Headquarters Daily report OCTOBER 20, 1994 *************************************************************************** REPORT NEGATIVE NO INPUT ATTACHED INPUT RECEIVED RECEIVED HEADQUARTERS X REGION I X REGION II X REGION III X REGION IV X *************************************************************************** PRIORITY ATTENTION REQUIRED MORNING REPORT - REGION I OCTOBER 20, 1994 Licensee/Facility: Notification: Northeast Utilities MR Number: 1-94-0116 Millstone 1 Date: 10/17/94 Waterford,Connecticut RI Dockets: 50-245 BWR/GE-3 Subject: EMERGENCY POWER SUPPLY (GAS TURBINE) START LOGIC Reportable Event Number: 27912 Discussion: During a review of an inadvertent Emergency Core Cooling System (ECCS) actuation, the licensee determined that the subsequent gas turbine trip was a result of the momentary start signal that did not seal-in. When the ECCS initiation start signal was removed prior to the completion of the start cycle, the gas turbine shifted from the emergency to the normal start sequence resulting in a "flame out" condition and a gas turbine trip. Operator action was required to restore the gas turbine to a standby condition. The licensee determined that the logic seal-in feature was inadvertently removed during a 1989 governor modification. The cause of the design change and retest errors are under review. The current gas turbine start logic configuration does not meet the IEEE 279 requirement that a protection system shall be designed such that once initiated, the protective action at the system level shall go to completion. The licensee determined that the gas turbine is operable as configured based on the ability of the gas turbine to respond to all currently analyzed transients. For a loss of offsite power concurrent with an event, the bus undervoltage start signal to the gas turbine, which does seal-in, provides redundant gas turbine initiation logic to assure the accident analyses are unaffected. However, the gas turbine availability is potentially vulnerable in a loss of coolant accident with a subsequent loss of normal power (generic issue addressed in NRC IN 93-17). Specifically, a transient which results in a momentary ECCS initiation signal during the gas turbine start cycle (first 48 seconds from initiation signal) and subsequently followed by a loss of normal power would result in the gas turbine "flame out" condition. The licensee is performing risk assessment reviews to determine if a plausible transient exists that would result in this condition and the relative severity of the outcome. Further remedial and/or corrective action will be based on the quantified risk assessed. Regional Action: Routine resident inspector follow-up. Contact: Larry Nicholson (610)337-5128 Paul Swetland (203)447-3179 PRIORITY ATTENTION REQUIRED MORNING REPORT - REGION I OCTOBER 20, 1994 Licensee/Facility: Notification: Philadelphia Electric Co. MR Number: 1-94-0117 Peach Bottom 2 Date: 10/20/94 Philadelphia,Pennsylvania SRI PC Dockets: 50-277 BWR/GE-4 Subject: Reactor Head Temperature Greater Than 212 Degree F During Natural Circulation Reportable Event Number: N/A Discussion: Control room operators, at Peach Bottom Unit 2, allowed reactor vessel head flange temperatures to exceed 212 F for about two hours late on October 16 and early October 17, 1994. The unit was nearing the end of a refueling outage. The reactor pressure test had been completed, coolant temperature as indicated on the vessel drain line was 193 F, the head vents were open, the reactor water clean up (RWCU) system was in operation and the B recirculation pump was operating. At 8:35 p.m. the operators removed the B recirculation pump (RCP) from service, placing the plant into a natural circulation mode of core cooling. The B RCP was secured to limit heat addition from pump operation. Reactor operators monitored reactor head flange temperatures and bottom head temperatures every 15 minutes while the pump was secured. The crew was attempting to start the A recirculation pump for a test run while flushing the B shutdown cooling loop. During this time, bottom head temperatures continued to decrease (152 F) and reactor head flange temperatures remained constant (205 F). After the A RCP failed to start, reactor flange temperature were increasing at a rate of 2 F per hour. The operator logging the temperatures did not notice this increase since bottom head temperatures continued to decrease. During shift turnover at 11:00 p.m., reactor flange temperature had reached 210 F and was increasing about 4 F every 15 minutes. The B loop of shutdown cooling was available to be placed in service at shift turnover, however, the operators did not recognize an immediate priority to start the system based on bottom head temperatures (128 F). At 12:30 a.m., the reactor operator noted a 1 psig pressure increase and realized that reactor head temperature was above 212 F. The control room supervisor directed that shutdown cooling be started immediately. When