Scope 3 Emissions and Product Data

A manufacturer that has worked seriously on its own operations usually finds those operations account for a small share of its total emissions. The rest sits in the value chain, in what the GHG Protocol calls Scope 3.

That is where the difficulty concentrates, because the emissions belong to activities the reporting company does not control and frequently cannot observe.

What Scope 3 covers

The GHG Protocol Corporate Value Chain Standard divides Scope 3 into fifteen categories, eight upstream and seven downstream. For most manufacturers two or three dominate and the remainder are immaterial.

Category 1 is where the passport is directly relevant. Categories 11 and 12 matter enormously for energy-using and durable products, and both depend on assumptions about behaviour that a passport can improve but not resolve.

Why spend-based estimation is the default

The most common method for Category 1 multiplies procurement spend by an emission factor per unit of currency in a given sector. It is defensible, it is fast, and it can be produced entirely from accounting data.

It also has a property that should trouble anybody using it to manage rather than to report.

Average-data methods using mass rather than spend are an improvement, since they at least respond to material substitution. They still cannot distinguish between two suppliers of the same material with very different production processes.

What supplier-specific data requires

Moving to supplier-specific figures means obtaining a product carbon footprint for purchased items, calculated to a consistent method with a stated boundary.

• A stated system boundary — cradle to gate is the usual choice, and the figure is meaningless without it.
• A consistent allocation method where a facility produces several products from shared inputs.
• Primary energy data for the supplier’s own operations rather than a sector average.
• Its own upstream treatment, which is where the supplier faces the same problem one tier further back.
• A reference period, because a figure from four years ago describes a plant that may have changed.

The fourth point is the recursive difficulty. Every supplier asked for a product footprint faces the same Scope 3 problem for its own inputs, which is why supplier-specific data thins rapidly as you move upstream.

The number usually goes up first

A finding that surprises programme sponsors is that replacing estimates with measurements frequently increases the reported footprint rather than decreasing it.

Sector average factors are averages, and a company using materials with above-average intensity has been under-reporting without knowing it. Discovering this is the point of the exercise, and it is worth setting the expectation before the first supplier-specific figures arrive.

The alternative framing is more useful: the first accurate number is a baseline, not a result. Comparing it to the previous estimate compares two different methods rather than two years of performance.

Where the passport fits

A product passport carrying a footprint figure with its boundary and method attached is exactly the input a Category 1 calculation needs, arriving through the product rather than through a separate survey.

This removes the annual survey cycle, in which a sustainability team emails several hundred suppliers, receives a minority of responses in inconsistent formats, and extrapolates the rest.

What to do with a mixed dataset

No inventory will be entirely supplier-specific for years, so the practical question is how to handle a dataset where some inputs are measured and most are estimated.

Report the mix explicitly. Stating what proportion of Category 1 by spend or by mass rests on supplier-specific data is more informative than the total, and it is the figure that shows progress year on year while the total is still moving for methodological reasons.

This also aligns with what CSRD reporting under the ESRS expects, since data quality and methodology disclosure sit alongside the quantitative result rather than behind it.