nav[aria-label="Primary Navigation"] { padding: 0; & ul { list-style: none; width: 100%; display: flex; flex-direction: row; justify-content: start; align-items: start; gap: 30px; padding: 0; & li { margin: 0; } & ul li { list-style: none; } } }

Child Item

Issues with RCDs not tripping

Jay.D87

Hi Guys

Just looking for other sparks opinions on this. Have had bit of a noodle scratcher last couple of weeks. Had my fluke 1652C calibrated on 22/11/17 have used it few times since.

Story is was called to emergency call out to RCD tripped. Lady said was happening intermittently. Took cover off board and realised dudes few week previous had wired an oven to it. Now they had connected 32mcb to RCD side so changed this due to earth leakage from oven powered up and no probs. So said I’d run fault loop impedance test elsewhere for peace of mind. Got good readings on all circuits.

Now here’s the odd part. Ran RCD Trip time test and no trip, on 1/2, 1 times and 5times. I even ramped her up to 500mA no joy. RCD not jammed as tried with test button. So I automatically assumed replace RCD. Went off got one and replaced. Still same issue so did bit of digging

I have an earth somewhere, as have 230V L-E at board. Although when I looked at ESB meter I can’t see the Neutralising done. I don’t even see any Earth on ESB side and I’m 90% sure there’s no earth rod either. I then thought my meter was faulty and was about to have it send back to guys who calibrated but out of curiousity I tested my apartment and same thing but when I tried test button on RCD and it’s jammed! So tried my girlfriends parents place and tripped on 1/2 times trip current! So I’ve checked a few premises I’ve worked on RCDS and RCBOS and perfect accurate trip times. So think will rule faulty tester out.

I thought the RCD I got faulty but had two and tried them. Rang whole sale and they said abb RCD wont be faulty and Defo not two in a row.
Now I’ve done a bit of reading into it and read people saying to unplug absolutely everything when on test and inductive loads can absorb the fault current and trick the tester into thinking there was no trip.

Has anyone come across this and if so any advice? Thanks in advance!

Find more posts tagged with

Quick Links

Comments

Bruthal

Tuco88 wrote: »

Its possible to experience a current greater than 30mA also with operation characteristic of the rcd so time being the factor.

Yea you would probably be better off with a 40ma shock bringing about a fast trip than a 29 - 30ma with a potentially much longer non trip time.

Tuco88

Its possible to experience a current greater than 30mA also with operation characteristic of the rcd so time being the factor.

Bruthal

2011 wrote: »

I had a few little “tingles” myself back in the day, never tripped an RCD that way though.

Yes its quite a nasty (feeling) shock perceived at the level needed to trip one. But they certainly do their intended job.

2011

I had a few little “tingles” myself back in the day, never tripped an RCD that way though.

Bruthal

I would make the point that a person coming in contact with a phase will not automatically trip an RCD.

30ma still has to flow to trip it, and 30ma is quite a nice shock. A person standing on dry ground contacting only the phase will be very unlikely to trip the RCD, but will also be unlikely to recieve much of a shock. Someone standing knee deep in wet concrete is a different story however.

I unintentionally tripped them a couple of times myself while in contact with a phase, but was also direct in contact with an earthed point too. Quite a perceived shock is felt before they will trip.

2011

RCDs always seem to cause confusion.
Sometimes it helps to go through the basics.
This video gives quite a good explanation:

Jay.D87

Tuco88 wrote: »

He could remove the load side of the rcd in the consumer unit and test, so taking all downstream faults out of the equation.

Yes going to try this Tuco as she’s been onto me again!

Tuco88

He could remove the load side of the rcd in the consumer unit and test, so taking all downstream faults out of the equation.

Skedaddle

I'm not an electrician but, work with some fairly techie areas and we once encountered an issue with this once where an "audiophile" had come up with the notion (based on a misunderstanding of something read online) that if he cross connected the neutrals and earths at sockets he would remove mains hum and get a "clean neutral". He had no idea how earthing systems worked at all.

He caused RCD trip problems and it was only flagged up as an issue when someone got shock and no RCD tripped. It was picked up by an electrician after a lot of working through every circuit and fitting. He found several sockets had been wired with links between the neutral and the earth terminals.

