No abstract

One explanation for wind turbine power degradation is insect roughness. Historical studies on insect-induced power degradation have used simulation methods which are either unrepresentative of actual insect roughness or too costly or time-consuming to be applied to wide-scale testing. Furthermore, the role of airfoil geometry in determining the relations between insect impingement locations and roughness sensitivity has not been studied.To link the effects of airfoil geometry, insect impingement locations, and roughness sensitivity, a simulation code was written to determine representative insect collection patterns for different airfoil shapes. Insect collection pattern data was then used to simulate roughness on an NREL S814 airfoil that was tested in a wind tunnel at Reynolds numbers between 1.6 × 10 6 and 4.0 × 10 6 . Results are compared to previous tests of a NACA 63 3 -418 airfoil.Increasing roughness height and density results in decreased maximum lift, lift curve slope, and lift-to-drag ratio. Increasing roughness height, density, or Reynolds number results in earlier bypass transition, with critical roughness Reynolds numbers lying within the historical range. Increased roughness sensitivity on the 25% thick NREL S814 is observed compared to the 18% thick NACA 63 3 -418.Blade-element-momentum analysis was used to calculate annual energy production losses of 4.9% and 6.8% for a NACA 63 3 -418 turbine and an NREL S814 turbine, respectively, operating with 200 µm roughness. These compare well to historical field measurements. iii ACKNOWLEDGEMENTS

Wind turbines experience significant power loss due to insect contamination on the blades. In order to estimate power losses, computational fluid dynamics software or empirical models are used to compute drag increases due to roughness‐induced boundary‐layer transition. These models require knowledge of the expected levels and location of insect contamination on the blade surface. This is generally unknown, making power loss predictions unreliable. The present study develops a computer simulation to predict this information and uses this tool to simulate insect impingement for a variety of turbine operating conditions. The simulation code combines an invsicid panel method with a Lagrangian insect particle simulation module to first solve for the turbine velocity field and then track insect paths through this field. The effects of blade thickness, angle of attack and insect size are studied for 2D airfoil sections. The simulations show that increasing blade thickness leads to a larger chordwise extent of insect impingement yet lower maximum levels of contamination. Increasing insect mass is also found to increase the chordwise extent of impingement. These results are consistent with previous wind tunnel results and theoretical predictions. The model is then applied to a representative 5‐MW turbine model to determine spanwise variation in contamination. Results agree qualitatively with field observation, suggesting that this technique may be used in the future to more accurately predict power losses on turbines. Copyright © 2015 John Wiley & Sons, Ltd.