Can't reset the RCD with any MCB's on - Faulty RCD?

[Geoff101]

We’ve had a few power glitches/cuts recently by our supplier. Each time, on power restoration our RCD has tripped. Irritatingly we were on holiday when it happened last - not good for frozen food!

When we had an extension 3 years ago some changes were made to the consumer unit and a new RCD was fitted into it. This was done by a qualified electrician with all the tests etc carried out. Since then I've never just been able to reset the RCD when tripped - I've always needed to switch all the MCBs off, reset the RCD and then reset the MCBs one by one.
I've now done some investigation trying to reset the RCD with just one of the MCBs in the on position and it appears that almost any individual circuit will prevent the RCD resetting if the MCB is left on.

We don't generally have any problems with other nuisance RCD trips - once it's reset everything is fine.

I haven't been round to remove everything that is plugged in, and some of the circuits that prevent resetting are lighting circuits anyway. I did try one experiment as an example - The oven circuit MCB, if on, prevented the RCD resetting. This circuit only has the oven connected to it. With the oven isolator switched off but the MCB on it allowed the RCD to reset - so clearly not an issue with the circuit itself. Switch the oven isolator on again and the RCD couldn't be reset. It's a fairly new built-in AEG dual oven and am pretty confident it's not faulty.

So the big question is, does this sound like it could be a faulty RCD which doesn't like being reset with any sort of load on it? In the case of the oven it obviously wasn't actually on and heating, just in standby with the electronics active. I'm guessing whatever earth leakage surge there is on power up is triggering the RCD - as it appears is happening in most the other circuits. Interestingly though, using the oven as an example the 'surge' is enough to prevent the RCD being reset, but clearly not enough to trigger it once reset and the MCB is put on.

I don't know how common it is for the RCD to be suspect like this - and if faulty it appears it has been from the start. Worth a swap to see?

Thank for any thoughts....

 

[DPG]

The cumulative leakage of the circuit may be close to the trip value of the RCD (this will be less than 30mA in practice). May be worth getting a spark to check the leakage with a sensitive clamp meter.

 

[Geoff101]

I agree it could be a factor. However, of the 10 circuits on the CU, 8 of the 10 each individually prevent the RCD resetting when only their MCB is on. Surely if it was cumulative thing, if each was leaking enough to trigger it individually then with all switched on there would be no hope of the holding even when reset? Or have I got it wrong?

 

[snowhead]

It's quite possibly a Neutral issue.

 

[Lucien Nunes]

What kind of neutral issue? A short would tend to cause tripping when load is highest, not when circuits are energised with minimal load on. High resistance would cause malfunction of the loads and/or overheating of faulty terminations etc.

The OP's systematic investigation and findings make suspect the RCD, and it is worth testing with an RCD tester or by substitution, along with the key electrical parameters of the installation. Many people assume their RCD is faulty for all sorts of spurious reasons, for which the cause usually turns out to be one of a number of standard problems such as an N-E short or aggregate leakage near threshold. None of those seem to fit here.

If the RCD is good, the only likely external influence is the presence of devices that create a leakage impulse when first energised. This can occur with suppression capacitors inside appliances etc. but then the behaviour would likely be specific to one or two circuits. One thing I would check first, if it has not already been checked, is the supply polarity. If all the MCBs are switching the neutral, the leakage transients from operation will be abnormal. Obviously this is also a dangerous situation as the MCBs will not offer fault protection to earth and single-pole functional switching will leave points live.

One way or another, electrical test equipment now needs to be brought to bear on the installation and I will be interested to hear the outcome.

 

[Wilko]

Hi - do you know what earthing system you're using at home - do you have a rod in the garden for example?

 

[Geoff101]

Hi
Thanks for all your thoughts - really helpful.
For a bit more background the house is a 30s extended semi and I guess was rewired in the 80s. We didn’t have this problem until 3 years ago when the kitchen was extended and some circuits were added/changed. A couple of spare ways in the existing mk sentry CU were used and a new RCD was fitted. The new and altered circuits were tested and I have the test cert from this. The issue of having to switch off the MCBs to reset the rcd was there from this point on. Never been a particular problem until recently as we have had quite a run of power outages and realised the consequences if we are away when it happens. On reflection, some circuits were unaltered by this work, with the old rcd in it used to reset fine with the MCBs on but this is no longer the case. These circuits may have been moved on the cu but nothing more than that. This makes me wonder even more if it is the new rcd which is the culprit.
I’m guessing there is a earth rod in the garden but I’m not entirely sure about that. We have had a new supply cable fitted from road to house within the last year as when they came to fit a smart meter they discovered bitumin leaking from the old cable and replaced it, the meter and head, main fuse. So everything on the supply side of the cu is new. And hasn’t changed anything.
The polarity is correct and the MCBs are indeed switching the live!
I wish there was an isolator so I could try a substitute rcd but there’s not.
I think I’ll need to get someone in but I would hopethey might start with a quick substitution as I’m feeling this could be a quick win..... or maybe not!

 

[Lucien Nunes]

Oops... my mention of polarity there was irrelevant as I was thinking about a problem you don't have (trip events when operating the MCBs).

The way forward is definitely electrical testing but I can't resist speculating. There's a scenario where both the RCD and the installation test OK, but a new RCD solves it, e.g a faulty RCD contact that arcs under the stray capacitive load of the wiring as it is being manually reset, but makes adequate contact once fully closed. In that case it would still pass ordinary RCD tests. As the installing electrician should have done these with all loads disconnected, it could have been faulty from new but wouldn't necessarily have misbehaved during commissioning.

 

[Geoff101]

Yes, Lucien, I think that perfectly describes my thoughts. Once set, everything works ok. We don’t have any spurious RCD trips under normal operations, but the RCD definitely works (I’ve occasionally N E tripped it when changing light fittings etc).
It’s definitely the actually resetting action that is the problem.
My plan is to see if I can get the spark who did the work for us last time to swap the rcd and see if that does it. If not I’ll ask him to investigate futher.......

 

[driverman]

Could it not be a neutral / earth fault?