Two high-power appliances on a single 40A RCD

Alcomax

You appear to believe that domestic rings and radials shall not have more than 3 socket outlets of 13A or actually two. This is curious to say the least, and is NOT supported by BS7671. In the circuit under discussion there is no intention that Ib is > In, because the intention is that reasonable loading is applied, ie. either the shower or the cooker. This is the intention of BS7671, and you have failed in your reading of BS7671 in that both at once is not the normal use, which you yourself emphasise. The fact that it can happen is why we fuse for overload protection, using a suitable CPD.  You also need to be aware that MCBs have a specification which includes a mechanical life, and breaking something like their rated current or a little more causes no more damage than in any switch. Breaking their ultimate breaking capacity (say 6kA) may cause some damage but not enough to require that they need to be replaced after clearing the odd short circuit fault. As a design criterion it is not normally considered that overload of a circuit must be impossible, just reasonably unlikely in terms of cable heating, and protection is supplied by the CPD. There is a possible danger that an overload of up to 30% or so may not open the CPD, either at all or take a long time, and this is the long slow case which is often discussed. This level may get the cable somewhat hot, shortening life, but never hot enough to burn it, which you may calculate.


Sparkingchip

That idea (old wives tale) about fuses and circuit Zs testing has never been true, because it is contrary to the physics. A short pulse of high current has very little heating effect anywhere, and is basically the adiabatic formula at low current and very short period. You can test this yourself again by calculation. It is proved by table lamps correctly fused every day (Filament lamp of course).


Everyone

Here is something which is useful for every electrician to try, and is easy if you have an electronic temperature probe or thermometer. Take a piece of 1mm T&E cable and a 13A plug and socket strip (say 4 sockets) Find a number of high power appliances and run them via your 1mm "extension cable" for various periods while monitoring the current and temperature. You may be very surprised by your results, particularly if you plot them as multiple curves of time against temperature (similar to the MCB curves in the regs with temperature on the Y axis, time on the X axis and a curve for each current. You may post the results, they are quite interesting. (Loads to about 20A are suitable, you may need to bodge the plug fuse if heating the cable takes too long). Take care not to start a fire etc.

Alcomax

You appear to believe that domestic rings and radials shall not have more than 3 socket outlets of 13A or actually two. This is curious to say the least, and is NOT supported by BS7671. In the circuit under discussion there is no intention that Ib is > In, because the intention is that reasonable loading is applied, ie. either the shower or the cooker. This is the intention of BS7671, and you have failed in your reading of BS7671 in that both at once is not the normal use, which you yourself emphasise. The fact that it can happen is why we fuse for overload protection, using a suitable CPD.  You also need to be aware that MCBs have a specification which includes a mechanical life, and breaking something like their rated current or a little more causes no more damage than in any switch. Breaking their ultimate breaking capacity (say 6kA) may cause some damage but not enough to require that they need to be replaced after clearing the odd short circuit fault. As a design criterion it is not normally considered that overload of a circuit must be impossible, just reasonably unlikely in terms of cable heating, and protection is supplied by the CPD. There is a possible danger that an overload of up to 30% or so may not open the CPD, either at all or take a long time, and this is the long slow case which is often discussed. This level may get the cable somewhat hot, shortening life, but never hot enough to burn it, which you may calculate.


Sparkingchip

That idea (old wives tale) about fuses and circuit Zs testing has never been true, because it is contrary to the physics. A short pulse of high current has very little heating effect anywhere, and is basically the adiabatic formula at low current and very short period. You can test this yourself again by calculation. It is proved by table lamps correctly fused every day (Filament lamp of course).


Everyone

Here is something which is useful for every electrician to try, and is easy if you have an electronic temperature probe or thermometer. Take a piece of 1mm T&E cable and a 13A plug and socket strip (say 4 sockets) Find a number of high power appliances and run them via your 1mm "extension cable" for various periods while monitoring the current and temperature. You may be very surprised by your results, particularly if you plot them as multiple curves of time against temperature (similar to the MCB curves in the regs with temperature on the Y axis, time on the X axis and a curve for each current. You may post the results, they are quite interesting. (Loads to about 20A are suitable, you may need to bodge the plug fuse if heating the cable takes too long). Take care not to start a fire etc.