Coal, Gas, Nuke, Water, Wind: The Green Power Grid in Our Future

by Richard K. Lester & David M. Hart

Clock Icon 34 Minute read

Unlocking energy efficiency—not only in buildings, but also in transportation and industry—will be the first wave of low-carbon energy innovation. However, the nation’s energy and climate goals cannot be achieved through efficiency improvements alone. America’s need for energy is too great, and the existing energy system too carbon-intensive, for an efficiency-only approach to work. Nor is natural gas-fired electric power a permanent solution, even though it emits substantially less carbon dioxide per kilowatt-hour than coal. While welcome at the moment, an excessive focus on gas could be unhelpful over the longer term if it deters investment in even lower-carbon alternatives. As a recent MIT study on the future of natural gas concluded, “though gas is frequently touted as a ‘bridge’ to the future, continuing effort is needed to prepare for that future, lest the gift of greater domestic gas resources turns out to be a bridge with no landing point on the far bank.”1

In the medium and long term we must unlock a second wave of innovation that enables low-carbon technologies to supplant natural gas and other fossil fuels for power generation. A central goal of this effort must be to improve the performance and lower the unit costs of electricity production from low-carbon generation technologies whose basic scientific and engineering characteristics are already well known. The options include solar photovoltaic and solar thermal technologies, wind (including offshore wind installations), advanced nuclear reactor and fuel cycle technologies, carbon capture and storage technologies for both coal and gas-fired power plants, geothermal, and biomass. Large-scale deployment of second-wave innovations that prove to be viable should begin about a decade from now and continue for several decades.

Much has been said about which electricity generation technologies ought to dominate in the low-carbon economy. We do not take a position on the eventual winners and losers, but instead envision an experimental approach to determining outcomes. Indeed, the terms “winner” and “loser” are inappropriate, since the likely outcome will be different mixes of options in different locations and for different uses. The nation’s portfolio of low-carbon generation alternatives is likely to include “all of the above.”

We include in this assessment distributed generation as well as central station generation. Almost all electricity consumed in the United States today is produced in central-station plants, mainly coal (45%), natural gas (23%), nuclear (20%), and hydro (7%). Historically, the economies of scale and cost efficiencies afforded by central station generation have far outweighed the drawbacks, which include large up-front costs and vulnerability to adverse events. Innovations on the horizon, which involve pricing, distribution, and other aspects of the electricity system as well as generation, may increase the role of distributed generation. But the traditional dominance of central station generation will not fade quickly—indeed, it could persist indefinitely.

A major institutional restructuring will be necessary to unlock second-wave innovations for electricity generation, whether central station or distributed. In particular, funding for the demonstration and early deployment of these innovations must be mobilized and allocated in new ways. As we argue in the next two sections, neither the market (including venture capitalists, some of the most creative participants in the market) nor the federal government has effectively supported these two stages of the innovation process in the energy realm. We then briefly review existing proposals to remedy this problem. The centerpiece of the chapter is our own proposal for a regionally based mechanism to fund demonstration projects and early deployment programs. Our proposal calls for upending established practices and injecting competition at multiple levels, building on the principles that we articulated earlier in this book.

Gaps in Financing Demonstration and Early Deployment

An innovation system that will hasten the transition to low-carbon energy must support all four stages of the innovation process: option creation, demonstration, early adoption, and improvement-in-use. One of its most important jobs is to ensure the availability of adequate funds. As a rough rule of thumb, the funding requirement increases tenfold when an innovation moves from one stage to the next (see figure 2.1). A well-functioning innovation system will also attend to the “gates” between the stages, winnowing out options that perform poorly. The order-of-magnitude jump in cost at each gate puts a premium on doing this job well.

While the cost of option creation will presumably be covered mainly by public funding, this stage is the least expensive of the four. At the other end of the process, the vast majority of funding for improvements-in-use will come from companies who are already producing the technologies that are being improved. The stages in between, demonstration and early deployment, logically require a mix of public and private funding. Just what the mix should be, and what rights the public and private partners ought to have with regard to winnowing out the options being evaluated, are hotly debated topics in both public policy and private investment circles. These issues have been so challenging to resolve that the passage through the two middle stages of the innovation process is sometimes referred to as the “valley of death.”2

The problem is difficult even for relatively small-scale innovations, such as software and electronics for managing energy use. But it is much harder for large-scale central station electricity generating technologies. When the budget of a single demonstration project (or what we will refer to as the “next few” post-demonstration projects) runs to hundreds of millions or billions of dollars, the valley of death looks more like a chasm. Costs for the mass production of distributed generation technologies (such as building manufacturing plants for solar modules) during the early adoption stage may also be in this range. (Solyndra, a thin-film solar photovoltaic manufacturing company, had to raise $970 million in equity finance, in addition to receiving $535 million in federal loan guarantees, even before it was floated on the stock market in 2010.3) Yet, unless these early costs are underwritten by somebody, the energy industry will not learn how to make low-carbon technologies affordably and operate them reliably.

