Fault Finding Conundrum

[funkydiver]

So the last couple of days we've been having random tripping RCD.

Before getting an electrician in, I'm trying to establish at least a cause, to save both time and money.

I've been trying to isolate appliances, but there doesn't appear to be anything specific causing it. Turn the vacuum on, trip. Reset, turn the vacuum on again in the same socket and it's fine. Turn the shower on, trip. Reset, and it's fine. Attached is a rough schematic of the wiring routing, we have 2 consumer units connected with armoured. Each with their own RCD.

It's ONLY Circuit 1 that isolates the RCD trips on. Note the shower is on circuit 2.

Any ideas?

 

[JohnW2]

Any ideas?

From what you say, I suspect that you may well have a neutral-earth fault somewhere - either in an appliance or the wiring. If that were the case, then 'isolating' appliances with a switch might not cure the problem, because the switches may not be double-pole (i.e. may not disconnect the neutral) - in such cases, unplugging or physically disconnecting are the only ways to fully 'isolate'.

Kind Regards, John

 

[funkydiver]

Thanks John, I have been doing that to appliances.

The thing that is confusing me, is the ability of the trip to cross from one consumer unit to the other. Consumer 2 never trips, also the RCD in consumer 1 is new and exhibiting that same behaviour.

 

[JohnW2]

... Consumer 2 never trips, also the RCD in consumer 1 is new and exhibiting that same behaviour.

Can you explain that a bit more?

Are all the circuits in both CUs protected by the CU's RCD, or do one or both of the CUs have some non-RCD-protected circuits?

Kind Regards, John

 

[funkydiver]

Can you explain that a bit more?

The schematic I attached on my first post shows it diagrammatically.

Are all circuits in both CUs protected by the CU's RCD, or do one or both of the CUs have some non-RCD-protected circuits?

As far as I'm aware, all individual circuits are protected by one of the CUs. I'm not aware of any circuit that's not fed via an RCD.

The feeds are shared before the CUs, and split near the meter.

Hope the picture helps to show the layout.

I do diagnostic testing as part of my job, but this one has me flummoxed.

I would presume that as the earth's are shared so the earth-neutral fault would be on CU1 as that's the one that's tripping, however by increasing the load in either CU1 or CU2 will replicate the fault. Am I looking in the wrong area?

 

[funkydiver]

Makes sense now.. from
https://www.diynot.com/diy/media/neutral_earth_trip.30200/

 

[JohnW2]

The schematic I attached on my first post shows it diagrammatically.

Yes, I saw that and understood the arrangement. However, I still don't really understand ...

Consumer 2 never trips, also the RCD in consumer 1 is new and exhibiting that same behaviour.

... if the 'same behaviour' that CU2 is exhibiting is not 'never tripping' (which obviously is not the case) what behaviour are you referring to?

As far as I'm aware, all individual circuits are protected by one of the CUs. I'm not aware of any circuit that's not fed via an RCD. ... The feeds are shared before the CUs, and split near the meter.....I would presume that as the earth's are shared so the earth-neutral fault would be on CU1 as that's the one that's tripping, however by increasing the load in either CU1 or CU2 will replicate the fault.

Yes, I think that an N-E fault would have to be on one of the circuits fed by the CU whose RCD trips. As for your "however..." .... now that you have read a bit about neutral-earth faults, you probably understand a bit better ... assuming that the situation is not complicated by some circuits which are not RCD protected, then, even if the load you increase is on CU which doesn't trip (whose RCD will not trip, since all of both L and N current from that CUs loads will go through its RCD) some of the neutral current which should have gone from that CU to the supply neutral will be able to 'go backwards' through the RCD in the other CU (the one whose RCD trips), thereby causing an imbalance in L and N currents (hence a 'trip') in that RCD. If the load increases on a circuit fed by the CU whose RCD does trip, then the mechanism is more obvious!

Kind Regards, John

 

[funkydiver]

Same behaviour means...

CU1 trips regardless of whether the additional load is applied on CU1 or CU2.

 

[EFLImpudence]

assuming that the situation is not complicated by some circuits which are not RCD protected, then, even if the load you increase is on CU which doesn't trip (whose RCD will not trip, since all of both L and N current from that CUs loads will go through its RCD) some of the neutral current which should have gone from that CU to the supply neutral will be able to 'go backwards' through the RCD in the other CU (the one whose RCD trips), thereby causing an imbalance in L and N currents (hence a 'trip') in that RCD. If the load increases on a circuit fed by the CU whose RCD does trip, then the mechanism is more obvious!

Assuming everything is wired and connected properly - or are we thinking it is not?

I don't follow that.

If that were the case then with a CU populated by RCBOs we would not be able to tell which circuit had the N-E fault.

 

[funkydiver]

To follow on... Thank you for your replies so far, it is becoming a lot clearer.

 

[EFLImpudence]

CU1 trips regardless of whether the additional load is applied on CU1 or CU2.

No. Not if the wiring is as it should be.

 

[JohnW2]

Assuming everything is wired and connected properly - or are we thinking it is not? I don't follow that.

I don't really follow your "not following"!

Do you not agree that if if there is a N-E fault (downstream of the RCD) on one of the circuits of CU B, then some of the current which should be flowing (from CU A loads) into the supply neutral to CU A will, instead, flow 'backwards' (to earth) through the RCD of CU B, thereby creating a current imbalance (and hence trip) in CU A's RCD?

If that were the case then with a CU populated by RCBOs we would not be able to tell which circuit had the N-E fault.

If, in the above, CU B was all-RCBOs, then it would surely just be the RCBO protecting the circuit which had the N-E fault that would trip (by the same mechanism described above), wouldn't it?

Kind Regards, John

 

[funkydiver]

No. Not if the wiring is as it should be.

The house itself used to be a Residential Home, and was tested annually, the armoured extension to the second CU was installed then as part of an extension build. The only changes since it was domesticated is the RCD has been changed to discount a faulty RCD and a new boiler installed, replacing an aging dinosaur that failed causing a leak into it's internal electrics a few months back (this was when the RCD was changed as the electrician that attended didn't/couldn't find a fault)

When I did my initial fault finding I didn't know as much as I do now, I'm going to have another decent look tomorrow morning when there isn't a plethora of people in the house.

 

[EFLImpudence]

I don't really follow your "not following"!
Do you not agree that if if there is a N-E fault (downstream of the RCD) on one of the circuits of CU B, then some of the current which should be flowing (from CU A loads) into the supply neutral to CU A will, instead, flow 'backwards' (to earth) through the RCD of CU B, thereby creating a current imbalance (and hence trip) in CU A's RCD?

I'm still thinking about it and getting confused, but I don't think so.
It's not the same as the diagram because there both circuits are covered by the same RCD.

If, in the above, CU B was all-RCBOs, then it would surely just be the RCBO protecting the circuit which had the N-E fault that would trip (by the same mechanism described above), wouldn't it?

Yes, but you are saying that an appliance elsewhere - hence anywhere - would cause it to trip.
Am I missing something?

 

[funkydiver]

both circuits are covered by the same RCD

I have two consumer units, each with their own RCD.