Cradle-to-Gate vs Cradle-to-Grave Boundaries

Boundary choice is the largest single reason two carbon figures for the same product disagree, and it is also the easiest to check. Before comparing any two footprints, establish where each one stops.

The difference is not marginal. For a product that consumes energy in use, a cradle-to-gate figure can represent a small fraction of the lifetime total, and presenting it as the product’s footprint is misleading even when every number in it is correct.

What each boundary covers

The fourth row deserves care. Crediting recovery to a next life is legitimate under an avoided burden convention and is exactly where double counting arises if the next product also claims recycled content benefit.

When is cradle-to-gate the right choice?

When the product is an input to something else and its downstream life is genuinely unknown to the producer. A steel mill does not know whether its coil becomes a car body or a filing cabinet, and inventing a use phase would be less accurate than omitting one.

This is why materials are conventionally reported to the gate and finished goods to the grave. The convention is sensible and becomes a problem only when the two are compared as though they measured the same thing.

When does the use phase dominate?

Wherever the product consumes energy over a long service life. For appliances, heating equipment, vehicles and industrial machinery, operational energy generally exceeds embodied impact by a wide margin.

For products that consume nothing — furniture, textiles, packaging, structural materials — embodied impact is most of the total and a gate figure captures a much larger share of it.

Stating a boundary precisely

The clearest convention comes from EN 15978, which labels life cycle stages A1 to A5, B1 to B7, C1 to C4 and D. Stating which are included removes ambiguity entirely.

A figure described as A1 to A3 is unambiguous; one described as cradle-to-gate is nearly so; one described as the product’s carbon footprint is not a boundary statement at all, and it is the most common form in marketing use.

Where the tricks live

• Comparing a competitor’s grave figure with your own gate figure.
• Excluding the use phase for a product that consumes energy, without saying so.
• Choosing a use phase assumption — hours, temperature, grid mix — that flatters the result.
• Using a short assumed service life, which reduces total use-phase impact.
• Applying end-of-life credit under avoided burden while claiming recycled content elsewhere.

The third and fourth are the subtle ones, because both are disclosed choices that a reader rarely checks. A use phase modelled on a low-carbon grid and a five-year service life can transform a comparison without a single incorrect number.

What boundary should a passport carry?

Whatever the applicable rule requires, stated explicitly, in a structured field rather than in prose. Regulation (EU) 2023/1542 specifies the boundary for battery carbon footprint declarations; the ESPR will do so per product group through delegated acts.

Where a product is an intermediate input, carrying a gate figure with its stage labels lets the downstream manufacturer extend the boundary correctly. That is the mechanism by which footprints compose along a supply chain rather than being recalculated at every step.

How footprints compose

This composition only works if every figure states its boundary and its allocation convention. Where suppliers report inconsistently, the assembler cannot add them, and the usual response is to fall back on generic database averages — which discards the supplier-specific data entirely.

What about transport?

Transport is the stage most often assumed to matter more than it does. For most manufactured goods, moving the finished product to market is a small share of the total, frequently smaller than a single material choice upstream.

The exceptions are heavy, low-value goods moved long distances by road, and anything moved by air. Air freight changes the arithmetic entirely, and a product routinely flown carries a distribution impact that can rival its manufacture.

This matters for boundary selection because a gate figure omits transport by definition. For a locally consumed product that is a minor omission; for an air-freighted one it removes a stage that a buyer would consider decisive.

Who chooses the boundary?

Ideally not the party publishing the result. Where a product category rule or a regulation specifies the boundary, the choice is removed and comparability follows. Where it does not, the publisher chooses, and the incentive runs one way.

That is the strongest argument for product category rules and for regulatory specification of boundaries. Regulation (EU) 2023/1542 does this for batteries, and delegated acts under Regulation (EU) 2024/1781 are expected to do it per product group, which will remove a large source of variation.

What does the use phase actually depend on?

Three assumptions, none of them properties of the product: how long it is used, how intensively, and what the energy supplying it emits. Change any one and the use phase moves substantially without the product changing at all.

Service life is the most consequential and the least evidenced. Manufacturers frequently model a life shorter than their products actually achieve, because a shorter assumed life reduces cumulative use-phase emissions in the study while also matching warranty periods rather than observed field data.

This is where per-item records eventually change practice. A manufacturer holding service histories for products in the field can evidence an actual service life rather than assuming one, which turns the most sensitive input in the study from a judgement into a measurement.

A practical rule

Report to the grave for anything sold to an end user, and to the gate for anything sold as an input, and label the stages in both cases. Then never compare two figures until you have confirmed they share a boundary.

That last habit catches most of the errors in this area, and it takes about ten seconds when the boundary is stated properly and is impossible when it is not.