Influence of Passive Design Parameters on Thermal Comfort

of an Office Space in a Building in Delhi

Sana Fatima Ali1; Lakshya Sharma2; Dibakar Rakshit3; and Bishwajit Bhattacharjee, M.ASCE4

Abstract: Energy efficiency measures in buildings, along with maintaining indoor thermal comfort, are increasingly drawing attention.
The most widely accepted indices, Predicted Mean Vote (PMV) and Percentage People Dissatisfied (PPD), have been used in this study
Downloaded from ascelibrary.org by University of Exeter on 05/06/20. Copyright ASCE. For personal use only; all rights reserved.

for evaluating thermal comfort of a small office room in Delhi. Various parameters influencing thermal comfort have been measured for
two months and the observations have been used for assessing the PMV and PPD, which have then been compared with the simulation results
from IDA ICE 4.7 Beta software, and found to be in agreement. Moreover, a comparison of the PMV values has been done with other indices
such as Tropical Summer Index (TSI) and Corrected Effective Temperature (CET). Parametric study for the office room has also been per-
formed to predict the shifts in the comfort levels by changing types and thicknesses of wall as well as roof insulations, provision of blinds, and
types of window glazing, using which, the best possible solution to enhance the thermal environment of the office room has been proposed.
DOI: 10.1061/(ASCE)AE.1943-5568.0000406. © 2020 American Society of Civil Engineers.
Author keywords: Predicted Mean Vote (PMV); Mean Radiant Temperature; Thermal comfort; Percentage People Dissatisfied (PPD);
IDA ICE Simulations; Passive Design Techniques.

Introduction conditioning systems, which is an imperative task of the HVAC de-

sign and energy engineers.
With increasing population, there is also a tremendous growth in
residential and commercial buildings. A study conducted by India’s
National Statistical Organization (NSO) in 2013 established that the Thermal Comfort
building sector utilizes approximately 40% of the electrical power. Thermal comfort is an equilibrium condition between humans and
Two major factors responsible for this are lighting and maintaining their surroundings, established when the generated heat by the me-
thermal comfort (air-conditioning) (CSE (Centre for Science and tabolism of humans is allowed to be dissipated into the environ-
Environment) 2014; Saxena et al. 2018; Sharma et al. 2018). ment. Several models for assessing thermal comfort have been
Energy efficiency measures in buildings, along with maintain- devised, giving the mathematical relationship between its values
ing indoor thermal comfort, increasingly draws attention. An and various factors affecting it, the most widely accepted of them
energy efficient building can be called effectual only when its being Fanger’s model of thermal comfort having Predicted Mean
occupants remain at thermal ease. If occupants of the building Vote (PMV) and the Percentage People Dissatisfied (PPD) indices,
are uncomfortable, this would not only be a risk to their health, which is also accepted as a standard for evaluating comfort levels
but also lead to a decrease in their productivity. Such buildings (ASHRAE 2010; ISO 2005). The PMV scale, ranging from −3.0
would use a major share of energy since the occupants would (cold) to 3.0 (hot), depends on six primary variables: metabolism
tend to opt for alternate methods of space conditioning to maintain (M), clothing insulation ( fcl), relative humidity (RH), temperature
a thermally acceptable environment. Thus, as far as possible, pas- (ta), velocity of air (var), and mean radiant temperature (tr). The
sive techniques should be implemented for lowering the energy standards (ASHRAE 2010; ISO 2005) recommend the allowable
consumption, along with sustaining the required thermal comfort. PMV range for thermal comfort to be −0.5 to +0.5 for an indoor
If the necessary requirement is not met even after this, then artificial space. Moreover, the PPD relates the PMV values with the satisfac-
methods need to be adopted such as heating, ventilation, and air- tion of the occupants. According to the standards, at least 80% of
the occupants need to feel thermally comfortable, the conditions
1 of which may vary from one individual to another (ASHRAE
Ph.D. Scholar, Centre for Energy Studies, Indian Institute of Technology
Delhi, Hauz Khas, New Delhi 110016, India. ORCID: https://orcid.org/0000
2010; ISO 2005).
-0003-0817-8041. Email: [email protected]
2
M. Tech. Alumnus, Centre for Energy Studies, Indian Institute of Tech-
Studies Based on Fanger’s Model
nology Delhi, Hauz Khas, New Delhi 110016, India. ORCID: https://orcid
.org/0000-0002-4090-8633. Email: [email protected] Various researches have been done on predicting the thermal com-
3
Associate Professor, Centre for Energy Studies, Indian Institute of Tech- fort of spaces using Fanger’s model-based PMV and PPD indices,
nology Delhi, Hauz Khas, New Delhi 110016, India (corresponding author). such as for a typical secondary school in Cyprus (Katafygiotou and
ORCID: https://orcid.org/0000-0002-8469-5924. Email: [email protected] Serghides 2014), traditional dwellings in Western Hunan (Liu et al.
4
Professor, Dept. of Civil Engineering, Indian Institute of Technology
2018), and fully air-conditioned and mixed-mode office buildings
Delhi, Hauz Khas, New Delhi 110016, India. Email: [email protected]
Note. This manuscript was submitted on April 23, 2019; approved on in Brazil (Forgiarini Rupp and Ghisi 2017), and for analyzing the
November 26, 2019; published online on April 22, 2020. Discussion period performance of air-conditioning systems in Iran (Pourshaghaghy
open until September 22, 2020; separate discussions must be submitted for and Omidvari 2012), Singapore (Azad et al. 2018), etc. Studies
individual papers. This paper is part of the Journal of Architectural Engi- have been carried out to find out a comfortable temperature range
neering, © ASCE, ISSN 1076-0431. for indoor spaces by making the use of Fanger’s comfort indices.

© ASCE 04020017-1 J. Archit. Eng.

J. Archit. Eng., 2020, 26(3): 04020017

 One such experimental as well as subjective study was for hospitals to analyze thermal comfort using the CFD approach, where 0.125
and shopping centers in Madagascar Island which provided com- was proposed as the optimum window-to-wall ratio (Ravikumar
fortable temperature range of 24.5°C–26.2°C (Nematchoua et al. and Prakash 2009). A CFD analysis was carried out to examine
2017). Another such study was held in a university classroom in an occupancy-based HVAC control system by evaluating the ther-
the Aegean region, which predicted the comfortable temperature mal comfort in terms of the PMV and PPD indices (Pazhoohesh
as 24.1°C for summer season and 26.6°C for winters, while as and Zhang 2018). While one study proposed that provision of par-
per the actual mean votes, the values were 23.2°C and 26.6°C re- tition walls improve the comfort conditions of a building, and
spectively. There was also a higher occupant satisfaction value ob- hence, should be promoted (Aryal and Leephakpreeda 2016).
served in the case of winters (93%) than in the summers (53%) These studies establish that the PMV index can be used as an
(Calis and Kuru 2017). A similar experiment as well as a survey effective tool for thermal comfort evaluation in a building and
study was conducted on predicting and comparing the thermal that individual effects of different passive techniques have also
comfort of the occupants for air-conditioned as well as naturally been studied in terms of the enhancement of comfort, however,
ventilated classrooms of a school in Japan, where it was found the scope of improvement in comfort could have been augmented
that the air-conditioned classrooms had a comfortable thermal en- by employing another passive technique in conjunction with
Downloaded from ascelibrary.org by University of Exeter on 05/06/20. Copyright ASCE. For personal use only; all rights reserved.

