Continental plugs don't have fuses, so why does UK?

[JohnW2]

The thing is the reason for the introduction of the ring circuit aka the shortage of copper, is as valid today as it was 70 years ago. Put simply long term there is insufficient copper in the world to go around. Consequently any move to radial circuits in a country that already has ring circuits is complete madness from a strategic viewpoint and should be resisted.

Whether or not ring final circuits result in a saving of copper is totally dependent upon the layout of the sockets - and the saving will often be minimal or non-existant. Indeed, in some situations, a ring final circuit can use more copper.

For a start, one can cut any 2.5m² ring final circuit into two 20A/25A radials, without changing the amount of cable/copper being used, and actually with a consdierable increase in amount of power available (40-50A, rather than 32A).

If one really feels the need to maintain a 32A total capacity for each circuit, one has to move to larger (usually 4mm²) cable. However, with some layouts of sockets (the extreme case being if they are all 'in a straight line'), the total cable length for a ring will be roughly double that required for a radial. In those situations, the radial will actually use less copper, since the total copper CSA of 4mm² cable is 9.5mm² (4+4+1.5), whereas the total CSA of 2.5m² cable of 6.5mm² (2.5+2.5+1.5) is considerably more than half that of 4mm² cable.

To take an example, with sockets in a straight line and the furthest 10m from the CU, the total amount of copper for a 2.5mm² ring final would be 130,000mm³ (about 1162g) whereas that for a 4mm² radial would be 95,000mm³ (about 849g).

With careful circuit design, one can often use less copper with radial circuits than with ring ones.

Kind Regards, John

 

[skotl]

Lots of comments on the stupidity of posting on an eight year old thread, and lots of comments pointing out the vast amount of invalid points made.

But nobody has picked up on the fact that this guy (it's always a guy, isn't it?) appears to have joined on 4th June 2013 specifically to flame this thread. And why this thread in particular.

Why would he do that? Bizarre behaviour, to say the least.

 

[jabuzzard]

Thing is sockets are almost never in a straight line and tend to go around rooms/floors. While we may have more sockets today than 70 years ago I doubt the calculation as to whether it saves copper has changed.

However what is certainly utter tosh is the idea that a radial circuit does not need fuses in the plugs to protect flex cords from overheating in fault conditions. Like I said a 3A flex drawing 20A is still a fire hazard.

The best system will be the one with the lowest number of electrocutions per one million of population and the lowest number of electrical fires per one million dwellings. On the electrocution front the UK beats the USA by an order of magnitude. My guess is the use of 110VAC and consequently higher currents along with unfused plugs means that electrical fires will also be more prominent in the USA as well.

 

[jabuzzard]

Neither the bayonet nor the edison screw is very safe in my opinion, yet I still read people on here complaining about how unsafe Wylex standard equipment is.

The bayonet socket is however safer than the Edison screw as there is no potential to wire it incorrectly. Also as I pointed out else where you could use MK Shockguard bulb holders with bayonet bulbs making them inherently safe, which is simply not possible with Edison screw.

I can see a future for ring final circuits but not necessarily how we use it at present. 32A is a lot of current for a modern domestic situation outside of a kitchen, and replacing these with a 20A breaker would lift the restrictions associated with the BS1363 assessory requirement, and ensure that whatever load you put on it, in whatever part of the ring, it would be compliant and safe.

No it won't because a 3A flex could still draw 20A and cause a fire. Any move to unfused plug in the UK would reduce safety. That other countries are less safe than the UK is not a reason to reduce safety in the UK.

Then there is the whole issue of safety of the sockets themselves. For example the standard Schuko socket is unshuttered and unpolarized which makes it less safe than BS1363 even without the fuse issue.

 

[JohnW2]

Thing is sockets are almost never in a straight line and tend to go around rooms/floors. While we may have more sockets today than 70 years ago I doubt the calculation as to whether it saves copper has changed.

Whatever the socket layout, if one uses two or more 2.5mm² (20A/25A) radials, I can't think of any situation in which it would involve more copper than a corresponding ring circuit.

However what is certainly utter tosh is the idea that a radial circuit does not need fuses in the plugs to protect flex cords from overheating in fault conditions. Like I said a 3A flex drawing 20A is still a fire hazard.

I don't disagree with that in concept. If the appliances don't contain fuses adequate to protect the supply flex and/or of you want to cover the (extremely rare) possibility of the flex developing a short along it's length (which would almost certainly result in rapid operation of the CPD), then the flex should be sized according to the protective device of the circuit. In countries that have 16A radials, that only requires 1.5mm² flex.

