Earth leakage on sockets

[captaincaveman]

This has messed my head up so much I’m not sure I can even explain it properly but hopefully someone out there may give me ideas of what to do next…


Had a call out to a house today with the owner (polish if it has any relevancy) saying he was getting small shocks off taps & was getting a reading on his socket tester which showed green green red which doesn’t have an explanation on the tester and the RCD was not tripping.

It is an 8 circuit setup over duel RCD. The owner said he had different mcbs set up for different sockets (1 for tv, 1 for comp room, 1 for lounge etc).

Although the ccu had only been in a year it had NO labelling (inc circuits), no installation certificate etc…

I tested the socket ring (32A) end to end r1, r2 & rn but with no continuity suggesting it was actually 2 radials.

I tested continuity (all unplugged) and found r1 & rn = 10 ohms, rn-r2 = 0.56 ohms, r1-r2 <999.

Checked sockets (which I had to break two of cos they were tiled in with broken screws) & found one had L/N reversed.
Found a socket with four spurs off it (to boiler switch, lounge socket*, upstairs, utility cupboard). The boiler switch had three spurs off it (to boiler, utility lights, upstairs socket). * which then had a spur to a lounge wall light

I did a alternative IR test on the boiler which failed (I believe due to the bonding becoming live)

I isolated the ‘faulty’ radial at ccu at L,N & E and tested the other sockets which showed the earth still live (50V). I isolated each & every circuit at a time at the mcbs but still got 50V on the other sockets. There was 50V getting back to the earth so I removed each cpc one at a time but still got 50V going back to the socket. Did the same with neutral & got the same result.

Alas I couldn’t think clearly any more & needed to eat so have left it for now.

I have 4 ideas at the moment:
1) two way landing lights using the cpc as live conductor. But would this affect the neutral?
2) another terminal wired with wrong polarity. But this wouldn’t make the cpc live? 3) A big screw going through the L/N & E
4) Water infiltration under the floor into a junction box.

Anyone come across anything similar with any tips and can save me having to test every circuit individually? It’ll take me ages just to label which circuit is which – especially as they seem to have this fault linking most of them together.

p.s. I replaced the RCD the ‘ring’ was on which wasn’t tripping, even at the button.
When I went to test the RCD I kept getting CON on my Megger (“wrong connection to instrument”)

 

[davesparks]

What did you use to test the voltage?

 

[Marvo]

This has messed my head up so much I’m not sure I can even explain it properly but hopefully someone out there may give me ideas of what to do next…


Had a call out to a house today with the owner (polish if it has any relevancy) saying he was getting small shocks off taps & was getting a reading on his socket tester which showed green green red which doesn’t have an explanation on the tester and the RCD was not tripping.

If he's getting shocks there's 2 faults. Firstly there's something floating outside the equipotential zone and not bonded, secondly there's probably an IR fault causing an earth leakage current assuming it's not functional leakage.

It is an 8 circuit setup over duel RCD. The owner said he had different mcbs set up for different sockets (1 for tv, 1 for comp room, 1 for lounge etc).

Although the ccu had only been in a year it had NO labelling (inc circuits), no installation certificate etc…

I tested the socket ring (32A) end to end r1, r2 & rn but with no continuity suggesting it was actually 2 radials.

I tested continuity (all unplugged) and found r1 & rn = 10 ohms, rn-r2 = 0.56 ohms, r1-r2 <999.

These test results make no sense.
I'd suggest you sort out the circuit first by either reestablishing the ring or making it into 2x radials on separate MCB's. Once you've done this then retest.

Checked sockets (which I had to break two of cos they were tiled in with broken screws) & found one had L/N reversed.
Found a socket with four spurs off it (to boiler switch, lounge socket*, upstairs, utility cupboard). The boiler switch had three spurs off it (to boiler, utility lights, upstairs socket). * which then had a spur to a lounge wall light

Sounds like you're going to be busy.

I did a alternative IR test on the boiler which failed (I believe due to the bonding becoming live)

I don't understand how bonding can become live. Fix the bonding.

I isolated the ‘faulty’ radial at ccu at L,N & E and tested the other sockets which showed the earth still live (50V). I isolated each & every circuit at a time at the mcbs but still got 50V on the other sockets. There was 50V getting back to the earth so I removed each cpc one at a time but still got 50V going back to the socket. Did the same with neutral & got the same result.

