PMV/PPD Comfort Index

What is the PMV/PPD model?

The PMV/PPD model is Fanger's heat-balance approach to predicting thermal comfort in buildings. PMV stands for Predicted Mean Vote, a number on a seven-point sensation scale from minus three (cold) through zero (neutral) to plus three (hot). PPD stands for Predicted Percentage Dissatisfied, derived from PMV by an empirical curve that shows the share of occupants expected to feel uncomfortable. The model has become the foundation of European standards (EN 16798-1) and ISO 7730, and it powers the comfort calculations in most building-simulation tools.

The decisive insight that reframes every design conversation: PPD never falls to zero. Even at a perfectly neutral PMV, approximately five percent of occupants remain dissatisfied. This is not a flaw in the model; it is the model telling you honestly that no single setpoint satisfies a full room of people with different clothing, metabolism, and position. Once understood, this shifts client expectations from getting the temperature right to keeping acceptable dissatisfaction low.

How does the seven-point sensation scale work?

The PMV scale runs from minus three to plus three, representing cold through neutral to hot:

Most practical standards restrict comfort to the band from minus one to plus one (cool to warm), where PMV stays close to neutral but dissatisfaction remains manageable.

What are the six comfort model inputs?

PMV combines six factors: four environmental measurements and two personal assumptions.

The last two inputs are where most uncertainty lives. The modeller assumes a single clothing value for the season (which varies by individual and preference) and assumes sedentary metabolic rate (which ignores active occupants). In a real office, clothing and activity vary, making the model's prediction an average rather than a precise forecast.

What does PPD tell us?

PPD (Predicted Percentage Dissatisfied) is a curve mapped from PMV. At a PMV of zero, PPD is approximately five percent: the irreducible minimum. As PMV moves away from neutral in either direction, PPD rises sharply. A PMV of minus one or plus one corresponds to roughly ten to fifteen percent dissatisfied; a PMV of minus two or plus two pushes PPD above thirty percent.

Standards therefore define comfort categories as bands of acceptable PMV that limit PPD to a target threshold. A tight category (Category I in EN 16798-1, suitable for sensitive spaces like hospitals) might require PPD <6 percent, allowing PMV only from minus 0.2 to plus 0.2. A relaxed category (Category III for less-sensitive buildings) might accept PPD <20 percent with a wider PMV band.

How do European standards use PMV?

EN 16798-1 (the current binding standard for building energy performance) defines four thermal-comfort categories. Each category specifies an acceptable PMV range that ensures a maximum dissatisfaction percentage. Category I (highest comfort) is used for sensitive spaces and special activities; Category II applies to normal office, residential, and retail use; Category III accepts wider variation and suits buildings with natural ventilation and individual control. The standard also references adaptive thermal comfort as an alternative for naturally ventilated buildings, reflecting the reality that static PMV overestimates dissatisfaction when occupants can open windows.

What are the model's limitations?

PMV was developed in the 1970s by Ole Fanger, calibrated on steady-state experiments with sedentary subjects in uniform clothing in controlled climate chambers. Three systematic biases emerge in real buildings:

First, the model assumes steady-state conditions (no transient changes in temperature, clothing, or activity), but real occupants experience dynamic comfort as they adjust between outdoor and indoor conditions, remove a jacket, or step into sunlight.

Second, the model systematically overpredicts dissatisfaction in naturally ventilated buildings. When a window is operable and drafts can be felt and addressed, occupants adapt more readily and tolerate wider temperature ranges than PMV predicts. This is precisely why the adaptive thermal comfort model exists; it accepts higher indoor temperatures in summer when outdoor conditions are warm and occupants expect variation.

Third, PMV rests on assumptions (clothing and metabolism) rather than measurements, introducing baseline uncertainty that any single forecast must acknowledge.

How does PMV differ from adaptive comfort?

Adaptive comfort applies Fanger's model to naturally ventilated buildings where occupants control windows, fans, and clothing. It recognizes that occupants in these spaces are more satisfied at a wider range of temperatures: typically 21–26°C in winter and extending to 27–28°C in summer, because they have agency to adapt. PMV alone applies best to climate-controlled buildings where mechanical systems maintain fixed conditions and occupants have little control, making them more sensitive to deviations. For Swedish passive houses and well-insulated new construction with mechanical ventilation, PMV guides design; for retrofit projects with natural ventilation or mixed-mode systems, adaptive comfort is more appropriate.

What about local discomfort?

PMV describes whole-body thermal balance, so a space can score well for overall comfort while cold downdraught from a large window, radiant-temperature asymmetry from a cold exterior wall, or warm floor underfoot makes it locally unacceptable. Standards therefore define separate criteria: draught rate (vertical air velocity <0.2 m/s near occupants), radiant-temperature asymmetry (wall and window surface temperatures within narrow bounds), and floor temperature (16–26°C depending on footwear and season). A room can have a PMV of zero and still fail local-discomfort criteria. In dense urban renovation where single-glazed windows or thin external walls exist, local discomfort often constrains design more tightly than whole-body PMV.