In a bout of man flu, I lay in bed wondering: If the MCB is supposed to protect the cable to the fitting, and the fuse in the plug is supposed to protect the cable to the appliance, how is the single bit of 2.5mm T+E run to a single socket spur adequately protected by anything?
You've basically got 2 bits of 2.5mm in parallel (usually unequal lengths), feeding a single bit of 2.5mm. 32A MCB.
Whereas if it was a standard radial circuit, you'd have just the single bit of 2.5mm, and you'd only be allowed a 20A MCB on it. (IIRC they even dropped the 2.5mm rating down to 18A, which is "correct" in accordance with the calculations, but then increased it back to 20A because that's what everyone does anyway).
I'm not worried. I just don't see the logic.
I have 10cm of 2.5mm T+E protected by a 40A MCB on the cooker circuit (feeding a socket that can only supply the gas hob ignition) and that seems very wrong too - but wiring a socket in 6mm is just too painful. However, the cooker circuit may yet drop back to 32A (because I suspect that's all it needs) which makes it no worse than the socket spur discussed above.
Odd as it may seem, overload protection *can* exist downstream as well as upstream.
In the case or a short circuit in the back of the socket, the 32A breaker will still protect the cable (ie interrupt the current before the cable overheats).
You have to consider the difference between overcurrent protection and fault protection.
If an MCB is used for overcurrent protection then the circuit has to be capable of running at the full current allowed by the MCB. For a B type MCB this is 1.45 times the rating of the MCB.
If the MCB is to be used only for fault current protection (ie a short circuit) then the cable only needs to be able to pass the short circuit current long enough to allow the MCB to trip within the required times when there is a fault.
A ring from 2.5 T&E is capable of meeting the 32 x 1.45 current overload protection that a B type MCB provides.
Now a spur from a ring is "self" overcurrent protecting as is is only allowed to feed 1 single socket or 1 double socket. The maximum normal current that you get down this spur is 26A (assuming a double socket loaded to the maximum) which is less than the 27A capacity of 2.5 T&E when clipped direct (or buried in plaster) so there is no overcurrent problem and the MCB is now only needed for short circuit protection to protect the spur. However if there is a short on the spur (nail, angle grinder etc) then the resistance reading at the end of the spur must be low enough to trip the MCB in the times required for a short circuit. The resistance reading is made up from the supply impedance plus the cable impedance, so the longer the length of the spur the greater the chance that the spur is not compliant and may not clear a short in the required time.
So on a 20A radial, the 20A is to protect against overloads (lots of "up to 13A" loads, potentially causing damage over time without ever tripping a 32A) whereas a short circuit would generate enough current to trip a 32A before the cable suffered harm?
Where can I find that calculation? I saw those MCB and fuse trip-time vs current graphs the posted the other week, but can't remember reading a shot circuit disconnect time requirement.
Do I need to change my 40A to a 32A, or leave it as it is?
NO. A 20 radial is a circuit with a 20A MCB. A spur from a 32A ring is a different thing. ( my other post did not make that one clear)
formatting link
the graphs.
I would change it. Far easier than doing calculations. However I would not use the 2.5 T&E for use with an electric cooker without making some very accurate calculations.
Its self protecting exactly as a spur on a 32A ring is (it only has a single
13 socket). I would change it too, most electricians would reject it as they don't know how to do the calculations and its just easier to use the regs than to explain why its OK.
Yes it is at the moment, but it will not be if the OP installs an electric cooker, and he has hinted that he may do so.
(the regs tell you what is OK and they give you the calculations, the regs and what is OK are inclusive not exclusive)
I have to take the bigger view. I might be able to show that a certain cooker is safe when installed with the 2.5 T&E, however the customer may change the cooker for a higher powered one or sell the house and a new owner may fit a higher powered cooker. I would rather see a cooker circuit correctly fused to the actual cable rating and not to some diversity calculations that mean nothing when a cooker is swapped.
You can do anything you like if you are competent to do the calcs, I believe that's in the regs somewhere. The majority of electricians mean the onsite guide when they refer to the regs.
