In short rocky your right, canisters with a propane - isobutane mix are a good way to go.
Fiona right PB mix conforms well to Raoult's Law, but in use that magic number of about 0C comes up quick in canister. What I've found is propane - butane is fine for warm weather. Propane - isobutane is good for cold weather. And for very cold weather you need tricks. (It's worth know that the ground temp can be alot lower then the air temp.)
The best of which I think is feeding the mix into the stove as a liquid (by turning the canister upside down after you have lit it(might flare - be careful)). But you need a remote stove with a preheating system.
There are lots of other tricks you can do which I sure you know and people will add....
Okay, so we know that propane has a higher partial pressure than butane, because it has a lower boiling point.
Q: What does this mean about the relative evaporation rates of the two gases in the mix?
A: Propane evaporates faster. So, as the canister is used, we end up with increasing amounts of butane. Which means that the pressure of the canister falls, for a given temperature, as the canister is used.
We know this is true from empirical evidence; our experience of using mixed gas cylinders in the field. The science provides the understanding, and backs up our original statements. I tried to explain the problem without delving into partial pressures and all that malarkey; although I'm perfectly happy to deal with them, I try to provide simpler explanations that everyone can understand.
If the cylinders are kept at constant temperature, then the pressure will fall with %age capacity in the same way. However:
i) a larger cylinder will provide more burns at a decent pressure from new, simply because its %age capacity doesn't fall as fast (for a given usage rate). It will, though, also give more burns at a lower pressure as it reaches the lower part of its capacity. ii) as John said, the larger cylinder has a higher thermal mass, and so will stay at a higher temperature during the burn, and will thus give marginally higher pressures than the smaller cylinder.
Another reason to carry one larger cylinder rather than two small ones is that the cylinders themselves weigh quite a bit; two 100g cylinders will weigh more than one 200g cylinder; a simple matter of surface area:volume ratios. Even if you get to the point where the bigger cylinder is as %age empty as the two smaller cylinders, you're carrying more dead weight around.
Granted, you can't start with a new cylinder if the temperature falls...
Reminded of this thread by a UKC thread, and finally got around to plotting the fall in cylinder pressure with cylinder capacity. It's interesting, and pretty damning.
Using the vapour pressure figures provided by Fiona (probably obtained from here, which, incidentally, has mistakenly used the percentage ratios as molar percentages, when I'm pretty sure that gas is supplied as mass percentages; I've corrected for that), I determined the cylinder pressure starting with a full cylinder, and then let it be used up, using each of the two gases in proportion to the partial vapour pressure they were contributing to the mix.
Pretty dramatic loss of performance, isn't it? By the time the cylinder is 50% empty, practically all the propane has gone, and you're burning a feeble butane pressure.
Yes, it's hard work using the blasted foot-poundal per cubic gramme-banana units. I only used those values because Fiona posted them, and I'd have much preferred to have used SI units.
40F = 4.4C And I've assumed no change in temperature of the canister as it's used (either from evaporation, or from conductive heating from the stove).
When looking for the dHvap figures, the first link quoted BTU/lb...
Blast: just tried another temperature and it revealed an error in the spreadsheet. Doesn't make a huge difference:
