Overheating: Specifying and Reporting to TM59, TM52 and Part O

Which rule applies: Part O, TM52 or TM59?

Part O applies to new residential buildings in England and requires TM59 for its dynamic route. TM52 is the adaptive comfort method for non-domestic spaces, with BB101 for schools and HTM 03-01 for healthcare.

Part O (Approved Document O, 2021 edition, on gov.uk) applies to new residential buildings in England and has been in force since 15 June 2022. TM59 is the CIBSE methodology that Part O requires for its dynamic route, and TM52 is the adaptive comfort method for non-domestic buildings. The three are connected but are not interchangeable, and a specification that cites the wrong one for the building type will produce the wrong overheating report.

Requirement O1 covers a dwelling, an institution or another building containing residential rooms, other than hotel rooms. It requires that reasonable provision is made to limit unwanted solar gains in summer and to provide an adequate means of removing excess heat. O1(2) says account must be taken of safety and enjoyment, and mechanical cooling should be used only where insufficient heat can be removed without it (paragraph 2.11), so passive measures come first.

TM52 is the adaptive thermal comfort basis that CIBSE has adopted for non-domestic overheating assessment, and TM59 builds on it for homes. A mixed-use building needs both: TM59 for the residential parts and TM52 or the client criteria for the commercial parts. State that in the specification and the reporting clause, so that each model is run against its own criteria. Part O sits alongside Part L: see the Part L 2026 guide for services engineers and the Avoice guide to Approved Document O.

When can the simplified method be used and when is dynamic modelling needed?

The simplified method uses ADO tables for glazing and free area by orientation, cross-ventilation and location risk. Dynamic modelling to TM59 is needed where the tables cannot be met or the method is restricted.

Approved Document O gives two routes to compliance: a simplified method, which is a set of tables for limited glazing and free area, and a dynamic thermal modelling method following TM59. The simplified method is quick to specify and check, and it is restricted in the situations that set its limits. The dynamic method gives a design-specific answer and is the usual route for larger or higher-risk schemes.

Simplified method values depend on cross-ventilation, orientation and whether the location is high or moderate risk. High-risk locations are the London postcodes listed in Appendix C. Cross-ventilation means openings on opposite façades. Free area means the geometric open area, calculated with a discharge coefficient of 0.62 (Appendix D). The simplified method is restricted where there is communal heating, so check the conditions in the Approved Document.

Table 1.3 sets minimum free area for ventilation openings. For high-risk locations it is the greater of 6% of the floor area or 70% of the glazing area, with 13% of floor area for bedrooms. For moderate-risk locations it is the greater of 9% of the floor area or 55% of the glazing area, with 4% for bedrooms. In high-risk locations, paragraph 1.9 also requires shading: external shutters with ventilation, glazing with a g-value of 0.4 or less and light transmittance of 0.7 or more, or overhangs with a 50 degree cut-off on due south façades.

The simplified method is attractive early, because it needs no model. It also gives the architect clear limits from the start. If the design cannot meet the tables, the dynamic method is the next step, so find out at RIBA Stage 2 rather than at Stage 4.

What do TM59 and TM52 require, and what should the report contain?

TM59 sets criteria for occupied hours above the comfort range and for bedroom night hours. TM52 uses three criteria, and a room fails on two. The report should list weather file, assumptions, results and mitigation.

The dynamic method must follow CIBSE TM59 (ADO paragraphs 2.3, 2.4 and 2.6). TM59 sets two criteria for naturally ventilated homes: no more than 3% of occupied hours above the comfort range, and for bedrooms no more than 32 night hours above 26 degrees Celsius, which is 1% of annual hours. The sample of units modelled should include the highest-risk ones. TM59 is dated 2017.

Part O fixes several modelling assumptions. No internal blinds or curtains are allowed (paragraph 2.8), no foliage (2.9), and fixed shading counts (2.7). Daytime window opening is gradual between 22 and 26 degrees. Bedroom windows are open all night if the temperature is above 23 degrees at 11pm. Ground-floor or easily accessible rooms are closed at night, and entrance doors are shut. The weather file is a point of care: the CIBSE position statement refers to 2020s medium emission DSY1, while a later CIBSE presentation refers to 2020s high emissions, 50th percentile DSY1. Check the weather file required by the current Approved Document and CIBSE guidance, and name it in the specification.

