A Systematic Review on the Studies of Thermal Comfort in Urban Residential Buildings in China

Abstract

There have been fruitful publications on thermal comfort of urban residential buildings in China. However, there is a lack of reviews on this topic to perform a comprehensive analysis and find opportunities to meet occupants’ thermal comfort needs while improving building energy efficiencies. This paper addresses this issue by presenting a systematic review on the advancements in research on thermal comfort in urban residential buildings in China. Firstly, two common thermal comfort research approaches, i.e., field studies and laboratory studies, are discussed. Secondly, eleven main thermal comfort evaluation indicators are summarized. Finally, this paper analyzes the thermal comfort survey data from different researchers, discusses the impacts of adaptive behaviors on human thermal comfort, and provides recommendations for future research on urban residential thermal comfort. It was found that people have higher and higher requirements for their indoor thermal environment as time goes by, especially in the winter; the thermoneutral temperature is higher in warmer climate regions in the summer but lower in the winter than in colder climate regions; the thermoneutral temperature tends to increase with the indoor air temperature due to an adaptation to the indoor thermal environment. The outcomes of this paper provide valuable information on thermal comfort behaviors of urban residents in different climate zones in China, which can serve as a resource for the academic community conducting future research on thermal comfort and assist policymakers in enhancing building energy efficiencies without compromising the occupants’ comfort.

1. Introduction

China has been experiencing rapid urbanization in recent years, with 59.2% of its population living in urban areas in 2018, and this is expected to reach 70.6% in 2030 and 80% in 2050 [1]. The rapid urbanization creates a climate warming effect, which has a negative impact on the thermal comfort of urban residents [2]. While a good indoor thermal environment has a positive impact on people’s work and life [3,4,5,6], an uncomfortable indoor thermal environment not only decreases the occupants’ learning and work efficiencies [7], but also poses a threat to human health [8,9]. The operation of a heating, ventilation, and air-conditioning (HVAC) system can help provide a comfortable indoor environment. However, it leads to a very high building energy consumption. In 2020, residential electricity consumption accounted for 16.7% of the total national electricity consumption in China [10], and building energy consumption accounted for 45.5% of the total national energy consumption [11]. Reducing building energy consumption while maintaining indoor thermal comfort is still very important. Therefore, there is a need to conduct research on occupants’ thermal comfort conditions in different climatic regions, develop relevant thermal comfort standards, and optimize the room temperature setpoints to reduce the energy needs from the heating, ventilation, and air-conditioning (HVAC) systems of residential buildings.

There are five climatic regions in China, which are severe cold, cold, hot summer and cold winter, hot summer and warm winter, and mild regions [12]. People in diverse climatic regions experience thermal comfort differently. The earliest publication found from the literature survey was from Xia et al. [13] in 1999, who conducted a survey on the thermal comfort of residential buildings in Beijing and found the neutral temperature of the residents to be 26.7 °C. Since then, studies on residential thermal comfort in urban areas have covered all the climatic regions in China with fruitful publications. Some studies were conducted in one city in a certain climatic region [13,14,15,16,17,18,19], and others were conducted in multiple cities or multiple climatic regions [20,21,22].

The research on the thermal comfort of urban residential buildings abroad started earlier than that in China. In 1970s, Aynsley conducted research on how to use natural airflow movement to improve thermal comfort in hot, humid, tropical housing in Sydney, Australia [23]. In 1999, Nicol et al. [24] conducted a thermal comfort survey in five cities located in different climatic regions in Pakistan and found that there was a clear relationship between indoor comfort and outdoor environmental conditions. In 2009, Bae and Chun [25] studied the indoor thermal environment and residents’ control behaviors on the heating and cooling system in Seoul, South Korea, and found that occupants preferred higher temperatures in the winter and lower temperatures in the summer over the past 25 years. In 2016, Singh et al. [26] performed an assessment on the thermal comfort status of existing pre-1945 residential buildings in Liège, Belgium, and found that modern comfort standards lead to an underestimation of the occupants’ thermal comfort in these buildings. In 2021, Rijal et al. [27] investigated the thermal comfort of three apartments in Japan and developed the corresponding adaptive model and found that the seasonal variation of the comfort temperature and regression coefficient of the adaptive model was smaller than that of a detached house.

Multiple literature surveys on this research topic have been found [28,29,30,31,32,33,34]. Enescu [28] conducted a detailed review on various thermal comfort indices and thermal comfort models in 2017. Wang et al. [29] reviewed the differences in the thermal comfort needs of different genders and ages in 2018. Sakiyam [30] presented a review on the thermal comfort in naturally ventilated buildings in 2020. Zhao et al. [31] conducted a literature survey on the theoretical basis of a thermal comfort model, some different classical thermal comfort models, including the PMV-PPD model, two-node model, multi-node model, as well as a machine learning algorithm-based model. Yao et al. [32] conducted a survey on thermal environment assessment methods and divided these into thermal balance methods, adaptive regression methods, and adaptive thermal balance methods as the three representative methods. They also analyzed the advantages and limitations of each method. Feng et al. [33] outlined the research methods and procedures of personal thermal comfort prediction, focusing on the experimental design, data collection and modeling, and the development of personal thermal comfort prediction models. de Dear et al. [34] introduced adaptive thermal comfort codes and standards in different countries and also analyzed the differences in adaptive comfort between different building types and found that the current standards are insufficient to reflect the thermal adaptations of occupants in different building types, especially in residential and educational buildings.

The above literature reviews provide valuable information on the advancements in thermal comfort research. However, they are not specifically for residential thermal comfort. In addition, these reviews did not address thermal comfort in different climate zones and seasons. Furthermore, thermal comfort sensations are affected by many factors such as culture, race, and more importantly, the type of air-conditioning system in residential buildings (e.g., split units and variable refrigerant flow (VRF) units are dominant in residential buildings in China). Finally, thermal comfort sensations in China are different from those of other countries, e.g., both the indoor operative temperature and neutral temperature in the cities of China were the lowest among the European, Chinese, and North American cities [35]. Therefore, there is a need to conduct a systematic review on the research and advancements in the thermal comfort of urban residential buildings in China. This paper first discusses the research methods of thermal comfort, including field research and laboratory research. Secondly the thermal comfort evaluation indices are evaluated. Finally, the relationships between the thermal neutral temperature and clothing thermal resistance, thermal neutral temperature, as well as outside air temperature, and the effects of adaptive behaviors on human thermal comfort are presented. The recommendations of the research on indoor thermal comfort in urban residential areas are also provided. Figure 1 shows the structure of this paper. This paper contributes to providing an in-depth analysis of urban residential thermal comfort under various climates and seasonal changes in China. It serves as a reference for conducting research on urban residential thermal comfort by comprehensively summarizing the research methods, evaluation indicators, and adaptive models. Additionally, it analyzes neutral temperatures, acceptable temperatures, and other data from various cities and climate zones in China. This provides a more comprehensive and specific understanding of thermal comfort and energy efficiency for both the academic community and policymakers.

2. Methods

In order to understand the research on thermal comfort in urban residential buildings in China, this paper collects all the available literature data from relevant databases including ScienceDirect, Google Scholar, Web of Science, CNKI, and Wanfang. The keywords used for searching are “China”, “residential building”, “thermal comfort”, “thermal sensation”, “thermal adaptation”, “thermal environment”, and “neutral temperature”. The selected keywords ““China” and “residential building” were applied in all the searches, while the other keywords were applied one at a time to ensure the maximum number of related papers were found. However, only those papers that conduct related research in thermal comfort of urban residential buildings in China with original contributions were selected. This review mainly focuses on the studies of thermal comfort under real conditions, especially field studies, so that the outcomes of the studies can be used to guide practical applications directly. Although CFD-related works are also important, they are out of the scope of this study and were therefore excluded. Figure 2 illustrates the selection criteria. A total of 468 research papers were identified in the first round of search. Figure 3a presents the publication trend regarding the number of papers. As can be observed, before 2012, only a few papers were published per year, and a rapid increase in the number of publications can be seen after 2012. The proportion of papers published from 2019 to 2023 accounts for 61% of the total publications, which means that with social and economic development, people are paying more and more attention to the indoor thermal environmental conditions, leading to more research outputs on thermal comfort. It also means that there is a need to review and update the research outcomes from the investigations to provide guidelines for air-conditioning control on indoor thermal comfort.

This paper also searched the relevant literature on theories, methods, norms, and standards related to thermal comfort, and collected data related to the urban population from relevant government websites. After careful examination, the literature that did not match the topic of the review and the literature that was of poor quality were excluded, leaving 133 publications in the last round. It was found that the literature mainly covers the investigations and analyses of indoor thermal sensations and thermal comfort, the behavior adjustments of occupants, and comparative studies with other researchers. Figure 3b presents the geographical distribution of the publications. It can be found that all the climatic regions are covered in the literature. The relevant cities investigated were mainly distributed in Heilongjiang, Beijing, Jiangsu, Henan, and Shaanxi, and the proportion of publications in those five provinces accounts for 51.4% of the total number of publications. Furthermore, the related research from some provinces is included in the associated climatic region and is not displayed in the figure.

3. Research Methods for Thermal Comfort

There are two main research methods for thermal comfort, i.e., laboratory-based studies and field studies [36]. Laboratory studies are conducted in artificially constructed laboratories, and this approach isolates the key environmental parameters that make up the indoor thermal environment, including the air temperature, mean radiant temperature, humidity, and air velocity. The relevant personal parameters are clothing thermal resistance and metabolic activity level. These environmental and personal parameters can be combined into a thermal environmental index that quantifies or predicts the subjective degree of thermal discomfort experienced by experimental subjects. Field studies were conducted under normal conditions in actual buildings and involved larger and more diverse samples of occupants rather than “paid-fee college students” [37]. Both methods have been conducted by the Chinese researchers and are reviewed in this paper.

