05000331/LER-2002-001
| Docket Number | |
| Event date: | |
|---|---|
| Report date: | |
| Reporting criterion: | 10 CFR 50.73(a)(2)(i)(B), Prohibited by Technical Specifications |
| 3312002001R00 - NRC Website | |
I. Description of Event:
The startup package was complete and preparations were being made to start the plant after a planned maintenance outage. On 3/8/02, At 0053, as part of securing Shutdown Cooling (SDC) and re-aligning to LPCI mode, the operator closed M02004, the "A" LPCI outboard inject valve, and within a few seconds secured the "C" RHR pump. Coolant temperature at this time was 155 degrees Fahrenheit.
Right after securing the RHR pump the Control Room received a low pressure header alarm. Since LPCI is required to be operable in Mode 3 and routine heat up from Cold Shutdown (Mode 4) conditions was in progress, the OSS declared the "A" LPCI subsystem inoperable due to a prolonged low- pressure condition. The required action for this condition is to fill and vent the system. At 0150, following the filling and venting, the "A" LPCI subsystem was declared operable. The delay in restoring LPCI to standby readiness condition due to depressurization consumed some of the available time to boil and is considered a contributing factor to this event.
After the filling/venting process, the next step was to open V19-0048. The operator opening V19-0048 felt binding in the valve and stopped with the valve slightly open. He then discovered that one of the four cap screws that secure the valve operator cover was missing and two others were loose. Two mechanics were dispatched to investigate. Upon removal of the cover and stem protector it was discovered that the yoke sleeve nut had come loose and had dropped onto the stem. The nut resting on the stem prevented the stem from traveling through the cover and the stem protector, thereby preventing the valve from opening. After installing the yoke stem nut, the mechanics told the Control Room that the valve could be opened but another half hour would be needed to complete needed repairs. At this time (about 0240), the Operations Shift Manager decided instead to shut M 9- 0048 and restore SDC operation. The cover was then reinstalled and the valve was closed.
At the time the decision was made to return to shut down cooling the time to boil was 36 minutes.
Without the ability to open the crosstie, LPCI could no longer be considered operable. Realignment of "A" RHR for SDC commenced. This required re-establishing the correct flow path for pump start, restarting RHRSW, and some SDC tagging activities. The "A" SDC was established at 0315. The reactor coolant bulk temperature exceeded 212 degrees Fahrenheit at 0313 and Mode 3 was entered for a period of approximately 6 minutes.
Subsequently, Mechanical Maintenance personnel removed the cover, then removed, smoothed, and reinstalled the nut. Later that day SDC mode of RHR was exited; V19-0048 successfully opened and plant startup resumed.
II. Cause of Event:
Failure of yoke sleeve nut setscrew to prevent the nut from loosening was the root cause of this event.
A lack of preventive maintenance was a potential cause but based on discussions with the vendor, no maintenance other than what was historically done was recommended. Additionally, poor design was considered as a potential cause (lack of securing the set screw) but it was concluded that the original design requirements were followed.
II. Cause of Event (continued) :
Technical Specifications permit entry into Mode 3 while still in SDC if RHR can be realigned to LPCI mode. Depressurization of RHR while securing from SDC delayed discovery of the V19-0048 problem.
Had this not occurred, a longer time to boil would have been available to restore SDC. This is therefore considered a contributing factor.
III. Assessment of Safety Consequences :
This event had no safety consequences. Although SDC was not re-established in time to prevent the mode change, it was started in a deliberate, controlled manner in accordance with procedures. Since both Core Spray subsystems were operable, no loss of safety function occurred. Forced circulation through the core was maintained by continuous operation of the running recirculation pump throughout the event. The failure of V19-0048 at no time prevented the restart of shutdown cooling. If V19-0048 could not have been closed, the motor-operated crosstie, MO-2010, could have been closed instead.
A review of similar valve designs was conducted and revealed twenty-three bevel gear manual operator valves similar to V19-0048. All of these valves are used as pressure boundary in safety-related applications. These valves are operated only for maintenance or surveillance testing. Only V19-0048 is routinely realigned as part of normal system operation. Failure of any of these valves has a potential to cause or extend an LCO, however, none has an active function to realign.
IV. Corrective Actions:
1. Mechanical Maintenance personnel removed the cover of V19-0048, then removed, smoothed, and reinstalled the nut. The nut setscrew was installed with Locktite to prevent loosening. The gears were lubricated and the cover was reinstalled with 1/4"-20 x 1" capscrews (vs. the 1/4"-20 x 7/8" removed) for thread engagement into the gearbox. (WO#A57740, completed 3/8/2002) 2. Evaluation of a more positive method of preventing nut rotation as an alternative to the setscrew for this type of valve will be performed. (AR#30532, due to be completed by 5/10/2002) 3. Follow-up maintenance will be performed as necessary on similar valves to prevent recurrence.
The maintenance will include removal of the operator cover and internal inspection that includes the security of the nut and its set screw and lubrication of the gears. Security of the valve hand wheel and proper engagement of the cover bolts should also be verified. This will be accomplished during normal, scheduled system windows and during the next refuel outage.
(AR#30494, due to be completed by 5/31/2003) 4. Procedural guidance is to be developed for securing SDC and preventing depressurization of the RHR system during plant startup. (AR#30106, due to be completed by 6/14/2002)
V. Additional Information:
V-19-0048 Maintenance History:
A review of maintenance history revealed that the last work performed on the operating mechanism of V19-0048 was in May 1985 when the stem was replaced. The valve was disassembled to correct a seat leakage problem in July 1985. In both of these cases the operator would have been disassembled. Neither of these work orders contains any instructions regarding the setscrew or any mention of problems with the setscrew. The Anchor Darling vendor manual for the valve and operator contains no mention of the locking device for the yoke stem nut in either the disassembly or assembly instructions. The setscrew does appear on the vendor drawing. One other related work order was the replacement of two socket head cap screws that hold the stem housing in place.
Anchor Darling has employed a number of methods to prevent loosening of the yoke sleeve nut. The setscrew as in V19-0048 is an early design that was later changed to staking the yoke sleeve nut to the yoke sleeve. Later still, the nut and bevel gear were drilled and a roll pin installed. They were not aware of cases in which the setscrew design actually failed to retain the yoke sleeve nut.
EMS System and Component Codes:
Low Pressure Coolant Injection System: BO V19-0048 RHR Crosstie Valve, Anchor Darling Valve Company (A391), Model 1889-3 This report is being submitted pursuant to 10CFR50.73(a)(2)(i)(B).