What Is a Functional Unit in a PCF and Why Choosing the Wrong One Makes Your Carbon Number Incomparable to Every Competitor's

Two manufacturers make the same steel bracket. Same material. Same weight. Same production process. One reports a Product Carbon Footprint of 3.2 kg CO₂e. The other reports 1.1 kg CO₂e. A procurement team assumes the second manufacturer has a dramatically lower-carbon product. They do not. They used a different functional unit.

That scenario is not hypothetical. It is one of the most common sources of misleading PCF comparisons in supply chain carbon reporting, and it happens because the functional unit — the foundational reference measure on which every PCF calculation is built — is treated as an administrative detail rather than the methodological decision it actually is.

ISO 14067:2018 requires that the carbon footprint of a product be expressed in relation to a quantified performance of the product system — for example, per kilogram of material, per litre of beverage, or per square metre of insulation. ISO 14067 places strict conditions on comparability: products can only be meaningfully compared if they use the same functional unit, system boundaries, data assumptions, and methodological choices. Without this alignment, comparisons can be misleading and are not considered compliant with the standard.

This blog explains what a functional unit is, what the standard requires, how it differs from a declared unit, how to choose it correctly for different product types, and the four errors that most commonly produce a PCF that looks legitimate but cannot actually be compared to any other manufacturer's number.

A functional unit is a quantified description of the function or service that a product provides. It is not a description of the product itself — it is a description of what the product does, expressed in measurable terms.

You can think of a PCF as a nutrition label for a product, with only one nutrient: GHG emissions. The serving size on that label is the functional unit — the reference measure to which all inputs and outputs are normalised. The PCF measures how many kilograms of CO₂ equivalent were released per unit of product as an emissions intensity.

The distinction between describing a product and describing its function sounds subtle but has significant practical consequences. Consider a light bulb. The product is a physical object with a specific wattage and physical dimensions. Its function is to produce light for a defined period. A PCF built on the functional unit of one light bulb captures the emissions of manufacturing that object but says nothing about how efficiently it provides light relative to alternatives. A PCF built on 1,000 hours of light output captures the emissions per unit of function delivered — which is what makes comparison between an LED and an incandescent bulb meaningful.

This type of example — providing a defined period of light output — is consistent with the approach ISO 14067:2018 requires for comparing lighting products. LED bulbs can appear carbon-intensive per unit of product due to higher embodied manufacturing carbon, but perform well per 10,000 hours of light output due to their long lifespan compared to alternatives.

Every emission, every material input, every transport leg in the PCF study is calculated and expressed relative to this reference measure. If the reference measure changes — even slightly — the entire PCF number changes with it.

ISO 14067:2018 introduces a distinction that many manufacturers encounter in practice but rarely understand clearly: the difference between a functional unit and a declared unit.

Comparisons based on declared units are permitted only for business-to-business purposes, and only under the strict conditions set out in normative Annex B — including that products belong to the same category, apply equivalent system boundaries, use consistent data quality and allocation assumptions, and address any additional functions explicitly. Any additional functions not accounted for in the comparison must be documented and explained. The full requirements are set out in Annex B of ISO 14067:2018, which is a normative annex and therefore mandatory, not illustrative guidance. The above is a summary only; practitioners must consult Annex B directly for the complete normative requirements.

The comparability requirements are set out in Clause 6 of ISO 14067:2018, which governs goal and scope definition, and in the normative Annex B, which practitioners must follow whenever a comparison between product systems is undertaken. These are not optional reference points — the standard states that if a comparison is undertaken, the requirements in Annex B shall be followed. A partial CFP based on a declared unit does not permit implied claims of overall environmental superiority.

The practical implication: if a customer asks for your PCF to compare it with a competitor's, the comparison is only valid if both PCFs use the same functional unit expressing the same function. If one manufacturer reports per kilogram of steel and another reports per metre of finished structural component, the numbers describe different things. A buyer who directly ranks those two numbers is comparing unlike with unlike — and the manufacturer with the lower number may actually have higher emissions per unit of function delivered.