Instead of chasing phantom voltages on live circuits and disconnecting CPC's which is probably going to result in someone getting a whack, rather concentrate on each circuit one by one, do the dead tests and fix the issues they show up.

Alas I couldn’t think clearly any more & needed to eat so have left it for now.

I have 4 ideas at the moment:
1) two way landing lights using the cpc as live conductor. But would this affect the neutral?
2) another terminal wired with wrong polarity. But this wouldn’t make the cpc live? 3) A big screw going through the L/N & E
4) Water infiltration under the floor into a junction box.

Anyone come across anything similar with any tips and can save me having to test every circuit individually? It’ll take me ages just to label which circuit is which – especially as they seem to have this fault linking most of them together.

p.s. I replaced the RCD the ‘ring’ was on which wasn’t tripping, even at the button.
When I went to test the RCD I kept getting CON on my Megger (“wrong connection to instrument”)

You're running yourself around in circles.

No, there's no quick way out. Do the dead tests and fix the issues on each circuit one by one. Check and test the bonding.

 

[PEG]

Hi dude,Pretty well covered^,have you any results for test prior to above? Incomer type,earth arrangement etc?

You may have got a bit ahead of yourself...a methodical approach is required,when faced with a myriad of issues,such as described.

Not jibing you,but you mentioned changing an RCD,but did not list the testing on that component.

The customers' nationality is of no consequence whatsoever. His possession and use of a suck-it-an-see,might be...:smartass:

 

[captaincaveman]

Thanks for your input guys.

To try and answer your points:

Davesparks, I tested with a voltage tester & also my megger.

Marvo,
P1) Quite possible
P2) I did separate the ‘ring’ onto 2x 20A mcbs & no, the readings made no sense (but nothing in that installation did).
P3) I’m already too busy, ergo me working on a Sunday at 5pm in the dark lol.
P4) I’m thinking a cpc has become live & working back to the ccu & bonding (then tracing back to the boiler?)
P5,6) I, of course, put the cpcs back ensuring no one was able to interfere while I tested. And you are right. I told him I would have to test each circuit one at a time (which will take a while, especially as there are so many radials & downlights and nothing is labelled)

Peg,
It’s tncs, 1361.
I tried to do it methodically but was coming across so many issues it kept taking me back to square one (and into the darkness – god, I hate winter!).
I tried doing an RCD trip test ½ & 1x but my megger was coming up ‘con’ which is wrong connection to instrument. I tried doing the RCD test button which did nothing so I swapped the RCDs around resulting in the test button now working (but not on the supposedly faulty one still - which I have swapped for a spare).
Nothing against poles, I just got the feeling his mate (if not him) had done the ccu & wondered if they had a habit of doing certain practices that might help explain this.
I guess it's back to square one testing each circuit. If he had this done on the cheap he's going to regret it now.

Just to add. He has two 2.5mm radials, one to his tv & one to his computer upstairs. Both on 32A.

Just something else I had to deal with ontop of the earth fault, the socket wired with reverse polarity, the faulty RCD, sockets with broken screws and glued in, 9 unmarked circuit (6 of which are sockets) & the multiple spurs off 1g sockets :dizzy2:

Incidentally, I tried identifying some circuits with a circuit finder but it was indicating several circuits including lights. Any ideas why this may be?

cheers again

 

[Deleted member 9648]

I think you are getting carried away with it. You are there to find out why the owner is getting shocks. That will mean verifying earth continuity on each circuit and then IR testing each circuit,earthing and bonding etc. All the other issues you come across such as broken rings etc are incidental and should be recorded and made known to the client but you are under no obligation to fix or investigate.

 

[rompling]

I think his nationality might be an issue - sounds like some of his countrymen maybe installed the circuits - part p they don't give a hoot
I have seen it loads of times especially in London.
Not racist but just my observations

As others have said start at the beginning and break it down into sections of testing , otherwise you can be chasing your tail
Don't forget to tell the customer this us going to be expensive - saves on the startled face when you present the bill !!

 

[Farmelectrics]

maybe recommend an eicr be carried out

 

[ruston]

Hi dude,Pretty well covered^,have you any results for test prior to above? Incomer type,earth arrangement etc?

You may have got a bit ahead of yourself...a methodical approach is required,when faced with a myriad of issues,such as described.

Not jibing you,but you mentioned changing an RCD,but did not list the testing on that component.

The customers' nationality is of no consequence whatsoever. His possession and use of a suck-it-an-see,might be...:smartass:

Yes, most customers have a socket and see.

