Variability assessment and forecasting of renewables: A review for solar, wind, wave and tidal resources

• Published in: Renewable & sustainable energy reviews 2015-04, Vol.44, p.356-375
• Main Authors: Widén, Joakim, Carpman, Nicole, Castellucci, Valeria, Lingfors, David, Olauson, Jon, Remouit, Flore, Bergkvist, Mikael, Grabbe, Mårten, Waters, Rafael
• Format: Article
• Language: English
• Subjects: Engineering Science with specialization in Science of Electricity; Engineering Science with specialization in Solid State Physics; Forecasting; Renewable energy; Solar energy; Teknisk fysik med inriktning mot elektricitetslära; Teknisk fysik med inriktning mot fasta tillståndets fysik; Tidal energy; Variability; Wave energy; Wind energy
• ISSN: 1364-0321; 1879-0690; 1879-0690
• DOI: 10.1016/j.rser.2014.12.019

Integrating variable and non-dispatchable renewable power generation into existing power systems will have consequences for their operation and future expansion. These impacts will depend on two factors: (1) the variability of the total renewable power generation on different time scales and (2) the possibilities of accurately forecasting these fluctuations. In this paper, previous research on variability assessment and forecasting of solar, wind, wave and tidal energy resources is reviewed. The aim is to summarize the state of knowledge in each area and to compare the approaches used for the respective resources. For temporal variability, methods and models used for assessing the variability are surveyed, as well as what is known about the variability at individual sites and for larger aggregates of sites. For forecasting, an overview of forecasting methods for the different resources is made, and selected forecasting methods are compared over different time horizons. An important finding is that it is hard to draw strong conclusions from the existing studies due to differences in approaches and presentation of results. There is a need for further, more coherent studies that analyze the variability for the different resources in comparable ways, using data with the same resolution, and for studies that evaluate the smoothing effect and complementarity of combinations of several renewable energy resources. For forecasting, future research should suggest ways to evaluate forecasts from different renewable energy sources in easily comparable ways, using data from the same locations or regions, with the same temporal and spatial resolution, and with comparable metrics for the forecasting errors.