Semantic search

On April 17. 2017, the staff identified a low refrigerant level in the Unit 1 train B essential chilled water (EC) system chiller during inspection. Operations personnel immediately declared EC chiller train B inoperable. On April 17, 2017, the leak was corrected and EC chiller train B was refilled with refrigerant to within the manufacturer's specifications. Operations personnel declared the system operable on April 18, 2017. The chiller had been inoperable since April 11. 2017, when the automatic purge system was placed into service. The direct cause of the low refrigerant level was leakage due to prior installation of a fitting on the automatic purge system filter without a plug.

During the 7-day period that EC chiller train B was inoperable, the supported low pressure safety injection (LPSI) system train B was inoperable. LPSI train A was also inoperable for approximately 17 minutes on April 13, 2017, during the performance of a routine surveillance test. This 17-minute period represented a condition that could have prevented the fulfillment of a safety function.

The cause of the leak was determined to be ineffective work instructions that did not identify the appropriate part number to be used during filter replacement. Corrective actions include revision of the work instructions. This change will ensure that the existing plug remains in place during filter element replacement. A leak test was also added to the work instructions to verify that no refrigerant leaks are present following maintenance.

On July 20, 2016, PVNGS received Unit 3 "B" train (3B) control room essential air filtration unit (AFU) carbon sample test results that exceeded the acceptance criteria of the Technical Specification (TS) Ventilation Filter Testing Program. The Unit 3 control room (CR) staff declared the AFU inoperable and entered TS Limiting Condition for Operation (LCO) 3.7.11, control room essential filtration system (CREFS). The carbon filter replacement and testing was completed, and the Unit 3 CR staff declared the 3B CREFS AFU operable on July 24, 2016.

The investigation determined the 3B CREFS AFU was inoperable since December 17, 2015, which exceeded the Required Action Completion Time for Conditions A and C of LCO 3.7.11 on December 24, 2015 and Condition E during movement of irradiated fuel. The direct cause of this event was exposure of the 3B CREFS AFU carbon filter to a high amount of volatile organic compounds (VOCs) during a CR renovation project. The apparent cause was a lack of knowledge and recognition by PVNGS personnel to identify and properly mitigate the effects of the project on the CREFS AFU resulting in inadequate guidance for controlling all potential sources of VOCs. Applicable change process and work control procedures have been revised to ensure flooring and furniture replacements that could impact the CREFS AFU are evaluated as potential sources of VOCs prior to performing work.

No previous similar events have been reported by PVNGS in the last three years.

On September 7, 2016, at 2131, with Unit 1 in Mode 1 and 100 percent power, the reactor was manually tnpped and reactor coolant pumps were secured due to a control malfunction which prevented closure of a pressurizer spray valve. This event initiated from an unsuccessful attempt to transfer a non-class 120VAC instrument power bus from its normal source to its emergency/alternate source. The transfer was attempted to facilitate an inspection of an electrical load center which was sprayed with water earlier in the day by a leaking sprinkler head following Fire Protection (FP) Department routine testing of a FP water line.

The cause of the spray valve malfunction was a failed pneumatic volume booster in the spray valve actuator system combined with a current to pressure converter (I/P) calibration offset which resulted from a voltage transient during the unsuccessful electrical transfer. These conditions caused the spray valve to stay approximately five percent open when it received a close demand from the spray valve controller. The I/P converter and pneumatic volume booster were replaced and the spray valve was returned to service. Additional corrective actions will adjust preventive maintenance frequency on the pneumatic volume boosters.

In the past three years, PVNGS has not reported a similar event to the NRC.

On April 10, 2016, at 2335, with Unit 1 in Mode 5, during a planned extent-of-condition inspection of the Unit 1 reactor coolant system (RCS), engineering personnel at the Palo Verde Nuclear Generating Station (PVNGS) identified white residue on a one-inch instrument nozzle on the reactor coolant pump 2B discharge pipe. Isotopic analysis confirmed the white residue resulted from leakage of RCS coolant and, at 0535 on April 11, 2016, engineering personnel determined that RCS pressure boundary leakage had occurred resulting in a condition prohibited by Technical Specification 3.4.14, RCS Operational Leakage.

The cause of the event was determined to be primary water stress corrosion cracking of the Alloy 600 instrument nozzle.

To correct the condition, the nozzle was repaired utilizing a Mechanical Nozzle Seal Assembly. A final repair of the nozzle will be addressed in the Corrective Action Program.

