Earthing a (very) remote building

[dennis]

It would require a knife. An alternative is to mount the box with the SWA terminations inside a box.

[Piers Finlayson]

They are, ta.

[Andy Wade]

That's exactly how I've done it a couple of times. M32 (IIRC) 'skintop' type gland (for 16 mm^2 2-core SWA) and 3:1 adhesive-lined heatshrink at the said point.

[Andy Wade]

That's why I was asking about the cable size and the earthing arrangements in the farmhouse. If the cable's big enough to give acceptable voltage drop at the far end then the chances are that a 32 A (say) MCB will give fault protection without need for an RCD at that end. But this should be checked in case the cable's tiddly or the Ze value's high (it's not unusual to find higher than normal Ze values on old farm installations).

[John Rumm]

Indeed, or the farm is already TT, in which case it could be almost anything.

[Tim Lamb]

Ummm... Electrical systems on small farms are a bit like *Topsy* they just grew!

Not strictly within the *industrial premises* category but yet likely to employ labour so beyond simple domestic.

I rather suspect that much of the system here would not meet a strict application of earth fault impedance criteria and can point to remote services still protected by re-wireable fuses.

Rich pickings for future time served electrical apprentices?

regards

[Andy Wade]

And, specifically, Section 705 of BS 7671 applies to any new or altered electrical work on agricultural and horticultural premises.

*All* new or altered circuits must be RCD protected, regardless of the type of earthing:

- 30 mA for sockets up to 32 A

- 100 mA for sockets over 32 A

- 300 mA for all other circuits.

Delayed RCDs don't appear to be precluded here, although the normal provisions for sockets up to 20 A (i.e. 13 A ones) will over-rule and demand a fast RCD.

I suspect that most farms in practice are a long way short of the present requirements.

[Andy Wade]

Hmm, I see what you mean. I hadn't spotted that there was a new version, so thanks for that.

[RJS]

I think that the copper thieves managed to strip out all possible earth routes.

Richard

[RJS]

Strangely enough, my ill-informed 'gut feeling' was that a locally established earth would be safer than trusting one imported from a remote location.

Indeed.

>

[RJS]

Thanks again to all

I think that you have devoted enough of your valuable time to this problem,= at least you have provided sufficient information for me to hassle people = on site about exporting earths and the risks associated with what I'll call= 'floating' earths which will, hopefully, make them think carefully about t= he problem.

Rgds

Richard

[Andrew Gabriel]

Funny this came up - just looking at doing exactly this.

The remote CU is plastic, but has a metalic DIN rail, under which the supply is fed before it goes through the RCD. I have sometimes in the past earthed the DIN rail in case anything comes adrift and touches it. In this case, it would have to be earthed to the cable sheath (exported earth) as the supply isn't RCD protected. However, I wouldn't want to expose this exported earth outside the CU, where all the earthing is TT, so there would be two earthing systems in the CU.

I have for the moment double insulated the non-RCD protected wires under the DIN rail, and maybe earthing the DIN rail is simply being too anal? Earthing it to the TT system would seem like a bad idea as the earth rod impedance is most unlikely to be able to sink enough current to clear supply short to it.

There is a slightly similar situation in my own CU, where everything is TN-C-S except the outdoor socket circuit which is TT and on its own RCBO (C20, 10mA), and whose earth is kept separate from the rest. However, the supply earth is not an exported one in this case.