We sent one coal sample to a lab in April. Can that result stand for the whole year's coal under CBAM?

No. The EU guide says a lab result may only be used for the batch the sample was taken from. A different method is a correlation: a yearly lab test shows how an easy measure, such as net calorific value for specific coal types, tracks carbon content. Ask your verifier how each batch's measure is taken.

The rule

The guide says that for each batch of material or fuel subject to analyses, a representative sample is required. It adds: analysis results shall only be used in calculations in respect of the batch from which the sample was taken.

So one April sample is a figure for the April lot. A later lot from another mine is a different batch.

A different route: a yearly correlation

The guide lets you work out a proxy for carbon content or the emission factor from a quick measure. For coal, the quick measure it names is net calorific value for specific coal types. For oils and gases it names density.

The pre-condition is that you can establish an empirical correlation at least once per year using lab analyses that meet the guide's requirements. In plain words, you test enough samples once a year to show how the quick measure lines up with the carbon content. You then analyse only once a year for the correlation, not for each batch. The guide says this reduces overall monitoring costs.

The guide does not spell out here how each batch's quick measure is obtained. Ask your verifier before you rely on it.

What the guide says about the lab work

• Methods based on corresponding International Organization for Standardization (ISO) standards. Where none exist, suitable European (EN) or national standards.
• The guide says labs shall be accredited under ISO/IEC 17025 (the International Electrotechnical Commission (IEC) joint standard for lab competence) for the relevant methods. Non-accredited labs may be used only where there is evidence that access to accredited labs is technically not feasible or would incur unreasonable costs. A lab counts as sufficiently competent only if it meets the guide's list, for example it is independent of the plant.

Do not mix this up with how often to test

How many analyses per year is a separate rule, with its own minimum table. This page is about which tonnes one result may stand for.

Example: which tonnes a single result covers

• Year's coal (example): 500 tonnes from mine A + 800 tonnes from mine B + 700 tonnes from mine C = 2,000 tonnes
• The April lot was 100 tonnes of mine A coal. The April sample covers 100 of 2,000 tonnes = 5%.
• The other 1,900 tonnes need their own lab results, a yearly correlation, or a fixed value from the ladder.

Example tonnes. A fixed value is one of the options on the guide's ladder of six methods.

What Senseible does here

Senseible reads fuel invoices and electricity bills into dated, fingerprinted records. It does not run lab tests, write a monitoring plan or verify a CBAM figure.

• Standard values, Type I and Type II
• How often coal, gas or limestone must go to a lab
• Supplier carbon figure on a coal invoice
• CBAM readiness for exporters

Sources: European Commission, CBAM Guidance No. 3, methods for calculating embedded emissions (14 August 2026)

Last reviewed 2026-11-07. Facts are from Guidance No. 3, section 4.6.1.4: the correlations paragraph, the requirements for laboratory analyses, and the requirements and recommendations for laboratories, dated 14 August 2026. The plain-words retelling of the correlation is mine. The note that the guide does not say how each batch's quick measure is obtained is my reading of the text, not a statement from the guide. The 5% share was checked again: 100 / 2,000, and 2,000 minus 100 is 1,900. The frequency table has its own live page.