Worked example: room selection for an overheating assessment

Overheating assessment room selection is a judgement about which rooms will fail first, and it is the part of a dynamic thermal model that an officer or a Building Control body is most likely to question. Model the wrong rooms and a scheme passes on paper and overheats in use. The CIBSE TM59 method does not require every room in a block to be simulated, but it does require the sample to capture the worst case, and the reasoning has to be written down.

What follows is an illustrative worked selection for a single residential block. The criteria quoted are those published in TM59 and in Approved Document O; the room choices are mine, to show the method.

Why a sample rather than every room

A 60-flat block might contain 200 habitable rooms. Simulating all of them is possible and occasionally sensible, but it produces a results table nobody reads and it hides the two or three rooms that actually drive the design.

TM59 allows an assessment of the worst-case rooms provided the selection is justified. The logic is that if the most exposed south-west facing single-aspect bedroom on the top floor passes, the sheltered north-facing bedroom two floors down will pass too. The burden is on the modeller to show that the chosen rooms bracket the range.

Where a scheme is being assessed for planning rather than for Building Regulations compliance, the sample also has to be defensible against a reviewer who did not design the building. An Overheating Risk Assessment that lists ten modelled rooms without saying why those ten were picked invites a request for more work.

Orientation, glazing ratio and exposure

Four variables do most of the sorting.

Orientation comes first. South and west facing rooms take the highest solar gains in the afternoon and evening, which is when bedrooms matter. A west-facing bedroom is usually worse than a south-facing one for the night-time criterion, because the peak gain lands shortly before the room is occupied.

Glazing ratio is the second. Express it as glazed area over internal floor area for each room type, and rank them. A studio with a full-height window wall will sit at the top.

Exposure is the third: single-aspect rooms cannot cross-ventilate, and a single-aspect flat with its only facade facing south-west is the classic failure. Dual-aspect and corner flats have options the model can use.

Floor level is the fourth, and it cuts both ways. Top-floor rooms gain through the roof and sit above any overshadowing from neighbouring buildings. Lower floors are shaded but often have less free area available because of security restrictions on openable windows at ground level.

Adjacency to risers and heat networks belongs on the list too. A small single-aspect bedroom next to a communal riser carrying heat network flow and return pipes can fail on distribution losses alone, regardless of glazing.

A worked selection for one block

Take an illustrative eight-storey block, 48 flats, double-loaded corridor, communal heat network, mechanical extract ventilation with openable windows for purge. The selection below is illustrative.

Six rooms, each chosen for a stated reason, covering both criteria and the corridor. A reviewer can disagree with one choice without the whole assessment collapsing.

Bedrooms assessed separately from living spaces

TM59 applies two tests to predominantly naturally ventilated homes, and they are not interchangeable. The first is the adaptive comfort test taken from TM52: for living rooms, kitchens and bedrooms, the number of occupied hours in which operative temperature exceeds the comfort threshold by one degree or more must not exceed 3 per cent of occupied hours between May and September.

The second applies to bedrooms only: operative temperature must not exceed 26 degrees for more than 1 per cent of annual hours between 10pm and 7am. Bedrooms therefore need the night-time profile modelled, with the occupancy and internal gains TM59 specifies rather than a generic template.

For homes that rely on mechanical ventilation and cannot open windows, the fixed-temperature approach applies instead, with all occupied rooms held below 26 degrees for all but 3 per cent of annual occupied hours. Where a scheme mixes strategies between blocks, say which block uses which and why.

Approved Document O sets its own modifications to the TM59 method for Building Regulations compliance, including the weather files to be used and limits on how far openable windows can be relied on where external noise or security constrains them. A planning-stage model built on different assumptions will need revisiting, so flag that in the submission rather than letting it surface at the start of construction. The sequence in the note on the first month after planning permission is a useful prompt for when to pick that up.

Communal corridors have their own criterion

Corridors are routinely left out of planning submissions and routinely come back as a condition. TM59 treats them separately because they are unoccupied circulation space with heat network distribution losses, no openable area in most designs, and a fire strategy that limits what can be done about it.

The published corridor test is 28 degrees for no more than 3 per cent of total annual hours. Hitting it in a block with a communal heat network usually means one or more of: pre-insulated pipework with higher specification, a lower network flow temperature, mechanical ventilation of the corridor, or moving the pipework out of the corridor entirely.

That last option is a design change, not a modelling tweak, which is why the corridor result should be produced early. The top floor corridor at the end of the longest riser run is the room to model.

Ventilation assumptions and the security constraint

The free area a model assumes is where most disputes land. A model that opens every window fully, day and night, on a ground-floor flat facing a public footpath is not describing how anyone will live.

State for each modelled room: the openable free area, the window type and restrictor position, whether night purge is assumed, and the justification. Where an acoustic constraint limits opening, the noise assessment and the overheating model have to agree with each other, and a discrepancy between the two is the easiest thing for a reviewer to spot. The same applies to the daylight evidence, because reducing glazing to fix overheating can push a room below the targets in a Daylight and Sunlight Assessment.

Internal blinds are not mitigation unless the model controls them on a defensible schedule, and external shading, deeper reveals, solar control glazing and balcony overhangs all do more. Keep the glazing specification consistent with the Energy Statement, because a g-value that differs between the two documents undermines both.

Writing down the selection

The assessment should carry a short room selection section that a reviewer can audit without opening the model. It needs the total number of habitable rooms by type, the variables used to rank them, the shortlist with a one-line reason each, the criteria applied to each room, and a statement that unmodelled rooms are less exposed on every variable used.

Add the weather file and location, the modelling software, the occupancy profiles, and the internal gains source. And say plainly where the sample is thin. A reviewer who can see the gap you have acknowledged is far easier to satisfy than one who finds it.

One honest uncertainty: councils differ on whether they want planning-stage overheating evidence at all, and some rely wholly on Building Regulations. In London, policy requires heat risk to be managed at planning stage, so the assessment is expected with the application. Check the local list before deciding scope.

This is general guidance, not legal advice; check the position with your local planning authority or a qualified professional before relying on it. Dynamic thermal modelling is specialist work and councils expect a suitably qualified author, so professional review is strongly advised.

PlanScope provides AI-assisted draft reports for planning applications. Reports should be reviewed before submission. We do not guarantee planning approval.