vironment, although it was inclined toward the slightly cool side the other.
of the thermal sensation scale, while the naturally ventilated class-
rooms were 3°C warmer than the air-conditioned classrooms, and
in this case too, it was observed that the PMV range lay between Studies on Alternative Thermal Comfort Models
−1 and 1 (Kwok and Chun 2003).
PMV has certain limitations as well, rendering it unsuitable to be
In addition, several researchers have carried out CFD analysis
applied everywhere, and thus, certain modifications have been
for envisaging the comfort conditions of buildings employing
made in the model to provide the best results in each fitting case.
Fanger’s model of thermal comfort, such as for a library in
In a study, the possible origins of the biasedness in the values of
Thailand, where it was found that the PMV values range from
−1.5 to 1 (Aryal and Leephakpreeda 2016). Another study carried PMV, both for air-conditioned and non air-conditioned rooms,
out the numerical study for an indoor stadium of the Athens 2004 based on its affecting parameters have been discussed, along with
Olympic Games, where different arrangement were provided for suggesting ways to modify it to reduce the difference between
the angle of the wall jets and the temperature of the incoming air, the PMV value and the actual mean vote (Humphreys and Fergus
and for which a 30° angle of wall jet and a stadium temperature Nicol 2002). A new design of the PMV model was developed in
of 16°C was proposed for the best thermal comfort conditions a research to compare the neutral temperatures with the real indoor
(Stamou et al. 2007). conditions, which included the indoor temperature and the indoor
relative humidity (Orosa and Oliveira 2011). An extension of the
PMV model was proposed in a study by introducing an expectancy
Studies on Energy Efficiency Measures and Their Impact factor, for non air-conditioned buildings in warm climates (Ole
on Thermal Comfort Fanger and Toftum 2002). Researchers have also used an adaptive
thermal comfort model that predicts the neutral operative tempera-
In order not only to benefit from maintaining the comfort, but also
ture based on the mean monthly outdoor temperature (Hwang et al.
to save energy, many investigations have been carried out to
2010). A review of literature on the PMV model and the adaptive
employ energy efficient measures in buildings and to analyze
thermal comfort model concluded that Fanger’s models cannot be
their impact on the thermal comfort. There has been a numerical
study on thermal comfort assessment in an enclosed courtyard, applied globally owing to different physiological factors of each in-
demonstrating the effect of wind and shading on the thermal com- dividual, and hence, an internationally acceptable model should be
fort, thus proposing a specific configuration and orientation of the developed which can accurately predict the thermal comfort of a
building, while also reporting that if a constant building tempera- larger number of people. However, since the perception of the com-
ture is assumed (25°C), adding galleries can be the most beneficial fort temperature is different for different individuals, it is also re-
(Berkovic et al. 2012). The PMV-PPD values have been estimated ported that a global model cannot be applied (Yau and Chew
for an office room in hot and humid climate for 14 windows–door 2014). An investigation on naturally ventilated buildings for differ-
opening arrangements, along with a survey study on the thermal ent climatic zones of north-east India concluded that a generalized
comfort sensation, which suggested that the occupants tend to comfort model is not viable for different regions, and hence, com-
open the windows when the outside wind velocity is high or the fort models based on the local parameters affecting the thermal en-
outside temperature is comparatively lower in order to feel satisfied vironment as perceived by people was developed considering
with their thermal environment (Daghigh et al. 2009). Natural ven- indoor and outdoor temperatures, clothing value, and relative hu-
tilation and building orientation were kept as the prime parameters midity (Singh et al. 2015). From the literature survey, it can be con-
for the study of their potential in achieving the optimum comfort- cluded that it is important to study the conditions in which the PMV
able environment in warm and humid climates in another research, model is valid and gives the best results, and also to provide a better
which demonstrated that by designing a climate responsive build- prospective to increase the thermal comfort along with the indoor
ing with proper natural ventilation, 9%–41% of improvements environment quality.
can be seen in terms of comfort for tropical regions, while it is The objective of this study is to analyze the thermal comfort
only 3%–14% in the case of subtropical climate (Haase and conditions in a building both by measurements and simulation-led
Amato 2009). Traditional dwellings of Western Hunan were ana- analysis, and to study the effects of various parameters on the
lyzed and optimized for the best thermal performance using thermal comfort. The PMV and PPD indices have been used as
10 mm polyurethane foam insulation layer for external wall, tools for assessing the indoor thermal comfort conditions, which
12 mm hollow layer ordinary insulating glass, 0.2 window-to-wall have also been compared with other comfort indices such as
ratio, and 80 mm reed foil for roof insulation, leading to energy TSI and CET. This is followed by a best possible solution prop-
savings as well (Liu et al. 2018). In another study, dimensions of osition for maintaining the thermal comfort of the occupants of
the openings of an office room in Tamil Nadu, India, were varied the office room.

© ASCE 04020017-2 J. Archit. Eng.

J. Archit. Eng., 2020, 26(3): 04020017

Downloaded from ascelibrary.org by University of Exeter on 05/06/20. Copyright ASCE. For personal use only; all rights reserved.

Fig. 1. Plan of the office room (metric units).

Methodology Table 1. Construction details of the office room

Specific
Experimental Setup Thermal heat
Thickness Density conductivity capacity
A small office room, located in New Delhi, has been selected for this Elements Layers (mm) (kg/m3) (W/m K) (J/kg K)
study. The design conditions include the time zone, latitude and lon- Walls Gypsum plaster 12 970 0.22 1,090
gitude, being GMT +5.5 h, N 28° 36′ and E 77° 12′ respectively. The Bricks 228.6 1,500 0.58 840
dimensions, in metric units, and plan of the room are shown in Fig. 1. Gypsum plaster 12 970 0.22 1,090
The Active Space Indoor Environmental Quality Sensor was
used for the measurements and data acquisition of desired parame- Floor Concrete slab 127 2,300 1.70 880
Gypsum plaster 12 970 0.22 1,090
ters, which was installed in December. The experiment lasted for a
Marble tiles 38.1 2,300 3.00 880
period of two months, from December to February, and the data
collected were filtered on the basis of the occupancy days of the of- Roof Gypsum plaster 12 970 0.22 1,090
fice, that is, all weekdays, and on the basis of the occupancy hours, Concrete slab 127 2,300 1.70 880
from 10:00 a.m. to 6:30 p.m. Cement mortar 100 1,648 0.72 920
slurry
Mud phuska 101.6 1,622 0.52 880
Measurements and Data Acquisition Brick tiles 38.1 1,892 0.79 880
The Active Space Indoor Environmental Quality Sensor is a WiFi Door Wood 38.1 2,300 0.14 500
enabled device, with industry sensors for examining the indoor en-
vironment of a space. The IEQ sensor was placed at the working
IDA Indoor Climate Energy, a building simulation software, is
desk of the occupant, as shown in Fig. 1, and the readings of tem-
established on a general system simulation platform with a modular
perature (ta) and velocity (va) of air, mean radiant temperature (tr),
system. The multidomain physical systems are described in the
and relative humidity (RH) were continuously measured and
logged into a workstation at intervals of every 1–2 min for the IDA using representative equations starting with a simulation
two months. The accuracy of the instrument in the measurements language Neutral Model Format (NMF–Neutral Model Format).
of the temperature, velocity, and relative humidity were ±0.3°C, To control the accuracy of the solution, the user defines tolerances,
±0.15 m/s, and ±2% RH respectively. However, the measurements thus allowing the isolation of numerical modeling approaches
from 10:00 a.m. to 6:30 p.m. on weekdays only (office hours) were (EQUA 2013).
taken into consideration for the analysis. To assess the thermal The building geometry with outer dimensions 7.62 × 6.1 m2 has
comfort conditions of the occupants, PMV and PPD values have been modeled, where two windows are located in the west direction
been obtained analytically for each of the measurements. and one window is located in the east direction. The room also has
two doors, where one opens in the north-west direction, while the
entrance lies in the north-eastern direction. Each surface of the
Simulation Study building has a thickness of 0.23 m, and the height of the room is
For analyzing the influence of various parameters on the office 3 m. Construction details and material properties of the room
room, a simulation study has been carried out in IDA ICE 4.7 and the properties of the single pane windows used (BIS 1987)
Beta software to model the room, as close to its actual working en- are listed in Tables 1 and 2. Figs. 2(a and b) show the north-western
vironment as possible. and north-eastern 3D views of the office room.

© ASCE 04020017-3 J. Archit. Eng.

J. Archit. Eng., 2020, 26(3): 04020017

 Since the office consisted of a large single room, so only one the RHS can be calculated by:
zone was created, for which, the number of occupants, that is,
seven, their schedule, along with the equipment and lighting sched- H = 3.96 × 10−8 × fcl × [(tcl + 273)4 − (tr + 273)4 ] − fcl hc (tcl − ta )
ule, that is, weekdays from 10:00 a.m. to 6:30 p.m., have been spec- (3)
ified as an input into the software. Eleven large PCs were present in
the office, while three tube-lights and four bulbs were present as Ec = 3.05 × [5.73 − 0.007 × (M − W ) − pa ] − 0.42[(M − W ) − 58.15]
equipment and lighting respectively. Several other system settings
need to be given as an input to the software, such as location setting (4)
and infiltration of the office room, along with the number and the
list of holidays. The same study period of December to February Cres = 0.0014 × M × (34 − ta ) (5)
has been provided as an input to the software to carry out the
simulations. Eres = 0.0173 × M × (5.87 − pa ) (6)
fcl can be obtained by:
Downloaded from ascelibrary.org by University of Exeter on 05/06/20. Copyright ASCE. For personal use only; all rights reserved.

