Electricity Scarcity Meets Aluminum Tariffs, and American Citizens Pick Up the Bill

---

Support CleanTechnica's work through a Substack subscription or on Stripe.

---

A collision between AI data centers being built—in the current AI bubble and with full throated support by the Trump Administration—and aluminum smelters for electricity is no longer theoretical. Utilities across the United States are facing binding constraints on generation and transmission. When presented with competing requests for hundreds of MW of firm power, they are choosing the customer that can pay more, scale faster, and accept shorter contracts. That customer is almost always a data center. Aluminum smelting, which depends on cheap electricity locked in for decades, loses in that comparison even before policy enters the picture.

Primary aluminum smelting is uniquely exposed to electricity pricing because electricity is not an input at the margin. It is the input. A modern smelter consumes power on the scale of a city and must operate continuously. Interruptions damage equipment and shutdowns destroy economics. To remain viable, smelters typically need long term power contracts at roughly $40 per MWh. AI data centers, by contrast, can tolerate higher prices, shift loads geographically, and monetize power into high margin digital services. That asymmetry is structural and it defines everything that follows.

The decline of US aluminum smelting did not begin with AI, renewables, or EVs. It has been unfolding for decades. In the mid twentieth century the United States was the world’s dominant aluminum producer, anchored by hydroelectric power and protected domestic markets. As globalization accelerated, smelting shifted toward regions with abundant low cost electricity and stable policy frameworks. Canada, Norway, Iceland, and the Gulf states built modern capacity while US plants aged. Rising power costs, environmental compliance requirements, and underinvestment eroded competitiveness long before today’s debates.

By the early 2020s, only four primary smelters remained operational in the United States. The remaining active smelters are in Indiana, northern New York, western Kentucky, and South Carolina, typically co located with legacy power infrastructure and long standing utility agreements that allow them to operate at all. Domestic primary production fell to roughly 700,000 tons per year while total US aluminum demand hovered around 5 million tons. Recycling expanded to fill the gap and now supplies roughly 72% of US aluminum consumption (like steel, the United States is a global leader in primary metal recycling, which is a strong positive). Net imports supply the remaining 14%, mostly from Canada. The system adapted to smelter decline through recycling and imports rather than revival of primary capacity.

In 2025, the administration sharply escalated tariffs on imported aluminum, citing national security, supply chain resilience, and domestic job protection. Section 232 tariffs rose to 25% and then to 50% for most sources. The stated intent was straightforward. Higher prices would support domestic producers and encourage investment in new capacity. Tariffs would level the playing field and bring smelting back.

In practice, tariffs were applied to a system that was already import dependent and constrained by electricity availability. Before the 2025 escalation, the US Midwest duty paid premium typically sat in the $450 to $550 per ton range, reflecting the delivered physical cost of aluminum in the United States above the global benchmark price. Within months of the tariff increases, that premium rose to roughly $1,300 to $1,900 per ton, a two and a half to four times increase over the prior baseline. Over the same period, the base LME aluminum price increased by roughly $200 to $300 per ton, on the order of 8% to 12%. The result was a sharp divergence between global aluminum prices and the delivered prices paid by US buyers, driven primarily by policy rather than underlying metal scarcity.

This price increase did improve cash margins for surviving smelters that already had tolerable power contracts. It did not address the binding constraints that prevented restarts or new builds. Electricity remained too expensive and too uncertain. Capital costs remained high. Policy risk increased rather than declined. Tariffs raised the price signal without changing the underlying feasibility of investment.

The first industries to feel the impact were those that buy large volumes of aluminum and operate on tight margins. Automotive manufacturing sits at the center of that group. Modern vehicles contain increasing amounts of aluminum to reduce weight and meet efficiency targets. EVs contain more aluminum than internal combustion vehicles because battery packs, structures, and thermal systems demand it. When delivered aluminum prices rise by $800 to $1,500 per ton, the added cost per vehicle ranges from $150 to $400 depending on design.

Automakers cannot fully absorb those increases. They respond by delaying model launches, reducing incentives, substituting steel where feasible, and passing costs to consumers. For EVs, where price parity remains fragile, aluminum inflation slows adoption and undermines competitiveness. Tariffs that were meant to support domestic manufacturing end up raising input costs for one of the country’s largest manufacturing sectors.

Grid infrastructure and electrification face similar pressure. Aluminum is embedded in transmission lines, substations, transformers, and renewable energy structures. Materials often account for 20% to 30% of transmission project costs. When aluminum and related metals become more expensive, utilities must either raise rate cases, delay projects, or scale them back. Regulatory approval takes time and cost recovery is partial. The result is slower grid expansion and higher electricity bills over time.

Renewable energy projects absorb these costs as well. Solar frames, wind components, and balance of system equipment all rely on aluminum. While aluminum is not the dominant cost driver for renewables, it is a meaningful one. Rising materials costs compound existing supply chain and permitting challenges. Deployment slows at the margin, especially where power purchase agreements were bid at tight prices.

Packaging and consumer goods also feel the impact, though the per unit effect is smaller. Beverage cans, appliances, and household products incorporate large volumes of aluminum. Higher costs are passed through gradually to consumers. Each item increases by cents rather than dollars, but the effect is broad and persistent.

Substitution provides only partial relief. Steel can replace aluminum in some automotive and construction applications, but weight penalties, corrosion risks, and performance limits constrain that option. Composites and plastics have already displaced aluminum in many uses and face sustainability and recycling challenges of their own. In critical applications such as power transmission, aerospace, and heat exchange, aluminum has no practical substitute. Higher prices do not eliminate demand. They shift costs through the system.

