7
Evaluating comfort with varying temperatures: a graphic design tool John Martin Evans Research Centre Habitat and Energy, Faculty of Architecture, Design and Urbanism, University of Buenos Aires, Pabello ´n 3, Piso 4, Ciudad Universitaria 1428, Argentina Abstract This paper considers the need to define comfort of indoor and outdoor spaces in relation to the daily variations of temperature. A graphical tool is presented, which indicates the daily swings of temperature, shown as a single point on a graph representing the average temperature and the maximum temperature swing. This point can be compared with the comfort zones for different activity levels, such as sedentary activity, sleeping, indoor and outdoor circulation according to the design proposals for different spaces. The graph allows the representation of climatic variables, the definition of comfort zones, the selection of bioclimatic design resources and the evaluation of indoor temperatures, measured in actual buildings or obtained from computer simulations. The development of the graph is explained and examples given with special emphasis on the use of thermal mass. # 2003 Published by Elsevier Science B.V. Keywords: Graphical tool; Comfort zones; Bioclimatic design; Temperature variation 1. Introduction The analysis of conditions required for thermal comfort and the selection of bioclimatic design resources involves many variables. Various graphic design tools have been developed for evaluating these conditions using charts that allow a visualisation of the conditions with two variables, typically temperature and relative or absolute humidity. Givoni’s psychometric chart (1967) and Olgyay’s biocli- matic chart [1] are probable the best known and most widely used examples and are still widely used today. The corrected effective temperature (Bedford, 1940) [2] and equatorial comfort index (Webb, 1960) [3] and predicted 4 h sweat rate use dry bulb (or globe) temperature and wet bulb tempera- ture as the defining variables for evaluating hot and humid conditions. For outdoor conditions, the wind chill index and index of thermal sensation use air temperature and wind speed to define an equivalent temperature. The incorporation of solar radiation as a variable in comfort indexes is less developed, due to the emphasis on indoor climate in many of the indices. The mean radiant temperature, more appropriate to indoor conditions is used more frequently in comfort indices such as the operative temperature (Winslow, et al., 1937) and predicted mean vote [4]. Most of these comfort indices tend to stress the steady- state conditions, often based on studies in climate chambers, rather than actual conditions found in outdoors and in many indoor spaces. These laboratory-based studies of thermal comfort used unvarying steady-state conditions. Question- naires used to obtain comfort votes usually relate to specific conditions at a fixed point in time rather than a response to the evolution of temperature [5]. In practice, the sensation of comfort responds to the changing conditions that are experienced during each day. These variations occur due to climatic, architectural and cultural factors as descriptions follow. (i) The periodic temperature variations vary in different climatic regions, so typical hot dry and continental climates have high temperature ranges, while warm humid climates and cold climates have lower swings. (ii) The swing of internal temperatures will also vary according to the thermal characteristics of the build- ing. Those with significant solar heat gains will have larger swings than those in well-insulated buildings shaded from direct sunlight. High-mass buildings will have lower swings than lightweight construction. (iii) The daily rhythm of activities will include the journey to work, exposed to the outdoor temperature and conditions in transport vehicles, the conditions in the office, factory or other work place and the return journey with different outdoor temperatures. Energy and Buildings 35 (2003) 87–93 E-mail address: [email protected] (J.M. Evans). 0378-7788/03/$ – see front matter # 2003 Published by Elsevier Science B.V. PII:S0378-7788(02)00083-X

Evaluating comfort with varying temperatures: a graphic design tool

Embed Size (px)

Citation preview

Page 1: Evaluating comfort with varying temperatures: a graphic design tool

Evaluating comfort with varying temperatures:a graphic design tool

John Martin EvansResearch Centre Habitat and Energy, Faculty of Architecture, Design and Urbanism, University of Buenos Aires,

Pabellon 3, Piso 4, Ciudad Universitaria 1428, Argentina

Abstract

This paper considers the need to define comfort of indoor and outdoor spaces in relation to the daily variations of temperature. A graphical

tool is presented, which indicates the daily swings of temperature, shown as a single point on a graph representing the average temperature and

the maximum temperature swing. This point can be compared with the comfort zones for different activity levels, such as sedentary activity,

sleeping, indoor and outdoor circulation according to the design proposals for different spaces. The graph allows the representation of climatic

variables, the definition of comfort zones, the selection of bioclimatic design resources and the evaluation of indoor temperatures, measured in

actual buildings or obtained from computer simulations. The development of the graph is explained and examples given with special emphasis

on the use of thermal mass.

