The Influence of Residential Behavior on Dwelling Energy Consumption and Comfort in Hot-Summer and Cold-Winter Zone of China—Taking Shanghai as an Example

Abstract

Human behavior plays a key role in building energy consumption, especially in residential buildings. Residents in the hot-summer and cold-winter zone of China tend to use energy in a “partial-time, part-of-the-space” manner. With the improvement in people’s living standards, the energy consumption of residential buildings in this region has been increasing. Exploring the building energy efficiency methods that suit the local residents’ energy use habits while ensuring building comfort levels is of great importance. This paper investigates the impact of living behavior on residential energy consumption and comfort in this region, focusing on personnel-in-room conditions, use of air-conditioning equipment, and window opening. Firstly, a questionnaire survey was conducted to analyze the energy use behaviors and the mentality of residents in the Jiangsu–Zhejiang–Shanghai Area of China. Then, based on the current situation of the household structure in Shanghai, 10 households were selected and monitored for a period of 8 months. Their energy use patterns were summarized to obtain a standard sample. Finally, the energy consumption simulation software DeST 2.0 was used to analyze the impact of each usage element on comfort and energy consumption. A dual-index evaluation of energy consumption and comfort is proposed for the first time. This paper found that respondents placed about a 6:4 important ratio on “comfort” and “energy efficiency” in their dwellings and that the importance placed on comfort increased from low-energy to high-energy households. Behaviors such as “gathering more in the living room” and “open windows as much as possible in the summer and appropriately in the winter” were more beneficial to both comfort and energy efficiency.

1. Introduction

According to the World Energy Statistics 2022 [1], China’s energy consumption has surpassed that of the United States and become the world’s largest energy consumer, accounting for 26.5% of total global energy consumption. With insufficient energy supply and increasing demand, China’s energy situation is not optimistic. Because of the different comfort requirements and energy use habits of residents of different countries in the indoor environment, the level of building energy consumption varies significantly [2]. Residents of developed countries, such as the United States, prefer mechanically controlled indoor environments with constant temperatures and humidity. In China, especially in non-heating areas, they are more inclined to natural adjustments, such as opening windows to improve indoor air quality, and use mechanical equipment only as an auxiliary means of temperature control [3]. This also makes the energy consumption of residential buildings in non-heating areas much lower than that of cities in developed countries with similar climates and scales.

At present, China needs to find a way to develop building energy efficiency that suits the usage habits of its residents and to minimize building energy consumption while ensuring building comfort. The varied living habits of people in different regions can greatly influence the energy consumption of residential buildings. China can be divided into five climate zones: the severe cold region, the cold region, the hot-summer and cold-winter region, the hot-summer and warm-winter region, and the mild region [4]. The hot-summer and cold-winter region mainly refers to the mid-low reaches of the Yangtze River and its surrounding areas. Its main climatic features are the sultry heat in summer and the wet cold in winter. The public’s demand for improving the living environment is very high, and it is a very critical part of China’s building energy efficiency work [5]. Households in this zone generally adopt a “partial-time, part-space” approach to energy use, so the energy consumption of residential buildings in these areas is much lower than that in northern heating areas. However, with the improvement in the standard of living, residents in the Jiangsu–Zhejiang–Shanghai Area have to use air-conditioning equipment for both heating in winter and cooling in summer to cope with the status quo of indoor stuffy heat in summer and wet cold in winter, which also greatly increases the energy consumption of residential buildings. For the energy use habits of residents in the hot-summer and cold-winter area, scholars have conducted in-depth studies [6]. However, fewer studies have focused on grouping the energy-using population in the hot-summer and cold-winter area and correlating their energy-using behaviors with building energy consumption and comfort. In conclusion, the impact of people’s behavior on residential energy consumption and comfort in the hot-summer and cold-winter area is particularly important. It not only provides new ideas for subsequent research but also offers theoretical support for the development of building energy-saving technology.

Therefore, this study is to explore the energy-saving model that meets the climatic conditions of the hot-summer and cold-winter region, the residents’ energy use behavior, and also ensures a certain level of comfort. It is necessary to investigate the impact of the region’s residential behavior on residential energy consumption and comfort. We will focus on the room conditions, air-conditioning equipment-use behaviors, and window-opening behaviors. We will also summarize the energy use patterns of a standard sample and simulate the effects of each energy use factor on energy consumption and comfort. Recommended energy use patterns are given at the end.

