Spotlight on Space Nuclear

Examining past hurdles

Space nuclear power and propulsion are technically challenging, but historically, technology alone has rarely been the limiting factor, according to Bhavya Lal, a professor of policy analysis at the RAND School of Public Policy and one of the authors of the 2025 report. The Soviet Union, after all, from 1970 to 1988 flew more than 30 satellites powered by fission reactors.

Instead, she has said in papers and public remarks, the hurdles facing past U.S. initiatives have been in the execution, beginning with the lack of a clear rationale for why specific missions required nuclear fission.

The SNAP-10A satellite in 1965, part of the Atomic Energy Commission’s Systems for Nuclear Auxiliary Power program, carried a 500-watt reactor that operated for 43 days before an unrelated electrical fault killed the craft. But with the growing availability of solar power for satellites, there was no clear need for fission power, and the fission portion of SNAP fizzled.

“Nuclear power in space is just like any other application in space, where if it meets the mission better than any other technology, then that’s why you use it,” says Sebastian Corbisiero, national technical director for the Department of Energy’s Space Reactor Program. His team is developing the reactor for SR-1.

Past program timelines were also mismatched with political cycles, inviting cancellations. Consider Project Rover/NERVA, a joint effort of NASA and the Atomic Energy Commission that tested multiple nuclear thermal reactors and rocket engines between 1959 and 1973. Nuclear-thermal propulsion uses a fission reactor to heat liquid hydrogen propellant and expel it from a nozzle, producing high thrust more efficiently than a chemical rocket.

The idea for Rover/NERVA was to develop a nuclear-powered upper stage for chemical rockets launching to the moon and more distant destinations. By swapping out the Saturn V’s third stage for nuclear propulsion, NASA estimated it could land up to 75% more payload on the moon than was possible with chemical rockets alone, according to a 1965 report.

But then-President Richard Nixon, who was already winding down the Apollo program and looking to cut spending, canceled Rover/NERVA in 1973.

Other programs depended on technological leaps that were too large, and often grew too costly, according to Sinacore — in part because of the lack of a coordinating authority to keep the budget under control.

For example, NASA’s Jupiter Icy Moons Orbiter (JIMO) program, proposed in the early 2000s, would have used a 200-kilowatt reactor — 400 times the power of SNAP-10A — in a nuclear-electric craft destined to explore the moons of Jupiter. At a talk at MIT in July, Lal argued that the lack of a clear lead among the three teams working on JIMO — NASA, DOE and the commercial team designing the spacecraft — contributed to the rapid increase of the mission’s projected cost. Indeed, when the program was canceled in 2005, NASA estimated a price tag of more than $20 billion.

Inside NASA’s strategy

NASA officials believe their plans for SR-1 and LR-1 have accounted for these previous failure modes, according to Sinacore.

Let’s start with the mission cases. Sinacore likens spacecraft with nuclear-electric propulsion, or NEP, to astronomical railroads, moving large cargo across deep space far more efficiently than chemical rockets. For a Mars mission like SR-1, this means less of the spacecraft’s mass is reserved for carrying propellant, opening up the potential for larger and more complex science payloads. On the lunar surface, he says, fission power is the only viable solution to keeping the Moon Base operating during the 354 hours of lunar night that occur every two weeks, when the moon’s surface gets no sun exposure. In NASA’s view, developing a handful of fission reactors is more practical than erecting fields of solar panels to shore up sunlight.

As for the political component, both SR-1 and LR-1 have clear deadlines and support from the Trump administration, stemming from the December executive order that also outlined a space nuclear strategy.

The White House followed up in April with another policy directive, National Space Technology Memorandum-3 (NSTM-3), that gave more detailed guidelines for federal agency coordination in reaching space nuclear power and propulsion goals.

These directives were the latest in a line of regulatory changes and executive orders directing federal agencies to pursue space nuclear applications, going back to the first Trump administration and continuing under former President Joe Biden, whose administration significantly revised the rulemaking pathways for new reactors.

Altogether, these changes “have helped streamline a lot of the regulatory paths that programs like DRACO have taken advantage of,” says Joel Sema, space market segment director for BWXT Advanced Technologies, which was contracted to develop the reactor for DRACO, the joint DARPA-NASA program.

