From reanalysis data to park optimization

摘要

In this work, we use the commercial WindSim CFD software to generate realistic wind resource climatologies (i.e. maps) for seven operational wind farms across three continents. The wind resource maps are generated using only atmospheric data provided by a global reanalysis dataset and are then used to estimate the actual power production of each wind farm's existing turbine layout. Then, using a park optimization module along with a basic suite of construction constraints, we estimate the most productive microsites within the wind farms and compare the optimized energy production with the original energy production estimate. Results yielded from this experimentation emphasize that park optimization efforts can generally lead to annual energy production increases of 3 to 6% and higher in some cases. On the other hand, challenges associated with imposing accurate optimization constraints were found to be significant.