Mathematical Model fcl = 1 + 1.29Icl ; If Icl ≤ 0.078 m2 K/W (7)

PMV fcl = 1.05 + 0.645Icl ; If Icl > 0.078 m2 K/W (8)

The equation given by P. Ole Fanger for finding the PMV value of a where, Icl can be found out from Annexure B of ASHRAE Standard
zone, based on the heat transfer balance equations, and adopted by 55 (ASHRAE 2010) for winter type of work clothing, from
standards (ASHRAE 2010; ISO 2005), has been used for obtaining where a constant value of 0.1953 m2 K/W has been taken into
the results as follows: consideration.
Here, hc and pa written in the previous equations can be attained
PMV = (0.303e−0.036M + 0.028) × L (1) from:
 √



hc = 2.38 × |tcl − ta |0.25 ; If (2.38 × |tcl − ta |0.25 ) > 12.1 × var
where, L = the thermal load on the body of an occupant, and can be (9)
found out from:
√


 √



hc = 12.1 × var ; If (2.38 × |tcl − ta |0.25 ) ≤ 12.1 × var (10)
L = (M − W ) − H − Ec − Cres − Eres (2)
RH
Here, W = 0 W/m2, whereas the value of M can be found out pa = 1000 × × ps (11)
100
from Annexure A of ASHRAE Standard 55 (ASHRAE 2010),
from where, a constant value of 63.8 W/m2 has been taken into  
consideration for normal office work. The rest of the terms on 16.6536 − 4030.183
ps = e ta +235 (12)
Table 2. Properties of windows of the office room var in the previous equations can be attained from:
 
Properties Values M
var = va + 0.005 − 58.15 (13)
Thickness (mm) 12 ADu × 58
Solar heat gain coefficient (SHGC) 0.68
Solar transmittance (τ) 0.60 Here, ADu is calculated from:
Visible transmittance (τvis) 0.74
Glazing U-value (W/m2 K) 1.90 ADu = (Ht.)0.725 × (Wt.)0.425 × 0.007184 (14)
Internal emissivity 0.84
External emissivity 0.84
The values of Ht. and Wt. have been expressed as 165 cm and
65 kg, taken as the average height and weight of a person.

(a) (b)

Fig. 2. 3D view of the office room: (a) north-west side; and (b) north-east side.

© ASCE 04020017-4 J. Archit. Eng.

J. Archit. Eng., 2020, 26(3): 04020017

 The surface temperature of clothing (°C), tcl, can be found out different passive strategies on the thermal comfort of the occupants
by iterative solution of the equation: through the software, the results provided by the software need to
be validated. Hence, the PMV and PPD indices received from the
tcl = tsk − Icl × 3.96 × 10−8 × fcl × [(tcl + 273)4 − (tr + 273)4 ]
actual conditions were compared with the results of the simulation
− Icl × fcl × hc × (tcl − tr ) (15) carried out for the base case.
The PMV values obtained analytically were observed to range
tsk = 35.7 − 0.028(M − W ) (16) from −1.21 to 1.08 and the PPD values ranged from 5% to
36.01%, which is not within the acceptable range given in the stan-
PPD dards (ASHRAE 2010; ISO 2005). Through the simulation study, it
The equation for finding PPD for each measurement depends upon was observed that the values of PMV for the study period range
the PMV values, and is shown as from −1.22 to 1.11, while the PPD values range from 5.00%
to 36.61%.
4
+0.2179 PMV2 )]
PPD = 100 − 95 × e[−(0.3353 PMV (17) There are three categories of thermal environment which is pre-
ferred by the occupants, as given in Table 3 (ASHRAE 2010; ISO
Downloaded from ascelibrary.org by University of Exeter on 05/06/20. Copyright ASCE. For personal use only; all rights reserved.

Validation of Base Case 2005). A graph between the frequency and the range of the PMV
as well as PPD values obtained through measurements have been
In order to successfully carry out a simulation study, it is imperative plotted, as shown in Fig. 3. Fig. 3(a) indicates that the maximum
to check whether the software provides the desired outcome or not. frequency of readings is between −0.5 and −0.2 PMV values,
Hence, to be able to carry out the analysis of the influence of the which indicate, though ideal, a slightly cooler environment.
While it can be observed that most of the values are falling in cat-
Table 3. Thermal environment categories egories A (−0.2 < PMV < =0.2), B (−0.5 < PMV < +0.5), and C
(−0.7 < PMV < 0.7) (ASHRAE 2010; ISO 2005), there are still in-
Category PMV PPD (%)
stances where the thermal environment lies even beyond the not
A −0.2 < PMV < 0.2 <6 ideal, but acceptable category, C. Similarly, Fig. 3(b) indicates
B −0.5 < PMV < 0.5 <10 that the frequency of PPD values lying below the acceptable
C −0.7 < PMV < 0.7 <15
range (ASHRAE 2010; ISO 2005) is significant, however, a

(a)

(b)

Fig. 3. Frequency plots for validation: (a) frequency of PMV index; and (b) frequency of PPD index.

© ASCE 04020017-5 J. Archit. Eng.

J. Archit. Eng., 2020, 26(3): 04020017

Downloaded from ascelibrary.org by University of Exeter on 05/06/20. Copyright ASCE. For personal use only; all rights reserved.

Fig. 4. Validation graph of simulation results with experimental results.

large percentage of people were dissatisfied by their thermal envi- Table 4. TSI values for various thermal sensations
ronment during the study period. Thermal sensation Range (°C) Optimum value (°C)
The results obtained through simulation too have been plotted in
Fig. 3, which show that the frequency of occurrence of the range of Slightly cool 19–25 22
Comfortable 25–30 27.5
PMV and PPD values is almost the same, with very minute differ-
Slightly warm 30–34 32
ences. This indicates that the simulation results were in close con-
formance to the actual measurements taken.
A graph between PMV and PPD obtained experimentally was conditions of northern India, shows the temperature of still air at
also compared with the graph obtained using simulation results, 50% relative humidity, which induces a similar thermal perception
shown in Fig. 4, which exhibited that the two curves matched as the provided environmental conditions. It considers the effects of
closely with each other. four environmental factors, that is, air temperature, humidity, air
In order to validate the simulation results, the percentage devia- motion, and radiation, on the thermal sensation of the occupants.
tion of the simulation results from the analytical values was deter- The ranges and values of these are given in Table 4, and their val-
mined using the root mean square percentage deviation for both ues can be obtained for each of the measurement from the follow-
PMV and PPD values. The percentage deviations were obtained ing (Sharma and Ali 1978):
using the following relations (Saxena et al. 2018):



















TSI = 0.308tw + 0.743tg − 2.06 va + 0.80 (20)
2
(ei )
e= (18) A normal distribution curve of the TSI values represented in
N
terms of PMV values has been plotted, along with the normal dis-
where tribution curve of the PMV values, shown in Fig. 5, to compare the
  two thermal comfort indices.
(Xi − Yi )
ei = × 100 (19) Since the comfortable range for TSI is 25°C–30°C, with the op-
Xi timum comfortable temperature being 27.5°C (Sharma and Ali
Using the above equation, the root mean square percentage de- 1978), while on the other hand, the ideal comfortable PMV range
viation obtained for the case of PMV values was 1.27, while for the is −0.5 to +0.5, with a neutral environment at 0 PMV value,
case of PPD values, the percentage deviation was found to be 1.07. hence the 3-sigma values of these means and ranges have been
This validated the simulation results with the actual environment, compared with each other to obtain TSI values in terms of PMV
rendering the software suitable to study the influence of various pa- values. Considering, 27.5°C as the mean comfortable temperature
rameters on the thermal comfort indices. for TSI, it would yield 25°C as −3-sigma and 30°C as +3-sigma.
Likewise, for PMV, 0 being the mean comfortable value, −0.5
would serve as −3-sigma limit and +0.5 would serve as
+3-sigma limit. Hence, for a particular TSI value obtained, corre-
Comparison with Other Indices
sponding PMV value can be obtained by comparing these values,
using the following:
Tropical Summer Index  
(27.5 − TSI) × (0 − (−0.5))
The Tropical Summer Index (TSI), developed at the Central Build- PMVTSI = − (21)
(27.5 − 25)
ing Research Institute, India, for hot and dry, and warm and humid

© ASCE 04020017-6 J. Archit. Eng.

