IR 05000129/2003022
| ML18101A630 | |
| Person / Time | |
|---|---|
| Site: | Salem, 05000129  |
| Issue date: | 04/07/1995 |
| From: | Jason White NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION I) |
| To: | |
| Shared Package | |
| ML18101A629 | List: |
| References | |
| 50-272-95-02, 50-272-95-2, 50-311-95-02, 50-311-95-2, NUDOCS 9504170013 | |
| Download: ML18101A630 (18) | |
Text
Report No License No Licensee:
Facility:
Dates:
Inspectors:
Approved:
U. S. NUCLEAR REGULATORY COMMISSION
REGION I
50-272/95-02 50-311/95-02 DPR-70 DPR-75 Public Service Electric and Gas Company P.O. Box 236 Hancocks Bridge, New Jersey 08038 Salem Nuclear Generating Station January 29, 1995 - March 22, 1995 Inspection Summary:
~-
Date This inspection report documents inspections to assure public health and safety during day and back shift hours of station activities, including:
operations, radiological controls, maintenance, surveillances, security, engineering, technical support, safety assessment and quality verification. A violation involving measures to assure configuration control, and several examples of deficient corrective action were identified in this perio Amplification is contained in the Executive Summary.
9504170013 950407 PDR ADOCK 05000272 Q
EXECUTIVE SUMMARY Salem Inspection Reports 50-272/95-02; 50-311/95-02 January 29, 1995 - March 22, 1995 OPERATIONS
{Module 71707) The licensee performed a safe post-refueling reactor startup for Salem Unit Subsequently, operations performed a safe shutdown of the Salem units to effect repair of inoperable solid state protection system power supplie The inspectors noted increased management awareness and involvement in the daily operation of the Salem units. Notwithstanding, the inspector identified two missed opportunities for consideration of the risk associated with the concurrent work on safety-related equipmen On February 24, 1995, Salem Unit 1 operators failed to comply with the Technical Specification 3.4.3 action statement requirements for an inoperable Power Operated Relief Valve (PORV).
The inspectors noted that the matter.was repetitive relative to a previous situation on March 24, 1994, when oper~tors also failed to comply with Technical Specification requirements for an inoperable Salem Unit 2 POR Consequently, corrective action effectiveness is considered a continuing.weakness. This most recent finding constitutes an apparent violation of 10 CFR 50, Appendix B requirements pertaining to correction action.
MAINTENANCE and SURVEILLANCE
{Modules 61726, 62703) Operations, Maintenance, Planning, and Radiation Protection demonstrated good coordination, thorough attention to detail, and excellent radiation worker practices in completing the No. 21 reactor coolant pump maintenance activitie A PSE&G team, assembled to identify the cause of control problems with main steam atmospheric relief valves (MSlOs}, performed a very thorough root cause analysis. The inspectors noted, however, that the team's efforts occurred only after a long history of continuing and repetitive problems with MSlO control at the Salem unit Inspectors determined that the licensee adequately and conservatively implemented the Technical Specification 4.6.1.5 requirement for surveillance of containment air temperature ENGINEERING and TECHNICAL SUPPORT (Modules 37551, 71707, 92903) Reactor engineering performed a well documented, technically sound, comprehensive, and timely evaluation of a Unit 1 radial flux tilt. Reactor engineering's technical expertise and support of operations contributed to safe plant performanc Engineering could not determine a root cause for recurring spurious alarms and test faults affecting the Safeguard Equipment Control system; however, engineering did aggressively pursue suspected electromagnetic interference ii
-*-
(EMI) or electrical noise as a possible cause, and is continuing to actively monitor and diagnose system performanc The inspectors identified several minor scaffolding discrepancies with no safety significance, but noted that a potential exists to adversely affect safety through ineffective control of scaffoldin Inspectors identified that failure to ensure pressurizer safety valve loop seal drains were properly installed and functioned effectively is an apparent violation of the IOCFRSO, Appendix B, Criterion V, which specifies that activities affecting safety are properly accomplishe Continuing weakness with corrective action effectiveness was identified relative to control of materials installed in safety-related reactor head vent valves, and failure to establish corrective measures in response to mis-operation of head vent 2RC4 These items were identified as examples of an apparent violation of lOCFRSO, Appendix B, Criterion XVI, "Corrective Action".
