05000250/LER-2007-002

LER-2007-002, Completion of Shutdown Required by Technical Specifications due to Inoperable Rod Position Indication for Two Control Rods in the Same Control Bank

Docket Number
Event date:	06-06-2007
Report date:	04-02-2008
Reporting criterion:	10 CFR 50.73(a)(2)(i)(A), Completion of TS Shutdown
Initial Reporting
ENS 43408	10 CFR 50.72(b)(2)(i), Tech Spec Required Shutdown
2502007002R01 - NRC Website

FACILITY NAME (1) DOCKET NUMBER (2) LER NUMBER (6) PAGE (3) Turkey Point Unit 3 05000250 NUMBERNUMBER

DESCRIPTION OF THE EVENT

On June 6, 2007, Turkey Point Unit 3 was operating at 100% power with no safety systems out of service.

At 0650, Operators observed that Rod Position Indication (RPI) for control rod F-4 in Control Bank C began to oscillate above and below 218 steps. At that time, Operations entered the off normal operating procedure for rod misalignment, declared control rod F-4 inoperable, and entered Technical Specification (TS) 3.1.3.1 Action d.1 to restore the inoperable rod to operable status. Operations confirmed that no control rod misalignment existed, and exited the 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> TS action.

Subsequently, at 0745 Operations entered the off normal operating procedure for rod position indication malfunction, declared the RPI for rod F-4 in Control Bank C inoperable and reviewed TS 3.1.3.2 for compliance. TS 3.1.3.2 allows continued operation with a maximum of one inoperable rod position indication per Control Bank provided the rod position is verified indirectly by the moveable incore detectors (i.e., flux map) every eight hours or power must be reduced to 75%. If more than one RPI is inoperable in the same Control Bank, entry into TS 3.0.3 is required to initiate a unit shutdown.

Prior to this event, there had been three RPI failures since September of 2006. Flux mapping was being performed for two control rods G-5 in Control Bank A and E-5 in Shutdown Bank B due to these two RPI failures that had recently occurred on May 1, 2007 for rod G-5 and on June 2, 2007 for rod E-5. The third RPI for rod M-6 in control Bank C had occurred September 1, 2006. Flux mapping was not being performed for rod M-6, since a Technical Specification change was approved by the Nuclear Regulatory Commission (NRC), to allow an alternate method for monitoring the rod's position, i.e., by verifying gripper coil parameters of the control rod drive mechanism to determine it has not changed state.

At approximately 0745, with inoperable RPIs for rods F-4 and M-6 in the same control bank C, unable to comply with TS 3.1.3.2 Action a., i.e., maximum of one analog rod position indication per bank inoperable, Operations entered 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> action in accordance with TS 3.0.3 to place Unit 3 in Hot Standby.

At approximately 0746, Operations initiated the Unit 3 shutdown. At 1152, reactor power was reduced below 25% and a manual reactor trip was performed in accordance with operating procedures. As such, at 1152, Operations exited TS 3.0.3. There were no abnormal indications observed during the duration of the Unit 3 shutdown. The RPI failures had no effect on the operation of any plant safety systems. There were no adverse effects on nuclear safety nor was the health and safety of the public compromised during this event.

This event was reported to the NRC on June 6, 2007 at 0932 pursuant to 10CFR50.72(b)(2)(i) due to initiation of a plant shutdown required by the plant Technical Specifications; and submitted event notification 43408. FPL condition report 2007-17324 was originated. This event is reportable pursuant to the requirements of 10CFR50.73(a)(2) (i)(A) due to a completion of a plant shutdown required by Technical Specifications. This LER is a supplement and supersedes the LER previously submitted to NRC by FPL letter L-2007-123 on August 6, 2007.

FACILITY NAME (1) DOCKET NUMBER (2) LER NUMBER (6) PAGE (3) Turkey Point Unit 3 05000250 NUMBER

BACKGROUND INFORMATION

The Turkey Point Unit 3 Reactor Vessel Closure Head (RVCH) was replaced during the Fall 2004 refueling outage. As part of the replacement effort, additional improvements were made to increase overall reactor vessel related system reliability, and enhance refueling/defueling operations. The RVCH was replaced with a new Integrated Head Assembly (IHA). This included all new IHA cable and connector assemblies for the Rod Position Indication, Control Rod Drive Mechanism, Core Exit Thermocouples, and Reactor Vessel Level Instrumentation System.

The RPI cable replacement included the following (Refer to Figure 1): a) replacement of cables surrounding the reactor cavity with new cable spliced into the existing cable, b) new bulkhead connector and panel on the refueling floor on the west end of the reactor cavity, c) new intermediate cables with connectors from the bulkhead connector panel to the RPI Seismic Plate Coil stack connectors. The existing Rod Position Indicator (RPI) coils were reused. The RPI coil stack seismic plate connectors were also replaced.

