MCNP modelling of a combined neutron/gamma counter

Abstract

A series of Monte Carlo neutron calculations for a combined gamma/passive neutron coincidence counter has been performed. This type of device, part of a suite of non-destructive assay instruments utilised for the enforcement of the Euratom nuclear safeguards within the European Union, is to be used for high accuracy measurements of the plutonium content of small samples of nuclear materials. The multi-purpose Monte Carlo N-particle (MCNP) code version 4B has been used to model in detail the neutron coincidence detector and to investigate the leakage self-multiplication of PuO 2 and mixed U-Pu oxide (MOX) reference samples used to calibrate the instrument. The MCNP calculations have been used together with a neutron coincidence counting interpretative model to determine characteristic parameters of the detector. A comparative study to both experimental and previous numerical results has been performed. Sensitivity curves of the variation of the detector's efficiency, ε, to, α, the ratio of (α,n) to spontaneous fission neutron emission rate and to f R, the reals coincidence gate utilisation factor, are presented. Sources of the inaccuracy in the calculations have not yet been fully investigated, because of the vast parameter space to be considered, but values of the coincidence gate utilisation factor derived directly from the MCNP data have been found to be overestimated by about 8.2%. Once bias-corrected, the trends of the real coincidence counts rate as a function of sample mass for three types of sample could be matched to experimental results within 0.33%. This result confirms the possible use of MCNP to calculate response trends accurately for a wide variety of source materials, given a limited experimental calibration set.