General Electric's wind energy division is trying to find a "Goldilocks" turbine design, one that's not too big and not too small.
Like other wind manufacturers, GE is benefiting from booming demand for wind turbines in Europe and in the U.S., even with the possibility of a renewable-energy tax credit lapsing later this year.
Still, the rapid expansion is being throttled by high prices of steel and other commodities, making wind power more expensive.
So instead of making bigger and bigger machines, manufacturers are trying to squeeze more energy from conventionally sized wind turbines.
That will be done by using alternative materials, better electronics, and shaping turbine blades to better capture the wind, said Stephane Renou, who manages research and development for General Electric's wind technology platform.
"The optimal point is changing...and going bigger is not the answer," Renou said. "Turbines in the two or three megawatt zone are the most efficient and the best cost per kilowatt."
A 2 megawatt or 3 megawatt wind turbine is still large. The tower on a 2.5 megawatt machine can stand nearly 330 feet high.
But making 5 megawatt or 6 megawatt turbines, as some off-shore turbine makers are doing, also means more raw materials, notably steel, which drives up the cost. The logistics of delivering and assembling these massive components can add to costs as well.
To get more power from the same footprint, GE is looking at a variety of technologies, said Renou, who oversees development at four GE wind research centers around the world.
"I see a lot more technology going into each of the components, especially the blades," he said. "The blades will look more funky and twisted to get better performance."
By adding more carbon composite to turbine blades, GE can add 16 feet to their length, which translates into a significant boost of energy.
Although it's four times more expensive than fiberglass, carbon also gives blades more flexibility, allowing them to operate at higher wind speeds.
GE is also working on electronics controls to optimize performance and sound level of entire wind farms, Renou said.
Limits in transmission line capacity are a barrier to both wind and solar energy. Wind farms and solar plants are typically best placed in remote areas, far from the areas on the coast where demand for electricity is highest.
One of GE's research teams is developing software for modeling how to best place turbines in a wind farm while another is working on the electronic controls to get wind power fed into the grid most effectively.
Overall, Renou said that wind technology is developing quickly and is getting more competitive on a price-per-watt basis with natural gas generators, which themselves are going up in price.
But perhaps just as significant, having a range of materials and technologies to work with gives GE more flexibility in how to assemble a turbine. Supply chain disruptions have contributed to product shortages and project delays.
"We're working on supply-chain flexibility by providing different technology options," Renou said. "It's all about having options at this point. We will structure things to have all the raw materials at the right costs."
One area that GE's wind labs is not pursuing aggressively is energy storage. A handful of companies and utilities are looking at truck-sized batteries or underground compressed-air storage to incorporate renewable energy more reliably.
But Renou said that storage attached to wind turbines is not likely to happen in the next two years. Instead, beefed-up transmission lines, along with smarter power-grid management, could push wind to make up 10 percent of power generation, up from less than 1 percent now.
"The grid is a fantastic source of energy storage. Wind variability can be handled by the grid and grid management," he said. "It's more about policy and grid development."
Update at 3:22 p.m. PT: Text of first caption corrected.