Scarcity of land sites applicable for wind turbines is pushing the technology offshore. By going offshore the expenses of nearly all components increases, which has triggered extensive research with the aim of decreasing component costs, and increasing reliability, as also maintenance costs increase as going offshore.
The focus of this thesis is to look at ways to reduce aerodynamic loads which can contribute in lowering structural fatigue loads and thereby component costs for offshore wind turbines. The loading is mainly reported as blade root flapwise bending moment on wind turbines intentionally made for bottom fixed offshore sites. The load reduction is investigated using a different wind turbine configuration with a downwind mounted rotor and further compared with the conventional upwind mounted rotor on a monopile tower. Blades on downwind mounted rotors are exposed to the fluctuating wake behind the towers, known as the tower shadow. The influence from the tower shadow on blade fatigue loads is investigated using three different types of towers; a full height truss type tower, a faring (airfoil shaped) tower and a monopile tower.
For reliable wind turbine simulations with downwind mounted rotors, an accurate tower shadow model is essential. Thorough investigations of the tower shadow is presented, including the detailed flow picture of the mean velocity deficit, unsteady and turbulent motions, as well as velocity spectra. The tower shadow is investigated in three different ways; using three dimensional physical model scale experiments, steady tower shadow models and two dimensional computational fluid dynamic (CFD) simulations.
By use of the tower shadow from the CFD simulations, an artificial increase in blade fatigue loading is seen as the transversal grid is made coarser. Although this is a ’safe-fail’ design (for the coarser grid), it should be kept in mind as this ’simulated’ safety factor from the coarse grid simulations could increase wind turbine costs.
A method for improving the accuracy of the steady tower shadow models (currently the most frequently used model in commercial software) through a preprocessing step, where the results are directly applicable in commercial software for full wind turbine simulations, is presented. This method will improve the reliability of the simulated results. The parameters of the steady tower shadow model and the turbulence intensity are fitted and calibrated with short CFD simulations of the relevant tower geometries. This method accounts for any deviation between the mean velocity deficit obtained from the steady tower shadow model and the CFD simulations, as well as the unsteady motions and turbulence due to the presence of the tower through the calibration of the turbulence intensity (maximum deviation of ±3 percent with respect to the tower shadow based on the CFD simulations, measured as blade fatigue loading).
The response measured as blade fatigue load show an increased loading for the blades on the downwind mounted rotors using the original blades, compared to the conventional upwind mounted rotor on a monopile tower. Introducing softer and lighter blades changed this result, with reductions in blade fatigue loading (compared to the upwind mounted rotor with the original blades) of three, four and five percent for the monopile, truss and fairing towers, respectively.