The further material is divided into several parts. Thus, Section 2 conducts a detailed and comprehensive literature review on the topic. Section 3 describes the materials and methods of the study. First, a questionnaire survey was conducted among the residents of the Jiangsu–Zhejiang–Shanghai Area. Then, 10 households in Shanghai were selected for data monitoring. Finally, the building energy simulation software DeST was used to calculate the experiments, and the dual indicators of energy consumption and comfort were used for evaluation. Section 4 presents the research results of each step. First, it collates and analyzes the results of the questionnaire research and summarizes the standard sample and its energy use pattern in the monitoring samples. Then, it collates the results of the software DeST to calculate the effect of each energy use factor on energy consumption and comfort and gives the proposed energy use pattern. Finally, an attempt is made to propose a comprehensive evaluation system of energy consumption and comfort for energy use patterns. Section 5 summarizes the conclusions of the study and suggests prospects for further research.

2. Literature Review

Scholars in many developed countries started their research on energy consumption in residential buildings very early and began to focus on the relationship between human behavior and energy consumption or comfort in residences in the 1980s. Warren and Parkins [7] conducted a 6-month survey of offices during the heating season and found that the outdoor temperature was the most influential factor affecting people’s window-opening behavior during this season. Kempton et al. [8] monitored the air-conditioning use patterns in eight American households and found significant differences in air-conditioning use patterns and energy consumption levels among them, which was the first study to categorize residential energy use behavior. Lutzenhiser [9] and Hitchcock [10] proposed a framework for people’s behavior in dwellings to describe their energy-using behavior. This framework includes the basic conditions of dwellings and the energy-using mindset of the residents. However, there is a lack of case studies to support them. Investigating residents’ energy use behaviors on a large scale, institutions and scholars in Japan are at the forefront of the world. The Architectural Institute of Japan [11] organized the Residential Energy Consumption Survey and Research Committee, which conducted a survey of 80 households in six regions of Japan within four years. This survey provided reliable data for the study of the relationship between residents’ energy use behavior and residential energy consumption. The study showed that the energy use behavior of residents in various regions is quite different, while the energy consumption in the Hokkaido region is much higher than that in the Kanto region. The climatic conditions of these two regions are similar to those of the cold region and the hot-summer and cold-winter region, respectively. This similarity provides a reference for the comparative study of residential energy use behavior and energy consumption in different regions of China. In addition, Olivia et al. [12] investigated the relationship between the residents’ energy use behavior and energy consumption in the Netherlands and showed that residents’ behavior has a significant impact on the level of energy consumption. Lyrian et al. [13] investigated the personnel behavior of low-energy residential buildings in Australia. It was found that the current behavioral description of people is not applicable to low-energy households, so the energy use model should be modified appropriately according to the actual situation. Yu et al. [14] proposed a new system of behavioral descriptions of people based on data mining technology. This system can precisely assess the energy-saving potential of a building by improving the accuracy of the description of people’s behavior. In the following year, Joao et al. [15] proposed a new energy consumption assessment model based on the Markov model. This model can more accurately predict the energy consumption of a building and the direction of energy consumption breakdown, which is very helpful for the development of building energy conservation. Alberto et al. [16] analyzed the data on the electricity bills of 130,000 households in Singapore to examine the trend of energy use among households with different income levels. In general, scholars in other countries around the world are more likely to evaluate personnel behavior from a sociological and economic perspective.

Chinese scholars started researching the relationship between people’s energy use behavior and residential building energy consumption or comfort relatively late. However, there have been a large number of studies in recent years that have gradually formed a scientific research system and focused on the behavior of air-conditioning equipment use and energy consumption during the air-conditioning and heating periods. They carried out further research in three directions: questionnaire research, data monitoring, and software simulation. Many Chinese scholars have studied the relationship between people’s behavior and residential energy consumption through a large number of questionnaire surveys and data measurements. Chen Shuqin [17] selected seven typical cities from five major climate zones in China and carried out a large number of studies. She found that the residents of Shanghai City adopted the “part-time, part-space” energy use pattern in both the winter and summer seasons. Chen Weihuang et al. [18] analyzed the influence of different factors on indoor thermal comfort and window-opening behavior. They used data monitoring and concluded that indoor and outdoor temperatures are the most important factors influencing residents’ window-opening behavior. In terms of software simulation, Zhang [19] took the lead in using DeST software, a building thermal environment simulation tool, to simulate the actual and regulated two energy consumption modes of a residence in Shanghai and analyze the impact of each influencing factor on energy consumption. This proves the reliability of DeST energy consumption simulation software and provides a reliable data simulation platform for subsequent scholars’ related research. After that, Zhang et al. subdivided the energy-using population and summarized the behavioral patterns of each group for simulation. Yang Shu [20] used statistical methods to conduct a detailed analysis of the correlation between energy-saving behaviors in residential buildings and family backgrounds. He argued that there are significant differences in the demographic factors affecting different types of energy-saving behaviors. This illustrates the necessity of discussing the association between personnel behavior and energy consumption by type. Cheng et al. [21] took a typical working household in Shanghai as a sample. They used data monitoring and DeST software simulation to analyze the correlation between residential energy use behavior and residential energy consumption in the summer. They found that personal behavioral patterns have a significant impact on energy consumption. This demonstrates the importance of categorically refining personnel-in-room conditions for research in this area. Many researchers in this field have since studied more refined definitions of presence situations and personnel behavior. Based on the current research on quantitative description and modeling of personnel behavior in buildings around the world, Sun [22] proposed a new method. She used an extended markup language schema (ob XML Schema) for a quantitative description of personnel behavior and also gave an analysis of the quantitative factors in personnel behavior that affect the energy consumption of residential buildings. On the basis of the above theories, the DeST simulation software platform has updated the personnel behavior module to improve the accuracy of the simulation results of other scholars.