 

[PEG]

Yes, most customers have a socket and see.

Hi fella,my mention of a "suck-it-an-see",was not a spelling error....it's just that we are very close...:28:

 

[ruston]

Hi fella,my mention of a "suck-it-an-see",was not a spelling error....it's just that we are very close...:28:

Lol, I knew what you were getting at, I/we have cocked this bodge up...........................:dunce2:

 

[telectrix]

This has messed my head up so much I’m not sure I can even explain it properly but hopefully someone out there may give me ideas of what to do next…


Had a call out to a house today with the owner (polish if it has any relevancy) saying he was getting small shocks off taps & was getting a reading on his socket tester which showed green green red which doesn’t have an explanation on the tester and the RCD was not tripping.

It is an 8 circuit setup over duel RCD. The owner said he had different mcbs set up for different sockets (1 for tv, 1 for comp room, 1 for lounge etc).

Although the ccu had only been in a year it had NO labelling (inc circuits), no installation certificate etc…

I tested the socket ring (32A) end to end r1, r2 & rn but with no continuity suggesting it was actually 2 radials.

I tested continuity (all unplugged) and found r1 & rn = 10 ohms, rn-r2 = 0.56 ohms, r1-r2 <999.

Checked sockets (which I had to break two of cos they were tiled in with broken screws) & found one had L/N reversed.
Found a socket with four spurs off it (to boiler switch, lounge socket*, upstairs, utility cupboard). The boiler switch had three spurs off it (to boiler, utility lights, upstairs socket). * which then had a spur to a lounge wall light

I did a alternative IR test on the boiler which failed (I believe due to the bonding becoming live)

I isolated the ‘faulty’ radial at ccu at L,N & E and tested the other sockets which showed the earth still live (50V). I isolated each & every circuit at a time at the mcbs but still got 50V on the other sockets. There was 50V getting back to the earth so I removed each cpc one at a time but still got 50V going back to the socket. Did the same with neutral & got the same result.

Alas I couldn’t think clearly any more & needed to eat so have left it for now.

I have 4 ideas at the moment:
1) two way landing lights using the cpc as live conductor. But would this affect the neutral?
2) another terminal wired with wrong polarity. But this wouldn’t make the cpc live? 3) A big screw going through the L/N & E
4) Water infiltration under the floor into a junction box.

Anyone come across anything similar with any tips and can save me having to test every circuit individually? It’ll take me ages just to label which circuit is which – especially as they seem to have this fault linking most of them together.

p.s. I replaced the RCD the ‘ring’ was on which wasn’t tripping, even at the button.
When I went to test the RCD I kept getting CON on my Megger (“wrong connection to instrument”)

has same green green red on socket tester today. fault causing this was a blown cable ( L-E short which had melted the cpc and the L wasn't much better). still got no earth on several sockets but socket tester show 3 greens. moral is not to trust a socket tester. they're only a basic tool. there's 55V on the cpc of 6 sockets till we dig up bedroom floor tomorrow. ( thee duff leg is isolated atm, btw).

 

[PEG]

has same green green red on socket tester today. fault causing this was a blown cable ( L-E short which had melted the cpc and the L wasn't much better). still got no earth on several sockets but socket tester show 3 greens. moral is not to trust a socket tester. they're only a basic tool. there's 55V on the cpc of 6 sockets till we dig up bedroom floor tomorrow. ( thee duff leg is isolated atm, btw).

Which backs up something i was told as a lad...if something/someone is not telling you the FULL story...they could be telling you ANY story :blush5:

Although it is surprising,the number of supposedly "authority" institutions,who routinely send out their operatives,with such devices.

 

[rich.250]

Is it TNCS by any chance?
Possible lost supply neutral and bonding has effectively become the neutral.....??
This may be completely out as you haven't mentioned proving the supply 1st.
Start from the beginning.

 

[captaincaveman]

It is tncs rich.250. Interesting point & one I haven't come across yet. I did do a polarity check on the incoming which I don't think showed anything wrong if memory serves. I will double check it when I go back tho. Where would the Neutral be lost? It would have to be between the incomer & neutral bar wouldn't it? cheers

 

[captaincaveman]

How did you get on Telectrix?

 

[rich.250]

It is tncs rich.250. Interesting point & one I haven't come across yet. I did do a polarity check on the incoming which I don't think showed anything wrong if memory serves. I will double check it when I go back tho. Where would the Neutral be lost? It would have to be between the incomer & neutral bar wouldn't it? cheers

If you loose a neutral it will normally be a connection on a overhead line.
The exact reason most DNOs will refuse to provide an earth now on overhead supplies.
Making nearly every new supply in my area TT!!

