- Email: [email protected]
Thermal adaptation in overheated residential buildings in severe cold area in China
Accepted Manuscript Title: Thermal adaptation in overheated residential buildings in severe cold area in China Author: Zhaojun Wang Yuchen Ji Jing Ren PII: DOI: Reference:
S0378-7788(16)31572-9 http://dx.doi.org/doi:10.1016/j.enbuild.2017.04.053 ENB 7549
To appear in:
ENB
Received date: Revised date: Accepted date:
16-11-2016 12-4-2017 19-4-2017
Please cite this article as: Z. Wang, Y. Ji, J. Ren, Thermal adaptation in overheated residential buildings in severe cold area in China, Energy and Buildings (2017), http://dx.doi.org/10.1016/j.enbuild.2017.04.053 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Thermal adaptation in overheated residential buildings in severe cold area in China
ip t
Zhaojun Wang1,*, Yuchen Ji1, Jing Ren1 1
School of Municipal and Environmental Engineering, Harbin Institute of Technology, Harbin,
cr
150090, China *
us
Corresponding email: [email protected]
Abstract
an
The winter in the severe cold area of China is long and cold. The mean outdoor temperature is
M
about -10.0°C during the winter in Harbin, while the indoor air temperature is often above 24°C. How does the indoor environment influence human thermal comfort and adaptation in such an
ed
overheated environment? A combined approach of spot-reading measurements and occupant interviews was adopted in
pt
nine residential buildings of five communities during the heating period in 2013-2014. Twenty
Ac ce
residents were chosen as respondents. Totally 308 valid questionnaires were collected. The heating periods were separated into three phases based on the outdoor temperature. The results show that the mean indoor air temperatures in the early-, mid- and late-heating periods were 23.6°C, 24.3°C and 25.0°C, respectively, which were larger than or close to the upper limit recommended by thermal comfort standards, and slightly higher than the related thermal neutral temperatures. With the heating process, the mean clothing insulation of residents decreased. Opening windows and reducing clothing were mainly taken by the residents to adapt to the overheated environment.
Page 1 of 49
The thermal neutral temperature has an upward tendency with the increasing indoor air temperature. On the other hand, overheating in residential buildings would make residents open windows, which may cause thermal discomfort and energy waste. Therefore, the lower limit of the
fully arouse residents’ adaptation and achieve sustainable building designs.
ip t
comfort indoor air temperature range should be suggested as the heating temperature, which could
cr
Keywords: Thermal comfort; Thermal adaptation; Overheating; Severe cold area; Residential
us
building
1. Introduction
an
The winter in the severe cold area of China is long and cold. The heating period lasts for 6
M
months from October until April of the next year in Harbin, a representative city in the area. The central heating system was operated throughout each day during the winter period.
ed
The mean outdoor temperature is about -10.0°C during the winter in Harbin, while the indoor
overheated buildings?
pt
air temperature is often above 24°C. And how do thermal comfort and adaptation perform in such
Ac ce
Nicol and Humphrey argued that people take various measures to adapt to their environment [1]. In the severe cold area of China, people adapt to the thermal environment mainly by changing clothing and opening windows.
Recently, overheating was found in different kinds of buildings during the heating season worldwide [2-8]. For example, in the United Kingdom, an increase of up to 1.38°C per decade in mean dwelling indoor temperatures in winter may have occurred from 1978 to 1996 [8]. Mumovic et al. found that there was evidence of overheating in winter in secondary schools in England while both teachers and students felt too warm in such an environment [2]. Kavgic et al.
Page 2 of 49
suggested that the air temperature was high in centralized heating residential houses with a great saving potential [3]. Cao et al. found that once people adapt to the warmer environment, they would lose adaptability to cold outdoor climate, which would cause energy wastes and health
ip t
problems [4]. Overly high indoor temperatures may induce a higher thermal neutral temperature [9]. Based on
cr
the adaptive theory [1, 10], thermal neutral temperature may increase with indoor temperatures
us
getting higher in space heated buildings. After adaptation to the thermal environment, people became more sensitive to the temperature drop, causing more cold discomfort [11]. Furthermore, it
an
may be easier and quicker for occupants to adapt to a well-controlled while for ‘spoiled’ occupants
M
it may be difficult [9]. Thus, the comfortable temperature ranges during winter would be higher and narrower [12]. On the other hand, in the centralized heated building, overly high temperature
ed
would cause more energy consumption [3, 4]. Based on the previous outcomes, our group found that the thermal neutral temperature changed
pt
when space heating began in residential buildings, and the thermal neutral temperatures in winter
Ac ce
and spring during space heating were different in teaching buildings [5-7]. The thermal neutral temperature in winter was lower than that in spring, and the indoor air temperature was overly high, so people felt uncomfortable [7]. During the long heating period, the outdoor temperatures fluctuate significantly, whereas the indoor temperatures stay almost constant in the severe cold area with central heating. Then, does the thermal neutral temperature change with different heating periods? Should the indoor design temperature change with different periods? Generally, the previous investigations were conducted in different periods or in different kinds
Page 3 of 49
of architectures. In addition, the time span was relatively small, and little focus was paid to thermal adaptation in overheated environments in severe cold climate in winter. Will residents’ thermal sensations change in the six-month heating period? To answer the above questions, a
ip t
continuous tracking investigation on residents was conducted during the period.
2. Methods
cr
In the winter of 2013-2014, a combined approach of spot-reading measurements and occupant
us
interviews was adopted. The measurement and questionnaires were processing at the same time. A digital self-recorded thermometer was placed in each subject’s main occupied room for
an
continuous measurement. Besides, we went to the apartments and measured indoor thermal
M
parameters every 2 or 3 weeks. The participants were asked to answer the questionnaires online once every week, and 308 valid questionnaires were received.
ed
2.1. Subjects Ten apartments were selected from nine buildings in five residential communities, with twenty
pt
participants. Any room surveyed with three internal walls and one exterior wall, where a radiator
Ac ce
was put under the exterior window. The information of the 10 apartments was given in Table 1. From Table 1, it is seen that most of the buildings investigated were constructed in the 90s. The structure of the building is brick and concrete with the thickness of 490mm without thermal insulation material. The apartments are equipped with central heating systems with radiators, except No 9 and 10 with floor heating. Because the temperature difference between living room and bedroom is very small with central heating system, we chose one room in a house to monitor the air temperature and relative humidity. Table 1 Building Information.
Page 4 of 49
The male to female ratio was almost 1:1. To compare with the youth group, participants in this study ranged in age from 28 to 72, with an average of 48.5. They had lived in Harbin above 30 years in average, completely accustomed to the local climate. Table 2 shows the backgrounds of
Table 2 Participants’ backgrounds.
cr
2.2. Measurement
ip t
participants.
globe temperature and wall inner surface temperature.
us
Thermal parameters involve indoor and outdoor air temperature, RH and air speed, as well as
an
The indoor air temperature and RH were measured in the continuous tracking. A self-recorded thermo-hygrometer was set 1.5m above the floor in each occupied room for continuous
M
measurement. In the manual tests, the globe temperature and air speed were measured close to the
ed
subjects at the room center, while the globe temperature and air speed were measured at the height of 0.6m. The surface temperatures were also tested manually. 5 points’ temperatures were
pt
measured in each wall inner surface, so did the center’s temperature in the floor and ceiling. In
Ac ce
addition, a thermo-hygrometer was placed in a box 1.5m above the ground and 10 m from the wall of any office building in a campus, recording the outdoor air temperature and RH continuously. The test instruments are shown in Table 3. Table 3 Test instruments and accuracy.
2.3. Subjective questionnaire The subjective survey was conducted through online questionnaires, which were filled in by respondents through computers every week. The subjective survey included the following: (1) The thermal responses of subjects, such as thermal sensation, comfort, preference and
Page 5 of 49
acceptability. ASHRAE 7-point scale was used in thermal sensation. Vote Scales were shown in Table 4. Table 4 Vote scales of thermal response.
closing windows, changing clothes and activity levels, having hot drink.
ip t
(2) The adaptive measures taken to improve the indoor thermal environment, such as opening or
cr
Through the measurement period, there were many residents who kept windows open for ventilation and improve thermal comfort. Therefore, the occupants were asked when they felt
us
warm or hot, what they would like to do. The lists include opening windows, taking off clothes,
3. Results and discussion of measurements
an
etc. They would like to choose what they used to.
M
3.1. Outdoor temperature and research phase division
ed
The mean outdoor temperature in winter (Oct. 17th to Apr. 10th) in Harbin was -9.4℃ in the Chinese standard [13], which is derived from the statistical data of outdoor temperature during the
pt
past twenty years. The outdoor temperatures during the investigation are given in Fig. 1.
Ac ce
According to the monitored data in this investigation, the daily mean outdoor temperature varied from -30℃ to 10℃ throughout the winter, with an average of -10℃ approximately. As a result, the temperature of -10℃ was recognized as a boundary of middle heating. Therefore, the whole heating period is divided into 3 phases (early heating period (EH), mid heating period (MH), and late heating period (LH)) according to the outdoor temperature in Harbin.
Fig. 1. Outdoor temperature and phase division of heating period. As Fig. 1 shows, the daily mean outdoor air temperature descended below -10℃ on 22 November 2013 and rebounded above -10℃ on 2 March 2014. So the two days were termed as the beginning and end of MH respectively. The mean outdoor temperature of EH and LH were 1.4
Page 6 of 49
℃and 2.2℃, which were close. The maximums of the two phases were 11℃ and 16℃, and the minimums were -8℃ and -12℃. The outdoor temperature of MH varied from -26℃ to 2℃, with an average of -16℃, which was the lowest in the periods.
ip t
3.2 Indoor air temperature and relative humidity According to statistic results, in four apartments of the ten apartments, the indoor air
cr
temperatures ranged from 20℃ to 25℃, while the indoor air temperatures varied from 25℃ to
us
30℃ in the other six apartments. For contrast analysis, No.2 and No.6 apartments were taken for an example, where the indoor air temperatures are shown in Fig. 2.
an
In No.2 apartment, the mean indoor air temperatures of the three phases were 24.37℃, 25.58℃
M
and 26.74℃, respectively. In 96.4% of the whole time, the indoor air temperatures were higher than 24℃, which is the upper limit of the thermal comfort zone in winter [14, 15]. In No.6
ed
apartment, the mean indoor air temperatures of the three phases were 23.64℃, 23.81℃ and 22.98 ℃, respectively. In 37.6% of the whole time, the indoor air temperatures were above 24℃.
Ac ce
pt
Generally, there existed overheating in some apartments.
Fig. 2. Hourly mean temperatures in No.2 and No.6 apartments.
