Hi Martin,

Did you find any supporting information that allows you to ignore the derating factor for short lengths of cable?

SSEN G81 guidance allows 15m of cable in a duct for every 100m before derating is applied. Not sure if this comes from the ERA 69-30 method.

The thinking is that from the open end of a pipe or duct there is some airflow for heat to escape and circulate, and throuhg circulation if both ends are open, as well as the conduction of heat along the metal conductors, though that  is not usually significant beyond more than a few tens of cable diameters. Near the end, cooling is part way between the extremely poorly cooled case of the middle of a very long duct with no air flow,  and the open cable case. Having said that, I do find 15m, which suggests a distance from the hottest spot to the outside of about 7.5 me to be a fair old way, and I'd expect some additional restraint requiring both ends being open to fresh air and the duct not being very full - is that mentioned anywhere ? Equally this is likely to be intended  for larger cables than I am used to so my mental rules of thumb may be a bit off.

Mike

Or it might be based on the logic that a short length of cable near the substation is easier to replace ... so a shorter lifespan is deemed acceptable. Just a theory.

   - Andy.

In some case maybe, but I'm not sure how you would change just the bit in the duct from the middle, or even the end, of a long length, unless at the time you laid the duct in you installed the space to have some joints. If you were planning  that far ahead, you may as well have changed cable size at that point too. Lifespan vs temperature is a very non-linear thing, or more accurately lifespans extrapolated from accelerated aging trials are - in reality we have not even had PVC installed long enough for cold running samples of it to have reached end of life. What we know is that once we get away from the running hot and steadily degrading state, other failure mechanisms that dont show in the accelerated testing can catch you. (estimate the mean time to JCB strike anyone ?)  Or after some decades. other factors now mean the cable is in the wrong place or the wrong size.

Mike

Very interesting.  Can i ask where the reference for the conduction being in the region of 10s cable diameters has come from?  Would be keen to read more about this.

Thanks,

Chris

Very interesting.  Can i ask where the reference for the conduction being in the region of 10s cable diameters has come from?  Would be keen to read more about this.

Thanks,

Chris

it is just a consequence of  the conductivity of copper (or aluminium) and the length of cross section to diameter. there is almost a one to one mapping between thermal behaviour and electrical resistance, in that the temperature drop is proportional to path length and inverse proportion to area. Note by the way this way of thinking this does not work for things that can convect ;-)

Basically solve for the easy case of a well-lagged half wire of your chosen diameter, and look at the thermal gradient either side of the lagged/unlagged transition . as you might hope, far from the junction the temperature is at the steady state value for the infinite lagged or finite unlagged case respectively.

Near the boundary, there is a cross-over where the amount of energy escaping down the wire to escape  from the hot side to the cool side depresses the temperature in the well insulated side and raises it on the cooler side. Temperature is not discontinuous.

basically to solve, you just need the temperature rise per watt per metre of dissipation in your 2 states, and the conductivity.

Miike