Approved Document O and its two compliance routes

Approved Document O compliance can be shown in one of two ways, and the choice you make at the start of a residential project shapes the elevations. Route one is the simplified method, a set of tabulated limits on glazing area and minimum opening free area. Route two is dynamic thermal modelling to CIBSE TM59. They do not always give the same answer for the same building, and switching between them late is expensive.

Part O of the Building Regulations came into force in June 2022 and applies to new residential buildings. Requirement O1 asks for reasonable provision to limit unwanted solar gains in summer and to provide adequate means of removing heat from the indoor environment. Two obligations, in that order, and the order matters because the document expects you to reduce the gain before you deal with the heat.

What Part O applies to

New residential buildings: dwellings, residential institutions and buildings containing rooms for residential purposes such as student accommodation and many forms of supported housing.

It does not apply to work on an existing building, and it does not apply on a material change of use. That gap is the reason an office converted to flats can be completed with no overheating check under the Building Regulations at all, even where the same flats built from new would fail the simplified method comfortably. Anyone relying on a conversion route should understand that planning policy, not Part O, is what will ask the overheating question.

Nothing in Part O removes the need for a planning-stage assessment where local policy requires one. In London in particular, the cooling hierarchy in the London Plan and the need for a TM59 assessment at application stage run alongside the Building Regulations, and an Overheating Risk Assessment prepared for planning is a separate document from the compliance evidence given to building control.

The simplified method and its limits

The simplified method sets maximum glazing areas and minimum free areas for openings, expressed as proportions of floor area, and varies them according to whether the building is in a location identified as higher risk and whether the room is cross-ventilated or single aspect. Look up the tables, check your glazing and your openable area, and if both comply the requirement is met.

Its attraction is speed and certainty. Its limits are real.

It gives no credit for external shading from neighbouring buildings, overhangs or brise soleil beyond what the tables assume. It gives no credit for thermal mass. It treats a single aspect south facing flat harshly, which is correct in principle and awkward in practice on a constrained urban site. And the free area requirement often collides with guarding, means of escape, security and acoustic requirements, so the opening that satisfies the table may not be buildable.

Single aspect flats with generous glazing are where the method usually fails. At that point the choice is to redesign, reduce the glazing, or move to modelling.

The dynamic modelling route and TM59

The modelling route requires a dynamic thermal model of the building assessed against the criteria in CIBSE TM59, using the design summer year weather file for the nearest location. For homes without mechanical cooling, TM59 sets two criteria: a limit on the proportion of occupied hours in living areas where the adaptive comfort threshold is exceeded across the summer months, and a separate and tighter limit on night-time hours above 26 degrees in bedrooms.

The bedroom criterion is the one that fails schemes. Daytime comfort in a living room can usually be managed with shading and purge ventilation. A bedroom that cannot shed heat overnight, because the window cannot be opened or because the room retains the day's gain, fails regardless of what happens at noon.

Modelling gives you credit for the things the simplified method ignores: g-values, external shading geometry, adjacent massing, ventilation strategies including mechanical ventilation with a summer bypass, exposed thermal mass and realistic internal gain profiles. It also exposes assumptions to scrutiny, so model inputs need to match the specification that is actually being built. Where the assessment relies on a blind, a shutter or a specific glazing product, that item becomes part of the compliance evidence.

Noise, security and the night purge problem

The provision that catches urban schemes is the interaction between ventilation and noise. Where the overheating strategy depends on windows being open at night, and external noise means they will be closed, the strategy does not work. Approved Document O addresses this by reference to internal noise levels in bedrooms at night, based on an average level and a maximum level for individual events, and where those would be exceeded with windows open the assessment has to be carried out with windows closed.

The consequences are immediate. A closed window assessment needs an alternative means of removing heat, which in practice means mechanical ventilation with a bypass, secure background ventilation through an acoustically attenuated route, or mechanical cooling as a last resort.

That is why the acoustic and the overheating work have to be done together. A Noise Impact Assessment that establishes facade levels and recommends closed window mitigation, produced after the overheating model was run on open windows, means both reports need doing again. On a site next to a main road or a railway, commission them as one package.

Security and guarding pull the same way. A ground floor bedroom window sized for night purge has to be secure, and an upper floor opening has to comply with guarding requirements, so the free area available is often less than the elevation suggests.

Evidence at completion

Part O is checked at completion as well as at design stage. The building control body will expect the as-built overheating strategy: the openings actually installed with their free areas, the glazing specification and g-values, the shading installed, and any mechanical ventilation or cooling provided.

Information also has to be given to the owner explaining how the strategy is meant to be operated, because a design that depends on opening windows at night depends on someone doing it. That document is easy to forget and awkward to produce retrospectively.

Two practical habits help. Record the free area calculation for each opening type on a schedule during design, and check the product data for the units actually ordered, since a window with a restrictor fitted does not deliver the free area of the same window without one.

Where the two routes give different answers

The simplified method and the modelling route are not calibrated to each other, and either can be the more generous depending on the building.

The pattern is that modelling helps where the site itself moderates the risk and hurts where the internal arrangement is the problem. A courtyard block in a city, overshadowed on three sides, is a modelling case. A detached house with south facing bedrooms under a warm roof is not going to be rescued by a model.

Choose the route at concept stage, and say which one you are using in the Energy Statement so that planning and building control evidence line up. Where a planning condition requires a TM59 assessment, the modelling route is effectively chosen for you.

Reforms of this kind are easier to track in one place, and the planning law timeline is where updates are gathered. For the wider procedural changes affecting residential schemes, see the notes on the 2026 NPPF structure and on Building Safety Regulator gateways.

This is general guidance, not legal advice; check the current edition of Approved Document O and the position with your building control body or a qualified professional before relying on it. PlanScope provides AI-assisted draft reports for planning applications. Reports should be reviewed before submission. We do not guarantee planning approval.