Condition monitoring of the nuclear graphite core using benchmarking techniques

Nuclear plant lifetime extension is an important goal within British nuclear industry. As a graphite core ages, there is an increased requirement to monitor the distortions within the core to permit safe continued operation of the station. In addition to existing monitoring and inspection, new methods of providing information relating to the core are being investigated. Two benchmarking methods are described in this paper to support the use of Fuel Grab Load Trace (FGLT) data, which has been shown to provide information relating to core distortions and brick cracking. The first method starts with the region of a single brick layer. Considering the similarities of each trace, those traces with their performance indexes out of the expected range are considered as abnormal candidates. The whole data of the interested region is considered in the second method. Recursive least square (RLS) algorithms are employed to identify peaks. As all the peaks have been identified, a potential crack can be classified according to its position.