There is, on the face of it, a clear logic behind President Donald Trump’s decision this month to raise the tariff on imported aluminum from 25 percent to 50 percent. He thinks the United States is too dependent on imports and that China is too dominant in the production of this essential metal. In theory, a tariff might spark renewed production in the U.S.

Unfortunately for Trump’s ambitions, the deeper you delve into the weird and wonderful world of aluminum, the more you realize there are physical limits that make a resurgence of U.S. production unlikely.

Raw supply isn’t the problem. There is more aluminum in the Earth’s crust than any other metal. But finding a way of extracting it and turning it into a usable form was something we achieved surprisingly recently. Until then, it was the most valuable metal on the planet.

In the mid-19th century, aluminum was so prized, it was worth more than its weight in gold. Napoleon III liked to impress his guests at banquets by swapping the standard gold plates with aluminum ones. When the Washington Monument was capped in aluminum in 1884, that little capstone was the single biggest piece of aluminum in the world.

What changed, a couple of years later, was the discovery of how to smelt aluminum in vast quantities using a form of electrolysis. The Hall-Héroult process, as it was called after its two inventors, transformed the world forever. All of a sudden, aluminum — far lighter than most other metals, not to mention more resistant to the kind of corrosion that plagues iron — was no longer confined to places like Napoleon III’s banquet hall.

It was thanks to this process that the Wright brothers were able to lift their plane off the ground at Kitty Hawk, North Carolina, in 1903 with an engine made of aluminum. It is thanks to abundant aluminum that we have powered flight today, not to mention the power lines that provide most of the world’s electricity, the bodies of many modern cars and the physical editions of newspapers, which are produced on printing plates made of, yes, aluminum.

Aluminum is still produced more or less the same way today, using an electrolytic reaction that consumes enormous amounts of power.

This metal is almost better thought of as a sort of battery — the product not just of rocks and ores from the ground but of inordinate amounts of electricity.

But it’s not just power that’s needed but continuous power. Unlike many other factories in the industrial world, you can’t easily turn aluminum plants off and on. A stoppage of just a few hours can cause molten metal to freeze, doing permanent damage to the entire production line. During World War II, the Nazis targeted Scotland’s Fort William smelter in the vain hope of causing an interruption that would disable the plant — and British airplane production — for the long haul.

At this stage, you’re perhaps wondering what relevance all this history has for 2025. The short answer is: rather a lot. Why is most of North America’s aluminum smelted in Canada? Because that’s where most of the hydroelectric plants are. Unlike gas-fired power stations, which tend to ramp up and down their output depending on prices, hydro plants usually provide the constant flow of power necessary for an uninterruptible customer such as an aluminum smelter. And the U.S. doesn’t have many obvious locations for new ones.

Now, in theory, the U.S. could still find ways to smelt more aluminum, especially if tariffs stay in place for years. And certainly, part of the reason many developed countries have scaled down their aluminum production is because China has scaled up so much in recent years, reducing the global price and making it ever harder to compete. That is a global problem with global consequences, felt across Europe as well as the Americas.

But there are three important catches to ramping up U.S. production.

The first is that investors are understandably nervous that the president might change his mind about the tariffs at any moment (thus obliterating the business case for a new smelter). The second is that it takes years to build these plants and connect them to the power grid, so even in the event that someone is brave enough to build a new one, American businesses and consumers will still have to fall back on imported aluminum — and more expensive cars, planes and everything else — for some time. The third catch is that it’s not altogether obvious that smelting more aluminum is in the best interests of the United States anyway.

The key issue here is that aluminum production sucks up power that might be useful elsewhere. Thanks to abundant shale oil and gas, the U.S. has plenty of energy at its disposal. Should it crack the nut of building nuclear power plants cheaply in the coming years, that would provide a new source of reliable power that’s well-suited for aluminum smelting.

But the U.S. also faces surging demand for more energy, much of which is coming from strategic industries that Washington also wants to promote in its economic competition with China. It takes a lot of power to run the data centers needed for artificial intelligence, build advanced semiconductors and develop a domestic battery or drone industry.

Diverting precious gigawatts to metal means raising prices for other users, which will inevitably make some desirable projects too expensive to green-light.

Higher tariffs will, in the short run, mean higher prices in the coming months. But they also raise deeper questions. Does the United States want to confront deindustrialization by restarting smelters and doubling down on old, albeit amazing, industrial processes? Or does it want to focus instead on building the technologies of the future?

Ed Conway is a journalist for Sky News and author of “Material World: The Six Raw Materials That Shape Modern Civilization.”