Subsystem 04 of 12 · Life support
Lunar Habitat ACT, the Autonomous Construction Testbed·Life support·Concept definition, revision A·August 2026·See it in the tour

Ninety eight percent of the water, and the oxygen gap that will not close

Water comes back almost completely. Oxygen does not, and the reason is stoichiometry rather than engineering, which is why the answer is a plant on the surface and not a better reactor.

98%water recovered
about 50%oxygen loop closed
1.65 kgdaily oxygen shortfall
Contents
  1. What four people need
  2. The air loop, step by step
  3. Why the oxygen loop stops at about half
  4. Water recovery
  5. Contaminants, and the thing that actually goes wrong
  6. Dual string, and what that actually means
  7. Mass and power
  8. What is not yet known
  9. Increment plan
  10. Sources and further reading
Section 01

What four people need

The crew figures below are the planning values NASA life support studies use for the International Space Station (ISS): 0.84 kg of oxygen a day per crew member and about 1 kg of carbon dioxide. NASA's Baseline Values and Assumptions Document, revision 2, gives 0.895 and 1.085 kg for an 82 kg crew member, about 7 percent more, and the in situ resource utilisation (ISRU) margin covers that difference. The water figures are anchored to the station's measured recovery. The cabin runs at NASA's exploration atmosphere, 56.5 kPa (8.2 psia) at 34 percent oxygen, chosen so that prebreathe before an extravehicular activity (EVA) is short.

ConsumablePer person per dayFour crewRecoveredResidual
Oxygenmetabolic0.84 kg3.36 kg1.71 kg1.65 kg
Potable waterdrink and food prep2.0 kg8.00 kg7.84 kg0.16 kg
Hygiene waterwash and flush1.6 kg6.40 kg6.27 kg0.13 kg
Food, dry massshipped and grown0.68 kg2.72 kg0.03 kg2.69 kg
Nitrogenleak makeup0.04 kg0.16 kg0.00 kg0.16 kg
Total5.16 kg20.64 kg15.85 kg4.79 kg

Of the 4.79 kg residual, 1.65 kg of oxygen comes from the ISRU plant and 3.14 kg a day comes from Earth. The oxygen recovered is 1.68 kg from Sabatier and 0.03 kg from the greenhouse; the food recovered is the greenhouse's dry matter. Carbon dioxide production runs at about 1.05 kg per person per day, so 4.2 kg for the crew. That is the input to the whole air revitalisation chain and the number the sieve beds are sized against.

Section 02

The air loop, step by step

Cabin air 56.5 kPa, 34% O2 4 bed sieve CO2 capture Sabatier CO2 + 4H2 Electrolyser H2O split Methane vent CH4 overboard O2 to cabin 1.68 kg/d ISRU makeup 1.65 kg/d 4.2 kg/d CO2 CH4 H2 top up
The methane vent is where the oxygen goes. Sabatier turns carbon dioxide and hydrogen into methane and water; the water is split back into oxygen and hydrogen, but the methane leaves carrying four hydrogen atoms with it, and those hydrogen atoms had to come from somewhere. The 1.68 and 1.65 kg a day figures are the station's 50 percent closure applied to four crew, and the same balance the systems page carries.
Section 03

Why the oxygen loop stops at about half

This is the single most misunderstood number in closed loop life support, so it is worth being precise about.

Sabatier runs CO₂ + 4H₂ → CH₄ + 2H₂O. The two water molecules are electrolysed back to 2H₂ + O₂. Every molecule of oxygen the electrolyser makes for the crew comes with two molecules of hydrogen. Reducing one carbon dioxide takes four, so the hydrogen on hand reaches only half the carbon dioxide the crew produce, and the methane carries two of every four away for good. The loop therefore recovers about half of the oxygen, which is where the station's system stands today. NASA states the exploration requirement as a minimum of 75 percent.

There is no reactor design that fixes this. It is the stoichiometry of the reaction. The options are to supply hydrogen from outside the loop, to crack the methane and recover its hydrogen, or to make oxygen from something other than cabin air.

