Tomorrow’s turbines, today, on SBU supercomputers

Now you see it: A turbulent wind field (red is faster winds, blue slower) churns around an offshore wind farm, as numerically simulated by Stony Brook University supercomputers.
By GREGORY ZELLER //

Supercomputers, artificial intelligence and next-generation turbine controls will work together to design the offshore wind farms of the future – a key step for the burgeoning national industry, and for Long Island’s emergence as a national wind-power leader.

Stony Brook University has landed a $1.1 million award from the National Offshore Wind Research & Development Consortium, earmarked for a College of Engineering and Applied Science study that will unite advanced technologies to create “high-fidelity simulations” of future offshore wind farms – perchance to determine optimal layouts, maximizing both output and efficiency.

With Long Island jockeying for position in the national offshore wind-a-thon, including the creation of a $20 million industry “training institute” and multiple offshore projects in the works, these are critical factors. And Fotis Sotiropoulos, dean of SBU’s engineering college and principle investigator of the study, aims to get them right.

His cutting-edge approach, incorporating parallel supercomputers and other advanced computational sciences, has already proven effective. Applying the algorithms – which consider “turbine wakes” and a host of other variables not included in simpler (and more common) engineering models – to land-based wind farms, Sotiropoulos previously prognosticated a better than 10 percent reduction in overall energy costs.

Fotis Sotiropoulos: First by land, now by sea.

“Our initial findings for land-based wind farms indicate that simulation-driven approaches, combined with advanced turbine controls, can result in levelized cost of energy reduced by more than 10 percent,” Sotiropoulos said. “I am confident that our research will help achieve similar results for offshore wind farms.

Or maybe even better: Factoring in changing wind resources, cutting-edge turbine controls, “sea states” and other advanced metrics, the SUNY Distinguished Professor of Civil Engineering’s AI-powered supercomputers could improve hypothetical wind-farm energy costs by as much as 15 percent, according to SBU, mostly by rearranging proposed assets.

The new research, including contributions from the University of Minnesota and a Texas-based satellite of New Mexico’s Sandia National Laboratories, could quickly prove useful beyond New York waters. Sotiropoulos et al plan to create industry-friendly simulation tools that can be customized to specific sites – a control co-design tack never before applied to offshore wind farms.

Carrie Bullen Hitt, executive director of the circa-2018 National Offshore Wind RDC, said the advanced computer simulations and their potentially universal application were right in the wheelhouse for the U.S. Department of Energy-funded, U.S. Department of Interior-directed, New York State Energy Research and Development Authority-administrated consortium.

“The consortium is working with partners like Stony Brook all across the nation to bring down the cost of developing this game-changing renewable resource while overcoming domestic market challenges,” Hitt said in a statement. “We look forward to working with Fotis Sotiropoulos and his team at Stony Brook on this innovation project.”