Earth leakage on high loads

[meerkatmag]

Ive been testing and have found lots of strange behaviour that I don't understand.

1. The lighting circuit and main socket ring seem to be connected somehow. When lighting circuit is on about 3v can be detected neutral/earth on socket ring. Goes to zero if I trip lighting breaker Same the other way round, with ring on and lights off there's a voltage on lighting neutral/earth. This is while all neutrals are disconnected from cu. Earths still connected. So I think there is a connection between live and earth and voltage is flowing back from earth to all the neutrals.
2. With power off and every device unplugged. I have a measurable resistance between neutral and earth everywhere on my ring. Slightly lower resistance on the middle floor. Ive opened every socket and can't find a smoking gun. I have separated the ring and both sides still have earth to neutral connection with slightly different resistance. If I turn lighting circuit on/off it affects the resistance on the socket ring. When I have more time i will spend longer disconnecting more parts of the ring and see if i can find where it's coming from.

 

[Avo Mk8]

Ideally testing really requires an understanding of the circuit under test, and a plan of what to measure where. I'm finding it a little difficult to unravel your findings, so grateful if you could clarify:

1. Are you saying that with both ring neutrals disconnected from the neutral bar in the CU, there is a voltage between those neutrals to earth when the lighting circuit is on, but not when lighting is off?
Assuming you are using a typical high impedance multimeter, that could be explained by capacitative coupling between the two circuits, if the lighting and ring cables run parallel or close to each other somewhere in the house. Often referred to as a 'ghost voltage'. It typically disappears when any load is placed on it. It is not necessarily indicative of a fault.

2. Is your initial statement about measurable resistance made with the relevant RCCB on or off? It would also be helpful if you gave values for the "measurable resistance" you've found, bearing in mind neutral and earth are connected together at supply 'source' (somewhere). Nevertheless there's often a small voltage between the two, depending on demand outside your property, which might affect multimeter resistance readings when the RCCB is on.

If (with everything off) you've taken both ring neutrals out of their bar in the CU, and you still have a very low resistance reading between them and earth, you surely have a fault or cross-connection somewhere on that ring, or something like a defunct immersion heater!

If your testing shows up specific low resistance readings between N and E with the circuit under test completely disconnected, and the cause is not obvious, get an electrician in!

 

[mainline]

Ive been testing and have found lots of strange behaviour that I don't understand.

1. The lighting circuit and main socket ring seem to be connected somehow. When lighting circuit is on about 3v can be detected neutral/earth on socket ring. Goes to zero if I trip lighting breaker Same the other way round, with ring on and lights off there's a voltage on lighting neutral/earth. This is while all neutrals are disconnected from cu. Earths still connected. So I think there is a connection between live and earth and voltage is flowing back from earth to all the neutrals.
2. With power off and every device unplugged. I have a measurable resistance between neutral and earth everywhere on my ring. Slightly lower resistance on the middle floor. Ive opened every socket and can't find a smoking gun. I have separated the ring and both sides still have earth to neutral connection with slightly different resistance. If I turn lighting circuit on/off it affects the resistance on the socket ring. When I have more time i will spend longer disconnecting more parts of the ring and see if i can find where it's coming from.

As @Avo Mk8

You need some proper test gear and someone who can use it.

Otherwise circles come to mind.

 

[meerkatmag]

Firstly thanks for sticking with me, clearly I'm getting to the point where I need an electrician. I think they will suggest a house rewire and I can't really afford it hence trying myself. But ultimately yes I totally get it in the end I will have to get someone really good in to look at this.


To answer previous questions just in case...

1. The ring neutrals are disconnected from the cu but the lighting neutral is also disconnected. Neither circuit had any current flowing without neutrals. But still voltage found its way across to the neutral of the ring, i presume via earth? But i didnt know about ghost voltage and it's only small so perhaps like you say it's within normal range. I do know about capacitive coupling, I thought that was when there was a missing earth?
2. With the ring neutrals and earth disconnected from cu I register in the range of 10 to 20 Mohms between earth and neutrals wires of the ring. I know the earth and neutral are connected at my supply so I have disconnected all the ring wires from cu to test them in isolation. All devices are unplugged but I do have some USB sockets that may be messing with this reading I suppose.

My multimeter isnt a really cheap one, I do work with electronics in my proper job so I do have some experience with the basics. But I don't believe it has a high voltage for resistance testing so it may be insufficient for this.

 

[newfutile]

Any neutral disconnected will become live though a load,maybe even if the load is off.
Capacitive coupling can happen between any conductors close to any other.
We test IR at 500 volts so getting a high reading at a few volts does not mean a lot.

 

[Avo Mk8]

meercatmag
Assuming your 10 to 20 Mohms are Megohms (as you say, measured with low voltage DC), that is probably due to superficial contamination in back boxes or humidity. It's only a matter of concern if a proper IR test at 500V revealed a value lower than one Megohm.
Those figures you report aren't a problem (within the limitations of a low voltage measurement). and would not lead to spurious tripping at a mere 10 to 20+ micro amps!

I would discourage you from roaming around the installation with an ordinary multimeter, you will probably get spurious readings from induced voltages on isolated circuits, and could inadvertently create a problem. If you are worried about the installation, get a competent electrician to do an Electrical Installation Condition Report; you will have the proper numbers from a calibrated tester, and know if there is a real issue. It's unlikely you'll need to rewire the house!

Regarding earth leakage and tripping that you queried in your first post, the evidence to me so far seems to confirm you just have an excess of leakage from small devices and accessories all fed from one RCCB.
The spec for a 30mA RCCB (or RCB) says it should trip somewhere between 15mA and 30mA leakage.
Your post #5 reports a 14mA steady leakage with washing m/c etc.
There seems to be little headroom.
These days such an installation might be wired with the kitchen on its own ring, large appliances on their own radials, probably a separate ring for each floor etc etc, and preferably RCBO's for each circuit. Adopting even minor upgrades to your installation would probably fix the tripping issue.
Just talk to a competent electrician!
Good luck.

 

[meerkatmag]

meercatmag
Assuming your 10 to 20 Mohms are Megohms (as you say, measured with low voltage DC), that is probably due to superficial contamination in back boxes or humidity. It's only a matter of concern if a proper IR test at 500V revealed a value lower than one Megohm.
Those figures you report aren't a problem (within the limitations of a low voltage measurement). and would not lead to spurious tripping at a mere 10 to 20+ micro amps!

I would discourage you from roaming around the installation with an ordinary multimeter, you will probably get spurious readings from induced voltages on isolated circuits, and could inadvertently create a problem. If you are worried about the installation, get a competent electrician to do an Electrical Installation Condition Report; you will have the proper numbers from a calibrated tester, and know if there is a real issue. It's unlikely you'll need to rewire the house!

Regarding earth leakage and tripping that you queried in your first post, the evidence to me so far seems to confirm you just have an excess of leakage from small devices and accessories all fed from one RCCB.
The spec for a 30mA RCCB (or RCB) says it should trip somewhere between 15mA and 30mA leakage.
Your post #5 reports a 14mA steady leakage with washing m/c etc.
There seems to be little headroom.
These days such an installation might be wired with the kitchen on its own ring, large appliances on their own radials, probably a separate ring for each floor etc etc, and preferably RCBO's for each circuit. Adopting even minor upgrades to your installation would probably fix the tripping issue.
Just talk to a competent electrician!
Good luck.

Thanks for your clear and detailed advice!