Abstract :
This paper presents an algorithm for the minimization of reusable water storage in multipurpose batch plants. The algorithm is based on a two-stage approach, with each stage focusing on a dedicated objective function. In the first stage, the objective is the minimization of freshwater, given the maximum reusable water storage capacity. After the minimum freshwater target has been set, it is then fixed and used as an input parameter in the second stage. In the second stage, central reusable water storage capacity is a variable that must be optimized, subject to the minimum freshwater target set in the first stage. The overall formulation is based on a continuous-time approach developed by Majozi and Zhu [Ind. Chem. Eng. Res. 2001, 40 (25), 5935-5949], using a so-called state sequence network, and is an extension of recently published work on wastewater minimization using central reusable water storage in batch plants [Majozi, Comput. Chem. Eng. 2005, 29 (7), 1631-1646]. The algorithm has been applied to a case study in which a >45% savings in freshwater demand was observed, compared to the case without the exploitation of water recycle and reuse. Moreover, a >60% reduction in reusable water storage capacity was observed, in comparison to the situation in which the central reusable water storage capacity is fixed a priori. It is worthy of mention that the presented methodology is only applicable to batch processes that are characterized by single contaminants. Furthermore, the mathematical formulation on which it is based is nonlinear (mixed-integer nonlinear programming, MINLP) and nonconvex, which implies that global optimality cannot be guaranteed, except in special cases.
Keywords :
Perturbation method , Tidal water table fluctuation , Secular term , Non-linearity