ML20057E674
| ML20057E674 | |
| Person / Time | |
|---|---|
| Issue date: | 09/23/1993 |
| From: | Cwalina G, Naidu K Office of Nuclear Reactor Regulation |
| To: | |
| Shared Package | |
| ML20057E671 | List: |
| References | |
| REF-QA-99900404 NUDOCS 9310130020 | |
| Download: ML20057E674 (8) | |
Text
.
ORGANIZATION:
Westinghouse Electric Corporation Forest Hills, Pennsylvania REPORT NO.:
99900404/93-01 CORRESPONDENCE Westinghouse Electric Corporation ADDRESS:
P.O. Box 355 Pittsburgh, Pennsylvania 15230 l
NUCLEAR INDUSTRY Nuclear steam supply system components and services l
ACTIVITY:
l INSPECTION August 16 - 17, 1993 l
CONDUCTED:
1 1
f45 C R:
i Kamalakar R. Naidu, Date Reactive Inspection Section-2 (RIS-2)
OTHER INSPECTORS:
F. H. Burrows, Electrical Engineering Branch l
APPROVED BY:
MM/1 7
8 de Gregory Cwalina, Section Chief 6at e '
f RIS-2, Vendor Inspection Branch INSPECTIOf4 BASES 10 CFR Part 21,10 CFR Part 50, Appendix B INSPECTION SCOPE Control of activities related to the re-manufacture of X-relays and other components installed in DB and DHP type 480 Volts breakers PLANT SITE All plants utilizing DB and DHP type 480 Volts APPLICABILITY:
circuit breakers 1
9310130020 931007 PDR GA999 ENVWEST 99900404 PDR
1 INSPECTION
SUMMARY
i 1.1 VIOLATIONS i
No violations were identified during this inspection.
1.2 NONCONFORMANCES No nonconformances were identified during this inspection.
l l
2 INSPECTION FINDINGS AND OTHER COMMENTS l
l 2.1 Backaround Information 1
On June 27, 1993, power to a motor control center (MCC-5) was lost at Connecticut Yankee Atomic Power Company's Haddam Neck nuclear power plant (Haddam Neck). On June 29, 1993, the Nuclear Regulatory Commission (NRC) dispatched an Augmented Inspection Team (AIT) to investigate the event.
The AIT determined that a possible root cause was a failure of the X-relay to reset (move to the de-energized position) in a DB-25 circuit breaker manufactured by Westinghouse Electric Corporation (H).
The X-relay is usually denoted by the symbol 52X in the electrical control circuit schematic diagram for the breaker. On receipt of a signal to close the breaker, the X-relay energizes and one set of its normally open contacts close to enable momentary energizing of the breaker's closing coil.
After the breaker closes, the same set of the breaker's X-relay contacts open to deenergize the closing coil even though the X-relay remains energized by the close signal.
If the close signal is still present when the breaker trips, the X-relay serves to inhibit repeated closure attempts until the close signal is removed.
Thus, it provides anti-pump protection to the breaker by preventing repeated breaker closure attempts when a continuous closure signal i
exists after a breaker trips.
i The armature assembly (see Figure 1) of an X-relay which fits inside a brass sleeve is surrounded by the relay's electromagnetic coil.
When the coil is energized, the plunger (moving core) is drawn up towards the top cap piece (stationary core) of the assembly and the latch arm operates the relay's contacts.
In its uppermost position an air gap is maintained between the bullet-shaped top of the plunger and the cavity in the cap piece with the shoulder of the plunger mating with the lip on the cap.
When the coil is de-energized, the moving core is designed to fall, by gravity, into its lowermost position.
The failure of the moving core to fall to its lowermost position is the possible f ailure mode of interest at Haddam Neck.
Possible causes of the failure mode of the X-relays discussed during the AIT are-Residual magnetism which can prevent the moving core from falling to its lower-most position.
2 i
I
.1
^
Mechanical binding due to dirt accumulation between the moving core and the brass sleeve.
Mechanical binding due to different coefficents of thermal expansion caused by heat dissipated by the energized coil-.
