A radioisotope thermoelectric generator (RTG) is an electrical power source that converts radioactive decay heat directly into electricity through the thermoelectric effect. A radioactive isotope supplies heat continuously, while solid-state conversion elements generate a voltage between hot and cold surfaces. Unlike a nuclear reactor, an RTG does not sustain a controlled nuclear chain reaction; unlike an ordinary battery, it does not obtain its energy from an electrochemical reaction.
RTGs are particularly important where sunlight is weak, intermittent, or unavailable and where maintenance or refueling is impractical. They have powered planetary landers, deep-space probes, lunar scientific instruments, and remote terrestrial equipment. Their electrical outputs are usually modest—typically from a few watts to several hundred watts per generator—but they can operate for decades. Their defining advantages are endurance, environmental independence, and the absence of moving parts in the conversion system.
The technology combines nuclear-material production, high-temperature materials engineering, thermal management, and spacecraft power-system design. Its limitations include low conversion efficiency, declining output, restricted fuel supplies, and the need to contain radioactive material during manufacture, transport, launch, operation, and disposal.
Physical principles and energy conversion
An RTG begins with energy released by unstable atomic nuclei. Depending on the isotope, decay produces charged particles, photons, or other radiation. Much of the useful decay energy is absorbed within the fuel and its surrounding structure and becomes heat. Energy carried away by neutrinos, where present, is not recoverable by the generator. The heat source therefore behaves as a continuously operating thermal reservoir whose output declines according to the isotope’s decay law.
Electricity is produced through the Seebeck effect, discovered by Thomas Johann Seebeck in 1821. When two dissimilar conducting materials form a circuit whose junctions are maintained at different temperatures, an electromotive force develops. Practical RTGs use many thermocouples, usually made from semiconductor materials, connected electrically in series and arranged thermally between the heat source and a cooler outer structure. For an approximately constant effective Seebeck coefficient , the open-circuit voltage of one couple is
where and are the hot- and cold-junction absolute temperatures. Actual performance requires accounting for the temperature dependence of material properties, electrical resistance, contact losses, and heat conducted through the converter without producing electricity.
The efficiency is constrained by thermodynamics and by the properties of thermoelectric materials. A commonly used dimensionless measure of material quality is
where is electrical conductivity, is thermal conductivity, and is absolute temperature. A useful material must conduct electricity well while limiting heat flow that bypasses conversion. These requirements conflict because charge carriers themselves can transport both electrical energy and heat. The design problem is therefore more demanding than simply finding an electrically conductive material.
Flight RTGs have generally converted roughly 4–7 percent of their initial thermal power into electricity, depending on design and operating conditions. A generator producing about 110 watts electrically can consequently release around 2,000 watts thermally. Most of that heat must leave through the generator housing and radiator fins or be deliberately transferred into the spacecraft.
The cold side is not necessarily cold in everyday terms: it may operate well above room temperature. What matters is a sustained temperature difference. In space, thermal radiation ultimately carries unused heat away; in an atmosphere, convection can also contribute. An RTG therefore needs both a dependable heat source and a heat-rejection system matched to its environment.
Radioisotope fuels and their properties
An RTG fuel must combine a suitable half-life, high heat output per unit mass, manageable radiation emissions, chemical stability, and practical availability. A short-lived isotope can provide high specific power but lose output too rapidly for a long mission. A very long-lived isotope changes little during service but may require an impractically large fuel mass. Radiation type also matters: penetrating gamma rays and energetic bremsstrahlung can impose heavy shielding requirements even when the thermal output is attractive.
Plutonium-238 is the principal fuel for United States space RTGs. Its half-life is approximately 87.7 years, and it decays mainly by alpha emission. Pure plutonium-238 produces about 0.57 watts of heat per gram. Its alpha particles are readily stopped within the fuel and surrounding containment, allowing a relatively compact system compared with fuels that emit substantial penetrating radiation. Nevertheless, actual fuel has neutron and photon emissions arising from minor isotopes, spontaneous fission, and reactions involving surrounding light elements; an RTG is not radiation-free.
