New House Build

[Taylortwocities]

Isn't the issue with using the armour as the earth solely one of being able to deliver the correct EFLI/disconnect times??

 

[JohnW2]

Regarding the use of cpc plus 'extra' cpc, the magnitising effects of the armouring, versus copper, mean you can't just add them together for calculation. It should be calculated as if the 3rd core alone exists in the cable.

I think we could probably do with some elaboration about that.

Kind Regards, John

 

[JohnW2]

Isn't the issue with using the armour as the earth solely one of being able to deliver the correct EFLI/disconnect times??

It is, indeed (and sometimes also it's adequacy as a main bonding conductor) ... but not just at the time of installation, also after many years, perhaps decades, of service.

Those who have concerns about relying on armour alone will (despite mfarrow's reassuring comments above, albeit relating to only 5 years' service) wonder about possible long-term corrosion of the armour (per Bernard's comments) and/or the long-term behaviour of the rather unusual way in which the armour is 'properly' terminated, which is more reminiscent of a plumbing fitting 'fouled' with bits of wire, than anything else electrical.

Having said that, the armour of SWA has clearly stood the test of time - at least, in the vast majority of cases, and I am sure that many will rely on the armour alone for a long time to come (regs allowing!).

Kind Regards, John

 

[mfarrow]

I think we could probably do with some elaboration about that.

So could I , but that's what written in GN8.

Yes, in terms of using armour as a cpc by itself, it's only the EFLI/resistance we need to worry about. It's when you start 'requiring' an extra cpc that it gets complicated.

Basically, if you need an R2 of 0.4 ohms, and you have SWA with resistance of 0.8 ohms, you shouldn't just add a separate cpc/core of 0.8 ohms. You are advised to use a separate cpc/core of 0.4 ohms, and forget the SWA exists. Assuming they act together, GN8 leads you to believe, is where things start to get complicated!

 

[JohnW2]

I think we could probably do with some elaboration about that.

So could I , but that's what written in GN8.

Interesting - I need to do some investigating!

Basically, if you need an R2 of 0.4 ohms, and you have SWA with resistance of 0.8 ohms, you shouldn't just add a separate cpc/core of 0.8 ohms. You are advised to use a separate cpc/core of 0.4 ohms, and forget the SWA exists. Assuming they act together, GN8 leads you to believe, is where things start to get complicated!

I really shouldn't even suggest this before I do my 'investigating', but (although, as always, I may be wrong) it sounds like total nonsense to suggest (if that's what GN8 is suggesting) that if you put the armour and a core in parallel, the effective contribution of the armour to the overall effective impedance is zero - particularly given that magnetic effects (such as you appear to have been alluding to) are not going to be very pronounced with essentially straight (albeit coaxial) conductors over sensible distances at 50 Hz.

If you were merely suggesting that the effective combined impedance (at 50Hz) of 0.8Ω's worth of steel armour in parallel with 0.8Ω's worth of copper core was fractionally more than 0.4Ω, I would probably be a lot less incredulous.

Kind Regards, John

 

[mfarrow]

Found it...

...it cannot be accurately predicted how the current will divide between what is effectively two parallel conductors (i.e. the core or separate conductor and the armouring) due to the magnetic effect of the armouring. It is therefore important that the additional core or separate CPC is sized as if it alone were to take the earth fault current. In other words, it is not permissible to simply add the csa of the two cores together.

I must admit I agree in finding the advice to be a little bit extreme. Surely working it out can't be too complicated, can it? I'm hoping one of you with an electrical engineering degree will let me know!

 

[JohnW2]

...it cannot be accurately predicted how the current will divide between what is effectively two parallel conductors (i.e. the core or separate conductor and the armouring) due to the magnetic effect of the armouring. It is therefore important that the additional core or separate CPC is sized as if it alone were to take the earth fault current. In other words, it is not permissible to simply add the csa of the two cores together.

I must admit I agree in finding the advice to be a little bit extreme.

Quite so. As I said, I really don't believe that, at 50Hz, the effect can be anything other than tiny.

Surely working it out can't be too complicated, can it? I'm hoping one of you with an electrical engineering degree will let me know!

Again, quite so. I'll see if I can make any sense of it.

Kind Regards, John

 

[JohnW2]

...it cannot be accurately predicted how the current will divide between what is effectively two parallel conductors (i.e. the core or separate conductor and the armouring) due to the magnetic effect of the armouring. It is therefore important that the additional core or separate CPC is sized as if it alone were to take the earth fault current. In other words, it is not permissible to simply add the csa of the two cores together.

... Surely working it out can't be too complicated, can it? I'm hoping one of you with an electrical engineering degree will let me know!

One other thing has occurred to me, which could be far more important that the effect on impedance (which I still think will be very trivial)...

... once one has established 'how the current will divide between the two parallel conductors' (which, as I've said, I think will probably be very close to the expected 'inversely proportional to CSAs'), when it comes to ascertaining the adequacy of the conductors in terms of fault protection, that is not totally straightforward. One certainly cannot just add together the 'effective CSAs' and then ascertain whether that 'total effective CSA' is adequate - one would have to undertake adiabatic calculations separately for the armour and core - since, even though the 'total effective CSA' might seem adequate, adiabatic calculation may show that one of the 'parallel' conductors' (armour and core) was not adequately protected.

