It’s no longer a question of whether to fly or not, whether to build or not. The question now is how, from what, and what is the most efficient and cheapest option. We’ve already written about technologies from our Institute regarding the possibility of building structures on the Moon using acid-alkaline geopolymer activation, rather than sintering, as others propose.
Today the question is how to make glass. After all, you certainly can’t bring it from Earth, and without windows, life on the Moon would be a life sentence in a dark prison. We have a solution for that too.
The Moon immediately strips away idle talk and leaves only what survives the test of mass, energy, and cost, because every decision here passes through a delivery-price filter of about $1,000,000 per kg, and that filter lets nothing superfluous through. At that point the familiar logic of construction disappears, because you can no longer operate on a “ship it and assemble it” basis, and only one question remains — what do you make glass from, if you can’t bring it? This is where things get interesting, because the answer lies literally underfoot, and it demands a different kind of engineering logic.
Regolith isn’t just dust — it’s a mixture of minerals, glassy phase, and fragments formed by billions of years of impacts, in a layer ranging from meters to tens of meters deep, with sharp, abrasive particles 60–80 µm in size, composed of silicon, aluminum, calcium, magnesium, and iron, with an FeO share of around 15–18%. There’s no water. No atmosphere. That means no classic concrete, no conventional binders, no wet processes in the usual sense. And most importantly, any kilogram that isn’t part of a critical system simply fails the economic barrier of delivery. At this point it becomes clear that there can only be one construction material — the one that’s already there on-site — and glass is no exception.
You can’t bring glass from Earth. Even taking the minimum volume needed for a base, the numbers quickly become astronomical. A base for 20 people requires 500–700 m³ of structures. If even 10% of that volume were replaced with glass — that’s 50–70 m³ — you’d need to ship 125,000–175,000 kg, roughly 150 tons. At a delivery cost of $1,000,000 per kg, that’s $125–175 billion for glass alone. Without counting equipment, people, or anything else. That’s where the conversation ends.
So the task is uncompromising: everything produced on the Moon must be made from local raw material. If not, there’s no project. If yes, everything else is just engineering.
Today the main approach under consideration is sintering regolith: heating to 1,100–1,200 °C, melting, cooling — yielding ceramic or glass at the output. It’s the classic scheme. But at scale, a systemic problem emerges.
Producing 1 m³ requires 400–550 kWh of energy. For a 500–700 m³ base, that’s 200,000–385,000 kWh. With available power of around 30–40 kW — typical of an early lunar-base power system — that means 8–14 months of continuous operation just to produce the material. And that assumes a significant share of energy is dedicated to construction, whereas in reality the resource is shared with life-support and thermal-stabilization systems, which extends the timeline further. On top of that, lunar night lasts 14 days, and solar generation drops to zero, requiring storage on the order of 10–11 MWh. The mass of such storage systems can run from tens to hundreds of tons depending on the storage technology, turning power supply into a separate problem in its own right.
But even if the energy question is solved, the mechanics remain. Glass is an amorphous mass with no internal structure. It withstands compression well but barely tolerates tension. Any microcrack in glass grows, because there are no internal barriers to stop it. On the Moon this is compounded by thermal cycling of around 290 °C and by micrometeorite damage that accumulates in the material. A pressurized volume works under tension, since pressure is inside and vacuum outside — creating a direct risk of failure.
Regolith, meanwhile, contains an amorphous glassy phase formed by impact processes, which potentially increases its reactivity but requires preparation and compositional control for a stable result. And this is where a different approach appears — one that doesn’t require full melting.
You take the same regolith. After sizing the fraction and adjusting the composition, an alkaline activator is added at 3–5%. This triggers a reaction that forms an aluminosilicate matrix without high temperatures. Energy consumption is 14–28 kWh per m³.
Let’s do the math. For a 500–700 m³ base, that’s 7,000–20,000 kWh. At 30–40 kW of power, that’s 1–3 months of operation, assuming a significant share of energy goes to construction. Compared to glass, the difference is about 93–97% in energy — more than an order-of-magnitude reduction.
The resulting material has a granular structure with a chemical binder. Unlike glass, cracks don’t propagate freely — they’re arrested at grain boundaries. This gives compressive strength of 60–90 MPa and tensile strength of 4–8 MPa, allowing it to be used as a load-bearing material and substantially reducing the need for additional framing elements, though these may still be required in certain zones.
Now back to glass. Yes, it can be made from regolith. This uses local heating to 1,100–1,200 °C, forming, and rapid heat removal through contact and radiation in vacuum. But energy consumption remains at the 400–550 kWh per m³ level.
The actual glass requirement for a 20-person base is limited to windows, optical elements, and protective layers, giving a benchmark of around 2–10 m³. That corresponds to 800–5,500 kWh, which fits within a localized production mode without requiring a dedicated power system.
This is where the logic comes full circle. The same starting material — regolith — yields two classes of product. In one case, melting produces glass at high energy cost. In the other, chemical activation produces a composite at a fraction of the cost. Both processes can run on the same energy and thermal equipment, operating in different modes — continuous low-energy and pulsed high-temperature.
ARBOK proposes a technological loop that unites these modes into a single system. Producing the composite costs 14–28 kWh per m³, versus 400–550 kWh for glass. For a 500–700 m³ base, that’s 7,000–20,000 kWh versus 200,000–385,000 kWh. At 30–40 kW of power, that’s 1–3 months versus 8–14 months. The energy difference reaches 93–97%, and the time difference is a multiple of that.
This means glass becomes a local product, produced in small batches for specific needs, while the main structure is formed from the composite. In such a system, no separate energy infrastructure is needed for glass, and construction timelines stay within a single mission.
And this is where the talk ends and the engineering begins.
Also published on: LinkedIn — Michael Vischmidt · Facebook — ARBOK Institute
