The wind energy sector is a frequent user of high-power simulations, with researchers aiming to optimize wind flows and energy production from the massive turbines. Now, researchers at GE are preparing to undertake a large wind energy simulation effort on the Summit supercomputer, aiming to advance the potential of offshore wind. Led by Jing Li, a research aerodynamics engineer at GE, the researchers will use Summit to conduct simulations hitherto computationally infeasible.
Despite being recently dethroned by RIKEN’s Fugaku system as the most powerful publicly ranked supercomputer as of the most recent Top500 list, Summit remains the runner-up, weighing in at a whopping 148.6 petaflops of computing power (and over 340 tons of iron). Summit’s 4,608 nodes — each equipped with two IBM Power9 CPUs and six Nvidia Volta GPUs — are supported by 250 petabytes of storage and Mellanox InfiniBand EDR 100 Gbps networking. And at 13 MW, Summit’s power footprint is on par with the peak output of around four or five onshore wind turbines.
“The Summit supercomputer will allow our GE team to run computations that would be otherwise impossible,” Li said. “This research could dramatically accelerate offshore wind power as the future of clean energy and our path to a more sustainable, safe environment.”
Specifically, the simulations will examine coastal low-level jets, which produce distinct wind velocity effects crucial to operating and optimizing offshore wind turbines. Offshore turbines, of course, face much different challenges and opportunities than onshore turbines: the winds are often much stronger, but the wind behaves differently and the tumultuous environment presents particular dangers.
A video of one of GE’s wind simulations from a previous project.
“We’re now able to study wind patterns that span hundreds of meters in height across tens of kilometers of territory down to the resolution of airflow over individual turbine blades,” Li said. “You simply couldn’t gather and run experiments on this volume and complexity of data without a supercomputer. These simulations allow us to characterize and understand poorly understood phenomena like coastal low-level jets in ways previously not possible.”
This undertaking will happen in collaboration with the ExaWind project at the National Renewable Energy Laboratory (NREL), which is aiming to bring wind energy simulation into the exascale era. “Scientists at NREL and [Oak Ridge National Laboratory] are part of a broader team that have built up a tremendous catalog of new software code and technical expertise with ExaWind,” Li said, “and we believe our project can discover critical new insights that support and validate this larger effort.”