Space heat sources generally use plutonium dioxide, , rather than metallic plutonium. This refractory ceramic has a high melting temperature and better chemical stability for containment applications. Fuel is consolidated into pellets and enclosed in protective cladding. The isotope composition, oxygen composition, microstructure, and fabrication history affect thermal behavior, radiation emissions, and response to mechanical loading.
Strontium-90 has been widely used in terrestrial RTGs, especially in former Soviet remote-power installations. It has a half-life of about 28.8 years and undergoes beta decay to yttrium-90, which subsequently decays to stable zirconium-90. The decay chain produces useful heat, but energetic beta particles generate bremsstrahlung radiation when slowed in surrounding materials. Shielding and fuel-form design therefore differ substantially from those of plutonium-238 systems. Strontium titanate has been used as a durable ceramic fuel form.
Other isotopes have been considered or used for specialized purposes. Polonium-210 provides very high specific thermal power but has a half-life of only about 138 days, making it unsuitable for multidecade missions. Americium-241, with a half-life of about 432 years, offers a much slower decline in heat production but substantially lower specific thermal power than plutonium-238 and different shielding requirements.
No isotope is optimal independently of the mission. A compact planetary spacecraft, an unattended terrestrial beacon, and an instrument intended to survive for many decades impose different priorities. Fuel selection determines not only generator mass and longevity but also manufacturing infrastructure, transport arrangements, radiation protection, and end-of-service obligations.
Generator construction and spacecraft integration
An RTG consists of several interacting subsystems: radioactive fuel, fuel containment, a heat-distribution structure, thermoelectric conversion elements, thermal insulation, an outer housing, and heat-rejection surfaces. Their arrangement must maintain a useful temperature gradient while allowing the fuel enclosure to withstand launch vibration, prolonged heating, and specified accident environments. Thermal, electrical, mechanical, and radiological requirements cannot be treated independently.
A major United States design lineage uses the General Purpose Heat Source (GPHS). Each module contains four plutonium-dioxide fuel pellets in individual iridium-alloy clads, protected by graphite structures. A module supplies roughly 250 watts of thermal power near the beginning of its service life. The graphite components provide mechanical protection and help manage severe heating during atmospheric reentry, while the cladding retains the fuel. Modular construction allows different generators to use different numbers of standardized heat-source units.
Surrounding insulation reduces unwanted heat leakage and directs a larger fraction of the thermal flow through the thermoelectric elements. High-temperature components must resist creep, diffusion, chemical interaction, sublimation, and changes in contact resistance over many years. Thermoelectric couples are connected to produce a useful combination of voltage and current, while their supports accommodate differential thermal expansion. Even small dimensional changes become important when they accumulate across numerous joints and repeated thermal cycles.
The outer structure rejects most of the fuel’s heat. In vacuum, the emitted radiative power depends approximately on
where is surface emissivity, is the Stefan–Boltzmann constant, is radiating area, and and represent surface and effective environmental temperatures. Fin geometry and surface coatings must therefore be considered alongside spacecraft orientation, nearby structures, and exposure to sunlight.
Electrical integration typically requires power electronics to regulate voltage, distribute power, and manage changing loads. A DC–DC converter can match generator output to a spacecraft bus. Rechargeable batteries may supplement the RTG during short periods of high demand, such as operating a transmitter, moving a rover, or activating a scientific instrument. The generator supplies average energy continuously; the battery helps meet peak power.
Waste heat may also be useful. Spacecraft can route some of it through fluid loops or conductive paths to keep instruments and electronics within operating limits. This must be distinguished from a radioisotope heater unit, which supplies heat directly without generating electricity. A spacecraft can carry heater units, an RTG, or both, depending on its thermal and electrical requirements.
Historical development and major design families
Practical RTGs emerged from the combination of radioisotope production and semiconductor thermoelectric research in the 1950s. In the United States, early development formed part of the Systems for Nuclear Auxiliary Power, or SNAP, program. SNAP included both radioisotope systems and reactors, so the program name alone does not identify a particular device as an RTG.
