Lunar Oxygen in Cryogenic Propellant Depots

Published June 14, 2026

Forget the brochure dome or the flag. The key lunar asset is a cryogenic oxygen transfer line at EML-1 (Earth-Moon Lagrange Point 1), where spacecraft stage cargo and propellant. Fill that line from Earth, and you pay full gravity tax on every kilogram. Fill it from the Moon, and the climb to the same spot is roughly one-fifth as hard. Same oxygen, same depot, fraction of the bill.

The High Ground, Chapter 6: Propellant Architecture

A tank of liquid oxygen on the Moon has to be held at 90 kelvin, ninety degrees above absolute zero. The liquid hydrogen stored beside it has to be held at 20. Those two liquids are the main way rockets store propellant, and a tank of either one left exposed on the lunar surface boils away. The gas that escapes is propellant someone paid to make or haul, lost before any engine burned it.

The High Ground's chapter on propellant does not treat oxygen as a load poured into one rocket for one mission. It treats a propellant depot as a utility on a network, a storage business, and a service whose job is to never run dry. Its claim is that keeping propellant stored and moving between vehicles is worth as much as the cost of making it, and that a store of propellant can earn money on its own.

Keeping the tank cold is the first problem. The fix under study is to bury the tanks. In simulation, a buried tank's boil-off rate fell below 0.05 percent per lunar month. The same research then tries to turn that rate into a yearly loss for a store of 1,000 tonnes, and two passages in it give different answers. The research has not settled which one is right.

The other way to hold the cold is to run cooling machines, and every machine on the Moon has to shed its waste heat. With no air or water around to carry heat off, the only way out is to radiate it into space. The book's research counts that as a hard limit on any lunar operation. The chapter brief adds the weight of the plant itself, every tank and machine needed to make and hold the oxygen, as a limit of the same kind.

Making the oxygen is still a laboratory job. Running electric current through melted lunar soil gives a yield of about 15 percent. Running it through a bath of molten salt gives over 40 percent. Reacting the soil with hydrogen gives about 1 percent. None of these runs at working scale on the lunar surface.

Liquid oxygen launched from Earth and landed on the Moon costs about $35,000 a kilogram once you count the whole chain: the fuel to lift the oxygen, the fuel to lift that fuel, and so on out of Earth's gravity. The Moon's escape velocity, the speed needed to leave it, is 2.4 kilometers per second, a fraction of Earth's. Moving propellant from the lunar surface to points around the Moon takes about one-sixth the delta-v of the same delivery launched from Earth. Delta-v is the total change in speed a rocket has to make for a trip, and it is the plainest measure of how hard the trip is.

The research puts the depot at Earth-Moon L2, a balance point in the gravity of the Earth and Moon on the Moon's far side. One orbital-transfer analysis has lunar-sourced propellant reaching a customer in a lunar orbit called near-rectilinear halo orbit at $28,800 to $29,100 a kilogram. Those savings only show up above 100 tons a year, with demand that holds for decades.

The other side of the ledger is in NASA's own studies. Its breakeven analyses have lunar propellant matching the cost of Earth-launched propellant only after about 35 years of operation. A NASA study of processing lunar ice, at a steady demand of 59 tonnes of propellant a year for 14 years, found the lunar propellant 97 percent more expensive than propellant from Earth, even with favorable assumptions. Both findings sit in the same research as the one-sixth climb.

In the research's plan for building out the space between Earth and the Moon, refueling depots are the critical middle layer. Systems built around depots are projected to cost $57 billion less over twenty years than systems that launch everything on very large rockets.

Chemical rockets have a floor. The book does not let their costs keep falling forever the way a factory's costs drop as it builds more units. Nothing in the research says lunar oxygen already undercuts oxygen from Earth.

The book's larger bet rests on that climb. Earth stays the forge, the place for capital, precision manufacturing, carbon chemistry, and training workers. The Moon, with its shallow gravity, becomes the source of oxygen, bulk building material, cold storage, and mass already parked high above Earth. The book argues that this split can reshape industry between Earth and the Moon even if the odds of full success stay low.

For now, a lunar depot is a tank at 90 kelvin, with burial under study, and the reported advantages show up only above 100 tons a year with demand that holds for decades, and lunar propellant costs what Earth's does only after about 35 years. Open The High Ground, Chapter 6, The Propellant Architecture, the chapter on depots, boil-off, and the limits of chemical rockets.