The comfort range in winter is 20-24℃ given by thermal comfort standards [14, 15]. For further analysis of overheating in apartments, Fig. 3 represents the distribution frequencies of indoor air temperatures of all the ten apartments from 15℃ to 33℃ at 1℃ interval. In each phase, the temperature frequency followed a normal distribution. With time processing, frequencies got larger in high temperatures. In the three phases, the frequencies of temperature above 24℃ were 47.5%, 57.1% and 57.0%, respectively. Therefore, it is further confirmed that overheating in apartments was significant.
Page 7 of 49
Fig. 3. Distribution frequency of indoor air temperature in 10 apartments in three phases. The mean indoor air temperatures and RH of all the apartments in three phases were figured out, shown in Table 5.
ip t
Table 5 Mean indoor air temperature and RH in three phases.
As known in Table 5, the mean air temperature during the whole period was 24.3℃. In LH, the
cr
mean indoor air temperature was at the maximum of 25.0℃. In the three phases, the mean air
us
temperatures were close to or beyond the upper limit of 24℃ recommended by standards. The mean RH ranged in 30%-50%.
an
The mean daily air temperatures and RH of ten apartments is given in Fig. 4.
M
Fig. 4. Mean daily air temperature and RH in apartments. As indicated in Fig. 4, the indoor air temperature became stable since the centralized heating
3.3. Operative temperature
ed
system applying, ranging in 21.5℃-26.1℃. The mean daily RH varied in 29.4%-60.5%.
pt
The indoor air temperature and globe temperature were recorded continuously every 5 min for
Ac ce
one week at the same site in an apartment. The mean radiant and operative temperatures were figured out by Equations (1) and (2) respectively, shown in Fig. 5.
tr = t g + 2.44 v(t g - ta )
(1)
t o = 0.5t a + 0.5t r
(2)
—
Where t r is the mean radiant temperature, ºC. tg is the globe temperature, ºC. ta is the air temperature, ºC. v is the air velocity, m/s.
Fig. 5. Contrast between air temperature and operative temperature. Fig. 5 demonstrated that the mean difference value between air temperature and operative
Page 8 of 49
temperature was 0.35℃. Cao et.al conducted investigations on thermal environment in Beijing and Shanghai, and found that the difference between air temperature and mean radiant temperature (MRT) was within ±0.5℃[16]. It is assumed that MRT equals air temperature, and the air
ip t
temperature was applied for evaluation. As a result, the air temperature was taken as an index for evaluating thermal comfort in the residential buildings.
cr
3.4. Other physical parameters
us
Table 6 gives the statistical results of environment parameters except air temperature and humidity, including air speed, globe temperature and surface temperature of envelope.
an
Table 6 Statistical results of the other physical parameters
M
In different periods, mean air speed was below 0.05m/s, satisfying human thermal requirements. Convection plays an important role in heat transfer because the residence mainly applied radiators
ed
for heating. Therefore, the globe temperature was close to the air temperatures, which further confirms that air temperature can be used as an evaluation index in this study.
Ac ce
4.1. Clothing insulation
pt
4. Results and discussion of questionnaire
The clothing insulation values of respondents in residential buildings in three phases are shown in Fig. 6.
Fig. 6. Mean clothing insulation in three phases.
As indicated in Fig. 6, the mean clothing insulations were 0.85, 0.78 and 0.69clo, and the standard deviations were 0.25, 0.19, 0.14clo, respectively. The clothing insulation decreased and got narrow with the heating period processing, when people had gradually adapted to the indoor air temperature. Clothing adjustments suggest that people can adapt to the indoor environment
Page 9 of 49
well through behavioral adjustment. 4.2. Thermal sensation The distribution of thermal sensation vote (TSV) of respondents in the residential buildings is
ip t
given in Fig. 7. The TSV frequency followed a normal distribution in each phase.
Fig. 7. Distribution frequency of TSV.
cr
With the space heating going on, the TSV gradually shifted to the warm side. In three phases,
us
only few respondents voted for -2 or -3, while more respondents felt warm (vote for +2 or +3).
evidence of overheating in residential buildings.
an
23.6% of respondents felt warm in MH, so did 21.0%, 10.0% in EH and LH. This is also a valid
M
Fig. 8 shows the air temperature distribution and clothing insulation at TSV of 0, +1, +2 and +3, to demonstrate a deep relationship between thermal sensation and air temperature.
ed
Fig. 8. Air temperature distribution and clothing insulation at different TSV. As seen in Fig. 8, at votes of +1 and +2, 50% of the indoor air temperatures were higher than 24
pt
℃. At the vote of +3, 95% of the temperatures were higher than 24℃.
Ac ce
Based on adaptive theory, when people get accustomed to a thermal environment, they may be less sensitive to the temperature to some extent. However, when the temperature goes beyond the adapted comfort range, human thermal sensation is aroused. As a result, when 50% of the indoor air temperatures were higher than 24℃, people may accept the temperature and voted for “neutral”, “slightly warm” or “warm”. But when 95% of the temperatures were higher than 24℃ beyond the adaptation range, people did not accept it and voted for “hot”. Meanwhile, there is no obvious discrepancy between clothing insulations at different votes. The metabolic rates were almost constant because the respondents were required not to do heavy
Page 10 of 49
activity. Generally, people felt warm mainly because of overheating indoors. 4.3. Thermal preference The distribution of thermal preference vote (TPV) is illustrated in Fig. 9. From EH to MH, the
ip t
votes of preferring temperature unchanged increased significantly, inferring that people adapted to the warm environment psychologically. The votes of preferring a higher temperature decreased by
us
Fig. 9. Distribution frequency of TPV.
cr
half, which indirectly suggesting overheating in the residential buildings.
4.4. Thermal Comfort
an
Fig. 10 gives the distribution of thermal comfort vote (TCV). The “comfort vote” (vote for “0”)
M
was more than 80% in the three periods and increased with heating processing. It indicates that residents adapted to the environment. Meanwhile, the “slightly discomfort” vote (vote for “1”)
ed
was more than 10%, indicating over-high temperature in the residential buildings.
Fig. 10. Distribution frequency of TCV.
Ac ce
results of TSV.
pt
Fig. 9 and 10 all assume that residents get accustomed to the environment, in accord with
4.5. Behavioral adjustment
The residents were interviewed: what do they often do when they feel warm or hot? The lists include opening windows, taking off clothes, etc. They may choose as they used to do. Fig. 11 shows the distribution of behavioral adjustments when respondents felt hot in residential buildings.
Fig. 11. Distribution frequency of adaptive adjustments in residential buildings. Through the heating period, the residents commonly improved the indoor environment by
Page 11 of 49
opening window in warm environment. Before and after the space heating, opening window could enhance natural ventilation and improve thermal comfort. However, in the heating period, opening window would cause energy waste for a hot indoor environment.
ip t
As seen in Fig. 11, 37.4% of the subjects opened windows when they felt hot in EH, so did 37.4% in MH and moreover 42.0% in LH. Therefore, the over high indoor air temperature led to a
cr
waste in winter.
an
5.1. Time scale of air temperature in thermal comfort model
us
5. Discussion of thermal comfort
Fig. 12. Indoor temperatures in a month.
M
Fig. 12 gives the daily mean indoor temperatures in a month, which changed a little from
ed
25.8°C to 26.9°C, with a standard deviation of 0.32°C, which means that the difference of mean temperatures between 7d and 30d is very similar. Therefore, the maximum time scale of 7d was
pt
adopted.
Ac ce
Residents’ mean thermal sensation (MTS) votes and the current temperature were linear fitted, as well as average air temperatures of 1d, 2d, 3d, 4d, 5d, 6d, 7d. The correlation coefficients of MTS and average air temperatures of different time scales were worked out, shown in Table 7. According to the statistics in Table 7, the correlation coefficients R2 of different time scales in each phase varied much, with distinct changing trends. By comparing the correlation coefficients R2 of different time scales, it is found that the correlation coefficient of 7d was usually the largest, which means that MTS has correlates with average air temperatures of 7d much well. This finding complies with conclusions in natural buildings [17]. Therefore, the average air temperatures of 7d were taken as an index for evaluating thermal environments in residential buildings.
Page 12 of 49
Table 7 Correlation coefficients R2 of different time scales. 5.2. Thermal neutral temperature and thermal adaptation We weighted the regression model by the number of samples falling within the each temperature bin.
cr
temperature in three phases (EH, MH and LH) are expressed as linear models.
ip t
In Fig. 13, the relationships of MTS and predicted mean votes (PMV) with indoor air
Fig. 13. Relationships of MTS and PMV with indoor air temperature in three phases.
us
Fig. 13 illustrates a comparison between MTS and PMV. There existed some discrepancy in
an
each period. To compare the difference between MTS and PMV, the two groups of each phase were analyzed with paired t-test. It is found that the p values of EH, MH and LH are 0.0184,
M
0.8667 and 0.1608, respectively, which means that there is a significant difference between MTS
ed
and PMV in EH, and MTS is not significantly different with PMV at the 0.05 level in MH and LH. Especially, the difference was the largest in EH period. With space heating processing, the
pt
difference became smaller in accord with PMV, which means that people get accustomed to the
Ac ce
warm environment. As known in Fig.1, there are about 30 days from EH to MH. So it is inferred that the adaptation period is about 30 days from non-heating to heating environments. Through Fig. 13, the thermal neutral temperatures in different phases were figured out, which are 21.6℃, 23.5℃ and 23.1℃, respectively. The neutral temperatures in winter and spring in classrooms were 22.6℃ and 21.7℃, respectively [7], which shows the same trend with neutral temperatures in residential buildings from EH to LH. Table 8 shows the comparison between thermal neutral temperatures and mean air temperatures of different phases, which suggests a significant discrepancy between the two temperatures. Table 8 Thermal neutral temperatures and mean air temperatures of different phases.
Page 13 of 49
After space heating began, the indoor air temperature increased significantly. Hence, most respondents did not adapted to the thermal environment, and felt warm. And this resulted in a low thermal neutral temperature, with a difference of 2.0℃. In MH, the outdoor temperatures
ip t
decreased, while the mean indoor air temperature went up by 0.7℃, and the neutral temperature is
cr
0.8℃ lower than the mean air temperature. Because the respondents built up some adaptation to
the indoor environment, the thermal neutral temperature increased by 1.9℃. In LH, the outdoor
us
temperatures rebounded, and the mean indoor air temperature still increased by 0.7℃. The thermal
an
neutral temperature was 1.9℃ lower than the mean air temperature.
Table 9 shows the slopes and confidence intervals of regression models for MTS and PMV. In
M
EH, the slope of MTS regression model was larger than PMV regression model. In MH and LH,
ed
the slopes of MTS regression model became lower than PMV regression model. In general, with heating underway, respondents’ thermal sensitivity decreased and adaptation to the warm
pt
environment formed.
