You published seven papers in the Journal of Plasma Physics, which is unusual because most private fusion companies keep their physics private and guard it closely. Why publish, and what did it solve for you?
Just to be clear, I don't do the math or science for the company, but I read every word of all seven papers looking for typographical errors, and I found some. One thing it made very obvious to me is that there's a whole lot about the universe that I don't understand. I've been in this sector for about nine years, mostly focused on the commercial and regulatory side. It cannot go without saying that the credibility of a fusion company from a technical and scientific standpoint is largely based, right, wrong or indifferent, on the peer-reviewed information that has been put out about the company. So we felt it was tremendously important to get our design basis and physics basis for a power plant into the scientific community and have it reviewed and published. Six of the seven papers were peer reviewed, one was an overview paper, and together they go into all the primary systems of a fusion power plant. We wanted to set a baseline showing that we had thought about every component.
We're in the business of developing a commercial product that customers can actually use to make money with, whether that's selling megawatts or powering a data center. So on your second question, there were a lot of preemptive steps around IP. There are protections in place to cover some of the concepts. We obviously did not go into the very fine detail of how we're going to do everything in those papers, but the publisher felt it was the new gold standard, that was their quote, for how you do this. We took the necessary precautions from an IP protection standpoint on any of the secret sauce in the different systems to make sure we were protected.
The TVA agreement is striking for a private fusion company, a signed agreement, a specific site, a filed regulatory application, a construction date. How did that relationship come together, and what did TVA's internal evaluation of the stellarator look like?
I'm about a week short of my third year anniversary with the company, and one of the first things I was asked to do when I got here was find a home for what we then called a risk reduction platform, the large-scale verification test bed that's now Infinity One. We looked around the country, focused on four states and 18 locations, and engaged Tennessee Valley Authority about areas where they had land or buildings. They said they were going to retire a coal plant in a few months and would be utilizing some of the infrastructure for another project, so it might be a good fit. At the same time, Governor Lee had just created a nuclear infrastructure fund to attract energy technology companies. We were actually the first awardee of that fund, and it helped offset some of the infrastructure costs for Infinity One.
So TVA got a front row seat into the creation of our team, both technical and commercial. Our CEO Ren Jia ran hydro business in Europe for years, was a nuclear power plant operator, and was the founder of one of the first small modular reactor companies. Below leadership, we have people who have built things, not just scientists and engineers, but builders. TVA got comfortable with how we worked together on Infinity One, and we started talking about whether there was enough space at the site for a power plant. Early last year they started an internal survey of their operational area, which actually touches seven states, and came up with a list of potential sites for a fusion power plant. Bull Run was on that list because we were already there, the plant was in decommissioning, and they had the grid interconnection, the water source, and 760-plus acres of land. They set up evaluation teams, brought in outside experts, took seats at the table, and got a crash course in what we were doing. That led to the letter of intent we signed with them last year to build one or more Infinity Two units, and on the same day we got our first commercial contract from TVA to start scoping the path to a final investment decision when Infinity One turns on. To be clear, we do not yet have a contract with TVA to build a fusion power plant, but TVA also hasn't given GE a contract to build an SMR at Clinch River, even though they're developing that site.
TVA wants construction at Bull Run as early as 2028. What does a brownfield site unlock that a greenfield fusion site never would, and what has to be true between now and 2028 for that date to hold?
What has to be true is that we continue to hit our developmental milestones. The technology, whether it's materials or physics, has to keep moving down the field toward a power plant we can actually go build. Some of those milestones are confidential and others aren't, but we have to stay on schedule. The 2028 date itself is the beginning of construction on the conventional island, not the fusion island. If you think about fusion power, it's a heat source. It creates heat, which creates steam in a steam generator, and once we understand the conditions of that output everything downstream is conventional technology. A steam turbine you can buy from Siemens, GE, Toshiba or Mitsubishi, and a generator set. We call that the conventional kit.
So as we develop our fusion power plant and get more detailed in the steam output, there's a high degree of certainty around what type of turbine and generator you need. The 2028 date is the start of construction of that conventional island. We will know our steam conditions by then so that TVA can move forward with at least the conventional line on the plant in 2028. The brownfield piece matters because a coal plant gives you grid interconnection, water, and a site already used for power generation. You inherit infrastructure that would take a decade to build from scratch.
You chose to go with TDEC rather than the NRC for the first fusion licensing application back in January. Why is that the right pathway, and why does it hold federally?
