Thought I found a nuisance trip but I didn’t. Any advice to my questions...?

[ppelec100]

So a customer had a nuisance trip and I spent 4 hours trying to find the fault (first hour identifying circuits as none labelled at board).

There was a ring circuit with each leg being fed by 32a breaker, I put them into the the same breaker.

Another ring had its leg connector blocked floating inside the CU so I connected that to it’s other leg in the MCB.

So moving on to the nuisance trip
The RCD which trips is feeding a shower circuit which is turned off when not in use apparently. It feeds downstairs 1 socket radial which does phone and WiFi. And it feeds the ring in front room back room which has so many things plugged into extensions and the boiler and heating water system seems to run off that ring too.

I told him to put the extension lead upstairs onto the other ring being fed by the other RCD and put it into a master plug RCD adaptor and he said the RCD adaptor is tripping which says the fridge is faulty right?

But he also says that the original RCD that was tripping is still tripping but not as much... I have a feeling it’s cos he has too much stuff plugged in which is leaking too much to earth on that ring which the RCD can only take 30ma.

He already paid me for four hours as I thought it was the fridge then he called today saying RCD in cu tripping.

My next step is to buy some sort of current leakage clamp meter to see how much is leaking on that ring through all his crap. If it is high I was thinking of splitting the ring in the middle creating 2 radials and obviously putting them on like 16 or 20a RCBO’s but this would be done when changing the rcd cu to rcbo cu which I recommended. Hopefully this would split the leakage current between the 2 rcbos if that is the problem?

But before I do that is there anything else I can do or test or check as I don’t really want to spend £100+ on a leakage current clamp meter just for this 1 job though I know I can use it on other fault finds but I’m not that busy like that.

Sorry for the long post and grateful for any help thanks.

 

[SparkyChick]

Did you ramp test the RCDs with nothing connected to them to establish whether they are conforming to specification (trip current around 25-26mA)?

When you say he has a lot of stuff plugged in... how much stuff? and perhaps more importantly, what? If it's a lot of electronics via plug-in adapters, many of them are class 2, so have no earth to create leakage.

If it's random tripping, and they can instantly reset the RCD and carry on with their day for a while, I would be asking if they've changed anything... hung any pictures, changes any accessories etc. And I would be tempted to run some long running insulation resistance tests on each circuit. Sometimes if there is a tiny pinhole in the insulation, a quick test won't reveal it but a test where you apply the test voltage for a significant time can result in a charge in the cable that can jump the pinhole resulting in a momentary drop in IR reading on the tester... it can be tricky to see with digital devices, but I've found this approach can reveal such issues quite reliably.

 

[Dartlec]

Lots to unpick there, not least the bit in passing about a 'ring circuit with each leg being fed by 32A breaker'. If that was legitimately the case, then god knows what other interlinks could be in place. A neutral in the wrong side of the board for example could easily cause RCD tripping problems.

You also mention another ring with a leg not connected? Neither of those sound like a well maintained electrical installation.

Finding that at the start of a job, for me the only real answer would be to carry out a proper EICR (at least in terms of practical tests) to confirm what is what. Anything else might well just end up wasting time chasing your own tail.

Start by not assuming anything, and check that the ring is a ring - and check R1+R2 at each point on it to ensure that what you and he think is on the ring is. I would probably test R1+Rn too just to ensure there are no weird cross connections on any circuits.

Then check the radial. Then the Shower. Make sure that you know what each L and N in the board is.

Only then once you know for sure what you are checking is it worth checking IR on the individual circuits - L-N and L-E/N-E, Do this with all appliances removed or isolated, so you can either rule in or out the fixed wiring.

Also ensure that for those tests, the CPC is connected to the MET - to rule out any possible earthing paths through building/pipework/etc

As mentioned, ramp testing the RCD with nothing else connected is a good way to see whether it may be getting over sensitive, though even so it usually takes a 'leaky' appliance or two to trigger it unless it's tripping at a ridiculously low current.

If the fixed wiring is then tested as OK, then you can move onto the appliances and narrow down which one it may be. A PAT tester can be useful for that, but you can also do most of the tests with a standard MFT, or any decent ohmmeter.

If you're lucky, testing N-E on an appliance will give a low reading and immediately suggest a culprit. If you're unlucky, all appliances will test fine dead, as the fault only occurs when they are running, or at a certain point in the running cycle.

Random intermittent faults can be a bugger to track down, and can sometimes be a case of trial and error. It's often a case of getting the client to write down immediately when things trip what they were using/doing - sometimes that then shows a pattern than they hadn't considered (recent case being the oven being on each time).

