IR 05000247/1986028
| ML20207N188 | |
| Person / Time | |
|---|---|
| Site: | Indian Point  |
| Issue date: | 12/23/1986 |
| From: | Norrholm L, Roxanne Summers NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION I) |
| To: | |
| Shared Package | |
| ML20207N157 | List: |
| References | |
| 50-247-86-28, NUDOCS 8701140040 | |
| Download: ML20207N188 (28) | |
Text
- - . . - . . - . _, _
.
.
.
'f-U.S. NUCLEAR REGULATORY COMMISSION DCS 50247-861020 861023 861106 861115
REGION I
Report N /86-28
- Docket N License N DPR-26 Licensee: Consolidated Edison Company of New York, In Irving Place New York, New York 10003 Facility Name: Indian Point Nuclear Generating Station, Unit 2 Inspection at: Buchanan, New-York l Inspection conducted: Ogaber 14,'1986 - November 21, 1986
- Inspectors
- L. Rossbach, Senior Resident Inspector P. Kelley, Resident-Inspector L. Briggs, Chief, Test Program Section F. Paulitz, Reactor Engineer N. Dudley, Lead Reactor Engineer (Examiner)
=1-s /> - 23- SS,
, Reviewed by: R. Summers, Pr A 1ect Engineer date
,
Reactor PPajects Section B, DRP r
Approved by: , wa a N holni, Chief /dat'e Leif flopProjects Section 28, ORP
.
'
Reactor
Inspection Summary: Inspection on October 14, 1986 - November 21, 1986 i (Report No. 50-247/86-28)
i~
Areas Inspected: Unit trips occurred on October 20, October 23, and November An unplanned shutdown occurred on November 15. The inspection effort focused on these events and associated component failures. Particular
- attention was given to the operation, modification, maintenance, and testing
,' of the auxiliary feedwater system. Also, an inspection of licensed operator recertifications was conducted. The inspection involved 320 hours0.0037 days <br />0.0889 hours <br />5.291005e-4 weeks <br />1.2176e-4 months <br /> by the
inspector "
8701140040 870105 PDR ADOCK 05000247
- G PDR
-
- . .--- .- - , . - . . . . - . - . - - . . . - , . . . - . . . - . . _ . . - - - - - . . . - - _ . .
.
.
Results: Several failures to properly test the auxiliary feedwater (AFW)
system were identified and resulted in a violation (Sections 4.1 and 6.6).
Other weaknesses in the testing of AFW components were also identified (Sections 4.2, 4.4 and 6.1). Concerns were raised about the impact on safety of deferring preventive maintenance from refueling outages (Section 5.1). The combined effect of component failures and scheduled maintenance created an increased risk of losing the AFW system (Section 5.2). A violation was also identified for not controlling partially completed AFW system modifications in accordance with procedures (Sections 4.1-and 4.3), and several AFW system design and operational concerns were raised (Section 3). The licensee recog-nized that AFW system reliability needs to be improved and volunteered to perform an AFW system functional assessment. Also, a meeting with the licensee is scheduled for December 9 in the Region I offices to discuss AFW system reli-ability, the licensee's corrective actions, and plans for the system functional assessmen Due to concerns about motor overcurrent protection, the licensee is reviewing overcurrent setpoints for all safety-related motors (Section 4.4).
The violations and concerns that were raised relating to the test program and control of modifications, the lack of oversight given to changes to the PM program, and the SNSC approval for plant restart on October 22 indicate weak-nesses in management oversight. An additional violation was identified for not submitting timely licensed operator recertifications to Region I. (Section 7)
L
.
.
DETAILS 1. Persons Contacted-Within this report period, interviews and discussions were conducted with members of the licensee management and staff to obtain the necessary information pertinent to the subjects being inspecte . Summary of Plant Events 2.1 Unit Trip on October 20 At 9:36 a.m. on October 20, reactor trip breaker B opened, tripping the unit from 100% power. The reactor trip breaker opened because normally energized relay SIAM-1X in the reactor protection circuit deenergized. The function of relay SIAM-1X is to send a trip signal to reactor trip breaker B when SIAM-1X senses a safety injection signal in train A of safety injection. No safety injection signal, which would have deenergized the relay, was present during this-even Following the unit trip, a 911ef valve (MS-52) lifted in the steam supply line to auxiliary feedwater (AFW) pump #22 and motor driven AFW pump #21 trippe The operator reset AFW pump #21, restarted it, and it continued running without further problem Motor driven AFW pump #23 started automatically and functioned normally. Steam generator level was maintained with AFW pumps #21 and #23. The plant was stabilized in hot shutdow While investigating the cause of the trip, the licensee found a loose connection in the circuit containing relay SIAM-1X. When the connection was tightened, the relay reenergized and closed. The licensee then tightened the accessible terminal block electrical connections in the reactor protection and safeguards relay panel The licensee also replaced relays SIAM-1X, SIAM-1Y, and SI-13X in the circuit that caused the trip although these relays were functioning properl On investigating the cause of the AFW pump #21 trip, the licensee discovered two overcurrent flags up: instantaneous overcurrent and long-term overcurrent. The Amptector overcurrent protection device was tested and found to be in calibration. The instantaneous overcurrent condition was not believed to have existed because there was no evidence of a short in the 480V circuit. The long-term overcurrent condition was also not initially believed to exis Another possible cause of the breaker opening was a low suction pressure trip provided for pump protection. The licensee verified that the suction pressure switches were calibrated properl The cause of the relief valve (MS-52) lifting in the steam supply line to AFW pump #22 was believed to be the inability of the pressure control valve (PCV 1139) to keep the steam pressure below the relief setpoint (700 +/- 20 psig) when the governor valve on the
q
.
turbine for AFW pump #22 was closed. The licensee reset the speed controller from 0% to 20% in order to open the governor valve and started the pump several times without a trip on overspeed or a relief valve lif A conference call was subsequently held between-the licensee and NRC Region I to discuss the causes and corrective actions for the trip and component failures. Between the time of the licensee's initial investigations and this conference call, the long-term overcurrent flag was again found to be indicating on the AFW pump #21 breaker with the breaker closed. As a result cf concerns expressed by Region I management related to the root cause analysis of the AFW pump #21 trip, the licensee committed to further testing of AFW pump
- 2 Region I also sent two specialist inspectors to the site to inspect this testing. The testing demonstrated that the motor current (560 amps) was within the the long-term overcurrent pickup band (600 amps +/- 10%). The long-term overcurrent setpoint was subsequently raised to 660 amps on AFW pumps #21 and #23. The licensee replaced the Amptector overcurrent protection device and sent it to its manufacturer for testin The Station Nuclear Safety Committee (SNSC) reviewed the cause of the trip, the component failures that occurred, and the corrective actions taken. The October 20 trip was the only event during this inspection period which required SNSC approval prior to restar SNSC approved restart and the unit was brought critical at 1:00 on October 2 Additional details and findings in the areas of operations, mainte-nance, modifications, and testing related to this trip and component failures are described later in this repor .2 Unit Trip on October 23 At 1:30 p.m. on October 23, the unit was manually tripped from 38%
power following the trip of #21 main boiler feedwater (MBF) pum Following the trip, steam-driven auxiliary feedwater (AFW) pump #22 tripped on overspeed. Both motor-driven AFW pumps started automatically and supplied adequate auxiliary feedwater. The plant was stabilized in hot shutdow Post-trip investigations fcund that the ring holding the check valve (BFD 1-1) in the discharge piping of #22 MBF pump had disengaged from the body of the valve. This condition would allow backflow of feedwater through #22 MBF pump whenever its discharge valve (BFD 2-22) was open. The circumstances of the #21 MBF pump trip were determined to be as follows:
_ _ _ _
---
.
