Burst discs: Yes they burst when they shouldn't

Everything is possible, except for skiing through a revolving door. Another heat cycle with another pressure change could be the straw that broke the fatigue camel's...er...burst disk.

Probable is another matter. As a rule, burst discs fail when tanks are full or being filled, when the pressure across the disc is at a maximum or rising or both. Or, as happened to me once, left in a hot car when full. (Full of Trimix. Of course. It had to be Trimix. Air is too cheap and quick to refill just in advance of a dive charter for it to have been air.) But in reality, once the fatigue life of the disk is at an end, any flexing in any direction can precipitate the failure.

 

While I have no idea about what hydro pressure is on your side of the pond, I have a decent idea about what it is over here: 150% of service pressure. So let's do the math.

500 psi = 35 bar.
Hydro pressure for a 232 bar tank is 348 bar. Let's round to 350.

PV=nRT, so P1/T1 = P2/T2. Thus, P2/P1 = T2/T1 or T1*P2/P1 = T2. Assume room temp, i.e. 20C, i.e. 293K. P2/P1 = 10, so T2= 2930K or 2660C.

So, to reach hydro pressure in a tank with 500psi, you need to heat it to 2660C or 4800F. I suspect that the tank material will melt long before that.

 

@Storker 5/3's working pressure for 3aa and 3al, but close enough. The special permit tanks are a little different.

Most steel melts around 1500c IIRC, but brass is less than 1000c so the valve will actually melt before a steel tank will, and I think aluminum is around 500c so long before that as well.