ML20215B968
| ML20215B968 | |
| Person / Time | |
|---|---|
| Site: | Hope Creek, Columbia, 05000000 |
| Issue date: | 09/26/1986 |
| From: | Driscoll H LOUIS ALLIS CORP. (DIV. OF MAGNETEK, INC.) |
| To: | Forney W NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION III) |
| References | |
| REF-PT21-86-341-000 PT21-86-341, PT21-86-341-000, NUDOCS 8610100010 | |
| Download: ML20215B968 (9) | |
Text
I I
h h% Nb ALouisAllis' A Dwso m u n d.*rc Drives & Systems,16555 West Ryerson Rd., New Berlin, WI 53151 414-782-0200 TWX 910-262-3115 26 September 1986
'f
, A '\\
-s rg[ 3q D
U.S. Nuclear Regulatory Commission 799 Roosevelt Road 5
Glen Ellyn, IL 60137 Attention:
William L. Forney - Section lA
Subject:
10CFR21 Investigative Report Refe rence :
Louis Allis 10CFR21 Notification August 22, 1986 Re:
Synchro-Start Speed Switch Type ESSB Gentlemen:
As provided for in the referenced 10CFR21 Notification, enclosed is a report of investigation and proposed corrective action as may be required for possible defective operation of Synchro-Start Type ESSB Speed Switches at facilities listed in the notification.
Based on the situation, Louis Allis Drives & Systems has taken the following actions:
1.
Notified the NRC by telephone on August 20, 1986 followed by written 10CFR21 Notification on August 22, 1986.
2.
Notified by copy of 10CFR21 Notification the users shown in the 10CFR21 Notification.
3.
Proposed tests in the attached report to determine and take corrective action if a problem exists.
If there are any further questions, please contact the writer.
nl Harlan Dri coll Military / Nuclear Products Louis Allis Drives & Systems HD/af Attachment 8610100010 860926
{DR ADOCK 05000 4
2 G86' rh C6/9
26 September 1986
}
- ~
Report of Investigation into Potential.
Defective. Operation of Synchro-Start Type ESSB Speed Switches at Nuclear Facilities
1.0 INTRODUCTION
This report is submitted as provided for in August 22, 1986 10CFR21 notification issued by Louis Allis Drives and Systems concerning potential defects in Synchro-Start speed switch type ESSB operation.
1.1 ' PURPOSE The purpose of this report is to detail the potential problem (s) found with IE qualified Synchro-Start speed switch type ESSB and to state recommended solutions for resolving the problems as they may affect 1E nuclear installations.
2.0 DESCRIPTION
OF PROBLEM 2.1 HISTORY Louis Allis D&S received a spare parts order from Colt Industries fik (Colt P.O.
482932-A, Revision 1, 6/12/86) for three ESSB Synchro-OI Start speed switches to be used for the Hope Creek nuclear power
' plant.
This type of unit had previously.been qualified by Colt Industries (Wyle Lab Report 47114-1) and dedicated for 1E ap-
.plications.
During the Louis Allis incoming inspection of the speed switches, it was noted that two of the switches exhibited higher than specified relay contact resistance and that when the speed switch was tapped, the relay contact resistance would vary.
Normal contact resistance would be less than 100 milliohms for r
2.1 HISTORY (Con t ' d)
J a new relay, and 50 to 500 m1111 ohms and steady for one in service. The' actual readings observed were approximately 5 to 20 ohms and erratic.
As a result of this situation, the Louis Allis OA Department rejected the switches and paid a visit to the Synchro-Start Company in Skokie, Illinois to investigate the observed problem and to perform a limited quality survey.
At the meeting Synchro-Start stated that they had been experiencing extensive on-going quality problems with R10 series relays for some time.
They gave Louis Allis a copy of a Potter & Brumfield Fail-ure Analysis (P&B Report 018609-2) made on 43 defective relays re-turned to Potter & Brumfield in January 1986.
Synchro-Start could not state or produce records to show whether the defective relays were from production testing or from field returns.
The P&B Re-port showed the following findings on a test sample of 13 relays picked by P&B from the 43 relays of similar type (R10):
A.
Intermittent contact closures caused by phenolic dust con-tamination and tarnish on surfaces of contacts.
(3 of sample of 13).
B.
High relay armature pull-in force required.
(4 of test sample of 13).
C.
Relays would dot operate because of open coil wires.
(5 of sample of 13).
D.
Stationary contact missing. (1 of sample of 13).
Synchro-start stated that as a result of the problems with P&B relays they have decided that for future production they will be changing to another type relay to be obtained from American Zet-tier.
