Optimal design of heat exchanger network with open-loop controllability

Abstract

Heat integration via the optimal synthesis of heat exchanger network (HEN) is an important issue in chemical process design, especially for the process that consumes large amount of energy. But the concept of optimality does not simply concern the cost-minimum in the design stage, since the highly integrated process may require unexpected additional utility cost due to poor operability. The controllability is a crucial aspect during the process design under uncertain environment and/or existence of external disturbances. In this research, the controllable heat exchanger network synthesis for proper operation is proposed and compared with the cost-optimal heat exchanger network. The match-network hyperstructure that includes all possible matches and configurations is considered to generate the network structure. For the controllability assessment, the variable gain matrix and the variable disturbance matrix of a heat exchanger are derived in terms of the structural variables. Three open-loop controllability measures, minimum singular value, condition number and disturbance condition number, are introduced and included with the MINLP optimization framework. To solve the proposed synthesis problem, the GBD algorithm is adapted and applied to three examples. The resultant heat exchanger networks have improved controllability measures at the expense of slightly increased total area cost. The controllable network structures feature more utility controlled exchangers or bypass streams. The trade-off between economics and controllability has been proved.