Earth leakage on high loads

[meerkatmag]

I'm new to this forum so apologies if this is in the wrong place. I'm a DIYer, not an electrician. I have had electricians out to look at this issue and they have failed to explain the problems so I've been looking into it myself.

My house has 2 rings, one is very large and covers almost all the plugs in the house, the other is just the kitchen. With the house on as normal, usual devices on and normal day to day running, the earth leakage is very minimal, in the range of 5ma to 10ma in total across the whole consumer unit. However if someone plugs a high load device in, such as a hairdryer or toaster the earth leakage jumps right up to 30ma or 40ma. Sometimes this will trip the RCD, sometimes it doesn't.

We have a computer setup in our loft office with various peripherals, speakers/monitors etc. If we turn all of this on at once a multi plug it almost always trips the RCD. The load when everything is on is minimal, like 2 or 3 amps. But when we turn the plug on the power supplies draw quite a bit of current momentarily and this is enough to trip the RCD pretty much every time. If we switch each peripheral on one at a time slowly the RCD doesn't trip and everything is fine.

If you're wondering how I know the leakage current, I purchase an earth leakage clamp meter and I have a clamp ammeter as well so I've been monitoring both across the two RCDs in my consumer unit. The earth leakage spikes happen across both of my rings, so if I run a kettle on either ring I get a 20ma approx spike which ever ring I plug it into.

If anyone has any ideas I would be very grateful, it's extremely annoying! If you need further info please let me know.
Magnus

 

[mainline]

I'm new to this forum so apologies if this is in the wrong place. I'm a DIYer, not an electrician. I have had electricians out to look at this issue and they have failed to explain the problems so I've been looking into it myself.

My house has 2 rings, one is very large and covers almost all the plugs in the house, the other is just the kitchen. With the house on as normal, usual devices on and normal day to day running, the earth leakage is very minimal, in the range of 5ma to 10ma in total across the whole consumer unit. However if someone plugs a high load device in, such as a hairdryer or toaster the earth leakage jumps right up to 30ma or 40ma. Sometimes this will trip the RCD, sometimes it doesn't.

We have a computer setup in our loft office with various peripherals, speakers/monitors etc. If we turn all of this on at once a multi plug it almost always trips the RCD. The load when everything is on is minimal, like 2 or 3 amps. But when we turn the plug on the power supplies draw quite a bit of current momentarily and this is enough to trip the RCD pretty much every time. If we switch each peripheral on one at a time slowly the RCD doesn't trip and everything is fine.

If you're wondering how I know the leakage current, I purchase an earth leakage clamp meter and I have a clamp ammeter as well so I've been monitoring both across the two RCDs in my consumer unit. The earth leakage spikes happen across both of my rings, so if I run a kettle on either ring I get a 20ma approx spike which ever ring I plug it into.

If anyone has any ideas I would be very grateful, it's extremely annoying! If you need further info please let me know.
Magnus

Have the ring finals been tested properly.

 

[Avo Mk8]

Just thinking aloud - could this suggest a spurious (maybe capacitative?) connection between N and E somewhere on the load side of the CU?
If all the computer setup is off and unplugged (eg to disconnect all the EMI filters in the equipment), does the kettle test still give the same value spike?

I have to say the effect of turning on all the technology simultaneously via an extension block, leading to tripping an rcd is not unexpected behaviour!

PS The allowable tolerance in tripping current of RCD's is -50%, so one tripping at 15mA is still within spec. If your background leakage is already 10mA, that is not really 'minimal'

 

[meerkatmag]

I'll try that check and report back. With all the computer equipment off I'll check if the kettle still causes the leakage.

We get quite a bit of leakage even with just the lighting circuit on. Every circuit seems to add to it.

On the question about circuit testing. When we had an electrician look at this many years ago he tested everything (I believe) and said that he couldn't find anything wrong and just blamed the devices we were plugging in.

