Subsystem 11 of 12 · Food
Lunar Habitat ACT, the Autonomous Construction Testbed·Food·Concept definition, revision A·August 2026·See it in the tour

Why the food loop is deliberately left open

Growing staple calories on the Moon costs more in power and volume than shipping them. So the bay grows what shipping does worst, and the rest is flown up. That is a decision, not a shortfall.

about 1%of calories grown
4.1 kgharvest per week
14 kWlighting load
Contents
  1. The calculation everybody skips
  2. What the bay actually delivers
  3. Why the bay glows magenta
  4. The bay is also part of the water and air loops
  5. What can go wrong in a closed greenhouse
  6. Crew time, which is the real cost
  7. What is not yet known
  8. Increment plan
  9. Sources and further reading
Section 01

The calculation everybody skips

Bioregenerative life support is usually argued for on closure. The argument that matters is cost per calorie delivered, and it goes the other way for staples.

A kilogram of wheat grain holds about 3,300 kcal. In NASA's Biomass Production Chamber at Kennedy Space Center, wheat under 510 to 930 micromoles of light per square metre per second made 23 to 40 grams of dry biomass per square metre a day and took 77 to 86 days to mature, and only part of that biomass is grain. NASA's own summary of that work is that at about 40 moles of light a day, roughly 50 square metres of crops feed one person, against 20 to 25 square metres to supply their oxygen. The bay is 22 square metres. Shipping the same kilogram costs one kilogram of landed mass.

The chamber figures make the point: the staples took two to five times the daily light of lettuce and months of growth, and then milling and cooking, before a calorie appeared. At current launch and landing costs, and at the power available, shipping wins for staples by a wide margin. It loses badly for anything with a short shelf life or a vitamin content that degrades, which is exactly what the bay is sized to grow.

Cropkcal per kg, as eatenDry yield in NASA's chamberShips well?Decision
Wheat, ricestaple3,300 to 3,60023 to 40 g/m²/dYes, indefinitelyShip it
Soy, pulsesprotein3,500 to 4,50010 to 16 g/m²/dYesShip it
Potatostaple77022 to 33 g/m²/dYes, monthsShip it
Lettuce, leafy greensfresh150 to 1706 to 8 g/m²/dNo, daysGrow it
Tomato, pepperfresh180 to 31013 to 20 g/m²/dNo, weeksGrow it
Herbsflavournot countednot testedNoGrow it
Radish, microgreensfast cycle160not testedNoGrow it
Section 02

What the bay actually delivers

Twenty two square metres over six tiers, on a staggered 28 day cycle so something is ready every few days rather than everything at once. Output is budgeted at 4.1 kg a week of fresh mass. By the United States Department of Agriculture (USDA) values for these crops that is about 120 kcal a day across four crew, about one percent of their food by mass and by calories, and a much larger share of their fresh vitamins.

Weekly harvest by crop Leafy greens 1.55 kg per week lettuce, chard, kale Tomato and pepper 1.05 kg per week the morale crops Radish and microgreens 0.68 kg per week fastest cycle, 21 days Herbs 0.22 kg per week disproportionate effect on food acceptance Dwarf bean 0.6 kg per week protein, and it fixes nitrogen 4.1 kg fresh mass a week, about 120 kcal a day across four crew.
From USDA values for lettuce, tomato, pepper, radish, basil and green beans, the mix comes to about 120 kcal a day. The herb yield looks trivial at 220 grams a week and is not. Food acceptance falls measurably over a long rotation on a monotonous diet, and the ability to change how something tastes is one of the few countermeasures available.
Section 03

Why the bay glows magenta

Chlorophyll absorbs most strongly in the deep red around 660 nm and the blue around 450 nm, and least in the green, which is why leaves look green. But a leaf still absorbs most of the green that falls on it, and McCree's measurements on 22 crop species put the quantum yield of green photons close to that of red once absorbed. The case for red and blue diodes is electrical, not botanical.

A 660 nm red diode can convert electricity to photons at up to 5.5 micromoles per joule, more than any other colour, and a white diode loses part of its energy in the phosphor that makes it white. Narrowband diodes at 660 and 450 nm therefore deliver the most photons per watt at wavelengths the plant uses well. The bay looks magenta because that is the sum of deep red and blue.

660 nmdeep red, photosynthesis peak
450 nmblue, morphology
14 kWtotal lighting draw
18 hphotoperiod, design; Veggie runs 16
A small amount of green is added back anyway

A little green is added, not for the plants but for the crew. NASA's Veggie unit on the station runs one green diode for every twelve red and three blue, about six percent of its light, at about 200 micromoles per square metre per second on a 16 hour day, and the Advanced Plant Habitat carries green and white banks alongside red and blue at up to 1,000 micromoles. Under pure red and blue the plants look black and grey and inspection for disease or stress is impossible. A few percent of green costs almost nothing and makes the bay somewhere people are willing to spend time.

