Impact of MCNP unresolved resonance probability-table treatment on uranium and plutonium benchmarks

Versions of MCNP up through and including MCNP 4B have not accurately modeled neutron self-shielding effects in the unresolved resonance energy region. Recently, a probability-table treatment has been incorporated into a developmental version of MCNP. MCNP results are presented for a variety of uranium and plutonium critical benchmarks, calculated with and without the probability-table treatment. The results from these calculations, along with their associated standard deviations, are presented. Four conclusions can be drawn from these results. First, not surprisingly, the only benchmarks that are substantially affected are those that have a significant fraction of their interactions within the unresolved resonance region of the principal uranium and plutonium isotopes that are present. For example, the unreflected metal spheres and the graphite-reflected IEU sphere have spectra that are too hard to produce large reactivity changes, while the fluoride and nitrate solutions have spectra that are too thermal. Second, the reactivity impact of the improvement is essentially negligible for 235U in these systems. The spectral peaks for all of the HEU benchmarks except GODIVA occur within the unresolved resonance region, but the probability-table treatment does not produce a statistically significant change in reactivity for any of them. Third, the probability-table method can produce substantial increases in reactivity for those benchmarks that include proportionately large amounts of 238U and high fluxes within the unresolved resonance region. Finally, the probability-table method also can produce significant reactivity changes for plutonium benchmarks with intermediate spectra. However, the magnitude of those changes is considerably smaller than for some of the cases with 238U. There are competing reactivity effects among the different plutonium isotopes, and in addition, the unresolved resonance range for the plutonium isotopes is not as broad as that for 238U