Current Calculations: Real load or MCB Rating?

[eveares]

Hi all, when calculating current/power aspect during cable size/voltage drop calculations do you base the current/power on a MCB's rating or the known maximum load.

I.e. Your running a supply up to a shed for a small 36W fluorescent light on its own 6A MCB in the shed.

Do you base the current demands used to calculate the voltage drop based on the 6 amps that the MCB can provide or the 0.156 amps the the light and circuit uses?

Regards.

 

[ericmark]

Where the maximum load is know then the maximum load. There are some odd ones like the ring final which is worked out at 20 amps in centre and the remainder even spread so with a 32A MCB ib = 26A and with 30A fuse ib = 25A.

However the impedance required to operate the trip is clearly based on trip so with a B6 then 7.67Ω still required.

Also a 36W HF fluorescent will use 36W but with a magnetic ballast it can be well above 36W as very voltage dependent. I got caught out fitting 110 volt fluorescent in a tunnel rated 58W so quick cal would seem 32 on a 16A supply so fitted 25 giving a little spare. When they tripped I used clamp-on to measure current nearly 25A. So I would double the tube size for calculations work on it being 72W not 36W.

 

[bernardgreen]

There are two "calculations"

[1]
Can the cable safely carry the current that the MCB will allow to flow before it trips. A 6 amp MCB will allow 7 amps for several hours. depending on type of MCB it might allow 10 amps for quite a long time when the circuit is overloaded. ( I have seen a 2 kW heater circa 8 amps running on a 6 amp MCB for 12 hours without the MCB tripping.

[2]
Are the voltage drops along the cable at maximum expected ( design ) load within permitted limits.

 

[ericmark]

I played with calculations on cable length and there are so many variables and the meters are not that accurate it is at best only an approximate figure.

From your other post you don't have a loop impedance meter so problem one is to get the information required to make the calculations.

Approximate figure you can draw about 200W from a B6 MCB supply at limits of loop impedance before the volt drop is exceeded.

For a job for a client all t's crossed and all i's dotted but in house really just two considerations.

1) Is it safe?
2) Does it work?

If I look at is it safe then the RCD is our friend the earth loop impedance can be huge and the RCD will still trip if volt drop is likely a problem then an active RCD will drop out if volts drop to low for it to work.

With all the meters loop impedance can be calculated for a new circuit and with that the local or total volt drop but one needs a start point with a reading and you have said in another post you don't have a meter to measure loop impedance.

Over the 7.67Ω the magnet part of trip will not operate so only the thermal part will work but in real terms with RCD protection not convinced the magnetic part is that important the thermal bit will still trip before any cable melts.

I looked at GU10 LED's and the voltage latitude is huge will still give out full light at around 185 volt so by selecting a lamp be it LED or HF florescent which has some form of switch mode supply in real terms the volt drop can be huge.

Some times we need to use common sense. If I was to install a cable for a gas cooker I would still use at least 6mm as I would plan for as some time in the future this being swapped for an electric cooker. To install the cable to supply the 2W for a gas cooker with a B32 MCB supplying an isolator which in turn supplies a fused connection unit with 2A fuse to supply gas cooker it would be expected for that FCU to be swapped for a cooker connection unit if moved to electric. The volt drop would be worked out for 32A even though supplying 2A as it is expected that the cooker will be changed in the future.

If we look at a garden shed and a light what you have to ask is what is the likelihood of this being added to in the future?

If to change the appliance or add to the circuit it required one to swap the MCB then likely the person swapping will check to see if it can be swapped before going ahead. But if already fitted with a MCB which is large enough then often people will swap the appliance without checking.

I am not saying that is what should be done clearly one should always check but likely most home owners will not bother.

So if supplying a shed with a length of cable where you know the volt drop will be exceeded taking the full 6A then either swap the MCB or fit a FCU with suitable fuse. OK not strictly required, but common sense says fit protective device to match the circuit parameters as well as appliance.

 

[JohnW2]

Where the maximum load is know then the maximum load. There are some odd ones like the ring final which is worked out at 20 amps in centre and the remainder even spread so with a 32A MCB ib = 26A and with 30A fuse ib = 25A.

In terms of VD calculations, what I find is 'odd' is that the apparent convention you quote only ever seems to be mentioned in relation to ring final circuits, whereas the considerations in relation to a distributed load on a sockets circuit seem to apply equally to radial circuits - i.e. one would expect that the same argument would allow one to calculate VD on a 32A radial circuit on the assumption that there was a 26A load at the end of the circuit.

With sockets circuits of any sort, one clearly has no choice but to calculate 'maximum VD' on the basis of the 'maximum permitted load' (even if there are 'concessions' in the calculation of that, due to the distributed nature of the load - as above). However, when there is a fixed-wired load, as you say, it seems appropriate to determine the VD in relation to the actual particular load - in the OP's (fairly extreme!) case, on the basis of a 0.156A load.