Also to note that the leakage happens on either ring, the small kitchen one and the large one. It's a 4 story victorian house so quite large distances on the big ring.

Probably unrelated but I get a 30ma current through my gas pipe and this is earthed to my cu.

 

[meerkatmag]

Right now with no computers on, no tv. Just washing machine, lights, fridge etc. This is 3.3amps draw There is 14.2ma leakage.

If I put kettle on it jumps to 22.6ma

 

[mainline]

Right now with no computers on, no tv. Just washing machine, lights, fridge etc. This is 3.3amps draw There is 14.2ma leakage.

If I put kettle on it jumps to 22.6ma

What is the earthing arrangement ?

Is it just the 2 rings, or does the lighting circuits do the same.

 

[meerkatmag]

What is the earthing arrangement ?

Is it just the 2 rings, or does the lighting circuits do the same.

There's 2xrcds. Left rcd has kitchen ring. Right rcd has main ring and lighting. We also have a car charger which has its own separate cu.

It looks like the earth line from cu goes to a box by the meter and here it's connected to an earth wire from the gas pipe and a wire that goes into the ground under the floor.

 

[mainline]

There's 2xrcds. Left rcd has kitchen ring. Right rcd has main ring and lighting. We also have a car charger which has its own separate cu.

It looks like the earth line from cu goes to a box by the meter and here it's connected to an earth wire from the gas pipe and a wire that goes into the ground under the floor.

I would get a decent electrician in to do some testing.

You could have a neutral to earth problem, but without test results it's all guesswork.

 

[Avo Mk8]

Right now with no computers on, no tv. Just washing machine, lights, fridge etc. This is 3.3amps draw There is 14.2ma leakage.

If I put kettle on it jumps to 22.6ma

But the kettle could be responsible for a real ~8mA earth leakage!?
Bit of a fiddle to measure it at the kettle, but could be done!

 

[meerkatmag]

But the kettle could be responsible for a real ~8mA earth leakage!?
Bit of a fiddle to measure it at the kettle, but could be done!

I'll try to measure it!

 

[meerkatmag]

But the kettle could be responsible for a real ~8mA earth leakage!?
Bit of a fiddle to measure it at the kettle, but could be done!

I modified an extension lead so I can clamp the line and neutral. Kettle did show around 8ma leakage slowly rising to 10ma then dropped to 0ma once it boiled. I'm going to try this in another house and see if this is the kettle or my house wiring.

 

[Avo Mk8]

I modified an extension lead so I can clamp the line and neutral. Kettle did show around 8ma leakage slowly rising to 10ma then dropped to 0ma once it boiled. I'm going to try this in another house and see if this is the kettle or my house wiring.

Useful check - thanks!
I'd bet on the kettle!

 

[mainline]

Useful check - thanks!
I'd bet on the kettle!

One leaky appliance isn't going to cause problems on the other rcd.

 

[James]

Classic Earth to Neutral Fault symptoms.
has anyone checked for it?

 

[mainline]

Classic Earth to Neutral Fault symptoms.

As #8

has anyone checked for it?

He's only got a clamp meter

 

[Avo Mk8]

One leaky appliance isn't going to cause problems on the other rcd.

???
Not suggesting that
Merely that leakage current as measured at the CU tails is going up by c8mA when the kettle is used because the kettle is directly responsible for that leakage current, and it's not because the load is increasing.

 

[meerkatmag]

???
Not suggesting that
Merely that leakage current as measured at the CU tails is going up by c8mA when the kettle is used because the kettle is directly responsible for that leakage current, and it's not because the load is increasing.

Yeah checked my daughters (new) hairdryer. 5ma at the plug, 5ma extra at the cu. So does seem to tally. If a single device can create 5ma leakage or even 8ma for the kettle then surely my readings of 15ma leakage current at base level is fairly normal? Turning my computers on seem to cause a spike of around 10 to 15ma so causes trip. Does this all sound normal then? Is it just to be expected? Can't be every device in my house is faulty so these amounts must be normal?

