Ring Final Circuit (RFC) versus Radial. Yes, again.

[7029 dave]

Radial.

Break in CPC = socket outlet with no CPC = risk of electric shock = RCD disconnect when imbalance occurs.

Break in ring final = no detection of fault as only one line conductor has slipped out = 2 x 2.5 mm cables with a best case scenario current carrying capacity of 27A (surface clipped) protected by a 32A MCB = Overloaded cable = House Fire = Death

Bit ott mainline, don't recall millions of caused by ring circuit.

 

[mainline]

Bit ott mainline, don't recall millions of caused by ring circuit.

Me neither, it's probably something to do with cables overheating.

 

[Pretty Mouth]

Great thread @pc1966 .

The eagle-eyed reader will note in the OSG that for the RFC the loads are assumed to be distributed, for the radial it is end-point.

It does say in the OSG that the loads are assumed to be distributed for RFCs, however I believe the RFC has actually been designed with the entire load placed at the exact mid-point of the ring:

For the volt drop calculation, I treat the ring as if it were 2 radials, equal length, sharing the load equally. So a 106m ring (wired in 2.5) with a 26A load becomes 2 X 53m radials each with a 13A load.

53 X 13 X 0.018 = 12.4V

Next apply the correction factor for operating temperature from P429 of BS7671, using

tp=70 deg
Ca, Cg, Cs, and Cd all =1
Ib = 13A
It = 20A (from reg 433.1.204)

Gives:

(300-(1-(13²/20²))X(40))/300 = 0.923

12.4 X 0.923 = 11.45V, very close to the 11.5V limit allowed

 

[timhoward]

Radial.

Break in CPC = socket outlet with no CPC = risk of electric shock = RCD disconnect when imbalance occurs.

Break in ring final = no detection of fault as only one line conductor has slipped out = 2 x 2.5 mm cables with a best case scenario current carrying capacity of 27A (surface clipped) protected by a 32A MCB = Overloaded cable = Possibility of a fire.

Agreed. Though it's astonishing what well terminated 2.5 sq mm can actually carry if it has to. e.g. the shower I referred to earlier.

It's difficult to talk objectively about this, which is why I nearly didn't make this point. Every house and loading is different.
And strange anomalies happen even then - e.g. I've known a bride and her 4 bridesmaids all trying to use top-end hair dryers at the same time cause a bit of fun. (It was rather an enjoyable call-out so maybe we should all fit radials and hope for another instance of this, but that's another story)

I've had thought that a mid-point loose N on a radial could be a slightly more dangerous prospect than any loose conductor on a ring, but I won't die on my sword over this discussion!

 

[mainline]

I've had thought that a mid-point loose N on a radial could be a slightly more dangerous prospect than any loose conductor on a ring, but I won't die on my sword over this discussion!

A loose connection on a radial is more than likely going to cause problems with appliances etc downstream.

 

[123]

Once a Ring is Infected with numerous spurs on spurs on spurs , it can no longer be called a Ring and must be called a Spider circuit which needs to added to the 19th Edition assuming Rings aren't outlawed before that

Poor installation and botched alterations ruin anything and everything though.

A house I tested last year had bathroom shower and cooker joined together in a JB somewhere. Upstairs sockets off the lighting circuit. Utility room off ensuite shower in a JB somewhere. Immersion heater and garage now a bar with electric heating tied in together. You can't fix stupid.

 

[nicebutdim]

Poor installation and botched alterations ruin anything and everything though.

That about sums up the ring/radial debate. Every argument involves examples of bodged installation and none are based on objection to well designed and installed circuits.

No one would argue that radial circuits are inherently unsafe when sockets are found to be spurred of a shower supply in 1.5 T&E, so I don't get why a circuit design should be condemned on the basis of questionable alterations, made long after initial installation.

 

[pc1966]

Break in ring final = no detection of fault as only one line conductor has slipped out = 2 x 2.5 mm cables with a best case scenario current carrying capacity of 27A (surface clipped) protected by a 32A MCB = Overloaded cable = Possibility of a fire.

