Investigation on the possibility to use fork detector for partial defect verification of spent LWR fuel assemblies. Final report on Task JNT A 1071 (BEL, FIN, SWE) of the Member States' Support Programme to IAEA Safeguards

The possibility to use a fork detector for partial defect verification of spent LWR fuel assemblies has been investigated in Task JNT A 1071 Partial Defect Test on Spent Fuel LWRs. The task was arranged as a joint task between the Finnish, Swedish and Belgian support programmes to IAEA safeguards. This task studied the prospects of both a conventional fork detector and an enhancement where the gross gamma and neutron signals of a conventional fork are combined with simultaneous gamma spectrometry using a CdZnTe detector. The fork method was investigated by measuring BWR and VVER-440 spent fuel assemblies and a fresh MOX mock-up assembly. Correction methods were developed to improve the analysis of measurement results. Also model calculations were performed to clarify the effect of the geometrical configuration of the defect. The investigations have shown that a general partial defect test based on the fork method is not possible without making use of operator's declared data. There exist configurations even with 50% of pins removed, which cannot be detected, either with the conventional fork or with the enhanced fork detector. Using the operator declared data cannot be avoided due to the influence of both the fuel design and the irradiation history to the measured signals. If operator's data are available and considered reliable, the detection limit of a partial defect is at about 20% of pins missing for BWR assemblies with the burnup 18 MWd/kg or higher. For developing a reliable, operator data independent partial defect verification device a totally different approach must be applied. (orig.)