ML20117B184
| ML20117B184 | |
| Person / Time | |
|---|---|
| Site: | LaSalle  |
| Issue date: | 09/25/1992 |
| From: | Galle D COMMONWEALTH EDISON CO. |
| To: | Davis A NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION III) |
| Shared Package | |
| ML20117B177 | List: |
| References | |
| CAL-RIII-92-11, NUDOCS 9212010266 | |
| Download: ML20117B184 (10) | |
Text
[.
' . ' ~ I. _ \ C::mmonwtalth Edison n / '1400 Opus Place 3 Downers Grove, Illinois 60515 September 25, 1992 q Mr. A. Bert Davis Regional Administrator U.S. Nuclear Regulatory Commission ;
799 Roosevelt Road - RIII Glen Ellyn, IL 60137
Subject:
LaSalle County Station Units I and 2 Response to Confirmatory Action Letter CAL-RIII-92-011 NRCJ1odutLttos 50-373_and_50-374
Reference:
(a) Confirmatory Action Letter (CAL-RIII-92-011) from A.
Bert Davis (NRC) to Cordell Reed (CECO) dated August 27, 1992
Dear Mr. Davis:
Confirmatory Action Letter (CAL) CAL-RIII-92-Oll addressed the scram and equipment failures at LaSalle Unit 2 which occurred on August 27, 1992.
The CAL reauested CECO to perform the following: investigate and determine the cause(s) of the equipment failures; place specific equipment in quaranti ne until released by the NRC's Augmented Inspection Team (AIT);
maintain evidence of our investigations and provide this information to the AIT team; evaluate the equipment failures and opea tor actions to determine any necessary actions; and to evaluate the applicability of the equipment failures to LaSalle Unit 1. Attachment A provides Commonwealth Edison Company's response to the CAL.
If there are any questions regarding this response, please contact JoAnn Shields, Nuclear Licensing Administrator, at (708)-515-7282.
Very truly yours,
) . b o M o_. 9 (26 h2 D. Galle Vice President-BNR Operations Attachment cc: Document Control Desk - NRR L.L. Siegel, Project Manager - NRR D.L. Hills, Senior Resident Inspector - LSCS 9212010266 921124 FDft ADOCK 05000374 PDR; P g gg
- ZNLD211W23
ATTACHMENT A RFSPONSE TO CONFIRMATORY ACTION LETTER CAL-RIII-92-011 Dftscriotion of the EY_ tat Reactor power was being reduced from 1100 Mwe ';o 850 Mwe at 120 Mwe/ hour using Reactor Recirculation Flow Control. At the time of the scram, actual flow control manipulations were briefly suspended to allow for xenon burnout.
On August 27, 1992 at 0305 hours0.00353 days <br />0.0847 hours <br />5.042989e-4 weeks <br />1.160525e-4 months <br />. Unit 2 experienced a Reactor Scram as a result of a Main Turbine Stop Valve (TSV) closure trip. The turbine trip was caused by a Thrust Bearing Hear Detector Turbine Trip signal to the Electro Hydraulic Control (EHC) System. As a result of the automatic scram signal, all control rods inserted to their full in position.
During the first seconds of the event, the Reactor Core Isolation Cooling (RCIC) System auto started due to a spurious Level 2 (-50 inches) initiation signal.
During the scram response, in an attempt to control reactor water level, the Motor Driven Reactor Feed Pump (MDRFf') was successfully started in preparation for tripping of the Turbine Driven Reactor Feed pumps (TDRFP). When attempting to shutdown the TDRFPs, all methods of tripping them initially failed including remote manual trip operation, High Reactor Level 8 automatic trip, or local mechanical trip operatien.
As a result of this failure, the Reactor Water level increased above the Level 8 High Level setpoint (+55.5 inches) resulting in a trip of the MDRFP and the RCIC System. The Outboard Main Steam Isole ion Valves (MSIV) were manually closed when the +73 inch reactor level administrative limit was reached. This limit is provided to prevent flooding in the steam lines outboard of the MSIVs (bottom of the Main Steam lines is at 108 inches). The level transient '
resulted in a maximum level of +13C inches.
The closure of the MSIVs resulted in TDRFP shutdown and also caused a loss of the Main Condenser as a heat sink.