J. Archit. Eng., 2020, 26(3): 04020017

 Corrected Effective Temperature Szokolay 2007):
Corrected Effective Temperature (CET) came from the concept of CET = (1.21tg − 0.21tw )/[1 + 0.029(tg − tw )] (22)
Effective Temperature (ET), developed by the American Society of
Heating and Ventilating Engineers (ASHVE), and is the tempera- For the purpose of comparison, a normal distribution curve has
ture of a still, saturated atmosphere, which would yield a similar ef- also been plotted for the PMV values and the CET values represented
fect in the absence of radiation as that produced by the given in terms of PMV values, as shown in Fig. 6. Similar to the procedure
environment. ASHVE then included the effects of air velocity in mentioned for the TSI values, the mean comfortable PMV value
the assessment of ET, while Vernon also considered the effect of along with its ±3-sigma values have been compared with the
heat from radiation. The values of the CET can be obtained for mean comfortable CET as 24.5°C, with the ±3-sigma values ranging
each of the measurement using the following (Auliciems and from 22°C to 27°C (Auliciems and Szokolay 2007). Hence, the CET
Downloaded from ascelibrary.org by University of Exeter on 05/06/20. Copyright ASCE. For personal use only; all rights reserved.

Fig. 5. Comparison graph of PMV and TSI results.

Fig. 6. Comparison graph of PMV and CET results.

© ASCE 04020017-7 J. Archit. Eng.

J. Archit. Eng., 2020, 26(3): 04020017

 values in terms of PMV values can be represented as graph between the PMV and CET in terms of PMV shows lesser
differences as compared to that with the TSI.
 
(24.5 − CET) × (0 − (−0.5))
PMVCET = − (23)
(24.5 − 22) Influence of Passive Parameters on Thermal Comfort
A simulation-led analysis, well validated with the experimental re-
Results and Discussion sults, has been carried out to observe the effect of different passive
technology parameters on the thermal comfort of the occupants.
Comparison with Other Indices For the purpose of this study, four parameters have been chosen,
that is, type and thickness of wall insulation, type and thickness
TSI of roof insulation, provision of blinds, and type of glazing.
The TSI values obtained lay in the range of 14.7°C–24.1°C,
suggesting that cooler conditions were persisting, which is in accor- Types and Thicknesses of Wall Insulation
dance with the fact that the experiment was performed during the
Downloaded from ascelibrary.org by University of Exeter on 05/06/20. Copyright ASCE. For personal use only; all rights reserved.

Five of the commonly used wall insulation materials have been

winter season. Fig. 5 illustrates a very small intersecting portion chosen for the analysis, that is, fiber glass, polyurethane foam, cel-
of the curves, while a larger portion of the TSI values lie in the re- lulose foam, mineral wool, and expanded polystyrene (EPS), with
gion of the thermal sensation of a cooler environment, which their properties listed in Table 5 (Greenspec 2016).
should be the case, since the study was done on a naturally condi- In the existing construction of the office exterior walls, a layer of
tioned building during the winter season. It shows a difference be- fiber glass with 20 mm thickness has been added to check its im-
tween the two curves which might be due to the fact that TSI takes pact on the variation in the comfort indices from the base case.
into account only environmental factors and the clothing insulation The results indicate that PMV values have reduced on both sides
of the occupant, as can be seen from Eq. (20), while PMV takes into of the thermal sensation scale, from −1.22 to −0.42 and from
account both environmental and physiological factors, since it also 1.11 to 0.99, as shown in Table 6, while the PPD value, though
includes the metabolic rate of the occupants. The difference is due has reduced from 36.61% to 25.90%, but is still depicting thermal
to the mean comfort perceived by the occupants, which might be discomfort to the occupants. This is due to the fact that the heat flow
because of the adaptation process and genetics of people, their be- is resisted by the insulating material.
havioral action, expectation and perception for their thermal envi- The thickness of the insulation has then been varied from 20 to
ronment, as well as their clothing and social condition. Since TSI 100 mm, with gap intervals of 20 mm each, to check their influence
was developed in Roorkee, the subjects belong to a tropical region on comfort levels. It was observed that with increasing thickness,
and are used to higher temperatures, hence, the temperature at minimum PMV value first increases, but after 40 mm of thickness,
which they would feel comfortable would be higher too. Thus, it starts decreasing, thus, moving away from the ideal thermal com-
for an index based on such subjects, for a comfortable range of fort range. A possible explanation for this could be that initially
PMV, their TSI should be higher. However, since the study was when thickness is increased, that thickness was still not sufficient
done in the winter months, the TSI came out low, making the ther- to resist the heat to flow from inside to the outside, thus increasing
mal environment uncomfortably cold for them and, hence, large the thermal sensation of the occupants toward a colder side. How-
differences in the two curves were obtained. ever, on increasing thickness beyond 40 mm, the heat flow could
have been adequately counteracted by the insulation, as depicted
CET by the decrease in the minimum PMV values toward a thermally
The CET values were found to be in the range of 18.2°C–27.3°C, neutral environment. On the other hand, it can be observed that
indicating that the environmental conditions ranged from a cooler the maximum PMV value constantly increases throughout the dif-
environment to a thermally neutral environment, while for most ferent thicknesses considered, hence, suggesting that as the thick-
of the time, cooler conditions persisted, agreeing with the experi- ness of the insulation is increased, it blocks the heat from the
mental conditions. A normal distribution curve of the CET values inside of the room to flow outside, resulting in maintaining a
represented in terms of PMV values was plotted, along with the warmer environment. It can also be observed that by the application
normal distribution curve of the PMV values, as shown in Fig. 6. of the fiber glass wall insulation, the maximum PMV value has re-
The figure establishes that a major portion of the curves intersect duced in all the cases, which could be because when the conditions
and match with each other, and indicates persistence of a cooler en- of a warmer ambient atmosphere persists, the insulation is able to
vironment. This conforms to the experimental conditions, where block the heat to enter inside the room. Likewise, when the thick-
the measurements were taken in the winter season for a naturally ness is increased, initially the PPD value reduced, suggesting that
conditioned building. It shows a difference between the two curves, the overall PMV range was closer to the ideal range for a comfort-
since similar to the case of TSI, CET also takes into account the able environment. However, when the thickness of the insulation is
environmental factors and the clothing insulation, as shown in increased beyond that, since the maximum PMV value is moving
Eq. (22). The difference in the two curves is due to the mean com-
fort sensed by the occupants. This could be because CET was de- Table 5. Thermal properties of insulation materials
veloped in the UK, while this study has been done for New Delhi,
India, hence the behavioral action, expectation, and perception of Fiber Polyurethane Cellulose Mineral Expanded
people for their thermal environment are location specific, and glass foam foam wool polystyrene
their adaptation process and genetics are different. The social con- Density (ρ) 48 30 57 40 15
dition and clothing of the occupants could also have an impact on (kg/m3)
this. Moreover, for the subjects who live in the temperate zones, Specific heat 843 1,800 2,020 840 1,300
slightly lower temperatures are comfortable for them since they (Cp) (J/kg-K)
are accustomed to such environments. Thus, for CET index, com- Thermal 0.04 0.026 0.035 0.038 0.038
conductivity
fortable PMV values can correspond to slightly lower temperatures.
(k) (W/m-K)
Consequently, for the study done during the winter season, the

© ASCE 04020017-8 J. Archit. Eng.

J. Archit. Eng., 2020, 26(3): 04020017

 away from the ideal range, it could have led to the increase in the as shown in Fig. 8. The shift of the lower limits of the PMV index
maximum value of PPD. Fig. 7 shows the PMV v/s PPD curve for from category C to category B is evident from the frequency plot,
each of the thickness considered for the same. as shown in Fig. 8(a), indicating that all the thicknesses are able to
A frequency plot of the PMV and PPD values has also been ob- block the heat flow through the room to the outside. It can also be
tained for the different thicknesses of the fiber glass wall insulation, seen that throughout the PMV range for category B, for each of the

Table 6. Thermal comfort indices with varying thicknesses of different wall insulations
Thickness

Insulation materials 20 mm 40 mm 60 mm 80 mm 100 mm

Fiber glass PMV −0.42–0.99 −0.45–0.97 −0.44–1.01 −0.42–1.03 −0.40–1.06
PPD (%) 25.90 24.82 26.43 27.53 28.53
Polyurethane foam PMV −0.47–0.95 −0.41–1.01 −0.39–1.05 −0.37–1.07 −0.36–1.09
Downloaded from ascelibrary.org by University of Exeter on 05/06/20. Copyright ASCE. For personal use only; all rights reserved.

