JHTDB-wind: sharing high-fidelity wind farm simulation data with the research community, and insights gained

The newly created JHTDB-wind (https://turbulence.idies.jhu.edu/datasets/windfarms) is a publicly accessible database containing large-eddy simulation (LES) fields from turbulent flows in wind farms. Such data are needed to elucidate complex flow physics such as wake evolution and entrainment, turbulence levels, and wind farm blockage effects. These flow phenomena determine wind power generation and structural loadings, and their understanding is vital in informing improved reduced models for farm level design and control. The database system builds on the framework of the Johns Hopkins Turbulence Database (JHTDB), which hosts DNS and LES datasets of canonical turbulent flows. JHTDB-wind stores the full space-time (4D) history of the flow and provides users the ability to access and query the data via web-based virtual sensor interfaces. A first dataset comprises LES results from a large wind farm with 6 x 10 turbines represented with the advanced filtered actuator line model, under conventionally neutral atmospheric conditions. This data set comprises one hour of flow field data (velocity, pressure, potential temperature, and others) stored every 0.5 seconds (i.e., 7200 3D fields are stored). Also stored are wind turbine data—including both turbine-level operational quantities and blade-level aerodynamic quantities. A second dataset is from LES of a 2 x 4 turbine wind farm evolving during a 24-hour diurnal cycle. A total of 172,800 3D fields is stored every 0.5 seconds. The presentation will discuss effects of added turbulence and flow blockage leading to increased power production in downstream turbines during early morning times in the diurnal cycle dataset. The presentation will also illustrate data access methods by visualizing temporal evolution of the pressure field for the 60-turbine wind farm.