I like the sound of earth neutral fault, something concrete to investigate. I guess I need an electrician to do better tests.

 

[Avo Mk8]

Does this all sound normal then? Is it just to be expected?

Every "electronic" device that involves switching power supplies, or even 'mechanical' switches, is fitted these days with a filter to prevent electromagnetic interference. It's a requirement of the ElectroMagnetic Compatibility standards that devices must comply with to earn the dreaded CEUK or equivalent mark, to be legitimately put on the market.
The filter includes capacitors across the mains, and from each supply wire to ground. This results in a small current to earth, which may be a few (or more!) mA.

Just one big ring and a kitchen ring, in a big house, with multiple devices, chargers, appliances, all relying on one or two rcd's, I would say could be a problem itself, not necessarily the wiring.

An electrician could easily check the fixed installation with a high voltage Insulation Resistance tester to prove the wiring wasn't contributing to the mA leakage. But you could do a crude check at the meter tails, by literally pulling out all plugs, switching off spurs etc etc, to see if there is a residual current.

 

[Pretty Mouth]

Yeah checked my daughters (new) hairdryer. 5ma at the plug, 5ma extra at the cu.

Out of interest, is this hairdryer a class 1 or class 2 item? Class 2 would have a 'square within a square' symbol somewhere on it, and possibly (though not definitely) a plastic earth pin on its plug.

 

[meerkatmag]

Out of interest, is this hairdryer a class 1 or class 2 item? Class 2 would have a 'square within a square' symbol somewhere on it, and possibly (though not definitely) a plastic earth pin on its plug.

Got the square within a square symbol, see attached. So class 2? And yes plastic earth. So how could there be earth leakage? I'm confused by that. Is my clamp meter wrong?

 

[meerkatmag]

Every "electronic" device that involves switching power supplies, or even 'mechanical' switches, is fitted these days with a filter to prevent electromagnetic interference. It's a requirement of the ElectroMagnetic Compatibility standards that devices must comply with to earn the dreaded CEUK or equivalent mark, to be legitimately put on the market.
The filter includes capacitors across the mains, and from each supply wire to ground. This results in a small current to earth, which may be a few (or more!) mA.

Just one big ring and a kitchen ring, in a big house, with multiple devices, chargers, appliances, all relying on one or two rcd's, I would say could be a problem itself, not necessarily the wiring.

An electrician could easily check the fixed installation with a high voltage Insulation Resistance tester to prove the wiring wasn't contributing to the mA leakage. But you could do a crude check at the meter tails, by literally pulling out all plugs, switching off spurs etc etc, to see if there is a residual current.

View attachment 121976

I'd love to get each floor separated into separate rings with rcds integrated into each breaker. Maybe that would be the solution. It's going to be really expensive job though.

I did find once some very old wiring with cloth insulation in a ceiling light fitting. That was replaced but it's not impossible there would still be some in other places.

I'll try to do the check you suggest checking for residual current.

 

[Pretty Mouth]

Got the square within a square symbol, see attached. So class 2? And yes plastic earth. So how could there be earth leakage? I'm confused by that. Is my clamp meter wrong?

I have a very similar clamp meter to yours, probably the same one re-branded. Whilst it has its uses, I'm not convinced of its accuracy at mA resolution. I suspect it falsely reads a certain amount of leakage, roughly proportional to the amount of current flowing through the clamped conductors, even where there is no actual leakage to earth. For this reason, I seldom use it for fault-finding in this way.

If I were asked to fault-find this problem, I'd start with an insulation resistance test, then perhaps RCD ramp tests under different loads, then look at the possibility of a faulty RCD.

 

[newfutile]

how are you checking this earth leakage? as in what conductors are you clamping.
most neutral to cpc faults cause instant trip of any (RCD) Residual Current Device ,however a double insulated device cant put current down a cpc.
is there any outside equipment may have got wet ?