In terms of thermal comfort evaluation, Yao et al. [23] carried out a study on the Danish scholar Povl Ole Fanger’s thermal comfort model. The Predicted Mean Vote (PMV) model is a standard for evaluating thermal comfort. Yao combined it with dynamic thermal simulation and computational fluid dynamics (CFD) simulation to create a free-running model for regulating thermal comfort in buildings. This is the first important study in China in this field. It is of great significance to the subsequent research on the relationship between human behavior and energy consumption or comfort. Since then, many researchers in this field have conducted a large number of investigations into the indoor comfort of people in different regions. Ji et al. [24] investigated the thermal comfort of respondents of different ages in non-air-conditioned environments in Jiangsu and Zhejiang in the summer. The actual temperature environment in Jiangsu and Zhejiang is mostly within the comfort range of the international standards 55-1992 [25] and ISO7730 [26]. These standards were created by the American Society of Heating, Refrigerating, and Air-Conditioning Engineers (ASHRAE). However, Ji found that 80% of the population said the upper limit of the tolerance temperature is 30 °C, which is much higher than the 26 °C that ASHRAE recommends. This indicates that residents in Jiangsu and Zhejiang have adapted to the local climate and improved their resistance to heat. It also suggests that the existing thermal comfort standards should be modified according to the actual situation. In 2012, the Ministry of Housing and Urban-Rural Development of China introduced the “Evaluation Standard for Indoor Thermal and Humid Environment of Civil Buildings GBT50785-2012” [27], which gives a clear definition and calculation method for indoor human thermal comfort and divides the indoor thermal and humid environments into three evaluation levels. On this basis, Zhu [28] proposed a more detailed rating of indoor thermal and humid environments in the same year and categorized the indoor environmental quality into five grades: “excellent, good, fair, poor, and poor”. It was also pointed out that “excellent” means that all people are highly satisfied with the environment; “good” means that most people are highly satisfied with the environment except for some sensitive people; “fair” means that the environment does not cause harm to people’s health but affects people’s comfort and has a high rate of dissatisfaction, but the standard can be relaxed if people stay in the same hot and humid environment for a long period of time; “bad” and “poor” are not suitable for people to stay for a long period of time. This study expands the specification and provides a detailed quantitative analysis of the thermal comfort situation with good data reliability. The thermal comfort evaluation system discussed later in this paper will be based on the results of this research.

In the past decades, international scholars have conducted a large number of studies on the correlation between energy use behavior and energy consumption or comfort. In general, it seems that the following stages have taken place: (1) proposing and revising the indoor personnel comfort evaluation system; (2) conducting a large-scale questionnaire survey and data measurement; (3) using the measured data and the first version of the DeST simulation software to verify and study the influence of factors on energy consumption; (4) using the new version of the DeST simulation software and a large amount of measured data to improve factors like personnel’s presence in the room and energy use behavior. However, the existing research still has the problems of insufficient segmentation of energy-using groups and a lack of comprehensive consideration of the dual-indicator evaluation of residential energy consumption and comfort. Furthermore, the data monitoring and energy consumption simulation part of the current study mostly focuses on the influence of single factors on energy consumption and lacks the linkage study of personnel-in-room conditions, air-conditioning energy use behavior, and window-opening behavior. Therefore, this paper is of great significance in studying the impact of residential behavior on residential energy consumption and comfort in hot-summer and cold-winter zones.

3. Materials and Methods

3.1. Study Areas

We selected the residents of downtown Shanghai as the study subjects. The climate of downtown Shanghai represents the typical climate of hot-summer and cold-winter areas, and the per capita Gross Domestic Product (GDP) has long been at the forefront, representing the level of economically developed high-density cities. Therefore, the relationship between energy use behavior and energy consumption or comfort of residents in downtown Shanghai can clearly reflect that of residents in high-density cities in hot-summer and cold-winter regions.