OptionClosureCostVerdict
Sabatier aloneflight proven, ISSabout 50%BaselineThe gap has to be filled somehow
Ship hydrogenfrom Earthabout 90%0.34 kg/day landedCheap in mass, but a permanent dependency. The crew make less carbon dioxide than the oxygen they take, so it never reaches 100
Methane pyrolysiscrack CH4 to carbon and H275%, NASA's minimumPlasma pyrolysis assembly and hydrogen separationThe path NASA is on, not yet flown
ISRU oxygenin situ resource utilisation, from regolith100% independent1 t plant, 14 kWThe architecture chosen here
Why ISRU rather than shipping hydrogen

Shipping hydrogen closes most of the loop for about 125 kg a year, which is trivial: reducing all 4.2 kg of carbon dioxide a day takes 0.76 kg of hydrogen, and electrolysis for the crew's oxygen gives back 0.42. The reason this base does not do that is that it is a construction testbed: the point is to demonstrate that the outpost can run on what is under it. A base that ships hydrogen is a very efficient camp. A base that makes its own oxygen is the beginning of an industry.

Section 04

Water recovery

Water is the success story. The International Space Station reached 98 percent recovery in 2023 with the brine processor assembly, installed in March 2021 as a technology demonstration; before it the station recovered 93 to 94 percent. That is the figure this design assumes. It is achieved in two stages: a distillation assembly that handles urine at up to 87 percent recovery, and a processor that recovers what the still leaves behind as brine.

Water recovery by stream Cabin condensate 5.9 kg/day essentially complete Urine distillate 4.1 kg/day 87 percent from the still Brine processor 0.6 kg/day the last 11 percent of urine Greenhouse transpiration 2.1 kg/day the bay is also a condenser Hygiene grey water 1.5 kg/day Sabatier product water 0.9 kg/day from the CO2 loop 15.1 kg recovered against 14.4 kg demand. The surplus buffers the reservoir.
The greenhouse contributes more than people expect. Plants transpire nearly all the water they take up, and the bay's condensers recover it clean, which is why the food and water loops cannot be sized independently of one another.
Section 05

Contaminants, and the thing that actually goes wrong

Closed volumes accumulate trace contaminants: ammonia from the crew, siloxanes from materials outgassing, volatile organics from the galley and the greenhouse, and formaldehyde from adhesives. On the station these are removed by a trace contaminant control system that runs continuously and is one of the most maintenance hungry items aboard.

The lunar case is worse in one respect and better in another. Worse, because the greenhouse adds a biological source of volatiles that a station does not have. Better, because the cabin can be partially vented and repressurised from ISRU oxygen without a resupply penalty, which is an option no orbital station has.

JSC 20584sets the limits, Johnson Space Center, revision C, June 2024
6exposure periods, 1 hour to 1,000 days
180 dthe exposure period a rotation is held to
not yet countedgreenhouse specific volatiles
Section 06

Dual string, and what that actually means

Every function in this chain exists twice, but not identically. Identical redundancy protects against random failure and does nothing at all against a design fault or a consumable that runs out, because both strings fail the same way at the same time.

FunctionPrimaryBackupBackup is different how
CO₂ removalcontinuous4 bed molecular sieveLithium hydroxide canistersConsumable, not regenerative. Immune to a sieve design fault.
Oxygen supplycontinuousSabatier and electrolysisHigh pressure O₂ storageStored gas, 96 hours at full crew
Water recoverycontinuousDistillation and brine340 L reservoirBuffer, not a process
Atmosphere pressurecontinuousN₂/O₂ makeup from storePartial vent and repressuriseUses ISRU oxygen, no resupply
Trace contaminantscontinuousCharcoal and catalytic oxidiserVent and repressuriseCrude, effective, always available
The 96 hour rule

Every consumable buffer on this base is sized to at least 96 hours, because that is the time to get a crew off the surface and home if something cannot be fixed. Ninety six hours is not a comfort margin, it is the length of the rescue.

Section 07

Mass and power

ElementMassPowerNote
Four bed molecular sievedual string412 kg1.24 kWRegenerated under vacuum, which is free here
Sabatier reactorsingle, with spare core118 kg0.42 kWRuns hot, contributes to the thermal load
Oxygen generation, electrolysisdual296 kg1.86 kWThe largest single electrical load in the rack
Water processor and distillerdual484 kg0.92 kWIncludes the brine processor
Trace contaminant controldual96 kg0.18 kW
Storage, tanks and plumbing530 kgnone340 L water, 96 h oxygen, nitrogen
Lithium hydroxide, 96 hconsumable78 kgnoneReplaced each resupply
Total2,014 kg4.62 kW$2.0 bn landed
Section 08

What is not yet known

Methane pyrolysis, which NASA needs to reach its 75 percent minimum, is not yet flight proven and is carried as a demonstration rather than a dependency. The crew figures are the station planning values; NASA's baseline document gives 0.895 kg of oxygen and 1.085 kg of carbon dioxide per crew member per day for an 82 kg crew member, about 7 percent more, and the table should be rerun at those values even though the ISRU margin covers them. The contaminant list for a habitat with a greenhouse has not been drawn up; JSC 20584 sets the limits but says nothing about which plant volatiles will appear or how often the charcoal beds will need changing. The oxygen gap is closed by the ISRU plant, whose maturity is the open question of the ISRU document.

