CIBSE TM59: How Overheating Risk in Homes Is Assessed

CIBSE TM59 is the design-stage methodology for assessing overheating risk in dwellings. It uses dynamic thermal models to test how a home will behave through a summer, rather than relying on a single worst-case temperature. The current edition, TM59 Overheating risk in dwellings: a design stage methodology, was published by CIBSE in July 2026 and runs to 39 pages¹.

The method matters because it is the route by which new homes in England demonstrate that they will not overheat. TM59 (2017) is still required for compliance with Part O, the overheating regulation for new homes¹. The 2026 edition is recommended by CIBSE for aligning with industry best practice, but the 2017 version remains the one that building control recognises for Part O purposes¹.

The criteria are specific. For bedrooms in naturally ventilated homes, TM59 sets a threshold of 26°C: a bedroom may only be warmer than 26°C for 1% of the year's sleeping hours, equivalent to 32 hours in any year between 10pm and 7am². Living rooms, kitchens and bedrooms are assessed separately against a comfort temperature under Criterion A⁵.

A consistent methodology: what TM59 standardises

Before TM59, overheating assessments in homes were inconsistent. Different assessors used different weather files, different internal gain assumptions and different comfort thresholds, so two homes with the same design could produce different verdicts. TM59 standardises the method: it specifies the modelling approach, the weather data and the pass or fail criteria, so that a result means the same thing whoever produces it.

The methodology uses dynamic thermal models rather than steady-state calculations¹. A dynamic model simulates the home hour by hour across a whole summer, accounting for thermal mass, solar gain through glazing, internal gains from occupants and appliances, ventilation rates and the way the building fabric responds to changing conditions. That is a heavier exercise than a simple check, but it captures the behaviour that matters: a home that is comfortable at 3pm may be unbearable at 11pm if it cannot purge stored heat.

TM59 places a strong emphasis on optimising passive design from the outset¹. The assessment is intended to shape the design, not to be a final hurdle. If a model shows a bedroom failing the criterion, the response is to change the design: add external shading, increase ventilation openings, adjust glazing ratios or orientation. The methodology also includes a three-stage overheating assessment strategy, and the 2026 edition adds updated ceiling fan modelling¹.

The scope is broad. TM59 is applicable to houses, apartments, sheltered accommodation, care homes, student accommodation, prisons, and bedrooms in hotels and hospitals¹. All homes are tested under unconstrained conditions, meaning the model does not assume occupants will necessarily open windows or deploy blinds in a particular way¹.

For a household, the practical value is that a home assessed under TM59 has been tested against a defined standard rather than a designer's judgement. That is a form of independence from the weather: the home's behaviour in a hot summer has been modelled before it was built, not discovered afterwards.

The overheating criteria: 26°C for no more than 3% of occupied hours

TM59 applies two distinct criteria, and a home must pass both. They test different things: Criterion A covers the daytime comfort of occupied living spaces, while Criterion B covers sleep.

Criterion A applies to living rooms, kitchens and bedrooms. It requires that the internal temperature does not exceed a defined comfort temperature by 1°C or more for more than 3% of occupied hours over the summer period, which runs from 1 May to 30 September⁵. The comfort temperature is not a fixed number; it adapts to the running mean outdoor temperature, which is why the criterion is expressed as an excess over a comfort temperature rather than a flat threshold.

Criterion B applies to bedrooms only. It requires that the internal temperature between 10pm and 7am shall not exceed 26°C for more than 1% of annual hours⁵. In the Welsh Government's factsheets, this is expressed as bedrooms being allowed to exceed 26°C for 1% of the year's sleeping hours before they are considered unsuitable, which is 32 hours in any year between 10pm and 7am².

The two criteria can produce different results on the same dwelling. Scottish Government analysis of overheating risk found that both dual-aspect flat typologies passed Criterion B for bedrooms, but both living rooms failed Criterion A, albeit to a limited extent⁵. That pattern is instructive: a home can be fine for sleeping and still be too warm in the living room during the day, or the reverse.

