Shed electrics.

[Colin Stamp]

Hi all,

I'd welcome any comments on my plan to get power to my forthcoming shed.

The shed itself will wooden, on a concrete slab and 3.6m X 6.6m. It's around 20m from the house, with around a 35m cable-run from the house CU to the shed CU. Initially it'll just be for storage but it'll get used as a workshop/office later on.

I have a 17th edition CU in the house with some spare non-RCD ways, so I'll feed the sub-main with an MCB in one of them. I'm hoping to squeeze a 40A sub-main out of 10mm SWA, but that doesn't leave much voltage drop for the final lighting circuit in the shed, so I may have to come down to 32A, which should be plenty anyway. I'll be trying to take the SWA all the way from CU to CU, but I'm guessing it's going to be about as flexible as a concrete lintel, so I may have to use T&E in earthed conduit for the bit inside the house.

The SWA will be directly buried 600mm deep under a slabbed path to the shed, alongside a 63mm duct for data cables. I'll put warning tape just under the slabs. Do I need any further down too?

The house is TN-C-S. After reading the regs and trawling the web, I'm a bit unsure about equipotential bonding. On the face of it there doesn't seem to be anything to bond, but the concrete slab is a bit of an unknown. Ideally, I'd like to use 2 core SWA with the armour as the CPC and no bonding conductor, but I'm not sure if that's acceptable with a bare slab. Would covering the slab with OSB help? Or would I need to embed a grid in the concrete and bond that?

If I do need a bonding conductor, I guess I can use 3-core 10mm SWA with the 3rd core as the bonding conductor? I'd really like to avoid TTing the shed if possible.

I'm sure I've managed to be overly anal and recklessly negligent both at the same time, so any comments to get me a sensible setup would be very welcome.

Cheers,

Colin.

[Brian Gaff]

While fully understanding the need for care and safety. I cannot help but observe that I've lived here since 1957, and the only new wiring to sheds was because of the eating of the old one by an unfortunate rat who managed to both singe and electrocute himself at the same time by eating the old rubber covered three core festooned on the fence!

grin.

Brian

[David WE Roberts]

Watching this with interest because we are about to do almost exactly the same. Estimated distances 15m in the house and 20m up the garden which seems to match you exactly. The electrician consulted a set of impressive looking books full of graphs and seemed to think we could get a 40A supply over 10mm. [Which would be just enough to run the 9.5kW electric shower we just took out of the bathroom. :-) ] Still undecide if we export the earth from the house (which is earthed to the outer of the supply cable) or use 2 core and earth locally at the shed.

I note with interest your questions about protection of the cable - the planned run will not initially be slabbed on the surface although some (or most) of it may be later. We will need to decide fairly quickly as we have a digger and a skip loader on hire for a week (delivered today, hire starts Monday) so that will include running in the services trench(es) from the back corner of the house to the shed.

So far we need to run mains water, main drainage, mains electric and data cables (Cat5E) up to the shed. Decided not to run gas up there ;-)

So the next couple of weeks are likely to be even busier than usual.

Cheers

Dave R

[Robin]

That makes me wonder if you are looking to become a "beds in sheds" landlord. But I'm sure no one here would dream of telling TPTB :)

[Andrew Gabriel]

I did something similar. I only exported the house earth to the end of the submain. The wiring in the shed is then a TT system with its own 8' earth rod (8.3 ohms), not connected to the house earth. All circuits in the shed are RCD protected (RCBOs), but the SWA submain isn't. I expect the shed will be used to power outdoor tools such as lawnmower, which was the main reason for using a TT system.

[John Rumm]

Covered in some detail here:

The shed itself will wooden, on a concrete slab and 3.6m X 6.6m. It's

You could use a cable like earthshield for the internal run.

No, once is enough.

A slab is not typically something you would bond. Whether it offers access to a local independent earth or not is another matter - concrete seems to vary a bit in conductance.

My preference in these cases is to do just that... use the source earth for the protection of the submain as far as the outbuilding, but then have that isolated with its own TT earth.

I'm sure I've managed to be overly anal and recklessly negligent both at

Certainly not the latter...

[Colin Stamp]

I'd thought of doing a TT system, but I'd been trying to avoid it to leave the option of a non-RCD circuit for a freezer and, if I'm honest, because I didn't like the thought of them.

I will be putting outside sockets on the shed though. I hadn't realized there was an advantage to TT systems for this. What have I missed? I'll probably end-up putting RCDs on all the circuits anyway, so if there's a decent advantage to TT, it might tip the balance.

Not sure I fancy driving in an 8' earth rod though!

Cheers,

Colin.

[Colin Stamp]

That's going to be some shed. I thought mine was going to be quite complicated because I'm lopping a corner off to clear a tree!

Cheers,

Colin.

[Colin Stamp]

getting me this far.

[Googles earthshield] Oh, I'd never heard of that. If I can get hold of some it could be just the job.

Righto. All the advice I've seen seems to suggest burying the tape sort-of mid way. I reckon just under the slabs is the most likely place to make sure it gets spotted.

Yep. I guess I'd be silly to rely on the slab as being any better than bare earth. I've hard of "equipotential planes" and I wondered if a bit of bonded weld mesh would help me avoid a TT system - if that's worth doing, which I'm starting to doubt.

