Prediction of Wind Turbine Performance and Wake Losses using Analysis Methods of Incremental Complexity

Abstract

This paper compares several methods for predicting wind turbine performance, as well as predicting wake interference in a wind farm. The analysis tools range from the full CFD calculations at the high complexity end to a traditional Blade Element Momentum based approach at the low complexity end. We perform full CFD calculations using the HELIOS (Helicopter Overset Simulations) code co-developed by the Army Aeroflightdynamics Directorate at the NASA Ames Research Center and researchers at the University of Wyoming. HELIOS provides a novel meshing approach to CFD with body conforming unstructured grids near the solid walls and solution adaptive Cartesian grids away from the walls. This approach is particularly suited for vortex wake dominated flows such as those observed in wind turbine flow fields, where effective wake capturing depends to a large extent on numerical dissipation. All the lower complexity approaches that we utilize are developed in-house at the University of Wyoming. They are, in increasing order of computational time, a traditional Blade Element Momentum (BEM) code (WYOBEM), a BEM code with free-vortex wake approach for inflow (UWake), and an actuator disk based CFD method (DiskFloW). The different analytical methods are first applied to a model of the horizontal axis two-bladed rotor of the NREL Phase VI Unsteady Aerodynamic Experiment and then to a model of the 3-bladed turbines of the experimental Sexbierum wind farm. We further compare wake deficits behind an isolated turbine using the subset of methodologies that allow for probing of flow fields. Finally, we apply two of the chosen methodologies for studying performance losses caused by wake interference. In particular, we analyze the performance losses and nature of wind deficit when two turbines are operating, one behind the other, in pure axial flow conditions.