This distinction matters particularly for manufacturers responding to customer PCF requests that feed into competitive tender decisions. If the customer has not specified a functional unit and different suppliers choose different reference measures, the comparison the customer makes will be invalid regardless of how accurately each supplier calculated its own number.

ISO 14067:2018 requires that the goal and scope of a PCF study include a precise description of the product and its functional unit — the quantified reference to which all inputs and outputs are related.

The standard sets out that a functional unit must satisfy three criteria simultaneously.

For example, instead of measuring a box of detergent, a company would measure the carbon footprint per 100 washes. A concentrated formula and a diluted formula both deliver 100 washes with different volumes of product — only the function-based unit reveals which is more carbon efficient.

The correct functional unit is different for every product category and context. There is no universal answer — but there are principles that apply across categories.

The functional unit must capture that performance difference. Insulation materials are the clearest example. Two insulation products might have the same mass but deliver dramatically different thermal resistance. A PCF expressed per kilogram makes the high-performance product look worse than the low-performance product. A PCF expressed per unit of thermal resistance delivered — per square metre at a defined R-value — correctly reflects which product achieves the insulation goal with fewer total lifecycle emissions.

The declared unit is often the most practical reference measure, because the downstream use is controlled by the customer's application. A steel coil supplied to an automotive stamping plant will be processed in a defined way producing a defined component. The declared unit — per tonne of steel coil at defined specification — is appropriate because the function is implicit in the specification. The critical requirement is that all competing suppliers use the same declared unit for any comparison to be valid, and that the comparison follows the requirements of normative Annex B.

The functional unit must account for lifespan. A heavy-duty industrial component designed for 20 years of service cannot be compared on a per-unit basis with a lighter alternative designed for five years of service. Per year of service, or per total service hours, is the correct basis. Other examples include the number of times a garment is worn, or driving a defined distance over a vehicle's expected service life.

The functional unit must normalise for use intensity. A cleaning product used at varying dilution rates cannot be compared per litre of product — it must be compared per number of cleaning applications achieved.

The functional unit is almost always governed by the applicable Product Category Rule (PCR) under EN 15804 or an equivalent sector standard. Following the PCR is mandatory for EPD purposes and strongly advisable for any PCF that may later be used as an EPD input.

This is the single most common functional unit error in manufacturer PCF reports. Reporting per kilogram or per unit when the competing products have meaningfully different performance characteristics — different lifespans, different efficiencies, different dosage requirements — makes the comparison look clean but produces a number that cannot support any valid claim of relative carbon performance.

A manufacturer of a high-efficiency hydraulic fluid reports its PCF per litre of product. A competitor's lower-efficiency fluid requires more frequent changes and a higher volume to achieve the same number of operational hours. Per litre, the competitor looks equivalent or even better. Per 1,000 operational hours of equipment life — the correct functional unit — the high-efficiency fluid's PCF is significantly lower. The per-litre figure is not wrong as a calculation. It is wrong as a basis for comparison.

The functional unit must be defined at the beginning of the PCF study, as part of the goal and scope definition, before data collection begins. When the functional unit is chosen after the fact — effectively working backwards from available data to find a reference unit that fits — it introduces selection bias. The practitioner, consciously or not, selects a unit that makes the result look favourable rather than one that accurately represents the function the product delivers.

A strong PCF reads like a good research paper: a clear question, a defined scope, transparent data, reproducible calculations, and a fair discussion of limitations. Defining the functional unit is the first of those steps — not an output of the analysis.

Manufacturers often calculate PCFs for multiple products in a range and present them as a comparative set. If the functional unit shifts between product variants within the same comparison — different units for different sizes, formats, or configurations — the comparison is invalid even if each individual calculation is correctly performed.

For example, a packaging manufacturer calculating PCFs for three film sizes cannot use per square metre for the smallest size, per roll for the medium size, and per kilogram for the largest size and then rank them. The ranking is meaningless because the reference measures are different. The correct approach is to define one functional unit that applies consistently across the entire comparison set.