Having said all that, how often have you heard of an appliance flex catching on fire, let alone doing harm to persons? - a theoretical risk, I agree, but an extremely small one. Fortunately, most faults tend to be very low impedance ones, so result in very rapid operation of a CPD, regardless of cable sizes.

Kind Regards, John

 

[plugwash]

Also in the unlikely event that you do have a cable fault that draws too much current for the cable to safely handle but too little current to blow a fuse then I suspect that the power being dissipated in the fault is going to be a far bigger concern than the overloaded cable supplying the fault.

(the above assumes a low impedance supply, high impedance supplies and/or supplies fitted with current limiting rather than overcurrent trips require more care).

 

[ericmark]

I think it's all down to volt drop and loop impedance where a ring from memory can use 106 meters of cable but a radial even with a 20A MCB feeding is well below the 53 meters from memory near to 25 meters.

So in a house where the ring was 50 meters splitting it into two radials would likely save copper but in a house where the ring is near the 106 meters mark likely one would need 4 radials to replace it.

Appendix 4 Correction for operating temperature gives the formula not for faint hearted needed to work out the maximum length. Table 4D5 gives volt drop but without the correction the results are somewhat out. At 18 mV/A/m without correction 26 meters for 25 amp radial and since half current in each cable and doubled up 104 meters with 25 amp ring. Because at 32 amp we are using less percentage of the cables total capacity than with 20A on a radial and our rules allow certain assumptions for example for calculation the ring is considered at 20A at centre and 12A even spread so we use 26A for calculations not 32A.

With a radial we still assume 20A at centre unless the MCB is less than 20A so with a 25A MCB we would work it out at 22.5A.

So although at first glance it may seem the ring does not save copper in real terms it did. I say did as with the new rules of 1/3 for beam penetration and socket heights being now fixed with rings one can end up with more cable used today. Much is down to design I know of one estate where the two rings were split side to side not up and down as this reduced the amount of cable used and so allowed just two rings rather than three to keep within the loop impedance allowed at the time.

I also know of one estate where half way through the build the beams were changed from the type with pre-drilled holes to solid wood which in turn extended the amount of cable required for the ring (1/3 rule) so three houses were built before it was realised the loop impedance had exceeded the magic 1.44 ohms figure.

Clearly when this post was started the 16th was in force and with 17th the volt drop and RCD requirement have altered the situation.

As to re-starting an old post I did it myself in error once due to an internet search. There is a problem in leaving old posts where the rules have changed and I think to remove all posts started before 2008 would be a good move so people do not read replies to pre-17 Edition questions.

 

[JohnW2]

I think it's all down to volt drop and loop impedance where a ring from memory can use 106 meters of cable but a radial even with a 20A MCB feeding is well below the 53 meters from memory near to 25 meters.

Sure (in principle - I'd have to check to see if your figures were exactly right), but that's just another way of admitting that radials can use less copper - higher VD means less copper!

So in a house where the ring was 50 meters splitting it into two radials would likely save copper but in a house where the ring is near the 106 meters mark likely one would need 4 radials to replace it.

Probably true, but are rings >50m really all that common in domestic premises? I would have thought that they would be pretty rare.

Also, don't forget that we're not comparing like with like. Two 2.5mm² radials can supply at least 40A (50A, if installation method is approipriate and you can find a 25A MCB), as compared with only 32A for a 2.5mm² ring. If you wanted a more valid comparison you would probably consider two 1.5mm² 16A (possibly even 20A) radials, hence the same capacity as one 2.5mm² 32A ring - which would represent a consdierable copper saving.

Kind Regards, John

 

[mfarrow]

Sorry, I didn't mean install a BS 546 on the un-restricted socket circuit, I was thinking more about exporting spurs to outbuildings and being able to provide a proper resettable breaker, or two. You could also have as many spurs as you want.

 

[ban-all-sheds]

But nobody has picked up on the fact that this guy (it's always a guy, isn't it?) appears to have joined on 4th June 2013 specifically to flame this thread. And why this thread in particular.

Why would he do that? Bizarre behaviour, to say the least.

Isn't it obvious why?

He thought he'd feel at home here...

Obviously this forum is for trolls and planks, I though it might be.

 

[ericmark]

Sure (in principle - I'd have to check to see if your figures were exactly right), but that's just another way of admitting that radials can use less copper - higher VD means less copper!

That is why I talked about the correction formula figures are not exact. Although near enough.

So in a house where the ring was 50 meters splitting it into two radials would likely save copper but in a house where the ring is near the 106 meters mark likely one would need 4 radials to replace it.

Probably true, but are rings >50m really all that common in domestic premises? I would have thought that they would be pretty rare.