# 2003 Published by Elsevier Science B.V.

Keywords: Graphical tool; Comfort zones; Bioclimatic design; Temperature variation

1. Introduction

The analysis of conditions required for thermal comfort

and the selection of bioclimatic design resources involves

many variables. Various graphic design tools have been

developed for evaluating these conditions using charts that

allow a visualisation of the conditions with two variables,

typically temperature and relative or absolute humidity.

Givoni’s psychometric chart (1967) and Olgyay’s biocli-

matic chart [1] are probable the best known and most widely

used examples and are still widely used today. The corrected

effective temperature (Bedford, 1940) [2] and equatorial

comfort index (Webb, 1960) [3] and predicted 4 h sweat rate

use dry bulb (or globe) temperature and wet bulb tempera-

ture as the defining variables for evaluating hot and humid

conditions.

For outdoor conditions, the wind chill index and index of

thermal sensation use air temperature and wind speed to

define an equivalent temperature. The incorporation of solar

radiation as a variable in comfort indexes is less developed,

due to the emphasis on indoor climate in many of the indices.

The mean radiant temperature, more appropriate to indoor

conditions is used more frequently in comfort indices such

as the operative temperature (Winslow, et al., 1937) and

predicted mean vote [4].

Most of these comfort indices tend to stress the steady-

state conditions, often based on studies in climate chambers,

rather than actual conditions found in outdoors and in many

indoor spaces. These laboratory-based studies of thermal

comfort used unvarying steady-state conditions. Question-

naires used to obtain comfort votes usually relate to specific

conditions at a fixed point in time rather than a response to

the evolution of temperature [5].

In practice, the sensation of comfort responds to the

changing conditions that are experienced during each day.

These variations occur due to climatic, architectural and

cultural factors as descriptions follow.

(i) The periodic temperature variations vary in different

climatic regions, so typical hot dry and continental

climates have high temperature ranges, while warm

humid climates and cold climates have lower swings.

(ii) The swing of internal temperatures will also vary

according to the thermal characteristics of the build-

ing. Those with significant solar heat gains will have

larger swings than those in well-insulated buildings

shaded from direct sunlight. High-mass buildings will

have lower swings than lightweight construction.

(iii) The daily rhythm of activities will include the journey

to work, exposed to the outdoor temperature and

conditions in transport vehicles, the conditions in the

office, factory or other work place and the return

journey with different outdoor temperatures.

Energy and Buildings 35 (2003) 87–93

E-mail address: [email protected] (J.M. Evans).

0378-7788/03/$ – see front matter # 2003 Published by Elsevier Science B.V.

PII: S 0 3 7 8 - 7 7 8 8 ( 0 2 ) 0 0 0 8 3 - X

Page 2: Evaluating comfort with varying temperatures: a graphic design tool

Adjustments in clothing and activity levels are possible

within certain limits, though a complete change of clothing

to respond to different environmental conditions is unusual,

except in extreme cases such as workers in refrigerated

storage or blast furnaces, participants in sports activities, etc.

Recent studies of thermal comfort have challenged the

emphasis on fixed conditions or limited temperature varia-

tions for defining thermal comfort (Nichol, 1995 [6]; [7]).

There is a clear evidence that users accept wider temperature

variations in buildings with limited or no artificial condi-

tioning, than in buildings with thermostatically controlled

temperatures (Humphreys, 1975 [8]; [9]; Nichols and Roaf,

1998 [10]).

