IR 05000275/1991038
| ML16341G364 | |
| Person / Time | |
|---|---|
| Site: | Diablo Canyon  |
| Issue date: | 10/25/1991 |
| From: | Morrill P NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION V) |
| To: | |
| Shared Package | |
| ML16341G363 | List: |
| References | |
| 50-275-91-38-MM, NUDOCS 9111260201 | |
| Download: ML16341G364 (40) | |
Text
'
U.
S.
NUCLEAR REGULATORY COMMISSION
REGION V
Report Nos.
50-275/91-38 and 50-323/91-38 Docket Nos.
50-275 and 50-323 License Nos.
Pacific Gas and Electric Company 77 Beale Street, Room 1451 San Francisco, California 94106 Facility Name:
Diablo Canyon Units 1 and
Meeting at:
Region V Office, Walnut Creek, California Report Prepared by:
B. J. Olson, Project Inspector Approved by orr~
,
ie React r Projects Section
a igne Meetin on October ll 1991 (Re ort Nos.
50-275/91-38 and 50-323/91-38)
A meeting was held in the Region V Office, Walnut Creek, California to discuss the licensee's assessment of recent problems with Unit 1 valve FCV-95 and Unit 2 valve CVCS-8378A.
9111260201 911106 PDR ADOCg 05000275 PDR
'
DETAILS l.
Neetin Attendees a.
Licensee Attendees W. Fujimoto, Vice President, Nuclear Technical Services R. Anderson, Manager, Nuclear Engineering and Construction Services B. Giffin, Manager, Maintenance Services K. Herman, Group Supervisor, Instrumentation and Controls Engineering B. Powers, Director, Mechanical Maintenance T. Grebel, Supervisor, Regulatory Compliance E. Connell, Assistant Project Engineer R. Cupp, Group Leader, Instrumentation and Controls Engineering K. Dalai, Group Leader, Piping and Valve Engineering N. Davido, Senior Engineer, Valve Engineering b.
NRC Attendees R.
Zimmerman, Director, Division of Reactor Safety and Projects S. Richards, Chief, Reactor Projects Branch D. Kirsch, Chief, React0r Safety Branch P. Norrill, Chief, Reactor Projects Section
H. Wong, Senior Resident Inspector P. Narbut, Reactor Inspector D. Corporandy, Reactor Inspector L. Tran, Reactor Inspector H. Freeman, Reactor Inspector B. Olson, Project Inspector 2.
Details Nr. Fujimoto opened the meeting by indicating that PGSE was prepared to discuss problems with Unit 1 valve FCV-95 and Unit 2 valve CVCS-8378A.
He added that the root causes of the valve problems had not yet been identified, but he did feel that the work to identify the root causes had been methodical and timely.
Nr. Fujimoto also indicated that he had been quite involved with work pertaining to the valve problems, and that the actions taken provide reasonable assurance that the valves are operable.
He stated that senior management is setting the tone to resolve problems.
Nr. Fujimoto turned the discussion over to Nr. Cupp.
Mr. Cupp provided a description and the failure history of FCV-95.
The valve is a four inch, flexible wedge, gate valve manufactured by Velan.
The valve is operated by a
DC motor and opens to admit steam to the turbine driven auxiliary feedwater pump.
FCV-95 automatically opens on low-low steam generator level or after a station blackout.
Until 1991, FCV-95 in Unit 1 failed to stroke open five times and opened successfully over two hundred times.
FCV-95 in Unit 2 had never failed to stroke open.
The licensee had previously concluded that FCV-95 in Unit 1 failed to open because of thermal
binding of the flexible wedge to the valve body. It was explained that the body of FCV-95 Onit I was made of carbon steel and the flexible wedge had been made of stainless steel.
The body and the flexible wedge of FCV-95 in Unit 2 were both made of carbon steel.
As a result of the thermal binding concern, the stainless steel wedge in the Unit I valve was replaced with a carbon steel wedge in Narch 1991.
Problems with FCV-95 were then thought to be resolved.
On August 23, 1991, FCV-95 in Unit I failed to open during a surveillance test.
Nr. Cupp explained that the DC motor tripped on thermal overload while trying to unseat the flexible wedge.
Nr. Cupp described how the surveillance test placed the system in a different configuration than would exist if the valve were to be required to open automatically.
The licensee formed a team to investigate the valve's failure to open.
Nr.
Zimmerman asked if the team was formed as a result of NRC interest in the matter.
Mr. Fujimoto replied that the team was forming shortly after the test failure and was not formed as a result of NRC interest.
Nr.
Cupp reviewed some of the actions subsequent to the test failure.
Diagnostic equipment was connected, and FCV-95 was stroked a number of times, each successfully.
Electrical engineering reviewed the operating characteristics of the motor and electrical supply and concluded that the motor and thermal overloads were operating properly when the valve failed to open.
