Boundaries, Stages & the Functional Unit

To count a building's carbon consistently, the industry breaks its whole life into standard stages, and knowing them by name is basic carbon fluency. Under the European building standard EN 15978 - the method most whole-life work follows - the life is divided into lettered modules. The product stage (A1-A3) covers raw material supply, transport to the factory and manufacture; this is the core of a material's embodied carbon and what most EPDs report. The construction stage (A4-A5) adds transport to site and the construction/installation process, including site energy and material wastage. Together, A1-A5 make up the up-front carbon - the emissions spent before anyone occupies the building, the irreversible burst this course keeps returning to.

The use stage (B1-B7) then spans the decades of occupation. It splits into the embodied side - B1 (in-use emissions from the fabric, such as refrigerant leaks), B2 maintenance, B3 repair, B4 replacement of components as they wear out, B5 refurbishment - and the operational side - B6 operational energy use and B7 operational water use. B4 replacement matters enormously for interiors and finishes, which are swapped many times over a structure's life, so a short-lived material's carbon is effectively multiplied. The end-of-life stage (C1-C4) covers demolition, transport, waste processing and disposal. Finally, module D sits outside the system boundary as a separate figure: the potential benefits (or loads) beyond the life - recycling, reuse or energy recovery that displaces future production. D is reported separately precisely so it is not netted off to flatter a headline number.

Why insist on this lettered scaffolding? Because it lets everyone count the same things in the same slots, so that assessments and EPDs can be added up and compared module by module. When you see "A1-A3" on a product declaration, you know exactly what it includes and, crucially, what it leaves out (transport to your site, installation, replacement, disposal). The modules are the shared language of building carbon - and most misleading comparisons come from people quoting different modules as if they were the same total.

The system boundary is the decision about which of those stages you actually include - where you start counting and where you stop - and it is the biggest single reason two honest carbon numbers can differ. A handful of standard boundaries recur, and you must be able to name them. Cradle-to-gate (A1-A3) counts from raw-material extraction to the factory gate - the material as made, before it leaves the plant. It is the most common boundary for product EPDs because it is what a manufacturer can control and declare, but it deliberately excludes everything after the gate: getting to your site, installing it, maintaining and replacing it, and disposing of it.

Cradle-to-gate-with-options adds selected later modules (often A4-A5, sometimes end of life). Cradle-to-grave counts the whole life, A1 through C4 - extraction to final disposal - and is the boundary a proper whole-life carbon assessment needs, because only then are you comparing the full consequence of a choice. Cradle-to-cradle goes further, crediting the material's next life through recycling or reuse (closely related to module D and to circular-economy thinking). A fifth, gate-to-gate, isolates a single process step. The point is not to memorise a glossary but to grasp that each boundary answers a different question, and a number is only interpretable once you know which one it used.

Here is where the danger lives. A cradle-to-gate figure for one product and a cradle-to-grave figure for another are not comparable - the second counts a whole life the first ignores - yet they are constantly placed side by side, sometimes innocently, sometimes to flatter a favoured option. A material can win at the gate and lose over the grave (because it wears out and needs replacing), or vice versa. This is also how greenwash hides: quote the flattering boundary, omit the rest, and let the reader assume it is the whole story. The professional reflex is simple and non-negotiable - before comparing any two carbon figures, confirm they share the same boundary and cover the same modules. If they do not, the comparison is invalid until they are put on the same basis, which is exactly the kind of work you defer to a specialist and the standards.

Even with matching stages and boundaries, a carbon figure is meaningless until you know what it is measured *per*. That is the job of the functional unit: the precise, quantified description of the function being delivered, against which all impacts are normalised. "200 kg CO2e" is not an answer; "200 kg CO2e per square metre of external wall providing a specified thermal performance and fire rating for a 60-year service life" is. The functional unit forces a fair comparison by pinning down not just the amount but the *job*: two wall systems can only be compared if they deliver the same function - the same performance, for the same duration. Compare per kilogram and you will favour dense materials unfairly; compare per unit of function and you compare like with like.

For construction products specifically, EPDs often use a declared unit (for example, per kilogram or per cubic metre of product) rather than a full functional unit, precisely because the manufacturer cannot know how much of the product your design will use or how long it must last. That is legitimate, but it throws the burden onto you: a declared-unit figure must be scaled by the quantity your design actually uses and, where relevant, by how often the product is replaced, before it becomes a meaningful contribution to your building. Reading "per kg" as "per project" is a classic error.

Inseparable from the functional unit is the reference study period - the length of time the assessment considers, typically 50 or 60 years for a building, though it varies. It matters because the study period governs how many times replaceable components (B4) are counted, how much operational carbon (B6) accumulates, and therefore the whole balance between up-front and lifetime emissions. Two identical buildings assessed over 50 versus 100 years will report very different totals, and neither is "wrong" - they answer different questions. So the rule that closes this lesson is a compound one: a carbon figure is only comparable when the stages, the boundary, the functional unit *and* the reference study period all match. Miss any one and you may be comparing apples with orchards. This is why credible comparisons are done to a common standard by someone who controls all four - and why your first question of any number is always "per what, from where to where, over how long?"

How does this discipline change what you actually do? First, it makes you a sharper reader. Handed a carbon claim - by a manufacturer, a competitor's case study, a rating submission - your reflex becomes to find the four edges before you react: which modules, which boundary, which functional or declared unit, which study period. Nine times out of ten the claim is fine but partial (cradle-to-gate only, say); occasionally it is misleading; either way you now know what it does and does not prove. That single habit will save you from more bad decisions than any amount of material trivia.

Second, it makes you a better briefer of assessments. When you commission a whole-life carbon assessment (2.4), you - with the specialist - set the goal and scope: you agree the boundary (usually cradle-to-grave for a real building decision), the reference study period, and the functional unit, up front, so the result answers your actual question and can be compared against benchmarks that use the same basis. An assessment scoped loosely produces a number that cannot be benchmarked or trusted; an assessment scoped tightly and transparently produces one that can. Agreeing these edges is a five-minute conversation that determines whether the whole exercise is worth anything.

Third, it keeps you honest. Because you understand that the boundary makes the number, you will resist the temptation - subtle and common - to quote the flattering scope for your own project. If your building's up-front carbon (A1-A5) is high, you do not bury it by only ever reporting a cradle-to-gate product figure; you report on the agreed whole-life basis and let the number stand. The standards (EN 15978 for buildings, EN 15804 for products) exist precisely to fix these edges so that everyone counts the same way, and the reason binding assessments go to a qualified specialist is that getting the boundaries, modules, units and period consistent - and defensible against a benchmark - is exacting work. Your role is to understand the edges well enough to demand they be set properly and read honestly. That is what turns "it depends" from an excuse into a discipline.