At the same time, wider temperature swings in indoor

spaces also achieve significant energy savings with eco-

nomic and environmental benefits. Wider temperature

ranges may also allow a reduction in heating and air con-

ditioning plant size.

The adoption of passive methods of temperature control in

buildings also requires a certain range of temperature varia-

tion. Solar gains will increase the indoor temperature, allow-

ing heat storage in high heat capacity materials used for indoor

surfaces. The higher the temperature swings the greater the

storage capacity of renewable energy. This range is limited by

the comfort requirements, though higher limits are possible in

intermittently used spaces such as sunspaces and glazed

circulation galleries than in conventional indoor spaces.

The increasing ease of use of numerical simulation allows

reliable data on hourly internal temperature variations. For

the analysis and comparison of these hourly temperature

outputs from thermal simulation models, the range is more

important than the individual temperatures at a specific time.

Measurements in existing buildings can be made using

miniature data loggers, which provide data on hourly tem-

perature variations, which also require simple methods of

comfort analysis for evaluating the effectiveness of design

alternatives or the definition of remedial passive or active

methods of improving indoor conditions.

This paper therefore presents a method to evaluate tem-

perature variations in relation to comfort requirements,

stressing the typical changes in thermal conditions through-

out the day.

2. Application

This paper presents a simple graph to define and compare

periodic temperature variations, which facilitates the follow-

ing tasks.

(i) Climatic data such as the monthly mean maximum and

minimum temperatures, representing the typical daily

variation of temperature in a month can be presented

and compared with desirable conditions.

(ii) Comfort zones can be defined in relation to different

activity levels and periodic temperature variations.

(iii) Indoor temperature variations can be compared and

evaluated, using measured values or data obtained

from thermal simulations.

(iv) Appropriate bioclimatic design strategies can be

selected to obtain desirable modifications of the

varying external temperatures to achieve comfortable

indoor conditions.

The graph shows the average daily temperature on the

horizontal axis and the average temperature swing on the

vertical axis (Fig. 1). A single point on the graph therefore

represents the variation of temperature during the day,

combining the average temperature and range.

3. Temperature ranges

Many researchers have presented graphs showing the

relation between the percentage of the population experien-

cing comfort and the effective temperature or the comfort

vote. These graphs have a typical curve reaching a maximum

of 85–90% comfortable or ‘satisfied’ and a reduction in the

percentage as temperatures drop or rise. Olgyay [1] shows a

series of curves for inhabitants of different towns, where the

range of temperature that produces a 10% drop in the

population experiencing comfort ranges from 3 to 4.5 8F(or 1.7–2.5 8C). The graph of predicted percentage dissatis-

fied (PPD) proposed by Fanger is expressed in terms of the

comfort vote ([11]; the method presented in this standard was

originally developed by P.O. Fanger). With a range from�0.5

to þ0.5 8C, slightly cool to slightly warm, the percentage

dissatisfied ranges from 5% to a maximum of 10%. This range

corresponds to a temperature variation of 3.8 8C for sedentary

activities, while an increased range, from �1 to þ1 allows a

variation of 8 8C. The range of�2 8C is recommended for the

comfort range by [7]. The range of comfort temperatures

expressed as standard effective temperatures range of 3.4 8C,

from 22.2 to 25.6 8C (Nishi and Gagge, 1977 [12]). Thus,

there is a fair degree of consensus that the comfort range for

sedentary activities is between 3.4 and 4 8C, except for the

earlier studies by Olgyay. In all cases, the level of clothing and

physical activity is assumed to remain constant.

However, in practice, activity levels, posture and clothing

can be adjusted to promote of maintain comfort. As the

temperature rises during the day, jackets or jerseys can be

removed, in more extreme conditions ties can be loosened,

shirtsleeves rolled up and neck scarves untied. The degree of

adaptation will depend of cultural traditions, employer’s

expectations and fashion. Ref. [9] shows that the temperature

of indoor comfort neutrality can vary from 18 to 28 8C as the

monthly mean outdoor temperature ranges from 2 to 30 8C.

