ML20210C789
| ML20210C789 | |
| Person / Time | |
|---|---|
| Site: | Vogtle  |
| Issue date: | 01/10/1987 |
| From: | Rice P GEORGIA POWER CO. |
| To: | Grace J NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION II) |
| References | |
| REF-PT21-87, REF-PT21-87-041-000 GN-1313, PT21-87-041-000, PT21-87-41, NUDOCS 8702090458 | |
| Download: ML20210C789 (7) | |
Text
me=9r I
$$NI?cI![NEbe=2 3413 404 724 *D N E - sas Geolgia Pbwer
, y p' u 4n fo 4 $,e no.na.
n V(e Pasdent m e,e ff "w
m January 10, 1987 United States Nuclear Regulatory Commission File: X7BG03-M134 Region II, Suite 2900 Log:
GN-1313 101 Marietta Street, Northwest Atlanta, Georgia 30323
Reference:
Vogtle Electric Generating Plant - Units 1 & 2; 50-424, 50-425; Solid State Protection System Relays Letter GN-1266, dated December 29, 1986 Attention:
Mr. J. Nelson Grace In previous correspondence, Georgia Power Company notified the NRC of a potentially reportable condition associated with the Solid State Protection System (SSPS) relays.
Georgia Power Company has completed its reportability evaluation and has determined that a reportable condition as defined by the reporting requirements of 10 CFR Parts 21 and 50.55(e) does exist.
Based upon NRC guidance in NUREG-0302, Revision 1, and other NRC correspondence, Georgia Power Company is reporting (this condition pursuant to the reporting requirements of 10CFR50.55e). A summary of our evaluation is attached.
Georgia Power Company initiated a two-part evaluation program regarding the adequacy of the SSPS relays in applications which exceed their published rating.
The initial part is based on engineering analysis and evaluation of test data for the currently installed relay configurations.
This evaluation is complete and has shown that the SSPS relays as currently installed on Vogtle Unit 1 are adequate for the first fuel cycle (approximately two years) and meet regulatory requirements.
The second part will be an assessment for adequacy of the SSPS relays for the long term.
The results of the long term assessment will be provided to the NRC no later than November 1,1987.
At this time it is not known that any changes will be required.
However, any corrective action identified as being required to be implemented prior to the end of the second cycle of operation will be completed prior to startup for the second cycle of operation.
Other required corrective actions will have specific dates by which they must be implemented.
Nf0$0$ NO 4
S hYn.
M
l 1
linited States Nuclear Regulatory Commission R$gion II, Suite 2900 101 Marietta Street, Northwest Atlanta, Georgia 30323 Page two i
This response contains no proprietary information and may be placed in the U NRC Public Document Room.
Yours truly, l
Yk l
P. D. Rice PDR/wkl l
Attachment xc:
U. S. Nuclear Regulatory Commission Document Control Desk Washington, D. C.
20555 H. G. Baker J. F. D'Amico J. P. O'Reilly W. D. Drinkard G. F. Head C. C. Garrett (OPC)
R. E. Conway D. Feig (GANE)
R. H. Pinson L. T. Gucwa B. M. Guthrie C. W. Hayes R. A. Thomas G. A. McCarley D. R. Altman R. W. McManus J. A. Bailey Sr. Resident (NRC)
G. Bockhold, Jr.
J. E. Joiner (TSLA)
NORMS l
EVALUATION OF A POTENTIALLY REPORTABLE CONDITION SOLID STATE PROTECTION SYSTEM RELAYS INITIAL REPORT:
On December 4, 1986, Mr. C. E. Belflower, Quality Assurance Site Manager-Operations, notified Mr. M. V. Sinkule of the USNRC Region II of a potentially reportable condition associated with 7
the solid state protection system relays.
In subsequent correspondence with the NRC, Georgia Power Company indicated that a final report on this issue would be submitted by January 16, 1987.
BACKGROUND INFORMATION:
In preparation for the Unit 1, Train A Engineered Safety Features Actuation System (ESFAS) test, the individual slave relays in the Solid State Protection System (SSPS) were tested to insure that they would function as intended, i.e.,
actuate the devices controlled by their output contacts.
Slave relay K622 was tested and all the devices controlled by the relay successfully actuated.
But, when the reset signal for relay K622 was initiated, it failed to reset.
A replacement slave relay for K622 resulted in the same failure mode.
Relays K624 and K748 also failed to reset when the reset signal for each of these relays was initiated during the ESFAS test.
Relays K622, K624 and K748, along with the other slave relays in the SSPS, are Potter and Brumfield (P & B) MDR rotary type 4121-1 relays.
