ML20045B336
| ML20045B336 | |
| Person / Time | |
|---|---|
| Site: | Trojan File:Portland General Electric icon.png |
| Issue date: | 06/10/1993 |
| From: | Masnik M Office of Nuclear Reactor Regulation |
| To: | Walt T PORTLAND GENERAL ELECTRIC CO. |
| References | |
| RTR-NUREG-CR-4674 NUDOCS 9306170205 | |
| Download: ML20045B336 (14) | |
Text
June 10, 1993 Docket No. 50-344 Mr. Tom Walt General Manager, Technical Functions Trojan Nuclear Plant 71760 Columbia River Highway Rainier, Oregon 97048
SUBJECT:
REVIEW 0F DRAFT 1992 ACCIDENT SEQUENCE PRECURSOR REPORT Enclosed is a copy of an accident sequence precursor (ASP) analysis of Trojan Nuclear Plant Licensee Event Report (LER), number 344/92-020 performed by the Oak Ridge National Laboratory (ORNL) under contract to the NRC Office for Analysis and Evaluation of Operational Data (AE00). The ASP analysis will be part of NUREG/CR-4674, Vol.17 & 18, " Precursors to Potential Severe Core Damage Accidents:
1992, A Status Report," to be published in September 1993.
t We are seeking written comments you may have on this preliminary analysis.
In particular, your response, which is strictly voluntary, should address the following three areas: the report characterization of the possible plant responses to the event, representation of plant safety equipment configuration, and analysis assumptions regarding equipment recovery. Any responses received by June 30, 1993, will be considered by the AEOD and ORNL r
in the final draft of NUREG/CR-4674. We appreciate any insight you may choose to provide. This request is covered by OMB clearance No. 315-0104 which expires May 31, 1995.
If you have any questions, please contact me at (301) t 504-1191.
Sincerely, Original signed by:
Michael Masnik, Senior Project Manager Non-Power Reactors and Decommissioning Project Directorate Division of Operating Reactor Support Office of Nuclear Reactor Regulation
Enclosure:
As stated cc w/ enclosure:
See next page DISTRIBUTION:[ASPLETER.MKW] (Webb-lb disk)
Docket Files' PDRs ONDD R/F BGrimes SWeiss MMasnik EHylton
$*bb 15' WD M
ACRS (10)
KPerkins, RV GHolahan RDudic ppy ONDD:EA ONDD:PM ONDD:D f /[4cn EHy MMasnik:lb SWeiss p/93 h / ///93
( //o/93 n
1 k
9306170205 930610 t
DR ADOCK 0500 4
I i
w
i..
s QE MCy n
n) f,[
B UNITED STATES 0
g* Q ^f NUCLEAR REGULATORY COMMISSION g
WASHINGTON. D.C. 2055. MOO 1 June 10, 1993 Docket No. 50-344 Mr. Tom Walt General Manager, Technical Functions Trojan Nuclear Plant 71760 Columbia River Highway Rainier, Oregon 97048
SUBJECT:
REVIEW 0F DRAFT 1992 ACCIDENT SEQUENCE PRECURSOR REPORT Enclosed is a copy of an accident sequence precursor (ASP) analysis of Trojan Nuclear Plant Licensee Event Report (LER), number 344/92-020 performed by the Oak Ridge National Laboratory (ORNL) under contract to the NRC Office for Analysis and Evaluation of Operational Data (AE00). The ASP analysis will be part of NUREG/CR-4674, Vol.17 & 18, " Precursors to Potential Severe Core Damage Accidents:
1992, A Status Report," to be published in September 1993.
We are seeking written comments you may have on this preliminary analysis.
In particular, your response, which is strictly voluntary, should address the following three areas:
the report characterization of the possible plant responses to the event, representation of plant safety equipment configuration, and analysis assumptions regarding equipment recovery. Any responses received by June 30, 1993, will be considered by the AE0D and ORNL in the final draft of NUREG/CR-4674. We appreciate any insight you may choose to provide. This request is covered by OMB clearance No. 315-0104 which expires May 31, 1995.
If you have any questions, please contact me at (301) 504-1191.