3.1. Field Studies

The purpose of the field studies is to investigate the thermal comfort in the real world [38], and the survey on indoor thermal comfort generally refers to the ASHRAE Standard 55 [39]. The survey objects include a variety of building types, and the survey content and results of different building types are quite different. For example, the thermal comfort feeling from the elderly in nursing homes is quite different from young people because of their decreased metabolic rate and tolerance to cold and hot environments [40]. The thermal sensation and neutral temperature perception also differs between men and women [41,42]. When selecting occupants for thermal comfort studies in indoor environments, a proper experimental design and standardized procedures should be followed [43].

3.1.1. Procedure of Field Studies

With this method, the research is carried out using a questionnaire plus on-site measurements [44]. During the on-site measurements, thermal environmental data such as indoor and outdoor air temperature, relative humidity, air flow velocity, and radiation temperature, and physical data such as the building structure and room size are collected. The questionnaire collects basic information on the occupants, their thermal sensations, clothing of the human body, habits of using air conditioners, etc. The surveys last from a few months to several years, and the scope of the survey is usually a city or climate region, although some large-scale thermal comfort studies include several climate regions. Statistical regression is used to analyze and process the collected data, and results of the regression equation for thermal sensation voting based on the indoor temperature, neutral temperature, and other data can be obtained. Figure 4 presents the flow chart of the approach based on field studies. It can be seen that after the research subjects are determined, the data needed from the questionnaire and field measurements are identified. Then, the questionnaires are released to obtain the thermal comfort sensation-related parameters, and field measurements are conducted to obtain the environmental parameters. Finally, data processing and analysis are performed. The design and release of the questionnaires are the key procedures of this approach. A reasonable setting of the questionnaire can help obtain people’s thermal sensations more accurately and reduce the error occurrence.

3.1.2. Survey

Thermal sensation is a subjective description of whether the surrounding environment is “cold” or “hot”. It cannot be directly measured and can only be obtained through questionnaires using a grade scale. ASHRAE’s seven-point scale grading is commonly adopted in thermal comfort research [39]. Most thermal comfort surveys use subjective questionnaires to directly obtain the thermal sensations of the subjects. The items in the questionnaire mainly include personal information such as gender, age, height, and body weight [45,46,47] and subjective survey parameters such as thermal sensation [46,47,48,49], wet sensation [15,50,51], air flow sensation [46,50,52], thermal comfort [14,15,50], thermal acceptability [47,53,54], thermal preference [14,15,17], etc. The survey also included other information such as clothing insulation [51,55,56], duration [55,56,57], economic condition [16,58], activity level [45,47,51], education level [16,58], etc. In order to study the impact of occupant behavior on thermal comfort, researchers also collect data on people’s adjustments to the thermal environment, such as window opening [53,54,59], air-conditioning usage [16,41,60], adaptation measures [14,16,17], and so on. Figure 5 presents the statistical data on the items included in the questionnaire from the researchers. It can be observed that most researchers mainly include subjects’ personal information, thermal sensation, thermal acceptability, thermal preference, clothing insulation, activity level, and adaptation measures, and few researchers includes other items in the survey.

The respondents filled out the questionnaire in a natural and psychologically relaxation state. Most thermal sensation surveys use the ASHRAE seven-point scale (−3, cold; −2, cool; −1, slightly cool; 0, neutral; 1, slightly warm; 2, warm; and 3, hot) [39]. A seven-point scale similar to the thermal sensation vote index was developed for the humid sensation and air flow sensation.

Table 1. Items included in the thermal comfort questionnaires.

Item	Subjective Evaluation	Ref.
Thermal sensation	−3, cold; −2, cool; −1, slightly cool; 0, neutral; 1, slightly warm; 2, warm; 3, hot	[54,57,58,59,61,62,63,64,65,66,67,68]
Humidity sensation	−3, very humid; −2, humid; −1, slightly humid; 0, neutral; 1, slightly dry; 2, dry; 3, very dry	[50,51]
Humidity sensation	−3, much too dry; −2, too dry; −1, slightly dry; 0, just right; +1, slightly humid; +2, too humid; +3, much too humid	[44,47,65]
Humidity sensation	3, too dry; −2, dry; −1, slightly dry; 0, just right; +1, slightly humid; +2, humid; +3, too humid	[60,67]
Wet sensation	−3, very wet; −2, wet; −1, slightly wet; 0, neutral; 1, slightly dry; 2, dry; 3, very dry	[15]
Air movement sensation or stuffy sensation	−3, very stuffy; −2, stuffy; −1, slightly stuffy; 0, neutral; 1, slightly windy; 2, windy; 3, very windy	[50,68]
Air movement sensation	−3, much too still; −2, too still; −1, slightly still; 0, comfortable and no breeze; +1, comfortable and slightly breezy; +2, too breezy; +3, much too breezy	[44,65]
Air movement sensation	−3, very breezy; −2, breezy; −1, slightly breezy; 0, neutral; 1, slightly stuffy; 2, stuffy; 3, very stuffy	[67]
Thermal comfort	−3, very uncomfortable; −2, uncomfortable; −1, slightly uncomfortable; 1, slightly comfortable; 2, comfortable; 3, very comfortable	[50,51]
Thermal comfort	0, comfort; 1, slightly discomfort; 2, discomfort; 3, high discomfort	[15]
Thermal comfort	0, comfortable; 1, slightly uncomfortable; 2, uncomfortable; +3, unbearable	[14,54]
Thermal comfort	−3, unbearable; −2, very uncomfortable; −1, uncomfortable; 0, slightly uncomfortable; 1, comfortable	[47]
Thermal acceptability	−2, totally unacceptable; 1, barely unacceptable; 1, barely acceptable; 2, totally acceptable	[47,50,51,68]
Thermal acceptability	−1, unacceptable; 1, acceptable	[15,54,61]
Thermal preference	−1, cooler; 0, no change; 1, warmer	[15,47,50,51,54]
Thermal preference	1, want warmer; 2, no change; 3, want cooler	[58]
Humidity preference	−1, lower; 0, no change; +1, higher	[44,65]
Humidity preference	−1, drier; 0, no change; 1, wetter	[47]
Wet preference	−1, wetter; 0, no change; 1, drier	[15]
Air movement preference	−1, lower; 0, no change; +1, higher	[40,44,58,65]
Adaptive strategies	Opening windows/doors, turning on fans, turning on A/C, none	[16]

lists the specific survey items found in the literature. It can be found that the scales for different parameters vary greatly among different studies. For the same thermal comfort survey item, different researchers use different evaluation indicators leading to different outcomes. Therefore, the most suitable evaluation indices should be selected in the research.

3.1.3. On-Site Measurement

Environmental parameters play important roles in thermal comfort sensations [22]. Field investigations usually use anemometers, temperature and humidity meters, and other equipment to obtain indoor and outdoor environmental parameters, usually including indoor and outdoor air temperatures, indoor and outdoor relative humidities, air velocity, black bulb temperature, mean radiant temperature, etc. Physical parameters, such as the building type, room orientation, basic building information, etc., are also recorded. Thanks to the continuous development of sensor technology, multiple parameters can be measured simultaneously using one instrument [41]. Care has to be taken while performing the field measurements. Firstly, a room is selected for measurement, such as the living room or bedroom. Secondly, the parameters to be measured are determined, such as the temperature/relative humidity and wind speed, and the corresponding measuring instruments are prepared, such as HOBO and Testo, having the desired measuring range and accuracy. Then, the measuring points are arranged and the measuring equipment is placed according to the geometric characteristics of the room. The measuring points should be reasonably set in height and density according to the actual needs to collect the data accurately. For example, Wang [41] selected three communities in Guangzhou City to conduct field measurements on the air temperature and relative humidity that lasted about two weeks for each residence. Monitoring points were placed in the bedroom, living room, and outdoors 1.1 m above the ground. The surface temperature on the exterior walls, interior walls, exterior windows, floors, and ceilings were also measured. It is worth noting that CO₂ is an important indicator affecting indoor air quality, and some studies [55] also measured the distribution of the indoor CO₂ concentration. Examples of the arrangement of the measurement points and measuring instruments are shown in Figure 6 and Figure 7.

Table 2. Items in the field measurement.

Item	Ref.
Air temperature	[45,46,47,48,49,52,53]
Relative humidity	[57,58,59,61,62,63,64,65,66]
Air velocity	[14,15,17,20,21,50,51]
Mean radiant temperature	[58,60,64,69]
Black bulb temperature	[17,46,53,56]
PMV–PPD index (thermal comfort meter)	[41,70]
Globe temperature	[14,15,21,47,50,51,71,72]
Carbon monoxide concentration	[55]
Envelope surface temperature	[14,49,57,66]
Building type	[20,21,55,58,59]
Room orientation	[20,21,41,57,73]
Basic building information	[20,21,49,55,57,73]

summarizes the items measured in the field studies. It can be found that the air temperature, relative humidity, air velocity, and mean radiant temperature are mostly measured during the on-site investigation.

Some researchers did not obtain the outdoor environmental parameters through on-site measurement, but from nearby weather stations or meteorological databases [16,55], which could be different from the local conditions. The air temperature, relative humidity, and air velocity are commonly measured by the researcher, while the rest of the parameters are not.

