Radials V Ring mains

[Pete999]

I haven't used a RFC for years! Separate 16 or 20 amp radials to each room for me. Several radials in a kitchen. All connected to double pole rcbos.

 

[Peatbog]

Welcome to the forum mate.
So in a typical 3 bedroom house in France, how many circuits are you looking at if it's all radial?

Your location says 'Surrey' but your flag is France. That's why I put France in my post.

I'm a UK electrician. I have perhaps been influence a bit by French electrics! They are generally a few years ahead of us. I used to use french 36 way stacking boxes, i.e 3 rows of 12. Been stymied by metal CU regulations as all french ones I have seen are plastic.

 

[Spoon]

I'm a UK electrician. I have perhaps been influence a bit by French electrics! They are generally a few years ahead of us. I used to use french 36 way stacking boxes, i.e 3 rows of 12. Been stymied by metal CU regulations as all french ones I have seen are plastic.

wow.... 36 ways... Thats one big CU for a house. How much are these things, fully loaded with RCBO's?

 

[Pete999]

I wouldn't do it either, now on a 25A is something I regularly do

Why a 23A when a 32A 4MM2 Or 20A 2.5mm2 are standard circuits for Radials?

 

[Pete999]

Hi all,

This may be a common topic and may already have been spoken about 100s of times so forgive me in advance.

I have always wired in ring mains for my sockets up until recently..

I came across an install where all sockets were wired in 2.5 radials protected by 20A RCBO they had wired 3 bedrooms all on separate 2.5 radials (Loft & 1st floor)
i did question why each room was separate and the response i got was "customers request due to computers"

I am hearing more and more people wiring 4mm radials protected by 32A as its "convenient"

whats peoples general thoughts on this?

A Ring Main is a distribution circuit Mate. Not a final circuit.

 

[Peatbog]

I'm a UK electrician. I have perhaps been influence a bit by French electrics! They are generally a few years ahead of us. I used to use french 36 way stacking boxes, i.e 3 rows of 12. Been stymied by metal CU regulations as all french ones I have seen are plastic.
[/QUOTE

wow.... 36 ways... Thats one big CU for a house. How much are these things, fully loaded with RCBO's?

I think even a small house in France will need 36 ways, and yes it does get expensive.

wow.... 36 ways... Thats one big CU for a house. How much are these things, fully loaded with RCBO's?

It's a fairly small one for a French house.

 

[Peatbog]

Just looked up cost of French CUs. The smallest box at Leroy Merlin (DIY shed) is
Legend 39 way box, 3 rows, 3 RCDs., 11 mcbs Euro 429.00

 

[Pete999]

Hi all,

This may be a common topic and may already have been spoken about 100s of times so forgive me in advance.

I have always wired in ring mains for my sockets up until recently..

I came across an install where all sockets were wired in 2.5 radials protected by 20A RCBO they had wired 3 bedrooms all on separate 2.5 radials (Loft & 1st floor)
i did question why each room was separate and the response i got was "customers request due to computers"

I am hearing more and more people wiring 4mm radials protected by 32A as its "convenient"

whats peoples general thoughts on this?

Nothing really, it's a standard circuit arrangement, best read appendix 15 Mate, to get the gen

 

[EricMark]

I can just about remember my Granddads house, the fuse box was under the stairs, and it has rows of fuses one for each socket, some 5 amp some 15 amp and compared with my dad's house it was huge.

The advantage of the ring final was a single 30 amp fuse supplied all the sockets in the house, so wiring much reduced and also the fuse box was normally just 4 fuses, lights, immersion, ring final and cooker.

I can see the point that with RCD protection we want more final circuits so each RCD has less on it, it does mean using RCBO in most cases as the way a consumer unit is designed we are in the main limited to two stand alone RCD's.

So having the house split side to side reduces the loop impedance and means without running cables up or down stairs you can continue if one circuit fails until the problem is rectified, however having tried using 16 and 20 amp MCB's there is a problem with in rush with type B, plug in a domestic welding set and the B32 MCB holds, but a B20 does not so one is forced to use a C20 instead. This in turn impacts on cable length, both for volt drop and loop impedance.

I have worked with ring finals for years the 1.44 OK now reduced but that figure is etched in my mind, and the one reel only of cable, OK was I think 86 meters now 106 meters but I really don't have to measure, I know by looking at the house if it is likely I am going to exceed the limits, so I test after and it all passes.

However with a 20 amp radial with a C20 I have got to work out the limits, I have just worked it out for this post, one is looking at 30 meters of cable as the limit. So where we could use two ring finals, looking at 6 radials. And easy to exceed 30 meters so one has to carefully work out routes, just too much hassle for my mind, ring finals as so much easier.

 

[Simon47]

As a bit of an off-the-wall idea ...
What if, we run radials, but then pair them up and link the ends of the CPCs to form a ring. It's part way to having high-integrity earthing, and would reduce loop impedance without impacting on the overcurrent protection for the radials.

 

[EricMark]

As a bit of an off-the-wall idea ...
What if, we run radials, but then pair them up and link the ends of the CPCs to form a ring. It's part way to having high-integrity earthing, and would reduce loop impedance without impacting on the overcurrent protection for the radials.

