Electric shower and RCD

[JohnW2]

In your case, discrimination will allow at least some of the installation to remain operable. The only time the upfront device will (should) operate is if there is a fault with the actual submain, in which case it has to be disconnected, but the amount of wiring is limited, and the likelyhood of faults is much less than the combination of all your final circuit wiring. ... You have taken steps to minimise (not eliminte!) inconvenience in the event of a fault.

Fair enough.

In passing, in addition to having 'one or more 30mA RCDs at the final CUs', I confess that in two or three cases I have final circuits in those CUs protected only by a SP RCBO. Since a N-E fault on one of those circuits would take out the up-front Type S 100mA RCD, would you code that as non-compliant (with 314.1)?

Yes I would! As mentioned above, a fault on one circuit would render the entire installation inoperable, due to inproper design. This could be remidied bt the use of the correct RCBOs, so inconvienience in the event of a fault has not been minimised. ... I would code 3 this as it's more of a recommendation, and doesn't effect electrical safety.

Interesting. Do you think that judgements about "inconvenience in the event of a fault" should take into account the probability of the type of fault in question? Only N-E faults on one of the (few) final circuits protected by SP RCBOs would cause such an inconvenience. I'm not sure how common 'sponteneous' N-E faults are (I don't count those due to touching N & E during electrical work ) in general but, FWIW, in 30+ years of living with dozens of RCD-protected circuits, I've yet to personally experience such a fault. If such N-E faults are rare, and if less than 10% of my final circuits are protected by SP RCBOs, couldn't it be argued that the risk of inconvenience is already 'minimal', hence perhaps doesn't actually need and further 'minimisation'?!

Anyway, what about my 'actual actual' situation? I don't have just one up-front TD RCD, I have one for each of three phases. Operation of one of them as a result of a a N-E fault on an associated SP RCBO-protected final circuit would therefore only result in a loss of about one third of my final circuits (less than would be lost by operation of one of two RCDs in a 'standard' dual-RCD CU. In this actual situation, would you still award my SP RCBOs a Code 3?

I also wonder how often you actually see DP RCBOs in domestic CUs when there is an up-front TD RCD. Admittedly my experience is pretty limited, but I can't recall ever having seen a DP RCBO in a domestic CU.

Kind Regards, John