Humphreys presents studies that show a variation of thermally

neutral temperatures ranging from 17 to 33 8C. If the extreme

values are discarded, the range becomes 18–30 8C, for

monthly mean outdoor temperatures between 10 and 32 8Cfor free-running buildings without artificial cooling.

88 J.M. Evans / Energy and Buildings 35 (2003) 87–93

Page 3: Evaluating comfort with varying temperatures: a graphic design tool

This variation of thermal neutrality is not inconsistent

with the narrower range of temperatures arising from the

comfort zone range or detailed models such as Fangers PMV

[13]. ISO 7730 can be used to show that temperatures

between 18 and 28 8C are within the comfort range of

PMV from �0.5 to þ0.5, when changes in activity rates,

dress and air movement are considered, consistent and

acceptable for sedentary office activity. Indeed the PMV

model was used as the main basis for defining the comfort

limits established in Section 4.

Field studies of acceptable temperature variations also

show similar variations. Busch [14] reports thermally accep-

table environments (90% voting between �1 and þ1) of 24–

26.5 8C for air conditioned buildings and 23.5–30 8C for

naturally ventilated office buildings in Bangkok. Karyono

[15] shows actual comfort votes between –1 and þ1 for

temperatures between 23.5 and 29 8C in offices in Indonesia.

In the cooler climate of Sydney, Australia, thermal prefer-

ences of 20–25 8C are reported for offices, though air

conditioning is not switched on until temperatures reach

26 8C. In England, office temperatures as low as 19 8C have

been proposed, based on field survey results [16].

4. Comfort zone

The ‘dynamic’ comfort zone is defined on this graph by a

four-sided rhomboid, which shows the combination of aver-

age temperatures and temperature ranges that fall within the

comfort range. The limits for sedentary activities, zone A in

Fig. 1, are based on the criteria described in Sections 4.1–4.4.

4.1. Minimum temperature

For sedentary office work, a minimum temperature of

18 8C is proposed, based on the recommendations and

results of Givoni [17], Fanger [13] and others. This is a

minimum acceptable temperature with no temperature var-

iation. Lower temperatures could be acceptable, but only

with increased levels of clothing insulation, which may not

be considered appropriate in many cultural situations or

practical for sedentary activities.

4.2. Increase in average temperature with increasing

temperature range

If temperature swings are present, as they are in most

buildings, a higher average temperature is needed to main-

tain comfort. With a sinusoidal daily variation of tempera-

ture, for 1 8C increase in the temperature range, the average

temperature must be increased by 0.5 8C. Thus the edge of

the comfort zone forms a diagonal line as one side of the

comfort triangle.

4.3. Maximum temperature range

A very large temperature range is also undesirable as the

adjustment of clothing, posture and activity levels have

Fig. 1. The comfort triangle with zones for sedentary comfort (A), right comfort (B), indoor circulation (C) and outdoor circulation (D).

J.M. Evans / Energy and Buildings 35 (2003) 87–93 89

Page 4: Evaluating comfort with varying temperatures: a graphic design tool

practical limits. Using Fangers model and different possible

clothing levels during the day, a maximum variation of about

8 8C was established for sedentary activity, however this

range will lower (down to about 6 8C) with higher tempera-

tures as levels of clothing insulation are lower with higher

temperatures. In winter, it will be possible to take off a jacket

or loosen a tie, but in summer there is less clothing to remove

(with decency!).

4.4. Maximum temperatures

The maximum comfort temperature without air move-

ment and with mean radiant temperature equal to air tem-

perature, is about 28 8C, with a slight variation according to

the humidity level, the users comfort expectations and the

average temperatures which the users experience. It should

be noted that with higher temperature ranges, the relative

humidity with maximum temperatures would decrease: for

example, with a temperature range between 22 and 28 8C,

the relative humidity at 28 8C will not exceed 70%. The

effect of humidity will be discussed later.

5. Comfort for other activities

For other activity levels and metabolic rates (MET), the

comfort range will vary according to the physical activity,

the typical clothing used, cultural traditions and comfort

expectations.