PVNGS reported similar events in licensee event report numbers 50-530/2015-001-00 (on June 5, 2015, when RCS pressure boundary leakage was identified on a Unit 3 RCP 2A suction pipe instrument nozzle) and 50-530/2013-001-00 (on December 6, 2013, when RCS pressure boundary leakage was identified on a Unit 3 reactor vessel bottom mounted instrument nozzle).

On August 13, 2015, at approximately 2106, the Unit 3 main steam isolation valve SGE-UV-181 (MSIV-181) B actuator train was declared inoperable and Technical Specification (TS) Limiting Condition for Operation (LCO) 3.7.2, Condition A, was entered due to a failed fitting on the air supply line. To correct the condition the failed fitting was replaced and an additional pipe support was installed on the air-line. Following retests, the MSIV-181 B actuator train was restored to operable status and LCO 3.7.2, Condition A, was exited on August 15, 2015, at approximately 1830.

A similar air-line fitting failure had occurred on the Unit 3 MSIV-181 B actuator train on May 19, 2015. The investigation of this condition following the second failure determined the MSIV-181 B actuator train air-line configuration was modified in the spring 2015 Unit 3 refueling outage and was inoperable from the time Unit 3 entered Mode 4 on May 1, 2015, at 0258, following the outage because the air-line tubing was not adequately supported following the design change.

The investigation of the two fitting failures is still in progress. The cause of this condition and any additional corrective actions will be reported in a supplement to this Licensee Event Report.

No similar conditions have been reported by PVNGS in the last three years.

On May 1, 2015, following completion of refueling activities, PVNGS Unit 3 entered Mode 4 and continued to Mode 3 in preparation for plant startup. On May 2, 2015, when plant conditions needed to test atmospheric dump valves (ADV) with steam were achieved, testing of ADVs was initiated. At 1739 on May 2, 2015, testing determined that ADV SGB-HV-178 (ADV-178) would not stroke more than approximately 13 percent open. Operations personnel declared ADV-178 inoperable and entered Technical Specification Limiting Condition of Operation (LCO) 3.7.4, Atmospheric Dump Valves, Condition A. An investigation determined ADV-178 was inoperable when Unit 3 entered Mode 4.

Inspection of ADV-178 determined internal sealing rings were improperly installed during maintenance performed in the refueling outage. The valve was repaired and tested and declared operable at 0853 on May 7, 2015. The causes of the event were human error by maintenance personnel and inadequacies with the procedure used to perform the valve maintenance. Corrective actions will revise work instructions to provide detailed guidance for valve re- assembly and to require verifications of proper re-assembly.

A similar event was reported in LER 50-529/2012-003-00 which resulted when testing in Mode 3 following refueling activities identified an inoperable steam supply valve for the steam driven auxiliary feedwater pump.

On November 6, 2014, at approximately 1116 Mountain Standard Time, Unit 2 was in Mode 1, 100 percent power, when control element assembly (CEA) 15 dropped fully into the core while the 88 other CEAs remained fully withdrawn.

Operations entered Technical Specification (TS) Limiting Condition for Operation (LCO) 3.1.5, Condition A, for one CEA misaligned from its group which requires a power reduction and restoration of CEA alignment. An initial power reduction was performed in accordance with TSs and attempts to repair the problem were initiated. The CEA could not be aligned within the 2 hour TS time limit and TS LCO 3.1.5, Condition C was entered at 1316 which required entry into Mode 3 within 6 hours. The power reduction was continued and the reactor was manually shutdown at 1636 to comply with TSs.

The direct cause of the event was a failed upper gripper coil on the control element drive mechanism (CEDM) for CEA 15.

The failed coil was replaced and Unit 2 was restarted and entered Mode 1 at 0332 on November 13, 2014. Operation of the coil at elevated temperatures accelerated thermal degradation of the coil insulation. A preventive maintenance strategy for establishing coil voltages and online CEDM coil monitoring, and adjustment, if necessary, was implemented.

In the previous 3 years, similar events related to malfunctions of control element drive mechanism control system equipment that resulted in a plant shutdown were reported in LERs 50-528/2011-004, 50-528/2011-005 and 50-530/2012-001.

On January 11, 2015, at 0024, Unit 2 received a plant computer monitoring system (RJ) alarm on point SASB22, indicating the setpoint for the bistable relay that compares pressures between Steam Generators (SGs) was approaching the technical (RK) and went unnoticed until late in the shift. Operators then verified the annunciator for SG differential pressure (DP) was not alarming and SG pressures were normal. The significance of the RJ alarm was not apparent because the value was displayed in units of voltage versus DP.

