Multi-Objective Finned-Tube Heat Exchanger Optimization Using a Genetic Algorithm

Abstract

Heat exchangers are a significant component in many industries, particularly in energy conversion systems. The design of heat exchangers itself is a complex process because it involves experience-based decisions, numerous variables and parameters, and some of them are competing with each other. Genetic Algorithms (GAs) are one of the first evolutionary algorithms which remains one of the most extensively used non-linear optimization methods today. This study explores the implementation of Non-Dominated Sorting Genetic Algorithm II (NSGA-II) for thermal design and optimization of a finned-tube heat exchanger. The chosen objective functions were minimizing the heat exchanger volume and minimizing the air side pressure drop. The decision variables for the design were tube outer diameter, number of tube rows, fin pitch, unit height, and unit width. The calculated parameters and estimated cost of both preliminary design and optimized design were also compared. The optimized design offered a bigger alternative design while meeting all the constraints according to standards and industrial needs. The optimization reduced annualized operational and maintenance costs by 228% and lowered air pressure drop by 413% with bigger heat exchanger volume of 12% compared to the preliminary design.