The switch gear, and the insulation of the cable must not be overheated. If you have maximum safe temperatures of 70C switchgear and 90C cable, then the 70C figure applies to the bit of cable attached to the switch or whatever.  But further along the cable route it may be grouped or in thermal lagging and there it may run at up to 90C. 

To be sure the copper of a cable of X mm2 stays below 70C , ensure loads are below the currents tabulated for thermoplastic insulation which are a bit lower than the currents for cables allowed to run hotter.

I thought you would have an emc question ;-)

Mike.

Thank you, Mike.

Your guidance is always significant. Tell me one thing. To 'ensure loads are below the currents tabulated for thermoplastic insulation which are a bit lower than the currents for cables allowed to run hotter', what should I do?

Hmm OK.

Let us consider a real example and see if that clears the fog.

Say you have a circuit that will run for long periods at 40A, and also for ease that the air ambient in the switch gear enclosure is the 30C that the tables assume.. (30C in the UK  - never ... ;-) )

you want to know will it be Ok  wired in 4mm² 4 core SWA.

this table based  on 4d4a suggests 35 A for the free air rating based on the copper reaching 70C. (as this table is for SWa with a PVC insulation) so at 35A or below , maybe the copper core reaches 70C, it should not exceed it.

the same chunk of metal (4mm cross sectiom 3 phase etc ) also in free space

reaches 90C at 44A according to this table based on 4E4A same cable contruction, same ambient air etc.

So these two tables together tell us that for the example of a 40A load is likely to get the 4mm² core to some temp above 70C  but below  90C

so although you could have used the 90 degree cable and it would be fine for 40A ,  the copper may run hot enough to damage the switchgear. so you should use 6mm cable instead, at least for the last bit connected to the switch gear. (or more simply, for the whole length) now it runs very cool ...

The swapping between the tables works because it is the same size chunk of copper in both examples, and as it happens the layout and so the thermal behaviour of the bits of plastic are similar. It is more daring to use example  tables for round cables to estimate the current rating of flat ones and things like that, and a bit of rounding in the direction to be safe is then a good idea.

The other common approach is to assume that temperature rise scales with power, so I²R, so if you increase the current by 10% to a first cut resistance is almost unaffected and temp rise increases by 110% times 110% or about 20%.

you can turn this round to estimate how much cooler an oversized cable runs - a 10% current reduction is a 20% power reduction so for example  a 25 degree rise becomes a 20 degree rise perhaps.

Mike.

Edit you can play the thermal game both ways -
If the cable is in a boiler room  at say 50C and the cable was PVC rated for 70C, then a 20 degree rise is all that is acceptable, half the  the 40 degree rise from 30 to 70 in the reference  tables. So we have to de-rate by  sqrt 2 and 35A amps becomes 25 amps.
if the cable is in a freezer room and always below say 0 degrees when the load is in use, you can say "but my insulation is rated for 90C, I can afford an 90 degree rise, instead of a 60 degree one (from 30 to 90)", so the heating can be 1.5 times higher so the current rating can  be sqrt (1.5) ~1.21 times higher,  for the same copper temperature so instead of 44A perhaps allow more like 54amps.

And then you can buy cables with a much higher insulation temperature. Cables to EN50306 WIRE are designed to run with core temperatures of up to  +120C. so you  see what look like very high current ratings

e.g.

and then you need to read the rest of the rather long document that explains the re-rating for groups, lower core temperatures various ambients etc

Mike.

Hmm OK.

Let us consider a real example and see if that clears the fog.

Say you have a circuit that will run for long periods at 40A, and also for ease that the air ambient in the switch gear enclosure is the 30C that the tables assume.. (30C in the UK  - never ... ;-) )

you want to know will it be Ok  wired in 4mm² 4 core SWA.

this table based  on 4d4a suggests 35 A for the free air rating based on the copper reaching 70C. (as this table is for SWa with a PVC insulation) so at 35A or below , maybe the copper core reaches 70C, it should not exceed it.

the same chunk of metal (4mm cross sectiom 3 phase etc ) also in free space

reaches 90C at 44A according to this table based on 4E4A same cable contruction, same ambient air etc.

So these two tables together tell us that for the example of a 40A load is likely to get the 4mm² core to some temp above 70C  but below  90C

so although you could have used the 90 degree cable and it would be fine for 40A ,  the copper may run hot enough to damage the switchgear. so you should use 6mm cable instead, at least for the last bit connected to the switch gear. (or more simply, for the whole length) now it runs very cool ...

The swapping between the tables works because it is the same size chunk of copper in both examples, and as it happens the layout and so the thermal behaviour of the bits of plastic are similar. It is more daring to use example  tables for round cables to estimate the current rating of flat ones and things like that, and a bit of rounding in the direction to be safe is then a good idea.

The other common approach is to assume that temperature rise scales with power, so I²R, so if you increase the current by 10% to a first cut resistance is almost unaffected and temp rise increases by 110% times 110% or about 20%.

you can turn this round to estimate how much cooler an oversized cable runs - a 10% current reduction is a 20% power reduction so for example  a 25 degree rise becomes a 20 degree rise perhaps.

Mike.

Edit you can play the thermal game both ways -
If the cable is in a boiler room  at say 50C and the cable was PVC rated for 70C, then a 20 degree rise is all that is acceptable, half the  the 40 degree rise from 30 to 70 in the reference  tables. So we have to de-rate by  sqrt 2 and 35A amps becomes 25 amps.
if the cable is in a freezer room and always below say 0 degrees when the load is in use, you can say "but my insulation is rated for 90C, I can afford an 90 degree rise, instead of a 60 degree one (from 30 to 90)", so the heating can be 1.5 times higher so the current rating can  be sqrt (1.5) ~1.21 times higher,  for the same copper temperature so instead of 44A perhaps allow more like 54amps.

And then you can buy cables with a much higher insulation temperature. Cables to EN50306 WIRE are designed to run with core temperatures of up to  +120C. so you  see what look like very high current ratings

e.g.

and then you need to read the rest of the rather long document that explains the re-rating for groups, lower core temperatures various ambients etc

Mike.

To Mike

Your explanation using specific figures makes me undersatand 512.1.5 remakably than before. I confirmed the relation between conductor operating tempperature and current-carrying capacity values.

Thank you again.