Overheating assessments get done against a fixed rulebook. The weather does not read it.
On 11 August 2026 the Met Office published provisional statistics showing the UK on course for its warmest summer on record for mean temperature. It’s worth reading the numbers properly, because they’re the backdrop to every Part O conversation you’ll have this autumn.
What the Met Office actually published
From 1 June to 10 August 2026, the UK mean temperature stood at 16.48°C — 1.88°C above the 1991–2020 average. The existing UK record is 16.12°C, set in 2025. To merely equal that record, the rest of August would have had to average 0.21°C below the long-term summer average. Anything warmer sets a new record, in a series that starts in 1884.
England and Wales are further ahead again:
| Area | Mean temp (1 Jun – 10 Aug 2026) | Difference from average | Current record |
|---|---|---|---|
| UK | 16.48°C | +1.88°C | 16.12°C (2025) |
| England | 18.19°C | +2.46°C | 17.44°C (2025) |
| Wales | 16.80°C | +2.17°C | 16.09°C (2025) |
Met Office senior scientist Mike Kendon noted this followed the second-warmest June and second-warmest July on record, with 38°C reached in June and 35°C in both late May and July. The 1976 record for average maximum temperature was also described as under threat.
Two things are worth separating here. A single hot summer is weather, not a methodology change. But the direction is consistent — and 2025 held the previous record, which tells you something about how quickly the reference points are moving.
Meanwhile, the compliance position hasn’t moved
Here’s the awkward part for anyone designing homes right now.
Approved Document O is still the 2021 edition. When MHCLG published the Future Homes Standard package on 24 March 2026, Part O was pulled out of it. Instead of an updated AD O alongside the new Parts L and F, the government announced a standalone full review of Part O, reflecting the volume of issues raised at consultation. That review has no published date, and because Part O sits outside the Future Homes Standard transitional window, it does not automatically share the 24 March 2027 in-force date that applies to Parts L and F.
The dynamic modelling methodology has moved on without it. CIBSE published a substantially revised TM59:2026 in July — a passive-first three-stage method with four assessment criteria and a rewritten bedroom night-time test. Approved Document O still points at the 2017 edition. If you run a dynamic assessment, you now have to be explicit about which edition you’ve used and why, because your Building Control body may not assume the same one you did.
Planning has moved too. Since 17 August 2026, NPPF policy CC3 requires development proposals to use design approaches which minimise risks from overheating. That’s a decision-making policy applying at planning — years ahead of any revised AD O.
So the practical position is a widening gap: the evidence base is getting hotter, planning has an overheating test, the modelling methodology has been rewritten, and the statutory guidance is five years old.
What this means on a live job
A few things worth doing now rather than in a year’s time:
- Don’t treat the simplified method as free. The Part O simplified method was written against a different climate baseline. Where a plot is borderline — south or west glazing, single aspect, urban site with restricted openable area — a dynamic assessment gives you evidence you can defend at both planning and Building Control.
- State your TM59 edition. Put it in the report front sheet. If you use TM59:2026, say so; if you use the 2017 edition because AD O references it, say that too. Ambiguity is what gets queried.
- Deal with overheating at planning, not after. Under CC3 a planning officer can now ask the question, and retrofitting a fix after permission — deeper reveals, external shading, changed window schedules — is expensive and often unpopular with the elevations.
- Watch the interaction with Part L. The FHS package already recognises this: where air conditioning or reduced g-value glazing is installed to comply with Part O, the Part L 2026 standards account for the resulting energy use, so you aren’t penalised twice for the same design decision. That’s a change in your favour — use it.
- Remember airtightness cuts both ways. Tighter fabric under FHS makes purge and night ventilation strategies more important, not less. A home that can’t dump heat overnight is a home that fails a night-time criterion.
The honest summary
Nothing in the Met Office release changes a single input in your model this week. What it changes is the risk profile of the assumption that a marginal pass today will still look marginal in ten years — and the likelihood that when the Part O review does land, it lands harder rather than softer.
If you’ve got a scheme where overheating is close to the line, that’s the one to model properly now.
If you’ve got a plot where overheating is close to the line — flats, single aspect, constrained urban sites, or anywhere a planning officer has already asked the question under CC3 — get in touch and we’ll tell you how we’d approach the Part O assessment and what evidence is worth putting in front of Building Control.
Sources: Warmest UK summer on record “increasingly likely” as temperatures stay well above average, Met Office, 11 August 2026 · The Future Homes and Buildings Standards — written statement, UK Parliament, 24 March 2026 · Government announces full review of Overheating Regulations, Elmhurst Energy, 9 April 2026 · National Planning Policy Framework, August 2026 (PDF), GOV.UK — policy CC3