Headquarters Daily Report OCTOBER 02, 1997 *************************************************************************** REPORT NEGATIVE NO INPUT ATTACHED INPUT RECEIVED RECEIVED HEADQUARTERS û REGION I û REGION II û REGION III û REGION IV û PRIORITY ATTENTION REQUIRED MORNING REPORT - REGION I OCTOBER 2, 1997 Licensee/Facility: Notification: Peco Energy Company. MR Number: 1-97-0053 Limerick 2 Date: 10/02/97 Philadelphia,Pennsylvania SRI PC Dockets: 50-353 BWR/GE-4 Subject: FAILURE OF A CONTROL ROD TO SCRAM USING THE SCRAM TEST SWITCHES Discussion: On September 25, 1997, during reactor protection system (RPS) functional testing, control rod 34-03 drifted when the RPS "B" scram solenoid pilot valves (SSPV) were de-energized. The control rod was originally at position 48, drifted less than one notch for a duration of 6 seconds and then settled back to position 48. RPS testing was stopped and an inspection of the associated hydraulic control unit (HCU) revealed a significant air leak on one of the "A" solenoid end caps. A plan was developed to repair the leaking SSPV end cap. During implementation of the repair plan, control rod 34-03 failed to scram using the test toggle switches. Control rod 34-03 was subsequently fully inserted by manually driving the control rod in, with no anomalies noted. The scram inlet and outlet valves functioned normally when air to the HCU was isolated and vented in preparation for testing. The SSPV assembly for control rod 34-03 was replaced and subsequent scram time testing was satisfactory. The licensee disassembled the failed SSPV and determined that all 4 diaphragms were hardened with some cracking noted. However, no evidence was found to indicate diaphragm chipping or plugged valve ports. As a follow-up action, scram time testing was performed on 18 Unit 2 additional control rods, 30%, of the 60 rods with SSPV diaphragms of the same type and age (BUNA-N in service for 4.5 years). The sample included all 7 SSPVs also found to be leaking. All additional rods tested, scrammed; however, two control rods with leaking SSPVs were found to be slow, marginally exceeding the average time specification to position 45. The licensee plans to evaluate industry information on leaking SSPVs and failures of control rods to scram, to determine if the Limerick failure is consistent with industry experience. to identify similar SSPV failures. The licensee plans to replace four of the Unit 2 SSPVs on October 2, 1997. Two of these are on the slow rods with leaking SSPVs. The SSPVs will be sent to GE for analysis. The licensee has implemented weekly monitoring of SSPVs for air leakage on Unit 1 and 2. The results of this monitoring will be used to initiate increased frequency scram testing. The schedule for replacement of the remaining Unit 2 SSPV Buna-N diaphragms is the week of November 8, 1997. The licensee postulated the failure of the control rod to scram was related to diaphragm hardening and the air leak. However, the failure mechanisms are under review. The failure was not predicted by previous surveillance testing and no anomalous reasons were identified suggesting a different reason for the failure. REGION I MORNING REPORT PAGE 2 OCTOBER 2, 1997 MR Number: 1-97-0053 (cont.) Regional Action: The resident inspector is following PECO Energy's failure analysis and evaluation efforts. Contact: Clifford Anderson (610)337-5227 Arthur Burritt (610)327-1344 _