Sincerely, hh
~)fng I a
Michael Masnik, Senior Project Manager Non-Power Reactors and Decommissioning Project Directorate Division of Operating Reactor Support Office of Nuclear Reactor Regulation
Enclosure:
As stated cc w/ enclosure:
See next page
1 l'
Trojan Nuclear Plant Mr. James E. Cross Portland General Electric Company Docket No. 50-344 i
1 cc:
Senior Resident Inspector U.S. Nuclear Regulatory Commission Trojan Nuclear Plant P. O. Box 250 Rainier, Oregon 97048 Mr. Michael J. Sykes, Chairman Board of C*unty Commissioners i
Columbia County St. Helens, Oregon 97501 Mr. David Stewart-Smith Oregon Department of Energy Salem, Oregon 97310 Regional Administrator, Region V U.S. Nuclear Regulatory Commission 1450 Maria Lane, Suite 210 Walnut Creek, California 94596 Mr. Tom Walt General Manager, Technical Functions Trojan Nuclear Plant
[
71760 Columbia River Highway l
Rainier, Oregon 97048 t
Mr. Lloyd K. Marbet 19142 S.E. Bakers Ferry Road Boring, Oregon 97009 Mr. Jerry Wilson Do It Yourself Committee 570 N.E. 53rd Hillsboro, Oregon 97124 Mr. Eugene Rosolie Northwest Environmental Advocates 302 Haseltine Building 133 S.W. 2nd Avenue Portland, Oregon 97204
{
h t
f t
t PRELBIINARY O" 'i B.20 LER Number 344/92-020 i
Event
Description:
Reactor Trip and AFW Pump Failure To Stan Date of Event:
July 22,1992 Plant:
Trojan B.20.1 Summary Erratic controller performance on one rnain feed pump and contro!!er failure on the other resulted in a reactor trip on low-low steam generator (SG) level. He contro!!er for the auxiliary feedwater (AFW) pump turbine also failed, rendering one of two safery-grade AFW pumps inoperable. De conditional core damage probability estimated for this event is 1.7 x 108 The relative significance of this event compared to other postulated events at Trojan is shown in Fig. B.40 ER 34492 020(LOFW + 1AFW) i e
i 1E-7 1E4 1E 5 1E-4 1F-3 1E-2
- s.,,
I Iv I
I I
n
- TRIP t'
- 360 h EP l
praxtraor cutoE ---!
h -
- LOOP Fig. B.40.
Relative event significance of LER 344/92-020 compared with other Trojan potential events B.20.2 Event Description Approximately two weeks prior to the event, operators noted that the automatic controller for the *A*
main feed pump was oscillating and placed the controller in manual. About two days prior to the event, operators observed that the 'B' feed pump was supplying 20,500 gpm while the "A* feed pump was supplying 10,000 gpm. While anempting to balance flows between the pumps, they experienced difficulty with the 'B' pemp controller and placed that controller in manual as well. On the day of the event, during troubleshooting of the 'B' main feed pump, the pump suddenly slowed to minimum speed.
LER NO: 344/92420
..V, B-345 PRELBIINARY
PRELIMINARY
^
Operators tripped the pump, initiating a turbine runback, but the reactor tripped a short time later on low-low SG level. He turbine-driven *A* AFW pump auto-started but tripped on overspeed.
Subsequent attempts to restart the pump were unsuccessful. De diesel-driven "B" AFW pump started correctly and provided cooling water to the SGs.
De cause of the "A" main feed pump controller failure was diagnosed to be a defective electronic component in the controller module. He *B* feed pump controller failed because of a misadjusted power supply. He "A" auxiliary feed pump failed because a defective ramp generator signal converter permitted the pump to overspeed and trip on each start attempt.
B.20.3 Additional Event-Related Information Trojan is equipped with two 100% capacity safety-related AFW pumps, each capable of supplying 880 gpm to any of the four SGs. One pump is powered by a steam turbine, and the other is powered by a diesel engine. A third, non-engineered safety feature (ESF) electric-motor-driven pump is available for use during plant stanups and shutdowns. His pump is operable from the control room and could have been used to provide flow to the SGs if both safety-related AFW pumps had failed.
