Research Article
Improvement of Performance Indicators of Multistage Flash Desalination Plant using Genetic Algorithms
Mongi Ben Ali1,2 and Lakdar Kairouani2
1Département Génie Mécanique, Institut supérieur des études technologiques de Nabeul, Campus universitaire Elmrazga, 8000 Tunisia
2Université de Tunis El Manar, Ecole Nationale d’Ingénieurs de Tunis, 05/UR/11-14 Unité de Recherche Energétique et Environnement, Tunis Belvédère, BP 37, 1002, Tunisia
Submitted: September 16, 2017; Revised: December 17, 2017; Accepted: December 23, 2017
Abstract
Multistage flash (MSF) desalination plants are energy intensive and it is, therefore, important to use operating parameters that lead to reduction of energy consumption and consequently reduction of fresh water production cost. In this study, an optimization of operating parameters of an actual MSF-BR desalination plant was performed using as objective the improvement of the main plant performance indicators. Four decision variables related to the operating conditions were chosen for optimization, i.e., the temperature of the heating steam, the cooling seawater flow rate, the brine recycle flow rate, and the make-up flow rate. These decision variables were subjected to constraints to ensure that maximum and minimum bounds were adhered. A multi-objective function that consists of the main plant performance indicators, i.e., the thermal performance ratio, the specific cooling water flow rate, the specific recirculating brine flow rate, and the specific feed flow rate, were used in the optimization problem. In order to achieve this we have used a multi-objective solver (gamultiobj) available in the MATLAB optimization toolbox. This solver uses genetic algorithms for finding the Pareto-optimal solutions. The optimization results reveal that a significant improvement of the performance indicators can be obtained if we use the optimal operating points given by solving the optimization problem.
Keywords
Multistage flash; Performance indicators; Matlab optimization solver; Genetic algorithms; Pareto-optimal solutions.
Downlaod Full Text >>