Kind Regards, John

 

[Risteard]

i.e. one would expect that the same argument would allow one to calculate VD on a 32A radial circuit on the assumption that there was a 26A load at the end of the circuit.

I don't see how as all of the current would be flowing in the same direction, so it could not be distributed.

 

[JohnW2]

i.e. one would expect that the same argument would allow one to calculate VD on a 32A radial circuit on the assumption that there was a 26A load at the end of the circuit.

I don't see how as all of the current would be flowing in the same direction, so it could not be distributed.

I don't understand your argument. The reason for the 26A quasi-design current has got absolutely nothing to do with 'what direction the current is flowing' - it's because 12A of the load is not at the end but, rather, evenly distributed along the length of the circuit (6A evenly distributed along each 'arm' of the ring) Indeed, from the point of view of VD calculations (which is what we are talking about), you could split a symmetrically-loaded 32A ring into two 16A radials, with 10A load at the end of each and 6A evenly distributed along the length of each of the circuits, and the calculations would be exactly the same.

Kind Regards, John

 

[PrenticeBoyofDerry]

Hi all, when calculating current/power aspect during cable size/voltage drop calculations do you base the current/power on a MCB's rating or the known maximum load.

The standard method is to add up the load the circuit will require, this load can then be calculated to give you the value of the design current.
for example, you can say you have for ease of explaining s circuit with 15 lamps on each at 60W that will give you 15X60 equaling 900 Watts.
This can then be divided be the nominal voltage, arguments sake we call it 230V.
So we have 900W/230V equaling 3.91A.
We can now select the protective device which normally would be the nearest size above the design current, so that would lead us to be selecting a 6A breaker.
We now take into consideration the size of the cable that can safely carry this current, this is not always straight forward as there are often a number of things to consider, which would include the reference method of installation/the method used and influence that will cause the cable not to dissipate heat as it would in an ideal situation.
This can be that the cable is within insulation along it's route, routed within conduit/trunking or in a space that lessen the heat dissipation or maybe has a higher air temperature. Once we have factored this in there is also grouping of cables to consider and the length of the circuit for voltage drop.
It could be possible that when these de-rating factors are calculated that you may require a larger CSA of cable than expected.
Link http://www.napitonline.com/downloads/CP 2 07 P 10-11 16th Cable Sizing 1 of 2.pdf

 

[JohnW2]

The standard method is to add up the load the circuit will require, this load can then be calculated to give you the value of the design current. for example, you can say you have for ease of explaining s circuit with 15 lamps on each at 60W that will give you 15X60 equaling 900 Watts. This can then be divided be the nominal voltage, arguments sake we call it 230V. So we have 900W/230V equaling 3.91A.

Indeed, but as I said before , one can only really do that in the case of 'fixed' hard-wired loads (maybe like the lights you mention). With sockets circuits, one has no choice but to assume that the circuit could be 'maximally loaded' (per rating of MCB/fuse) - although there is obviously then the complication that, in the case of a ring final circuit, one is allowed to use a cable with a current-carrying capacity of only 20A, although the circuit may be supplying 32A.

I say that the lighting example 'may be' an example of a'fixed hard-wired load' because there is obviously a bit of a 'socket-like problem' with some lighting, since the designer may have no control over what a user might put into a lampholder. That is, I presume, why the OSG still says that each lampholder should be considered as a load of at least 100W (before application of diversity) - although that has probably become a bit daft in the almost-post-incandescent era in which we live!.

Kind Regards, John

 

[ericmark]

That is, I presume, why the OSG still says that each lampholder should be considered as a load of at least 100W (before application of diversity) - although that has probably become a bit daft in the almost-post-incandescent era in which we live!.

Kind Regards, John

You beat me to it I was also thinking about the standard lamp size. We have the same problem with heat from lamps 422.3.1 starts at a 100W clearly a 7W LED does not produce the heat.

But there is a special GU10 fitting with a pin in centre which fits inside a hole on the LED lamp to stop tungsten from being fitted and a special bayonet fitting with three instead to two side pins again to stop tungsten being fitted with these versions then I would say the 100W is daft.

In the case quoted a florescent fitting also has a fixed tube size.

However over loaded lighting circuits are all too common with planetarium like spot light lighting where some one has seen a white ideal home kitchen with the spots fitted to reflect the light off the white surfaces then tried to emulate with dark surface which does not work.

 

[PrenticeBoyofDerry]

Indeed, but as I said before , one can only really do that in the case of 'fixed' hard-wired loads (maybe like the lights you mention). With sockets circuits, one has no choice but to assume that the circuit could be 'maximally loaded' (per rating of MCB/fuse) - although there is obviously then the complication that, in the case of a ring final circuit, one is allowed to use a cable with a current-carrying capacity of only 20A, although the circuit may be supplying 32A.