The licensee was investigating why #21 MBF pump would not produce full flow at full speed, a problem that had first occurred on October 18, 198 This investigation involved collecting main feedwater system data during plant startu Unknown to the operator, backflow was occurring through #22 MBF pump due to failed open check valve BF0 1-1. The speed control for #21 MBF pump was in manual. The operator closed #22 MBF pump discharge valve BFD 2-22, stopping the backflow, but increasing flow to the steam generators. The feedwater regulating valves automatically reduced flow, but with the feed pump in manual speed control, pump speed remained constant and the pump tripped on high discharge pressur Post-trip investigations of the steam-driven AFW pump #22 trip revealed that pressure control valve PCV 1139, which controls steam pressure, was not controlling properly. During post-trip testing the valve bound. The valve was disassembled and the internals were found scored and boun PCV 1139 was repaired and the valve pressure controller was replaced, reset, and tested. Additional testing was completed, which demonstrated the proper functioning of PCV 1139 and that AFW pump #22 would start automatically without overspeeding, tripping, or lifting the relie Conference calls were held between the licensee and NRC Region I during and after the post-trip investigations and corrective actions. During these calls, the licensee committed to: 1) address N%C concerns by analyzing the effects of relief valve MS-52 (AFW pump #22 steam supply relief) failing open; 2) addressing in procedures the actions that should be taken if MS-52 fails open; and, 3) evaluating the operation of the AFW pump #22 overspeed trip device. The licensee also committed to do an AFW system functional assessment to improve system reliabilit The unit was brought critical at 3:48 p.m. on October 2 Additional details and findings in the areas of operations, maintenance, modifications, and testirg related to this trip and component failures are described later in this repor .3 Unit Trip on November 6 At 2:34 p.m. on November 6, reactor trip breaker B opened, tripping the unit from 97% power. Safety systems responded normally to this event and the plant was stabilized in hot shutdow At the time of the trip a surveillance test was being performed on the low pressurizer pressure trip function of the reactor protection system. Post-trip investigations determined that a relay contact in the trip circuit was not closing. However, those contacts which provide the alarm and computer inputs were making proper contact to show relay status. As a result, there was no indication that one channel was in trip. When another relay was tripped as part of the
.
.
surveillance test, it made up the required 2 of 4 trip logic. All three relays in the trip circuit were replaced as part of the licensee's corrective action The unit remained down for other maintenance which included an inspection of other relays and the replacement of the motor on #23 AFW pum The unit was brought critical at 1:00 a.m. on November 9 and then returned to 97% power. Power was being limited by vibrations in the electrical generator end turn Additional details in the area of maintenance related to this trip and component failure are described later in this repor .4 Unplanned Shutdown on November 15 On November 13 at 10:50 p.m., the licensee was preparing to fill #21 accumulator. The operators can fill the accumulators with either the dedicated Accumulator Topping Pump or with a Safety Injection (SI)
Pum Since the Accumulator Topping Pump was out of service for repairs, the operators planned to use #22 SI pump. Prior to starting the pump, an NPO vented the pump to ensure the casing was full of water and not gas or air. The venting showed no signs of gas or air and the pump was then started from the CCR, but it tripped almost immediately. An NPO was unable to bar over the pump by hand. There-fore, per Technical Specification 3.3.A.2.a, the unit entered into a limiting condition for operation (LCO) which requires the plant to be shutdown using normal operating procedures in 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> unless the pump is repaired. #22 SI pump was last operated on November 11 to fill the accumulators and no problems were noted at that time. On November 14, the licensee removed the pump casing and inspected the pump internals. There were no visible signs of damage to the inter-nals or any signs of foreign material which would prevent the pump from rotating. A new internals package was installed in the pump, but due to a dimensional problem with the new internals, it had to be disassembled and restacked in order for the package to fit in the pump casing. Due to restacking the internals, the 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> LCO expired and the licensee commenced a plant shutdow The reactor was shut down at 3:45 a.m. on November 15. At 4:23 a.m., with the plant shutdown still in progress, the licensee tripped #22 main feedwater pump by closing its discharge valve. This is considered a normal evolution. After the main feedwater pump tripped, #21 auxiliary feedwater (AFW) pump did not start automatically as it should hav #23 AFW pump did start automatically. The operators manually started
- 21 AFW pump from the Central Control Room. At 10:50 p.m. on November 16, the licensee began to go into cold shutdown due to #22 SI pump being out of service. At 2:22 a.m. on November 17, #22 SI pump was declared operable by performing PT-Q20, Revision 0, SI Pump Functional Check, and the plant cooldown was terminated. Investiga-tions of the #21 AFW pump auto-start circuits identified loose
connections. These were tightened and the pump was returned to servic Additional details in the area of maintenance related to this unplanned shutdown are described later in this report.
3. Operations-3.1 Steam-Driven Auxiliary Feedwater Pump Auto Start Following the unit trip on October 20, the relief valve (MS-52)
lifted on the steam supply to auxiliary feedwater (AFW) pump #2 The licensee subsequently increased the setting of the AFW pump #22 speed changer from zero to 20% open. This was done to place a steam demand on the system and make it easier for the steam supply pressure control valve (PCV 1139) to keep steam pressure below the setpoint for MS-5 The inspector questioned staff from operations, technical support, and training about operation of the AFW system at the zero and 20%
speed changer settings. The licensee also demonstrated system operation for the inspector on the plant simulato Licensee staff understood that the governor valve for AFW pump #22 turbine was closed at the zero setting and partially open at the 20% settin It was also their understanding that the turbine would not be spinning when the speed changer was set at zero unless the governor valve was leaking. The plant simulator indicated about 500 RPM pump rotation after a pump start signal with the speed changer on zer Valve leakage was assumed by the simulator model to cause the rotation. The Technical Specifications and FSAR require that AFW pump #22 automatically start following station blackout or low-low level in two steam generators. The licensee's interpretation is that the auto-start requirement is met by the auto-open signal to PCV 1139 and that the pump need not be turning. Before returning to power on October 23, the licensee placed the speed changer at the 20% setting to keep MS-52 from lifting. This also assured that AFW pump #22 will be running following an auto-start signa Following the October 23 unit trip and trip of AFW pump #22, PCV 1139 was repaired and its controller adjusted so that it would regulate steam supply pressure properly. During post-maintenance testing, it was discovered that the pump actually spins at about 3300 RPM when the speed changer is set at zero. The normal speed changer position was, therefore, returned to zero and additional testing prior to plant startup on October 26 adequately demonstrated that AFW pump #22 will start upon receipt of an auto-start signa The inspector concluded from these events that the licensee did not fully understand how AFW pump #22 responds to an auto-start signa Contributing to this were the following: simulator modeling was incorrect; pump speed with the speed changer set at zero is not
.