5.1. Sleeping

Bedrooms will require a lower temperature range of about

5 8C as excessive variations will disturb sleep, but lower

temperatures are comfortable with warm bedding, permit-

ting temperatures as low as 10 8C. This is shown as zone B in

Fig. 1. When bedrooms are used for study and other activ-

ities, higher daytime temperatures will be needed and there-

fore higher average temperatures are required.

5.2. Interior circulation

For spaces such as stairs and passages, lower temperatures

are also possible due to three complementary factors: activ-

ity levels are higher, time spent in these spaces is limited,

and occupants expectations are more flexible. This is com-

fort zone is shown as zone C.

5.3. Outdoor circulation

Users of outdoor spaces are more tolerant of wider

temperature fluctuations, clothing insulation levels can vary

widely, as in cold conditions they can be increased with

overcoats, scarves and gloves, while in summer light cloth-

ing can be worn outdoors. Higher outdoor air movement in

hot conditions can also extend the comfort zone. Finally,

users of outdoor spaces can adjust activity levels over an

extended range to compensate for lower or higher tempera-

tures. Comfort zone D is established for outdoor spaces, with

appropriate activity levels and typical ranges of clothing

levels.

6. Comfort tolerances

Most users do not expect conditions to be completely

within the comfort zone all the time. For example, after a

winter walk in the cold, entering a cool room will feel

comfortable as the previous activity level was high, allowing

time to heat up the space. The following deviation from the

comfort range can therefore be considered in many circum-

stances.

6.1. Excessive temperature range

If the average temperature is comfortable but the range of

temperatures exceeds the comfort limit by 25% then dis-

comfort will be last for about 20% of the time, assuming an

approximately sinusoidal variation in temperature, typical of

the swings found in interiors of naturally conditioned build-

ings with high thermal inertia. When applying the graph to

analyse comfort, the time when this occurs and the times that

the spaces are use should be taken into account. For exam-

ple, in heavy weight buildings used only by day, the peak

temperature may occur in the evening when the buildings are

empty, while in light weight buildings peak internal tem-

peratures may occur a few hours after the outdoor peak when

day use buildings are still occupied. Minimum indoor

temperatures in light weight day use buildings occur a

few hours after the external minimum, corresponding with

sun-rise, so office buildings and schools will only experience

these minimum temperatures for the first hour of use.

6.2. High or low temperatures

Similarly, if the temperature range is acceptable, but the

temperature is about 1.5 8C too hot or too cold, discomfort

will similarly last for about 20% of the time.

6.3. Impact of humidity

Unlike the charts of Givoni [17] and Olgyay [1], this

comfort graph does not include the effect of humidity in an

explicit form. Evidently, high humidity combined with high

ambient temperatures will reduce the upper limit of the

comfort zone, as the evaporative capacity of the air is

reduced and sensible transpiration and skin moisture

increases.

However, in a typical warm humid climate, the average

external temperature range is in the order of 7–10 8C. When

the temperature varies from 23 to 30 8C, the minimum

relative humidity is only 65%, which coincides with the

90 J.M. Evans / Energy and Buildings 35 (2003) 87–93

Page 5: Evaluating comfort with varying temperatures: a graphic design tool

maximum temperature. At midday, the effect of the high

absolute humidity changes the upper comfort limit by

1.5 8C. The effect of humidity is therefore much less than

the effect of the 7 8C temperature variation throughout the

day.

7. Climate analysis

The relation between the typical external climatic condi-

tions and the requirements for comfort can be quickly

grasped using the chart. The comparison between the aver-

age daily external temperature variation and the desirable

internal comfort conditions indicates the bioclimatic strate-

gies that can be used to modify the external conditions

through the following passive design measures described in

Sections 7.1–7.7.

7.1. Air movement

If the average temperature is a few degrees above the

comfort zone and the temperature swing is <10 8C, the air

movement can be used to achieve comfort. Air movement

can reduce the apparent temperature by 2 8C and the tem-

perature swing inside a medium-mass building with good

solar protection will be less than the external swing. Ceiling

fans can be used to achieve air movement without exchange

of indoor and outdoor air, a useful resource at midday and on

hot afternoons when outdoor temperatures are above the

comfort zone.