Dimensional variations of the X-relay frame.
i On September 9, 1993, at W, in the presence of the NRC, a Haddam Neck engineer was able to cause the two parts to adhere to each other and prevent the moving core to fall freely by gravity by applying pressure to the plunger and the top cap piece of the armature while twisting them together. W and the licensee believe that residual magnetism or mechanical adherence are two of the probable causes for the parts sticking together.
During the AIT inspection, the (NRC) staff searched the failure history of W control relays.
The search revealed that there had been at least 27 reported failures of the X-relay during the past nine years with failure modes similar to the recent one at Haddam Neck.
The causes of these failures were reported to include dirt, aging, mechanical misalignment, or mechanical binding due to burrs. Corrective actions were usually replacement, repair or readjustment of the X-relay.
Two of the 27 reported failures are discussed in paragraphs 2.2.1 and 2.2.2 of this report.
2.2 Discussion of W Policy to Reoort Conditions Adverse to Safety W's Procedure OPR-19.0, " Identification And Reporting Conditions Adverse to Safety," Revision 4, of October 1,1992, establishes the policy for identifying and evaluating potential conditions adverse to safety, and reporting such conditions to the appropriate parties.
To accomplish this, W has established a Safety Review Committee to evaluate conditions adverse to safety and determine if they should be reported to responsible management.
When W becomes aware of a condition adverse to safety, W staff opens a potential item (PI) file on the subject and documents all the information including evaluations and the final disposition which may include informing the NRC that it is a 10 CFR Part 21 item. W personnel stated that they did not retain the records of such evaluations performed before 1991 in which W determined that it was not adverse to safety. Additionally, W does not track or trend information provided from failure history reports such as those provided by the Nuclear Plant Reliability Data System.
W personnel stated that they do not routinely receive information on failures experienced at non-W reactor sites.
Furthermore, if the failure mode of the breaker is not adverse to safety, there is no mechanism for W personnel to learn about the failure.
For instance, if a reactor trip breaker f ails to close because the X-relay failed, W does not considered it safety significant because the safety function of a reactor trip breaker is to open on demand.
The inspectors discussed with W personnel the bases for which the following two events were not reported as 10 CFR Part 21 items.
l 3
- ~ _ _ _ _ ;
i 2.2.1 Review of M Activity Associated with the Haddam Neck X-relay Failure in 1984:
License Event Report (LER)84-023 forwarded on November 28, 1984, I
from Haddam Neck to the NRC-reported six incidents. between July 1,1984, and t
August 24, 1984,.in which one W DB-25 and five DHP-250 type breakers failed to close when required.
Five of those failures were directly attributed to X-relay malfunctions.
The sixth breaker failure possibly resulted from a 1
X-relay malfunction. The main cause of the control relay malfunctions was i
stated to be dust or dirt accumulation on the moving core and its latch arm assembly.
Since the licensee concluded that the malfunctions presented a i
generic problem in the plant, the immediate action was to inspect and clean l
all X-relays, j
In a letter of November 5,1984, to Northeast Utilities, W stated _ that it was l
in the process of evaluating and issuing a significant Safety Hazard Report j
pursuant to the requirements of 10 CFR Part 21 on X-relays.
Records indicate j
that W did not submit a 10 CFR Part 21 report on the X-relays..W's letter l
dated April 29, 1985, to the Connecticut Yankee Atomic Power Company ~(CYAPC) l provided an evaluation of the six X-relay failures and ~ determined that the l
X-relay failed in only one safety-related breaker _ and therefore H did not consider it significant or a generic design problem. M also stated that the X-relays at Haddam Neck may not.have been. addressed in the plant's preventive maintenance program. M concluded that the maintenance program being implemented at the plants should be continued and that inspection and cleaning t
of the moving parts in the X-relay armature assembly should be included.
On September 2, 1993, the NRC called W to express concern over M's limited evaluation of the X-relay problem as documented in-its April 29, 1985, letter to CYAPC.
In the absence of other information, the NRC stated that regardless j
of the application of the failed X-relays (used in DB or DHP type breakers),
and whether they were used in safety or nonsafety-related applications, M should have evaluated the-potential consequences pursuant to the requirements of 10 CFR Part 21, particularly since W knew that X-relays were used in i
safety-related circuit breakers. A H representative stated that in 1985 there was no requirement to retain records of evaluations and therefore he could not determine if there were other reasons why W did not consider the failures were reportable under 10 CFR Part 21. The W representative stated that currently they perform more detailed evaluations.