B.20.4 Modeling Assumptions This event was modeled as a reactor trip with loss of feedwater and one AFW pump unavailable. A non-recovery probability of 0.34 was assumed for the main feedwater system, since the *A* feed pump was locally operable. De non-safety related AFW pump was assumed capable of providing SG cooling Q
following a manual start. It was further assumed that no procedures were available for this action.
v B.20.5 Analysis Results ne conditional probability of subsequent core damage estimated for this event is 1.7 x 10$. He dominant core damage sequences, highlighted on the following event tree in Fig. B.41, involve failure of the feedv.ater system, failure of AFW, and failure of feed-and-bleed cooling.
LER NO: 344/92 4 20 B-346 PRELIMINARY i
i
PRELBilNARY
^
FORV/
PORV/
IRV SRV HP1 HFR EBQ END MAL RB5 BAT i
NO STATB l
N i
M i
11 CD 12 m
M i
g y
13 m
f "
14 '
m nummuu K
f 15 m 0) u m
y 17 m
18 ATWS (1)OK for Osas D I
i Eg. B.41.
Dominant core damage sequences for LER 344/92-020, LER NO: 344/92-020 L
B447 PRELBIINARY 1
m
- e
i PRELIMINARY CONDITIONAL Catt DAMAGE PRutASILITY CALCULATIONS Event Identifier: 344/92-020 Event
Description:
Feedptap failures, reactor trip, and AN gmap failure Event Dates
- 07/22/92 Plant Trojan t
IMITIATIWG EVENT NON-RECOVERABLE INITIATING EVENT PtosABILITIES trams
[
1.0E @
$EQUENE CONDITIONAL PROSABILITY Stats End State / Initiator Probability CD tRANS 1.7E-05 Total 1.7E-05 ATVS trams 3.'E-05 Totet 3.4E 05 SEQUEW2 CONDITIONAL PRotA51LITIES (ProsABILITY OtDER)
I O
secpence End State Prob 5 Rec *'
17 trans -rt AFV NFW hpl(f/b)
CD 8.2E-06 9.7E-02 15 trens rt AF9 NFV -hpt(f/b) -hpr/-hpl pory.open CD 7.?E 06 1.2E-01 16 trens -rt AFV MFW hpf(f/b) hpr/-hpl CD 8.6E-07 1.2E 01 18 trans rt ATW5 3.4E-05 1.2E-01 1
- non recovery credit for edited case
.{
l SEQUENCE CONDIT10NAL PRORASILITIES (SEQUEk2 ORDER) 1 se wence End state Prob N see"
{
15 trem -rt AFw MFw -hpttf/b> -hpr/-hpi pory.open CD 7.7E-06 1.2E-01 16 trans -rt AFV MFW hpI(f/b) hpr/ hpl CD 4.6E-07 1.2E 01 17 trens rt AFW MFW hpitf/h)
CD S.2E-06 9.7E-02 18 trans rt ATWS 3.4E-05 1.2E-01
" non-recovery credit for edited cose Event Identiffer: 344/92-020 i
LER NO: 344/92-020 B-348 PRELIMINARY
PRELBilNARY
{
~
EEQUDICE M L:
c \\ esp \\1989\\pwrtseet. cap BRANCH MODEL:
c:\\ esp \\19895tro}en.sti Mt06 ABILITY FILE:
c:\\eep\\1989\\;mer bst1. pro i
i 50 Recovery Limit BRANCH FREQUENCIES /Pt08 ABILITIES l
Branch system non-Recov ope Fall trens 5.6E-06 1.0E+00 toop 1.K 05 3.6E-01 Loca 2.4E 06 4.3E-01 et 2.BE 06 1.2E 01 rt/1oco 0.M+00 1.0E +00 amers. power 2.9E-03 8.0E 01 AFW 3.3E 06 > 6.5E 03 3.4E-01 Branch modet:
1.0F.*
Train 1 cond Prob:
5.0E-02 > Failed j
Train 2 cond Prob:
5.0E-02
.l Train 3 Card Prob:
1.3E 01 ofu/emerg. power 2.5E-03 3.4E-01 MFW 1.0E+DO > 1.0E+00 T.DE-02 > 3.4E-01 l
Branch Model: 1.DF.1 t
Train 1 Cond Prob:
1.0E*00 > Faltell porv.or.erv.chett 4.0E-02 1.DE+00 porv.or.srv.reseet 2.M 02 1.1E-02 pory.or.srv. reseat /energ. power 2.0E-02 1.0E+00 seet.1oce 2.3E-01 1.0E+00 ep.rectst) 5.9E 01 1.0E+00 ep. rec 6.1E-02 1.0E+00 hpl 1.0E 03 S.4E 01
^
hpI(f/b) 1.0E-03 8.4E-01 1.0E-02 hpr/-hpf 1.5E 06 1.0E+00 1.0E-03 pory.open 1.0E-02 1.0E+00
'4.0E-06