7.2. Thermal mass

If the average temperature is within the comfort range

but the external swing is higher than the comfort range,

then internal thermal mass, combined with measures to

reduce solar heat gains, can be used to achieve comfort.

The internal temperature swing in a high-mass building

with solar protection can be less than a quarter than the

external swing.

7.3. Internal gains

If the average temperature is less than the comfort zone,

the internal gains in a normal dwelling will achieve an

increase of at least 3–4 8C, but in well-insulated houses

with limited ventilation, the increase can reach up to 10 8C.

In super-insulated buildings, even larger temperature

increases can be achieved.

7.4. Solar gains

Solar radiation through windows can also be used increase

the average internal temperature, but this will also increase

the temperature swing. This is due to the solar gains, which

are limited to daylight hours and coincide with the maximum

outdoor temperatures. High-mass well-insulated buildings

are therefore necessary in order to limit this temperature

swing.

7.5. Selective ventilation (cooling)

Using a combination of selective ventilation and thermal

mass the average internal temperature can be lowered by

about 3 8C, while the range is reduced by a factor of 50–

65%. The term selective implies the use of ventilation when

this is favourable at night to cool the interior of high-mass

buildings, but reducing the air change rate by day to avoid

undesirable heat gains. If the average temperature is above

the comfort limit and the daily range is >14 8C, the selective

ventilation can be used to lower the average temperature.

7.6. Selective ventilation (warming)

An opposite strategy can be used by day when night

temperatures are cool but day temperatures are comfortable

to warm. Internal mass will help to store the heat provided by

the warm air, improving conditions later when the outside air

temperature drops and the ventilation is reduced. Day

ventilation also helps to reduce internal humidity.

7.7. Passive solar systems

The graph also helps to explain the functioning of passive

solar systems such as the Trombe wall (Fig. 2). On a typical

sunny winter’s day when the outdoor air temperature varies

between 5 and 12 8C, the external conditions are well out-

side the comfort zone. But at the outer dark surface of the

Trombe wall, behind the glass, the average temperature and

the temperature range will increase significantly on sunny

days, reaching 25 8C and 40 8C, respectively. By day, the

surface temperature reaches 45 8C, but by night drops close

to 5 8C. However, on the inside of the same mass wall, the

high thermal inertia will dampen down the temperature

swing, to an average temperature of 25 8C and a range of

4 8C. If the internal design temperature is 21 8C, then the

surface temperature of this wall will never drop below this

level and during most of the day will provide a useful source

of heat to the interior.

The approximate equivalent external temperature Teq or

sol–air temperature can be calculated in the following

manner:

Teq ¼ Text þ Iare

where Text is the outdoor temperature, I the radiation

(W/m2), a the coefficient of absorption, and re is the external

surface resistance (including resistance of air cavity and

glass).

So, at midday, in a typical winter day in Buenos Aires,

with a radiation intensity of 490 W/m2 on a vertical north

facing surface and an external temperature of 5 8C, the

maximum temperature will be approximately 63 8C

J.M. Evans / Energy and Buildings 35 (2003) 87–93 91

Page 6: Evaluating comfort with varying temperatures: a graphic design tool

(Teq ¼ 5 þ 490 � 0:8 � 0:16). However, due to the sharp

peak of radiation at midday, the average temperature is only

25 8C and the effective range can be considered as 40 8C.

On the inside surface of the Trombe wall, the tempera-

tures will vary between 23 and 26 8C, with an average of

24.5 8C due to the moderating effect of the thermal mass.

Within a well-insulated room, temperatures will be slightly

lower due to heat losses 20–24 8C, with an average tem-

perature of 22 8C.

