We’re building nuclear spaceships again—this time for real

Phoebus 2A, the most powerful space nuclear reactor ever made, was fired up at Nevada Test Site on June 26, 1968. The test lasted 750 seconds and confirmed it could carry first humans to Mars. But Phoebus 2A did not take anyone to Mars. It was too large, it cost too much, and it didn’t mesh with Nixon’s idea that we had no business going anywhere further than low-Earth orbit.

But it wasn’t NASA that first called for rockets with nuclear engines. It was the military that wanted to use them for intercontinental ballistic missiles. And now, the military wants them again.

Nuclear-powered ICBMs

The work on nuclear thermal rockets (NTRs) started with the Rover program initiated by the US Air Force in the mid-1950s. The concept was simple on paper. Take tanks of liquid hydrogen and use turbopumps to feed this hydrogen through a nuclear reactor core to heat it up to very high temperatures and expel it through the nozzle to generate thrust. Instead of causing the gas to heat and expand by burning it in a combustion chamber, the gas was heated by coming into contact with a nuclear reactor.

The key advantage was fuel efficiency. “Specific impulse,” a measurement that’s something like the gas mileage of a rocket, could be calculated from the square root of the exhaust gas temperature divided by the molecular weight of the propellant. This meant the most efficient propellant for rockets was hydrogen because it had the lowest molecular weight.

In chemical rockets, hydrogen had to be mixed with an oxidizer, which increased the total molecular weight of the propellant but was necessary for combustion to happen. Nuclear rockets didn’t need combustion and could work with pure hydrogen, which made them at least twice as efficient. The Air Force wanted to efficiently deliver nuclear warheads to targets around the world.

The problem was that running stationary reactors on Earth was one thing; making them fly was quite another.

Space reactor challenge

Fuel rods made with uranium 235 oxide distributed in a metal or ceramic matrix comprise the core of a standard fission reactor. Fission happens when a slow-moving neutron is absorbed by a uranium 235 nucleus and splits it into two lighter nuclei, releasing huge amounts of energy and excess, very fast neutrons. These excess neutrons normally don’t trigger further fissions, as they move too fast to get absorbed by other uranium nuclei.

Starting a chain reaction that keeps the reactor going depends on slowing them down with a moderator, like water, that “moderates” their speed. This reaction is kept at moderate levels using control rods made of neutron-absorbing materials, usually boron or cadmium, that limit the number of neutrons that can trigger fission. Reactors are dialed up or down by moving the control rods in and out of the core.

Translating any of this to a flying reactor is a challenge. The first problem is the fuel. The hotter you make the exhaust gas, the more you increase specific impulse, so NTRs needed the core to operate at temperatures reaching 3,000 K—nearly 1,800 K higher than ground-based reactors. Manufacturing fuel rods that could survive such temperatures proved extremely difficult.

Then there was the hydrogen itself, which is extremely corrosive at these temperatures, especially when interacting with those few materials that are stable at 3,000 K. Finally, standard control rods had to go, too, because on the ground, they were gravitationally dropped into the core, and that wouldn’t work in flight.

Los Alamos Scientific Laboratory proposed a few promising NTR designs that addressed all these issues in 1955 and 1956, but the program really picked up pace after it was transferred to NASA and Atomic Energy Commission (AEC) in 1958, There, the idea was rebranded as NERVA, Nuclear Engine for Rocket Vehicle Applications. NASA and AEC, blessed with nearly unlimited budget, got busy building space reactors—lots of them.

Kiwi tries to fly

The first of those reactors was called Kiwi-A. The test done on July 1, 1959, proved that the concept worked, but there were devils in the details. Vibrations caused by the flow of hydrogen damaged the reactor after just five minutes of operation at a relatively meek 70 megawatts. The temperature reached 2,683 K, which caused hydrogen corrosion in the rods and expelled parts of the core through the nozzle, a problem known as “shedding.”

On the upside, rotating drums placed around the core that replaced standard control rods worked well. These were long tubes made with neutron-absorbing material that had one side covered with a coating that reflected the neutrons back into the core. The reactor was throttled up by rotating the drums so they faced the core with the reflective side and throttled down by turning the neutron-absorbing side toward the core.

Over 18 years, NASA, AEC, and industry contractors like Aerojet Corporation built and tested a total of 23 reactors. “The last engine in the Rover/NERVA program was the XE Prime. They tested it in a vacuum environment and brought it to TRL 6,” said Dr. Tabitha Dodson, a program manager at DARPA’s Tactical Technology Office. TRL 6 means “tech readiness level 6”—getting to 7 would mean putting a demonstration engine in space.

