Next Gen RTG: re-establishing thermoelectric power for deep space

What the work claims

The Next Generation RTG project claims it will deliver an unfueled, flight-qualified radioisotope thermoelectric generator based on the GPHS-RTG heritage design, by re-establishing silicon-germanium (SiGe) thermoelectric converter manufacturing capability that the US no longer possesses at production scale.¹ The project is a spaceflight system development effort within NASA's RPS Program, executed in partnership with Idaho National Laboratory (INL) and Battelle Energy Alliance. The primary technical objective is not to invent a new converter but to rebuild the manufacturing capability for a proven one.

The project will refurbish GPHS-RTG Flight Unit F-5 at INL, which is existing flight hardware from the heritage program. The paper states that the MMRTG is currently the only spaceflight-qualified RPS system available, and that the Next Gen RTG aims to provide a higher-power, vacuum-rated alternative for deep space missions that operate in hard vacuum (unlike MMRTG, which was designed to operate in planetary atmospheres as well as vacuum).¹

How it works

A radioisotope thermoelectric generator produces electricity from the heat of decaying plutonium-238. The heat source is the General Purpose Heat Source (GPHS), a standardized module containing pressed plutonium-238 dioxide fuel. The thermoelectric converter sits between the hot heat source and the cold of space, and the temperature difference across the converter drives electrons through the thermoelectric material, generating electric power directly with no moving parts.¹

The Next Gen RTG uses silicon-germanium (SiGe) unicouple thermoelectric converters, the same converter technology used in the GPHS-RTG that flew on Galileo, Ulysses, Cassini, and New Horizons. A SiGe unicouple is a pair of n-type and p-type semiconductor legs that generate voltage from the Seebeck effect across the hot-to-cold temperature gradient. The conversion efficiency of SiGe thermoelectrics is modest, in the range of 6 to 7 percent of thermal input converted to electricity, measured at the converter level under operating temperature. The rest of the thermal energy is radiated away as waste heat.

The project will refurbish GPHS-RTG Flight Unit F-5, a heritage flight unit held at INL. Refurbishment means inspecting, qualifying, and where necessary remanufacturing the converter and structural components to flight standard. The key manufacturing capability being re-established is the SiGe unicouple fabrication: growing the SiGe material, dicing it into legs, and assembling the unicouples with the hot and cold shoes and the graphite components that make up the converter stack. This capability was allowed to lapse after the GPHS-RTG production runs, and rebuilding it is the project's central technical challenge.

The strongest case

The strongest case is heritage. The GPHS-RTG is a proven design that flew four major deep space missions. Galileo used two GPHS-RTGs and reached Jupiter. Ulysses used one and reached a polar solar orbit. Cassini used three and orbited Saturn for 13 years. New Horizons used one and reached Pluto. The SiGe converter technology has flight heritage across decades and across the outer solar system. The Next Gen RTG is not inventing a new converter; it is rebuilding a proven one, which is a lower-risk approach than developing a new thermoelectric material or a new dynamic converter.¹

The decision to refurbish F-5 rather than build a completely new unit from scratch reduces both cost and schedule. F-5 is existing flight hardware with a known configuration. The structural components, the heat source assembly, and the housing can be inspected and qualified against existing acceptance criteria. The new manufacturing is concentrated in the SiGe unicouple converter stack, which is the capability that lapsed.

The vacuum-rated design is a real advantage for outer planet missions. MMRTG, the current system, was designed to operate in a planetary atmosphere (Mars surface) as well as vacuum, which imposes design constraints on the insulation and heat rejection. A vacuum-rated RTG optimized for deep space can achieve higher cold-side temperature differentials and thus higher conversion efficiency, because the only heat rejection path is radiation to space, with no convective losses.

Where a skeptic should push

The TRL honesty rule requires distinguishing what has flight heritage from what is being rebuilt. The GPHS-RTG design has flight heritage, but the manufacturing capability does not. The SiGe unicouple production line is not operating. The paper states that the project will re-establish this capability, which means it currently does not exist at the scale needed for flight production. A heritage design is only as good as the manufacturing capability that can produce it to flight standard, and that capability must be qualified, not assumed.¹

The 6 to 7 percent conversion efficiency of SiGe thermoelectrics is low. For a given thermal input from the plutonium-238 heat source, the electrical output is a small fraction of the heat. This means a large heat source mass is needed for a modest electrical output. The MMRTG, by comparison, uses lead telluride TAGS thermoelectrics with comparable efficiency. The Next Gen RTG's advantage over MMRTG is vacuum rating and potentially higher total power from more GPHS modules, not a step change in conversion efficiency. A dynamic converter (Stirling) could achieve 20 to 30 percent efficiency, but no dynamic RPS has flown in space, which is the subject of a separate analysis.

The F-5 refurbishment carries risk. A flight unit held in storage for decades may have degradation in non-metallic components, bond lines, or the thermoelectric material itself. The paper states the unit will be refurbished at INL, but the scope of refurbishment (inspect and qualify versus remanufacture) is not fully detailed. If F-5's condition requires more remanufacture than expected, the cost and schedule will grow.

Finally, the plutonium-238 supply is the elephant in the room. The Next Gen RTG is unfueled in its initial delivery, which means the fuel is a separate issue. US plutonium-238 production has been restarted at a low rate, and the inventory must be allocated across all future RPS missions. A flight-qualified but unfueled RTG is only useful if the fuel is available when the mission needs it.

What it means for power in space

The non-obvious implication is that the bottleneck is not the RTG design but the manufacturing capability. A heritage design that cannot be produced is not an option. The specific mechanism is the SiGe unicouple fabrication process: the material growth, dicing, and assembly steps that convert a proven design into flight hardware. For power in space, this means the decision about whether to invest in re-establishing thermoelectric manufacturing or in developing dynamic conversion (Stirling) is a decision about which manufacturing base to rebuild. The Next Gen RTG project bets on the proven, lower-efficiency thermoelectric base rather than the higher-efficiency but unflown dynamic base.¹

The genuine threat is fuel supply. The Next Gen RTG will be delivered unfueled. The plutonium-238 inventory is limited, and US production is at a low rate. If multiple deep space missions compete for the same fuel allocation, a flight-qualified RTG may sit without fuel, or a funded mission may be delayed waiting for fuel. The specific mechanism is the slow production ramp of plutonium-238, which constrains the number of RPS missions that can fly per decade regardless of how many RTGs are built.

The opportunity is that a vacuum-rated, higher-power thermoelectric RTG enables missions that MMRTG cannot support efficiently. Outer planet missions to the ice giants (Uranus, Neptune) or deep solar system probes operate in hard vacuum and at large solar distances where solar arrays are not viable. The Next Gen RTG, with its heritage design and vacuum rating, is the power system for the missions beyond Mars that need reliable, long-life power without moving parts. Re-establishing the SiGe manufacturing base keeps that mission class open.

The bottom line

What is solid is the heritage: the GPHS-RTG design is flight-proven across four outer planet missions, and the F-5 refurbishment leverages existing flight hardware. What is not established is the re-established SiGe manufacturing capability at production scale, which is the project's central technical objective and is not yet demonstrated. Confidence is moderate that the Next Gen RTG can deliver an unfueled, flight-qualified unit on the project's schedule, contingent on the SiGe unicouple manufacturing line being qualified. The reading would be strengthened by demonstrated converter production at flight quality and rate. It would be undercut if the F-5 refurbishment scope expands beyond inspection and qualification into significant remanufacture, or if the plutonium-238 supply cannot support a fueled flight unit when a mission requires it.