Ac ce
Table 9 Slopes and confidence intervals of regression models for MTS and PMV In a hot environment, human core temperature as well skin temperature is regulated to a higher level after ‘heat acclimation.’ [18]. More specifically, stronger physiological responses and more discomfort happened when people who acclimated to neutral- warm indoors were exposed in moderate cold environment than those adapted to the cold [19]. Therefore, the neutral temperature increased with the mean air temperature increasing in the heating period, which indicates that the residents got used to the indoor thermal environment gradually. On the other hand, the neutral temperatures were always below mean air temperatures, suggesting overheating in residential buildings, which might cause discomfort and energy wastes. From the view of health, once
Page 14 of 49
residents’ dependence to the indoor warm environment is built up, resistibility to the cold outdoor climate would be weakened. As a result, in addition to energy wastes, overheating would be against health. This finding further confirms the results presented by Cao et al in reference [4].
ip t
The residential buildings were constructed of brick and concrete. The performance of envelope insulation has been improved since 2005, so did the efficiency of central heating system. The
cr
indoor air temperature will keep climbing up in the newly-built buildings in the near future, which
5.3. Thermal adaptation between the elder and the youth
us
may induce a higher thermal neutral temperature.
an
An investigation in university dormitories was carried out in the same year to study the
M
college-students’ thermal comfort [11]. “The youth group” was referred to the students. We compared the mean thermal sensations between the youth in university dormitories and the elder
ed
in apartments, shown in Fig. 14. And the slopes and confidence intervals of the regression models are given in Table 10. The thermal neutral temperature for elder is higher than the temperature for
pt
the youth. Young people may have stronger adaptation to cool exposure.
Ac ce
The slope of MTS regression model for youth is the largest in EH, which suggests that young people are more sensitive to temperature upward. In MH and LH, the significant-reduced slope demonstrates that thermal sensitivity of young people decreased due to a stronger adaptation. From EH to FH, the difference of MTS between the youth and the elder got smaller. It may infer that thermal adaptation bridge the gap of age on thermal comfort. Fig. 14. Comparison of MTS between the youth and the elder in three phases. Table 10 Slopes and confidence intervals of regression models of MTS for youth and elder 5.4. Indoor air temperature and human health
Page 15 of 49
When temperatures fall into the comfort zone or beyond the zone, dryness and symptoms of sick building syndrome obviously became severe with temperature increase [20, 21]. On the contrary, a reasonable temperature drop could reduce symptoms of sick building syndrome
ip t
effectively [22]. A temperature drop of 2-3℃would double the perceived air quality [23]. Raised temperatures and air velocities, and the low indoor humidity levels in winter make the eye more
cr
sensitive to airborne particulates and other forms of pollution, which results in eye redness and
us
chronic eye ache [24, 25]. Moreover, a causal link between temperature and adiposity seems to be proved [8]. In addition, once people adapted to the overheated environment, their adaptability to
an
the cold climate would be weakened, which may cause health problems [4].
of comfort range may be healthier.
M
In general, over-high temperatures cause a lot of health problems. The lower limit temperature
ed
5.5. Energy analysis The increase rate of energy was worked out based on indoor air temperature. Fig. 15 shows
pt
distribution frequencies of air temperature in the whole heating period.
Ac ce
Fig. 15. Distribution frequency of indoor air temperature in 10 apartments.
An Equation for energy increase rate is given as follows:
∆N =
18
ti − tb × α i b − tw
∑ t i =1
(3)
Where ∆N is the energy increase rate, ti is the median of each interval, tb is the basic indoor air temperature, tw is the outdoor design temperature，equal to −24.2℃. According to Fig. 15 and Equation (3), 10.14% energy could be saved when the indoor temperature dropped from the operating conditions to 20℃. In general, in addition to energy increase, overheating in residential buildings would cause
Page 16 of 49
energy wastes for opening windows and discomfort for cold wind into the room. Moreover, it would weaken residents’ adaptation to the cold outdoor climate. Therefore, the indoor temperature should be kept at the lower limit of the comfort range in winter to keep residents’ adaptation and
ip t
realize sustainable building designs.
6. Conclusions
cr
(1) In EH, MH and LH, the mean indoor air temperatures were 23.6℃, 24.3℃, 25.0℃. And the
us
frequencies of air temperatures above 24℃ were 47.5%, 57.1% and 57.0%, respectively. This result shows that there did exist overheating in residential buildings.
an
(2) When 50% of the indoor air temperatures were higher than 24℃, people may accept the
M
temperature and voted for “neutral”, “slightly warm” or “warm”. But when 95% of the temperatures were higher than 24℃ beyond the adaptation range, people did not accept it and
ed
voted for “hot”.
(3) The thermal neutral temperature has an upward tendency with the increasing indoor air
pt
temperature, which indicates that the residents got used to the indoor thermal environment
Ac ce
gradually. On the other hand, the neutral temperatures were always below mean air temperatures, suggesting overheating in residential buildings.
(4) In the heating period, about 20% respondents felt warm (vote for +2 or +3) and above 30% respondents opened windows to release overheat. The clothing insulation decreased with the heating period processing.
(5) Due to high temperatures indoors, the residents commonly adapt to the indoor environment by opening window or taking off clothes. However, it would cause energy increase and discomfort. Therefore, the indoor temperature should be kept at the lower limit of the comfort
Page 17 of 49
range in winter to keep residents’ adaptation and realize sustainable building designs.
Acknowledgement All the participants are sincerely acknowledged in this research.
ip t
Funding The work presented in this paper was funded by the National Natural Science Foundation of
cr
China (No. 51278142).
us
References
[1]J.F. Nicol, M.A. Humphreys, Adaptive thermal comfort and sustainable thermal standards for
an
buildings, Energy and Building 34 (6) (2002) 563-572.
M
[2] D. Mumovic, J. Palmer, M. Davies, M. Orme, I. Ridley, T. Oreszczyn, C. Judd, R. Critchlow, H.A. Medina, G. Pilmoor, C. Pearson, P.Way, Winter indoor air quality, thermal comfort and
44(7) (2009) 1466-1477.
ed
acoustic performance of newly built secondary schools in England, Building and Environment
pt
[3] M Kavgic, Summerfield A, Mumovic D, Z.M. Stevanovic, V. Turanjanin, Z.Z. Stevanovic,
Ac ce
Characteristics of indoor temperatures over winter for Belgrade urban dwellings: Indications of thermal comfort and space heating energy demand, Energy and Buildings 47(4)(2012) 506-514.
[4] B. Cao, Y.X. Zhu, Q. Ouyang, X. Zhou, L. Huang, Field study of human thermal comfort and thermal adaptability during the summer and winter in Beijing, Energy and Buildings 43(5)(2011) 1051-1056. [5] Z.J. Wang, G. Wang, L.M. Lian, A field study of the thermal environment in residential buildings in Harbin, ASHRAE Transaction 109(2)(2003) 350-355.
Page 18 of 49
[6] Z.J. Wang, L. Zhang, J.N. Zhao, Y.N. He, A.X. Li, Thermal responses to different residential environments in Harbin, Building and Environment 46(11) (2011) 2170-2178. [7] Z.J. Wang, A.X. Li, J. Ren, Y.N. He, Thermal adaptation and thermal environment in
ip t
university classrooms and offices in Harbin, Energy and Buildings 77(7)(2011) 192-196. [8] A. Mavrogiannia, F. Johnsonb, M. Uccic, A. Marmotc, J. Wardleb, T. Oreszczyna, A.
cr
Summerfielda, Historic variations in winter indoor domestic temperatures and potential
us
implications for body weight gain, Indoor and Built Environment 22(2) (2013) 360-375. [9] M.H. Luo, R.J. de Dear, W.J. Ji, B. Cao, B.R. Lin, Q. Ouyang, Y.X. Zhu, The dynamics of
an
thermal comfort expectations: The problem, challenge and implication, Building and
M
Environment 95(1) (2016) 322-329.
[10] R.J. de Dear, G.S. Brager, Developing an adaptive model of thermal comfort and preference.
ed
ASHRAE Transactions 4 (1) (1998) 145–167. [11] H.R. Ning, Z.J. Wang, X.X Zhang, Y.C. Ji, Adaptive thermal comfort in university
pt
dormitories in the severe cold area of China, Building and Environment 99(5) (2016) 161-169.
Ac ce
[12] W.B. Meyer, Why indoor climates change: a case study: Climate Change 55(3) (2002) 395–407.
[13] The ministry of construction, Design code for Heating ventilation and air conditioning of civil buildings, GB50736-2012, (2012) (In Chinese). [14] ASHRAE.ANSI/ASHRAE standard 55 Thermal environmental conditions for human occupancy. Atlanta: ASHRAE, 2013. [15] CEN, Indoor environmental input parameters for design and assessment of energy performance of buildings–addressing indoor air quality, thermal environment, lighting and
Page 19 of 49
acoustics, EN 15251, European Committee for Standardization, Brussels, 2007. [16] B. Cao, Q. Ouyang, Y. Zhua, L. Huang, H. Hu, G. Deng, Development of a multivariate regression model for overall satisfaction in public buildings based on field studies in Beijing
ip t
and Shanghai, Building and Environment 47(1) (2012) 394–399. [17] C. Morgan, R. J. de Dear, Weather, clothing and thermal adaptation to indoor climate,
cr
Climate Research 24(3) (2003) 267-284.
us
[18] D. Wang, H. Zhang, E. Arens, C. Huizenga, Observations of upper-extremity skin temperature and corresponding overall-body thermal sensations and comfort, Building and
an
Environment 42(12) (2017) 3933-3943,
M
[19] M.H. Luo, B. Cao, Ouyang, Y.X. Zhu, Indoor human thermal adaptation: dynamic processes and weighting factors, Indoor Air 27(2) (2017) 273-281.
ed
[20] J.J.K. Jaakkola, O.P. Heinonen, Sick building syndrome, sensation of dryness and thermal comfort in relation to room temperature in an office building: Need for individual control of
pt
temperature, Environment International 15(1-6) (1989) 163-168.
Ac ce
[21] A.L. Krogstad, G. Swanbeck, L. Barregard, S. Hagberg, K.B. Rynell, A. Ran, et al., A prospective study of indoor climate problems at different temperatures in offices(in Swedish), Gothenburg: Volvo Truck Corporation. [22] D.P. Wyon, Sick buildings and the experimental approach, Environmental Technology 13(4) (1992) 313-322. [23] P.O. Fanger, What is IAQ? // Proceedings of the 10th International Conference on Indoor Air Quality and Climate. Beijing, 2005: 1-8. [24] C. Franck, Eye symptoms and signs in buildings with indoor climate problems (‘office eye
Page 20 of 49
syndrome’), Acta Ophthalmologica 64(3) (1986) 306-311. [25] N.M. Wyon, D.P. Wyon, Measurement of acute response to draught in the eye, Acta
Ac ce
pt
ed
M
an
us
cr
ip t
Ophthalmologica 65(4) (1987) 385-392.
