Challenge for screening of nuclear fuel debris by innovative spectral imaging and its verification by LIBS Mapping: (2) Hyperspectral imaging of simulated fuel debris samples in the visible and near-infrared ranges

Hyperspectral Imaging (HSI) is a non-invasive technique for rapid characterization, classification, and identification of materials based on their reflectance spectra. This study was conducted to investigate the applicability of HSI to efficiently sort fuel debris during the future large-scale removal of fuel debris from the Fukushima Daiichi Nuclear Power Plant (1F). A total of 10 simulated fuel debris samples with different compositions containing UO2 as nuclear fuel were prepared by powder sintering. HSI was performed on these samples in visible (350-1100 nm) and near-infrared (900-1700 nm) range to verify
whether the samples could be characterized, classified, and identified. As a result, different spectra were obtained for each sample. While the same composition system showed similar spectral responses, systematic spectral changes were observed depending on the UO2 ratio. The above results were observed in both ranges. Furthermore, classification maps generated by applying Spectral Angle Mapper (SAM) analysis to these spectra suggested that each sample could be identified.