Hoover Dam rose out of the frenzied efforts to combat the Great Depression. Similarly, the fields of heat-focusing mirrors in the California desert known as the Ivanpah Solar Power Facility might come to represent the frenzied efforts to keep the Great Recession from turning into a full blown depression.
"In past recessions, we had great projects come out," said Michael Splinter, executive chairman of Applied Materials, at the inaugural Advanced Research Projects Agency–Energy summit in 2010. "We need a Hoover Dam of solar; maybe we just surround the Hoover Dam with solar? What are we going to point to in 30 years and say: That came out of the Great Recession of 2009?"
Perhaps the legacy of this recession will be something a little less physically imposing, however. Perhaps it will be ARPA-E itself, the only agency to come out of the 2009 stimulus effort, even if the fledgling organization has yet to fund the invention of a technology as world-changing as the Internet.
ARPA-E started in 2009 with a budget of $400 million, about one third of what its intellectual predecessor, the Defense Advanced Research Projects Agency (DARPA) got for its start in 1962. With ambitions to instigate a second industrial revolution, the agency received proposals for some 3,700 would-be world-changing energy technologies and handed out $151 million to 37 of them, ranging from turning water and CO2 into fuel with nothing but sunlight to better batteries. The largest single award, for $9.1 million, went to Foro Energy to help develop laser drilling that could make it cheaper to tap Earth's heat to generate electricity.
ARPA–E's offices within the U.S. Department of Energy (DoE) are meant to feel more like a Silicon Valley start-up than a part of a sclerotic bureaucracy tasked primarily with minding nuclear weapons and their legacy. ARPA–E staff, including directors, serve three-year terms. The short time frame is meant to inspire the "fierce urgency of now," a quotation from Martin Luther King, Jr. that hangs on the wall of DoE HQ, a concrete block building on stilts with row after row of box windows. The plan was to be a government agency that did not fear risk—a bureaucracy without bureaucrats. "We wanted to be measured in our craziness," Arun Majumdar, ARPA-E’s first director, told me in 2013 after he had left the agency. "It is early and you want to create a reputation of solid foundations in technology which are risky but not wacky."
The question, even back in 2010 at the inaugural ARPA–E summit, was whether any of the proposed innovations were truly game-changers. “Business as usual and the pace of innovation is just not fast enough," Majumdar said at that time.
Defense industry mandarin Norman Augustine, for one, felt that the agency represented an inflection point. The former chairman and CEO of Lockheed Martin, who has chaired innumerable government reports, including the one that birthed ARPA–E, and a master of escape velocities, may have felt that the agency represented a shift from old thinking to new, but most of the first 37 projects represented ideas already circulating for years, such as turning algae into fuel or using carbon fiber to improve fuel-efficiency in cars. Thanks to ARPA-E, they were being pulled off the shelf, dusted off and funded anew, perhaps because of the enthusiasm for "shovel-ready" stimulus projects.

Five of the first 37 funded projects aimed to develop a cheaper way to capture the CO2 spewed by the nation's hundreds of coal-fired power plants. Unsatisfied with so few projects for such a critical technology, however, Majumdar created an entire program—Innovative Materials and Processes for Advanced Carbon Capture Technologies, or IMPACCT for short—to add another 15 projects aimed at reducing the cost of capturing of CO2. "We need to develop technologies to do fossil fuels cleanly," Steven Chu, then secretary of Energy, told me in 2010. Of course, none have accomplished that goal, yet.
Other would-be "home runs," to use Majumdar and Chu's preferred baseball metaphor, included Sun Catalytix's attempt to power homes using only sunlight, chemistry and water:

An attempt to make a giant battery out of liquid metals inspired by the electricity-heavy process of making aluminum from bauxite that would spawn the company Ambri:
A better way to make solar cells by pouring silicon like pancake batter to make wafers rather than sawing wafers out of big ingots. That innovation came from a company named 1366 after the amount of sunlight in watts that hits each square meter of the planet:
ARPA–E lacks the budget to tackle something as sprawling as the grid or as expensive as building a new kind of nuclear reactor. The technologies it has funded that are most likely to succeed include batteries to store the energy generated by nighttime winds and better photovoltaics to turn sunlight into electricity. In just six years the agency has also created a new research community focused on using microbes to turn carbon dioxide—the leading greenhouse gas causing climate change—into fuel, although such electrofuels remain a long way from life outside the science lab.

