Google Offers $1M Prize for Smaller, Smarter Inverters

Will the competition move the needle on power electronics for solar and data centers?

Photo Credit: XONOVETS via Shutterstock.com

While President Obama visited Silicon Valley to announce billions of dollars' worth of new green energy initiatives last Friday, local green booster Google dropped news of its own $1 million commitment to a technology not often associated with search engines and data centers: solar inverters.

Google’s Little Box Challenge is looking for inverters that come in much smaller packages than today’s models, according to its rather information-sparse launch page and blog post. Google isn’t giving any more specifics before the project's summer launch, a spokesperson told us Monday.

That gives would-be contest participants little to work with, beyond the fundamentals. Inverters use metals and semiconductor materials to convert direct current (DC) to alternating current (AC). Along with the cooling equipment and enclosure, this bill of materials can make up roughly one-third of the cost of each device. The concept of a smaller inverter implies the use of less of that stuff, and thus a lower cost to manufacture the devices, the Google representative said.

Solar inverter makers have been working to boost power conversion efficiencies as high as possible, squeezing every last drop of billable generation out of solar projects and reducing maintenance issues associated with cooling these high-voltage devices. Best-in-class devices now achieve 96.5 percent to 98.5 percent efficiency, according to MJ Shiao, director of GTM's solar research.

"It's really difficult to squeeze out that last little bit of efficiency when you're already at 98 percent," explained Shiao.

As for materials costs, inverter makers are making progress, and new semiconductor materials and designs are helping power electronics achieve better performance in smaller packages. However, most of the improvements are incremental. “If anyone has something revolutionary, they haven’t shown it in their product roadmaps,” said Shiao.

Fundamental design concepts might have to change to achieve the size reduction Google appears to be asking for, he noted. (For more on small-scale solar inverter technology, read GTM Research's report, The Microinverter and DC Optimizer Technology Landscape, 2014.)

Google has also invested in Transphorm, an ARPA-E-funded startup that’s using gallium nitride as a replacement for silicon in all manner of power electronics components. “Transphorm is interesting because it is more on the power semiconductor side, [focusing on] power conversion in general,” Shiao said. “It isn't an inverter manufacturer, but it is working with inverter manufacturers.”

GaN appears more suited to lower-power applications, which could qualify it for microinverters, but not central solar inverters. Silicon carbide is a more promising material for high-power applications, however. Some European inverter manufacturers are using silicon carbide diodes, and redesigning an entire system around the material could reduce the bill of materials by as much as 30 percent, said Shiao.

“The biggest question is how much of this is focused on solar inverters,” Shiao added, alluding to the fact that inverters are used in a variety of applications, not just solar.

For Google, “the one that seems to make the most sense is to have more efficient and cost-effective inverters for its data centers,” he said. We’ll await word from Google on just which applications the company is envisioning for these small inverters.

Google has invested more than $1 billion in wind and solar power projects, whether as a developer or as a purchaser of green energy. Now, there is some evidence that Google is moving deeper into solar, as the company hires people with experience in power electronics, noted Shiao.

Still, $1 million in prize money “is small change compared to R&D in the field right now,” he said. Last year, U.S. microinverter maker Enphase spent $34.5 million on R&D, and German inverter giant spent €102.5 million ($140 million).

Even if $1 million does bring a radical new innovation, product development and commercial deployment are still a challenge.

The big question, said Shiao, is whether it’s wise to concentrate on the size of the inverter, rather than on what it’s capable of doing. Inverter materials now make up less than 10 percent of overall solar system costs, meaning that any innovation in this area would likely have a negligible impact on overall cost. Adding smart features and functionality to inverters, on the other hand, could lower integration and installation costs, improve monitoring and maintenance, and lead to a slew of potential revenue-generating opportunities in grid balancing over time.

For now, we can only speculate what Google is planning. We'll find out more this summer.