.
recorded in surveillance tests; and pump governor adjustments are done by a vendor representative without any procedural controls on the' minimum pump spee The licensee has recently assigned a system engineer to develop and retain a detailed knowledge of the AFW system design and operatio This should prevent details of the system operation from being lost again. The licensee is also performing an AFW system functional assessment. Simulator modeling, testing, and the adequacy of procedural controls over pump minimum speed will be considered by this assessmen .2 Position Indication for Auxiliary Feedwater Steam Supply Valve Steam-driven auxiliary feedwater pump #22 started automatically following the unit trip on November 6. The inspector observed that only the green (closed) indicating light was on for steam supply valve PCV 1139 although the valve was in a partially open, modu-lating position. The correct indication for this valve as described in modification procedure EGP-85-30720 is for both the red and green indicating lights to be on when the valve is modulating. Although other indications are available to verify proper system operation, the present PCV 1139 indication is not a good human factors practic The inspector discussed this with operators and engineering manage-ment and verified that the licensee had identified this problem and given engineering the responsibility to correct i .3 Loss of Air Concerns for the Auxiliary Feedwater System Nitrogen gas bottles are used as a backup to the instrument air system supply to the auxiliary feedwater system because the instru-ment air compressors are stripped from their busses following a station blackout or safety injection. A system walkdown performed by the inspector and confirmed by the licensee identified that a nitrogen backup is not provided to operate the auxiliary feedwater pump #22 steam supply pressure control valve (PCV 1'.!?) or speed controller (HC 1118). The inspector discussed this with licensee representatives who stated that the need for such a backup air supply would be considered in the assessment of the auxiliary feedwater system. The inspector verified that procedures to operate this pump locally upon loss of instrument air are contained in abnormal opera-ting instruction A-27.1.9, Rev. 3, " Control Room Inaccessibility Safe Shutdown Control."
3.4 AFW Relief Valve Failure Analysis In a conference call with NRC Region I management on October 26, the
'
licensee committed to analyze the effects on AFW pump #22 of steam relief valve MS-52 failing open and include in procedures the oper-l ator actions to be taken if MS-52 fails open.
i i .
.
The licensee's analysis of MS-52 failing open concludes that AFW pump
- 22 would still deliver 320 gpm provided the pump's recirculation line was isolated. This is insufficient flow to meet the pump's technical specification operability requirements, but it does meet the minimum analyzed post-accident feedwater requirements. Operating procedure revisions are currently being prepared to describe operator *
actions to be taken if MS-52 fails ope .5 Steam Line Drains The inspector reviewed system operating procedure (SOP) 21.3, which describes starting the AFW system and observed the system being manually placed in service. In accordance with this SOP, water was drained from two drain valves (MS-94-16 and 17) in the AFW steam supply before startup. Since this is a manual operation, water would not be drained from this line before an automatic system start. The inspector discussed this with the licensee, who also confirmed that draining these low points is not done as a normal part of the operations preventive maintenance program. The licensee's AFW system assessment will review adding this to the plant operators' round .6 Emergency Operating Procedures Status Trees The operators and shift technical advisor (STA) use Emergency Operating Procedures (E0P's) in responding to plant trips. The STA is responsible for following plant parameters using the E0P status tree Following the October 23 plant trip, the inspector observed the STA report to the control room less than three minutes after the plant trip. The STA could not find a copy of the status tree About nine minutes after reporting to the control room, a staff engineer brought a copy for the ST The control room copy was located later. The same problem had occurred following the October 20 tri Prior to startup from the October 23 trip, the inspector verified that status trees were available in the control room, with the rest of the E0P's, in a clearly marked "EOP - Status Trees" binder. The status trees had previously been in an unmarked standard office folder. On October 27, the inspector asked the STA where the Status Trees were and he looked in the wrong place. The inspector discussed this with the generation support section which has responsibility for maintaining the E0P's. Generation support subsequently informed the STA's and operators of the new binder and the locatfor, for the status trees through shift turnovers and a memo which was required reading. The inspector considers that the corrective actions are now adequat r- :--
. .
3.7 Document Retrieval During this inspection many documents involving maintenance, madifications, tests, and design bases were requested and reviewed by the inspectors. The considerable time it took to retrieve these documents and the incompleteness of some of them were a delay in completing the inspection. The licensee's staff was also delayed in corrective actions by difficulties in document retrieva Due to previous problems with document retrieval, the licensee is microfilming all records in offsite storage and bringing them on site. Also, staff and their records located at the licensee's corporate office are being moved to the site and consolidated ther Although these actions should improve access to records, it will take at least a year to complete them. The licensee also recently revised the filing of post-maintenance test data so that a complete test package could be more easily retrieve . Modifications As part of the inspection of the auxiliary feedwater (AFW) system, the inspectors identified four modifications for review. 1) Pressure control valve PCV 1139, which regulates the steam supply to turbine-driven AFW pump #22, was replaced during the 1986 refueling outage under modification 85-3072 Failure of this valve to regulate steam pressure properly is the probable root cause of relief valve MS-52 lifting on October 20 and AFW pump #22 tripping on October 23. 2) The motor for AFW pump #21 was replaced during the 1984 refueling outage under modification 84-4077 This motor tripped on October 20. 3) Suction flow switches on AFW pumps
- 21 and #23 were replaced with suction pressure switches in October 1986 under modification 86-52652. A signal from this switch was initially a possible root cause for AFW pump #21 tripping on October 20. 4) Over-current trip devices were replaced on 480V breakers during the 1984 refueling outage under modification 82-15485. An overcurrent condition or malfunction of the overcurrent trip device are possible root causes for AFW pump #21 tripping on October 2 .1 PCV 1139 Replacement This modification involved the replacement of PCV 1139 and control This pressure control valve regulates the steam pressure to the turbine of AFW pump #22. Installed were a new valve, valve operator, positioner, and a control rack with mounted controls including solenoid valves and pressure controlle The inspectors examined the completed installation and reviewed the following documents:
-
Modification procedure MPC-85-30720 (the mechanical portion of the modification).
...
.
-
Modification procedure EGP-85-30720 (the electrical portion of the modification).
-
Maintenance procedure MP 13.53 for Work Permit 209051, mechanical portion of modification and for Work Permit 208330, the electrical portion of the modificatio Temporary procedure change 86-0032 to MP 13.53
-
Work Order 6407, installation of the modification
-
SAO-405 Modification Tracking Form for Work Order 6407
-
Post Maintenance Test (PMT) 6407 This modification was not completed before declaring the auxiliary feedwater system operational. The incomplete portion involved the addition of a block and bleed valve on an air line to the actuator for PCV 1139. Licensee procedures which control the return to service of partially completed modifications were not followed in that a report of installation was not completed and a startup authorization was not issued. This is a violation (50-247/86-28-01). A second instance of this procedural violation is described in para-graph 4.3 of this section which describes the installation of pressure switches for motor-driven AFW pumps #21 and #23. When the inspectors identified these two improperly controlled incomplete modifications to the licensee, the licensee stated that four addi-tional incomplete modifications on other systems may have been handled the same way. At the end of this inspection period, the licensee was reviewing partially completed modifications, completing what work was necessary to close them out, and revising the adminis-trative procedure for modifications (SA0 405).