The inspector determined that the licensee had taken appropriate measures to address reactor head leakage in response to Generic Letter 88-0 PLANT SUPPORT (Module 71707, 71750)
Licensee Event Report 94-15 identified that operators failed to collect grab samples of the Waste Gas Decay Tank as required by Technical Specifications when the gaseous effluent monitoring instrumentation was determined to be out of service. Though, technically a violation of regulatory requirements, enforcement discretion was applied in accordance with NRC policy; and the item is considered as a non-cited violatio The licensee acted properly relative to Emergency Classification notification in response to an Unusual Event due to low river water level that had the potential to affect circulating water and service water system SELF-ASSESSMENT AND QUALITY VERIFICATION Notwithstanding the violations identified in this report, PSE&G generally operated the Salem units safely. However, the continued manifestation of recurring equipment problems and ineffective corrective action indicate that PSE&G has not yet achieved any significant improvement in overall performanc Th~ licensee's failure to implement measures (procedures, directions, or drawings) to assure proper configuration following the safety relief valve loop seal modification during 2R7 demonstrates*weakness in work control The response to the recurring problems with the main steam atmospheric relief valves (MS-lOs) demonstrated recurrence of previously documented weak initial root cause investigatio As has also been the case in the past, in response to NRC questions and management recognition of the impact of the MS-10 problems, the licensee commissioned a multi-disciplinary team to perform a comprehensive engineering investigation of the cause of continued MSlO reliability problem While such effort is considered a positive step in attempting to resolve this long-standing issue, the licensee supposed that this item was previously resolved as a result of troubleshooting and engineering efforts following the April 7, 1994 tri i i i
EXECUTIVE SUMMARY.
SUMMARY O~ OPERATIONS Unit 1 began the period operating at 100% powe On February 3, the licensee initiated a shutdown to comply with plant Technical Specification requirements for inoperable solid state protection systems (SSPS).
On February 4, the licensee entered Mode 5 (Cold Shutdown).
On February 15, after completing modifications to SSPS, operators entered Mode 4 (Hot Shutdown).
The licensee maintained the unit in Mode 4 while resolving problems encountered with main steam atmospheric relief valves (MS-lOs).
On February 27, operators commenced a reactor startup, and on March 2, they increased power to 48%.
On March 3, operators reduced power to 28% for a bioshield entry to adjust RCP oil level On March 8, operators increased power to 100% and maintained the unit there until March 18, when operators reduced power to 70% in response to marsh grass fires with the potential to effect offsite power lines. Operators returned the unit to 100% power on March 1 The unit continued to operate at 100%
power until March 20, when power was reduced to 94%, due to a leak in the N llC feedwater heate Power remained at 94% through the end of the inspection perio Unit 2 began the period in Mode 3 (Hot Standby).
On February 1, the licensee commenced and safely completed a reactor startu On February 3, the licensee commenced a Technical Specification required shutdown from 1% power, for inoperable SSPS power supplies. The licensee placed the unit in Mode 5, completed modifications to SSPS, and commenced a plant startu On February 11, operators took the reactor critical and commenced power escalatio On February 19, the licensee decided to shutdown the unit (then at 47% power) to remove No. 21 reactor coolant pump (RCP) from service due to low seal water leakoff flo The licensee placed the plant in Mode 5, replaced the No. 1 seal, and commenced a plant startu On March 8, operators achieved reactor criticality and commenced power escalation. The unit operated at 90% power until March 18, when operators reduced power to 80% in response to marsh grass fires with the potential to affect offsite power lines. During the power reduction, operators encountered problems with manual control of No. 22 main feedwater pump, and further reduced power to 50% for repair of the pump controls. Operators returned the unit to 97% power on March 2 On March 22, operators reduced Unit 2 to 50% power in response to speed oscillations in the No. 21 main feedwater pum.0 OPERATIONS The inspectors verified that Public Service Electric and Gas (PSE&G) generally operated the facilities safely and in conformance with regulatory requirements except as noted in the details of this report. The inspectors evaluated PSE&G's management control by direct observation of activities, tours of the facilities, interviews and discussions with personnel, independent
verification of safety system status and Technical Specification compliance, and review of facility recoras. The inspectors performed normal and back-shift inspections, including 46 hours5.324074e-4 days <br />0.0128 hours <br />7.60582e-5 weeks <br />1.7503e-5 months <br /> of deep back-shift inspection.1 Solid State Protection System Forced Shutdown -
At 10:30 p.m. on February 1, 1995, the licensee declared the solid state protection system (SSPS) for both units inoperable, since a steam line break in the turbine building could cause the complete loss of the SSP Operators entered Technical Specification 3.0.3 for both units and began a shutdown of Unit 1 at 15 percent per hour. Operators maintained Unit 2 in Mode 2 (Startup) with the main steam isolation valves (MSIVs) shu The licensee determined that a steam line break could cause wiring to the turbine stop valves, the auto-stop oil pressure switches, and the reactor coolant pump (RCP) breaker position indication to short circuit. The short circuit would de-energize the 15 volt and 48 volt SSPS power supplies causing a reactor trip, and rendering both SSPS channels incapable of generating an automatic safety injection actuatio At 2:30 a.m. on February 2, the NRC granted the licensee enforcement discretion for 96 hours0.00111 days <br />0.0267 hours <br />1.587302e-4 weeks <br />3.6528e-5 months <br /> to allow the licensee to modify the SSP Operators terminated the shutdown on Unit 1 and returned power to 100%.
Salem planned to correct the inadequate electrical separatio In addition, the licensee established compensatory measures, including operator training on manual actuation of safety injection, a moratorium on engineered safety feature (ESF)
equipment maintenance until modification completion, maintaining Unit 1 in steady state conditions, and maintaining Unit 2 in Mode 2 with the MSIVs close At 5:22 a.m. on February 3, technicians de-energized a 15 volt power supply for Unit 1 SSPS train "A" to perform planned modifications. However, the redundant train "A" power supply simultaneously tripped and caused the loss of SSPS train "A".
