
Other hardware variations were of no conse-quence; many were noted on-breakors not supplied by GE to the station (refer to Item 3 below).
l 3.
Control of Manufacture and Maintenance i
Regarding the configuration of the Pilgrim Station 480-volt AK circuit breakers the following points should be considered.
peg, e -
p m
>;d :
,f.
- r~
Mr. Brian K.. Grimes L
Page 5 May'10, 1990' eF ATTACHMENT (Cont')
[
b 1)
The original plant equipment was purchased as commercial equipment in.1968/69 which would be consistent 'with industry practices at that time.
2)
Application of' this equipment to safety related service was under utility control as would be the case for other plants of this vintage.
3)
Twenty to twenty-five percent (20-25%) of the station breakers were supplied to the station by sources other than GE.
4)
GE was first contracted to perform maintenance on this e
equipment in 1984.
.5)
GE first supplied safety-related breakers to Pilgrim in 1986-87.
GE NE has not had control of this breaker equipment nor has GE ever been asked to standardize their configurations. At the beginning of the '90 maintenance work we were directed by BEco to note any and all hardware variations and correct those that could be done within the outage schedule.
However, our primary goal was - to assess the operability of the equipment and confirm for BECo its suitability for continued service.
This utility direction is consistent with current industry practice.
All maintenance performed by GE on the Pilgrim breakers (1984, 87 &
l
- 90) was accomplished in accordance with a BECo-approved GE' proce-dure.
The '84 and '87 work.wps governed by the GE QA Program with BEco QC overview.
The '90 work was under the BECo QA Program.
- There has never been a lack of formal. Quality Control on GE NE performed maintenance.
The procedures GE used each time reflect the 0.E.M.'s printed L
recommendations, proprietary design data, and field services' cumulative experience with this type equipment.
Never previously have we been-asked to compare the exact hardware configuration of L
this equipment.
We have only been asked to confirm the equipment's operability and reliability, n
The GE NE procedures have been continuously revised to reflect l
" Lessons Learned" and industry trends, including specific hardware checks on the complete equipment.
i i
l
,s
..