Electrical Circuits - Rings, Spurs & Radials

Factual Questions
1

A question for any Doper electricians, or those in the know. I’m not sure if this will make sense to the Americans, as it involves UK electrics.

Everything I’ve read says you cannot put more than one electrical socket on a spur coming off a ring mains. Why not? How is putting multiple sockets on a spur any different from a radial circuit?

As you might gather, I want to wire in some new sockets, but the socket that I was hoping in wiring into isn’t on the main ring and is a spur. (Unless it is itself a radial, I haven’t determined yet.)

2

Does this help?

3

Not really. It doesn’t explain why it’s not ok to spur.

I’m a typical Straight Doper. It’s never enough to tell me the rules, I need to know why they are the rules. :slight_smile:

4

I’m an American electrician and this is confusing to me.
Are these series parrallel circuits or am I missing something? What’s the deal with the ring? Looks like a series loop but you are also allowed to tap (spur) a parrallel branch circuit…
The other difference is that these circuits don’t have a nuetral like we’ve got here.
Further explanation please, Nutmagnet.

5

I’m guessing because you would get into some complex series parrallel circuits if you spurred more than once at each tap.
Strange things happen with the resistance and other variables in the circuit that would create an unstable environment for the electrons. :wink:
I’ts been 15+years since I’ve had this stuff in class.

6

Are all these sources assuming your spur cabling is only rated to 13A? I don’t see why you couldn’t run a big chunky 40A spur off your ring main with a few sockets on it, it can’t put any more load on it than several separate 13A spurs.

Maybe the regulation (if it is that) is there to prevent silly people running out a long chain of spurs off sockets and then running electo-plating units off them all?

7

Because, if there are 3 cables, it already has a spur connected and you can only have one spur per junction.

Look at the 3rd graphic on the linked page. See the box labelled “Spur from another socket”. See the cable where it joins the “other socket”? That makes 3 cables. That’s all the code allows. If you want to add a spur, you could use any of the other five sockets.

Uncommon Sense
No it’s not a series loop. Look at these graphics.

It’s just a loop (hence a “ring” main). Yes, there is no neutral. Both wires from the source are “hot” (110v each), providing 220V to the socket.

It’s purpose is to provide a shorter path for the power to travel to a new set of sockets. Look at the last graphic on the link above where they show adding to a radial to make a ring main. At the “consumer unit” the new wires are connected to the same terminals as the original set. This is to provide the same ground potential anywhere in the circuit, while providing a shorter (and more efficient/economical) path for the power.
Think of it as a branch circuit from the breaker - which is not legal in N.A.

8

Sorry, you misunderstand. I’m talking about what would effectively be extending the spur to a second(or third, fourth…) socket. So you’re not adding another spur.

I’m sure there may be good reasons for making it a rule of one and one only, I just wanted to know what they were. As it is, I’ve found another good source of info that says you can put more than one socket on the spur, as long as it doesn’t exceed the number of sockets on the ring. Which makes no sense at all to me.

9

I think you miss-understood that phrase. I read it as, you can install as many spurs as sockets (even more if you use rated junction boxes) and there is no limit to the number of sockets on that spur, only a limit to the number of spurs (one per each or one per j-box). If you could provide a link to the info that states a limit to the number of sockets on a spur, that would be nice.

10

I’ve gone back and re-read some of the sites I originally refered to and you may be right. Can’t find them all, and my main one was my DIY Manual, but I was reading;

Here. Which I interpreted as saying you can have as many on the spur as you like, as long as it’s less than the total on the ring.
And here Which I interpreted as being you’re only allowed one, which may not be correct.

I think maybe they could be phrased better. :slight_smile:

So, if I have it right, I can chain as many sockets as I wish onto a single spur, I just can’t have more than one spur coming off the original socket? I can live with that, but it still doesn’t explain why.

Thanks for your input!