The inspectors also reviewed the post-maintenance (modification)
tests for PCV 1139. The maintenance procedure for installation of this modification (MP 13.53) did not list the post-maintenance test as having been completed and accepted. The modification tracking form lists PMT G407 as the post-maintenance test. However, PMT 6407 was not a functional test of the modification. PMT 6407 included a valve PCV 1139 stroke test and an overspeed test of AFW pump #22, but it did not test to determine that the newly installed pressure controller was operating properly or that the AFW pump #22 system was still operable. During PMT 6407, which was run with the pump in recirculation only, the pump did not achieve more than 2600 rpm which is insufficient to achieve design flow. Also, while performing PMT 6407, the operator noted in the remarks section that PCV 1139 was not controlling pressure properly. Shortly after PMT 6407 was ccm-pleted, PCV 1139 failed open. It was repaired and retested, but the retest, another valve stroke and overspeed test, also was not a full functional test. On May 30, the quarterly surveillance test, PT-Q34 was performed which also identified that PCV 1139 was not controlling
_
,
.
pressure properly. The inspectors concluded that this modification was not properly tested before the AFW system was returned to service in that PCV 1139 was not demonstrated to be controlling steam pres-sure properly. This is a violation (50-247/86-28-02). Another instance of a violation involving the testing of AFW pump #22 is identified in Section 6.6 of this repor .2 AFW Pump #21 Motor Replacement This modification involved modifying the motor baseplate and installing a new AFW pump #21 motor manufactured by Reliance Electric in place of the original motor manufactured by Westinghous The inspectors examined the completed installation and reviewed the following documents:
-
Modification procedure MFI-84-40774
-
Safety evaluation NS-2-84-071
-
Modification completion form for MFI-84-40774, Rev. 1
-
Modification tracking form for MFI 84-40774
-
Station Nuclear Safety Committee minutes #844 and #858
-
Time-current characteristic curve for AFW pump breaker positions 6C and 9C, dated August 3, 1984 and October 24, 1986
-
Induction motor data sheet for Westinghouse and Reliance AFW pump motors
-
Post Maintenance Test (PMT) 2935 The full load name plate current for the original Westinghouse motor was 458 amps and for the new Reliance motor is 477 amps. The inspector expressed concern about the higher ampere rating of the replacement motor due to the possibility that an overcurrent condition caused this pump to trip on October 20. The original test requirements for this modification included a check of the 3 phase currents of the motor. Revision 1 to the modification deleted this test prior to installation. The safety evaluation for the original modification and test revision concluded that the Reliance motor was an equivalent replacement for the Westinghouse motor. The Station Nuclear Safety Coma,ittee (SNSC) approved the modification and test revision. No documentation was provided that the ampere ratings were considered in the safety evaluation and in SNSC reviews of the modification and revised test requirements. Licensee staff believe the current differences were considered by the relay protection staff to not be sufficiently significant to change the overcurrent
..
.
protection device (Amptector) settings. The inspector discussed with the licensee the potential importance of a change in motor characteristics and the need for better documentation that such changes are properly evaluate .3 AFW Pump Suction Pressure Switch Installation This modification involved replacing suction flow switches and time delay relays with suction pressure switches for motor-driven AFW pumps #21 and #23. The switches provide a trip signal to the motor breakers to protect the pumps in the event of low suction pressur The inspectors examined the completed installation and reviewed the following documents:
-
Modification procedure MPE-86-52652
-
Maintenance procedure MP-13.57
-
Work Orders 86-26400 and 86-26409
-
Post Maintenance Tests 26400, 26409, PC-EMil, and TPC-86-107T This modification was not completed before declaring the AFW pumps operational. The licensee is currently considering deleting the uncompleted portion of the modification, a second contact from the switches. Although the licensee has procedures which control the return to service of a partially completed modification, these procedures were not followed in that a report of installation was not completed and a start-up authorization was not issued. This is the second example of this procedural violation (50-247/86-28-01).
The inspectors reviewed the post-maintenance tests (Section 6.2 of this report) for this modification and concluded that the partially completed modification was properly tested before return to servic .4 Breaker Overcurrent Trip Device Replacement This modification involved the replacement of original breaker overcurrent trip devices with a new device (Amptector).
Due to the reactor trip of October 20, and the subsequent trip of motor-driven auxiliary feedwater pump #21, two region-based inspec-tors were dispatched to the site to observe a pump #21 test on October 22. The inspection is discussed in Section 6.3 of this report. A followup inspection by one region-based inspector from October 27-31 was conducted to review the modification of the 480 volt breaker protective trip elements and their setpoints. During that followup inspection, the following documents were reviewed:
-- .- - - . _ _ - _ _ _ -- _ - - _ _ _ . . _ __ - _ _ . . _ .
.
.
-
Modification Procedure ESG No. 82-15485 DB Breaker Trip Device Replacement, 9/23/82, Rev. 2, dated 6/11/84
-
Instructions for the Retrofit of DB Breakers I.B.33-850-6, August 1982
-
Addendum Instructions for Westinghouse Retrofit Kits for DB Breakers I.B.33-850-6
-
Selected test results for the above modification
-
PT-R46, Rev. 2, DB Circuit Breaker Overcurrent Test - 1984
-
Motor and pump data for:
-
Safety Injection Pump
-
-
Safety Injection Recirculation Pump
-
Service Water Pump
-
Containment Cooling Unit Fan
-
Containment Spray Pump
-
Auxiliary Feedwater Pump
-
Component Cooling Pump
-
Overcurrent Trip Setpoint for the above breakers
-
Engineering Support Request IP-1055, April 28, 1981, Retrofit of DB Breakers with DS Solid State Devices
-
Memorandum, July 9, 1984, IP-2, 480V Essential Motors Revised Settings
-
Temporary Procedure Change (TPC) 84-160T, 8/3/84, Change Pickup Currents
-
Calculation No. 69982-TA/1, 10/27/86, 480V DB Circuit Breaker Setting for the 400 HP Auxiliary Feedwater Pump Motors
-
Field Engineering Request (FER) No.96-229, Approved 10/29/86, Review Settings for All Safety-Related 480V Motors
-
Engineering Support Request (ESR) IP-60817,10/29/86, Same subject as FER 96-229
-
Maintenance Procedure (MP) 16.25, Rev. 2, Air Circuit Breaker DB 50 Inspection
-
Data Sheet from MP-16.25 for Auxiliary Feedwater Pump #21, breaker position 6C, SN 24Y4800 7x2, Work Permit 173487, MWR 13752, dated 7/13/84
.
.
The inspector discussed the DB breaker trip modification and setting-of the protective trips for the various breakers and also observed the overcurrent trip test for auxiliary feedwater pump #21 breake The inspector found that a functional test by tripping the breaker through the Amptector is not being performed after breaker PM and overcurrent trip test. PT-R46 only tests the Amptector calibration and does not provide a functional test to assure the breaker is tripped via the Amptector. The licensee has agreed to include this functional test in the procedure. This is an unresolved issue until the licensee provides a method to do the functional test and revises '
the procedure. (50-247/86-28-03)
The inspector reviewed existing overcurrent trip settings for safety-related equipment and determined that they may not be conservative. The licensee has initiated a review of the trip settings of all 480V safety-related motors on busses directly associated with the emergency diesel generators. The effects of expected pump loads coincident with a degraded voltage condition is to be considered. The licensee is expected to complete the review by December 17, 1986. By the end of this inspection period, the licensee had not identified any trip setpoints that needed resetting. This is an unresolved item until the licensee completes this review and revises any necessary trip setpoint (50-247/86-28-04)
5. Maintenance 5.1 Refueling Outage Preventive Maintenance Changes Portions of the auxiliary feedwater system preventive maintenance program were reviewed. As described below, preventive maintenance for several components of the auxiliary feedwater system was deferred from the last refueling outage. The inspection raised concerns that this may have an adverse impact on system reliability and plant safet Preventive maintenance for auxiliary feedwater pump #23 motor was rescheduled from the last refueling outage until the next refueling outage. About three months after the last refueling outage the motor failed. The motor was subsequently repaired under work order 86-27401 and then replaced under work order 86-2819 As discussed in inspection report 50-247/86-27, a preventive maintenance inspection of an auxiliary feedwater system check valve was planned in response to industry experiences with check valve failures leading to steam binding of auxiliary feedwater pumps. The licensee later decided not to do that inspectio .
