All-RCBO CUs - because we can?

[JohnW2]

The internal connections are not sheathed, just single insulated singles. ... EDIT: I suppose a third way would be to replace the internal cables with insulated and sheathed singles.

Not 'a third way' - the 'first' way that I mentioned at the start, and which you now seem to have grasped ...

... as I said at the start, if they are not worrying about the absence of 'fault protection' of the incoming tails "because they are insulated and sheathed" (which to my mind is nonsense) then they ought to be equally (nonsensically) satisfied by simply using insulated-and-sheathed conductors between main switch and RCD (rather than the 'solutions' they suggested).

However, as said, that seems like total nonsense to me. In the absence of an upstream (time-delayed) RCD, neither the conductors on the supply side nor load side of the main switch will have any 'fault protection' (in the usual and regulatory sense) =- and that remains true whether those conductors have one, two or 20 layers of covering!

As I said before, thankfully this is just the OSG speaking, not the regs!

Kind Regards, John

 

[SimonH2]

In terms of protection against shock, insulated and sheathed tails are deemed OK. If they weren't then we'd not be allowed to do much with any insulated and sheathed cable
We both agree that if the internal connections were replaced with insulated and sheathed singles then there'd be no problem - but I strongly suspect that few people would pick up on that and leave the manufacturer installed non-sheathed connections in place. As previously pointed out, failure of that single line of defence would make the entire installation live.

At which point we circle round to TN-C-S and failure of the PEN

 

[JohnW2]

In terms of protection against shock, insulated and sheathed tails are deemed OK. If they weren't then we'd not be allowed to do much with any insulated and sheathed cable

Indeed so - but as I keep saying, that has got absolutely nothing to do with 'fault protection' (or its absence) which is what the OSG was citing as 'the problem'.

If the OSG's only concern were that the conductors between main switch and RCD would normally only be single-insulated, why did they not say that, and suggest that the solution would be to use insulated-and-sheathed conductors for that pourpose?

We both agree that if the internal connections were replaced with insulated and sheathed singles then there'd be no problem

Indeed - that's what I initially said/implied, and you now agree.

... - but I strongly suspect that few people would pick up on that and leave the manufacturer installed non-sheathed connections in place.

As above, they could simply have suggested that explicitly - surely simpler (and much cheaper) than their actual suggestion of removing the main switch and replacing it with a TD RCD?

At which point we circle round to TN-C-S and failure of the PEN

If the installation is satisfactory, with all bonding in place (and intact), failure of a TN-C-S PEN should not lead to any shock hazard within the (equipotential zone) premises. Only if the main bonding 'melted' or someone were 'half in and half out' of the equipotential zone would there be any shock hazard.

Kind Regards, John

 

[ban-all-sheds]

As previously pointed out, failure of that single line of defence would make the entire installation live.

As evinced by your recent POTW, a failure of anything might not be necessary - poor workmanship could easily be enough.

 

[SimonH2]

If the installation is satisfactory, with all bonding in place (and intact), failure of a TN-C-S PEN should not lead to any shock hazard within the (equipotential zone) premises. Only if the main bonding 'melted' or someone were 'half in and half out' of the equipotential zone would there be any shock hazard.

Which should also be true of a TT installation.

But leads me on to something I've been thinking about at church. There's a metal oil pipe comes into the boiler room which I don't think is bonded. It's going to be earthed anyway as there's metallic elements all the way to the burner which is earthed. I'm thinking it should be bonded. But, outside it connects to a big metal tank sat on lumps of wood - and I'm wondering about loss of PEN conductor, particularly as it's overhead singles in the overhead supply. Do the nice damp blocks of wood make a good earth, or does the tank become live

As evinced by your recent POTW, a failure of anything might not be necessary - poor workmanship could easily be enough.

Indeed, and I can't help thinking that was in the mind of the OSG authors - but they weren't allowed to say it. They do specifically say that the risk of the rigid busbar in and all-RCBO board coming loose would be minimal - I guess someone would notice all the non-functioning circuits before enough terminals were loose enough for the bar to move.

 

[JohnW2]

Which should also be true of a TT installation.

Well, it would be true of a TT installation IF a TT installation had a "TN-C-C PEN" which could fail (which is what you were talking about, and I was responding to) - which it obviously doesn't!

But leads me on to something I've been thinking about at church. There's a metal oil pipe comes into the boiler room which I don't think is bonded. It's going to be earthed anyway as there's metallic elements all the way to the burner which is earthed. I'm thinking it should be bonded. But, outside it connects to a big metal tank sat on lumps of wood - and I'm wondering about loss of PEN conductor, particularly as it's overhead singles in the overhead supply. Do the nice damp blocks of wood make a good earth, or does the tank become live

We've discussed this, and variants thereof, before.

Whether or not, per regs, such a pipe requires main bonding depends on one's view as to whether it constitutes an extraneous-c-p. ELFI would (if one could disconnecting the pipe from 'incidental earthing) probably advise finding out by measuring the resistance from the pipe to MET. However, he would probably make that measurement when the wood blocks were dry (and conclude that it was not an extraneous-c-p, and therefore didn't need bonding) - at which point I would ask him how he knew that it would not know whether is would still be "no an extraneous-c-p" when the wood blocks were soaking wet!