On January 24, 2015, further questioning determined the setpoint for the bistable that monitors differential pressure between SGs had exceeded its allowable value. Channel B SG Pressure Difference-High was declared inoperable and SG Level 2-Low was placed in bypass per LCO 3.3.5, Condition A.

The direct cause of the event was setpoint drift of the SASB22 bistable relay caused by potentiometers that had not recently been wiped clean. The root cause was the lack of an annunciated alarm for SASB22 and associated alarm response procedure to ensure the alarm condition was promptly acknowledged, understood, and correctly addressed within TS time limitations.

No previous similar events have been reported to the NRC by PVNGS in the prior three years.

On April 10, 2014, a post-event review conducted in response to a proposed NRC non-cited violation concluded the actions to close and deactivate main steam isolation valve SGE-UV-170 (MSIV-170) to address an equipment malfunction on November 6, 2013, did not meet the operability requirements of Technical Specification (TS) Limiting Condition for Operation (LCO) 3.7.2. As a result, the requirement of LCO 3.7.2 Condition G to place the unit in Mode 2 within 6 hours was not completed. MSIV-170 was repaired and returned to service on November 9, 2013.

An 'investigation determined the cause of the condition prohibited by TSs was the operability determination (OD) was overly focused on the ability of the MSIV to perform its specified safety function and did not adequately consider the definition of operability and compliance with the TS. To prevent recurrence, PVNGS OD guidance will be revised to explicitly require evaluation of applicable LCOs to ensure the OD technical conclusions support compliance with the LCOs.

No similar events have been reported to the NRC by PVNGS in the prior three years.

On October 6, 2013, during a scheduled visual examination of the 61 bottom-mounted instrument (BMI) nozzles on the Unit 3 reactor vessel, white residue was identified around the annulus region of BMI nozzle 3. At 0219 on October 7, 2013, engineering personnel determined, based on past and present photographic evidence, that the white residue likely resulted from reactor coolant system (RCS) pressure boundary leakage.

The pressure boundary leakage resulted in a condition prohibited by Technical Specification 3.4.14, reportable pursuant to 10 CFR 50.73(a)(2)(i)(B), and a degraded principal safety barrier, reportable pursuant to 10 CFR 50.73(a)(2)(ii)(A).

Non-destructive examination (NDE) of BMI nozzle 3 identified axial cracking in a near-surface weld flaw in the J-groove weld that extended into the nozzle tube and ultimately resulted in the RCS leak path.

An extent of condition evaluation found no indication of unacceptable flaws or leakage in the remaining 60 BMI nozzle assemblies in Unit 3.

Additionally, no evidence was found that indicated leakage from BMI nozzle assemblies in Units 1 or 2.

The probable cause of the BMI nozzle 3 leakage was a near-surface weld flaw that came in contact with RCS water and allowed initiation of primary water stress corrosion cracking. The flaw was likely caused by a BMI nozzle J- groove weld defect that occurred during reactor vessel fabrication and was undetected by penetrant testing.

An American Society of Mechanical Engineers Code approved weld pad/half-nozzle repair was performed to establish a new BMI nozzle assembly weld configuration.

Corrective actions include increasing the frequency of BMI visual examinations so that one is performed each refueling outage.

No similar events have been reported to the NRC by PVNGS in the prior three years.

1

This LER reports the failure of components caused by a defect in ARD66OUR control relays reported by Westinghouse, pursuant to 10 CFR 21, on April 8, 2013. The defect resulted in the failure of five normally energized control relays at PVNGS in two different systems in Unit 1 and two different systems in Unit 2.

The relays were used in normally energized applications which are de-energized to position associated components to support the related safety function. The relays failed to change state when de-energized during testing.

The five failed relays were replaced. The cause was a change in the manufacturing process that occurred in May 2008. The defect was exhibited only on relays that failed in the manner described in the Westinghouse Part 21 report. An extensive testing program was completed to identify and replace ARD66OUR relays installed at PVNGS that exhibited the described failure mode. Westinghouse has modified the ARD66OUR relay plastic molding process to preclude this described failure mode. LER 05000529 / 2009-002-00 reported ARD66OUR relay failures that exhibited a similar failure mode.

