Débora C. Faria

Profit-based grassroots design and retrofit of water networks in process plants

In this paper, we present a methodology for the grassroots design and/or retrofit of water utilization systems using mathematical optimization to maximize net present value (NPV) and/or return of investment (ROI) instead of minimizing freshwater consumption. The examples show that the solutions where savings and/or profit are maximized can be different from those where freshwater is minimized. They also differ from each other when ROI or NPV are used. In addition, when the NPV objective is used, the optimum solutions also vary depending on the interest rate used to calculate the discount factor.

Novel bound contraction procedure for global optimization of bilinear MINLP problems with applications to water management problems

We propose a new method to obtain the global optimum of MINLP problems containing bilinearities. Our special method that contracts the bounds of one variable at a time allows reducing the gap between a linear lower bound and an upper bound obtained solving the original problem. Unlike some methods based on variable partitioning, our bound contraction procedure does not introduce new integers or intervals. We illustrate the method by applying it to water management problems.

On the appropriate architecture of the water/wastewater allocation problem in process plants

Abstract We discuss the definition of the water/wastewater allocation problem as it was originally defined by Takama et al (1980), how this concept was modified and simplified through time, as well as additional issues that are still not properly addressed. We review a few attempts where parts of our view were pointed out and we further investigate the impact that proper modeling has on predictions of freshwater consumption and total annual cost.

A new approach for the design of multicomponent water/wastewater networks

Abstract Water allocation problems have nonlinearities and non-convexities due to bilinear terms. To address this issue we propose to discretize one of the variables of the bilinear terms. As a result an MILP model is generated, which provides a lower bound. To reduce the gap between this lower bound and the upper bound (a feasible solution found using the original NLP model), an interval elimination procedure is proposed. As a result, the feasible space shrinks after each iteration and the global optimum is identified. We illustrate the methodology for minimum water allocation problems.

COMPARATIVE ANALYSIS OF DIFFERENT ASSUMPTIONS FOR THE DESIGN OF SINGLE-CONTAMINANT WATER NETWORKS

One common assumption used in the design of water/wastewater systems for single components is to fix the process outlet concentrations of the pollutant to its maximum allowed value. This converts the problem from one with nonlinear constraints into one with linear constraints. For problems minimizing freshwater consumption in single-contaminant systems, this assumption has been proven to lead to global optimality (Savelski and Bagajewicz, 2000). In this article, we investigate the effect of using this assumption in cases where it may not lead to global optimal solutions, namely when the number of connections is minimized and when the cost is minimized. We therefore show that the use of nonlinear models helps in dealing with degenerate solutions featuring the same freshwater consumption.