It holds federally because fission is literally the opposite of fusion. Fusion is really, really hard to start and easy to stop. Fission is splitting atoms in a chain reaction. It's happened naturally somewhere thousands of years ago and propagates until it runs out of fuel or gets moderated.
The consequences of failure of a device or safety system in fission are very different than in fusion. If any parameter of our device is off, it stops. It just quits.
The only radiological nuclei on our fuel side is tritium, which is in hundreds of thousands of exit signs across the United States and in many high-end watches. Our spent fuel from deuterium and tritium is helium-3 plus a neutron. So the source of the hazard is fundamentally different.
The Nuclear Regulatory Commission staff recognized that, the commissioners unanimously said this needs to be treated differently, more like a neutron source or nuclear medicine in a hospital, which is regulated under 10 CFR Part 30. That allows the 40 agreement states to regulate fusion, and Congress codified it in the Advance Act last year. Tennessee took the lead, the governor put money into the program for TDEC, and we submitted our application to operate Infinity One last month. They have six months by rule to turn it around, and the fee is $50,000. The flexibility this gives us is enormous. Under Part 30 with the agreement states, I do not need a design license, a construction permit, a site license, or a fuel design license from the NRC. In fission, a design license alone takes about 10 years and roughly $500 million. The BWRX-300 doesn't have that license yet in the US. That kind of constraint also limits flexibility, because if you find a better component a year before turning the plant on, you have to amend the license. This brings us to siting versatility. A coal plant doesn't have a nuclear site license, so you can't build a small modular reactor there without a lot of work. But fusion can site at all those retiring coal plants, and at the rural locations where data centers want to be, without going through a multi-year NRC site licensing activity.
Looking at the mid-2030s, when this technology is working, where does fusion actually fit in the energy mix and what is it displacing?
First, it's not a zero-sum game. There are enough megawatts to go around for any new source of energy. Even before AI, demand was increasing tremendously because of electrification of everything. How many devices are you wearing now that need a battery charger? You go to a concert with 100,000 people and they're all holding up a phone videoing it, and they all have to charge that phone later. Then AI accelerated demand on top of that. Until processors get more efficient, they're going to need more of them. So we don't have to displace anything, there's room for us to fit in.
Where we fit is firm baseload power. It's going to be very hard to build new coal in the United States, though there are life extensions out there for some plants. As a firm source of energy, that's where fusion will sit. With our fuel cost virtually zero over the life of the plant, I think we will become least-cost generation almost as soon as we get the first one up. Fusion is tremendously energy dense, three or four times more than fission, a few hundred times more than coal. We don't need acres and acres for an 800 megawatt plant, and our first-of-a-kind design is around 400 megawatts. As we learn more, we'll be able to remove some of the margin and get more energy out of the same footprint with the same machines. It's utility scale, not rooftop, not micro, not on the back of a truck. We're trying to solve one of the biggest gaps in energy demand.
Fusion seems to be chasing the same utility dollars as SMRs and AP1000s a decade from now. Do they compete or stack, and which one does the heaviest lifting?
A decade from now we should have two years of operation in our first plant. Up until that point, the AP1000s, they built them in Georgia, they may restart the project in South Carolina. The Georgia project was extensive, the first two units, so there is operational experience. But they are tremendously expensive, and that goes back to the regulatory requirements, which bleed into the supply chain requirements. I started my nuclear career as a supplier of pipe, valves, fittings and steel for the nuclear renaissance that was supposed to be imminent in 2006. Of the 30 reactors proposed, only a handful got built, and there was tremendous strain on the supply chain. There are only three or four companies in the United States that can supply qualified parts because of the quality assurance codes and standards for fission.
That has not been applied to fusion, and we don't expect it to be, because the consequences of failure and the hazard to the public are vast orders of magnitude less. That opens our aperture for common kit to virtually any supplier out there. There are some unique items, high-temperature superconducting tape crafted into cables for magnets, and external heaters called gyrotrons. There are two or three companies in the world that make gyrotrons, but if I'm going to order 150 of them in the next couple of years, I'd expect more companies to figure out how to make them. If I can bring an order book of multiple units to the market, the supply chain will ramp up very quickly compared to the challenges that faced the fission sector. The other differentiator that's sometimes lost is that we are a fusion OEM. We are not a build-own-operate-maintain company. We focus on designing the technology and delivering it through fusion direct partners, the best-in-class manufacturers. TVA's role is to operate, potentially own and maintain. We stay in our lane. Our model is the traditional technology OEM, like a Siemens or Westinghouse or GE selling equipment to generate power, and we'll be speaking with Siemens in the next couple of weeks as well.
