CBAM’s aluminium rules count carbon from anode consumption and rare process upsets, but not the electricity that actually dominates aluminium’s carbon footprint — and that one design choice reshapes the competitive map.

In this article:

• Which aluminium products CBAM covers (CN 7601 through 7616, with key exclusions)
• The three production routes — pre-baked anode, Søderberg, secondary/recycled — and how their emissions profiles differ
• The most counterintuitive rule in CBAM’s aluminium guidance: purchased electricity is excluded from the calculation
• Why PFC “anode effect” emissions deserve more monitoring attention than most producers give them
• Why alumina and bauxite sit outside CBAM’s boundary, and why that structurally favors recycled aluminium

If you produce, import, or trade aluminium into the EU, you’ve probably assumed CBAM will hit you hardest wherever your electricity is dirtiest. That’s reasonable — electricity is, by far, the biggest real-world driver of primary aluminium’s carbon footprint. It’s also wrong, as far as CBAM’s formula is concerned, and getting this right matters for how you budget and where you invest in decarbonization.

I run Sprih, where we build sustainability and compliance data infrastructure for industrial companies, and we’re building dedicated tooling for aluminium’s CBAM math because this sector has more nuance than almost any other CBAM category. Here’s what’s covered, how the three routes differ, and the one rule that should reshape how you think about your exposure.

Quick refresher: CBAM requires EU importers of certain carbon-intensive goods to buy certificates reflecting the carbon emitted in producing them, mirroring what EU manufacturers already pay under the EU Emissions Trading System. The reporting-only transitional period ended in 2025; since 1 January 2026, importers must calculate embedded emissions and surrender certificates. Aluminium is one of six covered sectors, alongside cement, iron and steel, fertilisers, hydrogen, and electricity.

Which Aluminium Products Are Covered

CBAM splits aluminium into two goods categories by CN (Combined Nomenclature) customs code.

Unwrought aluminium (CN 7601) — ingots, blocks, billets, and slabs, from primary or secondary production — is the foundational good. It’s a “simple good”: its inputs (alumina, anodes, scrap) count as raw materials with zero embedded emissions, so its footprint comes entirely from what happens inside the smelter or furnace.

Aluminium products (CN 7603–7616) cover essentially everything made from unwrought aluminium: bars, wire, sheet, foil, tube, structural parts, containers, cable, and other finished articles. These are “complex goods” — they carry forward the precursor’s embedded emissions plus whatever the forming process adds.

Quantities in the 7603–7616 range are reported in tonnes per your specific CN code, with no separate adjustment for non-aluminium components like coatings — classify carefully.

Three Production Routes, Three Different Emissions Profiles

CBAM’s methodology recognizes three ways aluminium gets made, and the route matters enormously for your emissions number.

Pre-baked anode smelting, the dominant primary technology, reduces alumina by electrolysis using individually cast carbon anode blocks replaced on a rotating schedule. Søderberg smelting, an older primary technology, uses a single continuous, self-baking anode fed with “green” anode paste. Secondary (recycled) aluminium starts from scrap: sorted by alloy type, de-coated and de-oiled, then re-melted in rotary, reverberatory, or induction furnaces before alloying and casting.

The practical difference is stark. Primary smelting’s direct emissions include CO₂ from anode/electrode consumption (the largest source), fuel combustion, flue-gas treatment, and — unique to electrolysis — PFC emissions from anode effects. Secondary aluminium carries none of the anode-consumption or PFC line items; its direct emissions come only from the fuel burned to melt scrap. In CBAM’s own worked example, anode and electrode consumption alone accounted for roughly 81% of a primary smelter’s direct emissions, with PFCs contributing another 8% — figures secondary simply doesn’t carry.

Primary smelting carries anode consumption and PFC emissions that secondary recycling simply doesn’t have.

The Rule Everyone Gets Wrong: CBAM Doesn’t Count Electricity for Aluminium

Here’s the part that surprises almost everyone I talk to, including people who’ve tracked CBAM closely in other sectors.

Aluminium goods sit on Annex II of the CBAM Regulation — the list of goods for which only direct emissions are counted. Indirect emissions, meaning the footprint of purchased electricity, are explicitly excluded from the embedded-emissions figure that determines your certificate obligation. This wasn’t always true: during the transitional period, both direct and indirect emissions were tracked and disclosed for aluminium. In the definitive period, only direct emissions flow into what an importer actually owes.

This matters because electrolytic smelting is one of the most electricity-hungry industrial processes anywhere — it’s why primary aluminium made on a coal-heavy grid has such a larger real-world footprint than metal made on hydro or gas power. Physically, electricity is the dominant input. But under CBAM’s current formula, that electricity-driven footprint isn’t what generates your certificate liability. Electricity still appears in every system-boundary diagram as an input — it just isn’t a counted emissions source; operators can track it voluntarily for internal purposes, but it doesn’t feed the CBAM number.