The licensee stopped the modification and restored train "A" to norma The licensee could not identify the root cause failure of the 15 volt power supply within the technical specification (TS) allowed outage tim At 11:00 a.m. on February 3, the licensee initiated a Unit 1 shutdown from 100% power to comply with TS 3.3.1.1. At 4:30 p.m. on February 3, the NRC rescinded the enforcement discretion for both units based upon the complications imposed by SSPS power supply failure At 10:30 p.m. on February 4, operators placed Unit 1 in Mode 5 (Cold Shutdown).
At 4:31 a.~.
on February 5, operators placed Unit 2 in Mode The inspector observed that operations performed a safe shutdown of Salem Unit Refer to NRC Inspection Report 50-272 and 50-311/95-03 for additional assessment of assessment of the SSPS modification and associated activitie.2 Risk Management During the inspection period, the inspector identified two missed licensee opportunities to consider the risk associated with the concurrent performance of work on multiple pieces of safety-related equipmen Specifically, on
March 7, 1995, Un-it 1 operators commenced a turbine driven auxiliary feedwater (TDAFW) pump surveillance prior to returning No. 12 residual heat removal (RHR) pump to servic The No. 12 RHR pump was out of service for preventative maintenance and was returned to service shortly after operators began the TDAFW pump surveillanc On March 9, 1995, operators authorized sandblasting immediately adjacent to No. 21 service water pump while No. 23 service water pump tagged out for maintenanc The inspector considered that the sandblasting and associated scaffolding had the potential to adversely affect No. 21 service water pump operability due to close proximity to interferences and grit at a time when No. 23 service water pump was not available for us The inspector discussed the risk perspective with the operator As a result, the aggregate risk was re-considered and activities that had the potential to directly affect the operability of No. 21 service water pump were curtailed until No. 23 service water pump was returne.3 Failure to Comply with Technical Specification Action Statement At 8:58 p.m. on February 24, with Salem Unit 1 in mode 3, an operator placed pressurizer Power Operated Relief Valve (PORV) 1PR2 in manual as required by procedure Sl.IC-CC.RC-0082, IPC-455K Pressurizer Pressure Contro Control room operators noted in the logs that they had entered Technical Specification 3.4.3, Action A, however, they did not close block valve 1PR7 within one hour.
Salem Unit 1 Technical Specification 3.4.3 Action A requires, in part, that in modes 1, 2 and 3, with one or more power PORVs inoperable, within one hour either restore the PORV to operable status or close the associated block valve At 7:10 p.m. on February 25, the Salem Unit 1 Nuclear Shift Supervisor realized that the PORV was in manual and immediately instructed the operator at the controls to close block valve 1PR The inspectors noted that the NRC previously identified a similar failure to adhere to Technical Specification 3.4.3 on March 25, 1994, involving Unit The corrective actions initiated for that instance appeared not to have be sufficient to prevent recurrence of this type of violation. Consequently, this latest example constitutes apparent non-conformance with the requirements of 10 CFR 50, Appendix B, Criterion XVI, which requires that measures be established to assure that corrective action is taken to preclude recurrence of significant conditions adverse to quality (EEI 50-272;311/95-02-01). MAINTENANCE AND SURVEILLANCE MAINTENANCE The inspectors observed portions of the following safety-related maintenance to verify that the licensee conducted the activities in accordance with approved procedures, Technical Specifications, and appropriate industrial codes and standards.
The inspector observed portions of the following activities:
Work Order{WO) or Design Unit Change Package CDCPl Description Salem 1 WO 950213241 12SJ134 - Safety Injection Pump to Cold Legs Valves Salem 1 WO 951126030 12 Service water pump strainer repair Salem 1 WO 950216203 Safety injection pump breaker repair Salem 2 WO 941226085 23 Charging pump The inspectors observed that the plant staff performed the maintenance effectively within the requirements of the station maintenance progra.1.1 Reactor Coolant Pump Seal Replacement On February 19, Salem Unit 2 control room operators shut the plant down from 47% power to remove No. 21 reactor coolant pump {RCP) from service due to low seal water leakoff flo Licensee attempts to identify and correct the low flow condition while in Mode 3 (Hot Standby) were unsuccessfu On February 22, operators placed the unit in Mode 5 (Cold Shutdown) to perform a RCP seal inspectio *
On February 23, the licensee placed the No. 21 RCP on the "backseat" to allow for seal inspection or replacement without operating at reduced RCS inventor Technicians found a small amount of debris in the No. 1 seal and an out of specification runout on the No. 1 runner retainer sleev On March 1, the licensee completed replacement of the No. 1 seal package and the retainer sleeve. At the end of the report period, the licensee had not completed an in-depth evaluation of the No. 1 seal and sleev The inspectors noted that plant staff safely performed the RCP "backseating" process. Operations, maintenance, planning, and radiation protection demonstrated good coordination, thorough attention to detail, and excellent radiation work practices in completing the No. 21 RCP maintenance activitie The inspector noted that No. 21 RCP low seal water flow was also a concern following the Salem Unit 2 seventh refueling outage in May 199 The licensee reduced pressure {< 1000 psig) to reseat/flush the seal at that tim Although seal return flow improved {> 1.0 gpm), operators noted low flow alarms in October 1993 (believed to be a spurious alarm coincident with component cooling water temperature manipulations), in April 1994 (instrumentation problem with corrective action scheduled for 2R8), and in October 1994 {related to component cooling water temperature variations).
