Optimization of water networks using bound contraction

In this paper, an optimization model is developed for the synthesis of water reuse networks in industrial processes, considering scenarios with and without pretreatment units. The model has a mixed-integer nonlinear programming (MINLP) formulation containing bilinear terms. The objectives are to reduce freshwater consumption in water networks, effluent generation, and total annual costs. The optimization model was coded in GAMS, and the bound contraction technique, which involves contracting the limits of one variable at a time, was employed. This technique enables the reduction of the distance between the lower and upper bounds obtained by solving the original problem. Two case studies, one involving a single contaminant and the other multiple contaminants, were used to test the applicability of the developed model and the optimization method, considering regeneration and pretreatment units. Results indicate that global optimal solutions are found for all the cases studied. The strategy employed contributes to both a reevaluation of existing water networks and the design of new ones, resulting in more efficient configurations.