- branch model file
- forced Event Identifier: 344/92-020 t
I i
i LER NO: 344/92-020 B-349 PRELBIINARY
l l
t PRELBfINARY i
LICENSEE EVENT REPORT (LER) l FACILITY NAMEi] e - Nuclear Plant DOCKET NO: 344 i
i TITLE: Reactor Ty,. ased by the Failure of the Controller on Main Feedwater Pump *B' Due to Electric Cenent Failures EVENT DATE: 07/22/92 LER #: 92-020-00 REPORT DATE: 08/21/92 OTHER FACILITIES INVOLVED: N/A DOCKET NO: 05000 l
t OPERATING MODE: 1 POWER LEVEL: 100 l
THIS REPORT IS SUBMITTED PURSUANT TO THE REQUIREMENTS OF 10 CFR SECTION:
50.73(a)(2)(iv)
LICENSEE CONTACT FOR THIS LER:
D. L. Claridge, Compliance Engineer TELEPHONE: (503) 556-5541 COMPONENT FAILURE DESCRIITION:
CAUSE: X SYSTEM:
JB COMPONENT:
TC MANUFACTURER:
W120 X
BA 65 W
D JB 65 m
REPORTABLE NPRDS: YES YES Q
YES
.{
J SUPPLEMENTAL REPORT EXPECTED: YES EXPECTED SUBMISSION DATE: 10/16/92 ABSTRACT:
On July 22,1992, the Trojan Nuclear Plant experienced a reactor trip from 100 percent power. De trip was caused by the loss of flow from the 'B' Main Feedwater Pump (MFP). At the time of the trip, both MFPs were in manual control due to oscillations in automatic flow control. Following the reactor trip, the *A* Auxiliary Feedwater Pump (AFW Pump) started, but tripped on overspeed shortly after starting.
Subsequent attempts to start the 'A' AFW Pump failed. He failure of the 'B' MFP was due to an electronic component failure on its goverr 'r's electronic control assembly (Woodward Model 8270). De control problems with the 'A' MFP were c wsed by a failed electronic componcat on its flow contro!!er (Westinghouse-Hagan Model 12 ). The 'A' AFW Pump failure was caused by a failed integrated circuit on the Ramp Generator Signal Converter in the Woodward Governor turbine startup control circuitry.
Corrective Actions included repairing and/or replacing the failed components and circuit boards. There l
were no safety consequences resulting from the component failures. De 'B' AFW Pump functioned as required to provide cooling water to the steam generators. Other plant systems functioned as expected, j
i LER NO: 344/92 4 20 l
~
B-350 PRELB11 NARY.
i I
l
k i
PRELBHNARY DESCRIPTION OF EVENT On July 22,1992, the Trojan Nuclear Plant was in Operational Mode 1 (Power Operation) at 100 percent l
power. At 0126 hours0.00146 days <br />0.035 hours <br />2.083333e-4 weeks <br />4.7943e-5 months <br /> a reactor trip occurred on low-low steam generator level due to a loss of l
feedwater flow from the 'B' Main Feedwater Pump (MFP) lSJ, P]. Both 'A' and 'B' MFPs were in manual at the time of the reactor trip. Manual control had been established on the 'A' MFP on July 6, 1992 due to oscillations in the controller while in aatomatic flow control. On July 20,1992, the 'B' MFP was placed in Manual when operating personnel were unable to maintain stable feedwater flows while in automatic. On July 22,1992, while in manual and with no operator action, the 'B' MFP slowed to its minimum speed and flow. Manual attempts to increase pump speed were unsuccessful, and the Control Operator tripped the 'B' MFP, which initiated a turbine runback. Approximately one minute j
later, the reactor tripped on low-low steam generator level.
