University of Southern California Dissertations and Theses

Improving thermal comfort in residential spaces in the wet tropical climate zones of India using passive cooling techniques: a study using computational design methods

(USC Thesis Other)

Improving thermal comfort in residential spaces in the wet tropical climate zones of India using passive cooling techniques: a study using computational design methods

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content
Copyright 2012 Priyanka Padmanabhan Nayar

IMPROVING THERMAL COMFORT IN RESIDENTIAL SPACES IN THE WET TROPICAL
CLIMATE ZONES OF INDIA USING PASSIVE COOLING TECHNIQUES: A STUDY USING
COMPUTATIONAL DESIGN METHODS

by

Priyanka Padmanabhan Nayar

A Thesis Presented to the
FACULTY OF THE USC SCHOOL OF ARCHITECTURE
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
MASTER OF BUILDING SCIENCE

December 2012

ii
ACKNOWLEDGEMENTS
A few pages doesn’t suffice to express my thanks to various people who had helped me shape this
research by being a part of it knowing or unknowingly. Though I have named only a few people
here, I have utmost gratitude to all others who had supported me in this research at one time or
other.
I would like to express my heartfelt gratitude to my committee members Prof. Ed Woll, Prof
Peter Simmonds and Prof. Pablo La Roche for their constant support and encouragement. Their
mentorship had been most valuable in shaping this research all through the way.
A special thanks to Prof. Doug Noble and Prof. Karen Kensek for providing helpful comments
and initial guidance for the research. Also, I would like to thank all the faculty members at the
USC School, of Architecture for their support. Prof. Anders Carlson and Prof. Marc Schiler had
been most helpful in providing with the instruments required for onsite data measurements and
related support.
I would also like to thank all my friends in the MBS department, who had always been there to
help with their insightful thoughts and comments. I also acknowledge and thank the support by
the staff of USC school of Architecture at various stages of my thesis submission process.
This research would have been impossible if not for my sister and my brother in law, Gadha and
Harikrishnan, who had opened their house for my study. I can’t thank you enough for your faith
in me and constant support in all my endeavors, especially for this research. Thank you for going
without air-conditioning as much as possible for the purpose of this research. Extra special thanks
for letting me put up tens of tiny blinking instruments, throughout your beautifully set home and

iii
ignoring the ugly patches that the scotch tapes left behind when they came off. I would also like
to thank my nephews, Nandu and Ramu and their grandparents, Mr Sanjayan Nair and Mrs. Sree
Kumari for being most amicable about the whole process. A special thanks to Mani, for helping
me put up and take down the instruments so many times.
Words are not enough to thank my husband, Ajith for being the constant source of
encouragement, support and unconditional love. This research wouldn’t have materialized
without your insights and feedback at every stage of the work. Thank you for everything and
more.
Last but not the least; I would also like to thank the rest of my family and all my friends for their
well wishes and support.

iv
TABLE OF CONTENTS
ACKNOWLEDGEMENTS II
LIST OF TABLES VIII
LIST OF FIGURES IX
ABSTRACT XX
EXTENDED ABSTRACT XXI
HYPOTHESIS XXI
METHODOLOGY XXI
RESULT XXII
CHAPTER 1: INTRODUCTION 1
1.1 THE STATE OF KERALA: A TROPICAL WET CLIMATE ZONE 2
1.2 THERMAL COMFORT 5
1.2.1 GRAPHIC COMFORT ZONE METHOD 7
1.2.2 COMPUTER MODEL METHOD 8
1.2.3 GRAPHICAL ELEVATED AIR SPEED MODEL 9
1.2.4 SET METHOD 10
1.2.5 ADAPTIVE MODEL OF THERMAL COMFORT 11
1.2.5.1 OPTIONAL METHOD FOR DETERMINING ACCEPTABLE
THERMAL CONDITIONS IN NATURALLY CONDITIONED
SPACES 12
CHAPTER 2: THE STUDY OF THERMAL COMFORT IN INDIAN CONTEXT 14
2.1 PMV-PPD METHOD 15
2.2 ADAPTIVE MODEL FOR THERMAL COMFORT. 17
2.2.1 USING MEAN MONTHLY OUTDOOR TEMPERATURE. 17
2.2.2 USING WEIGHTED RUNNING WEEKLY MEAN TEMPERATURE. 18
2.3 TROPICAL SUMMER INDEX (TSI) 19
2.4 COMPARISON OF THE FOUR METHODS OF DEFINING THERMAL
COMFORT 21
2.5 PSYCHOMETRIC CHART SHOWING THERMAL COMFORT ZONES
DEFINES BY FOUR METHODS FOR THIRUVANANTHAPURAM,
KERALA. 23
CHAPTER 3: RESEARCH METHODOLOGY 24
3.1 INTRODUCTION 24
3.1.1 PAST STUDY DONE ON THE TOPIC 24

v
3.2 STEPS INVOLVED IN METHODOLOGY 28
3.2.1 LIST OF PASSIVE COOLING STRATEGIES APPLICABLE TO
CLIMATE OF KERALA 28
3.2.2 BUILDING SELECTION AND DATA COLLECTION 33
3.2.3 SELECTING SOFTWARE TO SIMULATE THE BUILDING AND
CALIBRATING SIMULATED BUILDING 33
3.2.4 APPLYING STRATEGIES TO IMPROVE THERMAL COMFORT 34
3.2.5 MAKE A LIST OF SUGGESTIONS FOR IMPROVED THERMAL
COMFORT IN THIS CLIMATE 34
CHAPTER 4: CASE STUDY BUILDING 35
4.1 DESCRIPTION OF SELECTED BUILDING 35
4.2 HOBO DATA LOGGERS 39
4.2.1 RELATIVE CALIBRATION OF HOBOS 40
4.2.2 POSITIONING OF HOBOS 44
4.3 ANALYSIS OF COLLECTED DATA 48
4.3.1 ANALYSIS OF OUTDOOR DATA 48
4.3.2 ANAYSIS OF THE INDOOR DATA 50
4.3.2.1 THERMAL CONDITIONS IN THE SELECTED ROOM 51
4.3.2.2 THERMAL COMFORT IN THE SELECTED ROOM 55
CHAPTER 5:UNDERSTANDING THE HEAT TRANSFER PROCESSES
IN THE BUILDING 56
5.1 HEAT TRANSFER PROCESSES 58
5.1.1 CONDUCTION 58
5.1.2 CONVECTION 59
5.1.3 RADIATION 59
5.2 HEAT TRANSFER CALCULATIONS ASSUMING SIMPLE MODELS 60
5.2.1 EXPERIMENTAL DATA 61
5.3 HEAT TRANSFER THROUGH THE ROOF 63
5.4 HEAT TRANSFER THROUGH THE WALLS 68
5.5 DIRECT SOLAR HEAT GAIN THROUGH THE WINDOWS 73
5.6 PUTTING IT TOGETHER: A HEAT TRANSFER BUDGET
FOR THE ROOM 76
CHAPTER 6: BUILDING SIMULATION AND ANALYSIS 79
6.1 SOFTWARE SELECTION 79
6.2 WEATHER DATA FOR SIMULATION 79
6.3 THE DESIGN BUILDER MODEL 80

vi
6.3.1 MODEL CONSTRUCTION DETAILS 82
6.4 CALIBRATION OF THE MODEL 84
CHAPTER 7: APPLICATION OF STRATEGIES FOR IMPROVING
THERMAL COMFORT 93
7.1 STRATEGIES OF PASSIVE COOLING APPLIED 95
7.2 ANALYSIS OF EFFECTS OF APPLICATION OF PASSIVE COOLING
STRATEGY IN DESIGN BUILDER MODEL USING MEAN HOURLY
AVERAGED OUTDOOR WEATHER DATA 97
7.2.1 IMPROVED CROSS VENTILATION WITH WINDOWS OPENED FOR
24 HOURS. 97
7.2.2 IMPROVED CROSS VENTILATION WITH WINDOWS OPENED
ONLY DURING NIGHT TIME 99
7.2.3 WHITEWASHED FLAT ROOF 101
7.2.4 PROVIDING SHADE OVER THE CEILING WITH A
FLAT ROOF ABOVE. 104
7.2.5 PROVIDING A NON VENTILATED ATTIC 107
7.2.6 PROVIDING SHADE OVER THE CEILING WITH VENTILATED
PITCHED ROOF ABOVE. 110
7.2.7 PROVISION OF OPENING NEAR THE CEILING ON ONE WALL 113
7.2.8 PROVISION OF OPENING NEAR THE CEILING ON BOTH WALLS 115
7.2.9 SHADING WITH EXTERNAL LOUVERS 118
7.2.10 PROVIDING SINGLE GLAZED LOW E WINDOWS 120
7.2.11 PROVIDING DOUBLE GLAZED LOW E WINDOWS 122
7.2.12 PROVIDING SHADING WITH VERANDA 123
7.2.13 COMPARISON OF EFFECTIVENESS OF EACH STRATEGY
USING PSYCHOMETRIC CHART. 126
7.2.14 PLOTTING OUTPUT FOR SIMULATIONS USING ACTUAL
MEASURED WEATHER DATA AT SITE 126
7.2.14.1 COMFORT ZONE DEFINED WITHOUT AN UPPER RANGE FOR
HUMIDITY 126
7.2.14.2 COMFORT ZONE LIMITED TO 70% HUMIDITY AS
UPPER RANGE. 130
7.2.15 PLOTTING OUTPUT FOR SIMULATIONS USING TMY2 DATA 131
7.2.15.1 COMFORT ZONE DEFINED WITHOUT AN UPPER RANGE FOR
HUMIDITY 131
7.2.15.2 COMFORT ZONE LIMITED TO 70% HUMIDITY
AS UPPER RANGE 135
7.2.16 CATEGORIZING THE STRATEGIES ON THEIR EFFECTIVENESS

vii
ON IMPROVING THERMAL COMFORT 136
CHAPTER 8: PARAMETRIZING THE INDOOR TEMPERATURE IN TERMS OF
WINDOW OPENING AREA 141
CHAPTER 9: CONCLUSION 145
9.1 SUMMARY 145
9.2 LIMITATIONS 146
9.3 CONCLUSION 147
9.4 FUTURE WORK 149
BIBLIOGRAPHY 150
APPENDIX: SCHEDULES IN DESIGN BUILDER 152

viii
LIST OF TABLES
Table 1.1 ASHRAE Thermal sensation scale 5
Table 1.2 . Acceptable Thermal Environment for General Comfort4 7
Table 2.1 Comparison of advantages and disadvantages of various models of thermal
comfort. 22
Table 4.1 Classes of data for thermal comfort field research 45
4.2 Position of HOBOs on site 47
Table 5.1: Summary of the parameters used for the calculation of the heat transfer
through the roof. 65
Table 5.2: Summary of the parameters used for the calculation of the heat transfer
through the wall. 70
Table 7.1:Strategies in the decreasing order of their effectiveness( from the top) when
the simulations were run using the Actual measured weather data. Columns 2 , 3, 4
and 5 shows the percentage of time the room is in inside the comfort zone for the
whole year, January, April and June respectively. 137
Table 7.2: Similar to Table 7.1, except that the simulations were run using the TMY2
weather data. 138
Table 7.3: This table shows the improvement in the percentage of time the room was in
comfort zone over that of the base model (Reference) when the simulations were
run using the actual measured weather data. 138
Table 7.4: Similar to Table 7.3, except that the simulations were run using the TMY2
weather data. 139

ix
LIST OF FIGURES
Figure 1.1 The climate zones of India 3
Figure 1.2 Predicted percentage dissatisfied (PPD) as a function of predicted mean vote
(PMV).4 6
Figure 1.3 The Graphic Comfort Zone Method: Acceptable range of operative temperature
and humidity for spaces where metabolic rate is 1.1 met and clothing insulation values
are 0.5 and 1 clo4 8
Figure 1.4 Air speed required to offset increased air and radiant temperature4 10
Figure 1.5 Acceptable range of operative temperature and air speeds for the comfort zone
shown in Figure 1.3, at (example) humidity ratio 0.010.4 11
Figure 1.6 Acceptable operative temperature ranges for naturally conditioned spaces.4 12
Figure 2.1 Psychometric chart showing the distribution of outdoor air temperature for
Thiruvananthapuram, Kerala. 16
Figure 2.2 Acceptable Indoor Operative Temperature range for naturally conditioned spaces,
computed using the adaptive model of thermal comfort. The colored lines correspond to
mean monthly temperature of each month of the year for the city of Thiruvanathapuram,
Kerala 17
Figure 2.3 Adaptive Model of Thermal comfort based on weighted running weekly mean
temperature 18
Figure 2.4 The comfort zone marked in the psychometric chart with Tropical Summer index
lines 21
Figure 2.5 Psychometric chart showing the thermal comfort ranges defined by the four
methods for the city of Thiruvananthapuram 23
Figure 3.1 An example of traditional residence in Kerala 26
Figure 3.2 A Typical Traditional residence in Kerala
3
26
Figure 3.3 Examples of modern residences in Kerala 27
Figure 3.4 3D Chart of monthly Dry Bulb Temperature distribution of Trivandrum. 29

Figure 3.5 Monthly Dry Bulb Temperature distribution of Trivandrum. The grey band shows
the comfort range of temperature according to ASHREA PMV-PPD Method.
20
29

x
Figure 4.1: Plan of first floor (left) and second floor (right) of the modern residential building
in Thiruvananthapuram that was chosen for the case study. 37
Figure 4.2: Sectional detail of the residential building. 38
Figure 4.3 View from East (left) and South (right) of the building chosen for case study. 38
Figure 4.4 View of first floor living space and veranda Positions of the HOBOs are marked
with red circles in all pictures. 39
Figure 4.5: View of the second floor bedroom and position of the HOBO on the first floor
staircase landing. 39
Figure 4.6: Temperature and humidity measurements from ten HOBOs during the calibration
period, plotted against time. All the HOBOs are placed in the same environment, and the
difference in measured data between different HOBOs is an indication of the relative
calibration error. 41
Figure 4.7: Temperature and humidity measurements averaged over all the HOBOs at any
point in time, and maximum/minimum deviation from the mean 41
Figure 4.8: The relative error Δxi(t) of each HOBO from the mean value recorded by all
HOBOs at any point in time. The horizontal axis shows the time in days. 43
Figure 4.9: This figure indicates the position of HOBOs in the Ground Floor and First Floor. 47
Figure 4.10: Temperature (left) and relativity humidity (right) measured by HOBO 1 in
January, April and June plotted against the hour of the day. The thick dots correspond to
hourly temperature (or humidity) averaged over 30 days, while the error bars show the
minimum and maximum measured hourly temperature (or humidity) over the
measurement period. Hourly averaged temperature was used in order to avoid transient
fluctuations in the data due to everyday weather variations. Each data point correspond
to the hourly temperature measured by the HOBO averaged over a month. For eg. The
red dot corresponding to 5 am in temperature - time graph, represents the average of
temperature measured at 5 am for the whole month. 49
Figure 4.11: Mean hourly temperature (left) and relative humidity (right) from January
recorded by HOBOs placed at four different living spaces in the building (see legend).
The thick dots represent the mean values and the error bars correspond to the maximum
and minimum values of temperature/humidity measured over the measurement period.
Mean Temperature was used for the same reason as explained in Figure 4.10 50
Figure 4.12: Mean hourly temperature (left) and relative humidity (right) from January
recorded by HOBOs placed inside/outside bedroom1 (see legend). The thick dots
represent the mean values and the error bars correspond to the maximum and minimum
values of temperature/humidity measured over the measurement period. 51

xi
Figure 4.13: Mean hourly temperature (left) and relative humidity (right) from April recorded
by HOBOs placed inside/outside bedroom1 (see legend). The thick dots represent the
mean values and the error bars correspond to the maximum and minimum values of
temperature/humidity measured over the measurement period. 53
Figure 4.14: The temperature data measured by for H1, H2,H3 and H4 over a month in June
plotted against the hour of the day. The thick dots correspond to hourly temperatures
averaged over 30 days, while the error bars show the minimum and maximum measured
temperature over the 30 days. 54
Figure 4.15: Psychometric chart showing the thermal comfort ranges defined using four
methods. The blue dots correspond to the temperature and humidity data measured by
H3, at a height of 2' from FFL in January, April and June respectively. 55
Figure 5.1: The temperature data measured by four HOBOs over 18 days (27-Dec-2010 to 13
Jan-2011) plotted against the hour of the day. The thick dots correspond to hourly
temperatures averaged over 18 days, while the error bars show the minimum and
maximum measured temperature over the 18 days. 62
Figure 5.2: Schematic representation of the different modes of heat transfer through the roof.
Qsolar represents the heat gain of the roof from the direct solar radiation (short-wave),
Qrad-out represents the net long-wave radiation on the outer surface of the roof (the
difference between the heat gain of the roof from atmospheric radiation and the heat loss
due to radiation from the roof), Qrad-in represents the net long-wave radiation on the
inner surface of the roof, while Qc-out and Q c-in represent the convective heat transfer
at the outer and inner surface of the roof, respectively. 63
Figure 5.3: The top panel shows the heat transfer rate per unit area through the roof by means
of different heat transfer mechanisms: Direct solar radiation, net long-wavelength
radiation on the outer surface, convection on the outer surface, net long-wavelength
radiation on the inner surface, convection on the inner surface. Also plotted are the net
heat transfer on the outer surface, the net heat transfer on the inner surface and the sum
of these. Positive values indicate heat gain to the roof and negative values indicate heat
loss from the roof. The bottom panel shows the expected temperature of the roof T
roof
calculated from the heat transfer through the roof (thin black line) along with the
measured temperature of the ceiling (Cyan line). Also shown are the temperature of the
atmosphere and that of the room. 67
Figure 5.4: Schematic representation of the different modes of heat transfer through the walls.
Q
rad-out
represents the net long-wave radiation on the outer surface of the wall (the
difference between the heat gain of the wall from atmospheric radiation and the heat loss
due to radiation from the wall), Q
rad-in
represents the net long-wave radiation on the
inner surface of the wall, while Q
c-out
and Q
c-in
represent the convective heat transfer at
the outer and inner surface of the wall, respectively, and Q
cond
represents the heat
conduction through the wall. 68

xii
Figure 5.5: The top panel shows the heat transfer rate per unit area through the wall by means
of different heat transfer mechanisms: Net long-wavelength radiation on the outer
surface, convection on the outer surface, net long-wavelength radiation on the inner
surface, convection on the inner surface. Also plotted are the net heat transfer on the
outer surface, and the net heat transfer on the inner surface. Positive values indicate heat
gain to the wall and negative values indicate heat loss from the wall. The bottom panel
shows the expected temperature of the wall calculated from the heat transfer through the
wall (thin black line) along with the measured temperature of the wall (Green line).
Also shown are the temperature of the atmosphere and that of the room. 72
Figure 5.6: Schematic representation of the direct solar heat gain through the windows. Qsolar
represents the direct solar radiation (short-wave) through the window. The direct solar
radiation transfers heat energy to the floor, which in turn, radiates this energy back to the
room by means of long-wave radiation. Qrad-f represents the net heat transfer due to
long-wave radiation from/to the floor (difference between heat gain of the floor from
radiation from the room and the heat loss due to radiation from the floor), and Qconv-f
represents the convective heat transfer from the floor. 73
Figure 5.7: Top panel shows the heat transfer to and from the floor by different mechanisms:
long-wavelength radiation exchange between the floor and the air in the room,
convective heat transfer between the floor and the air in the room, and direct solar
radiation. Clearly the heat transfer during the day is dominated by the direct solar
radiation. The two peaks of the radiation are due to the radiation coming from three
different windows on the East and South walls. Positive values correspond to heat gain
by the floor and negative values to heat loss by the floor. The bottom panel shows the
calculated temperature of the floor along with the measured temperature of the
atmosphere, the ceiling, air inside the room and the inner surface of the South wall. 75
Figure 5.8: Top panel shows the heat transfer to and from the room by different mechanisms:
long-wavelength radiation exchange between the air inside and outside the room through
windows, radiation exchange with the floor, the roof and the walls, convective heat
transfer through the (open) windows, convective heat transfer with the surface of the
floor, the roof and the walls. (Note that the high heat gain from the floor from 8AM to
2PM is caused by the direct solar radiation absorbed by the floor). Positive values
indicate heat gain to the room and negative values indicate heat loss from the room. The
bottom panel shows the predicted ambient temperature of the room (thin black trace)
along with the measured ambient temperature of the room (blue traces). Also shown are
the predicted temperature of the floor, and the measured temperatures of the atmosphere,
the ceiling and the inner surface of the wall. 77
Figure 6.1 View of the model from the South direction 80
Figure 6.2 View of the model from the South direction 81
Figure 6.3: View of the model from the East direction. 81

xiii
Figure 6.4 Plan showing the Building Zones in DB in the first Floor(Left) and Second Floor
(Right) 82
Figure 6.5 Details of the wall construction in DB. 83
Figure 6.6 Details of roof construction in DB. 83
Figure 6.7 Details of glazing in the model in DB. 84
Figure 6.8 Temperature measured from the HOBO in bedroom1 on the second floor and the
simulated temperature for January plotted against the hours of the day. Also shown is the
outdoor temperature. 86
Figure 6.9 Temperature measured from the HOBO in bedroom1 on second floor and the
simulated temperature for April plotted against the hours of the day. Also shown is the
outdoor temperature. 86
Figure 6.10 Temperature measured from the HOBO in bedroom1 on the second floor and the
simulated temperature for June plotted against the hours of the day. 87
Figure 6.11 The top panel shows the temperature plotted against hour of the day, the middle
panel shows the heat balance plotted against hour of the day (the points above zero
indicate heat gain into the room and below zero indicate the heat loss from the room)
and the lower panel shows the air changes/hour inside the room plotted against hour for
the day. These plots are from January. 88
Figure 6.12The top panel shows the temperature plotted against hours of the day, the middle
panel shows the heat balance plotted against hour of the day (the points above zero
indicate heat gain into the room and below zero indicate the heat loss from the room)
and the lower panel shows the air changes/hour inside the room plotted against hour for
the day. These plots are from April. 91
Figure 6.13The top panel shows the temperature plotted against hours of the day, the middle
panel shows the heat balance plotted against hour of the day (the points above zero
indicate heat gain into the room and below zero indicate the heat loss from the room)
and the lower panel shows the air changes/hour inside the room plotted against hour for
the day. These plots are from June. 92
Figure 7.1 A plot showing the comparison of outdoor dry bulb temperature for a week of
January, obtained from three different weather files. The blue line shows the real time
weather data file (the weather data file for the year 2011)obtained from energy plus
website. The green line shows the TMY2 Data obtained from energy plus website which
shows the historic average dry bulb temperature over many years. The red line shows the
outdoor dry bulb temperature measured at site using Hobos. The measured weather data
was always much higher than the 95

xiv
Figure 7.2 : 3D view showing the position of the windows in the selected room. The windows
are open 24 hours for this simulation (though it does not show in the 3D view) 97
Figure 7.3: shows the comparison of indoor temperature in the baseline model and modified
model (windows open 24 hours) over 24 hours of the day in January. The red line
indicates the room temperature of the simulated model when the windows were opened
24 hours. The blue line indicates the the room temperature and the green line indicates
the outdoor dry bulb temperature. 98
Figure 7.4: is same as Figure 7.3, except that graph shows the values for the month of April. 99
Figure 7.5: is same as Figure 7.3, except that graph shows the values for the month of June. 99
Figure 7.6: shows the comparison of indoor temperature in the baseline model and modified
model (windows opened only during night time) over 24 hours of the day in January.
The red line indicates the room temperature of the simulated model when the windows
were opened only during night time. The blue line indicates the indoor dry bulb
temperature for the base model and the green line indicates the outdoor dry bulb
temperature. 100
Figure 7.7: is same as Figure 7.6, except that graph shows the values for the month of April. 100
Figure 7.8: is same as Figure 7.6, except that graph shows the values for the month of June. 101
Figure 7.9 shows the Design Builder material property window showing thermal absorptivity
and emissivity values provided on the outer surface of the white washed roof. 102
Figure 7.10: shows the comparison of indoor temperature in the baseline model and modified
model (with roof white washed) over 24 hours of the day in January. The red line
indicates the room temperature of the simulated model when the roof is white washed.
The blue line indicates the indoor dry bulb temperature for the base model and the green
line indicates the outdoor dry bulb temperature. 102
Figure 7.11: is same as Figure 7.10, except that graph shows the values for the month of April. 103
Figure 7.12: is same as Figure 7.10, except that graph shows the values for the month of June. 103
Figure 7.13: shows the heat gain and loss in the room through various elements such as walls,
roof, glazing etc. The horizontal; axis is the hours of the day and the vertical axis is the
heat balance in kBtu/hr. If the vertical axis is positive, it indicates the heat gain into the
room and negative values indicate heat loss from the room. The upper panel shows the
heat balance graph of the base model and the lower panel shows the heat balance plot of
the model with white washed roof. 104
Figure 7.14 : 3D view showing flat roof above the ceiling of the room. 105