The No. 21 RCP seal was originally scheduled for inspection during the recently completed refueling outage; however, emergent concerns with the No. 22 RCP seal assembly caused the licensee to redirect those resource The inspector
observed that while the licensee eagerly pursued the low seal water return flow problem when flow dropped below 1.0 gpm. the degraded condition existed since May 1993. Although the low seal flow ~id not present immediate safety concerns, the inspector concluded that the licensee did not perform timely or thorough root cause determination and corrective actio As a result, an unresolved equipment deficiency required operators to shut the plant down for seal repai.1.2 Main Steam Atmospheric Relief Valves On February 3, with atmospheric steam relief valve 13MS10 in automatic, a Salem Unit 1 operator attempted to close the valve by increasing the pressure setpoint of the controlle The valve responded erratically to the operator actions, then apparently shifted to manual control without operator action and failed open. After changing the pushbutton module, the operator was able to close the valv On February 11, a Salem Unit 2 operator found that the 22MS10 controller did not properly track pressure~ In addition, the controller did not properly demand valve opening in response to operator action On February 17, 13MS10 failed open agai The inspectors noted that, although the licensee was initially slow to assemble the resources to evaluate the cause of the MSlO valve control problems, they eventually mounted a multi-disciplinary team to examine the recent problems with 13MS10 and 22MS1 The team found maintenance, design, and refurbishment inadequacies caused the initial 13MS10 failure. These processes contributed to less than optimal calibrations, installation of incorrect parts (K3 relays), and inadequate preventive maintenance to preclude capacitor failure The team found that the 22MS10 failure resulted from calibration problems, and servo-station sensitivity that was not recognized or considered during modification The 13MS10 problem on February 17 resulted from low input impedance test equipment providing a parallel path for current flo The inspectors concluded that the team performed very thorough root cause analysis. The inspectors noted, however, that the team's efforts occurred only after a long history of problems with MSlO controller deficiencies at the Salem units, that had supposedly been corrected and resolved previousl The implication of these latest findings support previous assessments of weak and ineffective root cause and corrective action programs that continue to affect Salem operation.2 SURVEILLANCE The inspectors performed detailed technical procedure reviews, observed surveillances, and reviewed completed surveillance packages. The inspectors verified that plant staff did the surveillance tests in accordance with approved procedures, Technical Specifications and NRC regulations.
- The inspector reviewed the following surveillances:
Unit Procedure N. Test Salem 1 Sl.OP-PT.TRB-0001 Turbine Auto Trip Mechanisms Operational Test Salem 1 S 1. OP-ST. CS-0001 Inservice Testing - 11 Containment Spray Pump Salem 1 Sl.OP-ST.MS-0003 Steam Line Isolation and Response Time Testing Salem 1 Sl.OP-ST.RC-0008 Reactor Coolant System Water Inventory Balance Salem 1 SI. RE-ST. ZZ-0002 Shutdown Margin Calculation Salem 2 S2.IC-ST.SSP-0008 Solid State Protection S1stem Train A Functional Test The inspectors observed that plant staff did the surveillances safely, effectively proving operability of the associated system *
3.2.1 Containment Average Air Temperature
Each ~alem unit has 10 containment air temperature monitoring points. Salem Technical Specification (TS) 4.6.1.5 requires that the containment average air temperature be the arithmetical average of the temperatures at any five of the ten listed locations. The Salem plant computer averages all ten location The inspector also noted that one of the ten plant computer inputs for Unit 1 was taken at elevation 136' northeast vice elevation 78'-northeast as listed i n TS 4. 6. 1. 5.
The licensee noted that the Westinghouse Standard Technical Specification lists five measurement points for this surveillance, and requires that the containment average air temperature is determined from the arithmetic average of these five point The licensee concluded that the use of all ten measurement points, as opposed to any five of the ten, results in a more accurate determination of overall containment average air temperatur The licensee considered the input from Unit 1 elevation 136' a more conservative temperature than elevation 78'.
In 1992, licensing developed a proposal (never submitted to the NRC) to change TS 4.6.1.5 to require the arithmetic average of the temperatures from at least five of the ten locations. The proposal included removing the list of the monitoring locations from TS and placing it in the updated Final Safety Analysis Report (UFSAR).
The licensee never submitted the proposed change to the NRC.