2.2.2 Review of M Activity Associated with the X-relay Failure at Oconee Nuclear Power Station (0conee):
LER 92-002 of September 3, 1992, from Oconee reported that in June 1991 X-relays in the W DB-25 breakers in the Oconee emergency hydro units' field and field flashing circuitry failed..The cause of the specific failure mode was not determined and the nonsafety-grade relays were replaced with safety-grade relays.
Subsequently, on January 28,1992, a safety-grade X-relay failed to reset. As immediate corrective action the licensee inspected each X-relay to ensure that they did reset following each shutdown. A design change has now been implemented to replace the electro-mechanical anti-pump scheme provided by the X-relay with an electrical scheme.
j One of the X-relays from Oconee was sent to W for examination.
It was subsequently determined that the most likely cause of failure to reset was stray magnetic fields resulting from dc currents flowing near the relays.
In response to concerns associated with the X-relay failures at Oconee, y added a 4
, _. _ -.,. ~. _.
test to CD1 CEB-0108 (W X-relay dedication instruction) to ensure that the X-relays being dedicated as Class lE would drop out when de-energized af ter being energized for a period of time.
2.3 Review of the Purchase Order (PO) from Haddam Neck The inspectors reviewed the records pertaining to the recent (just prior to the June 27, 1993, event) procurement of the X-relays by Haddam Neck from W and determined the following:
I Haddam Neck issued Purchase Order (PO) No. 945288 of May 17, 1993, to W for the supply of seven 125 V dc X-relays, W part No. 33A2746G32 suitable for DB-50 and DB-25 type circuit breakers. The P0 stated that the X-relays were QA Category 1, intended for installation in a nuclear-grade motor control center and required that the packaging, shipping, storage and handling of the items were to be in accordance with W Specifications OPR 405-5 and WCAP 9245.
The Assembly and Test (A&T) division of W purchases commercial-grade spare parts, and dedicates and sells them to licensees as safety-grade items.
Records indicate that when Haddam Neck received the seven relays, they were inspected and the results were documented in Haddam Neck's Material Receipt Inspection Report (MRIR)92-399 of June 16, 1993. The team reviewed MRlR 92-399 and determined that seven X-relays were received, inspected and determined to be acceptable.
W's A&T department had issued a certificate of conformance (CoC) dated June 3, 1993, certifying that six 125 Vdc X-relays with serial No. 930.286-1 to -6 meet the drawing requirements of Revision 46 of Drawing 33A2746G32 and that the items were in compliance with W's NSD Quality Assurance Program Plan, WCAP 9245 and OPR 405-5.
A CoC of April 28, 1993, similarly certified one more X-relay, bringing the total of X-relays supplied to seven.
The inspectors reviewed the W quality assurance records and determined that the seven X-relays described above had been drawn from a batch of 24 supplied by Westinghouse Electric Supply Company (WESCo), Murraysville, Pennsylvania.
It is W A&T's policy to issue P0s to WESCo to procure commercial-grade items I
manufactured by W affiliated companies.
W A&T PO No. MA 13734M of August 11, 1989, to WESCo, Murraysville for 24 X-relays. WESCo in turn issued a P0 to W Greenwood, South Carolina to supply 24 X-relays.
In its P0, W specified that the X-relays should conform to Drawing No. 33A2746G32, Revision 46 "DB Single Pole Relay Assembly."
W A&T did not audit their supplier because the X-relays are purchased as commercial-grade items; it relies on its sister company to meet the requirements of the purchase order.
W's ALT quality control inspectors (QC) inspected the 24 X-relays received from W Greenwood through WESCo utilizing Engineering Control Instruction (ECI)
DAR-062185-01, Revision 01, and documented twelve adverse findings in Material Deficiency Report (MDR) 11206 of June 16, 1990.
A&T technicians corrected ten adverse findings in MDR 11206; the remaining two were accepted as-is.
