## Can a Million-Person Lunar City Actually Survive on Polar Water Ice?

The short answer, per new peer-reviewed math: no — not sustainably, and probably not even close. A new paper published in *Frontiers in Space Technologies* by Dr. Martin Elvis (Center for Astrophysics | Harvard and Smithsonian) and Dr. Jonathan McDowell finds that even under the most generous assumptions — one billion tons of available polar water ice and 98% water recycling efficiency, matching the best performance achieved aboard the International Space Station — a city of one million people on the Moon would exhaust the entire water supply in just over a century. Current best estimates of actual available water are roughly 30 times less than that billion-ton ceiling, which compresses the timeline proportionally. At realistic figures, even a small city runs dry within a decade or so. A village of around 1,000 people, or a town of around 10,000, is the scale where multi-century sustainability becomes plausible. That constraint has direct implications for every [cislunar space](https://orbital-intel.com/glossary/cislunar) architecture being planned today, from NASA's lunar surface programs to the ambitions of commercial operators targeting the lunar poles.

---

## The Water Budget: Where the Numbers Come From

Since 2013, a succession of lunar-orbiting missions has confirmed the presence of water ice in permanently shadowed craters near the Moon's poles — the so-called "cold traps" where crater floors have not received direct sunlight in approximately four billion years. Temperatures in these regions drop below 110 Kelvin (minus 262 degrees Fahrenheit), cold enough that water ice sublimates by no more than a millimeter per billion years even in hard vacuum.

The upper-bound estimate the scientific community has been working with is approximately one billion tons of accessible water. Elvis and McDowell use that figure as their generous starting point. Strip away recycling entirely, and a large urban population burns through it in a few years. Apply 98% recycling — the ISS standard, which represents the state of the art in closed-loop life support — and a one-million-person city gets roughly a century.

The harder blow is the current best scientific estimate of available water, which the authors put at roughly 30 times below the billion-ton ceiling. That multiplier compresses every timeline in the analysis by the same factor: what looked like a century of runway at optimistic estimates becomes a few years at realistic ones for large populations.

Current lunar water surveying techniques do not reach more than a few meters below the surface, while the regolith — the rubble-like surface rock layer — typically extends tens of meters down. Elvis notes that subsurface water in the cold traps represents the most promising avenue for changing the resource calculus, but accessing it requires prospecting infrastructure that does not yet exist.

---

## Power Is Not the Binding Constraint

One finding that cuts against the conventional framing of lunar settlement challenges: electricity generation appears tractable even without nuclear reactors. The rims of the permanently shadowed craters — sometimes loosely called "peaks of eternal light" — receive near-continuous solar illumination. Elvis and McDowell calculate that kilometer-tall towers fitted with photovoltaic arrays on those rims could generate three gigawatts of electrical power. The authors also note that silicon, the feedstock for solar panels, is abundant on the lunar surface, making local panel manufacturing a credible long-term option.

Three gigawatts is a meaningful figure. For context, that is enough to power a mid-sized terrestrial city's electrical grid. The authors suggest this power availability could even support AI data centers positioned near the illuminated peaks — an economic use case that could anchor early commercial activity and help establish what they call "the start of a true lunar economy."

The power picture does not solve the water problem, but it does reframe the debate. The binding constraint on lunar urbanization is not energy — it is [in-situ resource utilization (ISRU)](https://orbital-intel.com/glossary/isru) of water specifically.

---

## What Could Change the Math

Elvis identifies four potential mitigations, each with different feasibility profiles:

**Improved recycling efficiency.** Current ISS-grade recycling runs at 98%. A factor-of-five improvement would extend the water runway substantially, but the authors flag this as technically challenging — 98% is already close to thermodynamic limits for some process steps.

**Lower per-capita consumption.** Techniques like vertical farming reduce agricultural water demand relative to conventional methods. This is likely part of any serious lunar settlement architecture regardless of water constraints.

**Water importation.** Asteroids represent an external water source. Accessible near-Earth asteroids contain water in various forms, and several commercial and academic proposals have explored asteroid water mining. The [delta-v](https://orbital-intel.com/glossary/delta-v) cost of redirecting asteroidal water to lunar orbit is non-trivial but not prohibitive for water-rich near-Earth objects on favorable trajectories.

**Finding more water.** The authors treat this as the most promising near-term path. If subsurface water in the cold traps is substantially more abundant than surface surveys indicate, the entire resource picture improves. Drilling or seismic prospecting missions at the poles could dramatically update the estimate — in either direction.

