[borodi]

Reactive power losses due to the cable's inductance are probably the largest factor in the efficiency boost of using DC. Also you can use a somewhat higher voltage since you don't have to account for the AC peak, it does mean however that breaking a DC arc is harder since the is no 0 crossing.

[jacquesm]

Depends greatly on the material used and the cable construction. Typical: 60 strand aluminum (better skin effect properties than copper by the way) around a steel carrier. And yes, those are the largest factors in the boost to DC, but that's mostly because the resistance losses are there regardless so there isn't much else that you could improve on.

If you transmit a lot of power over very long distances then the higher the voltage the lower the current and DC gets rid of the skin losses so there's the case for HVDC transmission lines (which are extremely impressive feats of engineering, as are the substations).

Finally found a good picture of a cross section of a HV AC transmission cable:

https://en.wikipedia.org/wiki/Aluminium-conductor_steel-rein...

Based on that ruler that makes the AL wires about 3 mm each, and the skin depth at 50 Hz would be about 11.5 mm or so, so well within the range where the skin losses are extremely small (they are still there though, and when you're transferring Gigawatts every little bit helps).