shutdown cooling was started at 12:50 a.m., reactor flange temperatures had reached 230 F and bulk average temperature was noted to be 205 F. The inspectors noted that the control room operators did not control the cooldown evolution well. The control room supervisor did not establish priorities or controls to remain focused on the plant cooldown. Operators may have been distracted by other activities that prevented them from recognizing the increasing flange temperatures and the effects of stratification during natural circulation operations. Further, the operators did not follow the guidance in the cooldown procedure which emphasized stratification concerns. PECO Energy has taken actions to investigate this incident. PECO committed to perform a root cause analysis and establish interim corrective actions prior to performing the reactor start-up from the tenth refueling outage. Reactor startup is currently scheduled for October 20, 1994. Regional Action: The resident staff continues to monitoring licensee activities. A Special Inspection is being conducted, by the resident staff augmented by Region I personnel, to review this event and conduct 24-hour continuous monitoring during the reactor start-up. Contact: Clifford Anderson (610)337-5227 Wayne Schmidt (315)342-4041 PRIORITY ATTENTION REQUIRED MORNING REPORT - REGION I OCTOBER 20, 1994 Licensee/Facility: Notification: Philadelphia Electric Co. MR Number: 1-94-0118 Limerick 2 Date: 10/20/94 Philadelphia,Pennsylvania SRI PC Dockets: 50-353 BWR/GE-4 Subject: Reactor Scram on Low Level Due to Electrical Switching Error Reportable Event Number: N/A Discussion: On October 19, 1994, at 12:19 p.m., Unit 2 scrammed automatically from approximately 92 percent of rated power due to reactor vessel low level. The plant was coasting down for its third refueling outage, scheduled for January 1995. The low level condition was caused by the opening of all three feedwater pump minimum flow valves, which diverted water from the reactor vessel when control power was lost, due to the unintentional deenergization of safeguards bus D24. The bus became deenergized when an operator, who was securing the emergency diesel generator (EDG) after its monthly surveillance test, erroneously opened the normal off-site supply breaker to the bus followed by opening the EDG output breaker, thereby deenergizing and isolating the bus. The operator had intended to open the EDG output breaker and stop the EDG; the three switches are located vertically, with the normal supply breaker at the top and the EDG control switch at the bottom. Manually tripping the breakers removed the auto-close feature for the alternate supply to the bus. When the bus became deenergized, control power for the minimum flow valve control panel was lost, causing the minimum flow valves to fully open. A relay failure prevented the automatic switchover of the valve control power to its alternate supply, safeguards bus D23. Other problems encountered during and after the scram include: the B feedwater pump failed to automatically trip when required, the A recirculation pump motor/generator drive motor breaker tripped and should not have, an IRM failed to indicate full in, and a reactor water clean-up (RWCU) backwash inlet valve experienced electrical problems. Plant management is currently pursuing the cause of the problems and corrective actions; all work has been prioritized for the startup. Additionally, a few steam leaks are being repaired while the turbine is shut down. Regional Action: The resident inspectors responded to the control shortly after the scram, observed the recovery, attended the debriefs, and attended the status meetings. The residents will continue to monitor the startup and corrective actions for the event. Contact: Gene Barber (717)542-2134 Neil Perry (610)327-1344 PRIORITY ATTENTION REQUIRED MORNING REPORT - REGION III OCTOBER 20, 1994 Licensee/Facility: Notification: Commonwealth Edison Co. MR Number: 3-94-0188 La Salle 2 Date: 10/20/94 Marseilles,Illinois SRI VIA TELECON Dockets: 50-374 BWR/GE-5 Subject: UNIT 2 EHC TRANSIENT AND SUBSEQUENT REACTOR SCRAM Reportable Event Number: 27926 Discussion: On October 19, Unit 2 scrammed following a turbine trip on low EHC hydraulic pressure. The EHC transient occurred during adjustment to the 30V power supply for EHC. The regulated and the house power supplies were believed to be "fighting" and adjustments were being made to the regulated supply. Prior to the transient, control valve servo current oscillations were observed and there were reports of EHC pipe movement and an EHC fluid leak in the EHC pump room. Operators were shifting the running EHC pump in an attempt to stop the EHC leak when the turbine trip occurred. After the trip, a 3/4" EHC line was found to have sheared at the EHC skid, which was the cause of the low hydraulic pressure. The high flux reactor scram signal was generated due to the pressure increase when the turbine control