The loss of the Main Condenser as a heat sink required use of the Safety Relief Valves (SRV) for manual control of reactor pressure. During operation of 'A' and 'B' SRVs, remote position indication failed to show that the valves fully closed when demanded. Subsequent review showed that earlier in the event 'U' SRV had automatically cycled on reactor pressure as designed, with final position indicated as full closed. No "SRV Full Open" Alarm was seen by the operators during any SRV operation. Additional review of SRV tailpipe temperatures showed that the SRVs had closed.
After reactor water level was returned to the normal operating range and brought under control, an attempt was made to reestablish the Main Condenser as a heat sink. All the inboard MSIVs were closed and all outboard MSIVs were opened. Pressure was being equalized across the Inboard MSIVs. When the "A" Inboard MSIV was opened, a MSIV (Group 1) isolation High Steam flow signal was received resulting in closure of all five open MSIVs.
ZNLD2119/24
- . . = - - - .- - --. .- . . .. . -
ATTACHMENT A (Continued)
RESPONSE TO CONFIRMATORY ACTION LETTER CAL-RIII-92-011 Attempts to use the RCIC System to help control reactor pressure were made and the turbine tripped on high exhaust pressure on the first two start attempts.
The system was successfully started on the next attempt and operated normally to control pressure.
The MSIVs were subsequently successfully opened, and the plant stabilized in a hot thatdown condition. After stabilization, a decision was made to proceed to cold shutdown to complete investigations and repairs. During this event, several additional minor deviations from desired equipment performance were noted, and included in the investigations. These include:
1, 2E51-F066 RCIC Testable Check Valve Position Indication showed the valve to not be full closed-(RCIL Running Alarm).
- 2. Scram Annunciator "First Out" Indication did n't function.
- 3. High Drywell Temperature Alarm.
Renuelt:
- 1. Conduct an investigation to determine the cause of: (a) the failure of the Main feedwater Pumps (MFP) to trip. (b) the failure of the Safety Relief Valves (SRV) to reposition and/or failure of the SRV indicating circuitry, (c) the failure of the Reactor Core Isolation Cooling (RCIC) testable check valves or its indicating cirsuitry, (d) failure of the first out " Red" annunciator, (e) an unexpected trip of RCIC while in the pressure control mode, (f) the Group 1 isolation during attempts to equalize pressure-across the Main Steam Isolation Valves (MSIV), and (g) the turbine / reactor trip.
Response ta llal: Failure of the MFPs to trip.
Upon disassembly of the pumps, particulate matter was found in the oil and on the hydraulic disc dump valves of the 2A and 28 turbine driven reactor feed pump (TDRFP). Analysis of this oil by CECO's System Materials Analysis Department.(SMAD) indicated the presence of wood chips, silk, fiberglass, cellulose, and a clay-like material (aluminum and-sodium silicate), which ;
accumulated over the life of the plant.
Oil-to each TDRFP is. supplied by the main turbine lube. oil system. . The particulate matter in the disk dump valve assembly caused mechanical binding of the TDRFP disc dump valve, preventing the valve from repositioning, thereby-preventing the pump from tripping upon automatic, remote,:or local trip signals.
The oil system, including the valve ports,.was flushed and the reservoir filter was inspected. No particulate buildup was observed in the reservoir.
The.feedwater oil system will be flushed during every refuel outage to prevent the accumulation of particulate in low flow areas of the system.
ZNLD2119/25
j... ; o .s.u..e u. _ .% m-.2 - 4,+.+.s. a ATTACHMENT A (Continued)
PESPONSE TO CONFIRMATORY ACTION LETTER CAL-RIII-92-Oll The "A" feed pump dump valve had a 0.005 inch runout, which was discovered upon disassembly. The "B" SV-12 pilot valve experienced similar runout, 0.003.
inch. The dump valve and the pilot valve were replaced and satisfactorily tested. The "B" trip dump valve o-ring contributed to the restricted movement. The dump valves and pilot valves will be inspected during each Unit's next refuel outage. Further inspections will be performed as determined by LaSalle Management.
The trip circuit for each pump was tested during Unit 2 startup and within one week of startup. The circuitry will be tested two months after startup, and then will resume its normal six month test frequency.
Reiponse_lo_11bl: Failure of the SRV> to reposition and/or frilure of the SRV indicating circuitry.
Review of the event indicated that the SRVs operated properly but the low voltage differential transformer position indication (LVDT) for the A and B valves did not indicate properly. These two LVDTs were disassembled, and fretting was noted on the outer bush'.ng and along the lower portion of the LVDT shaft. Three more LVDTs were then inspected, with fretting noted on one of these. Based Upor these findings, all 18 Unit 2 LVDTs were inspected, with no further evidence of fretting observed. As a precautionary measure, all the ADS valves were given new LVDTs in addition to replacing those.with visible damage. Further evaluation revealed the fretting to be vibration induced.
