Derivation of uncertainties to be used with the MONK5 code for water moderated, low enriched, uranium systems

The MONK Monte Carlo neutronics code is used extensively within the UK for criticality safety assessment purposes. Because of this it is clearly important that the use of the code be properly validated for the range of situations it is likely to model. This paper reports the work carried out on the CEGB computer at Park Street in London to show that the version of the code known as MONK 5.3 is valid in the area of cylindrical pin, low enriched UO2, water moderated systems. The method of validation was to assemble a large number of relevant benchmark experiments and model them using the code. The experimental data were mainly for lattices whose critical parameters were obtained by extrapolation. The calculated values were then considered to be independent predictions by the code. The predictions were tested for normality and for correlation with the several materials used in the benchmarks. The materials investigated included stainless steel, boronated steel, lead, rubber and boron dissolved in the water moderator. Tests were also conducted to see if there were any systematics associated with the three different enrichments, or the three independent laboratories involved. The estimates of systematic and random errors will be used in future CEGB criticality safety submissions to quantify the uncertainty to be associated with the predicted reactivity of fissile assemblies calculated using the MONK 5.3 code