valves closed and turbine bypass was not available due to the loss of EHC. Six SRVs opened to control pressure and reclosed without problems. Control room level channels indicated level dropped to +6 inches. Operator response to the event appeared appropriate. Anomalies during the event: RCIC auto-started and fed the vessel, Reactor Recirc pumps tripped, and one channel of ARI actuated. All of these actuations should occur at -50 inches. (HPCS also actuates at -50 inches, but it did not start, indicating that there was some anomaly with the instrumentation.) During the fast transfer of house loads to offsite power, 2 of 3 running circulating water pumps tripped. Regional Action: The resident and regional inspectors responded to the scram and are following the licensee's investigation. Contact: H.B. CLAYTON (708)829-9602 PRIORITY ATTENTION REQUIRED MORNING REPORT - REGION III OCTOBER 20, 1994 Licensee/Facility: Notification: U.S. Steel Works MR Number: 3-94-0189 U.S. Steel Works Date: 10/19/94 Gary,Indiana TELECON ST. OF IN ON 10/19/20 Subject: RADIATION ALARM ON INCOMING SHIPMENT OF SCRAP METAL Reportable Event Number: N/A Discussion: Region III was informed by a representative of the Indiana State Department of Health that a truck hauling a dumpster of scrap metal caused the radiation monitoring system to alarm at U.S. Steel Works on October 19, 1994. Representatives from the State responded to the call and found two metal rods in the dumpster measuring a maximum of 1.5 mR/hr and 0.5 mR/hr respectively, near the surface of the devices using a sodium iodide detector. The rods were approximately 21 inches long and 1 inch in diameter. Using a portable gamma spectrometer, the State determined that the isotope in question was cobalt-60. There was no indication of removable contamination on the rods. The dumpster was transported to U.S. Steel Works by Ace Supply and Iron Company of Joliet, Illinois. According to an Ace representative, the dumpster contained some metal scrap from Ace Supply and Iron Company, however, most of the scrap originated at Blackford Corporation in West Chicago, IL. U.S. Steel Works rejected the load and arranged for transport of the dumpster to Ace Supply and Iron Company. The Illinois Department of Nuclear Safety was notified of the findings on October 19, 1994 and will have an inspector available to monitor all the contents of the dumpster when it arrives in Joliet on October 20. Regional Action: Region III will inform NMSS and State Programs of the incident. Contact: B.J. HOLT (708)829-9836 PRIORITY ATTENTION REQUIRED MORNING REPORT - REGION IV OCTOBER 20, 1994 Licensee/Facility: Notification: Pacific Gas & Electric Co. MR Number: 4-94-0125 Diablo Canyon 2 Date: 10/19/94 Avila Beach,California Resident Inspector Dockets: 50-323 PWR/W-4-LP Subject: LOSS OF RHR COOLING DURING LOAD SHED TESTING OF DIESEL GENERATOR BUS Reportable Event Number: N/A Discussion: On October 18, 1994, at 11:55 p.m., a loss of residual heat removal (RHR) occurred on Unit 2 during Diesel Generator 2-2 auto-start load transfer testing. RHR Pump 2-2 was running at the start of Diesel Generator 2-2 testing powered from 4 KV Bus H. Power to 4 KV Bus H is normally fed from an auxiliary transformer; however, during the test, Diesel Generator 2-2 is aligned to supply 4 KV to Bus H. The diesel generator testing being performed required opening the 4 KV Bus H auxiliary feeder breaker, measuring the time for the Diesel Generator 2-2 to auto-start and align to the 4 KV bus, and subsequent verification of the auto-start of required loads. When the test was performed, RHR Pump 2-2 was in service despite several precautions in the surveillance test procedure requiring the operator to review the effect of the testing on RHR pump operation. Additionally, a specific procedural step required securing RHR Pump 2-2 prior to commencement of the test since the loss of power would trip the pump. RHR Pump 2-2 was not shut down per this procedural step. During the test, the RHR pump lost power when the auxiliary feeder breaker was opened. Per design, the RHR pump did not auto-start after Diesel Generator 2-2 auto-started and aligned to supply power to the 4 KV bus. Approximately 6 minutes after the opening of the auxiliary feeder breaker, during a control board walkdown, the control operator involved with the testing noted that neither of the RHR pumps were running. At that point, an RHR pump was started to restore core cooling. During the 6 minutes that RHR was secured, core temperature increased from 96 degrees F to 102 degress F. The licensee made a 4-hour nonemergency report pursuant to 10 CFR 50.72(B)(2)(III). The licensee is evaluating the procedures and circumstances involved with this incident. Regional Action: The resident inspector is following the licensee's actions and will keep the Region advised of any further developments. Contact: G. Johnston (510)975-0304 M. Tschiltz (805)595-2354 D. Corporandy (510)975-0319