Optimization strategies and their application to heat exchanger network synthesis

Heat exchanger network synthesis (HENS) is a frequently encountered class of chemical engineering optimization targets. Besides its immediate relevance to practical, industrial application, the HENS problem presents an adequate complexity class to test new techniques of numerical optimization and benchmark them. This paper first sketches two established approaches to HEN optimization. The first one is based on thermodynamic insight, while the second one uses MINLP techniques for rigid optimization. Afterwards an alternative approach based on the paradigm of evolutionary optimization is introduced. The newly developed optimizer module is coupled to the well-known ASPEN PLUS simulator. The program package allows the optimization of parameters as well as network structure.