Wittbecker on Ali: Aluminum power blues persist

In January, Ukraine did not renew its transmission contract with Russia’s Gazprom.

Ukraine had been earning over $1 billion/year in transmission fees from the Russians for transporting Russian gas into Austria, Hungary, Moldova and Slovakia. Ukraine transshipments of gas for the Russians represented about 5% of European gas supply. The EU has done an excellent job replacing Russian gas with liquid natural gas from Norway and Qatar, amongst other origins.

So, you ask, “What does this impact aluminum power prices?”

Natural gas prices in the EU topped 50 euros per megawatt hour (mWh) on Jan. 1 when the agreement ended. They have now eased back to around 46 euros/mWh but remain elevated year on year. Natural gas prices in January 2024 were around 30 euros/mWh.

The higher gas prices reflect in day-ahead power prices across the key aluminum producing countries of Western Europe:

You will recall that Western European suffered a host of primary aluminum curtailments in the aftermath of the Ukraine war and Russia suspending natural gas exports to Northern Europe. Western European production peaked at 3.9 million tons in 2019 and will be around 3 million tons during 2025. High gas prices and corresponding day-ahead power problems were the root cause.

This problem is demonstrated by Aluminium Dunkerque, the largest industrial electricity consumer in France. On Jan. 3, the head of the company expressed his “need for long-term visibility on electricity prices that risk threatening the competitiveness of French industry.”

This year is crucial to Dunkerque, who is worried about the end of the Arenh system on Dec. 31. This system specific to France, allowed a cut-price tariff for EDF nuclear electricity for big consumers, i.e., 42 euros per megawatt hour, on part of their consumption. If the Arenh system is not replaced with something similar going forward, Dunkerque must go into the open market and face the “buzzsaw” of competition from datacenters.

Data centers: aluminum’s biggest competition for power

The aluminum industry worldwide has fought the problem of rapidly rising retail consumer demand for years. Regions that used to have “stranded power” are shrinking. “Stranded power” were situations where a large block of generation capacity existed with limited retail consumer demand to supply base load demand. This was the case in places like Brazil, Ghana and even Canada. Aluminum smelters represented the perfect solution to those assets. The smelter took a constant, large base load of electricity 24/7/365. Fast forward to 2025 and those situations are rare. That generation capacity has more demand than it can generate thanks to the population of many countries rising and retail power demand rising.

Now, the new threats are the data centers and AI. This demand seems insatiable and inelastic when it comes to the price tag of the power. According to Goldman Sachs, demand from data centers will grow 160% by 2030. Data centers will consume 8% of all US power consumption by 2030. Europe will experience similar growth, bolstered by unique EU decarbonization mandates pushing EV and broad electrification schemes. Electricity could represent 50% of the EU’s primary energy profile by 2030, compared to 20% in 2025.

What this means for aluminum power pricing: The Microsoft Precedent

A frightening example of what data center trend could mean is the recent example of the Microsoft- Constellation Energy deal in the US.

The parties agreed to a 20-year deal on power to facilitate the restart of the Three Mile Island Unit 1 in Pennsylvania. Constellation will invest $1.6 billion to reopen the unit and market estimates are that Microsoft will pay $100 per mWh for this power.

This price makes primary aluminum production untenable in the US. That price is at least 2x what most people would say is the practical limit what a world class primary smelter can tolerate at current London Metal Exchange prices.

What can aluminum (or steel) companies do in this environment?

Development of Behind the Meter (BTM) generating capacity is being looked at. The Chinese did this successfully for awhile in the past two decades, building in-situ, coal-fired capacity before Beijing decided to force them out of the energy business under the guise of decarbonization.

Companies such as Nucor are investing in small scale nuclear fusion reactor startups. Rio Tinto is doing due diligence in Finland, where substantial geothermal capacity is on the cusp of development. Both technologies offer the appeal of 24/7 base load. Wind/solar is still interesting as BTM solutions, but would require other base load energy to compensate for the lack of battery storage today.

One thing is clear. Baring a breakthrough in process technology that really slashes production costs OR (and this is a really big lift), we see a quantum leap in aluminum prices, it is hard to see anyone building aluminum capacity in Europe or North America given the trends in power cost.
The powers blues are hard to shake — and there is no solution imminent.

Canada should join the EU …the tariff could effectively put them there!

An article in The Economist Jan. 2, argued that Canada should join the EU. The logic was that the EU needs resources and Canada needs people. In other words, EU provides viable markets for its resources.

A move by the Trump Administration to tax Canadian imports at 25% would effectively push Canada into the EU aluminum market basis pure economic netbacks.

A 25% duty on aluminum levied on just the underlying LME cash price today (~$2,500 per metric ton) = $625 per metric ton or $0.28 per pound. Unless the US Midwest aluminum ingot premium expressed over LME cash would rise by that amount, Canada will look for alternatives.

Here is the math basis recent physical ingot premiums in the EU:

The current Midwest premium as quoted on the CME for calendar 2025 ranges from $0.2497 (January) to $0.2680 (December). This clearly does NOT reflect an assumption of the 25% duty coming into effect.

For Canada to justify maintaining its current level of exports to the US while incurring the 25% duty, Midwest premiums need to rise about $0.19 per pound. Otherwise, Canada will divert its production to the EU where it would be welcomed with open arms. Canadian metal is low-carbon (e.g., 2 metric tons of CO2 per ton of production) and would find a ready market in the EU, which stands on the verge of implementing its carbon border adjustment mechanism (CBAM) in January 2026.

Why is the market so nonchalant about the potential impact of the tariff?

The CME and the Midwest have been heating up over the past two weeks, moving from ~$0.20 per pound to the current level above. However, it’s well short of the threshold required to recoup the duty import.

When speaking with a respected physical trader recently, the comment was, “It is an extremely high-risk-versus-reward trade to add to a premium long at even these levels. Demand is still not that great and you are trying to handicap Trump’s mindset day by day. That is a dangerous play.”

What we are seeing are traders quietly positioning their physical premium longs within the US and not risking the effects of the duty on fresh imports. This is taking the form of metal placed into public warehouse, custom-cleared or placed into consignment deals at consumer mills. Either approach ensures that metal is not liable for the higher duties from either a universal tariff or one targeted against Canada.

Inflation with a capital ‘I’ when it comes to Canada

This uncertainty over the imposition of duties against Canada comes at incredible potential cost to the US economy.

Beyond the aluminum costs mentioned above, Canada’s total exports to the US are about USD $412 billion A 25% “haircut” for Canada would = USD $103 billion that would have to reckoned with. Some goods, like aluminum, could simply move elsewhere. However, some of the product flow may be inelastic and the burden of those duties would ripple through the supply chain to the ultimate end consumer.

Stay tuned for the next chapter.