A simultaneous optimization of a flexible heat exchanger network under uncertain conditions

Heat exchanger networks are important for energy recovery and cost reduction. Flexible networks can handle environmental disturbances, and is more practical for industrial processes than the conventional HEN design with fixed parameters. Previous research on flexible HEN synthesis has mainly applied sequential synthesis, in which the structure determination and area optimization are conducted separately, or other multi-period methods in which only a few extreme operating conditions are selected; however, these methods are not sufficiently accurate for solving nonconvex problems. In this study, a bi-level simultaneous optimization strategy is proposed for the synthesis of flexible HENs under the nominal and all the extreme operating conditions. A binary particle swarm optimization is used in the outer layer to optimize the structural variables, and an Alopex-based evolutionary algorithm is used in the inner layer to determine the heat exchanger areas. Via this method, the result of a convex problem can be directly obtained. Furthermore, to deal with nonconvex problems, a flexibility analysis is conducted to adjust the heat exchanger areas on each critical point. The results of two cases with lower total annual costs than those reported previously are included to demonstrate the efficiency of the proposed method.