A CFD Study of Hazardous Radioactive Pollutants Dispersion under UAE Environmental Conditions

Abstract

As demonstrated in the Fukushima catastrophe, a malfunction of a nuclear power plant may result in the release of elevated quantities of noble gases, iodine-131, and cesium-137. Consequently, the key components of decision-support mechanisms for emergency preparedness and response to hazardous nuclear incidents entail the evaluation of the potential danger to the population through the modeling of the dissemination of radionuclides into the surroundings. Therefore, a meticulous assessment of the environmental repercussions of the discharge is imperative and must be determined with confidence, particularly in the vicinity of both the source of the release and of the plant edifices, where direct consequences may impact the personnel involved in accessing the facility during exigent circumstances. Several computational software tools for atmospheric dispersion, based on either Gaussian plume or regional Lagrangian models, have been developed to evaluate accidental scenarios involving the release of radionuclides. However, these models' predictive accuracy can be limited, especially in the near-field region. This is because the associated physics of pollutants' dispersion may not be adequately captured, as these models either partially account for the nuclear plant's buildings (e.g., Gaussian plume with building effects accounted for) or ignore them altogether (e.g., standard Gaussian plume model and regional models). Additionally, it is crucial to consider key UAE environment characteristics, including the arid ground surface topology and the atmospheric boundary layer stability regime. To address these shortcomings, the present study aims to use computational fluid dynamics (CFD) modeling methodology to accurately account for such conditions and assess their influence on the discharge and dispersion of radionuclides from the nuclear power plant to the environment. Before examining radionuclide dispersion around the nuclear power plant's building, a validation and verification process of the CFD Reynolds-Averaged Navier–Stokes (RANS) turbulence models is first carried out. Accordingly, two test cases have been selected. The first case (CASE A) was used to assesses the performance of five RANS models in predicting the wind speed distribution, recirculation zone size, and turbulence level at different locations, for a flow around a building without pollutant emission. The second case (CASE H) involves a flow around the same building as in CASE A but with different ground boundary conditions and pollutant release. These test cases are investigated, assessed, and validated using the commercial CFD code StarCCM+. Out of the five turbulence models that were examined, the Lag Elliptic Blending (EB) k-ε model performs better at predicting the dispersion of radioactive pollutants due to its superior performance in capturing flow dynamics. Therefore, in agreement with similar studies, viable prediction of atmospheric dispersion depends on selecting the correct RANS based turbulence model. The final case, Barakah Nuclear Power Plant (BNPP), involves the numerical simulation of a flow around the four units to investigate the pollutant concentration of radionuclide due to two postulated releases. The current study affirms and underscores the significant influence exerted by buildings on the path of radioactive pollutants emanating from hypothetical cracks. Accordingly, it suggests that relying solely on classical Gaussian plume models for evacuation strategies may prove insufficient. Such models might overlook intricate airflow patterns resulting from the presence of buildings, consequently leading to erroneous forecasts regarding the dispersion and deposition rates of gas pollutants.

Date of Award	7 May 2024
Original language	American English
Supervisor	Yacine Addad (Supervisor)