Normally the trefoil (or quadrafoil with neutral - is that a word ? you know what I mean I hope ) groups are kept a bundle diameter or two apart on the tray/ rack or whatever , so there is air flow between and no big extra group of groups  derating factor - have you over done the derating  by mistake ?

(an old discussion thread may help)


How many bundles did you have in parallel, and are you assuming aluminium core or copper core singles. Perhaps start at 3 or  4 bundles at 500- 700a per bundle and see how that looks, then try one size up or one down from there.  The size and number of the transformer terminals can be also be used as a clue to what the makers are expecting.  Each bundle needs all 3 phases and  a neutral nearby, NOT a bundle of phase 1 and another bundle of phase 2 etc.

Come back if it makes no sense or does not seem to fit..

M.

Mike,


Thanks for the reply thats great.


I wasnt aware that you could use the 2 dia cable rule when in a trench (Table 4C6 makes no mention). Is table 4C6 assuming that all cable are grouped together with 50mm in between? 


This makes life a lot easier if you can disregard the grouping factor!


Thanks

After further looking I’m not sure that can be applied to cables in trenches (sorry I wasn’t clear). It seems the grouping factors make it near impossible to get realistic figures..

Hmm the trench may need to be a bit wider than you thought, or indeed the example width of  450mm -  you did not say the magic word 'trench' to begin with....


OK, look here   either derate the single core rating of your AWA or whatever by the amount in 4C6 relative to the one wire case, not the already de-rated trefoil case.


Or look at the wide spaced trefoil on tray case,  (more on that here) and realise that with enough room for air flow  it can becomes more or less the same.

another old thread on non-trefoil groups.


In the limit you could have multiple separated trenches, but do not, as it will make length matching a sod.

There are tricks involving backfill with specially graded sands and things to improve the dissipation, but see how close you are without that.


M.

Thanks Mike, Ive had a look at those links and still cannot find anything concrete.


Ive had a look at a couple of old jobs where an IDNO have sized the tails and it appears they do not use a grouping factor. 


An example is a 1500KVA TX, ACB set at 2125A supplied by 3 groups of 4 x 1C 500mm2 AWA cables in a trench between the sub and LV panel.


If each group can carry 731A in Trefoil, which gives a total capacity of 2139A.


I suppose the only way to get sensible sizes is to ensure that the trench has a degree of ventilation, which is possible due to being internal and also to specify a cable ladder to act as a heat sync. We could then potentially use table 4C5 and use the 2DE spacing to omit the grouping. Any kind of derating factor even at 0.8 would give ridcilous sizes it appears.

Okay, after further looking. 


For cables installed as per Installation method 117, with a ventilated trough. Which i would expect from an internal trough with an opening each end for connections to the TX and Switch panel, I can also specify a GRP ventilated lid. A 630mm cable from 4D3 can carry 771A horizontal flat spaced. This is the same figure if the cable was installed ladder horizontal flat spaced, I think therefore we can use the Trefoil method of the cable ladder as a reasonable substitute as there is no trefoil option in Installation method 117, this can then be combined with the grouping omission of 2De with neutrals in between.

Another factor to consider if a significant distance is involved is the energy cost of cable losses.

A loss of 1% is of course 15 KW. At current energy prices 15Kw will cost about £2 an hour.

For a long hour load, that could be about £16,000 a year. Spending even £30,000  on larger cables would be an excellent investment with a payback of less than two years.


Upsizing cables to reduce losses from say 2% down to 1% is well worth considering.


Is the transformer already existing and in in place ? If not, could it be placed closer to the load to save on the huge LV cables.