Adaptive Comfort Model: When Indoor Temperature Should Follow Outdoor Climate

Adaptive comfort is easy to misuse.

Many HVAC discussions treat indoor comfort as a fixed temperature target.

For example:

Set the room to 22°C.
Keep the office at 24°C.
Anything above 26°C is too warm.

That kind of thinking may be reasonable for fully air-conditioned buildings.

But it does not always fit naturally ventilated buildings.

In naturally ventilated spaces, occupants can adapt. They may open windows, use fans, adjust clothing, shift expectations, or accept a wider indoor temperature range when the outdoor climate is warmer.

That is the idea behind the adaptive comfort model.

It does not ask only:

“What is the indoor temperature?”

It asks a better question:

“Given the recent outdoor climate, is this indoor temperature still acceptable for occupants in a naturally ventilated building?”

The core formula

The calculator uses the ASHRAE 55 adaptive comfort relationship:

T_comf = 0.31 × T_outdoor + 17.8

Where:

T_comf = neutral indoor comfort temperature, °C
T_outdoor = running mean outdoor temperature, °C

The running mean outdoor temperature is not just a random outdoor temperature from one hour.

It is a recent outdoor temperature average, typically representing the thermal history that occupants have adapted to.

Once the neutral comfort temperature is calculated, the acceptable comfort band is applied.

For 80% acceptability:

Comfort Range = T_comf ± 3.5°C

For 90% acceptability:

Comfort Range = T_comf ± 2.5°C

The 80% band is wider.

The 90% band is stricter.

That means a space may pass the 80% comfort check but fail the 90% comfort check.

Why this matters

Adaptive comfort is useful because it recognizes that comfort is not always fixed.

For example, an indoor temperature of 27°C may feel too warm in a fully air-conditioned office where occupants expect tight temperature control.

But in a naturally ventilated building during warm weather, 27°C may be acceptable if occupants are adapted to the outdoor climate and have some control over their environment.

That does not mean “hot rooms are always fine.”

It means the comfort limit moves with the running mean outdoor temperature.

The model is especially useful for:

Naturally ventilated buildings
Mixed-mode buildings
Passive design studies
Low-energy cooling strategies
Tropical and temperate climates
Early comfort screening
Post-occupancy comfort checks

But it must be applied to the right building type.

Example: naturally ventilated office comfort check

Suppose a naturally ventilated office has:

Running Mean Outdoor Temperature = 28°C
Actual Indoor Temperature = 27°C
Acceptability Level = 80%

Step 1: Calculate the neutral comfort temperature.

T_comf = 0.31 × T_outdoor + 17.8
T_comf = 0.31 × 28 + 17.8
T_comf = 8.68 + 17.8
T_comf = 26.48°C

So the neutral comfort temperature is:

T_comf = 26.48°C

Step 2: Calculate the 80% comfort range.

For 80% acceptability:

Comfort Range = T_comf ± 3.5°C

Lower limit:

Lower Limit = 26.48 − 3.5
Lower Limit = 22.98°C

Upper limit:

Upper Limit = 26.48 + 3.5
Upper Limit = 29.98°C

So the acceptable range is:

22.98°C to 29.98°C

Step 3: Check the actual indoor temperature.

Actual Indoor Temperature = 27°C

Since 27°C is between 22.98°C and 29.98°C, the space is within the 80% adaptive comfort range.

Step 4: Calculate distance from the nearest comfort boundary.

Distance = min(T_indoor − Lower Limit, Upper Limit − T_indoor)

Substitute the values:

Distance = min(27 − 22.98, 29.98 − 27)
Distance = min(4.02, 2.98)
Distance = +2.98°C

The positive result means the indoor temperature is inside the comfort range.

So the result is:

Indoor temperature = 27°C
Status = Within 80% adaptive comfort range
Distance from nearest boundary = +2.98°C

That is a useful result because it prevents an overly rigid comfort judgment.

A fixed-temperature mindset might say:

27°C is too warm.

But the adaptive comfort model says:

For this running mean outdoor temperature and this building type, 27°C can still be acceptable.

What happens with 90% acceptability?

Now keep the same temperatures, but use the stricter 90% acceptability band.

Inputs:

Running Mean Outdoor Temperature = 28°C
Actual Indoor Temperature = 27°C
Acceptability Level = 90%

The neutral comfort temperature is unchanged:

T_comf = 26.48°C

For 90% acceptability:

Comfort Range = T_comf ± 2.5°C

Lower limit:

Lower Limit = 26.48 − 2.5
Lower Limit = 23.98°C

Upper limit:

Upper Limit = 26.48 + 2.5
Upper Limit = 28.98°C

The 90% comfort range is:

23.98°C to 28.98°C

The indoor temperature is still:

27°C

So it remains inside the range.

