According to the NASA website Energy from the Moon, that assuming we perfect the helium-3 ($^3\text{He}$) nuclear fusion technology required, that $^3\text{He}$

would generate only a very slight amount of radioactivity, equivalent in nature to that produced by hospitals in their nuclear medicine areas. When used in this plant,$^3\text{He}$ would have so much energy that it would require only 20 tons-less than one Shuttle load-to supply all the electricity used in the United States in a year.

Thus, it would be very beneficial for colonists on other worlds. So, my question is has there been proven sufficient $^3\text{He}$ reserves on the Moon (and other inner solar system worlds) that would make developing and implementing this power source viable?


3 Answers 3


This depends on your definition of "proven". What we know about Helium-3 is based on lunar soil samples collected at 9 different locations (6x Apollo, 3x Luna sample return missions). Extrapolation from that yields a figure of 2.47 million tons of Helium-3 stored in the surface layer of the Moon.
However, we also know that the He-3 is at a very low and variable concentration (1-15 ppb in samples). Traditionally, mining operations depend on a much more detailed survey of the area to be mined. Unlike mining on Earth, we expect He-3 to be basically everywhere on the Moon (it is supplied by the solar wind). The big question is, at which concentration does mining become profitable? This in turn depends on how much we need to spend to mine the He-3.

To gather 1 kg of He-3, we need to process 1-15 billion kg of rock. This requires thousands of tons of equipment (excavators, rock crushers, ovens) and lots of energy. On the other hand, one kg of He-3 might yield on the order of 3 GWh of energy.

This question addresses the issues with mining the Helium and generating energy from it in far more detail.

One theoretical design for a mining system is estimated to produce 33 tons of He-3 while consuming ~800 MWh per year, which would be profitable. A NASA report estimates that we can extract ~80 times more energy from He-3 that we'd spend mining it.

  • $\begingroup$ Wouldn't we be able to use data on solar wind trajectory combined with the orbital path of the moon to distinguish which places of the moon would have the highest exposure to them? Or am I over-simplifying/making stupid assumptions again? E.G. would a specific part of the moon be "closer" to a solar hotspot? $\endgroup$ Commented Jul 19, 2018 at 19:13

Although the exact value is not known, on various estimations (like in this paper), the local abundance of ${}^3\mathrm{He}$ is in the order of $10^{-5}$.

There is also a more close experimental data. On this NASA paper, we get roughly a 1mg ${}^3\mathrm{He}$ by heating a ton of Lunar regolith to $700 ^\circ \mathrm{K}$.

The paper also shows, a built up infrastructure would require $2253 \mathrm{GJ}$ energy to transfer ${}^3\mathrm{He}$ into the Earth what can produce $600000 \mathrm{GJ}$.

Note, the mining infrastructure on the Moon wouldn't be cheap. According to this analysis, a moon base could be built in \$35billion, while its operative costs would be \$7billion yearly. It is roughly like the Russian section of the ISS.

Maybe the development of the robotics and the private space industry could significantly lower the price.

  • 1
    $\begingroup$ Fantastic, way to go on improving an answer three and a half years later! $\endgroup$
    – called2voyage
    Commented Jun 14, 2017 at 19:00
  • $\begingroup$ Could you put in the units for that abundance value of 3He? And i note that the abundance is a ratio in that paper, so it doesn't actually indicate how much 3He there is. The second reference says 7 mg/tonne of regolith, which also seems to merit correction. $\endgroup$
    – kim holder
    Commented Jun 14, 2017 at 19:18
  • $\begingroup$ @called2voyage Thanks :-) I found this post accidentally (in the chat logs, searching for myself). $\endgroup$
    – peterh
    Commented Jun 14, 2017 at 19:19

I've heard it said that IF we get fusion going, lunar regolith would have a specific energy less than low grade coal.

Hobbe's answer gives some good cites so I'll use his numbers. 3GWh - that's 3 * 109 watts * 3600 seconds which is 1.08e13 joules. 1 to 15 billion kg of rock gives a specific energy of 10800 to 720 joules per kilogram.

Wikipedia gives the specific energy of coal as 24MJ/kg. 10800J/kg is about .00045 the specific energy of coal.

I'm not sure these He3 deposits woud be worth mining if they were on earth's surface.

  • $\begingroup$ Your comparison is not quite apt. There's a lot of spoil and tailings associated with coal mining. It's not just a lot, it's a whole lot, and it's rather toxic. On the other hand, your IF is spot on. We don't yet know if sustained fusion power is achievable, let alone profitable. That's a question our children or grandchildren will answer after the successor to the successor to ITER is put on line (IF that ever happens). $\endgroup$ Commented Apr 11, 2014 at 17:51
  • $\begingroup$ "There's a lot of spoil and tailings associated with coal mining." I need some numbers to be persuaded by this argument. Is the waste 99.955%? That would put earth coal dirt at the same specific energy as moon dirt. And that would be far short of what's needed to offset the extreme high cost of space transportation. $\endgroup$
    – HopDavid
    Commented Apr 11, 2014 at 18:25
  • $\begingroup$ Either way, fusion is still a pipe dream. He3 is still one of the most expensive things that exists. We have solar panels that work just fine, and cost near nothing to make in comparison. If you want dense power, you make an enormous solar city on the moon in an area of permanent daylight, then convert the local materials to methane using the electricity supplied. Then transport it. Easy. This is the simplest answer, and is why Musk is using it (not the moon bit, but in general). $\endgroup$
    – T. B.
    Commented Jun 14, 2017 at 18:59

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.