ID. 2622 - PLEA2014

Development of a Window Opening
Algorithm to Predict Occupant Behavior
in Japanese Houses
Hom B. Rijal, PhD
Michael A. Humphreys
J. Fergus Nicol
Tokyo City University
[email protected]
Oxford Brookes University
Oxford Brookes University
ABSTRACT
We investigated window opening behaviour and the thermal environment over a period of more than 3
years in the living rooms and bedrooms of dwellings in the Kanto region of Japan. We collected over
32,000 data-samples from 243 residents of 121 homes. The proportion of ‘open window’ in the free
running mode is significantly higher than that in the cooling and heating modes. The window opening is
related to the indoor or outdoor air temperature. Window opening behaviour as predicted by logistic
regression analysis is in agreement with the measured data. These findings can be applied to develop an
adaptive algorithm for window opening behaviour in Japanese residences.
INTRODUCTION
Natural ventilation from opening windows has been decreasing in houses in recent years because of
the increasing prevalence of mechanical ventilation and air-conditioning. However, temperature control
by opening and closing windows can reduce environmental impact by minimizing the period of the year
when air-conditioning is needed.
There has been more research into window-opening behaviour in offices (Rijal et al. 2007~2012,
Yun & Steemers 2007, Robinson & Haldi 2008, Kim et al. 2009, Haldi & Robinson 2010) and university
buildings (Suzuki et al. 2002, Umemiya & Yoshida 2004) than in dwellings (Dick & Thomas 1951,
Asawa et al. 2005, Kubota 2007, Rijal et al. 2013). The findings from research in offices and universities
cannot be assumed to apply to dwellings, where people’s behaviour is less constrained. There is
evidence that people respond differently in their own homes for a number of reasons, social, economic
and cultural (Oseland, 1995). Thus it was necessary to conduct research also on residential window
opening behaviour.
To explore window opening behavior and develop a window opening algorithm for Japanese
residences, thermal measurements were made and an occupant behavior surveys conducted over a period
of more than 3 years in the living rooms and bedrooms of dwellings in the Kanto region of Japan.
METHODOLOGY
Thermal comfort surveys and thermal measurements were conducted in 121 houses in Kanto region
(Kanagawa, Tokyo, Saitama and Chiba) of Japan from 2010 to 2013 (Table 1). The detail of surveys 1, 2
A Hom B. Rijal is an associate professor in the Department of Restoration Ecology & Built Environment, Tokyo City University, Japan.
Michael A. Humphreys is an associate of Oxford Brookes University and J. Fergus Nicol is emeritus professor at Department of
Architecture, Oxford Brookes University, UK..
30th INTERNATIONAL PLEA CONFERENCE
16-18 December 2014, CEPT University, Ahmedabad
1
and 4 can be found in Rijal & Yoshimura (2011), Katsuno et al. (2012) and Rijal et al. (2014)
respectively.
Indoor air temperature and relative humidity were measured in the living rooms and bedrooms,
away from direct sunlight, at ten minute intervals using a data logger (Fig. 1). The globe temperature
was also measured in the living room in surveys 3, 4 & 5. Outdoor air temperature and relative humidity
were obtained from the nearest meteorological station.
The number of subjects was 119 males and 124 females. Respondents completed the questionnaire
several times a day in the living rooms and twice in the bedrooms (“before go to bed” and “after wakeup from the bed”) (Table 2). The thermal comfort survey was conducted several times a day using sevenpoint thermal sensation scales (Table 2). The window opening behaviour was recorded in binary form (0
= window closed, 1 = window open). We have collected over 32,000 samples.
Table 1. Description of survey
Survey period
Survey
Surveyed room
Measured
variables*
Number of subjects
Number of
houses
Male
Female
Number of votes
Total
Living room
Bedroom
1
Start date
06-7-2010
End date
18-7-2011
Living, Bed
Ti, RHi
11
16
14
30
3299
2558
2
05-8-2011
06-9-2011
Living
Ti, RHi
55
52
57
109
2819
-
3
21-7-2011
08-5-2012
Living, Bed
Ti, RHi , Tg
14
11
12
23
463
984
4
25-7-2012
24-6-2013
Living, Bed
Ti, RHi , Tg
30
26
28
54
13083
7061
5
10-8-2013
03-10-2013
Living, Bed
Ti, RHi , Tg
11
14
13
27
936
1265
Ti: Indoor air temperature (°C), RHi: Indoor relative humidity (%), Tg: Indoor globe temperature (°C), *: Tg is measured only in the living
room.