These regulatory and policy changes have coincided with geopolitical competition, according to Jake Hecla, assistant professor in MIT’s departments of Aeronautics and Astronautics and Nuclear Science and Engineering. He points to China’s plan to place a nuclear reactor at the lunar south pole by the mid-2030s, and its cooperation with Russia to develop a 500-kW NEP space tug that could be used for space-based electronic warfare.

“We see a class of capabilities that are being developed, which cannot be met with solar power, and they’re being developed by our geopolitical adversaries,” Hecla says. “That provides significant mission pull.”

NSTM-3 also addressed the agency leadership issue, assigning the White House Office of Science Technology Policy as the overall coordinating agency. NASA followed suit in May, consolidating its space nuclear efforts under a newly created Division of Space Reactors led by Cynthia Simmons, formerly the deputy director of Goddard Space Flight Center. OSTP did not respond to multiple requests to comment for this story.

“I have a very clean line of accountability on Space Reactor-1 Freedom,” Sinacore says, “through Cynthia, to the new mission directorate, through to [NASA Administrator Jared Isaacman] himself.” That structure “puts a laser focus on accountability.”

Designing SR-1

Technologically, NASA believes SR-1 is in a sweet spot: ambitious enough to fulfill a unique mission, but consisting of existing technologies that can be built quickly. The DOE, for instance, has developed expertise in rapidly developing small reactors on the ground over just the past few years, according to Michael Goff, principal deputy assistant secretary for the department’s Office of Nuclear Energy.

“Some of these goals that we thought might have been obtainable a few years ago, now I think are very obtainable,” he says. “This is not just some technology that’s for next decade. It’s ready now, and it’s ready now for space applications.”

Fission reactors that will operate in space pose unique design challenges. Unlike their terrestrial counterparts, they must be optimized for low mass, different fuel types and higher temperatures, Hecla notes. They must also withstand the vibrations of a rocket launch and use specialized means of cooling and converting heat into electric power.

“It’s the difference between trying to make an F-150 and trying to build a Ferrari,” Hecla says.

One such challenge is dealing with excess heat. Terrestrial reactors can use evaporation with large cooling towers to clear excess heat leftover after generating electricity. In space, reactors have to rely on thermal radiation, shedding waste heat primarily as infrared light through large radiators.

“The nuclear core is frankly probably about a quarter or less of your total mass,” says Randy Bell, a senior project leader and space nuclear power and propulsion analyst with the Aerospace Corp., which operates a federally funded research and development center in Virginia. “Your radiator dominates the volume and mass,” especially as a reactor gets larger and produces more power.

But because of SR-1’s modest 20-kW reactor, “the heat rejection system actually is not one of the things that keeps me up at night,” Sinacore says, noting that NASA expects to be able to stick to well-understood and mature technologies for this particular mission.

What that system will look like for SR-1 is a reactor core that heats a gas — possibly helium or xenon — that will be compressed and used to turn a turbine. A heat exchanger will transfer heat from the gas to water within a heat pipe, a sealed tube that allows heated water vapor to travel its length to radiator panels at the far end, for thermal radiation into space. The cooled vapor then condenses and returns down the length of the pipe through a wicking material, possibly metal mesh, using only capillary action — no moving parts.

At a high level, the largest challenge isn’t a particular set of components, but in integrating the reactor with a spacecraft — and doing so on a short timetable. Here, Sinacore says NASA hopes to save time by repurposing the Power and Propulsion Element (PPE) from the now-canceled lunar Gateway space station.

“What makes the aggressive 2028 launch date achievable is the fact that PPE was nearly complete from a spacecraft bus perspective,” he says. “What that spacecraft bus gives me is the electric thrusters integrated with all the fuel systems and control systems.”

NASA Administrator Jared Isaacman told members of the House Science Committee in April the first two Gateway habitation modules had corrosion, but an agency spokesperson confirmed the PPE module does not have that issue.

NASA is working with contractor Intuitive Machines to identify the necessary PPE modifications for its new mission, Sinacore says. The spacecraft will need a more powerful X-band radio than the Ka-band planned for the Gateway, for instance. It also needs only three 12-kW Hall effect thrusters, so the agency can drop the additional four 6-kW thrusters required for lunar orbit.