In practice, I suspect that even 'soaking wet' blocks of wood would probably not provide a low enough impedance to earth to be able to result in a 'serious' electric shock (if pipe were touched simultaneously with something live), particularly if the wood blocks were not sitting directly oi bare earth - but I obviously cannot be sure of that.

If, by whatever process, one decides that bonding of the pipe is necessary, then, in a TN-C-S installation, one has the same potential hazard as with an outside tap - and the only real solution to that is to put an 'insulating section' into the pipe where it enters the building.

In fact, since (as in your case) the pipe will very often be 'incidentally earthed', the hazard with TN-C-S (in the event of a PEN failure) exists even if one does not bond it - so, again, if one is concerned about the very small risk, one should electrically isolate the outside pipe from the building's earthing system.

Kind Regards, John

 

[EFLImpudence]

I agree that an insulating section would be the best solution to avoid all the possibilities.

 

[JohnW2]

I agree that an insulating section would be the best solution to avoid all the possibilities.

Indeed. In fact, if it's TN-C-S and one is concerned about the possible outdoor hazard in the event of a 'lost neutral', then, given that the pipe will nearly always be 'incidentally earthed' (even if not bonded) then I think the only solution would be an insulating section (whether we are talking about an oil pipe, and outside tap or whatever).

Kind Regards, John

 

[SimonH2]

Well, it would be true of a TT installation IF a TT installation had a "TN-C-C PEN" which could fail (which is what you were talking about, and I was responding to) - which it obviously doesn't!

I probably wasn't clear. In practical terms, a L-E fault in a TT system with inadequate earthing/other protection would make any "earthed" item live; a loss of PEN conductor in a loaded TN-C-S system would also make any "earthed" item live (MET connected to line via the connected loads and the N-E connection in the cutout).
So from the PoV of users, there is no difference, all the "earthed" metalwork (including the proverbial outside tap) will be live. But other than for "things outside" it should all be "safe" thanks to bonding.

We've discussed this, and variants thereof, before.

I'm sure we have - endlessly

I agree that an insulating section would be the best solution to avoid all the possibilities.

Interestingly that may not be possible. I had a look on OfTec's website and plastic pipe can only be used when buried, and the oil supply cannot come into the building direct from underground but must come up above ground and have a fire valve with remote sensor before it enters the building. So the minimum is a bit of metal pipe sticking out of the wall with a valve on it - just like the outside tap.
So if plastic pipe were used, the metal pipe entering the building wouldn't need bonding, but since it would be incidentally earthed, would present the same hazard to someone outside as an outside tap - with the mitigation that unlike the outside water tap, it's not normal for someone to be touching the fire valve regularly.
In our case, the old steel tank (unbunded so not to current standards, and rusting) right next to the boiler room (also not to current standards) is connected with a rigid metal pipe. Yes, we do have plans (at some point) to replace the steel tank with a bunded plastic one suitably spaced from the building so both can meet current standards. I'll have to discuss with the boiler guy whether we can use plastic pipe so that at least the inside piping isn't an ECP.

I have thought of measuring the resistance to earth - disconnecting the oil supply pipe is easy enough (I'd let the boiler guy do that, possibly when he's next servicing the boiler). The substantial wooden blocks will be near enough permanently wet, and sat on near enough permanently damp soil with a considerable contact area (both to tank and soil). Probably better earthing than many earthing rods Would be interesting to know ...

 

[EFLImpudence]

We didn't say use plastic pipe, but if insulating sections can be fitted to gas pipes, then I presume it must be possible for oil.

 

[JohnW2]

I probably wasn't clear ...

Fair enough. If all you were saying was that, with any earthing arrangment, there is no shock hazard (due to touching two things) within a properly constituted equipotential zone (but could be a risk if that equipotential zone is exported outside of that zone - e.g. to a tap or pipe), then I obviously agree.

Interestingly that may not be possible. I had a look on OfTec's website and plastic pipe can only be used when buried, and the oil supply cannot come into the building direct from underground but must come up above ground and have a fire valve with remote sensor before it enters the building.

There must be a way of doing it (as with gas pipes). I said nothing about plastic pipe. There must be something available - ceramic inserts, perhaps?

Kind Regards, John
Edit: too slow again!

 

[SimonH2]

Looks like the rules for oil pipes are different to those for gas. Different medium, different risks, As I don't know the details of either set of rules then I can't really comment - but it did look rather like oil has to use plastic (only if buried), copper, or non-galved steel. And no completely buried joints - have to be accessible for inspection.
I suspect that latter one is because gas would seep out and eventually someone would smell it - whereas oil would just seep out and cause an environmental problem. Similarly there are requirements for boilers in some situation to have a drip tray and float switch - gas evaporates and disperses, oil will pool and be a long term fire hazard (as well as causing damage to decor etc).