PPD (%) 24.13 26.54 28.23 29.23 30.20

Cellulose foam PMV −0.48–0.93 −0.43–0.98 −0.41–1.01 −0.39–1.04 −0.37–1.06
PPD (%) 23.32 25.24 26.74 27.81 28.74
Mineral wool PMV −0.49–0.93 −0.46–0.97 −0.41–1.01 −0.39–1.04 −0.38–1.06
PPD (%) 23.29 25.13 26.67 27.87 28.79
Expanded polystyrene PMV −0.49–0.92 −0.44–0.98 −0.41–1.01 −0.39–1.04 −0.38–1.06
PPD (%) 23.23 25.19 26.72 27.92 28.76

(a) (b) (c)

(d) (e)

Fig. 7. PMV-PPD curves for different thicknesses of fiber glass wall insulation: (a) 0.02 m; (b) 0.04 m; (c) 0.06 m; (d) 0.08 m; and (e) 0.1 m.

© ASCE 04020017-9 J. Archit. Eng.

J. Archit. Eng., 2020, 26(3): 04020017

Downloaded from ascelibrary.org by University of Exeter on 05/06/20. Copyright ASCE. For personal use only; all rights reserved.

(a)

(b)

Fig. 8. Frequency plots for different thicknesses of fiber glass wall insulation: (a) frequency of PMV index; and (b) frequency of PPD index.

interval range considered, the frequency of PMV index is higher for of the plot has shifted from a cooler environment to a zone of
the fiber glass wall insulation of lower thickness, while beyond the warmer environment because of the restriction in the heat flow,
category B, the frequency of the higher thicknesses has increased. this has resulted in the occupants becoming dissatisfied with their
This again indicates that the higher thicknesses are able to suffi- thermal environment. It can be observed that till the PPD index
ciently resist the heat flow from the inside to outside, as compared lying in category B, there is a trend of having lower PPD values
to the material with lesser thickness. However, having a very warm for higher thicknesses of fiber glass wall insulation, which could
environment is also not acceptable to the occupants, and hence, be owing to the fact that the occupants were originally in a cooler
keeping too much thickness of the fiber glass insulation will not environment, and because of the obstruction to the heat flow by the
be very beneficial. For almost all of the thicknesses considered, material of higher thickness, the environment started to become
most of the PMV index frequency lies in category B, however, a slightly warm, thus providing satisfaction of the occupants with
large percentage of the PMV values have now shifted to the warmer their thermal environment. However, beyond a certain point,
side, beyond category C. dues to excessive heat trapped, the warmth inside the room started
Similarly, a frequency plot of the PPD index for different thick- to increase, thus making the occupants dissatisfied with the warmer
nesses has been obtained, as shown in Fig. 8(b). This plot too de- environment.
picts that maximum frequency of the PPD index lies in the category In a similar manner, analysis has been carried out for the remain-
B, however, a large percentage of the frequency also falls beyond ing four insulation materials, and their results are shown in Table 6,
the category C. The possible reason for this is that since the trend wherein similar trends were observed for each of the materials

© ASCE 04020017-10 J. Archit. Eng.

J. Archit. Eng., 2020, 26(3): 04020017

 with minute differences in the magnitudes of results in each case. It blocking the heat in the inside of the room from flowing outside, re-
was observed that the minimum PMV value constantly decreases sulting in the maintenance of a warmer environment. A similar var-
throughout the thicknesses considered, hence suggesting that the iation pattern can be seen in the case of minimum PMV values in the
wall insulation reduces the slight coldness in the thermal environ- context of a colder environment, as well as in the case of PPD values.
ment, and tends toward the ideal thermal comfort range, while the This implies that the thickness of the insulation material is a key fac-
maximum PMV constantly increases with increasing thickness, tor for maintaining a comfortable thermal environment in buildings,
thus, moving away from the ideal thermal comfort range. This sug- which is also suggested in the literature. Fig. 9 shows the PMV v/s
gests that as the thickness of the insulation is increased, it blocks PPD curves for each of the thicknesses considered for the same.
the heat flow from the inside of the room to outside, resulting in Similar to the case of fiber glass wall insulation, frequency plots
a warmer environment. Similarly, it was observed that the maximum of the PMV and PPD indices for the different thicknesses of fiber
PPD constantly increases with increasing thickness, indicating that a glass roof insulation are shown in Fig. 10. Fig. 10(a) suggests
greater percentage of people would start to feel discomfort if insula- that the environment zone has shifted from a negative PMV scale
tion with a larger thickness is provided. However, each case is still an to a positive PMV scale, with most of the frequency lying in cate-
improvement over the base case. Moreover, out of all the cases stud- gory B. There is a varying trend of frequency plots with the increas-
Downloaded from ascelibrary.org by University of Exeter on 05/06/20. Copyright ASCE. For personal use only; all rights reserved.

ied, 20 mm thick expanded polystyrene wall insulation yields a max- ing thickness, however, beyond the upper limit of category C, the
imum reduction in the PMV range, dropping it to −0.49 to 0.92 and frequency of the PMV values increases with increasing thickness,
the PPD reduction of up to 23.23%. indicating the presence of trapped heat. This plot also suggests
that mostly the thermal environment is of category B, however,
Types and Thicknesses of Roof Insulations more frequency of PMV is toward the positive side of the scale, in-
The same five materials mentioned previously have also been cho- dicating a warmer environment, while at some instances, the ther-
sen for the roof insulations to analyze their effects on the comfort mal environment was observed to be too warm, leading to an
levels of the occupants. Analogous to the method mentioned previ- increase in the frequency of the PMV beyond the category C, on
ously, first a layer of fiber glass roof insulation of 20 mm thickness the positive end of the scale.
has been included in the existing construction of the room. The re- Fig. 10(b) also indicates the persistence of the B-type thermal
sults imply that the PMV values have reduced on both the sides of environment for most of the time, be it on the cooler side of the
the thermal sensation scale, from −1.22 to −0.50 and from 1.11 to scale or the warmer side. It is to be noted, however, that the frequency
0.83, as shown in Table 7, while the PPD value has reduced from of the percentage of the people dissatisfied with their thermal environ-
36.61% to 19.51%. This implies that the heat flow to some extent is ment is still significant, extending beyond category C, due to the pres-
resisted by the insulating material, hence the range for PMV values ence of warmer conditions. Nonetheless, beyond category C, the
and the maximum PPD value have reduced. This proposes that al- frequency of the PPD index for each of the thickness considered
though the maximum PMV value does not lie in the range of the was still much less than the base case, suggesting that the insulation
ideal comfort levels, occupants are still satisfied with their thermal is being able to block the outside heat from entering the room.
environment. As the PMV values obtained are very close to the Simulations were also carried out for the remaining four insula-
ideal range with a little inclination toward the warmer side, the oc- tion materials, with varying thicknesses for the analysis, and their
cupants are still satisfied with their thermal environment since the results are shown in Table 7, where similar trends were observed
study has been conducted for the winter season, wherein cooler for each of the materials with minute differences in the magnitudes
conditions were prevailing. This can be observed from the fact of results in each case. With increasing thickness, it was observed
that the PPD value was just below the maximum allowable percent- that the maximum as well as the minimum PMV values were con-
age for the people feeling dissatisfied with their environment. stantly decreasing, thus, moving toward the ideal thermal comfort
For the analysis of this case as well, the same thicknesses of the range. This suggests that the provision of this insulation reduces
insulation have been considered. It was observed that with increas- the slight coldness as well as slight warmness in the thermal
ing thickness, the maximum PMV values obtained, though fluctu- environment, owing to the fact that, to some extent, the insulation
ating, lie well below the PMV value obtained in the base case, and confines the heat inside the room. Similarly, it was observed that
values corresponding to increased thicknesses, that is, 60–100 mm, the maximum PPD values showed either a slight decrease or re-
indicate that a zone of warmer environment has been created. This mained the same with increasing thickness, indicating that the per-
suggests that as the thickness of the insulation is increased, it is ception of the occupants for the thermal comfort was improving.

Table 7. Thermal comfort indices with varying thicknesses of different roof insulations
Thickness

Insulation materials 20 mm 40 mm 60 mm 80 mm 100 mm

Fiber glass PMV −0.50–0.83 −0.46–0.80 −0.64–1.04 −0.62–1.03 −0.61–1.03
PPD (%) 19.51 18.61 27.83 27.30 27.22
Polyurethane foam PMV −0.68–1.07 −0.64–1.04 −0.62–1.03 −0.60–1.02 −0.58–1.01
PPD (%) 29.25 27.92 27.26 27.69 26.38
Cellulose foam PMV −0.69–1.09 −0.63–1.08 −0.62–1.05 −0.60–1.01 −0.59–0.99
PPD (%) 30.13 29.94 28.47 26.56 25.94
Mineral wool PMV −0.68–1.09 −0.66–1.07 −0.64–1.03 −0.61–1.02 −0.59–1.02
PPD (%) 30.14 29.3 27.53 27.19 26.98
Expanded polystyrene PMV −0.69–1.09 −0.66–1.06 −0.64–1.04 −0.62–1.03 −0.61–1.03
PPD (%) 30.19 28.65 27.93 27.28 27.19

© ASCE 04020017-11 J. Archit. Eng.