“While this mission has an aggressive timeline, we’re confident in our ability to meet the launch date since we’ve been building and flying high-power systems for commercial customers for decades,” Chris Coker, vice president of civil space programs at Intuitive Machines, said in an emailed statement. “Our PPE hardware has undergone rigorous review and testing to meet NASA’s standards.”

Hall effect thrusters ionize atoms of a gas such as neon or xenon, then use an electric field to accelerate them out of the rear of the thruster to generate thrust. Although this thrust is low compared to that of chemical rockets, thrusters are so propellant efficient they can operate continuously for long periods, allowing deep space missions to pick up large changes in velocity over time with less propellant than chemical rockets.

The conceptual rendering of SR-1 released in March shows what the spacecraft could look like: A long, tapering bus with a rectangular cross section and the 20-kW reactor behind a boron carbide radiation shield at its head, the thinnest end of the spacecraft.

The heat pipe would run from the reactor to an array of titanium and composite radiator panels two-thirds of the length toward the aft of the craft. The rear is taken up by the repurposed PPE, which houses the electronics for controlling the reactor as well as the thrusters and propellant.

“You should start seeing hardware sub-scale tests, and then more assembly-level tests starting in early ’27,” Sinacore says.

The basic reactor and radiator design for LR-1 likely won’t differ too much from that of SR-1, he says, barring any major lessons learned in the deployment of the latter. But he expects there may be some different design tradeoffs when it comes to balancing the mass of “a significantly larger and heavier shield” to protect astronauts on the lunar surface from the reactor’s radiation, and keeping LR-1’s overall mass low enough that it can fit on a lander.

One of the things NASA hopes to learn from SR-1 is to “truly understand what the shielding needs to look like to ensure we are able to have those human-rated missions in the future,” Sinacore adds.

And while NASA is doing much of SR-1’s development internally, LR-1 is to be an industry-led effort, according to Sinacore. The agency in early August posted a notice seeking industry interest in providing space power and reactor capabilities.

The path forward

SR-1 is intended to usher in a series of nuclear-powered missions, Isaacman said in September remarks at the Air, Space & Cyber Conference.

“Now, SR-1 is just the beginning of what will become the nuclear NASA,” he said. “There will be an SR-2, an SR-3, and four as we progressively build the capability to reach deeper into the solar system and commission a new fission-powered American star fleet.”

MIT’s Hecla says success for SR-1 and LR-1 won’t be enough to guarantee a sustainable future for follow-on missions. Rather, long-term success requires further technological innovations. SR-1’s 20-kW reactor is sufficient for a pathfinder mission leveraging existing technologies, he says, but larger deep space missions — something like the canceled JIMO craft, for instance — will require reactors capable of generating at least 100 kW. Those, in turn, would require developing some truly novel technologies, particularly when it comes to the radiators.

“My belief is that things will move toward a droplet radiator concept,” Hecla says. In this design, instead of a sealed heat pipe, an emitter would eject tiny droplets of heated metal to be exposed to the vacuum of space, radiate heat and be reabsorbed by a receiver and cycled back through the radiator. “Millimeter-scale droplets have an enormous surface-area-to-volume ratio, and so they form a very effective radiating surface.”

And there are potential applications beyond deep space missions. Last year’s executive order also tasked the U.S. government with developing nuclear power for spacecraft orbiting Earth, Hecla notes, possibly for use in the Trump administration’s planned Golden Dome missile defense shield. But current reactor and radiator technology wouldn’t be practical, in his view, because they’d take up more size and mass and cost more than traditional solar arrays.

“Unless we build lighter radiators, we cannot compete with the amount of solar power available in low-Earth orbit,” Hecla says.

Another limiting factor is the availability of sufficiently large launch vehicles and — for LR-1 and other large surface systems — landers.

“It’s not that we don’t have the technology to do any of this. It’s that we’re not sure we have the ability to put it all together in a package that is small enough,” says Bell of the Aerospace Corp. “We could build a 100-kW system today. It would just be very large and difficult to land” on the moon.

The future might be more modular, he says, where the task is to get five 20-kW reactors working together rather than building one 100-kW system.

But for now, all eyes are on SR-1 and its 2028 launch target.

For the “first time in 60 years, the country has a space nuclear program with a real mission, a real deadline,” Lal said during her July presentation at MIT. “This is a moment.”