.
Until the last refueling outage, 480V breaker preventive maintenance was done during refueling outages. They are now done during plant operation. This requires taking safety components out of service, since spare breakers are not available. Although this is done within the Limiting Conditions of Operation of the facility's Technical Specifications, it affects plant risk by reducing safety system availabilit The inspector discussed his concerns about changes in the preventive maintenance program with the chairman of the Nuclear Facilities Safety Committee (NFSC). The NFSC is sponsoring the auxiliary feedwater system functional assessment to be done the first quarter of 1987. This will include an assessment of the preventive maintenance program, recent changes in that program, and the review and oversight given changes to the program. It appears that this assessment will address the inspector's concerns; the inspector will follow this assessment. (50-247/86-28-05)
5.2 AFW Train Unavailability Due to planned Maintenance As mentioned above, the deferral of preventive maintenance from the refueling outage reduces safety system availability. In addition to preventive maintenance, other scheduled maintenance was done on the AFW system in October 1986. For 21% of the time the reactor was critical in October, a motor-driven AFW train was out of servic This outage rate, together with the poor reliability of the system due to component failures, created an increased risk of losing the AFW syste .3 Loose Connections On October 20, reactor trip breaker 8 opened tripping the uni Post-trip investigations determined the cause of the trip to be a loose connection that caused relay SIAM-1X in the reactor protection circuit to deenergize and trip the uni Before returning to power, the licensee checked the tightness of 10,300 terminal block connec-tions in the reactor protection and safeguards relay racks. On November 15, while shutting the unit down, #21 auxiliary feedwater pump failed to start automatically due to a loose connection on a terminal block in the pump's auto-start circuit Before returning to power from this event, the licensee checked the tightness of 9865 connections in the supervisory panels. The inspectors observed portions of the licensee's post-trip check of connections and dis-cussed the results of both connection checks with the license Twenty-six connections out of the 10,300 checked initially were more than one-half turn loose and three screws were broken during the check. One-hundred and twenty connections out of the 9865 checked following the November 15 event were more than one-half turn loose
, and seven screws had stripped threads or were mis-sized. Connec-l tions have not previously been included in the licensee's preventive t
.
.
maintenance program, but as a result of these findings, the licensee plans to check additional connections the next time the plant is subcritical. Also, the licensee's engineering staff and quality assurance staff are developing recommendations for connection maintenance. In addition, specific preventive maintenance recom-mendations are being developed for the type of block (type 601
,
blocks manufactured by Kulka) on which screws brok A special inspection was conducted by NRC as a result of these loose connections. That inspection is reported separately in inspection report 50-247/86-3 .4 Failed Relays On October 23, while testing the safety injection logic relays, the licensee discovered low pressurizer pressure relay PC 457E to be inoperable. The relay was replaced. On November 6, the unit tripped while performing the monthly pressurizer pressure surveil-lance test. Post-trip investigations determined that the trip was caused by an open contact in either relay PC 456A or PC 474A. These relays and a third relay in the circuit (PC 455A) were replaced. All of the above relays are normally energized, AC, BF relays. The relays were original plant equipment and plant aging appears to be the cause of their failure. The relays have been sent to the manu-facturer for analysi The licensee has had almost no problems with AC, BF relays previously and most of those installed in the plant are original equipmen These relays are not included in a preventive maintenance progra As a result of these events, the licensee inspected BF relays, but did not identify any that needed replacement. The licensee's engi-neering staff is evaluating what further cctions could be taken to improve relay reliability, including replacing the relays and going to a solid state syste .5 Changeout of Three SI Relays Although the cause of the unit trip on October 20 was identified to be a loose connection, the licensee decided, conservatively, to change out three relays in the circuit even though the relays appeared to be functioning properly. The relays were SIAM-1X, SIAM-1Y and SI-13 These are BFD relay The inspectors observed portions of the changeout and reviewed the post-maintenance tes . - .- . .- - . _ . _ . _ - . ,. - -- ,. . - . - .
.
.
5.6 PCV 1139 Repair During testing of PCV 1139 (the AFW pump #22 steam control valve)
following the unit trip on October 23, the valve bound. The valve was disassembled and the plug and cage were found galled and boun The galling was attributed to weld spatter due to inadequate cleanliness controls during valve installation. The plug and cage were repaired by lapping and honing. The inspector observed portions of the repair. The vendor representative was present for the repai .7 Repair of Main Feed Pump Check Valves During post-trip investigations, the licensee discovered the ring holding check valve BFD 1-1 in the discharge of #22 main feed pump had disengaged from the body of the valve and was lying inside the valve bod This valve failure had contributed to the unit trip on October 2 The ring holding the check valve disk is attached to the valve body by two lock plates and bolts. The bolt is restrained by a lock wire. Both bolts had completely backed out. BFD1, the check valve on the #21 main feed pump discharge, was examined and one cf its bolts was partially backed out and had the head missin These valves are 20" Crane list 979 tilting disc check valves. Both valves were repaired by installing new bolts and securing them with welds instead of lock wire The inspector examined both valves and reviewed their maintenance history. The valves were nct included in a preventive maintenance program but had been inspected and repaired in December, 1982. At that time, new hinge pins were installed and missing lock pins and a pivot pin were replaced. This work was done in response to an t.'ormation bulletin from the manufacture .8 Repair of #22 Safety Injection Pump i #22 SI pump seized on November 13 while attempting to fill an l Accumulator. The pump casing was df sassembled and the internals package was removed per MP 7.62 " Removal and Installation of Internal Assembly of SI Pump." There were no visible signs as to l what caused the pump to seiz The inspectors observed the licen-l see's attempt to install a new internals package, which would not fit properly in the pump casing. The new internals 9ackage was disas-sembled and restacked. The internals package then fit properly in the pump casing. The pump was tested satisfactory using the SI pump
, quarterly test PT-Q29. During the reassembly, a representative from l
the Pacific Pump manufacturer was present. The insoectors examined the seized internals package which the licensee had partially disassembled. No conclusions have been developed by either the l
licensee or Pacific Pump as to the cause of its failure. This is an open item. (50-247/86-28-08)
'
.
.
6. Surveillance Tests Due to the problems encountered with the Auxiliary Feedwater (AFW) system this inspection period, extensive review of the licensee's AFW surveillance tests, post-maintenance tests (PMT) and post-modifications tests were performe .1 AFW Flow Control Valves PMT #29089 was performed on October 22 to determine if motor-driven auxiliary feedwater (AFW) pump #21 was tripping on overcurrent due to faulty Amptector settings on the pump breaker or an actual overcurrent condition. During the test it was noted that the AFW flows to the steam generators were not approximately 150 gpm to each generator as expected. The following AFW flows were recorded during the test: steam generator #21, - 300 gpm; steam generator #22, - 130 gpm; steam generator #23, - 170 gpm; and, steam generator #24, - 180 gpm. Also, the Central Control Room (CCR) controllers for AFW Flow'
Control Valves (FCV's) 406A, B, C, D valve positions were labled with the following settings for 150 gpm flow: FCV 406A, 0% closed (100% open); FCV 406B, 49% closed (51% open); FCV 406C, 63% closed (37% open); and, FCV 4060, 28% closed (72% open). The valves are set differently due to each valve having a different trim (internal).