Relays K622, K624 and K748 control a combination of three different loads:
ASCO solenoids (having a rated load current of 0.14 amperos),
Valcor series 200 or 300 solenoids (having respective rated load currents of 1.025 and 1.5 amperos), and P & B type HDR 137-8 auxiliary relays (having a rated load current of 0.082 amperes).
Those rated load currents are for 125 volt de and will increase on a proportional basis with an increase in dc voltage level.
P The SSPS MDR relay contacts are rated by the NSSS supplier (Westinghouso) to interrupt an inductive load current of 0.3 ampores at 143 volts dc.
Examination of the failed SSPS relays indicated that the failure of relays K622 and K748 to roset was due to the relay contact follower arms becoming embedded into the molted nylon rolay contact operating cam.
This mechanically provented the rolay rotor from rotating to the resot position.
The K622 contacts were observed to be in the open (or non-roso t) position.
The contacts for K748 woro observed to be in the closed position.
Review of the loads associated with each of the molted cams indicated that the failures (molting) had occurrod only with contacts that controlled Valcor solenoid valves.
Rolay K624 was examinod and datormined by test to be fully operable.
Its failure to reset during testing was found to be duo to the open circuiting of its roset circuit during the removal of tho original K622 relay and prior to tho installation of the replacement K622 relay.
PE0187012/SL.
f i
ENGINEERING EVALUATION:
Analysis indicates that the overheating of contact follower arm on relays K622 and K748 was induced by heat generated from arcing across the relay contact surfaces.
The arcing was due to the inability of the contact to interrupt the magnitude of the Valcor solenoids inductive load current.
The overheating damage to the cam surface of relay K748 was sufficient to allow its contacts to reclose to the reset position even though relay cam / rotor rotation to the reset position had not occurred.
l Based upon this analysis it has been determined that the root cause of the SSPS MDR relay failure to reset was due to the application of a relay contact in a circuit where its de inductive current interrupting capability was exceeded by the load current of the i
controlled device, in this case the Valcor solenoids.
i ANALYSIS OF SAFETY IMPLICATIONS To evaluate the safety implications of relay failures, the control circuits associated with a single MDR relay contact de-energizing a Valcor solenoid coil were analyzed.
The analyses included the control circuits of the failed slave relays K622 and K748 and those of other MDR relays within the SSPS and auxiliary relay panels which employ Valcor solenoid coil control circuits.
The solenoid valves associated with the MDR relays, including their l
corresponding normal and failure modes of operation were identified.
These solenoid valves are used to isolate the liquid, gas, and air sampling and monitoring lines within the containment from the corresponding lines outside the containment following an accident.
These valves are required to close and remain closed following an accident except for the containment hydrogen monitor system valves.
These valves are required to be operable following an accident to allow the monitoring of containment hydrogen concentration.
Failure of the slave relay contacts in the closed position would prevent opening of the contacts and positioning of the process valve to its safe position (closed).
Our analysis indicates that this failure mode could possibly occur during testing or operation of these slave relays and could potentially remain undetected during normal manual operation of the valves.
We have cotcluded that this condition could impact plant safety and is reportable under the provisions of 10 CFR 21 and 10 CPR 50.55(e).
BROADNESS REVIEW:
In order to assure that all other safety-related auxiliary relays, (Agastat, General Electric, Struthers-Dunn, Allen-Bradley, etc.) have the appropriate contact rating to handle de-onergization of required inductive loads, a broadness review was conducted.
This review identified two other situations where contacts have been utilized to interrupt inductive currents in excess of their rated capability.
These situations aret a.
Diesel Generator Building (DGB) IIVAC Systems The contact identified is a Potter and Brumfield R10 plug-in type relay.
This relay is actuated by a non-safety-related exhaust fan PE0187012/SL l
I l
to open the DGB wall dampers whenever the non-safety-related I
exhaust fan operates.
The DGB wall damper fail safe position is in the open position to assure ventilation flow when the diesel generator is running.
When the diesel generator starts, safety-related supply fans will operate, initiating a signal independent of the non-safety-related signal to open the DGB wall dampers.
This signal from safety-related equipment opens different contacts which assure the opening of the DGB wall dampers, even if the Potter and Drumfield R10 relay contact fails in the open or closed position.
Based on this, the condition will not impact the safety of the plant and is therefore not reportable.
b.
Main Steam Isolation Valve (MSIV) Soloniods The interrupting current requirements for the MSIV control circuit is significantly less than the Valcor solenoid control circuit (approximately 0.60 amperos versus 1.50 amperes).
No failures have been experienced during MSIV ESPAS testing, but the possible failure modes for the MSIV actuation contacts and their associated impact on plant safety were analyzed as follows:
MSIV Dump Solenoid Control Circuit There are two separate sets of contacts either of which can function to break the de inductive load of the MSIV fast close dump solenoid and causo the MSIV to closo.