xv
Figure 7.15 shows the comparison of indoor temperature in the baseline model and modified
model (flat shade over the ceiling) over 24 hours of the day in January. The red line
indicates the room temperature of the simulated model when the room ceiling is shaded
with a flat roof above. The blue line indicates the indoor dry bulb temperature for the
base model and the green line indicates the outdoor dry bulb temperature. 105
Figure 7.16 is same as Figure 7.15, except that graph shows the values for the month of April. 106
Figure 7.17 is same as Figure 7.15, except that graph shows the values for the month of June. 106
Figure 7.18: 3D view of the house with double roof ( unoccupied attic space) and no attic
ventilation 108
Figure 7.19 shows the comparison of indoor temperature in the baseline model and modified
model( pitched roof over room ceiling with minimal ventilation in the air space between
roof and ceiling) over 24 hours of the day in January. The red line indicates the room
temperature of the simulated model with the unventilated attic space above. The blue
line indicates the indoor dry bulb temperature for the base model and the green line
indicates the outdoor dry bulb temperature. 108
Figure 7.20 is same asFigure 7.19, except that graph shows the values for the month of April. 109
Figure 7.21 is same as Figure 7.19Figure 7.20, except that graph shows the values for the
month of June 109
Figure 7.22 shows the heat gain and loss in the room through various elements such as walls,
roof, glazing etc. The horizontal; axis is the hours of the day and the vertical axis is the
heat balance in kBtu/hr. If the vertical axis is positive, it indicates the heat gain into the
room and negative values indicate heat loss from the room. The upper panel shows the
heat balance graph of the base model. In the lower panels which shows the heat balance
plot of the model with shaded roof in January and April, the reduction in heat gain
through the roof can be clearly seen when compared to te heat balance plot of the base
model. . 110
Figure 7.23: 3D view of the house open sided pitched roof above the room ceiling. 111
Figure 7.24 shows the comparison of indoor temperature in the baseline model and modified
model over 24 hours of the day in January. The red line indicates the room temperature
of the simulated model when the pitched roof. Shading the ceiling of the room below has
open sides which facilitates ventilation. The blue line indicates the indoor dry bulb
temperature for the base model and the green line indicates the outdoor dry bulb
temperature. 111
Figure 7.25 is same as Figure 7.24, except that graph shows the values for the month of April. 112
Figure 7.26 is same as Figure 7.24, except that graph shows the values for the month of April. 112

xvi
Figure 7.27 3D view of the house showing the opening near the ceiling. 113
Figure 7.28 shows the comparison of indoor temperature in the baseline model and modified
model (opening near the ceiling in one wall) over 24 hours of the day in January. The
red line indicates the room temperature of the simulated model which has an opening
near the ceiling on one wall. The blue line indicates the indoor dry bulb temperature for
the base model and the green line indicates the outdoor dry bulb temperature. 114
Figure 7.29 is same as Figure 7.28, except that graph shows the values for the month of April. 114
Figure 7.30 is same as Figure 7.28 except that graph shows the values for the month of June. 115
Figure 7.31 3D view of the house showing the openings near the ceiling on both walls. 116
Figure 7.32 shows the comparison of indoor temperature in the baseline model and modified
model (openings near the ceiling on both walls) over 24 hours of the day in January. The
red line indicates the room temperature of the simulated model which has an opening
near the ceiling on both walls. The blue line indicates the indoor dry bulb temperature
for the base model and the green line indicates the outdoor dry bulb temperature. 116
Figure 7.33 is same as Figure 7.32, except that graph shows the values for the month of April. 117
Figure 7.34 is same as Figure 7.32Figure 7.28 except that graph shows the values for the
month of June. 117
Figure 7.35 3D view of the house showing the external shading on windows 118
Figure 7.36 shows the comparison of indoor temperature in the baseline model and modified
model (external louvers shading the ) over 24 hours of the day in January. The red line
indicates the room temperature of the simulated model which has external louvers
shading the windows. The blue line indicates the indoor dry bulb temperature for the
base model and the green line indicates the outdoor dry bulb temperature. 118
Figure 7.37 is same as Figure 7.36, except that graph shows the values for the month of April. 119
Figure 7.38 is same as Figure 7.36 except that graph shows the values for the month ofJune. 119
Figure 7.39 shows the comparison of indoor temperature in the baseline model and modified
model (single glazed Low E windows) over 24 hours of the day in January. The red line
indicates the room temperature of the simulated model which has Single glazed Low E
windows. The blue line indicates the indoor dry bulb temperature for the base model and
the green line indicates the outdoor dry bulb temperature. 120
Figure 7.40 same as Figure 7.39except that graph shows the values for the month of April. 121
Figure 7.41 same as Figure 7.39 except that graph shows the values for the month of June. 121

xvii
Figure 7.42 shows the comparison of indoor temperature in the baseline model and modified
model (single glazed Low E windows) over 24 hours of the day in January. The red line
indicates the room temperature of the simulated model which has double glazed Low E
windows. The blue line indicates the indoor dry bulb temperature for the base model and
the green line indicates the outdoor dry bulb temperature. 122
Figure 7.43 same as Figure 7.42 except that graph shows the values for the month of April. 122
Figure 7.44 same as Figure 7.42 except that graph shows the values for the month of June. 123
Figure 7.45 3D view of the house showing the effect of veranda using an 8’ wide overhang. 124
Figure 7.46 shows the comparison of indoor temperature in the baseline model and modified
model (with a 8’ wide projecting slab) over 24 hours of the day in January. The red line
indicates the room temperature of the simulated model which has 8’ wide projecting slab
outside the walls. The blue line indicates the indoor dry bulb temperature for the base
model and the green line indicates the outdoor dry bulb temperature. 124
Figure 7.47 same as Figure 7.46 except that graph shows the values for the month of April. 125
Figure 7.48 same as Figure 7.46 except that graph shows the values for the month of June 125
Figure 7.49 This figure shows a portion of psychometric chart zoomed out for better clarity.
The shaded region represents the range of temperature and humidity in which people are
comfortable. The non shaded areas are the areas outside comfort zone. Each of the
colored dots corresponds to a data point that represents a particular indoor temperature
and relative humidity. Each color shows the data points from the simulation when a
particular strategy is applied. The black cross (Reference) represent the data points from
the base model. The colored dots in this plot correspond to the data points from the
simulations using the actual measured weather data in site, for the January. 127
Figure 7.50: Same as Figure 7.49, but the dots correspond to data points for April. 128
Figure 7.51: Same as Figure 7.49, but the dots correspond to output data points for June. 128
Figure 7.52: This figure shows the percentage of time the room is in the comfort zone (when
the comfort zone is not limited by an upper limit of relative humidity) in the vertical axis
and each of the passive cooling strategies applied to the base model on the horizontal
axis. The three months is represented in three colors. For eg. When the room was cross
ventilated only during night time, 80% of the time, the room was comfortable in January
and June, whereas in April, the room was comfortable only 55 % of the time. Likewise
the effect of each strategy can be read from the graph. The percentage of time of time the
room fall is in the comfort zone is expressed as ‘comfort fraction’. The reference
represents the comfort conditions of the room in the base model. 129

xviii
Figure 7.53:Same as Figure 7.52, except that in this case comfort zone has an upper limit for
relative humidity. ie. A data point is considered inside the comfort range if it falls in the
shaded area below the 70 % relative humidity curve in the shaded region of the
psychometric chart in Figure 7.49, Figure 7.50 or Figure 7.51 . 131
Figure 7.54: The figure shows a portion of psychometric chart zoomed out for better clarity.
The shaded region represents the range of temperature and humidity in which people are
comfortable. The non shaded areas are the areas outside comfort zone. Each of the
colored dot correspond to a data point that represents a particular indoor temperature and
relative humidity. Each color shows the data points from the simulation when a
particular strategy is applied. The black cross represents the data points from the base
model. The colored dots in this plot represent the output data points from the simulations
using TMY2 weather data from, for January. 132
Figure 7.55: Same as Figure 7.54, but the dots correspond to output data points for April. 132
Figure 7.56: Same as Figure 7.54, but the dots correspond to output data points for June. 133
Figure 7.57: This figure is similar to Figure 7.52 and shows the percentage of time the room is
in the comfort zone (when the comfort zone is not limited by an upper limit of relative
humidity) in the vertical axis and each of the passive cooling strategies applied to the
base model on the horizontal axis. The three months is represented in three colors. 134
Figure 7.58: Same as Figure 7.57, except that comfort zone in this case has an upper limit for
relative humidity. ie. In this case a data point is considered inside the comfort range if it
falls in the shaded area below the 70 % relative humidity curve in the psychometric chart
in Figure 7.54, Figure 7.55 or Figure 7.56. 135
Figure 8.1 shows the exploded of the view room and the position of windows in the room. 142
Figure 8.2 : This figure shows the temperature in ˚F in the vertical axis and the hours of the
day in the horizontal axis. The different lines represent the hourly averaged indoor
temperature (except the orange line which is the outdoor temperature) of the room for 24
hours for the month of April. 142
Figure 8.3 This figure shows the rate of change of mean temperature difference between
indoors and outdoors with respect to the opening area of the window. The vertical axis
show the mean of the difference between outdoor temperature and indoor temperature.
The horizontal axis shows the area of window opening as a percentage of floor area of
the room. 144
Figure A1 Activity settings in Design Builder used for all simulations 151
Figure A2 Construction settings in Design Builder used for all simulations. 152
Figure A3 HVAC settings in Design Builder used for all simulations. Only heating and
cooling is turned off and Natural ventilation is on. 152

xix
Figure A 4 Light settings in Design Builder used for all simulations 153
Figure A5 Schedule for natural ventilation for all simulations. The space is naturally
ventilated 24 hours a day. The air movement and air changes in the room are controlled
by the closing and opening the windows. 153
Figure A6 General Window settings in Design Builder used for all simulations. The
percentage of opened glazing area and the duration during which it is opened is set
individually for each window. 154
Figure A7 Window settings for the south side window in the south east corner bedroom in the
second floor. 155
Figure A8 Schedule for opening and closing the south side window in the south east corner
bedroom in the second floor. 155
Figure A9 Window glazing data for simulations using single glazed Low E windows. 156
Figure A10 Window glazing data for simulations using double glazed Low E windows 156

xx
ABSTRACT
Kerala, the strip of land on the southwest coast of India falls under the category of wet tropical
climatic zone. Temperature in this coastal state by the Arabian Sea remains more or less constant
throughout the year, with no marked seasons, except the Monsoons. The traditional architecture
and construction practices of Kerala were well known for the use of natural and passive methods
for developing a comfortable thermal environment inside enclosed spaces. Over the years,
construction methods and materials have changed with technology and updated lifestyle
requirements. But remnants of some architectural design characteristics from the past linger on in
the current designs – mostly as design elements rather than as functional elements. The lack of
proper understanding of how the construction materials, techniques and architectural practices
interact with each other might produce negative effects in terms of thermal comfort level. This
motivates the need of a scientific study for improving thermal comfort with minimal mechanical
intervention. This study aims to investigate the factors (such as temperature, humidity and air
movement) that characterize thermal comfort in a particular house in Kerala. The goal is that the
study will make it possible to identify and suggest some modifications (in architecture,
construction materials etc.) to improve the thermal comfort inside the house. The study also
attempts to develop some general guidelines on improving thermal comfort using passive cooling
techniques in wet tropical climate zones of India. The study utilizes computer simulated models
of the building and parameterization tools within building information modeling (BIM) and
energy calculation software to calculate the thermal comfort parameters for various architectural
modifications. The study is particularly motivated by the current scenario of rising global
temperature and the increasing use of air conditioners in the developing world.

xxi
EXTENDED ABSTRACT
HYPOTHESIS – It is possible to improve the thermal comfort and to bring it closer to comfort
zones in enclosed residential spaces in the wet tropical climate zones of India, by using passive
cooling techniques.
METHODOLOGY
The first step is to identify the indexes that measure thermal comfort, as per accepted standards
(temperature, humidity, air movement, etc.). The importance of performing this study in Kerala is
motivated by climatology data and data showing the increased use of air conditioners in Kerala.
The next step is to identify the sample on which the hypothesis is being tested (i.e, a residence, as
opposed to commercial spaces, high rise buildings etc.) The reasons for selecting that sample:
one, with the recent emergence of an affluent middle class in India, this category of buildings has
an increasing trend of using air conditioners; two, since these spaces have fewer occupants, there
is a better chance of developing a thermal comfort zone acceptable to all of them. After having
decided on a building, temperature and humidity data are collected using appropriate instruments
for a specific period of time. The data analysis reveals the current thermal condition and helps to
identify the problems associated with it. These data are used to calibrate the computer simulated
model of the building against the original one, using standard techniques (e.g. ASHRAE, 2002-
ASHRAE Guideline 14 -2002).
Once the calibration is deemed acceptable, by systematically varying some parameters (e.g.,
construction materials, type of roof etc.) that affect the thermal comfort while keeping others
constant, the changes in the thermal comfort indexes (e.g., interior temperature and humidity) are
calculated using energy calculation software. The effect of each strategy was evaluated from the
fraction of time the building is in the comfort zone when the strategy was applied.

xxii
This exercise is used to identify the configurations producing acceptable range of thermal
comfort, from a list of design strategies in hot humid climate (that includes widely accepted
strategies and new suggestions). The hope is that this would help to characterize how different
architecture and materials interact with each other in developing a thermal comfort zone for that
particular building. It could also help quantify the effectiveness of the different strategies that we
consider. Using this as a baseline, an effort will be made to generalize the conclusions to similar
buildings in this particular climate zone.
RESULT
The end product is a calibrated simulation model of a building and the analysis that leads to
suggestions that could improve the thermal comfort inside the space.
Another outcome is a tabulated list of the architecture and construction techniques and materials
that would work together in that climate zone to bring the thermal comfort indexes to acceptable
limits by the use of passive cooling. This will contribute to providing guidelines for this category
of dwellings in wet tropical climate zones for reducing the carbon footprint of air-conditioners

1
CHAPTER 1: INTRODUCTION
India, with a total land area of 3,287,263 sq. km, encompasses a widely varying geography with
correspondingly varying climate zones. There are four major climate groups (which can be sub-
divided into 7 major climate types).
 Tropical Rainy Climate Group 
 Dry Climate Group 
 Humid Sub Topical Climate Group 
 Mountain Climate 
Tropical rainy climate is further subdivided into 1) Tropical monsoon rain forest and 2) Tropical
wet and dry or Savannah climate. The West Coastal Lowlands and the Western Ghats (the
mountain range along the western side of India), fall into the first subdivision. Due to widely
varying geographic features and climate zones, the habitat architecture in each region differs
widely. Traditional architecture of any region is significantly influenced by its landscape, climate
and culture. Also, the traditional styles fell in step with locally available labor and materials. In
the absence of modern technology to condition the enclosed spaces these steps ensured that the
spaces for living and circulation were tuned to human comfort. The advent of modern
technologies and materials provided new opportunities for molding spaces without the constraints
of the pre-modern era. This was a positive step in the development of habitats, providing better
solutions to overcome the limitations of traditional architecture. But, later, conditioned spaces
became an easy solution for providing thermal comfort inside buildings that were not well suited
to the environment. This trend not only created an imbalance in the natural flow of air and light in
built spaces, but has also added to the Greenhouse Gas emissions, which are causing one of the

2
biggest climatic challenges in human history. Much work has been done to assess the impact of
the building sector on Green House Gas Emissions. In India, the residential building sector alone
accounts for 12.5% of the total GHG emissions
1
. At the national level, an increase in surface
temperature of 0.4° has been noted.
2
The Third section of the National Action Plan on Climate
Change by the Government of India suggests a national mission for sustainable habitats
especially promoting energy efficiency in the residential sector.
1.1 THE STATE OF KERALA: A TROPICAL WET CLIMATE ZONE
The state of Kerala lies on the South West coastal strip of India flanked by the Western Ghats on
the East and the Indian Ocean on the West, between 7°N and 13°N. According the Koppen
classification, Kerala falls under the Tropical Wet Climate zone (Tropical monsoon rain forest
climate zone sub classification) in the Indian climate classification.
The strategic positioning of the Western Ghats range makes the climate of Kerala very different
from its neighbor states. Kerala has no marked seasons, other than the monsoons (which are
among the least understood climatological phenomena.) Kerala gets heavy rains from the the
South-West Monsoons (June to August) and from the North-East Monsoons (October and
November). The rest of the months are warm with temperatures ranging from 82°F to 90°F.

1
India Green House Emissions 2007, Ministry of Environment and Forest, Government of India.
2
National Action Plan on Climate Change, Government of India.

3

Figure 1.1 The climate zones of India
3

The temperatures peak from March to May which are the summer months. Being a coastal state
which receives the monsoons, the relative humidity is high: in the range of 63 to 85%. As with all
vernacular architecture designs, the traditional architecture of Kerala (dating from the early 10
th

century) followed a definite pattern that was in step with the climate and culture of the place.
Interestingly, research has supported the hypothesis that traditional buildings of Kerala do
provide an effective envelope to control the indoor temperature, thereby ensuring a controlled and
continuous airflow inside the building in order to achieve better thermal comfort even in summer
solely by passive control systems
4
.
Though the architecture and construction materials have changed significantly from the
traditional methods in accordance with changing needs of the community and technological

3
Picture Courtesy: Wikipedia, http://en.wikipedia.org/wiki/Climatic_regions_of_India
4
A.S. Dili, M.A. Naseer, T. Zacharia Vargese : Passive control methods of Kerala traditional architecture for
comfortable indoor environment: A comparative investigation during winter and summer,. Building and Environment
2010; 45:1134-1143

4
development, some traditional elements are retained. But mostly, when combined with modern
techniques and layout of the building, these serve as an aesthetic element rather than as an
effective feature for better conditioning of living spaces. For example, many houses in Kerala
boast of a courtyard, but most of them have sealed the open to sky area with large transparent
sheets of polycarbonate (mostly due to security reasons), which takes away the main ventilation
function of courtyards (though this still provides better day lighting to interior spaces). A
scientific, quantitative study is essential to better understand the effect of interaction of these
architecture elements and modern methods of construction.
The increasing trend of air conditioner use in the residential and commercial sectors is a point of
concern in terms of increased energy consumption and associated Greenhouse Gas Emissions.
The seasonal variations of Kerala have not changed drastically in the recent past. Living in a
region with typically high temperatures all the year round, the resident population has become
acclimatized to a level of high temperature and humidity with the help of some basic mechanical
ventilation equipment such as fans, whereas air conditioners were seen as a luxury. But the
economy boom in recent years has caused the development of an affluent middle class who can
afford air conditioners, resulting in their increased use in the residential sector. Now there is a
transitional stage from living in naturally ventilated spaces to fully conditioned spaces. If passive
solutions for better thermal comfort in residences are not explored at this point, this will soon be a
society totally accustomed to living in air conditioned spaces, without adaptability to the natural
environment of Kerala. The need to curb Greenhouse Gas emissions in the current world scenario
underlines the importance of a study exploring the possibility of improved thermal comfort
employing passive cooling techniques in Kerala.

5
1.2 THERMAL COMFORT
Thermal comfort is defined as “that condition of mind that expresses satisfaction with thermal
environment, related to air temperature, humidity and wind speed”.
5

4
The American Society of
Heating, Refrigerating and Air-conditioning Engineers (ASHRAE) describes thermal comfort as
'neutral' on a seven point scale ranging from hot to cold (see Table 1). “The science of thermal
comfort has been concerned with predicting what set of conditions (temperature, humidity,
airspeed etc.) corresponds most closely to this neutral feeling, and how tolerant people are of
deviations from it.”
6
Thermal comfort has been studied using human subjects as 'meters'; thermal
comfort is calibrated in “comfort zones”, the designation on a scale from hot to cold which best
describes the person's impression of the thermal environment.

Table 1.1 ASHRAE Thermal sensation scale
Errors are minimized by making these investigations in a climate chamber in which temperature
and humidity are controlled at steady values. Combining the results of these investigations and
theories of thermodynamics and physiology a formula to predict neutral conditions has been

5
ANSI/ASHREA Standard 55 - 2010
6
Krishnan A, Baker N,Yannas S, Szolkolay S V: Climate Responsive Architecture – A Design Handbook for Energy
Efficient Buildings.Tata McGraw-Hill Publishing Company Limited; 2001

6
devised. Ole Fanger proposed an expression for optimal thermal comfort involving metabolic
rate, clothing insulation and environmental conditions. The results have been presented in the
form of diagrams from which optimal comfort conditions can be read for a given metabolic rate
and clothing insulation level. Fanger further developed a method to predict the comfort vote
7
that
would arise from a given set of environmental conditions for given clothing and metabolic rate –
Predicted Mean Vote (PMV).
8

The PMV predicts the mean value of the votes of a large group of
people on the ASHRAE thermal sensation scale. PMV was further elaborated to predict the
percentage of population dissatisfied (PPD) with the environment (in terms of their comfort vote
on the ASHRAE scale). The PPD predicts the percentage of a large group of people likely to feel
either ‘too warm’ or ‘too cool’. The distribution of PPD is based purely on climate chamber
experiment observations.

Figure 1.2 Predicted percentage dissatisfied (PPD) as a function of predicted mean vote (PMV).
5

It can be understood from Figure 1.2 that percentage of people dissatisfied increases sharply as
PMV moves further from the neutral index of thermal comfort scale. ASHRAE defines the limits

7
Comfort Vote: The response of people to thermal environment in terms of thermal comfort scale.
8
Clear (Comfortable Low Energy Architecture). http://newlearn.info/learn/packages/clear/index.htmlng

7
of acceptable thermal environment for general comfort as when predicted mean vote falls in the
range between -0.5 and +0.5 and less than 10% of the population is dissatisfied. This is shown in
the table below.
PPD PMV Range
<10 –0.5 < PMV < +0.5
Table 1.2 . Acceptable Thermal Environment for General Comfort
5
ASHRAE addresses thermal comfort when the factors affecting it are in a steady state. The
majority of thermal comfort data available pertain to sedentary or near sedentary activity levels.
For conditioned spaces in which activity occurs ASHRAE suggests the following methods for
determining acceptable conditions in occupied spaces: the comfort zone is defined in terms of a
range of “operative temperature” that provides acceptable thermal conditions. Operative
temperature is defined as the average of the dry-bulb temperature and the mean radiant
temperature.
5

1.2.1 GRAPHIC COMFORT ZONE METHOD
This method provides a graph showing the acceptable range of operative temperature and
humidity for spaces where the occupants have activity levels that result in metabolic rates
between 1 and 1.3 met
9
and where clothing is worn that provides between 0.5 and 1 clo
10
of

9
met: is a “unit that describes the energy generated inside the body due to metabolic activity, defined as 18.4 Btu/h.ft2,
which is equal to energy produced per unit surface area of an average person(1.8m2) at rest.” (Definition by ASHRAE
55 Standard 2010). Metabolic rate below 1.3 corresponds to activities like sleeping, seated quite, writing, typing, etc.
Metabolic rate between 1.3 and 2 corresponds to activities such as walking about, cooking, house cleaning etc.
10
clo: is a “unit used that is used to express the thermal insulation provided by garments and clothing ensembles, where
1 clo = .88 ft2.h.°F/Btu.” (definition by ASHRAE 55 Standard 2010). A clothing insulation of 0.5 corresponds to
typical level of clothing worn when environment is warm and a clo value of 1 corresponds to typical clothing level in
cool environment. ASHRAE Standard 55- 2010 suggests methods to estimate clothing insulation level, based on tables
provided in the appendix.