That is a bit dramatic. If you look at the worst case, say 50% above CCC of one leg then you are looking at 50% extra (x 1.5) on just one conductor of that leg, so instead of I2R heating of 125% more (1.5 * 1.5 = 2.25) you are looking at more like 1.625 more heating, and from the usual 30C to 70C cable design that puts your conductor probably just above 95C.

Is this good for the cable? Clearly not, but it is far far away from starting a fire. In fact it is well below the 160C upper limit taken for adiabatic heating without cable damage (e.g. Table 54.3) so what you are looking at really is a loss of useful cable life.

I have no idea what the factor needed for the Arrhenius equation is for PVC cable but usually it is a halving of component life for every 10-20C increase in temperature (assuming nothing dramatic happens like a fire!) so taking the cable on the faulty RFC as running 25C above specification you are looking at probably a quarter of its nominal 25 year life.

But who can afford to run their RFC at 32A or so for a long time? At today's cost that is something like 2 grand a month electric bill!

So in reality that ~6 years cable life is going to stretch out over a couple of decades as probably the high temperature is only for tens of minutes per day, not 24/7

 

[pc1966]

It does say in the OSG that the loads are assumed to be distributed for RFCs, however I believe the RFC has actually been designed with the entire load placed at the exact mid-point of the ring:

Thanks, that is good to see the workings.

It is actually not a bad design assumption as you could reach 32A limit on just two double sockets, so they might be close to the mid-point in the worst case.

 

[mainline]

That is a bit dramatic. If you look at the worst case, say 50% above CCC of one leg then you are looking at 50% extra (x 1.5) on just one conductor of that leg, so instead of I2R heating of 125% more (1.5 * 1.5 = 2.25) you are looking at more like 1.625 more heating, and from the usual 30C to 70C cable design that puts your conductor probably just above 95C.

Is this good for the cable? Clearly not, but it is far far away from starting a fire. In fact it is well below the 160C upper limit taken for adiabatic heating without cable damage (e.g. Table 54.3) so what you are looking at really is a loss of useful cable life.

I have no idea what the factor needed for the Arrhenius equation is for PVC cable but usually it is a halving of component life for every 10-20C increase in temperature (assuming nothing dramatic happens like a fire!) so taking the cable on the faulty RFC as running 25C above specification you are looking at probably a quarter of its nominal 25 year life.

But who can afford to run their RFC at 32A or so for a long time? At today's cost that is something like 2 grand a month electric bill!

So in reality that ~6 years cable life is going to stretch out over a couple of decades as probably the high temperature is only for tens of minutes per day, not 24/7

That's a relief, no need to worry about overloaded broken ring finals causing a fire.

I would now assume that counts for 4 mm radials as well.

 

[Lucien Nunes]

Post No.28 above is important, because it debunks one of the most common incorrect arguments used against the 2.5sq.mm. 32A ring. It is often said that if one leg becomes disconnected or the circuit is badly arranged leaving all or most of the 32A load on one leg of 2.5, that cable presents a fire risk through overload. It does not.

As stated in the above post, the most likely result is a minor reduction in cable service life, and the worst reasonably likely result is a more significant reduction in cable service life. A length of 2.5 carrying any level of current that a 32A MCB will give it, will not burst into flames nor set anything else on fire. It will be hot to the touch and a bit floppy and that is all. It will withstand perhaps 80A before anything dramatic happens, and even then that is not generally fire.

To re-state that in practical terms, a defective ring of 2.5T+E with spurs off spurs, inappropriate load distribution and one leg open circuit, will support four 12A heaters plugged into a row of adjacent sockets anywhere until the 32A OCPD trips, without anything catching fire and probably without any further damage of any sort.

Fire does not start with marginally-overloaded conductors, which rise to an elevated but not extreme temperature and then reach equilibrium, never to get hotter. Much more likely to start a fire is the uncontrolled temperature that can be reached by arcing connections, since these can just go on getting hotter and hotter as their resistance rises. Where a connection fails high-resistance, the parallel paths available in the RFC can replace a potential arcing scenario with a marginal conductor overload one, in which case it probably reduces the risk of fire. It does then reduce the likelihood of fault detection, but the very fact that we do find burnt terminals still in service shows that detection is not by any means guaranteed in use.

