Interview: Pentagon's Jeff Waksman on Project Pele Microreactor

The US Department of Defense's (DOD) Project Pele program last year awarded BWX Technologies (BWXT) a $300 million contract to design, complete and deliver a full-scale transportable 1 to 5 megawatt equivalent microreactor prototype in 2024 for testing at the Idaho National Laboratory. The Pentagon is targeting 2027 for the deployment of the microreactor on a military installation site. DOD's Strategic Capabilities Office (SCO) Pele Program Manager Jeff Waksman spoke with Energy Intelligence's Jessica Sondgeroth to provide a progress update on the microreactor project and BWXT's TRIstructural-ISOtropic (Triso) fuel line. Waksman also delves into Pele's objectives and applications for the US military and beyond, including its implications for the commercialization of the advanced reactor industry.

Q: In June, the Pentagon's SCO awarded a $300 million contract to BWXT to complete and deliver a full-scale transportable high-temperature gas-cooled microreactor prototype in 2024 for testing at the Idaho National Laboratory. Where are you now with this project?

A: The next big regulatory hurdle is getting preliminary approval of the design by the Department of Energy [DOE]. In the DOE regulatory world, it’s called a PDSA [a Preliminary Documented Safety Analysis]. We are working to finalize the last details of the design over the next few months so that we can get DOE approval. Nuclear reactors are complicated integrated things. You can’t say that the design is done until literally every part of the design is done. Because there's always a risk that a change to one part of a reactor will cascade and affect any other part of it. So it's one of the things that makes nuclear reactors hard, but we think we're making good progress and hope to get that approval from the Department of Energy later this year.

Q: What are a couple of the bigger challenges when it comes to finalizing a design like this?

A: Most of the challenges come down to one of two areas. One is that we have an almost nonexistent supply chain in the United States because we haven't been building reactors outside the Navy for the last several decades. So for some of the materials or components, they're often custom one-off jobs. There are often very long, long lead times for some of these components. And in some cases, what you want to make, you simply can't make because there is nobody in the United States who can make it.

The other challenge comes from the fact that these high-temperature reactors operate in different physics regimes from the traditional light-water reactors that the United States currently operates. Because of that, we are often heavily reliant on models rather than data to demonstrate the safety of the system and that necessitates either larger safety margins, or larger margins of uncertainty in the design. For instance, if you have two materials that are going to be pressed up against each other in the core, and you're not exactly sure how they're going to expand or contract under those extreme conditions, you need to have margins in your design to make sure that nothing breaks when you're operating at extreme temperatures.

This is one of the real benefits that Project Pele will bring once this reactor actually operates: We'll have real-world data on how these materials operate under those conditions and it'll help validate and improve the models for any other future high-temperature gas reactor coming down the pike.

Q: What gave BWXT the edge over X-energy and Westinghouse in DOD’s contract award?

A: I can just talk broadly about what our grading metrics were. We were looking for a microreactor design that we had confidence we could build in a relatively short time. We were looking as much as possible for both high TRL [technology readiness level] materials and subcomponents, and a team has had experience building things like this before. We were not looking to build the most advanced reactor possible. There are certain things that you could do that would provide more power or more performance, but would've added additional risk in terms of cost and schedule. And at this stage, we are not looking for that. My goal on Project Pele is not to build the most perfect reactor, it is to build a microreactor that meets the requirements that we initially set out to deliver.

Q: It isn't just BWXT, of course, working on the project, you also have Northrop Grumman, Aeroject Rocketdyne, Rolls-Royce LibertyWorks and Torch Technologies. Can you tell us a little bit more about what these other companies are doing?

A: I think this is an important message for anyone who wants to explore nuclear power as a potential solution: Nuclear reactors are complicated and the hardest parts are not really the nuclear part, the hardest part of a nuclear reactor is the integration of the system. Even a company like BWXT, which has built a lot of nuclear reactors for the Navy, does not necessarily have all the expertise necessary for all the different parts of a microreactor system. For instance, power conversion systems in a high-temperature gas reactor require a turbine, so we wanted a company that has real expertise in turbines, namely Rolls-Royce. I don't believe that there is any nuclear company in the United States that can build a nuclear reactor entirely by itself; I think any successful company is going to need partners.

