Ring or Radial

[dingbat]

great cartoon , although i sure most electricians who have been 'doing it for 40 years' will fail to see the irony contained

Very true. But it usually only takes me about five minutes to convince them that they know nothing. Once 'broken' it is easier to re-educate them in good design practice. (Then they abandon all the good stuff once out of the classroom and go back to what they've always [not] known!) One such sparky took great delight in informing me that on a recent job he had only installed one ring circuit... it turned out to be for a single socket!
(And, as it happened, the principle circuits would have been well-suited to rings, having fixed, known, evenly-distributed loads, connected by socket-outlets! )

Irony? They think that's something to do with ferromagnetic effects.

 

[maguire]

On that last note how many sockets constitute a ring final ? I recently installed a new circuit in a hotel kitchen and it only had two outlets and as it had to be in conduit I did it in 2.5 singles, but the term "ring" gives rise to the idea that it must go around "somewhere" and back to the CU not in this case as it doubled back on itself, still a ring though as I describe a ring final as having a out and return conductor not technically a brilliant description I know. also if I had 4mm2 in the car I would have used that on a 32A, but there you go, so I think the whole debate moot.

 

[dingbat]

On that last note how many sockets constitute a ring final ? I recently installed a new circuit in a hotel kitchen and it only had two outlets and as it had to be in conduit I did it in 2.5 singles, but the term "ring" gives rise to the idea that it must go around "somewhere" and back to the CU not in this case as it doubled back on itself, still a ring though as I describe a ring final as having a out and return conductor not technically a brilliant description I know. also if I had 4mm2 in the car I would have used that on a 32A, but there you go, so I think the whole debate moot.

Watch the video: http://www.xtranormal.com/watch/6907375/

 

[ebee]

Both have their place.
Both have advantages and disadvantages.
It is up to the designer to produce an adequate solution.

One thing with a ring is that loads should be evenly distributed or mainly in the middle portion. I must not be run like a radial then have another cable comming from the far extreme back to the fuseway.
Ring finals have stood the test of time but are open to abuse too.

Ring mains are different to ring finals.

I have never worked on or created a ring main.

 

[bernardgreen]

Hi Bernard, ~~~~~~
FWIW, as discussed above, there may be situations where a ring main circuit is beneficial due to the nature of its use ( and the tolerance of a single conductor failure ), but until someone steps forward to say they have their own runway which they have DIY'd in their own DOMESTIC property, i think its a bit of pointless example.

I used it as an example of a system where tolerance of a single break was necessary. Modern systems of runway lighting don't get damaged very often and if they are the runway is closed.

But I still say that even in the domestic situations the ability to survive a single fault has benefits. Keeping power to appliances is the obvious one, the less obvious one is the reduced risk of a fire when a loose terminal on a radial has to carry the full load on that radial. I fully accept there will be an un-discovered fault but a fault that creates less of a hazard than the same fault on a radial.

With a radial if there is a single fault on the CPC then down stream sockets have no effective CPC. On a ring they will have a CPC

 

[ban-all-sheds]

So that's a failure of 1/3 of the conductors/connections, and with a radial it will become a danger if a 2nd fault occurs, with a ring it will not.

OTOH, the failure of either of the other 2/3 of the conductors will not become a danger with a radial and will immediately, with no need for a 2nd fault, become a danger with a ring.

The odds are not on your side, Bernard.

 

[mikhailfaradayski]

in the scenario of a critical electrical supply, such as runway lights, avionics, rail signalling systems, all manner of military applications, etc, etc (im sure there would be a list as long as your arm) then i would agree the system should have built in redundancy/tolerance to a single fault to prevent a much larger problem.
However, i imagine all of those scenarios are subject to regular maintenance and inspection. Those single faults being recognised and repaired within a reasonable timeframe.

This is not the case in a domestic situation, (yes i know about PIRs, but seriously....) the single fault will go un-noticed and un-repaired for the foreseeable future as the user of the installation will not see any symptoms of failure.

This in turn means that any leg of the RFC still connected could be overloaded, and the mcb will not provide close enough protection to guard against fire.

I would contend that although the single-fault-ring would continue to function, it has also become a fire hazard, how can that be good?

 

[Stoday]

RFCs are not intended to be fault tolerant. Put a nail in one cable and see how tolerant it is. The special case of a loose connection fault that was described would be better termed as "incompetent installation tolerant".

There seems to be some confusion with (real) ring mains.

Ring mains are normally run open circuit; the two "ends" through separate breakers or fuses. Sometimes from different busbars.
They have protective devices and/or links around the ring, often where loads are connected to the ring. These allow a faulty part of, or connection to, the ring to be isolated, usually manually but it can be automatic.

Hence ring mains are considered fault tolerant.

Ring finals (as defined in BS7671) on the other hand are connected to one protective device at source and have no means of switching to isolate a section or connection.

Hence ring finals are not fault tolerant.

 

[bernardgreen]

I would contend that although the single-fault-ring would continue to function, it has also become a fire hazard, how can that be good?

Not if the MCB rating is LESS than the current that can be safely carried by the single route of cable.

