Why does the UK use rings for sockets?

[bjouis]

I take it you're allowed to use radials for sockets?

So why does anyone bother with rings, seems like extra hassle to me.

 

[pc1966]

Oh this will open a can of worms when it comes to the arguments for and against ring final circuits!

Yes, you can use radial circuits and in the UK case with any number of sockets, subject the maximum typically of a 32A MCB feeding cable of adequate size.

Similar for a ring, but there you can use less cable and/or longer route, again no limit to sockets as such, but total is limited to typically 32A (some older systems would have a 20A fuse or MCB).

The key difference with the UK are really the fused plugs. That came from the idea of having a standard plug/socket for all normal loads but a choice of fuse, instead of the old arrangement of radials with different feed fuses each for a single socket, and different plug sizes so you could not plug a 2A appliance in to a 15A feed.
[automerge]1587590880[/automerge]
Advantages of the ring:

• Less copper for a given load / area
• No single point of failure for protective earth

Advantages of radial:

• Simpler idea
• Faults show up immediately

While for some it almost gets to a religious argument, in general if you only need a few sockets then a radial is best choice, if you are covering all rooms in one floor then a ring is probably better able to do it.
[automerge]1587591564[/automerge]
For the history try a search for "THE ORIGIN OF THE BS 1363 PLUG AND SOCKET OUTLET SYSTEM" as there is an IET publication from 2006 with details.

For a quick idea of what you can do with ring or radial in practice then the IET's On-Site Guide book has "Table 7.1(i)" giving the typical cable length limits for ring final circuits on page 65, and for radial final circuits on pages 68-73
[automerge]1587592017[/automerge]
In case you don't have that book, some numbers are:

• Ring final 2.5mm cable, 32A B-curve MCB, length 106m on TN-C-S
• Ring final 4mm cable, 32A B-curve MCB, length 171m
• Radial final 2.5mm cable, 25A B-curve MCB, length 33m
• Radial final 4mm cable, 32A B-curve MCB, length 43m

 

[static zap]

Ring lower impedance , less volt drop = less waste !
( Which nation likes efficiency - !)
Ring = RFC.

 

[Wilko]

Hi - on a lighter note - are you familiar with Lord of the Rings? One ring to rule them all .

 

[bjouis]

How does a ring use less copper?

A radial is basically a ring without the return cables. Confused!

 

[Ian1981]

Because we are British and don’t like change and have to do everything differently to the rest of the world

 

[telectrix]

How does a ring use less copper?

A radial is basically a ring without the return cables. Confused!

because to crate a 32A radial, you need to use 4mm or 6mm cable. and a radialcould well end up with the last socket alomost back to the board, so the extra leg of a ring might be only a few yards.

 

[pc1966]

How does a ring use less copper?

A radial is basically a ring without the return cables. Confused!

As already mentioned by @telectrix because the ring can use smaller cable for the same overall rating. Almost all final ring circuits in the UK use 2.5mm cross sectional area cables for a 32A rating, to do the same total current as a radial you need to use 4mm (saving in copper of typically over 30%) or even 6mm if the length is great.

The other thing you might not have realised is the UK difference due to fused plugs, here we can put on as many sockets as we want to either type of final circuit. Where as in most of the rest of the world the plugs are unfused so the supplying breaker has to be small enough to protect the final cable against fire, so often that makes for more expensive final cables as well as more radials per house as there is a limit on the number of sockets per radial.

Now the fact we can put on as many as we want per circuit does not mean we should! The higher total capacity (typically 32A) means better load diversity is possible, but even then in the UK you would often plan for more than one final circuit for several reasons:

• Fault resilience
• Expected load distribution
• Expected total load

As a "rule of thumb" you would be using in the UK one ring final per floor in a multi-story property, and even with a single floor flat you might plan for a separate ring for the kitchen where the majority of power-hungry appliances tend to live (washing machine, tumble dryer, dishwasher, etc).

Finally in the UK radial circuits are still very much alive and well for large fixed loads (such as cookers or larger air conditioning units), and for cases when you really don't want an unrelated fault to deny power to somthing critical (e.g. fridge-freezer might have its own breaker, or computer UPS supply in a home/office arrangement, etc).

 

[Vortigern]

I suppose we could go with the French method of radials. But hang on, it's a nightmare testing such circuits where there are branches and additions of various kinds and you end up with a radial circuit that is a tangle and very hard to determine what is connected to what. Hence the British like ring circuits because even where mistakes are made you have a beautifully mathematical way of deducing what has happened. The logic intrinsic in the ring circuit and testing methods are second to none throughout the world.

 

[telectrix]

to misquote a famous line... " the ring is dead: long live the ring"

 

[pc1966]

@Vortigern also raises the issue of circuit testing. A common complaint by folk looking at testing for the first time is the test procedure for ring final circuits looks unusual, and they think it is more complicated and time-consuming.

It is, but that misses the point! The ring test procedure also provides much more comprehensive fault coverage than it is easy to do with a radial circuit.

The loop impedances (end to end for L-L, N-N, and E-E) give you a good indication of bad joints if you see they are not matching (L & N should be around 0.05 ohm or less difference) or in ratio (typically E/L is 1.67 due to the 1.5mm earth versus 2.5mm phase size used with our T&E cable). This is also easy to do for an inspection as you can do it all at the consumer unit (fusebox).

The "figure of eight" test (L1 joined to E2 and L2 to E1) not only allows socket polarity to be verified (i.e. L and N are correct way round) but is also quite a good test of socket quality as each socket should have practically identical resistance (L to E in this case), and any showing an increase of 0.05 ohm or so might be poorly connected or tarnished contacts, etc. Now 0.05 ohm may not sound like much, but at 10A that is 5W dissipated!

 

[Mike Johnson]

No we are populating DB's with RCBO's do you think that radials will become the norm, what do the team think? can't find a smily that sit's on the fence.

 

[pc1966]

I don't really see it making a difference to the ring/radial debate.

But RCBO are a much better idea than several MCB + common RCD no mater what sort of final circuits you have!

 

[Mike Johnson]

Do you think that manufactures will standardise with the rest of the world and produce dual module RCBO's and get rid of the flying lead?

 

[Ian1981]

Do you think that manufactures will standardise with the rest of the world and produce dual module RCBO's and get rid of the flying lead?

Schneider produce rcbos with no neutral fly lead on their acti 9 isobar P DB’s, think the rcbo’s are only single pole tho.
DP rcbo’s are not necessarily required tho, be it a TN or TT arrangement, there’s no requirement to disconnect or switch the neutral during overcurrent conditions apart from those requirements detailed in part 7 of bs7671.