What Is TM59? The Two Overheating Tests, Explained

If you have been handed a TM59 report and told Plot 14 failed, you need two things: to know what was measured, and to know what to change. This covers both.

What TM59 actually is

CIBSE TM59 is a design methodology for the assessment of overheating risk in homes. It is not legislation. It is the method Approved Document O names when you take the dynamic thermal modelling route instead of the Simplified Method.

The useful thing about TM59 is how much it takes off your hands. It fixes the occupancy profiles, the internal heat gains and the window-opening behaviour. That matters because those assumptions are where general thermal modelling gets subjective: two modellers can reach two different answers about the same building if they each invent their own. TM59 removes that argument.

What is left is a description of the dwelling and an hourly simulation across a Design Summer Year.

The daytime test: hours of exceedance

The daytime test is inherited from CIBSE TM52 and applies to living rooms, kitchens and bedrooms.

> For occupied rooms, the number of hours during which the operative temperature exceeds the adaptive comfort threshold must not exceed 3% of occupied hours between May and September.

The important word is adaptive. The threshold is not a fixed number. It moves with the running mean outdoor temperature, on the basis that people tolerate warmer rooms during a warm spell and expect cooler ones after a cool week.

Daytime failures usually point at solar gain. Too much glass, the wrong orientation, or nothing shading it.

The night-time test: bedroom temperatures

The night-time test applies only to bedrooms and is absolute.

> The number of hours between 10pm and 7am during which the temperature exceeds 26°C must not exceed 1% of annual night-time hours.

No adaptive sliding scale. 26°C is 26°C.

This is the one that catches developments. A bedroom can clear the daytime test comfortably and still fail the night-time one, and when it does the cause is almost always that the room cannot shed the day's heat overnight.

Both must pass

There is no averaging and no trading one off against the other. Every assessed room has to pass the daytime test, and every bedroom has to pass the night-time one.

Reading a failure

This is the part most reports leave you to work out.

• The daytime test fails, the night-time test passes. A solar gain problem. Look at glazing area, g-value and shading on the offending elevation before you touch anything else.
• The night-time test fails, the daytime test passes. A night ventilation problem. The room gets rid of heat too slowly. Look at openable free area, whether the window has to be modelled closed for security, and whether there is any cross-ventilation to be had.
• Both fail. Usually a single-aspect dwelling with significant west or south glazing. Expect to change more than one thing.

The second case is worth dwelling on, because the instinctive fix, shrinking the windows, is the wrong one. Smaller windows reduce the gain the daytime test measures. They do not help a room that cannot purge heat at 2am, and they make the home darker for the sake of a test it was already passing.

Where this bites in practice

The limits that reduce your effective openable area are the ones people forget:

• Ground-floor bedrooms and other easily accessible rooms that are unoccupied at night have to be modelled with windows closed, for security.
• Fall protection, restrictors and guarding all cut equivalent area.
• Noise and air quality can rule out an open window on a particular facade.

Each of those is reasonable on its own. Stacked up on a single-aspect flat beside a main road, they are why the development needs mechanical purge ventilation.

The practical lesson

Run the assessment while the design can still change. A night-time failure found at RIBA Stage 2 is a conversation about ventilation strategy. The same failure found after the facade package is issued is a conversation about smaller windows.

AutoTHERMAL runs TM59 on the EnergyPlus engine with both tests assessed automatically. [Join the waitlist](/waitlist) for early access.