Wood vs Steel Embodied Carbon: How to Compare A1–A3 Fairly

A wood-versus-steel carbon comparison is useful only when both options perform the same job and are measured inside the same boundary. The answer is not an “exact” reduction that applies to every building. It is a project-specific estimate built from material quantities, declared carbon factors and documented assumptions.

That distinction matters. A published building study may report a lower result for mass timber, while another model produces a smaller advantage or a different ordering. Neither finding can be transferred to a new project by copying its headline percentage. Geography, structural layout, material grades, fire and acoustic assemblies, transport, fabrication, data sources and the treatment of biogenic carbon can all change the result.

Start with the question the calculation can answer

For an early material comparison, a practical question is:

“What are the estimated product-stage greenhouse-gas emissions of these two material inventories?”

Product stage means modules A1–A3:

• A1: raw-material supply;

• A2: transport to manufacturing;

• A3: manufacturing.

This boundary does not include transport to site, construction activity, use, maintenance, replacement, demolition, waste processing or benefits and loads beyond the system boundary. A1–A3 is therefore a useful starting point, not a whole-life carbon assessment.

Four conditions for a fair comparison

First, compare equivalent function. A cubic metre of timber and a tonne of steel are not interchangeable units of structural performance. The alternatives should be developed far enough to meet the same stated design brief. Loads, spans, stability, fire resistance, acoustics, durability and applicable codes still require qualified design.

Second, use complete material inventories for the scope being compared. Include the material groups that change with the structural option, not only the visually dominant material. A timber scheme may still contain concrete, reinforcing steel, connectors and protective layers. A steel scheme may use concrete slabs, fire protection and additional finishes.

Third, keep the life-cycle boundary identical. An A1–A3 result cannot be compared directly with an A1–A4 or whole-life result. If a source reports several scenarios, use the same scenario for both alternatives and label it clearly.

Fourth, use compatible data. Environmental product declarations and generic databases can use different product category rules, geography, electricity mixes, production routes, allocation methods and reference years. Record the factor, unit, source, date and declared module coverage for every material.

What one published building comparison found

A 2024 study indexed by the US Forest Service compared the built mass-timber structure of Adohi Hall with a functionally equivalent steel design. Within the study’s A1–A4 boundary, the reported embodied carbon was 198 kgCO₂e/m² for the mass-timber structure and 243 kgCO₂e/m² for the steel alternative. That is a 19% difference for this case and this method.

The same study reported 2,757 tonnes of CO₂e stored in the timber products. The storage figure was presented separately, which is the transparent approach. It should not be silently treated as an immediate, permanent offset against fossil emissions.

This example is evidence that a carefully defined timber design can outperform a matched steel option. It is not evidence that every timber building is 19% lower. The study also includes A4 transport, so its total cannot be reproduced by an A1–A3-only calculator.

Why published answers vary

A 2025 peer-reviewed comparison examined mass-timber and concrete results using five life-cycle assessment databases. The range was wide, and the treatment of biogenic carbon was one of the main reasons results differed. The paper shows why database choice is not a neutral background detail.

The same caution applies to steel. Production route, recycled content accounting, energy supply, product type and regional data can materially affect a factor. Selecting the lowest available declaration for one option and an average factor for the other would not be a balanced comparison.

A good article or calculator result should therefore show assumptions rather than compress them into a single authoritative-looking number.

How to make an A1–A3 estimate

For each material line, the basic calculation is:

Material quantity × A1–A3 carbon factor = estimated A1–A3 emissions

The quantities and factor units must match. For example, a factor expressed per kilogram must be paired with mass, while a declaration per cubic metre must be paired with volume for the declared product.

Add the material-line results to obtain a total for each alternative. Then compare totals only if their scope and function are equivalent:

Difference = baseline total − alternative total

Percentage difference = difference ÷ baseline total × 100

The percentage describes the two entered inventories. It does not become a general property of timber or steel.

What the FrameVerk calculator will and will not do

The FrameVerk Embodied Carbon Calculator is designed to estimate product-stage A1–A3 emissions from a user-entered material inventory. It can help keep quantities, factors and the baseline-versus-alternative comparison visible in one place. Where biogenic storage data is entered or displayed, it should remain separate from fossil and total indicators so the reader can see how the result was formed.

The calculator does not design a structure, decide whether two schemes are functionally equivalent, validate an EPD, predict service life, calculate transport or construction emissions, or certify compliance with a whole-life carbon method. Those decisions require project information and, where applicable, qualified professional review.

Place the calculator after the comparison method, where the reader understands what must be entered. The natural action is: “Compare your A1–A3 material inventory.”

A short quality check before using the result

Before presenting a percentage, ask:

• Do both alternatives satisfy the same defined function?

• Are all materially different components inside the inventory?

• Are both results A1–A3?

• Are factor sources, dates, regions and units recorded?

• Is biogenic carbon shown separately and interpreted cautiously?

• Are transport, construction, use and end-of-life exclusions visible?

If any answer is no, the result is not ready for a confident comparative claim. It may still be a useful working estimate, but it should be labelled as such.

Wood and steel can be compared meaningfully. The credible result is not the most dramatic number; it is the one another reader can trace from quantities to factors, reproduce, and challenge.

Sources

• US Forest Service Research and Development, “Comparison of Embodied Carbon Footprint of a Mass Timber Building Structure with a Steel Equivalent,” 2024: https://research.fs.usda.gov/treesearch/67831

• Duan et al., “Exploring embodied greenhouse gas (GHG) emissions of mass timber construction: A comparative study of life cycle assessment databases,” Resources, Conservation and Recycling, 2025: https://www.sciencedirect.com/science/article/pii/S0921344925003696

• WoodWorks, “Mass Timber Comparative LCA Series,” methodology and case-study collection, accessed 30 July 2026. WoodWorks is an industry-supported education organisation: https://www.woodworks.org/resources/mass-timber-comparative-lca-series/

• American Institute of Architects and Carbon Leadership Forum, “Embodied Carbon Toolkit for Architects,” accessed 30 July 2026: https://carbonleadershipforum.org/download/35250/