8
thermal insulation and humidity ratio is at or below 0.012.5

The limits of thermal comfort zone
are defined by the range specified in Table 1.2.

Figure 1.3 The Graphic Comfort Zone Method: Acceptable range of operative temperature and humidity for
spaces where metabolic rate is 1.1 met and clothing insulation values are 0.5 and 1 clo
5

1.2.2 COMPUTER MODEL METHOD
This method uses computer software (like the ASHRAE thermal comfort tool) to predict PMV
and PPD that is used to define thermal comfort zone in terms of operative temperature and
humidity in Figure 1.3. This is applicable in spaces where occupants have activity levels that
result in average metabolic rates between 1.0 and 2.0 met and where clothing insulation value is

9
1.5 clo or below. This method was used to define the thermal comfort environment for humidity
ratio above 0.012 in Figure 1.3.
The PMV model is calculated using the air temperature and mean radiant temperature along with
the applicable metabolic rate, clothing insulation, air speed (below 0.2 m/s), and humidity. If this
value is within the recommended range, the conditions are within the comfort zone.5
1.2.3 GRAPHICAL ELEVATED AIR SPEED MODEL
This method allows effects of elevated air speed to be included in the Graphical comfort zone
model (Figure 1.3) to increase the maximum operative temperature for acceptability under certain
conditions. The effect of increased air speed is more prominent for elevated activity level and
lighter clothing. The amount by which temperature may be increased is shown in Figure 1.4. The
starting point of these curves is the upper airspeed limit of PMV defined comfort zone, 0.2
m/s(40fpm) as described by the Computer Model Method. It can be applied in circumstances
where activity level is above 1.3 met or clothing insulation is less than 0.5 clo (not more than 0.7
clo) and required air speed is up to 0.8m/s (160 ft/min).5

10

Figure 1.4 Air speed required to offset increased air and radiant temperature
5
1.2.4 SET METHOD
This model enables air velocity effects on thermal comfort to be related across a wide range of air
temperature, radiant temperature and humidity ratio. The comfort zone in Figure 1.3 can be
extended to air speeds above 0.8m/s with this method. The following figure uses SET to extend
comfort zone of Figure 1.3 across a range of air speeds for the example humidity ratio of 0.010. 5

The PMV model was first used to calculate the operative temperature range of +/- 0.5 PMV at
0.15 m/s (30 fpm) in order to define upper PMV boundary zone. The curving comfort envelope
boundaries above .15m/s (30 fpm) were then defined by constant SET. The SET lines indicate
temperature/airspeed combinations at which skin heat loss is same as the 30fpm PMV comfort
zone boundary. The calculated value of SET can be obtained using ASHRAE Thermal Comfort
Tool.5

11

Figure 1.5 Acceptable range of operative temperature and air speeds for the comfort zone shown in Figure 1.3,
at (example) humidity ratio 0.010.
5

1.2.5 ADAPTIVE MODEL OF THERMAL COMFORT
Another approach to the study of thermal comfort is the field study approach in which thermal
comfort is analyzed in normal surroundings; all subjective factors excluded in the climate
chamber experiment are included and studied in this method. Field studies in warm humid
climates have shown that the PMV model predicts a warmer thermal sensation than what people
actually feel.
11
The field study approach constitutes the basis for the Adaptive Model which states
that there is an optimal temperature for a given indoor environment depending on the outdoor air
temperature. It takes note of the fact that humans can adapt and tolerate different temperatures
during different times of the year. Field studies have proved that occupants’ thermal responses in
naturally conditioned spaces depend partly on outdoor climate and may vary from that of

11
P.Ole Fanger, Jorn Toftum: Thermal Comfort in the Future – Expectation and Excellence, International Conference
'Moving Thermal Standards to 21st Century.'

12
occupants in air-conditioned buildings. The optimal temperature for a given time is determined by
considering the mean outdoor temperatures of each month of the year.
1.2.5.1 OPTIONAL METHOD FOR DETERMINING ACCEPTABLE THERMAL
CONDITIONS IN NATURALLY CONDITIONED SPACES
ASHRAE suggests this method for spaces where thermal conditions of the spaces are primarily
regulated by occupants through opening and closing of windows and there is no air conditioning
system, where it is permissible to use mechanical ventilation with unconditioned air and the
occupants are engaged in near sedentary activity (1 to 1.3 met). The occupants are free to adapt
their clothing to indoor or outdoor thermal conditions. In such conditions the allowable operative
temperatures can be determined from the Figure 1.6. 5

Figure 1.6 Acceptable operative temperature ranges for naturally conditioned spaces.
5

13
Figure 6 is based on an adaptive model of thermal comfort that is derived from a global database
of 21,000 measurements, which includes 4 continents and a range of climate zones.12 The 90%
acceptability limits is to be considered when higher thermal comfort is desired and 80%
acceptable limit is suggested for typical application. This method can be used when the mean
monthly outdoor temperature is between 50°F and 92.3°F. No effects of humidity or air speeds
are included when this method is used.
6

The effect of humidity on human comfort is significant, as various researches and studies have
proved. Higher temperatures feel more comfortable if humidity is lower. In a hot humid climate
(like Kerala) a definition of comfort range would not be accurate if the effects of humidity were
not included in defining comfort range.

12
Richard De Dear: Adaptive Thermal Comfort- Past, Present and Future

14
CHAPTER 2: THE STUDY OF THERMAL COMFORT IN
INDIAN CONTEXT
The study of thermal comfort in the Indian context has only a short history. Various thermal
comfort investigations have been performed in the recent years to define the comfort zone for
Indian climate. Indian building codes specify comfort temperatures in the range of 73.4 °F to
78.8°F
13
. But the results of the recent thermal comfort study by Indraganti
13
in residential
apartments in Hyderabad, India shows the comfort temperature in the range of 78°F to 90.5 °F. A
detailed thermal comfort study is required to define the comfort temperatures for Indian climate.
Since the goal of the current study is to quantify passive cooling strategies for residential spaces
in warm tropical climates, it is important to identify a temperature band in which a majority of the
population would feel comfortable. The air temperatures data obtained from the original building
and the air temperature from the simulated model incorporating passive strategies will be
compared to this thermal comfort band, to see if they fall within that range or how close they get
to the comfort range of temperature.
Over the years, various methods have been developed to quantify thermal comfort and to define a
comfort band of temperature for various climates. Though the comfort temperature range defined
by many of these methods varies from one another, it is beyond the scope of this work to do a
field study to make a definitive determination. Instead, in this study, four standard methods are
used to define the comfort temperature range for the climate zone of Kerala. By comparing the
comfort band temperatures suggested by these 4 methods, it is hoped that it will be possible to

13
Madhavi Indraganti: Adaptive use of natural ventilation for thermal comfort in Indian apartments; Building and
Environment, Vol 45 (2010) Page:1490–1507

15
identify the range of temperatures in which a majority of the population would feel comfortable
within a narrow error bar.
As discussed in chapter one there are several factors that affect human thermal comfort and can
be indicators of comfort. The three main climatic factors that affect human comfort are air
temperature, humidity and air velocity. Several attempts have been made to combine all or some
of these variables into a single value (thermal index) which can be used to evaluate how people
feel. There are several thermal indexes available. The following section defines the range of
'comfort' temperature for Indian climatic conditions using three indexes.
2.1 PMV-PPD METHOD
ASHRAE suggests the Graphical Method (discussed in Section 1.2.1) to determine acceptable
thermal condition in conditioned occupied spaces. The metabolic rate was taken as 1.0 which is
the activity level for quietly seated and light activities, as per ASHRAE standards. The clothing
level ranged from 0.4 to 0.84 depending on the gender and age of the occupants. There is not
much change in the winter and summer clothing level, due to cultural preferences. The maximum
and minimum clothing level calculated by Indraganti
13
for her research was used to define winter
and summer clothing level. ASHRAE Standard 55 suggested method of calculating the clothing
level was used to check these values.
Using Climate Consultant the outdoor dry bulb temperatures from the weather data of
Thiruvananthapuram, Kerala was plotted on the psychometric chart. Using the thermal comfort
tool comfort zone temperature was defined for activity level of 1 and clothing levels of 0.4 and
0.84.

16

Figure 2.1 Psychometric chart showing the distribution of outdoor air temperature for Thiruvananthapuram,
Kerala.
Figure 2.1 shows the psychometric chart with dry-bulb temperature on the horizontal axis and
humidity ratio on the vertical axis. The comfort temperature range where 90 % of the population
would be satisfied with the thermal environment (i.e. -0.5 ≤ PMV ≤ 0.5 and PPD = 10%) is
marked by the violet lines. It was calculated using the ASHRAE Thermal Comfort Tool, a
computer program suggested by ASHRAE. If the temperature within the occupied room is the
same as outdoor temperature (this can often be achieved by having all the windows open),
according to PMV-PPD method the space would be comfortable for the temperature range within
the boundary marked by the violet lines. It can be seen from the distribution of the outdoor air
temperatures throughout the year that, for most of the year, the outdoor temperature is outside the
thermal comfort range.

17
2.2 ADAPTIVE MODEL FOR THERMAL COMFORT.
2.2.1 USING MEAN MONTHLY OUTDOOR TEMPERATURE.
As discussed in Section 1.2.5, this method helps define the range of comfortable indoor operative
range temperature in terms of mean monthly outdoor air temperature.

Figure 2.2 Acceptable Indoor Operative Temperature range for naturally conditioned spaces, computed using
the adaptive model of thermal comfort. The colored lines correspond to mean monthly temperature of each
month of the year for the city of Thiruvanathapuram, Kerala
Figure 2.2 shows the acceptable indoor operative temperature range for naturally conditioned
spaces, with mean monthly outdoor temperature on the horizontal axis and indoor operative
temperature on the vertical axis. The colored lines indicated the range of indoor comfort
temperature range for Thiruvananthapuram, Kerala, corresponding to the mean outdoor air
temperature of each month. It can be seen that temperature range is on the higher side for the
summer months of April and May as compared to slightly cooler months of October and
November.

18
2.2.2 USING WEIGHTED RUNNING WEEKLY MEAN TEMPERATURE.
There has been criticism for the choice of mean monthly outdoor air temperature as the external
atmospheric parameter, as an average temperature over a month may not be the most trustworthy
basis. R.J de Dear
14
in his paper has presented a new function called the weighted running weekly
mean temperature (T
mmot
), which provides the outdoor temperature integrated over a weekly
timescale. This paper suggests that “the ideal outdoor temperature function for adaptive comfort
is a weighted running mean spanning eight days, including the current day, but the current day is
excluded for practical purposes.” The function for weekly outdoor temperature adaptive function
is expressed as follows.
T
mmot
= 0.34 T
(day -1)
+ 0.23 T
(day -2)
+ 0.16 T
(day -3)
+ 0.11 T
(day -4)
+0.8 T
(day -5)
+ 0.05 T
(day -6)
+ 0.03 T
(day -7)

Figure 2.3 Adaptive Model of Thermal comfort based on weighted running weekly mean temperature
Figure 2.3 shows the acceptable indoor operative temperatures for the city of
Thiruvananthapuram, Kerala for the months of December and January, in terms of T
mmot..
Figure

14
R.J de Dear: Adaptive Thermal Comfort in Building Management and Performance; Healthy Buildings, June 2006

19
2.3

has number of the days of the year on the horizontal axis and the indoor temperature on the
vertical axis. The figure shows 80% and 90% acceptability limits corresponding to the Tmmot of
each day for the city of Thiruvanathapuram , Kerala for the year 1990. The red line indicates the
mean outdoor temperature.
2.3 TROPICAL SUMMER INDEX (TSI)
This is a thermal index developed by Central Building Research Institute (CBRI), Roorkee, India
and is generally accepted for Indian conditions. It is defined as the temperature of still air in
which an individual feels comfortable at 50% relative humidity. The TSI is calculated from the
following formula.
15

TSI = 0.308Tw + 0.74Tg – 2.06V
1/2
+ 0.845
where Tw = wet bulb temperature
Tg = globe temperature
V = air velocity
The globe temperature term takes into account the effect of air temperature and radiant heat. The
globe temperature is measured using a globe thermometer and measures the Mean Radiant
Temperature (which is defined as the area weighted mean temperature of all the objects
surrounding the body). The relative humidity as measured by the wet bulb temperature and the air
velocity are also considered in the equation.
The equation for TSI can be further simplified to an approximate equation for any combination of
environmental variables as
TSI = 1/3 Tw + 3/4 Tg – 2V
1/2

15
P.C.Vargese: Building Construction; PHI Learning Private Limited, October 2008.

20
It has been demonstrated from substitution and comparison that values of TSI calculated from the
exact and approximate equations almost agree for over the temperature range of 75º F to 104º F.
16

By analyzing the thermal sensation noted down by the subjects, it has been concluded that the
range of the value of TSI for 'comfortable' thermal sensation' is 77°F – 86 °F, with the optimum
value at 81.5°F.

16
M.R.Sharma , Sharafat Ali :Tropical Summer Index – A Study of Thermal Comfort in Indian Subjects ;Building and
Environment, 1986,Vol 21,No 1, pp 11-24.

21

Figure 2.4 The comfort zone marked in the psychometric chart with Tropical Summer index lines
16

2.4 COMPARISON OF THE FOUR METHODS OF DEFINING
THERMAL COMFORT
As discussed earlier, each of the above discussed method has it's advantages and disadvantages.
It can be summarized as seen in Table 1.1.

22
ADVANTAGES DISADVANTAGES
GRAPHIC ZONE METHOD
Takes into account all factors affecting thermal
comfort .
All thermal comfort factors closely monitored.
Applicable to conditioned spaces only.
Gives very narrow band of comfort zone
temperature.
ADAPTIVE MODEL OF THERMAL COMFORT (Using mean monthly temperature)
Applicable to naturally ventilated spaces.
Gives wider range of comfort temperatures.
Applicable to conditioned spaces.
Uses mean monthly temperature ie it
doesn't take daily variation in
temperature in calculating comfort zones.
ADAPTIVE MODEL OF THERMAL COMFORT (Using weighted running weekly
mean temperature)
Applicable to naturally ventilated spaces.
Gives a better range of comfort temperatures as
it is correlated with outside temperature of
preceding days.
Deals purely with temperature (which is
just one thermal comfort factor).
TROPICAL SUMMER INDEX
Defined for Indian climate conditions.
Thermal comfort temperature zones in the
higher range.
Takes into account temperature, relative
humidity and wind speed.
The initial study used very limited
subject sample and limited to one climate
zone if India.
Table 2.1 Comparison of advantages and disadvantages of various models of thermal comfort.
For the purpose of this study, thermal comfort is defined by all four methods and if the
temperature falls within any of these range, the thermal environment accepted as comfortable.

23
2.5 PSYCHOMETRIC CHART SHOWING THERMAL COMFORT
ZONES DEFINES BY FOUR METHODS FOR
THIRUVANANTHAPURAM, KERALA.
Figure 2.5 shows the Psychometric chart with temperature on the Horizontal axis and Specific
Humidity on the vertical axis. The red dots indicate the outdoor dry bulb temperature for the city
of Thiruvanathapuram from 1
st
January, 2011 to 31
st
May, 2011.The various colored lines
indicate the range of thermal comfort as defined by the four methods discussed above for the
climate of Thiruvanathapuram.

Figure 2.5 Psychometric chart showing the thermal comfort ranges defined by the four methods for the city of
Thiruvananthapuram
The adaptive method with running weekly mean temperature has the widest comfort temperature
range whereas the TSI has higher temperatures falling within the comfort zone. The PPV/PMD
method has the narrowest range of comfort temperature.

24
CHAPTER 3: RESEARCH METHODOLOGY
3.1 INTRODUCTION
In the current context of increasing global temperatures, more and more research is being focused
on reducing carbon footprints, especially in developing countries like India. India had a strong
cultural tradition of building passively conditioned buildings and the potential of the design
solutions that the indigenous architecture offers has to be explored considering that recent studies
have shown that traditional passive design solutions were successful in maintaining a comfortable
dwelling space that worked in harmony with the environment
3.1.1 PAST STUDY DONE ON THE TOPIC
A study by Srivastav et al.
17
In a paper published in the International Journal of Low Carbon
Technologies in 2009, discusses the question of using traditional passive architecture strategies to
reduce the carbon emission in modern dwellings. They identify that, due to the changes in social
scenario caused by changed lifestyles, the requirements of the people, the materials available etc.
the ideal solution is not to revert to traditional strategies blindly but reinterpret them in a way that
suits the current set up.
The paper
17
also identifies the passive strategies that were incorporated in the traditional houses
in the hot humid climate of Kerala, as discussed below.
Most of the traditional dwellings are isolated dwellings surrounded by heavy vegetation (which
ensured external shading and protection from sun) and feature large open interior spaces. The
typical house features a central courtyard around which all the rooms are arranged. Deep

17
S. Srivastav, P.J.Jones: Use of traditional passive strategies to reduce the energy use and carbon emissions in modern
dwellings; International Journal of Low Carbon Technologies, Vol 4(2009) Page: 141 - 149

25
verandas run around the perimeter of the house as well as around the central courtyard. The Large
overhanging eaves of the external verandas protect the walls not only from the torrential rains, but
also from harsh sun. The house is oriented north – south to mitigate high heat gains from harsh
evening sun (when it is low). Built according to the Principles of traditional architecture called
Vaastusaasthra, it is a quadrangular building with four blocks arranged around the courtyard. The
house has a double roof system with a high pitched tiled roof (to drain the rainwater) on top and a
flat wooden ceiling for the rooms below it. This creates a large insulated air space in between that
is ventilated through gabled windows. (A study by N. Areemit et al
18
reports results for a passive
room-dehumidifying system using an existing attic space as a chamber, wood as a desiccant
material, and optimized ventilation and solar energy for system operation.) The courtyard is
usually sunk so that cooler air settles down (In the study by A.S Dili et all4, a temperature
difference of 10 ºF was noted between the floor level height and the bottom of courtyard). The
external verandas may be open or enclosed by wooden panels with slits to facilitate ventilation.
The general materials used for construction are wood (walls and ceilings) and laterite blocks
(external walls). The fenestration ratio is high and there are more openings in the direction of the
prevailing wind. Even though the depth of the building is large the courtyard offsets the
detrimental effect by providing cross ventilation.

18
N. Areemit, Y. Sakamoto: Numerical and experimental analysis of a passive room-dehumidifying system using the
sorption property of a wooden attic space; International Journal of Low Carbon Technologies, Vol 4( 2009) Page:141 -
149

26

Figure 3.1 An example of traditional residence in Kerala
19

Figure 3.2 A Typical Traditional residence in Kerala
4

19
Picture Courtesy: http://southindiatravelideas.blogspot.com

27
The modern dwellings in Kerala cannot be reduced to a single prototype as they vary widely in
layout, which is less a reflection of climatic sensitivity than of the owner’s income level. The
dwellings of middle income and low income groups (which constitute the majority of the
population), can be roughly considered as compact boxes which rarely have courtyards or proper
cross ventilation, due to lack of space. The building materials mostly used in construction are to
bricks or hollow concrete blocks (for walls), concrete (floors and slabs), clay tiles, and large
glazed windows with small overhangs above. Long verandas are absent from most of the designs
mainly due to lack of space.

Figure 3.3 Examples of modern residences in Kerala
20

The study by Srivastav et al.
17
have used environment performance modeling for an analysis of
thermal comfort in traditional and modern dwellings. A prototype of traditional building and that
of a modern building were modeled and comparative thermal performance assessed using HTB2
(Heat Transfer in Buildings Version2).

20
Picture Courtesy: http://homegallerydesign.com; http://kochi.olx.in

28
Though these authors concluded from their analysis that traditional passive strategies can be
effectively adapted to modern buildings to benefit comfort conditions in the houses, they
suggested further studies are required to find the combination of strategies that would fit into the
modern design scenario.
This thesis is based on this study and tries to take it to the next level which would define the
passive strategies (including the traditional strategies) that are most applicable to the current day
situation in the design and construction field in Kerala.
The modern building chosen for the current study is an existing building. Temperature and
relative humidity data has been collected from the residence using data loggers for a specific
period of time and this data is to be used to refine and validate the simulated model against the
original house. This refinement would improve the accuracy of the results.
3.2 STEPS INVOLVED IN METHODOLOGY

The following paragraphs discuss the various steps involved in the research methodology.
3.2.1 LIST OF PASSIVE COOLING STRATEGIES APPLICABLE TO
CLIMATE OF KERALA
Kerala's climate lacks seasonal variation in temperature and has high temperature and humidity
almost all the year round. There is very little diurnal variation in temperature (in the range of 5 –
7 degree Celsius) and the diurnal average is maintained over a period of several days. The
seasonal pattern is mainly characterized by period of high rainfall, in the months of July through
August and October through November. The night time temperatures are also high with high
relative humidity. The climate is outside comfort conditions for a large portion of the year.

29

Figure 3.4 3D Chart of monthly Dry Bulb Temperature distribution of Trivandrum.
21

Figure 3.5 Monthly Dry Bulb Temperature distribution of Trivandrum. The grey band shows the comfort range
of temperature according to ASHREA PMV-PPD Method.
21

21
Picture Courtesy: Climate Consultant 5

30
Givoni in his book 'Passive and Low Energy Cooling of Buildings' establishes a difference
between bioclimatic architecture and passive cooling systems. “Bioclimatic architecture involves
architectural design strategies and choice of materials aiming at providing comfort, while
minimizing the demand for energy used to cool a building. But passive cooling systems are
capable of transferring heat from a building to various natural heat sinks. Appropriate
architectural bioclimatic design is considered a precondition for application of passive cooling
systems and the two approaches supplement and reinforce one another.”
22

Several bioclimatic architectural strategies which would be effective in improving the thermal
comfort of naturally ventilated buildings in various climate zones have been suggested by many
studies. The strategies which can be considered applicable to an average residential building in
Kerala climate can be shortlisted as follows.
 ORIENTATION: The building should be oriented so as to minimize the solar heat gains
from the East and West directions. The optimum orientation would further ensure that the
building is aligned in such a way as to take advantage of prevailing wind direction.
 PLAN AND LAYOUT: Thin open plans that enhance the wind flow through the
openings and create better opportunities for cross ventilation would be ideal for this
climate. Including an internal courtyard could have a similar effect. The internal layout of
the spaces should be arranged to facilitate maximum wind movement. The ideal layout
would ensure that wind path would not bring in the internal heat gains from the areas
such as kitchens (where more heat is generated through cooking).