Table 2. Thermal sensation scale
No.
Now, how do you feel the air temperature?
1
Very cold
2
Cold
3
Slightly cold
4
Neutral (neither cold nor hot)
5
Slightly hot
6
Hot
7
Very hot
Figure 1 Details of the thermal measurement
RESULTS AND DISCUSSION
3.1 Distribution of indoor and outdoor temperatures during voting
Table 3 shows the mean and standard deviation of the indoor and outdoor air temperature in each
mode. Fig. 2 shows the monthly mean
outdoor and indoor air temperature in
FR mode in living room and bedroom.
The mean outdoor air temperatures
during the voting were 19.5 °C,
27.6 °C and 7.2 °C for FR, CL and HT
modes respectively (Fig. 2). The mean
indoor air temperatures at the time of
voting were 24.2 °C, 27.3 °C and
19.2 °C for FR, CL and HT modes
respectively.
The
Japanese
government recommends the indoor Figure 2 Monthly mean outdoor and indoor air temperature in
temperature settings of 20 °C in winter FR mode.
30th INTERNATIONAL PLEA CONFERENCE
16-18 December 2014, CEPT University, Ahmedabad
2
and 28 °C in summer respectively. The results showed that the mean indoor temperatures during heating
and cooling were close to the recommendation. The mean indoor and outdoor temperature difference
was 4.7 K, -0.3 K and 12.0 K for FR, CL and HT modes respectively. The results show that the seasonal
difference of the indoor air temperature is quite large, and that the data represent a wide range of outdoor
temperature.
Table 3 Indoor air temperature and proportion of open windows in various modes
Outdoor air temp. (°C)
Indoor air temp. (°C)
Window opening
Mode
Room
N
Mean
SD
13,454
20.4
7.7
Living
9,054
18.2
8.0
FR
Bed
All
22,508
19.5
7.9
CL
All
6,677
27.6
2.7
HT
All
2,982
7.2
4.2
N: Number of observation, SD: Standard deviation
N
13,361
Mean
24.8
SD
4.5
N
13,452
Mean
0.46
SD
0.50
9,000
22,361
6,407
2,960
23.3
24.2
27.3
19.2
5.6
5.0
2.0
2.9
9,110
22,562
6,648
3,050
0.30
0.39
0.03
0.00
0.46
0.49
0.16
0.03
3.2 Evaluation of window opening behaviour
3.2.1 Status of window opening (WO)
To understand the window opening behaviour, the mean proportions of ‘window opening (WO)’
are compared. Table 3 shows the mean and standard deviation of the windows open in each mode. The
mean WO is 0.39, 0.03 and 0.00 for FR, CL and HT modes respectively. The mean window opening in
living room is higher than in the bedroom (Table 3). Interestingly, the mean WO in UK office buildings
was 0.70 in NV mode and 0.04 in AC mode (Rijal et al. 2007). The mean window opening in Pakistan
office and commercial buildings was 0.33 in NV mode. The results showed that the mean windows open
is close to the Pakistan value and lower than the UK value. We shall limit the analysis to the FR mode.
Season, month and time of the day
Seasonal and monthly difference in proportion of windows open in FR mode is shown in Fig. 3.
The proportion of open windows (WO) is highest in summer and lowest in winter. The WO in autumn is
significantly higher than that in spring. This is possibly due to the fact that people are more adapted in
spring to the winter low temperature, and in autumn to the summer temperature. In reality, the indoor
and outdoor air temperatures in autumn are higher than in the spring (Fig. 3(b)).
Evidently, the proportion of open windows gradually increases towards the summer months (Fig.
3(c)). Conversely, it gently decreases towards the winter months as indoor or outdoor air temperature
varies (Figs. 2).
The data were divided into four groups, in ascending order of time. Interestingly, the proportion of
open windows gradually increases during the morning, and then decreases towards the evening (Fig.
4(a)). Most of occupants open the windows in the morning and shut them at night. These trends are
similar for all seasons (Fig. 4(b)).
Figure 3 The proportion of open windows, indoor and outdoor air temperature (at 95% confidence level)
30th INTERNATIONAL PLEA CONFERENCE
16-18 December 2014, CEPT University, Ahmedabad
3
Figure 4 Proportion of open windows at 95% confidence intervals of time of day in living room.
