Since the 1960s and until today, scores of satellites and probes launched by humans have explored Earth’s planetary neighbours, thousands of stars, the outer limits of the Solar System and the ceaseless wonders within and without. We have consumed the sights they have taken in and, through them, have built a vision of the cosmos as an unending void embedded with beauty, like a pristine glass marble impregnated with crystals frozen in space.
These missions were built, launched and manoeuvred across vast distances by, at their most essential, humanity’s dreams and determination. However, in space, these attributes count only as much as the mission’s engineering components – and never more so than the source of energy.
For the missions that don’t have to go far from Earth, where sunlight is abundant enough, the probes have been powered by solar panels. But human-launched missions have gone much farther – the Voyager 1 probe is just outside the Solar System – and we intend to keep taking more steps that way. In such cases, the preferred source of energy has been the radioisotope thermoelectric generator, or RTG.
Popular examples of probes powered by an RTG include the Cassini orbiter around Saturn, the New Horizon probe to the outer Solar System, the Curiosity rover on Mars and the veteran Voyager probes. There are dozens of other examples as well. In early 2016, it was rumoured that ISRO’s Chandrayaan 2 mission to the Moon would also be powered by an RTG; it has since been clarified that solar panels will do the job.
An RTG, unlike conventional generators we’re used to on Earth, has no moving parts. Instead, it converts the heat emitted by a radioactive material into electricity. RTGs often provide a few hundred watts at best, but this much is sufficient to power multiple spacefaring instruments. For example, the Curiosity rover has an RTG that converts the 2,000 W emitted by the decay of the 4.8 kg of plutonium-238 dioxide at its heart to produce 125 W of electric power. If this efficiency pales in comparison to a steam turbine’s (~45%) – the excess heat from the isotope can always be used to warm the rover when its surroundings become too cold.
The working principle
Because an RTG has no moving parts and doesn’t require regular maintenance, it is well suited for powering gadgets that can’t be attended to for long durations. Apart from missions to space, the Soviet Union also used RTGs to power a series of lighthouses it set up inside the Arctic Circle during the Cold War. Critics have expressed concern about these installations because of their susceptibility to theft – especially by terrorists looking for the strontium-90 isotope inside – since they are not continuously monitored.
To convert heat, or thermal energy, from the decaying radioactive isotope into electric energy, an RTG draws upon the Seebeck effect, named for Thomas Johann Seebeck, who discovered it in 1821. The effect is a type of thermoelectric effect (the others are the Peltier effect and the Thomson effect) in which an electric current is produced at the junction between two wires of different materials if they are at different temperatures.
This happens because the temperature difference causes electrons in the different materials to flow in different ways. As a result, a voltage is created across the junction that then leads to an electric current. A thermocouple uses such wires meeting each other at multiple points. Since the Seebeck effect works both ways, the thermocouple can be used to produce heat if a voltage is applied – or produce a current if heat is applied. In an RTG, the heat is supplied by radioactive decay.
Like all things about the American space programme, which was the first to use RTGs, development of the generators was commissioned by the American military. In 1947, an offshoot of the Manhattan Project that had been involved in building triggers for the plutonium bomb was set up in Ohio. It would later become known as Mound Laboratories. Two of its engineers, John Birden and Ken Jordan, patented the first RTG in 1958-1959 (as a “nuclear powered milliwatt generator”). In 1957, the US Army Signal Corps Research and Development Laboratories approached them for help with picking a suitable radionuclide to use in a thermoelectric generator.
According to a report published by the US Department of Energy in October 1960, this is how Mound went about shortlisting their candidates:
Initial elimination of isotopes as heat sources was made on the basis of half-life. Any isotope which had a half-life of less than 100 days or greater than 100 years was discarded. A few exceptions were made to insure not overlooking a likely isotope. Further eliminations were made of isotopes which (1) were gamma emitters only, (2) had radioactive transitions yielding gamma with energies greater than one million electron volts, (3) had transitions having an occurrence of gamma emission greater than ten per cent with energies over 0.1 million electron volts, (4) had decay schemes which involved daughter elements having any of the preceding gamma characteristics, or (5) had particle energies so slight that it would require more than one per cent conversion efficiency to give a 0.01 watt output. A literature search of the nuclides was conducted to eliminate isotopes which did not have desired nuclear properties.
Better engines, tougher problems
The shortlist had 47 isotopes. Plutonium-238 was among them. The report’s authors wrote that while plutonium-238 was produced in minute quantities as a byproduct in nuclear reactors, “the large amount of enriched uranium fuel used in reactors makes this scheme appear feasible”. Indeed, in February 2015, NASA reportedly had enough Pu-238 on its hands to fuel at least three more missions at Curiosity’s scale. It has since been established that one of those missions will be the Mars 2020 rover. Another could be the Europa Clipper mission slated for the mid-2020s.
But whether or not there will be a shortage, many space agencies have been working on advancing RTGs to make them more powerful and efficient. One such is the Stirling radioisotope generator – SRG.