The sheer size of the learning investment is a deterrent in both stages, but there are additional barriers specific to each stage. At the demonstration stage, private investors are deterred by a large array of uncertainties that make it hard to assess the odds of success. There are uncertainties not only about the technical and economic performance of the technology being demonstrated, but also about safety and environmental regulations, the future market price of competing fuels and the regulatory price that may be levied on carbon dioxide emissions. Another deterrent to private investors is the difficulty of appropriating the benefits of investments in demonstration projects. One of the main purposes of such projects is to disseminate credible information about the cost, safety, and reliability of the new technology as widely as possible to prospective providers, investors, and users. For private investors, this requirement reduces the value of proprietary knowledge or intellectual property that such projects produce.

The learning that takes place at the early adoption stage, by contrast, should be proprietary, occurring within the firms that are building and operating projects and among the firms in the supply chain. The chief goal in this stage is to drive down costs. Nonetheless, the process of doing so may be lengthy and expensive, especially for the “next few” projects built after the initial demonstration has shown a technology to be technically feasible. Innovators in this stage must compete head-to-head against well-entrenched competitors who have spent many years honing their technologies. The cost of power generated by the “next few” will often still be above that of the incumbents. The cost gap may persist for some time. (Indeed, in some cases it may never be fully closed. Some technologies will simply fail the affordability test and will need to be winnowed out as a result.)

Some observers have suggested that venture capital can bridge the “valley of death” in energy innovation as it has in many other industries since the mid-1970s. Venture capitalists are accustomed to taking technical risks, and they routinely build learning curves into their investment plans. In fact, “clean tech” venture investments rose rapidly during the past decade, from $230 million in 2002 to more than $4 billion in 2008.4 But most of this investment has been, by necessity, on a relatively small scale. A typical individual investment by such a fund rarely exceeds $10–$15 million. As a result, most “clean tech” investment has gone into distributed generation and smart-grid technologies, rather than large-scale central station generation technologies.

A second challenge for the venture capital model is exit. As the costs rise, venture investors will want to partner with or sell out to deeper-pocketed but more cautious investors. Yet the risk-return profile of investments in energy innovations in the early adoption stage is often unattractive for these “downstream” investors. This is because expected returns to innovation in commodity markets like electricity are fairly modest, but the risks and uncertainties are diverse and large. So downstream investors have often remained on the sidelines of large-scale clean tech, cutting off what has been a favorite exit route for venture investors in other high-technology industries. So far, for example, there have been few equivalents in the energy industry to the roles played by large pharmaceutical companies and information and communication companies as sources of scale-up capital.

(More details can be found in box 1 on “Why Venture Capital Plays Only a Limited Role in Energy Innovation.”)

Federal Support for Demonstration and Early Deployment: The Record

The failure of the market to adequately fund the demonstration and early deployment of low-carbon energy innovations is widely recognized. The federal government has frequently stepped in to try to fill these financing gaps in the past. Many of these attempts have failed. Troubled projects and perverse incentives have been all too common.

Box 1: Why Venture Capital Plays Only a Limited Role in Energy Innovation

The basic model of venture capital investing entails the formation of a fund managed by a small number of general partners and financed by third-party investors or “limited” partners (limited in both liability and management control). There is no upper limit on the size of these funds, but the partnership structure and the ability of individual general partners to manage no more than a certain number of active investments means that funds typically do not exceed a few hundred million dollars.

Most early-stage companies with innovative technologies or business models will never become profitable, but venture capitalists are willing to invest in these businesses anyway, despite their high technological and market risks, because they expect a small fraction to generate high returns in the end. A typical individual investment made by such a fund therefore rarely exceeds $10–$15 million. If the amount were much bigger, the fund would not be able to make enough investments to achieve reasonable risk diversification. This structure constrains venture capital to less capital-intensive businesses. Venture capitalists also prefer to invest in businesses whose commercial viability can be established in five years or less, allowing them to exit well before the life of the fund—typically 10 years—is over. This approach enables them to establish the record of successful performance needed to raise the next fund.

Exit is achieved through the public or private equity markets, or by selling the business to an established company in the sector. Downstream investors typically augment their capital with bank loans or other kinds of debt finance. Their access to these capital markets provides them with much more capital than venture capitalists have at their disposal, but it also means they have a much smaller appetite for risk.

An ideal venture investment—that is, one that enables rapid and profitable exit—is one where the technological risks, while large at the outset, can be resolved fully within a few years and with only a modest amount of capital. Many software-based businesses have these features, with Google and Facebook being among the most spectacular recent examples. Some innovative energy technologies and businesses also have these characteristics, but many more do not. Large central-station generating technologies like nuclear reactors clearly do not. Nor do many solar, wind, and storage technologies, where uncertainties over regulatory requirements and energy pricing policies as well as important technological and economic performance risks cannot be resolved in advance of utility-scale deployment or full-scale manufacturing, or both. These residual risks are likely to be too great for risk-averse downstream investors to assume, while the capital requirements for scale-up are too large for individual venture funds or even for syndicates of funds.

