Characterization of atmospheric turbulence using nacelle-lidar measurements and applications

Atmospheric turbulence, i.e., the random fluctuations of the wind, impacts both the power output and the structural load of wind turbines. Understanding and characterizing atmospheric turbulence are essential for the assessment of site conditions, to effectively use wind energy, and to evaluate the structural integrity of wind turbine components. Traditionally, anemometers mounted on meteorological masts are used to measure turbulence. However, with the increasing size of modern wind turbines, installing and operating meteorological masts that reach 200 m above the ground are very expensive and infeasible, especially offshore. In those cases, lidar is a cost-effective alternative to mast-mounted anemometers. Particularly, nacelle-based lidars have the advantage that they yaw with the wind turbine and track the inflow. Therefore, they have the potential to better characterize the flow that actually interacts with the turbine than in-situ anemometers on masts.

This thesis presents research on characterizing atmospheric turbulence using measurements from nacelle lidars and applications for wind turbine control and wake studies. The thesis objectives are (1) to investigate the methodologies to characterize atmospheric turbulence from nacelle-lidar measurements, (2) to study the impact of nacelle lidar scanning strategies on characterizing atmospheric turbulence, and (3) to make the best use of a single-beam lidar for wind turbine control and wake studies. Virtual and field measurements of nacelle lidars and aero-elastic wind turbine simulations are used to achieve these objectives.

The study in this thesis demonstrates that different nacelle lidars are able to measure the along-wind variance as well as a sonic anemometer, if the Doppler spectra of the radial velocities are used to account for the probe volume averaging effect. Nacelle lidars with more than six beams and two different beam opening angles are able to accurately estimate the Reynolds stress tensor following a least-square procedure. The best estimations of the six Reynolds stresses are found by a six-beam lidar measuring at a close focus distance with a large opening angle. These findings help lidar users to obtain accurate and detailed turbulence characteristics, which benefit wind turbine power performance assessments and optimizations, load validations and wind turbine design, and wind resource assessments in met-ocean applications.

Furthermore, the study shows that a single-beam lidar mounted in the wind turbine spinner achieves much more control benefits compared to the same lidar based on the nacelle. Being lower in cost and more flexible, the single-beam lidar in the spinner performs similarly to a four-beam nacelle lidar for feedforward pitch control. In addition, by using measurements from an inland wind farm, the study shows that the wake-induced power deficit of the downstream wind turbine decreases with increasing turbulence intensity, which is measured by a single-beam nacelle lidar on the upstream wind turbine. These findings demonstrate the potential of using a single-beam lidar for wind turbine control and wake studies.