An approach to radiation shielding evaluations using estimators by expected scoring

This paper is intended to demonstrate currently achieved capabilities and still remaining limitations of a designed Monte Carlo neutral particle transport simulation methodology based on unbiased estimators using variance reduction techniques. This methodology is purposed for shielding evaluations to safety analysis and support of the design process of spent fuel and radioactive waste equipment including casks, storage facilities, etc. The paper contains a not comprehensive overview of the methodology including substantive description of some of its principal algorithmic features and discussion of their efficient implementation issues: estimators by expected scoring, an exponential transform modification for convex shielding bodies, multicase computation techniques, and others. Those are performed several obtained with using the methodology computational radiation shielding evaluation cases which represents the main computationally problematic categories of neutral particle transport numerical simulation: self-shielded sources, strong-attenuation shielding, labyrinths, skyshine, and complex heterogeneous shields involving tiny holes and cavities. As a result, it is found that the methodology allows to obtain sufficiently accurate for many practical engineering applications results, including high-resolution dose rate mappings, by a night-long calculation on a dual-core processor-based machine.