What does generate liability is direct process emissions: anode and electrode consumption, fuel combustion, flue-gas treatment, and PFCs from anode effects — none of which proxy for grid cleanliness. They’re driven far more by smelting technology and process control than by electricity source.

The single most important, non-obvious fact in this sector: CBAM’s aluminium formula counts direct process emissions, not the electricity that physically drives most of the footprint.

In practice, a smelter running well-controlled equipment on a carbon-intensive grid isn’t automatically penalized by CBAM’s math the way its true footprint might suggest. It IS penalized, sometimes heavily, for sloppy process control that produces frequent anode effects, regardless of where its power comes from. That’s a different risk profile than most people expect — and a live policy question, since aluminium’s direct-only scope is debated and could change in future revisions.

Why PFC Monitoring Deserves More Attention

PFC emissions come from “anode effects” — brief upsets where alumina concentration in the bath drops too low and the bath itself starts to electrolyze, releasing CF₄ and C₂F₆. These aren’t a footnote: CF₄ is roughly 6,630 times as potent as CO₂, and C₂F₆ roughly 11,100 times, under CBAM’s conversion factors — so small leakage translates into large CO₂-equivalent figures.

CBAM offers two calculation methods — a “slope method” based on anode-effect minutes per cell-day, and an “overvoltage method” based on voltage disturbance — with technology-specific emission factors as the floor. Those vary widely: modern point-fed pre-bake cells sit at the low end, older side-worked pre-bake and horizontal-stud Søderberg cells sit meaningfully higher. The guidance recommends moving to installation-specific factors from field measurement, remeasured every three years or after major process changes, within roughly ±15% uncertainty.

Because PFCs are direct emissions with outsized GWP multipliers, precise anode-effect monitoring is one of the highest-leverage things a primary smelter can do to get its CBAM number right — and to avoid a default-value fallback that likely overstates real performance.

Alumina, Bauxite, and Scrap: Inside vs. Outside the Boundary

Two structural quirks matter before you build any internal cost model.

Alumina and bauxite sit entirely outside the CBAM boundary. Bauxite mining and Bayer-process alumina refining are both energy- and emissions-intensive, but neither is a CBAM good, and neither is tracked anywhere in the chain. CBAM’s boundary for primary aluminium starts at the smelter gate, not the mine or refinery. Anode production is treated the same way — even made on-site, anodes count as a zero-emissions raw material, not a tracked precursor.

Scrap is explicitly excluded (CN 7602 00) and, wherever used as an input, counts as a zero-embedded-emissions raw material. That’s a real advantage for secondary producers: the recycling route skips anode consumption and PFCs entirely, and the scrap feeding it carries no carbon burden into the calculation — unlike a metal precursor that arrives with its own embedded-emissions history.

CBAM’s aluminium value chain: the boundary starts at the smelter or furnace, not at the mine or refinery.

Together, this means secondary aluminium tends to show a structurally lower CBAM number regardless of where it’s produced — the advantage comes from the route, not the geography. Where a downstream product’s precursor could be either primary or secondary, CBAM requires them to be tracked separately, since they carry different embedded-emissions values — if you buy sheet, wire, or extrusions, ask your supplier which route their feedstock came from.

This also has real implications for major primary-aluminium exporting regions like the Gulf, where cheap electricity, not carbon intensity, has historically driven competitiveness. We cover that country-specific picture in our Middle East regional post.

Practical Guidance

• Primary smelters: invest in installation-specific, field-measured PFC factors instead of resting on the generic technology floor — the gap can be material.
• Buyers weighing primary vs. secondary: CBAM currently rewards recycling structurally, independent of where either producer sits.
• Downstream fabricators: track whether your precursor is primary or secondary, and get real supplier data instead of defaulting.
• Importers: don’t treat grid carbon intensity as a proxy for CBAM exposure in aluminium — ask for actual process data.

How Sprih Helps

Aluminium’s CBAM rules reward precision in exactly the places most producers aren’t used to measuring closely: anode-effect frequency and duration, technology-specific PFC factors, and whether your unwrought aluminium precursor was primary or secondary. Get those wrong and you either overpay by falling back on conservative default values, or file numbers you can’t defend under verification.

We’re building a dedicated CBAM Aluminium Studio inside Sprih’s SustainSense platform to handle exactly this: calculating PFC emissions under both the slope and overvoltage methods, tracking primary-versus-secondary precursor mass through downstream products, and keeping the documentation trail verification-ready. SustainSense indexes sustainability and compliance data across 120,000+ companies and 400,000+ reports, giving your team a defensible, audit-ready basis for the specific numbers CBAM asks for in this sector, rather than reconstructing them from scratch every cycle.

Sprih is an AI-native sustainability intelligence company helping enterprises turn climate and compliance data into competitive advantage.