Product Environmental Footprint (PEF)

Two life cycle assessments of similar products routinely produce results that cannot be compared, not because either is wrong but because they made different reasonable choices about boundary, allocation and data.

Product Environmental Footprint is the European Commission’s attempt to remove that freedom where comparison matters.

Why comparability needed fixing

ISO 14040 and 14044 define how to conduct a life cycle assessment competently. They deliberately leave many decisions to the practitioner, because the right choice depends on the study’s purpose.

That flexibility is appropriate for internal decision-making and fatal for public comparison. A practitioner can make defensible choices that produce a favourable result, and a competitor’s equally defensible choices produce a different one.

Sixteen impact categories

PEF assesses sixteen categories rather than reporting carbon alone, which is one of its more consequential design decisions.

Reporting sixteen categories exposes trade-offs that a carbon-only figure hides. A material change that reduces global warming potential while raising water use and ecotoxicity is a trade-off, and PEF makes it visible rather than allowing a favourable single number to be selected.

Category rules do the real work

The method alone would not produce comparability. What produces it is a Product Environmental Footprint Category Rule, developed per product group, fixing the choices that would otherwise vary.

• Functional unit — what the assessment is per, which is frequently a service rather than a physical unit.
• System boundary — which lifecycle stages are included and how end of life is modelled.
• Data quality requirements — how much must be primary data rather than database averages.
• Allocation rules — how impacts divide where a process produces several outputs.
• Benchmarks — a representative product for the category, against which results are positioned.

The functional unit is the choice with the largest effect and the least intuitive definition. For a washing machine it is not the appliance but a number of wash cycles at a stated load, which correctly places a durable machine against a fragile one.

The single score argument

PEF permits weighting the sixteen categories into one aggregated score, and this is the most contested part of the framework.

Both positions are reasonable. A consumer cannot act on sixteen numbers, and a single score embeds a judgement about how much water use is worth relative to climate impact that no measurement can settle. The framework’s answer is to make the weighting explicit and standard rather than leaving each company to choose its own.

Where PEF connects to the passport

PEF and the digital product passport were developed separately and converge naturally, because both need per-product environmental data expressed to a fixed method.

Where a delegated act under Regulation (EU) 2024/1781 requires environmental performance for a product group, a category rule is the obvious mechanism for defining how it is calculated. The passport then becomes the vehicle for publishing the result, with its method reference attached.

That connection also matters for green claims. Directive (EU) 2024/825 constrains unsubstantiated environmental claims, and a figure calculated to a published category rule is substantiated in a way a self-defined calculation is not.

What it costs to do

PEF is more demanding than a carbon footprint, and the difficulty is data rather than calculation.

Sixteen categories require inventory data on emissions, water, land and resource flows that a carbon-focused exercise never collected. Category rules also set primary data requirements, so database averages that satisfy a carbon calculation may not satisfy a PEF study for the same product.

The practical consequence is that a PEF study is a supply chain data exercise before it is an assessment exercise, which puts it on the same critical path as the passport programme itself and argues for doing them together rather than sequentially.

Sequencing them separately is the common and expensive mistake. A team that runs a supplier data collection round for the passport, then repeats it a year later for a footprint study, has asked the same suppliers the same kind of question twice and spent the goodwill that makes the second round work.