ISRU plant

Mission role and resource products

An ISRU plant encompasses the full chain of operations from excavation through beneficiation, processing, storage, and product delivery. For a lunar or Martian base, the core outputs fall into three categories. First, oxygen: used as breathing gas in habitats and extravehicular activity suits, as an oxidizer for rocket engines, and as feedstock for fuel cells. Second, water: for drinking, hygiene, agriculture, and as the primary feedstock for electrolysis to produce O₂ and H₂. Third, structural metals and construction materials: iron, titanium, aluminum, and other elements freed from reduced regolith, with potential use in additive manufacturing and habitat construction.

These outputs enable closed-loop or partially closed-loop surface mission architectures. Early human outposts on the Moon or Mars would still receive some supplies from Earth, but a functioning ISRU plant reduces that dependence progressively — with the long-term goal of providing virtually all propellant and life-support consumables from local sources.

Water extraction from regolith

Water on the Moon is concentrated in permanently shadowed regions (PSRs) at the poles, mainly as icy regolith at very low temperatures below 110 K. It also exists as chemically bound water or hydroxyl in minerals and glasses, and as adsorbed volatiles — water, ammonia, methane, and others — in cold traps. On Mars, water is found in subsurface ice and hydrated regolith.

Extraction typically proceeds in four steps. Robotic or human-tended excavators collect icy or hydrated regolith and deliver it to a processing reactor. The regolith is then heated — by solar thermal, resistive, or microwave means — to release water vapor and other volatiles; this approach is often called thermal mining. Released vapors are captured, cooled, and condensed into liquid water, which is then purified for potable use or electrolysis. Finally, electrolysis splits water into O₂ and H₂: oxygen is stored as gas or cryogenic liquid, while hydrogen is used as fuel or recycled to regolith-reduction reactors.

Several specialized hardware concepts have been studied. A solar-thermal mining approach redirects sunlight collected at crater rims via mirrors into PSRs to sublimate ices; vapor is collected inside a tent structure and condensed in cold traps. Hollow auger drill systems excavate icy regolith and deliver it to a dryer where heating releases water — the Lunar Auger Dryer ISRU (LADI) concept follows this approach. Laboratory work has also demonstrated microwave heating of lunar regolith simulant containing ice, validating a non-contact heating method suited to deep or cold environments.

An additional volatile source is the residual propellant remaining in descent stages after lunar landing. Proposed ISRU systems would convert leftover propellants into water, then combine in-situ oxygen with excess hydrogen to produce additional water — increasing overall resource recovery efficiency at early bases.

Oxygen extraction from lunar regolith

Lunar regolith is rich in oxygen chemically bound in silicates and oxides — for example, FeO in iron-bearing minerals. Three main electrochemical and thermochemical routes have been studied for extracting this oxygen.

Hydrogen reduction reacts iron oxides in regolith with H₂ gas at approximately 900–1,000 °C, producing iron metal and water vapor. The water is condensed, then electrolyzed to recover O₂ and regenerate H₂ for recycling. This closed loop also yields metallic iron as a by-product potentially useful for construction. ESA's PROSPECT payload team demonstrated hydrogen reduction on Apollo 11 and Apollo 16 soil samples (10084 and 60500), achieving measurable water production. NASA's RESOLVE/RPM payload (developed with the Canadian Space Agency) incorporates a hydrogen reduction demonstration in which regolith is heated from roughly 150 °C to 900 °C while H₂ flows through to extract oxygen as water, which is then captured and quantified. ESA conceptual designs project large-scale production on the order of 1,000 kg of O₂ per year.

Carbothermal reduction reacts regolith with methane or other carbon-bearing gases at high temperature to free oxygen from silicates and oxides. Downstream processing separates CO and CO₂ products and converts them to O₂ — for example via solid-oxide electrolysis — while carbon-bearing gases are recycled. NASA's ISRU Pilot Plant concept employs this approach. Hybrid architectures have been proposed that combine carbothermal reduction of dry regolith at peaks of eternal light with water extraction from icy regolith in nearby PSRs, either in parallel or series configurations.

Molten regolith electrolysis (MRE) directly melts regolith and applies electric current; metals are reduced at the cathode and oxygen evolves at the anode. It requires no additives and is considered the most direct method to produce both oxygen and metals on the Moon, though it demands very high operating temperatures and robust materials. Molten salt electrolysis (MSE) dissolves or suspends regolith in a molten salt, potentially achieving comparable outputs at somewhat lower temperatures. Both electrochemical approaches simultaneously yield oxygen and metallic elements useful for construction.

Mars ISRU: MOXIE and atmospheric oxygen

The Mars Oxygen In-Situ Resource Utilization Experiment (MOXIE), carried aboard NASA's Perseverance rover as part of the Mars 2020 mission, is the most concrete in-space ISRU demonstration conducted to date. MOXIE uses solid-oxide electrolysis to convert carbon dioxide from the Martian atmosphere — which is approximately 95% CO₂ — into oxygen. Its design requirements were to produce at least 6 g/hr of oxygen at 98% purity or better, and to complete at least 10 operating cycles on Mars meeting those targets.