 

[captaincaveman]

.... SO, I managed to get back to have a look at it today.

I did dead testing on all circuits (which wasn't easy as the house has radials criss-crossing to sockets all over the place & hidden sockets in cupboards conceiled with clothes etc, ) and all the circuits tested ok. I did however find doing a alternative IR test (LN-E 250v) on the boiler resulted in <1Mohm but this was disconnected during other tests anyway.

I'm still getting up to 40V on the earth though. I observed it starts as 0 with all circuits dead and goes up steadily to 40V as each circuit is reenergised.

Any suggestions my good fellows?

 

[HandySparks]

I'm still getting up to 40V on the earth though. I observed it starts as 0 with all circuits dead and goes up steadily to 40V as each circuit is reenergised.

40V between which two points exactly?

 

[PEG]

Hi,when you say "on the boiler",do you mean the fixed wiring to the boiler,with boiler disconnected,or.....

Also,what were you measuring between,when you observed the increasing voltage to earth? Ta

 

[captaincaveman]

40V between which two points exactly?

Neutral & earth

 

[HandySparks]

Neutral & earth

Yes, but where? Distribution board, accessory, where?

 

[captaincaveman]

Yes, but where? Distribution board, accessory, where?

Sorry, I'm tired. ...Through all socket points tested & also using a voltage pen on the cpc teminal in the ccu

 

[HandySparks]

Sorry, I'm tired. ...Through all socket points tested & also using a voltage pen on the cpc teminal in the ccu

OK, are you measuring 40V between the DB earth bar and the neutral bar?

If so, then there's a high resistance somewhere in the cabling or components between those two bars and the supplier's head where they should be joined (as this is supposed to be TN-C-S).

 

[captaincaveman]

OK, are you measuring 40V between the DB earth bar and the neutral bar?

If so, then there's a high resistance somewhere in the cabling or components between those two bars and the supplier's head where they should be joined (as this is supposed to be TN-C-S).

As in the main switch, Neutral tail or main cut out connection?

 

[HandySparks]

As in the main switch, Neutral tail or main cut out connection?

Yes, if you have a significant voltage between N and E in the board when under load.

Worth checking for volt drop between incoming neutral terminal on main switch and neutral busbar.

I'd be checking tightness of terminals on main switch, link to neutral busbar, and anything else you can get at that isn't sealed up. Careful of any parts that aren't isolated, of course.

 

[PEG]

Hi,so you have a potential,between your probes,a very short distance from where those two points, are one and the same.

If you draw that scenario out,identifying the parts concerned,a clearer view of what may be occurring,will reveal itself :smug:

 

[rich.250]

I'm sticking by my previous conclusion.
I'm thinking if the bonding was disconnected you would have 230v between the bars as I don't think you have a neutral.
Turn the main switch off and do a l-n loop test on the incoming side.
That will soon tell you.
Rather than guessing at different voltages.

 

[captaincaveman]

Yes, if you have a significant voltage between N and E in the board when under load.

Worth checking for volt drop between incoming neutral terminal on main switch and neutral busbar.

I'd be checking tightness of terminals on main switch, link to neutral busbar, and anything else you can get at that isn't sealed up. Careful of any parts that aren't isolated, of course.

Cheers for that handysparks. I will investigate that. The whole installation is a bit of a joke so it wouldn't surprise me if it were the case of oversheared tails or untightened terminals (although I did check all the ones on the N bar & RCDs, I didn't check at the main switch).

Cheers again for everyones input. Appreciated

 

[HandySparks]

I'm sticking by my previous conclusion.
I'm thinking if the bonding was disconnected you would have 230v between the bars as I don't think you have a neutral.
Turn the main switch off and do a l-n loop test on the incoming side.
That will soon tell you.
Rather than guessing at different voltages.

If the supply neutral is lost, how would that give rise to a significant voltage between neutral and earth bars in the DB, given that they're joined at the head?

 

[captaincaveman]

I'm sticking by my previous conclusion.
I'm thinking if the bonding was disconnected you would have 230v between the bars as I don't think you have a neutral.
Turn the main switch off and do a l-n loop test on the incoming side.
That will soon tell you.
Rather than guessing at different voltages.

Sorry rich. just seen this. I will put it on my list to check. cheers