Europe's americium RTG bet: production over power density

What the work claims

This is a primary program-status paper, not a single experiment: the team leading ESA's radioisotope power effort reports that americium-241 radioisotope systems have reached the point where "the program has evolved to a flight program," with an RTG the paper states remains on track to be flight ready by 2028, RHU flight delivery targeted for 2027, and an initial Stirling-engine radioisotope study (done with NASA Glenn) demonstrating 1.9 W/kg specific power in a feasibility study.¹ Everything in it is agency-program and university-bench work; nothing described has flown.

How it works

The physics choice is the isotope. Plutonium-238, used by virtually all flown radioisotope systems, has a half-life of 88 years; americium-241 has 432 years. A longer half-life means lower decay power per kilogram of isotope, so an Am-241 system is intrinsically lower in specific power than a Pu-238 system of the same mass. What Am-241 buys instead is availability: Europe has no Pu-238 production, but the UK's National Nuclear Laboratory extracts Am-241 from reprocessed and stored civil plutonium fuel, and the UK Space Agency announced in late 2020 the PuMA2 facility to scale production from lab scale (PuMA1) to 500 g per year or more, with PuMA2 stated to take about two years to become operational.¹

The generator architecture is deliberately conservative. The European Large Heat Source (ELHS) is a 200 W-thermal stack powering a modular bismuth-telluride thermoelectric generator: a 10 W-electric building block designed to scale to 50 W-electric, with an electrical specific power of 1 W/kg at generator level (measurement boundary: electrical output at the generator, against the full generator mass including the heat source).¹ A separate radioisotope heater unit (RHU) delivers 3 W-thermal at 15 W-thermal per kilogram (thermal boundary, not electrical).¹ The thermoelectric converters run between 220 °C and a 20 °C cold side, with core fuel temperature not expected to exceed 415 °C.¹ The paper distinguishes demonstrated from predicted: earlier RTG models targeted 10 W-electric and measured 9.3 W-electric; the current distributed-fuel design is predicted at 10 to 13 W-electric, with initial thermoelectric campaigns showing at least 10 W-electric measured at full generator level.¹

The safety engineering is where most of the hardware maturity sits. Four generations of RHU and ELHS heat sources have been produced and, in the paper's words, fully evaluated experimentally: vibration tests, vacuum and atmospheric operation, plasma re-entry heating of RHU aeroshells, and post-re-entry ground-impact testing (clads survived impact testing at University of Dayton Research Institute facilities, with recession meeting a requirement of less than 50 percent of aeroshell thickness).¹ The fuel form is an Am-241 plus uranium oxide solid solution in a consolidated pellet geometry, held in vented platinum-20-percent-rhodium cladding, inside a 3D carbon-carbon aeroshell stated to be mechanically stable above 3000 K; the ELHS is designed to tumble during re-entry to spread ablation and mechanical stress.¹

The strongest case for it

The steelman is that this is the only credible non-US path to sovereign radioisotope power, and it is engineered against the right bottleneck. The paper's schedule logic is coherent: systems production at scale by 2025, fuel production at scale online by 2026, launch qualification of RHUs and the 200 W ELHS by 2026, flight RHU delivery by 2027.¹ Crucially, the ENDURE program (kicked off in 2023 across ESA directorates after the late-2022 Council of Ministers) includes the launch safety authorisation process, launch-site and launcher adaptations, led by ArianeGroup, with the stated objective of operational production and launch certification from European territory in the latter half of the 2020s.¹ A 1 W/kg RTG is unimpressive on a spec sheet, but a certified, producible, launch-authorized 1 W/kg RTG is a mission enabler: Rosalind Franklin (2028) is listed to carry an americium RHU on its lander module, with the rover launched from US territory.¹

Where a skeptic should push

The single most load-bearing assumption is the schedule, and every element of it is a program statement, not a demonstrated fact. The 2028 flight-ready RTG depends on PuMA2 actually reaching 500 g per year (the paper was written while PuMA2 was still becoming operational), on launch qualification completing by 2026, and on a launch safety authorization process that has never yet been exercised for a European radioisotope launch. The half-life physics also cuts hard the other way: Am-241's 432-year half-life is exactly why the electrical specific power is 1 W/kg, roughly half of the 1.9 W/kg the program's own Stirling feasibility study (with NASA Glenn, a 200 W system) showed was possible.¹ Thermoelectrics trade half the specific power for zero moving parts; Stirling recovers it at the cost of a mechanism that must run for decades in vacuum. The paper also concedes the mission pull is still soft: "near term focus on RHUs," with Stirling and in-situ resource use cases in feasibility-study territory.¹ And the concrete flight anchor, Rosalind Franklin, has itself moved on the calendar before; a 2028 launch with RHU delivery in 2027 leaves almost no schedule slip.

What americium production changes for deep space power

The non-obvious implication is that for radioisotope power, the scarcest resource is not the isotope's energy content but the certified supply chain around it. Europe accepted a large penalty in decay power per kilogram of isotope (decay power scales inversely with half-life, so trading 88-year Pu-238 for 432-year Am-241 gives roughly five times less heat per kilogram) in exchange for an isotope it can extract from material already in storage, and then spent a decade on the part that actually gates flight: re-entry aeroshells that tumble and survive impact, vented cladding, and a launch safety authorization process treated as a first-class work item.¹ For anyone planning outer-planet or lunar-night missions, that reframes the procurement question from "what is the best W/kg on the market" to "whose safety case and production line will still exist in 2040."

The genuine threat is specific-power stagnation: if the thermoelectric path stays at 1 W/kg, Am-241 systems are confined to roles where watts per kilogram barely matter, namely heat (the 15 W-thermal per kg RHU line, where americium's abundance is a genuine advantage) and long-duration low-power science. The opportunity is the Stirling branch: the program's own study showed 1.9 W/kg from a 200 W-electric Am-241 Stirling system, and 241Am heat sources are stated in the paper as supporting Stirling conversion, lunar in-situ resource use, and life-support heat.¹ If Stirling survives life testing, the same production line that today justifies itself on heater units becomes a mid-power generator line, which is precisely the class of asset a permanent lunar surface presence needs to survive the 14.5-day night.

The bottom line

Established by the paper: bench hardware maturity of Am-241 heat sources through four evaluated generations, a full-generator thermoelectric measurement of at least 10 W-electric, and a coherent production-and-safety program structure (ENDURE) with named milestones. Claimed but not yet demonstrated: the 2027 to 2028 flight timeline, 500 g per year fuel production, and launch certification from European territory. Confidence: high that Europe will field Am-241 RHUs on a near-term mission (that is the program's stated near-term focus and the lowest-risk product); moderate that the thermoelectric RTG flies by 2028 as scheduled; low-to-moderate that Stirling radioisotope systems reach flight this decade. What would confirm the claim: a launched RHU and a published launch safety authorization; what would break it: PuMA2 schedule slip or a Rosalind Franklin delay past the 2027 RHU delivery date.