All the LVDTs will be inspected and replaced as necessary during every refuel outage.
Failure of the annunciator, which indicates that the SRVs were not closed, was due to a failed connector on the annunciator logic card. The faulty card.was replaced, and proper operation of the alarm us verified. However, when ,
tested in the mair,tenance shop, the card operated properly. At that: time, some corrosion on the card input pins was noted. As we have not been.able to repeat the failure, no further root cause investigation'is possible, ,
Resaonic to 1(cl: Failure of the RCIC testable check valves or its indicating circuitry.
Investigation revealed that the inboard testable check valve 2E51-F066 did not fully close during the shutdown of the RCIC sjstem, but that the valve's position indication functioncd properly. As the RCIC system was shut down, the outboard check valve fully closed, eliminacing backflow'through the system, and equalizing pressure across the inboard testable check valve.
To reduce frictional. forces in the indicating hinge pin stuffing box on the outboard testable check valve, 2E51-F066, packing rings were removed and replaced with graphite spacers. The original. concern on the valve remaining open due to opposing frictional forces had been resolved, allowing the valve to close on its own accord from approximately 601 open.
1 l
ZNLD211M.3
.k: ' b ATTACHMENT A (Continued)
RESPONSE TO CONFIRMATORY ACTION LETTER CAL-RIII-92-011 Responie_to E dl; Failure of the first out " Red" annunciator.
Troubleshooting determined that both red "first out" bulbs on panel 2H13-P603, window B308, "CHAN A2/B2 TSV NOT FL OPEN ALARM", were blown. Upon replacement of the bulbs, proper operation was verified. Review of the Hathaway alarm typer indicated that this was the first RPS alarm to annunciate.
A test was performed on the "first out" windows, and no other "first out" windows on Unit 2 were found inoperable.
Investigation revealed tnat surveillance procedure LES-AN-101 had been generated in 1988 to test these windows, but had not been entered in the station surveillance tracking system. This surveillance is now included in the tracking system, and will be performed annually on each unit.
Response to 1(ell i cetted trip of RCIC while in the pressure control mode.
Investigation into Vieis event determined that the RCIC system operated properly. Due to the high vessel water level, water entered the RCIC supply piping. When RCIC was restarted, water went through the turbine, and flashed in the exhaust line, causing RCIC to trip on high exhaust pressure. The flashing occurred on the first two attempts to start the system. On the third start, RCIC remained operating.
LaSalle is evaluating enhancing our operator training program to increase awareness of this mode of-operation.
Resuonse to 1(f>: Group 1 Isolation during attempts to equalize pressure
, across the MSIVs.
During recovery ac' ions after the scram, a Group 1 Primary Containment' - .
Isolation was received when an attempt was made to open MSIV 2B21-F022A with a '
760 psi differential pressure existing between the main steam lines and the -
reactor-vessel. '
The Group 1 isolation signal was generated by differential pressure (dp) switches, which monitor the main steam lines and actuate on a high flow condition. These dP switches are calibrated to actuate on 140% rated steam flow. The foer sets of dP switches (one set of four for each line) have the capability to isolate all-four main steam-lines and initiate the rest of the Group 1 Isolation when high flow is seen in any individual line. Prior to the event, all four inboard MSIVs were closed, and were holding reactor pressure.
All four outboard MSIVs were open, and the pressure-in the steam lines was approximately 120 psi. When the attempt was made to open 2B21-F022A, the flow instruments actuated, causing the Group 1 isolation signal. -
ZNLD211W27
~
f, ' 'i ATTACHMENT A (Continued)
RESPONSE TO CONFIRMATORY ACTION LETTER
- CAL-RIII-92-011 Procedural guidance states that the MSIVs should not be opened unless the-differential-pressure is less than 200 pside- The control room operator read the incorrect indicator when verifying pressure. When the attempt was made to open the MSIVs, the pressure was 760 psid'. This pressure is enough to actuate the high flow instrumentation, and cause the primary containment isolation.
All Group 1 isolation valves received their isolation signals. All velves not-closed prior to the event closed. The Primary Containment Isolation system functioned properly in response to an actual high flow condition, LaSalle will revise the operator training program to include all aspects.of.
the scram and equipment failures, including opening of MSIVs.