This didn’t mean “problem-free,” though. Shedding and fuel cracking issues persisted in all NERVA engines to various degrees. But what ultimately killed NERVA in 1973 was a shift in NASA’s goals away from deep space and toward low-Earth orbit. And NERVA wasn’t needed for that.

Nuclear Mars Express

It took over 40 years before NASA brought up nuclear propulsion again, first in the short-lived Jupiter Icy Moon Orbiter project and then in the design reference architecture for human exploration of Mars. Powering the latter missions with a compact reactor could cut down Mars transit by more than half, to three to four months versus the six to nine months predicted for chemical rocket engines. Less time in space meant less exposure to radiation for the astronauts and fewer supplies for the trip.

So, in 2017, NASA started a small-scale NTR research program. The budget was just a hair above $18 million, but it was something. Two years later, Congress passed an appropriation bill that granted $125 million for developing NTRs. Things were progressing, but they were mostly paper studies, followed by more paper studies, followed by even more paper studies.

And then on June 17, 2020, DARPA entered the chat and said, “We want a nuclear rocket.” Not just another paper study—a demonstrator.

Chasing Sputnik 2.0

DARPA’s website says it has always held to a singular mission of making investments in breakthrough technologies for national security. What does a nuclear-powered spaceship have to do with national security? The military’s perspective was hinted at by General James Dickinson, a US Space Command officer, in his testimony before Congress in April 2021.

He said that “Beijing is seeking space superiority through space attack systems” and mentioned intelligence gathered on the Shijian-17, a Chinese satellite fitted with a robotic arm that could be used for “grappling other satellites.” That may sound like a ridiculous stretch, but it was enough get a go-ahead for a nuclear spaceship.

And the apparent concern regarding hypothetical threats has continued. The purpose of the Demonstration Rocket for Agile Cislunar Operations (DRACO) project, stated in its environmental assessment, was to “provide space-based assets to deter strategic attacks by adversaries.” Dickinson’s worries about China were quoted in there as well.

“Let’s say you have a time-critical mission where you need to quickly go from A to B in cislunar space or you need to keep an eye on another country that is doing something near or around the Moon, and you need to move in very fast. With a platform like DRACO, you can do that,” said DARPA’s Dodson.

Two years after DARPA stepped in, the preliminary design phase was completed, and Lockheed won a half-billion-dollar contract to build DRACO. But DARPA wasn’t the only one paying. NASA chipped in as well. The two agencies made DRACO a joint project and split the bill 50-50.

Next-gen NERVA

Building DRACO, however, would run us up against a further problem: using it. “There is a series of regulatory and technical challenges,” said Kirk Shireman, the vice president of Lockheed Martin Space who oversees the DRACO project. For starters, firing nuclear engines in the open air somewhere in the Nevada desert was out of the question. Building facilities compliant with all the regulations alone would take years.

Then there was the fuel. NERVA reactors worked with highly enriched uranium used to build nuclear weapons. If anything went wrong at launch, roughly 700 kilograms of weapons-grade uranium would suddenly fall from the sky. And you only need around 25 kilograms of it to make a bomb.

That's why DRACO will use a new fuel called high-assay-low-enriched uranium (HALEU)—a fissile material made by blending the highly enriched uranium down to below 20 percent enrichment. “You can relax some security requirements by switching to HALEU,” said Joe Miller, the vice president of BWXT Technologies, a company specializing in naval reactors that Lockheed Martin chose to build the reactor for DRACO. And while making a bomb with HALEU is still possible under certain circumstances, it’s way harder than with highly enriched uranium, which was a must-have in all NERVA reactors.

Once the fuel was sorted out, BWXT went on to design the reactor itself. “Using HALEU drives the internal geometry of the reactor,” says Miller. To avoid reinventing the wheel, Miller’s team started with rummaging through huge piles of reports from the NERVA program. But compared to NERVA designs, his team used different channels to route the hydrogen through the reactor core and thermal management systems that transfer the heat to hydrogen.

Brown bag sessions

“Our chief engineer was a bit of a historian and a librarian, so he was digging all those reports out, scanning them, and integrating them into our design reviews. Lots of black and white photos. Lots of old graphs from testing. We learned from that. This was extremely relevant,” said Miller.

One of the key things BWXT found in the NERVA reports was the data on hydrogen-induced cracking of the reactor fuel. “We gave [the reports] to our young materials scientists, and they were able to use them as a springboard for the early design decisions they were making,” Miller said. The result, he said, was a coating that could withstand reactor temperatures without cracking. “We created our own internal formulation of the nuclear fuel I can’t really talk about in public,” he said.