3.2. Sequence of the Conducted Research

The research methodology and sequence of this paper are divided into the following three steps:

(1)

Firstly, a large-scale questionnaire survey was conducted on the residents of Jiangsu–Zhejiang–Shanghai Area in the form of “online-oriented and offline-supplemented”. The survey aimed to analyze the energy use behaviors, residential energy consumption, and energy use mentality of the respondents from different family backgrounds.

(2)

Then, based on the current situation of Shanghai’s household structure, 10 households were selected to carry out data monitoring for 8 months. The data monitoring included people’s presence in the room, window-opening behavior, air-conditioning behavior, and air-conditioning energy consumption. The energy use patterns (air-conditioning period, room presence, energy use behavior, tolerance temperature, and set temperature) of the energy-using population in the winter and summer were analyzed, and the reliability of the data was verified with the results of the questionnaire survey.

(3)

Lastly, the standard sample and its energy use pattern are summarized in the data monitoring samples. Using the software DeST, the effects of each energy use factor on energy consumption and comfort are calculated and compared, and for the first time, a dual-indicator evaluation of energy consumption and comfort is proposed. Recommended energy use patterns are given at the end.

3.3. Questionnaire Survey

In recent years, some scholars have conducted research on the actual energy consumption in hot-summer and cold-winter regions [29,30]. However, in view of the special characteristics of the personnel composition and family structure in the Jiangsu–Zhejiang–Shanghai Area, such as a large number of floating and single populations, the previous studies have not been able to summarize the background of the actual energy-using population in this area. This paper will focus on the above phenomena in the questionnaire research stage and obtain the energy use behavior, energy consumption, and energy use mentality of the residents from different family backgrounds in the Jiangsu–Zhejiang–Shanghai Area.

This questionnaire research adopts the form of “mainly online and supplemented by offline”. The online part was carried out in December 2021, and the offline part was conducted during community research, focusing on those under 20 and 60 years old. A pre-survey was conducted in August 2021 to streamline the questionnaire design and ensure that the online questionnaire could be completed within 5 min. The questionnaire is divided into two parts: basic information and energy use habits (

Table 1. Questionnaire projects and their essential element.

Investigative Projects	Essential Element
Basic information about the family	Origin, age, family structure, rooming-in status
Basic building information	Number of floors, floor area, building orientation
Household energy usage	Air-conditioning (heating) calculation period, tolerance temperature range, set temperature range, window-opening habits
Household energy consumption	High, medium, and low energy consumption groups
Energy consumption mentality	Weighting of emphasis on residential comfort and energy efficiency, emphasis on comfort in different rooms

).

3.4. Household Data Monitoring Surveys

In this paper, based on the preliminary research, 10 households in Shanghai were selected to carry out data monitoring for a period of 8 months, from June 2021 to January 2022. The energy use patterns of the population in the winter and summer seasons were summarized, and the reliability of the data was verified with the results of the questionnaire survey. The standard samples were selected as input conditions for the DeST software simulation.

3.4.1. Survey Tools

In this paper, smart home sensors were used on the basis of previous studies and supplemented with window-opening behavior. The new generation of smart home sensors has the advantages of compactness, aesthetics, and data networking. It can observe various data points in the sample home in real time, which provides a new option for in-home data monitoring. The study selected temperature and humidity sensors, door and window sensors, human body sensors, and air-conditioning power sensors for sample observation (Figure 1).

3.4.2. Situation of the Sample Families

The sample households have a reasonable hierarchical distribution in terms of household composition, energy consumption level, and form of cooling (heating) equipment, which can better represent typical households of cities in hot-summer and cold-winter regions (

Table 2. Sample households’ background, cooling and heating equipment information, building information, and installation of monitoring equipment.

No.	Family Structure	Composition	Energy Consumption Level	Cooling and Heating Equipment	Floor Area	Monitoring Equipment Installation 1,2,3
						Living Room	Master Bedroom	Children’s Room	Elderly Room
A1	Three-generation lineal household	Couple + elderly + child	Low	Split air conditioner	90	★☆▲△	★☆▲△	★☆▲△	★☆▲△
A2	Three-generation lineal household	Couple + elderly + child	Medium	Central air conditioning	150	★☆▲		★☆▲	★☆▲
B1	Standard nuclear household	Couple + child	Medium	Split air conditioner	120	★☆▲△	★☆▲△	★☆▲△
B2	Standard nuclear household	Couple + child	Medium	Split air conditioner	150	★☆▲△	★☆▲△	★☆▲△
C1	Couple household	Young couple	High	Split air conditioner	90	★☆▲△	★☆▲△
C2	Couple household	Young couple	Medium	Split air conditioner	60		★☆▲△
C3	Couple household	Middle-aged couple	Medium	Split air conditioner	90	★☆▲△	★☆▲△
C4	Couple household	Young couple	High	Central air conditioning	120	★☆▲	★☆▲
D1	One-person household	Youth	Medium	Split air conditioner	30		★☆▲△
D2	One-person household	Youth	High	Split air conditioner	30		★☆▲△

¹ Temperature and humidity sensor—★; human body sensor—☆; door and window sensor—▲; air-conditioning power sensor—△. ² Households A2 and C4 have central air conditioning and are unable to monitor air-conditioning power, so no air-conditioning power sensors were installed. ³ The master bedroom of household A2 did not have a device installed as the data could not be accessed due to a network failure.