The history of demonstration projects managed by DOE and its predecessors spans several decades. An early, and by many measures successful, federal investment was the Shippingport project of the 1950s, which demonstrated the use of pressurized-water nuclear reactor technology for commercial-scale power production (the technology was originally developed for naval propulsion).5 More recent projects have tended to be less successful. The Clinch River Breeder Reactor Project, for instance, cost U.S. taxpayers about $1.7 billion and never operated. The synthetic fuels demonstration program failed in its stated goal of producing a half million barrels a day of synthetic oil and gas from domestic sources, largely because the cost of its products far exceeded the market price of oil. The Yucca Mountain nuclear waste repository project has been mired in controversy since it was initiated in the late 1980s and is currently on hold. The saga of FutureGen, DOE’s erstwhile flagship project to demonstrate carbon capture and sequestration, is still unfolding, but seems to be headed down the same troubled path.

Federal post-demonstration subsidies have also been problematic. Rather than stimulating innovators to bring down the cost of new technologies as quickly as possible, they have often had the opposite effect. Open-ended government subsidies have rewarded firms not for innovating but simply for producing regardless of cost. Probably the most egregious example is the now decades-old program of corn ethanol subsidies, but there have been other cases where an idea that seemed promising at the outset spawned an entrenched constituency that later resisted all attempts to scale back support. The federal government has typically been unable to ratchet down subsidies in order to drive cost reductions, much less shut projects and programs down in a timely fashion when they have clearly failed to produce the expected results.

Why has the federal government’s track record in energy technology commercialization been so unsatisfactory? Excessive congressional involvement is one reason. Influential elected officials have sometimes interfered in technology selection and facility siting decisions and personnel appointments. Congressional pressure has also limited the ability of executive branch officials to adjust or terminate projects after conditions have changed. In other cases, Congress has forced the abandonment of projects even when conditions have not changed. The annual budgeting and appropriations process has invariably been a source of uncertainty for program managers, while policy shifts following congressional and presidential elections have added further to the lack of stability in funding and policies over the life of projects.

Other explanations focus on problems within the executive branch. Federal agencies including DOE have at times displayed a tendency to underestimate project costs (perhaps as a requirement to generate political support).6 Restrictive federal procurement regulations and bureaucratic rules governing personnel decisions, auditing requirements, and the use of federal facilities have also been implicated in cost overruns and poor management. MIT’s John Deutch notes that Congress has not granted DOE and other agencies the authority to attract individuals with the experience and skills to oversee major projects.7 DOE and its overseers on the appropriations committees have not been able to establish and adhere to multiyear budget planning.

Reform Proposals on the Table

The dilemma is evident. On the one hand, the prospects for successfully commercializing low-carbon electricity generation technologies, especially large-scale central-station technologies, will be very dim if the market is left to do this on its own. Some public role in sharing the costs and risks of demonstration projects, and in devising strategies for bridging the post-demonstration cost gap, seems essential. Yet it is difficult to avoid the conclusion that DOE is structurally unable to play this role effectively.

Several proposals have been advanced to break the impasse. Senator Jeff Bingaman, chairman of the Senate Energy and Natural Resources Committee, and ranking member Senator Lisa Murkowski have recently introduced legislation to create a new federal financing entity, the Clean Energy Deployment Administration (CEDA), that would give high-risk energy demonstration projects and deployment programs access to various forms of financing, including loans and loan guarantees. CEDA would be a semi-independent unit within DOE.8

Another proposal would go a bit further, creating a Green Bank as an independent, tax-exempt corporation that is wholly owned by the federal government. The Green Bank would be similar in structure to the Export-Import Bank. It would support diverse technologies and projects through debt financing and credit enhancement, giving priority to those projects that would contribute most effectively to reducing greenhouse gas emissions and oil imports. The Green Bank would be an independent, but government-owned, tax-exempt corporation with both public officials and experienced private individuals on its board.9

A third proposal, advanced by MIT’s Deutch, would establish an autonomous, quasi-public organization, the Energy Technology Corporation, specifically to finance and execute large-scale energy demonstration projects.10 It would have flexible hiring authority and follow commercial practices in its contracting. This corporation would be governed by an independent board of directors nominated by the president and confirmed by the Senate.

Finally, the American Energy Innovation Council, a group of leading business executives including General Electric CEO Jeff Immelt and Microsoft chairman Bill Gates, have proposed a public-private partnership to address the problem.11 Noting that America’s energy innovation system “lacks a mechanism to turn large-scale ideas or prototypes into commercial-scale facilities,” this group has called for the formation of an independent, federally chartered corporation, outside the federal government, that would be tasked with demonstrating new, large-scale energy technologies at commercial scale. This new organization would “offer project management services and technical resources to help accelerate and improve the design and construction of facilities. . . . It would work to enable fast-track siting and construction opportunities within utilities or public power agencies, on federal or military lands, or even overseas … in some cases.” It would report to a politically neutral, congressionally mandated Energy Strategy Board that would also be external to the federal government.

Though the details vary, all of these proposals are designed to overcome the limitations of DOE and to insulate innovation decisions from political forces to some degree. The new entities that have been proposed would be free of many of the most burdensome federal rules; they would also have much more flexibility in management and would be independent of the annual congressional budget cycle. But all of them would require a big one-time congressional appropriation to get started: $10 billion for CEDA; $10 billion for the Green Bank; $60 billion for the Energy Technology Corporation; and $20 billion for the public-private energy demonstration corporation proposed by the American Energy Innovation Council.

