Cable grouping

An interesting point made by one of my colleagues today.

If we consider a 36-way TPN distribution board that is full. How can we apply a grouping factor to those circuits at the point they leave the distribution board (the worst case position) without it making the cables so large that they can't be terminated if some of them are lighting or ring final circuits installed in trunking?

Clearly we could use sub-distribution, etc. but I have seen plenty of installations exactly like this with fairly typical 2.5sqmm or 4.0sqmm cables on the lighting and ring circuits.

Is it simply that the point of exiting the board is ignored and the main run of the cable is used for consideration of the grouping factor? If the cables are de-rated on the basis of where they come together at the board how does any installation ever practically make use of a 36-way TPN board without substantially over-sizing cables?

  • I think that they allow a lot less than 20 A after diversity. Our consumption over the past year has been an average of 2.4 A. I wouldn't fancy the bill if it had been 20 A.

  • After diversity maximum demand  is a lot lower than 20A by the time that your reach the larger substation. You may find a substation with 400A fuses on each of the 3 phases and 70 or so houses on each of those phases, so not much more than  one kW each on long term average, but of course the wiring and the transformer can take 100% overload without blinking for half an hour or so, which  is good, as so can the 400A fuses...

    And yes, the DNOs will have to do some fairly serious network re-inforcement, and quite quickly, if EVs and electric heat pumps etc take off at anything near the official predicted rate.

    Mike

  • Mike, you have reminded me what I read now... forgive me it was some time ago. It was actually based on 1kW per property not 20A!

  • Th tricky part I always find with these sorts of discussions is how you come up with a sensible set of calculations/reasoning in these scenarios without resorting to some pretty complex maths which is beyond many electricians (but not all electrical engineers - as amply demonstrated on occasion by this forum). It rapidly approaches full thermal modelling to determine the operating temperature of the conductors in multiple scenarios, and whist this is of course science-based practically it can appear more like witchcraft! 

    The grouping factor table (4C1) in BS7671 is fairly limited in its application (uniform groups of cables, equally loaded - how often does that happen?) seems to push to large sizes quite quickly, and once you stray from that table you're in withcraft territory.