Demolishing (some) character homes is the low-carbon option, actually.

Life-cycle analysis in a nutshell: shit in, shit out.

The Green Alliance UK recently published this very nice report Build up: the environmental case for new homes in sustainable locations. In addition to a little case study on up-zoning in New Zealand (p. 20), it includes an excellent section explaining why demolition of existing homes is sometimes right for the environment (p. 26). 

This is truly delightful to see, because the case for demolishing existing homes is one of my all time favourite topics of conversation – I made a thread about this last year, back when one of my lecturers at uni forced me to join Twitter and produce a thread so that he could grade my communication skills. Here I revisit that thread, but in formal prose with additional evidence from this new report. 

The NIMBYs of New Zealand have gotten really into talking about the environmental/carbon emissions impact of demolishing existing homes. It’s an argument that has become another tool in the toolbox for opposing new development in the ol’ leafy character suburbs.

They use lines like ‘the most eco-friendly building is one that already exists!’ if you consider life-cycle carbon emissions. The new frontier of emissions analysis! It’s no longer cool to only look at the *operational*, in-use emissions of a building. Now, you have to consider the *embodied*, up-front emissions too: all the carbon (and energy) that goes into building a building before you can use it. Concrete and glass aren't free. These materials cost us to produce and transport, and the emissions are locked-in before you even use the building.

And they’re wasted if you demolish a building too early – before its useful life-cycle is up. The argument goes: even if you replace an existing home with something modern and energy efficient (much lower operational carbon), the operational carbon savings don't out-weigh the wasted embodied carbon of the demolished house and the construction of the new house. So, with a life-cycle view, demolishing existing homes increases emissions. Probably true-ish.¹ NIMBYs rest case.

BUT! (there is a big but!!!) 

Good life-cycle analysis doesn’t cherry-pick parts of the life-cycle. Life-cycle analysis is absolute rubbish if your assumptions are bad (sorry but it’s true). Shit in, shit out. The NIMBY-argument assumes we’re trading like-for-like: one existing home replaced by one new home. Bad assumption! 🚩 🚩 We (the urban woke-sters) want to replace each home with lots of new homes!

Houses don’t exist alone in space – no house is an island – they’re part of a complex system: the city. Life-cycle analysis needs to consider the context: Aotearoa has a housing crisis caused by high demand and a persistent under-supply of houses. We need to build more homes (this is not optional) and they need to go somewhere – ideally somewhere where the life-cycle emissions of the whole city will be lowest. Which means we need to consider the life-cycle of the whole system. We have options – knock down houses and rebuild at higher density in inner-city suburbs, OR, build these new houses on the city fringe (this is the counterfactual!).

We have to assume that the embodied carbon-cost of new houses will be paid somewhere, but we can reduce it. Building at higher densities, ‘gentle density’ – terraces and walk-up apartments – reduces the carbon emissions per house by using resources more efficiently. Terraces and walk-ups share walls and roofs, and they tend to be smaller, which reduces material input (embodied carbon) and makes them more energy efficient (operational carbon). The higher the density, the lower carbon cost. To a point. For skyscrapers, the concrete and steel input (embodied carbon) needed to keep them upright gets too high – there is a sweet spot. 

Gentle density is generally better, but we could be building these terraces and walk-ups at the city fringe – why do we need to build like this in existing suburbs? Because density affects how efficiently we use infrastructure. When homes are closer together, it is cheaper – both in terms of carbon, and money – to service them with water, electricity etc., and connect them up with low carbon transport options (trains, buses, cycle paths). 

These efficiencies mean that the infrastructure material input for a city tends to increase at a much lower rate than the population (#sublinearscaling). Which means the embodied carbon of infrastructure tends to be lower per person in large, high density cities. For public transport solutions in particular, the denser the better – you need lots of people living close together to make a light rail worth building. But planners have to trade transport emissions off against the density sweet spot for buildings’ emissions. And they have to consider the carbon cost of all the other infrastructure – three waters, electricity lines, gas networks etc. Aaaaaand they need to consider impacts on affordability, which is a can of worms I intend to address in a future post.

Planners have to model (or at the very least, think hard about) all of these issues when making zoning decisions to find the perfect balance: the “goldilocks density”. Sadly, it’s been rare to see anyone attempt to model the transport/buildings emissions trade-off. But the Build up report does this. 

“To demonstrate the impacts of decisions around where to build, we imagine two scenarios in which 250 homes are added to an urban area. 

In the first, ‘demolish and densify’ scenario, 50 existing semi-detached homes are demolished and replaced with 300 flats. 

In the second ‘no demolition’ scenario, the same 250 homes are instead built on the urban fringe at a density of 15 dwellings per hectare. 

The first scenario results in no additional land take, providing 90 dwellings per hectare. The second scenario results in 17 additional hectares of greenfield land used for housing.”

“We have calculated the subsequent impact on carbon emissions of each scenario by modelling the embodied carbon expended in constructing the new semi-detached homes or flats, their heat requirements and the emissions from residents’ car use.

After 60 years, the ‘no demolition’ scenario results in 3,062 tonnes of extra carbon emitted to the atmosphere, compared to ‘demolish and densify’. The main effect causing this is the car dependency of the households on the urban fringe. This is the case even though we have assumed the switch to electric vehicles over the period into the early 2030s, in line with the Climate Change Committee’s (CCC’s) balanced net zero pathway. We also assume a change to electrified heating from 2030 and decarbonisation of power generation, in line with the CCC’s projections.”

Very nice, very tidy. I love to see it. 

In conclusion, I am not saying we have to demolish all the character homes, but some of them? I’d be into that. At the end of the day, I agree that character suburbs are nice and if there were no consequences to preserving them, I'd be all for it! But there are consequences, and they're severe: one of which is spiralling urban carbon emissions.

Side note:

The New Zealand Climate Change Commission recently published their second round of advice to the government on the next Emissions Reduction Plan, including a chapter on the built environment, in which they make a nice, simple recommendation that the government should:

“ensure planning systems are integrated and support building urban areas upwards and mixing uses in order to reduce emissions, while decreasing susceptibility to climate-related risks.”

Easy peasy!

The Climate Change Commission makes some strong assertions that:

“international studies have demonstrated that denser urban forms result in significantly fewer overall emissions – including embodied, enabled, and operational emissions. These studies show that greenhouse gas emissions per person in high density areas measured less than half of those in low density areas. Operational emissions from energy use – for either transport or buildings – are much more significant than embodied emissions, though embodied emissions are not insignificant.”

While I do believe all these statements are true, there is definitely still a dearth of research that comprehensively models urban life-cycle emissions at the neighbourhood, or city scale. Case and point: the commission only references three studies here. While all three do consider transport and building emissions together, one re-used the embodied carbon calculations from another of the three, and none of them consider the demolish-and-rebuild scenario - they all treat new development as essentially greenfield and merely compare density options.

Photo: Nick Sarvari via Unsplash

1. This is my professional opinion and I am technically a qualified environmental scientist. It does of course depend on what assumptions you make about the life-span of a new building, and the associated operational emissions. While it’s generally agreed that operational emissions tend to outweigh embodied emissions over the entire life-cycle of a building, when emissions are produced is also very important: we should probably worry more about embodied, upfront emissions because they are happening in the present (and the present is a time when we desperately need to be cutting emissions).