Optimization of venturi scrubbers using genetic algorithm

Abstract

Optimization of a venturi scrubber was carried out using a nondominated sorting genetic algorithm (NSGA). Two objective functions, namely, (a) maximization of the overall collection efficiency η _o and (b) minimization of the pressure drop Δp, were used in this study. Three decision variables, the liquid-gas flow ratio L/G, the gas velocity in the throat V_gth, and the aspect ratio Z were used. Optimal design curves (nondominated Pareto sets) were obtained for a pilot-scale scrubber. Values of the decision variables corresponding to optimum conditions on the Pareto set were obtained. It was found that the L/G ratio is a key decision variable that determines the uniformity of liquid distribution and the best values of L/G and Z are about 1.0 × 10^-3 and 2.5, respectively. In addition, V_gth was found to vary from about 40 to 100 m/s as the optimal ηo on the Pareto increased (as did Δp) from about 0.6 to 0.98. The effect of adding a fourth decision variable, the throat length L_o, was also studied. It was found that this leads to slightly lower pressure drops for the same collection efficiency than obtained with three decision variables. An optimum length correlation for the throat of the venturi scrubber was obtained as a function of operating conditions. This study illustrates the applicability of NSGAs in solving multiobjective optimization problems involving gas-solid separations.