RCD TD & Single pole RCBO

[AndyPRK]

So what would happen in this scenario ?

Wylex components.
TT installation


Main Supply into house >> 100mA TimeDelay RCD >> Switch fuses >> Consumer Units with single pole RCBO (e.g on socket circuit).

You get a N-E short on the socket circuit.

The RCBO trips, but can't isolate the N-E short.

Will the TD RCD then trip turning off the whole house? And presumably the main RCD can't be reset until the fault removed or that CU main switch turned off.

 

[Jackrae]

Unless the N-E fault exceeds 100mA for the duration of the time delay then I surmise it may (or should not) not trip

 

[JohnW2]

So what would happen in this scenario ?
Main Supply into house >> 100mA TimeDelay RCD >> Switch fuses >> Consumer Units with single pole RCBO (e.g on socket circuit). ... You get a N-E short on the socket circuit. ... The RCBO trips, but can't isolate the N-E short. ... Will the TD RCD then trip turning off the whole house? And presumably the main RCD can't be reset until the fault removed or that CU main switch turned off.

We've discussed this one a good few times before. When there is an up-front TD RCD, one should not, at least in theory, use SP RCBOs - for the very reason you describe.

Kind Regards, John

 

[JohnW2]

Unless the N-E fault exceeds 100mA for the duration of the time delay then I surmise it may (or should not) not trip

That's clearly true but, in reality, true N-E faults will usually result in at least that much (persistent) current imbalance in the RCD - if not immediately then subsequently when someone turns on load(s) in the installation..

Kind Regards, John

 

[Risteard]

Will the TD RCD then trip turning off the whole house?

It is highly probable that it will.

Use an RCBO which switches the neutral as well and then your S-type RCCB will discriminate correctly.

 

[AndyPRK]

Annoying, using RCBO's is actually worse in this case that dual RCD CU (in terms of keeping things up and running)

Thankfully haven't got many RCBO's

 

[JohnW2]

Annoying, using RCBO's is actually worse in this case that dual RCD CU (in terms of keeping things up and running). Thankfully haven't got many RCBO's

If all of your final circuits have RCD or RCBO protection then, AIUI, you will not usually** need to up-front Type S RCD, so you can rid yourself of that problem.

[ **I say 'not usually', because my (TT) installation appears to be an exception to that rule. All my final circuits do have protection (RCDs in most cases, but three SP RCBOs as well). As far as final circuits are concerned, I do not need the up-front TD RCD. However, in fact I do (or, at least, believe that I do, and hence do have!) up-front TD RCD to provide the required fault protection for some long distribution circuits. I therefore theoretically should not have those three SP RCBOs - but I do, and in many years none of them has (IIRC) operated, for any reason, let alone an N-E fault, so I don't regard it as a major issue! ]

Kind Regards, John

 

[Jackrae]

John

Without wishing to highjack the thread you might be able to answer a question in a similar vein

I have a TNS system with RCD on the main CU. From there a MCB feeder runs off to a lighting CU in which are installed a number of RCBO units feeding various lighting circuits (yes a big house). Occasionally, when a light bulb blows one of the respective RCBO units will trip. Question : what situations could cause the RCBO trip since there is obviously no connection between either L or N side of the failed filament. Could it be that the "twanging"" filament supports actually contact each other causing an instantaneous overload.

 

[AndyPRK]

Humm I could get rid of the initial RCD's (actually old 500mA Chilton) at the moment but annoyingly quick to operate!

Wanted to keep something there, as the cable isn't SWA to the remote CU's so should really have RCD protection (although its not really buried in walls)

I think I've made my mind up.

1) Use 100ma TD to a CU that has no RCBO's
2) Ditch RCD to CU that contains RCBO's

 

[AndyPRK]

Occasionally, when a light bulb blows one of the respective RCBO units will trip. Question : what situations could cause the RCBO trip since there is obviously no connection between either L or N side

Should have used a type C.

Lamps blowing can cause type B MCB's and RCBO's to trip.

 

[EFLImpudence]

Occasionally, when a light bulb blows one of the respective RCBO units will trip. Question : what situations could cause the RCBO trip

RCBOs react to overload as well as faults to earth.

since there is obviously no connection between either L or N side

…but there is a connection between L & N - the lamp.

 

[bernardgreen]

Could it be that the "twanging"" filament supports actually contact each other causing an instantaneous overload.

When the filament fails a bit of it vapourises and a plasma discharge is created in the metal vapour. This is effectively a short circuit between Live and Neutral and can be several tens ( if not hundreds) of amps in some designs of lamps. Sometimes with small capsule type lamps the plasma shatters the lamp before the MCB operates.

 

[Jackrae]

Thanks Bernard, that makes sense (to me at least)

 

[JohnW2]

Humm I could get rid of the initial RCD's (actually old 500mA Chilton) at the moment but annoyingly quick to operate! ... Wanted to keep something there, as the cable isn't SWA to the remote CU's so should really have RCD protection (although its not really buried in walls)

Perhaps the situation is a bit similar to my ('exception to the rule') one, then. As I see it, the concepts are no different for TT and TN, it's just that the devices differ. If the feed from meter to CU(s) is short and not particularly 'at risk', then that connection ('tails') needs no fault protection beyond that provided by the DNO's service fuse. If the distance is large (>3m generally gets quoted) and/or is 'at risk', then one has to provide upstream protection - usually a switch-fuse with TN, or the TD RCD with TT.

I think I've made my mind up.
1) Use 100ma TD to a CU that has no RCBO's
2) Ditch RCD to CU that contains RCBO's

That would certainly get (1) 'off the hook'. However, as I see it, as above, how acceptable (2) is depends on the length/routing of the path to CU (2). If it is long/'at risk' enough to need fault protection, then you shouldn't really 'ditch' the RCD.

Kind Regards, John

 

[JohnW2]

John ... Without wishing to highjack the thread you might be able to answer a question in a similar vein

Bernard beat me to it in explaining this to you. It's the overload part of an RCBO (i.e. the 'MCB' bit'), not the 'earth leakage' part, which can operate in response to the very high current which can flow for a very brief period (fraction of a second) whilst some types of lamps/bulbs are dying. The dame happens with MCBs.

I have a TNS system with RCD on the main CU. From there a MCB feeder runs off to a lighting CU in which are installed a number of RCBO units feeding various lighting circuits (yes a big house). Occasionally, when a light bulb blows one of the respective RCBO units will trip. Question : what situations could cause the RCBO trip since there is obviously no connection between either L or N side of the failed filament. Could it be that the "twanging"" filament supports actually contact each other causing an instantaneous overload.[/quote]