The functions of the different layers of the Trombe wall

and their action on the average temperature and temperature

range can be clearly identified (Fig. 2). The outer layer

reduced heat losses, the absorbing layer raises the tempera-

ture as high as possible during the hours of sunshine while

the mass wall dampens down the swing and delays the flow

of heat to the interior. Ideally, the internal surface has to

have an average temperature slightly above the indoor

design temperature to compensate for the heat losses due

to conduction (through the other walls, roof and floor) and

ventilation.

Functioning of a Trombe wall as a passive solar system

indicated by a series of lines that show the changing con-

ditions in different layers of the construction: (i) outdoor

air temperatures are low and temperature swings are

limited; (ii) on the absorbing surface behind the glass

average temperatures rise but swings also increase; (iii)

the thermal swing on the inside of the mass wall is lower

but the average temperature remains the same; (iv) the

indoor air temperature is lower than the surface of the mass

wall due to heat losses and ventilation.

8. Origins

The original version of the graph with a significant

variation in the form of the comfort zones was published

by Evans and de Schiller (1988) [18]. An adapted version of

this graph was first published in English last year in a short

presentation [19]. The importance of the temperature varia-

tion as a factor for selecting bioclimatic design resources

was recognised in the Mahoney’s Tables [20]. Here, the

variation of 10 8C was used as a critical limit. High tem-

peratures and low temperature ranges determine the need for

cross ventilation, while higher ranges indicate the need for

thermal mass as a climate moderator. The background to the

tables was reviewed recently in a PLEA paper [21].

9. Use of the graph

Graph has been used for a number of years as a teaching

tool in the undergraduate course on bioclimatic design and

the 1 year postgraduate bioclimatic design course. Students

have applied the graph to analyse the widely different

regional climates of Argentina and select bioclimatic design

resources. It has been found to be a useful didactic tool to

Fig. 2. The comfort triangle with temperatures in different layers of a Trombe wall: (i) external temperature, (ii) absorbing surface, (iii) internal surface, (iv)

internal air temperature.

92 J.M. Evans / Energy and Buildings 35 (2003) 87–93

Page 7: Evaluating comfort with varying temperatures: a graphic design tool

explain the way in which thermal inertia modifies the

external temperatures to achieve acceptable internal tem-

perature swings.

The graph was included in a number of successful sub-

missions to student competitions, including the PLEA Com-

petition in 1999 (1st and 2nd prizes) and the student

competitions organised by the Argentine Solar Energy

Society in Tucuman, 1999 (1st, 2nd and 3rd prizes) and

Resistencia, Chaco, 2000 (1st and 2nd prizes).

The following procedure is suggested for applying the

comfort triangles in bioclimatic design.

(i) Firstly, the monthly mean maximum and minimum

temperatures are obtained from standard meteorologi-

cal data.

(ii) The conditions for each month are defined as a series

of points on the graph, according to the average

temperature and the temperature range. The average of

the maximum and minimum can be used though this is

not exactly the average monthly temperature.

(iii) The relation between the external conditions and the

desirable comfort range can then be assessed, obtain-

ing the characteristic seasonal requirements for

comfort.

(iv) The appropriate bioclimatic design measures can then

be chosen to achieve a favourable modification of the

external climate. The importance of each measure will

depend on the number of months when each measure

applies and the distance between the points and the

comfort zone.

Once the bioclimatic strategies have be chosen and

incorporated in the building design, more detailed simula-

tions of the internal temperatures can be made. The results

can also be plotted on the comfort triangles graph to test the

effectiveness of the measures when integrated in building

design. Finally, existing buildings can be evaluated, using

indoor and outdoor temperature measurements for a series of

representative or extreme days, including periods in summer

and winter with clear and cloudy skies.

10. Conclusions

The comfort triangles graph is a simple tool for thermal

comfort analysis that complements existing methods of

bioclimatic design. The graph presents a new of visualising

the complex variation of temperature, comparing this with

desirable comfort conditions and choosing appropriate bio-

climatic design resources.

