Holland, Michigan already does this on a few downtown blocks using waste heat from a power plant. But what if every road in a snowy city used embedded electric heating cables — basically scaling up the same tech used in heated driveways? Let's run the numbers on Minneapolis.

The melting physics. To melt snow at a useful rate, you need to overcome three things: warming the snow to 0°C, supplying the latent heat of fusion (334 kJ/kg), and offsetting convective/radiative losses to a -10°C sky. Industry rule of thumb for snow-melting systems is roughly 300 W/m² at the pavement surface during a moderate snowfall (about 25 mm/hr). Heavy storms push that to 500 W/m².

Scaling to a city. Minneapolis has about 1,900 km of paved roads averaging 12 m wide — call it 23 million m² of pavement. At 300 W/m²:

23,000,000 m² × 300 W/m² = 6.9 GW

For reference, the entire state of Minnesota's average electrical demand is around 9 GW. Heating the roads alone during a snowstorm would nearly double statewide consumption. A 12-hour storm dumps 83 GWh of electrical energy into the pavement — about $8 million at $0.10/kWh, for one storm, in one city.

The cable itself. Mineral-insulated heating cable rated for embedment in concrete runs about 30 W/m at 240 V. To deliver 300 W/m², you space cables roughly 100 mm apart, giving 10 m of cable per m² — so Minneapolis needs 230,000 km of heating cable, enough to circle Earth nearly six times. At $15/m installed (and that's optimistic for trenching through existing asphalt), the capital cost is $3.5 billion before substations, transformers, or the grid upgrade.

The grid problem is worse than the bill. A 6.9 GW load that switches on suddenly when snow starts falling is a utility nightmare. You'd need dedicated 138 kV feeders into neighborhood substations sized for a load that runs maybe 200 hours per year — capacity factor under 3%. The transformers alone would cost more than the cables.

Materials fatigue. Asphalt expands ~12 × 10⁻⁶ per °C. Cycling pavement between -15°C and +5°C twenty times per winter induces strains of about 240 microstrain per cycle. Asphalt's fatigue limit is around 70 microstrain for long life — so you'd shred the road's service life from 20 years down to maybe 5. Concrete handles it better but cracks at the cable-pavement interface where thermal mismatch concentrates stress.

The clever alternative. Hydronic systems (warm glycol in PEX tubing) need only 60°C fluid and can use waste heat from data centers, sewers, or industrial processes. Iceland heats Reykjavík's sidewalks with geothermal return water that would otherwise be dumped. Stockholm's Sergels Torg uses district heating return at essentially zero marginal cost. The energy is already flowing — you just intercept it on its way to disposal.

The honest verdict. Electric road heating works beautifully for a bridge deck (where ice = lawsuits) or a hospital ramp. Citywide, you're proposing to burn the energy equivalent of a mid-sized nuclear plant to avoid plowing — when a $300,000 plow truck clears the same area in a few hours using 200 liters of diesel. The economics only flip when the heat is essentially free, which is why every successful large-scale system on Earth runs on waste heat, not resistance wire.