On the face of it, powering most spacecraft would appear to be a straightforward engineering problem. After all, with no clouds to obscure the sun, adorning a satellite with enough solar panels to supply its electrical needs seems like a no-brainer. Finding a way to support photovoltaic (PV) arrays of the proper size and making sure they’re properly oriented to maximize the amount of power harvested can be tricky, but having essentially unlimited energy streaming out from the sun greatly simplifies the overall problem.
Unfortunately, this really only holds for spacecraft operating relatively close to the sun. The tyranny of the inverse square law can’t be escaped, and out much beyond the orbit of Mars, the size that a PV array needs to be to capture useful amounts of the sun’s energy starts to make them prohibitive. That’s where radioisotope thermoelectric generators (RTGs) begin to make sense.
RTGs use the heat of decaying radioisotopes to generate electricity with thermocouples, and have powered spacecraft on missions to deep space for decades. Plutonium-238 has long been the fuel of choice for RTGs, but in the early 1990s, the Cold War-era stockpile of fuel was being depleted faster than it could be replenished. The lack of Pu-238 severely limited the number of deep space and planetary missions that NASA was able to support. Thankfully, recent developments at the Oak Ridge National Laboratory (ORNL) appear to have broken the bottleneck that had limited Pu-238 production. If it pays off, the deep space energy crisis may finally be over, and science far in the dark recesses of the solar system and beyond may be back on the table.
Hot and Ready
The development and use of RTGs for space missions closely parallels the build-up of space programs in the middle of the previous century. The first RTG was invented in 1954 in the “Atoms for Peace” era of efforts to find non-destructive ways to harness the power of the atom. The promise was great; essentially unlimited power with no moving parts, that could be scaled up or down to fit a huge range of applications, from powering remote terrestrial applications like lighthouses and remote weather stations to running implantable pacemakers with a power source that would outlive the patient.
It would not be until 1961 that the first RTG would go to space, aboard a Navy navigation satellite. The first of the deep-space missions to sport RTG power were the Pioneer missions in the early 1970s, which paved the way for the ultimate test of the RTG: Voyager 1 and Voyager 2. Each of those spacecraft uses three RTGs containing 4.5 kg of Pu-238, producing 480 watts of total power per vehicle at launch. More than forty years later, the RTGs are still working, their output greatly diminished by the passage of a fair fraction of the 87.7-year half-life of the fuel pellets and general degradation of the electrical system. But they still work, and probably will for at least another year or two.
The design of RTGs for space is fairly simple. Radioactive fuel is pressed into pellets that are covered with protective materials. The fuel is nestled into a container called the heat source, whose only job is to get hot. The heat source is slipped into another container, this one lined with thermocouples. The inside surface, in direct contact with the hot canister of decaying fuel, has the hot junctions of the thermocouple, while the cold junctions face out into the vacuum of space. The temperature difference is the key to creating electric power via the Seebeck effect, which is the same idea behind Peltier chips.
The Voyager mission’s RTGs were the “multi-hundred watt RTG” (MHW-RTG) that used silicon-germanium thermocouples, 312 per RTG. Later missions like Cassini and Galileo used a different design, the GPHS-RTG, or “general-purpose heat source RTG.” These RTGs were very similar to the MHW-RTG, with similar electrical design but a better, more efficient fuel package. The most recent RTGs are the “multi-mission RTGs” (MMRTGs) which have advanced thermocouples using lead telluride and an alloy called TAGS (tellurium, silver, germanium, and antimony). The Mars Curiosity rover has MMRTGs, as does the Mars 2020 Rover.
Robots to the Rescue
The nuclear alchemy used to produce Pu-238 and other radioisotopes is complex and extremely dangerous to conduct. Pu-238 was originally produced by bombarding uranium-238 with deuterium nuclei in a reactor, but later it was discovered that the production of Pu-239 for bomb cores yielded byproducts that could be more easily converted into Pu-238. Starting in the mid-1960s, all the Pu-238 for US civilian and military use was produced by neutron irradiation of neptunium-237 followed by chemical separation.
In 1988, the nuclear reactors at the Savannah River Site in South Carolina were turned off, and America’s Pu-238 spigot dried up. Even when the reactors were operating, production was a slow and hazardous process, resulting in only a few kilograms a year. Since 1993, NASA has sourced its Pu-238 from Russia, but they only managed to supply 16.5 kg of the stuff before they too shut down production.
In December of 2015, Oak Ridge National Laboratory in Tennessee produced the first Pu-238 in the US in nearly 30 years – 50 grams worth. The production process was laborious, with the bulk of the work going into making neptunium-237 pellets. The pellets were made by hand by adding Np-237 and aluminum powder and squeezing it into pellets suitable for neutron bombardment.
The best the lab could manage by hand was about 80 neptunium pellets a week, far short of the goal of 275 pellets a week. To achieve that level of production and ramp up from 50 grams of Pu-238 a year to 400 grams, ORNL has recently introduced an automated method of producing neptunium pellets. Details are hard to come by – plutonium production methods tend to be somewhat closely guarded for national security reasons – but if the neptunium pellets turn out to perform well in the reactors, ORNL could well be on the way to rebuilding the Pu-238 stockpile.
True, even if this automation advance proves itself, the total production capacity of Pu-238 in the US will still be under half a kilogram per year. But if the technology works, it can be replicated at both Los Alamos National Laboratory and at the Idaho National Lab, tripling the nation’s output and approaching NASA’s goal of 1.5 kilograms on plutonium-238 a year by 2025.
NASA appears optimistic that the deep-space energy crunch is nearing an end thanks to the new technology, but it’s still a long way from over. Nearly all of the 35 kilograms total inventory of Pu-238 that was available in 2015 has either been slated for future missions or is unsuitable for use in RTGs for deep-space.