[Enthalpy]

The resistivity of 99.999% aluminium drops by /1000 at 20K without most superconductor drawbacks.
  nvlpubs.nist.gov (5MB) p20 - lss.fnal.gov
Shall we chill to 20K the conductors of a 50GW 2000km ±960kV line to reduce the losses? An ambient temperature (RT) design is there:
  chemicalforums and next messages

The purer but smaller conductor costs as much metal and oil. At 1/3 the Carnot efficiency, coolers waste 1% of the carried electricity instead of 3% in Al at RT, saving around 0.1G€/year.

Feasible, in absolute terms:

• 50 plies in vacuum insulate the cold Al conductor.
• One more tube separates the vacuum from the oil. The external tube remains.
• Few vertical straps of Ti, duplex steel, glass fibers... hold the conductor.
• Bumps in other directions touch only during seisms.
• The helium in the line equals few hours of worldwide production.

But:

• Assembling vacuum and multilayer insulation in the fields makes trouble.
• Coolant helium must run quickly within the pure Al conductor.
• It takes a cryogenic station every 5km, not pumps and dissipators every 25km.
• If the vacuum is lost, electronics must swiftly protect the line.
• This protection can't easily prolong the operations accepting losses.
• A line for <50GW <2000km favors low-tech even more.

Put together, such a line is far less reliable than the RT design. But please don't extrapolate to electromagnets or motors.

Also: the RT and 20K designs aren't optimized. A bit more metal or oil at RT might well save as much electricity without all the vacuum and cryogenic fuss.

So the possibility exists, but I put it in sleep mode for now.

Marc Schaefer, aka Enthalpy