11

I am not familiar with the mechanical end of the connections overseas, but here in the US we have something called “box fill”. Where you may only have X amount of wires of X size in a j-box. It is possible that the reason they limit the number of spurs off of the ring is to minimize the number of wires that will inevitably be jammed into the box that aready contains the socket and the two or three pairs of wires. So, in essence they are controlling “box fill” in this manner.

It is also very possible and likely that there are only provisions for up to three pairs of wires in any j-box and or socket. Adding more wires will only cause mechanical stress to the terminals and or the socket connections.

So, basically it boils down to the amount of real estate that splices (spurs) take up. Me would think.

Where the hell is Casdave when you need him?

12

Uncommon sense
I had said, " Both wires from the source are “hot” (110v each), providing 220V to the socket.", which is wrong. This is the US/Canada method for supplying 220v. The UK supplies a single 230v hot and a neutral, just like our 110v circuits.

Futile
The more I look for info, the more conflicting info I find. It seems to be a point of confusion as to how many spurs per socket and how many sockets per spur.
Some sites say you “shouldn’t” put more than one spur at a socket, others say you “can’t”.

In no case though, could I find an explanation for the “limit” of one spur per socket. I suspect box fill is one consideration. I also suspect that it helps prevent confusion as to which part of the circuit is the ring and which is the spur.

13

It looks like UK wiring boxes use screw terminals rather than wire nuts, so there is indeed a hard limit of three cables per box.

It does all seem very confusing to any US “sparky” who’s accustomed to twisting six wires together under one wire nut. OTOH, for as many wires I’ve seen that were held to the other wires under a nut by some jelly that was on the electrician’s fingers after breakfast, a good one wire - one screw connection does seem more certain.

14

Let me present myself.

I’m a UK electrician, I have all my wiring regulations certificates.

Running spurs -

When you add additional sockets to any system, even if the rest of it is not part of any work you have done, the installer must take responsibility for the circuit right back to, and possibly including the supply box.(the supply box might already be up to its limits and an electrician would be expected to be able to work that out for themselves - thus a higher cpacity supply box might be needed-unlikely but possible)

This is because you are expected to assess the job and survey it properly before you start.

The OP is looking at what is considered good practice according to a Approved Code of Practice

Although this ACOP is known as the 16th Edition Wiring Regulations, the word regulation implies statutory law and is misleading.

The true regulations come under other legal instruments but these are very general things, whereas the ACOP is actually a from of good practice and much more specific.

Now the ACOP here(16th wiring regs) states that you have to run radials in a certain manner, but actually it isn’t true, you can run circuit any way you choose as long as the circuits comply with the law (such as Electricity at Work regulations, or Health and Safety at work act 1974)

Its entirely possible to comply with the law, without actually complying with the ACOP, however, non-compliance with ACOP can used as evidence of a lack of due care if you are prosecuted for failure to comply with the law.(ACOP have a quasi legal status)

Sorry about all that, now I’ll get to the point.

If you comply with the ACOP, you will almost certainly comply with the law, and the ACOP sets out its rules for installing radials and ring main systems.

Contractors prefer to have a set of simple golden practice rules to work to, rather than completely reassess and recalculate cable sizes and earth impedances for every single job, which in the end would only save a few pence in materials but cost a great deal in design time.

Your problem with a radial circuit could be(but not necessarily) this.

The ACOP allows you to calculate wire sizes and in a ring main, its possible to get away with using 1.5mm cable(even 1mm cable is possible sometimes according to the calculations), very few if any contractors would actually do this, they would almost certainly use 2.5mm cable, or even more in some circumstances.

If you ran a radial socket on 1.5mm cable, you would very likely contravene both ACOP and the law, because the earth resistance would be such that you probably could not achieve a low enough resistance to force protective devices like fuses to trip, operate or blow fast enough to protect the system in the event of a worst possible fault.So what I’m saying here is that 2.5mm cable for radials is a minimum in the real world.

You have to choose your protective device according to the time it takes to disconnect a faulty circuit and you must disconnect within certain time limits.
The system might well be designed and safe to run on rewirable fuses, but add a radial and you may end up needing to use sand filled cartridge fuses or circuit breakers(MCBs)

You would probably be ok with 2.5mm cable for a single radial, but add any more radial sockets after that and the earth resistances rise.

