Optimal allocation of distributed generation and capacitor banks using probabilistic generation models with correlations

The allocation of distributed generation and capacitor banks is critical for success in the planning of power grids. A methodology is developed for the optimal placement of distributed renewable generation (wind and photovoltaic powers) and capacitor banks is developed based on technical and economic parameters. In order to preserve the horoseasonal and stochastic dependence nature of the wind and solar power, the methodology uses a model that integrates the sequential Monte Carlo method and the diagonal band Copula model, integrating historical data of wind speed, solar radiation and feeder load from the region of study. An efficient algorithm based on Genetic Algorithms is proposed to implement the optimization. The algorithm validation demonstrates a reduction of up to 71.7% in annual losses of active power in the Bandeira feeder and 73.4% in the Recife feeder, with adequate voltage levels and a return on investment of 6-7 years.

Automated summary

This study develops a methodology for the optimal placement of distributed renewable generation and capacitor banks, integrating technical and economic parameters and using historical data and a model for validation. The algorithm validation demonstrates a significant reduction in active power losses and achieves proper voltage levels and investment returns.