Proponents of tariffs argue that these costs are a temporary price worth paying to rebuild domestic capacity. In theory, higher prices could support restarts of idled smelters and justify new investment. In reality, restart potential is limited, capital intensive, and still dependent on cheap long term power. Many idled plants are obsolete or partially dismantled. New smelters cost several billion dollars and require decades of confidence in power pricing and trade policy.

Capital markets are pricing materially higher risk in 2025, and that increase is directly linked to policy choices made by the Trump Administration. The escalation of trade tariffs, including the sharp increase in aluminum duties, has increased uncertainty around long term market access and price stability, which raises required returns for capital intensive industrial projects. Repeated use of executive authority to change trade rules has shortened investor time horizons and widened credit spreads for sectors exposed to policy reversal risk.

There is also a critical labor distinction that is often glossed over in discussions of smelter viability. The direct workforce inside operating aluminum smelters is heavily unionized, relatively stable, and largely composed of US citizens or permanent residents with long tenure and specialized operational experience. That workforce can keep an existing smelter running once it is operating.

Smelter restarts, major maintenance, retrofits, and new construction, however, rely far more on external and often non unionized labor pools. These include industrial construction crews, refractory specialists, electricians, welders, civil works contractors, and control system installers. In practice, parts of this supporting workforce have historically included undocumented immigrants or workers in mixed status households, particularly in construction, site preparation, and ancillary services. The Trump Administration’s aggressive immigration enforcement posture and expanded deportation actions have tightened these labor markets further, reducing availability, increasing wage pressure, and raising the risk of delays and work stoppages.

This raises project costs and lengthens schedules, which feeds directly into higher execution risk. For capital intensive projects with multi year timelines, this distinction matters. Even if a smelter can staff its unionized operations once running, uncertainty around the availability, legality, and cost of the broader maintenance and construction workforce further undermines investor confidence and pushes the effective cost of capital higher.

Compared with the 2010s, when stable trade policy, lower interest rates, and more predictable labor availability supported effective WACC levels in the 6% to 8% range, new US primary smelting projects in 2025 face effective WACC estimates in the 9% to 12% range or higher once policy volatility and labor constraints are fully priced. These risks compound over multi year build timelines. Even with higher aluminum prices, investors see more attractive risk adjusted returns in assets with shorter development cycles, lighter labor requirements, and clearer policy alignment, including AI data centers that are competing directly for the same electricity.

This combines to mean that restarting idled aluminum plants, something requiring hundreds of millions of dollars, or building new ones, something requiring billions, are very unlikely to reach final investment decision. The cost of capital alone raises the costs of these options by tens to hundreds of millions. This compounds the lack of competitively priced industrial electricity and the challenges with getting the skilled labor necessary for new plants. It’s much more likely that more primary aluminum smelters will close rather than new ones open, a situation directly attributable to the Trump Administration’s policies which were supposed to achieve the opposite.

The collision with AI does not end there. Data centers themselves require massive grid expansion. They need transmission, substations, transformers, and conductors to be built at unprecedented speed. Aluminum is embedded in all of that infrastructure. When tariffs raise the cost of those materials, the cost and timeline of data center electrification increase. The policy meant to assert industrial strength ends up slowing the infrastructure buildout required for digital growth.

Over time, these costs do not remain with producers. They are passed through to American consumers on staggered timelines across multiple sectors. In the automotive sector, higher aluminum input costs translate into vehicle price increases or slower price declines in the range of roughly $100 to $400 per vehicle, typically appearing within 6 to 18 months as model years turn over and incentive budgets tighten.

In beverage packaging, higher aluminum prices raise the cost of cans for soda and beer, adding on the order of a few cents per six pack or case. These increases tend to appear quickly, often within 3 to 9 months, as can contracts reset and packaging costs are passed through to beverage brands and then to retail prices.

In the electricity system, higher aluminum and metals costs raise the capital cost of transmission, substations, and grid equipment. Utilities recover these costs gradually through rate cases, leading to electricity bill increases that are smaller in any given year but persistent, often amounting to a few dollars per household per month over a 3 to 10 year horizon.

In construction, higher aluminum costs feed into windows, cladding, and structural components, contributing to incremental increases in housing costs that surface over multiple years as projects are repriced and delivered. Public budgets absorb part of the shock through subsidies and infrastructure support, which shifts costs into higher taxes or public debt over longer time frames. The incidence is delayed, spread across sectors, and difficult to trace to a single policy, but it is cumulative and unavoidable.

This pattern is regressive in its effects. Electricity and packaged goods costs represent a larger share of household spending for low income Americans, so even modest increases have disproportionate impacts. Higher beverage prices are visible and frequent, while higher vehicle prices delay access to EVs that offer lower operating costs, extending reliance on more expensive fuel and maintenance. Higher electricity rates affect all households as electrification advances and electricity becomes a larger share of total energy spending. What is framed as protection for domestic industry in practice functions as a broad cost transfer to consumers, delivered through higher prices and slower cost reductions, without producing a durable revival of domestic aluminum smelting capacity.

Beyond the current administration, the implications are strategic. Electrified economies are more productive, resilient, and competitive. They rely on abundant electricity, strong grids, and efficient materials supply chains. Policies that raise the cost of electrification inputs while failing to resolve core constraints slow the transition to an electrostate. Even if future leadership shifts course, the lost time and deferred investment compound.

A coherent strategy would need to align power policy, transmission buildout, trade stability, and targeted industrial support. It would need to recognize that electrons, capital, and skilled labor are the real bottlenecks. Blunt price levers applied to one part of the system ripple outward in unintended ways. The experience of 2025 shows that raising aluminum prices does not rebuild smelting, but it does raise costs across the economy.

Electricity will continue to flow to the highest value uses. Markets will keep choosing data centers over smelters when power is scarce. Unless policy addresses physical realities rather than price signals alone, Americans will continue paying more without securing the industrial future those costs were meant to buy.