Electrifying the summer house... gosh!

[Woland]

Dear all,

I have veen following anxiously the issue of electrical regulations, XVII ed. (part P?) or similar (seems like citing some obscure Criminal Law Act though...).

Now the problem: in 2000 I had my electrical system partially rewired and 'updated' to the current regulations of the times (Scotland). A CU with no RCD for light and power in the house, plus another CU with RCD (there were no split units at the time).

A 49Amp MAX armoured cable (should be 2.5mm^2, not certain but I'm certain it's 49 Amp MAX) which is some 40 metres long goes from the RCD-protected CU straight into the summer house to provide power.

I originally asked some supposed sparkies and the majority seemed to conclude that I could manage the work myself as it were a 'shed-like outhouse' but I', mot this very much convinced of that. If that's not the case I would still do the job myself according to current regulations and than waste a bit of money on the usual (un)qualified guy to let him fill a certificate of compliance/whatever.

Now...

1) I read somewhere (now well buried in one of the posts) that I'd better link the armoured cable, at the house to the non-RCD CU and then provide a split unit at the summer house end rather than keeping the current setup.

2) Anything else to comply with these apparently frightening regulations? (I would create ring circuits inside the summer house, have a RCD-protected and a non-RCD protected split unit if possible, etc). Note that I would use the summer house for woodworking involving a range of professional power tools

Thanks

E.

[Micky Savage]

Hi I would put a radial circuit in. That way you only have to isolate that circuit, if you have to work on it. Also if it does go down your whole house does not.

Ring rinals are on their way out. I know this because I am at Leeds College of Technology, I am studying for Level 3. We covered this very subject today. I am about to rewire my home and am putting in radial circuits in every room.

Kind Regards.

Micky Savage.

P.S. Testing is also a lot quicker on Radial circuits, so you won't have to pay the sparky as much for your testing.

[Woland]

It is already radial in the sense that there is one armoured cable from the 40A electronic fuse in the home CU to the summer house, so the total load should not exceed 40A x 240V.

It's just inside the summer house that I would place another CU on the receiving end of the armoured cable and create a local ring, the idea being that either the summer house CU electronic fuses deal with overloads/spikes/ whatever or in the worst case the 40A fuse in the house Cu deals with it.

W.

[meow2222]

Funny, considering the ring circuit has such safety advantages

[Micky Savage]

Funny, considering the ring circuit has such safety advantages

No not funny it's going to be fact.

Micky.

[Owain]

Part Pee does not apply in Scotland. Scottish Building Standards are different, but increasingly going down the line of being divorced from governmental responsibility and influenced by vested interests

Owain

[Tim S]

Micky Savage coughed up some electrons that declared:

Lecturers sometimes have their own slant on things and it may or may not align with reality :->

OK - rather than just state what I think, lets have some fun instead...

I would like to postulate that 32A final socket circuits are good for many applications because they doesn't keel over when you plug two 3kW appliances in. If you disagree, the debate is dead, if agree, please go to next step:

Next step: Can you construct a 32A radial circuit using 13A sockets?

Cheers

Tim

[Andy Wade]

Quite. AFAIK there's no intention whatever to abandon the faithful ring circuit, but there is something of a campaign going on to educate electricians about the radial alternatives. Hence Appendix 15 in the New Red Book.

Yes, of course - it's one of the standard circuits: 4 mm^2 cable for the main circuit, spurs in 2.5 subject to the same rules as for rings, max. floor area served 75 m^2.

[Andy Wade]

The 17th edition and Part P are quire separate things. The latter doesn't apply in Scotland.

Oh yes there were.

With that length of run the effective current rating of your submain will be limited by voltage drop and not by the current rating of the cable.

You might find

[Woland]

Was just a hypothesis

I had read that and I was expecting a serious voltage drop (no load reading doesn't count, obviously). So I set up the usual easy measurement with a 2 KW load (circular saw

• lights + other resistive loads) and... strange but true no drop!

Experiment repeated several time... same result. I am actually curious to find out why I'm not withnessing any voltage drop... maybe the load is too small?

W.

[fred]

Was just a hypothesis

I had read that and I was expecting a serious voltage drop (no load reading doesn't count, obviously). So I set up the usual easy measurement with a 2 KW load (circular saw

• lights + other resistive loads) and... strange but true no drop!

Experiment repeated several time... same result. I am actually curious to find out why I'm not withnessing any voltage drop... maybe the load is too small?