J. Archit. Eng., 2020, 26(3): 04020017

Downloaded from ascelibrary.org by University of Exeter on 05/06/20. Copyright ASCE. For personal use only; all rights reserved.

(a) (b) (c)

(d) (e)

Fig. 9. PMV-PPD curves for different thicknesses of fiber glass roof insulation: (a) 0.02 m (b) 0.04 m; (c) 0.06 m; (d) 0.08 m; and (e) 0.1 m.

Moreover, it is to be noted that each case is still an improvement a similar behavior as the previous case, but shifting the thermal envi-
over the base case. Furthermore, out of all the cases studied, ronment of the room towards warmer side, as shown in Fig. 11(b).
40 mm thick fiber glass roof insulation yields a maximum reduction A possible explanation for this could be that the internal blinds emit
in the PMV range, dropping it to −0.46–0.80 and the PPD reduc- the absorbed radiation inside the room only, thus making the internal
tion of up to 18.61%. environment warmer. Similar to the previous case, the PPD value has
decreased here as well, to a value of 23.26%, however, this value
Provision of Blinds crosses the threshold provided by the standards (ASHRAE 2010;
Another passive design strategy employed was the provision of ex- ISO 2005). Moreover, out of all the cases studied, provision of exter-
ternal and internal blinds to all the windows to analyze their effects nal blinds yields a maximum reduction in the PMV range, dropping to
on the thermal comfort levels, the properties (EQUA 2013) of −0.77–0.61 and a PPD reduction of up to 17.55.%.
which have been listed in Table 8. The frequency plots of the PMV index for different blinds, as
The results for the case of external blinds, as shown in shown in Fig. 12(a), suggest that though there were improvements
Fig. 11(a), exhibit that the PMV values extend from −0.77 to in the thermal environment in both the cases as compared to the
0.61, as against the base case, where the range is −1.22–1.11, indi- base case, the external blind shows better enhancement with respect
cating an improvement in the thermal perception of the occupants to the internal blinds. The provision of external blinds maintained
with respect to their environment toward the warmer and the colder the environment in a slightly cooler zone while the internal blinds
sides. This is due to the fact that the blinds provide shading from the led to a zone of warmer environment, owing to the reasons men-
direct solar radiations and can help improve the thermal environment tioned above. The provision of both types of blinds enhanced the
of a building by controlling heat gain or loss. The results also indicate thermal environment, but were still beyond the desired thermal en-
an appreciable decrease in the percentage of people feeling discom- vironment categories.
fort, with a PPD reduction from 36.61% to 17.52%, which is even Plotting the frequencies of the PPD index for both types of
within the acceptable limits, as specified by the standards (ASHRAE blinds, shown in Fig. 12(b), depict that the highest PPD frequencies
2010; ISO 2005). were found in category C, while a large proportion of the values
Whereas in case of internal blinds, results show that PMV values extended beyond the acceptable category, indicating that the PPD
are extending from –0.51 to 0.93, as against the base case; indicating values could be at the borderline of crossing category C, and

© ASCE 04020017-12 J. Archit. Eng.

J. Archit. Eng., 2020, 26(3): 04020017

Downloaded from ascelibrary.org by University of Exeter on 05/06/20. Copyright ASCE. For personal use only; all rights reserved.

(a)

(b)

Fig. 10. Frequency plots for different thicknesses of fiber glass roof insulation: (a) frequency of PMV index; and (b) frequency of PPD index.

Table 8. Properties of blinds analysis, the properties of which have been listed in Table 9
(Vitro Architectural Glass 2015; EQUA 2013; WINDOWS 2016).
External Internal
blinds blinds
Results for the case of double pane-clear glass windows, as
shown in Fig. 13(a), exhibit that PMV values are stretching from
Multiplier for g (solar gain factor) 0.14 0.65 −0.61 to 0.85, indicating an enhancement in the comfort level of
Multiplier for t (short-wave shading 0.09 0.16 occupants. In addition, the percentage of dissatisfied people has
coefficient)
Multiplier for U-value 1.0 1.0
also reduced, though it barely reaches the allowable threshold
Diffusion factor 1.0 1.0 value. Subsequently, the results of remaining types of glazing
were attained, shown in Figs. 13(b–d), wherein similar trends
were witnessed. The PMV and PPD values were found to have re-
hence, most of the percentages of people dissatisfied with their ther- duced in each case, hence, shrinking the curve toward an ideal
mal environment are extended beyond the C category, owing to the range. Moreover, out of all the cases studied, low-e glass windows
performance of the two blinds. yield a maximum reduction in the PMV range, dropping to −0.64–
0.77 and a PPD reduction of up to 17.53.%.
Types of Window Glazing Fig. 14 represents the frequency plots of the PMV and PPD
Glazing can affect the amount of heat transfer through windows of indices for the different types of window glazing. Similar to the
the buildings, thus affecting the thermal comfort of the occupants. observations made in the case of different types of blinds, it can
Windows with double pane-clear glass, triple pane-clear glass, be noticed that the thermal environment of the room has enhanced,
low-e glazing, and tinted glass have been considered for the as shown in Fig. 14(a), but PMV values beyond the acceptable

© ASCE 04020017-13 J. Archit. Eng.

J. Archit. Eng., 2020, 26(3): 04020017

 (a) (b)

Fig. 11. PMV-PPD curves for different blind types: (a) external blinds; and (b) internal blinds.
Downloaded from ascelibrary.org by University of Exeter on 05/06/20. Copyright ASCE. For personal use only; all rights reserved.

(a)

(b)

Fig. 12. Frequency plots for different blind types: (a) frequency of PMV index; and (b) frequency of PPD index.

thermal environment category still have large frequencies, owing to Maximization of Thermal Comfort for Given Parameters
windows being directly exposed to the sun’s radiation, and thus
The thermal environment of an enclosed space can be affected by
affecting the heat transfer through the room and creating a zone
different passive techniques, therefore, a combination of parame-
of warmer environment. The highest improvement can be seen ters studied previously have been used to obtain a maximized sol-
in the case of low-e glass, which still lies in the category C of ution to bring the thermal comfort index of the office room within
the thermal environments. However, considering the frequency the acceptable range defined by the standards (ASHRAE 2010; ISO
plot of the PPD index for the different types of glazing, shown 2005). The objective function for the maximization was
in Fig. 14(b), low-e glass also represents a thermal environment
extending beyond category C, as is the case with other types of fx = maximize PMV (if PMV ≤ 0) and minimize PMV(if PMV > 0)
glazing.
(24)

© ASCE 04020017-14 J. Archit. Eng.

J. Archit. Eng., 2020, 26(3): 04020017

 Using a combination of all the parameters considered, simula- the sun’s movement, but that resulted in a decrease in comfort.
tions were run and the PMV values were obtained for the cases. This was then replaced with internal blinds, which resulted in an
The combination that conformed closest to the ideal range of the improved thermal environment. Glazing types were also varied
comfortable thermal environment would be the best solution with this combination, which finally led to the best possible solu-
amongst the parameters studied. Some of the combinations of the tion of improved thermal comfort conditions, within the parameters
parameters that were run for the simulation for maximization studied above, by providing windows with low-e glass. This led to
have been shown in Table 10 for reference. a decrease in the PMV values, on both the negative and the positive
The types of roof insulations studied showed a similar trend sides of the PMV scale, with respect to the base case. The PMV
to each other, hence, out of these, fiber glass roof insulation was range obtained using the maximized solution was from −0.89
chosen for initial simulation with 40 mm thickness as it showed to 0.95, thus tending toward the ideal thermal comfort range,
maximum improvement, which was then combined with other along with a decrease in the percentage of thermally dissatisfied
types of wall insulations, and finally its combination with fiber occupants. Although the decrease in the values on the positive
glass wall insulation with 40 mm thickness was selected. In combi- side of the PMV scale is less in comparison with the decrease
nation, scheduling of external blinds was also done according to on the negative side, out of all the possible combinations ex-
Downloaded from ascelibrary.org by University of Exeter on 05/06/20. Copyright ASCE. For personal use only; all rights reserved.

