Fault current rating and disconnection time,  time tripping curve required of the protective device prevents this. 

...or directly using energy let-though for MCBs. The point though is that normally such things are calculated for the 50A device and 10mm² conductors on one hand, and the 6A device and 1mm² conductors on the other - but not other combinations if inter-circuit faults had to be considered. (FWTW a 6kA 50A B-type MCB has an energy let-though of anything up to 54,000 A²s - so can't easily be shown to protect conductors below about 2mm²).

   - Andy.

I don’t know what you are talking about . Use an RCD/ RCBO  if you are not confident in the MCB. You have to select the correct  CPD for your circuit. 

You have to select the correct  CPD for your circuit. 

Indeed - but that doesn't cover faults between conductors of different circuits - which was what I was trying to illustrate about BS 7671 (and engineering in general) not covering every eventuality and so nothing can really be considered totally faultless however hard we try.

Use an RCD/ RCBO  if you are not confident in the MCB

Er no. If the RCD is common to both circuits it won't even see the fault. If each circuit had its own RCBO then the L-N fault between circuits would be seen, but even then the guaranteed opening time (e.g. 40ms) would likely be far too slow to protect conductors against fault currents (in many conditions, k=115), fault currents above 575A would exceed the withstand of a 1mm² conductor (which is why RCDs are almost always backed up by OCPDs).

   - Andy.

Tell that to a TT earthing arrangement. I am unsure what your issue is, OCPD is not the same as fault protection.

and it is 400 ms for a TNC system before going to 5 seconds. 200 ms on a TT unless you have MPb and a max operation time of 300ms via RCD  will satisfy a TNC.

if you connect two circuits on the same phase via a ‘screw’ the OCPD will have to be relied on. Other than that your fault protection becomes what youd rely on. 
if you are worried about arcing faults fit an AFDD

Hi Mark. Just picking up on energy let-through of a device. For example, suppose a MCB is installed where the fault level is 10KA and the MCB has a energy let-through of 0.52 x10E6 A2s. And you want to know if a 6mm2 CPC with thermoplastic insulation be enough for this device? According to table 43.1 of BS 7671, the value of K is 115. So, we need to find the minimum cross-sectional area (S) of the CPC that satisfies this equation: S squared x 115 squared is greater than or equal to 0.52 x 10E6. If we solve for S, we get: S is greater than or equal to the square root of (0.52 x 10E6) divided by 115, which is about 6.27mm2. Therefore, a 6mm2 CPC would not be sufficient for this device, and we would need to use a 10mm2 CPC instead.

Thank ypu for the lesson in Adiabatic calculation  Of we are designing imaginary circuits… What if the fault current was 5kA and the CPD was rated at 6kA ?

Depends on the manufacturer energy let through table. different energy let-through values for various prospective fault current 

Exactly. Imaginary circuits can be designed to fail as an example

If the fault current is 5KA and your MCB rated short circuit capacity is 6 KA this means it’s adequately rated to interrupt fault currents safely. Icn - max can interrupt safely but may no longer be usable,  Ics max without loss of performance 

It’s kA the capital would be Amperes shortened name. The one Coulomb a second person. Cheers for the education. It is good to see there are people in the industry still interested. Enjoy ya night mate