One of the most interesting—to me, at least—sessions at the 2015 EmTech MIT conference was a panel on energy, involving experts in solar, battery storage, and nuclear energy. Indeed, I would say this, coupled with the artificial intelligence and robotics panels, were the panels I enjoyed the most.

Of them, again due to personal interest and background, the nuclear energy session led by Leslie Dewan of Transatomic Power was particularly interesting for the following reasons:

  • There is only one truly scalable zero carbon emission, dispatchable energy source that humanity has ever successfully deployed, and that is nuclear energy. Specifically, nuclear fission.
  • The U.S. stands on the cusp of a major decision: to carry on developing nuclear energy and pursuing one of the most promising avenues for zero-carbon energy productions, or close off this avenue and be left with much lower energy density and intermittent energy sources (solar and wind).
  • The world is not standing still: even if the U.S. and the developed economies of Western Europe and Japan decide to cease building new (and potentially safer) nuclear reactors, the rest of the world—particularly China and India—will continue to do so.

Now, several thoughts came to mind as I was listening to the panel, the Q&A, and later when I had time to consider what I had heard and synthesize it with what I knew from other sources. These are:

  • Cost
  • Regulatory Hurdles
  • Safety
  • Competition

Cost

Investing in disruptive—rather than incremental—energy-related startups is hard. And very expensive. One of the key takeaways from the panel was the sheer cost of getting a reactor design to the point where it would be ready for submission to the Nuclear Regulatory Commission: approximately $100 million as at 2007.1 Compare that against the very modest capital requirements of most startups that venture capitalists invest in, and it becomes clear that this is not a field for the fainthearted.

Moreover, the cost of submitting a design to the NRC is not the only cost that a startup in this field will face: it will need to obtain laboratory space, materials, hire engineers and scientists, engage experts to assist it in navigating the regulatory mazes that need to be traversed to get a nuclear reactor design approved. It will go through the inevitable false starts, unexpected problems, and changes of direction in response to new data that any new technology will face.2 Moreover, once the technology has been validated at the “laboratory” stage, it must then be scaled to a commercially viable reactor design. That, too, needless to say, costs money. Indeed, I would expect that financing a nuclear energy startup to commercial viability may require investments an order of magnitude greater than the $40 – 50 million that VCs typically invest to get their startups to a successful exit.

Now, part of that cost will likely be absorbed by research grants, government programs, and collaboration with utility companies, but that still leaves a very substantial amount to be covered by the startup’s equity capital, i.e. its VC investors. This is one of the reasons why many energy related startups fail—they do not have enough capital to adequately cover their costs until they begin generating revenue. This was especially true of many of the cleantech startups that received VC funding in the early 2000s, no small number of which ran into a “valley of death” between early stage investment to validate the technology and later stage investment to clear regulatory hurdles and establish commercial viability. Indeed, the difficulty of successfully funding a cleantech startup has been one of the reasons for pullback from “cleantech” VC investments among many of the VCs in Silicon Valley, and I would say a similar dynamic can be expected for nuclear energy startups: given the high cost of achieving commercial viability and the risk that the technology will not prove viable, it will be difficult to obtain VC funding.3

That being said, according to Ms. Dewan, there are approximately 55 nuclear energy related startups that have received a grand total of $1.6 billion in funding as of mid-2015. Now, that number is impressive but it is, without a doubt, a fraction of the total funding received by startups in other fields.

Peter Thiel’s Founders Fund has been one of the few VCs willing to invest in advanced nuclear reactors, having invested $2 million in Transatomic Power through its FF Science vehicle, as well as participating in a follow-on $2.5 million financing round with Acadia Woods Partners and Daniel Aegerter. The other key investors in other nuclear energy startups have included Bill Gates, Khosla Ventures, Charles River Ventures, Venrock, Presidio Partners (formerly CMEA Capital), and Mithril Capital Mangement.

Long lead time to minimum viable product

Ms. Dewan mentioned briefly that she anticipated that Transatomic Power would require four years from November 2015 to reach its first working prototype 520MW molten salt reactor (i.e. circa 2020), for a total of ten years from founding to first working prototype. Commercial deployment might not happen until eight years from November 2015 (i.e. circa 2024). Similarly, Terrapower and the China National Nuclear Corporation plan to build a 600MW working prototype reactor in China between 2018 and 2022, with commercial deployment of 1150MW commercial-scale reactors in Asia in the late 2020s. These timeframes suggests that the—likely optimistic—time from the startup’s founding to commercial deployment is likely to be in the range of ten to fifteen years (Terrapower was founded in 2008 and Transatomic Power in 2010). With the inevitable planning fallacy that usually arises even when very intelligent people are asked to estimate the timeframes for completing a project, I’m inclined to think that these timeframes may be optimistic.4

Nevertheless, even assuming that these timeframes are reasonably accurate, they still suggest one problem for VC investments in this space: the majority of nuclear energy startups will require more time to reach a minimum viable product (i.e. a prototype reactor) than the fund life of most VC funds. While a VC fund may exit an investment at some earlier stage in the development of the technology, this is not ideal. One should expect that some VCs will shy away from investing in this space due to the risk that they might reach the end of their fund life without a viable exit opportunity for their investments in nuclear energy startups.