plored, this was the best possible solution for improving the ther-
mal comfort conditions within the office room. Fig. 15 shows the
Table 9. Properties of window glazing PMV-PPD curve of the base case combined with the maximized
solution. Frequency plots of the PMV and PPD indices of the
Double pane Triple pane
windows windows Low maximized solution as well as the base case were also obtained,
with clear with clear emissivity Tinted as shown in Fig. 16(a), where it can be realized that most of the
glass glass glass glass frequencies lie in category C of the thermal environment, how-
Solar heat gain 0.76 0.68 0.62 0.60 ever, small percentages of frequencies are seen extending beyond
coefficient (SHGC) the acceptable thermal environments. Nonetheless, the combination
Solar transmittance (τ) 0.70 0.60 0.58 0.47 of parameters shows an enhancement of the thermal environment as
Visible transmittance 0.81 0.74 0.67 0.50 against the base case. Fig. 16(b) also suggests that the majority of fre-
(τvis) quencies lie within category C of the thermal environment, with the
Glazing value, 2.9 1.9 3.8 5.7 highest percentage of frequency observed to lie in category B. Some
U (W/m2-K) percentage of dissatisfied occupants still exist due to the PMV range
Internal emissivity 0.84 0.84 0.1 0.5
of the maximized solution not being within the ideal comfortable
External emissivity 0.84 0.84 0.1 0.5
limits (ASHRAE 2010; ISO 2005).

(a) (b)

(c) (d)

Fig. 13. PMV-PPD curves for window glazing types: (a) double pane windows with clear glass; (b) triple pane windows with clear glass; (c) tinted
window glass; and (d) low-e window glass.

© ASCE 04020017-15 J. Archit. Eng.

J. Archit. Eng., 2020, 26(3): 04020017

Downloaded from ascelibrary.org by University of Exeter on 05/06/20. Copyright ASCE. For personal use only; all rights reserved.

(a)

(b)

Fig. 14. Frequency plots for window glazing types: (a) frequency of PMV index; and (b) frequency of PPD index.

Table 10. Sample maximization results

Min. Max. PPD
S. No. Parameter combination PMV PMV (%)
1. 20 mm cellulose foam wall insulation +20 mm cellulose foam roof insulation + external blinds + double Pane −0.74 1.04 30.48
clear glass
2. 20 mm cellulose foam wall insulation +40 mm cellulose foam roof insulation + external blinds + double pane −0.69 1.01 29.54
clear glass
3. 20 mm cellulose foam wall insulation +40 mm polyurethane foam roof insulation + external blinds + double pane −0.59 0.98 29.14
clear glass
4. 20 mm cellulose foam wall insulation +40 mm polyurethane foam roof insulation + external blinds + triple pane −0.72 0.99 29.37
clear glass
5. 20 mm cellulose foam wall insulation +40 mm cellulose foam roof insulation + internal blinds + triple pane clear −0.85 1.05 30.93
glass
6. 60 mm mineral wool wall insulation +40 mm cellulose foam roof insulation + external blinds + double pane clear −0.51 1.08 31.06
glass
7. 60 mm mineral wool wall insulation +40 mm expanded polystyrene insulation + internal blinds + tinted glass −0.59 0.96 27.82
8. 80 mm polyurethane foam wall insulation +40 mm fiber glass roof insulation + external blinds + low-e glass −0.89 0.76 20.05
9. 100 mm cellulose foam wall insulation +20 mm cellulose foam roof insulation + internal blinds + triple pane −0.41 1.13 34.91
clear glass
10. 40 mm fiber glass wall insulation +40 mm fiber glass roof insulation + internall blinds + low-e glass (maximized −0.89 0.95 25.39
solution)

© ASCE 04020017-16 J. Archit. Eng.

J. Archit. Eng., 2020, 26(3): 04020017

Downloaded from ascelibrary.org by University of Exeter on 05/06/20. Copyright ASCE. For personal use only; all rights reserved.

Fig. 15. Comparison graph of PMV and PPD between base case and maximized solution.

(a)

(b)

Fig. 16. Frequency plots for maximized case: (a) frequency of PMV index; and (b) frequency of PPD index.

© ASCE 04020017-17 J. Archit. Eng.

J. Archit. Eng., 2020, 26(3): 04020017

 Conclusions adaption behavior of an occupant and its impact on the thermal
comfort. Optimization of the comfort requirement can be carried
PMV and PPD indices, being the most widely accepted (ASHRAE out by considering active devices as well as the passive condition-
2010; ISO 2005), have been used for the present study of obtaining ing of the building. Apart from this, since the comfort perception
thermal comfort indices within an office room in Delhi, both by varies from one location to another, a country-specific study can
measurements and by using software. The effects of various param- also be carried out.
eters on thermal comfort have also been studied, and the best pos-
sible solution for maintaining the comfort of occupants of the office
room has been proposed. The following inferences can be drawn
from the present work: Data Availability Statement
• PMV within the office room lies in the range −1.22 to 1.11, sig-
nifying a slightly cold environment for most of the instances, Some or all data, models, or code that support the findings of this
while PPD values range from 5% to 36.61%, depicting that in study are available from the corresponding author upon reasonable
some instances more than 20% of the people are dissatisfied request (data, experimental and simulation result dataset).
Downloaded from ascelibrary.org by University of Exeter on 05/06/20. Copyright ASCE. For personal use only; all rights reserved.

with their thermal environment.

• Experimental results are in agreement with the simulation
results, with the root mean square percentage deviations Acknowledgments
of 1.27 and 1.07 for the PMV as well as PPD values,
respectively. The authors are thankful to Global Evolutionary Energy Design
• A comparison of the PMV results with the TSI as well as CET (GEED), New Delhi, India, where the experimental work for the
exhibited that in both the cases most of the values lay in the zone study was conducted. The authors are also thankful to Yardi Soft-
of a cooler environment, which is in accordance with the fact ware Pvt. Ltd. for providing the funding and IIT Delhi for provid-
that the experimental study was done on a naturally conditioned ing necessary facilities and support for this research.
building during the winter season.
• A difference between the curves of TSI and PMV, and CET and
PMV have been observed, which might be because of the adap-
tation process and genetics of people, the location specific be- Notation
havioral action, expectation, and perception of people toward
their thermal environment and the social condition and clothing The following symbols are used in this paper:
of the occupants. ADu = dubois area (m2);
• The influence of various passive parameters on comfort indices CET = corrected effective temperature (°C);
was also analyzed, which showed that those parameters either Cres = heat exchange by convection in breathing (W/m2);
improved the overall thermal environment very slightly, or im- Ec = heat exchange by evaporation on skin (W/m2);
proved the thermal environment on one side of the scale, while Eres = evaporative heat exchange in breathing (W/m2);
degrading it on the other side of the scale. e = root mean square percentage deviation;
• Out of all the parameters studied for the room considered, fcl = clothing surface area factor;
maximum improvement in thermal comfort has been observed H = sensitive heat losses (W/m2);
in the case when a fiber glass roof insulation of 40 mm thickness Ht. = height of a person (cm);
is used, along with fiber glass wall insulation thickness of hc = convective heat transfer coefficient (W/m2 K);
40 mm, internal blinds with scheduling according to the sun’s Icl = clothing insulation (m2 K/W);
movement, and low-e glass used in windows. L = thermal load on the body of an occupant (W/m2);
• This maximized solution resulted in the PMV range of −0.89– M = metabolic rate (W/m2);
0.95, with percentage improvements of 27.05% on the N = number of readings;
colder side and 14.41% on the warmer side respectively, PMV = predicted mean vote;
which indicates that the solution has resulted in producing a PMVTSI = TSI values in terms of PMV;
thermal environment very close to the ideal range of thermal PMVCET = CET values in terms of PMV;
comfort. PPD = percentage people dissatisfied;
• The PPD value also drastically reduced from 36.61% to 25.39%, pa = water vapor partial pressure (kPa);
leading to a percentage improvement of 30.65% in the values of ps = partial vapor pressure of saturated air (kPa);
PPD, indicating a larger percentage of people being satisfied RH = relative humidity;
with their thermal environment. TSI = tropical summer index (°C);
ta = air temperature (°C);
tcl = surface temperature of clothing (°C);
Scope for Future Research tg = globe temperature (°C);
tr = mean radiant temperature (°C);
In addition to the work presented, obtaining the subjects’ vote for tsk = external skin temperature (°C);
their perception of their thermal environment is vital. Hence, actual tW = wet bulb temperature (°C);
mean votes of the occupants can be taken by carrying out a survey va = velocity of air (m/s);
study along with the experimental measurement of data for the var = relative velocity of air (m/s);
evaluation of comfort in enclosed spaces. Moreover, it can be W = effective mechanical power (W/m2);
investigated whether the thermal comfort indices and models Wt. = weight of a person (kg);
being used to assess the thermal environment of the occupants Xi = values of parameters obtained through simulations;
are being validated or not with the actual results. Furthermore, and
the adaptive model of comfort can be looked into to study the Yi = values of parameters obtained analytically.