3.2.3 Relationship between the open windows and air temperature
In FR mode the open window correlated better with the outdoor temperature than with the indoor
temperature (Table 4). The correlation coefficient for the living room is higher than for the bedroom.
From these observations, it can be inferred that the window opening is related to both indoor and outdoor
air temperatures.
Fig. 5 shows the proportion of open windows and the corresponding temperatures. The data were
divided into ten groups, in an ascending order of temperature. The proportion of the window opening
rises as the indoor globe or outdoor air temperature rises. The proportion of window opening in the
livingrooms is higer than in the bedrooms. When mean indoor air temperature is 27.1 °C, the proportion
of open windows is 0.63 in living room and 0.51 in bedroom (Fig. 5(a)).
When the mean outdoor air temperature is 24.3 °C, the proportion of windows open is 0.71 in
livingrooms and 0.58 in the bedrooms (Fig. 5(c)). These proportions are similar to the Pakistan study
(Rijal et al. 2008), and significantly lower than that of the UK study (Rijal et al. 2007). This is perhaps
because the indoor and outdoor air temperature in Japan and Pakistan are condiderably higher than that
in the UK.
Table 4. Correlation coefficients in FR mode
Room
Liivingroom
Bedroom
All
Items
Window：Ti
Window：To
Ti: To
0.58
0.62
0.87
13,289
13,382
13,352
Correlation coefficient (r)
0.46
0.50
0.89
Number of samples (N)
8,946
9,000
8,997
Correlation coefficient (r)
0.53
0.58
0.88
22,235
22,382
22,349
Correlation coefficient (r)
Number of samples (N)
Number of samples (N)
Ti： Indoor air temperature （°C）, To： Outdoor air temperature （°C）All correlations are significant (p<0.001)
Figure 5 Proportion of open windows with 95% confidence intervals at deciles of temperatures.
30th INTERNATIONAL PLEA CONFERENCE
16-18 December 2014, CEPT University, Ahmedabad
4
3.3 Potential of the open window
3.3.1 Indoor air temperature
Fig. 6 and Table 5 show the seasonal variation in indoor
air temperature for cases when windows are open and closed.
The mean indoor air temperature for the window open
condition is 27.7 °C in the livingroom which is significantly
higher by 5.3 K, than for the window closed condition. In UK
office buildings, the mean globe temperature for the window
open condition is 23.4 °C which is 1.2 K higher than when
the window is closed (Rijal et al. 2008a). Thus, the
temperature difference between the cases of open and closed Figure 6 Seasonal variation of indoor
window in residential buildings is higher than that of the air temperature for windows open and
office buildings. The temperature difference is highest in closed in FR mode.
autumn. In winter, the mean indoor air temperature for the
‘open window’ case is significantly lower than that of the ‘closed window’ case. The results showed that
window opening is an effective way to control the indoor thermal environment.
Table 5. Indoor air temperature for windows open and closed
Indoor air temperature Ti (°C)
Season
Window
Closed
Living room
N
Mean
SD
1,247
18.0
3.2
Winter
Open
Closed
55
13.1
4.5
2,025
21.8
2.6
493
24.2
2.2
1,038
27.5
2.4
Spring
Open
Closed
Summer
Open
3,685
29.0
2.2
Closed
2,926
22.9
3.3
Autumn
Open
1,820
26.5
3.0
Closed
7,236
22.4
4.0
All
Open
6,053
27.7
3.3
Bedroom
t*
Open-Closed
10.8
-4.9
19.2
2.5
19.0
1.5
37.1
3.5
82.1
5.3
N
Mean
SD
1,261
15.2
4.3
54
12.5
4.7
1,631
20.5
3.6
165
23.0
2.6
1,182
27.7
2.3
1,675
28.5
2.1
2,198
22.9
3.8
780
26.3
3.3
6,272
21.6
5.4
2,674
27.2
3.7
t*
Open-Closed
4.4
-2.6
-8.9
2.6
10.0
0.8
22.3
3.4
48.8
5.6
*All open/closed temperature differences are statistically significant (p<0.001)
3.3.2 Comfort temperature
The potential of the open window is further analyzed in the context of comfort temperature. The
comfort temperatures were obtained by the Griffiths’
method (Griffiths 1990, Nicol et al. 1994, Rijal et al.
2008, Humphreys et al. 2013, Rijal et al. 2014).