You'll probably have to hunt trough everything for inappropriate earth-neutral bonding
.
Apparently, connecting the neutral to the earth terminals on Schuko sockets was very common in former East Germany and parts of the old Eastern Bloc.

Bruthal

Risteard81 wrote: »

No. Because with TN-C-S although combined at the origin they are separate throughout the installation. If you have interconnections beyond that point then you essentially have a PEN conductor at that point. It won't always have this effect though and very much depends on the loads etc. and how they are dividing, and ultimately what the RCD is seeing through its coil.

The neutral earth fault has to be on the load side of the RCD to cause a divided fault path which affects the RCD. You said you seen a neutral earth fault on tails cause it. I am at a loss to see the difference between a neutral earth short on tails, and a neutral earth short at the neutralised point.

EDIT: Was on phone for first replies.
If we take a device plugged into a socket. A person contacts the phase and earth.

No neutral earth fault.

50ma flows through the person to the earth conductor in lead of device. RCD sees 50ma in on the phase. Nothing back on neutral. RCD trips

Now with neutral earth short on socket circuit.

50ma flows through the person to the earth conductor in lead of device. It comes to the neutral earth short and divides. 25ma through the earth, but the other 25ma back on the neutral via this earth neutral fault. RCD fails to trip from being less sensitive.

Neutral earth short in meter tails.

50ma flows through the person to the earth conductor in lead of device. RCD sees 50ma in on the phase. Nothing back on neutral (through rcd). RCD trips.

In the last scenario there, current through the person returns on the earth conductor and does not encounter any neutral path that can flow through the RCD neutral to desensitise it. A connection between neutral and earth at the tails, or neutralising point, or between main neutral and earth bars in the DB are the same thing, in reality, just at different points long the same conductors. Where it changes is when the fault is on the load side of the RCD.

Risteard81

Bruthal wrote: »

Risteard81 wrote: »

When it is upstream it can desensitise the RCD so that it never trips as too much of the current is dividing to cause a sufficient imbalance. I've seen this with a neutral/earth fault on the tails for example.

Anyone with a neutralised house has a desensitised RCD so?

No. Because with TN-C-S although combined at the origin they are separate throughout the installation. If you have interconnections beyond that point then you essentially have a PEN conductor at that point. It won't always have this effect though and very much depends on the loads etc. and how they are dividing, and ultimately what the RCD is seeing through its coil.

Certainly this phenomenon is probably more likely on a TT (Direct Earthed) system.

Bruthal

Risteard81 wrote: »

When it is upstream it can desensitise the RCD so that it never trips as too much of the current is dividing to cause a sufficient imbalance. I've seen this with a neutral/earth fault on the tails for example.

Anyone with a neutralised house has a desensitised RCD so?

Risteard81

Bruthal wrote: »

Risteard81 wrote: »

Have you checked for an upstream neutral/earth fault desensitising the RCD?

Doesn't have to be upstream of the test point for that to happen, just downstream of the RCD.

It would be expected to trip if any load plugged in. More so in neutralised installations.

When it is upstream it can desensitise the RCD so that it never trips as too much of the current is dividing to cause a sufficient imbalance. I've seen this with a neutral/earth fault on the tails for example.

Bruthal

Risteard81 wrote: »

Have you checked for an upstream neutral/earth fault desensitising the RCD?

Doesn't have to be upstream of the test point for that to happen, just downstream of the RCD.

It would be expected to trip if any load plugged in. More so in neutralised installations.

Jay.D87

Jay.D87 wrote: »

Disconnected main neutral and did continuity test across all earth to neutral doesn’t seem to be any fault

Will possibly try insulation resistance test though could show up something

Jay.D87

Risteard81 wrote: »

Have you checked for an upstream neutral/earth fault desensitising the RCD?

Disconnected main neutral and did continuity test across all earth to neutral doesn’t seem to be any fault

Jay.D87

Tuco88 wrote: »

Test the rcd at the DB, with all the downstream breakers open to focus on that alone. Is there a shower RCBO if so hows that working? And polarity good?

If the RCD is working with its test button thats a start. Cant be an S type if you got a brand new one so rule that out.

Hi Tuco

Yes tried that from top of RCD and didn’t trip. Tried from bottom of RCD and tripped all breakers were closed though. Didn’t try open I don’t think

Risteard81

Have you checked for an upstream neutral/earth fault desensitising the RCD?