Only once you've located the problem, would I say that individual RCBOs and radials may be worth considering as a solution to a known problem, rather than a possible fix.

I certainly wouldn't be changing the board as a 'fix and hope' effort, before working out what is going on as it then means you take responsibility for any future trips.

Likely things that may well be a cause of trips, in no particular order:

Anything with a heating element (Cooker, Kettle, toaster, iron, dishwasher, washing machine, etc)
Boilers (I wouldn't IR these though, unless L+N->E at 250V)
Fridges/Freezers
Outside sockets with leaks (or any outside electrics)
Rodent damage
Nails/screws going through cables where pictures or fittings have been mounted (not necessarily recently)

 

[bill01803]

I would suggest you should purchase the clamp meter, it is invaluable in the situation you describe, it will give you an instant guide to the problem, just put it around L+N (not around the cpc as this will not give you all the leakage).

 

[HappyHippyDad]

I would have started with some insulation resistance tests. Usually, an RCD is tripping due to a fault which an IR test will show up.

First of all a quick global test to earth. If that shows a fault (I.e <1Mohm) I would then be testing each curcuit individually. A tripping RCD usually shows a fault around 0.01Mohms.

If the global test was ok, then I would use the earth leakage clamp as I would suspect combined leakage from things plugged in (which is becoming more common). I would clamp the tails rather than the main earth as this will give a more accurate measurement.

Plus, as SC says, I would be testing the RCD. I would text this at the RCD with nothing connected. It's very rare to have a faulty RCD, but possible.

If all that fails, then I'd be starting a thread of my own on here!

 

[telectrix]

a quick way to measure leakage is to ramp test the RCD with all circuits and loads connected. then with all dissed. the difference will be the leakage .

 

[HappyHippyDad]

a quick way to measure leakage is to ramp test the RCD with all circuits and loads connected. then with all dissed. the difference will be the leakage .

I take it you don't have an earth leakage clamp meter .
You've always got a Heath Robinson solution.

 

[telectrix]

i do have a e/l clamp meter. however, my solution is for where A rcd FEEDS SEVERAL CIRCUITS, AND YOU CAN ONLY USE A METER ON 1 AT A TIME. every time i press "a" the caps lock comes on also. fat fingers, poor eyesight.

 

[ppelec100]

Did you ramp test the RCDs with nothing connected to them to establish whether they are conforming to specification (trip current around 25-26mA)?

When you say he has a lot of stuff plugged in... how much stuff? and perhaps more importantly, what? If it's a lot of electronics via plug-in adapters, many of them are class 2, so have no earth to create leakage.

If it's random tripping, and they can instantly reset the RCD and carry on with their day for a while, I would be asking if they've changed anything... hung any pictures, changes any accessories etc. And I would be tempted to run some long running insulation resistance tests on each circuit. Sometimes if there is a tiny pinhole in the insulation, a quick test won't reveal it but a test where you apply the test voltage for a significant time can result in a charge in the cable that can jump the pinhole resulting in a momentary drop in IR reading on the tester... it can be tricky to see with digital devices, but I've found this approach can reveal such issues quite reliably.

I should’ve mentioned that I did a ramp test and I think it tripped at 21ma which is seems okay?

He’s got stuff plugged in like sewing machine, pc, fish tank pump and filter, tv and some other stuff I didn’t take too much notice of at the time. His fridge seemed to be tripping the plug in RCD. Which is reducing how much the RCD at the CU trips because it’s on a plug RCD.

I did do some IR test with everything unplugged at 500v on the ring. Came out >1999Mohm.

Thanks for your reply

 

[ppelec100]

Lots to unpick there, not least the bit in passing about a 'ring circuit with each leg being fed by 32A breaker'. If that was legitimately the case, then god knows what other interlinks could be in place. A neutral in the wrong side of the board for example could easily cause RCD tripping problems.

You also mention another ring with a leg not connected? Neither of those sound like a well maintained electrical installation.

Finding that at the start of a job, for me the only real answer would be to carry out a proper EICR (at least in terms of practical tests) to confirm what is what. Anything else might well just end up wasting time chasing your own tail.

Start by not assuming anything, and check that the ring is a ring - and check R1+R2 at each point on it to ensure that what you and he think is on the ring is. I would probably test R1+Rn too just to ensure there are no weird cross connections on any circuits.

Then check the radial. Then the Shower. Make sure that you know what each L and N in the board is.

