Spot The Problem

[JohnW2]

If your changing a bulb in a plastic holder and accidentally touch the line pin then the rcbo will trip.

Whilst touching what else?

Presumably the floor? Ideally, a bonded pipe

A bonded pipe would certainly do, but you'd be unlucky if there was one of those in close proximity to a light fitting (particularly a ceiling one, which is what most of them are!).

As for the floor, I don't know what sort of houses people live in, or how they behave in them - maybe they walk about bare-footed on floors which are six inches deep in water. In the world (and house) in which I live, I'm not at all sure that a person wearing shoes and standing on a 'standard floor' would even feel anything if they touched (just) a live conductor, and I am sure that it's extremely (incredibly) unlikely that enough current would flow through them to cause an RCD to operate. I'm obviously not suggesting that anyone (deliberately) does the experiment, but I think it's close to being an urban myth!

Kind Regards, John

 

[Iggifer]

The ESC guide says only to fit an RCD/RCBO if IR tests at less than 1 mΩ

You mean 1MΩ. I presume.

It doesn't say ONLY if


On page 7 it says you should decline replacement of CU if <1M&#937; but ALSO if customer refuses fitting of CPC then fit an RCD.

Then, as you point to, page 9 says if resistance <1M&#937; fit RCD.

Typical.

You're quite right, on both points, I did of course mean 1M&#937; and I clearly skipped over that bit of P7

 

[AndyPRK]

Fair point.

I guess there is a band where someone could get a lethal shock and the rcd not operate.

I'd rather it be fitted though.

 

[Iggifer]

yea of course there is a chance of death/injury with any installation. Just because an RCD is fitted doesn't mean it will necessarily operate when it's needed, and it certainly won't stop you getting a decent belt

Class II accessories and fittings are of course the way to go with no cpc's though

 

[JohnW2]

Fair point. I guess there is a band where someone could get a lethal shock and the rcd not operate.

It's really more than a 'band' - there is never any guarantee that an RCD will prevent a shock proving lethal, no matter what the current through the person. Limiting the duration of currents >30mA through a person to a maximum of 300ms will prevent the shock proving fatal in most healthy people, but that's not a certainty - people vary, as do the routes current takes through their body. Once one considers 'unhealthy' people, all bets are off. There are people walking about (some seemingly totally healthy, with no knowledge of any problem) with hearts which are so electrically unstable that they are constantly at risk of spontaneously developing fatal disturbances of heart rhythm; in such people, just a mA or two of electrical shock can be the 'final straw' which results in their death.

However, that's not really much to do with the point I was making. In practice, it is all-but-impossible for enough current to flow through a person to trip an RCD (whether the shock proves fatal or not) if they are indoors, standing on floor or ladder and only touching an L conductor. The current through them would only be high enough to trip an RCD if there were simultaneously touching L and something (other than neutral) close to earth potential - the risk of which is obviously considerably reduced if there is no earth connection to, or in the vicinity of, a light fitting. Sure, they theoretically could touch an earthed pipe or radiator etc. at the same time as touching the live L of a lighting circuit, but that's incredibly unlikely. Perhaps most important, that very unlikely scenario is just as possible when the lighting circuit does have a CPC, so I don't really see it as a valid argument for fitting an RCD 'to compensate for' the absence of the CPC.

I'd rather it be fitted though.

Having an RCD in any situation is obviously theoretically better than not having it - but, as above, I think that the argument for including one to compensate for the absence of a CPC on a lighting circuit is, in itself, pretty (very?) weak. That's how I see it, anyway.

Kind Regards, John

 

[mfarrow]

including one to compensate for the absence of a CPC on a lighting circuit is, in itself, pretty (very?) weak.

Remember it's there to replace proper fault protection in the event of failure of basic insulation.

So, whilst the risk of someone touching something live is low in a sound circuit, with a live metal chandelier 2" above their heads, it will be more certain.

Having said that, I cannot think but wonder why 30mA is still chosen. There is very little leakage current on lighting circuits, unlike appliances, and therefore perhaps 10mA is more appropriate.

Most shocks I've had are live only, but I have had live to earth shocks through my hand. None has ever operated an RCD.

 

[JohnW2]

including one to compensate for the absence of a CPC on a lighting circuit is, in itself, pretty (very?) weak.

Remember it's there to replace proper fault protection in the event of failure of basic insulation. ... So, whilst the risk of someone touching something live is low in a sound circuit, with a live metal chandelier 2" above their heads, it will be more certain.

True, but also remember that, in the absence of a CPC connected to that metal chandelier, an RCD cannot provide 'additional protection' in the normal sense - the RCD will only operate if a path to earth external to the fitting (e.g. current through a person to something else at earth potential) arises.