I was rather surprised. There was an estate being built local that my son was working on and the firm like many others did not provide the electricians with test gear but had a guy who did all the inspection and testing who moved site to site.

The original beams were the type with built in holes so rule of 1/3 did not apply up wall through holes then down wall to next socket. Then the builder changed to solid wood beams so now the rule of 1/3 came in so with a 9 foot wide room an extra 4 foot was used for each drop. With 20 sockets this added around 50 foot (not all used 4 foot extra) this was enough to increase the ELI to 2 ohms so all the rings had to have the MCB's down rated to 20A in the three houses already wired and a new design (side to side rather than up and down) was adopted to reduce the cable being used.

After this I helped my son many times and often was asked to fit extra sockets. I was surprised to find many houses were already over the permitted 1.44 ohms so technically the ring could not be extended. So I would assume the size of the ring was getting near to the maximum.

Today the ELI is no longer the limiting factor it's volt drop. The RCD has changed things. However to measure the volt drop we in real terms have to calculate from the loop impedance for a 20A radial the loop impedance would be less than 0.575 add to that the incoming of course but as you start to measure you will be surprised how many fail.

 

[JohnW2]

Today the ELI is no longer the limiting factor it's volt drop. The RCD has changed things.

Except with TT installations (where you have no choice), you aren't really meant to rely on RCDs to achieve the required disconnection times - so, for TN installations, RCDs haven't really changed anything.

However to measure the volt drop we in real terms have to calculate from the loop impedance...

...yes, provided that the 'calculation' includes subtraction of Re and takes into account the smaller CSA of the CPC. It's obviously R1 (or Rn) that one is actually after.

Kind Regards, John

 

[mfarrow]

Does anyone remember the reg that John's on about, which does not permit RCDs on TN circuits to get around Zs limitations at the design stage?

I tried to find it ages ago but had little joy.

 

[JohnW2]

Does anyone remember the reg that John's on about, which does not permit RCDs on TN circuits to get around Zs limitations at the design stage? I tried to find it ages ago but had little joy.

I used to ask that question and tried quite enthusiastically to argue that it did not exist and, indeed, that a footnote to Table 41.1 implied that it was acceptable to rely upon an RCD to achieve required disconnection times with TN systems.

However, not many people seem to agree! The usual reg quoted is 415.1.2 (aided by 411.1 and 415.1.1 which talk of RCDs only as 'additional protection'). However, I would suggest that one might try to argue that even 415.1.2 allows an RCD and 'basic protection', but nothing more.

Furthermore, 531.3.1 specifically allows an RCD to be used in TN systems to achieve required disconnection times (if they can't be satisfied by an OPD) for "certain equipment in a certain part of teh installation". This tends to imply that they do not intend this to be a blanket approval of the use of an RCD for this purpose.

In any event, don't forget that there is still going to be the requirement to satisfy the disconnection times in the event of (L-N) 'faults', per 434 - since an RCD will not help in that situation (whether TT or TN). Hence RCDs could never be looked at as opening the floodgates to circuits with R1+R2 which were far too high to get required (L-E) disconnection times, since one doesn't have to go much further before R1+Rn would be too high in relation the the L-N faults. Hence, contrary to what some people think, one cannot use the fact that an installation is TT (hence reliant on RCD for protection against L-E faults) to enable one to have as high an R1+Rn as takes one's fancy!

Kind Regards, John

 

[jabuzzard]

[quote="JohnW2";p="2789087"
I don't disagree with that in concept. If the appliances don't contain fuses adequate to protect the supply flex and/or of you want to cover the (extremely rare) possibility of the flex developing a short along it's length (which would almost certainly result in rapid operation of the CPD), then the flex should be sized according to the protective device of the circuit. In countries that have 16A radials, that only requires 1.5mm² flex. [/quote]

Thing is you are roughly twice as likely to die from an electrical fire than from electrocution. Though all the numbers are really small. You are about as likely to die from electrocution as getting dressed in the morning and yes people really do die as a result of getting dressed. Generally they slip bang their head and/or break their neck and last figures I saw where ~20 people a year in the UK.

http://www.esc.org.uk/industry/policies-and-research/statistics/

That's 28 deaths from electrocution and 48 from electrical fires. Also large numbers of appliances don't have a fuse in of themselves. Also the notion that countries that have 16A radials use 1.5mm² flex for everything is also rubbish, they happily use 3A rated flex as well. Though apparently 20% of electrical fires would be prevented by an RCD.

Anyway there is ~ 20,000 electrical fires a year in England which is over half of all accidental domestic fires.

As far as I can see looking on the internet BS1363 plugs and sockets are no more expensive than the continental equivalent.