l Immediately after the reactor trip, the Turbine 4 riven Auxiliary Ferdwater Pump (the 'A' AFW Pump)
[BA, P) Auto-started. A few seconds later, the 'A' AFW Pump tripped on overspeed. Two subsequent j
attempts to restart the 'A' AFW Pump also resulted in overspeed trips. The Diesel-Driven Auxiliary Feedwater Pump (the 'B' AFW Pump) [BA, PJ operated properly to supply cooling water to the steam generators during the transient.
De reactor trip was a Reactor Protection System (RPS) [JE] actuation, and the initiation of Auxiliary Feedwater was an Engineered Safety Fea';ures (ESF) [JE] actuation. Both events were reported under 4
10 CFR 50.72(b)(2)(ii) on July 22,1992 at 0330 hours0.00382 days <br />0.0917 hours <br />5.456349e-4 weeks <br />1.25565e-4 months <br />, using the Emergency Notification System. His repon is being submitted to fulfill the requirements of 10 CFR 50.73(a)(2)(iv).
,,,g Sequence of Events 1
July 6,1992 Control Operator (CO) received ' Lube Oil Pressure 2350 hours0.0272 days <br />0.653 hours <br />0.00389 weeks <br />8.94175e-4 months <br /> IB alarms on the 'A' MFP and noticed the pump controller oscillating. CO p! aced controller in Manual and oscillations stopped. Troubleshooting efforts were initiated.
July 20,1992 CO received the 'B' MFP ' Shaft Coupling vibration i
1613 hours0.0187 days <br />0.448 hours <br />0.00267 weeks <br />6.137465e-4 months <br /> High' alarm. CO noted the 'A' MFP was running at approximately 10,000 gpm and the 'B' MFP was running at approximately 20,500 gpm. CO manua!!y increased the
'A' MFP flow in order to decrease the 'B' MFP flow. His cleared the high vibration alarm.
1634 hours0.0189 days <br />0.454 hours <br />0.0027 weeks <br />6.21737e-4 months <br /> CO was unable to maintain stable feedwater flows. CO placed the 'B' MFP in Manual.
i 1759 hours0.0204 days <br />0.489 hours <br />0.00291 weeks <br />6.692995e-4 months <br /> Troubleshooting determined that the 'B' MFP picked up excessive feedwater Gow while l
in Auto.
2100 hours0.0243 days <br />0.583 hours <br />0.00347 weeks <br />7.9905e-4 months <br /> While in Manual, the 'B' MFP flow had to be increased slowly over approxirrately a two-hour period to prevent flow from shifting over to the 'A' MFP. Stable Main l
a Feedwater conditions were established. De 'B' MFP was inspected, but no problems were noted.
1 LER NO: 344/92-020 B-351 PRELBHNARY 4
I
~
PRELIMINARY t
July 21,1992 Instrumentation and Control (l&C) personnel 1608 hours0.0186 days <br />0.447 hours <br />0.00266 weeks <br />6.11844e-4 months <br /> completed installation of monitoring equipment on MFP instrument control signals in the Westinghouse Hagan Controller racks.
1636 hours0.0189 days <br />0.454 hours <br />0.00271 weeks <br />6.22498e-4 months <br /> Started increasing feodwater flow on the 'B' MFP.
1721 hours0.0199 days <br />0.478 hours <br />0.00285 weeks <br />6.548405e-4 months <br /> Established a 195 psi differential pressure between the MFPs to allow monitoring of instrument speed control signals.
July 22,1992 he 'B' MFP slowed to minimum speed. Attempt was 0125 hours0.00145 days <br />0.0347 hours <br />2.066799e-4 weeks <br />4.75625e-5 months <br /> made to increase the 'B' MFP speed manually, but was unsuccessful, ne CO tripped the 'B' MFP, which initiated a turbine runback.