Page 21 of 49
Figure captions Fig. 1. Outdoor temperature and phase division of heating period. Fig. 2. Hourly mean temperatures in No.2 and No.6 apartments.
ip t
Fig. 3. Distribution frequency of indoor air temperature in 10 apartments in three phases. Fig. 4. Mean daily air temperature and RH in apartments.
cr
Fig. 5. Contrast between air temperature and operative temperature.
us
Fig. 6. Mean clothing insulation in three phases. Fig. 7. Distribution frequency of TSV.
an
Fig. 8. Air temperature distribution and clothing insulation at different TSV.
Fig. 10. Distribution frequency of TCV.
M
Fig. 9. Distribution frequency of TPV.
ed
Fig. 11. Distribution frequency of adaptive adjustments in residential buildings. Fig. 12. Indoor temperatures in a month
pt
Fig. 13. Relationships of MTS and PMV with indoor air temperature in three phases.
Ac ce
Fig. 14. Comparison of MTS between the youth and the elder in three phases. Fig. 15. Distribution frequency of indoor air temperature in 10 apartments.
Page 22 of 49
Table captions Table 1 Building Information. Table 2 Participants’ backgrounds. Table 3 Test instruments and accuracy.
ip t
Table 4 Vote scales of thermal response.
cr
Table 5 Mean indoor air temperature and RH in three phases.
Table 7 Correlation coefficients R2 of different time scales.
us
Table 6 Statistical results of the other physical parameters.
an
Table 8 Thermal neutral temperatures and mean air temperatures of different phases. Table 9 Slopes and confidence intervals of regression models for MTS and PMV.
Ac ce
pt
ed
M
Table 10 Slopes and confidence intervals of regression models for MTS for youth and elder.
Page 23 of 49
ip t cr
Ac ce
pt
ed
M
an
us
Fig. 1. Outdoor temperature and phase division of heating period.
Page 24 of 49
ip t cr us an
Ac ce
pt
ed
M
a. No.2 apartment
b. No.6 apartment
Fig. 2. Hourly mean temperatures in No.2 and No.6 apartments.
Page 25 of 49
ip t cr us
Ac ce
pt
ed
M
an
Fig. 3. Distribution frequency of indoor air temperature in 10 apartments in three phases.
Page 26 of 49
ip t cr
Ac ce
pt
ed
M
an
us
Fig. 4. Mean daily air temperature and RH in apartments.
Page 27 of 49
ip t cr us
Ac ce
pt
ed
M
an
Fig. 5. Contrast between air temperature and operative temperature.
Page 28 of 49
ip t cr us
Ac ce
pt
ed
M
an
Fig. 6. Mean clothing insulation in three phases.
Page 29 of 49
ip t cr
Ac ce
pt
ed
M
an
us
Fig. 7. Distribution frequency of TSV.
Page 30 of 49
ip t cr us
Ac ce
pt
ed
M
an
Fig. 8. Air temperature distribution and clothing insulation at different TSV.
Page 31 of 49
ip t cr
Ac ce
pt
ed
M
an
us
Fig. 9. Distribution frequency of TPV.
Page 32 of 49
ip t cr
Ac ce
pt
ed
M
an
us
Fig. 10. Distribution frequency of TCV.
Page 33 of 49
ip t cr
Ac ce
pt
ed
M
an
us
Fig. 11. Distribution frequency of adaptive adjustments in residential buildings.
Page 34 of 49
ip t cr
Ac ce
pt
ed
M
an
us
Fig. 12. Indoor temperatures in a month
Page 35 of 49
ip t
b) MH
M
an
us
cr
a) EH
ed
c) LH
Ac ce
pt
Fig. 13. Relationships of MTS and PMV with indoor air temperature in three phases.
Page 36 of 49
M
an
us
cr
b) MH
ip t
a) EH
ed
c) LH
Ac ce
pt
Fig. 14. Comparison of MTS between the youth and the elder in three phases.
Page 37 of 49
ip t cr us
Ac ce
pt
ed
M
an
Fig. 15. Distribution frequency of indoor air temperature in 10 apartments.
Page 38 of 49
Table 1 Building Information. No.
Construction year of
Floor
Room
level
Orientation
Function
of exterior
buildings
Floor area of 2
Number of
room / m
participant
wall
2001
7
Living Room
South
34.4
2
2
2001
7
Living Room
East
15.1
2
3
1999
4
Living Room
South
36.2
2
4
1999
3
Living Room
South
23.6
5
1996
5
Living Room
South
14.1
6
1996
5
Bedroom
South
11.1
2
7
1993
7
Living Room
South
32.4
2
8
2000
7
Living Room
South
9
2001
4
Living Room
West
10
2001
7
Living Room
Nortth
us
cr
ip t
1
2 1
2
41
3
32
2
Ac ce
pt
ed
M
an
34.6
Page 39 of 49
Table 2 Participants’ backgrounds. Number of participants Mean Std. Dev. Max Min
Years in Harbin 39.7 20.1 70 13
Ac ce
pt
ed
M
an
us
cr
ip t
20
Age 48.5 13.6 72 28
Page 40 of 49
Table 3 Test instruments and accuracy. Parameter
Range
Accuracy
Notes
Air temperature
-40~100℃
±0.5℃
Continuous
Relative humidity
0~100%
±3%
test
HWZY-1
Globe temperature
-50~100℃
±0.4℃
Testo425
Air speed
0~20m/s
±0.03m/s
Testo830-T
Surface
-30~400℃
±1.5℃
WSZY-1A
meter Globe thermometer Hot wire anemometer Infrared
1
temperature
Manual test
Manual test
Ac ce
pt
ed
M
an
us
thermometer
Manual test
ip t
Thermo-hygro
Type
cr
Instrument
Page 41 of 49
Table 4 Vote scales of thermal response. Scale
Thermal Sensation
-3 cold, -2 cool, -1 slightly cool, 0 neutral, +1 slightly warm, +2 warm, +3 hot
Thermal Preference
-1 cooler, 0 no change, +1 warmer
Thermal Comfort
0 comfortable, 1 slightly uncomfortable, 2 uncomfortable, +3 unbearable
Ac ce
pt
ed
M
an
us
cr
ip t
Vote
Page 42 of 49
Table 5 Mean indoor air temperature and RH in three phases. EH
MH
LH
Whole period
Mean air temperature (℃)
23.6
24.3
25.0
24.3
Mean RH (%)
46.9
34.6
34.7
37.2
Ac ce
pt
ed
M
an
us
cr
ip t
Parameter
Page 43 of 49
Table 6 Statistical results of the other physical parameters MH
LH
Mean
0.05
0.03
0.03
Std. Dev.
0.033
0.035
0.021
Max
0.12
0.21
0.09
Min
0.01
0.00
0.01
Mean
23.6
24.3
Std. Dev.
2.16
2.65
Max
28.7
35.1
Min
19.8
Mean
18.2
Surface temperature of exterior
Std. Dev.
3.91
window / ℃
Max
27.9
Min
12.2
Mean
24.6
2.16
28.1
cr
Globe temperature / ℃
20.5
20.3
17.7
23.2
3.47
4.04
27.8
30.7
13.0
15.8
21.6
23.2
24.0
2.42
2.75
2.01
us
Air speed / m/s
ip t
EH
an
Parameter
Std. Dev.
wall / ℃
Max
27.0
28.3
27.4
Min
13.4
16.9
19.6
Mean
22.8
23.8
24.4
Surface temperature of interior
Std. Dev.
2.65
3.08
2.68
wall / ℃
Max
28.6
30.2
36.5
Min
17.4
18.8
20.1
Mean
23.1
23.8
24.0
Std. Dev.
3.97
5.14
2.93
Max
31.5
36.0
30.0
Min
18.0
18.0
18.5
Mean
22.6
23.7
24.2
Std. Dev.
3.45
3.53
2.40
Max
30.0
31.5
28.5
Min
18.0
19.5
21.0
Ac ce
pt
ed
M
Surface temperature of exterior
Surface temperature of floor / ℃
Surface temperature of ceiling / ℃
Page 44 of 49
Table 7 Correlation coefficients R2 of different time scales. EH
MH
LH
Current 1d 2d 3d 4d 5d 6d 7d
0.6126 0.4972 0.5706 0.5029 0.4718 0.5016 0.5022 0.6009
0.3103 0.3894 0.4176 0.4874 0.3503 0.4488 0.3907 0.5017
0.4772 0.5131 0.5114 0.5089 0.3865 0.4801 0.5430 0.5743
Ac ce
pt
ed
M
an
us
cr
ip t
Time scale
Page 45 of 49
Table 8 Thermal neutral temperatures and mean air temperatures of different phases. EH
MH
LH
Average air temperature ta(℃)
23.6
24.3
25.0
Thermal neutral temperature tn(℃)
21.6
23.5
23.1
ta - tn (℃)
2.0
0.8
1.9
Ac ce
pt
ed
M
an
us
cr
ip t
Parameter
Page 46 of 49
Table 9 Slopes and confidence intervals of regression models for MTS and PMV MH
LH
0.2327
0.1456
0.1733
Lower limit
0.1819
0.1148
0.1358
Upper limit
0.2835
0.1765
0.2108
0.2108
0.1888
0.2143
Lower limit
0.1742
0.1676
0.1793
Upper limit
0.2473
0.2010
Slope MTS 95% confidence intervals Slope PMV
0.2494
Ac ce
pt
ed
M
an
us
cr
95% confidence intervals
ip t
EH
Page 47 of 49
Table 10 Slopes and confidence intervals of regression models for MTS for youth and elder
95% confidence interval
Lower limit Upper limit
Slope 95% confidence interval
Lower limit Upper limit
LH 0.1367 0.1038 0.1695 0.1733 0.1358 0.2108
Ac ce
pt
ed
M
an
us
cr
MTS(elder)
MH 0.1770 0.1503 0.2037 0.1456 0.1148 0.1765
ip t
Slope MTS(youth)
EH 0.4105 0.3699 0.4512 0.2327 0.1819 0.2835
Page 48 of 49
Highlights
Overheating in residential buildings in severe cold area of China were found.
2.
The heating periods were divided into three phases and three adaptive models were given.
3.
Time scale of air temperature in thermal comfort model were discussed.
4.
