Comparison of optimised germanium gamma spectrometry and multicollector inductively coupled plasma mass spectrometry for the determination of 134Cs, 137Cs and 154Eu single ratios in highly burnt UO2

Non-destructive and destructive methods have been compared to validate their corresponding assessed accuracies in the measurement of 134Cs/137Cs and 154Eu/137Cs isotopic concentration ratios in four spent UO2 fuel samples with very high (52 and 71 GWd/t) and ultra-high (91 and 126 GWd/t) burnup values, and about 10 (in the first three samples) and 4 years (in the latter sample) cooling time. The non-destructive technique tested was high-resolution gamma spectrometry using a high-purity germanium detector (HPGe) and a special tomographic station for the handling of highly radioactive 400 mm spent fuel segments that included a tungsten collimator, lead filter (to enhance the signal to Compton background ratio and reduce the dead time) and paraffin wax (to reduce neutron damage). The non-destructive determination of these isotopic concentration ratios has been particularly challenging for these segments because of the need to properly derive non-Gaussian gamma-peak areas and subtract the background from perturbing capture gammas produced by the intrinsic high-intensity neutron emissions from the spent fuel. Additionally, the activity distribution within each pin was determined tomographically to correct appropriately for self-attenuation and geometrical effects. The ratios obtained non-destructively showed a 1σ statistical error in the range 1.9–2.9%. The destructive technique used was a high-performance liquid chromatographic separation system, combined online to a multicollector inductively coupled plasma mass spectrometer (HPLC-MC-ICP-MS), for the analysis of dissolved fuel solutions. During the mass spectrometric analyses, special care was taken in the optimisation of the chromatographic separation for Eu and the interfering element Gd, as also in the mathematical correction of the 154Gd background from the 154Eu signal. The ratios obtained destructively are considerably more precise (1σ statistical error in the range 0.4–0.8% for most of the samples, but up to 2.8% for one sample). The HPGe gamma spectrometry can achieve a high degree of accuracy (agreement with HPLC-MC-ICP-MS within a few percent), only by virtue of the optimised setup, and the refined measurement strategy and data treatment employed.