© 2007 University of Sydney. All rights reserved.

Architectural Science Review

www.arch.usyd.edu.au/asr Volume 50.3, pp 224-233

Applying Natural Ventilation for Thermal

Comfort in Residential Buildings in Singapore

Wang Liping† and Wong Nyuk Hien

Department of Building, National University of Singapore, 4 Architecture Drive, Singapore 117566

†
Corresponding author: Email: g0301083@nus.edu.sg

Received 26 August 2006; accepted 23 February 2007

Abstract: This paper explores the potential of using natural ventilation for indoor thermal comfort in residential buildings in Singapore
based on climatic data analyses. A typical year of weather data was selected by statistical analyses of weather data over a five-year period
in Singapore. The characteristics of weather data for a typical year were analyzed with regard to temperatures, wind speeds, and wind
directions. Thermal comfort models for natural ventilation applicable for Singapore were used for the assessment of thermal conditions.
It was found that by applying natural ventilation, a thermally comfortable indoor environment can potentially be provided for a large
number of hours in the year with optimum of facade designs for residential buildings (negligible heat sources). The study also developed
facade design guidelines based on the required indoor air velocities for thermal comfort.
Keywords: Climatic data, Design guidelines, Fascade design, Indoor air quality, Indoor environmental quality, Natural ventilation, Residential buildings,
Thermal comfort, Modeling, Singapore, Wind

Introduction
There is a growing interest in the application of natural
ventilation in buildings due to the energy, indoor air quality and
environmental problems associated with mechanically ventilated
buildings. Various mechanical systems including HVAC systems
in residential and office buildings contribute substantially to the
energy consumption.
As the benefits of natural ventilation, including reducing
operation costs, improving indoor air quality and providing
satisfactory thermal comfort in certain climates, are recognized,
passive cooling of houses using natural ventilation has become an
attractive alternative to alleviate the associated problems with air-
conditioned buildings. The concept of natural ventilation is well
accepted and welcomed by people and designers in the world.
Although 86% of people in Singapore live in Housing and
Development Board (HDB) residential buildings, designed to be
naturally ventilated, residents tend to keep increasing the usage of
air-conditioning for indoor thermal environment. The statistic data
of energy usage in Singapore show that domestic energy consumption
rose up to 58% of total energy consumption in 1998 from 18%
of total energy consumption in 1988 and energy consumption
per person was doubled from 1991 to 2001. For tropical regions,
where the air temperature and relative humidity are generally high,
the effectiveness of natural ventilation is questionable. Therefore,
the investigation of natural ventilation applicability in hot-humid
climates is significant for providing guidance to optimize facade
designs. The purpose of this study is to test the availability of
natural ventilation for the hot-humid climate regions in Singapore
based on local climatic environment analyses.
Selection of Typical Year Data
For thermal analyses of local climate data, thermal conditions
of one typical year data set (8760 hours) of weather conditions are
required. Several kinds of typical weather data are currently available
and the criteria for constructing a typical year weather data vary from
one database to another.
Canadian Weather files for Energy Calculations (CWEC),
produced by Numerical Logics, contain hourly weather observations
representing an artificial one-year period specially designed for
building energy calculations. International Weather files for Energy
Calculations (IWEC) are derived from up to 18 years of DATSAV3
hourly weather data originally archived at the U.S. National Climatic
Data Center (NCDC). The weather data are supplemented by solar
radiation estimated on an hourly basis from earth-sun geometry, hourly
weather elements, and particularly the cloud amount information.
Solar and Wind Energy Resource Assessment (SWERA) developed
high quality information on solar and wind energy resources in 13
developing countries. Typical meteorological year 2 (TMY2) was
derived from a 1961-1990 period of record. The TMY2s are data
sets of hourly values of solar radiation and meteorological elements
for a one-year period. California climate zones 2 (CTZ2) updated
weather data for 16 California climate zones for use.
Wang Liping and Wong Nyuk Hien Natural Ventilation and Thermal Comfort 225

Typical meteorological year is based on an empirical approach No. of Max/Min (Dry bulb temp)
that selects individual months from a long period (30 years or
700
more). Data sets occurring with the maximum frequency may 1997
be chosen to be included in the typical year. Since there is only a 600 1998

limited amount of weather data available in Singapore, a different 500

1999
2000
criteria for construction of typical year weather data is applied. 2001
400

