Power generators (RTGs)

How RTGs work

RTGs are solid-state heat engines based on the Seebeck effect: when two dissimilar conductive materials are joined in a closed circuit and their junctions are held at different temperatures, a DC voltage is generated. In an RTG, the "hot junction" of each thermoelectric couple is placed near the radioisotope heat source, while the "cold junction" is anchored to a radiator exposed to the ambient environment — deep space, a planetary atmosphere, or the lunar vacuum. The temperature difference across the thermoelectric legs drives current; many couples are connected electrically in series and thermally in parallel to accumulate useful voltage and power.

The performance of a thermoelectric material is characterized by the dimensionless figure of merit ZT = S²σT/κ, where S is the Seebeck coefficient, σ is electrical conductivity, κ is thermal conductivity, and T is absolute temperature. Higher ZT allows a greater fraction of the available Carnot efficiency to be captured. In practice, real space RTG materials (silicon-germanium, lead telluride, TAGS alloys, bismuth telluride) achieve ZT values around 1 over limited temperature ranges, holding overall system conversion efficiency to single-digit percentages. Key loss mechanisms include finite ZT, side heat losses around the thermoelectric legs, electrical and thermal contact resistances at material interfaces, and long-term material degradation through diffusion, microcracking, and aging.

RTG electrical output declines over the mission lifetime for two compounding reasons. First, Pu-238 decays at approximately 0.79 % per year in thermal power output. Second, thermoelectric elements degrade independently; the Voyager RTGs, for example, experienced a total electrical output decline substantially larger than fuel decay alone would predict, dropping to roughly 67 % of original output after decades rather than the ~83 % expected from fuel decay alone.

Major RTG designs

U.S. space RTGs have evolved through several generations, each optimized for the missions and manufacturing technologies of their era.

Lunar surface applications

On the lunar surface, the most critical function of an RTG is providing continuous "survival power" through the lunar night — a period of approximately 14 Earth days during which temperatures drop to around −170 °C and solar panels produce no electricity. An RTG simultaneously supplies usable electrical power and waste heat, which can maintain electronics and mechanisms above their minimum operating temperatures without large battery banks or complex active thermal control.

The historical precedent for lunar RTG use is the SNAP-27, which powered the Apollo Lunar Surface Experiments Package (ALSEP) on five Apollo missions. Each SNAP-27 unit produced 70 We at beginning of mission and operated continuously for up to 5–8 years — far beyond its two-year design lifetime — until NASA shut down the ALSEP network in 1977.

For future lunar operations under the Artemis Program, RTGs are considered practical for low-power remote instruments, small robotic vehicles that must survive repeated lunar nights, decentralized power nodes for outlying assets, and hardware sited in permanently shadowed or weakly illuminated regions near the poles. A review focused on Artemis-era lunar systems concluded that Pu-238 and Am-241 RTGs are not practical as the primary power source for a crewed base, but are well suited for low-power instruments or remote surface assets where uninterrupted operation outweighs total output considerations.

Plutonium-238 fuel supply

Pu-238 supply has been a persistent constraint on RTG-powered spaceflight. U.S. production at the Savannah River Site ended in 1988, after which NASA relied on legacy stockpiles supplemented by approximately 30–40 kg purchased from Russia in the 1990s and 2000s. Russia stopped selling Pu-238 to the United States around 2009, and U.S. planning no longer assumes any future Russian supply.

Production restarted in 2013 at Oak Ridge National Laboratory (ORNL), and in December 2015 NASA and the Department of Energy announced the first new batch — 50 grams of Pu-238 — ending a roughly 30-year production gap. The Pu-238 Supply Project is a joint DOE–NASA effort involving ORNL, Idaho National Laboratory (INL), and Los Alamos National Laboratory (LANL). The process uses neptunium-237 from DOE stockpiles, fabricates irradiation targets, irradiates them in a reactor, and chemically separates the product.

The programmatic goal is to reach and sustain a production rate of approximately 1.5 kg of PuO₂ per year — roughly the scale needed to support one large RTG-powered flagship mission every few years alongside smaller uses. A single MMRTG requires on the order of 4–5 kg of Pu-238, meaning several years of new production per unit. INL states that the U.S. now has sufficient Pu-238 to fuel space missions scheduled for approximately the next decade, but emphasizes that domestic supply remains a limited resource requiring continued production to support additional or larger missions beyond that horizon.

Next-generation RTG development

NASA and DOE have pursued two primary next-generation radioisotope power system lines since the 2010s. The first is the enhanced MMRTG (eMMRTG) concept, which proposed replacing the MMRTG's legacy PbTe/TAGS thermoelectric couples with skutterudite (SKD) couples to raise conversion efficiency while retaining the MMRTG's mechanical and safety architecture. The eMMRTG remained at the concept and technology-development level; it did not proceed to hardware development or flight qualification.

The second and more active line is the Next Generation RTG Project, a spaceflight system project within NASA's Radioisotope Power Systems Program executed with INL. Its primary goals are to re-establish manufacturing capability for silicon-germanium (SiGe) unicouple thermoelectric converters, refurbish stored GPHS-RTG hardware, and deliver a high-power, vacuum-optimized RTG for deep-space missions. The project is structured in modular configurations:

Mod 0 involves refurbishment of a legacy GPHS-RTG qualification unit for near-term availability around 2024. Mod 1 re-establishes full production of the GPHS-RTG design with updated manufacturing of thermoelectric couples and Step-2 GPHS fuel forms, targeting availability around 2028 and a performance of up to 295 We at beginning of life and up to 210 We at end of design life (17 years after BOL, including up to 3 years of ground storage). Mod 2 aims for approximately 290 We at end of design life in the early 2030s, using Mod 1 as a retrofit-ready platform for higher-performance thermoelectric converter technologies.

L3Harris Technologies serves as prime contractor for the flight-qualified Next Generation RTG optimized for vacuum. In 2021, INL tasked L3Harris to re-establish key GPHS-RTG technologies and update the design for new deep-space missions. The system cleared its Critical Design Review (CDR) on 2 April 2026. The DOE contract runs through 2027, culminating in a production readiness review. L3Harris has indicated that flight units could support NASA deep-space probes starting in the early 2030s, including a notional Uranus orbiter using two Next Gen RTGs.

A 2017 NASA Next-Generation RTG study had identified eight candidate thermoelectric couple configurations and winnowed them to three recommended RTG families — the Segmented RTG (SRTG), Segmented-Modular RTG (SMRTG), and Hybrid-Segmented Modular RTG (HSMRTG) — sized to provide approximately 50–600 We using 2–16 GPHS modules, enabling application across Discovery-, New Frontiers-, and Flagship-class missions. Advanced RTG research projects theoretical system efficiencies of approximately 10.8 % using high-ZT thermoelectric materials, roughly 90 % higher than the state-of-the-art ~5–6 %, with next-generation U.S. programs targeting approximately 13 % efficiency.