Each AGR fuel assembly has two sets of brushes, one at the top of the assembly and one at the bottom. These brushes are designed to touch the surrounding graphite bricks and stabilise the movement of the fuel assembly within the reactor core. Fuel assemblies are periodically raised and lowered as part of the normal operation of the reactor, for example when refuelling is required. As the fuel assembly passes through the core as it is inserted or removed, the friction caused by the brushes contacting the graphite bricks varies due to small variations in the size of the channel bore, or as interfaces between bricks are passed. Consequently the load measured by the fuel hoist also varies, and this is captured in the form of the fuel grab load trace (FGLT).
Fuel grab load traces contain important information about the status of the graphite core, which experts at EDF Energy are able to extract by visual inspection. The analysis is complicated by the fact that both sets of brushes on the fuel assembly contribute simultaneously to the trace, and are separated by a distance of some 9 metres. This means that, whilst the bottom brush is mid-way through the core, the upper brush will be passing through equipment on the pile cap which will act to obscure some of the subtle features that are key to interpreting the trace.
Quintessa has developed LoTAS (Load Trace Analysis Software) for EDF Energy to help manage and analyse the fuel grab load traces. LoTAS is a web-based application, linked to a searchable database of FGLTs. When a new trace is received from a station, EDF Energy staff are able to upload the data via LoTAS to the database. LoTAS not only provides search and visualisation tools but is also able to analyse traces and automatically detect where the graphite brick interfaces are located on the trace. This information is reviewed and approved by staff before it can be used in further analysis to detect anomalies in traces. This is currently undertaken manually, but Quintessa is working with EDF Energy to develop mathematical algorithms that will automate the process. In future, these algorithms will be implemented within LoTAS, which will be able to alert EDF Energy staff to potential problems and possibly result in the need for a visual inspection of fuel channels.