RCD tripping

[John-SJW]

A neighbour in an apartment block in my road contacted me. They have an ML CU with two RCDs, which is about one to two years old. We had the power go off for a few seconds them come back on. I never noticed anything myelf as my lights were all off. Anyhow they contacted me saying they had no power. I went around finding one RCD had tripped. I switched on and all is fine again. They do have electric underflor heating and one electric wall heater. I was wondering why it tripped. Would a spike do this?

 

[happysteve]

Potentially, yes. Similar thing happens at my in-laws' house, when they have power cuts (fairly frequent, as there's a persistant fault on the underground supply that the DNO haven't managed to fix).

If power is cut off or restored very sharply, then high frequency components are introduced to the usual 50Hz waveform. Cables such as twin and earth have inherent capacitance. As the frequency of the waveform goes up (due to the sharp bits in the waveform due to the fault), the capacitive reactance (Xc) goes down, leading to leakage between live and earth. This can be detected by RCDs, which then trip.

It happened once while they were on holiday a couple of years back. Lost all the stuff in their freezer.

 

[telectrix]

Potentially, yes. Similar thing happens at my in-laws' house, when they have power cuts (fairly frequent, as there's a persistant fault on the underground supply that the DNO haven't managed to fix).

If power is cut off or restored very sharply, then high frequency components are introduced to the usual 50Hz waveform. Cables such as twin and earth have inherent capacitance. As the frequency of the waveform goes up (due to the sharp bits in the waveform due to the fault), the capacitive reactance (Xc) goes down, leading to leakage between live and earth. This can be detected by RCDs, which then trip.

It happened once while they were on holiday a couple of years back.

whitch is why our freezers are on a RCD free circuit.

 

[John-SJW]

Potentially, yes. Similar thing happens at my in-laws' house, when they have power cuts (fairly frequent, as there's a persistant fault on the underground supply that the DNO haven't managed to fix).

If power is cut off or restored very sharply, then high frequency components are introduced to the usual 50Hz waveform. Cables such as twin and earth have inherent capacitance. As the frequency of the waveform goes up (due to the sharp bits in the waveform due to the fault), the capacitive reactance (Xc) goes down, leading to leakage between live and earth. This can be detected by RCDs, which then trip.

It happened once while they were on holiday a couple of years back. Lost all the stuff in their freezer.

I assume this would happen to an RCBO?

 

[John-SJW]

whitch is why our freezers are on a RCD free circuit.

I know someone who did this. The fridge/freezer is on its own circuit back to the CU on a simple 3A or 6A mcb. He said the freezer had no metal parts that a user could touch, so from a users point of view was double insulated not needing an RCBO or RCD.

 

[John-SJW]

Potentially, yes. Similar thing happens at my in-laws' house, when they have power cuts (fairly frequent, as there's a persistant fault on the underground supply that the DNO haven't managed to fix).

If power is cut off or restored very sharply, then high frequency components are introduced to the usual 50Hz waveform. Cables such as twin and earth have inherent capacitance. As the frequency of the waveform goes up (due to the sharp bits in the waveform due to the fault), the capacitive reactance (Xc) goes down, leading to leakage between live and earth. This can be detected by RCDs, which then trip.

It happened once while they were on holiday a couple of years back. Lost all the stuff in their freezer.

Would the new SPUs solve this?

 

[Ziggie]

There is documented that a fridge freezer should be put on a separate circuit ( using a split CU. One bit of the CU for RCD protection and one without for such items.

i have a freezer in the farage and fitted an rcd which tripped when I plugged tools / lawnmower in.

interestingly, this didn’t happen using an RCD’d extension cable on a socket on the same circuit.

 

[telectrix]

our set up is 40A MCB (not RCD) in CU with 6.00mm T/E to sub in garage. RFC in garage (freezers, my bench, etc.|) . external IP66 RCD sockets for mowers etc.

 

[Lucien Nunes]

Would the new SPUs solve this?

