In short: UK buildings are designed against historical weather, but they will operate into the 2070s in a climate the record does not yet contain. Research projects heating demand falling up to 21% by 2080 while cooling demand rises sharply making climate resilient building design a present-day requirement, not a future one (Viganò et al., 2024).
Every building being designed in the UK today shares a quiet assumption: that the future will resemble the past. The weather files driving energy models are built from historical records of a statistical “typical year” assembled from decades already behind us. Yet the building itself will operate for fifty years or more, into the 2070s and beyond, in a climate the historical record does not contain.
That mismatch is no longer hypothetical, and it is now quantified. Research by a University of Brighton-led team including FREDS founder Marco Picco simulated the same UK building designs under current weather and under projected conditions for 2030, 2050 and 2080, across three locations spanning the country’s climate range: Exeter, Manchester and Aberdeen (Viganò, Rugani, Marengo and Picco, 2024, published in Architecture). The simulations ran on EnergyPlus and the Freds4Buildings platform, using future weather datasets from the Prometheus project and Meteonorm. The results describe, in numbers, what designing against yesterday’s weather costs and why climate resilience building design has stopped being optional.
What the simulations found
Heating falls gently. Across all three locations, heating demand declines steadily as the climate warms, reaching a reduction of around a fifth by the 2080 horizon. For a heating-dominated country this sounds like good news, and partially is.
Cooling rises sharply. The other side of the ledger dwarfs it. Where cooling systems were present, demand did not drift upward; it multiplied. In one configuration, annual cooling demand rose from 2,012 kWh to 8,254 kWh by 2080, a 310% increase; in another location the increase reached 700%. The larger non-domestic case told the same story at scale: heating down around 20% (92,170 → 73,482 kWh) while cooling climbed 203%. The UK’s defining building-physics assumption that summer takes care of itself is the assumption the projections retire.
Uncomfortable becomes uninhabitable-by-standard. For the conventional building without cooling, the research assessed overheating risk against the criteria used in UK practice: the share of occupied hours above the adaptive comfort threshold, with 3% of occupied hours as the pass/fail line, evaluated with the zone-by-zone approach implemented in Freds4Buildings. Under current weather: essentially no risk. Under future weather, the risk raises significantly. Exeter shows significant overheating as early as 2050, worsening by 2080; intermediate floors fail before the more sheltered zones; and even in cooler Aberdeen, hours of overheating risk climb steadily across every scenario. The same design, compliant against the past, fails against the future it will actually inhabit.
Notice what makes this finding operationally different from a general warning about warming: it has a date and a floor plan. The risk arrives within the design life of buildings being approved this year, and it arrives unevenly on top floors and high-gain zones first which is exactly the information a designer can act on.
How you simulate weather that hasn't happened
A fair question hangs over any 2080 result: where does future weather data come from, and why should it be believed? The team addressed the method directly in companion research (Reolon, Marengo and Picco, 2026), which sets out a transparent workflow for generating future EnergyPlus weather files.
The technique is called morphing: take a baseline weather file built from observed records, and adjust it by shift and stretch, variable by variable, month by month using climate anomalies from global climate model projections, then post-process so the file stays physically self-consistent (humidity coherent with temperature, and so on). The result is not a forecast of 14 July 2080; it is a typical year as the 2080s climate would shape it precisely the statistical object building simulation needs.
Benchmarked against established future-weather datasets (Prometheus, Meteonorm, Future Weather Generator), the workflow’s temperature series showed strong agreement, with correlation coefficients of 0.744–0.799. For Exeter, the morphed 2080s file shifts the annual maximum temperature from roughly 27.0°C to 31.6°C and lifts the winter minimum from −4.4°C to −2.5°C numbers that make the heating-down, cooling-up findings above feel less like model output and more like arithmetic.
The point of publishing the workflow is accountability: a future-weather claim is only as good as the method behind it, and a transparent, replicable method can be checked, criticised and improved. That is the standard future-performance claims should be held to by clients as much as by peer reviewers.
What climate resilience building design means in practice
Climate resilience building design is less a new discipline than a discipline applied to the right weather. The research converts directly into design behaviour. Four takeaways:
- Run every significant design against at least one future weather scenario. Not as a research exercise but as a stress test, the way an engineer load-tests a structure for conditions beyond today’s. A design that passes 2025 and fails 2050 is failing within its own mortgage term, and the failure is knowable now. With morphed weather files and fast simulation, the marginal cost of the 2050/2080 run has collapsed to minutes.
- Treat overheating risk assessment as a future-weather question by definition. Assessing overheating against historical files answers a question nobody is asking:Has this building overheated in the past? The regulatory direction (Approved Document O, TM59-style adaptive criteria) is toward exactly the occupied-hours analysis this research demonstrates; running it against projected weather is the version that protects the occupants who will actually live there in 2055.
- Design the passive defences from the start. The findings amplify everything the building-physics literature says about free cooling: the buildings that ride out the 2080 files are the ones with controlled solar gains, effective shading, thermal mass to buffer the peaks and night ventilation to discharge it. Those measures are cheap at concept stage and brutal to retrofit which is why the future-weather run belongs in the first week of design, not the last.
- Expect the cooling conversation, and get ahead of it. The projections imply many building types that historically ignored cooling will need a strategy: passive, mechanical or staged “cooling-ready” provision. Deciding that with numbers, early, beats discovering it through complaints in the building’s third decade.
The uncomfortable, useful conclusion
Nothing in this research is an argument against ambition; the same study’s net-zero findings, covered in a companion article, are genuinely encouraging. It is an argument against blind ambition: against certifying tomorrow’s buildings exclusively with yesterday’s weather. The gap between those two is now measurable: in our case study it is 21% of heating going, up to 700% of cooling arriving, the 3% overheating line crossed in Exeter by mid-century; and a measurable gap is a designable one.
The buildings that will still work in 2080 are being sketched this year. The weather they will face already exists as data. The only question is whether the design ever meets it.
Test your design against 2050 and 2080 weather. Future-weather simulation and overheating risk assessment are currently available through FREDS consultancy services. Talk to the team
Related reading
Frequently Asked Questions
How will climate change affect UK building energy performance?
What is a future weather file and how is it made?
Why use future weather instead of historical weather files?
What is climate resilience in building design?
References
Written by Marco Picco, PhD co-founder of FREDS4Buildings and a building-physics lecturer at the University of Lancashire, whose work on building energy simulation has been published in Energy & Buildings , Building and Environment and the IBPSA Building Simulation conference proceedings. About Marco Picco
References: Viganò, G.S.M., Rugani, R., Marengo, M. and Picco, M. (2024). Assessing the impact of climate change on building energy performance: a future-oriented analysis on the UK. Architecture, 4(4), 1201–1224. View paper · Reolon, M., Marengo, M. and Picco, M. (2026). A transparent workflow for future EnergyPlus Weather (EPW) files for building energy simulation. Academia Green Energy, 3. View paper
Try it on a live project. FREDS runs a full dynamic EnergyPlus simulation from your browser — describe the building and get validated results in minutes. Start a free simulation.
