Modelos de geração fotovoltaica e avaliação da influência nos nos índices de confiabilidade de sistemas de distribuição

Abstract

This work presents models of photovoltaic (PV) generation used to predict the power output of the grid connected PV systems. The modeling of the generation system takes into account the PV arrays models, inverters and a plant model, which has been associated to spatial smoothing. Regarding the PV array models, the focus is the performance evaluation based on the maximum power point, in which are adopted semi-empirical expressions, based on both explicit equations and information frequently available in the PV module datasheet and provided by manufacturers. From time series of ground measured irradiance and the local ambient temperature, methods that combine accuracy and simplicity were used to generate power output time series in minutely and hourly resolution, in order to evaluate the impact of variability of the solar photovoltaic resource and the connection of PV power plants (distributed generation and utility-scale) in the reliability indices of distribution systems. Case studies on the IEEE RBTS Bus 2 system and a real distribution system are presented and discussed, for a single power plant at different locations, taking into consideration the effect of using the selected PV generation models, the temporal resolution of the time series and the smoothing of the power output fluctuations.