Isothermal Reactor Network Synthesis Using Coupled NonDominated Sorting Genetic Algorithm-II (NSGAII) with Quasi Linear Programming (LP) Method

In this study a new and robust procedure is presented to solve synthesis of isothermal reactor networks (RNs) which considers more than one objective function. This method uses non-dominated sorting genetic algorithm II (NSGAII) to produce structural modification coupled with quasi linear programming (LP) method for handling continuous variables. The quasi LP consists of an LP by adding a search loop to find the best reactor conversions as well as split and recycle ratios which are much easier to solve. To prevent complexity and ensure optimum solution, only ideal continuous stirred tank reactors (CSTRs), plug flow reactors (PFRs) and PFR with recycle stream are considered in producing reactor networks. Also, to avoid differential equations which appear in design equations of PFR reactors, CSTRs in series are replaced for each PFR. Results show that the proposed method finds better solutions than those reported in the literature.