Cla

While I agree that there won't be any moisture inside the tank, there are ways that "propane frost" could form inside the valve.

There's something called the "Joule-Thompson Effect" which basically says that when a compressed gas is allowed to expand, there's an associated temperature drop in the gas.

Joule

You have a similar thing happening in an air compressor. The head on the air compressor will get quite hot because air heats up when it's compressed. And, on my little air compressor, the fins on the head get way too hot to touch.


What I'm thinking is that if the propane torch is outdoors where the ambient temperature is say -35 degrees Celsius, and propane vapour is expanding through a needle valve to atmospheric pressure, and the tip of the needle is insulated by a packing material that reduces heat transfer from the valve body to the needle, I wouldn't be too surprised if the needle in the valve could actually get as cold as -188 degrees Celsius, which is the freezing temperature of propane. In that case, "propane ice" would begin to form on the needle of the needle valve, thereby preventing it from closing properly.

Not saying that's what's happening, but since both the evaporation of propane absorbs heat, and the expansion of the propane vapour in the valve absorbs heat, and in winter there's not a lot of ambient heat outdoors to begin with, it may be possible to have propane freezing on the needle in the torch's valve causing exactly the same types of problems that water ice would.