The 2026 edition of TM59 introduces a new bedroom overheating criterion¹. The 26°C and 1% figures above are the thresholds published in the Welsh Government factsheets and the Scottish analysis, and they remain the reference points for the 2017 methodology that Part O relies on.

Category I limits for vulnerable occupants

TM59 is not a single pass or fail for every occupant. The methodology recognises that some people are more vulnerable to heat than others, and it provides for stricter limits where the occupants are known to be at higher risk. This is the Category I approach: a tighter comfort band applied to dwellings designed for vulnerable occupants.

The categories matter because the standard criteria are calibrated for a general population. Older people, very young children and people with certain health conditions regulate heat less effectively, and a bedroom that is merely warm for a healthy adult can be dangerous for them. Where a scheme is designed for sheltered accommodation or care homes, the assessment is expected to reflect that.

The scope of TM59 explicitly includes sheltered accommodation and care homes alongside houses, apartments, student accommodation, prisons, and bedrooms in hotels and hospitals¹. That breadth is deliberate: the methodology is not confined to mainstream housing, and the settings where occupants cannot easily adapt, such as care homes or prisons, are exactly where a stricter limit is warranted.

For a household, the practical point is that the criteria applied to a home depend on who is expected to live in it. A general-needs dwelling is assessed against the standard thresholds. A dwelling designed for vulnerable occupants is assessed against tighter ones. The same modelling engine produces both results; only the limit changes.

This connects to the wider question of who is at risk in a heatwave. The health consequences of indoor heat are not evenly distributed, and the design-stage assessment is one of the few points at which that is formally recognised before people move in.

Dynamic thermal modelling: how the assessment is done

The assessment is carried out using dynamic thermal models¹. In practice, a consultant builds a computer model of the dwelling, room by room, and simulates its thermal behaviour across a full year or a defined summer period using a standard weather file.

The model needs a great deal of input: the construction of walls, roofs and floors, the glazing type and area, the orientation of each facade, the internal gains from occupants and appliances, the ventilation strategy including window openings and any mechanical ventilation, and the shading from neighbouring buildings or overhangs. Each of these affects the result, which is why the methodology specifies them rather than leaving them to the assessor.

The dynamic thermal analysis method uses CIBSE TM59, Design methodology for the assessment of overheating risk in homes⁵. In London, residential developments undertake dynamic overheating modelling in line with the guidance and data sets in CIBSE TM59 and TM49 respectively, taking into account the associated Approved Document O requirements⁶. Non-residential developments use CIBSE TM52 and TM49 instead⁶.

The output is a report. The building control body should be provided with a report that demonstrates that the residential building passes CIBSE's TM59 assessment of overheating⁷. That report is the evidence that the design meets the requirement, and it is what a building control officer reviews.

For a household, the modelling stage is invisible: it happens before the home is built or converted. But it determines the shading, the window openings and the ventilation strategy that the finished home will have. A home that passes on paper should, in the weather the model assumed, stay within the criteria.

Climate scenarios: DSY1 and what 'passing' means

The weather file used in the model is as important as the building itself. TM59 uses Design Summer Year files, and the choice of file determines how hard the test is.

DSY1 is the Design Summer Year for the 2020s, high emissions, 50th percentile scenario⁶. It is intended to represent a moderately warm summer, covering 2011 to 2040⁵. It is the mildest of the available scenarios, and that is the crux of the criticism: CIBSE TM59 allows homes to pass using the mildest climate scenario, DSY1⁸.

The 2026 edition of TM59 addresses this by requiring 2050s weather data¹. That is a significant change. A home modelled against 2050s data is being tested against a hotter climate than one modelled against DSY1, and the difference in results can be substantial.