The inspector reviewed the licensee's implementation of TS Surveillance 4.6.1.5 and determined that averaging the ten temperatures represented a more conservative indication of containment temperature than require Notwithstanding, the licensee initiated an Incident Report, changed the surveillance to require averaging five temperatures, and initiated a review to assess the need to change TS 4.6..0 ENGINEERING Salem Unit 1 Radial Flux Tilt Evaluation On March 4, 1995, Salem Unit 1 operators increased power to 48%.
Operators performed Sl.OP-ST.NIS-0002, Power Distribution - Quadrant Power Tilt Ratio, and calculated a quadrant power tilt ratio (QPTR} of approximately 1.0 The QPTR is the ratio of the maximum upper and lower excore detector calibrated output to the average of the upper and lower excore detector calibrated outputs. Technical Specification 3.2.4 requires QPTR to be less than 1.02 before proceeding above 50% powe Reactor engineering obtained a flux map and determined the QPTR, based on incore detectors, to be 0.8% (compared to the 3% tilt calculated from excore detectors).
Reactor engineering analyzed this flux map and determined that the flux tilt was not abnormal. Reactor engineering attributed flux tilts to the previous full power core conditions, and expected quadrant tilts to return to the previous hot full power values following power escalatio Reactor engineering recommended updating the nuclear instrumentation, calculating the QPTR once per hour, limiting the power increase to 3% per hour, and performing additional flux maps at 75% and 95% powe In addition, reactor engineering requested Westinghouse to perform a core design evaluation of the cycle 12 flux map On March 6, Westinghouse determined that the safety analysis limits were expected to be met during the planned power escalatio Westinghouse concurred with PSE&G's expectations that the incore tilt would return to previous full power value On March 8, the plant reached full power and operators noted a shift in the flux tilt back to previous full power values. Operations performed QPTR's hourly and consistently calculated flux tilts <1%.
The inspector noted reactor engineering's quality operations support and thorough engineering evaluation. The inspector noted that the reactor engineer evaluation was well detailed and timel The inspector concluded that, in this instance, reactor engineering technical expertise, and conservative support of operations contributed to safe plant performanc.2 Scaffolding in Safety-Related Areas During the inspection period, the inspector noted several examples of the licensee's failure to adequately control scaffolding in safety-related areas in accordance with NC.NA-AP.ZZ-0023, Scaffolding and Transient Loads Contro The inspector observed minor deficiencies regarding adequate clearances, proper restraints, variance inspections, and timely remova Safety-related areas included the service water bays and 4kv vital electrical bus roo fl1e
inspector noted the operating shifts' timely response in addressing the deficiencies. The inspector did not observe any discrepancies that dir~ctly impacted or threatened nuclear safety presentl.3 Safeguard Equipment Control {SEC) Troubleshooting The inspectors documented in previous inspection reports (see NRC Inspection Reports 50-272/94-3I and 94-35) recurring problems with Unit I SEC degraded power supplies and frequent automatic test insertion (ATI) test faults and spurious alarm The inspector noted I2 instances of ATI test faults on IA SEC and I on IB SE On February 22, I995, maintenance replaced the IA SEC 24 volt power supply to restore the IA SEC to operability. This same power supply was replaced on January 6, I99 System engineering, together with nuclear engineering, have remained focused in their pursuit of SEC problem resolution. While engineering could not determine a definitive root cause, it was hypothesized that the spurious alarms and test faults were caused in part, by electromagnetic interference (EMI) or electrical noise. Accordingly, the licensee contracted an EMI specialist in mid-February to investigate the frequent ATI test fault Engineering, supported by the EMI specialist, determined that EMI levels in the SEC cabinet, although high enough to cause ATI alarms, do not impact the ability of the SEC to perform its designed safety functio Engineering is actively pursuing the EMI specialist's recommendations to improve the immunity of the ATI to EMI and to prevent future spurious ATI alarm System engineering plans to implement on-line monitoring to evaluate the potential for electrical bus disturbances affecting the ATI test circui In addition, engineering is in the final stages of preparing a design change package (DCP) to replace the SEC power supplies, implement needed EMI improvements, and install an ATI resistor modificatio Engineering concluded, based upon Unit 2 SEC fault-free operating experience, that the new power supplies, good EMI practices, and the ATI card modification resolved the problem of erroneous All alarms previously experienced on Unit Engineering is also working closely with the SEC vendor to improve EMI immunit The inspector noted that engineering determined that although power supply AC ripple voltage was found to be higher (4mv) than the vendor's acceptance criteria (lmv), it was still far below the point that the vendor stated that it could impact SEC operability (2.4 volts) for a 24 volt power suppl The inspector determined that while SEC test faults persist at a frequent periodicity, engineering is actively engaged in monitoring and diagnosing system performanc.4 Pressurizer Code Safety Valve Loop Seals On October 19, 1994, during Salem Unit 2 refueling outage 2R8, maintenance technicians prepared to remove pressurizer safety valve (PSV) 2PR3 for routine testin When they separated the faces of the inlet flange they encountered water leaking from the pipe connecting the pressurizer to the valv In response to the leaking water, an equipment operator found va~ve 2PR66 close *