The ELI and Volt drop limits are nearly the same, with a C20 MCB limit around 30 meters for ELI and 32 meters for volt drop.

The supply to many socket outlets traditionally you assume 20 amp at end or mid point, and rest even distributed. So design current for a 32A supply is 26 amp, be it 4 mm radial or 2.5 mm ring final. But the design current for a 20 amp 2.5 mm radial is still considered as 20 amp.

Clearly not written in stone, and as the circuit designed you could declare some other design current, for example a run of sockets provided to supply a cleaner where it was expected only one socket at a time is used could be 13A, however then it would be on a 16A MCB not a 20A one.

What we need to remember other than the British 13A socket the over current device is at the distribution unit, so if we returned to using 5 and 15 amp sockets, we would today have to use 6 and 16 amp MCB's to protect them, and if we have a standard lamp where the flex could be damaged, then that flex must be able to take the rating of the plug, so even a 60W lamp needs a 1.5 mm sq flex, the British system where we fuse in the plug so can use a 0.75 mm sq flex is to my mind a lot better, but it means non reversible plugs, and to get full use of the system the ring final with 2.5 mm sq or 4 mm sq cable with radial and the 32A MCB.

106 meters seems a lot of cable, but with each drop doubled up, and following the 1/3 rules for drilling beams, it is still easy to exceed, splitting the house side to side rather than up/down means in the main less cable is used, so volt drop is better and ELI is better, also means with a circuit failure the occupant does not need to run extension leads up/down stairs to keep things like freezers running, so far safer to split side to side than up/down.

However very few seem to split side to side, most seem to split up/down, never worked out why people spit up/down, maybe some one can give a good reason?

 

[Peatbog]

I can just about remember my Granddads house, the fuse box was under the stairs, and it has rows of fuses one for each socket, some 5 amp some 15 amp and compared with my dad's house it was huge.

The advantage of the ring final was a single 30 amp fuse supplied all the sockets in the house, so wiring much reduced and also the fuse box was normally just 4 fuses, lights, immersion, ring final and cooker.

I can see the point that with RCD protection we want more final circuits so each RCD has less on it, it does mean using RCBO in most cases as the way a consumer unit is designed we are in the main limited to two stand alone RCD's.

So having the house split side to side reduces the loop impedance and means without running cables up or down stairs you can continue if one circuit fails until the problem is rectified, however having tried using 16 and 20 amp MCB's there is a problem with in rush with type B, plug in a domestic welding set and the B32 MCB holds, but a B20 does not so one is forced to use a C20 instead. This in turn impacts on cable length, both for volt drop and loop impedance.

I have worked with ring finals for years the 1.44 OK now reduced but that figure is etched in my mind, and the one reel only of cable, OK was I think 86 meters now 106 meters but I really don't have to measure, I know by looking at the house if it is likely I am going to exceed the limits, so I test after and it all passes.

However with a 20 amp radial with a C20 I have got to work out the limits, I have just worked it out for this post, one is looking at 30 meters of cable as the limit. So where we could use two ring finals, looking at 6 radials. And easy to exceed 30 meters so one has to carefully work out routes, just too much hassle for my mind, ring finals as so much easier.

I must work in smaller houses than other people! I don't find 30m a problem. 5-10m up to a loft extension, then 20-25m around the room. Might put the leg to the room in 4mm2 if I was worried about the length.

My real problem with rings is that I so often find them working but dangerous. Often have breaks in the ring on one or more conductors. Unfused spurs off spurs.
Suspected figure of 8 (which I find really hard to identify and correct). It's then a real hard sell to tell a client that although it has been working for 40 years it is dangerous.

If you have a radial with a core popped out of the accessory then some of the sockets won't work and it's normally obvious which two accessories need looking at.

 

[EricMark]

I must agree that the ring can be faulty without it being apparent, but as long as you test the loop impedance (both to earth and neutral) before you start work, then during the work test to ensure there is a ring, then repeat the loop impedance test after work completed, although slim chance that ring was broken at that socket before you started work and was also rebroken as you put socket back, it is a very slim chance, so even with minimal testing ring finals should not be broken.

In use I suppose you can get rodent damage, but the 10 year test should spot that, 7/0.036 would cure all, it was only due to using metric cables was there ever a problem. Old design was and still is good.

 

[Pete999]

I must agree that the ring can be faulty without it being apparent, but as long as you test the loop impedance (both to earth and neutral) before you start work, then during the work test to ensure there is a ring, then repeat the loop impedance test after work completed, although slim chance that ring was broken at that socket before you started work and was also rebroken as you put socket back, it is a very slim chance, so even with minimal testing ring finals should not be broken.

In use I suppose you can get rodent damage, but the 10 year test should spot that, 7/0.036 would cure all, it was only due to using metric cables was there ever a problem. Old design was and still is good.

The old design used 7/029!!

 

[Simon47]

It wasn't the change to metricc, but the change to solid cores. You can be sure that the penny pinching would have happened even if we hadn't changed to metric.