Monte Carlo with variance reduction in and around near-critical configurations

We report on advances in the use of Monte Carlo variance reduction in the source-iteration solution of the eigenvalue problem. The context is the modelling of neutron transport in and around reactor cores and criticality safety architectures. The application requiring variance reduction is the estimation of differential responses within and outside the core or fissile configuration. The ex-fissile configuration problems focus on the dose outside a PWR spent fuel cask and on a variety of radiation responses around a GEN III PWR core. Comparison is made with an empirical approach involving analog Monte Carlo in the fissile zone and variance reduction outside. The in-core test looks at the neutron flux and the 96Zr(n,γ) reaction rate at the lower surface of a control rod in two VERA Core Physics Benchmark problems. Comparison is made with analog Monte Carlo. This paper provides support and technical background to a summary article that contains the principal results.