Number
In this study, year 2001 weather data has been tested as a typical
300
year data because it is the latest weather data available and has
minimum amount of missing data. This reduces the amount of 200

reconstruction and manipulation of data. To ensure that year 100

2001 weather data can be used as a typical year data, the testing
0
steps listed below were taken to test the data set. Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
• The number of instances where dry bulb temperature or Month
global radiation in a particular hour in a year exceedsFigure
the 1.Figure 1: The of
The number number
hourlyofinstances
hourly instances the dry
the dry bulb bulb temperature
temperature for each month of th
maximum or falls below the minimum of the other years yearisexceeds
for the
eachmaximum
month oforthe year
falls exceeds
below the maximum
the minimum of theor fallsyears.
other below the
counted and shown on Figures 1 and 2 respectively. Any minimum of the other years.
unusual traits in the hourly data of that year can be detected.
Year 2001 does not show any unusual characteristics. No. of Max/Min (Glob. rad)
• The differences in the amount of dry bulb temperature and
global radiation for each of the above instances are added 350
1997
separately for those above the maximum or below the 1998
280
minimum values. Year 2001 does not show any unusual 1999
2000
trends, as shown in Figures 3-6. 210 2001
Number

• The frequency of occurrence of particular wind conditions

for each month has been calculated. Figure 7 illustrates the 140
frequency distribution in July. Year 2001 does not show
any unusual trends. 70

• The year 1997 and year 1998 data show some unusual high
0
dry bulb temperature and wind velocity in some months. Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Therefore, it is concluded that the use of year 2001 as a Month
typical year is reasonably valid. Figure 2: The number of hourly instances the horizontal global
Figure 2. Thefor
radiation number
each ofmonth
hourlyofinstances
the yeartheexceeds
horizontal
the global
maximum radiation for each month of
or falls
the year exceeds thebelow
maximumthe or falls belowofthe
minimum theminimum
other of the other years.
years.
Thermal Comfort Studies
In order to evaluate applicability of natural ventilation in Cumul. Max (Dry Bulb Temp)
Singapore, the primary task is to select appropriate criteria to analyze
typical weather data. In the aspect of thermal performance, the 1000
1997
function and purpose of naturally ventilated residential buildings is 1998
800
to provide comfortable indoor thermal environment for residents, 1999

which including indoor air temperature, air velocity, mean radiant 2000
600 2001
temperature and relative humidity. Therefore, thermal comfort 1
к

for naturally ventilated buildings should be the idealized criteria 400

for evaluation.
It is important to know that the comfort expectations of a 200

tropical population and people from temperate or cold climate are

different. Nicol and Humphreys (1973) presented the results of 0
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
field studies in the UK, India, Iraq, and Singapore. It is noted that Month
temperatures well above 30������������������������������������
°C����������������������������������
are not considered uncomfortable Figure 3: The cumulative amount by which the dry bulb
in some cases. Figure 3. The cumulative amount by which the dry bulb temperature for each month of the
temperature for each month of the year exceeds the maximum of
Dutt, De Dear and Krishnan (1992) investigated thermal year exceeds the maximum of the other years.
the other years.
comfort in a naturally ventilated dining room of a students���������
’ hall���.
A tropicalized PMV index is proposed based on mean conditioned
in the survey (583 respondents) with mean air temperature 29.8���� °C��, Bangladesh. The results showed that people could endure high
mean radiant temperature 30.1���������������������������
°C�������������������������
and mean humidity 75%. temperature and very high humidity for comfort.
Thermal comfort analyses indicate that the mean indoor velocity Many studies indicate that thermal comfort predictions for an 19
0.47 m/s is associated with the predicted mean vote of 0.74, which air-conditioned room is not suitable for naturally ventilated rooms
amounts to three quarters of the way along between “neutral” and because of adaptive activities and different thermal perception. Baker
slightly warm” on the 7-point scale. Mallick (1996) conducted and Standeven (1996) suggested that adaptive activities/exercises are
a thermal comfort study of occupants living in urban housing in important to meet thermal satisfaction of the occupants. Thermal
 226 Architectural Science Review Volume 50, Number 3, September 2007

Cumul.
Cumul. MinMin (Dry
(Dry Bulb
Bulb Temp)
Temp) smaller for the air-conditioned buildings than for the naturally
1000 1000
ventilated buildings. Many published studies show that the human
19971997
thermal sensation in a naturally ventilated environment is much
19981998
800800 1999
better than that in a mechanically controlled thermal environment
1999
20002000 (Fanger & Toftum, 2002; Xia, Zhao & Jiang, 1999). De Dear
600600 20012001 and Brager (2002) believed that thermal perception of those who
live or work in naturally ventilated buildings –both preferences
к
к

