Universal, Collective and Nonlinear Structure of Wind Power Correlations

Abstract: Wind energy production is inherently variable due to the turbulent nature of atmospheric winds, posing challenges for grid stability and integration. Understanding the statistical properties of wind power fluctuations is important for improving renewable energy reliability. Previous studies have shown that aggregated wind power from multiple turbines exhibits persistent and intermittent fluctuations, linked to turbulence. In this talk, I report an analysis of wind power fluctuations from an 80-turbine farm spanning 20 km, using data collected at 10-minute intervals over a duration of five years. We find that wind power exhibits universal, collective, and nonlinear correlations that contribute to excess persistence and intermittency in aggregated production. A space-time correlation analysis reveals a transition from local decoherence to large-scale turbulence-driven scaling, while a bivariate analysis of power jumps uncovers nonlinear dependencies between turbine outputs. These effects lead to stronger intermittency in total farm output than expected from simple averaging, challenging conventional assumptions about power smoothing. These findings underscore the importance of nonlinear correlations already known to exist in power generation dynamics. By better characterizing these fluctuations, it is hoped that the results can inform strategies for grid management, storage optimization, and wind farm design. This work was performed with Dr. Samy Lakhal, JSPS Fellow at OIST and Mr. James Sardonia, Exus Renewables North America.