B11aonie to 1(91: Turbine / reactor trip The reactor scram was due to closure of the turbine stop valves, which closed due to the turbine trip. The turbine thrust bearing wear detector system actuated to trip the turbine.
Local and remote wear detector operations and turbine rotor thrust checks were performed. These tests identified a change the span between turbine end and generator end trip points had changed from the previcusly recorded span of 110 mils (-40 to +70) to 84 mils (-80 to + 4). This shift in the span, and not-thrust bearing failure is the actual cause of the trip. This test also showed' that the wear detector was able to consistently follow thrust collar position accurately.
A loose set screw attaching the lower coupling half to the bushing stem caused a shift in the calibration of the setpoint for the trip. The bushing drive coupling and stem were drilled to act;mmodate a roll pin, and the set screw wae re-applied.
The Unit I thrust bearing wear detector will be inspected in the next refuel-outage, and appropriate corrective actions will be implemented. .
Renun11:
- 2. Place.the MFP trip circuitry and mechanical actuator, the SRV's and their circuitry operated during the event, and the RCIC testable check valve and its circuitry in quarantine until released by the NRC's Augmented Inspection Team (AIT).
211portst_t.o_2:
The above equipment was quarantined until the-individual troubleshooting-action plans were approved, at which time the NRC AIT team released the specific quarantine.
ZNLD2119/28
, ..- - . . - - - - . . - - - - , -. - ~ _ _ , - - - : , . .
, -. - -. ~.. - -. -
e :,' ' .;
'l ATTACHMENT A .
(Continued)
RESPONSE TO CONFIRMATORY ACTION LETTER CAL-RIII-92-011 -
Request:
- 3. Maintain documentary evidence nf your investigation ~ effort and make this available to the AIT.
Reinonie_to.J:
Preliminary information from the scram, including computer data, operator log books, alarm typer output, etc. was provided to the NRC AiT team upon their arrival on site. Requests from the AIT team for information during the course of their investigation were fulfilled, and as CECO's troubleshooting and evaluations were completed, written reports detailing the findings were provided to the AIT team.
Reguest:
- 4. Evaluate these most recent equipment failures and operator actions in light of past equipment failures and operator performance to determine if additional actions are necessary, 823DDalt 10_4:
A review of Licensee Event Reports and reports of major. events which occurred over'the past two years was performed to determine whether previous events had common causes with the failures experienced on August 27, 1992. The root causes of the equipnient failures and any inappropriate ope.ator actions resulting from this scram were significantly different from any previous everts, and therefore, no additional actions are necessary.
Etquait: :
r
- 5. Evaluate the applicability of the equipment failures associated with the August 27, 1992 ev,^nt to LaSalle, Unit 1. ,
Reiponse to 5: >
(a) Failure of the'MFPs to trip. >
Pr?ner operation of the Unit 1 TDRFP disc durp valves was verified on September 3, 1992. The dump valves will be inspected during the next-refuel outage. The lube oil system will.be flushed during every.
refuel outage to prevent accumulation of particulate in the low flow areas of the system.
.0) Failure of the SRVs to' reposition and/or failure of-the SRV indicating circuitry.
' ZNLD211W29
ATTACHMENT A
-(Continued)
RESPONSE-TO CONFIRMATORY ACTION LETTER ,
CAL-RIII-92-011 The Unit 1 SRVs do not use LVDTs for position indication, so no further action is necessary.
(c) Failure of the RCIC testable check valves or--its indicating cir:uitry.
The Unit 1 RCIC inboard and outboard testable check valves will be inspected during LIR05, and each valve's actuator hinge pin packing j will be replaced. The indicator hinge pin packing and the necessity l
of the extended backstop will be evaluated. If the evaluation L determines that the packing or the backstop is not necessary, these items will be closed.
l (d) Failure of the first out " Red" annunciator.
The Unit 1 "first out" annunciator windows were tested on-September j 3, 1992, and all were operbble. Twc windows were found to have one i of the two bulbs burned oct. These bulbs were replaced and tested l satisfactorily. A similar surveillance will be performed annually.
(e) Unexpected trip of RCIC while in the pressure control. mode.
l Investigation revealed that RCIC operated properly, therefore. there .
l are-no ctions necessary on Unit 1.
l l (f) Group 1 isolation during attempts to equalize pressure across the l
The Unit 2 1 solation was oue to operator error, therefore, there are no actions necessary on Unit 1.
