Ever thought about ... ?

No, you are all considering this in the wrong way. The fault current available on a ring circuit is extremely unlikely to be several kA, not many domestic premises even reach 2kA at the intake. The point is that the electronics have very thin PCB tracks, and the fusing current is quite low. The semiconductors are small so also have very small fusing currents in wire bonds and silicon. The heat available from any of these failures is very small. I have a lot (probably a hundred wall warts) and have never had any flames, although some fail occasionally, usually at very inconvenient times. They rarely even make any smoke, although the cases are fairly well sealed and may not show it. Short circuit failure times will be similar to BS3036 wire fuses unless there is a fusible resistor, which has a higher resistance and therefore more power dissipation, so may be quicker. The cross-section of a track 3mm wide and 0.05mm thick is only 0.15mm2, (and that is a big track by electronics standards) which is similar to a 5A fuse wire. It would probably fail open circuit before tripping a B32. A wire bond is only 0.01mm or so in diameter with a correspondingly low fusing current. I suggest that the danger is not from fire at all, but this is easily proved by a few experiments, using deliberately damaged components. There are only a few grams of material in the electronics, and much of this is not inflammable at 400 degrees C, there is very little to burn to make a fire sufficient to burn the plastic case. There is also a very limited supply of oxygen (less than 1 gram) in these sealed up units which will also prevent much ignition.


Overall I think the transformer units are worse than the electronic ones, in that a reasonable overload can get the transformer very hot without failure, although many have a thermal switch in the windings. This could melt the case exposing mains voltages as well as providing more oxygen for fire.

No, you are all considering this in the wrong way. The fault current available on a ring circuit is extremely unlikely to be several kA, not many domestic premises even reach 2kA at the intake. The point is that the electronics have very thin PCB tracks, and the fusing current is quite low. The semiconductors are small so also have very small fusing currents in wire bonds and silicon. The heat available from any of these failures is very small. I have a lot (probably a hundred wall warts) and have never had any flames, although some fail occasionally, usually at very inconvenient times. They rarely even make any smoke, although the cases are fairly well sealed and may not show it. Short circuit failure times will be similar to BS3036 wire fuses unless there is a fusible resistor, which has a higher resistance and therefore more power dissipation, so may be quicker. The cross-section of a track 3mm wide and 0.05mm thick is only 0.15mm2, (and that is a big track by electronics standards) which is similar to a 5A fuse wire. It would probably fail open circuit before tripping a B32. A wire bond is only 0.01mm or so in diameter with a correspondingly low fusing current. I suggest that the danger is not from fire at all, but this is easily proved by a few experiments, using deliberately damaged components. There are only a few grams of material in the electronics, and much of this is not inflammable at 400 degrees C, there is very little to burn to make a fire sufficient to burn the plastic case. There is also a very limited supply of oxygen (less than 1 gram) in these sealed up units which will also prevent much ignition.


Overall I think the transformer units are worse than the electronic ones, in that a reasonable overload can get the transformer very hot without failure, although many have a thermal switch in the windings. This could melt the case exposing mains voltages as well as providing more oxygen for fire.