On the selection of design methodology for shell-and-tube heat exchangers optimization problems

Abstract

• Shell and tube heat exchangers are widespread and nowadays designed by software. • A wide number of optimization procedures are in use, leaving doubts on equipment real feasibility. • Used objective function and selection criteria are scarcely discussed in the literature. • The choice of the optimization objective is relevant. • Objective function strongly affects costs and engineering viability of designed equipment. In the last two decades research on shell-and-tube heat exchangers design has advanced focusing mainly on design optimization problems. While literature focused on applying a wide range of optimization algorithms, the attention to objective function selection has been comparatively scarcer, creating confusion in the designer, who is left without any guideline when undergoing this choice. Literature lacks clarity on reasons for choosing a specific function and authors never justify their choice. This results in equipment configurations which only apparently are optimized but may reveal practical drawbacks or, even, be unfeasible. As a contribute to fill this gap, in this paper a critical review and comparison of objective functions on a consistent basis is carried out, in order to show their effect on equipment configuration and assess respective merits and drawbacks. It is the first time that this kind of comparative analysis is carried out. Results show that the choice of an objective function strongly impacts on the obtained “optimal” design, that some frequently used functions are counterproductive and may lead to unpractical configurations from the engineering point of view. We show that thermodynamic-based objective functions may not lead to cost-effective design and that, in single-objective optimization, economic functions demonstrate their superiority. We also show that multiple objective optimization may lead to more balanced design and greater flexibility. The novelty and value of this study lies in the awareness coming from this additional knowledge allowing equipment designer and scholars to choose in a more effective manner the objective function to be paired with any optimization algorithm.