2.1 HISTORY (Cont'd) ivg.
The vendor survey of Synchro-Start also revealed that the Synchro-4 Start letter to Louis Allis of 4-3-85 that certified that Synchro-Start had tested the P&B RIO relay for a contact life exceeding 50,000 operations could not be substantiated by test documentation.
The speed switch had been qualified by Louis Allis D&S for 40 year life based on this letter and on the original Colt Qualification Report (Wyle Lab #47114-1).
The Synchro-Start Company has been suspended from the Louis Allis D&S Qualified Vendor List based on deficiencies noted during the August 12, 1986 Louis Allis lim-ited survey of their quality assurance program and it's implemen-tation as applied to the model ESSB speed switch.
3.0 INVESTIGATION OF SYNCHRO-START REPORTED RELAY DEFECTS 3.1 P&B FAILURE ANALYSIS REVIEW A review of the P&B Failure Analysis Report 018608-2 of January 9,
1986 covering the 43 defective relays by Synchro-Start to P&B has been made.
The report defects were traced by P&B tests to the following causes:
A.
Intermittent contact continuity -
Causes:
Phenolic particle contamination and tarnish at mating surfaces of contacts.
Particle contamination due to use of laminated phenolic in relays until August 1985.
P&B states that this material was the only UL and CSA approved material until 1984.
B.
High relay armature pull-in force required for closure -
Causes:
Corrosion in relay hinge area caused by excessive process fluids entering relay during speed switch manufacture..
~
r-3.1 P&B FAILURE ANALYSIS REVIEW (Cont'd)
C.
Relays do not operate -
Causes:
Open coils attributed by P&B to coil lug movement during user handling greater than allowable.020".
Relay design limits the lug movement before strain or breakage can occur.
3.2 LOUIS ALLIS ENGINEERING INVESTIGATION AND ANALYSIS The high relay contact resistance, up to 20 ohms on the Louis Allis rejected Speed Switches, was determined to be due to tarnish.
In an I
operating system, a resistance of this magnitude in series with 6K ohms or greater and with an applied coil voltage of 125VDC, will not be a significant factor and this contact resistance should, if due to tarnish, decrease with use (repeated contact closures).
The phenolic flaking problem (Synchro-Start rejects to Potter & Brumfield) which resulted in relay contacts becoming int.ermittent or open, was due to flakes falling onto or between contacts.
The orientation of the Speed Switch on the B.P'.S./ Louis Allis switchgear should minimize the problem in operating equipment.
4.0 AFFECTS ON 1E QUALIFICATION The Louis Allis qualification of the Synchro-start speed switch to approximately 40 years has been placed in jeopardy because Synchro-Start could not produce test data to verify their letter of 4-3-85 stating an indicated test contact life exceeding 50,000 cycles.
Potter & Brumfield also has been contacted and asked for confirming 9
test data for the claim of a contact life of 7.5 X 10 (oper-ations) life.
They could not produce test data to substantiate their published contacts life claim.
Thus, the only valid con-tact operational test data for the speed switch (Potter & Brumfield relay) contact is the 1,037 operations reported in the 1E Oual-ification report (Wyle Labs No. 47114-1) performed by Colt Indus-l tries.
This will reduce the expected life of the Hope Creek speed 4.0 AFFECTS ON 1E QUALIFICATION (Cont'd) switch to 5.5 years, and to 8.2 years on the WPPSS and Marble Hills contracts.
Colt. Industries has qualified the speed switch for its application to 6.9 years.
5.0 NUCLEAR FACILITY SITES AFFECTED Beloit Power Systems or it's successor, Louis Allis D&S, supplied the Synchro-Start speed switches through Colt Industries for the following sites.
Power Facilities Quantity 1.
Washington Public Power 2 Original Equipment Supply System (WPPSS) 2 Excess 1 Seismic Replacement 2.
Public Service Co. of 4 Original Equipment Indiana Marble Hill 3.
Public Service E&G Co.
4 Original Equipment Hope Creek #1 4 Excess 6 Spare Parts 23 Total The units designated as original equipment were mounted on BPS-Louis Allis built equipment and shipped to Colt Industries for shipment to the respective plant sites.
The units designated as excess were originally bought for panel installations but be-cause of contract cancellations became excess.
These devices were bought by Colt Industries for their use.
The one seismic replacement unit replaced per contract requirements for a device used for seismic testing.
The units designated as spare parts were supplied to Colt as spares for a specific plant site.
7 6.0 DETECTION OF DEFECTIVE DEVICES 6.1 SPEED SWITCHES AT UTILITIES It is recommended that the speed switches now in service or des-ignated for possible service be tested for proper relay contact closure and relay contact condition.
