Optimal integration of photovoltaic generators into urban and rural power distribution systems

In this paper, we present a strategy for integrating photovoltaic systems into power distribution networks to improve the technical, economic, and environmental aspects of these networks. Taking into account variations in photovoltaic generation and energy demand, the proposed strategy uses a mathematical model that is subject to the technical and operational constraints of the system under analysis. The proposed strategy employs a master-slave methodology based on metaheuristic algorithms, in which the master stage determines the optimal locations for the photovoltaic generators and the slave stage estimates the optimal power injection. We proposed three different configurations of the methodology, each using the Particle Swarm Optimization (PSO) algorithm for the slave stage and the following algorithms for the master stage: the Chu and Beasley's Genetic Algorithm (CBGA), the Monte Carlo (MC) algorithm, and the PSO algorithm. These proposed configurations were evaluated in two test scenarios tailored to the characteristics of two regions in Colombia. This allowed us to analyze the impact of the methodologies on grid -connected (urban network) and off -grid systems (rural network). According to the results, the configuration that used two PSO algorithms achieved the most favorable outcomes in the urban scenario, reducing energy losses by 32.88%, operating costs by 42.41%, and pollutant gas emissions by 40.37%. Similarly, the same configuration obtained the best results in the rural scenario, reducing energy losses by 22.12%, operating costs by 43.35%, and pollutant gas emissions by 41.94%. These results were obtained in a typical day of operation for each system.