Having said that, I cannot think but wonder why 30mA is still chosen. There is very little leakage current on lighting circuits, unlike appliances, and therefore perhaps 10mA is more appropriate.

That certainly used to be true, but one could obviously only take advantage of that if the lighting circuit had its own RCBO, since otherwise there would be leakage current from other circuits (and I'm not sure how easily obtainable, and/or affordable, 10mA RCBOs are). I say 'used to be true' because the growing appearance of switched mode power supplies on lighting circuits is presumably increasing the scope for leakage.

Most shocks I've had are live only, but I have had live to earth shocks through my hand. None has ever operated an RCD.

As I've commented before, I have not yet heard of a single case in which a shock has resulted in an RCD operating, although I guess it must have happened occasionally. I still don't believe that, in practice, truly 'live only' shocks (indoors) are ever going to trip a 30mA RCD, and probably not a 10mA one (which would still require a <23k&#937; 'incidental' path from the person to earth).

Kind Regards, John

 

[sparkwright]

Do we know if the metal switches and light fittings were fitted with an earth wire 'borrowed' from another circuit?

 

[JohnW2]

Do we know if the metal switches and light fittings were fitted with an earth wire 'borrowed' from another circuit?

In the case of Iggifer's OP, I don't think we know whether there are any metal switches or light fittings (earthed or not) - the discussion has been mainly about Iggifer's initially intended point, that to protect a lighting circuit with an RCBO to 'compensate for' the absence of a CPC in the circuit's wiring is (regardless of what ESC may think/say) pretty pointless, and certainly not anything approaching a proper (or even, it itself, acceptable) approach to the issue!

Kind Regards, John

 

[sparkwright]

He said later there were metal fittings.

 

[JohnW2]

He said later there were metal fittings.

Ah, so he did However, I don't think that whether or not they were earthed, from somewhere (which I very much doubt) has any real bearing on the question of whether adding the RCBO 'because of the absent CPC' would make a lot of sense. Adding any RCBO/RCD to any circuit obviously has the potential to 'improve safety' (to whatever extent!), but that's not what we're talking about.

Kind Regards, John

 

[sparkwright]

Yeah, I just thought I'd 'throw it in'.

 

[JohnW2]

Yeah, I just thought I'd 'throw it in'.

Fair enough. I suppose that, if one wanted to get pragmatic/philosophical about it, one could wonder whether, in practice, earthing the faceplate of a metal switch actually increases or decreases the net risk of a serious electric shock....

...so long as the faceplate is in place, earthing it certainly should prevent it becoming live (e.g. due to a trapped conductor or one which has come out of its terminal) for more than a very brief period of time, since a protective device should then operate. However, the chances of a serious shock resulting from that is usually going to be very small, because that would require simultaneous contact with something close to earth potential. On the other hand, once someone has (stupidly) taken the faceplate off without properly isolating the circuit, the presence of an earhed metal faceplate (probably being held in one hand, whilst the other hand 'meddles'!) surely must considerably increase the risk of a serious L-E through-body shock?

I'm obviously not suggesting that we shouldn't earth such switches, since the idea of having live faceplates (without any reason to know that they are live) is clearly not very nice at all. However, per above, it wouldn't surprise me at all if, in practice, that earthing actually increased the net number of serious shocks! ... one of those ironies!

Kind Regards, John

 

[mfarrow]

in the absence of a CPC connected to that metal chandelier, an RCD cannot provide 'additional protection' in the normal sense.

Failure of basic and/or fault protection is exactly what 'Additional Protection' by means of a 30mA RCD is there for.

 

[JohnW2]

in the absence of a CPC connected to that metal chandelier, an RCD cannot provide 'additional protection' in the normal sense.

Failure of basic and/or fault protection is exactly what 'Additional Protection' is there for.

Perhaps I should have been clearer (I was trying to be brief ).

When I wrote of " 'additional protection' in the normal sense", what I was talking about was that, in the normal situation of a Class I fitting/accessory with earthed exposed-conductive parts, a fault between L and that exposed-c-p of too high an impedance to result in operation of an OPD would usually result in an RCD (if there was one!) operating, removing the hazardous potential from the exposed-c-p, hopefully before anyone had a chance to touch it.

As I went on to say (in the rest of my sentence, which you didn't quote), in the absence of a CPC, the only way in which an RCD could/would operate was if a path from L to earth arose external to the fitting/accessory (e.g. through a person) - which, in practice, would be a pretty low risk. The exposed-c-p would, however, always be 'live'.

Kind Regards, John