The other companies in our competition, Westinghouse and X-energy, also recognized that they could not do this on their own and came with their own significant subcontractors. So when I see companies saying that they have these plans to build a nuclear reactor, one of the first questions to ask is: "Well, who are your partners? Who's going to help you build this? Because you will not have all that expertise in house."

I can't get into the details of what one company is doing or another. I can say broadly speaking, Rolls-Royce, their expertise is in the power conversion system. And where Northrop Grumman in particular is helping out is on the I&C [instrumentation and control] — all the computer control systems — they have a lot of experience doing that for a lot of different defense programs.

Q: The reactor system is designed for assembly onsite and operations within 72 hours with shutdown, cooldown, disconnection and removal in less than seven days. Can you walk us through the manpower that's needed to make that happen?

A: With large power systems that the DOD operates, such as their large diesel generators, you typically have a team that is assigned to that system and that moves with that system wherever it goes. So a team will be trained to transport, set up, operate and maintain the system. The size of the team envisioned for Pele has not been precisely set out and part of that will come down to how the Army, or any other service, wants to operate the system. But we envisioned it would probably be something on the order of around a dozen people to do setup, maintenance and the operation of the system. That includes the operating team, which would be two people at a time, probably on eight-hour shifts. So you need six people just to run the reactor for 24 straight hours, and then you'll also have people on the team who are experts in maintenance and fabrication. And then you also have some expert nuclear engineers to deal with more serious or complicated problems.

Q: Is there an autonomous operating system at work as well?

A: This reactor is designed to be highly autonomous; it should mostly run itself. There's some misconception that it's fully autonomous, but realistically no one in the near future is going to allow an unaccompanied operating nuclear reactor, there's always going to have to be someone there. We envision that at all times there will be two operators with the system, but there should be significantly less action required of these operators compared to more traditional light-water reactor systems. It's a much simpler, safer reactor with less excess reactivity in the core.

Q: How difficult would it be to power the reactor on and off for different locations and objectives?

A: The reason why we designed the system so that we could remove it in seven days is for really an emergency scenario where it's somewhere that is close to being overrun and you have to get the reactor out of there. It's not an efficient way to run the reactor. We would not recommend having the system moving every week. There are a lot of efficiencies with running a nuclear reactor relatively steadily, and the economic case for the reactor works a lot better if it stays in the same place for months at a time. The sort of systems that operate in the 1 MW-5 MW range are things like mobile hospitals and over-the-horizon radar systems and things like that tend to be relatively static. They can move, but they don't tend to move very often. So we don't anticipate that the system is going to be moving on a weekly basis. It's just less efficient that way.

Q: One of the potential uses DOD has considered for Project Pele is at forward-operating bases (Fobs). Is that something that DOD is still considering?

A: The term forward-operating bases has taken on a different meaning. When we think about forward-operating bases in the global war on terror era — where we were talking about these small Fobs in remote parts of Afghanistan — we would not anticipate putting a reactor there, nor would they have need for megawatt power. But a key thing to remember in the modern world is that everywhere in its own way is a forward-operating base. If we were to get into a real near-peer conflict, even our installations in the Continental US are not necessarily safe from disruption to their infrastructure. So we are focused on locations that are operationally important. But it will not be in the tactical zone. We don't anticipate using these in a tactical area. These microreactors would be a strategic asset to be used at the strategic level.

Q: I understand Project Pele is supposed to conduct kinetic impact tests to ensure that in the event of an attack, the fission products in the Triso fuel don't pose radiological problems. Can you tell me more about these tests?

