Up until about five years ago, cyber-security fears and electromagnetic interference concerns were keeping wireless technologies away from most US nuclear facilities. By H.M. Hashemian
Except for the Comanche Peak Nuclear Power Plant in Texas (above), which implemented a wireless backbone into the plant for voice and data communication and used wireless sensors for equipment condition monitoring through a joint project with the Electric Power Research Institute (EPRI), cyber-security fears and electromagnetic interference concerns were keeping wireless technologies away from most US nuclear facilities.
A typical wireless data acquisition system contains four major components: a computer that serves as the network controller, a local wireless access point, a remote wireless access point and a networkable remote data acquisition unit. Computer and access point vendors provide low-level software to connect each of the IT (networkable) components together, incorporating industry-appropriate security and file transfer protocols. Higher-level software is used to provide application specific instructions for data collection, scheduling, analysis, alarms, etc. The systems use the Wi-Fi protocol, based on the IEEE 802.11g standard, communicating at 2.4 GHz, with WPA2 security and AES encryption, which includes error detection and correction techniques built into the protocol.
"In the short-term, wireless systems are likely to be small, application-specific deployments for assets that are difficult or cost-prohibitive to monitor using traditional means."
In the short-term, wireless systems are likely to be small, application-specific deployments for assets that are difficult or cost-prohibitive to monitor using traditional means. More significant applications require more work on addressing issues including: electromagnetic compatibility (EMC), cyber-security, reliability, latency (transmission delay), cost, industry-wide scepticism of the technology, issues surrounding deployment of many sensors on existing network infrastructure, communication spectrum management, power and cabling concerns for transmitters/receivers.
With research funding from the US Department of Energy, Analysis and Measurement Services Corporation (AMS) has begun to investigate these issues. Three case studies are presented below.
High Flux Isotope Reactor: cooling fans
Oak Ridge National Laboratory has started monitoring HFIR cooling fans with wireless vibration sensors.
The sensors are conventional accelerometers that are connected with a few feet of wire to a wireless hub. At each 6-hour interval, 30 seconds of real-time vibration data is collected from each of the three sensors per motor (there are four motors for the four cooling tower fans). The data is then transmitted wirelessly from the remote data acquisition unit at the cooling towers to the HFIR server, around 250 feet away. Each time that data is collected, vibration statistics are calculated and on-going trends are updated. The server software application provides a "Green – Yellow – Red" visual alert for each motor.
Previously the equipment was monitored manually using handheld vibration measurement equipment. This was effective, but the accompanying labour effort and costs reduce the efficiency of true predictive maintenance and early detection of incipient failures. Wired vibration sensors could have been installed to provide for continuous online monitoring of the equipment, but there was no room in the existing conduits on the HFIR cooling tower structure to accommodate additional sensor wires.
Arkansas Nuclear One: containment fans
Arkansas Nuclear One (ANO) used to manually measure the vibration of its containment cooling fans once every nine months to verify that the vibration levels met the plant technical specification requirements. The plant can be forced to shut down within 72 hours if one or more of the fans exceeds its vibration limits.
AMS automated these measurements by installing a vibration monitoring system. The system includes a number of conventional accelerometers (mounted by magnet to the exterior of the tubes surrounding the fans) that were connected to a wireless hub with less than 8 metres of wire, a multichannel wireless transmitter, and a receiver placed near the containment penetration, 75 feet (23m) from the transmitter.
As a part of installation, the electromagnetic compatibility (EMC) of the wireless vibration monitoring system with the plant environment was established through pre-installation laboratory tests as well as EMC measurements at the plant.
Arkansas Nuclear One: RCP oil tanks
The lubrication for the reactor coolant pumps at ANO is maintained in a tank within the reactor containment. Through the 18-month operating cycle, the used oil from the reactor coolant pump is directed to another tank referred to as a collection tank, which sits inside a concrete basin in the containment. Currently there is no way for the plant to know the level of the oil in the collection tank. If the level is increasing in the tank in the event of a leak, it means that the oil is not leaking at some other point of the oil collection system. Oil collection system leaks could create a fire risk if oil were to come in contact with hot pipework.
AMS installed a wireless level detector to alleviate this problem in the summer of 2013. A sonar-like sensor uses sound waves to detect the surface of the oil in the tank and, by measuring the time for the sound to be reflected back to the sensor, can determine the level of the oil in the tank. The sensor is enclosed in a stainless steel box (pictured) attached to the tank and sends a signal over wire to a remote transmission unit 100 feet away, which then relays it wirelessly to the hub.
H.M. Hashemian, Ph.D, president, Analysis and Measurement Services Corporation, AMS Technology Center, 9119 Cross Park Drive, Knoxville, TN 37923, USA.