I think I can feel myself being persuaded...

question I'm left with there, is whether I need to extend the equipotential zone at-all. If I cover the slab with a wooden floor then I would assume not, but I guess I can't rule out someone (me even) making a change later on that brings in a local earth from somewhere.

but I guess I'll need 10mm of copper for any bonding conductors, hence using the third core of 3-core SWA. One thing I don't understand is that the CSA for the bonding conductor doesn't seem to be related to its length. I'd have thought its resistance would have been important and it'd need to be fatter if it was longer.

That's comforting :o)

Cheers,

Colin.

[John Rumm]

If you use a 100mA trip (or higher) dedicated circuit for the freezer is it very unlikely to nuisance trip.

Why so? Generally I quite like the idea that you control more of the variables.

With an outside socket on a TN-C-S system then you will be exporting the CPC outside of its equipotential zone. While it is unlikely that you will be using much class I kit plugged into those sockets, you would be quite vulnerable in some fault conditions if you did.

What sort of soil have you got? In these parts (clay soil) its easy to get a sub 10 ohm Ze with a 4' rod.

To drive it, just stick a socket drive adaptor and an old socket on the end of a SDS drill in hammer only mode, and power drive it;-)

[John Rumm]

That's just one brand name... there are other similar products. (and MICC/Pyro of course)

With slabs that makes sense - they will protect the tape nicely, and also would be the first things to be shifted if preparing to dig.

The things that would clearly need bonding would be other services entering via metal pipes etc. They have a clear capacity to bring a potential into your zone. The slab is a less obvious source of a potential, but its also not easy to mitigate it by bonding. Hence if its going to be exposed, its another reason I would prefer my local earth to be at that potential.

(I must do some experiments to see what insulation resistance exists between a slab and an earth probe)

>>> The question I'm left with there, is whether I need to extend the

Indeed - I would say the realistic options are either arrange things so that there is no need to maintain the EQ zone in the first place (but see my comments above about external sockets), or make the shed TT.

>>> I've run the CPC size calculation and the armour seems fine for that,

Or a larger CSA of armour to give the same effective resistance...

Its a good point... I am not sure what the design logic for the spec on main EQ bond CSAs was. It may be that they figured 10mm^2 was adequate for any typical situation within a property. Another reason it makes less sense to extend it too far.

Spoken like a true engineer ;-)

[Andrew Gabriel]

In theory, safer for using tools outdoors.

That was quite hard work. Someone suggested using an SDS drill but my 2.5kg one didn't make much impression on it. I ended up using a 12kg fence post hammer. The rod is in two 4' sections, and hammering it caused the screwed join to come loose no matter how tight you did it up, so I also had to keep a clockwise twisting force on it to keep it done up, and I wasn't very happy that there was no way to keep it screwed up jammed tight. However, the earth resistance came out so much lower than it needed to be, that even if the bottom half of the rod became completelty disconnected in the future, it will still be fine.

I've only used 4' rods before, and actually that would almost certainly have been fine here too.

You don't have to make all circuits TT or all TN. In my own house which is mainly TN-C-S, I have a separate outdoor socket circuit which is a TT circuit. However, you want to avoid mixing them where appliances plugged into different circuits might end up in the same area, with different potentials between exposed earthed parts. Most garden tools are double insulated though, so they won't have any exposed earthed parts.

[David WE Roberts]

Digger is for the extension to the rear of the house. The Mother Of All Sheds [uk.d-i-y passim] is already built, but as we are ripping up the drains at the back of the house and have a digger it seems churlish not to run services to the shed at the same time.

[Colin Stamp]

Yep. I think I've been persuaded.

That's not so bad if it can be done 4' at a time. I had visions of falling off step ladders and ending up in A&E complete with 8' rod embedded somewhere. They'd have awful trouble wheeling me through doorways...

I think I'll try out a 4' rod and see what I get.

Yep. The only thing I can think of that might have earthed metalwork is a lighting stand, currently being used from the TN-C-S downstairs ring. Perhaps I'd better check! At least there's RCDs now I've put in the new CU.

Cheers,

Colin.

[Colin Stamp]

There seems to be a standard method for that in appendix 13 of the regs. Making the electrode looks like a bit of a faff though, so I think I'll just assume the insulation resistance isn't good enough...

>>>>>

The more I think about it, the more I think TT is the way to go.

>>>>>

Yep, though I assume that won't be the cheapest way of getting 10mm of copper equivalent.

Cheers,

Colin.

[Colin Stamp]

Yep. I think having no RCD for the freezer is probably a bit over the top after all.

I think it's just the thought that you can't rely on the MCB to clear an earth fault, so you have to put all your faith in the RCD. Funnily enough, I pressed the test button on the three-month-old RCBO for my smoke alarms last week, and it didn't trip.

Ok, that makes sense now I come to think of it more carefully.

It's quite sandy here (Woburn Sands - I suppose the clue's in the name...)

I might get a 4' rod and see what sort of resistance I can get.

Sounds worth a try. I'm not sure my Chinese SDS drill will shift it past a few inches though. I might have to get out the sledge-hammer...

Cheers,

Colin.

[Grimly Curmudgeon]

Or two in parallel?

[John Rumm]

With TT, no rcd is often not an option since you can't get a low enough loop impedance to be able to clear a fault without one.

Yup, remembering to test TT systems from time to time is the main thing.

If you have not got equipment for testing it, then there is a procedure you can follow here :

> To drive it, just stick a socket drive adaptor and an old socket on the

Last one I did, went in easy enough with a club hammer. However your soil is different to ours.

[John Rumm]

Yup, done that.... ok if you stick them far enough away from each other (few m) to ensure their not in overlapping zones.

[Grimly Curmudgeon]

That ocurred to me just after sending - how close or not they would have to be.