Johns Hopkins University has recently released a free, comprehensive, public wind farm database developed through the university’s Ralph O’Connor Sustainable Energy Institute (ROSEI).

The Johns Hopkins Turbulence Databases – Wind (JHTDB-wind) provides massive amounts of data from computer simulations of turbulent flow in wind turbine arrays to help in wind farm design, analysis and operational increases to power output. This new publicly available resource supports everything from academic studies to practical wind farm planning and development projects.

Dennice Gayme and Charles Meneveau

“One of the problems in this field has been that people run massive computer simulations of turbulent flow in wind farms; a couple of papers get written by the people doing the simulations, and little else is done. These datasets are so large that it is difficult to share with others via traditional methods, limiting the potential broader impact of the data,” said Charles Meneveau, the Louis M. Sardella Professor in Mechanical Engineering and an associate researcher with ROSEI. “What we are doing instead is storing the data in a database, so that anyone can access it through the JHTDB-wind website using tools that we have developed to enable the data to stream directly into analysis programs. We’re hoping this will help bridge the previous gaps in shared knowledge that have plagued this research area.”

JHTDB-wind was created in close collaboration with the Institute for Data Intensive Engineering and Science (IDIES) and the access tools leverage those currently in use to access over a petabyte of data for fundamental studies of turbulent flows.

JHTDB-wind is currently comprised of two datasets. The first, titled “LES of large wind farm under conventionally neutral atmospheric conditions,” simulates a large wind farm under conditions where temperature differences across different heights in the atmosphere are negligibly small such as during sunset or on cloudy days. It functions as a baseline in terms of the atmospheric conditions.

“This set serves as a good general baseline so users can see how wind turbines in a big farm interact with each other under minimal effects from atmospheric stability conditions,” said Dennice Gayme, a professor of mechanical engineering and a core researcher with ROSEI.

The second dataset, titled “LES of large wind farm during a diurnal cycle,” simulates a smaller wind farm over 24 hours. During daylight, the sun heats the ground, creating convective conditions with rising air plumes that generate turbulence and interact with wind farms. At nighttime, temperatures drop, causing colder air to drift downward, reducing turbulence.

“If someone is particularly interested in a warm or cold part of the day, they can pinpoint those times and find the information they want,” Meneveau said.

The next datasets from JHTDB-wind are expected to be for offshore wind farms

Meneveau and Gayme are collaborating with Julie Lundquist, the Bloomberg Distinguished Professor of Atmospheric Science and Wind Energy, to add datasets that involve offshore wind farms, which bring complicating factors such as wave motions, and possibly turbine motion if they are placed on floating platforms. They will also explore the effect of very large turbines that reach high up and span various regions of the atmosphere.

“Including offshore options is essential for the next generation of wind farms because there is so much potential for wind and space in open waters,” Gayme said. “Charles, Julie and I are still very early in creating the set for that one, but the work is being supported by a new ROSEI SPARK award, and we are considering including a dataset that mimics conditions of a specific site off the coast of New England that has been targeted for a potential offshore wind farm.”

Meneveau compared JHTDB-wind’s place in wind farm flow analysis to that of streaming websites like Spotify in the music industry. Before the internet, people bought physical records and CDs to access music and would have to share physical copies with others. This was followed by the practice of downloading separate music files onto computers. These more cumbersome sharing practices have shifted to the current norm of directly streaming music on demand via user-friendly applications without the hassle of file transferring, management, and formatting.

JHTDB-wind, similarly, enables people to study wind farm flows and run small programs that calculate averages or extreme conditions, plotting data points without needing to keep data locally or downloading files.

“Just having data is not useful if you cannot access it, so the big innovation here is that you can easily play with the data,” Gayme said. “You aren’t just streaming it; you can work with it to learn more about whatever it is that you’re studying about wind farms. You can use the data on JHTDB-wind as your own little ‘remote’ lab.”