Toward optimizing the wind power generation and loads in Swedish conditions

Abstract

Over the years, the large eddy simulation technique has contributed greatly to the progress of the wind power industry by producing high-fidelity time evolution of the flow field around wind turbines. Yet, the substantial computational resources required for these simulations have driven researchers to explore ways to avoid or reduce these costs. This thesis presents three papers each contributing to this objective through different means. The first paper uses large eddy simulations to develop a fast-running 1D wake model that can predict the wake of a semi-infinite wind farm. A great agreement is achieved between the model predictions and simulations where the developed model can account for wind farm size, layout, and incoming turbulence level. Stiff and flexible actuator sector models are developed in papers II and III, respectively that can reduce the computation time of the simulation by allowing a larger time step size while maintaining an accuracy comparable to an actuator line model. Paper II investigates different aspects of the actuator sector model implementation by performing a parametric study. Based on this, a proposed model implementation agrees well with actuator line model simulations. In paper III, the flexible actuator sector model is developed and compared with its actuator line counterpart showing a good agreement concerning power values, damage equivalent loads, and wake flow. Two structural solvers of varying fidelity are examined, revealing the importance of the torsional degree of freedom and structural non-linearities to accurately calculate the power values and fatigue loads for large wind turbines with highly flexible blades.

Join the Licentiate presentation in Visby or on Zoom. If you want to listen on Zoom, please contact Mohammad Mehdi Mohammadi for the password.