Originally Posted by
JulesCW
[...] riding for several days without fuel in Hokkaido, Japan on one trip because canisters were rare in the region [...]
Yes, not much fun although I assume that you've been able to eat from convenience stores -- which is what we tended to do in Japan
Originally Posted by
MarcusT
Let us know what you decide and why.
My current thinking is to rely on a canister stove while bringing a
Kojin (alcohol) to use if I run out of gas. I came across this
remarkable post, comparing several upright canister stoves. They write that the Etek City is the most efficient (i.e. least fuel required to get to a rolling boil). Coincidentally, this is the stove we currently use, because it is one of the least expensive

.
So, I'll run a few tests at home in order to get a better grip on autonomy, and will try to improve my "technique" (which was -- turn the heat quite high; not bothering with the wind; and achieving a rolling boil). I plan to run tests to measure how fast the water temperature rises (I read somewhere that half of the fuel was required to move from 90 to 100C / 200-212F) and see what a partial wind screen does to efficiency, I hope to be able to meet my needs (1.5L/day) for 2 weeks or more on a single 8oz / 225g canister. Back of the envelope suggests that this would be possible in ideal conditions and almost certainly doable with a 16oz (450g) canister.
Among the things I want to learn : (1) figuring out how much gas is left in the canister. The float trick is nice, perhaps sufficient, but I'll also try ultra small digital scales (apparently popular among party goers...). Certainly more accurate, probably over doing things. (2) learning more about "warm-enough" / "hot enough". As in: tea shouldn't be prepared with boiling water; oatmeal and ramen are OK with warm water, etc etc. (3) figuring out how to evaluate water temperature. Typical advice is to look for the apparition of small and then larger air bubbles, and other physical changes occurring as the water gets warmer. Imprecise and implies that you open the lid and watch the whole process, which is counter productive and frankly boring / obsessive. There are temperature-sensitive stickers that might work. I'll probably try a cheapo (a few $, displays temperature up to 70C) and based on what I find, consider getting more accurate patches.
Interesting process