0126 hours0.00146 days <br />0.035 hours <br />2.083333e-4 weeks <br />4.7943e-5 months <br /> Reactor tripped on low-low steam generator water level. The 'A' AFW Pump Auto-started, then tripped on overspeed. De 'B' AFW Pump Auto-started. He 'A' AFW Pump tripped on overspeed on subsequent attempts to st:.a it.
Event Analysis f
The 'A' MFP Failure i
ne controller module and Manual / Auto Station (Westinghouse-Hagan Controller Model 124) UB, TC]
for the 'A' MFP were removed for bench testing. During the first two hours of the test period the l
controller was placed in automctic and a small step change signal was introduced. He module outputs i
remained satisfactory during this period. When a second step change was input to the controller, the n
output integrated off-scale high. He controller output would remain stable when it was placed in Manual. For the next several hours attempts to place the controller in Auto caused the output signal to integrate off-scale high.
The 'B' MFP Failure Troubleshooting was performed on the 'B' MFP governor controller (Woodward Governor Model 2301, Electronic Control Assembly Model 8270) UB, SC]. Several components on the signal convener DB, CNV], amplifier pB, AMP], and voltage switch DB, JS] modules of the Electronic Control Assembly were found to be overheating. Measurements taken during troubleshooting identified several failed componems. The component failures were determined to be associated with the power supplied to each module.
During troubleshooting, the Zener Diodes on the amplifier module were found to be hot to the touch.
Measurement of the output of the auctioneered power supplies DB, RJX] determined that the lead (primary) power supply was set at 24 VDC arxi the follower (backup) power supply at 19 VDC. For an auctioneered power supply, the lead power supply is supposed to be set only slightly higher than the follower to forward-bias the diode circuit. He lead power supply was adjusted to 19.3 VDC and approximately 17.7 VDC was measured at the Control Assembly. After this reduction, the Zener diodes were noticeably cooler.
LER NO: 344/92-020 B-352 PRELIMINARY
PRELIMINARY O
ne 'A' AFW Pump Failure The cause of the overspeed trip of the 'A' AFW Pump was originally diagnosed as a failure of the pump's Electric Overspeed Trip circuitry [BA, SC). However, subsequent testing (manually staning and controlling the pump locality at various speeds) verified that the electric overspeed trip circuitry was i
functioning normally. Fast starting the pump, however, produced an electric overspeed trip signal.
Additional testing was then conducted on the Woodward Governor control circuitry [BA,65]. A start up circuit (Ramp Generator Signal Converter) in the Woodward Governor programs turbine speed to prevent overspeed on turbine startup. It was determined that the Ramp Generator Signal Converter was not generating a ramp or idle speed signal. Without these signals present to control purnp startup, the turbine would quickly run up to full speed, overshooting to its overspeed trip setting. The failed Ramp Generator signn) Converter was returned to the vendor for repair and failure analysis.
CAUSE OF OCCURRENCE ne 'A' MFP Westinghouse-Hagan Controller failure was caused by a component failure. The erratic operation of the controller was due to a failed electrical component within the controller module. De particular failed compocent (or components) has not yet been identified, and is still under investigation.
De 'B' MFP governor electronic control assembly failure was caused by a misadjusted controller power supply. The 24 VDC power supply was set at 24 VDC when it should have been sat slightly higher than 19 VDC. Investigation to detmnine why the supply was set too high identified a Maintenance Request initiated in 1991 to replace a burned out indicating lamp in the lead power supply to the 'B' MFP m
Governar cor. troller circuit. While performing the replacement the maintenance personnel discovered that the lead power supply was not functioning properly. He power supply (Lambda Electronics Model LIS-10A-24-OV)[BA, RJX] was replaced on April 27,1991. However, the work instructions directed the personnel to verify that the power supply provided 24 VDC, and did not specify that the supply was to be Mjusted to slightly higher than 19 VDC (Output Voltage). The higher voltage and current conditions that the controller circuits were subjected to caused several of the electrical components to overheat and eventually fail. Further investigation as to the reason why the new power supply was not adjusted to the proper voltage is ongoing. He results of that investigation will be reported in a supplement to this report.