400400
as well as tolerances –are likely to extend over a wider range of
temperatures than are currently reflected in the old ASHRAE
200200
Standard 55 comfort zone. Nicol and Humphreys (2002) explained
0 0
the origin and development of the adaptive approach to thermal
Jan JanFebFebMar MarApr AprMayMayJun Jun Jul JulAugAugSepSepOct OctNovNovDecDec comfort. The adaptive principle is that people react in ways,
Month
Month which tend to restore their comfort if a change occurs such as to
Figure 4: The cumulative amount by which the dry bulb produce discomfort. They pointed out that the comfort zone may
temperature for each month of the year falls below the minimum of be as narrow as °C����������������������������������������������
������������������������������������������������
in a situation where there is no possibility
Figure
Figure 4. 4.TheThe cumulative
cumulative amount
amount by which
by other
which thethedrydry
bulbbulb temperature
temperature for for each
each month
month of the
the years. of of the
changing clothing or activity and where air movement cannot
yearyear
fallsfalls below
below thethe minimum
minimum of the
of the other
other years.
years.
be used. In situations where these adaptive opportunities are
Cumul.
Cumul. MaxMax (Glob.
(Glob. rad)rad) available and appropriate, the comfort zone may be considerably
30000
wider. Nicol (2004) conducted field studies in hot-humid tropical
30000
19971997 climates. It was found that the international standard for indoor
1998
1998 climate, ISO7730 based on Fanger’s predicted mean vote equation,
19991999
20000 20002000
does not adequately describe comfort conditions when buildings
20000
20012001 are not mechanically heated or cooled. ISO7730 overestimates
W/sq.m
W/sq.m

the occupant response on the ASHRAE scale at high temperatures

10000
and underestimates it at low temperatures.
10000
Current available models for thermal comfort prediction for
naturally ventilated buildings in Singapore are summarized in the
0
following two sections.
0
Jan JanFebFebMar MarApr AprMayMayJun JunJul JulAugAugSepSepOct OctNovNov
DecDec
Month
Month
ANSI/ASHRAE Standard 55-2004
Figure 5: The cumalative amount by which the horizontal global ASHRAE Standard (2004) provided the alternative method to
Figure
Figure 5. 5.TheThe
radiation cumulative
cumulative
for amount
amount
each month by
ofbythe which
which
year the the horizontal
horizontal
exceeds global
global
the maximum radiation
of thefor for
radiation each
each month
month
of the
of the yearyear exceeds
exceeds thethe maximum
maximum of the
of the other
other years.
years. predict acceptable thermal conditions for naturally conditioned
other years. spaces, developed from a global database of 21,000 measurements
taken primarily in office buildings. In the approach, thermal
Cumul.
Cumul. MinMin (Glob.
(Glob. rad)
rad) responses in naturally ventilated spaces are linked with the outdoor
30000
30000
climate and two sets of operative temperature limits - one for 80%
1997
1997 acceptability and one for 90% acceptability are included in the
1998
1998
1999
diagram “Acceptable operative temperature ranges for naturally
1999 21 21
20000 2000
2000
conditioned spaces” (ASHRAE Standard, 2004). The application
20000
2001
2001 of this method, operable windows that can be adjusted by the
W/sq.m
W/sq.m

occupants are required. Metabolic rates range from 1.0met to

10000
1.3met and occupants can adapt their clothing according to
10000
thermal conditions. The required indoor operative temperature
limits for 90% acceptability and 80% acceptability are listed in
0
Table 1, which is calculated based on outdoor air temperature of
0
JanJanFebFebMarMarApr AprMayMayJunJunJul JulAugAugSepSepOct OctNovNov
DecDec the typical year in Singapore. However, the estimation of thermal
Month
Month comfort using this method is limited with mean monthly outdoor
air condition.
Figure
Figure
Figure The cumulative
6:cumulative
6. 6.TheThe cumulative amount
amount
amount
by by
by
which
which
which
thethe
the horizontal
horizontal
horizontal global
global
globalfor for
radiation
radiation each
each month
month
of theradiation
year for
exceeds each
the month
minimum of
of the
the year
other exceeds
years.
of the year exceeds the minimum of the other years. the minimum of the Thermal Comfort Studies at the National University of
other years.
Singapore for Naturally Ventilated Residential Buildings
Feriadi (2003) developed a thermal comfort prediction chart
comfort zone for passive natural ventilation might be extended by and fuzzy thermal comfort model suitable for naturally ventilated
adaptive opportunities. Brager and De Dear (1998) pointed out that buildings in the tropics based on 1063 data collected through field
an adaptive model could eliminate the need to get the information surveys in Singapore and Indonesia. Wong and Khoo (2003)
of clothing patterns of the future, unknown occupants by taking conducted
22 22 a field study in classrooms in Singapore that were
account of the feedback loop between discomfort and purposive mechanically ventilated by fans. The results further confirmed
behavioral thermoregulation. De Dear and Brager (1998) found that the ASHRAE standard 55 is not applicable in free-running
that the biases between PMV and ASHRAE vote were generally buildings in the local climate. The adjusted PMV model by Fanger
Wang Liping and Wong Nyuk Hien Natural Ventilation and Thermal Comfort 227




 models to analyze typical climatic data in Singapore






with the aids of reasonable assumptions to investigate



 the potential of natural ventilation.







