A MINLP Study on Shell and Tube Heat Exchanger: Hybrid Branch and Bound/Meta-heuristics Approaches

Optimal design of shell and tube heat exchangers (STHE) has a crucial role in terms of the energy efficiency of plants. Therefore, it is currently one of the research topics of interest. In this study, a rigorous STHE design model strictly following TEMA standards and including various crucial constraints was established. Ten design variables of various kinds (continuous, integer, discrete, binary, and type) were selected for MINLP task. The handling of noncontinuous variables was attempted by means of three approaches including GA, PSO, and hybrid branch and bound methods over three case studies. Comparisons of the algorithms were performed using various performance criteria. Besides, the presence of noncontinuous variables in MINLP was analyzed. As a brief result, when type/binary/integer variables were included as optimization variables, very poor results were obtained, and hybrid PSO exhibited performance in each case superior to the others in terms of objective function value.