Building a space reactor is challenging, but at least it has already been done before. What hasn’t been done is building a spaceship around it.

The first nuclear spaceship

DRACO will be a medium-sized spacecraft, under 15 meters long with a diameter below 5.4 meters—dimensions dictated by the size of the standard payload fairing of the Vulcan Centaur rocket on which it will probably be launched. “We are familiar with liquid hydrogen, spacecraft systems engineering, and integration. We have the right skillset and the right people to build this thing,” said Shireman.

DRACO will work like NERVA-type rockets, with hydrogen tanks located at the head of the propulsion compartment, turbomachinery feeding this hydrogen through the core (fitted right behind them), but separated from the core by a radiation shield. The HALEU reactor will be surrounded by control drums and sit in front of an exhaust nozzle. Based on DARPA requirements, DRACO will have at least 700 seconds of specific impulse, which is over 300 seconds better than the RL-10, the best-performing chemical space engine we have.

“The main technical challenge here is working with liquid hydrogen stored at 20 K—very, very cold and really slippery molecules that like to slip out of wherever you put them,” Shireman said. For DRACO, Lockheed went for passive hydrogen cooling. The tanks will be thermally isolated to keep the Sun from heating them up. This way, the hydrogen should stay at 20 K for long enough to complete all tests. For longer missions, nuclear spaceships would need to rely on active cooling.

Test-driving DRACO

Because there is a nuclear reactor onboard, Lockheed and BWXT made sure that the risks of every potential catastrophic failure were brought to an absolute minimum and there was a contingency plan for every scenario.

What if the launch platform fails and DRACO crashes somewhere near its launchpad in Florida? That won’t be any more of a problem than a crash of a conventional engine, as the reactor will only be activated by its control drums after reaching a safe orbit at least 700 kilometers from Earth.

A crash into the ocean? This is a bit trickier because water is a moderator and would start the chain fission reaction, basically turning the reactor on regardless of what the control drums do. But DRACO is designed to prevent that, too. In such a case, neutron poison, a material that absorbs neutrons and immediately stops the reaction, would be deployed straight into the core.

The actual test drive will begin when DRACO reaches its target orbit. “First, we are going to do a series of checkouts, make sure all the sensors and actuators are working. Then, slowly, we are going to start powering the reactor up,” said Dodson. This will be a moment of truth for DRACO because the program does not include any ground tests with a powered reactor.

“Because the DRACO fuel uses uranium with lower enrichment than NERVA, we need to use more moderator. Also, we expect a phenomenon we call a negative temperature feedback, where a reactor powers down as it heats up. It’s one of the interesting unknowns in this project, and we are hoping to gather more data on how it works,” Dodson claims.

“It’s like a new performance car. You don’t take it out and run it at full throttle out of the gate. We are going to gradually move up the performance and finally, if we have opportunity to show something meaningful, perhaps we would go full power,” said Dr. Anthony Calomino, NASA’s Space Nuclear Technology portfolio manager. This “something meaningful” is specific impulse high enough to take humans to Mars. But that’s not all.

Lazy rivers

The problem with reaching destinations like the Moon or Mars has been that we can’t go there in a straight line. You don’t just point your conventional rocket at the Moon and fire away, Julius-Verne-style, expecting it to get there. “Such rockets can’t move entirely on their own. They use complex fractal orbits that go around Lagrange points, kind of riding gravitational eddy currents in cis-lunar space—‘lazy rivers,’ as I like to call them,” said Dodson.

Think of it like getting on a tiny boat in Liverpool with just enough fuel to reach the closest ocean current because you calculated this current will eventually wash you up in New York. That’s how we get around in space today. DRACO is intended to be the first step to powering through in nuclear space cruisers.

“There are civil applications as well,” said Calomino. “It’s about staging payloads that left Earth into lower orbits where a space tug can pick them up and ferry them to the Moon, back and forth.” Such nuclear space tugs, he suggested, would become the backbone of a new cis-lunar transportation system.

And perhaps the best thing about these space tugs is that the reactors can last for years. “We know there is water on a surface of the Moon. You can process this water to get hydrogen and use it to tank up your ship the way you fill up a car. The reactor itself is going to operate over a very long time,” Calomino said.

Topping up aside, there is another thing cars and nuclear spaceships have in common: We can supercharge them.

Supercharged nuclear spaceship

“My background is in hypersonic fluid dynamics, mostly in vehicles reentering back into the atmosphere. I attended the talks NASA gave about issues with going to Mars that even NTRs couldn’t solve,” said Ryan Gosse, a Herbert Wertheim College of Engineering professor of practice at the University of Florida. Gosse and his team figured they could solve some of these problems by tuning the NTR up with superchargers.