The first RTG launched into space was the SNAP-3B aboard Transit 4A on 29 June 1961. It produced approximately 2.7 watts of electrical power. This demonstrated that a small radioisotope heat source and thermoelectric converter could function on an orbiting spacecraft. Subsequent SNAP designs increased output and supported navigation satellites, planetary probes, and lunar instruments. The SNAP-19 family powered Nimbus III, the Pioneer 10 and 11 spacecraft, and the Viking Mars landers, with design changes for their different environments.
The Apollo program used SNAP-27 generators to power scientific equipment left on the Moon. Apollo 12, 14, 15, 16, and 17 deployed these generators with their lunar surface experiment packages. Each supplied roughly 70 watts electrically at deployment. Their use demonstrated a central advantage of radioisotope power: instruments could continue operating through the lunar night without depending on large energy-storage systems.
Later spacecraft adopted larger and more standardized designs. The Multi-Hundred Watt RTG used silicon–germanium thermoelectric elements and supplied about 160 watts electrically per unit near the start of service. Three such generators powered each Voyager spacecraft, launched in 1977. The GPHS-RTG subsequently combined modular heat sources with high-temperature thermoelectric conversion. Versions supplied roughly 285–300 watts initially and flew on Galileo, Ulysses, Cassini, and New Horizons. Differences in stated output can reflect fuel age, loading, and the point in mission preparation at which output was measured.
The Multi-Mission Radioisotope Thermoelectric Generator (MMRTG) was designed for operation both in space vacuum and in planetary atmospheres. It uses eight GPHS modules and produces about 110 watts electrically at the beginning of a mission from roughly 2 kilowatts of thermal power. Its mass is about 45 kilograms. Curiosity, launched in 2011, and Perseverance, launched in 2020, each carry one MMRTG.
These design families do not represent a simple progression toward maximum efficiency. Converter materials, fuel inventories, atmospheric compatibility, lifetime requirements, qualification history, and production continuity all influenced the choices. A well-characterized converter with extensive long-duration test data can be more suitable for a mission than a newer material with higher laboratory efficiency but uncertain behavior after years at elevated temperature.
Roles in planetary and deep-space exploration
RTGs are especially valuable where solar energy cannot provide a dependable power supply within acceptable limits of mass, area, and operating complexity. Solar irradiance decreases approximately with the inverse square of distance from the Sun. At Jupiter’s orbital distance, sunlight is only about one twenty-seventh as intense as near Earth; at Saturn, it is roughly one ninetieth. Spacecraft using solar cells can compensate with larger arrays and lower power consumption, but the engineering trade becomes increasingly demanding.
The Pioneer, Voyager, Galileo, Cassini, and New Horizons missions illustrate the range of radioisotope-powered exploration. RTGs enabled instruments, computers, heaters, and telecommunications equipment to operate during long cruises and encounters far from the Sun. New Horizons, launched in 2006, used a single GPHS-RTG during its 2015 Pluto encounter and subsequent exploration of the Kuiper belt. These missions required power availability that was largely independent of spacecraft orientation and local illumination.
On planetary surfaces, darkness and environmental variability can be as important as distance. Mars receives substantially more sunlight than the outer planets, but dust deposition, seasonal changes, atmospheric opacity, and night length affect solar generation. Curiosity and Perseverance use MMRTGs to maintain continuous background power and recharge batteries. This supports flexible scheduling, although rover activities remain constrained by the total energy available. An RTG does not provide unlimited mobility or allow every instrument to operate simultaneously.
Radioisotope power also supplies valuable heat. During cold nights, a rover must prevent equipment and fluids from falling below their allowable temperatures. Reusing generator heat reduces the electricity otherwise needed for resistance heaters. The balance is mission-specific: too much heat in a warm environment can be a problem, while too little during a cold period can threaten equipment survival.
Long-duration missions must accommodate steadily declining output. Voyager operations have progressively reduced electrical loads and retired instruments as available power diminished. The resulting mission strategy links power engineering directly to scientific priorities: instrument selection, communications schedules, thermal limits, and onboard redundancy all depend on the remaining wattage.