3.2. Laboratory Studies

In the field studies, the variations in solar radiation, indoor activities of occupants, air-conditioning operations, etc., have a disruptive effect on the assessment of thermal comfort. The laboratory studies can overcome this difficulty by controlling the environmental parameters in a test chamber [38]. In the environmental chamber, indoor environmental parameters can be precisely controlled, and the impact of each factor on human thermal perception can be separately identified. In addition, by wearing standard clothing, the difference in clothing thermal resistance among people can be eliminated to a certain extent [22]. However, laboratory studies also have obvious shortcomings. For example, the arrangement of tables, chairs, and furniture in the laboratory, the behavior of the subjects, and the control of the air-conditioning system are not representative of everyday life, and therefore, it cannot represent the behavior of most people. In addition, laboratory studies often involve only a small number of subjects, and the participants often share similar characteristics. Although the ratio of males and females was relatively balanced in almost all studies, young adults are more prevalent than children and the elderly [74]. The environmental parameters of the air-conditioned room can be changed according to the needs of the experiment. However, the subjects could only passively adapt to this change. Laboratory subjects who are not engaged in their daily work activities in a familiar environment may perceive and accept thermal environments atypically, which can affect the laboratory results [41]. Restricted by the size of the chamber, it is also difficult to conduct large-scale thermal comfort experiments. On the contrary, on-site surveys can be conducted on a very large scale, with a wide geographical range and a variety of survey samples. Therefore, when conducting laboratory studies, subjects should be properly selected to avoid a too concentrated age distribution and ensure a relatively balanced sex ratio to improve the representation of the results.

Table 3. Results from laboratory studies.

Year	Content	Conclusion	Ref.
2013	Differences in thermal adaptation to cold indoor environments between people in areas with/without heating systems	People will change their psychological expectations of the thermal environment according to the indoor environment	[75]
2014	Effects of air temperature on sleep quality and thermal comfort of sleepers	Human sleep quality is very sensitive to changes in air temperature, and current standards cannot ensure a comfortable sleep environment	[76]
2015	Investigating how controlled or uncontrolled thermal environments affect human thermal responses in the summer	About two-thirds of the subjects preferred manual controls rather than automatic controls	[77]
2016	Effect of personal control over air-conditioning system on occupant thermal comfort	The subjects’ perceived control ability to the thermal environment improves their thermal comfort; it is recommended to provide occupants with openable windows, fans, terminal controllable adjustment systems, etc., to control their thermal environment	[78]
2018	Analysis of the effects of long-term and short-term heat adaptation on human thermal regulation and perception	The high-temperature and high-humidity environment has adverse effects on the human body’s thermal perception, and the average skin temperature of the subjects has physiological adaptability to repeated thermal stimulations	[79]
2019	Comparative study on the thermal comfort of central air-conditioned buildings and buildings served by split units in South China; investigate their differences in occupants’ physiological and psychological responses under the same environmental condition; identify the thermal neutrality and acceptable temperature of air-conditioned buildings	There are no significant differences in the thermal sensations and neutral temperatures of the two buildings	[80]
2020	Influence of clothing thermal resistance on human thermal comfort	In cold environments, people are used to wearing more heavy clothing on the upper body than the lower body, and it is recommended to distribute the thermal resistance of clothing more evenly to the lower body to keep warm	[81]
2020	Comparing the thermal comfort of the thermal environment served by air conditioners and floor heating terminals under steady-state condition	There was no significant difference in thermal comfort between the thermal environment served by air conditioners and floor heating terminals under steady condition	[82]
2021	How the local thermal sensation affects the overall thermal sensation after turning on the air conditioning system	After turning on the air conditioner, the feet feel the coldest and have the greatest impact on the overall thermal sensation	[83]
2021	Effects of perceived control on indoor thermal comfort in the winter	Perceived control can improve residents’ thermal comfort in the winter	[56]
2022	Effects of different air-conditioning terminals on thermal comfort in a non-uniform environment	The radiant ceiling improves the thermal comfort of the forearm and the back of the hand obviously, and the radiant floor improves the thermal comfort of the calf and foot	[84]
2022	Indoor thermal comfort for different heating terminals (fan coil units and room air conditioners) under different operating strategies (continuous and intermittent operation)	Except for long-term shutdown, there were no significant differences between different control strategies, including continuous operation and intermittent operation	[85]
2023	Thermal adaptation in naturally ventilated and air-conditioned environments in different seasons	Subjects showed a lower level of heat adaptation under the air-conditioned environment in the winter	[86]
2023	Differences in thermal comfort between the young and the elderly	The elderly have a time lag and smaller changes in thermal responses than the young when air temperature changes	[87]

lists the typical laboratory research content and conclusions from the scholars.

3.3. Comparison between Field Studies and Laboratory Studies

The significant differences between these two approaches are the controllability, consistency, and stability of the physical environment. Laboratory studies have more precise control over the environmental variables and are able to hold certain parameters under constant conditions, while the parameters in field studies are within certain acceptable ranges but are not specifically controlled. Although researchers can change environmental parameters by turning on air conditioners, opening windows, etc., it is difficult to keep the consistency in terms of the range, control settings, and scenarios. Therefore, studies using climate chambers are often designed to precisely control the key variables. However, field studies may reflect realistic behavioral, operational, and system characteristics [33]. There are also some studies that combined the laboratory and field studies to improve the accuracy of their outcomes [56,82,85].

Table 4. Comparison of laboratory studies and field studies.

	Publication	Scale	Environmental Control	Sample Representation	Number of Subjects	Sample Type
Laboratory studies	Less	Small	Strict	Poor	Small	Few
Field studies	More	Large	Not strict	Excellent	Large	Multiple

presents a comparison of the laboratory studies and field studies. As can be seen from the table, although laboratory studies have better control over the environmental parameters than field studies, the number of papers on laboratory studies is less than that on field studies. This is due to the limitations on the climate chamber size and research objects, and only a small scale of research can be carried out by laboratory studies. There is also the problem of poor sample representation in laboratory studies, with smaller sample sizes and fewer sample types compared with field studies. Therefore, field studies should be chosen and accompanied by laboratory studies to improve the credibility of the research if needed.

Based on the above literature review, it can be found that field studies and laboratory studies are complementary, and each approach has its application areas. The selection of the approach should be based on the research objectives and the characteristics of the research object.

4. Thermal Comfort Evaluation Indices

Thermal comfort evaluation indices are commonly obtained through the combination of the objective physical environmental parameter measurements with occupants’ subjective feelings on thermal comfort. The selection of an appropriate thermal comfort evaluation index can accurately reflect the human thermal comfort under a specific environment, which is very important for thermal comfort research. This paper reviews the predicted mean vote (PMV)/predicted percent dissatisfied (PPD), thermal sensation vote (TSV), mean thermal sensation (MTS), thermal comfort vote (TCV), operative temperature (OT), neutral temperature (NT), thermal preference vote (TPV), thermal acceptability (TA), preferred temperature (PT), thermal adaptation, and other indices.

4.1. PMV/PPD

The PMV (predicted mean vote) and predicted percent dissatisfied (PPD) were developed by Fanger in 1970 [88]. The PMV and PPD represent the feelings of most people in the same environment and the percentage of people who are dissatisfied with the thermal environment. The PPD is a function of PMV. The PMV model was derived from experiments and is intended to be applied to airtight, air-conditioned buildings, where the occupants have no opportunity to adapt to the indoor thermal environment [50]. The calculation of PMV requires the use of four environmental parameters (air temperature, relative humidity, mean radiant temperature, and air velocity) and two separate variables (clothing thermal resistance and metabolic rate), and can be calculated using a computer program [21]. The PMV indicator adopts a seven-point scale (adopted in ASHRAE [39]) and uses language to describe people’s subjective feelings on the thermal environment. In a comfortable and neutral environment, the PMV is a value close to 0.

Many researchers conducted evaluations on the applications of the PMV and PPD in thermal comfort sensation in China. Some researchers have found that PMV values do not accurately predict the actual thermal response of occupants [89,90]. Some limitations of the PMV were found that deviate the PMV from actual thermal sensation: (1) The PMV model is established based on the standard clothing worn by the subjects in a uniform steady-state environment, without considering the influence of local cold and warm sensations [51]; (2) The PMV model ignores the ability of the human body to actively adjust and adapt to the thermal environment [91]. In reality, people are active participants in the environment and will change the indoor thermal comfort condition by adjusting their clothes, opening windows for ventilation, and turning on air conditioners; (3) The PMV cannot be applied to naturally ventilated buildings as it is developed under an air-conditioned space [92]. The following deviations were found by the researchers: (1) a large deviation when the human body deviates from thermoneutrality [93] and in evaluating the thermal comfort sensation for buildings served by split units [50]; (2) underestimation of people’s actual thermal sensations in Jiaozuo in cold climate region [68], people’s tolerance in hot environments in the summer, and people’s subjective thermal sensations when heated in the winter [94]; and (3) overestimation of subjects’ feelings of coldness [95] and the actual thermal sensations of the people when the operative temperature exceeds 27 °C [50]. The PMV index was found to be less effective in predicting thermal sensations in men and women [42], and the accuracy of the PMV model was inconsistent between the cold and warm regions for the elderly [96].

To expand the application of the PMV model to other climate regions, Yao et al. [97] proposed a theoretical adaptive model of the PMV (aPMV), which considered many factors such as cultural, climatic, social, psychological, and behavioral adaptations, and addressed the problem of deviations in the PMV from actual mean votes (AMV) by introducing an adaptive coefficient (λ). The aPMV can be calculated as below:

aPMV = PMV/(1 + λPMV)

(1)

The adaptive coefficient is based on field research and takes into account the local climate, culture, and social background, which are dimensionless. λ is positive in a warm or hot environment and negative in cold conditions. The aPMV can be applied to both warm and cold climates and is used in the current Chinese national standard “Evaluation Standard for Indoor Thermal and Humid Environments of Civil Buildings” [98]. Some researchers have applied the aPMV in India [36] and China [59,99], and it was proved to be effective.