Not necessarily. SPUs will chop the tops off excessive voltage peaks, but as happysteve explains, what tends to trip RCDs during spikes and brownouts is not the voltage itself but the sudden changes in voltage (dV/dt). The high rate of change causes high peak currents to flow through any capacitance to earth such as within interference suppressors, which the RCD detects as leakage (which technically it is.) Clearly if there is a high peak voltage with a high dV/dt, then the SPD might reduce it to the point where it doesn't cause a trip.

a freezer in the farage

Did you mean a greezer? A greezer in the Farage is worth...?

 

[telectrix]

Did you mean a greezer


thought Trump was keeping the Mexicans out.

 

[DPG]

I know someone who did this. The fridge/freezer is on its own circuit back to the CU on a simple 3A or 6A mcb. He said the freezer had no metal parts that a user could touch, so from a users point of view was double insulated not needing an RCBO or RCD.

But did the cable supplying the freezer socket need RCD protection?

 

[Ziggie]

It was written in edition 17 that a dedicated socket did not need an RCD as long as it was labelled.

this exception was removed in 18th.

 

[telectrix]

A greezer in the Farage is worth...?

1 Nigel.

 

[telectrix]

But did the cable supplying the freezer socket need RCD protection?

mine don't. all cabling in garage is surface. not as neat as maybe, but it works.

 

[happysteve]

Not necessarily. SPUs will chop the tops off excessive voltage peaks, but as happysteve explains, what tends to trip RCDs during spikes and brownouts is not the voltage itself but the sudden changes in voltage (dV/dt). The high rate of change causes high peak currents to flow through any capacitance to earth such as within interference suppressors, which the RCD detects as leakage (which technically it is.) Clearly if there is a high peak voltage with a high dV/dt, then the SPD might reduce it to the point where it doesn't cause a trip.

I'd been pondering this for a while and had come to the same conclusion... I almost added some vague thoughts along those lines but lacked the confidence.

So what would prevent this - RCD tripping in due to transients - happening?

(1) less capacitance by design (shorter cable runs/few circuits)
(2) some sort of device that had capacitance to earth before the RCD, so the excess current would flow to earth there rather than in the circuits after the RCD in the event of a transient
(3) a choke (i.e. inductor) to limit the flow of current as the frequency increases

(1) is I guess something to consider when designing from scratch.

I suspect there's some sort of filter that effectively does some sort of combination of (2) and/or (3)... it's just not an SPD. It's the sort of thing you'd buy from RS or Farnell.

 

[nicebutdim]

I assume this would happen to an RCBO?

Can anyone address this question?

Some ideas are running round my head (that I'm not foolish enough to share) and it would be interesting to learn the thoughts of more experienced minds.

 

[Lucien Nunes]

I suppose it's worth considering that any cables / appliances with built-in SPDs to earth, where no main SPD is fitted, will likely cause a trip on a spike. So perhaps there is more advantage to that than I first thought. Of course, it won't help much if any of the integrated varistors have a lower clamping voltage or faster response than the main ones.

Shunt capacitance at the origin, connected via the minimum possible inductance (e.g. feedthrough construction) will soften the edges but at the expense of needing to be pretty durable and hefty caps. Inductors that would provide tangible benefits would be more expensive still. I can't see it being taken up for general installation work, as it's a functionality upgrade rather than a safety one.

 

[John-SJW]

But did the cable supplying the freezer socket need RCD protection?

The cable dropped down in trunking from the CU, under the floor and up to the freezer to a low level concealed socket behind the fridge/freezer. It was not accessible and unlikely to have any future screws run into it, unless they were a foot long.

 

[John-SJW]

Clearly if there is a high peak voltage with a high dV/dt, then the SPD might reduce it to the point where it doesn't cause a trip.

As SPDs appear to be looking to be mandatory soon in new installations. Then they look like they will reduce the problem of RCDs tripping on power on-offs, but fully eliminate it.