
Key Takeaways
Calculating CBAM emissions is easy to misunderstand because the final result looks deceptively familiar: tonnes of CO2e per tonne of product. If a company already calculates product carbon footprints, Scope 1 and Scope 2 emissions or other product-level environmental metrics, it can be tempting to assume that one of those numbers can simply be reused for CBAM. That is not necessarily the case. CBAM has its own calculation methodology, system boundaries and rules for attributing emissions to goods, and the methodology for the definitive period changed in several respects from the transitional period.
The European Commission formalised the definitive-period calculation methodology in Commission Implementing Regulation (EU) 2025/2547 and followed it in August 2026 with detailed guidance for non-EU installation operators. The Commission has published a general calculation guide as well as separate guidance for cement, hydrogen, fertilisers, iron and steel, aluminium and electricity. The general methodology establishes the architecture of the calculation, while the sector guidance deals with the production routes and issues particular to each industry.
For most companies, the easiest way to understand CBAM calculation is not to start with the final formula. Start with the factory.
An EU importer may encounter CBAM at the level of an imported shipment, but actual embedded emissions are calculated using information from the installation where the goods were produced. This is an important distinction because the emissions associated with a product cannot usually be calculated reliably from customs information alone.
The installation operator first identifies the relevant production processes and production routes within the installation. Emissions are then attributed to those production processes according to the CBAM methodology. Where relevant precursor materials are used, their embedded emissions are incorporated as well. The resulting emissions are related to the quantity of goods produced to arrive at the specific embedded emissions of those goods.
Conceptually, the calculation looks something like this:
Emissions attributed to the production process + relevant embedded emissions from precursors ÷ quantity of goods produced = specific embedded emissions
The actual methodology is more detailed, and the treatment of different emissions depends on the sector and production route, but this basic structure explains why CBAM is not simply a matter of attaching an emission factor to an imported tonne.
Before calculating emissions, the operator needs to know exactly what is being produced and which CBAM methodology applies. CBAM currently covers selected goods in cement, iron and steel, aluminium, fertilisers, electricity and hydrogen. Within those sectors, different products can be produced through very different processes, and those differences affect the calculation.
Steel is the obvious example. Producing steel through an integrated blast furnace route is fundamentally different from producing it through an electric arc furnace. Aluminium, fertiliser and cement production similarly contain different processes and relevant inputs. A generic company-level emission factor cannot capture these distinctions accurately enough for an actual CBAM calculation.
The Commission’s 2026 sector-specific guidance is therefore the logical place to move once the relevant good has been identified. The general methodology tells you how CBAM calculations work; the sector guidance tells you how that architecture applies to the industrial process you actually operate.
The next task is deciding which emissions and activities belong inside the calculation. This sounds straightforward until you look at an actual factory. A single installation may produce several products, share utilities across production lines, generate intermediate materials internally and purchase other precursor materials from outside the site. Energy and emissions data may be measured for the entire installation even though CBAM needs emissions attributed to particular production processes.
The methodology therefore requires operators to establish appropriate production-process boundaries. The operator needs to understand which physical processes contribute to the CBAM good and how the relevant emissions should be attributed to them. Where a production process creates multiple goods or shares equipment and energy with other processes, allocation rules may become necessary.
This boundary-setting exercise is one of the reasons a CBAM calculation should not begin with a spreadsheet template. Before deciding which cells need numbers, the company needs to understand what those numbers are supposed to represent. A beautifully automated calculation built around the wrong production boundary will still produce the wrong answer.
Once the production process has been established, the operator determines the emissions attributable to it. In practical terms, this means identifying the relevant emission sources and the data used to quantify them. Depending on the installation, this may involve fuels consumed in furnaces or boilers, process emissions from chemical reactions, heating and cooling, electricity consumption and other sources covered by the applicable CBAM methodology.
Some emissions can be determined from direct measurement. Others are calculated from activity data such as fuel consumption multiplied by an appropriate emission factor. Industrial installations frequently use a combination of approaches because not every emission source is continuously measured.
The quality of the calculation therefore depends heavily on the underlying operational data. Fuel invoices, meter readings, production systems, laboratory analyses, material balances and emission factors can all feed into the calculation. The important requirement is that the operator can explain where the data came from, how it was used and why the methodology is appropriate.
CBAM distinguishes between direct and indirect emissions. Direct emissions broadly relate to emissions from the production processes used to manufacture the goods, including relevant emissions associated with heating and cooling. Indirect emissions relate to electricity consumed during production.
However, companies should be careful with the common shortcut that a CBAM calculation simply means adding direct and indirect emissions for every product. The scope of emissions included in the definitive regime depends on the sector. The treatment of indirect emissions has its own rules, and the Commission has continued technical work on how indirect emissions should operate within CBAM. The appropriate sector guidance therefore needs to be consulted rather than applying a generic Scope 1 plus Scope 2 formula.
This is also where familiar carbon-accounting terminology can cause confusion. Direct and indirect emissions under CBAM should be interpreted according to the CBAM methodology, not automatically mapped onto whatever Scope 1 and Scope 2 numbers a company already reports at corporate level. The concepts overlap, but the system boundaries and purpose of the calculations are different.