Design the ring for the eventuallity of a break into 2 radials and cable is protected.

Which is greatest fire risk, overheating cable or a loose terminal over heating ? Which gets hottest ?

 

[roberta1000]

This has been quite an interesting discussion if you snip the inevitable abuse. My tuppence worth...

Not if the MCB rating is LESS than the current that can be safely carried by the single route of cable.

I suppose so but I've never seen it done.

Design the ring for the eventuallity of a break into 2 radials and cable is protected.

Ditto.
During safety checks I've found countless cases where loose wires in a socket have effectively broken the ring. I'd be very surprised if any practicing spark hasn't found plenty of sockets that just come away in your hand as it were.

Which is greatest fire risk, overheating cable or a loose terminal over heating ? Which gets hottest ?

It depends I suppose...
But I've seen lots of sockets with the plugs fused into them due to bad connections.
In don't think I've seen many (maybe none in fact) 2.5 cables which had dangerously overheated on standard ring fusing. I've seen quite a few where an odd bit of stuff from the shed was used to hook up a washing machine or the like of course with the inevitable funny smell ...

Ahhh!! look at the time - where's the bottle? Happy New Year all!!

 

[ban-all-sheds]

Not if the MCB rating is LESS than the current that can be safely carried by the single route of cable.

But then you negate the whole raison d'etre of ring finals.

They exist solely to allow the use of a cable rated at less than the breaker - take that away and you've made the whole thing pointless.

Plus as the regulations currently stand you'd have to do it by using cable rated at ≥ 32A, because a 20A ring final is non-compliant, no matter what those NICEIC clowns say.

 

[bernardgreen]

Plus as the regulations currently stand you'd have to do it by using cable rated at ≥ 32A, because a 20A ring final is non-compliant, no matter what those NICEIC clowns say.

Can you please show me where it states ( not implies ) that a ring final with a 20 amp MCB is non -compliant. Please quote the words or provide a link to where it can be read.

 

[mikhailfaradayski]

I would contend that although the single-fault-ring would continue to function, it has also become a fire hazard, how can that be good?

Not if the MCB rating is LESS than the current that can be safely carried by the single route of cable.

oh, ok. If that were the case, maybe you have a point. When you have carried out such an install, how do you make note of this on your EIC?
Also, do you then follow the three step/figure of 8 test that you do for the kind of RFC recognised by BS7671?
Or is it more a case of 'plug in, three lights, we're good'?
If you do test as a RFC, do you find that method of testing preferable to the way you would test a radial?
(I thought you were talking about a standard RFC that is generally referred to in this context. i.e. a 30 or 32A fuse or mcb protecting 2.5/1.5mm T+E, as im aware, its the only one mentioned in the BRB )

Design the ring for the eventuallity of a break into 2 radials and cable is protected.

In a domestic, non-critical supply, how does this benefit anyone (lets forget runways and the like as i don't know any DIY'ers who have their own)

Which is greatest fire risk, overheating cable or a loose terminal over heating ? Which gets hottest ?

That would be a different topic to ring/radial. Personally, i would rather not have either in my or my customers houses.
I prefer to ensure that my connections are sound, and i inform my customers (via an EIC) that the installation should be re-inspected at 10 or 5 year interval, as appropriate

 

[ericmark]

Rings are best if correctly installed and maintained.
Question is then will it be correctly installed and maintained?
Unfortunate but the answer is likely no.
It takes a few minutes to test a ring final circuit if records are kept. A selected socket around central is carefully tested once with all records kept and the same socket it then tested every 6 months two mins and if all readings are the same for line / neutral and line / earth then we know nothing has gone open circuit or has a bad connection.
However the problem is records. Hands up when attending a premises to do a periodic inspection how often is one presented with the preceding reports?
Without the preceding report one has to start from scratch. Even with the preceding report how do you know that was done correct anyway? OK in a factory 3 electricians all know each others and have a degree of trust in each others work. But domestic and you haven't a clue who the last guy was and are you going to stake your reputation on his report no way. You start from scratch.
So how long did it take to go though whole test when doing your 2391? About 3/4 hour with all meters to hand and single circuit on a board in front of you? So spread around a house how long to do two ring mains with no labels on any sockets as to where feed from and no plans? To do as laid out in 2391 you would take 8 hours.
How long are you given? 1/2 a day? So what do you miss? And this is the problem. Unlike in house electricians who test when there are no break downs and have all the time in the world to test and record results those in the domestic trade have time constraints which means skipping things.
So in a firm with in house electricians ring final circuits and for that matter ring mains powering the step down transformers are fine.
But in a poorly maintained domestic premises forget it. Use radials. If they go faulty they stop working. They are repaired so no problem. Maybe in rented property where it is all tested on regular basis by in house electricians employed by land lord it may work! Like for example the county council! Oh sorry I've seen them houses, forget it stick with radials?

 

[1john]

Concur with ericmark, all circuits have flaws but ring final circuits have more, also when it comes to testing, there is so much less to do on a radial. I think it is kind of funny that once the ring final circuit has been abused and destroyed it ends up having to become a radial 9 times out of 10 unless serious disruptive repairs are made.