 
 Climatic Analyses


Singapore is situated at 1.2���������������������������
° latitude with relatively




 

 

 

 

 

 

 

 
 

 

!! 
high temperature ranging from 23°C������������������
to 34������������
°C����������
and high




 

 
 



 

 

 

 

 

 






relative humidity averagely 84% in the whole year. As
ZLQGVHSHHGUDQJHPV-XO
ZLQGVHSHHGUDQJHPV-XO typical year weather data represents the characteristics
Figure 7: The frequency of occurrence of particular wind conditions in July. of local climate, analyses of climatic parameters can
Figure
Figure7.7. The
Thefrequency
frequencyofofoccurrence
occurrenceofofparticular
particularwind
windconditions
conditionsininJuly.
July.
not only evaluate the feasibility of natural ventilation
for the certain location, but also put forward fa�����çade
and Toftum (2002), which incorporates two common forms design requirements to satisfy thermal comfort and minimize energy
of adaptation: namely, reducing activity pace and expectation, loads for the long term for naturally ventilated buildings.
still showed discrepancy in predicting actual thermal sensations,
especially at lower temperatures. Thermal comfort regression Wind Data Analyses
model Eq 1 (Wang & Wong, 2005) for naturally ventilated
residential buildings has been derived from 538 field survey data Thermal comfort chart (Feriadi, 2003) suggested making use
in Singapore of wind for natural cooling. Wind is a complex phenomenon
than can change in a short period both in wind speed and in wind
(1) direction. Figure 12 shows the selected ranges of wind speed

where Temp (���������������������������������������������
°C�������������������������������������������
) indicates the indoor air temperature and V distribution over a whole year. Figure 13 reveals that on average,
(m/s) refers to indoor air velocity measured at 1.2 m above the wind speed is low during the nighttime and early morning. The
ground. mean wind speed in the whole year is lower than 1m/s during the
The term PMV refers to the average (mean) response of a group time 22:00-6:00 in a day while ambient temperature is low. Wind
exposed to a given climatic condition rather than individual responses. speeds in the day time are mainly in the range of 1.5-3m/s. Figure
As relative humidity is highly correlated with dry bulb temperature, 13 also shows that there is high wind speed outside to compensate
the impacts of variation of dry bulb temperature on thermal comfort the high temperature at noontime. The day-night change of
can indirectly indicate the effects of relative humidity changes on wind speed profile is an important feature for promoting natural
thermal comfort. In another aspect, relative humidity in hot-humid ventilation in Singapore.
climate is generally in the high level (above 60%). Therefore, the Figure 14 shows wind rose graph for the whole year. 0��°
parameter relative humidity is not involved in the regression model. indicates that the wind comes from due north and 90° indicates
Individual factors in thermal comfort such as clothing insulation that wind comes from due east�������������������������������������
. It is evident that the prevailing
(clo) and personal activity (met) for residential buildings are normally wind directions in Singapore are north and south over the entire
standard in Singapore. Clo is around 0.34-0.5 as people tend to year. To illustrate the distribution of wind speed as a function
adjust their clothes at home for better thermal comfort and met is of wind direction, wind rose graphs are plotted for each month
equal to 1.0. Dry bulb temperature and air velocity are parameters of the typical year data. Wind rose graphs for January and May
for thermal comfort prediction. are demonstrated in Figure 15 and Figure 16. Wind direction
The following section utilizes the available thermal comfort in the range of 270������������������������������������������
°-30° dominates January, February, March,
2424

Table 1: Required indoor operative temperature limits for naturally ventilated spaces in Singapore base
on ASHRAE Standard 55-2004.
Month Jan Feb Mar Apr May Jun
Mean monthly outdoor air
26.74 26.97 27.48 27.75 28.52 28.16
temperature (°C)
MIN 22.66 22.70 22.85 22.93 23.30 23.12
80% acceptability
MAX 29.65 29.75 29.88 29.94 30.20 30.08
MIN 23.80 23.90 24.10 24.20 24.40 24.34
90% acceptability
MAX 28.72 28.80 28.90 28.95 29.20 29.08
Month Jul Aug Sep Oct Nov Dec
Mean monthly outdoor air
28.45 28.01 27.99 27.64 27.17 26.70
temperature (°C)
MIN 23.30 23.00 23.00 22.90 22.76 22.66
80% acceptability
MAX 30.20 30.00 30.00 29.91 29.80 29.65
MIN 24.40 24.30 24.30 24.16 23.98 23.80
90% acceptability
MAX 29.20 29.00 29.00 28.93 28.84 28.72
228 Architectural Science Review Volume 50, Number 3, September 2007

Figure 8: Hourly temperature and RH on Thermal comfort chart in February (modified from Feriadi, 2003).