22
Givoni, Baruch, Passive and Low Energy Cooling of Buildings:.John Wiley & Sons Inc; 1994

31
 ROOF: At latitude of 8°N, the sun is at a high altitude almost all the year round in Kerala
and hence it is reasonable to assume that strongest thermal impacts occur here in a
building of the scale of an average residence. A double roof with a ventilated space
between the roof and the ceiling below would be a preferable option in this type of
climate. The roof (preferably insulated) would act as the sun protection and the air gap
would act as the insulation. Using a light colored roof would reflect solar radiation and
help reduce the solar heat gain 
 WALLS: Changing the wall to low thermal mass material as opposed to high thermal
mass materials that are being currently used might help, by reducing the thermal lag, for
rapidly cooling at night. 
 OPENINGS AND WINDOWS: Maximizing fenestration ratio would be ideal for
ventilation in this climate. Cross ventilation should be ensured in each room. Smaller
openings in the windward side and larger openings towards the leeward side would
increase the air speed inside the space. Ventilating openings near the ceiling levels would
help provide an exhaust path for hot air. Quantifying the area of window openings and
ventilating openings needed for proper air movement needs to be explored. 
 VENTILATION: In order to facilitate ventilation as the primary cooling strategy the
building design should try to achieve high air speed and fast cooling of interiors. The
windows should direct the flow towards the areas and vertical zones which are inhabited. 
 SHADING: Due to the high ratio of openings on walls, it is important that windows and
openings are properly shaded so as to reduce direct internal solar heat gain. With high

32
altitude sun almost all the year round horizontal shading would mitigate the solar heat
gain through openings. 
 VERANDAHS: Traditional architecture used verandas to protect the walls from direct
solar radiation and act as transitional spaces to the inner living areas. The amount of heat
gain mitigated by providing verandas needs to be quantified. 
 COURTYARDS: Sunken courtyards were a major architectural feature of traditional
buildings that helped maintain the comfortable air temperature, air flow pattern and
improved ventilation by creating a stack effect. Though lack of availability of space may
hinder the implementation of courtyards in most residences, a small scale version can be
implemented by simulation to check the effects. 
 MATERIALS USED FOR CONSTRUCTION: The majority of the materials that are
currently used for construction are high thermal mass materials. Bricks/Concrete block
walls, reinforced concrete slab ceilings and floors and concrete block paving around the
residence are some examples of high thermal mass materials most commonly used in
construction. Also very little or no insulation is used in building envelope or roof design.
The effect of using low thermal mass materials or new methods of construction (for
example: hollow brick wall construction with air as insulator or light insulated panel
construction) should be evaluated to understand the percentage improvement this
provides. 
 HIGH PERFORMANCE GLAZING: Unlike traditional residences, the windows of
modern residences have a high percentage of glazing which would contribute towards the
heat gain in the space. By reducing the area of glazing and by using glass that reduces the

33
heat gain into the rooms, the thermal gain into the residential paces might be improved.
The percentage improvement that the change in glazing creates on a residential building
can be identified more accurately by simulating it in the model.
 VEGETATION: The presence of trees and other vegetation around the house have been
seen to bring down the indoor temperature in many cases. The effect of trees and other
vegetation in thermal comfort of the enclosed space would also be significant study.
The combination of different building strategies would be studied to quantify the effects of these
in various contexts.
3.2.2 BUILDING SELECTION AND DATA COLLECTION
An existing modern residential building in Kerala, with natural ventilation augmented by
mechanical air conditioning, is chosen as the case study which can be modeled in energy
calculation software and to which these strategies can be applied and the effects studied. The
temperature and relative humidity data inside the residence were measured for a specific period of
time using HOBO Data loggers. By analyzing this data the current thermal conditions inside the
residence can be assessed. This would serve to provide an example of how the current design
practices and materials perform in terms of thermal comfort.
3.2.3 SELECTING SOFTWARE TO SIMULATE THE BUILDING AND
CALIBRATING SIMULATED BUILDING
Among the many energy calculation software available, Design Builder version 3.5 Beta were
chosen to simulate the building, since the software is capable of simulating naturally ventilated
spaces. The initial step after the building is modeled in the software would be to validate it
against the original building by checking if it gives the same interior air temperatures as those

34
collected from the original building. If not, the model properties should be checked and revised
until the model is calibrated.
3.2.4 APPLYING STRATEGIES TO IMPROVE THERMAL COMFORT
Once the model gives results close enough to those of the original building, the listed strategies
for thermal comfort can be modeled and analyzed. The resultant comfort conditions would be
compared against the accepted thermal comfort criteria for Indian climate (as discussed in section
2.5). The effects of various strategies and their combination would be quantified and classified
according to the effectiveness.
Results would also be analyzed to see which combinations of the strategies would provide the
best thermal comfort options for the case study residence. This could help providing a set of
suggestions which would be applicable for this residence.
3.2.5 MAKE A LIST OF SUGGESTIONS FOR IMPROVED THERMAL
COMFORT IN THIS CLIMATE
These exercises would hopefully help to figure out which strategies would work more effective
than others in the climate of Kerala. These findings can be generalized to propose a list of
suggestions of passive cooling strategies for improved thermal comfort in small independent
residence in India.

35
CHAPTER 4: CASE STUDY BUILDING
In this chapter, we will discuss the building selected for monitoring and the analysis of the data
collected. The initial step was to decide on a building that was representative of a modern-day
building in Thiruvananthapuram, Kerala. The temperature and humidity of the building was
monitored using HOBO data loggers. HOBO data loggers, manufactured by the Onset Computer
Corporation are electronic instruments that records temperature and humidity at fixed intervals
over a period of time
23
. The data was to be collected the year round and analyzed for
representative months for summer and winter. By analyzing this data, conclusions can be made
about the current thermal environment inside the building.
4.1 DESCRIPTION OF SELECTED BUILDING
The building selected for case study is a 4500 sq ft building (see Figure 4.1), set on a 0.3 acre
plot, located in the city of Thiruvanathapuram, Kerala and built in the year 2007. The building
has a square-shaped foot print and front elevation faces the East-South-East (ESE) direction. The
site has a river front on the South side and has buildings on the North-West and North-East sides.
The building has deep foundation (cast in situ pile foundation) and the super structure is a framed
structure with reinforced columns and beams with non-load-bearing walls. The building has 8”
hollow concrete block masonry construction for walls, which are uninsulated. The walls have
cement-sand plastering on the outside and inside. The outside wall finish is rough with pale
yellow paint. The floor slab and roof slab are reinforced concrete slabs which are also
uninsulated. The inside of the roof slab (the ceiling) is white washed .The floor-to-floor height is
10 feet including the slab thickness of 1/3 feet. Above the flat concrete roof slab is a painted

23
http://www.onsetcomp.com/

36
Stainless Steel truss structure supporting terracotta (Clay based glazed/unglazed ceramic) tiles.
This structure partially covers the flat concrete roof. The windows and doors are painted hard
wood frames with single glazing. The first floor, with a total area of 2600 sq ft, has the drawing
room, the family living, two bedrooms and kitchen. The drawing room and family living space
(which are the common spaces) have a long veranda that partially encircles them. All the rooms
are well ventilated with large windows with placement that allows for cross ventilation. The
common spaces (drawing/dining/family living) have sliding doors that open up to provide
maximum air flow in those spaces. The staircase well has an opening in the ceiling, which
enhances the day lighting of the space below as well as improving the ventilation. The second
floor has 1600 sq ft area and includes an upper living space, a family room, two bedrooms and a
balcony facing the river.
The positioning of the rooms is such that all the bedrooms and common areas face the river. All
the spaces are naturally ventilated most of time, ie. The space is naturally ventilated during day
time all the year round and in the night time in the winter months. Air conditioners are used only
in bedrooms and only during the hot months of the year. The occupants of the house include 4
adults and 2 children. Although the windows in most of the rooms have been designed to enhance
cross ventilation, all the windows are not opened every day. Basic details about the occupants’
usage of various building elements have been noted; however, detailed study of the behavioral
pattern of the occupants has not been included in this study and is beyond the limit of the scope of
this project.

37

Figure 4.1: Plan of first floor (left) and second floor (right) of the modern residential building in
Thiruvananthapuram that was chosen for the case study.

38

Figure 4.2: Sectional detail of the residential building.

Figure 4.3 View from East (left) and South (right) of the building chosen for case study.

39

Figure 4.4 View of first floor living space and veranda Positions of the HOBOs are marked with red circles in all
pictures.

Figure 4.5: View of the second floor bedroom and position of the HOBO on the first floor staircase landing.
4.2 HOBO DATA LOGGERS
As discussed in Section 1.2, temperature and humidity are two of the most important factors
affecting thermal comfort. For this study, temperature and humidity of the selected residence
were monitored for a period of one year (2011).

40
HOBO data loggers manufactured by Onset Computer Corporation were employed for the study.
The HOBO U10-003 series, which supports Dry Bulb Temperature and Relative Humidity
measurements, was the choice for internal temperature and humidity monitoring.
The HOBO U10-003 loggers are two-channel Temperature/Relative Humidity Data Loggers with
10-bit resolution and capacity for 52,000 measurements. They have temperature range of -4
degrees Fahrenheit to 158 degrees Fahrenheit with an accuracy of +/-0.97-degrees Fahrenheit and
a relative humidity range of 25% to 95% with an accuracy of +/-3.5-percentage points.
24
A total
number of ten HOBOs was deployed for collecting data from the residence.
4.2.1 RELATIVE CALIBRATION OF HOBOS
The initial step before deploying the data loggers for monitoring temperature and humidity was to
calibrate all the HOBOs in order to understand the relative calibration error. Understanding the
relative error between various HOBOs is important, as those errors should be corrected for /
accounted for in the data analysis. For the comparative calibration of HOBOs the following
process was used: ten HOBOs were put together in a paper bag and kept together in varying
temperatures for a total period of 41 hrs. For the first 17 hrs the HOBOs were kept at room
temperature, the next 11 hrs inside a refrigerator and the next 13 hrs back at room temperature.
The data logging interval was set to 15min and temperature and relative humidity were both
measured. This data was used to understand the relative calibration accuracies.

24
http://www.onsetcomp.com/products/data-loggers/U10-data-loggers

41

Figure 4.6: Temperature and humidity measurements from ten HOBOs during the calibration period, plotted
against time. All the HOBOs are placed in the same environment, and the difference in measured data between
different HOBOs is an indication of the relative calibration error.

Figure 4.7: Temperature and humidity measurements averaged over all the HOBOs at any point in time, and
maximum/minimum deviation from the mean

42
Figure 4.6 shows the temperature and humidity measurements of the HOBOs during the
calibration period plotted against time. The following conclusions can be made from these plots:
It can be seen that temperature measurements from the ten HOBOs agree with each other within a
relative error of 3 ºF. The largest disagreement between different HOBOs occurs when the
temperature is above 80ºF or below 65ºF. The agreement is much better when temperature falls in
the range 70–80ºF.
Figure 4.7 shows the measurements averaged over all the HOBOs at any point in time, along with
the maximum/minimum deviation from the mean. If xi (t) denotes the temperature (or humidity)
measured by HOBO i (where i = 1 to 10) at time t, the mean x(t) and the error bar ε(t) can be
defined as

ε(t) = max
i
x
i
(t) − min
i
x
i
(t),
where N is the number of HOBOs (N = 10), while max
i
x
i
(t) and min
i
x
i
(t) denote the maximum
and minimum measurement among the ten HOBOs at time t. In simple terms, the error bar
represents the value of the maximum error measured by a particular Hobo at any point of time.
The equation above helps calculate the error bar for each Hobo at any time. Knowing the value of
error bar, we get an idea about the range in which the Hobo data is accurate.
As noted earlier, the error bars on temperature are the largest (4ºF), when the temperature is
above 80 deg F or below 65 ºF, whereas the error bars are the smallest (0.6ºF) when temperature

43
falls between 70 – 80ºF. In contrast, the error bars on humidity measurements stay more or less
fixed (7%). Nevertheless, we see that in the high humidity range (75 – 85%), the error bars are
reduced to 1%.
Figure 4.8: The relative error Δx
i
(t) of each HOBO from the mean value recorded by all HOBOs at any point in
time. The horizontal axis shows the time in days.
Figure 4.8 shows the relative error Δx
i
(t) of each measurement (temperature or humidity) from
the mean measurements, defined as
Δx
i
(t)= x
i
(t) – x(t).
By comparing Figure 4.8 with Figure 4.6, it can be seen that HOBO 5 shows a maximum
negative temperature bias from the mean value, when the temperature is above 85 ºF (i.e., it
recorded a lower temperature as compared to other HOBOs). On the other hand, when the
temperature was below 65 ºF, it shows a positive bias (i.e., it recorded higher temperature than

44
other HOBOs). The deviation from the mean value of temperature measurements is minimal
when the temperature is between 70ºF and 80ºF. In humidity measurements HOBO 6 has a
positive deviation as compared to other HOBOs (i.e., it recorded a higher humidity than all other
HOBOs throughout the recording period). The relative error in temperature measurements
between most of the HOBOs in room-temperature conditions is less than ±1ºF and the relative
error in humidity measurement is ±2% . Since this is well within the accuracies that we are
concerned about in this thesis, we will ignore these errors. It should be kept in mind in the future
discussions that uncertainties of ±1ºF in temperature and ±2% in relative humidity are present in
our data.
4.2.2 POSITIONING OF HOBOS
Physical Data collection is an important aspect in any study connected with thermal comfort
(TC). Three broad classes of TC field investigation can be identified based on the standard of
instrumentation and procedures used for indoors climatic measurements.
25
The description of the
field study classification is shown in Table 4.1.
In this case, the data collected for temperature and humidity would fall under Class III category.
The data was collected at a single height above the floor for the all of the spaces measured. Ten
HOBOs were deployed in collecting data from all rooms.

25
Brager, G.S., and De Dear, R.J. (1998), Thermal Adaptation in the Built Environment: A Literature Review: Energy
and Buildings, 27, 83-96
.

45
CLASS MEASUREMENT PROTOCOL RESULT DESCRIPTION
I All sensors and procedures are in 100%
compliance with ASHRAE Standard 55.
Simultaneous three heights of
measurement (0.1, 0.6 and 1.2 m).

Effects of non-uniformity (asymmetrical)
in the environment.
Suitable for detail examination of
localized discomfort in air-conditioned
spaces
II Measure all physical environmental
variables (air temperature., mean radiant
temperature, wind speed, relative
humidity, clothing, and metabolic rate)
At one height of measurement (0.8-1 m)

Allows an assessment of the impacts of
behavioral adjustment and control on
subjective responses.
Data can be converted to different thermal
indices such as operative temp, ET, and
SET.
III Simple measurement of indoor
temperature and possibly humidity.
At one height above the floor

Data does not necessarily allow
explanatory analysis.
Widest range of published data. The
majority of field studies used in the
derivation of early adaptive models.
Table 4.1 Classes of data for thermal comfort field research
26

Since the residence was of 4500 sq ft, it was planned that temperature and humidity of selected
areas would be monitored. The HOBOs recorded the values in all bedrooms (Bedroom 2 and
Bedroom 3 (first floor) and Bedroom 1 (second floor), living space and staircase well. It was
ensured that when the HOBOs were placed on surfaces such as wall and roof, they were separated
from the surface with a thermal break, so that the HOBOs measured the air temperature and not
the surface temperature. (However it was later discovered that, for the case of the HOBO placed
on the ceiling of bedroom 1, heat from the surface appears to have been conducted through the
thermal break and the HOBO appears to have measured the surface temperature of the ceiling
slab).

26
Henry Feriadi: Thermal comfort for naturally ventilated residential buildings in the tropical climate; Master Thesis,
National University of Singapore, 2003.

46
HOBOs 1,2,3,4 were placed in Bedroom 1 in the second floor (SF). This was one of the places
which the owners complained of heat. Air conditioning was switched on during certain recording
periods, and ceiling fan was used whenever the room was occupied. (Data from the times when
the air conditioning was switched on were discarded in the analysis). HOBO 1 was placed on the
outside wall at a height of 9 feet from finished floor level (FFL) and HOBO 2 on the inside of the
same wall at a height of 7 feet from FFL. This was done in order to understand the heat transfer to
the room through the wall. HOBO 3 was placed on the bedside table (to measure the air
temperature at the bed height) and HOBO 4 was placed on the ceiling (to get an idea of heat
transfer through the ceiling). After January 2011, HOBO 1 was placed on the balcony wall of the
first floor due to the physical difficulty in keeping it in the original position.
HOBO 5 was placed in the veranda that went around the family living in the ground floor. HOBO
7 monitored the family living space. HOBO 6 was placed in the Bedroom1 on the first floor (FF)
and HOBO 10 in Bedroom 2 (also on FF). HOBO 8 (at about 8.5 feet from finished floor level on
the first floor and HOBO 9 (at about 8.5 feet from the FFL of the first floor) were placed on the
staircase wall.
From here onwards, we will refer to HOBOi as Hi (H1, H2, H3 etc.).

47

Figure 4.9: This figure indicates the position of HOBOs in the Ground Floor and First Floor.
HOBO 1 On the outside wall of second floor. Measures outdoor air temperature. At a height
of 8’ from finished floor level (FFL).
HOBO 2 On the inside wall of Bedroom3 on Second Floor. Measures the air temperature of
the room. Placed at a height of 7’ from FFL.
HOBO 3 Near the bedside table in Bedroom3 on Second Floor. Measures the air temperature
of the room. Placed at a height of 2’ from FFL.
HOBO 4 On the ceiling in Bedroom3 on Second Floor. Measures the air temperature of the
room. Placed at a height of 9’ from FFL.
HOBO 5 On the verandah on the First Floor. Measures the outdoor air temperature. Placed at
a height of 7’ from FFL.
HOBO 6 Near the bedside table in Bedroom1 on First Floor. Measures the air temperature of
the room. Placed at a height of 2’ from FFL.
HOBO 7 On the furniture in Living Room on First Floor. Measures the air temperature of the
room. Placed at a height of 2’ from FFL.
HOBO 8 On the Staircase landing wall on First Floor. Measures the air temperature of the
room. Placed at a height of 2’ from FFL.
HOBO 9 On the Staircase landing wall on First Floor. Measures the air temperature of the
room. Placed at a height of 8’ from FFL.
HOBO 10 Near the bedside table in Bedroom2 on Second Floor. Measures the air temperature
of the room. Placed at a height of 2’ from FFL.
Table 4.2 Position of HOBOs on site

48

4.3 ANALYSIS OF COLLECTED DATA
Data has been collected for a period of 12 months, starting from 25th December, 2010 to 21st
December 2011. The diurnal variation of the temperature follows a similar pattern during winter
months, which is different during summer and monsoon months. The data from a typical winter
month (January), a typical summer month (April) and a typical monsoon month (June), has been
analyzed. These three months are representative of the temperature and humidity variation for the
major seasons in Thiruvananthapuram, Kerala. The temperature and humidity condition for all
other months fall close to or within the range specified by these selected months. In order to avoid
transient fluctuations in the data due to everyday weather variations, the hourly temperature
measured by each HOBO averaged over a month was taken as the hourly temperature of the
corresponding measurement point in the building.
4.3.1 ANALYSIS OF OUTDOOR DATA
This section covers the analysis of outdoor weather data measured at site. (The discussion of
various outdoor data weather files used in this project are discussed in detail in chapter 6 and
chapter7. This section purely deals with the analysis of weather data measured at site using
Hobo1). Outdoor temperature and humidity were collected using Hobo 5 for a period of one year.
The data for the three chosen months was analyzed. Figure 4.10 shows the mean hourly outdoor
temperature and relative humidity for the typical winter (January), summer (April) and monsoon
(June) months. The temperatures are higher in the month of April, which is the "summer" month
in Kerala. The winter month of January has the lowest temperature measured at 78˚F. The peak
temperature of April and January are close, but the lowest temperature for January is much lower
than April. The temperatures peaks in April and the highest mean temperature reaches almost

49
90˚F. The relative humidity is generally high (always > 73%) in the month of June. Though
temperatures are lower in June, the presence of relatively high humidity might push the thermal
comfort indexes out of the comfort zone. The diurnal variation in temperature is the highest in
January (in the range of 10˚F), whereas it is the lowest in June (in the range of 4˚F). In the month
of April, the diurnal variation is in the range of 5.5˚F.

Figure 4.10: Temperature (left) and relativity humidity (right) measured by HOBO 1 in January, April and
June plotted against the hour of the day. The thick dots correspond to hourly temperature (or humidity)
averaged over 30 days, while the error bars show the minimum and maximum measured hourly temperature (or
humidity) over the measurement period. Hourly averaged temperature was used in order to avoid transient
fluctuations in the data due to everyday weather variations. Each data point correspond to the hourly
temperature measured by the HOBO averaged over a month. For eg. The red dot corresponding to 5 am in
temperature - time graph, represents the average of temperature measured at 5 am for the whole month.

50
4.3.2 ANAYSIS OF THE INDOOR DATA

Figure 4.11: Mean hourly temperature (left) and relative humidity (right) from January recorded by HOBOs
placed at four different living spaces in the building (see legend). The thick dots represent the mean values and
the error bars correspond to the maximum and minimum values of temperature/humidity measured over the
measurement period. Mean Temperature was used for the same reason as explained in Figure 4.10
Figure 4.11 shows the mean hourly temperature and relative humidity from January recorded at
four different living spaces in the house – bedroom1 (second floor), bedroom 2 (first floor),
bedroom 3 (first floor) and the living room (first floor). It can be seen that bedroom 1 (H3)
consistently reports the highest temperature. This is consistent with the occupants’ observation
that bedroom1 on the second floor is the space that is the most uncomfortable during hot periods.
Thus, this room is chosen for detailed investigation. Since the temperatures measured in the other
rooms are below that measured in this room, it could be safely assumed that if the temperature in
this space could be brought closer to comfort zone temperatures (comfort zone temperatures as
defined Section 2.7 and Figure 2.5), the average temperature of the whole residence would come
closer to comfort zone temperatures.

51
4.3.2.1 THERMAL CONDITIONS IN THE SELECTED ROOM

Figure 4.12: Mean hourly temperature (left) and relative humidity (right) from January recorded by HOBOs
placed inside/outside bedroom1 (see legend). The thick dots represent the mean values and the error bars
correspond to the maximum and minimum values of temperature/humidity measured over the measurement
period.
Now, we proceed to discuss in detail the thermal environment in bedroom 1 (second floor) for
the winter month of January, the month of April and the monsoon month of June. The
conclusions from the analysis in this section are based on preliminary study of the data. By
understanding the heat exchange in the room in more detail by using computer simulations, more
solid conclusions are made.
Figure 4.12 shows the mean hourly temperature and humidity for January 2011. The Hobo H1
measures the outside temperature and the diurnal temperature variation is in the range of 10°F .
All the inside temperatures (H2, H3 and H4) peaked approximately 2.5 hours after the outside
temperatures peaked. This suggests a thermal lag, which could be due to high thermal-mass wall
and roof. H4 has the highest range of temperature variations. This is attributed to the fact that the
location of H4 is on the ceiling and the HOBO measured mainly the surface temperature of the
concrete roof rather than the air temperature. The inside temperatures (H2 and H3) did not

52
fluctuate as much as the outside temperatures and varied within a range of 6°F throughout the
day. This shows that the temperature variation inside the room is much small as compared to the
outside temperature. This suggests that the thermal mass in the building envelope is playing a role
in the heat transfer in the space. In chapter 6 and Chapter 5, the heat transfer process and the
effect of thermal mass of walls and roofs is discussed in detail.
A detailed interpretation for the temperature measured by HOBO 4 is given in Chapter 5.
Comparing the other two HOBOs, it can be seen that H2 (at a height of 2.5 feet) and H3 (at a
height of 7 feet) closely follow each other. This suggests that the temperature stratification (which
occurs when hot air becomes lighter and rises up where as cooler air sinks to the bottom of the
room, creating a stratification of temperature along the height of room) in the living space (from
Finished Floor Level to 7.5 feet height) is not important. There is a large variation in the outdoor
air temperature throughout the day. In January, it peaks around 2 pm and hits the lowest value at
7 am. The room temperature peaks almost two hours after the outdoor temperature peak, and the
temperature inside does not vary as much as outdoor temperature. This could be due to many
reasons. Though the space is naturally ventilated, the windows are closed at night and only few of
the windows are opened during day time. The number of air changes per hour (ACH) inside the
room might become lower at night due to the closed windows. During daytime also the room
might experience a lower ACH value if only the window on one side is open. Also the concrete
block walls act as a high thermal mass material that absorbs the heat from the room during day
time and releases it back during night. The roof, which is a concrete slab, also acts as thermal
mass. A more detailed look at which of these account more for thermal lag inside the space is
discussed in the following chapter. Relative humidity falls in the range of 70% to 80% throughout
the winter month.

53

Figure 4.13: Mean hourly temperature (left) and relative humidity (right) from April recorded by HOBOs
placed inside/outside bedroom1 (see legend). The thick dots represent the mean values and the error bars
correspond to the maximum and minimum values of temperature/humidity measured over the measurement
period.
Figure 4.13 shows the temperature and humidity plot for the room in April. It can be seen that the
temperature variation inside the space is different than that in the plot for January (Figure 26).The
diurnal variation of both outdoor and indoor temperature is less than that for January. The average
hourly temperature of the room rarely falls below the outdoor air temperature in April. The wind
velocity is slightly higher in the summer months and the ACH value could go up due to increased
airspeed during day time when the windows are open. Comparing the solar radiation on the roof
in summer and winter, it is seen that the values are almost double in April. These could be two
major reasons of increased room air temperature in April than in January. Relative humidity is
slightly lower than the winter month (65% to 75%).