The Stirling generator as such has been around since the 17th century, went through a period of disuse paralleling the rise of the steam engine and was finally revived in the pre-transistors period of the 20th century. Its working principle is like an internal combustion engine’s – except the Stirling has no internal combustion. That is, its heat source lies outside the engine.
According to Wikipedia, a Stirling generator works by the “cyclic compression and expansion of air or other gas at different temperatures, such that there is a net conversion of heat energy to mechanical work” (see GIF). As a result, it only requires a suitable working fluid – air often works – and two cylinders to be maintained at different temperatures. As with an RTG’s thermocouple, the heat source here is a radioactive isotope.
Because the working fluid isn’t being combusted, an SRG can be almost completely noiseless when on. Its other advantage is that, because its heat source is external to the engine itself and not inside it, its moving parts can last for much longer before they will have to be replaced.
NASA had been working on an advanced version of the SRG, using helium as a working fluid, until 2013, when the program was scrapped*. The principal reason? NASA’s budgets for interplanetary missions had been on a steady decline until then, so scrapping work on the advanced SRG freed up over $500 million (Rs 3,300 crore) that the agency could reallocate to keep some missions alive – as well as spared NASA the obligation to accommodate ASRGs on low-cost missions for testing.
This does complicate issues for NASA because, for the same amount of power, a multi-mission RTG uses four-times as much plutonium-238 as an SRG would. On the other hand, because an ASRG has moving parts while a multi-mission RTG doesn’t, missions will have to carry a spare in case the first one flops.
Finally, the US Department of Energy has a limited amount of plutonium-238, although it announced in November 2013 that it would produce more of the isotope from 2020, at 1 kg/year and a total outlay of $200 million. This means it will take a decade to build two more multi-mission RTGs.
Given these facts, how did NASA’s mission profile for the next decade change when it cancelled the ASRG programme?
The generator in our future
Van R. Kane, an ecologist at the University of Washington and a planetary exploration enthusiast, drew up this table in December 2013 based on decadal survey reports published by NASA. That the Saturn probe could ‘probably’ work with solar power is crucial: if it does, then NASA will have more of the isotope than it will need for missions in the next decade. If the Saturn probe will surely need a multi-mission RTG, then the agency will have less than it needs.
Kane further wrote on his blog that the Uranus orbiter (HORUS) wouldn’t be able to fly until the new production run of plutonium-238 had run for a while, and that the missions most impacted would likely be those under NASA’s Discovery umbrella: “The expectation had been that NASA would make at least one pair of ASRGs available for a Discovery mission. Engineers and scientists came up with clever ideas for ASRG-based missions – the comet hopper, a Titan lake lander, an orbiter to revisit Titan and Enceladus, a Uranus flyby and others. With [multi-mission RTGs] now the only option, NASA needs to hoard its supply of Pu-238.”
Unlike NASA – and ISRO, both of which are in countries that have access to domestic plutonium-238 – the European Space Agency (ESA) has to consider alternatives. One of them is americium-241 (one of the shortlisted candidates in the 1960 report).
According to Tim Tinsley, of the UK’s National Nuclear Laboratory, “When nuclear fuel is reprocessed, the plutonium is separated from the uranium and fission products and stored for reuse as fuel in civil nuclear reactors. Nuclear fuel that has been in a civil reactor will contain a range of plutonium isotopes including 241, which has a decay half-life of 14 years to americium-241. The long-term storage of civil-separated plutonium will therefore produce very isotopically pure americium-241 via this beta decay. Like the americium in used nuclear fuel, this americium is also considered by the nuclear industry to be a waste product that needs to be removed before the plutonium can be reused in nuclear fuel.”
Disadvantages: the radiation emitted by plutonium-238 is easier to shield against and its power density (0.5 W/g) is five-times higher. Advantages: americium-241 has a longer half-life, which means its power density decreases slower, and it is produced in a purer form (99.9% v. 80%). Biggest advantage: UK, and Europe, have access better access to americium-241 than to plutonium-238. Future Brexit deals could change this).
A.R. Sundararajan, a retired scientist at the Department of Atomic Energy (DAE), has written that the DAE has the requisite facilities to prepare plutonium-238 for ISRO to use in its own RTGs. It’s only that the need for using such generators hasn’t yet come up. Moreover, ISRO has not officially conceived of missions that would take Indian instruments to places in the Solar System where sunlight is intermittent, weak or unavailable. But it will. The Moon has already been visited and so has Mars. Follow-up missions to these bodies are in the works, as is an orbiter to Venus. Endeavours by other space agencies (e.g. JUNO) have demonstrated that solar panels will work all the way to Jupiter as well.
But once Saturn is in the crosshairs, or even the far side of the Moon, then RTGs and SRGs are the only existing way out (notwithstanding advances in low-light/low-temperature solar cells). The DAE has – or can manufacture – access to plutonium-238. ISRO can either develop the necessary technology indigenously, in which case it has to chalk out a long-term strategy and secure the necessary funding. Tinsley has written that ten years is an apposite period. Alternatively, ISRO could import it from the US under the 123 Agreement, although Sundararajan has said this might not be feasible.
But in every way, radioisotope generators are in our future.
*A residual programme continues in the wake but usable units are not expected before 2028.