Predicted mean vote and predicted percentage dissatisfied are the two indices at the centre of the analytical comfort method in Standard 55. This page explains what they measure, the heat balance reasoning behind them and where the method reaches its practical limits.
ASHRAE Standard 55, published by ASHRAE, a registered mark of that society, is currently issued as Standard 55-2023, Thermal Environmental Conditions for Human Occupancy. It provides methods for judging whether combinations of environmental and personal factors are likely to produce acceptable thermal conditions for a population of occupants. Its analytical method does not reduce comfort to a thermostat setting. Instead, it treats thermal sensation as the outcome of heat exchange between the human body and its surroundings.
Predicted mean vote, usually abbreviated to PMV, estimates the average thermal sensation vote that would be expected from a large group of people exposed to the same conditions. The word "predicted" matters because the result comes from a model rather than from a direct statement by every occupant. The word "mean" matters because individual responses are averaged. A person may feel cooler or warmer than the calculated central tendency without either the person or the calculation being wrong.
The model places thermal sensation on an ordered scale running from sensations of cold through neutrality to sensations of heat. A PMV result indicates where the average response is expected to lie on that scale. It is not itself an air temperature, a comfort-zone boundary or a percentage. It is a compact expression of the body's calculated thermal balance after the relevant environmental and personal inputs have been considered together.
That distinction is particularly important in Gulf buildings, where a single room temperature can coexist with markedly different radiant conditions, air movement and clothing choices. Two spaces showing the same dry-bulb reading may therefore produce different PMV results. Equally, changing air speed or the radiant environment may alter the predicted sensation even where the thermostat set-point remains unchanged.
Predicted percentage dissatisfied, abbreviated to PPD, converts the PMV result into an estimate of the proportion of a population likely to express thermal dissatisfaction. It does not identify which people will be dissatisfied, nor does it promise that the remaining occupants will report comfort. It expresses the statistical spread of human responses around the predicted average sensation.
The relationship between PMV and PPD reflects the fact that thermal preference is not uniform. Even when the average response is close to neutral, some people prefer a cooler environment and others prefer a warmer one. The model therefore does not imply that a condition can satisfy every occupant. Instead, it provides a reproducible basis for comparing design options, reviewing operating conditions and identifying whether a proposed environment falls within the analytical method used by the standard.
PPD should also not be confused with the percentage of complaints recorded by a facilities team. Complaint data are influenced by expectations, reporting culture, seating location, duration of exposure and the availability of personal control. The calculated index and observed feedback answer related but different questions. One predicts a population response under stated assumptions; the other records experiences in a particular building.
The analytical method begins with the body's need to maintain thermal equilibrium. Human metabolism generates heat. Some of that heat is converted into external work, but most must be released to the surroundings. Heat is exchanged by convection with the surrounding air, radiation with nearby surfaces, evaporation from the skin and respiration. Comfort becomes more likely when the body can lose heat at a rate consistent with its production without requiring strong physiological responses.
PMV uses this balance to estimate the thermal load on the body. A positive imbalance indicates that the body is tending to retain more heat than is comfortable, while a negative imbalance indicates excessive heat loss. The model then relates that calculated load to the average sensation expected from a population. This is why the method must consider more than temperature: each route of heat exchange can change independently.
The calculation is a model of steady or near-steady exposure. It assumes that the relevant conditions and activities can be represented by stable inputs over the period being assessed. Rapid transitions, highly variable work, direct solar exposure, intermittent equipment heat and strong local jets can make a single set of inputs less representative. In such circumstances, the calculation may still be informative, but its assumptions need to be stated rather than hidden.
The six principal inputs are air temperature, mean radiant temperature, air speed, humidity, clothing insulation and metabolic rate. Air temperature influences convective heat exchange. Mean radiant temperature represents the combined thermal effect of surrounding surfaces, including glazing, walls, ceilings and equipment. Air speed changes convective and evaporative heat loss. Humidity affects evaporative cooling, although its influence on sensation is often smaller than that of temperature or radiation under ordinary indoor conditions.
Clothing insulation describes the resistance that garments provide to heat loss. More insulating clothing generally shifts the balance towards warmth, while lighter clothing permits greater heat loss. Metabolic rate represents the heat generated by activity. A seated person carrying out light office work and a person performing physically demanding work cannot be assessed using the same activity assumption simply because they occupy the same room.
The inputs interact. Higher air speed may offset a warmer operative condition; a cooler radiant environment may make the same air temperature feel less warm; and a change in clothing may alter the expected sensation without any change to the building system. The model's value lies in combining these effects consistently. Its weakness appears when the selected inputs do not represent the people or conditions being assessed.
A whole-body PMV result does not capture every way in which an occupant can feel uncomfortable. Standard 55 therefore treats local thermal discomfort as a separate consideration. A person may have an acceptable overall heat balance but still experience an unpleasant draught at the neck, excessive warmth at the feet, a cool downdraught near glazing or one-sided radiant heating from a hot surface.
