Where did you obtain the information about the K2S2 of the foil being generally greater than that of a 1mm copper cpc Andy? 

Basically BS 8436 cable manufacturer's data - most of which seem to suggest that same sort of figures - 42,000 A²s for 1.0mm² or 1.5mm² cable and 60,000 A²s when used with 2.5mm² or 4.0mm² cable - compared with the normal calculation that gives 13,225 A²s for 1.0mm² c.p.c. in T&E and 29,756 A²s for a 1.5mm² - which suggest to me that BS 8436 cable is hugely superior to T&E in terms of withstand.

Yet there's all this fuss about only using certain values of B-type MCBs with BS 8436 cable while at the same time we stick 1.0mm² c.p.c. T&E on a C16 or 1.5mm² c.p.c. on a B32 in a 6kA or even 16kA situation without a second thought.

Probably we should be looking more closely at the T&E situation, but my general feeling is that if you'd be happy using T&E in a particular situation then using BS 8436 cable instead can only be an improvement.

   - Andy.

Thanks both. Atom have sent through clause 12 of the British Standard which seems to confirm all discussion here i.e. to safely protect the cable the mcb needs to have a maximum energy let through <42000 A2S and the circuit breaker should be type b, class 3, 60898 of correct value etc. Makes me feel a bit better anyway that I understand what I'm installing! As ever thanks for all your help.

Is it legitimate to confirm the max energy let through of the device is <42000 A2S for the measured fault current of the installation or do you think this should be verified as a generic figure from the manufacturer?

Interestingly enough I've always tried to swerve Type C's with T&E as I have noticed that the energy let through can be a bit tight for a near origin fault and sometimes even exceeds the CPC K2S2 at even modest fault levels - particularly with 1mm. But as you say…. seems to get a launched in at a rate of knots everywhere and no one seems too worried. Then again there were a lot of people on the Titanic and it still sank. I think I had an NICEIC area engineer actively encouraging it once on the basis that the 30mA RCD element of an RCBO would restrict the fault current to 30mA (which it wouldnt). 

I think it is good you are being thorough, many in your shoes would not bother to be.

But, do realize  you are starting to worry about a small effect - by all means note the PSSC while you have your meters out, but the whole “what is the the exact  fault current?”  thing  is not that black and white - tomorrow the street cable may be upgraded to make way for some EV chargers, or on the other hand the house holders may use  larger or smaller nail than the one in the BSI test. Or,  most likely, it will never get nailed, ever, in the life of the building.

And, if it does trip someone could just keep re-setting the breakers till it burns clear, or they may realize their error and remove the nail and yet may or may not get the cable repaired.

You cannot protect against all possible events however unlikely, just the reasonably credible ones.

And yes, thinking of the area engineer, not everyone with an official role talks sense all of the time. Some of us more than others , so let's hope it was a bad day, and not a fundamental mis-understanding about what is in an RCD.

Mike.

Hi Mike, thanks, for sure I do appreciate that the risk is small and possibly concerned with events beyond my reasonable control. I do find it comforting knowing that it was ok when you left it at least though! Particularly when as noted you're contemplating an installation that, although compliant, is a bit beyond the ordinary!

Regarding the PFC thing- when I come across reduced size ring final CPC's I do check the energy let through of the CPD's at onsite measured PEFC and compare this to the K2S2 of the CPC in an attempt to prove compliance. I also do this when I encounter higher pfc's on installations with T&E present. Would you consider that a redundant exercise given the discussed feasibility of network change? 

The same I suppose would apply to your measured earth fault loop readings on circuits if they were a bit on the edge. I've always based it on how it tests on the day particularly with existing installations. I've seen the design guide recommends designing for worst case scenarios from the network. i.e. a 16kA pfc and for example a 0.8 ohm ze for a tns system. In the real world this just isn't going to be practical though is it - and T&E would never be seen again - even in a domestic environment as >3kA some of the cpc sizes get a bit skinny (as table b7 in the onsite guide recognises).

Going back to the type c/rcbo thing - it seems to be a widely held misconception:

Seriously though. Thanks so much to you both for the discussion. It's great to be able to bounce these things off someone!