In 1827, Georg Ohm wrote down the law that still governs every wire on Earth: voltage equals current times resistance. Two centuries of engineering have not moved it an inch, and nothing ever will. Every electron the grid moves pays the toll that law sets.
There's a common assumption that transformers matter only for long-haul transmission, the hundreds of miles between a power plant and a distant city. Ohm's law says otherwise. Without a step-up transformer, high power can't survive even a short trip.
The math is unforgiving
Ohm's law has a direct consequence for anyone moving power: the energy a wire wastes as heat scales with the square of the current. Power is voltage times current, so for a fixed amount of power delivered, current falls as voltage rises. Put those together and losses fall with the square of voltage. Step voltage up 25x, from the 13.8 kV a generator produces to a 345 kV transmission line, and losses drop 625x.
Run the numbers on a realistic case: 100 MW over a typical overhead conductor.
At 13.8 kV, the voltage coming off a generator, the line loses more than a quarter of the power in the first 5 kilometers. By 19 kilometers, the line consumes everything. One hundred percent. Nothing arrives.
Step up to 69 kV and the same 100 MW travels roughly 24 kilometers before losses reach even 5 percent. At 138 kV, about 95 kilometers. At 345 kV, nearly 600.
This is not a long-distance problem. A power plant sitting next door to a data center still needs a step-up transformer, or the copper between them becomes the most expensive space heater ever built.
Simplified I²R model: loss fraction = P·R·d / V², single equivalent conductor at 0.1 Ω/km, unity power factor, no reactive or corona effects. Directionally accurate; real three-phase line design shifts the constants, not the conclusion.
Why generators are stuck at low voltage
If high voltage is so obviously better, why not just generate at 345 kV and skip the transformer? Because the machine can't do it.
Generator designers are squeezed from both directions. Push the output voltage higher and the copper windings inside the machine need thicker insulation. Insulation takes up space that should hold copper, traps heat, and invites internal arcing. Push voltage lower and current explodes instead: a 500 MW machine at 4 kV would have to move over 70,000 amps through its windings.
The engineering sweet spot lands at 13.8 to 24 kV for nearly every utility-scale machine, whether it burns gas, splits atoms, or catches wind. That's been the answer for decades, and there is no serious path around it.
Call it 20 kV. That's what the machine can make. Now consider what a wire can carry.
The gap nobody can engineer around
Take a 300 MW gas plant serving load 10 kilometers away. At 20 kV, that flow means roughly 8,700 amps. A stout overhead conductor carries about 1,000. Before you've discussed losses at all, you need nine or more parallel circuits just to move the current, each one bleeding energy the whole way. Utility planners never run this analysis, because 20 kV circuits top out in the single-digit megawatts over any real distance. Moving 300 MW at generation voltage isn't expensive. It's not an option.
The physics is unforgiving: hold the loss percentage fixed instead of the wire, and the conductor metal you need scales with the inverse square of voltage. Step from 20 kV to 345 kV, a 17x increase, and the same power travels through roughly 300 times less aluminum. The step-up transformer isn't saving a few percent in losses. It's the difference between one wire and a mine's worth of metal strung across the landscape.
So every plant faces the same gap: the machine physically cannot produce above roughly 24 kV, and power at 24 kV cannot travel more than a few kilometers without evaporating. The generator step-up transformer is the only bridge.
It scales all the way down to 200 meters
The same physics shows up on any campus microgrid, data center, or solar farm. Move 2 MW just 200 meters at 480 volts, the standard voltage for commercial equipment, and you're pushing 2,400 amps through a half dozen parallel runs of the heaviest copper feeder made, and voltage drop still fails your sensitive loads. Step up to 13.8 kV and the current falls to 84 amps on one modest cable, with a second transformer stepping back down at the building. Two transformers, two hundred meters.
This is why hospital campuses and data centers distribute at medium voltage between buildings, and why solar farms scatter step-up transformers across the site: inverters produce a few hundred volts, and that power can't even cross the farm without a lift. There is no scale at which electricity moves well at the voltage it's made or used. From 200 meters to 200 miles, the transformer is the toll every electron pays twice.
Wouldn't microgrids or superconductors fix this?
Two ideas come up whenever this problem is discussed, and both are worth taking seriously. Neither eliminates the transformer.
Microgrids shrink the distance, not the problem. Put generation next to the load and you shorten the wire, but the voltage mismatch stays. Solar panels and batteries produce power at a few hundred volts. Gas turbines produce it at around 13.8 kV. The equipment being served wants 480 volts here, 208 there. Every one of those boundaries is crossed by a transformer, and a campus full of distributed generation needs more of them per megawatt, not fewer: one at each solar inverter, each battery container, each generator, plus the tie back to the utility grid for the hours the local sources can't cover. A microgrid doesn't exit the transformer market. It places a bulk order.
Superconducting cables attack the losses, not the mismatch. A superconductor carries current with zero resistance, which would erase the losses in the chart above. The technology is real: pilot projects have run superconducting links under dense city centers. But the cable has to be held at liquid nitrogen temperatures along its entire length, which means cryogenic cooling plants, vacuum-jacketed pipe, and costs that only make sense for short urban runs of a kilometer or two. And even a superconducting line ends at ordinary equipment: a generator producing 13.8 kV on one end, loads built for 480 volts on the other. Zero resistance moves power without loss. It does not change voltage. Transformers still sit at both ends of the cable.
Every proposed workaround either shortens the wire or goes to extraordinary lengths to dodge Ohm's law, and either way a transformer still stands at every voltage boundary.
The cheapest thing you can't buy
Every megawatt of new generation, every data center, every factory expansion needs transformers on both ends of the wire: step-up at the source, step-down at the load. There is no workaround, no software fix, no clever routing. The physics allows exactly one solution, and it's a machine America has largely forgotten how to build at scale.
On a $500M+ generation facility, the step-up transformer is one of the smallest line items and routinely the longest-lead item in the entire project. Lead times now stretch years, not months. Most of the fleet serving the U.S. grid is imported. Prices keep climbing while demand from AI compute, electrification, and reindustrialization accelerates.
The grid is a machine for moving electrons at high voltage, and transformers are the only doors in and out. When you can't get the doors, it doesn't matter how much power you generate or how much load you're ready to serve.
The takeaway
Chips get faster. Batteries get cheaper. Ohm's law does not improve. It is the same law it was in 1827, and it will be the same law a century from now. The entire grid is a negotiation with that one equation, and the transformer is the only bargaining chip physics accepts.
Transformers are not a component of the grid. They are the precondition for having one. Every conversation about grid capacity, interconnection queues, or data center buildouts is, underneath, a conversation about transformer supply. The machine the whole grid was designed around is the machine we can no longer get.
Think of us as your Chief Transformer Officer.