Frequency distribution histogram of air temperature was used to analyze energy
cr
ip t
1.
Ac ce
pt
ed
M
an
us
consumption.
Page 49 of 49
S0378-7788(16)31572-9 http://dx.doi.org/doi:10.1016/j.enbuild.2017.04.053 ENB 7549
To appear in:
ENB
Received date: Revised date: Accepted date:
16-11-2016 12-4-2017 19-4-2017
Please cite this article as: Z. Wang, Y. Ji, J. Ren, Thermal adaptation in overheated residential buildings in severe cold area in China, Energy and Buildings (2017), http://dx.doi.org/10.1016/j.enbuild.2017.04.053 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Thermal adaptation in overheated residential buildings in severe cold area in China
ip t
Zhaojun Wang1,*, Yuchen Ji1, Jing Ren1 1
School of Municipal and Environmental Engineering, Harbin Institute of Technology, Harbin,
cr
150090, China *
us
Corresponding email: [email protected]
Abstract
an
The winter in the severe cold area of China is long and cold. The mean outdoor temperature is
M
about -10.0°C during the winter in Harbin, while the indoor air temperature is often above 24°C. How does the indoor environment influence human thermal comfort and adaptation in such an
ed
overheated environment? A combined approach of spot-reading measurements and occupant interviews was adopted in
pt
nine residential buildings of five communities during the heating period in 2013-2014. Twenty
Ac ce
residents were chosen as respondents. Totally 308 valid questionnaires were collected. The heating periods were separated into three phases based on the outdoor temperature. The results show that the mean indoor air temperatures in the early-, mid- and late-heating periods were 23.6°C, 24.3°C and 25.0°C, respectively, which were larger than or close to the upper limit recommended by thermal comfort standards, and slightly higher than the related thermal neutral temperatures. With the heating process, the mean clothing insulation of residents decreased. Opening windows and reducing clothing were mainly taken by the residents to adapt to the overheated environment.
Page 1 of 49
The thermal neutral temperature has an upward tendency with the increasing indoor air temperature. On the other hand, overheating in residential buildings would make residents open windows, which may cause thermal discomfort and energy waste. Therefore, the lower limit of the
fully arouse residents’ adaptation and achieve sustainable building designs.
ip t
comfort indoor air temperature range should be suggested as the heating temperature, which could
cr
Keywords: Thermal comfort; Thermal adaptation; Overheating; Severe cold area; Residential
us
building
1. Introduction
an
The winter in the severe cold area of China is long and cold. The heating period lasts for 6
M
months from October until April of the next year in Harbin, a representative city in the area. The central heating system was operated throughout each day during the winter period.
ed
The mean outdoor temperature is about -10.0°C during the winter in Harbin, while the indoor
overheated buildings?
pt
air temperature is often above 24°C. And how do thermal comfort and adaptation perform in such
Ac ce
Nicol and Humphrey argued that people take various measures to adapt to their environment [1]. In the severe cold area of China, people adapt to the thermal environment mainly by changing clothing and opening windows.
Recently, overheating was found in different kinds of buildings during the heating season worldwide [2-8]. For example, in the United Kingdom, an increase of up to 1.38°C per decade in mean dwelling indoor temperatures in winter may have occurred from 1978 to 1996 [8]. Mumovic et al. found that there was evidence of overheating in winter in secondary schools in England while both teachers and students felt too warm in such an environment [2]. Kavgic et al.
Page 2 of 49
suggested that the air temperature was high in centralized heating residential houses with a great saving potential [3]. Cao et al. found that once people adapt to the warmer environment, they would lose adaptability to cold outdoor climate, which would cause energy wastes and health
ip t
problems [4]. Overly high indoor temperatures may induce a higher thermal neutral temperature [9]. Based on
cr
the adaptive theory [1, 10], thermal neutral temperature may increase with indoor temperatures
us
getting higher in space heated buildings. After adaptation to the thermal environment, people became more sensitive to the temperature drop, causing more cold discomfort [11]. Furthermore, it
an
may be easier and quicker for occupants to adapt to a well-controlled while for ‘spoiled’ occupants
M
it may be difficult [9]. Thus, the comfortable temperature ranges during winter would be higher and narrower [12]. On the other hand, in the centralized heated building, overly high temperature
ed
would cause more energy consumption [3, 4]. Based on the previous outcomes, our group found that the thermal neutral temperature changed
pt
when space heating began in residential buildings, and the thermal neutral temperatures in winter
Ac ce
and spring during space heating were different in teaching buildings [5-7]. The thermal neutral temperature in winter was lower than that in spring, and the indoor air temperature was overly high, so people felt uncomfortable [7]. During the long heating period, the outdoor temperatures fluctuate significantly, whereas the indoor temperatures stay almost constant in the severe cold area with central heating. Then, does the thermal neutral temperature change with different heating periods? Should the indoor design temperature change with different periods? Generally, the previous investigations were conducted in different periods or in different kinds
Page 3 of 49
of architectures. In addition, the time span was relatively small, and little focus was paid to thermal adaptation in overheated environments in severe cold climate in winter. Will residents’ thermal sensations change in the six-month heating period? To answer the above questions, a
ip t
continuous tracking investigation on residents was conducted during the period.
2. Methods
cr
In the winter of 2013-2014, a combined approach of spot-reading measurements and occupant
us
interviews was adopted. The measurement and questionnaires were processing at the same time. A digital self-recorded thermometer was placed in each subject’s main occupied room for
an
continuous measurement. Besides, we went to the apartments and measured indoor thermal
M
parameters every 2 or 3 weeks. The participants were asked to answer the questionnaires online once every week, and 308 valid questionnaires were received.
ed
2.1. Subjects Ten apartments were selected from nine buildings in five residential communities, with twenty
pt
participants. Any room surveyed with three internal walls and one exterior wall, where a radiator
Ac ce
was put under the exterior window. The information of the 10 apartments was given in Table 1. From Table 1, it is seen that most of the buildings investigated were constructed in the 90s. The structure of the building is brick and concrete with the thickness of 490mm without thermal insulation material. The apartments are equipped with central heating systems with radiators, except No 9 and 10 with floor heating. Because the temperature difference between living room and bedroom is very small with central heating system, we chose one room in a house to monitor the air temperature and relative humidity. Table 1 Building Information.
Page 4 of 49
The male to female ratio was almost 1:1. To compare with the youth group, participants in this study ranged in age from 28 to 72, with an average of 48.5. They had lived in Harbin above 30 years in average, completely accustomed to the local climate. Table 2 shows the backgrounds of
Table 2 Participants’ backgrounds.
cr
2.2. Measurement
ip t
participants.
globe temperature and wall inner surface temperature.
us
Thermal parameters involve indoor and outdoor air temperature, RH and air speed, as well as
an
The indoor air temperature and RH were measured in the continuous tracking. A self-recorded thermo-hygrometer was set 1.5m above the floor in each occupied room for continuous
M
measurement. In the manual tests, the globe temperature and air speed were measured close to the
ed
subjects at the room center, while the globe temperature and air speed were measured at the height of 0.6m. The surface temperatures were also tested manually. 5 points’ temperatures were
pt
measured in each wall inner surface, so did the center’s temperature in the floor and ceiling. In
Ac ce
addition, a thermo-hygrometer was placed in a box 1.5m above the ground and 10 m from the wall of any office building in a campus, recording the outdoor air temperature and RH continuously. The test instruments are shown in Table 3. Table 3 Test instruments and accuracy.
2.3. Subjective questionnaire The subjective survey was conducted through online questionnaires, which were filled in by respondents through computers every week. The subjective survey included the following: (1) The thermal responses of subjects, such as thermal sensation, comfort, preference and
Page 5 of 49
acceptability. ASHRAE 7-point scale was used in thermal sensation. Vote Scales were shown in Table 4. Table 4 Vote scales of thermal response.
closing windows, changing clothes and activity levels, having hot drink.
ip t
(2) The adaptive measures taken to improve the indoor thermal environment, such as opening or
cr
Through the measurement period, there were many residents who kept windows open for ventilation and improve thermal comfort. Therefore, the occupants were asked when they felt
us
warm or hot, what they would like to do. The lists include opening windows, taking off clothes,
3. Results and discussion of measurements
an
etc. They would like to choose what they used to.
M
3.1. Outdoor temperature and research phase division
ed
The mean outdoor temperature in winter (Oct. 17th to Apr. 10th) in Harbin was -9.4℃ in the Chinese standard [13], which is derived from the statistical data of outdoor temperature during the
pt
past twenty years. The outdoor temperatures during the investigation are given in Fig. 1.
Ac ce
According to the monitored data in this investigation, the daily mean outdoor temperature varied from -30℃ to 10℃ throughout the winter, with an average of -10℃ approximately. As a result, the temperature of -10℃ was recognized as a boundary of middle heating. Therefore, the whole heating period is divided into 3 phases (early heating period (EH), mid heating period (MH), and late heating period (LH)) according to the outdoor temperature in Harbin.
Fig. 1. Outdoor temperature and phase division of heating period. As Fig. 1 shows, the daily mean outdoor air temperature descended below -10℃ on 22 November 2013 and rebounded above -10℃ on 2 March 2014. So the two days were termed as the beginning and end of MH respectively. The mean outdoor temperature of EH and LH were 1.4
Page 6 of 49
℃and 2.2℃, which were close. The maximums of the two phases were 11℃ and 16℃, and the minimums were -8℃ and -12℃. The outdoor temperature of MH varied from -26℃ to 2℃, with an average of -16℃, which was the lowest in the periods.
ip t
3.2 Indoor air temperature and relative humidity According to statistic results, in four apartments of the ten apartments, the indoor air
cr
temperatures ranged from 20℃ to 25℃, while the indoor air temperatures varied from 25℃ to
us
30℃ in the other six apartments. For contrast analysis, No.2 and No.6 apartments were taken for an example, where the indoor air temperatures are shown in Fig. 2.
an
In No.2 apartment, the mean indoor air temperatures of the three phases were 24.37℃, 25.58℃
M
and 26.74℃, respectively. In 96.4% of the whole time, the indoor air temperatures were higher than 24℃, which is the upper limit of the thermal comfort zone in winter [14, 15]. In No.6
ed
apartment, the mean indoor air temperatures of the three phases were 23.64℃, 23.81℃ and 22.98 ℃, respectively. In 37.6% of the whole time, the indoor air temperatures were above 24℃.
Ac ce
pt
Generally, there existed overheating in some apartments.
Fig. 2. Hourly mean temperatures in No.2 and No.6 apartments.