Tc = Ti + (4 - C) / a*
(1)
Tc is the comfort temperature by Griffiths’ method
(°C), Ti is the indoor air temperature (°C) and a* is
the regression coefficient (=0.50).
Fig. 7 and Table 6 show the seasonal variation in
comfort temperature with windows open and closed.
The mean comfort temperature for window open is
26.5 °C in living room which is 3.7 K higher than that
of the case of window closed. Brager et al. (2004) Figure 7 Seasonal variation of comfort
found 1.5 K higher comfort temperature for the people temperature for windows open and closed in
with an access to window operation than the group FR mode.
30th INTERNATIONAL PLEA CONFERENCE
16-18 December 2014, CEPT University, Ahmedabad
5
without in office buildings. The temperature difference is highest in autumn. In winter the mean comfort
temperature for the open window condition is significantly lower than for the window closed condition.
The results showed that window opening is effective to create the comfortable thermal environment.
Table 6. Comfort temperature for windows open and closed
Comfort temperature Tc (°C)
Season
Living room
Window
Closed
N
Mean
SD
1,247
19.5
2.6
t-value*
Open-Closed
14.6
-5.4
Winter
Open
Closed
55
14.1
4.6
2,025
22.1
2.3
492
23.8
2.2
1,038
26.5
2.1
Spring
-14.9
Open
Summe
r
Bedroom
Closed
-10.8
Open
3,683
27.3
2.1
Closed
2,926
23.5
2.8
Autumn
-31.0
Open
1,819
26.0
2.6
Closed
7,236
22.9
3.3
All
-69.1
Open
6,049
26.5
2.8
N
Mean
SD
1,260
17.1
3.8
54
13.3
4.6
1,631
21.0
3.3
165
23.2
2.6
1,179
26.5
2.2
1,675
27.3
2.0
2,196
23.3
3.3
779
25.9
2.9
6,266
22.1
4.5
2,673
26.3
3.2
1.7
0.8
2.6
3.7
t-value*
Open-Closed
7.1
-3.8
-8.1
2.2
-9.5
0.7
-19.2
2.6
-44.6
4.3
*All open/closed temperature differences are statistically significant (p<0.001)
3.4 Development of an algorithm to predict window opening behaviour
3.4.1 Logistic regression curves
In the previous section, we analyzed the window opening behaviour based on field data and
confirmed some general behavioural trends, but no attempt was made to predict the occupant behaviour
in housing (Rijal et al. 2013). Such predictions are needed for the thermal simulation of buildings.
Nicol and Humphreys (2004) made use of Probit analysis to predict occupant control behaviour in
NV buildings. For mathematical convenience they used a Logistic distribution in place of the Normal
distribution. The relationship between the probability of windows open (p) and the indoor or outdoor
temperature (T) is of the form:
logit(p) = log {p/(1–p)} = bT + c
(2)
(bT+c)
(bT+c)
p = exp
/{1+exp
}
(3)
and where exp (exponential function) is the base of natural logarithm, b is the regression coefficient
for T, and c the constant in the regression equation.
We have adopted the same method here, using SPSS version 19 for the calculations. The Logistic
regression equations, based on the indoor or outdoor temperature, are shown in Fig. 8. The following
regression equations were obtained in between the windows open and the indoor or outdoor air
temperature:
Living room
logit(p)=0.394Ti‐10.144（n= 13,289, R2=0.34, S.E.=0.007, p<0.001）
(4)
logit(p)=0.372Tg‐9.659（n=9,833, R2=0.29, S.E.=0.008, p<0.001）
2
logit(p)=0.258To‐5.675（n=13,382, R =0.38, S.E.=0.004, p<0.001）
Bedroom
logit(p)=0.291Ti‐8.100（n=8,946, R2=0.24, S.E.=0.008, p<0.001）
2
logit(p)=0.206To‐5.113（n=9,000, R =0.26, S.E.=0.005, p<0.001）
All data
logit(p)=0.349Ti‐9.235（n=22,235, R2=0.30, S.E.=0.005, p<0.001）
(5)
(6)
(7)
(8)
(9)
2
logit(p)=0.238To‐5.466（n=22,382, R =0.34, S.E.=0.003, p<0.001）
(10)
Ti: Indoor air temperature (°C), Tg: Globe temperature (°C), To: Outdoor air temperature (°C), n:
30th INTERNATIONAL PLEA CONFERENCE
16-18 December 2014, CEPT University, Ahmedabad
6
sample size, S.E.: Standard error, p: Significance level of the regression coefficient, R 2: Cox and Snell
R2.
