These vertical-axis turbines allow more efficient energy to be produced, in fewer square feet.

These vertical-axis turbines allow more efficient energy to be produced, in fewer square feet.

By Marcus Y. Woo,
Republished from Engineering & Science
Volume LXXIV, Number 2, Spring/Summer 2011



John Dabiri is a professor of aeronautics and bioengineering at Caltech.
John Dabiri is a professor of aeronautics and bioengineering at Caltech.
One day about five years ago, John Dabiri (MS ’03, PhD ’05) had a fishy idea. He was studying how air flows around solid structures—not unusual for an aeronautical engineer. In particular, he was trying to make wind turbines work efficiently amid the swirling gusts near buildings and skyscrapers, providing a source of renewable energy for cities. But as he played with the equations, he realized that they looked a lot like the ones that govern the flow of water through a school of swimming fish.

The arrangement of wind turbines is crucial for their efficiency, Dabiri says. Nature is often quite the engineer, and—mathematically, at least—the fluid dynamics around swimming fish are more or less optimized for efficiency. Once he saw the connection between fish schools and wind turbines, it seemed natural to put them together. Now, what began as a curiosity has become a new approach to wind power that offers a tenfold improvement over conventional wind farms.

These experiments will, for me, be the conclusive evidence that this approach works.
- John Dabiri, Professor of aeronautics and bioengineering

Because wind speeds are always changing, wind turbines produce only 25 to 30 percent of their maximum potential power output. But if every currently existing wind turbine were churning out as much power as possible, the United States would have the capacity to generate some 40 billion watts of wind power, which would account for 2 percent of the nation’s electricity. The maximum potential capacity of land-based wind power in the continental United States is estimated to be about 10 trillion watts, or terawatts (TW). Building wind farms on every suitable patch of land in the world could provide 75 to 100 TW. Considering that global power consumption was about 15 TW in 2008, wind could—in principle—power the entire planet.

The need for space

But one big problem with wind power is that conventional turbines—the ones that resemble huge propellers— need a lot of space. If these so-called horizontal-axis wind turbines are too close together, the wake behind the spinning blades interferes with adjacent turbines. To get the most out of each turbine, they have to be about 6 to 8 blade lengths apart and 20 blade lengths downwind of each other. With blades that can be 100 meters long, these turbines quickly occupy a lot of real estate.

Wind farms supply about 2.5 watts of power for every square meter of land. (See “Sustainable Energy—Without the Hot Air,” E&S 2010, No. 3.) If wind were to be the world’s sole source of energy, those wind farms would have to occupy a combined area equivalent to more than 60 percent of the United States. That’s clearly impractical, even without considering the minor difficulties: the wind doesn’t blow all the time, and some places can only muster a gentle breeze at best.

Wind power is generally considered a mature technology. In theory, wind turbines can convert 60 percent of wind energy into electricity. In practice, the best are already at 50 percent. But even though we seem to be pushing the limit, Dabiri is discovering that there’s still plenty of room for improvement.

Vertical-axis turbines are key

Dabiri’s fish-inspired wind farms use the lesser-known vertical-axis turbine, which looks a little like an eggbeater jutting out from the ground. When fish swim, they leave a horizontal row of regularly spaced vortices in their wakes; what would happen, he wondered, if he placed his downwind turbines in those vortices, and let them spin the turbines? In the spring of 2009, he assigned two grad students, Robert Whittlesey (MS ’09) and Sebastian Liska (MS ’09), to run a simple simulation of this arrangement as a class project. Astonishingly, they found that the turbines pumped out 10 times more energy per square meter.

“I play around with a lot of ideas, and the majority of them go to the scrap heap,” Dabiri says. “But after the students came back with such compelling results, I started to get excited that this could be a viable option.”

Individually, a vertical-axis turbine is less efficient than its monolithic cousin. But taken as a group, they can be positioned to squeeze as much power as possible from a given plot of land. Horizontal-axis turbines only capture the wind that blows through the circles swept by their blades, allowing precious energy to escape through the gaps between them. Verticalaxis turbines, on the other hand, can be bunched together until they’re almost touching, harnessing the energy of almost all the air that blows by.

At the beginning of 2010, Dabiri used some of his faculty start-up funds—which are provided to new faculty to build their labs— to buy a two-acre plot of land on the windy plains outside Lancaster, California. Here, at the Field Laboratory for Optimized Wind Energy (FLOWE), an array of half a dozen turbines has proven that Whittlesey’s and Liska’s results were right—and since then, the researchers have even improved on the fish-school models. “When we say we can increase the power output by an order of magnitude,” Dabiri says, “it’s not just a theoretical prediction.”

The key is that every turbine rotates in the opposite direction from its nearest neighbors. “That’s the secret sauce,” Whittlesey says. No one’s exactly sure why, but it may be that the opposing spins lower the local drag on each turbine, allowing it to whirl faster and generate more power.

Many advantages

Vertical-axis turbines have other advantages. They’re safer for birds. And instead of being 100-meter tall structures that would send Don Quixote into a tizzy, vertical-axis turbines are around 10 meters tall. Because they’re quieter and smaller, they can be distributed more widely and can be built closer to population centers. In fact, Dabiri is already working with the Los Angeles Unif ed School District to construct turbines at a new high school in San Pedro in 2012.

Other Caltech faculty members have gotten in on the action. Chemist Robert Grubbs is developing new materials to build stronger, lighter, and cheaper turbines and by manipulating structures at the nanoscale, Julia Greer is creating other materials for more durable blades. Aeronautical engineers Beverley McKeon and Mory Gharib (PhD ’83) are fine tuning turbines to control the airflow for maximum efficiency. And mechanical engineer Tim Colonius is running complex computer models of turbine wakes.

Meanwhile, the field tests continue. In one set of experiments, postdoc Matthias Kinzel is throwing fake snow into the whirling turbines. By taking pictures and video of the swirling flakes, he can measure exactly how the air flows and compare the physics with conventional turbines.

Even if Dabiri’s arrangements aren’t yet optimized, they’re still a vast improvement over the status quo—and more than good enough for commercial use. This summer, he’s building a few dozen more turbines at the test site, bumping the total to 42. “These experiments will, for me, be the conclusive evidence that this approach works,” Dabiri says. And there’s nothing fishy about that.