Wake Loss and Wind Farm Controls.

Characterizing and quantifying wind farm wake losses requires modeling the atmospheric conditions, turbine performance, and the impact on wind turbine wakes throughout a wind farm. This type of modeling can be computationally expensive, so this topic benefits from a hierarchy of models with varying levels of fidelity. High fidelity models employing computational fluid dynamics couple with turbine models of varying levels of fidelity to simulate the wake. Mid-fidelity models employ engineering models of the wake such as free vortex wake models, and they also couple with different fidelities of turbine models. At the lowest fidelity, analytical steady-state wake models provide a statistical representation of the wake while allowing for rapid simulations. Analytical models typically represent the turbine as a solid obstruction in the flow field.

This topic has high crossover with Aerodynamics. See that topic area for more information.

Software Stack

%%{init: { 'logLevel': 'debug', 'theme': 'forest' } }%% mindmap root("Wake Loss Modeling") ps(Engineering-fidelity Physics Models) OpenFAST FAST.Farm hfm(High Fidelity Models) AMR Wind Nalu Wind WindSE do(" ") ctrl(Controls) ROSCO FLORIS FLASC hercules ra(" ") cost(" ") wild(" ") om(" ")

At the core, OpenFAST provides turbine models of varying levels of fidelity making it a good central point for wind turbine wake modeling. It allows for controlling the wind turbine via a controller external to the software, so active control can be modeled, as well. Further, ROSCO provides a tool suite to automatically tune the controller parameters for a given turbine model.

In the high-fidelity regime, AMR-Wind and Nalu-Wind provide large eddy simulation models that can be coupled with OpenFAST. These can resolve the wake in great detail including tip vortices, wake mixing, and turbulence effects. Additionally, dynamic wake effects can be accurately captured. WindSE is a RANS model that includes it’s own turbine model. Hercules is a wind farm emulator that allows for wind farm developers and operators to investigate the potential impact of wind farm controls techniques.

In the mid-fidelity regime, FAST.Farm provides a free vortex wake model that can be coupled with OpenFAST. This model provides a computationally efficient method for simulating the performance of wind turbine blades with high flexibility that fall outside the scope of the blade element momentum theory.

In the low-fidelity regime, FLORIS provides a set of analytical steady-state wake models varying from simple top-hat wake shape (Jensen) to more complex statistical wake models (Gauss, Cumulative-Curl). Being a steady-state model, FLORIS allows modelers to capture instantaneous wake effects for a given atmospheric condition and control setting. FLASC is a post-processing tool that incorporates FLORIS to compare with field data.