Climate resilience in buildings: compare the service risk, not the latitude

There is a temptation to reason about climate resilience geographically: northern European buildings face cold, Gulf buildings face heat, and each is built accordingly. The reasoning is not wrong as a general observation about building stock and it is inadequate as an assessment method.

Exposure varies substantially within a region, within a city, and between adjacent buildings. Elevation determines flood exposure. Local drainage capacity determines whether heavy rainfall becomes a flood event. Orientation and shading determine solar gain. The condition and capacity of the building's own services determine how it responds when conditions exceed design assumptions. A building on higher ground with good drainage and recently renewed cooling has different exposure from a neighbouring building without those attributes, and the difference is not captured by their shared location.

The same applies in the other direction. Not every building in a hot climate is equally exposed to heat, because construction, shading, servicing strategy and occupancy pattern all mediate the effect. Treating all buildings in a climate zone as equivalent produces a picture that is neither accurate nor useful.

The framing that makes climate resilience comparable across buildings is the service, not the hazard. A building is not primarily at risk of a hazard; it is at risk of failing to provide what it exists to provide — a comfortable place to work, a safe place to recover, a controlled environment for a process, an uninterrupted service to the public.

That framing produces four questions, and they apply to any hazard.

Site hazard: what conditions can this specific site experience, based on its location, elevation, drainage and the local data available? This is established from site-level hazard information and, where appropriate, specialist assessment — not from a regional summary.

Asset vulnerability: how does this building respond to those conditions? Construction, envelope, servicing capacity, redundancy and the condition of the plant all determine whether a given condition causes a failure. A building with spare cooling capacity responds differently from one operating at its limit.

Occupancy sensitivity: how much does the building's use tolerate the disruption? A data centre, a hospital and a seasonal office have very different tolerances for an interruption of the same duration. The consequence depends on what the building is doing when the event occurs.

Operational consequence: what actually happens when the service is interrupted or degraded — how long, at what cost, with what knock-on effects, and for how long afterward? This is where the risk becomes a business question rather than a technical one.

Scenario analysis is the appropriate tool for assessing climate resilience, and it is easily misused by attaching probabilities to the scenarios and computing an expected outcome.

The misuse produces a precise-looking figure resting on probability estimates that are not available at building level. The more honest use is to describe a set of conditions and ask how the building's service performs under each, which identifies the conditions under which the building fails and the measures that would prevent it.

The output of that exercise is a set of decisions — a measure that is worth taking because it prevents a failure under a condition the building cannot tolerate, and a measure that is not worth taking because the condition is one the building handles comfortably. That is a more useful output than a risk score, and it can be acted on.

Resilience measures divide into those that change how the building is operated and those that change the building itself, and the two should be assessed against what they actually deliver.

Operating measures — setpoints within safe and comfortable limits, sequencing, maintenance regimes that keep plant performing, monitoring that detects degradation early — generally cost less and depend on consistent operation. Their limitation is that they are reversible in practice: a measure that depends on how people behave will decay without monitoring and accountability.

Capital measures — additional cooling capacity, drainage works, shading, storage, backup power, relocation of vulnerable functions — cost more and deliver durable change. They are also the measures where the assessment must be most careful, because the capital is committed against a condition that has not yet occurred and the sizing depends on assumptions about severity.

The two also differ in what they address, and conflating them is a common error. Measures that reduce emissions address the contribution to future climate change; measures that reduce exposure address the conditions the building will experience. A building can invest substantially in one and remain exposed to the other. Presenting an emissions measure as though it reduced a site's flood exposure misrepresents what has been achieved, and it can leave a real risk unaddressed while appearing to have been handled.

Where a question involves building services, structural capacity, drainage or fire safety, it requires the relevant qualified professionals, and any modelling of hazard or building response must be carried out by parties competent to do it. Nothing in this article substitutes for that.