Really strange RCD happenings

[dlt27]

Sorry this is a long one!
I have been carrying out testing and Inspecting for about 10 years now, but haven't come across what I found at my Mum and Dads house. I rewired it about 12 months ago. It is an old stone house and was a nightmare to do.(It's a TT system fed off overheads and a split load MK board with an RCBO on the smokes.
I got a phone call yesterday with my mum saying that Western Power had turned the power off and when they put it back on half the circuits wern't working. I haven't dived into it yet because I haven't had time and also I thought I'd put it out to you lot to see if you had any ideas.
I went around there turned of all the circuits and reset the (bad)RCD then when I put the kitchen sockets on and it tripped. I thought faulty appliance somewhere tried fridge and it tripped the RCD then tried kettle and it did the same. Tried unplugging everything then tried one at a time however then couldn't get it to trip. I thought adding a load to the circuit was obviously finding the fault and I would megger it the next day because didn't have tester.
I went around this morning and Dad said basically anything they turned on tripped the (Bad)RCD . Sometimes the lights, sometimes the immersion, however it was totally intermittent. The funny thing was even circuits on the 2nd (Good)RCD was tripping the 1st (Bad)RCD ie cooker, but not all the time.
I tried all trip tests on what I thought was the (Bad)RCD and couldn't get it to trip, not even with a ramp test. Test button worked though. I thought I've got it. I tried same tests on 2nd (Good)RCD and it all tripped correctly, however when I tripped it I found it tripped the (Bad)RCD and also the RCBO I have got on the smokes. So basically what I thought was the good RCD tripped the 2 RCD's and the RCBO. I also found the test button on the faulty (Bad)RCD also tripped the RCBO, but didn't trip the good RCD. I've checked all neutrals of RCD's to check not mixed up or in wrong neutral bar etc. All is correct. I've checked neutral coming in to property is neutral etc and found PFC to be 0.21 live to neutral. I am going tomorrow to carry out a proper test ie. Disconnect all Neutrals and Lives of circuits and then try RCD tests, meggering etc.
To summarise what I can't understand is how all circuits in the board are tripping the bad RCD and RCBO like a borrowed neutral and why only intermittantly. Also why does so called good RCD trip bad RCD and RCBO when carrying out tests, yet bad RCD won't trip when tested.
The final thing is I don't understand why this as only happened since Western power turned the power back on, but I can't find anything wrong with supply. Any ideas would be much appreciated.

 

[SPARTYKUS]

Hmmm...chewy.... Processing....

 

[Plonker 3]

What test results did you get when you tested the installation?

 

[davesparks]

There'll be a neutral to earth fault somewhere, probably of higher impedance.

 

[UKMeterman]

Have a google for RCD Handbook - December 2010 - Guidance Documents there is a very useful document by BREMA. Check for NE faults on the load side and also the supply still has the correct polarity (L N and E)!!

 

[dlt27]

Supply is correct polarity. Will check results when go around tomorrow. However all readings ie. Insulation Resistance were above 200M. Think I am going to have to do a full test tomorrow. Annoying part was due to cost of build and other things decided to use a board I already had rather than all RCBO's!.

 

[SPARTYKUS]

Supply is correct polarity. Will check results when go around tomorrow. However all readings ie. Insulation Resistance were above 200M. Think I am going to have to do a full test tomorrow. Annoying part was due to cost of build and other things decided to use a board I already had rather than all RCBO's!.

Not
convinced a board of RCBOs would help in this instance...

 

[Lucien Nunes]

The fault is likely to be an N-E fault downstream of the RCD, which trips the RCD whenever the N-E voltage is high enough to cause sufficient current to flow down the affected final's N to the fault to cause a trip. If the DNO have added or moved load on the mains nearby, the Ze and/or typical N-E voltage may have changed. So although in the past the loads within the installation might not have been enough to reach the tripping threshold, allowing the fault to remain hidden, the changed external conditions could have suddenly revealed it. The insulation test should find it, although it could be in an appliance so you might inadvertently disconnect it prior to testing.

 

[Octopus]

Not
convinced a board of RCBOs would help in this instance...

It might if only one RCBO was tripping!

 

[SPARTYKUS]

Maybe, but the situ he describes has two RCDS (RCCBs?) and a separate RCBO. And all three trip at the same time, sometimes. But I agree with the fact that a board of RCBOs can & usually will help narrow down the fault as you are increasing the granularity (yawn....) on the installation, albeit not clearing the fault unless they're double pole tounge_smile

 

[dlt27]

Why doesn't the (bad)RCD trip when carrying out testing on it. That is at 30 or 150mA of test current also doesn't trip on ramp test.

 

[dlt27]

I think you might be on the money with your post as it seems to be a neutral to earth fault. However would of thought it would not reset not be intermittent unless the n-e voltage is constantly changing. Also surely it would not affect the rcd not tripping when tested.

 

[Octopus]

I fault finded a job recently, the RCD wouldn't set, then it would, it wouldn't trip on the manual test button, then it would, when it held it wouldn't trip....

N-E fault on the socket circuit!

 

[snowhead]

Firstly forget you rewired it, assume it was an idiot that did it (no offense, yet).

Test as though you have no prior knowledge of the installation.

As Lucien said above, the fault may have been there all the time.

 

[guest119]

Central heating pump, washing machine, dishwasher, outside lighting all favourites to introduce a fault.

Boydy

 

[Lucien Nunes]

However would of thought it would not reset not be intermittent unless the n-e voltage is constantly changing

Which it is, usually increasing with load on both your own installation and others nearby. It's caused by voltage drop in the DNO's cable.

Also surely it would not affect the rcd not tripping when tested.

Another classic symptom of N-E faults as the fault can sometimes bypass some of the test current around the RCD.