05000397/LER-2003-007

LER-2003-007,

Docket Number
Event date:
Report date:
Reporting criterion:	10 CFR 50.73(a)(2)(iv)(A), System Actuation
3972003007R00 - NRC Website

REVISION

n 366A) (17)

Event Description

On June 30, 2003 at 0918 hours0.0106 days <br />0.255 hours <br />0.00152 weeks <br />3.49299e-4 months <br />, Columbia Generating Station (Columbia) experienced a main turbine trip followed by an automatic reactor scram. The plant was operating at approximately 79% power prior to the scram. The plant had recently commenced power operation on June 28, 2003, after completing the R16 refueling outage. As a result of the scram, all control rods fully inserted and the plant was brought to Mode 4.

Plant personnel collected data on the plant trip. It was noted that only a single main transformer differential relay target was found tripped, indicating that a true differential current fault condition probably did not occur. Additional troubleshooting inspected the current transformer (CT) circuits for non-design grounds (CTs are designed with a single point grounded). The circuits of all three main transformer phases that input into the 87TM/A, 87TM/13, and 87TM/C differential current relays were tested for grounds after the designed ground had been lifted. One of the circuits was still grounded indicating the presence of a non- design ground. Further troubleshooting revealed that the ground was on a wire in the CT circuit of the 'A' phase input to both the 87TM/A and 87TM/H relays.

Inspections of the CT circuit wiring revealed that the grounded wire and two other wires in the same conduit run showed surface damage due to fretting. There was no damage noted in other transformer wire bundles that were inspected. The fault was found in a vertical run of conduit.

The particular type of wire used was a thermoplastic insulated wire with a nylon outer jacket, rated at 90°C. Approximately one inch of the outer jacket was stripped, a portion of the thermoplastic insulation was damaged, and a small spot of copper conductor was exposed. The failed wire was not found to be pulled tight or forced against a conduit access box inner surface upon inspection. There is no record of any maintenance on these wires since original installation of the equipment.

Cause of Event

The root cause of this event is the use of non-jacketed cable (single conductor wiring) that was more prone to failure through fretting action. The insulation fretting damage was caused when the wire came into contact with the inner surface of a conduit access box while in relative motion. The relative motion was caused by vibration from the transformer and other adjacent equipment such as cooler group pumps and fans.

2615882 26158 R2

Safety Significance

This event posed no threat to the health and safety of the public or plant personnel. All safety equipment was available during this transient and performed as expected. Generator load rejection is bounded by the plant safety analysis. The plant trip was accompanied by a loss of one feed pump turbine due to a Reactor High Water Level 8 trip. Operations staff restarted the feed pump, and the balance of the plant shutdown to Mode 4 was uneventful.

Corrective Actions

The primary corrective action was to replace the CT conductors for main transformers TR-M1, M2 and M4 with jacketed cable. This provides a cable with heavier insulation that is less susceptible to fretting damage during installation and operation. In addition, the CT wiring for the normal transformers and the main generator was inspected, with no fretting damage observed.

Further Corrective Actions Current transformer wiring circuits that are: 1) associated with safety and non safety-related components; 2) may be subject to fretting wear; and 3) whose failure could prevent equipment operation or cause a spurious trip, will be analyzed, inspected, and repaired or replaced as necessary during the current operating cycle and upcoming R17 refueling outage.

Previous Similar Events

As previously discussed, on June 26, 2000, a similar event occurred at Columbia when a plant scram occurred due to a grounded CT (LER 2000-003-00). The cause of the ground was fretting at a tight elbow in the conduit. The previous corrective actions associated the June 2000 event failed to prevent the event discussed in this LER because there was a narrow focus on the failure. In the June 2000 event, bend areas were inspected and the failed wire was repaired. For the event discussed in this LER the failure was found in a vertical run of conduit that was not previously investigated.