Sub-main replacement

[Tony Bryer]

My flat is at the back of a converted house. The meters and a 60A switch-fuse are in the front hall and there is a 16mm2 T&E feed to my CU. This needs to be replaced as it runs through the adjacent flat in what will be an inconvenient location (so they will pay some of the cost).

The new routing involves feeding a cable under a suspended floor and because of finishes etc it cannot be clipped. Am I right in thinking that a PVC cable lying on the subfloor (as at present) is not going to be acceptable and it should therefore be SWA? And would there be any good reason to make it more than 16mm2 - flat is gas heated so not much elec used.

If it is to be SWA I'm proposing to get the cable in place myself and get a professional to do the terminating.

[Andy Wade]

The OSG sanctions the use of T&E in damp locations, but it shouldn't be used in wet conditions - anywhere liable to flooding, for instance. SWA would certainly be a more appropriate choice of cable for this application.

A big disadvantage of T&E is the small earth size (6 mm^2 in this case). Some questions:

- TN-S or TN-C-S/PME supply?

- type of fuse in the switch-fuse (fuse rating is 60 A, presumably)?

- length of run?

- is the main bonding connected at the meter end or the flat end?

[Stephen Dawson]

When we installed these for EDF we used to use 16mm t&e with a seperate 10mm earth fed from a 60a switchfuse.

Steve Dawson

[Andy Wade]

That's certainly a possibility. If you parallel up the CPC in the T&E cable with the separate 10 mm wire, and the two take substantially the same route then you have, in effect, a full-size 16 mm^2 CPC. That complies with Table 54G, so no earth-sizing calculation is required.

The voltage drop on 16 mm^2 single-phase at 60 A is 0.168 V/m (using Table 4D2B) so on a (say) 15 m run you've used up about a quarter of the allowable voltage drop. This needs to be taken into account in the overall voltage drop calculations. Ditto for the submain's contribution to the earth fault loop impedance, which will be about 0.04 ohm over the same length - not usually enough to matter, but it ought to be checked.

[Tony Bryer]

TN-S

60A fuse (cartridge IIRC - never been looked at in 25 years)

About 15m

[Andy Wade]

OK - but it would probably be prudent to allow for a later upgrade to PME.

Hmm, there's an interesting design problem there. Assuming it's a BS

1361 fuse the total earth fault loop impedance (Zs) required for 5 s disconnection is 0.73 ohm, yet the standard quoted max. external loop impedance (Ze) for TN-S is 0.8 ohm... In practice Ze is likely to be much lower (especially in London, which is where I think you are?) and there isn't likely to be a problem - but it would be useful to get a measured Ze value to be sure. Anyway, I'll proceed by assuming that Zs is just low enough to achieve 5 s disconnection. Applying the adiabatic equation shows that the 6 mm^2 CPC in the T&E cable is definitely not quite large enough. So if you stay with T&E I suggest a separate parallel 16 mm^2 CPC is run.

The current rating also needs to be considered. The clipped direct rating of 16 mm^2 T&E is OK at 85 A, but make sure that installation conditions such as thermal insulation or grouping with other cables doesn't bring it down below 60 A.

Voltage drop is not too likely to be a problem then, but needs checking (see my other post yesterday).

Again, if you want the ability to upgrade to PME, that rules out using the 6 mm^2 CPC.

[Christian McArdle]

I'm not entirely sure. However, cables in this location can be subject to rodent attack, which would be a good reason to use SWA.

I would use 3 core 16mm SWA, assuming the voltage drop and earth loop calculations pass. Pass the earth down both the armour AND an additional core. This provides protection, whilst giving a much more reliable copper conductor for your main earth. The steel armour can rust at the connection, although this can depend on the quality of the earth terminal used.

Christian.

[Tony Bryer]

That's the way I'll be going - thanks to you and Andy for the help. I was surprised, on checking, to find that 16mm2 3-core SWA is only slightly more expensive than 16mm2 T&E - £5.65 v. £4.86. Cost of glands to be added of course, but it will be a much better job.

[Andy Wade]

Seems like a good decision. Usually I would say 2-core would be fine, but in this case, with the slight concern over Zs, the lower resistance of the 3rd core is certainly beneficial.

Incidentally when using new-colour 3-core on single phase, recommended practice is to use the brown core for phase (obviously), the grey core (sleeved blue) for neutral and the black core (sleeved green/yellow) for the CPC. The idea is to dissociate the colour black from neutral.