CHAMPAIGN, Ill. – On a gusty October day, I find myself staring at my wind turbine-simulation results. I take a sip of coffee and smile: The results validate a promising hypothesis.
My research focuses on the design and control of wind farms. Two key factors contribute to the power a wind turbine generates: the incoming wind velocity and how fast the turbine blades rotate. Each turbine hosts its own internal computer that measures wind speed and assigns it a specific operating torque. The computer is designed to maximize the individual turbine’s power output.
But my research objective is to get all the turbines in a wind farm to work together as a collective. Early works have shown that each turbine acting in its own best interest will actually lead to less power generation than control schemes that focus on the farm as a whole.
My hypothesis about how to do this arose unexpectedly one day as I was driving to Chicago to visit my fiancée. For some reason, my GPS chose to take me off the main highway and onto country roads, and I found myself traveling through a wind farm. It was a lucky coincidence: A thick mist lay on the horizon and, thanks to the fog, I could see the turbulence fields each turbine generated in its wake.
Buccafusca uses computer models to test how the behavior of individual wind turbines influences the overall output of a wind farm.
Photo by Fred Zwicky
Delete
Edit embedded media in the Files Tab and re-insert as needed.
I knew that turbines generate a disturbance field, and that the turbulence from one can affect the turbines downwind. But I had never thought of the turbines as being coupled together – the impact of upstream turbulence directly correlates to the power generation of downstream turbines. The fog allowed me to visualize this coupling explicitly.
Back in the lab, I was able to take this insight to build better control systems. Now I can run high-fidelity simulations to test – and in this case, validate – each new idea.
What began as just an idea blossomed into an exploration of different methodologies to lessen the effects of upstream turbines. I’m now exploring a technique called “wake steering.” By intentionally misaligning turbines, I can design a controller to skew the wakes to one side, even avoiding some downstream effects altogether! This means that the power of the wind farm as a whole can be improved without necessarily changing any of the external components.
I gaze down and see that my mug is nearly empty. I enter a new set of simulation parameters, hit “RUN SIMULATION” and go to refill.