We explore the potential for high-resolution gamma-ray spectrometry to be integrated into the IDR process in order to enable more responsive control over the enrichment process. To this end, we are designing a new assay system for enrichment assessment, which will be combined with digital data acquisition and advanced ratio analysis techniques to deliver reliable real-time monitoring of uranium enrichment.
Our research on image processing is centred around combining hyperspectral data and high-resolution RGB imagery to detect defects and impurities in manufactured fuel pellets. We aim to develop new machine learning algorithms for anomaly detection, which together with our proposed hardware solutions will enable issues to be resolved in real time at the production line.
In addition to insights gained from gamma-ray spectrometry and image processing, we are constructing a digital twin of the IDR process, which will enable data-driven adaptive condition monitoring and response control. A virtual model of the process is derived from real-time sensor data, and a deep-learning framework is used for fault diagnosis and prognosis.
As well as normal manufacturing cost and quality control issues, nuclear fuel manufacturers have the added complication of handling uranium, a radioactive and potentially very dangerous material. The enrichment of the uranium is of particular interest as it is a key parameter when producing nuclear fuel. We have tested an underused gamma-radiation detector material in nuclear fuel manufacturing processes, for the purpose of improving the measurement performance in identifying the uranium enrichment of the nuclear fuels under production. The results of our recently published paper, highlight the excellent performance of strontium iodide compared to more common gamma-ray detection scintillators when measuring uranium enrichment in nuclear fuel pellets.
