Design and comparative parametric analysis using NSGA-II for multivariable constrained optimization of shell and tube heat exchangers

Abstract

Shell and tube heat exchangers (STHXs) have their pioneer use in industrial applications. The scope of this work is to design and optimize shell and tube heat exchanger of five distinct material combinations to comply with the requirements of effectiveness (heat recovery) and total cost as objective functions. For this purpose; thermal design is conducted according to TEMA specification in which tube outer diameter, length of tube, tube pitch, shell diameter, baffle spacing and baffle thickness are considered as primary design parameters. Ɛ-NTU method and Kern method are used to estimate tube side and shell side heat transfer coefficient and pressure drop respectively. For optimization, non-dominated sorting genetic algorithm (NSGA-II) is implemented in Python to attain a set of multiple optimum solutions called ‘Pareto optimal solutions’ stating the maximum effectiveness and the minimum total cost. The range of design parameters of pareto optimal solutions are also reported. Furthermore, responsiveness of total cost with respect to effectiveness of pareto optimal fronts are addressed. The percentage change of effectiveness and total cost are analyzed for 20% increase of input design parameters. Finally, a case study is explored where the optimized design parameters are selected from the pareto optimal fronts depicting the optimal cost about $11400 with effectiveness 0.68. After the different fundamental dimensions are obtained from the thermal design, a detailed mechanical design is also achieved using TEMA specification.