05000499/LER-2001-004
| Event date: | |
|---|---|
| Report date: | |
| 4992001004R00 - NRC Website | |
DESCRIPTION OF EVENT:
On May 8, 2001 Unit 2 was operating at 100% power.
During refueling outage 2RE08 in the spring of 2001 new digital positioners were installed on the Unit 2 Feedwater Regulating Valves (FRVs). The digital positioners have the ability to communicate their parameters in digital format through a protocol standard known as Hart. Communication is established by connecting a Hart adapter between a computer and the positioner demand signal leads. The computer (laptop) contained software (Cornerstone) to translate the Hart signal to support trending and troubleshooting. After 2RE08, some of the FRVs exhibited occasional instability. Condition Reports were written to address the instability of FRVs '2A' and '2D'. On May 8, 2001 monitoring of the digital positioners was initiated to obtain data for the resolution of the condition reports. Monitoring of the '2D' valve digital positioner was completed using a laptop computer and a Hart adapter.
A pre job brief for the troubleshooting work on '2A' FRV was conducted in accordance with work control procedures. During the pre-job brief, it was discussed that a short could close the valve. If this occurred, Operations would notify the troubleshooting team to immediately remove the laptop computer. Shorting of the valve signal was considered highly unlikely based on previous successful monitoring activities including the monitoring of the '2D' FRV performed earlier that day. Monitoring of the '2A' FRV positioner was established using the same laptop and Hart adapter used for the '2D' FRV. After about twelve minutes of monitoring, the team connected the laptop computer to a utility outlet to keep its batteries charged. Shortly after the laptop was plugged in, the team noticed that all signals from the positioner had fallen to zero but believed it was an indication issue. Unknown to the troubleshooting team, the Hart adapter was designed with one of its leads connected to the laptop computer ground.
This grounded Hart adapter lead was connected to the positive signal lead to the '2A' FRV positioner. The FRV control circuit is designed with the negative signal lead connected to plant ground in the 7300 Process Cabinets. When the laptop computer was plugged into the utility outlet, the AC adapter's ground plug connected the two FRV signal leads. This resulted in a loss of signal to the positioner, which caused the FRV to close and the water level in steam generator '2A' to rapidly decrease. The shorted signal leads also caused the loss of communication from the positioner to the laptop computer, resulting in the signal observed by the troubleshooting team.
The control room operators observed the loss of feedwater flow with a 100% indicated demand signal. Operations then transferred the '2A' FRV to manual control in an attempt to reopen '2A' FRV and instructed the troubleshooting team to remove the monitoring equipment. After the monitoring equipment was removed the signal was again available to the positioner and the digital positioner started to reboot. The positioner completed its' reboot and responded to the 100% demand signal. Feedwater flow increased, but the increased flow was insufficient to keep steam generator '2A' from reaching the LO-LO Level trip setpoint. Operations staff ordered a manual reactor trip in anticipation of a LO-LO Level automatic trip but the reactor tripped automatically just as the operator was about to manually trip the reactor. All control rods fully inserted. The Auxiliary Feedwater System actuated on low-low steam generator water level as expected. All safety related equipment operated as required.
CAUSE OF EVENT
Root Causes 1. Personnel did not recognize the laptop computer and Hart adapter as test equipment with the potential to introduce a ground path. The Hart adapter documentation was not obtained to verify vendor input that the adapter was not polarity sensitive. Prior testing was insufficient to detect the ground path.
2. The work control process did not provide adequate controls for working on reliability-significant equipment.
CORRECTIVE ACTIONS
1. Programmatic controls will be established for the use of test or data acquisition equipment when used to monitor in service reliability-significant equipment. This action will be completed by August 30, 2001.
2. The work control process will be revised to provide enhanced controls for working on reliability-significant plant equipment. This action will be completed by September 13, 2001.
ANALYSIS OF EVENT:
A notification was made to the Nuclear Regulatory Commission on May 8, 2001 at 1924 pursuant to 10CFR50.72(b)(2)(iv)(B) for an actuation of the Reactor Protection System and 10CFR50.72(b)(3)(iv)(A) for the actuation of specified systems.
The conditional core damage probability (CCDP) for a reactor trip is approximately 2.0E-7.
There were no structures, systems or components that were inoperable at the start of the event which contributed to the event.
ADDITIONAL INFORMATION:
Operating experience reviews did not find any cases where use of a laptop computer contributed to plant events. There have been numerous occasions where grounded test equipment has caused plant events. The use of a laptop was not considered as test equipment in regards to previous operating experience.
The Hart adapter was purchased based on engineering discussions with the valve positioner vendor. In these discussions, engineering requested information concerning whether the Hart adapter was polarity sensitive. The vendor indicated that polarity did not matter. After implementation of the FRV positioner modification during refueling outage 2RE08 and prior to startup, the Hart adapter was tested to verify that interchanging the Hart adapter leads had no effect on valve and communication performance. However, this testing was performed using a laptop computer that had a 2-prong AC connector. The laptop computer utilized in this prior testing was from a different manufacturer than the laptop used in this event which had a 3-prong AC connector. The third prong connects to ground. The prior test results were satisfactory and falsely proved/reinforced the vendor input that the Hart adapter was not polarity sensitive.