Surrogate optimization of coupled energy sources in a desalination microgrid based on solar PV and wind energy

Abstract

This paper proposes the use of new software called DesalinationPlant to solve the convoluted issue of calculating nominal capacities for an electric-to-water micro-grid comprised by an inexhaustible distributed energy resources; a decentralized storage system, and; a seawater desalination system. For this task, DesalinationPlant will use global stochastic black-box methods to optimize the amount of distributed resources and decentralized batteries coupled to a water desalination system. The final aim of this optimization is to maximize the profits of renewable micro-grid. DesalinationPlant is able to model very complex engineering systems, obtaining a feasible optimum in a limited time due to the small volume of evaluations needed to come to an optimum result. In this article, a complete desalination system associated with a photovoltaic plant, a wind power plant, electrochemical storage and a connection to the electric network was modeled to optimize the sizing of the distributed renewable resources. The optimization produces useful “optimal” values of the objective function, very “expensive” per se, in less than 100 evaluations. Therefore, it can be concluded that the method used, the DYCORS method is an excellent alternative to other black-box optimization methods such as genetic algorithms.