Actually, this is probably a picky point, but the O2 is'nt what's burning. The fire is caused by rapid oxidation of the fuel source, which requires oxygen, but the oxygen isn't the thing "oxidizing". High levels of O2 can just increase the rate at which oxidation can occur.
Here is an extract from the abstract of a paper by Littman, Church and Kinderman that says:
This is not happening because of a heat source. This is spontaneous chemical combustion along very similar lines to what happens when you expose sodium to water.
That's a PPRO2 of a bit more than 24 bar or in the case of the normal working pressure inside a regulator, it would mean it could catch fire if the oxide layer stopping it from catching fire is scraped or a crack develops in the reg when the tank has anything more than 50% O2 in it.
Theoretically, on the high-pressure side of the reg if even a small crack were to occur in the reg then it will start on fire, even using normal air.
My thinking is that this is probably what happened with the one regulator we know of where this happened. I'd be willing make a substantial bet that the reg was damaged and there was some kind of oxide free layer/crack/scrape inside.
That said. *WHY* would anyone use titanium to build a scuba regulator when it's a well know fact that this can happen? That's just silly. There are simply better choices of materials for scuba regulators and I for one would certainly advise against buying a titanium regulator for this reason.
R..
