Statistical Modeling in Renewable Energy Sector of the US

We developed a time series model to analyze the trend of consumption of renewable energy based on the historical statistics published by the US Energy Information Administration in August 2012. R statistical package was used to perform statistical analysis and generate graphs. We observed that there is not any significant difference in the pattern of production and consumption of the renewable energy in the US. The consumption is in skyrocketing trend. HoltWinters filtering and exponential smoothing of the data, point estimates, interval estimates, confidence band and prediction band were generated and analyzed. We also established a multiple regression model that best estimates the parameters for the creation of number of jobs in ethanol industry of Minnesota as per the statistics released by MN Department of Agriculture and National Renewable Energy Laboratory. We found that the number of jobs in ethanol industry is best estimated as the multiple regression function of the total number of ethanol stations in operation and the output impact due to the ethanol production. These models allow us to analyze and predict the trend of the renewable energy by different sectors in future. These works on energy modeling will help the energy planners, researchers and policy makers to better understand the numerical data and the future trend in renewable energy sector. The usual assumptions of normal distribution and constant error variance were followed during the research. Researchers can further make data analysis with non-parametric smoothing and bootstrap methods to come up with different findings.