Refurb vs New Build Carbon Comparisons Are Mostly a Waste of Time

We are asked to do this regularly now. A planning authority requires a comparison of embodied or whole life carbon between refurbishing an existing building and replacing it with a new one. The applicant wants to demolish. The planners want evidence that they have considered the alternative.

It is good that planners are asking about embodied carbon at all. Five years ago most of them were not. Westminster City Council adopted a formal Retrofit First policy in January 2026, and the City of London and Camden have been moving the same way. More boroughs are following.

First, planners (and those advising them) need to get their terms right.

• Embodied carbon is the greenhouse gas emissions and removals from the materials and construction processes across a building's whole life cycle.
• Whole life carbon is all of a building's emissions and removals, embodied and operational together, over its full life cycle including disposal. In standard terms that is modules A0 to A5, B1 to B7 (B8 optional), C1 to C4, with module D reported separately.

Both definitions follow LETI's 2023 guidance. We often see the two conflated, and client briefs are guilty of this too. Including operational energy makes a big difference to these comparisons. Either way, I think the exercise is close to pointless, and worse, it is easy to game.

Here is why.

The answer is known before the work starts

A new build will have higher embodied impacts than a refurbishment. Almost always. In my experience it holds about ninety eight times out of a hundred.

The reason is simple enough that you do not need an assessment to see it. The carbon in an existing structure has already been spent. The foundations, the frame, the substructure, all of it was emitted years or decades ago. Keeping them costs you nothing more in embodied terms. Replacing them means paying that cost again, in full, at today's carbon intensity, plus the demolition and disposal of what was already there. It depends on the building type, but the substructure and frame are typically 20 to 40% of a building's embodied impacts and 10 to 20% of whole life (including operational).

There are exceptions, but they are rare. If a scheme adds a significant number of extra floors, the existing structure may need so much underpinning and strengthening that it is lower carbon to knock down and start again. Of the fifty or so of these comparisons we have run, maybe one made that argument. Even then it rested on a claim from the structural engineers that was difficult to verify.

If the comparison covers whole life carbon, the new build may claim operational savings. That argument is increasingly weak, for reasons I come to below.

So when a planner asks for the comparison, they are commissioning a piece of work whose conclusion is known in advance. We run the numbers, the numbers say what everyone expected, and the report goes in the file. Or it is kept out of the application, because it does not show what was hoped.

The refurb scenario is invented, which is where it falls apart

This is the part that bothers me more.

To compare the two options you need a refurbishment scenario to compare against. But there is no refurbishment. The applicant wants to demolish. Nobody has designed the refurb, nobody has costed it, nobody has surveyed what could actually be retained. So the scenario has to be constructed for the purposes of the assessment.

That means somebody decides how much of the existing structure is assumed to be retained, what interventions the refurbishment would need, what its expected service life would be, and what energy standard it would be brought up to. That last one matters more than most people realise. Building regulations for refurbishment are far less demanding on energy efficiency than they are for new build. Every one of these is a judgement call, and every one of them moves the answer.

Assume the refurbishment only meets minimum regulations, which are far lower than what is achievable, plus extensive structural intervention, a shorter remaining service life and heavy fabric upgrades, and its whole life carbon climbs. Assume the new build has top end energy efficiency, a long service life and a lean structure, and its annualised figure falls. Do both and you can narrow a gap that ought to be enormous.

I am not saying this is routinely done in bad faith. Most of it is defensible modelling by people doing their best with no design information. But the pressure is real, particularly from paying clients. We have been asked to adjust a refurbishment baseline because it gave the wrong answer. We did not, and that building still got planning permission, by leaning on its social and economic value rather than its environmental case.

An assessment where the comparison case is invented by the party that wants a particular answer is not reliable evidence. It is a document that can be steered.

This is not only my view. In a 2024 commentary in Buildings and Cities, Colin Rose of UCL looked at the Marks & Spencer Oxford Street case. He argued that the refurbishment option it was measured against was a "straw man", and that the approach leaned towards justifying demolition. That scheme was eventually approved in late 2024, after a call-in, a refusal and a High Court challenge, with the embodied carbon given only moderate weight against it.

What planners actually want, and what would work better

I do not think planners are wrong to care. The instinct behind the request is right. Demolition is a carbon decision and it should be scrutinised.

The problem is that the comparison has been chosen as the mechanism, and it is the weakest available tool for the job. A few things would work better.

Ask for a retention assessment instead of a comparison. Rather than "prove new build is worse", ask "what could be retained, and why was it not?" That question cannot be gamed as easily, because it is about the existing building, which is a real thing that can be surveyed, rather than a hypothetical refurbishment that is not.

Require the assumptions to be stated on the face of the report. Energy efficiency standard, retention rate, assumed service life for both options, scope of fabric upgrade. If those four or five numbers are declared prominently, a reviewer can see immediately whether the scenario is reasonable. Most of the steering happens in assumptions that are currently buried in a model. It would be great to see industry bodies (RICs or UKGBC) develop some standardised guidance on this.

But I think the best solution is to remove the comparison all together and set whole life threshold limits. A whole life carbon limit for the new build, expressed per square metre, is a far harder target to argue your way around than a comparison against a scenario you invented yourself. For developments that involve demolition of existing buildings the target could be lower (say 10%-20%) representing the lost carbon from demolition

To be clear

None of this is an argument against assessing embodied carbon. It is an argument for assessing it in a way that produces a decision rather than a document.

Whole life carbon assessment is useful when it shapes a design: comparing structural options, testing material choices, finding the reductions that are actually available on a real project. Our guide to reducing embodied carbon in construction covers where those savings usually come from. That work changes buildings. A comparison against a refurbishment nobody intends to build changes very little, beyond adding cost and time to a planning application.

If planning authorities want demolition scrutinised, and they should, there are better ways to ask for it.

We run whole life carbon assessments on live projects, including embodied carbon assessments for planning submissions where this comparison is required. If you are new to the methodology, our five-minute guide to RICS Whole Life Carbon Assessment is a good place to start, or get in touch if you have a scheme where retention is in question.

If you want a feel for the numbers yourself first, our free OpenImpact Building tool gives a quick whole building carbon estimate.