Figure 9: Hourly temperature and RH on thermal comfort chart in May (modified from Feriadi, 2003).
 Wang Liping and Wong Nyuk Hien Natural Ventilation and Thermal Comfort 229

 -DQ

5HTXLUHG$LUYHORFLW\PV
)HE

0DU

$SU
 0D\

 -XQ

-XO

$XJ

6HS
 2FW
                       
1RY

7LPHLQDGD\ 'HF

Figure 10. Required

Figure10: Requiredaverage
averagemonthly
monthlyindoor
indoor air
air velocity in aa day.
day.
 -DQ

)HE

0DU

$SU
 0D\
&Y

 -XQ

-XO

$XJ

6HS
 2FW
                       
1RY

7LPHLQDGD\ 'HF

Figure Required
11:11.
Figure average
Required monthly
average Cv Cv
monthly distribution inina aday.
distribution day.

25.00% 100.00% 500 3.5

450 No. of hours
with wind 3

average wind speed (m/s)

20.00% 80.00% 400
No. of hours with wind
% culmu.in range

Average wind
in range 350 speed(m/s) 2.5
% in range

15.00% cumul. in range 60.00% 300 2

250
1.5
10.00% 40.00% 200
150 1
5.00% 20.00% 100
0.5
50
0.00% 0.00% 0 0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
0

0< ..<=0.4

0.4<..<=1

1<..<=2

2<..<=3

3<..<=4

4<..<=5

5<..<=6

6<..<=7

7<..<=7.8

Hour of the day

Figure
Figure 13: Average
13. Average occurrence
occurrence and speed and speed
of wind overof
24wind over
hours of 24in Year 2001.
a day
ranges for wind speed(m/s) hours of a day in year 2001.
Figure 12: Frequency
Figure of wind
12. Frequency speedspeed
of wind in Year 2001.2001.
in Year e

e e
e 

e e 25
e e
 e  PV

 e e
 PV
e 
 e
e e
e e  PV
 
 PV
 PV PV
e e  PV
e
e


e
e
 PV
 PV

e
e e
e
 PV
 PV
e 1RGLUHFWLRQQR
 PV
ZLQG 
e e
e e 1RGLUHFWLRQQR
e
ZLQG  26
Figure 14. Frequency of wind speed above selected values per direction in the year.
e e
Figure 14. Frequency of wind speed above selected values per direction in the year.
e 1RGLUHFWLRQQR
ZLQG 
Figure 14: Frequency of wind speed above selected values per direction in the year.
28

Figure 14. Frequency of wind speed above selected values per direction in the year.
230 Architectural Science Review Volume 50, Number 3, September 2007

Table 2: Percentage of hourly outdoor air Table 3: Percentage of hourly outdoor air Table 4: Percentage of hourly outdoor air
out of comfort zone in day or night with out of comfort zone in day or night with out of comfort zone in day or night with 0.6-
0.1m/s air velocity by ASHRAE-2004 0.2-0.6 m/s air velocity by ASHRAE-2004 1m/s air velocity by ASHRAE-2004 thermal
thermal comfort model. thermal comfort model. comfort model.
Month Day Night Month Day Night Month Day Night
Jan 35.48% 0.00% Jan 33.76% 0.00% Jan 27.0% 1.6%
Feb 36.19% 0.00% Feb 33.81% 0.00% Feb 13.4% 0.0%

Mar 42.15% 0.00% Mar 38.06% 0.00% Mar 26.1% 15.7%

Apr 45.1% 19.4%
Apr 43.33% 0.00% Apr 39.11% 0.00%
May 51.5% 31.9%
May 52.90% 1.79% May 46.88% 0.36%
Jun 38.2% 20.8%
Jun 44.89% 0.74% Jun 42.89% 0.00%
Jul 36.4% 21.1%
Jul 56.13% 0.00% Jul 51.40% 0.00%
Aug 40.1% 25.0%
Aug 40.65% 0.00% Aug 39.57% 0.00%
Sep 32.4% 19.7%
Sep 45.56% 0.00% Sep 44.22% 0.00% Oct 28.0% 16.1%
Oct 40.22% 0.00% Oct 38.28% 0.00% Nov 25.8% 6.4%
Nov 34.89% 0.00% Nov 33.78% 0.00% Dec 20.3% 3.1%
Dec 32.35% 0.00% Dec 31.36% 0.00% Average 32.1% 15.2%

