I'm not feeling the love.


What lengths are the various sections of the circuits, particularly the "home runs" (plural). Do you really need a capacity of 180 amps at the sockets? Why can't you just dog some of the outlets together in the MDB so it only needs a couple or three 20 amp rated home runs?


Can't it just have a single distribution circuit to a second distribution board or consumer unit? It seems waste of labour and materials if this is a installation that is going to be in place for many years.


Andy B.

Hi David


The DB is 3 phase for power and lighting supplied from 125A TPN tap off. I have around 2-3 MDBs for power and another 2-3 for lighting. 


Not worried about the lighting as the loads can be determined. The issue is with the power because since these are feeding sockets I cannot be certain about what could someone be connecting there so thought I should be using max value e.g. 20A

20 amp socket circuits are generally wired using 2.5 mm cables and 32 amp socket circuits using 4.0 mm cable, given the nominal loads of 5-6 amps per circuit you are designing for the whole design seems over the top.


As you have probably gathered I would not even be considering installing that MDB, so unless there is a convincing argument for installing it I would bin it and perhaps look at three 4.0 mm 32 amp socket circuits, one on each phase, then if that design doesn't work try dropping down to 20 amp circuits with 4.00 mm cable and if that doesn't work move the distribution board closer.


Andy B.

Because 9 x 4 x 2 = 72 mm2 of copper for some sockets without the CPCs or upgrading the size of the cable.

Andy, thank you for your comments. While I do agree in principle with your views about the cable sizes, the point is how to prove compliance with BS 7671 through valid electrical calculations. In general I understand that modular wiring system‘s main advantage is that it can significantly reduce the installation time as everything would be prefabricated and it would be like a ‘plug and play’  system. On the other hand, examples like this could maybe show that there are some disadvantages as well?


The home run cable (which would house the 9 circuits in singles and thus it would be 27 core) will be around 40m. I believe that voltage drop is not a problem. How to apply a reasonable grouping factor is and I feel the way BS 7671 is written, I end up with having to oversize the cabling.

If the circuits are never going to be unplugged again to reposition partitions or make other alterations is there really any advantage over running in a forty metre distribution circuit to another distribution board?


If you look in the text books by Brian Scaddan and others you will see examples of cable calculations where you end up needing 16 mm twin and earth to supply a 3 kW immersion heater, because of correction factors. In reality you use a smaller cable and route it differently along with other tweaks.


I changed a consumer unit in a two up, two down house. The full tally was living room, kitchen, bathroom, cellar and two bedrooms, the installation has a sixty amp main fuse and five 32 amp socket ring circuits spread across five rooms, where there had originally been one socket ring for the whole house, given that the customer had bought a dual RCD consumer unit for me to install the five circuits were split between two RCDs, two circuits would have been perfectly sufficient.


There comes a point where there generally ceases to be any advantage whatsoever in having additional circuits, in your installation it sounds like a socket circuit on each phase will do the trick, reducing the number of circuits from nine to three, in turn reducing the number of conductors in the home run significantly, even more so if it becomes a three phase distribution circuit to a distribution board.

I think that I can see the point of this: 3 circuits on each phase and the modular construction is rather like the wiring loom of a modern motor car.


Given that the total supply is 125 A (presumably with an OCPD somewhere) that is an average of 14 A per circuit. If some of them are rather lower because they are lighting circuits, they may be ignored - see Note 9 to Table 4C1.


So if that leaves 6 circuits in 4 mm² singles, you have a CCC of 32 A (RM B) x 0.57 = 18 A each. Allowing for diversity, perhaps the MCBs should be downgraded to 20 A?


Then what is realistically the current consumption of each circuit? A few PCs on desks? Surely not electric space heating!

Chris, your description is more or less the scenario. Only difference is that there are separate MDBs for power and lighting and as such, the grouping factor would be 0.5 and end up with the rating of the home run to be 16A. Is it ok to use the estimated design current for sockets? Since anyone could potentially plug anything into the sockets wouldn't I need to allow for the worst case scenario which would be 20A? I guess ideally I would have to change to 16A MCBs supplying the power instead.


But then again someone could argue that, as BS table 4B5 should be used instead where I d have 0.45 grouping factor (18 loaded cores)! This would make even the 6mm2 cable to fail on a 20A MCB ie 41x0.45=18.24 < 20!! (Note that I am leaving out the 0.95 derating factor for flexible cable that many times is used in modular wiring systems as a home run cable) What am I missing here? Is there another approach to calculate this and show compliance with BS? I just feel that when I try to follow BS strictly, I end up with way over-designed results..

Presumably your tap off has fuses, so why not fit 63A or 40A or not use all 3 phases? Then your maximum loading is controlled to a sensible level and your problem goes away using the methods above. Surely you don't really need 250 kVA for this installation? If you do modular wiring is not the correct choice. A large cable to a local DB from the tap off is the way to go, much simpler and cheaper as well.

Chris - are you designing a modular system in general terms (as if you were the manufacturer of the system, which could be then installed anywhere) or checking what's needed for a specific installation?


If the latter, could you use some expected load characteristics (e.g. W/m²) to put some sensible limits on the total load - and hence exclude the equivalent of some of the circuits from the calculation - perhaps as already suggested, enforced, by some upstream overload protection.


If the former it seems trickier - but are there any "manufacturer's instructions" that put any limits on how the system might be utilized (e.g. recommended limits on floor area served, total load and so on)?


I presume the system is something like this - https://www.theiscraft.co.uk/solution/modular-wiring/ (random Google find) - which begs the question is there any mileage in using a larger number of MDBs each with fewer 'ways' to reduce the grouping in each conduit run?


   - Andy.