RCD tripping - Have i found the cause?

[msheppard36]

Morning all

On xmas eve the RCD tripped. The CU has 2 RCD protected row of fuses. I flipped it back on and all ok. Then on Tuesday it tripped again. I have narrowed it down to the MCB that supplies a summerhouse. Yesterday I did some investigations. That MCB supplies power to the summerhouse and garage sockets. I removed the L and N for the sockets and the RCD tripped. I then reconnected the sockets and removed the summerhouse and all ok. So must be the summerhouse?

The SWA cable goes into the summerhouse into a socket. From that socket theres a T&E to an outside junction box that feeds some lights and an outside socket. When i disconnect that T&E, so the only cable connected to the socket is the SWA supply, all is ok. So I am guessing the problem must be the cable to the outside JB or moisture in the outside JB. The only doubt is that the outside JB appeared to be completely dry inside. I removed the cable feeding all the outside, and it still tripped, so I think I have ruled out everything and I am left with a problem with the cable to the outside JB. I was rained off yesterday so have not done anything else, but am I along the right lines? Cant see what else it can be.

Thanks in advance.

 

[Harry Bloomfield]

I was rained off yesterday so have not done anything else, but am I along the right lines? Cant see what else it can be.

Probably! It's always a good idea, to have a means to easily isolate both the L and N, at the point they emerge from indoors to an outdoor installation - basically, a double pole switched, spur unit.

 

[Taylortwocities]

On the right lines. Without proper test equipment you'll need to start removing branches off until you find the culprit.

From that socket theres a T&E to an outside junction box that feeds some lights

I would suspect the lights first. Start disconnecting one circuit at a time from the JB. Remember that "disconnecting" means both the live and the neutral. The RCD can trip if there is a fault on either.

 

[ebee]

Without a meter to verify anything it looks like you might be on the right track.
Just one caution - although you have turned a breaker off thereby preventing a L to E fault on that circuit you will not have disconnected the N to eliminate a N to E fault on that circuit (unless your particular Breaker switches the N along with L, fairly rare but suck breakers do actually exist) , an N to E fault can also cause trips to an RCD in many cases too, so could mislead your result meanings.

As Taylor 2 C above states quite rightly you need to ensure both Line and Nuetral are disconnected not just the Line conductor.

 

[EFLImpudence]

I removed the L and N for the sockets and the RCD tripped.

You probably touched earth with the neutral wire - which will trip the RCD if something else on the RCD circuits is using enough current.

 

[Adam_151]

You probably touched earth with the neutral wire - which will trip the RCD if something else on the RCD circuits is using enough current.

Thats what I read the OP as the first time, but I'm not sure thats what he meant to write, perhaps @msheppard36 could confirm, but I eventually parsed it as:

There being two outgoing cables on the CU way thats appears to be the issue: Summerhouse and garage sockets
He disconnected the garage sockets (L&N) at the board, and the RCD was still tripping
He then put the sockets back in the board and took the summer house out and there was no further tripping
He concluded that the issue must be in the summer house

@op Unfortunatly an insulation tester is really required to proceed further, you might be 'luckily' and the fault might show up on the K-Ohms scale of a multimeter (but it needs to be regarded as a test for badness; if you have a fault showing up that way, you definatly have a fault (provided you've not left the neutral connected to the supply, or any such error), but the lack of anything showing up does not indicate goodness (i.e. lack of a fault)

 

[ericmark]

As @Adam_151 has said, trial and error can help to some extent, but one can so easy get it wrong. My CU tested with my clamp on meter over the tails has around 26 mA of leakage, split between 14 RCBO's (MCB and RCD combined) no circuit is leaking enough on its own to cause a problem, however split into just two, if even, then 13 mA each, the guide lines are with no fault it should not exceed a third of the RCD rating, so should not exceed 9 mA.

So if you also have a high back ground leakage, even one extra 1 mA could be enough to tip the balance. On a new installation we often have no power, so only way to test is an insulation tester, and if we see 10MΩ we think that's OK. But we are testing with DC, and our supply is AC, so there is capacitive and inductive leaking, so the maths seems not to be working, 230/10000000 = 0.023 mA but put the clamp on over the line and neutral and common to see a few mA simply due to testing with DC.

For many years my clamp on would only measure 10 or 20 mA, so any leakage below 10 mA it would not show, only when I got a new clamp on that would measure 1 mA did I realise how close to the limit some circuits were. The PAT tester I had would show leakage, and I was surprised how many items had interference suppression that could cause a 1 mA of leakage, and it soon builds up.