On November 11, 1991, a quality assurance (QA) representative inspected the 5
l corrective action taken to repair the ten adverse conditions and use-as-is items and determined them acceptable.
Subsequently, the relay frames (Part No. 23A3609H01) were inspected to 1
commercial-grade dedication instruction (CDI) CEB-108 which had an additional requirement to perform metallurgical tests on the relay frames. Metallurigal l
reports from the material supplier and an independent test laboratory documented that the frames of all 24 X-relays were of incorrect material.
These frames along with two others were returned to the vendor. On March 12, j
1993, M A&T issued P0 No. MA-72697-M to Greenwood to supply replacement frames. These relay frames were used to replace the ones that had been supplied with the X-relays.
Following the frame replacements, all 24 X-relays were determined to be acceptable, and stored at H's A&T storage facility.
The inspectors reviewed the records during the inspection and determined that A&T has established accountability for the acceptable X-relays.
2.4 Observation of Activities The NRC team, accompanied by W staff, toured the A&T facilities and observed the following:
A&T personnel demonstrated the operation of DB-25 and DB-50 circuit breakers; the inspectors were able to observe the contact movement of i
the X-relays when the breakers operated. Also, the test to measure the X-relay dropout voltage was demonstrated. H stated that residual magnetism is the most likely cause of the X-relay's failing to reset and the addition of a brass shim between the relay plunger's shoulder and cap piece lip to increase the armature assembly's air gap is being consider as corrective action.
The reactor trip breakers (RTB) from the Point Beach Nuclear power plant were being refurbished.
A&T personnel showed the inspectors the cubicle assembly which was used to replace AKR type circuit breakers manufactured by General Electric l
Company with H DB-416 type breakers for the Palo Verde Nuclear Generating Station, (Palo Verde) Unit 2.
l Each set of four RTBs installed at Palo Verde Station Units 1, 2 and 3 consisted of a pair each of M's D5-206 and GE's AKR type circuit breakers.
Palo Verde issued a P0 to M A&T to supply AKD switchgear with four DS-416 type circuit breakers to replace the existing RTBs. H l
informed the inspectors that it had completed the manufacture and l
l testing of one AKD switchgear for installation at Palo Verde's Unit 2 station and had shipped it with all the qualification reports.
6 I
I W demonstrated to the team its Digitrip RMS type overcurrent trip i
- devices,
,j I
W showed the team A&T's automated storage system. The storage area of acceptable components appeared to be clean and free from rodents and debris. A&T personnel control the climatic conditions inside the storage area and access to it.
In addition to the above demonstrations, W stated that it had replaced the EC I,
type overcurrent devices manuf actured by General Electric Company on AK2-25, i
-50 and -75 type GE circuit breakers at the Niagara Mohawk and Monticello nuclear power plants with its Amptector 1 A type overcurrent protection gilj devices.
i I
l 3.
PERSONS CONTACTED Individuals contacted during the inspection are listed below:
NAME TITLE George Dillon Manager, Nuclear Services Division l
1 Tom Moser Manager, Replacement Component Services (RCS)
Dick Miller W Nuclear Safety Joe Jelovich Manager, Power Systems Engineering (PSE)
T.M. Wambaugh Quality Assurance Engineer J.J. Evans Quality Assurance Manager Jeff Black RCS Audit Coordinator
+ Dale Rygg Manager, RCS Strategic Operations.
Dave Riffe Engineer, RCS/PSE
+ WAS PRESENT ONLY AT THE EXIT MEETING ON 8/17/93 l
l 7
1 s ', ) ~ 's
/
/
l N
\\
{l e 1
,' p
\\ \\
\\ '
' s%.,_ -
e
_ !. /
i l
i i
,/
'g.
y CAP PIECE (STATIONARY CORE)
/
\\
\\
3
\\
UP L
J e
\\
C'3
/ R
' PLUNGER (MOVING CORE) f g
gi j
l l
.\\'
r3
/
N T
i 1
\\
1
\\
g LATCH ARM I
N s
g 1
X
\\
l i
i
/
/
I f
I I
i l
l t
A. Sub-Assembly B. Cross Section Figure 1. X-relay Magnetic Core