A framework for analyzing and optimizing renewable energy portfolios

With growing penetration of renewable energy sources in power grids, it is increasingly important to reduce the renewable power forecasting error and variability to maintain balance of grid load and supply and participate in wholesale energy markets. Power from weather-dependent renewable sources like wind and solar are particularly subject to variability and forecasting error. In this study we propose an innovative framework for analyzing the renewable generators at a given location and constructing energy portfolios that minimize the variability and forecasting error of the overall power output. The framework's key innovations are (1) its ease of automated implementation and (2) its ability to work even without any geographical diversity. We have implemented this framework for wind turbines and solar photovoltaics, and successfully executed it for a location in eastern USA. The results from this experiment have been quite promising and they demonstrate that both renewable power forecasting error and variability can be reduced by 40% with our framework.