A Regionally Based Demonstration and Early Deployment Funding Mechanism

We agree with the thrust of these proposals, though we do not think they go far enough. It will of course be quite difficult to secure a congressional appropriation for tens of billions of dollars in today’s budget environment. More important, these proposals do not adequately embody the foundational principles for a durable and productive energy innovation system. The most important of these principles include: creating space for new entrants (both new firms and existing firms from other industries), providing a reasonably steady flow of investment capital to innovators over long periods, ensuring timely and rigorous down-selection at each stage of the innovation process, accommodating regional differences, and matching the size of the challenge. In the remaining sections of this chapter, we put forth our alternative.12

A starting point for this exposition is EPRI, formerly known as the Electric Power Research Institute, a nonprofit R&D organization serving the electric utility industry. EPRI was established by electric utilities in 1972 after a series of massive power blackouts prompted concerns about the industry’s performance. It is a membership organization with broad representation within the electric utility sector. Over the years EPRI’s research agenda has been set by its professional staff in conjunction with their technical counterparts in utility companies. The research agenda is approved by the organization’s board of directors, who are elected by the members. EPRI has made important research contributions although it has shrunk considerably since utility contributions to support its work were curtailed in the wake of the restructuring of the 1990s.13

The key features of EPRI from our perspective are its membership base and professional staff. EPRI members are the users of its work: the organization provides them with a unique opportunity to articulate their collective needs and informs them of technical trends. These features distinguish EPRI from the demonstration and early deployment financing models we described above. But EPRI also has weaknesses, including a lack of guaranteed funding and a near-monopoly position in its field. Moreover, the electric power industry has not pushed the innovation frontier with the vigor the climate challenge demands.

The scheme we propose has a set of EPRI-like membership organizations at its core. We call them regional innovation investment boards (RIIBs). As the name suggests, we envision creating several of these boards around the country, rather than having only a single one. The RIIBs would decide which large-scale demonstration projects and early adoption programs to support. A federal agency would serve as a gatekeeper to certify the acceptability of investments that could be proposed to the RIIBs. RIIB funding would derive from a surcharge on electricity sales. The surcharge funds would be allocated to the RIIBs by a set of state-level trustee organizations. The RIIBs would have to compete to win enough support from enough trustees to fund their portfolios. This regionally based demonstration and early deployment funding mechanism fulfills our principles more fully than the other proposals that we have considered.

Regional Innovation Investment Boards

Recommendation: Congress should authorize the creation of a network of regional innovation investment boards that would invest in demonstration projects and early deployment programs for low-carbon energy technologies.

RIIBs are at the cutting edge of the energy innovation system that we envisage. Each board’s membership would comprise the electric power companies in that region, including “smart integrator” utilities, transmission companies, vertically integrated utilities (in those parts of the country where they remain in place), independent power producers, providers of grid services, and demand response and energy service firms. The membership would thus reflect the perspectives of innovation users with close knowledge of the electric power system and the requirements of the electricity market. Some of the members would be potential buyers and operators of the innovations being demonstrated and deployed; others would be indirect users of them. For example, an independent power producer and a smart integrator might both be represented on a board that is supporting a project to demonstrate a new central station generation technology. The IPP would be an eventual buyer of the technology, while the smart integrator might ultimately incorporate the technology into its network. RIIB members would elect the governing board, which would have ultimate authority over innovation investment decisions. To limit conflicts of interest, representatives of entities proposing projects to an RIIB would not be eligible to serve on its governing board.

Projects that might be funded by the RIIBs would include first-of-a-kind large-scale demonstration projects and “next few” post-demonstration projects. A variety of investment instruments could be used, including loans, loan guarantees, interest buy-downs, or equity infusions. Investments by the RIIBs would leverage larger amounts of private-sector funding for these projects. The RIIBs might also invest in programs designed to accelerate the early take-up of innovations in distributed generation. The investment instruments used for this purpose could include large-scale user subsidy programs such as feed-in tariffs, rebates, or subsidized loan programs.

Many different kinds of teams might propose projects and programs for RIIB support. In some cases a single entity, such as a technology vendor, could propose a demonstration project. A state government could propose an early adoption program. Alternatively, the project team might be a consortium of established and entrepreneurial vendors, other members of the technology supply chain, independent power producers, and national or academic laboratories. Those proposing projects would seek RIIB funding not as the sole source of finance but rather to augment their own investments and to lower their risks. RIIBs would choose among competing project proposals based on the strength of the project team, the quality of project management, the attractiveness of the technology, the extent of self-funding by the proposers, and so on.

To seek RIIB financing, projects would first have to be certified by the federal gatekeeper organization (criteria for certification are discussed below). Certified projects could be proposed to one or more RIIBs for funding. RIIBs could operate independently, or they could collaborate with each other on investment decisions. They would be obligated to commit all funds received in a given year by the end of the following year; in other words, they would not be permitted to accumulate funds. However, funding commitments to projects could be spread over time, with future payments made contingent on the satisfactory performance of the project team.

