[rotorwrench]

I've seen more than a few model A engines run on wood gas which is the product of burning wood. Carbon monoxide is on the ragged edge between inert and combustible. It depends on how much in unburned hydrocarbon is still present in the car exhaust.

If a person welds with CO2 mix or argon, then that works fine. A friend of mine and I were in his dad's shop when he started to modify a new pickup bed mounted fuel tank. The moment he struck an arc on that tank, it blew up and hovered for a few seconds before banging back down on the floor. It was rectangular but turned into a cylinder shape real quick. It launched my friends dad and he ended up in the bathroom right behind him. Luckily, he wasn't injured anywhere but his ears rang for a day or so after. The tank had never had fuel in it but the slosh sealer was cured in there and likely left the combustible fumes & materials that caused it to blow. Ever since then, I try to fill them with water if possible. CO2 would be my next choice. For model A tanks, I use CO2. Argon is more expensive.