On October 8, 2012, at 13:30, when Unit 2 was in Mode 5 during refueling outage 2R17, Shutdown Cooling System (SDC) Train A was declared inoperable in accordance with Technical Specification 3.4.7 due to a leak on a Low Pressure Safety Injection (LPSI) system Train A drain pipe during SDC operation. The leakage source was a weld defect on the LPSI cold leg 1A injection pipe drain connection upstream of drain valve SIA-V908. The leakage was first discovered on October 7, 2012, at 22:00 when water on the floor adjacent to the LPSI cold leg 1A injection pipe was first found, but not identified as leakage through the drain pipe weld until insulation was removed on October 8, 2012.

A configuration control problem in the early 1990s allowed contact between the drain pipe and a pipe hanger when the SDC system was in operation. This resulted in a weld defect being introduced due to the high cyclic stresses from the pipe/hanger contact. The configuration control problem was corrected in May 1993; but, the weld defect propagated slowly during periods of SDC operations until the leak occurred in the 2R17 outage.

The cause was determined to be inadequate guidance to ensure temporary fittings on safety-related fluid systems were removed prior to placing the system in service. To prevent recurrence, procedures were revised to require that systems with capped pipe ends be returned to their design configuration following maintenance.

On May 9, 2012, a review of the procedural guidance for inoperable essential ventilation equipment by an investigation team concluded that a condition prohibited by Technical Specifications (TS) had occurred in unit 2 on January 22, 2012, when the TS requirements for electrical distribution equipment were not met.

On January 22, 2012, during surveillance testing in unit 2, a control relay did not change state as expected and the essential ventilation dampers controlled by this relay did not change position. Operations declared the affected essential air handling units inoperable. The TSs do not contain specific Limiting Conditions for Operation (LCO) requirements for the affected essential ventilation equipment and no TS LCO Conditions were entered. Operations followed existing procedural guidance for inoperable ventilation equipment. The procedural guidance required operations to verify that there was no loss of safety function and that normal room cooling was available.

The cause of the event was an inadequate procedure that was based on an incorrect understanding of the relationship between essential ventilation equipment that provides a support function to electrical distribution equipment and operability of the electrical distribution equipment. As corrective action, the affected procedure was revised.

Similar reportable events have not occurred in the past three years.

On August 6, 2011, at approximately 1119 Mountain Standard Time, during the performance of surveillance test 40ST-9SF01, Control Element Assembly (CEA) Operability Checks, the Unit 1 reactor tripped due to an automatic actuation of the reactor protection system. The actuation occurred when CEA 37 of shutdown group B fully inserted causing a deviation of greater than 9.9 inches from other CEAs in its subgroup. This deviation generated a CEA calculator penalty factor and actuated the low departure from nucleate boiling ratio and the high lOcal power density trips on all four core protection calculator channels. Following the reactor trip, plant operators observed CEA 16 of regulating group 5 did not indicate that it had fully inserted. While operators prepared to borate the reactor coolant system, full insertion indication for CEA 16 was received approximately 1.5 minutes following the reactor trip.

The cause for the reactor trip was determined to be a loose terminal lug on the CEA power switch assembly which developed a high resistance connection and led to the improper operation of the control element drive mechanism upper gripper coil. As an immediate corrective action, the terminal lug was replaced and retested satisfactorily. Troubleshooting of CEA 16 did not identify any problems. Unit 1 returned to service and entered Mode 1 at 2140 on August 10, 2011.

No similar conditions have been reported by Palo Verde in the past three years.

On January 19, 2011, at approximately 1840 Mountain Standard Time, Unit 3 was at 100 percent power at normal operating temperature and normal operating pressure when the main feedwater pump A minimum flow recirculation valve (mini-flow valve) failed open causing a percentage of feedwater flow to be diverted to the condenser. Subsequently, Unit 3 experienced a reactor power cutback when main feedwater pump B tripped on low suction pressure. The combination of the main feedwater pump B trip and the main feedwater pump A mini-flow valve failure caused both steam generator water levels to lower causing an automatic reactor trip.

Both steam generator levels continued to drop which initiated an auxiliary feedwater actuation signal. The plant was stabilized in Mode 3.

The failed open mini-flow valve was caused by a failed diaphragm in a pneumatic 3-way precision relay within the mini-flow valve control loop. The relay was replaced and Unit 3 was returned to 100 percent power on January 23, 2011, at approximately 0100.

No similar conditions have been reported by Palo Verde in the past three years.