November and December with comparable high wind speed���. ventilated buildings. It is recommended that large openings should
Wind direction in the range of 120������������������������������
° - 210° dominates May, June, be located in north and south sides for cross ventilation while
July, August and September������������������������������������������
. In April, the wind speed is relatively appropriate building construction materials should be provided
low and there is no prevailing wind direction. In October, the for east and west external walls to avoid high solar heat gains.
wind direction is southwest (i.e. 150����������������������������
°-210°, 240°-270°)����������
. Unique
monsoon conditions throughout the whole year with north or Thermal Comfort Estimations based on
south direction provides naturally optimum fa�����������������
çade designs for
naturally ventilated buildings�������������������������������������
. Situated so close to the equator, Thermal Comfort Models
Singapore experiences more intensive solar radiation occurring on
east and west facing fa������������������������������������������
çades more than on north and south facing ASHRAE Standard
façades. ����������������������������������������������������������
This characteristic of the wind profile makes it possible The ASHRAE Standard 55-2004 relates the outdoor air
e
for Singapore to achieve satisfactory thermal comfort in naturally
 temperature with comfortable operative temperature range. When
e e

e

e  e
e e
e  PV
 
e e
e  e
 PV
e 
e
e e  PV


    PV

PV PV
e e e  PV
e 

e
e
 PV
 PV
e
e
e
e
 PV
 PV
1RGLUHFWLRQQR
e  PV
ZLQG 
e e
e e 1RGLUHFWLRQQR
e
Figure 15. Frequency of wind speed above selected values per direction (January).
ZLQG 

Figure 15. Frequency of wind

e speed above selected values per direction (January).
e
e 1RGLUHFWLRQQR
ZLQG 
Figure 15: Frequency of wind speed above selected values per direction (January).

Figure 15. Frequency of wind speed above selected values per direction (January).
Wang Liping and Wong Nyuk Hien Natural Ventilation and Thermal Comfort 231

Table 5: Percentage of hourly outdoor air out Table 6: Percentage of hourly outdoor Table 7: Percentage of hourly outdoor
of neutral zone in day or night by thermal air out 6:
Table of neutral zoneof
Percentage in hourly
day or night by
outdoor air out of neutral zone in day or night by
comfort chart. thermal comfort regression model with
air out of neutral zone in day or night by thermal comfort regression model with
0.1 m/s indoor
thermal comfortairregression
velocity. model with 0.4 m/s indoor air velocity.
% out of neutral comfort zone 0.1 m/s indoor air velocity.
Month Day Night Month Day Night Month Day Night
Jan 27.0% 1.6% Jan 24.1% 0.0% Jan 21.9% 0.0%
Feb 13.4% 0.0% Feb 27.1% 0.0% Feb 24.7% 0.0%
Mar 26.1% 15.7% Mar 34.3% 0.4% Mar 30.8% 0.0%
Apr 45.1% 19.4% Apr 37.2% 0.8% Apr 33.1% 0.1%
May 51.5% 31.9%
May 46.5% 10.9% May 43.0% 6.0%
Jun 38.2% 20.8%
Jun 44.6% 5.8% Jun 40.0% 2.7%
Jul 36.4% 21.1%
Jul 49.7% 6.2% Jul 45.0% 2.0%
Aug 40.1% 25.0%
Aug 41.7% 5.6% Aug 35.8% 2.2%
Sep 32.4% 19.7%
Sep 41.9% 1.7% Sep 37.4% 0.3%
Oct 28.0% 16.1%
Nov 25.8% 6.4% Oct 37.1% 0.5% Oct 32.9% 0.0%

Dec 20.3% 3.1% Nov 30.6% 0.4% Nov 25.6% 0.1%

Total 32.1% 15.2% Dec 25.7% 0.0% Dec 22.8% 0.0%

the difference between radiant and air temperature is less than 4����
°C��, as the temperature difference between MRT (mean radiant
the operative temperature is calculated by the following (2), temperature) and ambient temperature should be less than 2���
°C�
(2) for indoor thermal comfort (Chrenko, 1953), which can be met
by applying appropriate construction materials and shading in
where =air temperature naturally ventilated residential buildings.
= radiant temperature With the two assumptions, the percentage of estimated
A= 0.5 (<0.2m/s); 0.6(0.2-0.6m/s); 0.7(0.6-1m/s) hourly indoor air condition out of neutral comfort zone (80%)
for each month is calculated. When indoor air velocity is less
In this approach, two assumptions are adopted. The outdoor than 0.2 m/s, the 80% acceptable thermal comfort condition
air temperature is assumed equal to the indoor air temperature,
e
can be provided for about 73% of the whole year. The indoor
which can be met for indoor heat sources indoor environment
 condition outranges acceptable thermal comfort condition
higher thane
are negligible. MRT is 2�����������������������������������
°C���������������������������������
ambient temperature ise
shown in Table 2. When indoor air velocity is 0.2-0.6

e e
e
  PV
e  e
e
e  

e
 PV
e e
  PV
e e e e

 PV
e   e
  PV
PV
 PV
e
e


e
e
  PV
PV
e e  PV
e e
 PV
e 1RGLUHFWLRQQR
e e
ZLQG 
PV
e e
e 1RGLUHFWLRQQR
ZLQG 
Figure 16. Frequency of wind speed above selected values per direction (May).
e e
Figure 16. Frequency of wind speed above selected values per direction (May).
e 1RGLUHFWLRQQR
ZLQG 
Figure 16: Frequency of wind speed above selected values per direction (May).