).

Figure 2 shows the per capita electricity cost of each household in the past Civil year, while those of C2, C3, and C4 users are not listed because they cannot be counted for the whole year. It can be seen that households’ electricity consumption trends are basically the same, with peaks in July–September and December–February. However, the air-conditioning and heating periods of each household are different. The air-conditioning period of A1, A2, and C1 households starts relatively late, and the end of this period is relatively late in A2 and C1. The heating periods for the households are relatively close to each other. In addition, most of the households are more resistant to cold during the heating period, while all households are less resistant to heat during the air-conditioning period.

3.4.3. Climatic Conditions

This paper compares the outdoor temperature data for a typical meteorological year in Shanghai with the measured data in 2021. The data are provided by the software DeST and are recorded hour-by-hour for summer and winter. We analyzed the difference in climatic conditions between this research monitoring period and the typical meteorological year (the average data of the last 10 years). It can be seen that the summer climate of the measured year is hotter than that of the typical meteorological year in the same period, and the temperature of more than 30 °C is more than twice as high as that of the previous years. The winter climate of the measured year is warmer than that of the typical meteorological year in the same period, and the extreme cold weather is significantly less than that of the typical meteorological year (

Table 3. Percentage of hot weather temperatures in the summer of the measured year versus the typical meteorological year of the same period.

Temperature	Percentage of Summer
	Typical Meteorological Year	Measurement Year
>34 °C	0.54%	3.03%
>32 °C	3.13%	12.86%
>30 °C	11.14%	27.99%
>28 °C	28.31%	53.71%
>26 °C	54.39%	82.52%

and

Table 4. Percentage of cold weather temperatures in the winter of the measured year versus the typical meteorological year of the same period.

Temperature	Percentage of Winter
	Typical Meteorological Year	Measurement Year
<0 °C	8.53%	0.81%
<2 °C	18.41%	5.85%
<4 °C	33.53%	16.53%
<6 °C	50.87%	33.67%
<8 °C	64.85%	56.52%

). In summary, the measured year is hotter in the summer and slightly warmer in the winter than the same period, which may result in higher measured cooling energy consumption and lower heating energy consumption in this study compared to previous studies.

3.5. Residential Energy Simulation and Analysis

Based on the actual energy use patterns (air-conditioning period, presence in the room, energy use behavior, tolerance temperature, and set temperature) from the preliminary data monitoring for residents in the Jiangsu–Zhejiang–Shanghai Area in the winter and summer, which were confirmed as reasonable by questionnaire research results, we chose a typical household A1 in the data monitoring and simulated its energy use pattern using the software DeST-h. We simulated the effects of each energy use factor on energy consumption and comfort to give the proposed energy use patterns.

Energy consumption and comfort conflict in the context of architectural research. Energy consumption is mostly used to maintain the indoor physical environment or electrical equipment and is closely related to the subjective perception of the user and the comfort of the physical environment. For example, a higher set temperature for indoor air conditioning in the winter usually means an increase in heating energy consumption. Lowering the energy consumption decreases the indoor air temperature, as does the comfort level.

In this paper, a total of six experiments were designed for three groups. In order to understand the effects of different influencing factors on energy consumption and comfort, the research was carried out in six aspects, including the presence of people in the room, the ventilation mode, the calculation period of air conditioning (heating), the tolerance temperature, the set temperature, and the air-conditioning operation mode.

The parameters to be entered in the software were building plan and geographic information, thermal parameters of the building envelope, meteorological parameters, indoor heat generation (people in the room), ventilation mode, air-conditioning and heating calculation period, tolerance temperature, set temperature, air-conditioning operation mode, and so on. Among them, the thermal parameters of the building envelope referred to the code setting, and the meteorological parameters referred to the data of a typical meteorological year. Indoor heat generation consists of two parts: indoor lighting heat generation and indoor personnel and equipment heat generation. The indoor lighting heat generation referred to in the code was set to 0.0141 kWh/m²·d. As for the indoor personnel and equipment heat generation part, this paper takes into account that the hour-by-hour room presence rate and activity state are different. Therefore, we present the concept of “hour-by-hour personnel thermal resistance” [31], which is calculated as in Equation (1):

M_i = R_ai × m_ai + R_bi × m_bi (i = 0, 1, 2, … , 23)

(1)

In the formula, M_i denotes personnel thermal resistance (W/m²); R_ai denotes the probability of being active in the room (%); R_bi denotes the probability of being rested in the room (%); m_ai denotes metabolic rate per unit of activity (W/m²); and m_bi denotes metabolic rate per unit of rest (W/m²).

