Fitting two RCDs one after the other

[EFLImpudence]

Best have three, then.

This is getting silly.

 

[JohnW2]

Depends on your barrel. Heat could be a failure mode, or humidity.

Indeed, that's why I mentioned it, but I still think the chances of two of them suffering the same fate (in response to likley environmental influences) must be extremely low.

Also, how do you know when there's a latent failure? Both could fail and everyone be blissfully unaware until they are next tested. So you're just changing from the time until one faults, to the time until two fails.

Again indeed - but if one does test both fairly regularly, the hope would obviously be that one would detect (and deal with) a fault in the first to fail before the second one had failed.

As I implied, with (essentially untestable) MCBs, there is more of a problem. However,, even without any testing, in both cases (RCDs and MCBs), even if the two RCDs (or MCBs) are subjected to the same environmental influences, the chances of both failing within any period of time will presumably be less than the chance of one of them failing (unless the 'environmental influences' are really extreme - like a house fire!).

Kind Regards, John

 

[JohnW2]

Classic example might be a large(ish) house with a main consumer unit in the vicinity of the incoming mains feeder/meter and a sub-distribution unit for lighting systems located elsewhere in the property. RCD in main unit protects against typical ground faults and RCBO in sub-DBunit protects against outgoing lighting circuit faults, particularly excessive current when a lamp fails. Yes it is possible that either RCBO or RCD might trip on an imbalance but my experience has been that the outgoing RCBO only trips on a lamp failure (as might be expected).

I think you are confusing the issue, because, on your own admission, you are talking primarily about the over-current function of the RCBO, rather than its residual current function. In other words, what you say would be equally true if the RCBO were an MCB.

Kind Regards, John

 

[Jackrae]

If you can't confuse the issue, where's the fun
All I was endeavouring to illustrate was an application where having two devices in series might be beneficial.

 

[JohnW2]

All I was endeavouring to illustrate was an application where having two devices in series might be beneficial.

Yes, but, in terms of "devices" in general, there is very often no "might" about it - my point was that having two "devices" in series is very often essential - when you have an RCD, then you must have some sort of OPD 'device' (MCB, RCBO or fuse) in series with it.

The discussion is specifically about having two RCDs (with same IΔn) in series - in that case, so long as the 'first' one is functioning satisfactorily, there is (from the residual current point-of-view) never any point in having a second one in series, so the only reason to have a second one (from that POV) would be for redundancy.

I've also mentioned one other reason why one might want to use a RCD socket, plug or FCU on an already-RCD-protected circuit, but that has nothing to do with that socket/plug/FCU being a "residual current device" - something with just 'NVR' functionality would serve the same purpose.

Kind Regards, John

 

[Jackrae]

Surprised no-one asked the OP when they last paid for their 'loved one's" system to be professionally tested and set up a standing order for such tests to be conducted at set intervals over and above those generally considered as adequate.

Or maybe even considered having 10mA RCDs installed on 'high-risk' circuits.

 

[JohnW2]

Surprised no-one asked the OP when they last paid for their 'loved one's" system to be professionally tested and set up a standing order for such tests to be conducted at set intervals over and above those generally considered as adequate.

Yes, a valid suggestion, as maybe yet a further level of reassurance. However, if it is felt that there is a need for RCD redundancy, that perceived need would probably remain even if tests were undertaken fairly frequently (realistically, probably no more often than once or twice per year), since a faulty RCD (which became faulty after the most recent testing) could otherwise remain in service for an appreciable period of time.

Or maybe even considered having 10mA RCDs installed on 'high-risk' circuits.

That's often suggested but one has to remember that an RCD does not limit the magnitude of shock current, and I doubt that shock currents will often be between 10mA and 30mA. In practice, shock currents are likely to be much higher than 30mA, so that even a 30mA RCD will operate. The only real potential benefit is that, for modest shock currents (i.e. not a lot above 30mA), a 10mA RCD should operate a bit more quickly, but even a 30mA is deemed to operate rapidly enough to provide adequate protection.

One also has to factor in the potential hazards that could theoretically result from 'nuisance tripping' with a 10mA device, particularly given the amount of 'leaky' equipment and appliances that are around ... not to mention my personal uncertainties as to how many lives actually have been saved by RCDs, of any IΔn!

Moving from electric shocks to clearing L-E faults before anyone has a chance to get a shock, I again very much doubt that a significant number of such faults result in a fault current under 30mA. Indeed, such a fault current will commonly be thousands of times higher than that (which is what we had to rely upon, for ADS, before we had RCDs).

Kind Regards, John