Operating on Mars
What the environment forces on the design, the system sized for a crew of six, and the habitat requirements it has to meet.
Conditions and what they decide
| Condition | Value | Design consequence |
|---|---|---|
| Liquid water | Not stable at surface pressure and temperature | Every wet stage runs inside the pressurised habitat |
| Gravity | 3.7 m/s², 0.37 g | Flow through a bed scales with g at a fixed head: 2.65× the filter area, or a taller head |
| Temperature | About 20 °C down to −153 °C | Heated, insulated bioreactor; the reference strain is grown at 37 °C |
| Perchlorate | 0.4–0.6 % at Phoenix; 0.5–1 % in soils generally; also at Gale Crater | Raw regolith is never filter medium; it passes Stage 1 first |
| Regolith | Basaltic | Basaltic volcanic scoria as the bench analogue |
| Dust | 0.1 mg/m³, 24-hour average, below 10 µm | Regolith handling sealed from cabin air and the water path |
| Water source | Shallow ground ice | The habitat mines ice; this system treats water after use |
| Planetary protection | COSPAR policy | Stage 1 is closed; nothing living is vented or discharged |
Sources for every row are on the references page.
Sized for six crew
Reproducible with rwl:
pipx install rwl
rwl size --crew 6 --base early --explainOr move the crew and base live in the sizing simulator.
| Quantity | Early base | Mature base |
|---|---|---|
| Hygiene greywater | 33.2 L/day | 61.0 L/day |
| Continuous flow | 1.38 L/h | 2.54 L/h |
| Filter area, Earth g | 0.0035–0.014 m² | 0.0064–0.025 m² |
| Filter area, Mars g | 0.009–0.037 m² | 0.017–0.067 m² |
| Bench-size columns, Mars g | 2–8 | 4–15 |
| Bed volume, Mars g | 4–15 L | 7–28 L |
| Trash | 8.3 kg/day | 8.3 kg/day |
| Char, dry basis | 0.82 kg/day | 0.82 kg/day |
| Retort heat | 5.0–7.6 MJ/day | 5.0–7.6 MJ/day |
Greywater and trash rates come from NASA's Life Support Baseline Values document; filter rates are those of slow sand filtration, 0.1–0.4 m/h; char assumes food and paper are 47.5 % of trash and slow pyrolysis leaves 25 % of its dry mass as char.
The filter is small. A six-person base needs two to eight columns the size of the bench build, in parallel. That argues for modules: identical sealed cartridges swapped when spent, instead of one large bed dug out by hand.
Gravity: area or head
Flow through a packed bed follows Darcy's law, and the bed's hydraulic conductivity is proportional to g. At 0.37 g the same column passes 37.7 % of its Earth flow. There are two ways back:
| Earth | Mars, same flow | |
|---|---|---|
| More area, same head | 1 column | 2.65 columns |
| Same area, more head | 0.20 m of water over the bed | 1.22 m of water over the bed |
With a free-draining outlet the driving gradient is (standing water + bed) ÷ bed, so the whole of it has to grow 2.65×, not just the water on top. A 1.2 m water column per cartridge is impractical inside a habitat; parallel cartridges are not. That is why the design adds area. The bench column's own conductivity, and the same bed at 0.37 g, come out of a timed run with rwl permeability.
Retort heat
| Six crew, per day | Amount | Basis |
|---|---|---|
| Feed, food and paper | 3.96 kg | 47.5 % of 8.34 kg trash |
| Moisture in the feed | 0.69 kg | Plate waste 70 %, paper 10.2 %, unopened food 0.2 % |
| Pyrolysing the dry feed | 2.6–5.2 MJ | 0.8–1.6 MJ/kg near 500 °C |
| Water to steam at 500 °C | 2.3 MJ | 3.40 MJ/kg from 20 °C at 1 atm |
| Total | 5.0–7.6 MJ | 57–88 W averaged over a day |
Moisture is from BVAD Table 4-28, pyrolysis energy from Daugaard and Brown, steam from NIST. The pyrolysis figures are for wood and crop residues in a fast reactor, not mixed food waste in a slow retort; they set the order of magnitude, not a design value, and wall losses are not included. Either way it is tens of watts on average, run as a batch.