© ASCE 04020017-18 J. Archit. Eng.

J. Archit. Eng., 2020, 26(3): 04020017

 References Kwok, A. G., and C. Chun. 2003. “Thermal comfort in Japanese schools.”
Sol. Energy 74 (3): 245–252. https://doi.org/10.1016/S0038-092X(03)
Aryal, P., and T. Leephakpreeda. 2016. “Effects of partition on thermal 00147-6.
comfort, indoor air quality, energy consumption, and perception in air- Liu, S., C. Huang, Y. Liu, J. Shen, and Z. Li. 2018. “Retrofitting traditional
conditioned buildings.” J. Sol. Energy Eng. 138 (5): 051005. https://doi western hunan dwellings with passive strategies based on indoor ther-
.org/10.1115/1.4034072. mal environment.” J. Archit. Eng. 24 (3): 04018017. https://doi.org
ASHRAE (American Society of Heating, Refrigerating and /10.1061/(ASCE)AE.1943-5568.0000316.
Air-conditioning Engineers). 2010. Thermal environmental conditions Nematchoua, M. K., P. Ricciardi, S. Reiter, S. Asadi, and C. M. H. Demers.
for human occupancy. ANSI/ASHRAE 55-2010. Atlanta, GA: 2017. “Thermal comfort and comparison of some parameters coming
ASHRAE. from hospitals and shopping centers under natural ventilation: The
Auliciems, A., and S. V. Szokolay. 2007. Passive and low energy architec- case of Madagascar Island.” J. Build. Eng. 13: 196–206.. https://doi
ture international: Design tools and techniques—Note 3. Brisbane, .org/10.1016/j.jobe.2017.07.014.
Australia: Dept. of Architecture, Univ. of Queensland. Ole Fanger, P., and J. Tøftum. 2002. “Extension of the PMV model to
Azad, A. S., et al. 2018. “Evaluation of thermal comfort criteria of an active non-air-conditioned buildings in warm climates.” Energy Build.
chilled beam system in tropical climate: A comparative study.” Build. 34 (6): 533–536. https://doi.org/10.1016/S0378-7788(02)00003-8.
Downloaded from ascelibrary.org by University of Exeter on 05/06/20. Copyright ASCE. For personal use only; all rights reserved.

Environ. 145: 196–212. https://doi.org/10.1016/j.buildenv.2018.09 Orosa, J. A., and A. C. Oliveira. 2011. “A new thermal comfort approach
.025. comparing adaptive and PMV models.” Renewable Energy 36 (3): 951–
Berkovic, S., A. Yezioro, and A. Bitan. 2012. “Study of thermal comfort in 956. https://doi.org/10.1016/j.renene.2010.09.013.
courtyards in a hot arid climate.” Sol. Energy 86 (5): 1173–1186. https:// Pazhoohesh, M., and C. Zhang. 2018. “Investigating occupancy-driven air-
doi.org/10.1016/j.solener.2012.01.010. conditioning control based on thermal comfort level.” J. Archit. Eng.
BIS (Bureau of Indian Standards). 1987. Handbook on functional require- 24 (2): 04018003. https://doi.org/10.1061/(ASCE)AE.1943-5568
ments of a building (other than industrial buildings) (Parts 1–4). SP 41. .0000295.
New Delhi: BIS. Pourshaghaghy, A., and M. Omidvari. 2012. “Examination of thermal com-
Calis, G., and M. Kuru. 2017. “Assessing user thermal sensation in the fort in a hospital using PMV–PPD model.” Appl. Ergonomics 43 (6):
Aegean region against standards.” Sustainable Cities Soc. 29: 77–85. 1089–1095. https://doi.org/10.1016/j.apergo.2012.03.010.
https://doi.org/10.1016/j.scs.2016.11.013. Ravikumar, P., and D. Prakash. 2009. “Analysis of thermal comfort in an
CSE (Centre for Science and Environment). 2014. 02 energy and buildings. office room by varying the dimensions of the windows on adjacent
New Delhi: CSE. walls using CFD: A case study based on numerical simulation.”
Daghigh, R., N. M. Adam, and B. B. Sahari. 2009. “Thermal comfort of an Build. Simul. 2 (3): 187–196. https://doi.org/10.1007/s12273-009
air-conditioned office through different windows-door opening arrange- -9317-7.
ments.” Build. Serv. Eng. Res. Technol. 30 (1): 49–63. https://doi.org/10 Saxena, R., K. Biplab, and D. Rakshit. 2018. “Quantitative assessment of
.1177/0143624408099448. phase change material utilization for building cooling load abatement
EQUA. 2013. User Manual, IDA Indoor Climate and Energy, Version 4.6. in composite climatic condition.” J. Sol. Energy Eng. 140 (1):
Solna, Sweden: EQUA simulation AB.
011001. https://doi.org/10.1115/1.4038047.
Forgiarini Rupp, R., and E. Ghisi. 2017. “Predicting thermal comfort in of-
Sharma, L., S. F. Ali, and D. Rakshit. 2018. “Performance evaluation of a
fice buildings in a Brazilian temperate and humid climate.” Energy
top lighting light-pipe in buildings and estimating energy saving poten-
Build. 144: 152–166. https://doi.org/10.1016/j.enbuild.2017.03.039.
tial.” Energy Build. 179: 57–72. https://doi.org/10.1016/j.enbuild.2018
Greenspec. 2016. “Insulation materials and their thermal properties.”
.09.022.
Accessed February 9, 2017. http://www.greenspec.co.uk/building
Sharma, M. R., and S. Ali. 1978. “Evaluation of thermal stress—tropical
-design/insulation-materials-thermal-properties/.
summer index.” Build. Dig. 127: 1–5.
Haase, M., and A. Amato. 2009. “An investigation of the potential for nat-
Singh, M. K., S. Mahapatra, and J. Teller. 2015. “Development of thermal
ural ventilation and building orientation to achieve thermal comfort in
warm and humid climates.” Sol. Energy 86 (3): 389–399. https://doi comfort models for various climatic zones of North-East India.”
.org/10.1016/j.solener.2008.08.015. Sustainable Cities Soc. 14: 133–145. https://doi.org/10.1016/j.scs
Humphreys, M. A., and J. Fergus Nicol. 2002. “The validity of ISO-PMV .2014.08.011.
for predicting comfort votes in every-day thermal environments.” Stamou, A. I., I. Katsiris, and A. Schaelin. 2007. “Evaluation of thermal
Energy Build. 34 (6): 667–684. https://doi.org/10.1016/S0378 comfort in indoor stadiums of the Athens 2004 olympic games with
-7788(02)00018-X. CFD models: Case of nikea indoor stadium.” J. Archit. Eng. 13 (3):
Hwang, R.-L., T.-P. Lin, M.-J. Cheng, and J.-H. Lo. 2010. “Adaptive com- 130–135. https://doi.org/10.1061/(ASCE)1076-0431(2007)13:3(130).
fort model for tree-shaded outdoors in Taiwan.” Build. Environ. 48 (8): Vitro Architectural Glass. 2015. “How Low-e Glass Works.” Accessed
1873–1879. https://doi.org/10.1016/j.buildenv.2010.02.021. February 15, 2017. http://glassed.vitroglazings.com/topics/how-low-e
ISO (International Organization for Standardization). 2005. Ergonomics of -glass-works.
the thermal environment—analytical determination and interpretation WINDOWS. 2016. “Windows for High-performance Commercial
of thermal comfort using calculation of the PMV and PPD indices Buildings.” Accessed January 29, 2017. https://www
and local thermal comfort criteria. Switzerland: ISO. .commercialwindows.org/tints.php.
Katafygiotou, M. C., and D. K. Serghides. 2014. “Thermal comfort of a Yau, Y., and B. Chew. 2014. “A review on predicted mean vote and adap-
typical secondary school building in Cyprus.” Sustainable Cities Soc. tive thermal comfort models.” Build. Serv. Eng. Res. Technol. 35 (1):
13: 303–312. https://doi.org/10.1016/j.scs.2014.03.004. 23–35. https://doi.org/10.1177/0143624412465200.

© ASCE 04020017-19 J. Archit. Eng.

J. Archit. Eng., 2020, 26(3): 04020017