Section 09

Increment plan

Increment 1, open loop with buffers

Sieve, oxygen store and a water processor, with consumables sized for the first rotation. Closure is around 60 percent and the resupply bill is high, which is acceptable for a first crew.

Increment 2, full closure and ISRU tie in

Sabatier, the brine processor and the ISRU oxygen feed. This is the step that takes water to 98 percent and closes the oxygen gap without shipping hydrogen.

Increment 3, greenhouse integration

The bay comes online as a real contributor rather than a science experiment: transpiration recovery into the water loop and a measurable oxygen contribution.

Increment 4, methane pyrolysis trial

Cracking the Sabatier methane to recover its hydrogen would take the loop toward NASA's 75 percent minimum. It is carried as a technology demonstration rather than a dependency.

Sources

Sources and further reading

Every figure in this document traces to one of these. Agency documents and peer reviewed work first, reporting only where it is the primary record.

NASA, Spacecraft Oxygen Recovery (SCOR) project nasa.govOxygen recovery from carbon dioxide presently about 50 percent with Sabatier; exploration goal above 75 percent
NASA, Reliable and efficient electrochemical recovery of O2 from metabolic CO2 at the ISS, 2024 https://ntrs.nasa.gov/citations/20240002161States the station loop is limited to 50 percent and the exploration minimum is 75 percent
NASA, NASA achieves water recovery milestone on International Space Station, June 2023 nasa.govThe 98 percent figure, and the 93 to 94 percent before the brine processor
Status of ISS water management and recovery, ICES 2023 https://ntrs.nasa.gov/citations/20230006217Urine processor at up to 87 percent recovery; brine processor installed March 2021 as a demonstration
NASA, Marshall Space Flight Center ECLSS fact sheet, 2025 nasa.govSabatier fed by the oxygen generation assembly hydrogen; 98 percent of sweat and urine reclaimed
Jones, Moon base life support design depends on launch cost, crew size and mission duration, ICES 2019 https://ntrs.nasa.gov/citations/20190027610The 0.84 kg per crew member per day oxygen planning figure
NASA, Life support baseline values and assumptions document, revision 2, 2022 https://ntrs.nasa.gov/citations/20210024855Table 3 31: 0.895 kg oxygen and 1.085 kg carbon dioxide per crew member per day for an 82 kg crew member
NASA, Spacecraft maximum allowable concentrations for airborne contaminants, JSC 20584 revision C, June 2024 https://ntrs.nasa.gov/citations/20240007713The contaminant limits, over six exposure periods from 1 hour to 1,000 days
NASA, Recommendations for exploration spacecraft internal atmospheres, TP 2010 216134 nasa.govThe working group's recommendation of 8.0 psia and 32 percent oxygen for surface habitats with a 60 minute in suit prebreathe; NASA's reference surface habitat now carries 8.2 psia and 34 percent, the 56.5 kPa figure used here
Samplatsky et al., Development and integration of the flight Sabatier assembly on the ISS, 41st International Conference on Environmental Systems, AIAA 2011 5151, 2011 doi.orgFlight heritage for the carbon dioxide reduction stage, on the station since October 2010
Burke et al., Internal layout of a lunar surface habitat, ASCEND 2022, NASA NTRS 20220013669 https://ntrs.nasa.gov/citations/20220013669The reference surface habitat runs at 8.2 psi and 34 percent oxygen, or 10.2 psi and 26.5 percent, and generates high pressure oxygen for its rover
Reliability and lifetime estimation of bioregenerative life support system based on the 370 day closed human experiment of Lunar Palace 1 and Monte Carlo simulation, Acta Astronautica, 2022 https://www.sciencedirect.com/science/article/abs/pii/S0094576522006294The greenhouse contribution to closure, and the failure record of a closed loop that actually ran