Abstract
Currently, research on the effect of operating parameters and their optimization on the performance and cost-efficiency of NOx removal using active coke remains limited. Therefore, this study designs a simulation model using COMSOL software to analyze the denitrification process in a pipe packed with active coke particles. The effects of key operating parameters are investigated. The results indicate that, varying the temperature from 110°C to 150°C, the flow velocity from 3 to 5 m/s, the pipe length from 800 to 3000 mm, or the inlet NO concentration from 0.02 to 0.04 mol/m³, affects the denitrification efficiency by 42.23%, 19.39%, 43.6%, and 3.54%, respectively. In contrast, increasing the pipe diameter from 50 to 80 mm results in a negligible change of 0.41%, but directly increases the flue gas processing capacity from 0.0059 to 0.0151 m³/s. Additionally, a neural network-based predictive model is developed and trained using simulation results. A genetic algorithm is then applied to minimize operating costs and maintain effective denitrification. Key parameters, including temperature, flow velocity, and pipe length, are optimized, resulting in a reduction of operating costs by up to 25% while ensuring compliance with denitrification efficiency standards.