Figure 16. Frequency of wind speed above selected values per direction (May).
232 Architectural Science Review Volume 50, Number 3, September 2007

Table 8: Percentage of hourly outdoor Regression Model

air out of neutral zone in day or night by The assumption that outdoor air temperature is equal to indoor
thermal comfort regression model with air temperature is employed when thermal regression model (Wang
0.7 m/s indoor air velocity. & Wong, 2005) is applied for climatic analyses. If the indoor air
0.7m/s Day Night velocity is 0.1 m/s, indoor air condition in neutral comfort zone
can be about 61% of total hours in the year. If indoor air velocity
Jan 18.7% 0.0% is increased to 0.4m/s, indoor air condition within neutral comfort
zone can be 66%. When the indoor air velocity reaches to 0.7m/s,
Feb 21.6% 0.0%
about 72% in the whole year, indoor air conditions are in neutral
Mar 26.3% 0.0% zone. The percentages of hourly outdoor air condition out of
neutral comfort zone are illustrated in Tables 6, 7, 8 respectively
Apr 28.1% 0.0%
for various indoor velocities.
May 38.4% 2.0% The required air velocity for the indoor thermal comfort
(consider the neutral condition PMV=0) was calculated based on
Jun 31.9% 0.5%
the regression model. The monthly average required indoor air
Jul 39.1% 0.4% velocity for thermal comfort has been shown in Figure 10. High
air velocity (0.8-1.2 m/s) is frequently required to reach thermal
Aug 30.1% 0.3%
comfort at noontime (11:00-14:00). According to the required air
Sep 31.9% 0.0% velocity and outdoor wind speed, the required velocity coefficient Cv
in Singapore for each month in typical year was calculated, as shown
Oct 26.5% 0.0%
in Figure 11. Higher Cv requirements occur in April (0.66) and
Nov 21.8% 0.0% June (0.61) around 9:00 am in the morning. During the nighttime
and early morning (after 21:00 or before 7:00), the required Cv for
Dec 19.2% 0.0%
thermal comfort is much lower around 0.1-0.2. However, there
are 154 hours in the typical year, in which air velocity is required
but there is no wind speed outside. During these hours, thermal
comfort is achieved by other means - appropriate construction
m/s, the 80% acceptable thermal comfort condition can be materials and shading devices or mechanical ventilation if needed.
provided for about 75% of the year. The indoor condition The required air velocities and air velocity coefficients for indoor
outranges acceptable thermal comfort condition is shown in thermal comfort obtained from climate analyses can be used for
Table 3. When indoor air velocity is lager than 0.6m/s, the fa�������������������������������������������������������������
çade designs for naturally ventilated buildings in Singapore.
80% acceptable thermal comfort condition can be provided for
about 78% of the year with natural ventilation. The indoor Conclusion
condition outranges acceptable thermal comfort condition is
shown in Table 4. From the above analyses on climate data, indoor environment
has very high possibility to achieve satisfied thermal comfort by
Thermal Comfort Chart purely natural ventilation in Singapore with smart fa���������������
çade design in
In this approach, the outdoor air temperature is assumed residential building (with negligible heat sources)�����������������
. The estimated
equal to indoor air temperature and outdoor relative humidity is percentages in a year using three different methods are 73-77% in the
assumed equal to indoor relative humidity. The two hypotheses range of an 80% acceptable comfort zone (ASHRAE standard 55),
can be met for no heat sources indoor environment such as 53% (thermal comfort chart, Feriadi, 2003) and 66-72% (regression
residential buildings. Conditions of the air decided by dry-bulb model, Wang & Wong, 2005) inside a neutral zone. Comparing
air temperature and relative humidity for outdoor climate have with the other two thermal comfort models, thermal comfort chart
been registered on the thermal comfort chart (Feriadi, 2003) to (Feriadi, 2003) is strict for indoor comfort prediction. During those
investigate the possibility of thermal comfort through natural hours outside comfort zone, thermal comfort is required by other
ventilation in Singapore within each month. means–construction materials and shading devices or mechanical
Figures 8 and 9 illustrate the hourly temperature and RH on fan if needed.
Thermal comfort chart in February and May. It can be observed There are two strategies to make indoor environment inside neutral
that weather conditions during the period of nighttime and rainy zone and achieve satisfactory thermal comfort based on the above
daytime are mainly on the top left of thermal comfort chart, in analyses. Firstly, to compensate high temperatures, increase the range
which thermal comfort is easy to be satisfied with low air velocity. of neutral comfort zone by providing higher wind speed (more than
Table 5 shows the percentage of hourly outdoor air condition out 1m/s). Secondly, with suitable fa��������������������������������������
çade designs - qualified construction
of neutral comfort zone (1m/s air velocity) in day or night. From materials and shading devices, indoor temperature could be reduced
the results, it can be seen that thermal comfort could be naturally by 2-3°C�����������������������������������������������������������
comparing with outdoor temperature. In addition, thermal
satisfied within 52.7% of total hours in the year during which comfort zone is much wider than neutral zone (Feriadi, 2003).
outdoor air conditions are inside neutral comfort zone. The rest Nicol and Humphreys (2002) has pointed out comfort zone for
47.3% of total hours in the year are outside neutral comfort zone. natural ventilation may be °C������������������������������
around the neutral zone in a
The percentage out of neutral comfort zone is highest in May up situation where the possibility of changing clothing or activity
to 83.4%. The lowest percentage 13.4% occurs in February. is small, and no much air movement is provided. In situations
Wang Liping and Wong Nyuk Hien
where these adaptive opportunities are available and appropriate,
the comfort zone may be considerably wider. Therefore, indoor
environment is applicable to achieve satisfactory thermal comfort
by natural ventilation in Singapore with smart fa����������������
çade designs in
residential building for energy saving and sustainability.
References
ASHRAE (2004). Thermal Environmental Conditions for Human
Occupant. Atlanta: American Society of Heating, Refrigeration and
Air-conditioning Engineers.
Baker, N., & Standeven, M. (1996). Thermal comfort for free-running
buildings. Energy and Buildings, 23, 175-182.
Brager, G.S., & De Dear, R.J. (1998). Thermal
��������������������������������
adaptation in the built
environment: a literature review. Energy and Buildings, 27, 83-96.
De Dear, R.J., & Brager, G.S. (1998). Developing
��������������������������������
an adaptive model of
thermal comfort and preference: Field studies of thermal comfort and
adaptation. ASHRAE Technical Data Bulletin, 14(1), 27-49; also in
ASHRAE Transactions, 104(1), 145-167.
De Dear, R.J., & Brager, G.S. (2002). Thermal comfort in naturally
ventilated buildings: revisions to ASHRAE Standard 55. Energy and
Buildings, 34, 549-561.
Chrenko, F.A. (1953). Heated ceilings and comfort. Journal of the Institute
of Heating and Ventilating Engineers, 20, 375-396; 21, 145-154.
Dutt, A.J., de Dear R.J., & Parthiphan, K. (1992). Full scale and
model investigation of natural ventilation and thermal comfort in a
building. Journal of Wind Engineering and Industrial Aerodynamics,
44, 2599-2609.
Natural Ventilation and Thermal Comfort 233
Fanger, P.O., & Toftum, J. (2002). Extension of the PMV model to non-
air-conditioned buildings in warm climates. Energy and Buildings,
34, 533-536.
Feriadi, H. (2003). Thermal comfort for natural ventilated residential
buildings in the tropical climate. Unpublished PhD Dissertation,
National University of Singapore.
Mallick, F.H. (1996). Thermal comfort and building design in the tropical
climate. Energy and Buildings, 23, 161-167.
Nicol, J.F., & Humphreys, M.A. (1973). Thermal comfort as part of a
self-regulating system. Building Research and Practice (J CIB), 6(3),
191-197.
Nicol, J.F., & Humphreys, M.A. (2002). Adaptive thermal comfort and
sustainable thermal standards for buildings. Energy and Buildings,
34, 563-572.
Nicol, F. (2004). Adaptive thermal comfort standards in the hot-humid
tropics. Energy and Buildings, 36(7), 628-637.
Wang L., & Wong, N.H. (2005). Thermal analysis of climate environments
based on weather data in Singapore for naturally ventilated buildings.
Unpublished paper presented at the 10th International Conference
on Indoor Air Quality and Climate, Beijing, 2005.
Wong, N.H., & Khoo, S.S. (2003). Thermal comfort in classrooms in
the tropics. Energy and Buildings, 35, 337-351.
Xia, Y.Z., Zhao, R.Y., & Jiang, Y. (1999). Thermal comfort in naturally
ventilated houses in Beijing. Heating, Ventilation, Air-conditioning
and Refrigeration, 29(4), 1-5 (in Chinese).