The comfort range in winter is 20-24℃ given by thermal comfort standards [14, 15]. For further analysis of overheating in apartments, Fig. 3 represents the distribution frequencies of indoor air temperatures of all the ten apartments from 15℃ to 33℃ at 1℃ interval. In each phase, the temperature frequency followed a normal distribution. With time processing, frequencies got larger in high temperatures. In the three phases, the frequencies of temperature above 24℃ were 47.5%, 57.1% and 57.0%, respectively. Therefore, it is further confirmed that overheating in apartments was significant.
Page 7 of 49
Fig. 3. Distribution frequency of indoor air temperature in 10 apartments in three phases. The mean indoor air temperatures and RH of all the apartments in three phases were figured out, shown in Table 5.
ip t
Table 5 Mean indoor air temperature and RH in three phases.
As known in Table 5, the mean air temperature during the whole period was 24.3℃. In LH, the
cr
mean indoor air temperature was at the maximum of 25.0℃. In the three phases, the mean air
us
temperatures were close to or beyond the upper limit of 24℃ recommended by standards. The mean RH ranged in 30%-50%.
an
The mean daily air temperatures and RH of ten apartments is given in Fig. 4.
M
Fig. 4. Mean daily air temperature and RH in apartments. As indicated in Fig. 4, the indoor air temperature became stable since the centralized heating
3.3. Operative temperature
ed
system applying, ranging in 21.5℃-26.1℃. The mean daily RH varied in 29.4%-60.5%.
pt
The indoor air temperature and globe temperature were recorded continuously every 5 min for
Ac ce
one week at the same site in an apartment. The mean radiant and operative temperatures were figured out by Equations (1) and (2) respectively, shown in Fig. 5.
tr = t g + 2.44 v(t g - ta )
(1)
t o = 0.5t a + 0.5t r
(2)
—
Where t r is the mean radiant temperature, ºC. tg is the globe temperature, ºC. ta is the air temperature, ºC. v is the air velocity, m/s.
Fig. 5. Contrast between air temperature and operative temperature. Fig. 5 demonstrated that the mean difference value between air temperature and operative
Page 8 of 49
temperature was 0.35℃. Cao et.al conducted investigations on thermal environment in Beijing and Shanghai, and found that the difference between air temperature and mean radiant temperature (MRT) was within ±0.5℃[16]. It is assumed that MRT equals air temperature, and the air
ip t
temperature was applied for evaluation. As a result, the air temperature was taken as an index for evaluating thermal comfort in the residential buildings.
cr
3.4. Other physical parameters
us
Table 6 gives the statistical results of environment parameters except air temperature and humidity, including air speed, globe temperature and surface temperature of envelope.
an
Table 6 Statistical results of the other physical parameters
M
In different periods, mean air speed was below 0.05m/s, satisfying human thermal requirements. Convection plays an important role in heat transfer because the residence mainly applied radiators
ed
for heating. Therefore, the globe temperature was close to the air temperatures, which further confirms that air temperature can be used as an evaluation index in this study.
Ac ce
4.1. Clothing insulation
pt
4. Results and discussion of questionnaire
The clothing insulation values of respondents in residential buildings in three phases are shown in Fig. 6.
Fig. 6. Mean clothing insulation in three phases.
As indicated in Fig. 6, the mean clothing insulations were 0.85, 0.78 and 0.69clo, and the standard deviations were 0.25, 0.19, 0.14clo, respectively. The clothing insulation decreased and got narrow with the heating period processing, when people had gradually adapted to the indoor air temperature. Clothing adjustments suggest that people can adapt to the indoor environment
Page 9 of 49
well through behavioral adjustment. 4.2. Thermal sensation The distribution of thermal sensation vote (TSV) of respondents in the residential buildings is
ip t
given in Fig. 7. The TSV frequency followed a normal distribution in each phase.
Fig. 7. Distribution frequency of TSV.
cr
With the space heating going on, the TSV gradually shifted to the warm side. In three phases,
us
only few respondents voted for -2 or -3, while more respondents felt warm (vote for +2 or +3).
evidence of overheating in residential buildings.
an
23.6% of respondents felt warm in MH, so did 21.0%, 10.0% in EH and LH. This is also a valid
M
Fig. 8 shows the air temperature distribution and clothing insulation at TSV of 0, +1, +2 and +3, to demonstrate a deep relationship between thermal sensation and air temperature.
ed
Fig. 8. Air temperature distribution and clothing insulation at different TSV. As seen in Fig. 8, at votes of +1 and +2, 50% of the indoor air temperatures were higher than 24
pt
℃. At the vote of +3, 95% of the temperatures were higher than 24℃.
Ac ce
Based on adaptive theory, when people get accustomed to a thermal environment, they may be less sensitive to the temperature to some extent. However, when the temperature goes beyond the adapted comfort range, human thermal sensation is aroused. As a result, when 50% of the indoor air temperatures were higher than 24℃, people may accept the temperature and voted for “neutral”, “slightly warm” or “warm”. But when 95% of the temperatures were higher than 24℃ beyond the adaptation range, people did not accept it and voted for “hot”. Meanwhile, there is no obvious discrepancy between clothing insulations at different votes. The metabolic rates were almost constant because the respondents were required not to do heavy
Page 10 of 49
activity. Generally, people felt warm mainly because of overheating indoors. 4.3. Thermal preference The distribution of thermal preference vote (TPV) is illustrated in Fig. 9. From EH to MH, the
ip t
votes of preferring temperature unchanged increased significantly, inferring that people adapted to the warm environment psychologically. The votes of preferring a higher temperature decreased by
us
Fig. 9. Distribution frequency of TPV.
cr
half, which indirectly suggesting overheating in the residential buildings.
4.4. Thermal Comfort
an
Fig. 10 gives the distribution of thermal comfort vote (TCV). The “comfort vote” (vote for “0”)
M
was more than 80% in the three periods and increased with heating processing. It indicates that residents adapted to the environment. Meanwhile, the “slightly discomfort” vote (vote for “1”)
ed
was more than 10%, indicating over-high temperature in the residential buildings.
Fig. 10. Distribution frequency of TCV.
Ac ce
results of TSV.
pt
Fig. 9 and 10 all assume that residents get accustomed to the environment, in accord with
4.5. Behavioral adjustment
The residents were interviewed: what do they often do when they feel warm or hot? The lists include opening windows, taking off clothes, etc. They may choose as they used to do. Fig. 11 shows the distribution of behavioral adjustments when respondents felt hot in residential buildings.
Fig. 11. Distribution frequency of adaptive adjustments in residential buildings. Through the heating period, the residents commonly improved the indoor environment by
Page 11 of 49
opening window in warm environment. Before and after the space heating, opening window could enhance natural ventilation and improve thermal comfort. However, in the heating period, opening window would cause energy waste for a hot indoor environment.
ip t
As seen in Fig. 11, 37.4% of the subjects opened windows when they felt hot in EH, so did 37.4% in MH and moreover 42.0% in LH. Therefore, the over high indoor air temperature led to a
cr
waste in winter.
an
5.1. Time scale of air temperature in thermal comfort model
us
5. Discussion of thermal comfort
Fig. 12. Indoor temperatures in a month.
M
Fig. 12 gives the daily mean indoor temperatures in a month, which changed a little from
ed
25.8°C to 26.9°C, with a standard deviation of 0.32°C, which means that the difference of mean temperatures between 7d and 30d is very similar. Therefore, the maximum time scale of 7d was
pt
adopted.
Ac ce
Residents’ mean thermal sensation (MTS) votes and the current temperature were linear fitted, as well as average air temperatures of 1d, 2d, 3d, 4d, 5d, 6d, 7d. The correlation coefficients of MTS and average air temperatures of different time scales were worked out, shown in Table 7. According to the statistics in Table 7, the correlation coefficients R2 of different time scales in each phase varied much, with distinct changing trends. By comparing the correlation coefficients R2 of different time scales, it is found that the correlation coefficient of 7d was usually the largest, which means that MTS has correlates with average air temperatures of 7d much well. This finding complies with conclusions in natural buildings [17]. Therefore, the average air temperatures of 7d were taken as an index for evaluating thermal environments in residential buildings.
Page 12 of 49
Table 7 Correlation coefficients R2 of different time scales. 5.2. Thermal neutral temperature and thermal adaptation We weighted the regression model by the number of samples falling within the each temperature bin.
cr
temperature in three phases (EH, MH and LH) are expressed as linear models.
ip t
In Fig. 13, the relationships of MTS and predicted mean votes (PMV) with indoor air
Fig. 13. Relationships of MTS and PMV with indoor air temperature in three phases.
us
Fig. 13 illustrates a comparison between MTS and PMV. There existed some discrepancy in
an
each period. To compare the difference between MTS and PMV, the two groups of each phase were analyzed with paired t-test. It is found that the p values of EH, MH and LH are 0.0184,
M
0.8667 and 0.1608, respectively, which means that there is a significant difference between MTS
ed
and PMV in EH, and MTS is not significantly different with PMV at the 0.05 level in MH and LH. Especially, the difference was the largest in EH period. With space heating processing, the
pt
difference became smaller in accord with PMV, which means that people get accustomed to the
Ac ce
warm environment. As known in Fig.1, there are about 30 days from EH to MH. So it is inferred that the adaptation period is about 30 days from non-heating to heating environments. Through Fig. 13, the thermal neutral temperatures in different phases were figured out, which are 21.6℃, 23.5℃ and 23.1℃, respectively. The neutral temperatures in winter and spring in classrooms were 22.6℃ and 21.7℃, respectively [7], which shows the same trend with neutral temperatures in residential buildings from EH to LH. Table 8 shows the comparison between thermal neutral temperatures and mean air temperatures of different phases, which suggests a significant discrepancy between the two temperatures. Table 8 Thermal neutral temperatures and mean air temperatures of different phases.
Page 13 of 49
After space heating began, the indoor air temperature increased significantly. Hence, most respondents did not adapted to the thermal environment, and felt warm. And this resulted in a low thermal neutral temperature, with a difference of 2.0℃. In MH, the outdoor temperatures
ip t
decreased, while the mean indoor air temperature went up by 0.7℃, and the neutral temperature is
cr
0.8℃ lower than the mean air temperature. Because the respondents built up some adaptation to
the indoor environment, the thermal neutral temperature increased by 1.9℃. In LH, the outdoor
us
temperatures rebounded, and the mean indoor air temperature still increased by 0.7℃. The thermal
an
neutral temperature was 1.9℃ lower than the mean air temperature.
Table 9 shows the slopes and confidence intervals of regression models for MTS and PMV. In
M
EH, the slope of MTS regression model was larger than PMV regression model. In MH and LH,
ed
the slopes of MTS regression model became lower than PMV regression model. In general, with heating underway, respondents’ thermal sensitivity decreased and adaptation to the warm
pt
environment formed.
