Nuclear startups should stop planning to sell energy
The nuclear sector needs to move away from utilities
This week’s economic turmoil has been a nearly constant topic of conversation among my friends and colleagues. It’s got me thinking more about alternative types of investments, like energy. The way I’m most used to thinking about energy as a sector is through the lens of nuclear energy startups.
Nuclear Fission Market Map
It’s good to be back on campus at the University of Chicago! Classes at the Booth School of Business don’t pick up again until the end of the month, but I came back early to help my colleagues who are just starting in the MBA program prepare for recruiting.
One of the startups I mentioned in my market map last fall, Valar Atomics, made headlines last week when it joined several state Attorney Generals and other nuclear reactor companies in a lawsuit against the Nuclear Regulatory Commission (NRC). At issue is basically the scope of the NRC’s authority — can it regulate all types of commercial nuclear plant, or no?
Valar Atomics
That led me to read up on the firm more generally, which raised a monster seed round earlier this year. According to TechCrunch, Riot Ventures led the $19 million round; AlleyCorp, Initialized Capital, Day One Ventures, and Steel Atlas participated.
Valar Atomics is planning to do something really curious with its reactors. They’re not going to generate power directly.
Isaiah Taylor, the founder and CEO, detailed all of this in a rather long post on X, which is audacious — by publicizing his idea, he adds material execution risk. A key issue he’s got to worry about now is whether or not he can implement his vision faster than his competitors.
That led me down a rabbit hole into the nuclear industry more broadly.
Nuclear industry history
The American nuclear industry has its roots in World War II. I’m not going to get technical in the history of it, but Destin Sandlin is in the process of creating a long-form YouTube series about nuclear power in the US on his channel SmarterEveryDay. Destin’s an engineer by training and an excellent science communicator (this is not the first time I’ve written about his content). The first video in his series about nuclear power is live and can be found here. It’s a walkthrough of EBR-I, the first fission reactor to generate usable electricity.
During the Cold War, the number of nuclear power plants grew rapidly, before levelling off as that era wrapped up.
The causes of this appear to have been:
a declining growth rate in electricity use
cost overruns in construction reduced the appetite for new facilities
a massive municipal bond default from a key customer
During the end of the 20th century, more than 120 reactor orders were cancelled. All these reactors were fission reactors, which create energy by splitting atoms and turning heavy elements into lighter elements.
Throughout this time period, we’ve also been “just a few years away” from commercially viable fusion reactors, which create energy by using heat from nuclear fusion, where lighter elements become heavier elements.
The overwhelming majority of nuclear reactors in the US are used for power production. Notable exceptions in the US are government-owned, and include:
Naval reactors, nuclear power plants used to drive Navy ships and submarines (and also generate power for onboard consumption)
More than 500 reactor cores have been brought into operation by the US Navy
Nuclear powered spacecraft
The US only flew one reactor, SNAP-10A, in the 1960s
The US today uses radioisotope thermal generators to power spacecraft — which are not reactors in the traditional sense, but do generate heat from radioactive decay which is converted into power
The USSR flew about 40 reactors to space during the Cold War
Breeder reactors produce more nuclear products than they start with
To my knowledge, there are no active breeder reactors in the US
Nuclear power is an interesting place to play
I think nuclear startups are interesting for three main reasons:
If everything works out, startups in this space will not only make a lot of money for investors, but they will also make humanity in general (and the US in particular) a better place to live
The engineering is devilishly difficult, which attracts interesting, passionate, and competent people
There are really big regulatory challenges, which can turn into compelling and defensible moats
These are similar to the reasons I find myself intrigued by many deep tech sectors.
I want to double-click on that last point. The NRC runs a remarkably thorough process to make sure that the general public remains safe when new nuclear reactors are built and brough online. It is burdensome for reactor owners. This is probably a good thing for reactor owners, as it provides them with a real competitive moat — even if it moves about as fast as molasses and is expensive to comply with.
That’d be really awesome if, and only if, nuclear reactor operators could sell their power at whatever price the market would bear — and that price is high enough to justify reactor construction and operating costs.
That’s not the case right now.
Nuclear reactors produce utilities
Nuclear rectors generate power at a cost of around $0.10/kilowatt-hour. Lots of this cost is attributable to the immense expenses of the regulatory process and reactor construction. Other methods in use today, like fracking, cost less than half that.
In theory, the government could just price the electricity from different sources differently, but that’d simply function like a subsidy to the nuclear industry. This might make more nuclear power available, but it seems unlikely to incentivize the design and development of cheaper power-producing nuclear reactors.
Even if the costs of nuclear power were competitive, electricity prices aren’t really a function of the free market.
This is because of just how critical to society electrical power has become.
The government, on several different levels, regulates what utilities vendors can charge for things like electrical power. There are different regulators for the government through the Department of Energy at the federal level, state public utility commissions, and the Federal Energy Regulatory Commission — which regulates interstate electricity rates.
Nuclear reactors have the potential to produce commodities
What the nuclear industry needs in order to take off is to sell products where it has more pricing power. That’s not to say it should be doing non-repeatable things, but it’s got to stop selling utilities, and start selling commodities.
When I talk about commodities, I’m using the term in a classical sense. I mean a material or product that is substantively fungible — it won’t matter to buyers who produced it.
Utilities and commodities can both work at venture scale because these sorts of products can have a very low marginal cost. The best example here is probably that a super low marginal cost is a large part of what makes SaaS investments so lucrative to VC investors.
Commodities are far more compelling as an investor because the company has more pricing power due to less government regulation.
Valar Atomics’s website says the firm has a focus on “grid-independent products: hydrogen, data center power, heavy industrial power, and clean hydrocarbon fuels.”
It’s starting to make this transition by focusing on providing electrical power separate from the grid — which may be a play to avoid utility regulators.
The company’s plan to produce hydrogen and hydrocarbons is really what’s piqued my interest, because the government regulations on those sorts of products are much more similar to taxes than price caps.
The other thing Valar Atomics is doing that’s noteworthy here is that they’re making moves towards commoditizing the reactor itself. By building them in large quantities, the plan appears to be that they’re turning the power sources into, if not a commodity, at the very least a high-value, high-throughput product (like an iPhone).
When we talk about what it means to reindustrialize at the national level, we have to acknowledge that this is going to require new sources of materials, new facilities, new processes, and new services. There’s potential to integrate nuclear reactors very closely with energy-intensive sites that provide these things.
The end game, and I think Valar Atomics is looking to contribute to this as part of its long-term plan, is using nuclear reactors and their byproducts directly to enable reindustrialization at scale.
This is a microcosm of the bull case I envision for the nuclear sector in the coming ten to twenty years.
If you invest in startups, like my writing, and want to chat about possible career opportunities, don’t hesitate to be in touch. The best way to get ahold of me is via email: