Benchmarks & Targets

A benchmark is a reference value for embodied carbon per unit of floor area - kgCO2e/m2 - for a defined type of building, drawn from studies of many real projects. It answers the orienting question your own number cannot: is this typical, or heavy, or light, for a building of this kind? Benchmarks are usually published as a spread with bands - a business-as-usual value around which most buildings cluster, and progressively lower values for good practice and best practice - so that placing your building against them tells you not just a pass or fail but roughly how far from the leading edge you are. That placement is the whole point: it converts a bare figure into a judgement and a direction.

The catch, and it is the defining catch of this whole lesson, is that a kgCO2e/m2 number is meaningless without its boundary and its basis. Three things must match before two figures can be compared. The life-cycle scope: does the benchmark count only up-front carbon (A1-A5), or whole-life including replacement and end-of-life? A whole-life figure and an up-front figure for the same building differ enormously, and comparing across them is nonsense. The elements included: structure only, or structure plus facade, finishes, services and external works? And the floor-area definition: gross, net, which measurement convention? Two honestly produced figures for the identical building can differ by a factor of two purely from these choices.

So the first discipline of using benchmarks is forensic: before you compare your number to any benchmark, confirm that your boundary matches its boundary. A building that looks excellent against a structure-only benchmark may be ordinary against a whole-building one; a whole-life target compared to your up-front number will flatter or damn you for no real reason. Reputable benchmark sources state their boundary precisely for exactly this reason, and the skill is to read that boundary first and the number second. A kgCO2e/m2 quoted with no stated boundary is not a benchmark you can use - it is a number you should distrust. This is also why benchmarks are published as bands rather than single lines: the spread acknowledges that real buildings vary and that data is uncertain, and it lets you read your building as a position within a range rather than a pass or fail against a false point. Treat a benchmark, in other words, the same way you learned to treat your own result in the last lesson - as a reference with a scope and a spread, not a hard number.

Where a benchmark describes what buildings do, a target prescribes what yours will achieve, and targets come in two flavours worth distinguishing. An absolute target is a fixed kgCO2e/m2 value the building must come in under - the approach behind the well-known band-based frameworks (LETI and the RIBA 2030 Climate Challenge are widely cited examples) that publish tightening limits for different building types and years, so a project can declare it will meet, say, the current best-practice band. A relative target is a percentage reduction against a defined baseline - "thirty per cent below business-as-usual for this type" - which is useful when a credible absolute figure for your exact context does not exist, as is often the case in India, because it only needs you to establish your own baseline honestly.

Rating bands stitch benchmarks and targets together into a ladder. Rather than a single line, frameworks typically define several tiers - business-as-usual, good, and best practice, sometimes lettered or starred - each a lower kgCO2e/m2 threshold, so a project can aim at a band and know what level of ambition it represents. These bands are genuinely useful as design orientation: they tell you not just whether you passed but how hard you pushed, and they tighten over time, so a best-practice building today is tomorrow's baseline. Some green-building rating systems and a growing number of regulations are beginning to reference such thresholds, turning what were voluntary targets into requirements - a direction of travel designers should expect to accelerate.

Two honesties must travel with any target. First, a target inherits all the boundary-dependence of the benchmark it is built from: a target of "below X kgCO2e/m2" is meaningless unless X's scope, elements and area basis are defined and matched by your assessment. Second, a target is a design commitment, not a reporting formality - its value lies in being set early and used to drive decisions, not announced at the end. A target set at concept and tracked down through the stages changes buildings; a target quoted in a completion report changes nothing. And the specific band values any framework publishes are region- and standard-specific and revised over time, so the figure you adopt must be the current one for your context, read from the source, not an illustrative number from a course.

The most common and most damaging error with benchmarks is assuming they are universal. They are not: a benchmark is a photograph of a particular population of buildings, built with a particular region's materials and energy, assessed to a particular standard - and it does not travel to a different context without distortion. Most of the widely cited embodied-carbon benchmarks and band frameworks were developed in Europe or similar contexts, and applying them unadjusted to an Indian project is a real methodological mistake, not a minor approximation. The reasons run deep. The material mix differs - Indian construction leans heavily on concrete, brick and blockwork where some benchmark populations assume more steel or timber. The energy grid differs - a coal-heavier grid raises the carbon of the same materials, so a "good" European figure may be physically hard to hit with Indian-made materials at the same design. And climate and typology differ - the buildings, their servicing and their expected life are not the same.

Standards-dependence compounds this. Benchmarks are computed to a method - which life-cycle stages, which boundary, which data vintage - and a figure produced under one standard cannot be laid against a figure produced under another without reconciliation. Even the floor-area convention varies between jurisdictions. So a number that sounds authoritative - "the benchmark for offices is X" - is only authoritative within its own region, type and standard, and using it outside that frame produces a comparison that is precise and wrong. The honest user of benchmarks always asks three questions before adopting one: where was this population, to what standard was it assessed, and does that match my project and my assessment?

For Indian practice this has a constructive implication rather than a despairing one. Where a credible local, boundary-matched benchmark exists, use it. Where one does not yet - a common situation, as Indian embodied-carbon benchmarking is still maturing - the honest move is often to establish your own project baseline (a business-as-usual version of this building, assessed consistently) and set a relative reduction target against it, rather than borrowing an ill-fitting foreign absolute. That keeps the comparison internally valid even when external benchmarks are thin, and it sidesteps the false precision of a transplanted number. Module 10.3 returns to the Indian benchmarking picture; the principle here is simply that benchmarks are local, and pretending otherwise is a common road to greenwash.

Putting it together, setting a carbon target for a real project is a short, disciplined sequence. First, fix the boundary - decide and write down what your assessment will count (which stages, which elements, which area basis), because everything downstream must share it. Second, establish the reference - find a current, local, boundary-matched benchmark for your building type if one exists, and if it does not, build your own business-as-usual baseline by assessing a conventional version of the project consistently. Third, set the target - choose a value below the reference that is ambitious but achievable for this project's constraints, expressed either as an absolute kgCO2e/m2 or as a percentage reduction against your baseline, and aligned to a recognised band if a credible one fits your context. Fourth, break it down - apportion the target across the big elements (a structure sub-target, an envelope sub-target) so the team knows where the carbon must come from.

Then the target has to be used, not filed. Set it at concept, when the levers are open, and track the building's estimate against it at every stage - measure, compare to target, reduce the hotspots, re-measure - so the target actively pulls the design down rather than merely judging it at the end. Expect to revisit the target as the design and data firm up, and be honest about that: a target missed for good reasons, documented, is more credible than a target quietly relaxed. The whole mechanism from Lesson 3.1 - the living assessment loop - exists to serve this target, and a target without that loop behind it is an aspiration, not a plan.

Finally, the honesty clauses. State your target with its boundary and its reference, so it means something to anyone who reads it. Do not adopt an illustrative figure (including any in this course) as your target - derive it from a current, local source or your own baseline. And keep the course's firm line: the binding, declared figure your project ultimately reports against the target belongs to a qualified assessor working to the recognised standards with verified data. Your job is to set an honest, boundary-defined, early target and to drive the design to meet it; theirs is to certify the result. Set well and used hard, a carbon target is the single most effective way to turn measurement into a lower-carbon building.