TM52 criteria

TM52 uses three criteria: He, no more than 3% of occupied hours with a temperature difference of 1 K or more above the adaptive limit; We, a daily weighted exceedance of no more than 6; and Tupp, an upper limit of no more than 4 K above the limit. A room fails if it fails any two or more of the criteria.

How do the results flow into the specification?

Every mitigation assumed in the model, including glazing g-value, shading, free area, restrictors, MVHR bypass and pipework insulation, needs a clause or schedule entry so it cannot be lost in value engineering.

An overheating result is only useful if each mitigation measure assumed in the model appears in the specification. Glazing g-value, shading, free area, restrictor openings, MVHR summer bypass and pipework insulation are the usual ones. A measure that exists only in the model report can be value engineered out without anyone noticing.

Section 3 of Part O adds constraints that limit how openings can be used. If noise in a bedroom exceeds 40 dB LAeq,T over eight hours or 55 dB LAFmax more than 10 times a night, windows may be assumed closed between 11pm and 7am. Secure openings only are assumed on ground-floor bedrooms, such as lockable louvres or grilles. Where the model assumed open windows at night and the acoustic report says they cannot be, the specification needs a different solution, and the acoustician and the overheating modeller should have exchanged those results before issue.

Regulation 40B requires information to be given to the owner within 5 days of completion, and Part O Section 4 together with the Home User Guide includes the staying cool in hot weather advice. Add that to the O&M and handover clauses: see the commissioning, soft landings and O&M guide.

How should overheating be coordinated across the design team?

Architect, services engineer and acoustician share the decisions. Check simplified limits at Stage 2, model at Stage 3, finalise at Stage 4 and re-assess substitutions through change control.

Overheating is solved by decisions that belong to several people: glazing and shading by the architect, ventilation by the services engineer and noise limits by the acoustician. The specification and reporting have to join these up, and the best time to do it is early, because the strongest levers, orientation, glazing ratio and shading, are fixed by RIBA Stage 3.

Where the contractor designs part of the glazing or ventilation, the overheating model must be re-run on the contractor's proposals. Write that into the performance specification and the contractor's design portion requirements: see the guide to contractor design portions. For mixed-use or higher-risk schemes, the model forms part of the compliance evidence in the golden thread.

The government has said it will carry out a full technical review of Approved Document O, with no timetable published at the time of writing. Cite the edition in the specification so that a later revision triggers a review, and check ADO before each new project.

Questions engineers ask about overheating

When did Part O come into force?

Approved Document O in its 2021 edition was published on 15 December 2021 and came into force on 15 June 2022. It applies to new residential buildings in England.

What is the difference between TM52 and TM59?

TM52 is the CIBSE adaptive comfort method used for non-domestic buildings. TM59 is the methodology for homes that Part O requires for its dynamic modelling route.

What are the TM59 criteria?

For naturally ventilated homes, TM59 allows no more than 3 percent of occupied summer hours above the comfort range, and bedrooms no more than 32 night hours above 26 degrees Celsius, which is 1 percent of annual hours.

Can I use the simplified method in Part O?

Yes, where its conditions are met. It sets glazing and free area limits by orientation, cross-ventilation and location risk, and it is restricted where there is communal heating. Otherwise use the dynamic method.

Can mechanical cooling be used to comply?

Only where insufficient heat can be removed without it. Part O requires passive measures to be used first and asks that account is taken of safety and enjoyment.

Which weather file does the dynamic method use?

CIBSE guidance has referred to 2020s DSY1 files, and the emissions scenario named differs between CIBSE documents. Check the current Approved Document and CIBSE guidance when the model is commissioned.

Where Avoice fits

Overheating results fail between the model and the specification: a g-value in the report that does not match the glazing clause, or a bypass in the model that the ventilation section never mentions.

Avoice writes building services specifications alongside architectural ones, so glazing, shading and ventilation clauses can be checked in the same workspace. The office master holds the practice's overheating reporting clause as a governed library entry, and practice rules can require that the standard and weather file are named. Avoice MEP and structural specifications are structured to CAWS or Uniclass 2015, the same as architectural specifications.

The Avoice Regulations library includes Approved Document O, and the Building Regs Tracker records which requirements apply to a project. The pre-issue Technical Review compares the specification against schedules and drawings, so a g-value that differs between the window schedule and the specification is flagged before issue. Software retrieves, compares and flags. The services engineer and the modeller decide the strategy and the result.