Uma abordagem metaheurística para alocação de chaves e geração distribuída em rede de distribuição de média tensão considerando formação de microrredes com fronteiras flexíveis

Abstract

Distributed generation have grown in Brazil and in the world, transforming the distribution power grids into active networks. Microgrids are practical solutions to the distributed energy resources (DER) and local load integration, which can operate in parallel with the grid or in island mode. Island formation allows the recovery of customers during contingency scenarios, improving the system's reliability indexes. The active networks will be formed by multiple microgrids in the future. For the formation of the microgrids is necessary the installation of electrical switches and DERs. The location and amount of switches and DERs is crucial for the formation of microgrids. For the allocation of these components, it was applied the metaheuristic Genetic Algorithm with an objective function that defines the cost of the power losses, the investment, violations of the voltage limits and the Energy Interruption Cost (EIC) reliability index by using single-objective optimization approach. The interruptions represent a large part of the total system costs, so it's been proposed the formation of microgrids with dynamic boundaries to improve the EIC index. The study was developed on the simulator DIgSILENT PowerFactory through a DPL (DIgSILENT Programming Language) script, and the performance was tested on the IEEE-33 bus radial system. It was evaluated a total of 1373000 combinations during 50 executions of the main script, and on each execution it was extracted the best combination. Most of the individual objectives are confliting, for example, the solution with minimal Total System Cost (TSC) differs from the solution that presents the minimal EIC, showing that investments are necessary to increase the system's reliability. By using a multiobjective analysis through Pareto's Front, comparing the tradeoffs between the objectives, the results show that it's possible, with a relatively low investment, to reduce the values of TSC and EIC by 94.329% and 98.16%, respectively. This work show the importance of the reliability indexes on the cost analysis of power grids and proposes a solution with a efficient usage of the allocated grid resources through the microgrid formation with dynamic boundaries.