Precursors are one of the reasons CBAM calculations can become significantly more complicated than measuring emissions inside the final production process. A CBAM good may be manufactured using another material whose embedded emissions must also be included in the calculation. In that situation, the emissions embodied in the precursor effectively travel into the calculation for the downstream good.
Consider a simplified production chain in which material A is produced and then consumed as a relevant precursor in the production of CBAM good B. The emissions associated with producing B are not necessarily limited to the emissions occurring during the final transformation from A to B. Where the methodology identifies A as a relevant precursor, its embedded emissions also contribute to the embedded emissions of B.
This creates a supply-chain data problem inside what initially looks like a factory-level calculation. If the precursor is produced within the same installation, the operator may already control the necessary information. If it is purchased from another installation, obtaining the required emissions information can involve another producer and another underlying calculation. The Commission’s definitive-period methodology also introduced changes concerning precursor calculations and multifunctional production processes, which is another reason transitional-period calculation models should not simply be carried into 2026 unchanged.
After the relevant emissions have been attributed to the production process and precursor emissions have been incorporated where required, the result still needs to be related to production. CBAM does this through the activity level of the production process, essentially connecting the emissions attributed to the process with the quantity of goods produced.
This is what allows the calculation to move from an absolute emissions number to a specific embedded-emissions value, typically expressed in tonnes of CO2e per tonne of product. If a production process has 100,000 tCO2e of relevant embedded emissions associated with 50,000 tonnes of output, the simplified specific embedded-emissions result would be 2 tCO2e per tonne. Real calculations may require additional treatment depending on the production route, precursors and applicable methodology, but the division by activity level is the basic bridge between factory emissions and the carbon intensity of the good.
Production quantities therefore matter just as much as emissions data. A company can have extremely sophisticated carbon measurements and still produce a poor CBAM result if the production volumes used to allocate those emissions are inconsistent, incomplete or mapped to the wrong reporting period.
Suppose an installation produces 20,000 tonnes of a CBAM good during the relevant period. After applying the appropriate CBAM methodology, the operator determines that 24,000 tCO2e of direct emissions are attributable to the relevant production process. Assume for simplicity that the applicable calculation also requires 6,000 tCO2e of embedded emissions associated with relevant precursors and that we are leaving indirect emissions out of this illustrative example.
The calculation would be:
Attributed direct emissions: 24,000 tCO2e
Relevant precursor embedded emissions: 6,000 tCO2e
Total embedded emissions: 30,000 tCO2e
Activity level: 20,000 tonnes of product
The specific embedded emissions would therefore be:
30,000 ÷ 20,000 = 1.5 tCO2e per tonne of product
If an EU importer purchased 1,000 tonnes of those goods, the embedded emissions associated with that quantity would, in this simplified example, be 1,500 tCO2e.
This is deliberately not a sector-specific CBAM calculation. The purpose is to show the flow of information. An actual calculation needs to follow the relevant production-route rules, system boundaries, precursor requirements and treatment of direct and indirect emissions prescribed for the goods concerned.
This distinction deserves particular attention because companies increasingly have existing PCF programmes. A product carbon footprint generally takes a life-cycle perspective and can include emissions from raw-material extraction, upstream production, transport, manufacturing and potentially later life-cycle stages depending on the methodology and boundary being used. CBAM is designed for a different purpose: creating a carbon-price mechanism for certain imported goods that corresponds to the carbon constraint faced by EU production.
As a result, CBAM system boundaries are narrower than a conventional life-cycle product carbon footprint. Not every emission included in a PCF belongs in a CBAM calculation, and the way precursor emissions and production processes are treated follows CBAM’s own rules. An ISO 14067 or TfS product carbon footprint can contain useful underlying information, but the final PCF number should not automatically be copied into a CBAM field.
For companies already collecting supplier PCFs, this does not mean starting from zero. Much of the underlying energy, production and emissions information may overlap. The useful exercise is to map existing data to CBAM requirements and identify what can legitimately be reused, what needs to be recalculated and what information is missing.
The definitive-period methodology places considerable importance on how installation operators monitor the information used to calculate embedded emissions. The Commission’s 2026 guidance includes minimum elements for the Monitoring Plan and the operator’s emissions report, and these documents subsequently support verification.
A monitoring plan should explain the production processes, emission sources, measurement and calculation methods, data sources, allocation approaches and controls used to determine the reported emissions. In other words, it explains how the company knows that the final number is what it says it is. The emissions report then contains the results generated through that monitoring system.
This is why building the calculation only at year-end is a fragile approach. If the underlying fuel, electricity, production and precursor data has not been maintained consistently during the year, a company may find itself reconstructing assumptions months after the relevant production occurred. A continuous monitoring process is usually easier to verify because the calculation retains its connection to the operational records that produced it.
Real factories rarely produce perfect datasets. Meters fail, suppliers do not respond, invoices arrive late, production data is corrected and information may exist at a different level of aggregation from the one required for CBAM. The calculation methodology therefore has to operate in a world where some information is imperfect.