A: We have access to good computer modeling software of what a kinetic attack does to things like a reactor core. And what we need to do is to validate those models with some physical testing. So we're not going to build a full Pele system and blow it up, but what we will do is do smaller-scale physical testing using the real materials and real threats in order to validate the data. And unfortunately, I can't get more into that because the actual details and the tests are all classified.

Q: How often would it need refueling?

A: A requirement for the program is that the reactor has to operate for at least three years at full power. For the first system, that’s what it will be.

Q: Pele requires high-assay low-enriched uranium (Haleu), correct? How much Haleu would be needed for each loading?

A: I can't give you a precise number there, but I can say that the order of magnitude of the quantity of uranium that we're using is in the hundreds of kilograms.

Q: Beyond remote locations, the military is looking for an efficient power resource to support energy-intensive weapon systems. Can you say more about this?

A: Until the invention of the nuclear-powered submarine, there was no such thing as a submarine. There were things called submarines in World War I and World War II, but they were really just boats that could briefly go underwater. They don't look like the circular submarines that we're used to, and that’s because it took until you had this huge source of power that didn't require air in order to be able to have a real submarine.

Now the DOD operates in an extremely energy-constrained mode, we are incredibly energy limited and we are incredibly vulnerable to energy disruptions. We are very limited in what we can do or even think about because of that limit on energy, in particularly remote and islanded locations. If there were large amounts of power, it opens the aperture to all sorts of ideas, whether that is new radar systems or new directed energy systems, or anything else, it just gives planners a new way to think.

Back when I was at NASA, the science experiments that went into orbit operated on incredibly tiny amounts of power; all the big space missions you've heard of have operated on order of magnitude, dozens of watts or a hundred watts. One time the science leadership was consulted on what would they do if they had kilowatts, hundreds of kilowatts available, and the scientists in the room had their minds blown: "It never occurred to us that we could have a hundred kilowatts to do a science experiment in space." I think that will be what military planners will have if they suddenly enter a world of energy abundance.

Q: Interesting. Where else outside of the military do you envision this type of microreactor could be deployed and used?

A: For folks that have reached out to us trying to learn more about Pele, one application is very remote communities. There are a lot of remote communities, particularly in Alaska that are not connected to a larger grid. They pay very high prices for very dirty energy right now.

Another application, believe it or not, is oil and gas and mining. This is an area that Canada has been very interested in because oil and gas mining and drilling are very energy intensive and tend to be in very remote areas where it's very expensive to get energy. A lot of them have actually made net-zero public commitments and there's no way to do that without bringing some nuclear reactors to bear.

The third application is critical, private-sector infrastructure, things like servers. If Google or Microsoft have servers that have to operate 24/7 no matter what, a microreactor would be able to provide that. However, I don't think microreactors will ever be cost-competitive with the larger public grid in the Continental United States. It's never going to be 10¢ a kilowatt hour, so it'll only be for locations where you're not easily connected to the grid or where you're willing to pay extra for resiliency.

Q: Regarding the Triso fuel, BWXT noted that the contract covers the manufacturing of Triso fuel for the core for Project Pele and for additional reactors, and coded particle fuel for NASA. Can you tell us what the “additional reactors” is in that?

A: We are working with BWXT on developing more advanced variants of Triso. So the variant of Triso that we chose for Pele is AGR [advanced gas reactor] Triso, which is the version that is already qualified, so it has the highest TRL. However, there are more advanced forms of Triso that can withstand even higher temperatures. We have been working with NASA to help develop fuel that might be of use for one of the space systems that NASA or Darpa [Defense Advanced Research Projects Agency] is looking to do.

Because this is a commercial Triso line, any other company can come and buy BWXT's Triso fuel. There are other companies that have initiated work on their own Triso manufacturing facilities, but none are likely to produce the fuel for at least another five years.

Q: I know BWXT has made Triso in the past, so what did it take to restart those production lines?

A: BWXT actually still had some of the old equipment and we brought in a lot of the old personnel. We had to buy additional equipment in order to increase the throughput, but then the key thing was to requalify equipment and to train all of the staff so that they're qualified. It's a delicate process. It's very complicated to make high-quality Triso, but as of December, that line is open for business and anyone who wants qualified Triso can buy it.

