Automated design of discontinuous water systems

Both conceptual and mathematical methods have been established for the minimization of freshwater consumption and wastewater generation in the process industries. Although the methods have seen successful industrial application, most studies have focused on continuous processes. For discontinuous processes, the design of water systems is characterized not only by flowrate and concentration, but also time. Water recovery is limited through time constraints as water is used in different time periods. Also, storage capacity for water re-use is necessary to link different time periods. In this paper, a new design method has been developed for discontinuous water systems considering time constraints and the network design in which minimum cost is systematically identified. The resulting optimization problem to be solved is a mixed integer non-linear programme (MINLP). The design methodology accommodates design complications and practical issues. A decomposed approach is suggested to provide a reliable and robust starting point for solving the MINLP problem.