The adaptive approach
The adaptive approach to thermal comfort starts, not from a consideration of the heat exchange between man and the environment, but from the observation that there are a range of actions that we can and does take in order to achieve thermal comfort. An adaptive principle is at work stating that:
If a change occurs such as to produce discomfort, people react in ways that tend to restore comfort
The centre of our temperature regulation is the temperature of the brain. This is used to control the equilibrium between us and the environment using physiological responses or behavioural actions which tend to maintain the brain temperature within close limits. If a change occurs, in the environment or elsewhere, causing the brain temperature to deviate from the close limits, physiological responses occur. If these are not enough to prevent discomfort then an action may be taken which will tend to restore it to these limits.
The types of behavioural action which can be taken are:
* modification of the internal heat generation
* modification of the rate of body heat loss
* modification of the thermal environment
* selecting a different environment
Modifying the internal heat generation can be achieved unconsciously with raised muscular tension or, in a more extreme situation, the shivering reflex, or consciously, for instance through jumping about in the cold to increase metabolic heat or having a siesta in the warm to reduce it.
The rate of body heat loss can be changed unconsciously through vasoregulation or sweating or consciously by such actions as changing ones clothing , cuddling up or by taking a cooling drink.
Modifying the thermal environment can be achieved through lighting a fire, opening a window (controls), or in the longer term by insulating the loft or moving house.
Selecting a different environment can be achieved within a room by moving closer to the fire or catching the breeze from a window, between rooms in the same house with different temperatures or by moving house or visiting a friend.
These are only examples of the actions which can be taken, and if we are always free to take the necessary actions then thermal discomfort should not be a problem.
Constraints
There are usually constraints which limit our ability to take actions to avoid discomfort. For instance climate, cost and fashion. If we have no direct control over the environment (for instance when the heating engineer sets the temperature for everyone) this can increase the likelihood of discomfort. In addition many of the actions we could take to improve comfort have a distinct time limitation - building a new house, changing clothing, visiting a friend and so on all need time to complete. Actions are also limited in how successful they can be: removing a garment can only compensate for a limited change in temperature.
In the dynamic relationship with the environment, constraints are the key to deciding whether a particular temperature is comfortable, and whether comfort can be achieved.
The implication of the adaptive principal is that given sufficient time, people will find ways in which to adapt to any temperature so long as it does not pose a threat of heat stroke or hypothermia. Discomfort will arise where temperatures:
* change too fast for adaptation to take place
* are outside normally accepted limits
* are unexpected
* are outside individual control
Changes which take place too fast will not allow preventative action to be taken. Within a single day the change may be too much to be compensated for by clothing changes. The variation in the environment is the cause of discomfort. It is important to make the distinction between imposed variations which can cause discomfort and chosen variation which can reduce it. Chosen variations may be the way in which the person keeps comfortable. Changes can even cause pleasure - see any beach on a sunny day when people dive in the sea to cool off after sun bathing.
The range of temperatures which any society normally encounters is limited. A temperature outside normal experience may mean that people become uncomfortable because they do not have the experience to deal with such conditions. To define the comfort limits these normal limits need to be known.
In any particular situation people have expectations about the thermal conditions they will meet. These are usually based on experience. The success of the thermal strategy they take will depend on the accuracy of this prediction. Arriving for a concert, for instance, one might find the hall colder than expected and be unable to get comfortable. Discomfort would then be unavoidable - if one wanted to hear the concert. Note here that it is not only the strategy of the concertgoer which is at fault but also that of the hall manager.
In any situation an individual will have a desired temperature. If the individual cannot control the temperature (and if they can’t move to a better location) then discomfort can occur. One of the problems for the environmental engineer is to deal with a multi-occupied space such as an office or a factory, where there are variations in the needs of different occupants. The heat-balance model of thermal comfort answers this by providing the temperature with the lowest PPD for the assumed clothing and activity. The adaptive model favours some measure of control being at the disposal of each individual.
Evidence of adaptation
Whilst the individual processes of adaptation are complex, the form of the process is that of a feed-back system. Because of this the end results can seem very simple.
Comfort temperatures reflect average temperatures
First consider the effect of adaptation on the comfort temperature. Given time and opportunity people will try to suit themselves to the average temperature they experience (note that some of the adaptation is expressed in the temperature itself, some in the person's adaptation to it). So we would expect the comfort temperature to be close to the average temperature they experience. Using the results of field surveys, Humphreys (1981) [10,5] has shown this to be very close to the real situation. The results of his survey are shown in Fig 4.1
Figure 4.1. the variation of comfort temperature with the mean temperature experienced in a number of surveys (after Humphreys)
The points in figure 4.1 each represent a field survey of thermal comfort. In each case Humphreys has calculated the comfort temperature predicted by the survey. He has plotted these against the average temperature experienced by the subjects in the survey. The slope of the regression line in fig 1 is close to unity. Further analysis shows a difference in the temperature which occupants of heated and cooled buildings find comfortable compared to those in buildings with no heating or cooling. Humphreys discusses this in his paper. Nevertheless the relationship between average comfort temperature and average temperature experienced is clear to see.