Q: Does that include the Haleu component?

A: BWXT can make Triso at any enrichment level. In fact, they’re already starting to make the fuel for the Pele core on that line.

Q: I know Pele has access to military stocks of highly enriched uranium (HEU) for downblending into Haleu. But if a commercial vendor were to come to BWXT, they would still have to procure their own Haleu, correct?

A: Most of the enriched uranium is controlled by the NNSA [National Nuclear Security Administration]. So it's really everyone's responsibility to either acquire their own enriched uranium from the NNSA or else procure it through the markets, whether through Centrus or Urenco or whoever else. BWXT doesn't have stockpiles of uranium that they can give to people.

Q: You touched on NASA. I recently wrote about their plans for a microreactor to power a lunar base and their work with the Pentagon’s Darpa to develop nuclear thermal propulsion. Is that what you're talking about when it comes to Darpa and NASA’s potential use of Triso fuel?

A: The Darpa and NASA teams are still looking at a variety of fuel options. I don't think the design is fully settled, but they're very likely going to go with some form of encapsulated fuel because, on a space rocket, the fuel efficiency of the rocket has a direct relationship to the temperature that you operate in the core. The hotter the core, the more efficient the rocket. So the higher temperatures your fuel could withstand the better your spacecraft will be. It's going to take some form of encapsulated fuel in order to get to those crazy high temperatures.

Q: What do you say to concerns about safety, security and nonproliferation risks associated with this fuel in both military and space applications?

A: I would say that Haleu Triso fuel is the most proliferation-proof fuel you could conceive of. For one thing, the fuel is broken up into millions of tiny little particles that are each wrapped in a very, very tough silicon carbide material. It's very difficult to recycle. Second of all, 20% fuel is actually fairly ideal from a proliferation perspective because, while it's a higher uranium enrichment level, you do produce less plutonium with a higher enriched core. The easiest way to produce plutonium is with a natural uranium core.

People should keep in mind that these reactors are going to be under the control of the US military in large strategic installations, surrounded by a lot of people with guns. So if I were trying to divert nuclear material for some sort of terrorist activity, I can't think of a less attractive target than a Pele reactor.

Q: Because Project Pele has access to Haleu before a lot of the other commercial vendors developing advanced reactors, as a first mover with respect to Haleu, what does this mean for a wider commercial case for Haleu production?

A: First, I would emphasize that the amount of Haleu that we're using is very small, you could redirect all the fuel used for Haleu and have a negligible impact on private-sector demand. But second of all, I think Pele is a big initiating factor for uranium because prior to the Russian invasion of Ukraine, the commercial sector in the United States was happy to buy most of their uranium from overseas. They were happy to buy a lot of it from Russia. The only justification for US enrichment at that time was national security applications that require unobligated fuel. That would be the weapons program and the Navy and now Pele.

Both the nuclear weapons program and the Navy are good on their fuel for the next couple of decades. So the existence of a project like Pele gives a reason to invest in unobligated, American-made enriched fuel.

Now a lot of momentum has come behind this since the Russian invasion of Ukraine, because suddenly everyone's decided it's a problem to buy all of your uranium from Russia and there has been a lot of action by Congress and the White House, as I know you're aware, so I do think that Pele has helped to drive this somewhat, but obviously not as much as Russia invading Ukraine.

Q: I know that covers our time. Is there anything else that you would like to add, Jeff, that you feel we didn't touch on?

A: We have a very strong team and we've been working very quickly to achieve what we have. We are at the point now where we are ordering hardware for the Pele reactor and as I mentioned, we are making the fuel for the Pele core. It's often difficult in the public sphere to be able to tell the difference between what projects are real and which are PowerPoint. We want to emphasize that this is real, we have the hardware, we are going to begin constructing it this year, and we do hope to be producing electrons soon.