Embodied Carbon in Construction Materials

For most of the last thirty years, reducing a building’s carbon meant reducing the energy used to run it. That work succeeded well enough that the arithmetic changed: in a highly efficient new building, the emissions from making it can exceed those from operating it for decades.

Embodied carbon is also spent before anyone moves in. Operational emissions can be reduced later by a retrofit or a cleaner grid; embodied emissions are settled at construction and cannot be revisited.

What counts as embodied carbon?

The definition is broad, which is exactly why claims need to state their boundary. Two figures for the same building can differ by a factor of two purely from which life cycle stages were counted.

Upfront carbon, the term used most often in policy discussion, generally means A1 to A5: everything emitted before the building is handed over. It is the figure most amenable to design decisions and the one most member state rules target.

Where does it actually sit in a building?

Overwhelmingly in the structure and substructure. Concrete, steel and the foundations they sit on typically dominate, with facades, finishes and services making smaller contributions that vary more between building types.

The practical consequence is that structural decisions taken early determine most of the outcome. A material substitution made at concept stage moves the number; a specification change made at fit-out does not.

Why generic data is the problem

Most embodied carbon assessments use average figures per material from national or industry databases. Those averages are useful for benchmarking and nearly useless for reduction, because they cannot distinguish between two suppliers of the same material.

Steel is the clearest illustration: electric arc furnace and blast furnace routes differ by roughly a factor of four, and an average conceals that entirely. A designer using averages sees no benefit from specifying the better supplier, so there is no commercial reward for being better.

Where does supplier-specific data come from?

From Environmental Product Declarations, which are third-party verified documents produced to a product category rule. They are the established route and they have two practical weaknesses.

They are documents rather than data, so using a hundred of them in a model means transcription. And they are produced per product per manufacturer at real cost, so smaller producers frequently have none, which biases specification toward larger suppliers regardless of actual performance.

How the passport changes this

A construction product passport carrying a declared footprint as structured data removes the transcription step and makes supplier comparison mechanical rather than manual. That is the difference between an assessment done once for compliance and an assessment used during design.

It also carries the boundary and method alongside the number, which is what makes two figures comparable at all. A footprint without its stage coverage stated is not a quantity that can be added to another one.

What is being regulated?

Several member states have introduced or announced embodied carbon limits for new buildings, ahead of any EU-wide requirement. The Construction Products Regulation recast, Regulation (EU) 2024/3110, brings environmental declarations into the harmonised framework, and the ESPR provides the passport mechanism.

The direction is consistent even though the instruments are separate: declared, comparable, product-level environmental data becoming mandatory rather than voluntary.

What can actually reduce it?

1. Build less: reuse an existing structure rather than replacing it.
2. Use less material: efficient structural design beats material substitution in most cases.
3. Choose lower-carbon suppliers of the same material, which requires supplier-specific data.
4. Substitute materials where structurally appropriate, with honest accounting of the trade-offs.
5. Design for future deconstruction so the embodied carbon is recoverable rather than lost.

The order matters and is frequently inverted in practice. Substitution attracts the most attention and sits fourth, because the largest reductions come from not building and from using less, both of which are structural decisions rather than procurement ones.

What about reuse?

Reused structural elements carry almost no new embodied carbon, which makes reuse the single most effective intervention available. The barrier is rarely the material and almost always the evidence.

A steel section removed from an existing building cannot be specified without knowing its grade, properties and history, and that information was usually never recorded in a retrievable form. This is precisely the gap a material passport addresses, as covered in material passports for buildings.