Calculation mode A is the standard calculation mode with a standard set of personnel thermal resistance mode, ventilation mode, air-conditioning (heating) calculation period, tolerance temperature mode, set temperature mode, and air-conditioning operation mode. In the setting of the standard group, the personnel thermal resistance mode corresponding to the actual room presence of household A1 was set as “personnel thermal resistance mode 1”. The ventilation mode was characterized by the frequency of air changes, and the ventilation mode corresponding to the actual frequency of air changes was set as “ventilation mode 1”. The air-conditioning and heating calculation period of the sample households was obtained based on the empirical analysis and validation of the research results. Based on the measured tolerance temperature and set temperature of each room of Household A1 in each time period, the tolerance temperature mode and the set temperature mode were set as “tolerance temperature mode 1” and “set temperature mode 1”, respectively.

In addition, considering that the probability of air-conditioning use is also related to indoor temperature and humidity conditions, directly incorporating the actual probability of use into the calculations may lead to smaller experimental results. Therefore, this paper assumed an air-conditioning operation mode based on the actual use probability of air conditioning and set it as “air-conditioning operation mode 1”. The energy efficiency ratios for cooling and heating air conditioning refer to the recommended values of 3.1 and 2.5 in conventional codes.

In order to save energy consumption while ensuring comfort and to explore a balanced development model for them, this paper adopted a dual-evaluation index of energy consumption and comfort. PMV (Predicted Mean Vote) [32] is usually used internationally to describe the overall thermal sensation of people in hot and humid environments. Based on this, we refer to the refinement of the standard by relevant scholars in the previous literature [28], and we present the detailed class division standards and definitions in

Table 5. Predicted mean vote grading criteria.

Grading	Grading Criteria	Characteristics
Excellent	|PMV| ≤ 0.5	90% of the population is satisfied and suitable for human life.
Good	0.5 ≤ |PMV| ≤ 1	75% of the population is satisfied with the environment, except for sensitive people.
Average	1 ≤ |PMV| ≤ 1.5	The environment is not harmful to health, but it affects comfort and has a high dissatisfaction rate.
Not good	1.5 ≤ |PMV| ≤ 2	Prolonged stays are harmful to health and seriously affect comfort, leading to a very high rate of dissatisfaction.
Bad	|PMV| > 2	Harmful to human health, extremely poor comfort, and a very high rate of dissatisfaction.

. In this paper, we use the look-up table method to obtain the time-by-time PMV values of each room in the air-conditioning period (heating period). The data are derived from “Determination of PMV and PPD indexes of medium thermal environments and the provisions of thermal comfort conditions”. In this paper, we regard “excellent” and “good” as “compliance”, and we calculate the comfort compliance rate in the air-conditioning (heating) period as an evaluation index of indoor comfort. We also use the power consumption of air conditioning per unit area at the peak of energy usage during the air-conditioning and heating periods as an evaluation index of energy consumption (

Table 6. Evaluation criteria for energy consumption and comfort.

Grading	Comfort	Energy Consumption
★★★★★	More than 50% of “excellent, good”	Less than 6 kWh/m2 throughout the year
★★★★	50% of “excellent, good”	6.0–6.2 kWh/m2 throughout the year
★★★	49% of “excellent, good”	6.3 kWh/m2 throughout the year
★★	48% of “excellent, good”	6.4–7.2 kWh/m2 throughout the year
★	Less than 48% of “excellent, good”	Greater than 7.2 kWh/m2 throughout the year

★ are used here as a representation of the comfort and energy consumption evaluation grade, one ★ to five ★, with evaluation from one to five gradually rising.

).

Experiment 1 examines the effect of room presence on energy consumption and comfort with a different distribution of people in the room for calculation mode A versus calculation mode B. Experiment 2 explores the effect of ventilation patterns on energy consumption and comfort with different window openings for calculation mode A versus calculation modes C1 and C2. Experiment 3 simulates the effect of air-conditioning (heating) calculation periods on energy consumption and comfort, with different lengths of air-conditioning and heating periods for calculation mode A versus calculation modes D1 and D2. Experiment 4 looks at how personnel air-conditioning tolerance temperatures affect energy use and comfort. Tolerance temperature ranges for mode A and modes E1-4 are different. Experiment 5 investigates the effect of personnel air-conditioning set temperatures on energy consumption and comfort, with different set temperature ranges for calculation mode A versus calculation modes G1-G4. Experiment 6 explores the effect of personnel air-conditioning operation modes on energy consumption and comfort, with different probabilities of air-conditioning usage time for calculation mode A versus calculation modes H1 and H2.

