SUMMARY

During the reactor licensing process, emergency planning is foundational to the design and operation of nuclear facilities, ensuring that response and mitigation strategies have been implemented in a worst-case scenario. However, current regulatory requirements of emergency planning zones (EPZ) were historically developed when bespoke dose and risk-based frameworks were not fully evolved, resulting in the utilization of generalized distance-based siting boundaries. However, advances in modern simulation tools have enabled the ability to construct site-specific dose and risk-informed frameworks.
Combined with the improved inherent safety design of advanced reactor concepts, the historical methods for EPZ designation may no longer align with the realized risk spectrum. The aim of this study is to investigate the computation of EPZs of advanced reactors two separate computational plume models: (1) a straight-line Gaussian plume (GPM) and (2) a semi-Lagrangian Particle in Cell (PIC). These atmospheric dispersion models were used as a basis for comparison in two case studies. The first was conducted as a benchmark that mirrored the approach taken by the NRC SOARCA study at Peach Bottom Nuclear Generating Station and another investigated the end-of-cycle inventory of an advanced INL Heat Pipe Design A microreactor concept.
Results demonstrated that in the farfield case for an existing large BWR, GPM and PIC estimate similar mean peak dose quantities and, using a 0.01 Sv threshold, both place the EPZ limit far beyond current regulations. However advanced reactors whose source terms are not expected to reach beyond the nearfield, PIC modeling the lack of specific nearfield dispersal modeling could result in prohibitively high dose distributions and thus inaccurate EPZ regions. On the other hand, both GPM and PIC display significant reduction in EPZ sizing to within the nearfield region and with further reductions in source term remove the need for one entirely.