On May 7, 2010, Palo Verde discovered that a calculation used for the closing force required for Main Steam Isolation Bypass Valves (MSIBVs) was non-conservative. During the Component Design Basis Review (CDBR), it was determined that the closing force would be inadequate to fully close the MSIBVs upon the receipt of a Main Steam Isolation Signal when Steam Generator (SG) pressure is greater than 700 psi. This condition renders the MSIBVs inoperable when SG pressure is above 700 psi. At the time of discovery, Unit 1 was in Mode 5 and the MSIBVs were not required to be operable; Unit 2 and Unit 3 were in Mode 1, and the MSIBVs were required to be operable. Both Unit 2 and Unit 3 entered the applicable Technical Specification (TS) Limited Condition for Operation (LCO) 3.6.3, Containment Isolation Valves. As an immediate corrective action, Unit 2 and Unit 3 ensured the MSIBVs were closed, with their penetration flowpath isolated to comply with the conditions of TS LCO 3.6.3.

Administrative barriers that existed at the time of the calculation revision (December 2000) were unsuccessful in preventing the error due to ineffective reviews and the lack of a questioning attitude. Calculations with the potential for this error were reviewed and evaluated. The extent of condition evaluation determined the error was limited to safety related Air Operated Valve (AOV) gate valves.

No similar events involving a non-conservative calculation for an AOV have been reported by PVNGS in the last three years.

On September 27, 2008, at approximately 21:51 hours, Mountain Standard Time, Unit 3 control room operators initiated a manual reactor trip from approximately 34% rated thermal power in response to high vibrations on the main turbine. The main turbine vibrations were experienced during a required power reduction resulting from a steam generator chemistry excursion. The event was considered an uncomplicated reactor trip. No automatic engineered safety feature (ESF) actuations occurred during the event and none were required. All safety related buses remained energized from normal offsite power during and following the reactor trip.

The cause of the reportable event was the decision by control room operators to initiate a manual reactor trip in response to the main turbine high vibrations. The operators chose to initiate the reactor trip earlier than intended in the planned shutdown in response to the high turbine vibrations. The reactor trip was not necessary to mitigate the consequences of the event.

There have been manual reactor trips in the past at Palo Verde Nuclear Generating Station but none with causes similar to this event. As such, the corrective actions for those events would not have prevented this event.

On February 1, 2008, during performance of the Component Design Bases Review initiative, engineering personnel determined that surveillance test procedure (STP), "Remote Shutdown Disconnect Switch and Control Circuit Operability," was not adequate to meet the Technical Specification (TS) Surveillance Requirement (SR).

TS SR 3.0.3 was entered for the affected components, and risk assessments were performed which supported continued operability until such time as the components could be adequately tested.

The direct cause of this condition was that the STP did not ensure that each circuit was verified to meet the requirements of the TS SR. The preliminary root cause analysis is that 10 CFR 50, Appendix R experienced personnel did not provide input during the development of the STP. Corrective actions include performance of new and revised STPs to adequately demonstrate compliance with the TS SR within the schedule permitted by TS SR 3.0.3.

There have been two previous similar occurrences of failure to meet TS SRs due to inadequate STPs in the past three years. Corrective actions taken as a result of those conditions would not have prevented this condition.

On August 11, 2007, Palo Verde Units 1, 2 and 3 were in Operating Mode 1 (Power Operations), at approximately 100 percent rated thermal power, when a void was discovered in the Unit 1 Containment Spray (CS) "B" header, which was not in compliance with Technical Specification (TS) Surveillance Requirement (SR) 3.6.6.2.

The direct cause of the void in the CS header was that the Unit 1 "B" CS Header was not properly vented during the fill and vent process on July 1, 2007. The root cause of this condition was ineffective use of operating fundamentals during the pre-job briefing, resulting in failure to identify the potential for air entrapment during the fill and vent process on July 1, 2007.

Subsequent investigation determined that the Surveillance Test Procedure (STP) forall three units did not adequately satisfy the requirements of the TS SR to demonstrate that the CS headers were full of water.

The direct cause for the identified failure to properly meet the TS SR to verify the CS header is full was a procedural deficiency in that STP 40ST-9S113, "LPSI and CS System Alignment Verification" did not adequately demonstrate compliance with TS SR 3.6.6.2. Ultrasonic Testing (UT) was performed on all CS headers in all three units to verify that they were full of water, and a small void in Unit 3 "B" header was detected and filled and vented. UT was subsequently added to the affected STP to ensure compliance with TS SR 3.6.6.2.

System containment sump piping due to a failure to translate the design intent to have the line filled with water into start-up, surveillance, and operating procedures.