Gosse’s idea was based on using a wave rotor. “In cars, it’s called a compressor or a supercharger,” explained Gosse. In his NTR concept, a wave rotor is fitted between the reactor core outlet and the exhaust nozzle to further increase the exhaust gas temperature.

“The limiting factor for the NTR is the temperature of the reactor core. Today, this is roughly 3,000 K, which gives you around 900 seconds of specific impulse,” Gosse said. A wave rotor, according to his calculations, should bump this up to 1,400 seconds—twice as much as DRACO. Gosse and his team proposed this concept to NIAC, a NASA program funding innovative early-stage ideas, and in 2023 got the funding to make a detailed feasibility assessment.

But the wave rotor is not the only unique thing about Gosse’s spaceship. The real magic starts when the NTR engine has finished its burn. The ship will turn around, flying nozzle first. It will then switch the reactor into power plant mode by rerouting its heated hydrogen away from the nozzle and into a closed loop with power-generating turbines and use the electricity to power a specific form of ion thruster that is attached at the opposite end of the spacecraft. They will increase the specific impulse from 1,400 to over 10,000 seconds.

Getting big and staying cool

A bi-modal propulsion system like this was first hinted at toward the end of the NERVA program. There were two problems, though.

First, electric thrusters have always been used to drive small, unmanned spacecraft. Scaling them up to accommodate the thousands of megawatts generated by nuclear reactors would require huge spaceships. “The current electric thrusters can go to like 100 kilowatts. If you try to use them in our spacecraft, you’re going to need so many it wouldn’t be practical. It’s not a trivial problem like, ‘well, just get a thousand 100 kilowatt thrusters and that’s it,’” said Gosse. “So we are looking at magnetoplasmadynamic (MPD) thrusters, which have much higher energy density and have been demonstrated to work up to a megawatt.”

The second problem is cooling. The NTR doesn’t have waste heat problems because the hydrogen works as coolant for the reactor and is then expelled from the ship. In the nuclear electric propulsion (NEP) mode, the coolant flows in a closed loop, which means heat accumulates in the spacecraft. This is why all NEP designs have huge radiators. In NASA’s reference chemical-NEP architecture, the radiator alone had to be over 2,000 square meters. Gosse’s bi-modal wave rotor ship would need a radiator that is five times bigger.

It would be seriously fast, though. “A reference NTP spacecraft should make it to Mars in 297 days and weigh more than 600 tons. Chemical/NEP design would need 382 days weighing 418 tons,” said Gosse.  His bi-modal wave rotor concept is fast enough to launch when Mars and Earth are closer to each other and reach Mars in just 45 days with a mass of 530 tons. “Flying a bit slower, doing a 65-day trip, we can go as low as 273 tons,” Gosse said.

Baby steps

But this idea won’t be tested on DRACO. “The crawl-walk-run strategy is what we really want to implement here,” Calomino said. “The primary thing is to get the NTP engine up and running, get some confidence, understand the reactor, get some resilience on this reactor, so let’s focus on that. Let’s get that done.”

Once we know it works, there would be time to evaluate whether it makes sense to add the complexity of MPD thrusters. When you do both electric and thermal nuclear propulsion, you have two systems with different requirements, even if they feed off the same reactor. Then you need to add the mass of both and stack that up against using just one system and giving it more fuel. Adding complexity also adds risk.

For some within the DOD, a lot is riding on a successful demonstration of the simple system. “Think of the Navy. The best way to get around with heavy payload through the oceans is using huge battleships with large engines. Nuclear propulsion being the best option. The same is true for space. Right now, the Department of Defense does not have such capabilities,” Dodson said. “But once we have them, our ships could move through space like they do through the oceans.”

Leaving behind the issues with that statement (the Navy never had nuclear-powered battleships, and moving through space is very unlike moving through the ocean), the question is whether we need nuclear space battleships in the first place.

The key reason we don’t fly NTRs today is they have never been an enabling technology for anything we tried to do. Each time their supporters said something couldn’t be done without nuclear rockets, they were proven wrong. Nuclear warheads? Done with chemical rockets. Moon landing? Done with chemical rockets. Hunting Chinese grappling satellites? In 2021, Russia destroyed a satellite using a chemically powered missile launched from the ground.

Giant space tugs cruising between the Earth, the Moon, and Mars? Our need for them remains an open question. The question of whether we will one day need nuclear space battleships to keep them safe is even more remote. But some of the people involved are definitely thinking long-term.

“DRACO really has great potential for the future, for the world. This could really open up something. It’s setting on a path that maybe your grandchildren are going to finish. We hope to make a little history,” Shireman said.