About 5% of all electricity generated in the US is lost as heat in transmission and distribution lines. That sounds small until you do the math: US generation is roughly 4,000 TWh per year, so we're losing about 200 TWh annually — enough to power all of Mexico. At average wholesale prices, that's around $6–8 billion per year, literally warming the atmosphere for no reason. Superconductors have zero electrical resistance. What if we just... replaced all the wires?

The physics is real, but the thermodynamics are brutal. High-temperature superconductors (HTS) like YBCO (yttrium barium copper oxide) lose their resistance below about 93 K (−180°C). "High temperature" is relative — that's still colder than anywhere on Earth's surface. Every meter of superconducting cable needs to be kept inside a cryostat, continuously cooled by liquid nitrogen (boiling point: 77 K). The cable is no longer a wire; it's a thermos.

Let's run the numbers on cooling cost. The US has approximately 160,000 miles (257,000 km) of high-voltage transmission lines. A typical HTS cable system dissipates about 1–2 watts of heat per meter into the cryostat from the ambient environment, even with good vacuum insulation. Taking 1.5 W/m:

Cooling load = 257,000 km × 1000 m/km × 1.5 W/m
             = 385 MW (thermal)

But here's the catch: removing heat at 77 K is thermodynamically expensive. A cryocooler's coefficient of performance (COP) is roughly T_cold / (T_hot − T_cold). At 77 K with a 300 K ambient:

COP_Carnot = 77 / (300 − 77) = 0.345
Real COP   ≈ 0.345 × 0.30 (practical efficiency) ≈ 0.10

So for every watt of heat removed at 77 K, we spend about 10 watts of electrical power. Our 385 MW thermal load demands ~3.85 GW of electrical power just to keep the cables cold. That's roughly the output of four large nuclear plants — dedicated entirely to refrigeration.

Compare that to the 5% transmission loss we're trying to eliminate. Five percent of average US load (~450 GW) is about 22.5 GW. So the cooling penalty eats up 3.85 / 22.5 ≈ 17% of the savings. We still net roughly 18.6 GW — not bad! The energy math actually works in our favor.

The cost math is where it falls apart. Current HTS cable costs around $50–200 per kiloamp-meter. A single high-capacity transmission line might carry 3,000 A. At a conservative $80/kA·m, that's $240/m, or $240 million per thousand kilometers — just for the superconductor tape, ignoring the cryostat, vacuum jacket, cooling stations every 5–10 km, and installation. Replacing all 257,000 km of US transmission lines would cost on the order of $60 trillion, roughly three times US GDP. And that's before you build thousands of refrigeration stations and a liquid nitrogen supply chain that dwarfs anything existing.

Where it does make sense: dense urban cores. Superconducting cables can carry 5–10× the power of conventional copper in the same cross-section. ConEdison tested a 138 kV HTS cable in Manhattan in 2010. When right-of-way costs $10,000+ per meter and you can't dig wider trenches, the economics flip. A few hundred meters of superconducting cable can defer building an entire new substation.

There's also a tantalizing dark horse: MgB₂ (magnesium diboride) superconducts at 39 K and costs a fraction of YBCO. It needs helium-temperature cryogenics — harder — but the raw material is dirt cheap. If someone cracks reliable, affordable 20 K cryocoolers, the map changes.