RTGs nevertheless are not the default choice for all distant missions. Juno, launched in 2011, demonstrated solar-powered operation at Jupiter using large arrays and a carefully designed power budget. Selection depends on trajectory, eclipse duration, radiation environment, required power, spacecraft geometry, and fuel availability. Radioisotope power is most compelling when its endurance and environmental independence outweigh the costs and constraints of carrying a nuclear heat source.
Terrestrial, marine, and specialized applications
Before the expansion of reliable solar equipment and modern batteries, RTGs provided power at sites where fuel delivery or maintenance was exceptionally difficult. Applications included navigation aids, weather stations, unattended monitoring systems, and subsea equipment. Their advantage was not high power output but the ability to supply modest continuous loads for years without combustion, sunlight, or routine human attendance.
The Soviet Union deployed substantial numbers of strontium-90-powered generators at remote lighthouses and navigation installations, particularly along Arctic and other sparsely populated coasts. At these sites, extended winter darkness and difficult transport made conventional power supplies costly to maintain. A terrestrial generator could accept greater shielding mass than a spacecraft, making strontium-90 a more practical option despite its radiation characteristics.
United States terrestrial programs also developed radioisotope generators for remote and marine service. The SNAP-7 series, for example, included strontium-90-powered systems for navigation and weather-related applications. Underwater use introduced requirements different from those of spacecraft: the enclosure had to resist hydrostatic pressure, corrosion, and long exposure to seawater. Water could provide an effective heat sink, but the generator’s cold-side temperature and thermal interfaces still required careful design.
Power demands strongly shaped these installations. A continuously operating radio beacon or sensor might require relatively little average power, while a flashing light or transmitter imposed intermittent peaks. As in spacecraft, electrical storage could smooth the difference between continuous generation and changing demand. High reliability also depended on components outside the RTG itself, including cables, seals, switching equipment, and the powered instrument.
Miniature plutonium-238 thermoelectric generators were also used historically in some implanted cardiac pacemakers. Introduced around the end of the 1960s and beginning of the 1970s, these systems sought very long service life from a small continuous source. The later development of reliable lithium batteries reduced their practical role. Such devices represent a specialized historical use and differ greatly in scale, packaging, and output from spacecraft generators.
Terrestrial RTGs create an enduring stewardship obligation. An installation may become obsolete while its fuel remains radioactive and thermally active. Following the dissolution of the Soviet Union, recovery and replacement programs addressed many remote generators that were no longer adequately maintained or secured. Modern solar-battery systems can replace RTGs in numerous surface applications, although the comparison depends on local darkness, temperature, access, and load. Retirement must account for the radioactive source separately from the ordinary electrical equipment.
Safety, accident behavior, and environmental protection
RTG safety is primarily a problem of radiation protection, material containment, and accident resistance. An RTG is not a reactor and does not operate through a self-sustaining fission chain reaction. Its heat output cannot be shut off electrically, however: radioactive decay continues regardless of whether a spacecraft is operating. Safety provisions must therefore remain effective during handling, launch delays, mission failure, and long-term storage.
For plutonium-238 systems, intact containment limits exposure and prevents dispersal. Alpha particles have short ranges and do not penetrate the outer layer of skin, but plutonium-containing particles can present a serious internal hazard if inhaled or otherwise taken into the body. External neutron and photon radiation must also be considered during ground operations. Handling arrangements consequently include controlled facilities, monitoring, appropriate shielding, and procedures for managing contamination.
Space heat-source designs employ multiple barriers, including ceramic fuel, metallic cladding, and graphite protective structures. Qualification programs investigate mechanical impact, vibration, elevated temperatures, fire, and atmospheric reentry conditions. These barriers are intended to reduce the probability and consequences of fuel release across defined accident environments; they are not a claim that every conceivable accident leaves every component undamaged.
Several historical events influenced this approach. In 1964, a launch failure involving a Transit satellite carrying a SNAP-9A generator led to atmospheric reentry and dispersal of its plutonium fuel, consistent with the dispersal-oriented design philosophy then used. Later United States space heat sources emphasized fuel containment through accident and reentry conditions. In 1968, the SNAP-19 generators aboard the failed Nimbus B-1 launch were recovered from the ocean with their fuel contained.
