Researchers from the University of Texas at Dallas, including G. Valerio Iungo, Romit Maulik, S. Ashwin Renganathan, and Stefano Letizia, have conducted a study to better understand and characterize the wakes generated by onshore wind turbines. Their work, published in the journal Applied Energy, focuses on the variability of mean velocity and turbulence intensity within these wakes, which are influenced by various factors such as incoming wind characteristics, turbine blade aerodynamics, and control settings.
Wind turbine wakes are areas of reduced wind speed and increased turbulence downstream of a turbine, resulting from the extraction of kinetic energy from the wind. Understanding and predicting the behavior of these wakes is crucial for optimizing wind farm performance and minimizing the impact of turbulence on turbine components. The researchers used LiDAR (Light Detection and Ranging) measurements to capture detailed data on the wake velocity fields generated by isolated wind turbines at an onshore wind farm.
To analyze the extensive dataset collected by the LiDAR, the researchers employed a machine-learning technique called proper orthogonal decomposition (POD). This method reduces the dimensionality of the data by identifying the most significant patterns or modes within the wake velocity fields. By focusing on just five key POD modes, the researchers could effectively capture the variability of the wake characteristics without being overwhelmed by the sheer volume of data.
Next, the researchers applied a clustering algorithm known as k-means to group the LiDAR data into distinct categories based on the coefficients of the POD modes. This approach allowed them to identify the most representative examples of wind turbine wakes without relying on predefined thresholds for wind and turbine parameters, which can sometimes introduce bias. By analyzing the clustered data alongside corresponding SCADA (Supervisory Control and Data Acquisition) and meteorological data, the researchers could study the variability of wake velocity deficit, wake extent, and wake-added turbulence intensity under different conditions.
The findings of this study contribute to a better understanding of how wind turbine wakes behave under various operating conditions and atmospheric stability regimes. This knowledge can be applied practically in the energy sector to optimize wind farm layouts, improve turbine control strategies, and enhance overall wind farm performance. By minimizing the negative impacts of wakes, such as reduced wind speeds and increased turbulence, wind farm operators can maximize energy production and extend the lifespan of turbine components.
In summary, the researchers from the University of Texas at Dallas have demonstrated the value of machine-learning techniques in analyzing and characterizing wind turbine wakes. Their work provides valuable insights for the energy industry, paving the way for more efficient and effective wind farm operations. The research was published in the journal Applied Energy, offering a robust foundation for further studies in this critical area of wind energy research.
This article is based on research available at arXiv.