On February 16, 2007, Unit 2 was In Mode 1 at approximately 100 percent rated thermal power, running the High Pressure Safety Injection (HPSI) 2A pump in support of activities to determine the leak rate of oil from the pump bearings. The oil leak rate was found to be greater than acceptable and at approximately 13:01 Mountain Standard Time (MST), Operations personnel declared the HPSI 2A pump inoperable and entered Technical Specification (TS) Limiting Condition for Operation (LCO) 3.5.3, Condition B, placing the plant in a 72 hour action. Corrective actions could not restore the pump to an operable status within the TS LCO allowed time. On February 19, 2007, at approximately 14:56 MST, Operations commenced a plant shutdown and at 16:45 MST, the reactor was manually tripped from approximately 20 percent power and the unit entered Mode 3. No TS LCO allowed action times were exceeded.

The oil leaks were corrected and on February 22, 2007, at approximately 05:02 MST Unit 2 declared HPSI 2A operable. The corrective actions were also implemented in the other five HPSI pumps on site.

In the past three years, there were two events reported for a TS shutdown that became necessary when attempted repairs could not be performed within the completion time for the TS LCO required action.

4.

On December 22, 2006, the Palo Verde Nuclear Generating Station (PVNGS) received a Final Significance Determination letter from the Nuclear Regulatory Commission (NRC) for Apparent Violations received during an NRC Heat Exchanger Performance inspection. The NRC concluded in the letter that the Unit 2 Essential Cooling Water, Train B heat exchanger was Inoperable and a Technical Specification (TS) 3.7.7 violation for a period of 78 days occurred, ending on September 27, 2003 when Unit 2 was shutdown for a refueling outage (U2R11). The subject heat exchanger was restored to Operable during the refueling outage. Essential Cooling Water, Train A heat exchanger was functional for the 78 day period that EW 2B was considered inoperable.

As a result of other heat exchanger problems, in May 2006, Arizona Public Service (APS) initiated an investigation of the Essential Cooling Water heat exchangers and the chemistry controls for the Spray Pond System. The APS investigation identified system problems with chemistry control, equipment maintenance and organizational weakness which allowed the problems to develop and remain uncorrected for an extended period. A number of corrective actions were implement or are being implement which include restoring effective chemistry controls of the system, procedure changes, personnel training, and equipment modification.

In the past three years, there were no similar events reported.

On July 16, 2006 at approximately 14:52 Mountain Standard Time (MST) Unit 2 was operating at 100 percent power, Model (power operations), when an Area Operator (AO) notified the control room that door C-A-06 was found open with no compensatory measures established.

Door C-A-06 is a watertight fire door that functions as the train separation barrier between Auxiliary Feedwater (AF) pump rooms 'A' and 'B.' When door C-A-06 is open with the unit operating in Mode 1, compensatory action must be taken in order to maintain both trains of AF operable. The AO who discovered door C-A-06 open and unattended, closed the door immediately. Based on a review of security computer transaction logs, control room personnel determined that door C-A-06 was open with no compensatory actions in place for approximately 4 hours and 20 minutes. As such, AF trains 'A' and 'B' were considered inoperable between 10:32 and 14:52 (MST). Based on further review and personnel interviews, an investigation concluded that a Fire Department Emergency Services Officer (ESO) had failed to close C-A-06 after leaving the AF pump room.

In the past three years, there were two similar events reported (LER 50-529/2005-003-00, and 50-530-2006-001-00).

On August 12, 2005 at 0019 Palo Verde Unit 1 was in Mode 1 (Power Operations), operating at approximately 94 percent power, when Control Room personnel commenced a reactor shutdown required by Technical Specification 3.8.1.

Prior to the event, on August 9, 2005, Emergency Diesel Generator "B" failed to maintain proper steady state output voltage during the performance of a routine monthly surveillance test.

Engineering and Maintenance personnel were unable to identify and correct the cause of the fluctuating generator output voltage within the 72 hour required action completion time associated with Limiting Condition for Operation 3.8.1 which would have expired August 12, 2005 at 0420. At 0216 on August 12, 2005 Control Room personnel manually tripped the reactor from 23 percent reactor power. At 0217, Palo Verde Unit 1 entered Mode 3 (Hot Standby), completing a reactor shutdown required by Technical Specification 3.8.1. At 2237 on August 12, 2005, Unit 1 entered Mode 5 (Cold Shutdown) and exited LCO 3.8.1.

All times in this report are approximate and Mountain Standard Time unless noted otherwise.

In the past three years, Palo Verde reported reactor shutdowns required by Technical Specifications but none associated with the same root cause.