Remember a 30 mA trip must not trip at 15 mA and must trip at 30 mA, in the main they trip at around 25 mA, but one would only see that with a ramp test, my old tester was simply does not trip at 15 does trip at 30.

I changed my meter as wanted to measure DC amps, not to measure AC in 1 mA increments, but thought if changing it, then may as well get one which can. I have been surprised to see how close to the wind some installations were sailing. Which is of course why today we no longer use twin RCD consumer units, most today are all RCBO.

 

[JohnW2]

You probably touched earth with the neutral wire - which will trip the RCD if something else on the RCD circuits is using enough current.

You probably realise this, but we've been through this before ... an N-E fault will/can result in an RCD tripping even if the 'something else using current' is on a circuit protected by a different RCD.

Kind Regards, John

 

[EFLImpudence]

I was merely informing the OP what might happen after he wrote.

I have narrowed it down to the MCB that supplies a summerhouse. Yesterday I did some investigations. That MCB supplies power to the summerhouse and garage sockets. I removed the L and N for the sockets and the RCD tripped.

Whether, as Adam wrote, the OP meant something else I don't know.


However,

You probably realise this, but we've been through this before ... an N-E fault will/can result in an RCD tripping even if the 'something else using current' is on a circuit protected by a different RCD.

I am not sure I do remember that. Please remind me how that works.





Following logic, would that not mean that all RCBOs would trip with every fault?

 

[bernardgreen]

The fault may not be in the summer house wiring even though the MCB for that circuit appears to be involved.
.

 

[EFLImpudence]

an N-E fault will/can result in an RCD tripping even if the 'something else using current' is on a circuit protected by a different RCD.

 

[msheppard36]

There being two outgoing cables on the CU way thats appears to be the issue: Summerhouse and garage sockets
He disconnected the garage sockets (L&N) at the board, and the RCD was still tripping
He then put the sockets back in the board and took the summer house out and there was no further tripping
He concluded that the issue must be in the summer house

Yes, this is what I meant. I have disconnected the L and N from the cable that supplies everything outside and the RCD still trips so I think it can only be the cable between socket and JB. Cant quite see how it would have become damaged but maybe? The only other thing was that I noticed the JB is attached to the outside of the summer house which is clad in shiplap and water seems to have got down into the hole where the cable enters the inside of the summerhouse. It has been extremely wet and blowy here and I guess its possible moisture has found its way into something, but not sure. Anyway, thanks for all the replies and help. I know a sparky so if I need to i will ask him to help me.

 

[JohnW2]

I am not sure I do remember that. Please remind me how that works.

Some of the N current from a load protected by 'the other RCD' goes 'backwards' through the RCD protecting the circuit with the N-E fault,hence creating a current imbalance in that RCD.

Following logic, would that not mean that all RCBOs would trip with every fault?

No. As above, no matter how many RCDs or RCBOs there are, the only one which will trip is the one protecting the circuit with the N-E fault (the L and N currents in all the otheer RCDs/RCBOs are 'balanced).

Kind Regards, John

 

[EFLImpudence]

You 'corrected' me by saying the same as I did.

The 'something else using current' could be anywhere; not just on the RCD circuits.

You probably touched earth with the neutral wire - which will trip the RCD if something else on the RCD circuits is using enough current.

So, if there are no non-RCD circuits...

 

[JohnW2]

You 'corrected' me by saying the same as I did.

I didn't 'correct' you. Starting my reply with "You probably realise this, but ...", I merely pointed out that the load causing an RCD to trip (when it is protecting a circuit with an N-E fault) does not necessarily have to be on a circuit protected by the same ('tripping') RCD, since I was not totally clear as to what you meant by ".... something else on the RCD circuits" (I wondered whether the final 's' of that may have been unintentional).

The 'something else using current' could be anywhere; not just on the RCD circuits.

Sure. The current in any circuit can cause the RCD (or RCBO) protecting a circuit with an N-E fault to trip (by the mechanism I described).

In fact, as we've also discussed in the past, I don't think the current even needs to be in the same installation - I need to re-think to be sure, but I think that a load in a neighbouring property can also do it, since part of the neutral current from that load will also 'go backwards' to earth though thee RCD protecting the faulty circuit.

So, if there are no non-RCD circuits...

Exactly - as above. Also as above, probably also if ther are no current-drawing loads at all in the installation (but some 'next door')!

Kind Regards, John