Shutdown dose rate analysis with CAD geometry, Cartesian/tetrahedral mesh, and advanced variance reduction

Abstract

In fusion energy systems (FES) high-energy neutrons born from burning plasma activate system components to form radionuclides. The biological dose rate that results from photons emitted by these radionuclides after shutdown-the shutdown dose rate (SDR)-must be quantified for maintenance planning. This can be done using the Rigorous Two-Step (R2S) method, which involves separate neutron and photon transport calculations, coupled by a nuclear inventory analysis code. The geometric complexity and highly-attenuating configuration of FES motivates the use of CAD geometry and advanced variance reduction for this analysis. An R2S workflow has been created with the new capability of performing SDR analysis directly from CAD geometry with Cartesian or tetrahedral meshes and with biased photon source sampling, enabling the use of the Consistent Adjoint Driven Importance Sampling (CADIS) variance reduction technique. This workflow has been validated with the Frascati Neutron Generator (FNG)-ITER SDR benchmark using both Cartesian and tetrahedral meshes and both unbiased and biased photon source sampling. All results are within 20.4% of experimental values, which constitutes satisfactory agreement. Photon transport using CADIS is demonstrated to yield speedups as high as 8.5 • 10 5 for problems

Implementation and application of PyNE sub-voxel R2S for shutdown dose rate analysis

“…In the photon transport process, the source sampling methods also differ between voxel R2S and sub-voxel R2S. In sub-voxel R2S, the source photons are uniformly sampled in each sub-voxel whereas in voxel R2S photons are uniformly sourced anywhere in the voxel [10]. Biased photon source sampling methods addressed in [10] are also implemented for sub-voxel R2S.…”

“…There are several mesh-based R2S implementations for calculating the SDR, such as MCR2S [7], R2Smesh [8], R2SUNED [9], PyNE R2S [10,11] and R2S with cosRMC [12]. All these validated mesh-based R2S implementations, with unique features, advantages and disadvantages, are reliable methods for calculating the SDR.…”

“…All these validated mesh-based R2S implementations, with unique features, advantages and disadvantages, are reliable methods for calculating the SDR. For example, PyNE R2S can reduce the modeling efforts to calculate the SDR for fusion systems due to its ability to perform SDR analysis directly on CAD geometry with Cartesian or tetrahedral meshes and support advanced variance reduction [10,13]. PyNE R2S supports multiple Direct Accelerated Geometry Monte Carlo (DAGMC) embedded Monte Carlo particle transport codes, such as DAG-OpenMC [11] and DAG-MCNP5 [10], with not much modification to the source code except for the subroutine for photon source sampling.…”

“…For example, PyNE R2S can reduce the modeling efforts to calculate the SDR for fusion systems due to its ability to perform SDR analysis directly on CAD geometry with Cartesian or tetrahedral meshes and support advanced variance reduction [10,13]. PyNE R2S supports multiple Direct Accelerated Geometry Monte Carlo (DAGMC) embedded Monte Carlo particle transport codes, such as DAG-OpenMC [11] and DAG-MCNP5 [10], with not much modification to the source code except for the subroutine for photon source sampling. The interface functions and codes of PyNE R2S can be easily transferred or applied to other Monte Carlo codes with mesh-tally ability, such as cosRMC [12].…”

“…However, this is currently unavailable for users not involved with ITER; Monte Carlo codes such as OpenMC, cosRMC. R2S with an unstructured mesh that fits the geometry exactly, which is available in both MCR2S-UM [18] and PyNE R2S [10], requires much higher computational resources in terms of memory usage and CPU time.…”