A novel approach to parametric instrumental neutron activation analysis at MURR has been established. In particular, a detailed MCNP5 steady-state model of the MURR core was developed. The model, which was based on the most recent continuous-energy neutron data from the ENDF and JEFF libraries, was used to compute the local continuous-energy neutron flux distribution. By coupling the computed flux spectrum to the energy-dependent (n, γ) cross-sections for a range of nuclides, their intrinsic reaction rates were predicted in irradiation channel ROW2. The model was initially benchmarked by measuring the intrinsic (n, γ) reaction rates for a set of mostly dilute single-element standards in ROW2.

Results show that the model predicts the absolute reaction rates of many nuclides including those with high epithermal sensitivity (e.g., Au-197 and Zr-96), and non-1/v nuclides (e.g. Lu-176) within ±5% of the measured values. Using predicted (n, γ) reaction-rates characterized as the parameter π_theo, trace-elemental concentrations were determined in NIST standard reference materials, bovine liver, obsidian and coal fly ash. The agreements with the certified values were generally within ±5%. The new methodology has produced agreements with the certified values that are better for a greater number of elements than k₀. The model was also combined with MONTEBURNS and ORIGEN to test the feasibility of Mo-99 production at MURR from fissioning LEU. Results from a 5-gram low-enriched uranium target show predictions of Mo-99 end-of-irradiation yields are within 3% of the measured value. This dissertation entails a complete study of the MCNP5 model and the new neutron activation analysis method.