A new method to process algebraic equation systems used to model a MSF desalination plant

This work presents a new method, called heuristic selection (HS), to process algebraic equation systems (AEs). These systems are part of all chemical processes models that respect the state space models form. For given AEs, the proposed method produces a set of decision variables, a set of subsystems, a set of variables for iteration, a set of equations for verification, and all of the possible calculation sequences. The method allows assigning degrees of preference to each variable. These degrees are considered when the method has to select an output variable for a given equation. The goal of the method is to select the decision variables and the calculation sequence that either minimizes the size of the subsystems or minimizes the set of variables of iteration and, at the same time, selects the output variables in order to maximize the sum of the corresponding degree of preference. Finally, this method was used to develop a dynamic simulator for a MSF desalination plant.

A new method to process algebraic equation systems (AEs) was presented. This method produces a set of decision variables, a set of subsystems, a set of variables for iteration, a set of equations for verification, and the calculation sequence. These tasks are performed simultaneously during a single stage, which is an important improvement over traditional methods. In fact, those methods are generally designed to perform just one or a few number of those tasks, and they have to be sequentially applied along several stages. The proposed method allows assigning degrees of preference to each variable. These degrees are considered when the method has to select an output variable for a given equation. The goal of the method is to select the decision variables and the calculation sequence that either minimizes the size of the subsystems or minimizes the set of variables of iteration and, at the same time, selects the output variables in order to maximize the sum of the corresponding degree of preference. The method was used to determine the calculation sequence needed to solve the model of a MSF desalination plant simulated by Tarifa and Scenna [5]. In that work, a dynamic simulator was developed. Its model was formed by a large AEs system combined with another large ODEs system. The application of the HS method to that model produced a design of a robust, fast strategy to solve it. Acknowledgments This research is supported by Consejo Nacional de Investigaciones Científicas y Técnicas de Argentina (CONICET), Universidad Nacional del Jujuy (UNJu), International Center for Water and Energy Systems ICWES and Institute Foundation for Water Science and Technology, Abu Dhabi (IFFWSAT).