In addition, in Experiment 4, taking calculation mode A as the standard mode, under the condition that other parameters are the same, the influence of tolerance temperature on energy consumption and comfort can be expressed by the influence factor [33], which is shown in Equations (2) and (3):

J_fe = (E_f − E_a) ÷ E_a

(2)

J_fp = (P_f − P_a) ÷ P_a

(3)

In the formula, J_fe denotes the energy consumption impact factor (%); E_f denotes the cooling (heating) energy consumption corresponding to each tolerance temperature (kWh/m²); E_a denotes the cooling (heating) energy consumption of the standard mode (kWh/m²); J_fp denotes the comfort impact factor (%); P_f denotes the cooling and heating comfort compliance rate corresponding to each tolerance temperature (%); and P_a denotes the cooling (heating) comfort compliance rate of the standard mode (%).

In Experiment 5, the influence of the set temperature on energy consumption and comfort can be expressed by the influence factor, which is shown in Equations (4) and (5):

J_ge = (E_g − E_a) ÷ E_a

(4)

J_gp = (P_g − P_a) ÷ P_a

(5)

In the formula, J_ge denotes the energy consumption impact factor (%); E_g denotes the cooling (heating) energy consumption corresponding to each set temperature (kWh/m²); J_gp denotes the comfort impact factor (%); and P_g denotes the cooling and heating comfort compliance rate corresponding to each set temperature (%).

4. Results and Discussion

4.1. Questionnaire Results

The questionnaire survey received a total of 293 questionnaires (239 online and 54 offline), of which 267 were valid (234 online and 33 offline). This study adopts the case deletion method and the partial deletion method to deal with the data, so the statistical sample size may be different. According to the data statistics, the respondents are widely spread, with a good distribution of age groups and family structures. The results of the survey are well-represented.

4.1.1. Basic Information on Respondents

The interviewed groups all live in the Jiangsu–Zhejiang–Shanghai Area. Forty-nine percentage of the respondents are from the Jiangsu–Zhejiang–Shanghai Area, with a similar background in energy use habits, while 21% of the respondents are from the northern regions, with living habits that are very different from those of the residents of the Jiangsu–Zhejiang–Shanghai Area. In terms of age, the main group of respondents is young people aged 21–40 years old, accounting for 50%, while the older and teenager groups account for a smaller proportion of 15%. The structure of the respondents’ families has the largest proportion of standard nuclear families, and the building area tends to increase accordingly with the increase in the number of permanent residents in the family.

The study of room presence omitted sleeping hours and was divided into three time periods: morning (6 a.m. to 12 p.m.), afternoon (12 p.m. to 18 p.m.), and evening (18 p.m. to 24 p.m.). The living room was used most frequently throughout the day, with peak usage during the day and the bedroom at night. Couples and children were more active outside during the day; the elderly were less active outside and were the family members most closely associated with the indoor thermal and humid environment. According to the literature [34], the main energy use times during the air-conditioning period are concentrated in the afternoon and evening, while during the heating period, it mainly occurs in the evening. It can be hypothesized that the presence of people in the room is one of the influencing factors that affect residential energy consumption.

4.1.2. Analysis of Energy Use

The start and end of the summer air-conditioning period for the surveyed households were respectively concentrated in late June to late July and September, while about 8% of the respondents did not turn on the air-conditioning during the summer. The start and end of the winter heating period for the surveyed households were respectively concentrated in late November to early December and late February. About 24% of the respondents do not turn on their heating equipment in the winter, but almost none of them are from northern China.

According to the results of the questionnaire survey, as shown in Figure 3 and Figure 4, the tolerated and set temperatures for the air-conditioning period in the surveyed households were 29–30 °C and 26 °C, respectively. The research shows that the residents of the Jiangsu–Zhejiang–Shanghai Area already have strong thermal resistance, and their tolerance temperature has risen to 29–30 °C. This high tolerance temperature greatly reduces the frequency of use of air-conditioning equipment and subsequently decreases residential energy consumption in the summer. In contrast, the tolerance and set temperatures of the surveyed households during the heating period were below 13 °C and above 20 °C, respectively. The aggregated data by age group show that the set and tolerance temperatures of the teenagers under 20 years old are significantly lower than those of other groups by 1–2 °C. This indicates that the teenagers are less heat-resistant than the other groups, and it also reflects that the age group is an influencing factor in the indoor set and tolerance temperatures of the residents during the air-conditioning period. By summarizing the data for the low, medium, and high energy consumption groups, we find that there is no significant difference in the air-conditioning set temperature of these groups. However, the tolerance temperature shows a significant difference. The heat tolerance of high-energy users is significantly lower than that of low-energy users, which has great potential for energy savings.

