Should we get material from the Moon?

Order-of-magnitude lower cost of raw materials to LEO.

Getting anything to orbit requires so much energy that rockets are essentially huge fuel tanks with a small payload at the tip. 90% of the mass of a rocket comes from its propellant! This is awkward because we already have a huge amount of useful material in orbit; some call it the Moon. Given how hard it is to get stuff into space, does it make sense to source building materials from the Moon?

The Moon has long been a dream of space industrialists, the abundant raw materials and lack of zoning restrictions make it a great place to build. When launch costs were higher, some suggested that we use industry on the Moon help assemble large structures in orbit. Since then, launch costs have fallen dramatically, making the math for lunar industry dubious. But I think with careful design, sourcing from the Moon can still make sense. The key is to use cheap terrestrial launch capabilities to our advantage and focus on sending exceedingly simple products like bulk regolith.

For a fair comparison, let’s look at the cost to send material to low-Earth orbit (LEO) from either the Moon or Earth. I expect most of cislunar industry will operate here regardless.

How do we get material from the Moon to LEO? Some purely electrical means of propulsion is important, since chemical propellant is hard to get on the Moon. In the past, people have suggested mass drivers, but this would involve a large rail system and careful engineering in a remote and unforgiving environment.

As it turns out, the company Spinlaunch has a better option; they’re building what is essentially a catapult. The speed of the projectile is close to the Moon’s escape velocity and working in a low gravity, near-vacuum environment eliminates many of their technical challenges. It’s more compact than a mass driver and easier to assemble.

So let’s assume we’re building a Spinlaunch system on the Moon and getting equipment there via Starship. If the Spinlaunch system is 100 tons, we can load it on a single Starship. It will take about 10 flights (mostly refueling flights to LEO) to boost that payload to the Moon.

Assuming a launch cost of $100/kg, that’s $100 million to get the equipment there. Spinlaunch’s payload is 10 metric tons, so assuming we can get 10,000 launches out of the tether, that gives us a cost of $1/kg.

How much energy does this require? Spinlaunch requires 100 MWh per spin-up, but that’s far more than the kinetic energy imparted on the projectile (~5 MWh). Spinlaunch recoups some of that energy running an electric generator off the tether post-launch, so let’s assume the round trip efficiency of charging and discharging the tether is 95%¹. That means each launch costs us 10 MWh, 5 for the projectile and 5 from system losses.

On to the power source. Instead of trying to be clever, let’s just launch 1000 KRUSTY nuclear reactors split over 2 Starship launches. Over 10 years that should provide enough power for all of the lunar launches with roughly 3 launches per day. Assuming 100 kg per reactor, that brings our total price to $3/kg. Add in the cost of things like thrusters, robotic assembly, mining equipment, and repair robots and let’s round it off to $10/kg.

This is already an order of magnitude cheaper than the $100/kg launch costs we assumed at the beginning. With tricks like beaming microwave power to the Moon, space tethers, double-sided tethers, or sourcing materials from lunar regolith² I think we can get another order of magnitude reduction in cost.

Notice that every cost is based on Starship launch costs, so the ratio between the two will stay the same. If launch costs fall to $10/kg, then the cost of this plan will also fall by an order of magnitude.

What does this buy us? Cheap lunar material means that we don’t need to launch bulky equipment from Earth. Large metal structures, radiators, solar panels, and other heavy items can be manufactured in LEO from lunar regolith; this dramatically lowers the total mass we need to send from Earth. Terrestrial launch can focus on sending complex items like computer chips and scarce resources like methane.

Particularly interesting are the opportunities to get oxygen from the Moon and assemble atmospheric scoops from lunar material. Liquid oxygen takes up 70% of Starship’s mass, by sourcing it in space the number of refueling flights can fall from 10 to about 2.

These benefits make it cheaper to build industry on the Moon, sparking a virtuous cycle. As the industry bootstraps itself, it will take on a larger role in projects going to Mars and other parts of the solar system.

While the math pencils out, we’ll have to wait decades for the technology that makes this possible. Energy costs are a major problem; even with the optimistic case I presented here, lunar energy costs $1/kWh. Compare to Earth, where that number approaching $0.01/kWh. This is a critical, since lunar industry will need a lot of energy. One option is to bootstrap to more energy; perhaps solar energy could be used to create more solar, or nuclear energy could power a factory that collects fissile material³.

Progress will be slow, but significant. Government forays to the Moon will kick off a self-reinforcing loop that can lower terrestrial launch costs by another order of magnitude at least. Once industry is established in space, people will laugh at how much effort we took to reach orbit when everything we needed was already there.

EDIT: a nice discussion on lunar catapults.

1. This is high for normal electric motors and generators, but the application benefits greatly from high efficiency. In the cold of space, it may make sense to use superconducting motors or cryogenic aluminum. One alternative is to have two tethers and exchange momentum between them directly. The company Astro Mechanica plans to use liquid methane cooled motors to achieve ~98% efficient energy transfer between two turbines. The tether also acts as a nice energy storage system, smoothing out load from variable sources of energy on the Moon like solar or thermal.

2. Heck, it might even make sense to make the tether larger and slower and make it out of high-strength glass fiber derived from regolith.

3. A Thorium nuclear reactor is particularly interesting here since Thorium is more abundant and doesn’t require isotopic enrichment. See here for more.