 

[mainline]

I'd love to get each floor separated into separate rings with rcds integrated into each breaker. Maybe that would be the solution. It's going to be really expensive job though.

I did find once some very old wiring with cloth insulation in a ceiling light fitting. That was replaced but it's not impossible there would still be some in other places.

I'll try to do the check you suggest checking for residual current.

You could consider splitting the ring finals using rcbos

 

[HappyHippyDad]

I feel I may well be missing something here (Or just not read the thread thoroughly), but... Isn't it simply that there is too much earth leakage and the single RCD trips? Or even if it's a dual RCD, it may well be too much? A 30mA RCD trips at anywhere between 19mA and 23mA usually.

If it's an RCBO board then I'd be thinking more a N-E fault.

What Consumer Unit do you have @meerkatmag . Single RCD, Dual RCD, RCBO's? A picture would be great.

 

[littlespark]

Right at the start @HappyHippyDad the OP said there was 2 RCDs… so a dual RCD board.

Money is on just too much “normal” Earth leakage across several circuits.
Everything tests fine, but cumulative leakage knocks it over.

It could be possible that continual trips have weakened the rcd…. So a 30mA rcd that would normally trip at 19 to 23 like you say, may be as little as 10mA if faulty.
RCD tester with ramp test would confirm.

 

[meerkatmag]

Right at the start @HappyHippyDad the OP said there was 2 RCDs… so a dual RCD board.

Money is on just too much “normal” Earth leakage across several circuits.
Everything tests fine, but cumulative leakage knocks it over.

It could be possible that continual trips have weakened the rcd…. So a 30mA rcd that would normally trip at 19 to 23 like you say, may be as little as 10mA if faulty.
RCD tester with ramp test would confirm.

Hi thanks for all the help by the way!

It's a dual RCD CU by Harok. We had this fitted about 10 years ago but the house wiring wasn't changed at the time. Wiring in house is old Red/Black easily 20 or more years old.

The RCD started tripping randomly right after the CU was fitted 10 years ago and has got gradually worse over the last 10 years. Possibly because we have modernised the house and fitted more electronics. The kids have grown up and now have computers etc. The problem was improved greatly by moving as many devices from the house Ring to the Kitchen ring. Spreading the load more evenly between the two. I moved the washing machine/dryer and freezer to the kitchen ring and that made a huge reduction in random trips.

Since then it mostly only trips when you plug in a high power device, especially multiple devices using a 4 gang all at once. So the idea that it's just too many devices on a large ring and/or an old weakened RCD does sound likely.

I've been doing more tests this morning and as far as I can tell everything seems to be reasonably normal. I've been using the hairdryer to test earth leakage. As I know this device doesn't have an earth it literally can't leak. I think my earth clamp is inaccurate as previously stated. The higher the overall current going through the wires the more leakage it shows. I've tried using the meter just on the earth cable from the CU and this seems more consistent.

With this setup I get about 2 to 3 ma leakage from the hairdryer. About 10 to 15ma of base level leakage with average house load all on (that's across both RCDs). Interestingly if I turn everything off (breakers/RCDs/Main) I still get about 2ma of current in the earth cable. Is that normal? As I turn the breakers on - the current actually drops in the earth cable, before devices start turning on and it returns to higher amounts. Is that the capacitors draining to earth after losing power maybe?

 

[westward10]

Are these two ring finals protected by the same RCD?
A picture of your consumer unit might help.