The boundaries of the RIIBs would need to be carefully drawn. To avoid excessively fragmented decision making, no more than ten such boards should be created. One approach would be to build on existing groupings of states with prior experience working together on regional energy and environmental issues—for example, the New England states, the Northwest states, and the Southwest states. A somewhat different approach would be to use the boundaries of the eight North American Electric Reliability Corporation (NERC) regions. Each NERC region is relatively autonomous and has a distinctive character that reflects its history and demand profile.14

Mapping the RIIBs onto the NERC regions or some other regional structure would have the benefit of steering innovation activities toward the energy transition pathways most suitable and attractive to each region of the country. Over time, each RIIB might be expected to specialize in innovations of particular interest to its region—for example carbon capture and storage in the coal-dependent Midwest, nuclear in the Southeast, solar in the Southwest, and offshore wind in the Northeast. A RIIB might also tend to invest in projects located in its own region, though it would not be limited in this regard.

State Trustees

Recommendation: Federal and state legislation should establish an innovation surcharge on retail sales of electricity and create trustee organizations in every state to allocate the proceeds to the RIIBs.

Funding for the RIIBs would derive from an innovation surcharge on retail sales of electricity. Individual RIIBs would not have an entitlement to funding, however. Rather, disbursing the proceeds from the surcharge would be the responsibility of a trustee organization in each state. After a start-up period in which each trustee would be required to direct all surcharge funds collected in its state to the RIIB in its region, trustees would be free to allocate funds to one or more RIIBs in other regions. The allocation would be based on each trustee’s assessment of which RIIB portfolios of demonstration projects and early adoption programs most closely matched its needs. The trustee organizations would be more broadly representative of stakeholders in the electricity system than the RIIBs themselves. Their members might include a variety of business groups, government organizations and officials, environmental and labor groups, and technical experts. Depending on the state, trustee board representatives might be appointed or elected. Trustees would have to allocate all proceeds from the surcharge to RIIBs within a year of receipt.

We suggest an innovation surcharge of 0.2 cents per kilowatt hour. This would add roughly 2 percent to the average retail price of electricity in the United States, or about $1.80 per month to the average residential electricity bill, and would generate roughly $8 billion per year. (As noted, these funds would leverage private sector funds such that the total investment induced might be at least two to three times larger.) This funding stream is similar in scale to that proposed by the President’s Council of Advisors on Science and Technology to support federal energy RD&D.15 Federal legislation would be required to enact the proposed surcharge. It might justifiably be enlarged. Revenues from other sources, such as revenue raised by eliminating tax incentives for oil and gas production or by imposing a carbon tax, could be added as well.

A Federal Gatekeeper

Recommendation: A federal gatekeeper should certify proposals for demonstration projects and early adoption programs before they become eligible for RIIB investment.

The third new institution in the regionally based funding scheme that we propose is a federal gatekeeper organization that would certify, decertify, and recertify projects and programs. Every proposal would have to be certified before it could be presented to the RIIBs. The gatekeeper would make sure that RIIB investments were supporting national purposes. All proposals would have to demonstrate the potential to lead to significant reductions in carbon emissions at a declining unit cost over time. The gatekeeper would apply this criterion to all proposals in determining their eligibility for RIIB funding, regardless of the nature of the innovation under consideration. The gatekeeper would not determine whether a specific proposal should receive funding, nor would it rank innovations or evaluate the organizational capabilities of the proposing organizations. These tasks would be undertaken by the RIIBs themselves.

Projects or programs would need to meet a minimum size threshold (say $500 million of public and private funding combined) to qualify for RIIB funding. The objective for first-of-a-kind projects would be to (1) demonstrate technical feasibility at full scale and (2) generate and disseminate data on the technology’s cost, reliability, and environmental performance. “Next few” post-demonstration projects would need to incorporate information from previous demonstrations with the aim of making continued rapid progress toward competitiveness with incumbent technologies. Large-scale user subsidy programs, similarly, would need to show their impact on technological progress down the learning curve.

The gatekeeper would be responsible for monitoring progress. Proposals would be judged in part on how effectively they put pressure on innovators to exploit learning to reduce costs. Public subsidies would have to decline steadily as experience with the innovation is gained. Certification would be granted only for a limited period—five years, say—and could be withdrawn at that time if progress proved too slow. The gatekeeper would also track projects and programs that target the same innovations to guard against duplication and overlap. At the same time, it might take into consideration the value of pursuing several different approaches in parallel as circumstances warrant.

The gatekeeper would need to be a technically competent organization capable of representing the broad public interest. It could be part of an existing federal agency such as DOE or it could be a stand-alone unit. It would need to be capable of assessing the technical progress of projects and programs. It would also need to be able to evaluate the potential of scale economies and future learning opportunities. Finally, it would have to have a global perspective and be knowledgeable about developments overseas, so that RIIB investments would not simply duplicate work being done elsewhere.

Pros and Cons of the RIIB Approach

A regionally based funding mechanism offers several advantages over current practice. It would create a dedicated new stream of funding for what has been a chronically under-resourced part of the energy innovation system. It would avoid the stop-and-go pattern associated with the federal appropriations process, where decisions are too often driven by short-term energy price volatility, technology fads, and election results, and would instead generate the steady, predictable flow of funds needed to make credible multiyear investment commitments. The innovation surcharge would generate roughly $8 billion per year and would induce as much as $20 billion per year of total investment in demonstration and early-adoption activities—enough to have a major impact on the nation’s energy innovation challenge, without increasing the federal budget deficit.