As a result, water had collected in the inlet piping, configured by original plant design, to permit formation of a loop seal containing approximately ten gallons of wate During the previous refueling outage, PSE&G installed a modification, Design Change Package (DCP) 2EC-3190, in response to NUREG 0737, Item II.D.1, to address the concern that operation of the PSVs (2PR3, 2PR4, and 2PR5) with a loop seal would result in stress levels in excess of Code allowable on the structural welds and discharge pipin The modification replaced the PSV internals with internals designed to operata in a steam only environmen In addition, the modification installed drain lines for the loop seals. The drain lines from each of the PSVs had individual isolation valves, and a common isolation valve (2PR66) in the common drain line heade The licensee concluded that Salem Unit 2 had operated from the end of refueling outage 2R7 (May 1993) until the beginning of refueling outage 2R8 with the loop seals not draine Engineering reviewed existing analyses and initially concluded that, although an analysis did not exist to cover operation of the PSVs under the exact conditions that existed between 2R7 and 2R8, analyses with sufficient similarity existed to provide reasonable assurance that the PSVs and their associated discharge piping would have performed their safety function if they had been challenged.. They concluded that demonstrating the valves had been operable required a detailed analysi Notwithstanding, more recent evaluation indicates that operation of the PSVs under design conditions would have resulted in loads on the discharge piping in excess of Code allowable. Nuclear engineering initiated the detailed analysis, scheduled for completion in April 1995, to determine the safety significance of this matter and whether operability would have been affecte Salem management tasked the operations staff with determining the root cause for mispositioning the valve. Operations scheduled the task for completton on March 26, 199 The inspectors, with the assistance of operations staff, determined that the modificati~n package (2EC-3190) contained a requirement to add the valves installed as part of the modification, including 2PR66, to the database used to control valve lineups. The modification required that operations place 2PR66 in the open position for normal operatio The inspector noted that 2PR66 was added to the database on May 4, 1993, and was assigned to lineup RC MECH 001 on May 18, 199 The inspector noted that the operations staff performed lineup RC MECH 001 on May 10, 199 The operations indicated that they typically controlled the lineup of components affected by modifications during a refueling outage through the use of auxiliary lineups. However, the operations staff could find no evidence that operators had performed an auxiliary lineup for 2PR66 but did find several examples of completed auxiliary lineups associated with other 2R7 modification The operations staff reviewed the work order that implemented modification 2EC-3190, and could find no evidence of a post-modification test to verify that the loop drain performed its intended functio The operations staff
also demonstrated that the database currently contained 2PR66 assigned to lineup RC MECH 001, and that operators had correctly positioned the valve at the conclusion of the 2R8 refueling outag The inspectors noted that the difficulty experienced with the PSVs leaking past the seat at the end of 2R8 further demonstrated that operators correctly positioned 2PR66 after 2R8, and the lack of seat leakage after 2R7 tended to support the conclusion that 2PR66 was erroneously left in a closed position for the entire cycle following the 2R7 outage, and not detected until after the completion of the 2R8 outag The inspectors concluded that PSE&G had operated Salem Unit 2 from at power between the end of refueling outage 2R7 in May 1993, and the beginning of refueling outage 2R8 in October 1994, with an unanalyzed configuration associated with the pressurizer safety valve The licensee is currently performing a detailed evaluation to determine if the valves were operable in that period in view of the different type of valve internals and other configuration changes that were installed to support removal of the loop seals. Failure to ensure that modification 2EC-3190 was satisfactorily completed is an apparent violation of 10 CFR 50, Appendix B, Criterion (EEi 50-272;311/95-02-02).
Associated with this apparent violation is the continuing weakness in the licensee's process and program for root cause and corrective action effectiveness as demonstrated by the fact that, while the licensee ensured that operators opened 2PR66 prior to startup after 2R8, there was no assessment made to determine the causes of the circumstances that led to the as found condition, no consideration of the potential for other component or configuration problems, or no determination of the adequacy of implementation and performance of post-modification/installation testing for other modifications completed during 2R. 5 Contro 1 of Reactor Head Vent Materia 1, Parts*, and Components During a review of information in the Managed Maintenance Information System (MMIS) data base, the inspector noted that MMIS listed the open and closed limit switches for position indication of the reactor vessel head vent valves as not safety-related, not environmentally qualified, not seismically qualified, and not requiring quality control. The inspector noted that 10 CFR 50.44 C (3) (iii) and the guidelines of NUREG 0737 Item 11.B.1 require operability of the reactor head vent valves and that the reactor head ven~
limit switches provide continuous positive valve indication during plant operatio In addition, the limit switches are required to be seismically and environmentally qualified in accordance with IEEE 344-197 The inspector found that on June 7, 1994, Procurement Engineering had initiated Discrepancy Evaluation Form (DEF) DES-94-0007 The importance of the need for qualified components was recognized in the DEF by indicating that two reactor head vent valves were on the same circuit; and consequently, a shorted limit switch would open the circuit breaker supplying both valves