There’s also human nature to consider, few folks would actually take the trouble to survey the circuit properly, and work out exactly if there have been any previous alterations, which might mean you were adding a radial socket, onto another radial socket and so on.

The idea that maybe a socket has two sets of wires going to it, therefore it must be part of a ring main is misguided and could be dangerous.

You can run lots of radials if you want, but then you have to arrange protection devices, and you could do this by using a fused spur, as the 40 or 60amp fuse in the supply box may well not be good enough protection.

There is hardly any need to run a radial anyway, all you do is decide where you intend to break into the ring, which will be at some particular socket, you disconnect one set of wires from that socket, and instead you run another set of wires from it up to your new socket, and you then return back down from the new socket and put the disconnected wire and the return into either another socket next to the break in point, or you use a junction box.

15

This part is not correct, or rather, you cannot rely on it to be entirely true in all circumstance,

Given that the ring main could be as low as 1.5mm cable, but there are two sets of cable going to each socket in the ring main, the current carrying capacity might be adequate for that task, but for a radial a single 1.5mm cable would extremely likely not to be enough.

You would use at least 2.5mm cable, and imagine you were putting a single socket some distance away, like in your garage, you would then quite likely need 4mm calbe or even more, again its to achieve a low earth circuit resistance.

You have to assume that any socket may be called upon to deliver 13Amp, even if you don’t intend to use it to that extent, because another person might do so.

How long is the cable run going to be?

16

Thanks for taking the time to reply, Casdave, excellent info. I will take this away and digest.

The cabling I’m thinking of adding won’t be more than another 6m, max. But the problem is the socket I was hoping to connect to appears to either be at the end of a spur, or the end of a radial. It’s not on a ring. My main worry is that if I chain new sockets off this it will overload the existing cabling.

Obviously I need to do a proper survey of the layout this weekend. I suspect it’s a bit of a jumble and the last owner did a bit of wiring that appears on first look to take the long way around everything.

Am I right in thinking that if I choose to connect everything up through a 13amp trip (or fused) switch at the original socket I can pretty much do whatever I like? Obviously that would limit my new sockets to a combined max of 13amps, but the worse that could happen is that the switch trips.

17

Hopefully.

Thanks casdave, well done.
Here in the US our typical feeder (ring) is always larger than the branch circuits (spurs). It would never occur to me to run, say, a 1.5mm feeder and then a larger mm branch circuit. It’s always the reverse here, you run a 4mm feeder and then the branch circuits are smaller. The catch is that you MUST fuse the entire circuit at the main panel for the smallest guage wire in the run.(Unless I install fuses at some point in the circuit to protect the 1.5mm spur.)
So, if I have a 4mm ring and 1.5 mm spurs, I have to fuse the whole shebang at the 1.5 mm ampacity. I’m assuming that’s the case for you as well?

18

OK, I’m an American do-it-yourselfer who does simple electrical projects.

My understanding of the way US 110v circuits work (at least beyond the circuit breaker box, which is the only place I feel comfortable working) is that there is a hot (black) lead and a neutral (white) lead, which is connected to an earth ground somewhere down the line. Current is supplied through the hot leads and returns through the neutral. There is also usually a ground wire, also connected to earth ground, but there should be no current flowing through this ground wire. Most appliances only use a two prong plug (hot and neutral), and those that use a three prong plug (hot, neutral and ground) have the ground only as a safety feature.

Common household circuits are either 15 ampere or 20 ampere. Under the code, a 15A circuit must use at least 14 gauge wire, and a 20 amp circuit must use at least 12 gauge (I don’t know what that translates to in mm). However, it is common for all or parts of 15A circuit to be wired with 12 gague wire, so wire gauge is no indicator. The way to tell is to look at the capacity of the circuit breaker.