By putting project and program selection in the hands of potential innovation users—the RIIBs—rather than in the hands of government officials or elected representatives, this new institutional framework would be much more responsive to feedback from the market and from technical results than the existing system. At the same time, the public interest would continue to be strongly represented by the federal gatekeeper and state trustee organizations.

The institutional framework we propose would also introduce multiple levels of competition into the innovation process. In the past, demonstration projects have been selected through a highly centralized and sometimes arbitrary process, in which individual congressional champions or national laboratories have often played influential roles. In the new arrangement, project teams, once certified, would compete with each other for funds from one or more RIIBs to design, construct, and operate demonstration and post-demonstration plants, or to implement early adoption programs. The RIIBs, in turn, would compete with one another to secure support for their portfolios from the state trustees. A regional board with a portfolio deemed promising by multiple trustees would see its investment budget swell, while those with less promising portfolios would shrink. This more decentralized scheme would also allow new entrants who lack connections to the existing federal R&D structure to get a better hearing for their ideas than they do at present.

Finally, our scheme would create opportunities for regional needs and preferences to be expressed in the energy innovation system, and would give states a direct stake in innovation outcomes.

Our scheme also has several potential drawbacks. The electric power industry has not been known for its vision and imagination, and assigning investment responsibilities to its representatives might increase the likelihood of cautious, incremental decision-making. The gatekeeper is designed to ward off this possibility and focus the RIIBs on significant reductions in carbon emissions at declining unit cost. At the same time, the injection of conservative user perspectives at the demonstration and early adoption stages has the virtue of constraining politically popular projects and programs that are not consistent with the overall goal of reducing risks to a level that would allow private investors to fund future projects on their own.

A second potential drawback involves resistance from entrenched interests. The DOE laboratories, for instance, have long enjoyed a privileged position in the U.S. energy innovation system. They exert a strong influence over DOE resource allocation decisions and receive by far the largest share of DOE RD&D funds.16 In our proposed framework the laboratories would have to compete with industry, universities, and others for a share of the RIIB energy innovation funds, although the total amount of funding available would be much greater. (The labs’ traditional role in support of other DOE mission areas, such as national security, basic science, and environmental quality, would not be affected.) States might also have difficulty accepting the regional approach. There could be pressure on the trustees to spend funds collected within the state only on projects within the state. However, the proposed scheme would not permit the trustees to fund individual projects. Rather, they would be required to allocate their funds at the portfolio level, with each RIIB portfolio typically comprised of multiple projects distributed across states both inside and outside the RIIB’s region. In addition, the RIIBs would be membership organizations and therefore less susceptible to this kind of pressure.

Perhaps the most serious objection to the scheme we propose is that it would entail the creation of many new organizations and take time to set up. There will be a learning curve for each of these organizations. That is unavoidable, but the process should be complete in a period of years, while the payoffs will occur over several decades. Competition among the RIIBs and discipline provided by the gatekeeper should minimize the risk of administrative inefficiency over the medium-term.

An illustration of how this scheme would work can be found in box 2 on “The RIIB-Gatekeeper-Trustee Innovation System in Practice: The Case of Nuclear Power.”

Box 2: The RIIB—Gatekeeper—Trustee Innovation System in Practice: The Case of Nuclear Power

To see how the regionally based system for demonstration and early deployment might work in practice, consider the case of nuclear power, which—perhaps more than any other major energy technology—has depended on and been influenced by direct federal involvement in the past. High capital costs for nuclear power plants and the lack of a resolution to the problem of nuclear waste disposal have been the main obstacles to a revitalization of the U.S. nuclear power industry in recent years. More recently, the Fukushima nuclear accident has brought a renewed focus on safety concerns, with uncertain implications for the industry’s near-term prospects in the United States and elsewhere. Yet many innovations in nuclear reactor and fuel cycle technologies are under development that have the potential to contribute to safety, environmental, economic. and security objectives.

Historically, the federal government has played a leading role in commercializing nuclear power reactor and fuel cycle technologies. In recent years its primary focus has been to promote, in collaboration with the nuclear industry, the development and deployment of advanced light water reactors. Relevant federal policies have included regulatory reforms, such as the more streamlined combined construction and operating license procedure set in place by the U.S. Nuclear Regulatory Commission two decades ago; federal cost matching to encourage private firms to seek approval for advanced reactors under these new licensing procedures; energy production tax credits covering the first 6000 MWe of new nuclear capacity; federal regulatory insurance to cover the cost of regulatory-induced delays in the operation of the first six new nuclear power plants; and federal loan guarantees to support advanced nuclear power plant construction.

In addition, since the early 1980s the federal government has been legally responsible for the final disposition of spent fuel discharged by current as well as any future nuclear power plants. The government has also provided significant funds for the development of advanced nuclear energy technologies, including sodium-cooled fast reactors and gas-cooled high-temperature reactors, as well as a range of new fuel cycle technologies, especially for spent fuel reprocessing, actinide separation, and transmutation. Although the government has played no role in developing advanced uranium enrichment technologies since the privatization of federal enrichment operations in the 1990s, in recent years it has offered loan guarantees and other funding to first-of-a-kind enrichment plants planned by private firms.