Figure 1: RPI Coil Stack Cable Connectors New Intermediate cable with Connectors Bulkhead (Angled) CRDMConnector Extension Tube Bulkhead Panel RPI Seismic Plate (Existing) RPI Coil Stack New Bulkhead � (Existing) Connector Pig-Tail (spliced to existing) New RPI Connector New RPI Pig-tails (spliced to coil leads) The rod position detector is a linear variable transformer consisting of primary and secondary coils alternately stacked on stainless steel support tube. The Rod Control Cluster Assembly (RCCA) drive rod serves as the "core" of the transformer. The vertical position of the drive rod changes the primary to secondary coupling and produces a unique A.C. analog secondary voltage. The output voltage is an analog signal directly proportional to the position of the control rod.

FACILITY NAME (1) DOCKET NUMBER (2) LER NUMBER (6) PAGE (3)

EVENT ANALYSIS

There have been four Turkey Point Unit 3 RPI failures since September 1, 2006 resulting in inoperable RPIs for rods M-6 in control Bank C, G-5 in Control Bank A, E-5 in Shutdown Bank B, and F-4 in Control Bank C. The number and chronology of failures was considered highly abnormal and potentially indicative of common cause failure. The symptoms of failure were similar amongst the four RPI detectors consisting of sudden erratic rod position indication without intentional rod movement.

After testing the RPI cables and connectors it was determined that the cause of the erratic RPI indication was the insulation breakdown of the connector insert on the reactor head. Based on part numbers, vendor documentation, insert material color, and laboratory testing of the failed dielectric from the M-6 RPI coil stack connector, it was determined that the RPI coil stack seismic plate connectors utilized neoprene rubber inserts, instead of silicone rubber. The connector insert, which should provide insulation between connector pins and connector body, was found to be conductive across its exposed face. All the failures identified were at the RPI coil stack seismic plate connection with the seismic plate-mounted male connectors being the failure initiator/propagator. For these failed connectors, all tests showed evidence of a migration of neoprene material to the mating cable connectors with silicone rubber inserts.

The RPI coil stack seismic plate connectors with neoprene inserts are a subcomponent of the RPI coil stack pigtail assembly. These assemblies were fabricated and installed during the Fall 2004 outage for the replacement of the reactor vessel head. The RPI cables and connectors installed in the vicinity of the reactor vessel must be able to maintain their physical and electrical insulation properties over many years under high temperature environmental conditions. The degradation of the neoprene rubber insert from the pigtail assembly connector contaminated and permeated the silicon rubber insert of the intermediate cable mating connector. The contamination of the intermediate cable connector resulted in a breakdown of its silicon rubber insert and caused it to become conductive. This conductivity resulted in shorting of conductor dielectric thus causing the erratic RPI detector oscillations.

CAUSE OF THE EVENT

There are two root causes for this event. One root cause is that the IHA Vendor Quality Assurance Program implementation failed to ensure the proper connectors specified by the IHA vendor and FPL purchase order were used in the fabrication of the RPI pigtail assemblies supplied to FPL. The other is that the FPL nuclear material management technical reviewer failed to review design basis documentation and take action to ensure proper engineering documentation was prepared for the new stock code part per FPL quality instructions.

FACILITY NAME (1) DOCKET NUMBER (2) LER NUMBER (6) PAGE (3) Turkey Point Unit 3 05000250

ANALYSIS OF SAFETY SIGNIFICANCE

Unit 3 was shutdown in accordance to plant procedures due to multiple RPI failures. The multiple RPI failures had no adverse impact on the ability of the operators to shutdown the reactor. There were no instances of rod misalignment.

ADDITIONAL INFORMATION: EXTENT OF CONDITION Based on the insulation resistance checks of all 45 RPI circuits, only the four RPIs identified previously were found with unacceptable resistance readings. The remaining 41 RPIs were capable of performing their function at the time of the Unit 3 shutdown. Only the RPI seismic plate stack coil pigtail assembly connectors had the neoprene inserts. The Unit 4 installed RPI seismic plate stack coil pigtail assembly connectors had neoprene inserts. Other head cable connector inserts were silicone rubber, which is the proper material for the application. The condition report for this event is 2007-17324.

CORRECTIVE ACTIONS

Turkey Point Unit 3 was shutdown and the RPI seismic plate coil stack connectors were removed.

Because of parts unavailability and the long lead time to procure new connectors, the RPI coil stack pigtail assembly wires have been spliced to the RPI intermediate head cables as a corrective action which eliminates the failure mode by eliminating the unqualified portion of the RPI system.

Turkey Point Unit 4 was proactively shutdown on July 22, 2007 to remove these connectors and splice the RPI coil stack pigtail assembly wires to the RPI intermediate head cables.

FPL personnel have investigated the IHA vendor's quality program. FPL Quality Assurance findings were incorporated into the IHA vendor's corrective action program for process resolution.

FPL nuclear material management technical reviewers were trained and qualified to FPL's Quality instructions.

SIMILAR EVENTS: There is no record of past occurrences of this type of event at Turkey Point.