Synthesis of heat exchanger networks featuring a minimum number of constrained-size shells of 1–2 type

In chemical plants requiring a high number of heat exchangers, standard sizes are established so that most of the services can be satisfied through arrays of a limited number of different standard-type units. If such an industrial practice is not taken into account during the heat recovery network synthesis task, the optimality of the proposed design could be doubtful. This paper addresses the HENS problem allocating multiple constrained-size shells rather than a single one to accomplish a heat match. Two cases are considered: (a) pure countercurrent exchangers and (b) shell-and-tube exchangers of 1–2 type. The neighbor-based HENS framework of Galli and Cerdá (1998) has been generalized in order to adopt a more realistic fixed-cost target, i.e. the overall number of constrained-size shells. Therefore, new 0–1 variables have been defined to stand for the additional shells needed to get both, a shell size below the specified upper bound, and simultaneously, an FT correction factor above the threshold value everywhere. The resulting MILP problem formulation is now able to find network structures reaching the heat recovery target, under some structural constraints on the network design specified by the user, at near-minimum capital cost. The proposed algorithmic approach has been successfully applied to the solution of a couple of example problems and produced significant capital cost savings compared with prior HENS techniques.