I say that the lighting example 'may be' an example of a'fixed hard-wired load' because there is obviously a bit of a 'socket-like problem' with some lighting, since the designer may have no control over what a user might put into a lampholder. That is, I presume, why the OSG still says that each lampholder should be considered as a load of at least 100W (before application of diversity) - although that has probably become a bit daft in the almost-post-incandescent era in which we live!.

I am in total agreement with your comments, my post was purely to offer an example on the method used to design a circuit. In most cases when it comes to socket and lighting circuits, we do indeed need to consider that there is very much likely to be load changes and provision can be made for this. And the nuts and bolts of a socket circuit would suggest that the circuit should not exceed 7kW and we can then work backward from that, also the recommendation was no more that 12 lamps on circuit but with the introduction of the much lower energy efficient lamps over recent times we need to make exception for this.

 

[ericmark]

I am in total agreement with your comments, my post was purely to offer an example on the method used to design a circuit. In most cases when it comes to socket and lighting circuits, we do indeed need to consider that there is very much likely to be load changes and provision can be made for this. And the nuts and bolts of a socket circuit would suggest that the circuit should not exceed 7kW and we can then work backward from that, also the recommendation was no more that 12 lamps on circuit but with the introduction of the much lower energy efficient lamps over recent times we need to make exception for this.

7kW I assume refers to ring final this is an odd one out as volt drop worked out at 26A but ELI at full 32A.

With a supply to a socket or even a cooker connection or fused connection unit I would consider VD worked out by fuse or MCB size as it is likely the appliance can change without an electrician doing the work and using a meter to test.

12 lamps at 100W = 5.2A since was a 5A fuse this makes sense and now with 6A MCB it would be 14 lamps again lamps are an odd one as no diversity they may be all on together.

So that's one using Design current for circuit Ib and two using size of fuse or MCB. What we are talking about is using common sense rather than sticking to regulations and guides to the regulations.

If we were to fit a light in the loo at bottom of garden placed where the muck cart can empty it then you have a long run of cable and little chance of anyone wanting to use the building for any other use while being used as a loo so running 1mm cable and fitting a 60W bulb even if supplied from 6A MCB you would calculate at 100W for volt drop.

However with a garden shed in same position now the muck cart no longer calls will likely have at some point a socket fitted or security light and so it would make sense to work out volt drop on the size of the MCB.

But the regulations clearly have VD worked out on the design current so 190 meters of 1mm² cable would allow two bulbs (200W total) with a volt drop under 6.9 volt (around 6.3 volt) and be just within the 7.6Ω required to trip a B6 MCB with a Ze of 0.35Ω. The limiting factor is the loop impedance rather than the bulb size so using a 4A MCB would allow even more cable and with a 1A MCB it gets silly.

This raises the question if running a cable like that would it not be better to have a FCU at origin and actually fit a 3 amp fuse and then the loop impedance is 16.4Ω it's seems a daft exercise with a 35W lamp.

I will guess the poster is not thinking about a florescent lamp, but intends to fit some thing completely different and wants the argument to say what he intends is OK. Which is why I am saying use common sense. If he wanted to fit a gas cooker in his shed at bottom of garden feed with a B32 MCB with 4mm² 100 meters long then yes the VD and Ze would be within limits for the gas cooker, but swap to an electric and VD would only allow a 10.5A draw so for that job it would be unsuitable. One would use 10mm² cable as one would expect at some time for the gas cooker to be swapped to electric other wise one would not have used a B32 MCB.

 

[JohnW2]

7kW I assume refers to ring final this is an odd one out as volt drop worked out at 26A but ELI at full 32A.

VD is obviously crucially dependent upon current. However, current only effects ELI slightly (increased temperature due to current flow increasing resistivity), and AFAIAA one normally does not explicitly determine ELI 'for any particular current'.

If one is calculating Zs from BS7671 mV/A/m figures, then one will be implicitly assuming the 'worst case' of the conductors being at 70° - but whether conductor temp would actually get anything like that high in a ring final loaded to 32A is a very different matter.

As for measured, rather than calculated, Zs, my impression is that people just 'measure it', with the conductors probably usually fairly 'cold', and do not attempt to work out what it might rise to if the circuit were fully loaded with a certain current - but maybe I'm wrong!

12 lamps at 100W = 5.2A since was a 5A fuse this makes sense and now with 6A MCB it would be 14 lamps again lamps are an odd one as no diversity they may be all on together.

If you accept the guidance of the OSG, having calculated on the basis of 100W per lampholder, you are then allowed to apply 66% diversity to the total (in domestic premises).

Kind Regards, John