The valve manufacturer and other licensees were contacted, and an industry search of similar-operating failures was being conducted.
After questioning by Nr. Zimmerman, Nr. Fujimoto indicated that PGSE would specifically ask the vendor if this type of operating failure had been reported before.
Nr. Fujimoto also indicated that the conditions and the results of the successful tests subsequent to the failure would be provided to the HRC.
Mr. Cupp then reviewed some of the potential failure mechanisms.
One of the mechanisms described was bonnet pressurization.
When FCY-95 is shut in a normal configuration, there is an approximate differential pressure across the flexible wedge of 800 psi.
The differential pressure results in the upstream side of the wedge unseating, allowing steam into the bonnet.
Prior to the surveillance test, the steam supply to FCV-95 was isolated.
After the steam was isolated, the upstream side of the wedge could have reseated and trapped pressure in the bonnet.
Mr. Cupp explained that calculations indicated that the motor would have to develop an additional 5,000 to 7,000 pounds of thrust to overcome the trapped pressure in the bonnet.
While this amount of additional thrust was not enough to trip the motor, bonnet pressurization might have contributed to the valve's failure to open.
Mr. Cupp added that FCY-95 in Unit 2 had a slight leak prior to the current outage, and this leak was considered to have been enough to prevent bonnet pressurization.
While refurbishing the valve during the current outage, a hole was drilled in the upstream side of the wedge of the Unit 2 valve to ensure that any bonnet pressure would be released.
Nr. Cupp indicated that additional instrumentation will be added to the valves in both Units in a effort to confirm if bonnet pressurization is occurring and to determine any difference in the operation of the valves between Units.
Additionally, FCY-95 in Unit I is being tested once a week to ensure operability.
The valve is considered operable as the failure on
August 23, 1991, occurred with the system in a non-standard configuration, and the valve has always operated when called upon to open automatically.
Also, all subsequent valve operations have been successful.
Mr. Zimmerman indicated that PGIIE's approach to the problem appears to be methodical with a broad perspective.
He added that had this approach been used during the past failures of the valve to operate, the problem may have been solved.
Mr. Fujimoto replied that Nr. Zimmerman's comments were valid because of the most recent failure.
He added that PGIIE believed that the problem had been isolated to thermal binding and had been solved, but this was not the case.
Finally, Nr. Zimmerman cautioned the licensee not to view other equipment problems narrowly and not to observe a number of repeat fai lures before a team is organized to resolve the prob'1em.
Nr. Powers described the problem with valve CVCS-8378A installed in Unit 2.
The valve is a three inch diameter check valve, manufactured by Velan, and is the first off reactor coolant system check valve in the alternate charging line.
The valve developed a body to bonnet gasket leak and was opened for repair on August 23, 1991, during Unit 2's refueling outage.
The valve was found to be stuck open by a mechanic and an engineer.
An attempt was made to shut the valve by hand, but the attempt was unsuccessful.
Due to area radiation levels, the valve bonnet was reinstalled, and the mechanic and engineer exited the area.
On August 18, 1991, after installation of temporary lead shielding, valve CVCS-8378A was reopened and found to be shut.
Mr. Powers explained that Velan had previously submitted a report to the HRC as required by 10 CFR Part 21.
The Part 21 report indicated that two and one-half, three, and four inch check valves may jam in the open position because an accumulation of tolerances may allow the valve's bushings to rotate and prevent the valve from shutting.
Mhile PGIKE expected the tolerance problem to have resulted in valve CVCS-8378A sticking open, inspection did not indicate this to be the cause.
Nr.
Powers also indicated that the valve internals appeared to be in good condition, but the valve moved with a slight resistance.
After the valve internals were wiped down, there was no resistance to movement.
The valve internals were reinstalled in the valve, and a flow rate of 95 gallons per minute was pumped through the valve for 30 minutes.
A subsequent radiograph indicated no apparent problems with valve operation.
Nr. Zimmerman asked what the safety function of the valve was and how often the valve was used to pass flow.
He was told that the valve would shut for an upstream charging line leak, but even if the valve had been stuck open, there were other valves that could have been shut had a leak developed.
Mr. Powers told Nr. Zirrmerman that valve CVCS-8378A is normally only used during a quarterly surveillance test, but it had been used prior to the refueling outage as a result of a leak that had developed in the normal charging line.
Nr. Powers also stated that PGIIE plans to radiograph the valve during the next refueling outage.
Mr.
Zimmerman indicated that the NRC would follow the licensee's long term actions to monitor the valv The meeting concluded after Nr. Connell described the inspection effort conducted at Diablo Canyon as a result of the Velan
CFR Part
report.
Nr. Connell indicated that 26 Unit 1 and 19 Unit 2 valves had been inspected.