---

## Industry Implications: Scaling Ambitions vs. Physical Limits

[Blue Origin](https://orbital-intel.com/companies/blue-origin)'s Jeff Bezos and [SpaceX](https://orbital-intel.com/companies/spacex)'s Elon Musk have both publicly articulated visions of lunar-scale human habitation — Bezos advocating moving heavy industry to the Moon, Musk describing "self-growing cities." The Elvis-McDowell analysis does not invalidate early base operations or research outposts, but it does place a hard physical boundary on the urbanization narrative that has been driving parts of the commercial lunar investment thesis.

For operators and investors evaluating lunar programs, the practical takeaway is that settlement architecture at the village-to-small-town scale — the paper cites 1,000 to 10,000 people as sustainable for several centuries or more under conservative assumptions — is supportable with known water resources. Scaling beyond that requires either a major upward revision in the water inventory (which is possible but unproven) or external supply chains from asteroid resources or Earth.

[Commercial Lunar Payload Services (CLPS)](https://orbital-intel.com/glossary/clps) missions already manifest water-prospecting instruments among their science payloads. The Elvis-McDowell paper effectively raises the stakes on what those instruments find — and on the depth of future subsurface surveys. A mission that can characterize water content at 20-30 meter depth rather than a few meters would be scientifically and commercially transformative.

The paper also implicitly challenges the timeline compression in some commercial lunar roadmaps. Building habitat infrastructure before the water inventory is adequately characterized is building on an unvalidated foundation. The 30-times uncertainty in the current best water estimates is not a small error bar — it is the difference between a sustainable outpost and a logistical crisis.

---

## Key Takeaways

- **Hard ceiling:** At 98% water recycling efficiency and one billion tons of available polar water ice, a one-million-person lunar city exhausts all water in just over 100 years — per Elvis & McDowell in *Frontiers in Space Technologies*.
- **Realistic baseline is far worse:** Current best scientific estimates put available water at roughly 30 times below the one-billion-ton figure, compressing every timeline proportionally.
- **Sustainable scale:** A lunar village of ~1,000 or town of ~10,000 people is supportable for several centuries or more under conservative resource assumptions.
- **Power is not the problem:** Three gigawatts of solar generation is achievable from crater-rim photovoltaic towers without nuclear reactors, and in-situ silicon enables local panel manufacturing.
- **The fix is subsurface prospecting:** Current surveys penetrate only a few meters; regolith extends tens of meters and may conceal substantially more water — or may not.
- **Asteroid import and advanced recycling** are identified mitigations but carry significant technical and logistical uncertainty.
- **Investor implication:** Lunar settlement architectures sized beyond small towns require validated water inventories that do not yet exist. CLPS and future prospecting missions are load-bearing for the commercial lunar thesis.

---

## Frequently Asked Questions

**How much water is thought to exist at the lunar poles?**
The upper-bound estimate used by Elvis and McDowell is approximately one billion tons, based on data from lunar-orbiting missions that have been mapping polar cold traps since 2013. However, the authors note that current best estimates are roughly 30 times lower than that ceiling. Surveys are limited to a few meters of surface depth, leaving deeper deposits uncharacterized.

**Why can't a large lunar city just recycle all its water?**
Even at 98% recycling efficiency — the level achieved on the ISS, which represents the current state of the art — a one-million-person city still consumes water faster than it can be recovered. The 2% loss per cycle compounds at urban population scales. A factor-of-five improvement in recycling efficiency is theoretically possible but technically demanding.

**What settlement scale is actually sustainable with known lunar water?**
Per the Elvis-McDowell analysis, a lunar village of around 1,000 people or a town of around 10,000 people could be sustained for several centuries or more under conservative water assumptions. Larger populations require either substantially more water than current surveys indicate, dramatically improved recycling, or imported supply.

**Could asteroid water solve the lunar water problem?**
Potentially, yes. Water-bearing near-Earth asteroids exist, and some are on trajectories that make water extraction and delivery to cislunar space conceptually feasible. The delta-v economics depend heavily on the specific object and trajectory. This remains a long-term option rather than a near-term solution — no operational asteroid water mining capability exists today.

**Does the lunar power situation present similar constraints?**
No — the paper finds power is a solvable problem. Crater rims near the poles receive near-continuous solar illumination, and the authors calculate that kilometer-tall photovoltaic tower arrays on those rims could generate three gigawatts of electricity. With silicon available in the lunar regolith, in-situ solar panel manufacturing is considered a credible long-term option.

**What missions could update the water inventory estimate?**
Subsurface drilling or seismic prospecting at the lunar poles, penetrating to depths of tens of meters rather than the current few-meter limit, would substantially change the resource picture. Several CLPS payloads carry water-detection instruments, but dedicated deep-prospecting missions would be needed to definitively characterize the cold-trap inventory at depth.