The right response is not to silently fill the gap with the most convenient number. Companies need documented approaches for handling missing data and should understand when conservative assumptions, substitute data or other prescribed methods are permitted under the applicable CBAM rules. The treatment can depend on the nature of the missing information and the calculation methodology involved.
This is also why data-quality controls matter before verification. A useful CBAM workflow should flag missing values, unexpected changes, inconsistent units and unusual emission factors when the data enters the system. Finding a suspicious electricity value while the plant team still remembers what happened that month is much easier than explaining it to a verifier the following year.
For actual emissions to be used in a CBAM declaration, they must ultimately be verified by an accredited CBAM verifier. The verifier reviews the monitoring approach, calculations and supporting evidence at the non-EU installation before issuing the verification report. The Commission’s timeline allows accredited verifiers to register in the CBAM Registry from September 2026, with the first verification reports issued from January 2027.
That changes what “good calculation” means. A mathematically correct result is not enough if nobody can establish where the inputs came from. The company should be able to move from the final specific embedded-emissions value back through the calculation to the activity level, attributed emissions, precursor information, emission factors, meter readings and other source evidence supporting it.
A useful test is whether a knowledgeable person who did not build the original spreadsheet could reproduce the logic. If the answer depends on finding the employee who remembers why cell G42 was manually overwritten in March, the calculation is not yet in particularly good shape.
Calculating embedded emissions is important, but it is not the same thing as calculating the final number of CBAM certificates that must be surrendered. Several additional elements sit between the emissions calculation and the ultimate financial obligation.
The definitive CBAM calculation considers embedded emissions, the applicable free-allocation adjustment and, where relevant, a carbon price effectively paid in the country of production. The CBAM certificate price itself is linked to EU ETS auction prices. In 2026, the Commission publishes quarterly certificate prices based on the weighted average auction clearing price of EU ETS allowances; from 2027, prices move to a weekly calculation.
This distinction matters when companies model CBAM cost. Multiplying a supplier’s embedded emissions by today’s EU ETS price can be a useful approximation, but it is not automatically the company’s final CBAM liability. Embedded emissions are one major input into a broader regulatory calculation.
There is plenty in the CBAM calculation process that software can handle well. Production quantities can be imported from operational systems, energy and fuel data can be mapped to calculation periods, approved emission factors can be maintained centrally, precursor information can be linked to the goods consuming those precursors, and calculation logic can be applied consistently across repeated reporting periods. Automated checks can also flag missing data, unit mismatches and unusual movements before they become verification problems.
The harder parts are the places where the calculation meets the physical factory. Someone still needs to understand whether the production boundary makes sense, whether an allocation method reflects how shared equipment is actually used, whether a precursor has been classified correctly and whether an unusual data point reflects an error or a real change in operations. Those are not data-entry problems. They require knowledge of the production process and the regulatory methodology.
A sensible division of work is therefore to automate repeatable calculation and data-management tasks while escalating methodological decisions and exceptions to people who understand them. The objective should not be to remove humans from CBAM calculation. It should be to stop using human time for arithmetic, copying data between files and repeatedly reconstructing the same calculation.
For a company preparing actual CBAM emissions, the process can be organised into a fairly logical sequence. First identify the CBAM goods and relevant production routes, then establish the installation and production-process boundaries. Map the relevant emission sources, energy inputs and production quantities, identify any relevant precursors and determine where their emissions information will come from. Once the underlying data model is clear, apply the appropriate calculation methodology, calculate the specific embedded emissions and run data-quality checks before preparing the information for verification.
The sequence matters because many calculation problems are actually mapping problems. If a company cannot reliably connect a product to a production route, an energy meter to the processes it serves or a purchased precursor to the goods in which it is consumed, adding a more sophisticated carbon-calculation engine will not solve the underlying problem.
For manufacturers with relatively simple production, much of this may remain manageable using internal systems and controlled spreadsheets. As the number of installations, products, precursors and production routes grows, however, the challenge increasingly shifts from performing an individual calculation to maintaining a reliable calculation system over time.
CBAM calculations are ultimately regulatory calculations, and companies need to perform them correctly for that reason. But once the information exists, limiting its use to the declaration would be a missed opportunity.
A manufacturer calculating embedded emissions across production routes can begin to see which facilities, energy sources, precursors and processes drive the carbon intensity of its products. An importer can compare the embedded emissions of alternative suppliers and understand how those differences affect future CBAM exposure. Procurement teams can begin incorporating carbon cost into sourcing decisions rather than discovering it after the purchase has already been made.
This is where the calculation becomes more interesting than the report. Two suppliers may quote the same price for the same tonne of material while producing it with very different embedded emissions. As CBAM progressively attaches a financial consequence to those emissions, the cheaper tonne at the factory gate may not remain the cheaper tonne when it reaches the European customer.
Getting the calculation right is therefore necessary for compliance. Understanding what the calculation is telling you is where it starts becoming useful.
