NASA’s Decades-Long Quest to Build a Deep Space Nuclear Reactor

NASA has struggled for half-a-century to take a nuclear reactor from the drawing board into space. This year it successfully tested the first new nuclear reactor design in the US in over 40 years.
May 7, 2018, 2:59pm
Image: NASA

On Tuesday, NASA announced that it completed its first full-power test of the Kilopower reactor, an experimental portable nuclear power plant that the agency hopes will one day power permanent settlements on Mars and the Moon. It’s a major milestone in the quest to establish a permanent human presence on other celestial bodies and launch energy intensive robotic missions to the outer solar system.

About the size of a refrigerator, each Kilopower reactor will be capable of delivering up to 10 kilowatts of energy when they are ready to be deployed in space. To put this in perspective, this means each Kilopower reactor produces enough energy to power about 10 average American homes on Earth. NASA estimates that an array of four Kilopower reactors would be enough to sustain a decent-sized outpost on the moon or Mars for about a decade.


Although the Kilopower reactor has been in development since 2012, testing on a 1- kilowatt prototype only began in November. The testing program, called the Kilopower Reactor Using Stirling Technology (KRUSTY), was undertaken at the Nevada National Security Site and meant to study the integrity of the reactor design in conditions similar to those in deep space. After placing the reactor in a vacuum chamber, researchers subjected the device to all kinds of critical failure scenarios, like power reduction, engine failure, and heat pipe failure. The experiments culminated in a full-power test that kept the reactor online for 28 hours.

Everything worked flawlessly, even when the reactor was dealing with several failures at once.

“We put the system through its paces,” Marc Gibson, the lead Kilopower engineer at NASA Glenn, said in a statement. “We understand the reactor very well, and this test proved that the system works the way we designed it to work. No matter what environment we expose it to, the reactor performs very well.”

This is great news for NASA, which has struggled since its inception to integrate nuclear reactors and space exploration. After billions of dollars spent on nuclear reactor designs that never made it to orbit, much less the moon, the KRUSTY program was the first successful test of a nuclear reactor intended for space applications over half-a-century. In fact, it is the first new operable fission reactor concept designed in the US in over 40 years. It heralds the beginning of a nuclear-powered era of space exploration, which will open the door to previously impossible missions into deep space, whether this is powering settlements on Mars or using robotic submarines to explore the icy oceans of Jovian moons.


On the moon and Mars, the most likely candidates for future human settlements, sources of energy are hard to come by. Unless Mars was once home to organic life, we’re not going to find any oil or coal there since these fossil fuels are mostly formed from dead plant matter. (Although Mars may harbor pockets of methane gas.) If Mars ever had an ocean or rivers, they have long since dried up, so no hydropower. The moon has no atmosphere and thus no way to harvest wind energy. Parts of the face of the moon can be shaded from the sun for stretches lasting up to 14 days, whereas Mars’ distance from the sun makes it hard to harvest enough solar power on the surface to power a rover, much less an entire settlement.

“When we start sending astronauts for long stays on the Moon and to other planets, that’s going to require a new class of power that we’ve never needed before,” Gibson said.

An artist's depiction of a Kilopower reactor on the lunar surface. Image: NASA

Shipping oil, coal, or the components for vast solar arrays to the moon or Mars is far too resource intensive to be practical—that stuff is heavy, and every pound sent to space costs about $10,000 to get there. NASA needed a relatively lightweight, yet highly efficient energy source if it ever hoped to establish a long-term human presence on the moon or Mars.

Enter uranium-235, the radioactive material that powers nuclear power plants around the world and is one of the most energy dense materials known on Earth. In fact, just one pound of enriched uranium fuel can produce as much energy as about 3 million pounds of coal.

Read More: What is Uranium?

NASA only needs a few dozen kilowatts worth of power for its initial lunar or Martian outposts. Since most nuclear power plants on Earth are designed to produce hundreds of kilowatts electricity, this meant that Gibson and his colleagues would have to come up with a new fission reactor design. Not only would it have to be compact enough to fit in a rocket fairing, it would also have to be powerful enough to host a settlement and designed in such a way that it could withstand the harsh space environment. If Gibson and his colleagues were successful, it would be an unprecedented feat of nuclear engineering.


In the last half century, over 40 nuclear reactors have been sent to space and almost all of them were Soviet. The only nuclear reactor that the US has ever sent to space was the SNAP-10A, a reactor developed by the Atomic Energy Agency and launched in 1965 to test its feasibility as a power supply for the CORONA satellite spy program. The reactor was able to produce 500 watts of power and was in low earth orbit for 43 days before the Air Force decommissioned the satellite. It is still in orbit as a piece of space junk and is predicted to remain there for another 4,000 years. Nevertheless, nuclear power continued to play a major role in US space exploration, even if reactors had been tossed to the wayside.

Since the 1960s, almost every major NASA mission into deep space has relied on electricity produced by radioisotope thermoelectric generators (RTGs). These types of electric generators have no moving parts and don’t rely on nuclear fission. Instead, an array of electric conductors harvest the heat released by the natural decay of radioactive material (usually plutonium) and convert this heat into electrical energy.

RTGs powered the Voyager 1, the first spacecraft to enter interstellar space; they powered the Cassini mission to Saturn for 20 years; they powered the New Horizons mission to Pluto and beyond; and continue to power a number of satellites and rovers on and around Mars. They are an ideal power supply for spacecraft insofar as they can produce a steady supply of electrical energy for decades with a relatively small amount of fuel. RTGs are, however, limited to an output of a few hundred watts. While this is enough to meet the demands of most spacecraft, which need to take measurements and communicate with Earth, more demanding robotic missions, like sending a submarine into the oceans of Europa, will demand a lot more energy.