In a practical study, Li et al. [20] calculated the PPD distribution of the occupants in a residential building using the PPD model and compared the distribution of the actual percentage of dissatisfaction (APD) calculated from the thermal sensation vote. It was found that they agree well with each other, confirming that the PPD model can be successfully applied to residential buildings.

Although new thermal comfort models have been proposed, the PMV model is still the most widely used. Because there are various defects in the PMV model, researchers should carefully select multiple models for a comparative analysis to obtain convincing results. The modified PMV models have been proved to have an improved practicability and accuracy, and future research should continuously focus on the improvement of the PMV model.

4.2. TSV

The thermal sensation vote (TSV) is the thermal sensation of the human body at a certain temperature. When conducting thermal sensation experiments, researchers set up some voting options for subjects to fill in their own thermal sensations. It uses the same seven-point scale as recommended by ASHRAE [39].

Most researchers regard the TSV as an evaluation index of thermal comfort. Xu et al. [17] analyzed the thermal sensations of traditional residential buildings in Nanjing using the TSV and found that the environment in the summer is more comfortable than in the winter. Some researchers compared the TSV with the PMV and came to the following conclusions: (1) The TSV of persons in hot and cold environments was closer to neutral than predicted by the PMV, and the human thermal perception in non-air-conditioned environments was not as sensitive as predicted by the PMV model [94]; (2) The subjective TSV value is much larger than PMV value, and the comfortable temperature range of human beings in the actual thermal environment is wider than predicted [49]; (3) The TSV was always higher than the PMV under the same temperature in the winter, regardless of whether radiant floor heating or radiator heating was used. Differences between the PMV and TSV have also been seen in other studies [50,72], and researchers often choose one of the methods for thermal sensation studies, leading to variations in the conclusions from different researchers. In the future, the thermal comfort models need to be improved to reduce the prediction bias and provide a comparative analysis of different indices.

4.3. MTS

The mean thermal sensation (MTS) is the mean thermal sensation of the human body in a certain temperature range. The correlation of the regression equation obtained using the TSV is relatively low, and therefore, the regression equation does not have the statistical significance for predicting the thermal sensations of the occupants. The MTS can be used to better predict the average thermal sensation of the human body than using the regression equation [41]. Yang et al. [15] conducted a linear regression analysis on the MTS of the residents of three residential buildings and the indoor operative temperature resulted in an R-squared greater than 0.9, proving that the regression equation is effective in predicting the variations in thermal sensations. Researchers tended to compare the results of the MTS with the PMV. For example, Han et al. [58] found that the PMV model predicted value was not equivalent to the MTS obtained from the questionnaire. Ning et al. [100] found that the MTS was always greater than the predicted value of the PMV in the dormitories of Harbin University, and the seasonal deviation between the MTS and PMV varied greatly. Yan et al. [21] found that the MTS-based neutral temperature was lower than the PMV-based neutral temperature in the winter, and the opposite was found in the summer. Han et al. [64] found that the neutral operative temperature calculated using the PMV was lower than that calculated using the MTS, and they attributed this difference to human thermal adaptation as people can adjust their clothes, open windows, and turn on fans to adapt to the thermal environment.

4.4. TCV

The physiological and psychological sensations of the individual affect his/her thermal comfort feeling, which are almost impossible to express numerically [101]. In order to overcome this problem, the thermal comfort vote (TCV) was proposed to evaluate the degree of thermal comfort. Some researchers in China have carried out corresponding investigations. For example, Wang et al. [14] set up a thermal comfort scale of 0 to 3 in their survey, with 0 representing comfort and 3 representing discomfort. Some other researchers set up the thermal comfort scale from −3 to +3, which is more detailed in describing the thermal comfort of the human body [50]. Many researchers often choose to set up the thermal comfort vote by themselves, but it should be noted that the thermal comfort vote should not be set to be too precise, so as to prevent the subjects from distinguishing the subtle differences between the thermal comfort sensory scales.

4.5. Operative Temperature

The operative temperature was proposed by Winslow et al. in 1937 [102]. It is a synthetic temperature obtained by comprehensively considering the influence of the air temperature and the mean radiation temperature on the thermal sensation of the human body [103]. When the relative humidity is within the thermal comfort range and the indoor air velocity is low, it is recommended to use the operative temperature as the thermal comfort index [104]. The operative temperature can be calculated as below:

$${\mathrm{t}}_{\mathrm{o}\mathrm{p}}=\frac{{\mathrm{h}}_{\mathrm{c}}{\mathrm{t}}_{\mathrm{a}}+{\mathrm{h}}_{\mathrm{r}}{\mathrm{t}}_{\mathrm{r}}}{{\mathrm{h}}_{\mathrm{c}}+{\mathrm{h}}_{\mathrm{r}}}$$

(2)

where h_c is the convective heat transfer coefficient, in Wm². °C; h_r is the radiative heat transfer coefficient, in W/m². t_op is the operative temperature, in °C; t_a is the air temperature, in °C; and t_r is the mean radiant temperature, in °C.

When the air velocity is low (<0.2 m/s), the operative temperature can be calculated as the average value of the air temperature and mean radiant temperature [39]:

$${\mathrm{t}}_{\mathrm{o}\mathrm{p}}=\frac{{\mathrm{t}}_{\mathrm{a}}+{\mathrm{t}}_{\mathrm{r}}}{2}$$

(3)

In field studies, the operative temperature is used as a substitute for air temperature to describe the indoor thermal environmental conditions, to calculate the MTS or TSV regression equations, neutral temperature, acceptable temperature, etc. Researchers usually use the average of the air temperature and mean radiant temperature when calculating the operative temperature [15]. In some indoor environments, the air temperature is very close to the mean radiant temperature, and the air temperature can also be directly used as an index to evaluate the thermal comfort of residential buildings [14,73].

4.6. Neutral Temperature

The neutral temperature is the temperature at which people feel neither warm nor cold [60]. The neutral temperature can be divided into the measured thermoneutral temperature and predicted thermoneutral temperature. The thermoneutral temperature is highly related to the thermal comfort, expectations, and acceptability of the people [105]. The thermoneutral temperature is not constant and has seasonal variations [100,106].

The temperature frequency method (Bin method) [104] is most commonly used to calculate the thermal neutral temperature [15]. The measured thermoneutral temperature uses the temperature frequency method to divide the indoor operative temperature into several temperature intervals (0.5 °C or 1 °C for each interval). The central temperature or average temperature of each operative temperature interval [107] is used as the independent variable, and the average subjective MTS in each temperature range is the dependent variable to develop the regression model as:

MTS = a + b·t_op

(4)

where a and b are the regression constant and slope, respectively. The operative temperature at MTS = 0 is the measured thermoneutral temperature [51,52,53,62,68,69,70].

The linear slope in this regression equation describes subjects’ sensitivities to room temperature [108]. The theoretical thermal neutral temperature (predicted thermoneutral temperature [69]) is calculated similarly by using the predicted PMV instead of the MTS and developing the regression model between the PMV and operative temperature with an interval of 0.5 °C. The operative temperature at a PMV = 0 is the predicted thermoneutral temperature [69,70,107].

The measured thermoneutral temperature is not equal to the predicted thermoneutral temperature. Yan and Yang [69] found that in the mixed cooling mode, the measured thermoneutral temperature is higher than the predicted one, while Yang et al. [107], Qu et al. [70], and Gong et al. [62] reached the opposite conclusion that the predicted neutral temperature is higher than the measured neutral temperature. Yang and Zhang [60] studied the thermal comfort of naturally ventilated buildings and air-conditioned buildings and found that the neutral temperature of naturally ventilated buildings is higher than that of air-conditioned buildings, and the occupants of naturally ventilated buildings seem to be more tolerant to high temperatures.

4.7. Thermal Preference Vote

The thermal preference vote (TPV) is obtained by conducting surveys on the subject’s thermal preference by setting options in the questionnaire. Three-point scales were typically used by the Chinese researchers. For example, from −1 to +1 [14,15], or from +1 to +3 [58]. Through thermal preference voting, occupants can express their subjective opinions on whether the indoor thermal conditions meet the requirements. When the subjects were satisfied with the thermal environment, they voted “no change”, indicating that they wanted to keep the current indoor environment, otherwise they voted to expect “colder” or “warmer” [105]. Some studies regard the TSV of the three central values of the ASHRAE scale (TSV = −1, 0, 1) as satisfactory or acceptable [41,105], while thermal preference voting can more directly describe the thermal expectations of the human body. Yang et al. [15] found that residents’ thermal preferences are biased toward a warmer indoor environment. Cao et al. [72] found that compared with Beijing, people in Harbin are more accustomed to the warm indoor environment in the winter. The thermal preferences differ between urban and rural residents, and the expectations of residents in urban areas are relatively higher than those in rural areas [58], which may be due to the differences in lifestyles and economic levels.

4.8. Thermal Acceptability

One of the important purposes of thermal comfort research is to identify the thermal environment conditions that can be accepted by most occupants [39]. There are different approaches to consider the thermal environment as satisfactory. The first approach is to use the PMV range to determine the acceptability of the thermal environment. When −0.5 ≤ PMV ≤ 0.5, PPD ≤ 10%, it means that more than 90% of the people are satisfied with the thermal environment; when 0.85 ≤ PMV ≤ 0.85, PPD ≤ 20%, it means that more than 80% of the people are satisfied with the thermal environment [15,93]. The second approach considers the three central values of the ASHRAE seven-point scale (−1, 0, 1) as satisfactory or acceptable [109], while the thermal sensation voting values (−3, −2, + 2, +3) are considered as unacceptable [103].

