ML20071K688

Special Rept:On 940519,pilot Operated Relief Valve Opened for Approx 5 Seconds During Functional Testing of Replacement Reactor Trip Module in Reactor Protection Sys

ML20071K688
Person / Time
Site:	Davis Besse
Issue date:	07/27/1994
From:	Stetz J CENTERIOR ENERGY
To:	NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
References
2336, NUDOCS 9408010290
Download: ML20071K688 (3)

Text

. \

CENTERIOR ENERGY 0200 0 9 Ne Bowe<cro W Adress incependence OH P O 6'e 9400; 76 4473100 Cweed OH 44:0: 4601 Docket Number 50-346 License Number NPF-3 Serial Number 2336 July 27, 1994 United States Nuclear Regulatory Commission Document Control Desk Vashington, D. C. 20555

Subject:

Events Surrounding the Pilot Operated Relief Valve Actuation on May 19, 1994 Gentiemen:

On May 19, 1994, at Davis Besse Nuclear Power Station (DBNPS), the Pilot Operated Relief Valve (PORV) opened for approximately 5 seconds.

The PORV opened during functional testing of a replacement Reactor Trip Module (RTM) in the Reactor Protection System (RPS). Although the PORV 10 not part of the RPS, a pressure signal to the PORV actuation control circuit originates from an RPS pressure sensor.

This event was reported in the Monthly Operating Report for May 1994.

The purpose of this letter is to provide a status of investigation of this event and to provide additional information since this event was of regulatory interest.

On May 19, 1994, the surveillance test Channel Functional Test of Reactor Trip Breaker "B", RPS Channel 1 Reactor Trip Module Logic, and Anticipatory Reactor Trip System Channel 1 Output Logic (DB-MI-03011) was being performed to meet Technical Specification requirements. At approximately 1000 hours0.0116 days <br />0.278 hours <br />0.00165 weeks <br />3.805e-4 months <br />, RPS Channel 1 was manually tripped as required by the test procedure. Status lights on RPS channel 2 did not indicate the expected RPS Channel 1 trip. Investigation showed that a set of contacts in the RPS Channel 1 RTM had not opened when the associated relay coil was de-energized. Operations declared RPS channel 1 inoperable at 1008 hours0.0117 days <br />0.28 hours <br />0.00167 weeks <br />3.83544e-4 months <br /> and performed required actions in accordance with Technical Specification 3.3.1.1. A Maintenance Vork Order (MV0) was implemented to replace the malfunctioning RPS channel 1 RTH vith a spare RTH from varehouse stock.

9408010290 940727 PDR ADOCK 05000346 I S PDR O

Operotag Compomes Chweiond t mctne litummahng <

1 v

I0'ed0 Id504

Docket Number 50-346

. License Number NPF-3 l

Serial Number 2336 Page 2 The spare RTM drawn from stock was originally purchased from Bailey Controls by the Sacramento Municipal Utility District (SMUD) for the Rancho Seco RPS system. Three RTMs along with a number of other ,

Bailey module types were purchased from SMUD. SMUD originally bought I and maintained the modules as safety related under their 10 CFR 50, Appendix B Ouality Assurance program. However, the modules were purchased by Toledo Edison as commercial grade to simplify the procurement process. The dedication process at DBNPS for the RTMs purchased included verification of part number, visual inspection and i vork bench functional testing at the module level. This was completed acceptably and the RTMs were placed in varchouse stock on December 10, 1993.

The replacement RTM vas installed and post maintenance functional testing was initiated in accordance with Channel Functional Test of RPS Channel 1 (DB-MI-03001). At approximately 1755 hours0.0203 days <br />0.488 hours <br />0.0029 weeks <br />6.677775e-4 months <br /> with reactor power at approximately 99 percent, during the course of testing, the Power Range Test Module Switch was placed in the Test Operate position and the RPS Direct Current (DC) Power Supply Breaker subsequently opened. During this time, RPS Channel 1 vas selected to supply RCS pressure signal input to the Non-Nuclear Instrumentation (NNI). When power was lost, the pressure signal spiked high, then failed lov and the PORV opened for approximately 5 seconds. This caused a slight (approximately 35 psi) pressure decrease in the RCS. Following the PORV closure, operators placed the pressurizer heaters in manual, entered the Pressurizer System Abnormal Operation procedure (DB-0P-02513) and verified the PORV closed. NNI pressure and flow inputs from the RPS vere svapped to RPS Channel 2 and at 1845 hours0.0214 days <br />0.513 hours <br />0.00305 weeks <br />7.020225e-4 months <br /> DB-0P-02513 was exited.

