Skip to Content

Can ARPA-E Solve Energy Problems?

The young agency is popular, but its short-term research programs aren’t enough.
March 5, 2012

Republicans and Democrats in Congress don’t agree on much, especially when it comes to the U.S. Department of Energy, but they agree that the department’s Advanced Research Projects Agency for Energy (ARPA-E), is a good thing. Last year, when they were cutting every program in sight, they actually voted to increase the agency’s funding by 50 percent. The bipartisan support was clear last week at the agency’s third Innovation Summit, attended by a mix of liberal and conservative politicians and business leaders.

ARPA-E has been popular in large part because it’s inexpensive—for about the same amount the government gave to failed solar-panel maker Solyndra in the form of a loan guarantee, ARPA-E has funded 180 projects. But how important is the agency for solving major energy challenges like volatile oil prices and climate change? At the summit, several speakers warned that the sort of short-term, two-to-three-year funding that the agency provides isn’t enough to address long-term energy problems.

From its beginning, ARPA-E has set relatively modest goals for itself, acknowledging its limits.

“In the energy sector, it takes time for an innovation to go all the way and scale and make a big difference in the commercial market,” said Arun Majumdar, ARPA-E’s director, in a press conference. “That takes about 10, 15, maybe 20 years. Who knows?” He pointed to other measures of success. The agency has helped increase private investment in energy, he said, noting that 11 of the projects ARPA-E funded—with about $40 million—led to more than $200 million in private-sector funding. He also noted the success of Envia, which has demonstrated a large increase in battery-storage capacity.

During a panel discussion with Energy Secretary Steven Chu, Microsoft founder and chairman Bill Gates warned that energy innovation moves a lot slower than innovation in software. “The IT revolution is the exception that kind of warped people’s minds about how quickly things can work,” he said. “If you underestimate how hard it is, that’s part of why we can end up underfunding the kind of innovative work that needs to go on.” He said that energy innovations in the past have taken 50 to 60 years to make an impact.

Chu said that ARPA-E fills a specific role—in some cases, innovations “can find a way to market very quickly.” But he said that there is still a need for DOE’s longer-term research funding—including Innovation Hubs that focus on a problem for a decade—as well as financing for larger projects, such as loan guarantees.

But some of these larger and longer-term projects are under fire, as can be seen in the congressional investigation of the DOE’s loan to Solyndra. “No one can deny that the reaction to Solyndra has had a damping effect on government finance programs for companies,” Chu said. But he also said that failure was expected. Congress knew there would be failures when it authorized the loan program; that’s why it appropriated $10 billion to cover losses. “It’s extremely unfortunate what happened with Solyndra—a half a billion dollar loss,” he said. “But I would be personally very surprised if we were to lose a third of that appropriated money.”

Gates expects the failure rate for energy innovation funding more broadly will be “well over 90 percent.” And he said there’s “no clear mapping between the amount you spend on R&D and the amount you get out”—it’s possible that the innovations will come from companies that are already well funded now. “But it’s more likely that the underfunding is delaying the rate of progress,” he said. “This is a very complex set of technologies, and so we need literally thousands of companies trying these things to increase the odds that we will have the 10 or 20 approaches that will get us the magic solution.”

Keep Reading

Most Popular

Large language models can do jaw-dropping things. But nobody knows exactly why.

And that's a problem. Figuring it out is one of the biggest scientific puzzles of our time and a crucial step towards controlling more powerful future models.

OpenAI teases an amazing new generative video model called Sora

The firm is sharing Sora with a small group of safety testers but the rest of us will have to wait to learn more.

The problem with plug-in hybrids? Their drivers.

Plug-in hybrids are often sold as a transition to EVs, but new data from Europe shows we’re still underestimating the emissions they produce.

Google DeepMind’s new generative model makes Super Mario–like games from scratch

Genie learns how to control games by watching hours and hours of video. It could help train next-gen robots too.

Stay connected

Illustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at with a list of newsletters you’d like to receive.