After the breaker, the circuits can split and branch any darn way they please. Several outlets or appliances can be linked in series, with the cable connecting one to the other to the other (that is to say, all of the hot leads connected in a series, all of the neutral leads in a series, and the whole darn thing grounded–connecting appliances neutral to hot is a no-no). Or they can be in a star pattern, with all wired up to a central point. Or they can be some combination of these with branches going off in any direction. As long as the wire is the minimum gauge, it is kosher.

First off, from the U.S perspective, is what I have described accurate.

Second, from what the UK explanations and wiring diagrams linked seem to show, it is different there, with the three wires being required to be run in a ring from the “consumer unit” (is that the circuit breaker/fuse box?). How does that work? Are there problems with connecting powered wires around in a ring? Are the wires the same, with hot, neutral and safety ground? Am I missing something?

Thanks.

19

Pretty accurate.

Correct, wire guage is no indicator, but you must size the circuit breaker to the smallest conductor. You CAN put a 15a breaker on a 12 guage circuit if you want to but the BIGGEST breaker you can install would be a 20a.

If you intend to use the same size wire and follow the code for splicing and box fill.
There are rules for tapping larger circtuits if you intend to use smaller wires to do so. Fuses and wire lengths come into play at this point.

Technically, they’re not linked in series, but in parrallel. The wires appear to be in series but since the path travels from the hot to the nuetral you actually have a ladder circuit and not a daisy chain.

I would say you know enough to be dangerous. :wink:

20

The spur cable has to be able to carry full short circuit fault current until the protective device operates, in that sense, the spur cable has to deal with a great deal more than 40 amps for a very short time.

You definately would not want to run a 40amp radial out of a ring main, it would mean you would have to increase you cable sizes in the ring main, to at least 4mm.

If you wanted to do this, you would have a dedicated fuse in the box and run straight out from that.

You are hinting here at what we call diversity.

This is the assumption that no matter how many consumer points there are on a system, and no matter what their potential output, the reality is that this capacity will not be called upon all at the same time.
In theory you could literally have 13amp outlets side to side all the way around a room, you are not going to use more than a few at a time, and then you are most unlikely to be drawing the full 13amps from all those you are using.

This is actually the only reason so many dodgy installations survive, but when something happens like house ownership changes, or maybe the house owner does something not routine, like doing some construction alterations and therefore has a lot more devices plugged in than usual, this is when problems can arise.

I’ve been looking at my copy of ‘The On-site Guide’ and this leads me to the conclusion that you can safely use a fused spur, you must use 2.5mm cable.

Make sure you have some idea of how many sockets are supplied by the circuit fuse, and some idea of what the total demand will be, because if it happens that this circuit also supplies your kitchen, which is a high demand area, you may find that it would be better to split your ring main into two separate circuits so that there are two separate fuses.

Your problem here is not the one socket you want to install, its the alterations that someone else has done, which may already not be adequate.
This means that even if you didn’t install this extra socket, you still may need to do something about it, and if you have a fire and the insurance assessor comes around… you know what I’m saying here.

You can get away with a lot of sins if you replace the fuses with RCD (do not mistake these for MCB which operate on thermal principles rather than the electronic ones of the RCD)

To the US electrical Dopers.

One or two huge differances in our system compared to yours, our plugs have fuses within them, so what we do is have fuses typically 40 but maybe up to 60 amps in the supply box, and the appliance plug fuse will be rated between 3amps and 13amps.
It means that all the wiring in the house must be able to handle a fault current that a 40amp or more fuse will allow before it disconnects.
The supply unit fuse is only there to protect the permanent wiring in the house, and not the appliances that may be plugged in.

We also use a non-current carrying conductor, its the earth wire, and the neutral is actually joined to this right back at the supply box.
The idea is that any leakage to earth will cause a massive current flow and so blow the fuse, the greater this current flow, the faster the fuse blows, so earth impedances are extremely important.
What complicates things a little is that the earth wire in our twin and earth cable(Romex ?) used to be smaller than the main conductors, so that 2.5mm cable would actually have 1.5 mmm earth wire, and this would increase the earht impedance. There is an awful lot of this stuff still around.