Table 5. Calculation method for acceptability.

Method	Information	Ref.
Direct inquiry	Set thermal acceptability voting in the questionnaire, ask the subjects whether the current thermal environment is acceptable (for example, 1 means acceptable, −1 is unacceptable), and find the statistical results from the questionnaire	[51,58,61,64,68,72,103]
Thermal sensation	The middle three values (−1, 0, 1) of the ASHRAE seven-point scale are considered as satisfactory or acceptable, while the thermal sensation votes (−3, −2, +2, +3) are considered as unacceptable	[41,45,57,61,62,66,103,105]
Thermal expectation	The subjects who voted “unchanged” (the thermal expectation vote is 0) are satisfied with the thermal environment they are in, the proportion of the people to the total population are calculated	[60,110]

show the specific methods.

The thermal acceptance rates obtained from different methods are not consistent. Some researchers found that the thermal acceptance rates obtained using the direct inquiry and thermal expectation methods were the highest and lowest, respectively [60,69,110].

The temperature range that the occupants can accept for the indoor thermal environment is considered as the acceptable temperature range. ASHRAE Standard 55 stipulates that the thermal environment acceptable to 80% or more of the subjects is an acceptable environment [39], which was employed by the researchers in China [111]. The calculation methods of the acceptable temperature range include the thermal sensation method and direct inquiry method, as shown in

Table 6. Calculation method for acceptable temperature range.

Method	Info.	Ref.
Thermal sensation	At a 0.5 °C interval, the proportion of people who feel cold (−2, −3) or warm (2, 3) in the temperature range were calculated and fit as a polynomial curve; the operative temperature between two intersection points of the 20% unsatisfactory straight line and the polynomial curve is considered as the acceptable temperature range	[45,64,69]
Direct inquiry	Acceptance is voted as +1 for thermally acceptable and −1 for thermally unacceptable, and a regression analysis is performed on the thermally unacceptable rate and operative temperature at intervals of 0.5 °C to find the thermally acceptable temperature range	[61]

.

The indices, including PMV, TSV, and MTS, can all be used in the thermal sensation method. Gong et al. [62] adopted the MTS and PMV to calculate the acceptable temperature range and found that the acceptable temperature range corresponding to the MTS was larger than that of the PMV. Xu et al. [17] substituted the TSV values (−0.5 to +1.2 in the winter and −1.4 to +0.5 in the summer) into the linear regression equation of the TSV and indoor operative temperature to determine the 80% acceptability temperature range.

Different calculation methods lead to variations in the acceptance rate. Therefore, the discrepancy among different methods should be studied to find the optimal method for calculating the thermal acceptability.

4.9. Preferred Temperature

The thermal sensation method and direct inquiry method can also be used to calculate the preferred temperature. In the thermal sensation method, the proportions of people who vote for cold (−2, −3) or warm (+2, +3) at each 0.5 °C interval of the operative temperature range calculated were used to develop regression models with the operative temperature. The temperature at the intersection points of the two fitting curves was considered as the preferred temperature [60,69,107]. The direct inquiry method also adopts the 0.5 °C interval and calculates the proportions of people who voted for warm (+1) or cold (−1) at each interval to develop regression models with the operative temperature. Similarly, the temperature at the intersection point of the two fitting curves was considered as the preferred temperature [56,61,62,70,103]. However, Yan and Yang [69] found a great difference between the preferred temperatures obtained using the two methods.

Many researchers conducted analyses on the preferred temperature and neutral temperature and found they differed from each other in the same climate region and reached the following conclusions: (1) Subjects prefer a cooler indoor thermal environment than thermally neutral in a hot summer environment [69]; (2) No matter whether in a severely cold region or cold region, the thermoneutral temperature in the summer is higher than the preferred temperature, and people generally expect the temperature to be lower than the neutral temperature [107]; (3) Whether it is a naturally ventilated building or an empty building, the preferred temperature in the summer is lower than the neutral temperature [60,103].

Other scholars have conducted analyses in different climate regions and seasons. For example, Yang et al. [15] found that residents in arid regions have higher preferred temperatures than the thermoneutral temperature, and Qu et al. [70] found that residents in cold regions also have higher preferred temperatures than the thermoneutral temperature. Xu et al. [17] found that the residents have a higher preferred temperature than the neutral temperature in the winter and lower than the neutral temperature in the summer, which indicates that people in warm climates prefer to feel slightly cooler than neutral, while people in cold climates prefer to feel slightly warmer. Residents in different locations also have different preferred temperatures. For example, Zhang et al. [61] found that urban residents expect an indoor thermal environment that is lower than their current indoor temperature, while rural residents expect an indoor thermal environment that is higher than their current indoor temperature.

4.10. Thermal Adaptation

There is a clear relationship between indoor comfort and outdoor environmental conditions [24]. The adaptive model was proposed by de Dear and Brager [108] in 1998 and was included in the ASHRAE 55 standard [38]. The adaptive model is based on field studies in naturally ventilated buildings to develop a linear regression equation between the indoor operative temperature and the outdoor temperature. It was found that the indoor operative temperature increases with the increase in the outdoor temperature [112].

Researchers from China also conducted a lot of investigations to develop adaptive models for different regions and different types of building [20,21,44,45,46,50,59,66,67,70,71,103,113]. A study in East China showed that there are differences in thermal comfort models in different climate regions [21]. Thermal adaptation plays an important role in people’s adjustment of their own comfort feeling. Thermal adaptation can alleviate the discomfort caused by temperature deviation and expand the acceptable temperature range of people [20]. The thermal comfort model will help to improve the thermal design of buildings and the energy conservation operation of building equipment, and it can also be used for climate analyses during the design phases of buildings [70].

Based on the literature survey, it was found that some researchers only focus on a certain city or a certain climate zone when calculating the adaptive model, which leads to the narrow application of the obtained adaptive model. In the future, the research should be expanded to a larger scale and focus on developing more adaptive models which can be applied to more regions and building types.

4.11. Other

Other than the indices mentioned in the above section, and there are also many other factors that can affect the thermal sensation of the human body, e.g., humidity and air flow. Some researchers refer to the seven-point scale thermal sensation index and set up a similar seven-point scale wet sensation vote and seven-point air flow sensation vote [44,50,60,65], as well as preferred humidity and preferred airflow sensations [44,65], to obtain subjects’ wet perceptions and airflow perceptions of the environment. Yan et al. [103] obtained residents’ preferred air velocities by referring to the method to calculate the preferred temperature.

Operative temperature cannot be used directly for indoor air temperature control. Therefore, some researchers have proposed new temperature indices as substitutes to the operative temperature, for example, the effective temperature (ET*), an equivalent dry bulb temperature proposed by Gagge et al. [114], and standard effective temperature (SET*), proposed by Gonzalez et al. [115]. They were also employed by the Chinese researchers. For example, Xia et al. [13] used the ET* instead of the air temperature or operative temperature to obtain the thermoneutral temperature. Zhang et al. [116] used the SET* instead of the operative temperature to study the thermal comfort of naturally ventilated buildings in hot and humid areas. Yan et al. [50] conducted a study using both the operative temperature and SET* and found that the neutral temperature obtained using SET* was 2–3 °C lower than that obtained using the operative temperature. However, these new indexes also have some flaws. The calculation of the ET* or SET* assumes that the skin heat loss under the standard environment and the actual environment are equal. However, after verification, it is found that the same average skin temperature and skin moisture cannot be achieved under the two environments [117]. Therefore, further research is needed to address this problem.

5. Discussion and Analysis

5.1. Thermal Comfort Field Investigation

Figure 8 summarizes the number of studies in different climate regions for the field investigation method from the literature survey. Detailed information of the results is listed in

Table A1. Summary of the field investigation from the literature survey.