Section 04

The bay is also part of the water and air loops

Crew 4.2 kg/d CO2 Greenhouse 22 m² Condensers 2.1 kg/d H2O O2 to cabin 0.03 kg/d Nutrient loop 340 L CO2 transpiration photosynthesis
Plants transpire nearly all the water they take up: NASA's chamber measured about 2 litres per square metre a day for lettuce and 5 for wheat, so the bay cycles tens of litres a day through its condensers, most of it straight back to the nutrient loop. The 2.1 kg a day shown is the share credited to the cabin water loop, a design allocation. The oxygen figure is 0.03 kg a day, the same as the site's systems page: 4.1 kg of fresh lettuce a week is only about 30 grams of dry matter a day, and photosynthesis makes roughly a gram of oxygen for each gram of dry matter. That is about one percent of what four crew breathe, not a quarter.
Section 05

What can go wrong in a closed greenhouse

A closed agricultural system has failure modes an open field does not, and the Yuegong 1 run at Beihang University, 370 days from May 2017 with two crews of four rotating through, and Biosphere 2 are the evidence base for most of them.

FailureOnsetEffectMitigation
Pathogen introductionany crew transferDaysCan take a whole tierTier isolation, seed sterilisation, no soil
Nutrient imbalancegradual driftWeeksYield falls before it is visibleContinuous electrical conductivity (EC) and pH, weekly assay
Ethylene accumulationclosed volumeWeeksPremature senescence across every cropCatalytic oxidiser in the bay loop
Pollination failureno insectsImmediateNo fruit from tomato or pepperManual pollination, scheduled crew task
Root zone anoxiapump failureHoursDeep water culture is unforgivingDual pumps, dissolved oxygen alarm
Light degradationdiode ageingYearsSlow yield lossOutput measured, not assumed. Diodes replaced on measurement.
The lesson from Biosphere 2

Biosphere 2's oxygen fell from 21 percent to 14 percent over the first 16 months of closure, and the CO₂ that should have risen to match did not. Severinghaus and Broecker traced it with carbon isotopes: microbes were respiring the organic rich soils, and the CO₂ they produced was being absorbed by the structure's curing concrete, which broke the carbon accounting. The lesson is not about concrete. It is that a closed system will find an unmodelled sink, and that the only defence is measuring every flow rather than inferring any of them.

Section 06

Crew time, which is the real cost

The bay draws 14 kW and takes 1.5 crew hours a day. On a base where four people have twenty productive hours between them, 1.5 hours is seven and a half percent of the entire labour budget, which is a larger cost than the electricity.

It is spent anyway, and not only for the food. Analog and orbital crews consistently rate plant care among the most valued activities available to them, and the review by Odeh and Guy collects the evidence on mood and cognition in long isolation. The bay is carried as a habitability measure that happens to also produce lettuce.

Section 07

What is not yet known

The bay's contribution to oxygen and water is modelled, not measured; increment four instruments every flow so the next base's greenhouse sizing is defensible. The 4.1 kg a week is a budget, not a yield: NASA's chamber grew 6 to 8 grams of dry lettuce per square metre a day at 17 moles of light, and 22 square metres at that rate would give several times the budgeted fresh mass, so the crop mix and the crew time, not the area, set the figure. The bay's light level and photoperiod have not been fixed against the flight record, which is Veggie at about 200 micromoles for 16 hours and the Advanced Plant Habitat at up to 1,000. The 14 kW lighting load is an allocation. Whatever the mix, the calorie share stays near one percent, which is the point of section 01. A closed system will find an unmodelled sink, as Biosphere 2 did, and the only defence is measuring every flow rather than inferring any of them.

Section 08

Increment plan

Increment 1, no bay

All food is shipped. The volume is landed but fitted out as stowage, because a greenhouse with nobody to tend it is dead mass.

Increment 2, two tiers

Fast cycle crops only, radish and microgreens, proving the nutrient loop and the pathogen protocol before anything slower is planted.

Increment 3, six tiers at full yield

The staggered 28 day cycle running properly, and the condensate recovery plumbed into the water loop.

Increment 4, closure measurement

Every flow instrumented, so the contribution to oxygen and water is measured rather than modelled. This is what makes the next base's greenhouse sizing defensible.

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.

A case for supporting human long term survival on the Moon: the Lunar Palace 365 mission, Acta Astronautica 2024The 370 day closed bioregenerative life support system (BLSS) run at Beihang University
NASA's contributions to vertical farming, NASA technical memorandum 2020Biomass Production Chamber yields, light levels, water use and the area per person figures
Wheeler, agriculture for space: people and places paving the way, Open Agriculture 2017The history behind the chamber data
Microbiological and nutritional analysis of lettuce crops grown on the International Space Station (ISS), Frontiers in Plant Science 2020Veggie light level, spectrum ratio and photoperiod
Hardware validation of the Advanced Plant Habitat on ISS, Frontiers in Plant Science 2020The 1,000 micromole light level and the five diode banks
Kusuma, Pattison and Bugbee, from physics to fixtures to food: current and potential LED efficacy, Horticulture Research 2020The 5.5 micromole per joule limit for 660 nm red
Beyond red and blue: the action of green wavelengths on plant physiology, Environmental and Experimental Botany 2025The McCree quantum yield of green photons
Severinghaus and others, oxygen loss in Biosphere 2, Eos 1994The unmodelled sink lesson
USDA FoodData CentralCalorie values per kilogram for the crops in the table and the harvest chart
Odeh and Guy, gardening for therapeutic people plant interactions during long duration space missions, Open Agriculture 2017The evidence on mood and cognition in section 06