Few of these policies have been successful. The federal loan guarantee program, potentially the most important of the government’s promotional measures, has been largely ineffective so far. Seventeen utilities have submitted applications for these federal loan guarantees, on behalf of more than 20 advanced light water reactor projects. But in the five years since the loan guarantee program was first introduced, DOE has approved guarantees for only one new project, and most of the utilities have either withdrawn from the program or suspended their plans.a The economic viability of those projects was adversely affected by the decline in natural gas prices and by the unwillingness of the Congress to enact a carbon price. In some cases, however, DOE’s loan guarantee office and the Office of Management and Budget were simply unable to agree on acceptable terms with the utilities.

The prospects for nuclear development have also been adversely affected by problems in the government’s program for high-level waste disposal. The latest setback was the Obama administration’s decision to halt work on the Yucca Mountain nuclear waste repository project, which for the last twenty years has been the sole focus of federal spent-fuel disposal activities. The current impasse over Yucca Mountain, the lack of an alternative disposal plan, and the continued inability of the federal government to fulfill its contractual obligations to remove spent fuel from commercial power reactor sites—a long-running irritant in relations between the nuclear utilities and the government—together convey an impression of government ineptitude and ineffectiveness in a domain vital to the future of the nuclear industry.

Over the past year DOE has begun to promote the accelerated development, licensing, and deployment of small, modular nuclear power plants, both light water reactor (LWR)-based concepts and non-LWR designs. With congressional support, DOE is proposing to transfer some funds from other nuclear energy programs to this new initiative. But it has not yet indicated how it will allocate available federal funds among design certification and licensing activities, demonstration projects, deployment incentives, or the creation of reactor manufacturing capacity. Nor has it said how it will direct these funds among several privately developed small reactor concepts.

The arrangements we advocate in this chapter would transform radically the U.S. approach to nuclear energy innovation. DOE and its laboratories would continue to conduct R&D in nuclear reactor and fuel cycle technologies, including R&D in the important areas of safety and regulatory development. But their role in commercializing new nuclear technologies would be diminished. RIIBs, responding to proposals by nuclear innovators, would be the primary decision-makers on key strategic questions: how to manage the financial risks of new advanced light water reactor power plant construction; what the appropriate division of effort should be between the implementation of incremental advances in large light water reactors versus the commercialization of small modular reactors; which particular reactor technologies and commercial groupings are most deserving of demonstration and “next few” investment; how to allocate investment among different approaches to new reactor and fuel cycle technologies, funding human capital development, and so on.

Ironically, the federal high-level nuclear waste disposal program, which is almost certainly the most egregious example of a nonperforming federal investment in nuclear innovation, has been financed for decades by a revenue-raising mechanism—a 0.1 cent per kilowatt hour fee on nuclear electricity paid into a federal Nuclear Waste Fund—with some similarities to the arrangement we advocate here. Some $25 billion has now accrued in the fund (including interest). But the differences are important. All decisions regarding the use of the Nuclear Waste Fund have been centralized within the DOE and overseen by Congress. Federal spending in this area has been subject to extraordinary constraints, such as the nearly twenty-five-year-long statutory prohibition on investigating disposal technologies and geologies other than those at Yucca Mountain—a provision that actively discouraged innovation in the field of nuclear waste disposal. Congressional oversight, exercised through the annual appropriations process, has routinely been politicized as opposition to the Yucca Mountain project has mounted, and DOE has frequently complained about its inability to implement plans for the site because of congressionally imposed spending restrictions. The nuclear utilities and their ratepayers have had essentially no recourse in the face of these problems, other than to sue the federal government for its failure to meet its contractual obligations to remove their spent fuel. We concur with the recommendation, first made more than thirty years ago, to move responsibility for implementing nuclear waste management and disposal out of DOE and into an independent government authority.b

a. For a nuclear industry perspective on the federal loan guarantee program, see “Credit Subsidy Costs for New Nuclear Power Projects Receiving Department of Energy (DOE) Loan Guarantees: An Analysis of DOE’s Methodology and Major Assumptions,” Nuclear Energy Institute White Paper, August 2010.

b. The latest recommendations along these lines have been made in a recent MIT study, The Future of Nuclear Fuel Cycle (Summary Report, Massachusetts Institute of Technology, 2010). The original recommendation was made in Mason Willrich and Richard K. Lester, Radioactive Waste: Management and Regulation (New York: Free Press, 1977). A scheme similar to the one proposed here for energy innovation might also be considered for managing high-level nuclear waste.


The problem with past energy technology demonstration projects was not that they failed but that at some point the goal became to avoid failure. For the leaders of these high-profile projects and their supporters in and outside government the costs of failure were too great, so failure had to be avoided at all costs. But some of the strategies for preventing failure themselves proved costly, including driving out other alternatives prematurely, refusing to recognize legitimate problems until long after they arose, and failing to acknowledge that key assumptions were no longer valid. And, of course, these projects also generated a constellation of opponents, whose goal became precisely to cause their failure, and to prevent them from producing anything useful. In this environment, the most important goals of the innovation process—generating new information and learning quickly about the strengths and weaknesses of alternative approaches—were undermined.