By 2011, ARPA–E's investment in inventing the future of energy had spread to 121 projects in seven core areas with catchy acronyms, like Batteries for Electrical Energy Storage in Transportation or BEEST. "Just like Intel inside your laptops, I hope you have BEEST inside your electric cars in the future," Majumdar told the crowd at the agency's second summit. "How do we win the future? We invent affordable clean technology."
ARPA-E projects are rife with acronyms and out-there program names like electrofuels or REBELS (Reliable Electricity Based on ELectrochemical Systems). Majumdar himself helped come up with PETRO, or Plants Engineered to Replace Oil, in a flurry of e-mails over a weekend. But the agency has no direct ethanol program and Sen. Lamar Alexander of Tennessee (R), one of the political fathers of ARPA–E, used the summit stage in 2011 to call for the abolishment of never-ending subsidies for ethanol from corn and other "mature" energy sources like coal, gas and oil.
By 2011, Majumdar and Chu could extol the more than $200 million in private funding that followed ARPA–E's investment in certain of these technologies, such as Ambri's batteries or 1366's wafer-making technique. The Department of Defense became a major customer that year—if not the major customer—of some of these energy innovations, too. "Changing the way we produce and use energy is fundamentally about improving the national security of this country," said Ray Mabus, Secretary of the Navy in 2011, noting the Navy's history of fuel switches—from wind to coal in the 19th century and coal to oil supplemented by nuclear over the course of the 20th century. "I am confident—as we lead again in changing the way we power our ships and aircraft—that the naysayers who say, 'It's too expensive; the technology is not there,' are going to be proven wrong again."
Progress will more likely be measured in carbon dioxide molecules not emitted into the air and barrels of oil not imported but it might take as long as 20 years for any of that to not show up. DARPA took 10 years to lay the foundations of what would become the Internet, and it required several decades after that for the Web to conquer the world. In the absence of an energy tech equivalent of the Internet, ARPA–E has had to focus on some nearer term wins, like harnessing the flood of cheap natural gas released by hydraulic fracturing, or fracking. Even as early as 2011 Majumdar was scheming to tackle problems with natural gas. "We need this to be a bridge to somewhere rather than a bridge to nowhere," cautioned Ernest Moniz, the physicist who would succeed Chu as Energy secretary, at the 2011 summit. "Somewhere is zero carbon, which would be coal and natural gas with carbon capture and storage, and renewables."
Better batteries—a constant in ARPA–E funding—are also part of the mix, and the focus of the agency's highest profile failure: Envia. At the 2012 summit, Majumdar and others extolled the start-up, which was founded in a public library in Palo Alto, Calif., and one of the 37 initial awardees, for achieving power densities of 400 watt-hours per kilogram, results that were independently verified by Naval Surface Warfare Center in Crane, Ind. The battery maker signed a deal with GM to help power the Chevy Volt or another electric car but ultimately failed to deliver. In the same vein Sun Catalytix had to shift focus from the artificial leaf to flow batteries in order to survive, and several of the would-be CO2 capture techniques also dropped out. "I don't call them failures, I call them opportunities to learn," Majumdar told me in 2012.
On the other hand, Tesla’s Elon Musk used the ARPA–E stage in 2013 to foreshadow the repayment of his company's loan from the federal government nine years early and with interest. "The DoE loan to Tesla should be viewed as a pretty significant success," Musk said. "If people are going to attack DoE for Solyndra, for goodness sake, then there should be some praise for DoE with success."
The liquid-metal battery concept became a company known as Ambri, now churning out commercial batteries at its new factory in Massachusetts; 1366 also built a manufacturing facility in the state. Google has taken ownership of a bid to build kites from carbon fiber that would harvest the energy in the steady winds of the stratosphere, known as Makani Power after a Hawaiian word for wind.

Still, ARPA–E's current ambitions seems to have shrunk, from funding electrofuels to better air conditioners and better windows. "A 50 percent improvement in how much fuel you need to run an air conditioner is not incremental, but it is not sexy," admitted the agency's second director Cheryl Martin in 2013. "Once you demonstrate that it's possible, then the world changes."
These arguably unsexy wins serve perhaps to safeguard ARPA–E's current $300 million or so a year of funding, already a far cry from the $1 billion per year recommended by ARPA–E's intellectual founders or, more recently, the corporate heavyweights of the American Energy Innovation Council, including Augustine. All told, ARPA–E has invested a total of roughly $1.1 billion in more than 400 projects.
Even sustaining that level of support looks challenging. "Hard choices are going to have to be made about what types of energy investment provides the best return with the broadest impact," Randy Weber a Republican congressman from southeast Texas and chairman of the Subcommittee on Energy of the House Science, Space and Technology Committee, told the crowd at ARPA–E's most recent summit.
The vision of a clean energy future is little better than a mirage without the resources to invest in it. Without sustained support for innovation, when the next oil shock hits the U.S. we will be unprepared to deal with it—again. "Every time the price of oil goes up we panic, and when it goes down we hit the snooze button," Chu said back in 2011. "Let's take a longer term, more measured approach."
ARPA–E has not had enough time to be judged a success or a failure, although some are already willing to make that call. One is Fred Smith, CEO of FedEx, who at the 2012 summit said, "Pound for pound, dollar for dollar, it's hard to find a more effective thing government has done than ARPA–E."
The game is not changed, nor is the world. Global oil consumption exceeds 90 million barrels per day and civilization burns more than seven billion metric tons of coal each year—both numbers have grown in the short span of ARPA–E's existence. As a result, nearly 40 billion metric tons of CO2 are invisibly spewed into the air each year and its atmospheric concentration has hit levels not seen in the entire existence of our species: 400 parts-per-million. U.S. energy security has been delivered by fracking for oil and natural gas in North America as well as mandatory efficiency measures aimed at making cars and trucks burn less fuel per kilometer. Jobs in manufacturing continue to dwindle in this country although the general economy has recovered from the worst of the Great Recession.
The arc of an energy transition may be long, but it bends toward clean. The International Energy Agency noted that in 2014, for the first time in 40 years, pollution from the energy sector did not grow, even as the global economy did. That's primarily thanks to China burning less coal. And even though India hopes to burn more, ARPA–E may yet help with that. "Maybe we should have the ARPA–E exhibit in India?" Secretary of Energy Moniz mused on stage at this year's summit.
"An exhibit like this in India would be swamped," responded Ratan Tata, chairman of the Tata Group and a billionaire businessman who sells everything from trucks to tea. "People are hungry for ideas and they don't exist in India."
"Maybe we have an action item," Moniz said. "I think it could really have an impact."
That is also the hope of ARPA–E's new director, chemist Ellen Williams, formerly of the University of Maryland and oil giant BP. Her goal is to expand the impact of ARPA–E, whether that be improving the efficiency of the internal combustion engine, a cheaper battery being made in a garage in Harlem or even something weirder yet.

Martin, the agency’s previous director, has a different metric for impact. "In 2060 someone else will say if we were a success," she told me in 2013. Success may prove a long time coming but if it comes, it will come as a clean energy future that ARPA–E helped invent.