Year	City	Climate Region	Season	Note	Average Outdoor Temperature (°C)	Average Indoor Temperature (°C)	Neutral Temperature (°C)	Preferred Temperature (°C)	80% Acceptability Temperature (°C)	Average Clothing Thermal Resistance	Ref.
2018	Tianjin	Cold	Summer	--	21.3	--	25.4	--	21.0–27.3	--	[16]
2009	Hunan	Hot summer and cold winter	Winter	--	--	12.1	14.0	--	--	2.0	[58]
2006	Harbin	Severe cold	Winter	Men	--	--	20.9	--	--	1.33	[41]
2006	Harbin	Severe cold	Winter	Women	--	--	21.9	--	--	1.42	[41]
2014	Beijing	Cold	Winter	Central heating	--	21.3	20.7	--	--	0.86	[105]
2014	Beijing	Cold	Winter	Independent heating	--	21.2	18.1	--	--	0.87	[105]
2014	Shanghai	Hot summer and cold winter	Winter	--	--	16.5	17.2	--	--	1.45	[105]
2022	Inner Mongolia	Severe cold	Winter	--	−1.8	26.9	24.8	--	21.9–27.5	0.43	[15]
2018	--	Severe cold	--	--	--	--	20.2	--	--	--	[20]
2018	Xi’an	Cold	--	--	--	--	19	--	--	--	[20]
2018	--	Hot summer cold winter	--	--	--	--	18.5	--	--	--	[20]
2018	--	Hot summer and warm winter	--	--	--	--	21	--	--	--	[20]
2018	Kunming	Mild	--	--	--	--	22	--	--	--	[20]
2017	Harbin	Severe cold	Winter	Early heating	−10	24.3	21.6	--	--	0.85	[14]
2017	Harbin	Severe cold	Winter	Mid heating	−10	24.3	23.5	--	--	0.78	[14]
2017	Harbin	Severe cold	Winter	Late heating	−10	24.3	23.1	--	--	0.69	[14]
2017	--	Severe cold	Winter	--	−19.7	19.2	19.5	--	--	1.37	[21]
2017	--	Cold	Winter	--	−2.3	16.3	20.8	--	--	1.41	[21]
2017	--	Hot summer and cold winter	Winter	--	5.9	15.7	18.2	--	--	1.26	[21]
2017	--	Hot summer and warm winter	Winter	--	15.3	16.5	19.7	--	--	1.22	[21]
2017	--	Severe cold	Summer	--	26.8	24.4	25.4	--	--	0.58	[21]
2017	--	Cold	Summer	--	30.7	28.6	27.2	--	--	0.34	[21]
2017	--	Hot summer cold winter	Summer	--	32.1	29.4	27.6	--	--	0.32	[21]
2017	--	Hot summer and warm winter	Summer	--	31.7	29.7	27.5	--	--	0.34	[21]
2018	Nanjing	Hot summer and cold winter	Summer	--	27.3	--	28	--	22–30.1	0.46	[17]
2018	Nanjing	Hot summer and cold winter	Winter	--	2.4	--	15.8	--	10.6–28.5	2.07	[17]
2010	Harbin	Severe cold	Summer	--	27.0	26.9	23.7	--	21.5–31.0	--	[44]
2022	Jiaozuo	Cold	Summer	Air conditioned	32.6	27.7	28.7	--	24.3–30.0	--	[50]
2022	Jiaozuo	Cold	Summer	Naturally ventilated	30.2	29.6	26.9	--	16.2–29.9	--	[50]
2022	Jiaozuo	Cold	Summer	Hybrid mode	31.5	28.6	27.7	--	23.7–30.1	--	[50]
2021	Beijing	Cold	Winter	Transition period	12.0	19.5	24.6	--	--	1.0	[55]
2021	Beijing	Cold	Winter	Heating period	5.2	22.9	21.9	--	--	0.9	[55]
2017	--	Hot summer and cold winter	Spring	--	20.05	20.42	21.1	--	--	0.69	[59]
2017	--	Hot summer and cold winter	Summer	--	29.57	28.98	24.3	--	--	0.26	[59]
2017	--	Hot summer and cold winter	Autumn	--	20.31	21.07	23.8	--	--	0.60	[59]
2017	--	Hot summer and cold winter	Winter	--	9.57	11.78	21.1	--	--	1.30	[59]
2016	Harbin	Severe cold	Winter	--	−8.9	23.9	22.0	--	--	0.96	[72]
2016	Beijing	Cold	Winter	--	3.6	21.3	22.0	--	--	1.1	[72]
2016	Shanghai	Hot summer and cold winter	Winter	--	9.9	16.4	22.7	--	--	0.99	[72]
2011	Harbin	Severe cold	Winter	Before heating	15.9	20.6	25.1	--	17.5–24.0	0.77	[65]
2011	Harbin	Severe cold	Winter	During heating	−3.6	21.6	20.4	--	19.0–26.5	0.88	[65]
2008	--	Hot summer and cold winter	Summer	Naturally ventilated	--	33.0	28.3	27.9	25.0–31.6	0.28	[60]
2008	--	Hot summer and cold winter	Summer	Air conditioned	--	28.5	27.7	27.3	25.1–30.3	0.32	[60]
2023	Guangzhou	Hot summer and warm winter	Summer	--	--	--	28.4	--	27.1–29.6	0.5	[49]
2022	Jiaozuo	Cold	Winter	Radiator heating	5.4	21.1	19.8	--	17.0–29.4	1.15	[51]
2022	Jiaozuo	Cold	Winter	Floor heating	5.4	21.5	19.2	--	15.9–34.5	0.89	[51]
2019	North	--	Winter	--	−5.5	18.6	20.2	--	--	1.35	[45]
2019	South	--	Winter	--	8.1	11.9	17.1	--	--	1.70	[45]
2020	Nanjing	Hot summer and cold winter	Summer	--	--	--	27.6	--	--	--	[46]
2020	Nanjing	Hot summer and cold winter	Winter	--	--	--	14.4	--	--	--	[46]
2007	--	--	Summer	--	--	29.8	28.6	--	--	0.54	[64]
2014	Xi’an	Cold	Transition season	--	13.9	18.1	21.3	23.9	Lower limit: 11.4	1.19	[70]
2015	Yinchuan	Cold	Summer	--	29.2	28.9	25.9	21.7	Upper limit: 28.5	0.3	[103]
2012	Jiaozuo	Cold	Summer	--	29.5	30.2	Measured: 27.7 Predicted: 25.4	27.4	Thermal sensation: 19.8–28.8 Direct inquiry: 20.7–29.2	0.258	[69]
2021	Bayannur	Cold	Winter	City	--	26.8	24.8	25.3	20.9–28.0	0.43	[61]
2021	Bayannur	Cold	Winter	County	--	19.4	20.4	22.1	17.7–23.3	0.43	[62]
2020	Guilin	Hot summer& cold winter	Summer	--	28.3	30.8	Predicted: 25.4 Measured: 24.8	25.8	Predicted: 22.8–28.0 Measured: 21.0–28.5	--	[62]
2021	Jiaozuo	Cold	Summer	--	30.4	28.6	27.6	--	Upper limit: 29.9	0.29	[68]
2019	Xi’an	Cold	Winter	--	--	--	21.0	--	--	0.5	[63]
2021	Temperate continental	Cold	Summer	--	--	--	24.0	--	--	--	[57]
2021	Monsoon climate	Cold	Transition season	--	--	--	20.9	--	--	--	[57]
2021	Temperate continental	Cold	Summer	--	--	--	25.7	--	--	--	[57]
2021	Monsoon climate	Cold	Transition season	--	--	--	22.1	--	--	--	[57]
2021	Temperate marine	Cold	Summer	--	--	--	26.2	--	--	--	[57]
2021	Monsoon climate	Cold	Transition season	--	--	--	22.3	--	--	--	[57]
2022	Temperate marine	Hot summer & cold winter	Winter	--	9.37	11.47	Measured: 13.6 Predicted: 14.2	--	--	1.65	[133]
2022	Monsoon climate	Hot summer & cold winter	Transition season	--	--	--	19.9–26.9	--	16.3–30.7	0.69	[66]
2022	Temperate continental	Hot summer & cold winter	Transition season	--	--	--	20.1–29.0	--	16.1–30.7	0.96	[66]
2022	Desert climate	Severe cold	Summer	--	27.1	28.1	Measured: 24.9 Predicted: 25.2	24.6	23.8–26.9	0.287	[107]
2022	Temperate continental	Cold	Winter	Before heating	12	19.9	24.6	--	20.5–29.2	1.1	[73]
2022	Desert climate	Cold	Winter	After heating	5.2	22.7	21.9	--	20.7–27.3	0.9	[73]
1999	Yancheng	Cold	Summer	--	--	28.6	26.7	--	Upper limit: 30	0.31	[15]
2006	Shanghai	Hot summer & cold winter	Summer	--	--	--	22.5	--	14.7–29.8	--	[71]
2014	Wuhan	Cold	Winter	District heating	−2.1	20.4	22.0	--	--	0.92	[118]
2014	Baotou	Cold	Winter	Independent heating	−2.1	18.8	18.6	--	--	0.92	[118]
2018	Beijing	Severe cold	Winter	Early heating	1.4	23.6	21.6	--	--	0.74	[53]
2018	Beijing	Severe cold	Winter	Mid heating	−16	24.3	23.5	--	--	0.74	[53]
2018	Beijing	Severe cold	Winter	Late heating	2.2	25.0	23.1	--	--	0.74	[53]
2016	Shanghai	Severe cold	Winter	--	−11.0	20.6	21.3	--	17.2–25.4	0.89	[52]
2016	Beijing	Cold	Winter	--	−5.6	20.9	20.3	--	15.4–24.1	0.86	[52]
2016	Beijing	Cold	Winter	--	5.8	17.2	20.4	--	9.8–26.2	1.18	[52]
2021	Harbin	Hot summer & cold winter	Winter	--	2.4	--	16.9	15.8	--	--	[56]
2021	Harbin	Hot summer & cold winter	Winter	--	2.5	--	18.6	19.7	--	--	[56]
2013	Harbin	Cold	Winter	--	--	--	--	18.9	13.6–32.4	1.44	[67]
2013	Baotou	Cold	Summer	--	--	--	--	23.3	13.6–32.4	0.46	[67]
2019	Yinchuan	Severe cold	Winter	--	--	24.0	23.0	--	18.3–27.3	0.79	[54]

of Appendix A.

Figure 9 displays the proportion of the studies on each index from the literature survey. It can be observed that the TSV, operative temperature (top), thermal acceptability (TA), MTS, and PMV are the most commonly used thermal indices, which appear in 79% of all the thermal comfort surveys.

When conducting the field studies, it is usually necessary to select multiple thermal comfort indices to obtain more reasonable conclusions. Some researchers [20] selected the PPD and APD to calculate the thermal acceptability of the indoor environment and found that the APDs in most districts are very close to the PPDs, which can be successfully used in residential buildings. Some other researchers [14,105] used both the PMV and TSV to measure indoor thermal sensations and found there are some differences. Since the same thermal comfort index can be obtained using different methods, it is necessary to select multiple thermal comfort indices for a cross-comparison to verify the accuracy of the results.

