Current carrying capacity of XLPE/SWA cables

Correct, although Thermosetting cables (4E4A) are rated at 90 degrees there is very little equipment that can be connected at this temperature as you rightfully pointed out. On Amtech, there is a drop down to run these cables at 70 degrees, but Amtech simply uses table 4D4A for the calculation which is a shame as the thermosetting cables can dissipate heat better than thermoplastic and therefore could get a higher current carrying capacity in theory.

the 90c thing is most use for a mixed route - so the ends may be better cooled than the middle for example.

Also although the cores may be at 90 in the centre of the bundle, once the cable is split to singles, they are much better cooled, so in practice the terminal temperature has more to do with the environment. This is especially significant for thinner cables that have more surface area relative to cross-section.

I think that this matter encapsulates the conflict between, on the one hand, being strictly compliant, and therefore conservative; and on the other hand using engineering judgement.


I don't know what terminals are rated to 90 deg C, but I suppose that an industrial oven might be. I assume that in industrial installations, cables may run near their max temperature (so don't touch them else you will be off to A&E) but in domestic, commercial, and light industrial (i.e. small units in an industrial estate) I doubt that cables get nowhere near 70 deg C.


As Mike has pointed out, cables bunched in conduit or trunking might well get to 90 deg C, but even a few inches of freedom in a panel would allow them to cool down.


So if you err on the safe side, the worst thing that you have done is wasted a little copper and money.


ETA: if this is in the context of an EICR, then I think that you do have to make a judgement call (C2 or C3 - I don't think that a non-compliance can be ignored) but back it up with reasoning in the Observations section.

Take a look at table 6 in BS EN 61439 - that should tell you that the allowable temperature rise on terminals is 70K (usually the ambient is 20C) - so compliant switchgear could easily be operating at 90C terminal temperature when at full load - which will clearly melt PVC.


XLPE insulated copper connected to compliant switchgear can happily run at 90C - you would expect the external surface temperature of the cable to be around 80C


The problem usually arises at the load end - and often it can be shown that the actual connected load is usually less then the protective device setting (often significantly) and therefore usually well below the 90C rating of the conductor - even if the load ratio is much closer, loads often also don't operate continuously - so the cable time constant factor also becomes important (basically the point at where a change in load causes change in temperature (we usually work it out to a 66% change in temperature following a change in load ie you are working out the change from Time, T0 to T1 following the load step change)


For the vast majority of cases the system is neither fully loaded nor constantly loaded - but it is a common mistake made by designers and then picked up by testers - where it becomes a shitstorm of acrimony and argument based entirely on "What Iffery" that has no bearing on reality


Regards


OMS

And if you believe that line from whatever cowboy film it was said at a card game that "a Smith and Wesson beats four aces"  (*) then in much the same way a

 

.. storm of acrimony and argument based entirely on "What Iffery"

conjecture can sometimes  be beaten  by real measurements, in the form of those stickers that change colour to indicate peak temperature, as a sort of high tide marker- only really good on large diameter stuff, but then on large diameter lugs and bussbars,  the cost of rework and or over sizing becomes  most significant. Looking at the stickers every xx months reassures everyone it has not been overheating yet.


* I'm not suggesting that people should be shot for their electrical work,  well only a few anyway.

I agree absolutely with OMS above. There is far too much nonsense about "what if" going on and until someone can show this situation as "correct" they should remain silent. A temperature a couple of degrees different from expectation is not a major problem. Real difficulties are nothing like that simple, which should be fully understood by everyone, including the HSE!


David CEng MIET

As an aside, there may be nothing wrong with a 306A cable protected by a 355A fuse if overload protection is not required

Lots of situations 50. Consider a heater with a full load current of say 20amps. Providing the cable can carry 20 amps then it is not possible to overload the circuit and thus the design can be based on fault current protection by employing the adiabatic formula.

In what situation would overload not be required? The only thing I can think of is maybe life safety equipment, but I’d still want the cable sized to breaker I think

or anywhere where overload protection is provided downstream - the common example is industrial motor circuits where the control gear next to the motor provides overload protection so the overcurrent protection at the DB only has to provide fault protection - and is frequently over-sized compared with the cable rating to ensure discrimination with the motor's overload.


The same principle can even be found in simple domestics - e.g. spurring off a 32A socket circuit using 20A cable to feed one single or one double socket.


  - Andy.