54

Figure 4.14: The temperature data measured by for H1, H2,H3 and H4 over a month in June plotted against
the hour of the day. The thick dots correspond to hourly temperatures averaged over 30 days, while the error
bars show the minimum and maximum measured temperature over the 30 days.
From Figure 4.14, which shows the plot of hourly averaged temperature and humidity against the
hours of the day in June, it can be seen that the diurnal variation range of hourly temperatures is
very small (84°F to 86°F); the diurnal variation of room temperature is also within the range of
2°F difference ( 84.5 to 87°F). The relative humidity is higher than April. This month has high
temperatures and high Relative Humidity. The room temperature follows the same pattern as in
April. The highest wind speed occurs in June. During the day time when the windows are open,
the increased wind speed may increase the ACH, there by bringing the outdoor and indoor air
temperature closer. The temperature gain from the roof and the walls peaks in the late afternoons,
which could result in the higher room temperature during late afternoon. Further investigations
carried out using energy simulation software are discussed in Chapter 6 which explains the
various heat transfer process in the room.

55
4.3.2.2 THERMAL COMFORT IN THE SELECTED ROOM

Figure 4.15: Psychometric chart showing the thermal comfort ranges defined using four methods. The blue dots
correspond to the temperature and humidity data measured by H3, at a height of 2' from FFL in January, April
and June respectively.
Figure 4.15 shows the psychometric plot with comfort zones ranges for the four comfort zone
methods marked. Maximum number of points falls within the comfort zone in January when the
temperatures are cooler. Even in April 20% of points are within the comfort zone. In the monsoon
month, though relative humidity is high, the temperatures are cooler and hence, at least 50% of
the time, the temperatures are within the comfort zone (defined by TSI method) . The
temperatures fall within the comfort zone around 60% of the time during the winter month, 50%
in the monsoon months, and 20% of the time during the "summer" month of Apri

56
CHAPTER 5: UNDERSTANDING THE HEAT TRANSFER
PROCESSES IN THE BUILDING
In this chapter we will try to understand the various heat transfer processes that happen in the one
preselected zone (bedroom 1) of the building by performing analytical/numerical calculations
assuming simple models of the system (that is, the room). The emphasis will be on qualitatively
understanding the measured temperature data in terms of the heat transfer processes involved.
Results of the more accurate calculations of the full complex model by the aid of the simulation
software Design Builder are reported in Chapter 6 and 7.This section is intended to understand
the basic heat transfer mechanisms in the room, to help develop the model more accurately in
Design Builder. The calculated heat transfer values may not be accurate as the heat transfer
process in a dynamic system is extremely complicated. A more accurate values for heat transfer is
obtained using Design Builder in chapter 6 and 7, as the Energy Plus engine in Design Builder is
capable of calculating the complicated heat transfer mechanisms.
Although the general principles of heat transfer remain the same, the analysis of heat flow under
dynamic (rapidly changing) conditions is more complex than under steady state (fairly stable over
time) conditions. Steady state conditions mean that the temperature at each point is independent
of time. ie. In a heat transfer occurring in steady state conditions, the factors that affect the rate of
hear transfer is assumed to be constant all the time. Hence the steady state heat transfer
calculation has fewer variables than dynamic state heat transfer calculation. The major
determinant of steady state heat flow is thermal resistance (which is a measure of a temperature
difference by which an object or material resists a heat flow). In a heat transfer involving
dynamic conditions, all the factors that affect the heat transfer changes with time. In real life all
heat transfer occur in dynamic conditions. An analysis of heat flow occurring under dynamic

57
condition involves more variables, including thermal capacitance which is the ability of a body to
store thermal energy.
In this section we try to understand the heat transfer that occur inside the room, for a specific
period of time, say 24 hours. By comparing the indoor temperature obtained from the calculations
to the actual measured indoor temperature, we can check if the calculations are correct. This
process would help understand the methods of heat transfer in the chosen room in detail. Also this
knowledge would guide in developing the model in computer simulation software.
In order to have accurate values for indoor temperature, for a period of 24 hours, we need to
consider the heat transfer occurring under dynamic conditions. This is because the factors
affecting the heat transfer changes with time in reality. Hence the heat transfer equations used are
Differential Equations. Differential equations of heat transfer accurately calculate the rate of heat
transfer into the room by taking into account the rate at which all the parameter changes over
time. There are various sources which gives the dynamic equations for various modes of heat
transfer. These differential equations can be solved to get the rate at which heat is transferred into
the room at any point of time and using that value the indoor temperature can be calculated. In
this section, the differential equations for heat transfer have been solved using the software
Mathematica. The final results are plotted against time for 24 hours in Figure 5.3,Figure
5.5,.Figure 5.7 and Figure 5.8 .
In certain cases heat transfer equations under static cases can be applied, even though the heat
transfer is occurring under dynamic conditions. For example when the thickness of the material is
extremely small or when the thermal conductivity of the mass is extremely high, the heat transfer
equations occurring can be assumes as steady state.

58
In the following section, the steady state heat transfer equation has also been added, just for
reference purpose. The calculations in this chapter have been done using dynamic equations for
heat transfer.
5.1 HEAT TRANSFER PROCESSES
Before we proceed to calculate the heat transfer processes in the room, let us recapitulate the
basics of heat transfer by summarizing the different modes of heat transfer. There are there modes
through which heat is transferred to and from the room: conduction, convection and radiation.
5.1.1 CONDUCTION
Conduction is a mode of transfer of energy from the more energetic to the less energetic particles
of a substance due to interactions between particles. Thus, heat flows from the region at a higher
temperature to a region at a lower temperature.
In the context of buildings, heat conduction can happen through walls, roof etc. For the case of a
one-dimensional plane wall having a temperature distribution T(x), the heat transfer rate (heat
energy transferred per unit time per unit area) of conduction is given by the Fourier’s law:
dQ / dt = − k (dT / dx)
where k is called the thermal conductivity.
The steady state equation for conduction can be written as
q = U A Δt
where q = hourly heat transferred through the material/ hour
U = U factor of the material
Δt = temperature difference between inside and outside

59
5.1.2 CONVECTION
Convection is the mode of transfer of energy by means of collective or bulk movement molecules
within fluids. Here also energy flows from a region of higher temperature to a region of lower
temperature. Convection can take place only in fluids (as opposed to solids).
Heat transport by means of convection can happen in several situations in a building: for
example, chimneys, windows etc. facilitate heat transfer by means of convection. The heat
transfer rate of convection from a surface of temperature Ts to a fluid of temperature T
0
is given
by the Newton’s law of cooling:
dQ / dt = h (Ts − T
0
)
where h is called the convection heat transfer coefficient.
The heat transfer due to convection in a steady state system can be written as
q = 1.08 V Δt
where q= heat transferred / hour
V= airflow rate in cfm (cubic foot/min)
Δt = temperature difference between inside and outside
5.1.3 RADIATION
Thermal radiation is the energy emitted by matter that is at a finite temperature (above absolute
zero). The energy is transported by electromagnetic waves. While the transfer of heat energy by
conduction or convection requires the presence of a material medium, radiation does not require a
medium.

60
The most common source of heat gain in buildings is radiation gain from the Sun. Additionally,
the building itself absorbs heat from the Sun and re-radiates. If a body has a surface temperature
of Ts, the energy emitted by the surface can be calculated from the Stefan-Boltzmann law:
E = ε σ T
s
4

where σ = 5.67 ×10-8 W/m2/K4 is a fundamental constant, called the Stefan-Boltzmann
constant, and ε is a property of the surface, called the emissivity (0 ≤ ε ≤ 1).
The radiation that is incident on a unit area of a surface is called the irradiation G. A portion of
the irradiation may be absorbed by the surface, thereby increasing the thermal energy of the
material. The rate at which radiant energy is absorbed per unit area of the surface is given by:
G
abs
= α G
where α is also a property of the surface, called absorptivity (0 ≤ α ≤ 1).
In a steady state heat transfer system, the heat transfer through radiation is measured mostly as
mean radiant temperature. Mean radiant temperature (MRT) is simply the area weighted
mean temperature of all the objects surrounding the body. For easy calculations, the mean radiant
temperature is usually expressed as the surface temperature. The operative temperature which
affects the human comfort in an enclosed space is defined as sum of indoor dry bulb temperature
and half of the mean radiant temperature.
5.2 HEAT TRANSFER CALCULATIONS ASSUMING SIMPLE
MODELS
In this Section, we perform some calculations for the heat transfer processes that happen through
different elements constituting the room, such as walls, roof, windows etc. We will divide the
room into simple and idealized systems so that idealized models of heat transfer can be applied.
As mentioned earlier, the emphasis would be to make qualitative estimates. The temperature at

61
different parts of the room calculated using the heat transfer calculations will be compared with
the measured temperature data from the room. A good (qualitative) agreement between the
calculated and the measured data will provide some confidence in the models of heat transfer that
we employ. Using these models, we can then estimate the relative significance of the heat
gain/loss to the room by means of different mechanisms. This, in turn, will help to understand the
results of simulations reported in CHAPTER 6: .
5.2.1 EXPERIMENTAL DATA
Temperature relevant to heat transfer calculations in the room has been measured by four HOBOs
placed at different parts of the room. As discussed earlier, in order to avoid transient fluctuations
in the data due to everyday weather variations, the hourly temperature measured by each HOBO
averaged over 18 days starting from 27-Dec-2010 to 13 Jan-2011 was taken as the hourly
temperature of the corresponding measurement point in the room. The mean hourly data of each
HOBO is plotted in Figure 5.1 along with the minimum and maximum hourly temperature
measured over the 18-day period mentioned above. It can be seen that the daily variations in the
temperature can be considerable, while the mean temperature is a smoothly varying function of
time. These mean temperatures will be used for the heat transfer calculations in the following
sections.

62

Figure 5.1: The temperature data measured by four HOBOs over 18 days (27-Dec-2010 to 13 Jan-2011) plotted
against the hour of the day. The thick dots correspond to hourly temperatures averaged over 18 days, while the
error bars show the minimum and maximum measured temperature over the 18 days.
In this section we calculate the heat transfer through the roof. A schematic representation of the
different modes of heat transfer through the roof is given in Figure 5.2. In principle, the heat
conduction between the outer and inner surfaces of the reinforced-concrete roof also needs to be
accounted for. However, since the thickness of the roof is quite small (L = 0.1 m) as compared to
the area (A = 15.1 m2) and since the thermal conductivity of reinforced-concrete is quite high (k
≈ 1 W/m/K), the temperature gradients within the concrete roof are rather small. This property
will simplify the treatment of the problem.

63
5.3 HEAT TRANSFER THROUGH THE ROOF

Figure 5.2: Schematic representation of the different modes of heat transfer through the roof. Qsolar represents
the heat gain of the roof from the direct solar radiation (short-wave), Qrad-out represents the net long-wave
radiation on the outer surface of the roof (the difference between the heat gain of the roof from atmospheric
radiation and the heat loss due to radiation from the roof), Qrad-in represents the net long-wave radiation on
the inner surface of the roof, while Qc-out and Q c-in represent the convective heat transfer at the outer and
inner surface of the roof, respectively.
The heat transfer processes that we consider here are:
 Qsolar : The heat gain per unit time per unit area of the roof from the direct solar
radiation. This is estimated from the average hourly statistics of the direct normal solar
radiation for the month of January collected by ISHRAE
28
.
dQ
solar
/ dt = α
rsw
S.
Above, α
rsw
is the absorptivity of the outer surface of the roof to direct solar radiation,
and S is the energy density (W/m2) of the direct solar radiation after taking into account
the different angles of the incident solar radiation over a day and the effect of shadows on
the roof.

64
 Q
rad-out
: The net long-wave radiation per unit time per unit area on the outer surface of the
roof. This is the difference between the heat gain of the roof from atmospheric radiation
and the heat loss due to radiation from the roof. If T
sur
is the atmospheric temperature and
T
roof
is the temperature of the roof, the net long-wave (infra-red) radiation is given by
dQ
rad-
out / dt = α
r
σ T
sur
4
− ε
r
σ T
roof
4
,
where α
r
and ε
r
are the absorptivity and the emissivity of the concrete roof to long-wave
radiation, and σ is the Stefan-Boltzmann constant.
 Q
rad-in
: The net long-wave radiation per unit time per unit area on the inner surface of the
roof. Same as the above, except that this is the radiative heat transfer at the inner surface
of the roof.
dQ
rad-in
/ dt = α
r
σ T
room
4
− ε
r
σ T
roof
4
.
Above, Troom is the temperature of the air in the room.
 Q
c-out
: convective heat transfer per unit time per unit area at the outer surface of the roof.
This is due to the collective movement of air molecules over the outer surface of the roof,
and can be calculated as:
dQ
c-out
/ dt = h
ro
(T
sur
− T
roof
),
where h
ro
is the convective heat transfer coefficient of the outer surface of the roof.
 Q
c-in
: convective heat transfer per unit time per unit area at the inner surface of the roof.
Same as the above, except that this is the convective heat transfer at the inner surface of
the roof.
dQ
c-in
/ dt = h
ri
(T
room
− T
roof
),
where h
ri
is the convective heat transfer coefficient of the inner surface of the roof.

65
The net heat gain of the roof per unit time is the sum of the heat gain through all the different heat
transfer processes described above:
dQ
net /
dt = A ( dQ
solar
/ dt + dQ
rad
-out / dt + dQ
rad-in
/ dt + dQ
c-out
/ dt + dQ
c-in
/ dt ).
This will produce a temperature change of dTroof / dt in the roof, which is given by
dT
roof
/ dt = (dQ
net
/ dt) / (ρ
c
c
c
V
c
)
where ρ
c
is the density, c
c
is the specific heat capacity, and V
c
is the volume of the concrete roof.
The (approximate) values used for the parameters governing the heat transfer processes are given
in Table I. The mean hourly temperature measured by the HOBO-1 was assumed to be the
temperature of the atmosphere T
sur
, and the mean hourly temperature measured by the HOBO-3
taken as the temperature of the room T
room
(see Figure 5.1).
PARAMETER VALUE
Area of the roof (A) 15.1 m
2
Thickness of the roof (L) 0.1 m
Thermal conductivity of (dense) concrete (k) 1 W/m/K
Absorptivity of the outer surface for
direct solar radiation (α
rsw
)
0.5
Absorptivity of the inner and outer surfaces for indirect
(long-wave) solar radiation α
r
0.9
Emissivity of the inner and outer surfaces (ε
r
) 0.9
Specific heat capacity of (dense) concrete (c
c
) 880 J/kg/K
Density of concrete (ρ
c
) 2320 kg/m
3
Convective heat transfer coefficient: inner surface (h
ri
) 1.5 W/m
2
/K
Convective heat transfer coefficient: outer surface (h
ro
) 10 W/m
2
/K
Table 5.1: Summary of the parameters used for the calculation of the heat transfer through the roof.
By solving the differential equations for dQ
net /
dt and dT
roof
/ dt in Mathematica, we get the values
for heat transfer (heat loss and heat gain) into the room and the temperature of the roof, for the
period of 24 hours.

66
The expected temperature of the roof T
roof
calculated from the heat transfer through the roof is
plotted in Figure 5.3 (lower panel) along with the measured temperature of the roof, the
atmosphere and the room. It can be seen that the calculated value of the roof temperature (thin
black line) qualitatively reproduces the measured roof temperature (Cyan line) of the roof. The
top panel of Figure 2 shows the calculated heat transfer rate per unit area through the roof by
means of different heat transfer mechanisms. The following conclusions can be made from these
results:
Heat gain to the roof is almost entirely caused by the direct solar radiation on the roof. The roof is
contributing negatively towards the thermal comfort inside the room. It can be seen that, during a
significant part of the day (starting from 11 AM), the roof loses heat to the room causing the
room to gain heat (see the trace “Net transfer (in)” in the upper panel of Figure 5.3).
The time delay between the outside temperature and the temperature of the ceiling is due to the
fact that the partial shades on the ceiling blocks the direct solar radiation until around 11 AM.
This would also imply that shading the entire roof will significantly reduce the heat gain through
the roof. The maximum heat gain into the space though the roof occurs between 1pm and 2pm
and the maximum value for heat gain into the space from calculations is 250 W/m
2
and the
maximum value for the heat loss through the roof is 80W/m
2
. The heat gain/loss through the roof
at any point of time can be read from the graph (Figure 5.3).
The minor disagreement between the calculation and the measured data could be due to a number
of reasons:

67
The direct solar radiation is calculated from historical data of the average direct solar radiation for
the month of January, which can be considerably different from the mean direct solar radiation in
the particular 18 days that we have considered.
We have neglected the temperature gradients within the roof. Note that the roof is partially
shaded and the HOBO is placed on a region which is not covered by the shade.

Figure 5.3: The top panel shows the heat transfer rate per unit area through the roof by means of different heat
transfer mechanisms: Direct solar radiation, net long-wavelength radiation on the outer surface, convection on
the outer surface, net long-wavelength radiation on the inner surface, convection on the inner surface. Also
plotted are the net heat transfer on the outer surface, the net heat transfer on the inner surface and the sum of
these. Positive values indicate heat gain to the roof and negative values indicate heat loss from the roof. The
bottom panel shows the expected temperature of the roof T
roof
calculated from the heat transfer through the
roof (thin black line) along with the measured temperature of the ceiling (Cyan line). Also shown are the
temperature of the atmosphere and that of the room.

68
5.4 HEAT TRANSFER THROUGH THE WALLS
In this section, we calculate the heat transfer through the walls. A schematic representation of the
different modes of heat transfer through the wall is given in Figure 5.4. Unlike the case of the
roof, here we need to consider the temperature gradients inside the wall. This is due to the fact
that the wall is made out of hollow concrete blocks which are considerably thick (L = 0.25 m) and
have a rather low value of thermal conductivity (k ≈ 0.19 W/m/K). Thus we need to solve for the
conduction through the wall also using the heat diffusion equation.

Figure 5.4: Schematic representation of the different modes of heat transfer through the walls. Q
rad-out

represents the net long-wave radiation on the outer surface of the wall (the difference between the heat gain of
the wall from atmospheric radiation and the heat loss due to radiation from the wall), Q
rad-in
represents the net
long-wave radiation on the inner surface of the wall, while Q
c-out
and Q
c-in
represent the convective heat transfer
at the outer and inner surface of the wall, respectively, and Q
cond
represents the heat conduction through the
wall.
 Q
rad-in
: The net long-wave radiation per unit time per unit area on the inner surface of
the wall. Same as the above, except that this is the radiative heat transfer at the inner surface of
the wall.
dQ
rad-in
/ dt = α
w
σ T
room
4 −
ε
w
σ T
wall-in
4
.
Above, T
room
is the temperature of the air in the room.

69
 Q
c-out
: convective heat transfer per unit time per unit area at the outer surface of the
wall. This is due to the collective movement of air molecules over the outer surface of the wall,
and can be calculated as:
dQ
c-out
/ dt = h
wo
(T
sur
−
T
wall-out
),
where h
wo
is the convective heat transfer coefficient of the outer surface of the wall.
 Q
c-in
: convective heat transfer per unit time per unit area at the inner surface of the
wall. Same as the above, except that this is the convective heat transfer at the inner surface of
the wall.
dQ
c-in
/ dt = h
wi
(T
room
−
T
wall-in
),
where h
wi
is the convective heat transfer coefficient of the inner surface of the wall.
The temperature distribution T(x,t) (as a function of the position x in the wall and time t) as a
result of heat conduction can be calculated from the heat diffusion equation
27

k
w
∂
2
T(x,t) / ∂x
2
= ρ
w
c
w
∂T(x,t) / ∂t
where ∂T(x,t) / ∂t denotes a partial derivative of the temperature with respect to time t, and T(x,t) /
∂x denotes a partial derivative with respect to the distance x from the outer surface of the wall.
Also, k
w,
ρ
w
, c
w
are the thermal conductivity, density and the specific heat capacity of the wall,
respectively. The heat diffusion equation is solved with the following boundary conditions :
−
k
w
∂T(x,t) / ∂x = dQ
rad-out
/ dt + dQ
c-out
/ dt , at x = 0,
−
k
w
∂T(x,t) / ∂x =
−
(dQ
rad-in
/ dt + dQ
c-in
/ dt), at x = L,
and the initial condition:
T(x,t) = T
0
(x), at t = 0.

27
F. P. Incropera and D. P. De Witt, Fundamentals of Heat and Mass Transfer, Jonh Wiley & Sons (1996).

70
The initial temperature distribution in the wall T
0
(x) is assumed to be linearly distributed with
T
0
(x) = T
sur
at x = 0 (outer surface of the wall) and T
0
(x) = T
room
at x = L (inner surface of the
wall). The values used for the parameters governing the heat transfer processes are summarized in
Table 5.2.
PARAMETER VALUE
Thickness of the wall (L) 0.2 m
Thermal conductivity of the lightweight concrete block (k
w
) 0.19 W/m/K
Absorptivity of the outer surface for
direct solar radiation (α
wsw
)
0.5
Absorptivity of the inner and outer surfaces for indirect (long-
wave) solar radiation α
w
0.9
Emissivity of the inner and outer surfaces (ε
w
) 0.9
Specific heat capacity of light-weight concrete block (c
c
) 1000 J/kg/K
Density of concrete (ρ
c
) 1000 kg/m
3
Convective heat transfer coefficient: inner surface (h
wi
) 2.5 W/m
2
/K
Convective heat transfer coefficient: outer surface (h
wo
) 10.5 W/m
2
/K
Table 5.2: Summary of the parameters used for the calculation of the heat transfer through the wall.
The expected temperature of the wall T(x,t) calculated from the heat transfer through the wall is
plotted in Figure 34 (lower panel) along with the measured temperature data. The green line
corresponds to the temperature measured by HOBO-2 placed on the inner surface of the wall,
which we take as the temperature of the inner surface. The red line corresponds to the
temperature measured by HOBO-1 placed on the outer surface of the wall, which we take as the
temperature of the outer surface. Note that, actually these HOBOs measure a combination of the
temperature of the wall and that of the air. This introduces a minor ambiguity in the data and can
potentially cause disagreements between the calculated and measured temperature. In spite of
these limitations, the calculated temperature on the inner and outer surface of the wall
qualitatively agrees with the measured temperature.

71
The upper panel of Figure 5.4 shows the heat transfer rates per unit area by means of different
mechanisms. The following conclusions can be made by comparing this plot with the plot
showing the heat transfer through the roof (lower panel of Figure 5.3):
1. The heat gain through the wall (max transfer ~ 10 W/m
2
) is considerably small as compared
to that through the roof (max transfer ~ 200 W/m
2
).
2. The wall acts as a thermal mass thus contributing positively to the thermal comfort inside
the room: during the day time the wall gains heat from the room (providing a cooling effect to
the room) and during the night time the wall loses heat to the room (providing a heating effect
to the room). This is quite the opposite of the effect of the roof.

72

Figure 5.5: The top panel shows the heat transfer rate per unit area through the wall by means of different heat
transfer mechanisms: Net long-wavelength radiation on the outer surface, convection on the outer surface, net
long-wavelength radiation on the inner surface, convection on the inner surface. Also plotted are the net heat
transfer on the outer surface, and the net heat transfer on the inner surface. Positive values indicate heat gain to
the wall and negative values indicate heat loss from the wall. The bottom panel shows the expected temperature
of the wall calculated from the heat transfer through the wall (thin black line) along with the measured
temperature of the wall (Green line). Also shown are the temperature of the atmosphere and that of the room.

73
5.5 DIRECT SOLAR HEAT GAIN THROUGH THE WINDOWS
In this section, we estimate the direct solar heat gain through the windows. Since the absorptivity
of the air towards direct solar radiation (short-wave) is poor, the heat gain of the room from direct
solar radiation happens through secondary processes: materials in the room (such as the floor)
absorb solar radiation and gain heat energy. These materials re-radiate their internal heat energy
back to the room in the form of long-wave (infra-red) radiation, which gets trapped by the air
inside the room. This phenomenon is the well known greenhouse effect.

