Six principles for energy innovation

Decades of experience must inform future initiatives, urge Gabriel Chan and colleagues.
Large floating solar farm project under construction on a lake

A solar farm floats on a lake that formed after the collapse of a deep coal mine in Huainan, China.Credit: Kevin Frayer/Getty

Last month, the European Union marked the tenth year of its Strategic Energy Technology Plan. It is one of many policy initiatives worldwide to accelerate innovation in energy technologies to reduce greenhouse-gas emissions. As the window of opportunity to avert dangerous climate change closes, we urgently need to take stock of these initiatives — what works and why?

Public investments in energy research, development and demonstration (RD&D) have risen since the low levels of the mid-1990s and early 2000s. In 2016, member countries of the Organisation for Economic Co-operation and Development spent US$16.6 billion on energy RD&D, compared with $10 billion in 2000 (adjusted for purchasing power parity). In October, the United Kingdom set out its Clean Growth Strategy to invest more than £2.5 billion ($3.3 billion) in low-carbon innovation between 2015 and 2021. In 2015, the EU and 22 nations pledged to double their investment in energy RD&D under the Mission Innovation adjunct to the Paris climate agreement. However, the overall goal might be out of reach given the proposed 35% cut in US President Donald Trump’s 2018 budget for energy RD&D.

Different nations are pursuing various strategies and creating new types of institution (see Supplementary Information). For example, the Advanced Research Projects Agency-Energy (ARPA-E) run by the US Department of Energy (DOE) targets grants at key technologies such as affordable energy storage. The DOE Energy Innovation Hubs form research teams to work on technologies such as nuclear-reactor modelling.

The United Kingdom has set up the Energy Technologies Institute (ETI), a public–private partnership to accelerate the development of low-carbon technologies. It also launched the Catapult programme, which aims to build bridges between universities and industry, and sustainability advisory services that are run by bodies such as the Carbon Trust. And China is reforming the Chinese Academy of Sciences and its national labs, as well as creating larger lab facilities.

At the international level, the United Nations Framework Convention on Climate Change (UNFCCC) Technology Mechanism enables technology development and transfer in developing countries to support the Paris agreement. Since 2013, the World Bank Group has opened seven climate-innovation centres in developing countries such as Kenya. The centres provide seed financing, policy guidance, networking and technical training. The Nairobi centre, for example, advises start-ups such as Futurepump, which is developing solar-powered water pumps.

Most of these bodies can claim successes. But a comprehensive global assessment of energy-innovation programmes is needed to learn from collective experience and to establish best practices. As a starting point, here we distil six principles to guide public initiatives for energy innovation. These are drawn from the scholarly literature and from third-party assessments of experience in UK, US and multilateral institutions.

Guiding principles

Give researchers and technical experts autonomy and influence over funding decisions. Active scientists are better placed than managers to spot bold but risky opportunities. For instance, US national laboratories lead the development of a subset of projects that comprise 4% of their current budgets. Yet these projects produce more high-impact publications and commercially viable technologies than do those that are controlled by DOE headquarters (see ‘Expert benefits’)1.

Source: Ref. 1

Such decentralized funding at the National Renewable Energy Laboratory in Golden, Colorado, has supported the development of more cost-effective methods of cultivating algae for biofuels, as well as groundbreaking research on perovskite-based solar cells2.

Public labs that conduct energy RD&D should allocate a significant fraction of their budgets, say 10%, to internally selected projects. They need the flexibility to adjust goals as research proceeds. Funding institutions could follow the approach of ARPA-E and employ technical experts as programme managers to direct funds and to modify or cut projects as they progress3.

Incorporate technology transfer in research organizations. Public institutions that fund or perform energy RD&D must collaborate with private owners of energy infrastructure, as well as those that produce, deploy and operate new energy technologies. Otherwise, research can remain in silos and might never be put into practice. Formal technology-transfer programmes should be set up to build connections. This requires strong institutional backing. When political and financial support wanes, technology-transfer rates fall1.

Formal programmes for technology transfer have built on the work of DOE national laboratories4. For example, since 1994, one-fifth of all new patents in advanced energy-storage systems for vehicles cite at least one DOE-granted patent5. Strategies are needed to innovate faster. Research universities have shown the value of sustained collaboration through a diversity of channels6. Sandia National Laboratories, which has facilities in New Mexico and California, has seen the value of giving researchers up to three years’ leave to work in the private sector and commercialize technologies. Pilot programmes should be scaled up. For example, at California’s Lawrence Berkeley National Lab, the Cyclotron Road and Visiting Entrepreneurial Research Fellows programmes lower barriers to collaboration and provide facilities, expertise and funding to entrepreneurs.

Focus demonstration projects on learning. Many viable technologies stumble at the demonstration stage when they reach the ‘valley of death’. Companies are reluctant to finance pilot projects for new, risky technologies, such as carbon capture and storage (CCS). This makes it impossible to scale them up without public support. Demonstration projects are expensive and can be harshly judged. For example, the US Synthetic Fuels Corporation fostered technologies in the 1980s to create liquid fuels from substitutes such as coal. The failure of the programme to meet its goal of reducing oil imports has been used to argue against public investments in demonstration projects that aim to pick winners. Yet the programme created useful knowledge: technology trialled at the corporation’s Cool Water plant is being considered for use in CCS7.

Policymakers should set goals for demonstration projects on the basis of the knowledge they will generate about the cost and performance of future technologies7. Other important features include: an exit strategy to halt projects that miss milestones; design that acknowledges the possibility of failure while keeping other options open; involvement of a broad pool of private actors; and mechanisms to track and disseminate the knowledge produced8.