The licensee uses flow data obtained from PMT-63, " Motor-Driven Auxiliary Boiler Feed Pump Flow Test," to determine the settings for the FCV 406 controllers to deliver 150 gpm to each steam generato PMT-63 was performed on December 22, 1984 with the steam generator pressures at 950 psi Per this test, the following positions for the FCV 406 valves were recorded: FCV 406A, 15% closed (85% open);
FCV 406B, 60% closed (40% open); FCV 406C, 63% closed (37% open);
FCV 4060, 37.5% closed (62.5% open).
The design condition of the AFW system is to deliver 150 gpm to each steam generator with steam generator pressure at 1100 psig. The licensee cannot functionally check the AFW flow with steam generator pressure at 1100 psig, due to the steam generator safety valves lifting. Therefore, the licensee user pressure drops, pump curves, design pressures, and flow coefficients, Cv, to extrapolate the FCV 406 settings for 150 gpm flow at 1100 psig steam generator pressur This setting ensures operator action is not immediately required to obtain a minimum flow rate to the steam generators upon automatic initiation of AF The licensee also performs PT-R7A, " Motor Driven Auxiliary Boiler Feed Pump Full Flow Test" each refueling period. The test checks and adjusts recirculation flow, the low pump suction flow alarm and the low flow pump trip setpoints. FCVs 406A, B, C and D are adjusted to obtain 150 gpm flow to each steam generator and their valve positions anc pump discharge pressures are recorded. The
___ . _ _ _ _ _ _ _ _ _ _ - _
.
pressure / flow data obtained from the test is checked against a flow versus steam generator pressure curve for acceptance. This test was last scheduled on January 14, 1986, the beginning of the 1986 refueling outage. During that performance of the test. #21 AFW pump did not meet the flow requirements, #21 and #23 AFW pumps did not trip on low suction flow trip signals, and the recirculation lines for both pumps indicated flow with the recirculation valves shu The low flow from #21 AFW pump caused it to fail the operability acceptance criteria of the tes The settings for FCV 406A, B, C and D, as obtained from PT-R7A, are as follows: FCV 406A, 0% closed (100% open); FCV 406B, 30% closed (70% open); FCV 406C, 66% closed (34% open); FCV 406D, 41% closed (59% open). It is noted again that #21 AFW pump failed its portion of the flow acceptance criteria with the FCV's in these position PT-R7A was performed again on January 14, 1986 for AFW pump #21 only. During this performance, FCV 406A was adjusted in order to meet the flow criteria. AFW pump #21 passed the test this time, but the FCV 406A and B valve settings were not recorded on PT-R7A due to the test engineer deleting this requirement. PT-R7A was performed again on October 22, 1986 to verify the AFW flow rates. During this performance, the following FCV 406 valve positions were recorded to obtain a minimum of 167 gpm flow rate to each steam generator: FCV 406A, 30.5% closed (69.5% open); FCV 406B, 43% closed (57% open);
FCV 406C, 61% closed (39% open); FCV 406D, 28.5% closed (71.5%
open). The licensee is currently using these results for the FCV 406 settings and has stated that PT-R7A test data will be used from now on in determining the valve (controller) position It is noted that on December 7, 1984 the licensee notified the NRC that the FCV 406 valves were discovered set to deliver 60 gpm instead of 150 gpm, as required. The licensee had, at that time, reset the valves so they would deliver 150 gpm flow rate. This was the subject of a special resident inspector report (50-247/84-34).
The continuing difficulty the licensee has had in setting these valves properly, indicates that the licensee's previous corrective actions were inadequat The licensee does have an Engineering Support Request (ESR)
- IP-50806 under evaluation, where AFW flow woulo be an input to the FCV 406 controllers. With this input, the FCV 406 valves would constantly deliver 150 gpm flow without regard to steam generator pressure or valve internal .2 AFW Suction Pressure Switches j As described above, PMT #29089 was performed on October 22, 1986 to I determine the cause of #21 AFW pump tripping. One parameter moni-tored during the test was AFW pump suction pressure. A modification, which is partially completed, was performed at the beginning of October 1986. The motor-driven AFW pumps (AFW pumps #21 and #23)
.
.
low suction flow instruments were replaced with low suction pressure instruments which provide pump protection by tripping the pumps at 9 psig suction pressure. It was thought that the suction pressure may be approaching the trip setpoint during pump start or that the instruments may be set incorrectl During PMT #29089, the normal pump suction pressure was approximately 40 psig with the pumps secured. During pump startup and after the pumps have been running, the lowest suction pressure recorded was 30 psig. PMT's 26400 and
'
26409 for AFW pumps #21 and #23, respectively, were reviewed for the installation and testing of the low suction pressure switches. Both suction pressure switches have been satisfactorily calibrated, leak tested in place and functionally tested in place, including tripping the pump's 480V breakers. The licensee is currently evaluating if they will complete the modification packag .3 AFW Breaker Overcurrent Protection As previously discussed, PMT #29089 was performed to check on the possibility of AFW pump #21 tripping on an actual overcurrent con-dition or because of a faulty overcurrent trip device (Amptector).
Upon initial investigation after #21 AFW pump tripped on October 20, it was found that the long delay and instantaneous overcurrent flags were up on the Amptector overcurrent trip device installed on #21 AFW pump 480V breaker. The licensee believes it~is not logically pos-sible for both flags to be up simultaneousl Prior to performing the PMT #29089 test, the Amptector device on #21 AFW pump breaker was changed ou For the test, 480V bus voltage was reduced to 460V to assume worst case conditions. At 460V, the low bus voltage alarm annunciates in the CCR. With the plant in hot shutdown, low levels in a steam generator were simulated to auto-matically start AFW pumps #21 and #23. Ammeter readings were taken with a portable device connected to the three phases leading to AFW pump #21. The 480V bus voltage drop due to starting the pumps was also recorde Due to a reactor trip at 9:36 a.m., October 20, and the subsequent trip of AFW #21 (electric) and the lifting of the steam relief valve on AFW pump #22 (steam driven), two region-based inspectors were dispatched to investigate and evaluate the reason for the trip of the motor-driven AFW pump (#21) and to witness a test of the pump to verify its operabilit Each of the two motor-driven AFW pumps (#21 and #23) are 100% capacity pumps (400 gpm each). The steam-driven AFW pump (#22) is a 200% capacity (800 gpm) pum When the inspectors arrived on site, a copy of the test procedure was provided to and reviewed by the inspectors. The inspectors asked several questions concerning conduct of the test and locations of test personnel. All questions were answered by the licensee. In addition the inspectors toured the various areas that personnel would be stationed and discussed data to be take .
. . -
_ -
.
.
The licensee informed the inspectors that following the October 20 trip, positive flow indication had been obsersed on AFW pump #21 and
.
that the pump trip had occurred sometime bet'<een initial start and four minutes into the. event when it was noticed that flow from AFW pump #21 had stopped. At that time, the operator reset the AFW pump
and started it. It ran without further problem Subsequent investigation by the licensee indicated that both the long time and instantaneous overcurrent trip flags (indicators) were up on the breaker for AFW pump #21, indicating that the breaker had tripped on overcurrent. However, both flags should not be indicating. The Itcensee checked calibration of the Amptector overcurrent device and found it correct. The licensee also checked the low suction pressure trip for the #21 AFW pump and found it correct. The licensee also stated on October 21-that AFW pump #21 had been started and run five times subsequent to its October 20 trip without tripping; however, '
the long time overcurrent trip flag had indicated during that tim The Amptector for AFW pump #21 was replaced with a new device and set at the same overcurrent trip value The testing of AFW pump #21 was to be conducted under Work Order NP-86-29089, designed to duplicate the conditions that caused the AFW pump #21 motor breaker to trip. The following were the initial conditions for the test:
-
Motor bus voltage was reduced from 480V to 460V. Degraded grid voltage alarm was receive AFW pump discharge flow was set at 300 gpm per pump (150 gpm perS/G).