You're obviously using special zero-ohm oxygen-free speaker cable. Supplier/price?

;)

[Andy Wade]

A saw's not much good as a test load because the current drawn will vary wildly with the mechanical load. Use a 3 kW fan heater or similar.

How are you measuring? Is the cable perhaps bigger than you think?

The resistance of a 40 m run of 2.5 2-core will be about 0.6 ohm. If you were to allocate the whole 3% permitted drop (for lighting) to the submain on the assumption that there'll be little more in the short wiring in the summerhouse, the max load will be 6.9 V / 0.6 ohm which is only 11.5 amps.

What equipment do you propose to use in the summerhouse? For a 32 A supply you really need a 10 mm^2 cable, or 6 mm^2 (4 at a pinch) for a

16 A feed.

[Tim S]

Andy Wade coughed up some electrons that declared:

Thanks Andy - it was really for Mickey. Anyway, I'll follow through on the thought I was hoping to provoke:

Assuming I'm reading Table 4D2A (IEE regs, 17th) correctly, 4mm2 cable derates to 25A capacity in thermal insulation and 30A in conduit in wall, so quickly becomes unsuitable for a 32A radial circuit.

A BS1363 socket has a stated terminal capacity of 3 x 2.5mm2, or 2 x 4mm2 conductors per terminal, with occasionally a manufacturer allowing 1 x

6mm2.

So, my argument to Mickey and his lecturer:

Suppporting a 32A radial circuit in practise is difficult. A 20A radial is easy enough, but personally, I find them conceptually less useful for general purpose use, where loading is unpredictable and the possibility of wanting a load > 20A on a particular circuit is quite likely.

Ergo, I don't think radials are the way to go, except in limited cases where

20A is known to be adequate.

Cheers

Tim

[John Rumm]

Well leaving aside any issues of why you would be paying anyone else to test your circuits for you, one needs to consider the circumstance.

If you have just installed a new ring final circuit, then it is no more difficult or time consuming to test than a radial. The time that a ring circuit becomes more difficult to test is when you are attempting to discover and reverse engineer its topology after a period alterations etc.

IME the general fault tolerance in real world situations is better with rings than radials since the most typical faults have a lesser impact:

[meow2222]

Begs the question of how you're measuring it. Basic physics tells us there will be a v drop.

NT

[unopened]

No it doesn't. The OP is not indulging in petitio principii. It might raise the question, but it doesn't beg it.

And to sound off on ring circuits vis-a-vis radials: weren't rings in part devised to cope with post-WW2 copper shortages - wouldn't it be 'greener' to continue using rings as radial usage would increase copper usage.

And another thing...with all the campaigning to minimise the use of 'standby' rather than physically 'off' switches, and encourage people to unplug mobile phone chargers and the suchlike when not in use, won't the requirements for RCDs significantly increase base-load power consumption? All the RCDs I've come across dissipate a few watts. Replacing all the rewireable fuses in my consumer unit with RCBOs; or indeed replacing sockets with ones that have RCD protection build in will increase consumption.

Regards,

Sid

[John Rumm]

They were... although

Prolly better to use whichever is most appropriate for the particular circumstances. Depending on layout a ring might use more copper than a

4mm^2 radial anyway.

Can't say I have ever noticed a RCD getting warm, but then again I have never measured their power consumption. Either way, you can't expect joined up thinking from the government can you?

[Dave Liquorice]

There is a difference between "dissipate" and "consume". The key is does and RCD get warm when it is not passing any load? I don't think they do so they don't "consume" any power in their own right. They simply "dissipate" a bit when carrying a load, just like the cables do as well. No increase in base load.

[unopened]

..."does and [sic] RCD get warm when it is not passing any load?" - Yes! - At least every single one I have used has done so. They get even warmer when passing load. My _guess_ is that all the ones I have come across use a solenoid to keep the contact closed, and some power is used in the solenoid. It isn't much - somewhere between 0.5 and 5 watts is my estimate when not under load, but multiply that up by the number of RCDs across the country and it is a lot of power being used. I have no idea why they warm up under load. I hope its not ohmic heating across the contacts!

Cheers,

Sid

[The Natural Philosopher]

NO, they use the solenoid to break the trip.

IIRC they wind live and neutral together round a solenoid, the idea being the currents cancel..any imbalance causes a net magnetic field and trips the switch. You need a fair number of turns to get the sensitivity, so there is an inevitable resistance, especially at high load currents.