Can we make oxygen on the Moon instead of bringing it from Earth?

Why make oxygen on the Moon?

Space missions carry almost everything they need from Earth. That works for short visits. It becomes a serious burden when people, landers and spacecraft need a steady supply of oxygen.

In-situ resource utilisation, usually shortened to ISRU, means making useful products from material found at the destination. On the Moon, that could mean taking oxygen from lunar soil or water ice. The oxygen could support breathing, but the larger demand is likely to be rocket oxidiser. A rocket needs much more oxygen than fuel by mass.

The idea sounds obvious: use local material and launch less from Earth. The difficulty is that the mine and factory must also be launched. They need excavators, reactors, power, tanks, controls and replacement parts. If that equipment is heavy or unreliable, bringing oxygen from Earth may still be the better choice.

There are two main ways to get lunar oxygen.

The first is to heat ordinary lunar soil and remove oxygen that is chemically locked inside its minerals. NASA and Sierra Space tested an automated reactor at about 1,800°C in a vacuum chamber. NASA reported repeated operation and rated the reactor at Technology Readiness Level 6.[1]

The second is to mine water ice near the lunar poles. Heat releases the water, and electrolysis can split it into oxygen and hydrogen. This gives two useful products, but only if a real ice deposit is rich enough and can be mined reliably.

The best method depends on what is actually present at the landing site. A mission that mainly needs oxygen may favour ordinary soil. A mission that also needs water and hydrogen may accept the harder polar mine.

How much soil must the mine process?

We modelled the soil needed to produce one tonne of oxygen. The answer changes sharply with the amount recovered from each kilogram of soil.

For ordinary soil, we tested oxygen yields of 10%, 20% and 30%. These are Lyon Industries assumptions used to show the sensitivity; they are not flight-test results. For polar ice, we used NASA test values of 4–6% water content and 43–56% water recovery.[2]

The middle examples give a useful comparison. At 20% oxygen yield, ordinary soil needs about 5,000 kg of material. At 5% water and 50% recovery, the ice method needs about 45,000 kg. That is nine times more digging for the same amount of oxygen.

This does not prove that ordinary soil is always better. Water and hydrogen have value of their own. It does show why a polar mine needs good measurements of the deposit before anyone designs a large plant.

Does the machine produce more mass than it consumes?

A small plant may be easy to land but produce little oxygen. A larger plant may make more oxygen but cost much more launch mass. Spare parts also count because they must arrive from Earth.

Annual oxygen needed to break even = plant mass and lifetime spares ÷ useful operating years.

Take the middle example. A 1,000 kg plant, used for five years with 100 kg of spares each year, must make about 353 kg of oxygen per year just to repay that delivered mass. It must make more than this before the mission receives a real mass saving.

This first calculation leaves out the power plant, storage tanks and transport equipment. A real mission must add them. That will raise the production target.

Start with a small oxygen plant and measure everything.

The first plant should use ordinary lunar soil unless prospecting finds a clearly better ice deposit. Its job is not to supply a base. Its job is to prove that the full process can work without constant help from Earth.

The plant should dig and weigh the soil, run the reactor, remove waste, measure the oxygen, and store it. It should also record its electricity use, stopped time, lost oxygen and replacement parts.

Lunar PlanetVac collected, moved and sorted lunar soil during the Blue Ghost mission in March 2025.[3] That is useful proof for one part of the job. The next test must connect soil handling to oxygen production and storage.

What could change the answer?

A rich, continuous and shallow ice deposit could make water mining much more attractive. The answer would also change if a mission needs large amounts of hydrogen or water, not only oxygen.

Ordinary-soil processing could lose its advantage if the hot reactor, seals or soil-handling equipment need frequent replacement. Oxygen lost during cooling, transfer or storage could also remove the expected saving.

We therefore need measured deposit data and a complete pilot run. Until then, lunar ISRU is promising engineering work, but it is not yet a proven way to reduce mission mass.