A regression coefficient of 0.349 is obtained when the indoor air temperature is the predictor. This
is higher than that obtained when the outdoor air temperature is used. In the Gifu region of Japan (Rijal
et al. 2013), regression coefficients of 0.248 and 0.210 respectively were obtained with indoor or
outdoor temperature. In Pakistan (Rijal et al. 2008) and in UK (Rijal et al. 2007) studies, regression
coefficients of 0.176 and 0.354 respectively were obtained with indoor globe temperature is the predictor.
In Kyoto (Majima et al. 2007) and UK (Rijal et al. 2007) data returned the regression coefficients of
0.119 and 0.181 respectively with outdoor air temperature is the predictor. The regression coefficient in
the living room is slightly higer than the bedroom. The predicted window opening is well matched with
measured values (Fig. 8).
Living room
Bedroom
All
All
0.8
0.6
0.4
0.2
0.0
0
Globe temp.
Indoor air temp.
0.8
0.6
Living room
Bedroom
All
All
0.8
0.6
0.4
0.4
0.2
0.2
0.0
5 10 15 20 25 30 35
Indoor air temperature (C)
(c) Outdoor air temp.
1.0
Proportion of windows open
(b) Temperature (Living room)
1.0
Proportion of windows open
Proportion of windows open
(a) Indoor air temp.
1.0
0.0
0
5
10 15 20 25 30 35
Temperature (C)
0
5 10 15 20 25 30 35
Outdoor air temperature (oC)
Figure 8 Comparison of measured (open circular dots) and predicted value (curved line) in NV mode.
Measured values were grouped for every 1 °C for indoor air temperature and for every 2 °C for outdoor
air temperature. The grouped data for samples less than 100 are not shown.
CONCLUSIONS
We have investigated the window opening behaviour and corresponding thermal environment over
a period of more than 3 years in the living rooms and bedrooms of dwellings in the Kanto region of
Japan and the following results were found:
1. The proportion of the window opening in the free running mode is significantly higher than that
of the cooling or heating modes.
2. The window opening is related to the indoor and outdoor air temperature in the free running
mode.
3. The window opening behaviour is predicted based on indoor and outdoor air temperature using
logistic regression analysis. The predicted window opening matched well with that of the
measured value.
ACKNOWLEDGEMENTS
We would like to thanks to all people who participated in the survey, to Kawamoto Industries, Ltd,
Japan for their cooperation and to all students for data entry. This research was supported by Grant-inAid for Scientific Research (C) Number 24560726 and (B) Number 25289200.
REFERENCES
Asawa, T., Umehoshi, N., Takezawa, T. and T. Shimizu. 2005. Analysis of the behavioral characteristics
of both window opening and air conditioning use at detached houses: Relationship between outdoor
microclimate and residents’ living open to outdoor Part 2. J. Environ., Eng., AIJ, 593: 87-94.
Brager, G.S., Paliaga, G. and R. de Dear. 2004. Operable windows, personal control, and occupant
comfort. ASHRAE Transactions, 110(2): 17-33.
Dick, J.B. and D.A. Thomas. 1951. Ventilating research in occupied houses. JIHVE 19: 306-326.
Griffiths, I.D. (1990). Thermal comfort in buildings with passive solar features: Field studies. Report to
the Commission of the European Communities, EN3S-090, UK.
30th INTERNATIONAL PLEA CONFERENCE
16-18 December 2014, CEPT University, Ahmedabad
7
Haldi, F. and D. Robinson. 2010. On the unification of thermal perception and adaptive actions. Building
and Environment, 45(11): 2440-2457.
Humphreys, M.A., Rijal, H.B. and J.F. Nicol. 2013. Updating the adaptive relation between climate and
comfort indoors; new insights and an extended database. Building and Environment, 63(5): 40-55.
Katsuno, J., Rijal, H.B. and S. Kikuchi. 2012. Investigation of the comfort temperature and adaptive
model in Japanese houses in summer. Proceedings of 7th Conference: The changing context of
comfort in an unpredictable world, Windsor, UK, 12-15 April 2012. London: Network for Comfort
and Energy Use in Buildings.
Kim, J., Kawaguchi, T. and S. Tanabe. 2009. The behavioral characteristic of office workers use of
natural ventilation window. J. Environ. Eng., AIJ, 74(643): 1075-1082.