Figure 5.6: Schematic representation of the direct solar heat gain through the windows. Qsolar represents the
direct solar radiation (short-wave) through the window. The direct solar radiation transfers heat energy to the
floor, which in turn, radiates this energy back to the room by means of long-wave radiation. Qrad-f represents
the net heat transfer due to long-wave radiation from/to the floor (difference between heat gain of the floor from
radiation from the room and the heat loss due to radiation from the floor), and Qconv-f represents the
convective heat transfer from the floor.
A schematic diagram of solar heat gain through the windows is given in Figure 5.6. Here we will
assume that the entire solar radiation that enter the room incidents on the floor. The heat transfer
processes that we consider here are:

74
 Q
solar
: The heat gain of the floor per unit time from the direct solar radiation. This is
estimated from the average hourly statistics of the direct normal solar radiation for the month of
January collected by ISHRAE
28

dQ
solar
/ dt = A
g
α
fsw
S.
Above, α
fsw
is the absorptivity of the floor surface to direct solar radiation, S is the energy density
of the direct solar radiation after taking into account the different angles of the incident solar
radiation and the position and geometry of the windows, and A
g
is the total area of the windows.
 Q
rad-f
: The net heat gain from the long-wave radiation from/to the floor per unit time.
This is the difference between the heat gain of the floor from radiation from the room and the
heat loss due to radiation from the floor. If T
room
is the ambient temperature of the room and
T
floor
is the temperature of the floor, the net long-wave (infra-red) radiation is given by
dQ
rad-f
/ dt = A
f
( α
f
σ T
room
4 −
ε
f
σ T
floor
4
),
where α
f
and ε
f
are the absorptivity and the emissivity of the floor to long-wave radiation, A
f
is the
total surface area of the floor and σ is the Stefan-Boltzmann constant.
 Q
conv-f
: convective heat transfer from/to the floor per unit time. The convective heat
transfer can be calculated as:
dQ
conv-f
/ dt = A
f
h
f
(T
room
−
T
floor
),
where h
f
is the convective heat transfer coefficient of the surface of the floor.
The net heat gain of the floor per unit time is the sum of the heat gain through all the different
heat transfer processes described above:
dQ
net
/ dt = dQ
solar
/ dt + dQ
rad-f
/ dt + dQ
conv-f
/ dt.
This will produce a temperature change of dT
floor
/ dt in the floor, which is related to the heat gain
dQ
net
/ dt by
dT
floor
/ dt = (dQ
net
/ dt) / (ρ
c
c
c
V
c
)

75
where ρ
c
is the density, c
c
is the specific heat capacity, and V
c
is the volume of the concrete floor.
The expected heat transfer to the floor from direct solar radiation is shown in the top panel of
Figure 5.7 and the calculated temperature of the floor is shown in the bottom panel. The heat gain
from direct solar radiation through windows is peaked at two distinct times: around 9-10 AM
(radiation through the East window) and at around 4-5 PM (radiation through the South window).

Figure 5.7: Top panel shows the heat transfer to and from the floor by different mechanisms: long-wavelength
radiation exchange between the floor and the air in the room, convective heat transfer between the floor and the
air in the room, and direct solar radiation. Clearly the heat transfer during the day is dominated by the direct
solar radiation. The two peaks of the radiation are due to the radiation coming from three different windows on
the East and South walls. Positive values correspond to heat gain by the floor and negative values to heat loss by
the floor. The bottom panel shows the calculated temperature of the floor along with the measured temperature
of the atmosphere, the ceiling, air inside the room and the inner surface of the South wall.

76
5.6 PUTTING IT TOGETHER: A HEAT TRANSFER BUDGET FOR
THE ROOM
In this section, we combine the heat transfer estimates through different elements (that we
calculated in the earlier sections) to provide a “heat transfer budget” for the room. That is, we try
to “predict” the temperature in the room by combining the heat gain/loss to the room through the
different elements.
Figure 5.8 shows a summary of the heat transfer budget. The top panel shows the calculated heat
transfer to/from the room by means of different mechanisms and various envelope elements (roof,
wall, windows etc.). It can be seen that, in the forenoon the heat gain to the room is dominated by
the radiative and convective transfer from the floor. (The floor absorbs the direct solar radiation
coming through the windows and releases the stored internal heat energy by convection and long-
wave radiation to the room). In the afternoon, on the other hand, the heat gain to the room is
dominated by the radiative and convective transfer from the hot ceiling (heat energy gained from
direct solar radiation on the roof). Most of the time (except for a few hours in the afternoon) the
walls act as a heat sink, providing a cooling effect to the room. The convective heat transfer
through the windows is quite inefficient. This is partly due to the fact that the window is open
only during the day hours (only one window is open even during the day hours).
It can be seen from Figure 5.8 that the maximum heat gain into the space occurs through the roof
( ~360 W) and the second highest heat gain occurs through the direct solar heat gains through the
windows(~160 W). The maximum heat loss occurs through the walls (~ 80 W). In CHAPTER 7:
the building is simulated in an energy calculation software and more accurate and detailed values
for heat transfer into the room are obtained.

77

Figure 5.8: Top panel shows the heat transfer to and from the room by different mechanisms: long-wavelength
radiation exchange between the air inside and outside the room through windows, radiation exchange with the
floor, the roof and the walls, convective heat transfer through the (open) windows, convective heat transfer with
the surface of the floor, the roof and the walls. (Note that the high heat gain from the floor from 8AM to 2PM is
caused by the direct solar radiation absorbed by the floor). Positive values indicate heat gain to the room and
negative values indicate heat loss from the room. The bottom panel shows the predicted ambient temperature of
the room (thin black trace) along with the measured ambient temperature of the room (blue traces). Also shown
are the predicted temperature of the floor, and the measured temperatures of the atmosphere, the ceiling and
the inner surface of the wall.
The bottom panel of Figure 5.8 shows the predicted ambient temperature (thin black trace) of the
room by considering the heat transfer processes mentioned above. The measured ambient
temperature in the room is also plotted (blue trace). Here also the calculated temperature agrees

78
qualitatively with the measured ambient temperature in the room. The difference between the
predicted and measured values of the ambient temperature can be explained in the following way:
there exists a temperature gradient in the room (higher temperatures at larger heights). The
calculated ambient temperature is the mean temperature of the entire room without considering
this temperature gradient. On the other hand, the measured temperature corresponds to the
temperature at a particular height (0.5 above the floor). Note that the average temperature of the
room is expected to be somewhere between the measured temperature at a height of 0.5 m and the
temperature of the ceiling. This is what we observe.
This simple heat transfer budget suggests that the dominant source of heat gain to the room is the
direct solar radiation falling on the roof and entering the room through the windows. Providing
external shades to the roof and windows would help reducing the heat gain of the room.
Improving natural ventilation at night by opening the windows will help the room to efficiently
loss heat during the early hours of the night. The actual heat transfer rate occurring in the space is
calculated in more detail through computer simulated models in chapter 7. The effectiveness of
each of these modifications is also quantified there.

79
CHAPTER 6: BUILDING SIMULATION AND ANALYSIS
6.1 SOFTWARE SELECTION
Design Builder is a software tool for investigating the energy, carbon, lighting and comfort
performance of the building. The particular version of the software selected for this simulation
was Design Builder Beta X (from here onwards, referred to as DB). This software was chosen
based on following considerations.
DB has an extremely user friendly interface, that helps model all geometries.
DB could model and simulate naturally ventilated spaces.
DB is one of the few software that can model temperature floats.
The weather tool option of DB was used to develop the weather data for the simulation, based on
the actual measured data.
6.2 WEATHER DATA FOR SIMULATION
DB uses epw weather format for weather files. The weather file for Thiruvanthapuram was
downloaded from the Energy Plus Energy Simulation Software website. The weather data
developed by the Indian Society of Heating, Refrigerating and Air-Conditioning Engineers
(ISHRAE), is obtained from the TMY2 weather data that consists of the historical average of all
weather data parameters averaged over many years. Real Time weather data for the year 2011
was also obtained from the Energy Plus website
28
upon request. Comparing the outdoor
temperature data obtained from external sources (ISHRAE and Energy Plus) with the data
measured on-site using HOBOs (Figure 7.1), it was seen that the measured outdoor temperature
varied significantly from the 1990 data, but was consistent with the year-2011 data within a range
of 2 degree Celsius. Also it was seen that the TMY 2 weather data had a larger diurnal variation

80
in all months when compared to the Real time weather data obtained from Energy Plus website
and weather data measured on site. Since the year 2011 data obtained from Energy Plus website
had data missing over certain days of the months, the measured outdoor temperature and
humidity data (using HOBOs) was used to create a new weather file. Other parameters such as
wind velocity, cloud cover, direct solar radiation etc. was taken from the weather file of 1990,
since they were not measured on-site (nor were they available from Energy Plus). The fact that
we use historical data for parameters other than temperature and humidity is expected to cause
some level of error in our analysis.
6.3 THE DESIGN BUILDER MODEL
The model was developed in Design Builder with all the elements such as walls, roofs, glazing
etc. having the characteristics and thermal parameters as close to the original building as possible.

Figure 6.1 View of the model from the South direction

81

Figure 6.2 View of the model from the South direction

Since the original building had more than 15 rooms (including the toilets and storage rooms).
some of the rooms were lumped together in the model and the walls separating these spaces from
the other spaces were defined as adiabatic. If a wall is defined as adiabatic the software assumes
that the temperature on either side of that wall is same and there is no net heat transfer between
the two spaces. This approach was used in order to simplify the simulation, as too many zones
would complicate the simulation results.

Figure 6.3: View of the model from the East direction.

82

Figure 6.4 Plan showing the Building Zones in DB in the first Floor(Left) and Second Floor (Right)
6.3.1 MODEL CONSTRUCTION DETAILS
This section explains the details of the construction assigned when the model was developed in
DB.
The external walls were modeled as aerated concrete blocks with cement sand plastering on both
sides. Figure 6.5 shows the thermal parameters and the surface properties of the walls.

83

Figure 6.5 Details of the wall construction in DB.

Figure 6.6 Details of roof construction in DB.

84
The roof and floor slabs were 4 inch thick reinforced concrete slabs. The roof had cement-sand
plastering on the outer side and surface properties of the plaster were changed to be the same as
the outside surface properties of the roof in the original building. The windows were single glazed
with painted window frames. The partial truss with terracotta tiles above the flat roof was also
modeled. Figure 6.6 and Figure 6.7give the details of the roofs, floors and window glazing. More
details of envelope parameters and schedules are given in Appendix.
In DB, the activity level inside the rooms as well as the hours of operation can be scheduled. For
example, it can be scheduled what percentage of windows open at what time of the day. Such
details have been collected from the original residence and the schedules in the DB model have
been set as close to the original model as possible
.
Figure 6.7 Details of glazing in the model in DB.
6.4 CALIBRATION OF THE MODEL
The model developed in DB was calibrated against the temperature and humidity measured from
the original building. This was done to ensure that the software was running the simulations
correctly and to gauge the error bars in the results. In order to calibrate the model, the weather

85
data was fed to the model, and the temperature inside the modeled building was extracted, and
was compared against the measured temperature inside the building. As described in the earlier
section, the weather data for the year 2011 was incomplete (only temperature and humidity
measurements were available for the year 2011; all other parameters such as wind speed, solar
radiation, cloud cover etc. were given by averaged historical data collected by ISHRAE). Thus,
the weather file was unable to record the transient daily weather fluctuations in 2011 accurately.
However, we expect that the trends in the hourly weather (averaged over several days) should be
very similar every year. Thus, we averaged the hourly records of various parameters
(temperature, humidity, solar radiation etc.) over one month, and used these averaged
measurements in the weather file. In summary, we have 24 data points (corresponding to average
temperature) for each month and these were compared against the similarly averaged temperature
measurements from the inside of the building (bedroom 1).
The temperature of bedroom 1 on the second floor in the simulated model and the measured
temperature in the room were compared for the three months of the year: January (the winter
month), April (the summer month) and June (the monsoon month). As discussed earlier, these
three months represent the weather conditions during the three major seasons in Kerala and all the
other months follow the pattern of one of the three months.

86

Figure 6.8 Temperature measured from the HOBO in bedroom1 on the second floor and the simulated
temperature for January plotted against the hours of the day. Also shown is the outdoor temperature.

Figure 6.9 Temperature measured from the HOBO in bedroom1 on second floor and the simulated temperature
for April plotted against the hours of the day. Also shown is the outdoor temperature.

87

Figure 6.10 Temperature measured from the HOBO in bedroom1 on the second floor and the simulated
temperature for June plotted against the hours of the day.
Figure 6.8,Figure 6.9 and Figure 6.10 show the room temperature plotted against hours of the day
for bedroom1 for three different months (January, April and June). The blue line represents the
actual measured room temperature, the red line represents the DB simulated room temperature
and the green line represents the outdoor temperature. As discussed earlier, since there are many
unknown factors in the weather data used for simulation (such as the wind speed and solar
radiation), an accurate match with the measured values was not expected. When the simulated
values followed the same trend of the measured values over the course of the day and fell within a
range of 1 to 1.5ºF of the measured temperature, the calibration of the model was deemed
acceptable.

88

Figure 6.11 The top panel shows the temperature plotted against hour of the day, the middle panel shows the
heat balance plotted against hour of the day (the points above zero indicate heat gain into the room and below
zero indicate the heat loss from the room) and the lower panel shows the air changes/hour inside the room
plotted against hour for the day. These plots are from January.
Figure 6.11 shows the DB output for the month of January where temperature, heat balance and
air changes per hour in the room are plotted against time. It can be seen that the maximum heat
gain into the room occurs through the roof during daytime. The second highest heat gain source is
through the windows and glazing. The heat gain through convection through windows is small.
This could be due to the fact that only a single window is opened during day time (from 8 to
18:30) and all the windows are closed during night time. The high thermal mass walls and floors
act as a heat sink during day time, creating a cooling effect and releases heat into room at night.
The cooling effect of the walls and floors can be calculated as follows:

89
Let us consider the heat transfer in the room at any point of time during the day, say at 17:00 on
11
th
of January. It can be seen that the indoor air temperature is lower than the outdoor
temperature at that point of time. It means that the hot air from outside that enters the room
through open windows is being cooled by some elements in the room. From the Design Builder
heat balance graph, it can be seen that the walls and floor are the elements that absorb the heat
from the room. The following paragraph describes the steps to compute the cooling effect of the
walls and floors (in kBtu/hr)
Net heat gain into the space = Heat gain from roof (Q
roof
) + Heat gain through glazing (Q
glazing
)
+ Heat gain through convection (through open windows and infiltration; Q
convection
) + Heat gain
from the occupants. (Q
occupants
) (eq. 6.1)
Heat gain through any element of the envelope can be found using the equation
Q = U × A × ΔT, (eq. 6.2)
where
Q = hourly heat loss /gain through specific component (Btu/h)
U = conductivity of the element. ( Btu/(h ˚F sq ft))
A = Area of the element (sq ft)
ΔT = temperature difference between the exterior and interior of the element (˚F)
By substituting the values of U, A and ΔT in eq. (6.2, we get
Q
roof
= 0.897 × 243.5 × 2.94 = 650 Btu/h.
The convective heat transfer through the open windows can be calculated as
Q
convection
= CFM × 1.08 × ΔT
air
(eq. 6.3)

where CFM is volume of air change in cubic foot/min = (ACH × Room volume)/60 = ( 5.75 ×
2374.12)/60 = 227.5CFM.
Thus, the convective heat transfer can be calculated from eq.(6.3) as

90
Q
convection
= 227.5 × 1.08 × 1.2 = 294.84 Btu/hr (eq. 6.4)
The heat gain through is obtained from the Design Builder as
Q
glazing
= 435.2 Btu/hr (eq. 6.5)
Q
occupancy
is usually considered as 250 Btu/person. Design Builder calculates the occupancy load
by using detailed heat balance equations by taking into consideration the metabolic factor and
clothing factor, the surface area of the human body (a fixed value depending on male, female or
child) and the variations in the room temperature. In this case at 17:00, Design Builder has given
the total occupancy heat gain load as 241 kBtu.
The net heat gain into the space at the selected point of time can be calculated using eqs.(6.1) to
(6.5) as,

Q
net
= 1.84 kBtu/sq ft
Thus, the minimum cooling effect (Q
cooling
) provided by walls and floors = 1.84 kBtu/sq ft. Note
that, the value of Q
cooling
through the walls and floor calculated by Design Builder is 1.82 kBtu/sq
ft. Hence it can be concluded that the heat gain/loss values calculated by Design Builder is fairly
accurate. For all further discussions the heat gain/loss through each element as calculated by
Design Builder is used.

91

Figure 6.12The top panel shows the temperature plotted against hours of the day, the middle panel shows the
heat balance plotted against hour of the day (the points above zero indicate heat gain into the room and below
zero indicate the heat loss from the room) and the lower panel shows the air changes/hour inside the room
plotted against hour for the day. These plots are from April.
It can be seen from Figure 6.12 that the heat gain and heat losses inside the spaces are similar to
those in January, but there is an increase in air changes inside the room on account of increased
air velocity. The direct solar radiation through the windows is reduced in summer, since the sun is
at a higher angle and the overhangs provide shading to the windows most of the time.
Figure 1.13 shows that there is lesser heat gain through the roof in June than in April which could
be on account of reduced direct solar radiation due to the monsoon. But the heat gain through the
roof is higher than in January. This could be explained by the fact that the partial truss with tiles
above the concrete roof provides shading more time during January when the sun is at a slightly

92
lower angle and in June, this shading effect is reduced on account of high solar altitude. The air
changes per hour in the room are highest in June, although only single window is opened all
through the year, due to the increased outdoor air velocity in June.

Figure 6.13The top panel shows the temperature plotted against hours of the day, the middle panel shows the
heat balance plotted against hour of the day (the points above zero indicate heat gain into the room and below
zero indicate the heat loss from the room) and the lower panel shows the air changes/hour inside the room
plotted against hour for the day. These plots are from June.

93
CHAPTER 7: APPLICATION OF STRATEGIES FOR
IMPROVING THERMAL COMFORT
In this section, the effects of applying various strategies of passive cooling in the simulated
building model are discussed. Selected strategies of passive cooling were applied to the selected
room in the baseline model and the effects on the indoor temperature and humidity were
analyzed.
In order to understand the diurnal effects of each selected strategy, the simulations were
performed using the mean hourly outdoor weather data averaged over a month. The indoor
temperature of the modified simulated model is compared against that of the baseline model over
three different months – January, April and June. As a result of these comparisons, the average
difference in indoor temperature due to application of each of the selected strategies can be
clearly understood. In order to understand how the thermal comfort conditions in the room
change when each strategy is applied to the model, the temperature and relative humidity from
the simulation runs (with various strategies applied) are plotted in the psychometric chart. The
number of points that fall inside the comfort zone in each modified model is compared to the
number of points inside the comfort zone in the baseline model. If the number of points inside the
comfort zone in the modified model is larger than that in the baseline model, it means that the
percentage of time the room is inside the comfort zone higher in the modified model. In that case,
it can be concluded that the passive cooling strategy has created a positive impact in the room in
terms of thermal comfort.
Each simulation had been performed using three different weather data file.

94
The first set of simulations, as discussed above uses the mean hourly outdoor weather data. The
output shows 24 data points corresponding to each hour of a typical day. This clearly
demonstrates the effect of each strategy on inside room temperature over a diurnal period when
compared to the baseline model.
Since mean hourly average excludes the extreme temperatures and relative humidity, those data
points are excluded when plotted in the psychometric chart. In order to see how the room
performs when the strategies are applied for the entire range of temperature for each month, the
models are simulated with the outdoor measured weather data. This weather file uses the outdoor
dry bulb temperature (DBT) and relative humidity measured at site for the entire month. The
output of the simulations is the indoor temperature and relative humidity for each hour of the day
and for all the days in the selected months. This would include (24 x 30/31) data points in the
psychometric chart.
The third set of simulations were run using the energy plus weather file (.epw) which is based on
the data obtained from the 'typical meteorological year '(TMY)weather data file, from the Energy
Plus website of the US Department of Energy
28
“A typical meteorological year (TMY) is a
collation of selected weather data for a specific location, generated from a data bank much longer
than a year in duration and it gives annual averages of weather data parameters that are consistent
with the long-term averages for the location”. Since the TMY data is the historical data averaged
over many years,it represents the data for an average year and does not include any extreme
conditions for the location. However, TMY data is the most common weather data used by
designers .This weather data was used for the third set of simulations and the output data was
plotted in the psychometric chart. This was compared with the baseline model output, in order to

28
http://apps1.eere.energy.gov/buildings/energyplus/cfm/weather_data.cfm

95
understand which strategy creates the most positive impact in terms of thermal comfort. This
would invariably give different results than the simulations using measured outdoor data weather
file. A comparison of the difference in the outdoor temperature from the TMY file and the data
measured on site is shown is the Figure 7.1). In conclusion, a comparison is made in the
effectiveness of strategies of passive cooling when different weather files are used for simulation.

Figure 7.1 A plot showing the comparison of outdoor dry bulb temperature for a week of January, obtained
from three different weather files. The blue line shows the real time weather data file (the weather data file for
the year 2011)obtained from energy plus website. The green line shows the TMY2 Data obtained from energy
plus website which shows the historic average dry bulb temperature over many years. The red line shows the
outdoor dry bulb temperature measured at site using Hobos. The measured weather data was always much
higher than the

7.1 STRATEGIES OF PASSIVE COOLING APPLIED
The following selected strategies were applied:
 Improved cross ventilation: In the baseline model only a single window was opened
during day time and all windows were closed during night time. The effect of cross
ventilation was checked by opening the windows located in the adjacent walls. Two
scenarios were simulated where the room was only night ventilated, ie all the windows

96
were opened only in night time and when the room was ventilated 24 hours, ie all the
windows were opened 24 hours.
 Effect of Roof: .Three different scenarios including placing a flat shade 1' above the
concrete roof, a pitched roof with limited ventilation and with increased ventilation and
white washing the outer surface of the concrete roof was simulated. The resultant
temperature and humidity data from the room was analyzed.
 Providing vents near the ceiling of the room: Providing opening near the ceiling of the
room creates a vent for hot air(especially if there is a stratification of air in the room) to
escape out, which would result in cooler air moving into the room through the windows
which is at a lower height. This effect was simulated in the model and the effect of
providing openings of varying sizes on a single wall and on two adjacent walls was
studied.
 Reducing heat gain through windows: Since the second highest heat gain into the space
was through the windows, some selected strategies such as providing an external shading
for the windows and using Low E glass for glazing was applied and the results were
analyzed.
 Shading of walls through a veranda: A simple veranda (8' wide horizontal projection) was
modeled in DB around the rooms so as to provide shade for the wall. Though walls were
not a major source of heat gain into the space, this strategy was used to study the effect of
a verandah.

97
7.2 ANALYSIS OF EFFECTS OF APPLICATION OF PASSIVE
COOLING STRATEGY IN DESIGN BUILDER MODEL USING
MEAN HOURLY AVERAGED OUTDOOR WEATHER DATA
This section discusses the effect of the passive cooling strategies in the indoor room temperature
during a typical day 24 hour period. The comfort zone conditions of the room and percentage of
time the room falls inside the comfort zone when the strategies are applied is discussed in section
7.2.13.

Figure 7.2 : 3D view showing the position of the windows in the selected room. The windows are open 24 hours
for this simulation (though it does not show in the 3D view)
7.2.1 IMPROVED CROSS VENTILATION WITH WINDOWS OPENED FOR
24 HOURS.
The chosen room had windows in adjacent walls. The window opening size on the windward size
was smaller than the window opening size on the windward side. The operating schedule was set
to 24 hours. This means that all the windows in the room are open 24 hours a day. The

98
comparison of the indoor temperature in the modified model and the baseline model is shown in
the Figure 7.3, Figure 7.4 and Figure 7.5.
From the figures, it can be seen that when the room is cross ventilated 24 hours, the room
temperature follows the outdoor air temperature very closely. This has a positive effect during
night time as the indoor room temperatures are lowered during night time for all three months
during daytime in April and May, the indoor air temperature was higher than outdoor air
temperature in the base case. Cross ventilating the space 24 hours created a slight improvement
by reducing the indoor temperatures and bringing it close to outdoor temperature. But this
strategy did not have a positive impact in January as it increased the indoor temperature more
than that in the base case. But in general, it can be said that the application of this strategy
brought down the indoor temperature a large portion of the time.