There is a second scenario in use. DSY2 represents 2003, a year with a very intense single warm spell⁶. The two files test different things: DSY1 tests a moderately warm summer, DSY2 tests a short, extreme heat event. A home that passes on DSY1 may still fail on DSY2, and the choice of file is therefore a policy decision as much as a technical one.

The Welsh Government's factsheets make the gap explicit. Research by Hayles et al. (2022) projected that the percentage of hours that would exceed CIBSE guidelines for thermal comfort would be far higher and more dangerous than the TM59 allowance³. That research modelled six locations across Wales over 22 July to 31 August³.

For a household, the meaning of a pass depends on the weather file behind it. A pass on DSY1 is a pass against a moderate summer. A pass on 2050s data, as the 2026 edition requires, is a stronger statement¹.

Reporting requirements and validation on real projects

TM59 is not a theoretical standard. It has been applied to real schemes, and the results have been mixed, which is itself useful evidence about how demanding the criteria are.

The Scottish Government's overheating risk analysis used dynamic thermal analysis following the CIBSE TM59 (2017) methodology⁵. The findings were that both dual-aspect flat typologies passed Criterion B for bedrooms, but both living rooms failed Criterion A, albeit to a limited extent⁵. This is a real project result, not a modelled hypothetical, and it shows that the two criteria can diverge sharply within the same dwelling.

The reporting requirement is straightforward. The building control body should be provided with a report demonstrating that the residential building passes the TM59 assessment⁷. In London, the energy assessment guidance requires residential developments to undertake dynamic overheating modelling in line with CIBSE TM59 and TM49, taking into account Approved Document O requirements⁶.

The 2017 edition remains the compliance route. TM59 (2017) is still required for compliance with Part O¹. The 2026 edition is recommended by CIBSE for aligning with industry best practice, but a designer working on a Part O submission needs the 2017 methodology¹.

For a household buying a new home, the practical implication is that the home should have been assessed, and the assessment should have produced a report. Whether that report used the 2017 or 2026 edition depends on the regulatory route the developer took.

TM59 in practice: Welsh dwelling climate research and future development

Wales has produced some of the most detailed public research on how TM59 thresholds compare with projected real-world conditions, and the findings are a caution about relying on a single pass.

The Welsh Government's factsheets on summertime overheating in highly insulated homes draw on research by Hayles et al. (2022), which modelled six locations across Wales over 22 July to 31 August³. The research projected that the percentage of hours that would exceed CIBSE guidelines for thermal comfort would be far higher and more dangerous than the TM59 allowance³. In other words, the TM59 threshold, applied to the Welsh climate under projected conditions, would be breached far more often than the methodology's pass would suggest.

That finding does not invalidate TM59. It shows that the standard is a design-stage tool calibrated to a particular climate scenario, and that the gap between a pass and real-world comfort widens as the climate warms. The Welsh research is one of the clearest public statements of that gap.

Wales is also revising its building regulations. The 2025 consultation version of Approved Document L Volume 1: Dwellings is for use in Wales⁹. The Welsh Government's earlier Part L review consultation proposed referencing the BRE guide for the design of low temperature Domestic Heating Systems (FB 59) within the final Approved Document L, and including references such as BS 7593 (2019) and CIBSE/ADE CP1 Heat Networks: Code of practice for the UK (2015)¹⁰. Those are heating and heat network references rather than overheating ones, but they show the direction of travel: a more integrated set of standards for Welsh dwellings.

Scotland has its own route. The Scottish Government's building standards policy sets out proposed changes to energy standards and associated topics⁵, and the new build heat standard factsheet covers the move away from direct emissions heating¹¹. Scotland's eBuildingStandards service is free¹². Northern Ireland publishes energy performance certificate statistics separately¹³.

For a household, the territorial picture matters. TM59 is a UK-wide methodology, but the regulatory route that requires it differs: Part O in England, the Welsh Approved Document L, Scottish building standards, and Northern Ireland's own arrangements. The criteria are the same; the compliance path is not.