Between landing in February 2021 and the conclusion of its mission on 30 September 2024, MOXIE completed 16 oxygen-production runs. Total oxygen produced across all runs was 122 g. Peak performance reached 12 g/hr at 98% purity or better — twice NASA's original goal. MOXIE's 16th and final run produced 9.8 g of oxygen. The instrument satisfied all success criteria and experienced no failures.

A detailed post-mission summary noted that MOXIE's design translated from Earth laboratory to Mars with no degradation in performance. The instrument demonstrated robustness against thermal cycling, Martian dust, and seasonal changes in atmospheric density and temperature. Dust was determined to be of little concern for a full-scale ISRU oxygen plant — an important design insight for future systems. Impurity levels in the produced oxygen were described as unmeasurably low, far exceeding the 98% purity requirement.

MOXIE is explicitly a technology demonstrator, not a plant: it has no integrated fuel synthesis, liquefaction, storage, or large-scale production capacity. Future Mars ISRU plants would combine atmospheric CO₂ processing with water extraction from subsurface ice or hydrated regolith, and would synthesize methane and oxygen as ascent vehicle propellant in addition to providing breathable air for crews. The MOXIE results provide benchmarks — production rates, purity figures, and operational lessons — for scaling to larger systems.

Regolith-based 3D printing and construction

ISRU plants are not limited to chemical processing; they also encompass additive manufacturing and construction from local regolith, reducing or eliminating the need to launch structural materials from Earth. As of 2024, no true regolith-based habitat has been constructed on the Moon or Mars — all demonstrated elements are Earth-based analogs or small-scale laboratory components.

NASA has supported Texas-based construction company ICON to develop a large-scale 3D printing platform called Olympus, designed to print landing pads, roads, and eventually pressurized habitats using native lunar or Martian soil. Olympus employs a laser-based process called Laser Vitreous Multi-material Transformation, in which a high-powered laser melts regolith, which then cools into a ceramic-like solid. In February 2025, ICON flew the Duneflow experiment on a Blue Origin reusable rocket under NASA's Flight Opportunities program to test how simulant behaves in lunar gravity and to advance the laser-based processing system. Separately, NASA's MMPACT project targets autonomous construction capabilities for lunar bases, with early projections indicating that a 30-metre-diameter lunar landing pad would require several thousand kilograms of lunar regolith.

ESA and architecture firm Foster + Partners have produced structural elements from lunar regolith simulant, including a 1,500 kg building block. Their favored hybrid approach places a conventional habitat module on the surface and surrounds it with a 3D-printed layer of compacted or sintered regolith for radiation and micrometeoroid protection. The EU Horizon 2020 RegoLight project demonstrated solar-driven sintering — fabricating bricks layer-by-layer using concentrated sunlight, without imported binders. The German MOONRISE project mounted a customized laser on a prototype lunar rover to melt regolith into spherical shapes, pointing toward mobile robotic systems capable of driving across the surface and sintering structural features in place.

The Redwire Regolith Print study aimed to demonstrate in-space manufacturing on the International Space Station using a fused-deposition-like process with feedstock of metal oxides and binder derived from lunar regolith simulant, with printed test plates returned to Earth for materials analysis. A 2024 review of in-situ additive manufacturing with lunar regolith concluded that regolith-based geopolymers and sintered structures are promising but face challenges in binder availability, achieving adequate mechanical properties, and scaling to habitat-level structures in the lunar environment.

Technology readiness status (2024)

Technology readiness levels (TRL) run from 1 (basic concept) to 9 (mission-proven). TRL 6 denotes a system or subsystem prototype demonstrated in a relevant environment; TRL 8 requires the actual system completed and qualified in its final form under expected conditions; TRL 9 means the system has proven through successful mission operations.

As of 2024, integrated ISRU plants for lunar or Martian resource extraction and processing are generally at TRL 4–6. Bench-scale reactors for processes such as molten regolith electrolysis, carbothermal reduction, and hydrogen reduction have been demonstrated with lunar regolith simulants in laboratory vacuum and thermal conditions, corresponding to TRL 4–5. Prototype testbeds that integrate regolith feed systems, reactors, and basic product handling in ground facilities using simulant and vacuum chambers reach TRL 5–6. No complete end-to-end plant — covering excavation, beneficiation, processing, storage, and distribution — has been operated on the Moon, and none has been flight-qualified to TRL 8.

Selected subsystems stand at higher readiness. Resource-characterization instruments such as spectrometers, radars, and cameras that have flown on lunar missions can reach TRL 8–9 individually. For Mars, the MOXIE solid-oxide electrolysis unit achieved TRL 7–8 by operating successfully in the actual Martian environment. However, MOXIE is a single processing subsystem; a complete Mars ISRU plant integrating Sabatier reactors, electrolysis, liquefaction, and large-scale storage remains at approximately TRL 3–5 as an integrated system, since no full chain has flown and operated on Mars.