rendering them both inoperable. Nothwithstanding, though the DEF acknowledged that the switches appeared to be improperly classified, there was no concern indicated relative to operability or safety significanc In response to the DEF, on June 16, 1994, the licensee initiated a Bill Of Materials (BOM} change to reflect the requirement that the limit switches be obtained seismically and environmentally qualified. It was not until February 1995 that the plant staff determined that non-qualified limit switches had been installed in Salem Unit 1 reactor head vent valves 1RC41 and 1RC4 All other head vent valve limit switches were found to be appropriately qualifie The licensee also issued a work order for I&C action to replace the limit switches during the next Salem Unit 1 refueling outage, presently scheduled for September 199 The licensee determined that the manufacturer's part number for the non-qualified limit switch was the same as the part number for the qualified limit switch, and determined that the limit switches installed in 1RC41 and 1RC43 were manufactured to the same standards as limit switches obtained as safety related parts; the exception being that qualified parts are certified by testing. Consequently, the licensee concluded that installation of the non-qualified limit switches constituted a loss of quality (as discussed in NRC Generic Letter 91-18}, but not a loss of operabilit The inspector reviewed an MMIS Bill of Material Validation report, documenting engineering assessment of Purchase Class 4 codes (commercial grade, non-safety related} assigned to safety related component Engineering initiated the assessment as a result of identification that, in August 1994~ a rheostat designated for use in safety-related Salem battery chargers had been obtained as a commercial grade componen The report documented that, in a review of approximately 500 Purchase Class 4 components, 76% were appropriately classified. The remaining 24% required further engineering evaluation to permit use in safety-related applications, or were inappropriately classifie The report further recommended that an additional 497 items should be reviewed in depth to determine if they are acceptable for use in safety-related applications. Engineering expected to begin the review in March and complete the review by June 30, 199 *
The inspector determined that relative to the DEF, the engineering staff did not have sufficient bases to conclude that no operability or safety concern existed, particularly since the DEF recognized that a short of a single limit switch could affect operability of two reactor head vent valve Further, the engineering staff apparently did not determine that non-qualified switches had been actually installed in 1RC41 and 1RC43 until February 199 Consequently, the inspector determined that the licensee did not take adequate or timely corrective action in response to the deficiency identified in the DE The inspector noted that a similar problem was identified the April 7, 1994, Unit 1 loss of circulating wate In that case, the NRC issued a Notice of Violatirin that addressed the installation of incorrect parts in PORVs 2PR1 and 2PR2, and installation of a summator module for high steam flow setpoint with incorrect identification and an incorrect electronic part. The corrective
action for those findings apparently was insufficient to assure critical evaluation and review of the potential for other instances of the installation of incorrect parts in safety related application Failure to take adequate corrective actions relative to the identification of non-qualified limit switches on the safety-related reactor head vent valves is an apparent violation of the requirements of 10 CFR 50, Appendix B, Criterion XVI, Corrective Action (EEi 50-272;311/95-02-03). Reactor Head Vent Valves The reactor head vent valves were installed as one of the action items of NUREG-0737, "Clarification of TMI Action Plan Requirements."
For each Salem unit, four solenoid-operated valves provide redundant flow paths for post-accident venting of hydrogen from the reactor vessel hea The valves are not used during power operation, however, they are part of the reactor coolant system (RCS) pressure boundar On July 6, 1994, Unit 2 was in cold shutdown when operators attempted to stroke reactor head vent valve 2RC4 When the valve's open position indication light did not illuminate, operators initiated a work reques During maintenance to repair the suspected indication problem, the,licensee discovered that the valve had actually failed to stroke open (through use of vendor supplied diagnostic equipment).
In a memorandum dated July 7, 1994, maintenance engineering informed operations that the "most probable" cause of the failure was boric acid that may have solidified around or in the valve's pilot plug. According to the memorandum, when RCS temperature was increased above 180°F, the valve could be opene Subsequent testing demonstrated that the valve stroked within specifications and would pass the appropriate flow for the given plant condition The memorandum recommended the valve be returned for normal use, and that the need for additional preventive maintenance should be evaluate There were, however, no recommendations for addressing the generic implications or impact of this finding on nuclear safet No discrepancy report was initiated for evaluation of this safety-related component failur The inspector reviewed incident reports from that period, and was not able to find documentation of 2RC40's failure to stroke ope Similarly, a licensee review of completed maintenance work packages from the 1994 summer outage found no documentation that specifically addressed the failure of 2RC40 to strok Salem Administrative Procedure NC.NA-AP.ZZ-0009(Q), "Work Control Process," requires workers to report incorrect operation of safety-related components to their supervisors. Supervisors are required to evaluate such problems against criteria for initiation of an incident report contained in Attachment 1 of NC.NA-AP.ZZ-0006(Q), "Incident Report/Reportable Event Program and Quality/Safety Concerns Reporting System."