De 'A' AFW Pump failure was caused by a faulty Ramp Generator Signal Converter in the Woodward Gosernor control circuitry. This converter had failed on June 1,1992, and had been replaced on June 3,1992, with a new converter. He failure of the new converter on July 22,1992, was detennined to be the result of a failed integrated circuit on the printed circuit board. According to the verdor, the failure did not appear to be the result of an externa! cause, but was probably due to a premature failure of the integrated circuit due to am internal flaw on the chip.
LER NO: 344/92-020 B-353 PRELIMINARY
PRELBilNARY CORRECTIVE ACTIONS Actions Completed i
1.
He failed components on the 'B' MFP governor electronic control assembly were replaced, the power supply was adjusted, the governor was tested satisfactorily, and the 'B' MFP was returned.
c toservice.
C 2.
De Westinghouse-Hagan Controller Module for the 'A' MFP was replaced, the control circuitry was tested satisfactorily, and the 'A' MFP was returned to service.
I t
3.
De failed 'A' AFW Pump Ramp Generator Signal Convener was replaced and tested successfully.
Power supply voltages were checked when the circuit was replaced and found to be correct.
l i
Actions in Progress d
1.
Increased frequency testing of the Turbine Driven Auxiliary Feedwater Pump has been implemented.
4 De increased testing frequency will continue for six months as follows: once per week for the first i
month, twice per month for the next two months, and once per month for the last three months.
His testing is intended to monitor for premature component failure by verifying proper operation j
i of the Ramp Generator signal Convener, and will be done under fast start conditions.
t Actions to be Taken 1.
Investigation to determine the root cause of the inadequate work instructions for replacing the failed
~
{
power supply on the 'B' MFP Governor controller circuit will, be completed and the results reponed in a supplement to this report by October 16,1992. Corrective actions deemed necessary j
as a result of this investigation will also be reponed in the supplement.
i i
ANALYSIS OF SAFETY CONSEQUENCES AND IMPLICATIONS t
i Dere were no safety consequences resulting from the failure of the 'B' MFP or the 'A' AFW Pump.
_i Following the reactor trip, safety systems functioned as required, with the exception of the 'A' AFW Pump. However, the *B' AFW Pump staned and ran normally, and supplied cooling water to the steam generators, as required. Also, it would have been possible to manually stan and control locally the 'A' AFW Pump, had it been needed. In addition, the efectric-driven auxiliary feedwater pump [BA, PJ could have been staned, if required. Derefore, the subject failures did not prevent the steam p serators (SB, SG) from receiving an adequate supply of cooling water.
I PREVIOUS SIMILAR EVENTS i
1 A reactor trip occurred on June 5,1992, due to a steam generator high-high water level signal caused 4
by a failed controller for the 'B' Main Feedwater Regulating Valve ISJ, FCV). His event was reponed on LER 92-14, dated July 6,1992. De failure in that event was a manual pushbutton, in which the LER NO: 344/92-020 i
i B-354 PRELIMINARY t
i a
+
-w-c r
PRELBiINARY r
D switch contacts would intermittently stick closed, which caused the ' increase flow
- signal to stay in until the steam generator water level reached its high-high setpoint. His was a mechanical failure, not an electronic circuit failure, and was due to a combination of a manufacturing defect and age-relaxation of the switch contacts
- spring.
Another reactor trip event, as reported in LER 88 43, dated December 13, 1988, also involved the controller for the 'B' Main Feedwater Regulating Valve. hat event was not due to a solid state circuit failure, but a failed capacitor in the controller's power supply. He failed power supply caused the valve to fail open.
In addition to the above two events, the 'A' AFW Pump Ramp Generator Signal Converter that failed ton July 22,1992, was the replacement for the converter that had failed on June 1,1992, during testing i
of the 'A' AFW Pump. He June 1,1992, failure, which also causel the turbine to trip on startup, was caused by a component failure. The component that failed then was also an integrated circuit, but not i
l the same one that failed on July 22,1992.
l h
i J
e
?~
em I
i i
[
i i
LER NO: Mt/92-020 PRELBIINARY l
B-355
-