Plumbing and earthing

[Dan S. MacAbre]

I've a question about the earthing connections on our domestic plumbing. I can see that there is a wire in parallel to our water meter, but I'm not sure that this can be the main earth because the supply pipe is plastic. I can also see a wire going into the wall under the sink, so my assumption here is that it mainly uses the electrical earth. But there are speedfit connectors under there, which I expect are not conductive, so there are parts (e.g. pipes feeding taps) that are isolated from the rest. I'm not worried about this (although perhaps I should be?), since all this was supposedly done by professionals; but I thought I'd ask here if every single section of pipe is meant to be individually earthed, or if that would not really be practical. Does it even make sense to earth, for example, braided tap tails?

[John Rumm]

Generally nothing to worry about, if all we are discussing is equipotential bonding.

Given many of them seem to flounder a little on this particular subject, that is not always an assurance of anything!

Short answer - no.

Not really.

Firstly, this is not about earthing as such, but equipotential bonding - different things, with different objectives that work in different ways. However they do often use the same green/yellow insulated wire, and there are situations where the systems interact.

In its simplest terms, earthing is designed to prevent electrical fires, and limit the maximum duration of a an electric shock.

Equipotential bonding is designed to limit the magnitude of an electric shock (i.e. reduce the current that can flow into the victim, by limiting the voltage (or potential difference) across their body).

EQ bonding is used in places where you are likely to be more unacceptable to the influence of a shock (e.g. when wet / naked). The primary question ask when deciding what needs bonding, is "Is this conductive part actually capable of introducing a potential into the area". So a length of metal water pipe that comes from elsewhere in the house to your bath tap, certainly could, and hence needs bonding. A short section of copper "show work" pipe on the end of a feed in plastic, does not, since it can't bring a potential in from outside the zone.

Have a read through this for a fuller explanation:

[Robin]

Your water pipes shouldn't be used to earth your wiring (though they commonly were in the past). You should have something separate at your consumer unit/fusebox.

What I think you are looking at is Main Bonding and Supplementary Bonding. I think you may well be OK on both but suggest you look at

and ask again if you still have questions - when people who know better than me can help.

As for earthing, you can see what things typically look like at

If that leaves you with questions about just what you've got it'd help if you could share [photos somewhere

[Dan S. MacAbre]

Well, I can tell from some of things I've found that they weren't the best available.

Brilliant. Thanks.

[Dan S. MacAbre]

I'm assuming it is the other way around - the mains earth is being used to earth the water pipes. The wire is coming out of original plaster (whoever did the kitchen didn't plaster over anything that wasn't on show), so it has been there since the mid 80's.

Thanks for the links. I'll read them, but I'm not going to worry about what's going on here.

[newshound]

I used to have a kitchen sink that would sometimes give you a tingle. It didn't after I added bonding, but neither did that trip the RCD. I never found out what was causing it.

[Dan S. MacAbre]

I had a metal light switch that did that. I found that the earthing wire in the lighting daisy-chain (they hadn't bothered to close the loop at one of the ends) wasn't connected in one of the roses, which meant that the rest of them further down the chain weren't earthed. I measured an induced voltage of 70V on the faceplate. Fixing the non-connected earth wire solved the problem. It only gave a 'tingle', and was soon discharged, but it's surprising what you can find in a house when you start digging :-)

[Jack Harry Teesdale]

Very true, one house i bought of early 1970's vintage on a large housing development, had the ring main wired into 2 different fuses in the same fusebox (yes they were fuses).

I wondered why i couldn't isolate the ring main by withdrawing just the one fuse.

In converstion later with a next door neighbour he said he had exactly the same problem!

[John Rumm]

There was a time that it was not uncommon to join two 15A radial circuits like this to make a single 30A ring.

[John Rumm]

Lighting circuits are rarely wired as ring, so it would be very unusual to have the loop closed.

Yup, floating wires can pic up stray induced voltages. IME It only takes a *tiny* current to pass to let you sense it when you touch. (even if you don't get a shock or tingle, you can often feel a kind of 50Hz vibration as you lightly slide your finger tips over the surface of the metal).

[Dan S. MacAbre]

Interesting - I thought it was an oversight.

Yes, it was generally felt when feeling for the switch in the dark. I sometimes thought it was just the feeling of the brushed-effect metal. It was only when I measured a voltage there (and it immediately dropped to zero through the meter) that I decided to do something about it.

[Jack Harry Teesdale]

I don't think that was the intention in this case. I think it was wired by an older sparks who didn't really understand the theory of the 30A ring main.

[RJH]

+1. About as close as I'm going to get to understanding :-)

[ARW]

Especially if you buy Fred West's old house.

[David Wade]

The idea was not to provide an earth but to tie all exposed metal to be at the same potential (voltage). Google "equipotential bonding".

However the wiring regulations changed at the beginning of this year and such things may not be required where there are sections of plastic pipe.

Dave

[tabbypurr]

What matters is that you've got an effective earth. Is there such a thing from the incomer next to the fusebox? If not, expect to truly need to sort it. If there is then the waterpipe connections are going to only be equipotential bonding, which is a minor safety matter.

NT

[Adam Funk]

I mentioned here a few years ago that in the process of removing an electric shower, I found my volt stick lighting up at the load end of the only 10 mm^2 cable, which I was absolutely sure was disconnected close to the CU. (Of course, I double- & triple-checked.)

The only explanation was current induced from the socket circuit cable that ran adjacent to the shower cable for several metres.

[ARW]

If required then the absence of main equipotential bonding is a code 2 on an EICR.

[tabbypurr]

Yet in terms of actual safety, the benefit is minimal.