 

[meerkatmag]

Are these two ring finals protected by the same RCD?
A picture of your consumer unit might help.

each ring is on a different RCD. Pic Attached

 

[Avo Mk8]

Meerkatmag
You ask some questions in your posts that I'm afraid we don't always respond to.
You might be interested to look at the paper below on RCD selection and testing. A bit technical, but gives some idea of 'normal' leakage values and what should be the aim in design.
Also shows some waveforms, drawing attention to the complexity of this subject, and why a more basic tester would not give "true rms" accurate results.
Only if you're interested, i'm not trying to make a point, other than I suppose that to make sense of measurements in this domain, one probably needs accurate calibrated specialist test equipment!

https://www.doepke.co.uk/downloads/Techpubs/Techpub-19.pdf

 

[Bob Geldoff1234]

Have any of your 4 gang adaptors got surge protection in them? I have had a few which can show 0.1 meg ohms down to earth. They could cause the rcd to trip?

 

[mainline]

A

Yeah checked my daughters (new) hairdryer. 5ma at the plug, 5ma extra at the cu. So does seem to tally. If a single device can create 5ma leakage or even 8ma for the kettle then surely my readings of 15ma leakage current at base level is fairly normal? Turning my computers on seem to cause a spike of around 10 to 15ma so causes trip. Does this all sound normal then? Is it just to be expected? Can't be every device in my house is faulty so these amounts must be normal?

I like the sound of earth neutral fault, something concrete to investigate. I guess I need an electrician to do better tests.

A class 2 appliance shouldn’t cause a rise in earth leakage not unless there’s a problem with neutral current being diverted.

The problem may not even be in your property, without proper testing it’s all guess work.

What adds up is that your kettle causes more leakage than the hair dryer so seems to tally with the theory more load more leakage

 

[brianmoooore]

The first thing to do is to invite Mr. Bainbridge back to carry out another EICR. Interesting that he recommended a retest after 5 years instead of the normal 10 for domestic. This might indicate there were some concerns about the wiring and that it might be failing.
I assume that the EV point has been installed since 2014, since the incorrect make of the RCD and its poor fitting should render a report 'unsatisfactory.'

 

[meerkatmag]

Thanks for all the help, even though I haven't got to the bottom of this. I feel like I understand what's going on a bit more. I think what I'm going to do is get an electrician to fit rcbos to each circuit and then have all the wiring tested.

I have a feeling that there's not a lot actually wrong, I'm just running too many devices that leak to earth. That kettle for example I've tested a number of times and does seem to leak 10ma, so that's going in the bin.

If there is an earth neutral fault I think I need an electrician I can't find it myself.

 

[mainline]

Thanks for all the help, even though I haven't got to the bottom of this. I feel like I understand what's going on a bit more. I think what I'm going to do is get an electrician to fit rcbos to each circuit and then have all the wiring tested.

I have a feeling that there's not a lot actually wrong, I'm just running too many devices that leak to earth. That kettle for example I've tested a number of times and does seem to leak 10ma, so that's going in the bin.

If there is an earth neutral fault I think I need an electrician I can't find it myself.

Wait and get the electrician to insulation test the kettle.

Whichever way, come back and tell us the outcome.

 

[meerkatmag]

Wait and get the electrician to insulation test the kettle.

Whichever way, come back and tell us the outcome.

I will report back I promise. It's been very interesting investigating it all. One small update, I found 2 x surge protector 4 gang extensions and 1 ikea USB 4 gang hidden away in bedrooms. So they might have been creating some leakage. They are all binned now. I get extremely weird readings from my earth wire from CU. The line and neutral clamped gives relatively sensible readings - I've noticed that the placement of the wires in the clamp need to be exactly central and not twisted otherwise you get wild readings. The earth wire from CU shows higher leakage than the live-neutral, and it seems to react to load much more than the live-neutral. Also I measure 90ma on my gas pipe yesterday. All very interesting (and annoying).

 

[newfutile]

If it's a tnc-s pme supply you would expect to protective bonding conductors to carry some current.
Have you tested your meter on a fixed small load?

 

[brianmoooore]

Could a slightly resistive joint in the neutral supply to the property be the cause of this? The higher the load in the property, the more the neutral and earth terminal at the consumer terminals will move above true earth, giving rise to the current in the gas bond.