Vestas uses supercomputers to improve efficiency of wind power
Vestas Wind Systems is looking to improve the efficiency of its wind farms by using supercomputers to optimize site location and operation, with a goal to make wind power cost-competitive with fossil-fuel power generation. The Danish energy company partnered with IBM to use Firestorm supercomputers to analyze petabytes of data, including weather reports, tidal phases, geospatial and sensor data, satellite images, deforestation maps, and weather models, to improve site selection and optimize power output over the lifetime of the turbines.
Since turbines are stationery and operate for decades, even small improvements in daily performance can result in large efficiency gains over the life of the system.
Vestas hopes the performance improvements will drive greater adoption of wind farms over conventional power plants that burn coal, natural gas and diesel fuel. The environmental impact of raising use of regenerative wind power could be significant over time. According to the American Wind Association, increasing wind capacity to 20 percent by 2030 could reduce CO2 emissions by 7,600 million tons in the U.S. It also has the potential to reduce water consumption in the electrical sector by four trillion gallons and lower consumer demand for natural gas by 12 percent.
Once turbines are installed, the computers will be used to predict performance, tweak aerodynamics and determine schedules for taking turbines offline for maintenance. Firestorm supercomputers are capable of performing 150 trillion calculations per second – enough compute power to support Vestas’ future plans to analyze more than 20 petabytes of diverse weather data per project within four years.
Image credit: Vestas