The inspector determined that the failure of 2RC40 met the procedure's criteria (Attachment 1, "Examples of Off-normal Events," Item 8b) and the subsequent failure to enter the finding as an incident report circumvented an established quality deficiency reporting and corrective action syste As a
..
result, the systems engineering organization was not involved, and the safety impact, generic implications, and root cause were not adequately evaluate In addition, there was no review by enginel ing managemen Since July 1994, all four reactor head vent valves in Unit 2 have been replaced (due to leakage problems) and, therefore, the immediate safety significance of this issue is lo However, there has been no determination if this failure constitutes a potential common mode failure that may effect all reactor head vent valves over tim In addition, no assessment has been made to determine the impact on plant safety should one or more of these valves fail, undetecte The failure to identify the deficiency and effect corrective actions are apparent violations of Salem Administrative Procedure, NC.NA-AP.22-0006(q) and 10 CFR 50 Appendix B Criterion XVI, "Corrective Action" (EEI 50-272;311/95-02-04). Review of Enforcement Discretion Requests The inspector reviewed ten licensee requests for enforcement discretion, covering the period from October 1987 through January 1994. Typically, the licensee initiated the requests to allow maintenance or troubleshooting activities to continue without changing modes, e.g., replacement of an individual battery cell for the IC 125VDC battery, testing of No. 13 auxiliary feedwater pump, replacement of the motor for No. 22 containment fan coil unit motor, replacement of degraded piping in the core spray system, degassing of No. 1 station power transformer insulation oil, and conducting a special test of main steam isolation valve One request asked for relief from continuing a shutdown because the licensee anticipated imminent issuance of a waiver of compliance that would obviate the need to shutdow For each request, the inspector determined that until the NRC granted enforcement discretion the licensee complied with the applicable Technical Specification conditions for operation and associated action statements, and verified that the licensee complied with the reporting requirements of 10CFR50.72 and 10CFR50.7 The inspector concluded that the licensee had properly operated the units and met reporting requirement The inspector also reviewed a justification for continued operation (JCO) the licensee submitted June 17, 1993 in response to control rod malfunction On May 27, 1993, with a startup of Unit 2 reactor in progress, a control rod cluster withdrew when given an insert demand signa On June 4, the licensee completed their investigation into the anomaly and concluded there was a potential unreviewed safety question with respect to rod contro In response, Unit 2 operators inserted all control rod They also submitted a JCO for Unit 1, which was operating at 100% power. The inspector determined that, as in the instances of enforcement discretion requests, the licensee operated the plants in accordance with TS while the investigation was in progress and while awaiting NRC review and approval of the JC The inspector concluded that the licensee operated the units safely and the licensee complied with the reporting requirements of 10CFR50.72 and IOCFRS0.73.
'*
..
14 Reactor Head Leakage Detection System On March 17, 1988, the NRC issued Generic Letter {GL) 88-05, "Boric Acid Corrosion Of Carbon Steel Reactor Pressure Boundary Components in PWR Plants,"
in response to several incidents where leaking reactor coolant caused significant corrosion problem The GL discussed the effects of concentrated boric acid solution or boric acid crystals, formed by evaporation of leaking coolant, on reactor coolant pressure boundary component In many instances, licensees had detected the existence of leaks, but had not evaluated their potential impact on plant safety or taken timely corrective actio In a response letter, dated May 27, 1988, PSE&G described enhanced monitoring techniques and procedures, and specific inspection criteria to address this concer The letter also described a design change intended to improve the detection of small reactor coolant leaks. A reactor head leakage detection system, currently referred to as the main coolant system leakage air particulate monitor {MCSLAPM), was installed as an "experimental system" to provide continuous control room indication of radiological conditions-above the reactor head, inside the control rod drive mechanism {CRDM) ventilation shroud. Although not explicitly described in the licensee's May 1988 letter, the MCSLAPM was principally intended for evaluating the effectiveness of temporary clamps installed on thermocouple and spare control rod drive column The inspector reviewed the licensee's GL response and discussed the MCSLAPM system with Radiation Controls and Engineering personne Based on interviews with those responsible for the system, MCSLAPM has had no routine calibration or surveillance since installation in mid-198 Corrective maintenance had been performed on at least one occasion after the system had obviously faile However, licensee personnel believe that this system has provided qualitative trend information, based on past instances when the MCSLAPM indicated increased activity, coincident with indications from the safety related containment air monitoring syste Currently, a radiation monitoring system upgrade program is being implemented at Salem, which proposes removal of the MCSLAPM system and installation of a sample pump for the CRDM ventilation plenu The proposed sample pump would be different from the MCSLAPM because it would require a containment entry for operation and would not provide on-1 ine informatio The inspector noted that the experimental system mentioned in the licensee's GL response letter is currently in use at Sale Despite the fact that the MCSLAPM has not been maintained or tested' as a calibrated system, trend information from the MCSLAPM could help operators differentiate between leakage from the reactor head and leakage elsewhere in containmen The system is not classified as safety-related, and the inspector was unable to find additional information regarding the MCSLAPM on the Salem docket or in the Final Safety Analysis Repor The inspector noted that the temporary clamps installed in 1988 have since been replaced by seal weld Also, the central issue of GL 88-05 {i.e., detection of boric acid corrosion) has been satisfactorily addressed through procedural enhancements, plant walkdowns by system engineering personnel, and additional checks for boric acid