Ono contact is from the SSPS which actuates on main steam isolation actuation, and the other contact actuates on manual MSIV fast close initiation.
The same MDR relay model is utilized for both functions.
Tests have boon conducted which demonstrato the MDR relay contact interruption capability for this application.
Westinghouse's engineering ovaluation of these test results has concluded that the use of a single SSPS MDR contact in the MSIV control circuit will moor the requirements of the intended service for a minimum of one fuel cycle (approximately two years).
MSIV Exorciso Solenoid Circuit This circuit is similar to the dump solenoid circuit in that there are two separate sets of contacts olther of which can function to break the de inductivo load of the MSIV exorciso solenoid, one contact functions to remove tho (slow closo) rostriction orifico from tho (fast closo) dump lino on completion of the MSIV exorcise close tout (valvo roopon).
The other contact functions to abort l
an exorciso tost (romove the rostriction orifico from the fast closo dump lino) upon occurronce of an SSPS or manually initiated MSIV fast closure.
l j
PE0187012/SL !
4 The possibility exists that the normal contacts..used to remove the J~
restriction orifice after the test is complete could become locked closed similar to the' K748 failure.
However, the'MDR type test abort contacts and the exercise test complete contacta are wired in series.
Therefore, when an SSPS or manually initiated MSIV fast closure signal is received, the test abort contacts will open and the MSIVs will fast close as designed.
This condition does not impact plant safety since an emergency or manual MSIV close signal will cause the restriction orifices to be removed from the dump lines.
The contacts used to abort the tests are on MDR type relays and are the same type relays as analyzed above for the fast dump lines.
Westinghouse's engineering evaluation of test results has concluded that the MDR type test abort contacts will meet the l
requirements of their intended service for at least the first fuel cycle (approximately two years).
EVALUATION OF QUALITY ASSURANCE PROGRAM BREAKDOWN The root cause was l
determined to be insufficient attention to the inductive current interruptive rating of the contacts.
Specifically, insufficient attention was paid to the magnitude of the current interruption requirement for "in-line" process solenoid valvo coil control circuits.
Since the large majority of the solenoids in the plant are used for "non-in-line" pilot (instrument air) switching service, their coil current interruption requirements are typically well within the l
SSPS MDR relay contact's current interruption rating.
The broadness I
review based on the root cause revealed only two additional cases where corrective actions are required.
These were the DGB HVAC dampers and the MSIV exercise test complete solenoid.
The condition is therefore considered an isolated case and does not constitute a significant breakdown in Bechtel's Quality Assurance Program.
CONCLUSION:
Based on the results of the above analysis, the plant safety could potentially have been adversely impacted had the condition gone uncorrected.
Therefore, Georgia Power Company;has concluded that a reportable condition as defined by the critetia of 10 CPR 50.55(e) and 10 CFR 21 does exist.
Based on the guidancJ in i
NUREG-0302, Revision 1, concerning duplicate reporting of an event, Georgia Power Company is reporting this event per the criteria of 10 CFR 50.55(e).
CORRECTIVE ACTION:
The corrective action involves using two MDR relay contacts in series and the addition of a surge suppression rectifier connected across each Valcor solenoid coil.
This confiouration was l
bench tested (200 cycles each for 5 individual NOR type relays at 141 volts and 1.8 amps) to confirm that the relay contacts would interrupt the inductive circuit and reset (recione) as required.
l The implementation of this design modification for the reportable-i condition in Unit 1 was completed on December 8, 1986 through the following Field Change Requests (FCRs):
l E-FCRB-1534N, 1536N, 1550N 1552N through 1574N l
1580N through 1582N l
1584N through 1601N
[
7 The same circuit configuration will be utilized in Unit 2 for the
. similar Unit 2 solenoid valve circuits, unless the evaluation of long term i:apabilities result in a different des 1gn resoiution. Thd Unit 2 corrective i
action will be tracked through PCW action items #2B1609-0001 and 0002.
s These actions have been analyzed and tested to be adequate for at least the first fuel cycle (approximately two years) and meet regulatory requirements.
An ongoing engineering evaluation of the long term capabilities of these actions is proceeding.
The results of the long term assessment are being tracked on a Westinghouse Interface Action List by WIT Log Number 338-01 and will be provided to the NRC no later than November 1,
1987.
At this time it is not known that any changes will be required.
However, any corrective action identified as being required to be implemented prior to the end of the second cycle of operation will be completed prior to startup for the second cycle of operation.
Similarly, any corrective actions required for Unit 2 will be implemented prior to Unit 2 fuel load. Other required corrective actions will have specified dates by which they must be implemented.
PE0187012/SL