ARIMA-based Forecasts for the Share of Renewable Energy Sources: The Case Study of Germany

Abstract

Renewable energy sources are the better alternative to the traditional fossil-based generation. However, the generation from renewables is discontinuous due to their high dependency on environmental conditions. This makes their integration into our modern grid very challenging and necessitates suitable forecasting models. In this paper, the problem of generating forecasts for the percentage of renewable energy sources is studied. To this end, motivated from our previous work and the lessons learnt, a new set of ARIMA-based models for each month of the year is proposed. A finer analysis for the identification of the exogenous variables is carried out. The improved methodology of this paper contributes to enhanced predictions, which is showed to be not exceeding 10% for the considered years and the case study in Germany.