5.2. Acceptable Temperature Range

The 80% acceptable temperature ranges in different seasons from the literature are presented in Figure 10, with the curve at the top showing the upper limit, and the one at the bottom showing the lower limit of the acceptable temperature range. It can be found that the average lower and upper limits of the acceptable temperature in the summer are 21.3 °C and 29.7 °C, respectively. The average lower and upper limits of acceptable temperature in the winter are 17.0 °C and 27.6 °C, respectively. The acceptable temperature in the summer is higher than that in the winter, owing to the difference in the outdoor air temperature. For example, Lhasa is located in the cold area, where the outdoor temperature can be very low in the summer. Its acceptable temperature range is 13.6–32.4 °C, which is large, with a lower limit as low as 13.6 °C in the summer [67], due to the large diurnal temperature difference. For another city, Guangzhou, located in the hot summer and warm winter region, its acceptable temperature range is 27.1–29.6 °C, with only 2.5 °C in difference [49], due to the high temperature and low diurnal temperature difference in the summer.

5.3. Yearly Variation in Neutral Temperature

The variation in the neutral temperature with time in different climate regions is presented in Figure 11. Since it is greatly affected by the seasons [100,118], they are separately presented for the winter and summer, respectively. It can be seen from the figure that in the summer, the neutral temperature tends to increase with time in hot summer and cold winter regions. However, it is not obvious in cold regions. In the winter, the neutral temperature in severe cold and cold regions tends to increase with time, because in severe cold and cold regions, the climate conditions in the winter are poor. With the improvement of living standards, people have higher and higher requirements for their indoor thermal environment. At the same time, in those regions, central heating is provided, and therefore, people have the ability to control the indoor environment in the winter, leading to an increase in the thermoneutral temperature with time. However, in hot summer and cold winter regions, people have poor control over the indoor environment, and there is no obvious trend in the variation of thermoneutral temperature over time. Meanwhile, it can be found that the thermoneutral temperature in the summer is higher than that in the winter due to the difference in the outdoor air temperature.

5.4. Variation in Neutral Temperature with Climate Region

China has a large geographical span, and the environmental conditions of different climate zones vary greatly. In order to study the influence of different climate zones on the neutral temperature, the average neutral temperatures of different climate zones in the summer and winter from the literature survey are obtained and listed in

Table 7. The average neutral temperature in different climate regions.

Climate Region	Average Neutral Temperature (°C)
	Summer	Winter
Severe cold	24.7	22.4
Cold	26.6	21.2
Hot summer and cold winter	26.4	17.3
Hot summer and warm winter	28.0	19.7

.

It can be found that there is a very small difference in the thermoneutral temperature in different climate zones in the summer. However, there is a great difference in the winter. This is because the temperature difference across China is small in the summer. Even in Harbin, located in the severe cold region, its average temperature in the summer is above 20 °C [44]. However, in the winter, the temperature varies greatly among different regions in China. For example, the outdoor temperature in Harbin can be as low as −20 °C [53], while in Guangzhou, located in the hot summer and warm winter region, its ambient temperature can be well above 0 °C, resulting in a large difference in the neutral temperature in the winter.

In the summer, the thermoneutral temperature is higher in warmer climate regions. This is due to the human body’s adaptability to the thermal environment. As people adapt to the warm ambient environment, their tendency to accept a higher thermoneutral temperature also increases. However, in the winter, the thermoneutral temperature is lower in the warmer climate zone. This is due to the buildings in severe cold and cold areas generally being equipped with a central heating system, and therefore, the indoor temperature is higher than that in other climate regions, where there is no central heating system. Due to the adaptation of occupants to the thermal environment, the neutral temperatures in severe cold and cold climate regions are higher than in other regions. In hot summers and warm winter regions, the outdoor temperatures are higher than those in the hot summers and cold winter climate regions, leading to a higher neutral temperature.

People’s thermal adaptation varies with time, season, and climate [79]. It can be seen from the table that in the same climate zone, the neutral temperature in the summer is greater than that in the winter, especially in hot summer and cold winter regions and hot summer and warm winter regions. Some other studies also reach the same conclusions [17,46,59]. This is because of the temperatures in different seasons and the difference in clothing thermal resistance. The seasonal differences in neutral temperatures indicate that the outdoor climate affects human adaptability, and it is suggested to maintain the indoor temperature in the winter lower than that in spring, which is a comfortable and energy-saving adjustment strategy [106].

In general, the thermoneutral temperature is associated with but not completely affected by the climate zone. It is also affected by the heating, ventilation, and air-conditioning system in the building.

5.5. Variation in Neutral Temperature with Clothing Thermal Resistance

The thermoneutral temperature is affected by clothing thermal resistance. Figure 12 presents the variations in the neutral temperature with clothing thermal resistance based on the results from the literature.

It can be observed that the clothing thermal resistance has a stronger correlation with the thermal neutral temperature in the winter. However, the correlation in the summer is weaker. This is because people wear light and thin clothes (0.25–0.5 clo.) in the summer, and the difference in clothing thermal resistance is small, so it has little impact on the thermoneutral temperature. In the winter, the outdoor temperature varies greatly in different climate zones. The outdoor temperatures in severe cold and cold areas are much lower than those in hot summer and cold winter areas. The thermal resistance of people’s clothing (0.5–2.0 clo.) varies greatly, which is greatly affected by the outdoor temperature.

5.6. Variation in Neutral Temperature with Indoor Air Temperature

Figure 13 presents the variation in the neutral temperature with the indoor air temperature. It can be observed that whether it is winter or summer, the thermoneutral temperature tends to increase with the increase in indoor air temperature, reflecting the adaptability of the human body to the indoor thermal environment. When the indoor temperature is high, people will adopt a series of adaptive behaviors to adapt to the indoor environment, leading to a high thermoneutral temperature.

5.7. Variation in Clothing Thermal Resistance with Indoor and Outdoor Air Temperatures

Clothing thermal resistance is affected by the air temperature [21]. Figure 14a,b present the variations in clothing thermal resistance with the indoor and outdoor air temperatures, respectively. It can be observed that the clothing thermal resistance decreases with the increase in indoor and outdoor temperatures. The clothing thermal resistance has a strong correlation with the indoor temperature. Because people live indoors most of the time, when the indoor temperature is too high, people will choose to wear thinner clothes, and when the indoor temperature is lower, they will wear thicker clothes. The correlation between clothing thermal resistance and outdoor temperature is much weaker, as people spend over 90% of their time indoors. Therefore, the outdoor air temperature most of the time has an indirect impact on the occupants. People can turn the air conditioners and other equipment used to control the indoor air temperature on/off, and the impact of the outdoor temperature on the space is greatly weakened. When the outside temperature is too low, people will turn on the air conditioner or put on clothes to prevent themselves from being in uncomfortable conditions. When analyzing indoor thermal comfort, the influence of clothing thermal resistance cannot be ignored [119].

5.8. Adaptive Model

A large number of field studies have shown that the outdoor climate has a great influence on the thermoneutral temperature of naturally ventilated buildings [44], and the thermal adaptability varies with climate regions [21]. Based on data from the field studies, the adaptive model is developed as a linear regression equation between the neutral temperature and outdoor air temperature.

Table 8. Summary of the adaptive models.

Year	Location	Climate Region	Season	Adaptive Model	Ref.
2017	--	Severe cold	--	tn 1 = 0.121 tout 2 + 21.488	[21]
2017	--	Cold	--	tn 1 = 0.271 tout 2 + 20.014	[21]
2017	--	Hot summer and cold winter	--	tn 1 = 0.326 tout 2 + 16.862	[21]
2017	--	Hot summer and warm winter	--	tn 1 = 0.554 tout 2 + 10.578	[21]
2010	Harbin	Severe cold	Summer	tn 1 = 0.486 tout 2 + 11.802	[44]
2022	Jiaozuo	Cold	Summer	tn 1 = 0.271 tout 2 + 20.014	[50]
2017	--	Hot summer and cold winter	Winter	tn 1 = 0.709 trm 3 + 8.25	[59]
2019	North	--	--	tn 1 = 0.438 tout 2 + 13.56	[45]
2020	Nanjing	Hot summer and cold winter	--	tn 1 = 0.7954 tout 2 + 4.8167	[46]
2014	Xi’an	Cold	Transition season	tn 1 = 0.3322 tout 2 + 14.263	[70]
2015	Yinchuan	Cold	Summer	tn 1 = 0.33 tout 2 + 18.8	[103]
2021	Temperate continental monsoon climate	Cold	--	tn 1 = 0.2904 tout 2 + 14.367	[57]
2021	Temperate maritime monsoon climate	Cold	--	tn 1 = 0.1604 tout 2 + 21.626	[57]
2021	Temperate continental desert climate	Cold	--	tn 1 = 0.4190 tout 2 + 11.2520	[57]
2022	Shanghai	Hot summer and cold winter	Transition season	tn 1 = 0.3681 tout 2 + 17.986	[66]
2022	Wuhan	Hot summer and cold winter	Transition season	tn 1 = 0.3792 tout 2 + 17.671	[66]
2006	Shanghai	Hot summer and cold winter	Summer	tn 1 = 0.42 tout 2 + 15.12	[71]
2013	Lhasa	Cold	Summer	tn 1 = 0.474 tout 2 + 13.8	[67]

¹ Neutral temperature. ² Outdoor air temperature. ³ Daily running average outdoor air temperature.

lists the adaptive models developed by different researchers. It can be observed that the adaptive models vary with seasons and climate regions. The adaptive models in the literature are developed for a certain city or for a certain climate region. These adaptive model regions vary widely and cannot be shared between each other.

Figure 15 presents the correlation between the neutral temperature and average outdoor temperature in different climate zones based on the survey data in the literature. It can be observed that the regression equation for the hot summer and cold winter region has the largest slope, and the one for the severe cold region has the smallest slope, which indicates that the warmer the outdoor climate region, the smaller the annual temperature difference and the more sensitive the neutral temperature to the variations in the outdoor temperature. This is consistent with the conclusion obtained by Yan et al., which is based on the data from different climatic regions in East China [21].

