Enhancement of micrometeorological methods for an improved understanding of the behaviour of wind turbines

In order to show the interaction between the flow in the atmospheric boundary layer and a wind turbine, the aeroelastic code FAST was coupled with the Large-Eddy simulation model PALM. With the help of these coupled models, information about the wind turbine at different atmospheric currents can be obtained, e.g. the loads, torques and performance of the turbine can be investigated. Various turbine models (Figure 1) were implemented with the aim of reducing the computational time without loss of information.

Figure 1: Schematic representation of various turbine models: The ALM represents the individual rotor blades as lines and therefore requires a small time step to completely display the movement of the rotor. In the ADMR model, the forces of the rotor blades are averaged over the entire rotor area, this saves computing time, but is less accurate. The ASM method is a compromise of the other two models.

Figure 2 shows the first results of a 5-megawatt turbine of the National Renewable Energy Laboratory (NREL), USA. The generator power was calculated in a laminar flow at 8 meters per second with different turbine models. The SWIRL model is a variation of the ASM in which the wind speeds are taken in front of the rotor and not, as it is usually the case, in the area crossed by the rotor. The wind speeds at the positions of the rotor are then calculated via induction in FAST. The coupling is to be validated with data from a 3.5-megawatt turbine from a measurement campaign that took place as part of the CompactWind project.

Figure 2: Generator power of NREL's 5-megawatt turbine, calculated for laminar flow at 8m/s. Different turbine models were used: The models ALM, ASM and SWIRL (solid lines) were calculated in the coupled model, i.e. the wind speed data were calculated with PALM and the forces at the turbine with FAST. For comparison, the ADMR model was used, which is implemented in PALM and simulates without FAST. On the other hand, FAST was calculated without the coupling to PALM as a reference.

In the future, the possibilities for deriving footprints from Large-Eddy simulations with an integrated Lagrangian particle model are to be expanded. In micrometeorology, the "footprint" is the area at the interface between earth and atmosphere that influences the turbulence characteristics observed at the measuring point. The footprint calculation using PALM simulations is to be extended so that it will be possible to determine the "footprint" not only for one point but for an entire rotor area. Such a method would be of great use for site characterization, e.g. also to identify causes of turbulence at an early stage or to check whether a later modification of the environment of the wind turbine (e.g. construction of a building, planting of a wind deflector, etc.) has effects on the wind conditions at the turbine site.