Incentivize international collaboration. International cooperation can accelerate innovation beyond the capabilities of a single nation. Pooling costs enables projects of greater scale, lessens duplication and integrates regional specializations. But more needs to be known about how to do this effectively. Few multilateral collaborations stretch beyond holding meetings and issuing joint statements. Deeper collaborations range from loosely coordinated pledges for domestic actions, such as Mission Innovation, to shared platforms for technology development, such as the International Energy Agency (IEA) Technology Collaboration Programmes. Some partnerships achieve integrated cooperative RD&D — 35 countries are involved in the ITER project to build the world’s largest magnetic fusion device in southern France.

It can be fruitful for nations that have specific technical expertise to partner with those that are keen to exploit it. For example, the U.S.–China Clean Energy Research Center has helped US companies such as 3M to test technologies in China to improve the energy efficiency of buildings. The pace and scale of construction in China meant that US companies learnt more about real-world effectiveness than they would have done working only in the United States.

Barriers remain: collaborators must negotiate rights before outcomes are known, partners may lack trust, and domestic political support can fluctuate. Face-to-face interactions, long-term strategies and well-designed management plans are essential9.

Adopt an adaptive learning strategy. Lessons must be drawn from a diverse range of experiences because energy innovation occurs in many different industrial and funding contexts. Efforts vary in their primary goals, such as competitiveness, security and environmental protection, as well as in their implementation strategies.

Mechanisms for evaluating and adapting programmes should be designed into institutions from the start. There are many ways to measure innovation-policy outcomes: from the money invested or the number of papers, citations, patents and start-ups that are generated, to economic measures such as productivity and qualitative measures that can be assessed through surveys. Public agencies should store and track data on operations and outcomes and release them to independent researchers.

New groups of experts might be needed. For example, the UK Behavioural Insights Team, created in 2010, incorporates findings from behavioural psychology into policies that encourage the use of energy-efficient heating and lighting systems. International institutions such as the IEA, the International Renewable Energy Agency, the UNFCCC Technology Mechanism and the World Bank should help governments to learn from others and develop strategies for adapting energy-innovation programmes.

Keep funding stable and predictable. Government funding for energy innovation is, in many cases, volatile. For example, between 1990 and 2017, US political shifts meant that each year, on average, one in five DOE technology areas saw a budget increase or decrease of greater than 30% (see ‘Volatile funding’ and Fluctuations in funding erode the cost-effectiveness of programmes by precluding strategic, sustained investments that are high risk but potentially high reward. A slashed budget for renewables in the 1990s led to the loss of decades of experience during layoffs at the US National Renewable Energy Laboratory.

Source: Ref. 10

Institutions for energy innovation have evolved just as erratically. In the United Kingdom, each prime minister since 2000 has focused on a different strategy. Tony Blair created the Carbon Trust, Gordon Brown the ETI, David Cameron the Catapult programme, and Theresa May has created a Faraday Challenge for batteries as part of the Industrial Strategy Challenge Fund. Although experimentation has benefits, there are also costs. Learning how to work with new programmes and people takes time and effort. For example, early engagers with the Carbon Trust applied for grants and incubator support, only to see the programme’s scope limited in 2011 to providing advice and certification services.

Rather than overhauling institutions for energy innovation with different political cycles, existing programmes should be continuously evaluated and updated. New programmes should be set up only if they fill needs that are not currently met.

Let’s learn from experience to accelerate the transition to a cleaner, safer and more affordable energy system.

doi: 10.1038/d41586-017-07761-0
Nature Briefing

Sign up for the daily Nature Briefing email newsletter

Stay up to date with what matters in science and why, handpicked from Nature and other publications worldwide.

Sign Up


  1. 1.

    Anadon, L. D., Chan, G., Bin-Nun, A. Y. & Narayanamurti, V. Nature Energy 1, 16117 (2016).

  2. 2.

    US Department of Energy ASCAC Subcommittee. First Report to the Advanced Scientific Computing Advisory Committee (11 April 2017); available at

  3. 3.

    National Academies of Sciences, Engineering, and Medicine. An Assessment of ARPA-E (National Academies Press, 2017); available at

  4. 4.

    Chan, G. Essays on Energy Technology Innovation Policy. PhD thesis, Ch. 2, Harvard Univ. (2015).

  5. 5.

    Ruegg, R. & Thomas, P. Linkages of DOE’s Energy Storage R&D to Batteries and Ultracapacitors for Hybrid, Plug-In Hybrid, and Electric Vehicles (2008); available at

  6. 6.

    Mowery, D. C., Nelson, R. R., Sampat, B. N. & Ziedonis, A. A. Ivory Tower and Industrial Innovation (Stanford Univ. Press, 2004).

  7. 7.

    Anadon, L. D. & Nemet, G. F. In Energy Technology Innovation: Learning from Historical Successes and Failures (eds Grubler, A. & Wilson, C.) Ch. 19 (Cambridge Univ. Press, 2013).

  8. 8.

    Nemet, G. F., Kraus, M. & Zipperer, V. Discussion Papers of DIW Berlin 1601 (2017); available at

  9. 9.

    Lewis, J. I. Energy Policy 69, 546–554 (2014).

  10. 10.

    Anadon, L. D., Chan, G. & Lee, A. In Transforming U.S. Energy Innovation (eds Anadon, L. D., Bunn, M. & Narayanamurti, V.) Ch. 2 (Cambridge Univ. Press, 2014).

Download references

Supplementary Information

  1. Supplementary Information