Many energy technologies that could prove transformative over the coming decades begin as ideas or testing too uncertain for private investment. The Advanced Research Projects Agency-Energy (ARPA-E) was built to fill that gap. Created in 2009 and modeled on the Pentagon’s well-known Defense Advanced Research Projects Agency (DARPA), the agency funds high-risk, high-reward research and asks one question of every idea: If it works, will it matter?
The agency’s design reflects that uncertainty. Technical experts run programs on three-year terms, projects answer to milestones, and off-ramps exist for projects that stall. Its portfolio is broad by design, from agriculture and buildings to grids and transportation, but it focuses narrowly on the earliest, riskiest stage of development.
I recently co-authored a paper with Claire C. Cody of Clean Tomorrow examining what those bets have delivered. The broader evidence is encouraging: a recent National Academies analysis found that roughly 40 percent of ARPA-E awards were followed by a significant rise in patenting in the same technology area, a sign that the agency draws others into the fields it helps open. Our case studies offer some anecdotal confirmation of how that happens in practice.
Houston-based geothermal startup Fervo Energy, for example, set out to borrow the horizontal drilling and fracturing techniques of the shale boom to engineer next-generation geothermal wells. It lacked a way to see how water moved through fractures deep underground, and in 2019, a $1.2 million ARPA-E award helped it build that sensing technology. Fervo developed most of its core technology with private money that has enabled it to construct a multi-billion-dollar orderbook today, while the agency funded a missing catalytic component.
The same approach reaches well beyond geothermal. In 2012, ARPA-E ran its open call for ideas outside its standing programs. HexaTech, a small North Carolina manufacturer of aluminum nitride, a semiconductor material that handles higher voltages and temperatures than silicon, decided to apply. ARPA-E awarded the company $2.8 million in 2012 to research and develop a novel power switching device, ultimately expanding the knowledge base for next-generation high-voltage semiconductor designs. In both cases, a hard technical problem stood between a promising idea and a viable product, and that is where ARPA-E entered the picture.
The point is not that every bet pays off. Breakthroughs rarely come from the safest or most straightforward research. Modest, targeted public investment at the riskiest stage can clear a path, and private capital can take it from there. ARPA-E’s best investments follow that model, and its value depends on staying focused on that gap.
