Optimization of Attila4MC Mesh Parameters in Large-scale Models for Nuclear Decommissioning Planning

The anticipated increase in nuclear decommissioning in the coming decades requires innovative approaches to maintain worker exposure as low as reasonably achievable. Occupational dose, an important component of cost-benefit analysis and work planning in decommissioning, can be estimated using radiation transport codes. Monte Carlo N-Particle (MCNP) is a robust, well-established code that has been used to model a breadth of geometries and source terms. Attila4MC offers a graphical user interface for users to build and run MCNP simulations and create an unstructured mesh of computer-aided design (CAD) models with tunable meshing parameters, including element edge length bounds and curvature refinement features. To understand how these parameters might be optimized for a large-scale model for dose estimation, a building containing gloveboxes, a hot cell, ventilation, robotic characterization tools, and operators was modeled in a CAD program. Source terms from available literature were applied to the equipment, and the operator dose was tracked for several exposure geometries. Mesh parameters, including maximum edge length (MEL) bounds, curvature refinement part selection, d/h ratio, and minimum edge length, were varied, and the resulting dose estimates were compared. The upper MEL bound had little effect on the estimated dose rate, but the varying the lower bound resulted in a 40% change in dose rate compared to the default case. Curvature refinement increased the MCNP figure of merit very slightly, about 2.6% when applied globally, but increased by over 31% when applied to only selected parts within the model. Both minimum edge length and d/h ratio showed a maximum change in dose rate of 10% compared to the default case for the values investigated in this study. Finally, the dose rate results suggest that the use of robotic or remote characterization methods may reduce occupational dose to workers by several orders of magnitude for the modeled scenario.