 

[mainline]

Post No.28 above is important, because it debunks one of the most common incorrect arguments used against the 2.5sq.mm. 32A ring. It is often said that if one leg becomes disconnected or the circuit is badly arranged leaving all or most of the 32A load on one leg of 2.5, that cable presents a fire risk through overload. It does not.

Even in a properly connected joint, when heated could cause a fire risk.

Much more likely to start a fire is the uncontrolled temperature that can be reached by arcing connections, since these can just go on getting hotter and hotter as their resistance rises.

Where a connection fails high-resistance, the parallel paths available in the RFC can replace a potential arcing scenario with a marginal conductor overload one, in which case it probably reduces the risk of fire. It does then reduce the likelihood of fault detection, but the very fact that we do find burnt terminals still in service shows that detection is not by any means guaranteed in use.

Is this where the debate of AFDDs when used in an RFC may not detect the series arc fault.

 

[pc1966]

That's a relief, no need to worry about overloaded broken ring finals causing a fire.

Again, over the last 6 decades, how often do you hear a spark saying "oh, found another burnt out RFC today"?

Burnt out plugs/sockets, yes, but that is independent of the RFC/radial choice.

I would now assume that counts for 4 mm radials as well.

It would.

But there is a slight difference in designing to meet the regs (choice of cable, install method, and OCPD, etc) and worrying about a fault developing on a design that meets the regs.

 

[pc1966]

Incidentally, did John Ward not do a video once that tried to burn out 2.5mm cable? Maybe testing the quality of Wagos, etc?

I vaguely remember it taking something quite large like 60 odd amps, etc.

 

[pc1966]

Is this where the debate of AFDDs when used in an RFC may not detect the series arc fault.

As others have pointed out, the AFDD not detecting a series arc in the RFC is down to such an arc being practically impossible to start (as only ten or so volts difference if the ring breaks) and if on microscopic gap it does, it is not big enough to be a serious risk.

From your own experience, how often do you find burnt out joints, and of those you have, how many were on sockets of either system versus showers?

 

[mainline]

Again, over the last 6 decades, how often do you hear a spark saying "oh, found another burnt out RFC today"?

Burnt out plugs/sockets, yes, but that is independent of the RFC/radial choice.

It would.

But there is a slight difference in designing to meet the regs (choice of cable, install method, and OCPD, etc) and worrying about a fault developing on a design that meets the regs.

But surly in the design process the risk of a particular fault developing in the circuit choice has to be taken into account.

 

[Dustydazzler]

Incidentally, did John Ward not do a video once that tried to burn out 2.5mm cable? Maybe testing the quality of Wagos, etc?

I vaguely remember it taking something quite large like 60 odd amps, etc.

He did , I think he got around 60-70amps up a 2.5 before it started to get warm and around 80-90amps before it started smoking / melting. Yes it was a short length clipped direct but it shows how much overloading a cable clipped direct can actually take

 

[Dustydazzler]

I think Bic Nundy did a similar test again with a bit of old 2.5, iirc he did it with different types of cable pvc / lsf etc
Again kept pumping more amps up it until it started smoking. again it took a serious overload to get it to smoke

 

[mainline]

As others have pointed out, the AFDD not detecting a series arc in the RFC is down to such an arc being practically impossible to start (as only ten or so volts difference if the ring breaks) and if on microscopic gap it does, it is not big enough to be a serious risk.

From your own experience, how often do you find burnt out joints, and of those you have, how many were on sockets of either system versus showers?

Rare at sockets to be honest, but many JBS and shower pull cord types, mostly due to not using ones that have decent terminations.

 

[Lucien Nunes]

As mentioned above, a 'series arc' in one leg of a ring won't really arc because there's hardly any voltage across it. If it doesn't arc, the AFDD won't find it. If it does arc, the AFDD will react and disconnect it. A loose terminal in a socket will arc on a ring just as as on a radial and the AFDD will find that too.