After this occurrence, the original failed RTM was repaired by Toledo Edison personnel by replacing the Main Trip Relay Circuit Board (PC2).

• When the repaired RTH vas reinstalled in RPS Channel 1 for testing, a malfunction was revealed in the channel bypass circuitry in this module. Status lights on the RTM indicated the channel was bypassed even though the channel bypass key switch was not in the bypass position. A replacement Channel Bypass Circuit Board (PC5) was not '

I available in spare parts stock but was obtained from another spare RTH and installed in the original RTM. Post maintenance functional testing, followed by testing to satisfy Technical Specification Surveillance requirements was completed at approximately 1035 hours0.012 days <br />0.288 hours <br />0.00171 weeks <br />3.938175e-4 months <br /> on May 20, 1994. RPS Channel 1 was declared operable at approximately 1120 hours0.013 days <br />0.311 hours <br />0.00185 weeks <br />4.2616e-4 months <br />.

The reason for actuation of the PORV vas traced to the RTH that was obtained from stock that had been purchased from SHUD. Review of the maintenance history for this module revealed that it was a spare module at Rancho Seco and had never been installed in the plant.

Detailed visual inspection of the PC2 board from this module revealed a difference in configuration when compared to a PC2 board from a Toledo Edison RTM. Review of the modification history of the PC2 board revealed that a change to the board in 1971, prior to Rancho Seco commencing operation, was not impicmented on this spare module.  ;

1 Docket Number 50-346

, License Number NPF-3 Serial Number 2336

- Page 3 "The other two RTMs purchased from SMUD had the modification implemented. The purpose of the module modification was to alleviate a cross-bridging condition in the relay by relocating a resistor so that it functioned in a current limiting mode should cross bridging occur. Cross bridging causes a momentary shorting of the -15 volt power supply to ground. Since the modification had not been made, actuation of the relay forced an excessive current draw from the DC -

power supply, causing the power supply breaker to open and the resultant opening of the PORV.

Investigations as a result of this event have been conducted on the modification history of the SMUD modules purchased by Toledo Edison and other Babcock and Vilcox Owners Group (BV0G) utilities. The BV0G Instrument and Control Committee representatives vere notified of the circinnstances of the DBNPS event. The BV0G Instrument and Control Committee reviewed all Preliminary Safety Concerns (PSC) from the original design of the RPS until the time of the purchase from SMUD.

This cross-bridging was not considered a PSC when the modification was +

implemented in 1971 because de-energization of a power supply is fail-safeanpthereforenotasafetyconcern. Further investigation has been requested by the BV0G that vill provide complete design change information from B&V to each affected BV0G utility for the RPS modules. This vill provide detailed information from which the modules that were spare at SMUD can be reverifi3d as the current  !

design. Enhanced inspection of the SMUD modules based on this l information should prevent further recurrence.

In this event, the RPS logic was not required to actuate and did not j actuate. The PORV actuation control circuit receives a pressure signal from the RPS pressure sensor but is not part of the RPS. The I event was determined to not be reportable under the criteria of 10 CFR f 50.73. This event was also not considered to be reportable under the criteria of 10 CFR, Part 21. The failure mechanism that was corrected by the design change vould not have created a substantial safety hazard because a deenergized RPS channel by design would go to a tripped state. l 1

Should you have any questions or require additional information, I please contact Mr. Villiam T. O'Connor, Manager - Regulatory Affairs, f at (419) 249-2366.

Very truly yours,

&bf John P. Stetz Vice President - Nuclear Davis-Besse Nuclear Power Station DLH/ eld cc: J. B. Martin, Regional Administrator, NRC Region III S. Stasek, DB-1 NRC Senior Resident Inspector R. J. Stransky, NRC Project Manager Utility Radiological Safety Board

. _ _