Interaction Between the Atmospheric Boundary Layer and Wind Energy: from Continental-Scale to Turbine-Scale

Wind turbines and groups of wind turbines, or “wind plants”, interact with the complex and heterogeneous boundary layer of the atmosphere. We define the boundary layer as the portion of the atmosphere directly influenced by the surface, and this layer exhibits variability on a range of temporal and spatial scales. While early developments in wind energy could ignore some of this variability, recent work demonstrates that improved understanding of atmosphere-turbine interactions leads to the discovery of new ways to approach turbine technology development as well as processes such as performance validation and turbine operations. This interaction with the atmosphere occurs at several spatial and temporal scales from continental-scale to turbine-scale. Understanding atmospheric variability over continental-scales and across plants can facilitate reliance on wind energy as a baseload energy source on the electrical grid. On turbine scales, understanding the atmosphere’s contribution to the variability in power production can improve the accuracy of power production estimates as we continue to implement more wind energy onto the grid. Wind speed and directional variability within a plant will affect wind turbine wakes within the plants and among neighboring plants, and a deeper knowledge of these variations can help mitigate effects of wakes and possibly even allow the manipulation of these wakes for increased production. Herein, I present the extent of my PhD work, in which I studied outstanding questions at these scales at the intersections of wind energy and atmospheric science.My work consists of four distinct projects. At the coarsest scales, I analyze the separation between wind plant sites needed for statistical independence in order to reduce variability for grid-integration of wind. Site data from three datasets spanning continents, durations and time resolution include 45 years of hourly wind speed data from over 100 sites in Canada, 4 years of five-minute wind speed data from 14 sites in the US Pacific Northwest, and one year of five-minute wind power generation data from 29 wind farms in southeastern Australia. I find similarities between these datasets in which correlations that fall to zero with increasing station separation distance, and the higher the high-pass cut-off frequency, the smaller the station separation required to achieve de-correlation. Shifting to atmospheric interaction on turbine-scales, I use 2.5 months of upwind tower and turbine data to understand how power production varies with different atmospheric stability and turbulence regimes. At lower wind speeds, periods of unstable and more turbulent conditions produce more power than periods of stable and less turbulent conditions, while at wind speeds closer to rated wind speed, periods of unstable and more turbulent conditions produce less power than periods of stable and less turbulent conditions. Using these new, stability- and turbulence-specific power curves to calculate annual energy production (AEP) estimates results in smaller AEPs than if calculated using no stability and turbulence filters, which could have implications for manufacturers and operators. In my third project, I address the problem of expensive power production validation. Rather than erecting towers to provide upwind wind measurements, I explore the utility of using nacelle-mounted anemometers for power curve verification studies. I calculate empirical nacelle transfer functions (NTFs) with upwind tower and turbine measurements. The fifth-order and second-order NTFs show a linear relationship between upwind wind speed and nacelle wind speed at wind speeds less than about 9 m s–1, but this relationship becomes non-linear at wind speeds higher than about 9 m s–1. The use of NTFs results in AEPs within 1 % of an AEP using upwind wind speeds. Additionally, during periods of unstable conditions as well as during more turbulent conditions, the nacelle-mounted anemometer u

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.