Another effect which can be predicted from the adaptive process is the independence of the average comfort vote from the average temperature over a period of time. The average comfort vote might be expected to be close to neutral, but the mean temperature to have little effect on it. Fig 4.2 demonstrates the small effect of the average temperature on the average comfort vote. With very large variations of temperature, the range may fall outside the adaptive range of the population. Extreme temperatures will then give rise to discomfort where the adaptive opportunity is not sufficient (Nicol and Raja 1997).
Clothing level reflects temperature.
The fact that adaptation is taking place will give rise to observable effects.
In temperate climates one of the most powerful of adaptive processes open to an individual is the use of clothing. This is reflected in the changes people make to their clothing at different temperatures. Humphreys (1972) made use of this relationship in his work on school children. He was able to deduce the comfort temperature from observations of the number of children in minimum clothing in a class. The close relationship between clothing and classroom temperature was remarkable.
The clothing changes were more closely related to a moving-average of the temperatures they had experienced in the immediate past than they were to the temperature at the time of the observation. In other words the clothing changes were taking time to reflect the temperature changes. In a similar way he was able to show (Humphreys 1978a ) that clothing changes from day to day had a half-life of about a day - again people were taking time to adapt to changing conditions. For a fuller discussion of such effect see Nicol (1992).
Similar
effects have been found among office workers in Pakistan and the UK. (Nicol
and Raja 1999), McCartney
and Nicol 1998). Here clothing changes were as well correlated with
outdoor as with indoor temperature, and changed with outdoor temperature even
among subjects in air conditioned buildings with a constant indoor temperature.
Indoor temperatures
are affected by adaptive measures.
The designer rarely has accurate knowledge of the indoor temperatures which a population will experience. Outdoor conditions are more easily obtained from meteorological data. In a development of his work correlating comfort temperature with indoor temperature, Humphreys (1978) correlated comfort temperature with outdoor temperature and indoor temperature with outdoor temperature for surveys conducted in free-running buildings. The results are shown in figure 4.3. The relationship between mean indoor (Tm) and outdoor (To) temperatures are particularly interesting as evidence of adaptive action taken by building occupants. The equation for the line of best fit is:
Tm = 0.55To + 14.1 (4.1)
The indoor temperature is changing at only half the rate of the outdoor temperature. At temperatures in excess of about 31oC (at which point Tm = To) the indoor temperature is below outdoor temperature, at temperatures below this figure indoor temperature exceeds outdoor temperature by an increasing amount.
|
|
| This graph shows how the mean comfort temperature varies with the mean indoor temperature. Each point in the graph is the mean value for a whole survey. | |
Even in buildings with no heating or cooling plant, the occupants' efforts to achieve a comfortable environment reduces changes in indoor temperatures below those in outdoor temperature. They achieve cooling in hot conditions and warming in cool conditions.
Comfort temperature reflects design temperature in heated and cooled buildings.
Humphreys (1978) found that comfort temperature in heated and air-conditioned buildings were about 2K higher than the world average in American studies and 2K lower than average in the Soviet Union. UK studies showed comfort temperatures 1K lower than average. This might seem to be a confirmation of the correctness of design temperatures since it reflected a difference in standards. However there is no real reason why these differences between one population and another should occur except that they are differences in expectation. If American workers go to the office expecting 23oC and European workers expect 19oC then each will dress to suit this expectation. There is a cultural element to the difference (American workers expect to work in shirt sleeves, European workers in a suit, or a pullover) but the standards were driving the comfort temperature rather than the other way around. In other words static comfort standards work and continue to be used because people adapt to them.
Setting Comfort standards using the adaptive model.
Comfort standards based on adaptive assumptions will be more than simply a temperature to aim at. The standard will need to reflect the interactions between comfort and environment in its formulation. Such concepts as predictability, constraints, variety, adaptive opportunity and control will need to be incorporated into the standard.