Regulatory hurdles

The process of getting a nuclear reactor design approved NRC has been described as difficult, to say the least. Construction of a nuclear reactor requires at least two NRC approvals:

  • Design certification: Approval of a nuclear power plant’s design, independent of any application to construct or operate a power plant based on that design.
  • Early site permit: A permit approving a specific site for the construction of a nuclear power plant, independent of any construction permit that may subsequently be granted. The early site permit does not authrize construction of a nuclear power plant on the site.
  • Construction permit: A permit approving the construction of a nuclear power plant at a specific site, which is issued prior to the issuance of an operating license permitting the operator to operate the completed nuclear power plant.
  • Operating license: A license permitting an operator to operate a completed nuclear power plant. This can be combined with a construction permit in a “combined license” to allow an entity to construct and operate a nuclear power plant on a specific site.

For most nuclear startups, the design certification is the key regulatory hurdle they must overcome. As both Ms. Dewan and the Government Accountability Office have noted, the cost of a design certification application can easily exceed $100 million. The cost is by no means the only problem, the time required to obtain design certification is another problem. Once the startup obtains design certification, it then needs to obtain site permits, construction permits and operating licenses to build and operate a prototype reactor based on its design. Each subsequent nuclear reactor built from that startup’s design must in turn obtain separate site permits, construction permits and operating licenses. Inevitably, the licensing process adds significantly to the risks that investors (whether VCs or strategic investors) must factor into their investment thesis and expected returns.

While the Department of Energy, through the Gateway for Accelerated Innovation in Nuclear program can assist nucler startups to navigate the regulatory process and provide financing (typically loan guarantees) to cover the costs of license applications, it remains to be seen whether the assistance provided will suffice to improve the experience (and streamline the process) for nuclear startups.

Indeed, the difficulty of complying with the NRC is one of the reasons why Terrapower in particular is building its prototype reactor in China, while Transatomic is exploring options to deploy commercial reactors based on its molten salt reactor design in countries with more favorable regulatory environments, including China and, possibly, Singapore. This has profound implications for the development of the U.S. nuclear energy sector, particularly in terms of developing the expertise necessary to construct and operate advanced nuclear reactors (an entirely different skillset from designing such reactors).

Safety

Beyond the regulatory hurdles that a nuclear startup must navigate, there is the ever present issue of safety. This encompasses the safety of the design, i.e. the existence of failsafes that will contain the radioactive material in the event of a disaster, the safety of the materials used to construct the design (notably risks relating to fatigue, corrosion, and weaknesses induced by bombardment of the materials by high energy neutrons generated from the fission process), and the safety of the site in the event of natural disasters and terrorism.

While some of these safety issues are more properly the responsibility of the utility company that is operating the nuclear reactor, the design and materials safety issues are ones that most nuclear startups will have to address.

Competition

The competition to develop advanced nuclear energy is ongoing. As Ms. Dewan noted, there are more than 50 startups worldwide working on nuclear related technologies. There are also several different reactor designs being explored:

  • Pebble bed reactors—HTR-PM—being developed in China by Tsinghua University, China Nuclear Engineering Corporation and China Huaneng Group
  • Traveling wave reactors being developed by Terrapower
  • Molten salt reactors such as the one being developed by Transatomic Power
  • Supercritical water-cooled reactors being developed in China, Canada, and the European Union
  • Sodium-cooled fast reactors being developed in the European Union and India
  • Gas-cooled fast reactors being developed in the European Union
  • Reactors based on the thorium fuel cycle being explored in India and China

It is entirely possible, though, that there is enough room in the world for multiple reactor designs. Certainly there may be site-specific considerations that result in one design being superior to others, but with the continued need for stable baseload power that is also zero carbon emission, it seems to me that there is room for several different reactor designs to coexist. Indeed, it seems that India in particular may find it preferable to develop reactors based on the thorium fuel cycle due to its abundant sources of thorium, while other nations that are not so endowed may find reactors based on the uranium fuel cycle preferable. The problem, from the perspective of an investor in a nuclear startup, is that this suggests that a successful nuclear startup, despite spending significant sums to get its design certified in multiple countries with their own certification processes and safety standards, might not be able to obtain enough contracts from utilities to generate returns sufficient to justify the risk of investing in such startups.

One final potential risk to advanced nuclear fission reactors startups is the—slight but non-zero—possibility that commercially viable nuclear fusion might be developed at a time when advanced nuclear fission reactors are themselves just beginning to reach commercial viability. I will admit that I find this possibility highly unlikely, but it is worth considering. Moreover, it strikes me that if nuclear fusion became commercially viable in the next two to three decades, those nations that have not already heavily invested in advanced nuclear fission reactors might be the best positioned to “leapfrog” directly to fusion reactors. From my perspective, as someone not invested in any of the current crop of nuclear startups, the prospect of viable nuclear fusion reactors within my lifetime is exciting. It could be the key to managing human carbon emissions and its impact on health and the climate5 without compromising on our energy demands.

Conclusion

I remain convinced that nuclear energy is one of the “energy miracles” humanity will need to achieve in order to build a sustainable, high-technology civilization. There are, however, serious difficulties with traditional VC funding of nuclear startups. One thought that I have had recently has been on the role of truly patient capital—such as family offices and sovereign wealth funds—in the funding of nuclear startups and other technologies with very long lead times to minimum viable products. It seems to me to be an interesting area to explore in greater detail, outside of this post.