Kubota, T. 2007. A field survey on usage of air-conditioners and windows in apartment houses in Johor
Bahru city. J. Environ. Eng., AIJ, No. 616: 83-89.
Majima, M., Umemiya, N., Yoshida, H. and H.B. Rijal. 2007. Thermal comfort of traditional narrow
alleys in an urban area. J. Environ., Eng., AIJ, 622: 41-48.
Nicol, F., Jamy, G.N., Sykes, O., Humphreys, M., Roaf, S. and M. Hancock. 1994. A survey of thermal
comfort in Pakistan toward new indoor temperature standards. Oxford Brookes University, Oxford
England.
Nicol, J.F. and M.A. Humphreys. 2004. A stochastic approach to thermal comfort – occupant behavior
and energy use in buildings. ASHRAE Transactions, 110(2): 554-568.
Oseland, N. 1995. Predicted and reported thermal sensation in climate chambers, offices and homes.
Energy and Buildings, 23(2): 105-116.
Rijal, H.B., Tuohy, P., Humphreys, M.A., Nicol, J.F., Samuel, A. and J. Clarke. 2007. Using results from
field surveys to predict the effect of open windows on thermal comfort and energy use in buildings.
Energy and Buildings, 39(7): 823-836.
Rijal, H.B., Tuohy, P., Humphreys, M.A., Nicol, J.F., Samuel, A., Raja, I.A. and J. Clarke. 2008.
Development of adaptive algorithms for the operation of windows, fans and doors to predict thermal
comfort and energy use in Pakistani buildings. ASHRAE Transactions, 114 (2): 555-573.
Rijal, H.B., Tuohy, P., Nicol, F., Humphreys, M.A., Samuel, A. and J. Clarke. 2008a. Development of
an adaptive window-opening algorithm to predict the thermal comfort, energy use and overheating
in buildings. Journal of Building Performance Simulation, 1 (1): 17-30.
Rijal, H.B., Humphreys, M.A. and J.F. Nicol. 2009. Understanding occupant behaviour: the use of
controls in mixed-mode office buildings. Building Research and Information, 37 (4): 381-396.
Rijal, H.B., Tuohy, P., Humphreys, M.A., Nicol, J.F. and A. Samuel. 2011. An algorithm to represent
occupant use of windows and fans including situation-specific motivations and constraints. Building
Simulation, 4 (2): 117-134.
Rijal, H.B. and S. Yoshimura. 2011. Investigation of comfort temperature in naturally ventilated and air
condition mode in Japanese houses. The 4th International Conference on Human-Environment
System (ICHES 2011), pp. 259-264, 3-6 October, Sapporo, Japan.
Rijal, H.B., Tuohy, P., Humphreys, M.A., Nicol, J.F. and A. Samuel. 2012. Considering the impact of
situation-specific motivations and constraints in the design of naturally ventilated and hybrid
buildings. Architectural Science Review, 55(1): 35–48.
Rijal, H.B., Honjo, M., Kobayashi, R. and T. Nakaya. 2013. Investigation of comfort temperature,
adaptive model and the window opening behaviour in Japanese houses. Architectural Science
Review, 56(1): 54–69.
Rijal, H.B., Humphreys, M.A. and J.F. Nicol. 2014. Development of the adaptive model for thermal
comfort in Japanese houses. Proceedings of 8th Windsor Conference: Counting the Cost of Comfort
in a changing world, Cumberland Lodge, Windsor, UK, 10-13 April 2014. London: Network for
Comfort and Energy Use in Buildings, http://nceub.org.uk
Robinson, D. and F. Haldi. 2008. An integrated adaptive model for overheating risk prediction. Journal
of Building Performance Simulation, 1 (1): 43-55.
Suzuki, T., Umemiya, N. and H. Yoshida. 2002. Structure of the causes of window opening and closing
behavior from summer to autumn. J. Archit. Plann. Environ. Eng., AIJ, 556: 91-98.
Umemiya, N. and H. Yoshida. 2004. Survey on relationship between the ratio of window-opening area
and the indoor and outdoor climate mainly for the season of natural ventilation. Transaction of AirConditioning and Sanitary Engineers of Japan, 92: 19-28.
Yun, G.Y. and K. Steemers. 2007. Time-dependent occupant behaviour models of window control in
summer. Building and Environment, 43(9): 1471-1482.
30th INTERNATIONAL PLEA CONFERENCE
16-18 December 2014, CEPT University, Ahmedabad
8

Download Report