GA BASED GLOBAL OPTIMAL DESIGN PARAMETERS FOR CONSECUTIVE REACTIONS IN SERIALLY CONNECTED CONTINUOUS STIRRED REACTORS

Abstract

This article uses Genetic Algorithm (GA) for the global design optimization of consecutive reactions taking place in continuous stirred tank reactors (CSTRs) connected in series. GA based optimal design determines the optimum number of CSTRs in series to achieve the maximum conversion, fractional yield and selectivity of the desired product at minimum total production cost. The results indicate that the optimum number of reactors required is two CSTRs in series in order to minimize the total cost, operating cost in particular, involved in carrying out the reaction. By employing reactors in series layout, the capital cost involved is reduced due to the decrease in reactor volume requirements. From the results of the optimization, it is found that there is 52.33% reduction in the volume needed for the reaction with the ratio of kinetic constants equal to one. The study has also been carried out at various ratios of reaction rate constants. GA based optimization is found to be an efficient technique as the results indicate the highest degree of accuracy and precision in comparison with the traditional gradient based method. Thus GA is more effective method in predicting the global design parameters for the process equipments such as chemical reactors used in industries.


JMDMES Vol.2(1) 2003: 15-25