The full blurb is below, but the TLDR is: RCD that tested perfectly (ramp test to 27mA can't remember the numbers for trip times but all was good), nothing I can find that would be causing the tripping, someone replaced the RCD with an apparently identical one and the tripping stops... I'm confused by this.
The full story in all its boring detail:
RCD at a customers house would trip out at random times, several times per day. Testing the RCD it seemed to be working perfectly, 27mA on the ramp test and I can't remember exactly what the other numbers were, but all good.
As one of the socket circuits had tripped the mcb once I figured that's probably the problem circuit, so I set about the usual tests, find no insulation problem with the circuit (although I did find it had a broken line ring so I found and fixed that), and found no insulation problems with any of the equipment plugged in on that circuit either. While looking for the broken line I stumbled across a fused spur behind which the cpc had come out and was hovering close to the line terminal, no visible sign of arcing/contact but I make a mental note of "maybe this is the problem?"
I then, with the MS and RCD off and the mcbs on, connected the busbar and neutrals and insulation tested again, with all equipment that would normally be plugged in connected (at 250V only to ensure nothing gets damaged). The total result was about 0.1Mohm (almost instantly went up over 1Mohm and further, but that's the cold reading the moment the test started), and remember this is with all equipment plugged in before you're alarmed by that low reading. Anyway the leakage on that should be 240V/0.1 to 0.13 Mohm is about 2mA... at tops a tenth of the RCD trip current.
We'd already ruled out it being any particular item being plugged in/turned on manually, so my thoughts at this stage are that it's a piece of equipment that decides of its own when to become active, such as the fridge/freezer/boiler/immersion heater. So I force the fridge/freezer on using the quick-cool or w/e you call it buttons, no trip. Then turn off the boiler+immersion heater circuit with everything else on and go around looking for any equipment they may plug in that wasn't in when I checked the IR. When I'm out of anything I can think to check it hasn't tripped off, so I advise that they keep the boiler+immersion heater circuit off and if the tripping stops get the boiler checked out.
Of course it does trip again. At this point I advise that they try turning the circuits off one at a time and keeping each one off until either the RCD trips or the tripping appears to have stopped and keep a record of it, then I'd at least have a clue where to look for the problem, and perhaps put the guilty circuit into an RCBO instead if the problem can't be found.
Anyway after not hearing from them for a while I call and am told that another electrician simply replaced the RCD with one that I'm told is identical (at least that's what they told the homeowner, not sure the homeowner would be able to tell if it wasn't, I tried asking if it says 100mA instead of 30mA and they got a bit confused so I didn't press the issue). The tripping has apparently stopped... but you may recall from the start of this ramble that the RCD is the first thing I checked and it tested perfectly!
Does anyone have any idea what could be causing an RCD that tests well to nuisance trip? Or what else may be going on? Or am I right to wonder if the other guy put in a 100mA RCD?
The full story in all its boring detail:
RCD at a customers house would trip out at random times, several times per day. Testing the RCD it seemed to be working perfectly, 27mA on the ramp test and I can't remember exactly what the other numbers were, but all good.
As one of the socket circuits had tripped the mcb once I figured that's probably the problem circuit, so I set about the usual tests, find no insulation problem with the circuit (although I did find it had a broken line ring so I found and fixed that), and found no insulation problems with any of the equipment plugged in on that circuit either. While looking for the broken line I stumbled across a fused spur behind which the cpc had come out and was hovering close to the line terminal, no visible sign of arcing/contact but I make a mental note of "maybe this is the problem?"
I then, with the MS and RCD off and the mcbs on, connected the busbar and neutrals and insulation tested again, with all equipment that would normally be plugged in connected (at 250V only to ensure nothing gets damaged). The total result was about 0.1Mohm (almost instantly went up over 1Mohm and further, but that's the cold reading the moment the test started), and remember this is with all equipment plugged in before you're alarmed by that low reading. Anyway the leakage on that should be 240V/0.1 to 0.13 Mohm is about 2mA... at tops a tenth of the RCD trip current.
We'd already ruled out it being any particular item being plugged in/turned on manually, so my thoughts at this stage are that it's a piece of equipment that decides of its own when to become active, such as the fridge/freezer/boiler/immersion heater. So I force the fridge/freezer on using the quick-cool or w/e you call it buttons, no trip. Then turn off the boiler+immersion heater circuit with everything else on and go around looking for any equipment they may plug in that wasn't in when I checked the IR. When I'm out of anything I can think to check it hasn't tripped off, so I advise that they keep the boiler+immersion heater circuit off and if the tripping stops get the boiler checked out.
Of course it does trip again. At this point I advise that they try turning the circuits off one at a time and keeping each one off until either the RCD trips or the tripping appears to have stopped and keep a record of it, then I'd at least have a clue where to look for the problem, and perhaps put the guilty circuit into an RCBO instead if the problem can't be found.
Anyway after not hearing from them for a while I call and am told that another electrician simply replaced the RCD with one that I'm told is identical (at least that's what they told the homeowner, not sure the homeowner would be able to tell if it wasn't, I tried asking if it says 100mA instead of 30mA and they got a bit confused so I didn't press the issue). The tripping has apparently stopped... but you may recall from the start of this ramble that the RCD is the first thing I checked and it tested perfectly!
Does anyone have any idea what could be causing an RCD that tests well to nuisance trip? Or what else may be going on? Or am I right to wonder if the other guy put in a 100mA RCD?