Wind turbine wake inflow over a heterogeneous forest - comparison between measurement and LES simulation

Abstract

Abstract In this work a new step in understanding the wind turbine (WT) wake behavior on forested areas is made. For this analysis, a pair of real scale WTs located on a site with moderately complex terrain and heterogeneous forest is simulated using Large Eddy Simulation (LES). This simulation is compared with met mast and power output measurements of two WTs in Ryningsnas, Sweden, considering near neutral stratification in the atmospheric boundary layer (ABL). Three validation steps are followed; first, the undisturbed wind profile is compared with met mast data and another similar LES code. Then, the wake for each WT wake impacting on the met mast at different directions is addressed. A feature of this pair of WTs is that these have different hub heights, but the same rated power and rotor diameter, which helps provide insight into how the tip clearance over the forest affects the operation and wake characteristics. Finally, power output deficits when the WTs are operating in each others wakes are compared to observed power deficits. For these simulations SOWFA, the Open FOAM project for wind farms simulation in ABL, is used. In this code, three new additions are made; the forest model, the mesh modification for complex terrain and the representation of the WT using an actuator disc model with local force adaptation for wind farm flows. The simulation results show a good performance on quantitatively and qualitatively capturing the velocity in the wake, but for TKE the simulation underestimates the magnitude, and fails to match the measured structure of the wake for one of two WTs. The power deficit on the impacted WTs is well captured, despite the complexity related to turbines with different hub heights. This study makes one of the first steps on validating LES simulations for wind farms in forest.