Enough arm waving from me, I'll put some real numbers in
Of course 140A is going to cook a 1mm before a 100A fuse operates but but I suppose its just a rather big overload
I'l admit that at the time, I was thinking of a large fault current welding the breaker shut. if you take the current required to operate a BS1361 in 0.1s, that is 1800A and calculate the I²t, you'll get 324,000 (a larger current will result in faster time, and lower I²t value, calculate it for 0.2, 0.4 second etc to prove that they are larger - the I²t tends towards an asymptote value as times get quicker)
Anyway, for a cable to be worth using afterwards, the I²t needs to not exceed the k²S² of the cable, for a loaded T+E at 70 allowed to reach 160 during a fault, k is 115, which gives a minimum size of 4.9mm² to be perfectly fine after the fault, so your cooker circuits and shower circuits will be fine, if the cable isn't loaded, the K value is 143, and anything above 3.9mm will be fine to use again, as to preventing fires, rather than saving cables, if we assume a k value of 166 which is for an unloaded thermosetting rubber cable allowed to reach 220degrees during a fault, which is roughly the temperature paper ignites at (though we can asusme the pvc insulation wont be too well afterwards!) this gives a value of s of 3.42mm².
This is where we get into guesswork again, the temperatures we are refering to are the temperatures of the the copper core and in this case for a tenth of a second, it would be reasonable to assume that stored materials resting against the cable would not be exposed to a temperature equal to this, though how you'd go about estimating numbers for this, I'm not sure!