That I agree with.. diversity is a PITA. It doesn't even work when someone swaps a few single plugs for doubles in a ring and then plugs in a few fan heaters at one end. That sort of things makes rings run out of spec. I suppose its regarded as unlikely that some will use 9kW of heating in a room, even one they are trying to dry out after the recent floods. ;-)
The problem being that there is no way to know, it might be a big room with the doors and windows open with a gale blowing through, they would stay on then. The reason why I brought it up is because I have seen it happen BTW. Not everyone would think about running an extension lead from somewhere else to make sure there wasn't a problem. They would leave it running until the room was dry or it all failed.
This is one of those occasions where the responsibility for overload protection and fault protection can be split. The fault protection (i.e. very high short circuit currents) must always be at the origin of the circuit, however the overload protection can be at the load end in some cases. (others would include a 3A drop wire to a pendent fixing on a 6A protected lighting circuit for example - the largest bulb you can get won't come close to 3A so no chance of overload))
An unfused spur can power a total of one double or single socket. The diverse load of a double socket is taken as being 20A. (That is lower than the current carrying capacity of 2.5mm^2 T&E in all but the most adverse installation methods). Even the theoretically possible 26A is just under the maximum rating when clipped direct or buried in masonry. So the conditions at the load end (i.e. only one double socket) enforce the overload protection.
The next question concerns the fault protection. i.e. what happens when you nail through the cable or some other drastic fault occurs. Here you will get a fault current that limited only by the the resistance of the wires themselves and that of the supply and earth connections. This is one of the reasons for their being a maximum cable length specified for most circuits - to ensure the so called "Earth Loop Impedance" (i.e. round trip resistance from supply through circuit wires, and to earth) can't get too high.
A 32A breaker will typically need as much as 160A to open "instantly" i.e. on the magnetic part of its trip response. Instant in this case means 0.1 secs or less.
See charts here:
formatting link
say you work out that your prospective fault current is going to be
200A, you now need to assess[1] what happens when you try and stick that down the skimpy 1.5mm^2 earth wire. Needless to say that will get hot, and quickly. Its also going to be heating so fast that the natural heat losses to its surroundings are going to be negligible during the time scale we are interested in. So we have what can be though of as adiabatic heating.
What you need to check is that it won't melt before its done its job and tripped the MCB. For this you use what is called the adiabatic equation: s = sqrt( I^2 x t ) / k, where s = the minimum cross sectional area of copper required in the conductor, and k is a factor specific to the the type of cable (115 in the case of PVC T&E).
So 200A, a check of the MCB charts says this is plenty to open the breaker in 0.1 secs. So our sum becomes s = ( 200 x 200 x 0.1 ) / 115 =
0.56mm^2, which is significantly less than the 1.5mm^2 we actually have an hence is ok.
[1] In reality you can skip this stage since these are "standard" circuit designs where as long as you obay the length limits, you know the design "works".
Current carrying capacities are listed here:
formatting link
I'm not worried. I just don't see the logic.
Well the overload protection is achieved by the 13A maximum load on the socket. The only other question is will it be adequately fault protected. Assuming the rest of the circuit is a more substantial cable, the the loop impedance ought to be fairly low to that point. Hence it seems reasonable to assume you will get fault current in excess of
200A. That will open the 40A circuit breaker in 0.1 secs, and we know from the sum above that the wire will handle it.
Indeed. In fact due to the nature of the load presented by cookers, even a very powerful cooker with a theoretical peak load over 60A will usually be fine on a 32A circuit. (cooker diversity is calculated as 10A plus 30% of the remainder). So a 60A cooker would need a circuit provisioned for 10A + 0.3 x 50 = 25A, or 30A if there is also a socket on the cooker point.
So you don't think they got the compromise between cost and safety on ring mains wrong then? That would be why they added radials with breakers that actually match the cables because they didn't need to. Who knows, sometime in the future when all the old electricians have retired rings will finally die.
Rings have notable safety benefits over radials, namely that the CPC is doubled up over two paths.
Also, show me a practical 32A radial with multiple socket drops noting the BS standard restrictions on terminal capacities of socket and FCU accessories.
They also have several safety problems. They can have latent faults that the householder will only find out about when its too late. Things like broken earths (which negates your above argument) can just be there undetected for years unless you have regular inspections and real continuity tests. This isn't a solution though as the act of doing the test will make the faults more likely to occur.
Why do you need a 32A radial? I don't think you will find them in the OSG.
4mm will do the job with many accessories and you could always crimp in a joint if the accessory is to small. I wouldn't bother with a 32A radial to replace a ring.
HomeOwnersHub website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.