Retrofit versus Rebuild

The commonest error is to compare the two options on the wrong number. A shiny new building will usually run more efficiently than a retrofitted old one, so if you look only at annual operating energy, rebuild seems to win. But that ignores the far larger figure: the embodied carbon of construction - and demolition. The right yardstick is whole-life carbon: embodied plus operational, over the building's whole life (Module 3).

Count it properly and the two paths look very different. A rebuild carries: the carbon of demolishing the existing building and disposing of the waste; the full embodied carbon of the new structure, envelope and finishes; and then its operational carbon over time. A retrofit carries only: the (much smaller) embodied carbon of the upgrade measures - insulation, glazing, a heat pump; and its operational carbon, which is now low too. The rebuild's embodied carbon is typically several times that of a deep retrofit for a comparable performance outcome.

So the honest comparison plots both paths over time, as in the figure. The rebuild starts with a huge upfront carbon spike and then descends slowly on a shallow operational slope. The retrofit starts with a small spike and descends on a slightly steeper (less efficient) slope. The two lines may eventually cross - the point where the rebuild's better efficiency finally repays its upfront debt - but that carbon payback often lands decades out, sometimes beyond the building's likely life, and always after the critical decade in which the extra carbon did its worst atmospheric damage. Emitting a tonne now is worse than emitting it in 2050; the timing is not a footnote, it is the point.

The decarbonising grid tilts the comparison further toward retrofit. The new building's advantage is mostly operational efficiency - but as electricity gets cleaner, the carbon value of that operational advantage shrinks year on year, so the rebuild's slow repayment slows even more and the crossover retreats into the distance. Meanwhile the rebuild's embodied carbon was spent at today's dirtier emissions intensity and can never be recovered. In other words, betting on a rebuild is betting that operational carbon will keep mattering as much as it does now, when the whole direction of the energy transition is to make it matter less - which is exactly the wrong bet.

Retrofit wins in the great majority of ordinary cases: where the existing structure is sound or economically repairable, where its geometry can serve the need (as-is or adapted), and where a deep retrofit can bring performance up to a good standard. That covers most homes, offices, schools and commercial buildings. The upfront carbon saved by keeping the structure is so large, and lands at such a critical time, that it usually dominates the comparison. This is why 'retrofit first' is emerging as policy and best-practice default - most visibly in London, where major demolitions increasingly have to justify themselves against a retrofit option and a whole-life carbon assessment rather than proceeding by right.

Rebuild genuinely wins in a narrower set of honest cases, and it is worth being precise about them so the exception is not abused. When the existing structure is unsafe or so degraded that repair costs more carbon and money than replacement. When its geometry is hopeless for any viable use - floor heights, grid or plan that no adaptation can rescue. When the site must carry far more density (say a two-storey building where policy and need demand ten), so a like-for-like retrofit cannot meet the actual requirement and the extra floor area's value, carbon-wise, justifies rebuilding. And, occasionally, when the existing building is so extraordinarily energy-hungry and the rebuild so efficient that the carbon genuinely pays back quickly. These are real; they are also less common than demolition-by-default assumes.

The test for an exception is honesty. 'The structure is unsafe' must mean a real engineering finding, not a convenient survey. 'We need more density' must be a genuine requirement, not a yield-maximising preference dressed as necessity. And even a justified rebuild carries an obligation: build the replacement to a genuinely low-carbon, low-embodied standard, and design it for its own eventual disassembly and reuse (Module 4), so the carbon you are choosing to spend buys the longest, best-designed life you can give it.

Turn all of this into a repeatable process you can run on any project, as in the framework figure. Start by assuming retrofit - that is the default, and it sets where the burden of proof sits. Then work through the tests.

1. Assess the structure. Is it safe and sound, or economically
   repairable? If genuinely not -> rebuild may be justified.
2. Test the geometry/use. Can it serve the need, as-is or adapted
   (change of use, extension)? If truly not -> rebuild candidate.
3. Compare whole-life carbon. Retrofit embodied+operational vs
   rebuild demolition+embodied+operational, over ~60 years.
4. Compare whole-life cost, and the softer values - character,
   disruption, speed, community, planning risk.
5. Default to RETROFIT unless a rebuild is clearly justified on
   an HONEST reading of 1-4. If you rebuild, build it low-carbon
   and design for disassembly.

Several things make the comparison fair rather than rigged. Use a proper whole-life carbon assessment over a realistic study period (commonly ~60 years) - and count the demolition and waste of the rebuild option, which are routinely 'forgotten'. Compare a genuinely deep retrofit against the rebuild, not a token one, so you are not straw-manning reuse. Be explicit about the study period and assumptions, because they swing the result. And weigh the timing: give appropriate weight to upfront carbon spent now versus operational carbon spread over decades.

A subtlety worth internalising: the tests are ordered for a reason. Structure and geometry come first because they are pass/fail gates - if the frame is genuinely unsafe or the geometry truly cannot serve any viable use, the carbon comparison is moot and you are into justified-rebuild territory. Only if the building clears those gates do you run the whole-life carbon and cost numbers, and only then the softer values. Running the tests in that order stops you from wasting effort modelling a building that cannot be kept, and, more importantly, stops a marginal cost or yield argument from overriding the fundamental question of whether reuse is physically sound - which it usually is.

Beyond carbon, bring in whole-life cost (including demolition, finance and the value of earlier occupation) and the harder-to-quantify factors - heritage and character, disruption to occupants, programme, community attachment, and planning risk, which increasingly runs against needless demolition. Run this framework and you replace a gut call with a defensible decision. That is what a client, a planner and the planet are entitled to.

Because so much rides on this decision, it is a magnet for motivated reasoning and quiet greenwashing (Module 10). The pressures usually push toward demolition: new-build is simpler to design and finance, developers may prefer the yield of a bigger new building, and 'it wasn't worth saving' is an easy story to tell after the fact. Your professional value is to keep the comparison fair when incentives are pulling it out of shape.

Watch for the classic ways the maths gets rigged. Omitting demolition carbon from the rebuild side flatters it enormously. Comparing a deep new-build against a shallow retrofit straw-mans reuse - compare like for like. Choosing a study period that happens to be just long enough for the rebuild to pay back, or long enough to bury the upfront spike. Overstating the existing building's problems with a pessimistic survey, or understating what a good retrofit could achieve. And ignoring the timing of emissions, treating a tonne now as equal to a tonne in forty years. A trustworthy comparison names its assumptions, counts everything on both sides, and would survive being shown to a sceptic.

The honest bottom line is not 'never demolish'. It is that demolition should be a justified exception, not a default - decided by a fair whole-life comparison, with the burden of proof on the rebuild. Get into the habit of asking, on every project that starts with 'we'll knock it down and start again', a simple question: have we honestly tested whether we could keep it? That question, asked early and answered rigorously, is one of the most powerful carbon interventions an architect can make - and it is the practical heart of the entire existing-building imperative this module has built.