In this work an approach is presented for the synthesis of near cost optimal heat exchanger networks with a high controllability. The presented methodology is based on a sequential approach utilizing different algorithms and mathematical models. The procedure is carried out in three steps. A genetic algorithm is used to produce a high diversity of promising structures that suffice the needs for cost reduction and controllability characteristics. For selected structures controllability evaluations are carried out, which are mainly based on relative gain array and condition number assessments. Therefore, a linear model is utilized which was further used to calculate the area adjustments and nominal bypasses for complete disturbance rejection for perfect control. As a last step simplified mathematical solution models are used to further extend insight gained for decision-making. These models are used to carry out dynamic simulations for testing the impact of different possible decentralized feedback control structure designs developed by previous methods. The application of the developed approach will be shown using an example from literature.
