Pinched wire

[plugwash]

I'm not quite sure what you have in mind in your (2).

Earthing systems usually end up bonded to other metal objects which are either burried in the ground or run between properties.

When neutral and earth are combined these "other metal objects" provide parallel paths back to the supply and therefore end up carrying a proportion of the neutral current. Depending on the details of the "other metal objects" are this can be a bad thing.

Afaict this is the reason for minimum size requirements on main bonding conductors in TN-C-S supplied properties and why TN-C-S supplies are forbidden from supplying petrol stations.

 

[JohnW2]

I'm not quite sure what you have in mind in your (2).

Earthing systems usually end up bonded to other metal objects which are either burried in the ground or run between properties.

Indeed so.

When neutral and earth are combined these "other metal objects" provide parallel paths back to the supply and therefore end up carrying a proportion of the neutral current. Depending on the details of the "other metal objects" are this can be a bad thing.

That surely should read "When neutral and earth are connected (within the premises)" - so what you say is (as you more-or-less go on to imply) equally true of any TN-C-S installation as it would be of an installation which had CNE conductors within in? I therefore don't see that your point (2) really was a specific reason for not have combined neutral and earth conductors within an installation. However, your (1) and (3) remain valid.

Kind Regards, John

 

[plugwash]

That surely should read "When neutral and earth are connected (within the premises)" - so what you say is (as you more-or-less go on to imply) equally true of any TN-C-S installation as it would be of an installation which had CNE conductors within in?

CNE conductors can certainly cause currents to flow in ways that were not intended whether or not they are within the premises.

However it gets worse when you take the CNE cores into the premises. With TN-C-S most of the "paralell path currents" will flow through the main bonding and you will see little in the way of voltage gradiants within the installation itself. On the other hand with TN-C you will get voltage gradiants and hence currents taking unplanned paths happening within the installation itself.

For example you plug two computers (which have mains earth and signal ground connected together as is typical in desktop PCs) into sockets that happen to be on different circuits, then connect a serial cable between them. In a TN-C system some proportion of the neutral current will start flowing down the ground wire of the serial cable.

 

[JohnW2]

Interconnection of earths can happen within an installation too. For example you plug two computers (which have mains earth and signal ground connected together as is typical in desktop PCs) into sockets that happen to be on different circuits, then connect a serial cable between them. In a TN-C system some proportion of the neutral current will start flowing down the ground wire of the serial cable.

I am not really sure what you are saying here, or what is its relevance to CNE conductors. Why should significant current flow through a connection between two earths (neither of which have significant current flowing through them), even if they are on different final circuits - both N and E of those final circuits are, after all, joined at the CU (even if through an RCD). It's not as if there is a N-E connection in the computers. .. and, anyway, there are all sorts of 'interconnections of earths' within most installations. I think I need a bit of help in understanding what you are suggesting.

Kind Regards, John
It has occurred to me that, although you didn't say so, you might have been talking about the hypothetical situation in which there were CNE conductors within the installation (hence there would be a connection of N and E at the computer, as well as everywhere else), rather than making a 'general' comment. In the case of an installation with CNEs, I would agree with you.

 

[JohnW2]

However it gets worse when you take the CNE cores into the premises. With TN-C-S most of the "paralell path currents" will flow through the main bonding and you will see little in the way of voltage gradiants within the installation itself. On the other hand with TN-C you will get voltage gradiants and hence currents taking unplanned paths happening within the installation itself.

I'm not sure that I fully understand that. Most 'parallel path currents' will flow through the main bonding whether its TN-C-S or T-N-C. Don't forget that we are discussing your point (2), namely "2: Currents in the neutral can create voltage gradiants in the earth which can push current down paths that were not originally intended to be current carrying" - which I still don't really understand.

As a different issue, I agree that having CNE conductors within an installation renders it less than a strictly equipotential zone, since exposed-conductive parts are then connected to something which is subject to VDs in the CNE conductor, and therefore may differ. However, if one sticks within the VD guidance of BS7671, then (in the absence of faults) the greatest potential difference one could see between two exposed-c-ps as a result of this would only be 5.75V - hardly enough to represent a hazard. However, I do agree that any 'interconnection' of those earths/exposed-c-ps could then result in currents (part of the neutral current) flowing through those connections, which is perhaps what you meant by your point (2)?

I think I need to again reiterate that I am in no way arguing for CNE conductors within installations, which is clearly uauslly bad news for several reasons. Rather, I'm exploring what those reasons are.

Kind Regards, John

 

[plugwash]

However, if one sticks within the VD guidance of BS7671, then (in the absence of faults) the greatest potential difference one could see between two exposed-c-ps as a result of this would only be 5.75V - hardly enough to represent a hazard.

Not a shock hazard to humans no.

However the "source resistance" of that 5.75V difference is likely to be well under 1 ohm. So if those two "exposed CPs" are connected together (for example by a signal cable or just by them touching each other) then currents measured in amps could easilly flow through the interconnection. Several amps through a signal cable could easilly be enough to melt it.

 

[JohnW2]

Not a shock hazard to humans no. However the "source resistance" of that 5.75V difference is likely to be well under 1 ohm. So if those two "exposed CPs" are connected together (for example by a signal cable or just by them touching each other) then currents measured in amps could easilly flow through the interconnection. Several amps through a signal cable could easilly be enough to melt it.

Yes, that's true. In fact, it's probably potentially worse than you imply. If, for the purpose of simplicity, the Rn of the two final circuits (from point of interconnection to CU) in question were equal, and the impedance of the interconnection negligible, then the neutral current of the loaded circuit would be shared equally between the two neutrals, hence half of the load current going through the interconnection (signal cable or whatever). So, if the load current were, say, 32A, then 16A would flow through the interconnection.

Kind Regards, John