To start with the `comfort temperature' which we define as the temperature at which there is the least probability of discomfort, or at which satisfaction with the environment is most likely. The value of the comfort temperature will vary at the very least with climate and season. The value of the comfort temperature in free-running buildings can be deduced from a graph such as that shown as figure 4.4. Humphreys (1978) found that the best outdoor temperature predictor for the comfort temperature was the mean of the monthly mean minimum and the monthly mean maximum temperatures. The prediction was improved significantly by the inclusion of a term for the annual maximum temperature, but we do not know why this should be, but it suggests a climatic effect is involved.
 |
|
Effect of outdoor temperature on indoor neutral (comfort) temperature The indoor comfort temperature also changes with outdoor temperature especially in buildings which are free-running (neither being heated nor cooled) - filled circles and Line A |
In a building that is not free-running the comfort temperature is decided by social and economic factors and only slightly by climatic ones. For instance people in America and Europe have a different comfort temperatures for broadly comparable populations. So the comfort temperature in such circumstances will require research among the local population. These variations occur not just between different populations, but within the same population between economic or social groups. Such variations in comfort temperature are more difficult to formalise, and can be explained as the need to provide variety and control so that people can choose for themselves.
The comfort temperature is not the only temperature which people can find comfortable. Clearly there are allowable variations around it which will not cause discomfort. The amount of variation allowable will be time-dependent. This is because the longer people have to adapt the further they can change without significantly increased discomfort. Thus we might find that ±2K was the maximum allowable within-day variation with a maximum within-week variation of, say, ±5K. Dynamic temperature standards would change the way in which the designer investigates a building. The dynamic thermal characteristics of free-running buildings as well as the steady-state characteristics would be incorporated in design.
Another factor which needs clarification is the variability of temperature (and other factors) within a room. A model which seeks to explain thermal comfort needs to take in to account the variations in conditions within a space, and the constraints on the ability of the occupants to make use of this variability. Many existing models of room temperature assume a single `room characteristic temperature' without defining how it might vary from place to place within the room. In conditions where people are able to move around, variability may be a key factor in user satisfaction.
Research needed
There is a need for research on several fronts before such temperature standards can be formulated.
Modelling the adaptive processes. We tend to talk of the adaptive `model' of thermal comfort, as if it were a defined thing in the same way as the heat balance model. It is amenable to such an analysis, though many of the parameters will be more obviously probabilistic than they are in the heat balance model. However the work of defining and calibrating the model will require a considerable amount of research. It will probably require fairly sophisticated time-series treatment to define the ways in which adaptation takes place.
Relating to the heat exchange model - facts needing investigation are listed in section 3.5 above: the dynamic nature of clothing, metabolic rate and people's interaction with the environment all need investigation, as do people's social and psychological attitudes to their thermal environment.
The calibration of the model requires wide ranging work in the field. In one sense the adaptive model is (or will be) immensely complex. In another, because it is a feed-back system, its outcomes are remarkably simple: the relation between comfort and mean temperature, the limits within which comfort can be achieved over different time-periods, are more amenable to definition than a complex function dependent on clothing and metabolic rate. The problem is that whole new ways of looking at and thinking about our thermal environment is the starting point in the process.
APPENDIX
Adaptive actions
Some actions in response to cold:
Vasoconstriction (reduces blood flow to the surface tissues)
Increasing muscle tension and shivering (generates more heat in the muscles)
Curling up or cuddling up (reducing the surface area available for heat loss)
Increasing the level of activity (generates body heat)
Adding clothing (reduces the rate of heat loss per unit area)
Turning up the thermostat or lighting a fire (usually raises the room temperature)
Finding a warmer spot in the house or going to bed (select a warmer environment)
Visiting a friend or going to the library (hoping for a warmer environment)
Complaining to the management (hoping someone else will raise the temperature)
Insulating the loft or the wall cavities (hoping to raise the indoor temperatures)
Improving the windows and doors (to raise temperatures/reduce draughts)
Building a new house (planning to have a warmer room temperature)
Emigrating (seeking a warmer place long-term)
Acclimatising (letting body and mind become more resistant to cold stress)
Some conceivable actions in response to heat:
Vasodilation (increases blood flow to surface tissues)
Sweating (evaporative cooling)
Adopting an open posture (increases the area available for heat loss)
Taking off some clothing (increases heat loss)
Reducing the level of activity (reduces bodily heat production)
Having a beer (induces sweating, and increases heat loss)
Drinking a cup of tea (induces sweating, more than compensating for its heat)
Eating less (reduces body heat production)
Adopting the siesta routine (matches the activity to the thermal environment)
Turning on the air-conditioner (lowers the air-temperature)
Switching on a fan (increases air movement, increasing heat loss)
Opening a window (reduces indoor temperature and increases breeze)
Finding a cool spot or visiting a friend (hoping for a cooler temperature)
Going for a swim (selects a cooler environment)
Building a better building (long term way of finding a cooler spot)
Emigrating (long-term way of finding a cooler place)
Acclimatising (letting body and mind adjust so that heat is less stressful)