Ac ce
Table 9 Slopes and confidence intervals of regression models for MTS and PMV In a hot environment, human core temperature as well skin temperature is regulated to a higher level after ‘heat acclimation.’ [18]. More specifically, stronger physiological responses and more discomfort happened when people who acclimated to neutral- warm indoors were exposed in moderate cold environment than those adapted to the cold [19]. Therefore, the neutral temperature increased with the mean air temperature increasing in the heating period, which indicates that the residents got used to the indoor thermal environment gradually. On the other hand, the neutral temperatures were always below mean air temperatures, suggesting overheating in residential buildings, which might cause discomfort and energy wastes. From the view of health, once
Page 14 of 49
residents’ dependence to the indoor warm environment is built up, resistibility to the cold outdoor climate would be weakened. As a result, in addition to energy wastes, overheating would be against health. This finding further confirms the results presented by Cao et al in reference [4].
ip t
The residential buildings were constructed of brick and concrete. The performance of envelope insulation has been improved since 2005, so did the efficiency of central heating system. The
cr
indoor air temperature will keep climbing up in the newly-built buildings in the near future, which
5.3. Thermal adaptation between the elder and the youth
us
may induce a higher thermal neutral temperature.
an
An investigation in university dormitories was carried out in the same year to study the
M
college-students’ thermal comfort [11]. “The youth group” was referred to the students. We compared the mean thermal sensations between the youth in university dormitories and the elder
ed
in apartments, shown in Fig. 14. And the slopes and confidence intervals of the regression models are given in Table 10. The thermal neutral temperature for elder is higher than the temperature for
pt
the youth. Young people may have stronger adaptation to cool exposure.
Ac ce
The slope of MTS regression model for youth is the largest in EH, which suggests that young people are more sensitive to temperature upward. In MH and LH, the significant-reduced slope demonstrates that thermal sensitivity of young people decreased due to a stronger adaptation. From EH to FH, the difference of MTS between the youth and the elder got smaller. It may infer that thermal adaptation bridge the gap of age on thermal comfort. Fig. 14. Comparison of MTS between the youth and the elder in three phases. Table 10 Slopes and confidence intervals of regression models of MTS for youth and elder 5.4. Indoor air temperature and human health
Page 15 of 49
When temperatures fall into the comfort zone or beyond the zone, dryness and symptoms of sick building syndrome obviously became severe with temperature increase [20, 21]. On the contrary, a reasonable temperature drop could reduce symptoms of sick building syndrome
ip t
effectively [22]. A temperature drop of 2-3℃would double the perceived air quality [23]. Raised temperatures and air velocities, and the low indoor humidity levels in winter make the eye more
cr
sensitive to airborne particulates and other forms of pollution, which results in eye redness and
us
chronic eye ache [24, 25]. Moreover, a causal link between temperature and adiposity seems to be proved [8]. In addition, once people adapted to the overheated environment, their adaptability to
an
the cold climate would be weakened, which may cause health problems [4].
of comfort range may be healthier.
M
In general, over-high temperatures cause a lot of health problems. The lower limit temperature
ed
5.5. Energy analysis The increase rate of energy was worked out based on indoor air temperature. Fig. 15 shows
pt
distribution frequencies of air temperature in the whole heating period.
Ac ce
Fig. 15. Distribution frequency of indoor air temperature in 10 apartments.
An Equation for energy increase rate is given as follows:
∆N =
18
ti − tb × α i b − tw
∑ t i =1
(3)
Where ∆N is the energy increase rate, ti is the median of each interval, tb is the basic indoor air temperature, tw is the outdoor design temperature，equal to −24.2℃. According to Fig. 15 and Equation (3), 10.14% energy could be saved when the indoor temperature dropped from the operating conditions to 20℃. In general, in addition to energy increase, overheating in residential buildings would cause
Page 16 of 49
energy wastes for opening windows and discomfort for cold wind into the room. Moreover, it would weaken residents’ adaptation to the cold outdoor climate. Therefore, the indoor temperature should be kept at the lower limit of the comfort range in winter to keep residents’ adaptation and
ip t
realize sustainable building designs.
6. Conclusions
cr
(1) In EH, MH and LH, the mean indoor air temperatures were 23.6℃, 24.3℃, 25.0℃. And the
us
frequencies of air temperatures above 24℃ were 47.5%, 57.1% and 57.0%, respectively. This result shows that there did exist overheating in residential buildings.
an
(2) When 50% of the indoor air temperatures were higher than 24℃, people may accept the
M
temperature and voted for “neutral”, “slightly warm” or “warm”. But when 95% of the temperatures were higher than 24℃ beyond the adaptation range, people did not accept it and
ed
voted for “hot”.
(3) The thermal neutral temperature has an upward tendency with the increasing indoor air
pt
temperature, which indicates that the residents got used to the indoor thermal environment
Ac ce
gradually. On the other hand, the neutral temperatures were always below mean air temperatures, suggesting overheating in residential buildings.
(4) In the heating period, about 20% respondents felt warm (vote for +2 or +3) and above 30% respondents opened windows to release overheat. The clothing insulation decreased with the heating period processing.
(5) Due to high temperatures indoors, the residents commonly adapt to the indoor environment by opening window or taking off clothes. However, it would cause energy increase and discomfort. Therefore, the indoor temperature should be kept at the lower limit of the comfort
Page 17 of 49
range in winter to keep residents’ adaptation and realize sustainable building designs.
Acknowledgement All the participants are sincerely acknowledged in this research.
ip t
Funding The work presented in this paper was funded by the National Natural Science Foundation of
cr
China (No. 51278142).
us
References
[1]J.F. Nicol, M.A. Humphreys, Adaptive thermal comfort and sustainable thermal standards for
an
buildings, Energy and Building 34 (6) (2002) 563-572.
M
[2] D. Mumovic, J. Palmer, M. Davies, M. Orme, I. Ridley, T. Oreszczyn, C. Judd, R. Critchlow, H.A. Medina, G. Pilmoor, C. Pearson, P.Way, Winter indoor air quality, thermal comfort and
44(7) (2009) 1466-1477.
ed
acoustic performance of newly built secondary schools in England, Building and Environment
pt
[3] M Kavgic, Summerfield A, Mumovic D, Z.M. Stevanovic, V. Turanjanin, Z.Z. Stevanovic,
Ac ce
Characteristics of indoor temperatures over winter for Belgrade urban dwellings: Indications of thermal comfort and space heating energy demand, Energy and Buildings 47(4)(2012) 506-514.
[4] B. Cao, Y.X. Zhu, Q. Ouyang, X. Zhou, L. Huang, Field study of human thermal comfort and thermal adaptability during the summer and winter in Beijing, Energy and Buildings 43(5)(2011) 1051-1056. [5] Z.J. Wang, G. Wang, L.M. Lian, A field study of the thermal environment in residential buildings in Harbin, ASHRAE Transaction 109(2)(2003) 350-355.
Page 18 of 49
[6] Z.J. Wang, L. Zhang, J.N. Zhao, Y.N. He, A.X. Li, Thermal responses to different residential environments in Harbin, Building and Environment 46(11) (2011) 2170-2178. [7] Z.J. Wang, A.X. Li, J. Ren, Y.N. He, Thermal adaptation and thermal environment in
ip t
university classrooms and offices in Harbin, Energy and Buildings 77(7)(2011) 192-196. [8] A. Mavrogiannia, F. Johnsonb, M. Uccic, A. Marmotc, J. Wardleb, T. Oreszczyna, A.
cr
Summerfielda, Historic variations in winter indoor domestic temperatures and potential
us
implications for body weight gain, Indoor and Built Environment 22(2) (2013) 360-375. [9] M.H. Luo, R.J. de Dear, W.J. Ji, B. Cao, B.R. Lin, Q. Ouyang, Y.X. Zhu, The dynamics of
an
thermal comfort expectations: The problem, challenge and implication, Building and
M
Environment 95(1) (2016) 322-329.
[10] R.J. de Dear, G.S. Brager, Developing an adaptive model of thermal comfort and preference.
ed
ASHRAE Transactions 4 (1) (1998) 145–167. [11] H.R. Ning, Z.J. Wang, X.X Zhang, Y.C. Ji, Adaptive thermal comfort in university
pt
dormitories in the severe cold area of China, Building and Environment 99(5) (2016) 161-169.
Ac ce
[12] W.B. Meyer, Why indoor climates change: a case study: Climate Change 55(3) (2002) 395–407.
[13] The ministry of construction, Design code for Heating ventilation and air conditioning of civil buildings, GB50736-2012, (2012) (In Chinese). [14] ASHRAE.ANSI/ASHRAE standard 55 Thermal environmental conditions for human occupancy. Atlanta: ASHRAE, 2013. [15] CEN, Indoor environmental input parameters for design and assessment of energy performance of buildings–addressing indoor air quality, thermal environment, lighting and
Page 19 of 49
acoustics, EN 15251, European Committee for Standardization, Brussels, 2007. [16] B. Cao, Q. Ouyang, Y. Zhua, L. Huang, H. Hu, G. Deng, Development of a multivariate regression model for overall satisfaction in public buildings based on field studies in Beijing
ip t
and Shanghai, Building and Environment 47(1) (2012) 394–399. [17] C. Morgan, R. J. de Dear, Weather, clothing and thermal adaptation to indoor climate,
cr
Climate Research 24(3) (2003) 267-284.
us
[18] D. Wang, H. Zhang, E. Arens, C. Huizenga, Observations of upper-extremity skin temperature and corresponding overall-body thermal sensations and comfort, Building and
an
Environment 42(12) (2017) 3933-3943,
M
[19] M.H. Luo, B. Cao, Ouyang, Y.X. Zhu, Indoor human thermal adaptation: dynamic processes and weighting factors, Indoor Air 27(2) (2017) 273-281.
ed
[20] J.J.K. Jaakkola, O.P. Heinonen, Sick building syndrome, sensation of dryness and thermal comfort in relation to room temperature in an office building: Need for individual control of
pt
temperature, Environment International 15(1-6) (1989) 163-168.
Ac ce
[21] A.L. Krogstad, G. Swanbeck, L. Barregard, S. Hagberg, K.B. Rynell, A. Ran, et al., A prospective study of indoor climate problems at different temperatures in offices(in Swedish), Gothenburg: Volvo Truck Corporation. [22] D.P. Wyon, Sick buildings and the experimental approach, Environmental Technology 13(4) (1992) 313-322. [23] P.O. Fanger, What is IAQ? // Proceedings of the 10th International Conference on Indoor Air Quality and Climate. Beijing, 2005: 1-8. [24] C. Franck, Eye symptoms and signs in buildings with indoor climate problems (‘office eye
Page 20 of 49
syndrome’), Acta Ophthalmologica 64(3) (1986) 306-311. [25] N.M. Wyon, D.P. Wyon, Measurement of acute response to draught in the eye, Acta
Ac ce
pt
ed
M
an
us
cr
ip t
Ophthalmologica 65(4) (1987) 385-392.
