An experimental investigation of wind turbine aerodynamic interaction
Abstract
Wind turbine installed capacity in the United States has seen an exponential growth over the last decade and mostly coming in the form of large wind farm installations. The wind farms themselves too have been increasing in size, incorporating more wind turbines in larger areas than ever before. Wind turbines have become a viable component in the overall energy makeup of the United States due to improved economics where energy prices have risen and production costs dropped. For a fixed cost, the effectiveness of a wind turbine financially is highly related to its performance. Considering the size of current wind farms, a minor performance improvement will result in large additional sums of revenue. A problem that has received attention with wind farms is that while the fixed costs of the development do get spread out further to reduce the installed cost of each wind turbine, the wind turbines have been observed to perform at lower performance values in the wind farm setting. Development work is performed to predict maximum theoretical levels of performance describing a wind turbine. This work is extended to include predictions of the limiting power for a two-rotor, counter-rotating wind turbine configuration. Wind farm performance losses are also modeled for the dominant modes of interaction when operating a wind turbine within the wake of upstream machines; covering single-, multiple-, and lateral-wake scenarios. A year's worth of wind speed data are analyzed to reveal seasonal trends of the wind turbine input condition. Wind turbine performance is simulated using this data and compared amongst small and large wind turbines. Predictions of wind farm wake models are compared to data generated using the Purdue University Micro Reconfigurable Wind Farm facility. This facility contains four small wind turbines in an in-field experimental setting where wake scenarios are created and performance comparisons measured. Model validation is obtained using the experimental results which provides insight into the model's assumptions' range of effectiveness, and resultant predicted wake behavior. The physical mechanisms of wake operation power losses are also observed from the data showing the relative contribution of the wake loss constituents in different wind farm configurations.
Degree
Ph.D.
Advisors
Fleeter, Purdue University.
Subject Area
Mechanical engineering|Energy
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.