(g) Turbine / reactor trio. l l
l The turbine titrust bearing wear detector assembly will be-irspected ;
j during LIR05, and the appropriate corrective actions will be taken, i
Re.quelt: ,
- 6. Provide within 30 days to NRC Region III a documented evaluation of the above issues including corrective actions you have.taken or plan to take.
L Re.sponse to 6:
l l
This letter constitutes'LaSalle's 30 day formal report regarding this event.
l ZNLD2119/30
ATTACHMENT A (Continued)
RESPONSE TO CONFIRMATORY ACTION LETTER CAL-RIII-92-Oll-B1Quelt:
W6 further understand that reactor startup (power operation) will not occur untti you have informed the Regional Administrator or his designee of the results of your investigation and corrective actions.
Renone:
On September 3, 1992, at approxiw 'ely 1000 hours0.0116 days <br />0.278 hours <br />0.00165 weeks <br />3.805e-4 months <br />, G.J. Diederich, LaSalle Station Ma,..ger discussed reactor .*artup with T.O. Martin, Region III-Deputy virector, Division of Reactor Safety. At that time, Mr. Martin stated that the restriction from reactor startup had been lifted, and that LaSalle coul return Unit 2 to power operation.
SUMMARY
l In order to minimize the potentiai for recurrence, LaSalle has taken or will tr.ke the following actions:
The Unit 2 TDRFP control oil system was flushed. Each Unit's system will be flushed during every refuel outage.
The dump valves were replaced on both the Unit 2 TDRFPs. Each Unit's dump valves will be inspected during each unit's-next refuel out69e. Future inspections will be performed as determined by LaSalle Management.
The Unit 2 TDRFP trip circuitry for each pump was satisfactorily tested during Unit 2 startup ano one week after startup. The circuitry will be tested 2 months after startup, and then revert to the-normal six month test frequency.
The Cleanliness Control Procedu.a, LAP-300-16, will be reviewed for f' enhancements.
All 18 Unit 2 LVDTs were inspected for fretting. The three LVDTs showing signs of fretting were replaced. All LVDTs on ADS valves were replaced.
The internals on the Unit 2 LVDTs will be inspected during each refuel outage.
The SRV logi. cult on Unit 2 was tested and a card containing a suspect
. contact was ceo, .<d.
The actuator hinge pin packing on the Unit 2'RCIC outboard testable check-valve was replaced. The actuator hinge pin packing on .the Unit 1 RCIC =
outboard testable check will be replaced'during LlR05. The indicator hinge pin packing for both Unit 1 valves will be evaluated during L1R05. L
-ZNLD211F31
4 c 1.,
1 ATTACHMENT A l (Continued) i RESPONSE TO CONFIRMATORY ACTION LLTTER j CAL-RIII-92-011 l An extended backstop will be added to the Unit 2 RCIC valyc during L2R05, !
and the backstop will be evaluated for the Unit 1 RCIC vahes, ,
The "first out" red annunciator systems for both Unit 1 and 2 were tested. Only the Unit 2 window involved in the scram was found ,
inoparable, e surveillance procedure to test these windows annually was entered into the station surveillt.nce program.
A review is in progress to ensure all appropriate surveillances have been entered into the tracking system. The processing of new surveillance requests to ensure tracking system revision is being reviewed for -
- enhancements.
The LaSalle operator trainir.g program will be revised to include training :
on the Unit 2 scram and equipment failures.
The Unit 2 thrust bearing assembly was disassembled and inspected. Clutch !
face irregularities-were dressed up. The clutch spring was retorqued. . l The lower ball bearing was exchanged with the upper bearing. The bushing i drive coupling and stem were drilled to accommodate a roll pin, and the set screw was re-applied.
The Unit 1 thrust bearing wear detector will be inspected during the LIR05 outage, and the appropriate corrective actions taken.
The Unit 2 drywell temperature monitoring system was modified prior to Unit 2 startup to install a cap on the end of the conduit in which the temperature sensor is located to delay sensor heat-up due to radiant heat. A computer algorithm change is being reviewed for both Unit I and 2 to ensure reliable information is annunciated in the control room.
The LaSalle operating training pregram will be evaluated to enhance- ,
training on the impori Ice of accurate, updated information provided in 10 l CFR 50,72 notificatice l
7NLD211W32 u sA
I
[ f 1
_ . . . . _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ . . _ _ . _ . _ _ _ _ _ _ _ _ _ _ _ _ . . . _ _ _ _ _ _ _ _ _ _ _ _ . _ . _ _ . - _ _ _ _ . _ _ _ _ _ _ _ _