Figure 7.3: shows the comparison of indoor temperature in the baseline model and modified model (windows
open 24 hours) over 24 hours of the day in January. The red line indicates the room temperature of the
simulated model when the windows were opened 24 hours. The blue line indicates the the room temperature and
the green line indicates the outdoor dry bulb temperature.

99

Figure 7.4: is same as Figure 7.3, except that graph shows the values for the month of April.

Figure 7.5: is same as Figure 7.3, except that graph shows the values for the month of June.
7.2.2 IMPROVED CROSS VENTILATION WITH WINDOWS OPENED ONLY
DURING NIGHT TIME
In this case, the windows in the model were opened only during night time. All the windows were
closed during day time.

100

Figure 7.6: shows the comparison of indoor temperature in the baseline model and modified model (windows
opened only during night time) over 24 hours of the day in January. The red line indicates the room
temperature of the simulated model when the windows were opened only during night time. The blue line
indicates the indoor dry bulb temperature for the base model and the green line indicates the outdoor dry bulb
temperature.

Figure 7.7: is same as Figure 7.6, except that graph shows the values for the month of April.
.As in the case before, Figure 7.6, Figure 7.7 and Figure 7.8 show the comparison of indoor air
temperature of the modified model and the baseline model for the months of January. April and
June. It can be seen that the indoor air temperature is lowered during night time in all the three
cases . Except in June the daytime indoor temperature is lower than that of baseline model. In

101
conclusion it can be said when the room is cross ventilated at night time, the indoor temperature
is lower than that of the baseline model for the majority of the time.

Figure 7.8: is same as Figure 7.6, except that graph shows the values for the month of June.
7.2.3 WHITEWASHED FLAT ROOF
The roof of the selected room is a flat reinforced concrete slab. Four strategies of passive cooling
were applied to the roof elements and the effects of each were noted.
In order to understand the effect of whitewash, the roof properties were changed on the outer
surface of the roof. The thermal emissivity
29
, solar absorptivity
29
and visible absorptivity
29
values
for newly whitewashed surface were provided for the outer surface of the roof.

29
B.Givoni, Man,Climate and Architecure, Applied Science Publishers LTD, 1976

102

Figure 7.9 shows the Design Builder material property window showing thermal absorptivity and emissivity
values provided on the outer surface of the white washed roof.

Figure 7.10: shows the comparison of indoor temperature in the baseline model and modified model (with roof
white washed) over 24 hours of the day in January. The red line indicates the room temperature of the simulated
model when the roof is white washed. The blue line indicates the indoor dry bulb temperature for the base
model and the green line indicates the outdoor dry bulb temperature.

103

Figure 7.11: is same as Figure 7.10, except that graph shows the values for the month of April.

Figure 7.12: is same as Figure 7.10, except that graph shows the values for the month of June.
By comparing the indoor dry bulb temperature in the base model with that in the modified model
in Figure 7.10, Figure 7.11 and Figure 7.12, it can be seen that the temperatures are always lower
for the model with the white washed roof. Due to the lower absorptivity of the whitewashed

104
surface the heat gain through the roof is minimized during daytime. Due to the high emissivity
value of the white washed roof surface, it radiated more heat to night sky, consequently reducing
the temperatures during day time and night time. The heat balance graph from Design Builder
(Figure 7.13 ) clearly illustrates the reduced heat gain through the roof during daytime and
increased heat loss during night time.

Figure 7.13: shows the heat gain and loss in the room through various elements such as walls, roof, glazing etc.
The horizontal; axis is the hours of the day and the vertical axis is the heat balance in kBtu/hr. If the vertical
axis is positive, it indicates the heat gain into the room and negative values indicate heat loss from the room. The
upper panel shows the heat balance graph of the base model and the lower panel shows the heat balance plot of
the model with white washed roof.
7.2.4 PROVIDING SHADE OVER THE CEILING WITH A FLAT ROOF
ABOVE.
A flat roof was provided above the ceiling of the room at a height of 12” from the ceiling height.
This would create a shade over the ceiling and thereby reduce the direct solar radiation hitting the

105
outer surface of the ceiling. The air space in between the flat roof and the ceiling of the room is
ventilated as the sides of the air space are open.

Figure 7.14 : 3D view showing flat roof above the ceiling of the room.

Figure 7.15 shows the comparison of indoor temperature in the baseline model and modified model (flat shade
over the ceiling) over 24 hours of the day in January. The red line indicates the room temperature of the
simulated model when the room ceiling is shaded with a flat roof above. The blue line indicates the indoor dry
bulb temperature for the base model and the green line indicates the outdoor dry bulb temperature.

106

Figure 7.16 is same as Figure 7.15, except that graph shows the values for the month of April.

Figure 7.17 is same as Figure 7.15, except that graph shows the values for the month of June.

107
By comparing Figure 7.15,Figure 7.16 and Figure 7.17, it can be seen that the indoor air
temperature of room is lower than the base model in April and June where as in January, the
temperature is close to the base model.

7.2.5 PROVIDING A NON VENTILATED ATTIC
Another strategy of passive cooling applied was providing a pitched roof shade over the ceiling
the room, thereby creating a double roof and a non occupied attic space. In the initial case, the
attic space had only very minimal ventilation. No openings were provided on the sides or top of
the pitched roof .
It can be seen from Figure 7.19,Figure 7.20 and Figure 7.21 that the indoor dry bulb temperature
of the modified model is lower than the base model in all the cases. The heat gain through the
roof is minimized as the direct solar radiation on the roof is prevented by the pitched shade above.
Out of all the three months, the indoor temperature drop in January due to the application of this
passive strategy is minimal where as there is more substantial temperature drop in April and June.
This method is effective as it prevents the direct solar radiation hitting the roof and consequently
reduces the heat gain through the roof. Comparing the heat balance graph of the base model (top
panel of Figure 7.22 ) to the heat balance graph in the modified model (middle and bottom panels
of Figure 7.22), it can be seen that the heat gain into the space through the roof in the second case
is reduced in the modified model.

108
.
Figure 7.18: 3D view of the house with double roof ( unoccupied attic space) and no attic ventilation

Figure 7.19 shows the comparison of indoor temperature in the baseline model and modified model( pitched roof
over room ceiling with minimal ventilation in the air space between roof and ceiling) over 24 hours of the day in
January. The red line indicates the room temperature of the simulated model with the unventilated attic space
above. The blue line indicates the indoor dry bulb temperature for the base model and the green line indicates
the outdoor dry bulb temperature.

109

Figure 7.20 is same as Figure 7.19, except that graph shows the values for the month of April.

Figure 7.21 is same as Figure 7.19Figure 7.20, except that graph shows the values for the month of June
.

110

Figure 7.22 shows the heat gain and loss in the room through various elements such as walls, roof, glazing etc.
The horizontal; axis is the hours of the day and the vertical axis is the heat balance in kBtu/hr. If the vertical
axis is positive, it indicates the heat gain into the room and negative values indicate heat loss from the room. The
upper panel shows the heat balance graph of the base model. In the lower panels which shows the heat balance
plot of the model with shaded roof in January and April, the reduction in heat gain through the roof can be
clearly seen when compared to te heat balance plot of the base model. .
7.2.6 PROVIDING SHADE OVER THE CEILING WITH VENTILATED
PITCHED ROOF ABOVE.
In this case the ceiling of the room was protected from direct solar radiation with a pitched roof
which was open on either ends facilitating constant air movement in the space between the roof
with tiles and the concrete ceiling of the room below. The increased air movement in the space
between the pitched roof tile and the ceiling of the room, would ideally help dissipated the heat in
the attic space and in turn should reduce the heat in the room.

111

Figure 7.23: 3D view of the house open sided pitched roof above the room ceiling.

Figure 7.24 shows the comparison of indoor temperature in the baseline model and modified model over 24
hours of the day in January. The red line indicates the room temperature of the simulated model when the
pitched roof. Shading the ceiling of the room below has open sides which facilitates ventilation. The blue line
indicates the indoor dry bulb temperature for the base model and the green line indicates the outdoor dry bulb
temperature.

112

Figure 7.25 is same as Figure 7.24, except that graph shows the values for the month of April.

Figure 7.26 is same as Figure 7.24, except that graph shows the values for the month of April.

113
By comparing Figure 7.24,Figure 7.25 and Figure 7.26, it can be seen that the indoor temperature
of the room is lower than that in the base model in January, April and June. The reduction in
temperature is very minimal in January. More prominent temperature difference can be seen in
April and June.
7.2.7 PROVISION OF OPENING NEAR THE CEILING ON ONE WALL
An opening of size 3’x 1’ was provided on one wall near the ceiling of the room. It is expected
that the hot air that rises to the top of the room would be flushed out through this opening. This
would create a lower pressure in the room, which would pull in the cooler air from outside
through the other windows. Ideally, this strategy would help create a continuous air movement in
the space and would help lower the air temperature in the room, when the outside air is cooler
than the indoor room temperature.

Figure 7.27 3D view of the house showing the opening near the ceiling.

114

Figure 7.28 shows the comparison of indoor temperature in the baseline model and modified model (opening
near the ceiling in one wall) over 24 hours of the day in January. The red line indicates the room temperature of
the simulated model which has an opening near the ceiling on one wall. The blue line indicates the indoor dry
bulb temperature for the base model and the green line indicates the outdoor dry bulb temperature.

Figure 7.29 is same as Figure 7.28, except that graph shows the values for the month of April.

115

Figure 7.30 is same as Figure 7.28 except that graph shows the values for the month of June.
By comparing Figure 7.28, Figure 7.29 and Figure 7.30, it can be seen that the indoor temperature
is lower than that in the base model for all the three months, although the difference is minimal in
January.
7.2.8 PROVISION OF OPENING NEAR THE CEILING ON BOTH WALLS
In this case openings of size 6’x 1’ were provided on two adjacent walls. This strategy was
employed to understand the effect of increased size of opening near the ceiling of the room. By
comparing the indoor room temperature with that of base case (Figure 7.32, Figure 7.33 and
Figure 7.34), it can be seen that the indoor air temperature is lower than the base model in April
and June. But the increased size of the opening has not made a significant improvement over the
case before, when the room had a single opening near the ceiling. Also, in the case discussed
above, in January, the indoor air temperature was very close to the base model. But in this case
(increased area of opening near the ceiling) in January, the indoor air temperature goes above that
of the base model during day time. The increased ventilation in the room due to increased area of

116
openings in two walls brings in the air from outside at a faster rate. Since the outside air
temperature is higher than the room temperature in January, the indoor room temperature became
higher than that in the base case (where day time ventilation rate was lower).

Figure 7.31 3D view of the house showing the openings near the ceiling on both walls.

Figure 7.32 shows the comparison of indoor temperature in the baseline model and modified model (openings
near the ceiling on both walls) over 24 hours of the day in January. The red line indicates the room temperature
of the simulated model which has an opening near the ceiling on both walls. The blue line indicates the indoor
dry bulb temperature for the base model and the green line indicates the outdoor dry bulb temperature.

117

Figure 7.33 is same as Figure 7.32, except that graph shows the values for the month of April.

Figure 7.34 is same as Figure 7.32Figure 7.28 except that graph shows the values for the month of June.

118
7.2.9 SHADING WITH EXTERNAL LOUVERS
In this case the windows are shaded with external louvers which would ideally reduce the direct
solar radiation into the room though the windows and glazing.

Figure 7.35 3D view of the house showing the external shading on windows

Figure 7.36 shows the comparison of indoor temperature in the baseline model and modified model (external
louvers shading the ) over 24 hours of the day in January. The red line indicates the room temperature of the
simulated model which has external louvers shading the windows. The blue line indicates the indoor dry bulb
temperature for the base model and the green line indicates the outdoor dry bulb temperature.

119

Figure 7.37 is same as Figure 7.36, except that graph shows the values for the month of April.

Figure 7.38 is same as Figure 7.36 except that graph shows the values for the month ofJune.
By comparing the Figure 7.36, Figure 7.37 and Figure 7.38, it can be seen that indoor air
temperature is reduced during day time in January and April. From the heat gain graph, it can be

120
clearly seen that the heat gain through the glazing is significantly lower in this case. But there is
not much difference in indoor temperature in June. This is due to the fact that June is the
monsoon month and the direct solar radiation is low due to the cloudy skies. Since the direct solar
radiation is already low, there is not much effect on indoor room temperature when shading is
provided for windows in June.
7.2.10 PROVIDING SINGLE GLAZED LOW E WINDOWS
In this case the windows are provided with Low E single glazing. A Low-E coating is a
microscopically thin, virtually invisible, metal or metallic oxide layer deposited directly on the
surface of one or more of the panes of glass. Low E coating absorbs the long wave radiation from
outside and minimizes the infrared radiation from the warm pane of the glass to inside.

Figure 7.39 shows the comparison of indoor temperature in the baseline model and modified model (single
glazed Low E windows) over 24 hours of the day in January. The red line indicates the room temperature of the
simulated model which has Single glazed Low E windows. The blue line indicates the indoor dry bulb
temperature for the base model and the green line indicates the outdoor dry bulb temperature.

121

Figure 7.40 same as Figure 7.39except that graph shows the values for the month of April.

Figure 7.41 same as Figure 7.39 except that graph shows the values for the month of June.
From Figure 7.39,Figure 7.40 and Figure 7.41 it can be seen that single glazed low E coating
doesn’t make much difference in the indoor temperature in January and April when compared to
that of the base model. In June, the indoor temperature of the modified model is lower than that in
the base model.

122
7.2.11 PROVIDING DOUBLE GLAZED LOW E WINDOWS
In this case room has double glazing with Low E coating in the outer pane. In warm climates,
Low e coating is provided on the outer pane of the window.

Figure 7.42 shows the comparison of indoor temperature in the baseline model and modified model (single
glazed Low E windows) over 24 hours of the day in January. The red line indicates the room temperature of the
simulated model which has double glazed Low E windows. The blue line indicates the indoor dry bulb
temperature for the base model and the green line indicates the outdoor dry bulb temperature.

Figure 7.43 same as Figure 7.42 except that graph shows the values for the month of April.

123

Figure 7.44 same as Figure 7.42 except that graph shows the values for the month of June.
From Figure 7.42,Figure 7.43 and Figure 7.44 it can be seen that providing double glazed window
with Low E coating actually increased indoor temperature in January and April over the base
model. One explanation is that the short wave radiation entering the room is absorbed by the
furniture and reradiated as long wave radiation. The Low E coating might be reflecting this long
wave radiation back into the space and thereby increasing the temperature of the room.
7.2.12 PROVIDING SHADING WITH VERANDA
In this case a veranda 8’ wide was modeled as a projecting slab at the ceiling height. The effect of
shading the walls and windows with a veranda was studied in this case.

124

Figure 7.45 3D view of the house showing the effect of veranda using an 8’ wide overhang.

Figure 7.46 shows the comparison of indoor temperature in the baseline model and modified model (with a 8’
wide projecting slab) over 24 hours of the day in January. The red line indicates the room temperature of the
simulated model which has 8’ wide projecting slab outside the walls. The blue line indicates the indoor dry bulb
temperature for the base model and the green line indicates the outdoor dry bulb temperature.

125

Figure 7.47 same as Figure 7.46 except that graph shows the values for the month of April.

Figure 7.48 same as Figure 7.46 except that graph shows the values for the month of June
From Figure 7.46, Figure 7.47 and Figure 7.48, it can be seen that this strategy reduced the indoor
temperature when compared to that of the base model in all the three months, but not by a
significant amount.

126
7.2.13 COMPARISON OF EFFECTIVENESS OF EACH STRATEGY USING
PSYCHOMETRIC CHART.
In this section, the effectiveness of each strategy is compared against the base model. This is done
by plotting the indoor temperature and relative humidity in the psychometric chart and checking
what percentage of time the room falls inside the comfort zone range in each case. As discussed
earlier in section 2.5, the adaptive method using running mean and mean monthly average does
not give an upper limit for humidity. Tropical Summer Index also does not specify an upper limit
for humidity. But when the comfort range of temperature (25˚C to 30˚C) is marked in the
psychometric chart (Figure 2.10), the authors of the paper “Tropical Summer Index- a Study of
thermal Comfort of Indian subjects” limit the comfort zone range within 70% of relative
humidity, citing that extreme relative humidity values may be undesirable. Since an upper limit
for relative humidity could not be defined, for the purpose of defining thermal comfort range (and
checking what percentage of time the indoor temperature and humidity falls in the comfort zone
range) for this study, two separate checks are done. In one case it is checked what percentage of
points fall inside the comfort zone range if the upper limit of humidity is limited as 70%. In the
other case it is checked what percentage of points fall inside the comfort zone range if no upper
limit of humidity is defined.
7.2.14 PLOTTING OUTPUT FOR SIMULATIONS USING ACTUAL
MEASURED WEATHER DATA AT SITE
7.2.14.1 COMFORT ZONE DEFINED WITHOUT AN UPPER RANGE FOR HUMIDITY
Following figures show the indoor temperature and humidity plotted in the psychometric chart for
the simulations using the actual measured outdoor temperature at site. For better clarification in
visualization, the region marked as the comfort zone range is zoomed out. The shaded region is
the comfort zone range, ie the range of temperature and humidity in which people are

127
comfortable. In this case an upper limit of relative humidity is not defined. The non shaded
regions are outside the comfort zone. Each of the colored dots corresponds to a data point that
represents a particular indoor temperature and relative humidity. Each color shows the data points
from the simulation when a particular strategy is applied. The black cross represents the data
points from the simulation of the base model. By comparing the number of points of a single
color that fall inside the shaded region to the total number of points in the same color, the
percentage of time the room is inside the comfort zone can be calculated for each strategy.

Figure 7.49 This figure shows a portion of psychometric chart zoomed out for better clarity. The shaded region
represents the range of temperature and humidity in which people are comfortable. The non shaded areas are
the areas outside comfort zone. Each of the colored dots corresponds to a data point that represents a particular
indoor temperature and relative humidity. Each color shows the data points from the simulation when a
particular strategy is applied. The black cross (Reference) represent the data points from the base model. The
colored dots in this plot correspond to the data points from the simulations using the actual measured weather
data in site, for the January.

128

Figure 7.50: Same as Figure 7.49, but the dots correspond to data points for April.

Figure 7.51: Same as Figure 7.49, but the dots correspond to output data points for June.

129

Figure 7.52: This figure shows the percentage of time the room is in the comfort zone (when the comfort zone is
not limited by an upper limit of relative humidity) in the vertical axis and each of the passive cooling strategies
applied to the base model on the horizontal axis. The three months is represented in three colors. For eg. When
the room was cross ventilated only during night time, 80% of the time, the room was comfortable in January
and June, whereas in April, the room was comfortable only 55 % of the time. Likewise the effect of each strategy
can be read from the graph. The percentage of time of time the room fall is in the comfort zone is expressed as
‘comfort fraction’. The reference represents the comfort conditions of the room in the base model.
In Figure 7.52, the percentage of time the room falls inside the comfort zone (the percentage of
time people feel comfortable inside the room) is expressed as comfort percentage. It is the ratio of
the number of data points that fall inside the shaded region (comfort zone) to the total number of
data points in each simulation case in Figure 7.49,Figure 7.50 and Figure 7.51 The comfort
percentage is plotted in the vertical axis and the various strategies employed to the base model to
improve the thermal comfort in the room is plotted in the horizontal axis. This figure shows the
percentage of time the room was in comfort zone when the different cases of the model ( ie when
various strategies of passive cooling applied) was simulated in Design Builder. The first point in

130
the horizontal axis is the reference which is the base model. It can be seen that when the base
model was simulated, the room was inside the comfort zone 70% of the time for January, 73% of
the time for June and 30% of the time for April. By comparing the comfort fraction of the room
when various strategies are applied to the base model, it can be seen that most of the strategies
improves the thermal environment of the room ie the room is more comfortable than in the base
model. The most effective strategy in making the room comfortable for large percentage of time
seems to be cross ventilation during night time.
7.2.14.2 COMFORT ZONE LIMITED TO 70% HUMIDITY AS UPPER RANGE.
Figure 7.53 shows the percentage of time the room is inside the comfort zone when the upper
limit comfort zone range is defined as 70% relative humidity range.ie, only the data points that
fall in the shaded area below the 70 % relative humidity curve in the psychometric charts in
Figure 7.49, Figure 7.50 or Figure 7.51 would be considered as within comfort range. Since the
climate of Thiruvanathapauram has relatively high humidity all the year round, very few data
points from the simulations fall inside the comfort range when the comfort range is limited by
70% relative humidity. In the monsoon month of July, almost all the time the room is outside the
comfort zone. Even in January, when the space is cross ventilated 24 hours a day, the room is
comfortable only 3.2% of the time. The percentage of time the room is comfortable never goes
beyond 3.5% in any month for any strategy applied.

131

Figure 7.53:Same as Figure 7.52, except that in this case comfort zone has an upper limit for relative humidity.
ie. A data point is considered inside the comfort range if it falls in the shaded area below the 70 % relative
humidity curve in the shaded region of the psychometric chart in Figure 7.49, Figure 7.50 or Figure 7.51 .
7.2.15 PLOTTING OUTPUT FOR SIMULATIONS USING TMY2 DATA
7.2.15.1 COMFORT ZONE DEFINED WITHOUT AN UPPER RANGE FOR HUMIDITY
In this section the indoor temperature and humidity data from the simulations using the TMY2
data from energy plus website (the historical average outdoor weather data) are plotted in the
psychometric chart. As in the case above the percentage of time the room is comfortable is
computed by seeing what percentage of the total number dots (that represent indoor temperature
and humidity for every hour of the day for January, April and June) fall in the shaded region in
the psychometric chart.

132

Figure 7.54: The figure shows a portion of psychometric chart zoomed out for better clarity. The shaded region
represents the range of temperature and humidity in which people are comfortable. The non shaded areas are
the areas outside comfort zone. Each of the colored dot correspond to a data point that represents a particular
indoor temperature and relative humidity. Each color shows the data points from the simulation when a
particular strategy is applied. The black cross represents the data points from the base model. The colored dots
in this plot represent the output data points from the simulations using TMY2 weather data from, for January.

Figure 7.55: Same as Figure 7.54, but the dots correspond to output data points for April.

133

Figure 7.56: Same as Figure 7.54, but the dots correspond to output data points for June.
In this case the data points are spread over a wider range in the psychometric chart as compared
to data points from the simulations with actual measured weather data.
In Figure 7.57, it can be seen that the room in the base model (labeled ‘Reference’) has lower
comfort fraction in all the three months than in Figure 7.52. The main difference in the comfort
fraction for these simulation output is due to the fact that the range of outdoor temperature is
much larger here than the actual measured outdoor temperature. Here the room in the base model
is comfortable 10% of the time in January, 20% of the time in June and40% of the time in April.
The percentage of time the models with each of the strategies applied fall in the comfort zone is
quite different from the values in Figure 7.52.In this case the strategy that makes the rooms most
comfortable in all the months is the provision of openings near the ceiling in both walls.

134

Figure 7.57: This figure is similar to Figure 7.52 and shows the percentage of time the room is in the comfort
zone (when the comfort zone is not limited by an upper limit of relative humidity) in the vertical axis and each of
the passive cooling strategies applied to the base model on the horizontal axis. The three months is represented
in three colors.

135
7.2.15.2 COMFORT ZONE LIMITED TO 70% HUMIDITY AS UPPER RANGE

Figure 7.58: Same as Figure 7.57, except that comfort zone in this case has an upper limit for relative humidity.
ie. In this case a data point is considered inside the comfort range if it falls in the shaded area below the 70 %
relative humidity curve in the psychometric chart in Figure 7.54, Figure 7.55 or Figure 7.56.
As explained in Section 7.2.14.2 , the figure above shows the comfort fraction for room for
various strategies applied, if the upper limit of relative humidity for comfort zone is considered as
70% in the psychometric chart. Again, it can be seen that the percentage of time the room falls
inside the comfort zone reduces greatly if the upper limit for relative humidity is fixed at 70%. In
the monsoon month of June the room is outside the comfort zone most of the time, irrespective of
any of the strategies applied.