For large-scale technologies, developed in government-led and government-financed projects, this kind of pathology is an ever-present risk. Yet the rapid development and deployment of new large-scale electric power technologies will be essential to the low-carbon energy transition. So a critical task is to devise an innovation system in which, even for these large-scale technologies, multiple pathways can be pursued and failure is tolerable. In this chapter we have suggested one such scheme. Undoubtedly others can also be devised and we encourage further explorations along these lines. The most important goal is to create an institutional structure that can accommodate and promote diversity, experimentation and competition in the innovation process—even for large-scale technologies and even during the downstream stages of demonstration and early adoption.

The second wave of innovation is not only about large-scale electric power technologies, however. There are many other innovations that should be pursued in the second wave. For these, too, new institutional structures are needed to unlock the full creativity and competitiveness of America’s innovators.



1. MIT Energy Initiative, The Future of Natural Gas: An Interdisciplinary MIT Study, 2011, 69, available at As we discussed in chapter 1, substitution of gas for coal has the potential to reduce carbon emissions per unit of electricity output by 50% or more. We endorse using electricity from new as well as existing gas-fired power plants to substitute for old coal plants in the next decade or two.

2. This colorful but imprecise term invites confusion with another commonly cited “valley of death,” where the latter refers to the shortage of seed capital needed to translate the results of laboratory research into a sufficiently convincing business concept to attract angel or early-stage venture capital investors.

3. Shikhar Ghosh and Ramana Nanda, “Venture Capital Investment in the Clean Energy Sector,” MIT Industrial Performance Center Working Paper, August 1, 2010.

4. Ibid., 8.

5. Ashley Finan, unpublished Ph.D. thesis, Department of Nuclear Science and Engineering, Massachusetts Institute of Technology, forthcoming 2011. As Finan notes, the legacy of the Shippingport project has been much debated. The project had initially been planned as one of several parallel government-funded demonstrations of alternative nuclear power reactor technologies. But federal budget constraints, political conflicts over the proper roles for government and private industry in the power sector, and the greater maturity of pressurized-water reactor (PWR) technology thanks to its already-extensive role in naval propulsion left Shippingport as the only major government-funded nuclear reactor demonstration. The success of the project, which was closely supervised by Admiral Hyman Rickover, helped establish an unassailable lead for PWR technology for commercial power applications.

6. See, for example, Linda Cohen and Roger Noll, The Technology Pork Barrel (Washington, DC: Brookings Institution Press, 1991).

7. John M. Deutch, “An Energy Technology Corporation will Improve the Federal Government’s Efforts to Accelerate Energy Innovation,” Hamilton Project Discussion Paper 2011-05, Brookings Institution, Washington, DC, May 2011.

8. Jeff Bingaman, “An Energy Agenda for the Next Congress,” Issues in Science and Technology, Spring 2011, 35–42.

9. John Podesta and Karen Kornbluh, “The Green Bank,” Center for American Progress, May 21, 2009, available at

10. John Deutch, “An Energy Technology Corporation.” Somewhat similar proposals have been implemented elsewhere. Sustainable Development Technology Canada, a non-profit foundation, operates two funds totaling about $1 billion designed to fill the financing gap between early stage research and commercial deployment. One of the funds is focused on financing first-of-a-kind large-scale demonstration facilities for next-generation renewable fuels. The organization was created and is funded by the Canadian government and is accountable to Parliament. See

11. American Energy Innovation Council, A Business Plan for America’s Energy Future, available at

12. Some of the ideas on which this proposal is based were originally suggested by Paul Romer, “Implementing a National Technology Strategy with Self-Organizing Industry Investment Boards,” Brookings Papers: Microeconomics 2 (1993): 345–390.

13. Today EPRI receives just $180 million per year in membership revenues, a tiny fraction (one-twentieth of one percent) of electric power industry sales—and not nearly enough to fund a serious program of technology demonstrations. Moreover, most of the funds are used to support near-term engineering work aimed at improvements in operations and maintenance.

14. Each NERC regional coordinating council is comprised of all the investor-owned utilities, public power authorities, independent power producers, and large energy users in the region. So there would be a strong overlap between the membership of the NERC regions and the membership of the RIIBs. But the two functions—system reliability coordination and long-term innovation investment—are very different, so there might not be much benefit to combining them in a single organization. A possible additional complication is that three of the NERC regions, the Western Electricity Coordinating Council, the Midwest Reliability Organization, and the Northeast Power Coordinating Council, extend into Canada, and two others, Texas and Florida, are single-state regions.

15. President’s Council of Advisors on Science and Technology (PCAST), Report to the President on Accelerating the Pace of Change in Energy Technologies Through an Integrated Federal Energy Policy, Executive Office of the President, November 2010, available at, p. viii. PCAST proposes that $10 billion in energy RD&D funding (out of a proposed expansion of $16 billion) be derived from new funding streams.

16. Excluding Recovery Act (stimulus) funding the DOE laboratories in 2010 accounted for 75% of total DOE spending on science and 45% of spending on applied research, development and demonstration (John M. Deutch, “An Energy Technology Corporation,” 13). The laboratories’ efforts in energy technology commercialization have long been criticized, although there is evidence that this aspect of their performance has improved. (See Adam Jaffe and Josh Lerner, “Reinventing public R&D: Patent policy and the commercialization of national laboratory technologies,” RAND Journal of Economics 32, no.1 (Spring 2001): 167–198).

Related Reads