Multiple Turbine Wakes

The central goal of the present research was to study single and multiple interacting wind turbine wakes using both full-scale lidar experiments and high fidelity CFD numerical approaches.Firstly, single wake dynamics have been studied experimentally using full-scale (nacelle based) pulsed lidar measurements conducted on a stall regulated 500 kW turbine at the DTU Wind Energy, Risø campus test site. As part of the experimental analysis, basic Dynamic Wake Meandering modeling assumptions were validated. A wake center tracking algorithm was used to estimate the measured wake advection velocity and to obtain an estimate of the wake expansion in a fixed frame of reference. A comparison of selected datasets from the campaign showed good far wake agreements of mean wake expansion with Actuator Line CFD computations and simpler engineering models. An empirical relationship, relating maximum wake induction and wake advection velocity, is derived and linked to the characteristics of a spherical vortex structure. Additionally, a new empirical model for single wake expansion is proposed based on an initial wake expansion in the pressure driven flow regime and a spatial gradient computed from the large scale lateral velocities, and thus inspired by the basic assumption behind the Dynamic Wake Meandering model.Secondly, the impact of the atmospheric stability on wind turbine wake deficit is studied experimentally and numerically. The measurements collected from the previous pulsed lidar campaign was reused as part of the experimental analysis. An inflow wind sector of 30° is selected based on both a wind resource and a lidar data assessment. Wake measurements are averaged within a mean wind speed bin of 1 m/s and classified according to atmospheric stability using 3 different approaches: the Obukhov length, the Bulk-Richardson and the Froude number approach. Three test cases are subsequently defined covering various atmospheric conditions. Simulations based on the EllipSys3D ABL flow solver are carried out using Large Eddy Simulation and Actuator disc rotor modeling.The turbulence properties of the incoming wind are adapted to the thermal stratification using a newly developed spectral tensor, which includes buoyancy effects. Discrepancies are discussed as basis for future model development and improvement. Moreover, the impact of atmospheric stability and terrain on large/small scale wake flow characteristics was investigated.Later, wake interaction resulting from two stall regulated turbines aligned with the incoming wind were studied experimentally and numerically. The experimental work was based on a new dedicated full-scale measurement campaign involving 3 nacelle mounted Continuous Wave scanning lidars. A thorough analysis and interpretation of the measurements was performed to overcome either the lack or the poor calibration of relevant turbine operational sensors, as well as other uncertainties inherent to wake resolving from full-scale experiments. The numerical work was based on the in-house EllipSys3D CFD flow solver, using Large Eddy Simulation and fully turbulent inflow, where the rotors are modeled using the Actuator Disc technique. A mutual validation of the CFD model with the measurements is proposed for a selected dataset where wake interactions occur. An excellent agreement between measurement and simulation is seen in both the fixed and the meandering frame of reference. A benchmark of several wake accumulation models is performed as a basis for the subsequent development of an engineering model for wake interaction.Finally, the validated numerical CFD model is used as part of a parametric study where wake interaction is studied in a generic way, under several turbine spacings, mean wind speeds and turbulence intensities and in the fixed and the moving frame of reference of the wake. The analysis revealed that the industry widely used quadratic summation of single wake deficits for modeling the resulting double wake deficit is only relevant at high turbine thrust coefficients. For high wind speed and low thrust coefficient, linear summation should be primarily used. The first iteration of a new engineering model capable of modeling the overlapped wake deficit is formulated and its performance is tested again double, triple and quadruple wake deficits. Good performance in the prediction of both the maximum merged wake deficit and wake width is observed.