earticle

영어

Background: Field atmospheric diffusion tracer experiments are commonly used methods for verifying the effectiveness of atmospheric diffusion models in nuclear accident consequence assessment systems. China Institute for Radiation Protection (CIRP) is conducting a study to verify the effectiveness of the nuclear accident consequence assessment system. This study proposes to conduct field atmospheric diffusion tracer experiments in different environmental conditions in China and use the experimental data to evaluate the effectiveness of model simulation in the China Nuclear Accident Consequence Assessment System (NACADOS). Materials and Methods: Some atmospheric diffusion tracer experiments have been conducted at a relatively flat coastal plant site. To evaluate the applicability of diffusion models in the Realtime On-line DecisiOn Support of Java version (JRODOS) system at China’s nuclear power plant sites, and the effectiveness of the model in the China NACADOS, the diffusion model Risø Mesoscale PUFF model (RIMPUFF) from JRODOS and a three-dimensional particle diffusion model from NACADOS were selected to simulate the sulfur hexafluoride (SF6) tracer experiment in the CIRP effectiveness verification research project. Similar to the currently commonly used methods for evaluating the effectiveness of atmospheric models, the simulation results were compared with the experimental results using various indicators such as fit factor (FACT), mean error (BIAS), and other statistical measures. Results and Discussion: The BIAS, mean absolute error, root mean square error, and normalised mean square error show that the model in JRODOS is better. The FACT2 and FACT3.5 also indicate that the simulation performance of the JRODOS models is better, but the FACT5 and FACT10 indicate that the NACADOS model is slightly better. Overall, the difference in model simulation results between the two systems is not particularly significant. Conclusion: For the currently conducted atmospheric diffusion tracer experiments in flat terrain conditions, the performance of the diffusion model in JRODOS is slightly better, and the difference in simulation results between the two systems is not particularly significant.