The construction industry is fast arriving at the consensus that a retrofit-first approach to development is foundational to achieving significant decarbonisation.
As grids get cleaner and operational energy standards tighten, upfront embodied carbon accounts for an increasing proportion of lifetime emissions. A building’s structure is typically the worst offender, but is also the element most likely to be reused in retrofit, allowing for more major alterations to facades or interiors without a huge carbon penalty. So retrofit is good, and should be encouraged, applauded, and incentivised by policy.
This is the right starting point, but more detail brings more nuance. The real carbon impact of a retrofit project can only be assessed by comparing it to what would happen over a given time period if the proposed retrofit didn’t take place. In other words, by comparing to a baseline.
Carbon performance baselines are worthy of consideration in new-build construction too, but the options are even more varied in a retrofit context. With varied baseline options comes increased opportunity for subjectivity, confusion, and even willful manipulation.
As a simple example, we could divide the potential baselines against which to compare a proposed retrofit into 3 categories:
- Demolition & new build – if without a retrofit, the existing building would be demolished and rebuilt. Using this baseline scenario will typically look favourably on a proposed retrofit.
- No interventions – if without a retrofit, nothing would change. This might be the case if the existing space is working perfectly well. Using this baseline scenario will encourage a retrofit proposal to pay back the upfront emissions within an acceptable timeframe.
- Light touch retrofit - if the proposal is a deep retrofit, it could be that the baseline alternative is a lighter touch fabric or services upgrades.
Whether a proposed retrofit has a positive, negative or marginal carbon impact depends strongly on which of these (or various other) baseline scenarios is chosen. A thorough analysis is required to properly understand the right course of action, with as much importance attached to finding the right baseline as to the details of the proposed scheme. Building designers should always have these considerations in mind during concept stage, but this approach is also relevant to policy makers and planners, who have the power to set precedents and block inappropriate proposals.
A relatively extreme example, not uncommon in London, is retrofit (often bordering on new-build) behind a façade retained for conservation reasons. Many of these projects benefit from the sustainability marketing behind retention, while in truth they emit far more carbon over the critical time period than a do-nothing scenario, even with improved fabric and heat pumps.
A simplified carbon payback analysis from one of these projects is shown in Figure 1. The proposal was to retrofit a row of Victorian terraced houses, along with the addition of a top storey and a basement. The facades were to be retained, but the extent of the work meant a large amount of new concrete and steel structure. Fabric upgrades and new elements achieving the EnerPHit Standard meant that heat pumps would work comfortably with the low heating demand. The existing dwellings were perfectly functional and fit-for-purpose, but had inefficient fabric and ran on gas boilers.
Despite the operational carbon improvement, it takes 43 years to pay back the upfront carbon emissions. By then it will be nearly 2070, and the likelihood is that climate change will have pushed beyond various devastating tipping points (or more optimistically, we will have been living in a net zero world for decades).
It’s possible to make the proposed retrofit look even worse by assuming that heat pump technology will improve over time, allowing installation more widely without any significant fabric intervention. Figure 2 shows a simplified version of this scenario, with a heat pump installed in the existing dwelling in year 20. In this case, the carbon payback of the proposed upfront emissions stretches towards a century. A similar result would be achieved by comparing the proposal to other lighter touch and more carbon conscious interventions.
These graphs show resoundingly that the scheme performs poorly from a carbon perspective, and is unlikely to align with local or national emission reduction commitments. This is largely caused by a new and particularly carbon-intensive basement, rather than what we would typically regard as retrofit. A failure to carry out this scenario analysis, or at least to attach sufficient weight to the do-nothing baseline, mean the development is likely to get the go-ahead from planners. The exemplar operational energy performance also serves to mask the truth, taking advantage of an industry that often prefers to stay in the decades-old comfort zone of U-values and double glazing.
Other retrofit projects of existing fit-for-purpose buildings will have a much shorter carbon payback period, particularly those where decarbonisation is a primary objective. However, it’s still worth keeping an eye on any extensions and basements, as these could easily push the payback period into a debatable category.
In some cases, the true baseline scenario is even less clear. The most high-profile example of baseline uncertainty is the M&S Oxford Street project, where demolition and new-build was justified by comparison with a seemingly light touch retrofit, with a surprisingly high energy use of 257 kWh/m2yr. The proposed scheme’s upfront carbon emissions and subsequent low operational emissions were therefore compared to this, rather than a much more energy efficient upgrade. The approach was challenged by industry figures including member of Construction Carbon’s Technical Advisory Panel Simon Sturgis, resulting in the proposal ultimately being rejected by Secretary of State Michael Gove. The fact that the development had successfully made it through planning up to this point demonstrates a lack of unified thinking and policy in this area.
A proper scenario analysis is required to truly understand the carbon impact of any proposed intervention to an existing building. At least 3 scenarios will be relevant in many cases: demolition & new build, no intervention, or retrofit. Further complexity is added by the fact that retrofit can be subdivided into ‘deep’ and ‘light touch’, and that many retrofits are accompanied by elements of new-build. A more subjective decision on whether an existing building is or isn’t fit-for-purpose can add yet another layer of uncertainty.
Combining these considerations makes for a confusing web of alternatives and a lot of work for a competent LCA consultant. Fortunately, real-life options are often much more obvious than in some of the scenarios described here, particularly if financial constraints limit the available pathways.
It is nevertheless important that consultants and designers present all available options and their carbon implications to clients, allowing informed decisions to be made. Increasingly robust embodied carbon benchmark data means that a simple scenario analysis can be performed without undertaking a full lifecycle carbon assessment. The difference between options will often be big enough to absorb large data uncertainty, so high levels of accuracy need not be a major concern during early-stage analysis.
A similar approach could be used by policy makers and planners. Other than in exceptional cases such as the M&S example, careful scenario analysis and carbon comparison doesn’t appear to play a part in planning decisions. Working this into policy would ensure that low carbon options over an appropriate timeframe are prioritised. The value of the prescriptive analysis requirements that typically come with formalised rules would need to be balanced against the highly project-specific nature of many of these considerations in order to incentivise genuinely sustainable redevelopment.