Given this post, showing a huge 4.4" wide cable: Original reddit post
Also a Electric Engineering StackExchange question: Electrical StackExchange question
And somewhat knowing that current mostly flows on the outside rim of the inner copper core surrounded by the white XLPE, I'm assuming is due to magnetic forces.
Question is, why is it still necessary to have the very center core be copper as well, instead of filling with the cheap plastic material similar to the surround, like XLPE or something similar, and saving the cost of copper, or does that create its own problems? (Let's assume the center core material would be 100% EM free)
Best Answer
There exist high-voltage aerial conductors with hollow cores. These aren't quite XLPE-insulated cables, but they illustrate that someone, somewhere, does make power conductors with hollow cores.
From the Westinghouse T&D Book, Chapter 3 "Characteristics of Aerial Lines", we have:
From what I can tell, these hollow overhead conductors are meant to reduce corona losses, not so much to reduce skin effect losses.
As far as insulated cables go - per the original question - the following points apply.
I have some data for 87/150 kV, copper, 1,600 mm², XLPE single core cable. This is a really big cable, comparable to the 1,750mm² examples you first looked at.
Resistance values for this cable are quoted as: DC resistance 0.0113 Ω/km at 20° C (about 0.0144 Ω/km at 90° C), and 50Hz AC resistance 0.0159 Ω/km at 90° C. So the difference between the DC and AC resistance is only 10% - the skin effect only adds about 10% to the resistance of this cable. (Note this is at 50 Hz. The skin effect doesn't really do much at this low frequency.)
At best, if you eliminate the skin effect entirely, you will only save 10% of your losses, which are pretty small already.
(Meanwhile, the inductance of this cable is 0.122 Ω/km at 50 Hz. That's 10 times more than the resistance.)
If the cable has a hollow core, it will have to be bigger to carry the same current. This means the cable needs more space to bend around corners.
The nominal diameter of the above cable is 99.8mm (call it 100mm.) The minimum bending radius for most cables is 18 × the diameter. So you need about two metres to make a 90 degree turn.
If it were hollow cored, it might need to be 150mm outer diameter to carry the same current. So you would now need three metres to make a 90 degree turn. That means any cable pits would have to be three metres wide. It's difficult to get cable pits in that size, and even more difficult to get lids for them! (Especially if you want to be able to drive a car or truck over those lids.)
When installing a hollow-cored power cable, you'd have to be careful not to bend it too sharply, or it might collapse on itself (like a copper pipe.)
To summarise, making hollow-core power cables (for 50 / 60 Hz) wouldn't give you much improvement in skin effect (10% less resistance at best), they would need more space to install, and more care during installation. That doesn't sound like a winning combination.
The situation is very different for higher frequencies, where it's normal to have special cable designs that reduce the skin effect, i.e. Litz Wire.
On a third track - regarding the "bi-metallic" cables mentioned by @mickkk -
These are quite common in overhead power lines. A common construction is "ACSR", aluminium conductor, steel reinforced.
Example from the Olex Australia Aerial catalogue:
The conductor is constructed with steel wires on the inside and aluminium wires on the outside.
The steel is there for mechanical purposes - to increase the tensile strength of the conductor. That means you can hang longer spans, so you need fewer towers, and you save money.
Any reduction in skin effect is purely a bonus side-effect of the improved mechanical strength.