Embedding Carbon in the Workflow

The central tension of carbon in the workflow is that the two things you want - influence over the carbon and certainty about the number - move in opposite directions across a project. At the very start, at brief and concept, your influence is enormous: you can still reuse a building instead of demolishing it, change the whole structural system, halve the floor area, or decide the project should not be built at all. But your certainty is low - you have massing and intent, not quantities, so any carbon figure is a coarse estimate. By the end, at detailed design and construction, it is the reverse: you can produce a precise, data-backed number because everything is specified, but your influence has collapsed to minor substitutions because the form, structure and envelope are fixed.

A workflow that only assesses at the end sits at the worst corner of that trade-off: maximum certainty, minimum influence. It knows the answer precisely and can do nothing about it. The whole point of embedding carbon is to also work at the other corner - to make coarse, fast estimates early, when they are uncertain but can still change the outcome by large factors. A rough concept-stage estimate that says "this scheme is structure-heavy" is worth more than a precise handover figure, because you can still act on it.

So the discipline is assess early and assess often: a first rough carbon read at concept, updated at each stage as the design and data sharpen, rather than a single assessment at the end. Early estimates are deliberately approximate - order-of-magnitude, using default or benchmark data - and that is fine, because their job is to steer between options (reuse vs new, concrete vs timber, this grid vs that), not to report a final figure. As the design develops, the estimates get more precise and the decisions they inform get finer. The binding whole-life number still comes at the end, from a qualified LCA specialist working to the recognised standards - but by then it should confirm a low-carbon design that early, iterated estimates already shaped, not deliver a verdict on a design that ignored carbon until it was too late.

Embedding carbon means being concrete about which question gets asked at which stage and who owns the answer. At the brief, before design starts, the question is whether carbon is even a goal: agree a whole-life carbon target or at least an ambition with the client, decide reuse of any existing building is the default to beat, and write carbon into the brief so it is not optional later. At concept, when massing and the reuse-or-new decision are live, make a first rough estimate - even a back-of-envelope one driven by floor area and structural type - to compare fundamentally different options and kill carbon-heavy directions early. At scheme design, when the structural system, grid and envelope firm up, run a proper option comparison: this is where the biggest measurable reductions are locked in, so it deserves real effort. At detailed design, refine the assessment with actual specifications and verified EPD data, and use it to choose between products and mixes. At construction, verify against the as-built reality - substitutions, wastage and procurement changes can move the number - and capture the final data.

Ownership matters as much as timing. In a small practice the architect or a carbon-literate team member may carry the early estimates, calling in an LCA or carbon specialist for the detailed assessment and the binding numbers. In a larger project the specialist may be engaged from early on. Either way, someone must be named as responsible, or carbon quietly falls off everyone's plate. The structural engineer is a key partner because structure is usually the biggest chunk; the cost consultant increasingly reports carbon alongside cost; the contractor holds the as-built truth.

The practical trick is to right-size the effort to the stage: fast and rough when the design is fluid, thorough and data-backed when it is fixed. A common mistake is demanding a full, precise LCA at concept - it is slow, spuriously precise given how much will still change, and it burns the goodwill that keeps carbon in the process. Match the depth of assessment to how much the design can still move, and carbon assessment stays light enough to actually happen at every stage.

Carbon assessment survives in a workflow only if it rides on data the team is already producing rather than demanding a parallel, hand-built exercise. This is where BIM (building information modelling) earns its place. A model that carries geometry and, increasingly, material and quantity information is most of what a carbon estimate needs: extract the quantities of concrete, steel, masonry, timber, glazing and finishes, attach a carbon factor (a benchmark early, a verified EPD figure later) to each, and you have an embodied-carbon estimate that updates as the model changes. Several carbon tools plug directly into common BIM platforms, or read a quantity take-off exported from them, so the estimate can refresh with the design instead of lagging weeks behind it.

The prize is live feedback: change the structural grid or swap a slab type in the model, and see the carbon move, the same way you already see cost or area move. That turns carbon from a report into a design dashboard. It also removes the biggest excuse for skipping early assessment - that it is too slow - because the quantities are already there. In practices without deep BIM, the same logic works more manually: a spreadsheet linked to a quantity take-off and a benchmark carbon dataset can give a serviceable early estimate; the tools matter less than the habit of estimating.

Two cautions keep this honest. First, garbage in, garbage out - a carbon number from a model is only as good as the quantities and factors behind it, and early models are incomplete, so treat model-derived figures as estimates whose uncertainty you state, not as precise truth (Module 3.3). Automating the arithmetic does not automate the judgement. Second, the tool does not choose the standard or the data - what boundary is drawn, which life-cycle stages are counted, and which EPDs are used are decisions that follow the recognised methods and a qualified specialist, not defaults buried in software. Used well, BIM and carbon tools make assessment fast enough to be routine; they do not replace the method, the data discipline or the specialist who signs off the binding whole-life figure.

The habit that ties it all together is iteration: treating carbon as a number you watch and improve across the whole design, exactly as you already watch cost, floor area and programme. No competent practice designs a building, freezes it, and only then checks whether it is affordable - cost is tracked and steered from the first sketch, with rough figures early and precise ones late. Carbon deserves identical treatment. A live carbon estimate, revisited at every design review beside the cost plan and the area schedule, keeps the team steering toward a lower-carbon outcome instead of discovering the outcome at the end.

In practice this looks like a short carbon line in the design-review conversation: where are we against the target, what moved the number since last time, and what is the next lever. It looks like option studies that report carbon alongside cost and buildability, so the low-carbon option is visible and arguable rather than invisible. It looks like a designer who, offered a heavier structure or a more carbon-intensive finish, asks "what does that do to our carbon?" as reflexively as "what does that do to our budget?" The number does not have to be precise to do this work - a consistent, roughly-right estimate that everyone trusts to compare options is far more useful than an exact figure that arrives too late.

Iterating also guards against the classic trap of a single heroic low-carbon decision undone by a hundred small later ones - a lean timber structure quietly loaded with carbon-heavy finishes, or a reuse project buried under a high-carbon fit-out. Only a carbon number that stays live through detailed design catches that drift. And it makes the final, binding assessment a formality rather than a verdict: when the specialist runs the whole-life carbon to the recognised standards at the end, it should confirm the low-carbon building the iterated estimates already produced. The aim is not a better report; it is a lower number, arrived at by designing with carbon in view the whole way through - the discipline the rest of this module, and course, is meant to make second nature.