Risteard81

Have you checked for an upstream neutral/earth fault desensitising the RCD?

Tuco88

Test the rcd at the DB, with all the downstream breakers open to focus on that alone. Is there a shower RCBO if so hows that working? And polarity good?

If the RCD is working with its test button thats a start. Cant be an S type if you got a brand new one so rule that out.

Jay.D87

Bruthal wrote: »

If you are testing via sockets, make sure they are wired properly, in case someone connected sockets with neutral linked to earth terminal in lieu of a proper earth or something mad like that.

Yes I threw my eye on it, very messy board hard to get proper look. But seems to be done right feed into bottom Live RCD feed out top to Busbar.
Then N from main neutral bar to bottom and neutral from top to RCD neutral bar.

Coincidentally I happened to check her extractor 3 a fuse blown but opened socket and they had LIVE in N and neutral in L!!

I did a few continuity tests at board and couldn’t find any links. I’m driven demented with the place now at this stage

Bruthal

If you are testing via sockets, make sure they are wired properly, in case someone connected sockets with neutral linked to earth terminal in lieu of a proper earth or something mad like that.

Jay.D87

Bruthal wrote: »

Did you do a loop impedance test?

I did, good readings all circuits

Jay.D87

Jay.D87 wrote: »

Cheers Bruthal so maybe hammering the earth rod in is best course of action?

Btw both of these premises are old inner city houses, so god knows condition of ESB cables

Is it worth while contacting them? Or will I get laughed at? Problem with the earth rod aswel is it’s concrete jungle no garden or greenery and she doesn’t want me running 10sq through house to back garden.

I’m always wary of hammering the rod down at the front of those houses without plans or knowing where the service pipes are ðŸ˜…

Bruthal

Did you do a loop impedance test?

Jay.D87

Bruthal wrote: »

Most assume a voltage reading of 230v L to E means there is a good earth path...

But, a digital multimeter showing 230v L to E does not show/prove there is a good earth, or any real earth path.

If you test from L to earth through a 1k resistor, the meter will still show 99.99% of the supply voltage because of the very high meter input impedance in voltage test setting, even though 1k ohm is practically an open circuit in terms of an earth fault path.

Cheers Bruthal so maybe hammering the earth rod in is best course of action?

Jay.D87

Tuco88 wrote: »

Just a few questions.

1) Will the RCD operate on its own test button?
2) Is it a 30mA Rcd, Not 300mA?
3) Is it an AC or A type RCD and are you testing on the correct one?

Does sound very odd.

Yea Tuco,

Was thinking everything you’ve mentioned. AC RCD 30mA rang whole sale and made sure and he laughed at me asking was I smoking something at lunch cheeky bugger.

It will operate with test button. Just really weird all RCDs and rcbos I’ve tested since are perfect so it has to be something in the two properties I checked. I never got around to checking the supply impedance mind you.

And I have the customer contacting me know worried I will go back and smash an earth rod in and see will that help.

The only explaination I have at the mo is either the ESB neutral is damaged or else there is something or and appliance in circuit causing interference with the test. I think I read phone chargers can as they have capacitors in them that can absorb the fault current and throw the fluke off. I just couldn’t check that day as she had two young kids in both bedrooms and I didn’t want to be hassling her about going into the rooms.

Bruthal

Jay.D87 wrote: »

I have an earth somewhere, as have 230V L-E at board. Although when I looked at ESB meter I can’t see the Neutralising done. I don’t even see any Earth on ESB side and I’m 90% sure there’s no earth rod either.

Most assume a voltage reading of 230v L to E means there is a good earth path...

But, a digital multimeter showing 230v L to E does not show/prove there is a good earth, or any real earth path.

If you test from L to earth through a 1k resistor, the meter will still show 99.99% of the supply voltage because of the very high meter input impedance in voltage test setting, even though 1k ohm is practically an open circuit in terms of an earth fault path.

Tuco88

Just a few questions.

1) Will the RCD operate on its own test button?
2) Is it a 30mA Rcd, Not 300mA?
3) Is it an AC or A type RCD and are you testing on the correct one?

Does sound very odd.

Skedaddle

Ok..