The following readings were to be taken during the test:
-
AFW pump #21 motor current (by clamp on Ammeter locally).
-
AFW pump #21 suction pressure (local).
-
AFW pump motor bus voltage (in control room).
-
AFW pump flow (control room).
NRC and licensee personnel were stationed to monitor the desired parameters in the control room, switchgear room and at the AFW pump . Test Results The bus voltage dropped to 425 volts during the motor start and returned to 450 volts. Pump suction pressure dropped from 38 psig and remained at 28 psig. Motor current steadied at 560 amps. All phases were balanced. AFW pump #21 flow was 450 gpm, not 300 gp _ _ _ _ _ -
. .
.
The test of AFW pump #21 was successful. No abnormal electrical conditions were noted except that motor phase current was somewhat higher than expected as discussed belo . Analysis AFW pump #21 motor name plate rating is 400 HP, 440 volts, 477 amperes, service factor 1.15, 3552 rpm, Design B, Code F temperature rise 40 degrees C. The motor also has a rating at 80 degrees C, of 460 HP, 440 volt, 549 amper The insulation is ENDURASEAL and the motor manufacturer is Reliance Electri The pump is an Ingersoll Rand type 3HMTA, nine stage, horizontal split case centrifugal, rated at 400 gpm at 1350 psi.
The pump discharge flow during the test was about 1 percent greater than rated. This reflected to the motor input is about 15 percent. The motor voltage reduction was about 5 percent below rated. This 20% increase in motor load is in line with the increased current from design of 477 amperes to about 560 ampere The trip setting for the long time trip was 4.95 secondary amperes as calibrated July 13, 1984. This represents 594 primary amperes. Stated accuracy of the Amptector is +-
10% which corresponds to +-0.5 amperes at the 5 ampere setting. The minimum level of secondary current that could possibly cause Amptector action is 4.5 amperes which represent 558 primary amperes. The time calibration at 2x was 246 seconds and at 6x was 29.9 seconds. The original secondary current trip setting was 5 amperes which repre-sents 1.25% of rated motor curren . Corrective Action The licensee changed the trip setpoint for AFW pump motor
- 21 and #23 from the 1.25 percent overload, 600 primary amperes, 5 secondary amperes to 1.35 percent overload, 660 primary amperes, 5.5 secondary ampere . Finding The last maintenance for the AFW pump #21 motor breaker was August 25, 1984. A preventive maintenance (PM) was scheduled for September 25, 1986 which was not done. Both AFW pump #21 and #23 motor breakers were scheduled for PM the week of October 27, 1986. At that time the trip set-tings will be rechecked. The time delay trip time setting will also be checked. The licensee's change of the trip setpoint appears to be acceptabl ..
.
The inspector also reviewed the modification which installed the Amptectors and the basis for the overcurrent trip set-points for other safety-related motors. That review identi-fied concerns with the adequacy of other overcurrent trip setpoints as described in Section 4.4 of this repor The root cause of the AFW pump #21 breaker trip could not be determined with certainty, although most evidence indicated faulty operation of the Amptector overcurrent detection device. During discussions with the Instrumen-tation and Control Engineer, it was agreed that the suspect Amptector would be returned to the manufacturer for analysis to determine if it caused the trip of AFW pump #21 breake The inspectors also observed the recalibration of both AFW pump #21 and #23 Amptectors performed via PT-R46. The inspectors noted some confusion concerning the desired setpoint of the Amptectors since the data sheet specified a tolerance band vice a specifically desired setpoint. The inspector discussed this with the licensee who agreed that more specific instruction would be incorporated into the procedure before its next scheduled use. This item is unresolved pending revision of PT-R46 (247/86-28-06).
6.4 No. 22 Steam-Driven Auxiliary Boiler Feed Pump Pressure Control Valve (PCV) 1139 After the reactor trip on October 20, 1986 it was noticed that MS-52 was lifting. Upon investigation it was determined that PCV 1139 cycling was causing MS-52 to lift. The current PCV 1139 was installed per Modification #85-30720 during the 1986 refueling outag The former PCV 1139 exhibited excessive seat leakage. PCV 1139 is supposed to maintain downstream steam pressure at 600 psi The MS-52 relief valve located downstream of PCV 1139 protects the turbine for the steam driven auxiliary feedwater pump (AFW pump #22)
- from overpressure. Upon investigating the reason for PCV 1139 not l maintaining downstream steam pressure properly, the post modification
! test PMT 6407, was reviewed. The PMT cycled PCV 1139 using the hand-
,
wheel and also automatically using air. The valve welds were also l leak checked. PT-V6, Auxiliary Boiler Feed Pump Turbine Manual
'
Overspeed Trip, was performed on March 12, 1986 as part of PMT 640 This :est was performed with the pump coupled to the turbine. For
,
this test the mechanical overspeed trip lever is actuated. This test l was performed satisfactorily, although in the comments portion of the test it was noted that PCV 1139 did not control steam pressure. The valve would either fully open or fully close, l
l i
a
.
PT-V8, Auxiliary Boiler Feed Pump Turbine Mechanical Overspeed Trip Test was performed as part of PMT 6407 on March 14, 1986. For this test, the pump was uncoupled from the turbine. The turbine is brought up to speed and the setting at which the turbine trips or overspeed is recorded. This test was performed satisfactorily. The PMT did not test the controlling aspect of the valve, although the PMT was considered to be satisfactor Remarks written on the PMT indicate that the valve did not control pressure satisfactoril On October 25, during troubleshooting of PCV 1139, the valve stuck when the valve internals bound the valve. The controller for the valve was also replaced at this time. After repairs were made, PMT 29112 and PMT 68 were performe PMT 29112 cycled the valre using the handwheel and automatically with ai PMT 68 verified that PCV 1139 controlled pressure properly by preventing MS-52 from lifting, preventing the steam-driven AFW pump from tripping on overspeed, and the flow and pressure requirements of the pump were met. It took several attempts to set the PCV 1139 controller properly so MS-52 would not lift and the pump would not trip on overspeed. The tests were also performed with AFW pump #22 speed controller in various positions. The test was run satisfactorily with the PCV 1139 controller set to control pressure at 525 psig with the speed controller set at 0% ope PT-R22A, Steam Driven Auxiliary Feedwater Pump System Test, was performed after PMT-68. PT-R22A performs the full functional test of AFW pump #22. The test was completed satisfactorily. Although the pump trips that were observed during testing were due to tripping of the overspeed trip lever, observations of the pump during these tests indicate that the pump did not overspeed. The licensee has committed to evaluate the functioning of the overspeed trip devic PT-Q34 is the quarterly AFW pump #22 functional test. During the performance of this test on May 30, 1986, it was noted in the comments section that PCV 1139 needed to be adjusted. Deficiency tag #86125 was also installed on the valv From the time PCV-1139 was installed and considered operable during the 1986 refueling outage until the events in October, 1986, the controlling aspect of PCV-1139 was not tested satisfactorily. This is'a violation and was described previously in Section .5 AFW Pump #22 Steam Relief Valve-MS-52 After the October 20, 1986 reactor trip, MS-52 was discovered to be lifting due to PCV 1139 as described above. This valve had also lifted following a unit trip on September 16 and an engineering study was planned to be completed by December 15. After the October 23 reactor and AFW pump #22 trips, it was determined through in place testing of the valve that the lift setpoint was 665 psig. The licensee attempted to reset the Crosby valve to 700 psig, it's normal setpoint, but was unsuccessful. Adjustments to the PCV 1139 controller prevented the further lifting of MS-52 as determined by i PMT 68 and PT-R22A as described above.
i
!
