There's this ingenious little hack video: "How to make tritium nuclear battery at home".

The idea is very simple: "tritium keychain lights" - tubes with tritium and luminophor - placed between solar batteries, they provide light, converted by the batteries to electricity.

The battery in the video provides very little electricity, but it looks like the solution could be easily scaled up - it only uses two tiny keychain lights, $33/pc, and cheapest, dollar-store "garden solar light" solar batteries.

Half-life of Tritium is 12.3 years, so it would not perform nearly as well as 238-Pu with half-life of about 87,7 years on long-term missions like Voyager, but the simplicity, low mass and accessibility are so far superior, it seems like a tritium battery based on thin film solars, of the same mass as a typical Pu-238 based RTG, on timescale of ~25 years - missions like MSL Curiosity rover - could outperform the Plutonium counterpart by strides, and in fraction of the price.

Why isn't this kind of power source not more popular?

  • 2
    $\begingroup$ Probably because the mass won't be low. This video of a similar device (youtube.com/watch?v=KKdzhPiOqqg) talks about 1 microwatt of power output. You'd need 500 million of them to equal the initial power output of an RTG. $\endgroup$ – Hobbes Jan 31 '17 at 18:04
  • 1
    $\begingroup$ @Hobbes: That's an artifact of the amateur-level hack. Proper tritium betavoltaic devices are about 6% effective. That's about 1/3 keV per atom (!) = 2E26 eV/mole = 30 MJ/mole. Now for practical reasons you'll want to avoid tritium gas, but paraffin is roughly CH2 repeated. That's 7 grams per mole of tritium, so 4 MJ/gram (plus 60 MJ of heat). Of course, that's released in ~1 gigasecond so you have about 4mW/gram. 100 kg of paraffin gets you a decent 400W. And that's ignoring secondary use of the heat. $\endgroup$ – MSalters Feb 2 '17 at 21:56
  • 1
    $\begingroup$ Interesting. How opaque is paraffin? I.e. how thick can the paraffin layer be before you start losing efficiency because light is absorbed again before it gets to the PV panel? $\endgroup$ – Hobbes Feb 3 '17 at 9:15

In a keychain light, you have maybe 5 cm3 of tritium, or 0,00045 grams of 3H. This makes a tritium light cheap even though Tritium costs \$30,000/kg to make. If 5 cm3 can generate 1 $\mu$W, you need 224.5 kg to make 500 W, or about as much as has ever been produced in the world.

This would cost $6.7 million at current prices, which is unrealistic because you'd need to invest in new production facilities to make this much.

With 224 kg of tritium alone (plus solar panels and a pressure vessel), your device would be much heavier than an RTG (a 500W RTG is in the region of 80 kg).


Among other issues, tritium is hard to come by ($$). For a long time the US was not producing it at all. Only about 500 pounds have ever been produced in the US.

  • 1
    $\begingroup$ ...meanwhile, China sells it in keychain pendants, dime a dozen. (well, $400 a dozen...) $\endgroup$ – SF. Jan 31 '17 at 22:19
  • $\begingroup$ US nuclear energy policy is as crazy and political as its space policy. $\endgroup$ – Organic Marble Jan 31 '17 at 22:53

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.