Page 21 of 49
Figure captions Fig. 1. Outdoor temperature and phase division of heating period. Fig. 2. Hourly mean temperatures in No.2 and No.6 apartments.
ip t
Fig. 3. Distribution frequency of indoor air temperature in 10 apartments in three phases. Fig. 4. Mean daily air temperature and RH in apartments.
cr
Fig. 5. Contrast between air temperature and operative temperature.
us
Fig. 6. Mean clothing insulation in three phases. Fig. 7. Distribution frequency of TSV.
an
Fig. 8. Air temperature distribution and clothing insulation at different TSV.
Fig. 10. Distribution frequency of TCV.
M
Fig. 9. Distribution frequency of TPV.
ed
Fig. 11. Distribution frequency of adaptive adjustments in residential buildings. Fig. 12. Indoor temperatures in a month
pt
Fig. 13. Relationships of MTS and PMV with indoor air temperature in three phases.
Ac ce
Fig. 14. Comparison of MTS between the youth and the elder in three phases. Fig. 15. Distribution frequency of indoor air temperature in 10 apartments.
Page 22 of 49
Table captions Table 1 Building Information. Table 2 Participants’ backgrounds. Table 3 Test instruments and accuracy.
ip t
Table 4 Vote scales of thermal response.
cr
Table 5 Mean indoor air temperature and RH in three phases.
Table 7 Correlation coefficients R2 of different time scales.
us
Table 6 Statistical results of the other physical parameters.
an
Table 8 Thermal neutral temperatures and mean air temperatures of different phases. Table 9 Slopes and confidence intervals of regression models for MTS and PMV.
Ac ce
pt
ed
M
Table 10 Slopes and confidence intervals of regression models for MTS for youth and elder.
Page 23 of 49
ip t cr
Ac ce
pt
ed
M
an
us
Fig. 1. Outdoor temperature and phase division of heating period.
Page 24 of 49
ip t cr us an
Ac ce
pt
ed
M
a. No.2 apartment
b. No.6 apartment
Fig. 2. Hourly mean temperatures in No.2 and No.6 apartments.
Page 25 of 49
ip t cr us
Ac ce
pt
ed
M
an
Fig. 3. Distribution frequency of indoor air temperature in 10 apartments in three phases.
Page 26 of 49
ip t cr
Ac ce
pt
ed
M
an
us
Fig. 4. Mean daily air temperature and RH in apartments.
Page 27 of 49
ip t cr us
Ac ce
pt
ed
M
an
Fig. 5. Contrast between air temperature and operative temperature.
Page 28 of 49
ip t cr us
Ac ce
pt
ed
M
an
Fig. 6. Mean clothing insulation in three phases.
Page 29 of 49
ip t cr
Ac ce
pt
ed
M
an
us
Fig. 7. Distribution frequency of TSV.
Page 30 of 49
ip t cr us
Ac ce
pt
ed
M
an
Fig. 8. Air temperature distribution and clothing insulation at different TSV.
Page 31 of 49
ip t cr
Ac ce
pt
ed
M
an
us
Fig. 9. Distribution frequency of TPV.
Page 32 of 49
ip t cr
Ac ce
pt
ed
M
an
us
Fig. 10. Distribution frequency of TCV.
Page 33 of 49
ip t cr
Ac ce
pt
ed
M
an
us
Fig. 11. Distribution frequency of adaptive adjustments in residential buildings.
Page 34 of 49
ip t cr
Ac ce
pt
ed
M
an
us
Fig. 12. Indoor temperatures in a month
Page 35 of 49
ip t
b) MH
M
an
us
cr
a) EH
ed
c) LH
Ac ce
pt
Fig. 13. Relationships of MTS and PMV with indoor air temperature in three phases.
Page 36 of 49
M
an
us
cr
b) MH
ip t
a) EH
ed
c) LH
Ac ce
pt
Fig. 14. Comparison of MTS between the youth and the elder in three phases.
Page 37 of 49
ip t cr us
Ac ce
pt
ed
M
an
Fig. 15. Distribution frequency of indoor air temperature in 10 apartments.
Page 38 of 49
Table 1 Building Information. No.
Construction year of
Floor
Room
level
Orientation
Function
of exterior
buildings
Floor area of 2
Number of
room / m
participant
wall
2001
7
Living Room
South
34.4
2
2
2001
7
Living Room
East
15.1
2
3
1999
4
Living Room
South
36.2
2
4
1999
3
Living Room
South
23.6
5
1996
5
Living Room
South
14.1
6
1996
5
Bedroom
South
11.1
2
7
1993
7
Living Room
South
32.4
2
8
2000
7
Living Room
South
9
2001
4
Living Room
West
10
2001
7
Living Room
Nortth
us
cr
ip t
1
2 1
2
41
3
32
2
Ac ce
pt
ed
M
an
34.6
Page 39 of 49
Table 2 Participants’ backgrounds. Number of participants Mean Std. Dev. Max Min
Years in Harbin 39.7 20.1 70 13
Ac ce
pt
ed
M
an
us
cr
ip t
20
Age 48.5 13.6 72 28
Page 40 of 49
Table 3 Test instruments and accuracy. Parameter
Range
Accuracy
Notes
Air temperature
-40~100℃
±0.5℃
Continuous
Relative humidity
0~100%
±3%
test
HWZY-1
Globe temperature
-50~100℃
±0.4℃
Testo425
Air speed
0~20m/s
±0.03m/s
Testo830-T
Surface
-30~400℃
±1.5℃
WSZY-1A
meter Globe thermometer Hot wire anemometer Infrared
1
temperature
Manual test
Manual test
Ac ce
pt
ed
M
an
us
thermometer
Manual test
ip t
Thermo-hygro
Type
cr
Instrument
Page 41 of 49
Table 4 Vote scales of thermal response. Scale
Thermal Sensation
-3 cold, -2 cool, -1 slightly cool, 0 neutral, +1 slightly warm, +2 warm, +3 hot
Thermal Preference
-1 cooler, 0 no change, +1 warmer
Thermal Comfort
0 comfortable, 1 slightly uncomfortable, 2 uncomfortable, +3 unbearable
Ac ce
pt
ed
M
an
us
cr
ip t
Vote
Page 42 of 49
Table 5 Mean indoor air temperature and RH in three phases. EH
MH
LH
Whole period
Mean air temperature (℃)
23.6
24.3
25.0
24.3
Mean RH (%)
46.9
34.6
34.7
37.2
Ac ce
pt
ed
M
an
us
cr
ip t
Parameter
Page 43 of 49
Table 6 Statistical results of the other physical parameters MH
LH
Mean
0.05
0.03
0.03
Std. Dev.
0.033
0.035
0.021
Max
0.12
0.21
0.09
Min
0.01
0.00
0.01
Mean
23.6
24.3
Std. Dev.
2.16
2.65
Max
28.7
35.1
Min
19.8
Mean
18.2
Surface temperature of exterior
Std. Dev.
3.91
window / ℃
Max
27.9
Min
12.2
Mean
24.6
2.16
28.1
cr
Globe temperature / ℃
20.5
20.3
17.7
23.2
3.47
4.04
27.8
30.7
13.0
15.8
21.6
23.2
24.0
2.42
2.75
2.01
us
Air speed / m/s
ip t
EH
an
Parameter
Std. Dev.
wall / ℃
Max
27.0
28.3
27.4
Min
13.4
16.9
19.6
Mean
22.8
23.8
24.4
Surface temperature of interior
Std. Dev.
2.65
3.08
2.68
wall / ℃
Max
28.6
30.2
36.5
Min
17.4
18.8
20.1
Mean
23.1
23.8
24.0
Std. Dev.
3.97
5.14
2.93
Max
31.5
36.0
30.0
Min
18.0
18.0
18.5
Mean
22.6
23.7
24.2
Std. Dev.
3.45
3.53
2.40
Max
30.0
31.5
28.5
Min
18.0
19.5
21.0
Ac ce
pt
ed
M
Surface temperature of exterior
Surface temperature of floor / ℃
Surface temperature of ceiling / ℃
Page 44 of 49
Table 7 Correlation coefficients R2 of different time scales. EH
MH
LH
Current 1d 2d 3d 4d 5d 6d 7d
0.6126 0.4972 0.5706 0.5029 0.4718 0.5016 0.5022 0.6009
0.3103 0.3894 0.4176 0.4874 0.3503 0.4488 0.3907 0.5017
0.4772 0.5131 0.5114 0.5089 0.3865 0.4801 0.5430 0.5743
Ac ce
pt
ed
M
an
us
cr
ip t
Time scale
Page 45 of 49
Table 8 Thermal neutral temperatures and mean air temperatures of different phases. EH
MH
LH
Average air temperature ta(℃)
23.6
24.3
25.0
Thermal neutral temperature tn(℃)
21.6
23.5
23.1
ta - tn (℃)
2.0
0.8
1.9
Ac ce
pt
ed
M
an
us
cr
ip t
Parameter
Page 46 of 49
Table 9 Slopes and confidence intervals of regression models for MTS and PMV MH
LH
0.2327
0.1456
0.1733
Lower limit
0.1819
0.1148
0.1358
Upper limit
0.2835
0.1765
0.2108
0.2108
0.1888
0.2143
Lower limit
0.1742
0.1676
0.1793
Upper limit
0.2473
0.2010
Slope MTS 95% confidence intervals Slope PMV
0.2494
Ac ce
pt
ed
M
an
us
cr
95% confidence intervals
ip t
EH
Page 47 of 49
Table 10 Slopes and confidence intervals of regression models for MTS for youth and elder
95% confidence interval
Lower limit Upper limit
Slope 95% confidence interval
Lower limit Upper limit
LH 0.1367 0.1038 0.1695 0.1733 0.1358 0.2108
Ac ce
pt
ed
M
an
us
cr
MTS(elder)
MH 0.1770 0.1503 0.2037 0.1456 0.1148 0.1765
ip t
Slope MTS(youth)
EH 0.4105 0.3699 0.4512 0.2327 0.1819 0.2835
Page 48 of 49
Highlights
Overheating in residential buildings in severe cold area of China were found.
2.
The heating periods were divided into three phases and three adaptive models were given.
3.
Time scale of air temperature in thermal comfort model were discussed.
4.
Frequency distribution histogram of air temperature was used to analyze energy
cr
ip t
1.
Ac ce
pt
ed
M
an
us
consumption.
Page 49 of 49