Kind Regards, John

 

[EFLImpudence]

Wouldn't this only apply to large cables with very high currents because, with PVC at least, the armour IS suitably sized for a CPC up to 2 core of 95mm², 3 core - 185mm² and 4 core - 400mm²?

XPLE is considerably smaller.

 

[JohnW2]

Wouldn't this only apply to large cables with very high currents because, with PVC at least, the armour IS suitably sized for a CPC up to 2 core of 95mm², 3 core - 185mm² and 4 core - 400mm²? XPLE is considerably smaller.

That is, indeed, what the tables say (I haven't checked by calculation) - although, as you imply, with XPLE there are problems right down to 10mm².

We (at least, I) don't know what precedes the snippet of GN8 that mfarrow has quoted - so I guess they may have pointed out that, at least with PVC cable, the issue they are mentioning (in the quoted material) is not likely to arise, since the armour (or a core) alone will usually be adequate as a CPC.

... so it could well be that the point mfarrow has raised is a bit of a red herring, certainly for domestic installations, and probably for a lot beyond that. In fact, even in this thread, no-one has suggested using armour+core (rather than just armour) because of the functional inadequacy of armour alone, so I'm not totally sure why mfarrow raised the issue!

Kind Regards, John

 

[mfarrow]

so it could well be that the point mfarrow has raised is a bit of a red herring

No, the values in the table are where using Table 54.7. You can of course use the adiabatic, and then many of the 2c cables, with many circuit arrangements, will comply.

Also, anyone bought any BS 6346 cable recently? The standard cable I've seen over the last half-decade or more has been BS 5467. Therefore I do not believe the "doesn't apply to domestics" statement is true.

no-one has suggested using armour+core (rather than just armour) because of the functional inadequacy of armour alone, so I'm not totally sure why mfarrow raised the issue!

I'm more comfortable to have one of the cores (as well as armour) as CPC

 

[JohnW2]

so it could well be that the point mfarrow has raised is a bit of a red herring

No, the values in the table are where using Table 54.7. You can of course use the adiabatic, and then many of the 2c cables, with many circuit arrangements, will comply.

Exactly - so I don't really understand your point. As you say, if one is not lazy, and actually does adiabatic calculations, the armour alone will often prove to be adequate, at least for cable sizes likely to be used domestically, even with XPLE - hence underlining EFLI's point that it will rarely be necessary (at least domestically) to consider using a core to 'top up' armour of inadequate CSA.

Also, anyone bought any BS 6346 cable recently? The standard cable I've seen over the last half-decade or more has been BS 5467. Therefore I do not believe the "doesn't apply to domestics" statement is true.

See above. If one does the calculations, armour alone is usually/often going to be adequate in domestic situations, even with XPLE.

no-one has suggested using armour+core (rather than just armour) because of the functional inadequacy of armour alone, so I'm not totally sure why mfarrow raised the issue!

I'm more comfortable to have one of the cores (as well as armour) as CPC

You appear to have totally misunderstood what people have been saying. I certainly haven't suggested in this thread (and I don't think anyone else has) using armour+core because armour alone would be 'functionally adequate' (i.e. have adequate CSA) - indeed, the whole point of this thread is that the OP wanted to know whether he should be considering using a CPC core, even though armour alone would be functionally adequate (at least, at the time of installation). The reason I am 'more comfortable' with armour+core (in a situation in which armour alone is functionally satisfactory at the time of installation) is as a 'belt and braces' measure in case 'anything happens' to the armour (and/or its terminations) over years or decades.

Kind Regards, John

 

[Taylortwocities]

His original intention was to use 2-core SWA and a separate earth cable.
Your advice (that I share) was not to use a separate earth cable,

I'm more comfortable to have one of the cores (as well as armour) as CPC

That should have been enough...

 

[JohnW2]

His original intention was to use 2-core SWA and a separate earth cable. Your advice (that I share) was not to use a separate earth cable,

I'm more comfortable to have one of the cores (as well as armour) as CPC ...

It wasn't quite like that. The OP acknowledged that armour alone could be used as the CPC, but said that he had "read that this was not recommended", hence his question about whether he should use a 'separate earth'. I assume that we would all agree that a 'separate earth cable' is not the way to go, so the discussion naturally shifted to the choice between armour alone and using 3-core cable.

That should have been enough...

From my standpoint, my comment you quote above would, indeed, be enough, since it fully expresses my feelings. However, I need not tell you that many electricians (not you?) would favour/suggest advise the third option (neither 'separate earth cable' nor my 'armour+core'), namely to just rely on the armour, as originally mentioned in the OP. That's what caused most of the subsequent discussion - and, as for the complication introduced by mfarrow, I still think that is probably not usually relevant in a domestic setting, since no-one has suggested that the armour alone would not be adequate (at least, at the time of installation).

Kind Regards, John