5.9. Adaptive Behavior

Thermal adaptation in the built environment can be attributed to three distinct processes: behavioral adjustment, physiological adaptation, and psychological habituation or expectation [91]. The human body’s subjective evaluation of the thermal environment is the result of the joint actions. When the indoor thermal environment is not within the comfortable range that the human body adapts to, people will first adopt different adaptive behavior measures to improve their current thermal comfort feeling, and then gradually change their psychological expectations and physiological adjustments over time to adapt to different thermal environments [15,61]. Both the indoor and outdoor climate can affect the thermal adaptation of the human body [118]. When the human body is stimulated by the external environment, according to the intensity, type (cold stimulus or heat stimulus), and duration of the environmental stimulus, the occupants tend to adopt corresponding physiological, psychological, and behavioral measures to adapt to the environment so as to feel comfortable in the current environment [18]. Age also affects the thermal adaptation of the human body. Wang et al. [14] found that the thermoneutral temperature of the elderly is higher than those of the young, and the young have a stronger ability to adapt to cold temperatures.

Personal control over the indoor environment can alter human thermal comfort sensation [120], and human behavior regulation cannot be ignored in thermal comfort research. People can adjust the human body heat balance in various ways, such as changing clothes, changing the activity levels, turning the air conditioner or fan on/off, etc. [32,121]. The adaptive behaviors of occupants not only play an important role in restoring thermal comfort conditions, but also play a role in creating a comfortable indoor thermal environment [21]. Luo et al. [105] divided the surveyed apartments into three groups based on the availability of personal control. They found that people with a higher degree of personal control tended to accept a wider range of indoor thermal environments. For the apartments that have individual control over the heating system, the neutral temperature was 2.6 °C lower than those without personal control. Kumar et al. [32] found that occupants’ preferred behavioral adaptations were to turn on ceiling fans (about 100%), followed by opening doors (about 80%) and windows (about 50%) to restore thermal comfort conditions. Yang et al. [15] found that through various behavioral adjustments, urban residents can recover thermal comfort conditions from uncomfortable environments more easily than residents in towns and rural areas. Yan [103] showed that in naturally ventilated buildings, people can better adapt to their environment through adaptive behaviors such as opening windows and using fans.

The behavior of occupants to improve the indoor environment plays an important role in saving building energy [122]. Positive behavioral adaptations can help reduce energy consumption, e.g., proper air-conditioning system control can lead to the improvement of residents’ thermal perceptions and building energy efficiencies [123]. The window opening behavior has a significant impact on the indoor environment and building energy consumption [124]. In the presence of a central heating system, people tend to open windows when the indoor temperature becomes too high, resulting in wasted energy [14,100]. In hot and humid regions, the electricity consumption of the air-conditioning system accounts for a major proportion in the summer. The cooling energy consumption will decrease by about 20% for every 1 °C increase in the temperature setpoint [49].

Window opening is an effective way to adjust the indoor thermal environment and indoor air quality. It can effectively affect the indoor thermal environment parameters [125] and the thermal comfort of the occupants [120] and can also reduce energy usage [126]. Operable windows enable thermal comfort over a wider range of indoor temperatures compared with non-operable windows [108]. Rijal et al. [127] showed that the probability of a window being opened depends on the three operating modes, i.e., free running (FR), heating (HT), or cooling (CL). In the FR mode, the chances of the windows being opened are much higher than in the CL or HT mode. Yin et al. [55] proposed that the possibility and duration of window opening in the heating period is significantly lower than that in the transition period, and ventilation is the main reason for window opening. Most occupants from the survey by Wang et al. [44] chose to open windows to enhance indoor ventilation and improve the indoor air quality.

The use of air conditioners can quickly and effectively Improve the indoor thermal environment to alleviate the thermal discomfort caused by high temperatures [50], but the economic burden caused by the use of air conditioners also limits people’s arbitrariness in using air conditioners, thus weakening people’s ability to control the environment [69]. The utilization rate of air conditioners increases with the increase in the outdoor temperature [68].

The use of electric heaters can increase the indoor air temperature, but it will cause the indoor air to be dry. Wang et al. [65] found that few people use electric heaters, and many occupants want to use air humidifiers to increase the indoor humidity. Cao et al. [94] found that once people adapt to a warm environment in the winter, they will lose their ability to adapt to cold climates, resulting in discomfort and energy wastage.

People can adapt to the cold environment through certain physiological adjustments. The four mechanisms that contribute to thermoregulation are sweating, shivering, vasodilation, and vasoconstriction [128]. When the body is exposed to cold ambient temperatures, the skin surface blood flow rate and skin temperature decrease, thereby preventing excessive body heat dissipation [41]. When a person is in a hot environment, the body starts to sweat and vasodilation occurs, which increases blood flow to cool the body [129].

In addition to physical regulation, psychological regulation also plays an important role. Psychological factors include many aspects, such as the cultural background, educational level, economic status, religious beliefs, visual impact, emotional state, and ability to control the environment [56]. Psychological adaptation can speed up the process of thermal adaptation to changes in outdoor climate conditions [130]. When living in a cold environment without indoor heating for a long time, the occupants will reduce their psychological expectations for the environment [75]. However, long-term thermal expectations in a comfortable indoor environment will increase, resulting in becoming pickier about the thermal environment [131]. Ning et al. [100] found that due to long-term exposure to artificially heated environments, the thermal adaptation of the subjects to cold climates gradually weakened. The improvement in living standards and popularity of the air-conditioning system led to a reduced tolerance to the natural ventilation environment of the occupants [50].

5.10. Implications and Limitations of the Findings

5.10.1. Implications of the Findings

The outcomes of this research can serve as a guide for the operation of air-conditioning systems in urban residential buildings, air-conditioning system design in new buildings, and for the policy makers in developing energy efficiency strategies without sacrificing the thermal comfort of the residents.

For example, the use of natural ventilation during the operation phase can effectively reduce building energy consumption and improve residents’ comfort. Further research should be conducted in relevant directions to use natural ventilation as an alternative to maintain the space’s thermal comfort.

Except for the severe cold region, the neutral temperatures in the summer are 0.4–2.0 °C higher than 26 °C (as required by the energy saving design standard in China), which means that the temperature setpoint in those climatic regions can be raised by 0.4–2.0 °C. As every 1 °C increase in the temperature setpoint leads to 20% cooling energy savings [49], an estimated 8–40% in cooling energy savings can be achieved. In addition, the neutral temperature in the hot summer and cold winter region is 0.7 °C lower than 18 °C (as required by the energy saving design standard in China), which means that the temperature setpoint in this type of climatic regions can be decreased by 0.7 °C, leading to about 14% in heating energy savings [132].

5.10.2. Limitations of the Findings

The number of papers is not evenly distributed across all the climate zones, which could lead to some bias in the conclusions. For example, few papers were found for the mild region and hot summer and warm winter region. Meanwhile, the adaptive models vary in different cities and climate zones. Further research should be conducted in each city to find the most suitable adaptive model to improve the thermal comfort of residents before planning the energy strategies.

6. Conclusions

This paper presents a systematic review of the research on thermal comfort of urban residential buildings in China. The influencing factors, research methods, and evaluation indicators are reviewed, and the results from previous researchers are summarized. The following conclusions can be made:

(1)

More investigations were conducted through field studies than laboratory studies. The field studies are strong in collecting actual and large occupant samples but weak in controlling the environmental parameters, which can be compensated by conducting laboratory studies.

(2)

The thermal sensation vote, operative temperature, thermal acceptance, mean thermal sensation, and predicted mean vote are the most commonly used thermal indices. The evaluation of the thermal comfort sensation might vary with different indices, and therefore, it is recommended to use multiple indices for evaluation.

(3)

People have higher and higher requirements for the indoor thermal environment. In severe cold and cold regions, the thermal neutral temperature in the winter tends to increase with time, and people have higher requirements for the indoor thermal environment in the winter.

(4)

The thermoneutral temperature varies in different seasons. There is little difference in the thermoneutral temperatures of different climate zones in the summer, but they vary greatly in the winter. In the summer, the thermoneutral temperature is higher in warmer climate zones and it is the contrary in the winter. Whether it is winter or summer, the thermoneutral temperature tends to increase with the indoor air temperature due to an adaptation to the indoor thermal environment.

(5)

The thermoneutral temperature of the human body is affected by the clothing thermal resistance. The correlation between the clothing thermal resistance and the thermoneutral temperature is weak in the summer and strong in the winter.

(6)

In the summer, the thermoneutral temperature is higher in warmer climate regions. However, in the winter, the thermoneutral temperature is lower in warmer climate regions.

(7)

In the same climate region, the neutral temperature in the summer is greater than the neutral temperature in the winter, especially in hot summer and cold winter regions and hot summer and warm winter regions.

6.1. Future Research Directions

Future research directions could focus on studying adaptive thermal comfort in automatically controlled indoor environments by taking into account the ambient weather conditions and occupants’ behaviors. Moreover, large-scale, thermal comfort studies can be conducted through big data-mining approaches.

6.2. Challenges

The challenges of the research lie in (1) how to obtain high-quality data to conduct large-scale surveys, and (2) how to take into account the impacts of differences in the social cultures and living habits on thermal comfort to more comprehensively understand the needs of residents.

6.3. Broader Implications of the Study

The outcomes of this research can serve as a guide for operating air-conditioning systems in urban residential buildings, designing air-conditioning systems for new buildings, and assisting policymakers in developing energy efficiency strategies that prioritize residents’ thermal comfort.