One check valve in the feedwater system of Unit 1 had been found with the tolerance problem that could have caused the valve to stick open.
The internals of the feedwater check valve were replaced even though the valve was not found in the open position.
Nr. Zimmerman thanked the PG&E personnel for their presentation and the meeting was adjourne Recent Unit
FCV-95 Failure HPF
INTRODUCTION
~ Overview
~ Recent Event Chronology
~ Previous Events
~ Updated Industry Experience
~ Potential Failure Mechanisms Additional Monitoring
~
Ongoing HIT Activities
.
N0.2 MAIN STEAM LEAD MAIN STEAM ISOLATION VALVE EXHAUST TO ATMOSPHERE FCV-37 M
I"5l66 FCV-95-FCV"l.52 l"5l67 FCV-38 M
STM STM STM TRP TRP TRP TERRY TURBINE N0.3 MAIN STEAM LEAD MAIN. STEAM ISOLATION VALVE TUR8 I NK STEAM SUPPLY
OVERlllEW
~ 9/26/91 FCV-95 Stroke Failure During Surveillance Testing
~
5 Previous Failures
~ Engineering and Plant HIT Team Formed Previous FCV-95 Experience Updated Industry Experience Special FCV-95 Diagnostics Evaluating Potential Failure Mechanisms
~
No Specific Failure Mechanism Identified
~ FCV-95 Has Not Failed During Monthly and Quarterly AFW Functional Testing for Operability
PREVIOUS FAILURES CORRECTI VE ACTIONS
~
Issuance of Operations Gu'idance to Close Valve "Hot" (5/24/90)
'
Unit
Disk Material Change From Stainless Steel to Carbon Steel (3/91)
IDENTIFICATION OF POTENTIAL FAILURE MECHANISMS VALVE ISSUES HIGH PACKING LOADS THERMAL BINDING BONNET PRESSURATION SEAT RING STEM GROWTH FCV-9S Fals to Open ACTUATOR ISSUES GEAR BOX DEFICIENCY TORQUE SWITCH ADEQUATE POWER SETTING INTERMITTENT OPERATION VALVE SPECIFIC DIFFERENCES WEDGING VAIVE l ACTUATOR MISALIGNMENT ELECTRICAL ISSUES THERMAL OVERLOAD PERFORMANCE MOTOR PERFORMANCE SYSTEM ISSUES CORRECT P AND FLOW IPING WOUCED STRESS
NORMAL VALVE OPENING CHARACTERISTICS Thrust MOV Motor Trip 14,000 Pounds Packing Friction Disk Unseating Stem Unbalance
{With Flow Only}
Total "Normal Force" 600 1,200 Pounds 1,000 3,000 Pounds 1,000 Pounds 1,600 5,200 Pounds
ACTUATOR/ELECTRICAL POTENTIAL FAILURE MECHANISMS EVALUATED
~
Thermal Overload Operating Characteristics
~
Improper Motor Operating Characteristics
~
Gear Box Deficiency
~
Torque Switch Function
~
'
SYSTEM POTENTIAL FAILURE MECHANlSMS EVALUATED
~ System Pressure
~ Elow
~ Differential Pressure Across Valve
~ Torque Switch Setting
~ Piping Induced Stresses
VALVEPOTENTIAL FAILURE MECHANISMS EVALUATED
~
Disk Microwelding
~
High Packing Loads
~
Disk Thermal Binding
~.Stem Growth Thermal Binding
~
Valve Specific Differences
~
Bonnet Pressurization
~
Wedging
ADDITIONAL MONITORING DURING SURVEILLANCE TESTS
~
Performed on Both Units
~
Strain Gages, Thermocouples, and Pressure Transducers will be Installed on the Valve and Piping
~
Data will be Recorded During Planned Surveillance Tests
~
Data Reduction will Assist in Confirmation of Root and Contributing Causes
HIT TEAM ONGOING ACTIVITIES
~ Industry Experience Other Utilities Vendors Other
~ In-Depth Review of All Activities Prior to Identified Failures
~ Diagnostic Testing Results
~
Finalize Analytical Evaluations
~ Complete HIT Team Evaluations and Recommendations
Alternate Charging First Off RCS Check Valve CVCS-2-8378A Stuck Open me11
'
SHING CIRCE 3C58 l -ROE t -CNER 5 - AXING'%OVER ST%)
4 -KKMX~h%R t4fl 5 - ~~AL6ASKET 6 -OSC 7 - HANGER 8 - BURY 9 -HANGER PN 6-W BaT I -UX:K lÃAKET Q -DISC l4JT PN HANGER HNCXET lS -ROLL PN l6 SEAT t
lt
e-nsc moec sam 7-l4M4KR mcxnmare 4-SUSH848 MHNC3ER bRJCKET o-~~8m
~aca~m
~o-m(max'ncscaasvaa 1 &LOCKMNKET