Lunar Water Study: 1M-Settler City Runs Dry in 2.5 Years

For space and defense planners banking on cislunar propellant logistics, the September 14 study is a hard constraint, not an academic abstraction. The Moon's water may be trillions of pounds in glass beads, but only an estimated 1 billion tons is accessible enough to model. If 1 million residents exhaust that in under 2.5 years, the feasible scale of any long-duration lunar garrison or settlement drops dramatically.

On Monday, September 14, 2026, the journal Frontiers in Space Technologies published a peer-reviewed study that quietly upends one of the most ambitious claims in commercial spaceflight. Led by Martin Elvis, senior astrophysicist at the Harvard and Smithsonian Center for Astrophysics, and his colleague Jonathan McDowell, now honorary professor at Durham University's Space Research Centre, the analysis modeled whether the Moon's water could sustain a large permanent human population. The trigger was Elon Musk's February 2026 announcement that SpaceX had shifted from Mars to 'building a self-growing city on the Moon' in less than ten years. Elvis and McDowell asked a simple question: how big a city could that water support? The answer, by their calculations, is surprisingly small.

“

The trigger was Elon Musk's February 2026 announcement that SpaceX had shifted from Mars to 'building a self-growing city on the Moon' in less than ten years.

Taking one billion tons of water as a generous estimate of what could ultimately be extracted from the lunar surface, the authors calculated consumption across drinking, hygiene, agriculture, rocket propellant, and breathing oxygen. A lunar city of one million people - roughly the population of Stockholm - would finish that entire supply in under two and a half years under normal-use assumptions. The gap between perception and reality is enormous. That same one million-person settlement would stretch the water to one hundred years only by achieving and maintaining the 98% water recovery rate currently demonstrated aboard the International Space Station, and even that scenario requires what Elvis described as people being 'very nice' about water: accepting 'rather dubious toilets' and not doing laundry. Such austerity, the authors argue, is unlikely to be sustainable in a permanent civilian city.

The strategic sting comes from the dual-use nature of lunar water. Water is not merely hydration; split into hydrogen and oxygen it becomes breathable air and rocket propellant, making it the fundamental energy and logistics currency of cislunar operations. Trillions of pounds of water locked in tiny glass beads across the regolith sounds like abundance, but the study's one billion tons is a generous accessible budget, not all of which can be economically harvested or recycled. For national space agencies and commercial operators that have bet on in-situ resource utilization to avoid launching propellant from Earth, the finding imposes a hard ceiling. It also reframes the return-to-the-Moon era from one of settlement scale to one of remote logistical maintenance, where water recovery, extraction efficiency, and electrical power for regolith heating become the decisive constraints.