Nuclear fission is the process of splitting the nucleus of an atom, usually with a neutron, which releases a tremendous amount of energy. At the same time, the split knocks other neutrons loose from the nucleus, which themselves split other atoms. This results in a cascading process of nuclear fission and a constant supply of heat energy. This process has been used to make the most destructive weapons ever created, as well as providing a relatively clean source of energy for civilian applications. (The downside of nuclear energy is that we still don’t know how to properly dispose of the waste, which remains radioactive for tens of thousands of years.) In a nuclear reactor, the heat released by splitting the nucleus of an atom—usually uranium-235—is used to heat water and generate steam, which turns a turbine to generate electricity.

While this works well enough on Earth, conventional power plants are too big to build on the moon or Mars, and definitely wouldn’t fit inside a spacecraft. Smaller fission reactors are found on naval vessels like aircraft carriers and submarines, but these types of reactors also use steam to generate electricity. This precludes them from being adapted to space applications because they require a lot of water to work and water is a precious commodity in space.

NASA considered various reactor designs for space applications after launching its only reactor in 1965, but most of these designs were considered too expensive to make and would require too much of a runway to ever be made in time for particular missions. The SP-100 reactor prototype developed in the 1980s was terminated after a decade and more than a billion dollars was spent on research. NASA’s Project Prometheus began in the early 2000s to investigate the use of nuclear reactors in space for propulsion and powering flight systems, but was terminated after nearly $400 million was spent on research and development.

The Kilopower reactor is put in a vacuum chamber for the KRUSTY project. Image: NASA


After decades of failing to take a reactor design from paper into orbit, NASA joined forces with the Department of Energy and Lockheed Martin in 2008 to pursue a nuclear power source called an advanced Stirling radioisotope generator (ASRG) that would be four times as efficient as RTGs. The efficiency gains result from using a Stirling converter, a type of engine that relies on the heating and cooling of a gas to convert heat energy into electricity. The basic idea was that energy released by natural radioactive decay would be used to heat a gas, which would then expand. This pressure created by the expanding gas would be used to power an electric generator to produce electricity for a spacecraft. It was a novel design, but the research program on ASRG was mothballed in 2012 after costs began to drift tens of millions of dollars over budget without an end in sight.

That same year, however, researchers at NASA Glenn and the DOE’s Los Alamos National Laboratory achieved a breakthrough in a different type of nuclear power source that also relied on a Stirling converter. Keeping with the Simpsons themed acronyms, the Demonstration Using Flattop Fissions (DUFF) used nuclear fission, rather than natural radioactive decay, as an energy source for a Stirling converter. The energy produced during fission is used to heat a sodium fluid in a “heat pipe,” and this heat is used to drive a Stirling converter which converts the heat energy into the mechanical motion needed to drive an electrical generator.


Although DUFF only produced about 24 watts of power during its tests—not enough to power most light bulbs—it was the first successful test of a nuclear reactor intended for space applications made by NASA in nearly half-a-century. Its design prefigured Kilopower, which built on insights gleaned from the experiment to create a working prototype.

The Kilowatt reactor prototype that has been tested over the last few months uses two Stirling converters (the reactors that will eventually be sent to space will have eight converters) to produce up to 1 kilowatt of power. It’s core is a solid piece of uranium-235 that is about the size of a paper towel roll and a rod of boron carbide acts as a neutron moderator to control the bombardment of the uranium by neutrons. Before flight, the boron carbide would be fully inserted into the reactor to prevent a fission reaction. Once it is extracted, however, the fission reaction will begin and cannot be stopped completely, although the rate of fission—and hence heat output—can be controlled by the depth of the boron rod in the reactor. The heat from the fission reaction is used to heat liquid sodium, which transfers this heat to eight Stirling engines that convert the heat into mechanical motion to drive electric generators and produce electricity. The StirlingStriling engines used in the prototype were repurposed from the failed ASRG experiments, although McClure said custom Stirling engines will be developed for missions in the future.


One of the most staggering things about this design, however, is that this working nuclear reactor prototype was produced for only $18 million, an order of magnitude less than previous designs that never made it beyond the drawing board.

“When we got started, we looked at all the projects that had gone on before us,” Patrick McClure, the lead Kilopower engineer at Los Alamos National Laboratory told me on the phone. “We felt they made some mistakes by trying to build what I’m going to call a ‘sporty’ reactor. By sporty I mean they’re trying to get a lot of power for as low of a weight as possible. Some of their design choices were really pushing the bounds of engineering.”

An artist conception of four Kilopower units on Mars. NASA estimates that four reactors could power an initial outpost for a decade. Image: NASA

Rather than trying to reinvent the wheel, McClure said he and his colleagues working on Kilopower opted for more simple, cost effective reactor designs.

“We’d much rather have a reactor that’s real then one that’s a great design, but only exists on paper," McClure said. "We don’t want this to die.”

McClure told me that he and his colleagues are excited that the reactor works, but that there’s still a lot of work to do before it gets to space, much less a lunar or Martian colony. In addition to a safety review by NASA and an “ independent, interagency nuclear safety review panel,” it will need to be integrated into a specific mission so that the prototype can be designed to accommodate that mission’s needs. He said the Kilopower team is working with NASA management to determine the future of Kilopower, which will need to be tested in space before it's brought to the surface of the moon or Mars.

“We're excited and relieved that we got the test done, but anxious about where we go next,” McClure said. “Our ultimate goal is to get a nuclear reactor back in space, and if NASA’s ever going to use the reactor, you have to prove to the folks doing missions that it will really work in space.”

“There's clearly a desire to try to put one on the moon, but that might not happen until the mid-2020s,” McClure added. “The team would like to get one in space before then, but in the meantime we’ve got to convince some people.”