For the kinds of tanks used in upper stages, what is the ratio of the weight of the tank to the weight of the hydrogen? Are larger tanks more efficient?

This is for considering transport of hydrogen to be used as fuel in combination with local oxygen on the moon - so, quite hypothetical but I want to postulate such a system as accurately as I can. I presume the kinds of tanks used in for instance the Centaur would be the thing to use?

  • 3
    $\begingroup$ Centaur, or the S-II and S-IVB stages of the Saturn V, are representative tanks that include rocket engines but carry both hydrogen and liquid oxygen; the propellant mass fraction is between 86% and 92.4% for those stages. The US Space Shuttle's external fuel tank would be an example of a LH/LOX tank without engines; its propellant mass fraction is 96.5% in its later, super-lightweight versions. $\endgroup$ Dec 15, 2014 at 6:16
  • 1
    $\begingroup$ The design should take into account the micrometeoroid hazard (Whipple shield thickness), boiloff rate (number of MLI layers, mass of sunshade/active cooling system), and stress due to launch-time acceleration and internal pressure. Nominal structural material is usually Al-Li 2050 or Aluminum 2219-T87, tanks are cylindrical with $\sqrt{2}/2$ domes. This comment doesn't qualify as a full answer. $\endgroup$ Dec 15, 2014 at 8:23

1 Answer 1


Even with hydrogen, the tank weight is quite small compared to the content weight. For some examples, you can look here (you may need to do some math to get the actual numbers):


You are right that the heat loss is less of a problem for larger tanks, but structural integrity, especially when facing the aerodynamic forces of launch, becomes a larger problem.

Transporting liquid hydrogen to the moon is probably not a good option. Apart from the high tank volume and weight (compared to other fluids) it also boils away very quickly under thermal radiation. On a path to the moon, the spaceship has to coast for several days, all the while losing fuel.

This usually works the following way: The tank is kept at a certain pressure. If this pressure is exceeded, a valve opens to vent some gas. This way the liquid is constantly boiling: If a bit of heat enters the tank, a bit of hydrogen evaporates and escapes through the valve. The amount of evaporated hydrogen is the entered heat over the specific evaporation heat of hydrogen at its current pressure.

Let's say we set the valve to a higher pressure. Then the hydrogen would keep boiling, and thereby pressurizing the tank, until that higher pressure is reached and the valve opens again. Now we have a bit higher temperature on the tank because the boiling temperature rises with pressure (up to a certain point). Therefore there is less of a temperature difference between the cold hydrogen inside and the solar radiation on the outside and less heat will enter the tank. However, we need to build a thicker tank, and this is generally not worth it. As seen in the document linked above, hydrogen tanks are designed to operate at a modest pressure level of a few bar. (and even this pressure is not because of the temperature in the tank, but because of the pressure requirement of the pump inlet)

The general idea to bring fuel to the moon may be viable, although a lot of energy would be needed to generate sufficient amounts of oxygen for propulsion.

That is unless there is abundant water on the moon, in which case generating fuel and oxidizer would be simpler, but that is still under discussion.

  • $\begingroup$ I was thinking of putting boil-off rates in another question. I haven't found anything yet that gives me a clear sense of how that works. I found a paper on proposed 'zero boil off' storage on Earth. hydrogenresearch.org/nrm_nov05/fsec-baik-zbo-nov05.pdf Don't know if it can be adapted. $\endgroup$
    – kim holder
    Dec 15, 2014 at 0:44
  • $\begingroup$ The tank would need to be massively insulated and refrigerated. It will not make sense to put something like this on a spacecraft. I will add a small explanation of boil-off to the answer. $\endgroup$ Dec 15, 2014 at 0:53
  • 1
    $\begingroup$ That is an interesting tome you gave the link to. I am really going to have to learn to search around in NASA archives. $\endgroup$
    – kim holder
    Dec 15, 2014 at 3:53

Your Answer

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

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