I've thought about using the cargo quadricycle to take my propane tank to the hardware store for a refill but didn't do it because it felt hard to do safely.
That didn't worry this person, who has put the tank (about 1/3 the size I have) on a furniture dolly improvised into a trailer. #CarryShitOlympics #BikeTooter
@glightly You thought pool noodles were effective, you should see the wide berth drivers give you with a couple propane tanks.
2020-10-01 https://twitter.com/enobacon/status/1311768274216972288
The only issue is see is that the tanks aren't chained (although there are retaining walls), meaning metal gas bottles may be thrown around during a big enough collision, but that's mostly a blunt trauma risk really.
Metal propane tanks have a very large safety factor, as propane only needs 1220 kilopascals (177 psi) of pressure to keep it liquid at 37.8 °C, so explosions don't happen unless you have a tank that has been leaking gas for ages, leaving a lingering gas cloud and a source of ignition.
I would guess that the main risk comes from blows to the valve (which enjoys a significant but incomplete protection from that collar).
I think you are right on limited consequences of a fire. I would expect though that an ignition source should be pretty easy to find in a car accident: the exhaust systems of all involved cars are likely got enough.
That seems unlikely.
The autoignition temperature of propane is 470°C; https://en.wikipedia.org/wiki/Propane?useskin=monobook
According to this site, exhaust systems do get hot, with typical exhaust systems heating up to 300-500°C (only the upper 30°C in the range is hot enough), although catholic converters can heat up to 750°C; https://cargearguru.com/how-hot-does-car-exhaust-get/
But really, I don't believe surface temperature is applicable to atmospheric heating, plus in a big accident, anyone with sense is going to stop their vehicle engine, which would allow for some cooling time before a nice flammable cloud can form.
It's only really if the escaping plume meets an ignition source before it disperses that a problem will result (metal against metal action may cause sparks as the valve is being shattered, but it seems that you'd need to be really unlucky for that to happen).
According to this site (https://propanehq.com/how-explosive-is-propane/), the explosive mix between propane and air is approximately 2.1% to 9.5%, so it looks like even if the escaping plume does get lit, that'll most likely end up burning rapidly instead of exploding (although if the tank is overheated to the point that the propane starts boiling and ruptures the tank, the pressure difference is an explosion; https://en.wikipedia.org/wiki/Boiling_liquid_expanding_vapor_explosion?useskin=monobook but it appears that small tanks don't really have the capacity to allow for the required heating before the contents are exhausted, while large tanks certainly do).