The method has been tested in the teaching environment

over a number of years at the graduate and postgraduate

levels. It has been found particularly useful for visualising

and explaining passive design strategies involving periodic

heat flow, such as night ventilation, thermal inertia, direct

and indirect solar gains, etc. The wide range of temperatures

proposed follows the adaptive approach to thermal comfort,

based on field studies, and is also consistent with models

based on climate chamber studies.

References

[1] V. Olgyay, Design with Climate, Princeton University Press,

Princeton, NJ, 1963.

[2] T. Bedford, Environmental Warmth and its Measurement, Medical

Research Council, War memorandum No. 17, HMSO, 1940.

[3] C. Webb, Thermal discomfort in an equatorial climate, IHVE Journal

27 (1960) 297–304.

[4] C.E.A. Winslow, L.P. Herrington, A.P. Gagge, Physiological

reactions to environmental temperature, American Journal of

Physiology 120 (1937) 1–22.

[5] F. Nichol, M. Humphreys, O. Sykes, S. Roaf, Standards for Thermal

Comfort, Indoor Temperatures for the 21st Century, E & FN Spon,

London, 1995.

[6] J.F. Nichol, Thermal comfort and temperature standards in Pakistan,

in: F. Nichol, M. Humphreys, O. Sykes, S. Roaf (Eds.), Standards for

Thermal Comfort, Indoor Temperatures for the 21st Century, E & FN

Spon, London, 1995, pp. 149–156.

[7] A. Auliciems, S.V. Szokolay, Thermal Comfort, PLEA Note 3,

Passive and Low Energy Architecture Design Tools and Techniques,

University of Queensland, Brisbane, 1997.

[8] M.A. Humphreys, Field Studies of Thermal Comfort Compared and

Applied, Building Research Establishment, Current Paper, (76/75),

Watford, 1975.

[9] A. Auliciems, Towards a psycho-physical mode of thermal

perception, International Journal of Biometeorology 25 (1981)

109–122.

[10] F. Nichol, S. Roaf, Pioneering new indoor temperature standards: the

Pakistan Project, Energy and Buildings 23 (1996) 169–174.

[11] International Standards Organization, ISO, Method to Evaluate

Thermal Comfort, ISO Standard 7730:1984, Geneva.

[12] Y. Nishi, A.P. Gagge, Effective temperature scale for hypo- and

hyperbaric environments, Aviation, Space and Environmental

Medicine (1997) 97.

[13] P.O. Fanger, Thermal Comfort, Danish Technical Press, Copenhagen,

1970.

[14] J.F. Busch, Thermal comfort in Thai Air-Conditioned and Naturally

Ventilated Offices, in: F. Nichol, M. Humphreys, O. Skyes, S. Roaf

(Eds.), Standards for Thermal Comfort, Indoor Temperatures for the

21st Century, E & FN Spon London, 1995.

[15] T.H. Karyono, Higher PMV causes higher energy consumption in air

conditioned buildings: a case study in Jakarta, Indonesia, in: Nicol,

Humphreys, Sykes, Roaf (Eds.), Standards for Thermal Comfort, Indoor

Temperatures for the 21st Century, E & FN Spon, London, 1995.

[16] M. Humphreys, Outdoor temperatures and comfort indoors, Building

Research and Practice 6 (2) (1978) 92–105.

[17] B. Givoni, Man, Climate and Comfort, Elsevier, London, 1967.

[18] J.M. Evans, S. de Schiller, Diseno Bioambiental y Arquitectura

Solar, EUDEBA, Buenos Aires, 1988.

[19] J.M. Evans, Comfort triangles: analysis of temperature variations and

design strategies in passive architecture, in: K. Steemers, S. Yannas

(Eds.), Proceedings of PLEA 2000 on the Architecture, City,

Environment, James and James, London, 2000.

[20] O.H. Koenigberger, C. Mahoney, J.M. Evans, Climate and House

Design, United Nations, New York, 1970.

[21] J.M. Evans, 30 years of the Mahoney’s Tables, in: S. Szokolay (Ed.),

Proceedings of PLEA’99, University of Queensland, Brisbane, 1999.

J.M. Evans / Energy and Buildings 35 (2003) 87–93 93