Aerospace and Defense

Pu-238: Fuel for Deep Space Journeys

02 February 2017

The U.S. space agency NASA is particular when it comes to the fuel used to power its longest space missions – only plutonium-238 will do.

Yet the Cold War-era fuel that powered the New Horizons journey to Pluto and will drive the Mars 2020 Rover mission is in short supply, as the United States hasn’t produced any since the late 1980s.

The agency has been working with the Department of Energy to produce more of the vital substance, and scientists at the Oak Ridge National Laboratory (ORNL) have produced two 50 gram (1.8 ounce) batches, with another planned for later in 2016.

ORNL plans to scale up production to 1.5 kilograms (3.3 pounds) per year in the early 2020s. The Pu-238 generates heat that is converted to electricity by a multi-mission radioisotope thermoelectric generator (MMRTG). The units measure about 64 by 66 centimeters (25 by 26 inches), weigh approximately 45 kg (94 lb) and contains 4.8 kg (10.6 lb) of Pu-238.

Concept of NASA Mars 2020 Rover.Concept of NASA Mars 2020 Rover.The Mars Rover 2020 mission is the only mission currently slated to use the nuclear fuel, but with the multi-year lead time needed to produce sufficient fuel, NASA decided it should start producing Pu-238 sooner rather than later.

Pu-238: The Ideal Space Fuel?

The MMRTG’s that Pu-238 powers provide NASA with reliable, relatively safe, and maintenance-free power for space missions. Although it’s expensive to produce and dangerous to use, engineers are familiar with Pu-238. Indeed, they have been using it to fuel space missions for more decades. (The unit in Voyager 1 will mark its 40th anniversary in early September.)

“It’s very nice because it has a very high energy density,” says Casey Dreier, director of space policy at the Planetary Society. “It’s also relatively safe for a radioactive isotope. It’s stable, so they can make it in a ceramic form that is not easily destroyed and not easily vaporized, and it’s very, very hot, so it gives off a ton of energy.”

It’s fortunate that Pu-238 does emit a lot of energy, as the process of turning it into electricity is not very efficient “at the few percent to an upper limit of about 5%,” says Ralph McNutt, a space physicist at the Johns Hopkins Applied Physics Laboratory.

Another Pu-238 advantage is that the Cold War-era infrastructure to manage, process, and move it is already in place “so you don’t have to create a whole new everything for it,” Dreier says.

Handling Pu-238 is relatively safe, compared to other radioactive isotopes. Dreier says it’s primarily an alpha particle emitter, and “it’s little helium atoms without nuclei can be stopped with a piece of paper. You can hold it in your hand and you wouldn’t get a dangerous dose. You would burn your hand, but you wouldn’t get a dangerous dose of radiation.”

Casey Dreier, Planetary SocietyCasey Dreier, Planetary SocietyPu-238 throws off a lot of heat, which is good news for spacecraft that need to keep operating at low temperatures.

“I can make use of that heat in order to keep the spacecraft itself from freezing, to keep the electronics at a usable temperature, and also from keeping fuel lines and oxidizer lines from freezing up,” McNutt says.

Solar is the other main option for powering a spacecraft. It’s an excellent solution for the International Space Station, for example, which is powered by an acre of solar panels. The current outer limit for more distant solar-powered missions appears to be Jupiter; beyond that and solar radiation is too weak to produce sufficient power. Pu-238 also keeps producing energy in a Mars dust storm; dust-covered solar panels do not.

It Starts with Neptunium-237

The process of producing Pu-238 is elaborate and involves three of the United States’ national laboratories. It starts with obtaining neptunium oxide feedstock from the Idaho National Laboratory (INL), says Bob Wham, a chemical engineer who leads the Plutonium-238 for Space Missions Project for the Nuclear Security and Isotope Technology Division at ORNL.

Wham’s team at ORNL then conditions the neptunium oxide to remove some of the radioactive decay daughters (protactinium-233) so they can get as high an assay as possible. Next, the neptunium is blended with aluminum powder and pressed into pellets. The pellets are loaded in an aluminum tube backfilled with helium.

Aluminum is an ideal medium for the powder and tube as “it’s transparent to neutrons and it gives very high thermal conductivity,” Wham says.

Once the tubes pass inspection, they are placed into ORNL’s high flux isotope reactor and irradiated until 85% of the plutonium produced from the process is Pu-238 (the rest is plutonium-239). The process fissions a small percentage of the neptunium. That means the irradiated material needs to age (cool down) for up to a year before it’s moved from the reactor pool to the hot cells.

