Super-targeting of non-minimal WHEN systems

Abstract

Energy conservation and emission reduction have become critical issues due to increasing environmental policy pressures and high energy prices. Work and heat exchange network (WHEN) synthesis is a promising approach to enhance the energy efficiency of energy-intensive processes. WHEN considering single-stage compression/expansion can be divided into two classes by structure complexity: minimal structure and non-minimal structure. The synthesis problem of minimal structure has been addressed in our previous work, where the global optimal solution can be guaranteed. In contrast, the compressors/expanders are not at the beginning or end of the streams' paths in the non-minimal structure and therefore the inlet/outlet temperatures are now variables of the problem. Instead of solving this complicated problem directly, we propose a super-targeting model to screen and scope the solution space for primal promising options before detailed structure design which can be designed using regular HEN synthesis methods. The super-target is estimated by regulating the thermodynamic paths of hot and cold streams where the objective function uses classical HEN super-target to evaluate the costs of the heat exchanger network. Benefiting from the unimodal nature of non-minimal WHEN super-targeting problems, we develop two deterministic algorithms: one relies on derivatives and one is derivative-free. Case studies show the efficiency of the proposed algorithm. The resulting TAC can be 23.1% lower than the literature report and the solution of super-targeting problems are proven to be globally optimal. The derivative optimization algorithm has been found to exhibit superior efficiency in solving problems of relatively large scale.