The Coming Power Crunch: America Is About to Run Out of Electricity
Everyone’s screaming about inflation, tariffs, and who’s going to control Congress after next year’s midterms.
Fine. Let them argue.
The real bottleneck that will shape the next decade of American growth is hiding in plain sight, and almost nobody on CNBC is talking about it yet.
We are running out of electricity. And we’re running out fast.
For two straight decades, U.S. electricity demand was essentially flat. Total consumption sat at roughly 4,000 terawatt-hours a year and refused to budge. We swapped incandescent bulbs for LEDs, slapped better insulation on houses, and quietly shipped aluminum smelters and chemical plants to Asia. Mission accomplished: same GDP, less juice.
That twenty-year vacation from growth is now officially over.
Three unstoppable forces just showed up at the same time:
- Electrification of everything. Cars, trucks, home heating, industrial heat—the whole economy is switching from molecules to electrons. The EIA thinks EVs alone will add 130 terawatt-hours of new demand by 2030. That’s like plugging in the country of Spain.
- Population and economic growth. The U.S. is adding roughly 2 million people a year, and most of them want air-conditioning, refrigerators, and 800-hp electric pickups.
- Artificial intelligence. This is the elephant that became a herd of blue whales overnight. Data centers already consume about 4–4.5% of all U.S. electricity—roughly 170–180 TWh. Goldman Sachs, the IEA, and every utility CEO with a pulse now say that number triples or quadruples by 2030. Pick your favorite forecast; they all land between 500 and 750 additional terawatt-hours just for servers and GPUs. For context, the entire United Kingdom uses about 300 TWh a year. We’re about to add two or three UKs worth of demand just so ChatGPT can keep getting smarter.
The grid that has to deliver all this was mostly built between the 1950s and 1970s. Half the big transmission lines and power transformers in America are now 40–60 years old. They’re like 1965 Corvettes—beautiful in their day, but try pushing 1,000 horsepower through one on a 100-degree afternoon and watch what happens. Losses skyrocket, voltage sags, and in extreme cases things literally melt.
Renewable projects routinely wait five to seven years in interconnection queues. Hyperscalers can pour concrete and rack servers in 18–24 months, then sit on their hands for another three to five years waiting for the local utility to scrape together enough transformers and transmission capacity. Northern Virginia—now the largest data-center market on Earth—has requests in the queue that would require more power than the entire state currently generates.
This is no longer a future risk. It’s a today problem. In 2024 and 2025 we’ve already seen rolling brownouts in California, emergency pleas in Texas, and utilities from Georgia to Illinois telling big customers “please stop growing for a minute.” Power availability has quietly become the number-one site-selection criterion for any factory or data center worth nine figures or more.
Whenever you get a structural shortage this large—housing in the 2000s, semiconductors in 2021–22, labor today—two things happen: prices go up, and someone makes an absolute fortune solving it.
Electricity is the mother of all structural shortages.
The U.S. will need to generate, move, and store roughly 50–100% more power by the middle of the 2030s than we do today. That is the biggest buildout since Eisenhower sent the interstates across the continent and the Army Corps dammed everything that looked like a river.
A handful of companies are standing dead-center in the flow of capital and steel that is about to be unleashed:
- GE Vernova (GEV) – gas turbines, wind turbines, grid software, high-voltage equipment. They’re becoming the “picks-and-shovels” play for whatever fuel source wins.
- Eaton (ETN) – the bottleneck inside the bottleneck. They make the transformers and switchgear that everyone is fighting over. Lead times that used to be 20–30 weeks are now 150–200 weeks in some cases.
- Quanta Services (PWR) – the largest specialty contractor in North America for transmission and substation work. Their backlog is at all-time highs and still growing double-digits.
- Regulated utilities – the ultimate “boring but beautiful” beneficiaries. They get a guaranteed return on every dollar they spend upgrading the system, and the regulators have no choice but to let them spend.
This isn’t the green trade, the fossil-fuel trade, or the AI trade in isolation. It’s the “America needs twice as much electricity in ten years and has no realistic way to get there yet” trade. Fuel source debates will rage, but electrons are agnostic—they’ll flow through whatever hardware gets built first and fastest.
The United States is about to launch the biggest power-construction boom since the postwar era. Nuclear plants, natural-gas combined-cycle units, solar farms the size of counties, geothermal, maybe even a few fusion pilots—all of it is coming. And before a single electron from any of them reaches your wall socket, someone has to upgrade the wires.
The only real questions left are how fast we move and who gets the contracts.
I’ll be breaking this down piece by piece in the coming weeks: the nuclear renaissance, the natural-gas bridge, the grid-infrastructure bonanza, and the handful of publicly traded companies best positioned to profit from each.
Stay tuned. The lights aren’t going out—but the margins are about to light up.
The opinions voiced in this material are for general information only and are not intended to provide specific advice or recommendations for any individual.
All investing involves risk including loss of principal.
The economic forecasts set forth in this material may not develop as predicted and there can be no guarantee that strategies promoted will be successful.