- Reaction score
- 2,441
One of the oddest jobs I've ever had today. Called by a plumber I know who had attended after a leak through the kitchen ceiling from bathroom above - and got a shock from the bathroom tap (and his 'pen glowed on everything')
Turned up to check and, sure enough, voltage of 16V AC showing on the (plumbed) towel radiator, the basin tap and the bath taps - but not on other pipework. This disappeared when the socket circuit was isolated.
Kewtech pen more or less lit up anywhere in the bathroom, including the walls - and the carpet of the adjacent room!
There was no supplementary bonding in place, and a late 60s metal Wylex board with plug in MCBs, There was gas bonding, but no water bonding - however the incoming water was seemingly in some sort of plastic (Plumber didn't know it's name but has seen it before used in properties of this age)
My first thoughts was that perhaps a junction box was sitting in water...
However the plumber struggled to find any noticeable leak, other than a fairly small but constant run from the shower screen when shower was on...
Lots of head scratching later and discovered the problem, but not one I would have guessed at.
Finally cut a hole in the kitchen ceiling below, to find a damp patch (but not floods of water) right where the cabling ran up in the stud wall to a socket in the adjacent bedroom.
Located the relevant socket in the back of a built in chest of drawers, and the bottom wooden joist of the stud wall was sopping wet. However, the socket itself was bone dry with no sign of water damage. Also no sign of damage to the cable sheath where it ran through the wet wood.
However, it it looks like in some way the flow from the shower had run under the tiles and soaked the joist of the stud wall opposite, enough to allow current to transfer to the heating pipes that also ran through the wall.
This then seemingly transferred to the radiator, and all the other metal fittings in the bathroom.
Interestingly, the basin tap and the radiator were both connected with short plastic tails to copper - I can only assume that the water itself was conductive enough to allow enough of a current to give a noticeable reading. Does show that short 'insulating' pieces don't negate the need for bonding.
Pulling the cabling down into the ceiling below and reconnecting it removed the stray voltage so that clearly was the problem. Whether the leak was resolved or not is fortunately down to the plumber!
It was an emergency job with no warning after a long week, so my brain wasn't fully in gear, but it has raised some interesting issues for me...
My initial thought is that supplementary bonding the pipes in the adjacent boiler cupboard (which should have been done at some point since there is no RCD) would have made this safer, but I'm actually not 100% on that in this particular case...
Since the incoming water is actually in plastic, under current Regs there would be no need for equipotential bonding it - and had supplementary bonding been in place, it's possible that this voltage would have been transferred to the rest of the house, since I imagine the ~16V would not have been enough to trip the MCB.
Am I right in suspecting that?
The best protection would obviously be RCD protection, but it's interesting to note that the circuit that caused the issue in this case is not a bathroom one or even one that ran through a bathroom, so would not necessarily be considered on an EICR as a C2 due to lack of RCD protection (because it covered downstairs sockets in this case it would, but if it had only been upstairs sockets then I'd have considered it a C3.)
I guess it's also possible that the current was below 30mA, but I experienced it myself and it was definitely more than a static tingle.
I've raised the topic of a new consumer unit with the elderly houseowner and she seemed keen so hopefully we will resolve that at some point. And the plumber will ensure that the leak is sorted too.
Is this a situation that others have run into? Am I missing something blindingly obvious?
The leak seems to have been generated by the British Gas man servicing the boiler in the morning, and possibly running the shower for a while (which the homeowner rarely uses) and causing the leak this time, but there were tide marks on the stud wallpaper that suggested it had happened before....
Pics to follow if anyone's interested....
Turned up to check and, sure enough, voltage of 16V AC showing on the (plumbed) towel radiator, the basin tap and the bath taps - but not on other pipework. This disappeared when the socket circuit was isolated.
Kewtech pen more or less lit up anywhere in the bathroom, including the walls - and the carpet of the adjacent room!
There was no supplementary bonding in place, and a late 60s metal Wylex board with plug in MCBs, There was gas bonding, but no water bonding - however the incoming water was seemingly in some sort of plastic (Plumber didn't know it's name but has seen it before used in properties of this age)
My first thoughts was that perhaps a junction box was sitting in water...
However the plumber struggled to find any noticeable leak, other than a fairly small but constant run from the shower screen when shower was on...
Lots of head scratching later and discovered the problem, but not one I would have guessed at.
Finally cut a hole in the kitchen ceiling below, to find a damp patch (but not floods of water) right where the cabling ran up in the stud wall to a socket in the adjacent bedroom.
Located the relevant socket in the back of a built in chest of drawers, and the bottom wooden joist of the stud wall was sopping wet. However, the socket itself was bone dry with no sign of water damage. Also no sign of damage to the cable sheath where it ran through the wet wood.
However, it it looks like in some way the flow from the shower had run under the tiles and soaked the joist of the stud wall opposite, enough to allow current to transfer to the heating pipes that also ran through the wall.
This then seemingly transferred to the radiator, and all the other metal fittings in the bathroom.
Interestingly, the basin tap and the radiator were both connected with short plastic tails to copper - I can only assume that the water itself was conductive enough to allow enough of a current to give a noticeable reading. Does show that short 'insulating' pieces don't negate the need for bonding.
Pulling the cabling down into the ceiling below and reconnecting it removed the stray voltage so that clearly was the problem. Whether the leak was resolved or not is fortunately down to the plumber!
It was an emergency job with no warning after a long week, so my brain wasn't fully in gear, but it has raised some interesting issues for me...
My initial thought is that supplementary bonding the pipes in the adjacent boiler cupboard (which should have been done at some point since there is no RCD) would have made this safer, but I'm actually not 100% on that in this particular case...
Since the incoming water is actually in plastic, under current Regs there would be no need for equipotential bonding it - and had supplementary bonding been in place, it's possible that this voltage would have been transferred to the rest of the house, since I imagine the ~16V would not have been enough to trip the MCB.
Am I right in suspecting that?
The best protection would obviously be RCD protection, but it's interesting to note that the circuit that caused the issue in this case is not a bathroom one or even one that ran through a bathroom, so would not necessarily be considered on an EICR as a C2 due to lack of RCD protection (because it covered downstairs sockets in this case it would, but if it had only been upstairs sockets then I'd have considered it a C3.)
I guess it's also possible that the current was below 30mA, but I experienced it myself and it was definitely more than a static tingle.
I've raised the topic of a new consumer unit with the elderly houseowner and she seemed keen so hopefully we will resolve that at some point. And the plumber will ensure that the leak is sorted too.
Is this a situation that others have run into? Am I missing something blindingly obvious?
The leak seems to have been generated by the British Gas man servicing the boiler in the morning, and possibly running the shower for a while (which the homeowner rarely uses) and causing the leak this time, but there were tide marks on the stud wallpaper that suggested it had happened before....
Pics to follow if anyone's interested....