Oxygen is a side product
Full reduction, with acetate as electron donor:
ClO₄⁻ + CH₃COO⁻ + H⁺ → Cl⁻ + 2 CO₂ + 2 H₂O
| Per tonne of regolith | 0.4 % perchlorate | 1 % perchlorate |
|---|---|---|
| Perchlorate | 4.0 kg | 10.0 kg |
| Acetate, full reduction | 2.4 kg | 5.9 kg |
| O₂ if every molecule were captured | 1.3 kg | 3.2 kg |
Why the design does not count on it
The bacteria respire that oxygen themselves. Six crew breathe 5.37 kg of O₂ a day; even perfect capture would take 1.7–4.2 tonnes of regolith treated daily. Stage 1 exists to detoxify.
Chloride
Every perchlorate reduced leaves one chloride ion behind: 1.43–3.56 kg per tonne of regolith. Diluting that to the 250 mg/L taste limit for chloride would take 5.7–14.3 m³ of water per tonne. Dilution is not an option on Mars, so the rinse is kept small and the brine goes to a brine processor of the kind already demonstrated on the ISS for urine brine.
What Mars supplies, what Earth supplies
One bench-size cartridge, taking the RWL-001 media schedule and counting every mineral layer as treated regolith:
rwl cartridge --crew 6| Item | Amount | Comes from |
|---|---|---|
| Treated regolith | 1.73 kg | Mars, Stage 1 |
| Char | 0.16 kg | Mars, Stage 3; a day of trash makes enough for 5–6 cartridges |
| Acetate to detoxify the regolith | 4.1–10.3 g | Earth, until a waste-derived donor works |
| Cartridge shell | 1 | Earth, refilled on site |
| Chloride left to rinse | 2.5–6.2 g | Brine processor |
The acetate is the stoichiometric minimum; bacteria also spend some donor on growth. For scale, the ISS water processor relies on expendable multifiltration beds that need about 472 kg of resupply a year. This design does not replace those beds and does not claim to extend their life by any measured amount. Its own media, by mass, is made on Mars.
The stages at mission scale
Stage 1, bioreactor. Closed, heated and anaerobic, treating regolith in batches with Azospira suillum PS and Dechloromonas aromatica RCB. The donor is supplied as acetate or lactate until a waste-derived donor is shown to work. Treated regolith is rinsed of chloride.
Stage 2, filter modules. Sealed cartridges with the RWL-001 layering, in parallel under gravity. Effluent goes to the main water processor, never straight to the crew.
Stage 3, pyrolysis. A closed retort for food and paper. Char goes to Stage 2, off-gas is scrubbed, and retort heat warms the Stage 1 jacket.
Flows are drawn in RWL-003 and placement in RWL-004, on the drawings page.
Habitat requirements
NASA-STD-3001 Volume 2, Revision E.
| Requirement | Asks for | Response |
|---|---|---|
| V2 6026 Potable Water Quality | Safe potable water at the point of use | Not claimed: this is pre-treatment, potability is the main processor's job |
| V2 6046 Water Quality Monitoring | Monitor water and alert crew | Photographic turbidity as a low-cost trend check, not a certified sensor |
| V2 6051 Water Contamination Control | Keep water free of contamination, including dust | Regolith handling and water path physically separate |
| V2 6052 Particulate Matter | Total dust < 3 mg/m³, respirable < 1 mg/m³ | Closed regolith transfer; cartridges never opened in the cabin |
| V2 6153 Celestial Dust Monitoring | Monitor Martian dust and alert crew | Dust monitor at the handling glovebox |
| V2 6053 Lunar Dust Contamination | Lunar dust < 0.3 mg/m³ | Lunar only; the stricter 0.1 mg/m³ Martian limit is used |
| V2 6050 Atmosphere Contamination Limit | Gases below maximum allowable concentrations | Retort off-gas scrubbed and sent to life support |
| V2 6004 Carbon Dioxide Levels | ppCO₂ ≤ 3 mmHg, one-hour average | Bioreactor and retort CO₂ ducted to life support |
Open questions
- Whether the retort's rejected heat covers the Stage 1 jacket, which depends on a reactor not yet designed
- Whether the pyrolysis liquid can serve as electron donor
- The rinse volume that gets chloride out of the regolith pores, which is not measured
- Filtration with real Martian fines at 0.37 g, which is modelled, not measured