- .. - .- -- __._._, , --- ,-
.
.
The inspectors also reviewed the test history of valve MS-52. . It was bench tested in October, 1982, under Work Order 82-5097 and Test Procedure PTQ-35-1410. The relief lifted at 690 lb. and had no leakage. It was bench tested again in August, 1984, under Work Order 84-14525 and Test Procedure PT-R35. The relief lifted at 695 lb, and had no leakage. The valve has since been removed from the test program. The licensee agreed to include a review of the decision not to test this valve in their AFW system assessmen .6 System Test After AFW Speed Changer Repositioned After the October 20, 1986 reactor trip and with the above-described problems, the operations department placed the AFW pump #22 speed changer at the 20% open position in order to prevent the MS-52 from lifting. The licensee believed that MS-52 was lifting due to the pressure spike caused by PCV 1139 opening and filling the short pipe between PCV 1139 and the governor valve before the PCV 1139 control-ler could react to the increased steam pressure. The licensee believed that, with the governor valve set at 20's open, this would act as a steam relief path through the AFW turbine and would prevent MS-52 from lifting. Shortly after this was accomplished, a test was performed in which PCV 1139 was opened and MS-52 did not lift. This test, however, was not a formal test and was not done by a written and reviewed test procedure. This is the second example of a viola-tion for not properly. testing AFW pump #22. (247/86-28-02). The other instance is described in Section 4.1 of this report. During the manual reactor trip on October 23, 1986, MS-52 lifted and AFW pump #22 tripped on overspeed with the governor set at 20%. Tests were performed on October 25-26, 1986 with the governor at various position It was determined that with the speed changer set at 0%
and with PCV 1139 open, AFW pump #22 would spin at approximately 3300 rp This fact was not known prior to performing these tests. The speed changer is currently positioned at 0%.
6.7 Tests Following November 15 Failure of AFW Pump to Auto Start During the controlled plant shutdown on November 14-15, 1986, due to one inoperable SI pump, the licensee shut #22 main boiler feedwater (MBF) pump discharge valve in order to trip the #22 MBF pump. Upon doing so, #22 MBF pump tripped, but #21 AFW pump did not automati-cally start as designed on a trip of any MEF pump. #23 AFW pump did j start automatically as designed. The licensee started #21 AFW pump
! from the CCR manually. After the plant was shut down, the licensee
- set up the same #22 MBF pump status as existea when the problem first
! .ppeared. The operators shut the #22 MBF pump discharge valve, l causing the #22 MBF pump to trip, but again #21 AFW pump did not l start automatically. Upon investigation, it was determined that a l loose wire on a terminal block in a CCR panel prevented this auto-l matic actuation of #21 AFW pump. The wire was tightened and PMT l 29528 was successfully performed by automatically starting the AFW pump.
i i
l E
___ . _.__ _ _
, _
,
6.8 No. 23 AFW Pump Vibrations The licensee performs quarterly surveillance test PT-Q27, Motor-Driven AFW Pump Functional Test to determine pump operability. The July 22, 1986 performance of PT-Q27 indicated higher than normal vibrations acting horizontally on the #23 AFW pump motor. Although the motor vibrations were high, 4.0 mils and 3.6 mils at the inboard and outboard bearings respectively, the pump itself met the oper-ability criteria. The operability criteria per ASME Section XI for action is greater than 1.5 mils measured horizontally on the pum The Section XI vibration measured valve for the #23 pump was .36 mil During the next performance of PT-Q27 on OctoLcr 22, 1986, the inboard and outboard motor bearing vibrations for #23 AFW pump were 7.4 mils and 3.6 mils, respectively. The Section XI vibration criteria for the pump was measured as .19 mils with the specification .
for action of greater than 1.5 mils. Due to these vibrations, the licensee decided to change out the motor bearings. After the change-out, the inboard and outboard motor bearing vibration levels were mils and 1.6 mils, respectively, which were still considered hig The licensee then changed out the motor with a Reliance motor. The previous motor was a Westinghous PT-Q27 was performed again as PMT 29238 and the results were satisfactor . Licensed Operator Recertifications An inspection was conducted to verify facility compliance with 10 CFR 55.31 (e), which requires a licensed operator, who has not been actively performing the functions of an operator or senior operator for a period of four months or longer, to demonstrate to the Commission.that his understanding of facility operations is satisfactory prior to resuming active licensed duties. The inspection was a follow-up to items identified in inspection reports 50-247/86-15 and 50-247/86-26. The inspection involved a review of one year of Central Control Room Logs from October, 1985 through September, 1986, a check of training records, and discussions with plant personne As a result of the review, seven persons were identified who sigred onte the logs- for a position requiring a license and had not signed onto the logs in the previous four months. Four individuals' resumption of licensed duties do not constitute a violation of 10 CFR 55.31 (e) in that: one individual had been issued a Senior Reactor Operator license three months prior to signing onto the logs, two individuals had been assigned to other duties during an outage which allowed the individuals to actively function in positions associated with the daily operation of the facility, and one individual had been acting as the operations manage ,
One individual returned to licensed duties after being assigned as the Support Facility Supervisor. The Support Facility Supervisor, who is in charge of the Unit 1 facility, is not considered to be actively performing the functions of a senior operator as defined in 10 CFR 55.4. No certifi-cation for resumption of licensed duties was submitted to the Commission prior to this individual resuming licensed dutie One individual returned to licensed duties on April 18, 1986, after being )
assigned to the Training Department. His assignment did not include teaching plant systems, performing plant walk throughs or instructing on the simulator. On June 18, 1986 a certification was submitted requesting approval for this individual to resume licensed duties. The certification was approved on June 24, 1986, two months after the individual had resumed licensed duties. The approval of the request to resume licensed duties was based in part on completion of a written and oral examination. The written and oral examinations were completed on June 17, 198 The facility was unable to identify what position the seventh individual held for the four months prior to his resuming licensed dutie The licensee has not adequately screened licensed operators prior to their resuming licensed duties after extended absences, and has not submitted proper certification to the Commission. This is a violation of 10 CFR 55.31 (e) (247/86-28-07).
The inspector also reviewed selected individuals' training records to verify the training and experience specified on some 10 CFR 55.31 (e)
certifications and on some selected renewal applications. The training records supported the information supplied on the certifications and application . Unresolved Items Unresolved items are those for which further information is required to determine whether the item is acceptable or a violation. Unresolved items are discussed in Paragraphs 4.4 and . Exit Interview At periodic intervals during the course of the inspection, meetings were held with senior facility management to discuss the inspection scope and findings. An exit interview was held with licensee management at the end of the reporting period. The licensee did not identify any 2.790 materia l l
l
__ _. _ _ _ . - , . _ . . _ __ __ _ . , _ . ,