Globally Optimal Design Of Double Pipe Heat Exchangers Using Local Properties And Discretized Models

The design of heat exchangers in general, including double pipe, is typically performed using methods based on simplifying assumptions (e.g. uniform overall heat transfer coefficient based on averaged physical properties, mostly). However, these assumptions may sometimes result in considerable under or oversizing, especially when ample variations of the physical properties with temperature take place. Aiming at circumventing this limitation, we present a novel MILP (globally optimal) formulation of the optimization of hairpin double-pipe heat exchangers using rigorous models and discretization. Numerical results illustrate the importance of the utilization of the proposed approach in relation to the traditional analytical models, showing to what extent one can undersize or oversize severely the heat exchanger.