Metaheuristics Applied to the Optimal Renewable Microgrid Sizing to Supply Remote Communities

Abstract

Remote communities are small conglomerations of loads that are far from major energy centers and need alternative forms of generation. They are usually supplied by diesel generation due to storage and generation convenience. However, due to high operating costs and environmental concerns, other forms of generation have been sought. Renewable microgrids have shown promising potential in various applications nowadays with the development of technology. However, these solutions require careful analysis for implementation due to high initial investment costs and varying climatic resources. Thus, this paper presents two optimization methods for sizing a renewable microgrid using Genetic Algorithm and Variable Neighborhood Search metaheuristic, considering the uncertainties through Monte Carlo simulation. The application of the proposed method was evaluated through a case study for a real remote community in Canada. Both methods successfully sized the microgrid for autonomous operation with adequate annualized investment costs, avoiding the use of diesel generation.