BETA : a system for automated intelligent analysis of fuel grab load trace data for graphite core condition monitoring

West, Graeme and Mcarthur, Stephen and Towle, Dave; Neighbour, Gareth, ed. (2009) BETA : a system for automated intelligent analysis of fuel grab load trace data for graphite core condition monitoring. In: Securing the Safe Performance of Graphite Reactor Cores. Royal Society of Chemistry, GBR, pp. 79-87. ISBN 9781847559135

Full text not available in this repository.Request a copy

Abstract

A key leg of the safety case required for operation of the Advanced Gas-cooled Reactor (AGR) stations is monitoring, particularly relating to the graphite core. It is not sufficient just to gather the monitoring data, but it must be analysed to extract the necessary information relating to the current condition of the core. The British Energy Trace Analysis (BETA) system has been developed to provide automated intelligent support to the analysis of Fuel Grab Load Trace (FGLT) data. FGLT data is routinely gathered during reactor core refuelling and can provide some information relating to the condition of the reactor core in addition to the inspections carried out during planned reactor outages. The process of developing the BETA software from a research prototype into a support tool which is installed on the British Energy network and accessible from anywhere in the company is described. The particular issues which have been addressed include the design, implementation, verification and validation of the software in terms of the core functionality, but also in the knowledge it contains in order to undertake its automated assessment of new refuelling event data. The second version of BETA is also described, highlighting the move to a web-based system with all the information relating to FGLT stored in a single location and describing the enhanced analysis that the BETA software undertakes. Finally, a forward look to the next version of BETA is provided, one which will integrate with other sources of condition monitoring data, and one which will have the ability to learn automatically as it is presented with new data.

CORE (COnnecting REpositories)