Apollo 13 carried a SNAP-27 fuel assembly intended for lunar use. After the mission was aborted in 1970, the lunar module reentered over the South Pacific, carrying the fuel assembly into a remote ocean region. The event illustrates that safety analysis must cover hardware that never reaches its planned operating location, including return trajectories and disposal environments.
Modern launch assessments examine accident probabilities, potential release mechanisms, atmospheric transport, exposure pathways, and possible environmental consequences. In the United States, these assessments involve agencies responsible for the mission, nuclear-material management, and launch authorization. International principles adopted by the United Nations in 1992 address the use of nuclear power sources in outer space, including safety assessment and notification responsibilities.
Terrestrial safety additionally depends on physical security and record keeping. A disused generator can remain hazardous after its electrical equipment fails or its markings deteriorate. Recovery, secure storage, and eventual disposition of the source are therefore integral parts of the installation’s life cycle, rather than optional activities after power generation ends.
Lifetime, fuel supply, and technological alternatives
The thermal output of an RTG decreases predictably as its fuel decays. For a single dominant isotope,
where is initial thermal power and is the half-life. Plutonium-238 therefore loses about 0.79 percent of its thermal output annually. After 30 years, approximately 79 percent of the initial decay heat remains. This predictable decline is one reason it is well suited to multidecade operation.
Electrical output declines for additional reasons. Thermoelectric materials and electrical contacts can degrade, insulation can change, and the hot-side temperature falls as the fuel cools. The resulting reduction in conversion efficiency means that electrical power can decline faster than thermal power alone would suggest. Lifetime predictions must therefore combine nuclear decay calculations with long-duration testing and models of material behavior. The date of fuel production and the interval before launch also affect beginning-of-mission output.
Plutonium-238 supply is a specialized industrial constraint. It is generally produced by irradiating neptunium-237 in a reactor, followed by chemical processing that separates the newly formed plutonium. Fuel fabrication then requires conversion into suitable oxide, pellet production, encapsulation, testing, and assembly. Plutonium-238 is not interchangeable with the plutonium isotopic mixtures used for other nuclear purposes. Production capacity depends on target fabrication, irradiation space, radiochemical facilities, skilled personnel, and sustained quality control.
Following the end of earlier domestic production, the United States relied on existing inventories and purchases from Russia before restarting production activities. New production at Oak Ridge National Laboratory was demonstrated in 2015, within a program involving several Department of Energy facilities. Expanding usable supply involves more than producing the isotope: chemical recovery, heat-source fabrication, and qualification must all meet mission requirements. Fuel availability can consequently influence which missions are selected and when they can launch.
One development path seeks better thermoelectric materials. Skutterudites and other engineered compounds have attracted interest because their electrical and thermal transport properties can improve conversion efficiency. Higher efficiency could reduce fuel demand for a given electrical output. Laboratory performance alone is insufficient, however; a flight converter must also withstand years of high temperature, radiation, launch loads, and contact-material interactions.
An alternative is a radioisotope-powered Stirling engine coupled to an electrical alternator. Such a system is a radioisotope generator but not an RTG, because conversion involves a cyclic heat engine rather than solely thermoelectric elements. Dynamic converters can offer substantially higher efficiency, with accompanying requirements for moving-part reliability, vibration control, and fault management. The United States Advanced Stirling Radioisotope Generator program was canceled in 2013 before a flight deployment, although research on dynamic conversion continued.
Other alternatives occupy different power ranges and environments. Solar arrays paired with energy storage avoid radioisotope-fuel constraints where illumination is adequate. Fission systems can supply much higher power but introduce reactor-specific control, shielding, and deployment requirements. Americium-based heat sources offer another route for long-lived radioisotope systems, particularly where plutonium-238 production is unavailable. Their lower specific thermal power changes the mass, shielding, and converter trade-offs rather than providing a direct substitute with identical performance.