Statistical Modeling of Solar Energy

Abstract

Renewable energy comprises solar, wind, tidal, biomass and geothermal energies. Use of renewable energy resources as a substitute for fossil fuels inevitably reduce environmental footprint. Therefore, integration of renewable energy to the power grid, smart grid planning and grid-storage preparations are some of the major concerns in all developing countries. However, unpredictability in renewable energy resources makes the situation challenging. In light of this, the present study aims to develop a solar energy forecasting model to estimate future energy supply for a smooth integration of solar energy to the current electric grids. A suite of eight probability models, namely exponential, gamma, normal, lognormal, logistic, log-logistic, Rayleigh and Weibull distributions are used. While the model parameters are estimated from the maximum likelihood estimation method, the performance of the candidate distributions is tested using three goodness of fit tests: Akaike information criterion, Chi-square criterion, and K-S minimum distance criterion. Based on the sample data obtained from the Charanka Solar Park, Gujarat, it is observed that the Weibull model provides the best representation to the observed solar radiations. The study concludes with the analysis of forecasted solar energy and its possible role in replacing thermal energy resources.