Ralph McNutt, Johns Hopkins Applied Physics Laboratory.Ralph McNutt, Johns Hopkins Applied Physics Laboratory.In the hot cells, engineers dissolve the aluminum away from the neptunium and plutonium oxides, and then separate the neptunium from the plutonium and fission products. Only about 14-15% of the neptunium is converted to plutonium, according to Wham. The rest can be reused.

Once they’ve finished processing it (which can take up to three years), ORNL ships the Pu-238 to Los Alamos National Laboratory (LANL) in New Mexico. LANL takes the Pu-238 powder and performs additional conditioning steps, presses it into pellets and puts the pellets into an iridium clad. After passing inspection, the clads are shipped to INL, which assembles and tests the MMRTGs for NASA. The complete cycle, from neptunium-237 to ready-to-use MMRTGs, takes several years.

Pu-238 as a Last-Resort Fuel

As suitable a fuel as Pu-238 is for certain space missions, it’s not without downsides. The decision to use nuclear fuel adds significant cost to a mission and can delay it for years while it goes through the qualification process.

The United States National Space Policy states that nuclear power systems can be used if they “significantly enhance space exploration or operational capabilities,” so it’s incumbent upon mission planners to look for another power source. Advances in using solar technology in space allowed Barry Goldstein, project director of NASA’s planned Europa mission, to consider it as an alternative to a Pu-238 powered mission.

The Europa team performed a “significant number of tests, on various solar cells, to see if they would survive the combined impacts of the radiation dose, as well as the temperature extremes” that the mission would require, Goldstein says. Their calculations, and data from NASA’s solar-powered Juno mission, showed that “solar power would work very well for our mission,” he says.

Bob Wham, ORNLBob Wham, ORNLDespite its cost and dangerous nature, there’s no real alternative to Pu-238 for some space missions. It’s not for lack of trying, says McNutt. “People have really looked hard at this trying to figure out if there are alternatives, and a lot of money has been spent, and a lot of effort.”

Many of the thousands of known radioisotopes have been investigated, but all have been found wanting. The European Space Agency has been investigating americium-241 as a possible Pu-238 substitute. McNutt says that Am-241 is more readily available than Pu-238 but lacks its energy density.

Engineers need sure things to power billion-dollar spacecraft, Dreier says, and are understandably conservative about what energy sources power them. As a result, NASA has invested in the development of more efficient radioisotope generators.

Although it produced a small amount of the material, Dreier says the initial demonstration of a new process for creating plutonium-238 is “a very good step in progress to getting full production back online.”

To contact the author of this article, email

Powered by CR4, the Engineering Community

Discussion – 11 comments

By posting a comment you confirm that you have read and accept our Posting Rules and Terms of Use.
Re: Pu-238: Fuel for Deep Space Journeys
2017-Feb-02 6:19 AM

"..“it’s little helium atoms without nuclei can be stopped with a piece of paper...'

Hmmm. Nope. Little helium atoms without electrons, sure. Electrons are not nuclei.

Re: Pu-238: Fuel for Deep Space Journeys
2017-May-05 4:24 PM

Pu-238 conversion to energy only 5% thermal efficiency? Not good.

Why not do something exotic smart, and use a thermal radioactive source to heat silicon (or other essentially inert material) to 1414 °C, (or even higher), several container materials already proven OK, including SiC, C, W, Ta, etc. (you physics geniuses would have to work out the likelihood of neutron or alpha activation of the thermal storage media, and the container) Next the light emitted at that temperature is well suited to TPV devices that have already reached above 20% TE (in their sleep mode). Some say we will have TPV conversion efficiency near 80% in the not distant future. Band selection is key.

Do not bother with attempting thermopile in addition to, as it will just siphon off more energy than it produces.

Re: Pu-238: Fuel for Deep Space Journeys
In reply to #2
2017-May-05 10:57 PM

I think it is the mass and volume constraints of going to space that lead to a low efficiency. My guess is that the heat sink is the weak link....being able to keep components to a temperature that affords some reliability without having the radiator be gargantuan must be tricky.


By the way, what is this alpha activation of which you speak?

Neutron production by pu238 is not very strong.

Re: Pu-238: Fuel for Deep Space Journeys
In reply to #3
2017-May-08 9:21 AM

Probably that was a questionable statement, but suppose you accelerated an alpha to near light speed. I makes a good cue ball for some other nucleus at that point, and could result in all manner of havoc. Do not ask me how it would be accelerated to near c.

I had second thoughts about including that statement. I agree the neutrons are not too strong from it.

