An AI upgrade could improve wind turbine efficiency, study finds

A new artificial intelligence software upgrade for wind turbines has been developed that could improve efficiency enough to power Albania for a year.
Christopher McFadden
A minor AI upgrade could boost turbine efficiency.
A minor AI upgrade could boost turbine efficiency.


A new artificial intelligence has been developed that could boost wind turbine efficiency by around 0.3 percent, New Scientist reported. While this doesn't sound like a lot, if all wind turbines worldwide received the software upgrade, it would generate enough extra power to provide electricity to a modestly-sized country.

The main problem with wind turbines (besides their arguable sustainability) is that they need to face directly into the wind to provide maximum output. At present, this is very tricky to achieve as wind turbine nacelles are heavy and unwieldy things.

Most wind turbines align themselves with the wind through sophisticated control systems and aerodynamic design. The turbine's rotor, consisting of several blades, is mounted on a horizontal or vertical axis that can be adjusted to face the wind. Sensors, such as anemometers and wind vanes, detect the wind direction and speed, sending this information to the turbine's control system. Based on these inputs, the control system commands the yaw drive, which rotates the entire turbine nacelle to face the wind. This continuous monitoring and adjustment ensure that the turbine maximizes energy capture by maintaining an optimal alignment with the wind direction.

But this is not foolproof, and in most cases, wind turbines are rarely, if ever, positioned directly facing the prevailing wind. The new AI was built to improve this significantly by combining the wind turbines existing reactive systems with a reliable predictive element computed by the AI. Rolling out this technique globally could dramatically increase electricity production, amounting to 5 terawatt-hours per year. New Scientist points out that this is equivalent to the amount of electricity Albania consumes annually or 1.7 million typical UK households.

This is interesting and comes in light of other software improvements to wind farms that could improve efficiency by small increments. But, if implemented wholesale, those small savings all add up. For example, engineers from MIT developed a means of enabling wind turbines to work as a team rather than as individual units. Currently, most wind turbine farms are deployed so that leading turbines can affect others downwind from generated wind wakes.

By developing a new algorithm, they found that the increase in energy output from a given installation was about 1.2 percent overall and 3 percent for optimal wind speeds. The algorithm can be utilized at any wind farm, and due to the increasing number of wind farms being established to meet ambitious climate objectives, its deployment potential is expanding quickly.

According to the researchers, applying the 1.2 percent energy increase to all of the world's existing wind farms would be equivalent to adding over 3,600 new wind turbines. This would generate enough power for about 3 million homes and result in a total gain of almost a billion dollars per year for power producers.

Add Interesting Engineering to your Google News feed.
Add Interesting Engineering to your Google News feed.
message circleSHOW COMMENT (1)chevron
Job Board