136

7.2.16 CATEGORIZING THE STRATEGIES ON THEIR EFFECTIVENESS ON
IMPROVING THERMAL COMFORT
This section discusses the effectiveness of all the strategies applied to the base model, on basis of
their effectiveness in improving thermal comfort inside the room. By plotting all the data points
for the entire three months in the psychometric chart for various strategies and comparing it with
that of the base model, the percentage of time each strategy improved the thermal comfort inside
the room can be understood. The strategies are arranged according to their effectiveness, which is
assessed by the percentage improvement of thermal comfort in the room. When the upper limit of
comfort zone was defined as 70% relative humidity in psychometric charts discussed earlier, it
was seen from Figure 7.53 and Figure 7.58 the percentage of time the room was comfortable was
very minimal (the location being a tropical climate.). Hence the strategies are listed based on their
effectiveness based on the percentage of time the room is sin comfort zone, when there is no
upper limit of humidity for comfort zone.
The percentage of time the room is in the comfort zone varies widely, depending on the outdoor
weather file used. Hence the two tables showing percentage improvement of thermal comfort
inside the room when the simulations were done using two different weather files are shown.
Table 7.1 and Table 7.2 shows the list of strategies in the decreasing order of their effectiveness
(with the most effective strategy at the top), when the simulations were run using the real
measured weather data and TMY2 data respectively.
The first column list the base model and the various passive cooling strategies applied to the
model. The second column represents the percentage of time in the entire year the room is inside
the comfort zone when each strategy is applied to the base model. In all the tables ‘reference’

137
represent the data from the base model. The next three columns show the percentage of time of
each month (January, April and June respectively) the room is inside the comfort zone when each
strategy is applied.
Table 7.3and Table 7.4 shows the improvement in the percentage of time the room was in
comfort zone over that of the base model. Table 7.3 corresponds to simulations performed using
the measured weather data on site, while Table 7.4 corresponds to TMY2 weather data.
Modification
Full
year

January April June
Cross vent 24 hrs 72 74 60 84
Cross vent at night 71 81 56 82
Opening near ceiling (both walls) 66 72 50 80
Pitched roof (min vent) 64 70 42 84
Flat shaded roof 64 70 42 84
Opening near ceiling (one wall) 63 71 41 82
Pitched roof (ends open) 62 68 40 82
Whitewashed roof 60 75 34 78
Reference( base model) 55 69 31 74
Window: shading with ext louvres 55 66 30 74
Window: single glazed low E 55 68 30 73
With verandah 54 69 30 71
Window: double glazed low E 50 63 25 70
Table 7.1:Strategies in the decreasing order of their effectiveness( from the top) when the simulations were run
using the Actual measured weather data. Columns 2 , 3, 4 and 5 shows the percentage of time the room is in
inside the comfort zone for the whole year, January, April and June respectively.

138

Modification
Full
year January April June
Opening near ceiling (both walls) 71 96 48 84
Flat shaded roof 43 72 48 26
Cross vent 24 hrs 39 16 39 50
Cross vent at night 37 1 41 49
Window: double glazed low E 36 0 54 33
Opening near ceiling (one wall) 36 8 50 33
Pitched roof (min vent) 30 2 48 25
Pitched roof (ends open) 29 2 46 25
Whitewashed roof 27 3 44 20
Reference (base model) 26 8 41 19
With verandah 26 6 41 19
Window: shading with ext louvres 24 4 39 18
Window: single glazed low E 23 9 36 15
Table 7.2: Similar to Table 7.1, except that the simulations were run using the TMY2 weather data.

Modification Full year
Cross vent 24 hrs 17
Cross vent at night 16
Opening near ceiling (both walls) 11
Pitched roof (min vent) 9
Flat shaded roof 9
Opening near ceiling (one wall) 8
Pitched roof (ends open) 7
Whitewashed roof 5
Reference ( base model) 0
Window: shading with ext louvres 0
Window: single glazed low E 0
With verandah 1
Window: double glazed low E -5
Table 7.3: This table shows the improvement in the percentage of time the room was in comfort zone over that
of the base model (Reference) when the simulations were run using the actual measured weather data.

139

Modification Full year
Opening near ceiling (both walls) 46
Flat shaded roof 17
Cross vent 24 hrs 13
Cross vent at night 11
Window: double glazed low E 10
Opening near ceiling (one wall) 10
Pitched roof (min vent) 5
Pitched roof (ends open) 4
Whitewashed roof 1
Reference( base model) 0
With verandah 0
Window: shading with ext louvres -2
Window: single glazed low E -3
Table 7.4: Similar to Table 7.3, except that the simulations were run using the TMY2 weather data.
In either case it can be seen that the provision of a veranda and shading the windows with
external louvers did not make any improvement in the base model. This might be due to the fact
that the base model has inside diffusing blinds in place and hence the heat gain through the
glazing was not significant. The double glazed windows provided a positive improvement in the
simulation using TMY2 weather data, where as it performed poorly in the simulations that use
actual measured weather data. This can be explained by the fact that the TMY2 data has some
extremely low outdoor temperatures, in which case a double glazed window would help retain the
heat and would improve the thermal comfort in the room. On the other hand when the outdoor
temperature is high, as in the case of real measured weather data, this would negatively affect the
thermal comfort of the space.

140
In general it can be said that the strategy that is most effective seems to be cross ventilation. The
strategy that creates the second highest positive impact in the space is providing openings near
the ceilings followed by strategies that include different variations of roof shading.

141
CHAPTER 8: PARAMETRIZING THE INDOOR
TEMPERATURE IN TERMS OFWINDOW OPENING AREA
In this section, we deal with parametrizing one of the strategies from the list of effective
strategies in the last chapter. Cross ventilation proved as one of the most effective strategies in
improving the thermal comfort of the selected room. Cross ventilation increased the air changes
in the room. By increasing the cross ventilation through the room, the temperature of the room
would eventually become the same as outdoor temperature. This would create a positive impact
on thermal comfort if the outdoor temperature is lower than indoor temperature.
In order to understand how the indoor temperature varies with the area of window opening in a
naturally ventilated space in Thiruvanathapuram, the selected room was simulated in Design
Builder with different window opening areas and the corresponding indoor temperature was
noted.
Windows with equal size was provided on two adjacent walls of the room. The area of the
windows was varied as a fraction of floor area. For eg. In the first case, the area of window
opening was taken as 10% of the total floor area. Similarly nine simulation were run with window
areas being 15%, 20%, 25%, 30%, 35%, 40%, 45% and 50% of floor area (FA). The simulations
were also run for the case when the windows were shut and when a single window was opened (ie
there is ventilation but no cross ventilation). The simulations were run using the hourly averaged
weather data (discussed in Chapter 6, section 1.4) for the summer month of April, for a period of
24 hours. The indoor temperature of each case was compared against the outdoor temperature.

142

Figure 8.1 shows the exploded of the view room and the position of windows in the room.
Figure 8.2 shows the temperature on the vertical axis and the hours of the day on the horizontal
axis.

Figure 8.2 : This figure shows the temperature in ˚F in the vertical axis and the hours of the day in the
horizontal axis. The different lines represent the hourly averaged indoor temperature (except the orange line
which is the outdoor temperature) of the room for 24 hours for the month of April.

143
It can be seen that when all the windows were closed the indoor temperature was larger than
outdoor temperature by 3˚F during the day and 5˚F during the night. When a single window was
opened, the indoor temperature reduced by almost 1.5 ˚F. For the rest of the cases, both the
windows were opened 24 hours. When the area of the windows was 10% of the floor area, the
temperature of the room was very close to the outdoor temperature. Subsequent increase in
window sizes reduced the indoor temperature at night to some extent. The maximum indoor
temperature during day time equaled the maximum outdoor temperature even the window area
was 10% of the floor area. The indoor minimum temperature reduced when the window area sizes
was increased, until the window size was 30% of floor area. When the window opening was
increased beyond 30% of floor area, there was no further decrease in temperature, as the indoor
minimum temperature was almost same as the outdoor minimum temperature.
The difference between the indoor temperature and the outdoor temperature was calculated for 24
hours for each case. The mean of the temperature difference was compared to see the pattern in
which the temperature changed when the area of window opening was increased. This is
illustrated in Figure 8.3, where the area of window opening is plotted against the mean of
temperature difference.

144

Figure 8.3 This figure shows the rate of change of mean temperature difference between indoors and outdoors
with respect to the opening area of the window. The vertical axis show the mean of the difference between
outdoor temperature and indoor temperature. The horizontal axis shows the area of window opening as a
percentage of floor area of the room.
It can be concluded that as far as reduction in indoor temperature through cross ventilation is
concerned, if the window area is at least 10% of floor area, the effect is most prominent
.Increasing the window area beyond 10% slightly reduces the temperature. This effect is not seen
when the window area is greater than 35% of floor area.

145
CHAPTER 9: CONCLUSION
9.1 SUMMARY
Kerala, the southwestern state of India that falls under the category of wet tropical climate zone,
has been famous for employing passive cooling techniques in traditional architecture for
improving thermal comfort in residences. Over the years there have been changes in construction
practices, techniques and materials. Many studies have suggested that modern residences
performed poorly in terms of thermal comfort when compared to traditional residences. The
current research focused on analyzing the effect of selected passive cooling techniques on thermal
comfort of the room in a modern residence in Thiruvanathapuram, Kerala. As an initial step, a
thermal comfort study was carried out, in order to define a thermal comfort range for the climate
of Kerala. The temperature and humidity data from the residence and the site was collected for a
period of one year .The data was analyzed for 3 specific months of the year. This helped in
developing an understanding of the indoor thermal environment of the house. The residence was
simulated in the energy analysis software Design Builder. The indoor temperature in the
simulated model was compared with that of the original measured data from the site. When the
indoor temperature from the simulated model came within a range of 1 to 1.5° F, the model was
deemed calibrated. This simulated model was considered as the base model against which all
other modified models were compared. A list of different strategies of passive cooling methods,
which would be effective in the climate of Kerala, was composed. Each of this strategy was
applied to the base model and the effect in the thermal comfort of the room was compared against
that in the base model. The process was repeated using two different weather files- the weather
file with data measured on site for the year 2011 and the epw weather file from the Energy Plus
website which gives the historical average weather data for Kerala. The percentage of time the

146
room was inside the comfort zone when each strategy was applied and for the base model was
calculated. The strategies were listed in the order of the effectiveness in improving the thermal
comfort of the room. As a final step, one of the most effective strategies of passive cooling was
selected and a parameterization study was carried out in the simulated model.
9.2 LIMITATIONS
The accuracy of final results of the research is limited due to the various assumptions made for
the purpose of the study. Some of the assumptions made in this study are discussed in this
section.
The indoor temperature and the relative humidity data measured using the HOBOs are within the
range of the error as shown in the calibration process of the HOBOs (Section 4.2.1). Due to
limited availability of data collection tools, wind speed and incident solar radiation data was not
measured on site. For the purpose of developing weather data file for the year 2011, the historic
average wind velocity and solar radiation data obtained from the Energy Plus website were used.
The simulations run using this weather file is within the range of errors of this assumption.
Though care had been taken to thermally insulate the HOBOs from the surface on which they
have been mounted, one of the HOBOs (H1) recorded the surface temperature of the concrete
ceiling rather than the air temperature. In the simulated model in Design Builder, most of the
internal partitions of the rooms were considered as adiabatic in order to simplify the model in the
energy calculation software. But in reality there is heat transfer occurring between various rooms.
In applying various strategies of passive cooling techniques, most simplified version of the
strategy was applied to the base model in simulations. For example, in the case of shading the
walls using a veranda, a simple 8’ wide horizontal overhang was used as the verandah. The effect

147
of the veranda might be slightly different if a pitched shade with a slight opening at the top was
applied.
Also, the effect of all the strategies of passive cooling have been worked out on the basis of their
effect on the thermal comfort in the rooms of one residence that has been simulated. The results
might be slightly skewed if a different residence is simulated. By repeating the process with more
residences located in Kerala and by comparing the results with that of the current study, it can be
checked if the simulations results of the majority of the cases support the results of this study.
9.3 CONCLUSION
Since the climate of Thiruvanathapuram is hot and humid, it is difficult to bring the indoor
temperatures inside the comfort zone (as defined by the commonly employed methods) all the
year round. However, many of the strategies of passive cooling provided an improvement in
thermal comfort as compared to the base case, where no passive cooling strategy was applied.
As explained in Section 7.2.14 and Section 7.2.15, the simulation was run using two different sets
of weather data. The results were slightly different in both cases. The reasons for the difference in
output are explained in CHAPTER 7: . In both the cases when the upper limit of thermal comfort
zone is limited by 70% relative humidity, the percentage of time the room was inside the comfort
zone was extremely small. In case of simulation using actual measured weather data, the
percentage of time the room was inside the comfort zone was below 5% irrespective of the
passive cooling strategy use. But when the simulation was run using the TMY2 weather data, the
thermal comfort of the room was improved. This is mainly because the TMY2 weather data had
lower temperature and humidity than the actual measured weather data. By analyzing the results

148
of the simulation (using actual measured weather data and TMY2 weather data) without defining
an upper limit of humidity, the following conclusions could be made.
1. In both the cases, the strategies that seem to be most effective (in terms of bringing the
room in the comfort zone majority of the time in a year) are cross ventilation, providing
opening near the ceilings and various methods of shading the roof. White washing also
produced a significant improvement in the thermal comfort of the room when compared
to the base case.
2. Singled glazed Low E glazing windows and double glazed Low E windows were not
effective in improving thermal comfort. Shading the windows with external louvers also
did not make much improvement in thermal comfort of the room as compared to the base
case.
3. Since the walls were not a major source of heat gain into the room in this house, shading
the walls with verandah was not very effective in improving the thermal comfort of the
room. The results might be different if a more ingenious version of the veranda is used
for simulation; for example, In the place of the 8’ wide horizontal projection as the
verandah roof, a sloping roof with an opening at the top which would help to create a
continuous air movement through the space below, could be used.
4. The parameterization study on cross ventilation aspect of the room, suggests that for this
climate, the effect of cross ventilation was highest when area of the window opening was
between 10 % and 30 % of the floor area. When the area of window opening was beyond
35% of the floor area, no further improvement in the thermal conditions pf the rooms was
observed.

149
5. All the conclusions from this study is based on the data and simulation done on a single
residence in Thiruvananthapuram. Though all the conclusions hold true in the case of this
residence, similar studies should be carried out in other buildings in this climate in order
to generalize the conclusions for all residential buildings.
9.4 FUTURE WORK
This study establishes a method to check the effectiveness of the various strategies of passive
cooling techniques in a chosen building. In order to generalize the conclusions of this study for
residential class of buildings in this climate, similar study should be carried out in more buildings
in Thiruvanathapuram If all the simulation yield similar results, the conclusions of this study can
be generalized for this climate.
Also, as a next step to this research, the list of passive cooling strategies can be expanded and the
effect of the combination of these strategies on the thermal comfort of the room can be studied.
Another step to enhance the results of this study would be to do a parametric study of the most
effective strategies of passive cooling.

150
BIBLIOGRAPHY
ANSI/ASHREA Standard 55 – 2010: Thermal Environmental Conditions for Human Occupancy,
American Society of Heating, Refrigerating and Air-conditioning Engineers, 2010.
“Architectur”, Home Gallery Design, accessed on 29
th
November 2011,
http://homegallerydesign.com
Areemit, N and Sakamoto,Y: Numerical and experimental analysis of a passive room-
dehumidifying system using the sorption property of a wooden attic space; International Journal
of Low Carbon Technologies, Vol 4( 2009) Page:141 - 149
Brager, G.S., and De Dear, R.J. (1998), Thermal Adaptation in the Built Environment: A
Literature Review: Energy and Buildings, 27, 83-96.
“Climate Consultant 5”, Energy Design Tools, accessed 23rd September 2011,
http://www.energy-design-tools.aud.ucla.edu/
“Climatic Regions of India”, Wikipedia, accesed 25
th
August 2011,
http://en.wikipedia.org/wiki/Climatic_regions_of_India
“Data Logger”, Onset HOBO Data Loggers, accessed 5
th
September 2011,
http://www.onsetcomp.com/
De Dear, R.J: Adaptive Thermal Comfort in Building Management and Performance; Healthy
Buildings, June 2006
De Dear, R.J: Adaptive Thermal Comfort- Past, Present and Future
Dili, A.S and Naseer, M.A.and Vargese, T. Zacharia : Passive control methods of Kerala
traditional architecture for comfortable indoor environment: A comparative investigation during
winter and summer,. Building and Environment 2010; 45:1134-1143
Fanger, P.Ole and Toftum, John: Thermal Comfort in the Future – Expectation and Excellence,
International Conference 'Moving Thermal Standards to 21st Century.'
Feriadi, Henry: Thermal comfort for naturally ventilated residential buildings in the tropical
climate; Master Thesis, National University of Singapore, 2003.
Givoni, B: Man, Climate and Architecture, Applied Science Publishers Ltd, 1976
Givoni, Baruch, Passive and Low Energy Cooling of Buildings: John Wiley & Sons Inc; 1994
Government of India: National Action Plan on Climate Change.
“Houses for Rent”, OLX, accessed 29
th
November 2011, http://kochi.olx.in

151
Incropera, F. P. and De Witt, D. P.: Fundamentals of Heat and Mass Transfer, John Wiley &
Sons, 1996.
Indraganti, Madhavi: Adaptive use of natural ventilation for thermal comfort in Indian
apartments; Building and Environment, Vol 45 (2010) Page:1490–1507
Krishnan, A and Baker, N and Yannas, S and Szolkolay S V: Climate Responsive Architecture –
A Design Handbook for Energy Efficient Buildings,Tata McGraw-Hill Publishing Company
Limited; 2001
Ministry of Environment and Forest, Government of India: India Green House Emissions, 2007.
Sharma, M.R.and Sharafat, Ali :Tropical Summer Index – A Study of Thermal Comfort in Indian
Subjects ;Building and Environment, 1986,Vol 21,No 1, pp 11-24.
Srivastav,S, and Jones, P.J.: Use of traditional passive strategies to reduce the energy use and
carbon emissions in modern dwellings; International Journal of Low Carbon Technologies, Vol
4(2009) Page: 141 – 149
“Temperature Data Loggers and Sensors”, Products, Onset HOBO Data Loggers, accessed
5
th
September 2011, http://www.onsetcomp.com/products/data-loggers/U10-data-loggers
“Thermal comfort”, Clear (Comfortable Low Energy Architecture), accessed 17
th
September
2011, http://www.new-learn.info/packages/clear/thermal/index.html
“Traditional Kerala House”,16
th
January 2011( 10:20 p.m),South India Travel Points Blog,
http://southindiatravelideas.blogspot.com
Vargese, P.C.: Building Construction; PHI Learning Private Limited, October 2008.
“Weather Data”, EnergyPlus Energy Simulation Software, accessed 25
th
November 2011,
http://apps1.eere.energy.gov/buildings/energyplus/cfm/weather_data.cfm

152
APPENDIX: SCHEDULES IN DESIGN BUILDER
This section shows the various settings used in Design Builder for the base case scenario and for
some of the cases when passive cooling strategy was applied.

Figure A1 Activity settings in Design Builder used for all simulations

153

Figure A2 Construction settings in Design Builder used for all simulations.

Figure A3 HVAC settings in Design Builder used for all simulations. Only heating and cooling is turned off and
Natural ventilation is on.

154

Figure A 4 Light settings in Design Builder used for all simulations

Figure A5 Schedule for natural ventilation for all simulations. The space is naturally ventilated 24 hours a day.
The air movement and air changes in the room are controlled by the closing and opening the windows.

155

Figure A6 General Window settings in Design Builder used for all simulations. The percentage of opened
glazing area and the duration during which it is opened is set individually for each window.

156

Figure A7 Window settings for the south side window in the south east corner bedroom in the second floor.

Figure A8 Schedule for opening and closing the south side window in the south east corner bedroom in the
second floor.

157

Figure A9 Window glazing data for simulations using single glazed Low E windows.

Figure A10 Window glazing data for simulations using double glazed Low E windows

Abstract (if available)

Abstract Kerala, the strip of land on the southwest coast of India, falls under the category of wet tropical climatic zone. Though traditional architecture and construction practices of Kerala were well known for the use of natural and passive methods for developing a comfortable thermal environment inside enclosed spaces, the modern construction methods and materials are not successful in creating a comfortable indoor environment, mostly due of the lack of understanding of how these elements interact with each other. This motivated the need of a scientific study for improving thermal comfort with minimal mechanical intervention. This study aimed to investigate the factors (such as temperature, humidity and air movement) that characterize thermal comfort in a particular house in Kerala and to identify and suggest some modifications (in architecture, construction materials etc.) to improve the thermal comfort inside the house. The study also attempts to develop some general guidelines for improving thermal comfort using passive cooling techniques in residences in wet tropical climate zones of India. ❧ A study on thermal comfort indices was first made in order to define the thermal comfort range appropriate for this climate (to mark in the psychrometric chart). As the next step, the temperature and humidity data was collected from a modern residence in Kerala for a year and this was used to understand the thermal conditions of the house. The building was simulated in the energy calculation software Design Builder, and was calibrated against the measured data. From a chosen list of passive cooling strategies, each one was applied to the base model by systematically varying some parameters (e.g., construction materials, type of roof etc.) that affect the thermal comfort while keeping others constant, and the changes in the thermal comfort indexes (e.g., interior temperature and humidity) are calculated. The effect of each strategy was evaluated from the fraction of time the building is in the comfort zone when the strategy was applied. The three most effective passive cooling strategies for the climate was found to be: 1) cross ventilating the space, 2) provision of openings near the ceiling, 3) different variations of providing shading to the building roof. The end result is a tabulated list of all passive cooling strategies tested, in the order of their effectiveness on the thermal comfort in the building.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Solar thermal cooling and heating: a year-round thermal comfort strategy using a hybrid solar absorption chiller and hydronic heating scheme

PDF

Airflow investigation of fabric membrane forms: a fluid dynamic analysis for thermal comfort

PDF

Air flow through louvered openings: effect of louver slats on air movement inside a space

PDF

Occupant-aware energy management: energy saving and comfort outcomes achievable through application of cooling setpoint adjustments

PDF

Tiny house in the desert: a study in indoor comfort using moveable insulation and thermal storage

PDF

Night flushing and thermal mass: maximizing natural ventilation for energy conservation through architectural features

PDF

Developing a data-driven model of overall thermal sensation based on the use of human physiological information in a built environment

PDF

A parametric study of the thermal performance of green roofs in different climates through energy modeling

PDF

A high-performance SuperWall: designed for a small residence at Joshua Tree National Park

PDF

Mitigating the urban heat island effect: thermal performance of shade-tree planting in downtown Los Angeles

PDF

Thermal performance of a precast roof assembly: achieving comfort using dynamic insulation and photovoltaics in an extreme climate designed for a small residence at Joshua Tree National Park

PDF

Net-zero cultural urbanism: the implementation of traditional and cultural net-zero urban housing design in Lagos

PDF

ctrl+z: exploring the effects of undoing retrofits to pre-war buildings in Los Angeles

PDF

Microclimate and building energy performance

Asset Metadata

Creator Padmanabhan Nayar, Priyanka (author)

Core Title Improving thermal comfort in residential spaces in the wet tropical climate zones of India using passive cooling techniques: a study using computational design methods

School School of Architecture

Degree Master of Building Science

Degree Program Building Science

Publication Date 08/02/2012

Defense Date 05/09/2012

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag effectiveness of passive cooling,hot humid climate,India,Kerala,OAI-PMH Harvest,passive cooling strategies,psychrometric chart,thermal comfort

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Woll, Edward (committee chair), La Roche, Pablo (committee member), Simmonds, Peter (committee member)

Creator Email priyanka.pnayar@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-85637

Unique identifier UC11289493

Identifier usctheses-c3-85637 (legacy record id)

Legacy Identifier etd-Padmanabha-1121.pdf

Dmrecord 85637

Document Type Thesis

Rights Padmanabhan Nayar, Priyanka

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA