Hide/Show Apps

Numerical simulations of wind turbine wake interactions using actuator line and les models

Download
2019
Önel, Hüseyin Can
Wind is one of the most promising renewable energy resources of the future. After years of optimization studies, Horizontal Axis Wind Turbines shine out as the most efficient type and have been the only model used in large scale commercial wind farms. Layout planning plays an important role in getting the most power out of a wind farm as much as turbine blade design. Most important parameter in this planning phase is the inevitable wake generated by rotors and its impact on other wind turbines which results in power loss. Wake is a higly complex structure whose effects are immensely sensitive to boundary conditions. This situation raises a problem that is difficult to model, especially in wind farms where dozens of turbines are in operation simultaneously. Various analytical and empirical data based simplified wake models have been developed, but they all have a limited use in practical applications. With rapid advancements in computer technologies, Computational Fluid Dynamics offers high fidelity methods for wind turbine and wake simulations. One of these methods is the Actuator Line Model, where turbine blades are represented as distributed volumetric forces without the need of boundary layer resolution, hence saving some significant computational resource. In this study, accuracy and feasibility of Actuator Line Model in wind energy applications is evaluated. Model's sensitivity to several simulation parameters are assessed and two in-line turbines are simulated using best performing values. OpenFOAM software is used for Actuator Line Model implementation and Navier-Stokes solutions. Results are validated with Blade Element Momentum theory solutions and the model performed well in estimating turbine power production. Turbulent and vortical structures in the wake are best captured with LES turbulence model, velocity deficit curves and field variable plots are presented. Also, the model is capable of handling non-homogeneous conditions under atmospheric boundary layer flow.