Draught concerns unwanted local cooling caused by air movement, particularly where the flow is intermittent, turbulent or directed at sensitive parts of the body. Vertical air temperature difference concerns stratification between lower and upper parts of the occupied zone. Floor surface temperature addresses discomfort transmitted through the feet. Radiant asymmetry concerns unequal radiant exposure, such as a warm ceiling, cool window or strongly heated wall.
These mechanisms are not interchangeable with PMV. Increasing the room temperature to address a local draught may make the wider space too warm, while lowering the set-point to respond to solar discomfort may overcool occupants away from the façade. Diagnosis therefore requires attention to the spatial pattern of conditions, the location of occupants and the type of complaint rather than reliance on a single central reading.
Clothing insulation is one of the most difficult inputs to estimate. Occupants may wear business clothing, traditional dress, uniforms or layered casual clothing in the same zone. Garments can also be adjusted during the day. A calculation based on a nominal ensemble may therefore describe no individual exactly. A sensible assessment records the assumed clothing basis and considers whether the population is sufficiently uniform for one value to be representative.
Metabolic rate presents a similar difficulty. Job title is not a reliable substitute for observed activity. A person may alternate between seated work, walking, lifting and standing. The model requires an estimate of the activity during the assessed period, so a mixed task may need separate scenarios or a time-weighted interpretation rather than one convenient label.
Instrumentation and spatial selection also matter. Mean radiant temperature cannot always be inferred from air temperature, especially near external glazing or large internal heat sources. Air speed may vary substantially across a workstation. Measurements taken in an unoccupied central position may miss the conditions at the person's chair. Airflow measurement, system balancing and commissioning methods belong to a separate building-services resource and are not reproduced here.
The standard is consequently best understood as a structured decision method. It makes assumptions visible, combines relevant variables and produces indices that can be compared consistently. It does not replace occupant feedback, engineering judgement or investigation of local conditions. A calculated result is only as credible as the measurements, observations and assumptions entered into it.
Dubai Municipality's Technical Guidelines for Indoor Air Quality for Healthy Life, DM-HSD-GU119-IAQ Version 4 of 11 December 2024, section 9-8-5 Table 3, gives dry bulb 22.5 to 25.5 degrees Celsius, relative humidity 20 to 60 per cent and air movement 0.2 to 0.3 metres per second, to be delivered for 95 per cent of the year. Those figures are the guideline's own thermal and physical comfort ranges; they should not be presented as the comfort-zone data of Standard 55.
The distinction prevents two common errors. The first is to treat a local guideline range as though it reproduces the analytical comfort boundaries in the standard. The second is to assume that remaining inside a dry-bulb range proves that PMV, PPD and local discomfort requirements have been addressed. A temperature range is a useful operational reference, but the analytical method requires the full set of environmental and personal inputs.
Edition years matter here more than they usually do, because UAE instruments reference these standards in two incompatible ways. Al Sa'fat, Dubai's green building system, whose Silver Sa'fa requirements are mandatory for new buildings, requires the latest edition of Standards 62.1, 62.2 and 170 and deliberately attaches no year, so the applicable requirements move as the standards are revised. Estidama's Pearl Building Rating System, Version 1.0 of April 2010, does the opposite: it names specific editions — the 2007 edition of Standard 62.1, the 2007 edition of Standard 62.2, the 2004 edition of Standard 55 and the 2007 edition of Standard 90.1 — so the requirements it imposes are frozen at those editions regardless of what has been published since. Dubai Municipality's Technical Guidelines for Indoor Air Quality for Healthy Life, Version 4 of 11 December 2024, references Standard 62.1 without giving an edition year. Anyone reading a requirement should therefore establish which instrument imposes it and whether that instrument names a year, before establishing what the current edition says.
Al Sa'fat §401.01 (2nd edition, January 2023) requires the latest edition and attaches no year — Estidama PBRS Version 1.0 (April 2010) names fixed editions
No. PMV is a predicted average thermal sensation derived from heat-balance inputs. Air temperature is only one of those inputs. Radiant temperature, air speed, humidity, clothing and activity can change the result even when the thermostat reading stays constant.
No. The model describes the expected average response of a population and recognises that individual preferences vary. PPD exists because a neutral average does not remove the spread of responses around that average.
It can contribute to the investigation, but a whole-zone PMV result may miss local radiant asymmetry, downdraught or a local air jet. The assessment should examine the actual workstation and the relevant local discomfort mechanism.
They directly affect body heat production and heat loss. An apparently precise PMV result can be misleading when the assumed clothing or metabolic rate does not represent the occupants. Stating the assumptions allows the result to be reviewed and tested against real use.
This is an independent information resource. It is not affiliated with, endorsed by, or connected to ASHRAE. ASHRAE and the names of its standards are the trademarks of their respective owners and are used here only to identify the standards described.