Only then once you know for sure what you are checking is it worth checking IR on the individual circuits - L-N and L-E/N-E, Do this with all appliances removed or isolated, so you can either rule in or out the fixed wiring.

Also ensure that for those tests, the CPC is connected to the MET - to rule out any possible earthing paths through building/pipework/etc

As mentioned, ramp testing the RCD with nothing else connected is a good way to see whether it may be getting over sensitive, though even so it usually takes a 'leaky' appliance or two to trigger it unless it's tripping at a ridiculously low current.

If the fixed wiring is then tested as OK, then you can move onto the appliances and narrow down which one it may be. A PAT tester can be useful for that, but you can also do most of the tests with a standard MFT, or any decent ohmmeter.

If you're lucky, testing N-E on an appliance will give a low reading and immediately suggest a culprit. If you're unlucky, all appliances will test fine dead, as the fault only occurs when they are running, or at a certain point in the running cycle.

Random intermittent faults can be a bugger to track down, and can sometimes be a case of trial and error. It's often a case of getting the client to write down immediately when things trip what they were using/doing - sometimes that then shows a pattern than they hadn't considered (recent case being the oven being on each time).

Only once you've located the problem, would I say that individual RCBOs and radials may be worth considering as a solution to a known problem, rather than a possible fix.

I certainly wouldn't be changing the board as a 'fix and hope' effort, before working out what is going on as it then means you take responsibility for any future trips.

Likely things that may well be a cause of trips, in no particular order:

Anything with a heating element (Cooker, Kettle, toaster, iron, dishwasher, washing machine, etc)
Boilers (I wouldn't IR these though, unless L+N->E at 250V)
Fridges/Freezers
Outside sockets with leaks (or any outside electrics)
Rodent damage
Nails/screws going through cables where pictures or fittings have been mounted (not necessarily recently)

The CU had been changed by someone not competent but the rest of the house wiring seems okay you know the old school stuff house rewires.

I did do R1+R2 on the circuits being fed by the tripping RCD and they were okay. IR tested at the board all seemed okay >1999Mohm

Ramp test was done and it tripped 21ma.

I did do the pat test thing on the fridge which gave a low reading I cannot remember what, which is why I suggested the plug in RCD to him to rule it out. Which it has done but there’s still something else which seemed like it was partnering up with mr fridge.

Yeah I really was thinking I don’t want to change the CU and have extra headache, at least if I can’t find it without changing the CU I can walk away without it being my headache lol.

With the cooker kettle stuff etc they’re on a MCB being fed by the other RCD that doesn’t trip. Again outside electrics on the fine RCD.

Rodent damage is what is could be but I’m not sure as IR test is clear. I hope not though. Will have to look out for pins pictures etc.

Thanks for your reply

 

[ppelec100]

I would suggest you should purchase the clamp meter, it is invaluable in the situation you describe, it will give you an instant guide to the problem, just put it around L+N (not around the cpc as this will not give you all the leakage).

Yeah I think if he decides he still wants me to carry on looking I will buy one. Which one do you use?

I was looking at the TIS560 as it has nice big jaws and it has good reviews seems pretty accurate too.

 

[ppelec100]

I would have started with some insulation resistance tests. Usually, an RCD is tripping due to a fault which an IR test will show up.

First of all a quick global test to earth. If that shows a fault (I.e <1Mohm) I would then be testing each curcuit individually. A tripping RCD usually shows a fault around 0.01Mohms.

If the global test was ok, then I would use the earth leakage clamp as I would suspect combined leakage from things plugged in (which is becoming more common). I would clamp the tails rather than the main earth as this will give a more accurate measurement.

Plus, as SC says, I would be testing the RCD. I would text this at the RCD with nothing connected. It's very rare to have a faulty RCD, but possible.

If all that fails, then I'd be starting a thread of my own on here!

Hi thanks for your reply. Ramp test and IR test have been done. My next step is the earth leakage meter like some of you suggested.

 

[ppelec100]

a quick way to measure leakage is to ramp test the RCD with all circuits and loads connected. then with all dissed. the difference will be the leakage .

Tried this but I think it might have been with only the fridge in and it was still 21ma (same as with everything disconnected). Will have to do it again with all his stuff plugged in.

Thanks for your reply

 

[ppelec100]

I take it you don't have an earth leakage clamp meter .
You've always got a Heath Robinson solution.

 

[HappyHippyDad]

When we quote another's text it means we're replying specifically to what that person said. In this case telectrix, not yourself.