Distance from the nearest boundary:

Distance = min(27 − 23.98, 28.98 − 27)
Distance = min(3.02, 1.98)
Distance = +1.98°C

The space still passes, but the margin is smaller.

That is the practical difference between 80% and 90% acceptability.

The 90% band is not just a label.

It narrows the acceptable temperature range and leaves less room for drift.

What happens if indoor temperature rises?

Now suppose the indoor temperature rises to 30°C while the running mean outdoor temperature stays at 28°C.

Inputs:

Running Mean Outdoor Temperature = 28°C
Actual Indoor Temperature = 30°C
Acceptability Level = 80%

Neutral comfort temperature:

T_comf = 26.48°C

80% comfort range:

Lower Limit = 22.98°C
Upper Limit = 29.98°C

Now compare:

Actual Indoor Temperature = 30°C
Upper Limit = 29.98°C

The indoor temperature is slightly above the upper comfort limit.

Distance from comfort range:

Distance = Upper Limit − T_indoor
Distance = 29.98 − 30
Distance = -0.02°C

The negative result means the indoor condition is outside the acceptable range.

This is where the model becomes useful for design decisions.

The result is not simply “30°C is always bad” or “30°C is always acceptable.”

The result is:

At this running mean outdoor temperature and acceptability level, 30°C is just outside the 80% adaptive comfort range.

That is a much more precise engineering statement.

Common engineering mistake: using adaptive comfort for fully air-conditioned spaces

The biggest mistake is applying the adaptive comfort model to the wrong building type.

The adaptive model is intended for naturally ventilated buildings where occupants can interact with the environment.

That usually means people can do things like:

Open windows
Use fans
Adjust clothing
Change local airflow
Experience outdoor climate variation
Adapt expectations to seasonal conditions

A sealed, fully air-conditioned office is different.

If occupants have limited control and expect mechanical cooling to maintain a narrow setpoint, adaptive comfort may not be the right model.

For mechanically cooled spaces, PMV/PPD methods or standard HVAC comfort criteria may be more appropriate.

The formula may still produce a number.

But the engineering interpretation can be wrong.

Another mistake: using one outdoor temperature instead of running mean temperature

The input is running mean outdoor temperature.

That matters.

Using a single afternoon peak temperature can distort the result.

Using one random weather value can also distort the result.

The adaptive model is based on recent outdoor thermal history, not one isolated moment.

For example:

Outdoor peak today = 34°C
Running mean outdoor temperature = 28°C

Those are not the same input.

If the engineer uses the peak temperature instead of the running mean value, the comfort temperature shifts too high and the acceptable band may look more permissive than it should.

That can lead to an overly optimistic comfort conclusion.

Another mistake: ignoring humidity and air movement

The basic adaptive comfort formula does not directly include humidity.

It also does not fully describe local air movement, radiant asymmetry, solar exposure, or internal heat gains.

That matters in real buildings.

A room can be inside the calculated adaptive temperature range and still feel uncomfortable if:

Humidity is very high
Air movement is too low
Direct sun hits occupants
Radiant surfaces are hot
Internal equipment gains are high
People cannot open windows
Fans are not available
The space is densely occupied

Adaptive comfort is a powerful screening method.

It is not a full replacement for detailed thermal comfort analysis.

Practical design checks

Before accepting an adaptive comfort result, ask:

1. Is the building naturally ventilated or mixed-mode?
2. Do occupants have real control over windows, fans, or airflow?
3. Is the outdoor input a running mean temperature, not a single peak value?
4. Is the selected acceptability level 80% or 90%?
5. Is the indoor temperature an operative comfort temperature, not only a random air sensor value?
6. Is humidity high enough to create discomfort even inside the temperature band?
7. Are solar exposure and radiant temperature effects important?
8. Is the outdoor running mean temperature within the valid model range?

These checks matter because the adaptive comfort model is simple, but the application is not automatic.

Practical engineering takeaway

The adaptive comfort model changes the way engineers think about comfort.

Instead of treating indoor temperature as one fixed target, it links acceptable indoor temperature to recent outdoor climate.

The main formula is:

T_comf = 0.31 × T_outdoor + 17.8

Then the acceptable range is applied:

80% acceptability: T_comf ± 3.5°C
90% acceptability: T_comf ± 2.5°C

The result helps answer a practical design question:

Can this naturally ventilated space be considered thermally acceptable under these outdoor conditions?

Used correctly, the model can support passive design, mixed-mode operation, wider comfort bands, and lower mechanical cooling energy.

Used incorrectly, it can justify uncomfortable spaces with the wrong formula.

The key is to apply it only where the assumptions match the building.

For a quick first-pass check, you can use the Adaptive Comfort Model Calculator.

It calculates neutral comfort temperature, acceptable comfort range, and distance from the comfort boundary based on running mean outdoor temperature, actual indoor temperature, and the selected ASHRAE 55 acceptability level.