A functional unit that describes a claimed performance outcome — per 10 years of corrosion protection or per 50,000 cycles of mechanical operation — is only valid if that performance claim is backed by documented testing data. If the product's claimed service life is an assumption rather than a verified specification, the functional unit describing that service life is unverifiable, and the PCF built on it cannot be confirmed by a third-party reviewer.

PCF verification evaluates the robustness of data sources, assumptions, system boundaries, emission factors, and calculation methodologies, and reviews primary and secondary data for completeness, transparency, and consistency with defined functional or declared units. A functional unit based on an unverified performance claim will not withstand that review.

The fix is to either use a functional unit based on verified, documented performance specifications, or to use a physical declared unit with explicit documentation that no comparative performance assertion is being made.

An increasing number of large manufacturers and procurement bodies are now specifying the functional unit they require in their PCF requests. This is the correct approach and it resolves the comparison problem at source — all competing suppliers calculate against the same reference, and the resulting numbers are directly comparable.

When a customer specifies the functional unit: use it exactly as stated. Do not adjust it to match your existing PCF methodology. If your current PCF uses a different functional unit, the correct response is to recalculate against the customer's specified unit, not to submit your existing number with a note explaining it was calculated differently.

When a customer does not specify a functional unit: ask before calculating. Specifically ask what the PCF data will be used for — internal carbon inventory, direct comparison with other suppliers, regulatory submission, or EPD input. The answer determines the correct reference measure:

Choosing the wrong functional unit invalidates comparisons. There is no way to retroactively fix a comparison that was built on different reference measures — the calculation must be repeated.

Under ISO 14067:2018, the functional unit must be explicitly documented in the PCF methodology statement. This documentation is not optional and is not satisfied by a brief mention in a footnote.

The methodology statement must specify: the precise definition of the functional unit; the quantified reference value; the reasoning for why that unit was chosen over alternatives; and, for any comparative study, confirmation that the same functional unit was applied consistently across all products being compared.

Under ISO 14067:2018, CFP results are often not sufficient by themselves for direct comparison of products unless consistent system boundaries, data quality, and methodological choices are applied across all compared products — and any such comparison must follow the requirements of Annex B. Importantly, the standard addresses only the climate change impact category and does not address all environmental indicators — meaning a PCF result alone cannot support a claim of overall environmental superiority.

That prohibition on implied superiority is directly connected to the functional unit question. A manufacturer who reports the lowest carbon number in a supplier comparison but achieved it by using a per-unit reference measure while competitors used a per-function measure is implying superiority on a basis that does not withstand methodological scrutiny.

The EU's Empowering Consumers for the Green Transition Directive (Directive (EU) 2024/825) entered into force on 26 March 2024. Member States were required to transpose it into national law by 27 March 2026, and the rules will apply across the EU from 27 September 2026. It bans generic or unsubstantiated environmental claims — including claims based on carbon data that cannot be verified through a defined, transparent, and consistently applied methodology. The functional unit is the foundation of that methodology.

Note on the Green Claims Directive: The separate proposed Green Claims Directive was suspended in June 2025. The European Commission announced its intention to withdraw the proposal, which prompted cancellation of the final scheduled trilogue negotiations. However, the Commission has since clarified that the proposal has not been formally withdrawn. Its legislative status remains unresolved and on hold with no confirmed timeline for revival, and it has not entered into force. It should not be cited as current binding law.

The functional unit is not an administrative box to tick before the real work of a PCF calculation begins. It is the decision that determines what the PCF number actually describes, who it can be compared with, whether it can be used in a regulatory submission, and whether it will survive third-party verification.

Getting it right requires asking, before any data is collected: what function does this product deliver, who will use this PCF number, and what will they compare it against? The answers to those three questions determine the correct functional unit. Everything else in the PCF follows from there.

A carbon number without a clearly defined, correctly chosen, and consistently applied functional unit is a number that describes something — but not necessarily what a customer, regulator, or auditor needs it to describe. In a market where PCF comparisons are increasingly feeding procurement decisions, regulatory submissions, and public carbon claims, that distinction is no longer academic.