Crossing the “Valley of Death”: How to Design and Run Successful Clean-Energy Demonstration Projects

Transformational clean-energy innovations are required to achieve the nation’s economic, environmental, and national security goals. Smart grids that can integrate massive distributed resources, power plants that can capture and sequester carbon emissions, and other advanced technologies must be demonstrated at scale before they can be fully commercialized. Public-private partnerships are needed to cross this “valley of death” between prototype and commercialization and strengthen investor confidence in the affordability, reliability, and practicality of such innovations. But demonstration projects pose many policy and management challenges, and the historical record of the Department of Energy running them is uneven. On July 26, the Information Technology and Innovation Foundation (ITIF) hosted a panel to discuss a new report assessing recent federal efforts to overcome these challenges and consider how this record might be extended and improved upon in the future.

After welcoming remarks by William Bonvillian, the former director of MIT’s Washington office and the event’s moderator, David Hart, senior fellow at ITIF and lead author of the report, opened the event by underlining the necessity of clean-energy demonstration projects. “If we’re going to have large-scale change in our energy system, as many of us believe is necessary to address the challenges of our time, especially climate change, we’re going to need more demonstration projects,” Hart said. The conventional wisdom holds that “big demonstration projects in this country are doomed to failure basically for political reasons, that once the money starts flowing to a particular locality, regardless of what the technical results are, it’s very hard to pull the plug on it.” But successful projects like the Petra Nova plant appear to contradict this narrative, Hart argued. Indeed, he continued, demonstration projects are necessary because the sheer complexity of energy systems means that not every new system can succeed—viable technologies may not make it to widescale deployment if private actors have doubts about a new clean-energy technology.

Hart outlined several of the recommendations in his paper, including creating a robust demonstration project portfolio, improving information-sharing, placing private actors at the forefront of public-private partnerships, accommodating projects’ varying risk profiles with a wider range of cost-sharing ratios, and being prepared to terminate non-viable projects. Given the difficulty of setting up an organization with the capabilities of the Department of Energy (DOE) and the findings in the paper suggesting that DOE can be effective at running demonstration projects, Hart concluded that DOE’s leadership in this area is crucial.

Joseph Hezir, principal at the Energy Future Initiative, echoed Hart in arguing that “there really is an important federal role in the demonstration stage of the [energy] innovation process. But that role really should be one that’s defined and determined on an analytical process and not purely on an ideological process.” Hezir continued with some recommendations building on the suggestions in the paper for improving the clean energy demonstration project process. To begin, Hezir warned that demonstration projects should be designed with market risk and policy risk in mind, citing as a cautionary tale the Clinch River Breeder Reactor, which faced policy shifts on uranium reprocessing rules as well as market shifts that made reprocessing less economically attractive. “In thinking about what a future demonstration program might be,” Hezir stated, “we need to think very carefully about funding certainty.” He concluded by calling for a single organization inside DOE to coordinate demonstration projects regardless of what type of power source they use, encouraging the creation of “exit strategies” to pull support for nonviable projects, and reiterating Hart’s call for follow-on support through tax credits or guaranteed government purchases.

Christopher Smith, Baker Institute Advisory Board Fellow in Energy Studies at Rice University, followed Hart and Hezir. Smith previously served as the assistant secretary for fossil energy at DOE, where he oversaw the department’s fossil energy R&D program and carbon capture and sequestration (CCS) projects. He discussed his experience at the Kemper Project, a CCS-enabled coal plant, which despite running into the industry-wide “buzzsaw” of plunging natural gas prices provided valuable experience and data in CCS technologies. “You walk around the site with the engineers who built that plant, and you’re there with your boots on and your hard hat,” said Smith, “and they can point to ten things they’ll do next time to peel off 5 percent here, 10 percent here, 30 percent here. So, by building these projects, you get better at doing them.” Coal plants with CCS face a challenge that wind, solar, and other renewables do not, said Smith, in that CCS coal plants have to compete with non-CCS coal plants under a “regulatory scheme where you can emit as much CO₂ in the environment as you want to and it’s essentially free.” He concluded that demonstration projects like Kemper are crucial to drive down the cost of emissions-mitigating technologies like CCS so that they will be competitive if and when the United States does put a cost on carbon.

Samuel Thernstrom, founder and executive director at the Energy Innovation Reform Project, rounded out the panel by emphasizing the importance of demonstration projects to bringing transformative energy programs to maturity. To solve for some of the weaknesses of the demonstration project process, Thernstrom reiterated other speakers by calling for follow-on support with large-scale purchase agreements from giant agencies like the Department of Defense. He critiqued the paper by arguing it underestimated the challenges of “insulating these programs from politics… let’s work on it, but let’s keep our expectations in that area modest.” Echoing Hart’s observations that replicating DOE’s capabilities would present an overwhelming challenge, Thernstrom advocated for reforming existing institutions, like DOE and the National Labs, rather than inventing new structural arrangements to separate demonstration projects and politics. Thernstrom also argued that our definitions of success and failure in evaluating demonstration projects may be too limited, asking “should we think of [the Synthetic Fuel Corporation] today as a terrible pork barrel, or as a project that actually advanced a lot of important science?” He concluded by arguing for market-based programs like reverse auctions, in which companies seeking to demonstration a new clean energy technology would bid to show the least need for public support in a cost-sharing arrangement.

Transforming the U.S. energy sector will require the ingenuity of the market and the risk-absorbing capacities of the federal government. Clean energy demonstration projects are a case-in-point; without them, no amount of basic R&D funding will convince a company with a steady income stream to be a first-mover. Given the broader social, environmental, and economic benefits of energy innovation, the valley of death represents a textbook market failure. The paper and panel raised important questions regarding how best to address that market failure, a discussion that will prove crucial to energy policy in coming years.