(Contact resistance).
6.1.1 RECOMMENDED TESTS 6.1.2 OFF LINE SPEED SWITCHES (PLANNED SHUTDOWN)
This is the preferred test method since it allows use of a type of test ohmmeter designed to accurately check relay contact resis-tance under current carrying conditions.
Proceed as follows:
1.
Disconnect all leads to the speed switch except those to the tachometer output.
(The test may be made as a bench test without the tachemeters connected for an RPM indicator.
If this is done, connect a 4.0K ohm, 5 watt resistive load a-cross the tachometer output terminals of the speed switch).
2.
Connect a 125VDC supply to the speed switch plus and minus supply voltage terminals.
3.
Connect a signal generator capable of simulating the tach-ometer relay signal (for voltage and frequency output) to the tach relay signal generator terminals of the speed switch.
4.
Connect a low resistance reading current sourcing ohmmeter (Valhalla 4300ATC or equivalent) across the two speed switch terminals associated with the relay set to energize at the lowest RPM setting for the switch.
5.
Operate the speed switch by increasing signal generator fre-quency to lowest RPM setting of the speed switch.
(Note that the speed switch interprets a two hertz output as equivalent to one RPM). The relay should energize closing its contacts and causing the ohmmeter to indicate a steady reading of 500 m1111 ohms or less.
Switch the signal generator signal to the speed awitch on and off to obtain 10 readings.
All 6.1.2 OFF LINE SPEED SWITCHES (PLANNED SHUTDOWN)
(Cont'd) 5.
(Cont'd) -
readings should be stable after' contact closure and 500 milli-ohms or less.
The following are defect indications:
(1)
Ohmmeter reads more than 500 milliohms and/or erratic readings are obtained.
(2) _ Relay fails to close contacts.
This is an indication of an open relay coil or a severely corroded relay-hinge.
(3)
Relay operates intermittently.
This is an indication of a corroded relay hinge.
6.
Repeat steps 1 thru 5 with signal generator set to operate speed switch for each set of relay contacts (lowest to highest RPM).
6.1.3 TEST PROCEDURE FOR ON LINE SPEED SWITCHES Synchro-Start type ESSB speed switches mounted on site in an op-erational-equipment may be tested during generator start-up.
Proceed as follows:
(1)
Connect a calibrated 5 digit digital voltmeter (Fluke 884,0A or equivalent) across the two contact terminals associated with the relay contact set, that is to close at the speed switches lowest" RPM setting.
Observe the meter to read approximately 125VDC.
Disconnect voltmeter.
(2)
Monitor the tachometer during generator start-up.
When the low-est speed switch RPM setting is reached or surpassed, the relay will energize closing its contacts and energizing the relay circuit in series with the contact.
Connect the volt-meter.
It should now read less than 0.lVDC.
(3)
Repeat the above procedure ten times for each set of con-tacts on the speed switch progressing from the lowest re-maining RPM setting contacts to the highest.
Voltage readings should stabilize at less than 0.1 volts. (Note: If the upper speed switch setting is an overspeed condition it may not be possible to test the highest RPM contacts under normal operating conditions.
In this case, test as in Section i
l
- i
'6.1.3 TEST PROCEDURE FOR ON LINE SPEED SWITCHES (Cont'd)
.(3)
(Cont'd) -
6.1.2 for off line switches). Failure indications include any condition where a steady DC voltage of greater than 0.1 VDC is read during speed switch relay closures.
7.0 CONCLUSION
S AND CORRECTIVE ACTION 7.1 REPAIR / REPLACEMENT OF SPEED SWITCHES If it is concluded, per recommended tests in Section 6.0 above, that a defective speed switch (per Section 5.0) exists, the switch should be returned to Louis Allis.
Louis Allis will work with its customer for a suitable solution.
Louis Allis will provide testing per Section 6.0 of this report for any returned speed switches specified in Section 5.0.
A return authorization number for re-turns for the purpose of tests will be supplied upon request.
7.2 LOUIS ALLIS OUALIFICATION OF SYNCHRO-START MODEL ESSB SPEED SWITCH Because supporting test data for 50,000 relay contact test oper-ations is unavailable, the 1E Qualification Report speed switch qualified life for Hope Creek, Marble Hill and WPPSS is being rescinded.
A new qualified life of 5.5 years for Hope Creek and 6.9 years for Marble Hill and WPPSS will be established based on Wyle R (47114-1) 1,037 test operations.
Harlan Driscoll
-Quality Assurance Manager Military / Nuclear Products Louis Allis Drives & Systems 16555 W. Ryerson Road i
New Berlin, WI 53151 414-782-0200 _
. _., -. ~
-