Why have a heat sink at all when you want the thing to be hot? The Thermo-Photovoltaic elements absorb energy at their active spectral band, and reflect back the rest, by design. In space, it should in fact be easier to make thermal silicon tech work better. If you are talking about radiative cooling of CB components, in space, that is a horse of a totally different color. One wants to use high impedance circuitry in space to minimize power consumption in the first place. I expected that all board cooling is done radiatively.

Re: Pu-238: Fuel for Deep Space Journeys
In reply to #4
2017-May-09 5:59 PM

"...Why have a heat sink at all when you want the thing to be hot? The Thermo-Photovoltaic elements absorb energy at their active spectral band, and reflect back the rest, by design. In space, it should in fact be easier to make thermal silicon tech work better..."


Why have a heat sink at all? - b/c keeping everything organized and doing its intended job becomes much more difficult when sporting solid phase for key components becomes exceedingly passé and subject to thermodiscrimination.


Assuming the absorber/converter is semi-decent at absorbing the frequencies it can use and reflecting most of the rest back, the emitter is going to continue to increase in temperature. This will lead to the absorber/converter receiving more and more radiation. Normal real world inefficiencies in the converter would yield more and more added heat. Some form of catastrophic self disassembly would be a likely outcome.

Re: Pu-238: Fuel for Deep Space Journeys
In reply to #5
2017-May-10 10:32 AM

I am not wishing to start an argument with you, but here I go anyway.

(1) nuclear fuel source is regulatable to X heat output.

(2) X raises temperature of the silicon (in the case of zero G) to just sub-melting, since liquid containment at zero G becomes only slightly more complicated.

(3) Y heat output is couple out of the silicon (or other suitable high temperature "black body")

(4) Z portion of Y is converted to electricity in TPV panels (that include the necessary band selection optics). (one still has to apply the thermal efficiency of the TPV to Z).

(5) Y-Z heat output of the silicon is immediately returned in the band selection optics to the source.

(6) Heat (actual selected light) not converted to electricity heats up the TPV (actually increasing net efficiency (to some limit)). This quantity is τ (tau), which once radiative equilibrium is surpassed between the silicon (or other) and the TPV leads to the following: Y- Z is reflected heat into the storage medium (true), but there is still the quantity τ to be absorbed and converted or radiated into space. If there is a coaxial material to the TPV that can absorb and utilize a portion of τ, then the residual heat is further reduced to long wave thermal radiation peak (to somewhere in the infrared). Call this quantity ξ that will be coupled out the top (and bottom if you can decide up and down in space) of the secondary converter. Supposedly, in this way the system is regulated as to temperature (especially considering any turn-down on the reactor pile), and will operate at a stable set of temperatures indefinitely (until the fuel runs out).

Re: Pu-238: Fuel for Deep Space Journeys
In reply to #6
2017-May-11 11:21 AM

There is no significant negative (or positive) temperature reactivity associated with decay heat. These are not reactors in the sense there is no neutron population controlling the power.

Decay heat is what is beingutilized. Basically things heat up until the energy output matches the energy input. In insulating vacuum of space, for objects of that aren't very big, a significant continual heat input can lead to very high equilibrium temps.

Re: Pu-238: Fuel for Deep Space Journeys
In reply to #7
2017-May-11 12:19 PM

Okay, so we are talking decay heat, and there is no apparent way to moderate the rate of decay, so that leaves us back in the swamp? I expect the only way to further control that might be to expend some of that energy (heat) as radiation out the back of the craft, and to thus provide a minuscule thrust boost integration over time.

Re: Pu-238: Fuel for Deep Space Journeys
In reply to #8
2017-May-11 10:13 PM

Hey, I thought you might find this interesting. Not sure which post would be most appropriate...,but this one was at least close.

Re: Pu-238: Fuel for Deep Space Journeys
In reply to #9
2017-May-12 9:26 AM

I looked the first few pages, and I find that sort of thing fascinating, especially when it matches iron content of the real object.

When I first saw it, I though O2 + Si → SiO2, then I realized you are talking about billions of degrees and millions of grams/cc. I bit more drastic than any conditions I hope to be doing any actual testing on any time soon, (except with a very good telescope). Awesome sauce! Thank you, and happy morning!

Re: Pu-238: Fuel for Deep Space Journeys
In reply to #10
2017-May-12 2:32 PM

So the 'explosive' part in the title, had be thinking the 'burning' was of the combustion type. And then I started reading and realized they were talking about stars, yet it was still very interesting. I'm glad you had a similar experience.

Engineering Newsletter Signup
Get the Engineering360
Stay up to date on:
Our flagship newsletter covers all the technologies engineers need for new product development across disciplines and industries.