An artificial intelligence (AI)-based tool has been developed by U.S. National Renewable Energy Laboratory (NREL) researchers to boost the efficiency and reduce the environmental impact of future wind farms.

The Wind Plant Graph Neural Network (WPGNN) was trained on simulations of more than 250,000 randomly generated wind plant layouts under various atmospheric conditions, plant designs and turbine operations. The method used these simulation inputs to determine the optimal design of a wind plant, with an emphasis on wake steering strategies.

The amount of energy a wind energy facility produces can be optimized by controlling the wake moving from an upstream turbine away from a downstream turbine. The WPGNN tool was used to evaluate the impacts wake steering would have on land use, cost and revenue.

The analysis considered a nationwide deployment of 6,862 plant buildouts in the U.S. with a cumulative 721 GW of generated power, with the goal of reducing 95% of carbon emissions from the energy sector by 2050. The wake steering scenarios described in Nature Energy were determined to reduce land requirements for future wind plants by 18% on average and by as much as 60% in some instances. Land savings total about 13,000 km², equivalent to 28% of the wind energy footprint nationwide.

In addition to reducing the cost of energy from wind systems, the AI-based program can help engineers design for a larger concentration of turbines in a smaller footprint. Such solutions provide increased flexibility from a site-planning perspective and deliver economies of scale for larger projects.