Refined Monte Carlo analysis of the H.B. Robinson-2 reactor pressure vessel dosimetry benchmark

Refined analysis, based on use of the Monte Carlo code MCNPX-2.4.0, is presented for the “H.B. Robinson-2 pressure vessel dosimetry benchmark”, which is a part of the Radiation Shielding and Dosimetry Experiments Database (SINBAD). First, the performance of the Monte Carlo methodology is reassessed relative to the reported deterministic results obtained with DORT. Thereby, the analysis is accompanied by a quantitative evaluation of the optimal energy cut-off value for each of the in- and ex-vessel dosimeters that were employed. Second, a more realistic definition of the neutron source is implemented than proposed in the benchmark. Thus, the current procedure for power-to-neutron-source-strength conversion, as also for explicitly considering the burnup-dependent fuel assembly-wise average fission neutron spectrum, is found to affect the calculated values significantly.In addition to the modelling refinements made, different approaches are tested for deriving the dosimeter activities, such that the neutron source time-evolution and the activity decay can be taken into account more accurately. Finally, in order to achieve a certain assessment of uncertainties, several sensitivity studies are carried out, e.g. with respect to the nuclear data used for the dosimeters, as also to the assumed physical location of the dosimeters. In spite of some apparent degradation in the prediction of experimental results when refining the Monte Carlo modelling, the final calculation/experiment (C/E) ratios for the measured dosimeter activities remain significantly improved (in general within 15% of unity), relative to the deterministic results.

► Activation of in- and ex-vessel radiometric dosimeters is studied with MCNPX.
► Influences of neutron source definition and cross-section libraries are examined.
► 237Np(n,f) energy cut-off is set at 10 eV to cover the reaction completely.
► Different methods for deriving activities from reaction rates are compared.
► Uncertainties are evaluated and are below 10%, final C/E ratios being within 15%.