derating multiple trefoils

There are a number of things that affect the rating, and make the choice to use one or other 'book value' more or less inaccurate. Ultimately it is all about the cooling. (well, assuming the run is not long enough that it is all about voltage drop instead)

Where are these bundles of 3 or 4 cables going to be mounted? I assume some sort of tray, ladder or basket but not too close to a wall or ceiling that will trap hot air.. Are the bundles spaced sideways by more than a bundle diameter apart, so air can flow between ?  Are the 2 layers  arranged so that the top layer is heated by the hot air rising from the cables below, or are they offset so that does not happen?

In general bundles side by side are better cooled than the same spacing one above the other,  but in real situations any ventilation or lack of it can have a huge effect.

I'm assuming these are not armoured cables, sa it looks like your 607amp  figure comes from Table 4E1A ,but if it had been AWA and table 4E3A moves you to 612A per core before grouping factors, again for 30C ambient air and  copper at 90C. (the core to core spacing increases with the AWA, so it is fatter and cools marginally better...)

The maker's figure of 775A for starters makes me wonder if their assumptions are different - perhaps they assume an allowance for a lower air temperature (20C perhaps) or a different mounting method.

There are some tables by ERA that consider cable ratings that are based on a load of tests don in the 1970-s and 80s and these do suggest that many of the IET regs values  can safely be pushed a bit. Certainly the next to nearest neighbour effect in a horizontal direction is far less than vertically, and for larger cables can be safely forgotten Smaller cables at smaller spacing not so...

I think your 6 bundle assumption is conservative and safe, and will run cool at full load, the makers 4 bundles may be a bit more gung ho, though it may well work out in practice, perhaps check their assumptions against the actual situation to see how much slack there really is.

Mike.

IMHO, it also depends a bit on the location.

In a warm climate I would adopt a conservative approach. Remember that the outside air may regularly reach 30 degrees, and that 40 degrees is likely in well ventilated switchroom or substation. Peak loads will tend to be in the hottest weather and might tend to exceed the design figures.

In the UK I might push the cables a little harder, to the manufactures ratings when these exceed other guidelines. Outside air in the UK seldom reaches 30 degrees, and moreover peak loads tend to be in the colder weather.

There is a lot of difference between say 2,000 amps in a heatwave with 40 degrees ambient and the same 2,000 amps on a cold winter day at around freezing point.

Finally with energy prices increasing, one should consider not just "how many amps can the cables carry" but also what is the annual cost of the cable losses. Reducing the losses by use of larger or more numerous cables may be worthwhile.

Saving about one volt at these sort of currents can amount to as much as 10kw. At current UK prices that is about £5 an hour. For a long hour load that could be in the region of £25,000 a year.

And of course the loading - is the full load on 24/7 or is it much lower with occasional busts of full load ?

Mike.

Another factor to note is the spacing of the cleats, this has to be designed in consideration of the cable bursting strength under fault conditions.

Jaymack

Also IEC60364 has no 0.97 flexible cable de-rating factor....

BS 7671 does however - as I recall it's a relatively recent introduction (17th AMD1?) and accounts (I think) for different manufacturing tolerances allowed for class 5 conductors.

  - Andy.

I should add that installation is cable ladder spaced 2De. Ambient 41C

Data centers so load is pretty constant at about unity P.F.

I have just posted a follow up query on use of quads and derating thereof

Spacing is 2De with 300mm between ladders

I largely agree with the comments above (I think you missed Jaymack's point that you'll need to consider longitudinal spacing of cleats along each cable/trefoil to design out bursting due to magnetic effects of through fault currents).

I would suggest, as Mike did above, that you get a copy of ERA* report 69-30 parts III and V, which might have more appropriate arrangements for what you're trying to achieve. It's often my go to for these situations.

*Now RINA of course

You will also need to consider asymmetric loading due to the magnetic fields.

If the cables are installed outside and exposed to the Sun, for a continuous load I would advise you to consider the heating effect of insolation and/or radiation from hot surfaces (e.g. reflection from a concrete floor). Or to shade the cables.

However, for a project such as you suggest, with multiple large supplies and presumably a substantial budget, I might be inclined to go for a detailed calculation per IEC 60287. In practice that means specialist software (or a consultant that has it), and you do need to understand what you're doing, but it frees you from the by-necessity generic tables and you can see exactly what's going on.