In terms of window-opening behavior in the summer, 80% of the interviewed households often open the windows in the morning and evening. In winter, the frequency of window openings is significantly reduced, and more people choose to warm their homes with equipment, which in turn raises heating energy consumption. By summarizing the window-opening situation of respondents in different energy consumption groups, it can be seen from Figure 5 that the high energy consumption group has a higher rate of not opening windows. It indicates that the high energy consumption group prefers mechanical temperature control and has a higher energy savings potential.

4.1.3. Analysis of Energy Consumption and Energy Use Mentality

Fifty-nine percentage of the interviewees are medium energy users, and the proportion of low and high energy users is 26% and 15%, respectively. The data conform to the law of normal distribution and are of some reference significance. The proportion of energy consumption groups in different household structures also varies. One-person households have the largest percentage of low-energy users; households with children (the elderly) have the smallest percentage of high-energy users; and standard nuclear households are a relatively weak group in terms of energy conservation awareness.

The overall weight of respondents on “comfort” and “energy efficiency” is about 6:4. Moreover, the weight of comfort increases from low energy consumption to high energy consumption in households, indicating that the higher the energy consumption, the more attention is paid to indoor comfort. In addition, the importance of room comfort for the respondent group, in descending order, is master bedroom, living room, kitchen, and second bedroom. The importance attached to comfort in different rooms varies among households with different family structures. When there are elderly people and children at home, the importance attached to comfort in the living room and the second bedroom rises significantly.

4.2. Analysis of Personnel Behavior in Sample Households

In this study, we organized several important indicators of the energy use patterns of sample households during the air-conditioning period and the heating period, including the air-conditioning period, room presence, energy use behaviors, tolerance temperatures, and set temperatures. We also categorized and summarized the analysis by room function and household structure.

4.2.1. Analysis of Personnel Behavior in the Summer

Based on the frequency of air-conditioning use, this paper focused on the monitoring period from 1 July 2021 to 30 September 2021 and determined the start and end of the air-conditioning period based on the average start and end of the air-conditioning period for all households. The measured average air-conditioning period for households is from 3 July to 15 September, which is within the interval of the questionnaire study, indicating that the air-conditioning period is relatively reasonable.

As illustrated in Figure 6, the room presence rate is low during the daytime, and the highest rates are found between 5:00–9:00 and 21:00–24:00, indicating that most of the households are at home during this period, which is the most likely time for energy use behaviors to occur. The peak hours for room activities are 6:00–10:00 and 18:00–23:00, and the main hours for resting in the room are 0:00–7:00. The hours corresponding to these two states of being in the room are the most important hours for indoor comfort, and due to the different activity states, the comfort conditions are also different.

A summary of the data by room function shows that the presence rate of the living room exceeds 60% except for the early morning hours, making it the most central functional space in the home, and as the resident population increases, the presence rate of the living room grows accordingly. The master bedroom and children’s bedroom have a lower occupancy rate from 6 to 17 p.m., while the elderly bedroom maintains a higher presence rate throughout the day because the master and children go out to work (study) during the day while the elderly stay at home, which is consistent with the results of the questionnaire survey.

In this paper, we propose a new concept, “active behavior”, which refers to the behaviors of households that help to change the indoor environment by opening windows and using air conditioners when they are inside the room. As indicated in Figure 7, the average probability of active behavior for the sample households is about 60%, with about 35% opening windows and 25% using the air conditioner. About half of the time, households do not engage in any active behavior, and when they do, they mainly open windows, and air conditioning is only an auxiliary means of cooling. This is also an important reason why the energy consumption of air conditioners in the Jiangsu–Zhejiang–Shanghai Area is much lower than that of developed countries. From the perspective of window-opening behavior, the window-opening rate is significantly higher during the day than at night and continues to decline after a slight increase in the evening, and the window-opening pattern can be summarized as “often open in the morning and evening”. This is consistent with the results of the questionnaire survey. The trend of air-conditioning behavior is just the opposite of window-opening behavior. The air-conditioning behavior peaks in the evening, and the air-conditioning usage rate exceeds the window-opening rate after 20:00 p.m. The summer air-conditioning usage pattern of households can be summarized as “often in the afternoon and evening”. In terms of energy use behavior, the probability of opening windows in the master bedroom is significantly lower than that in other rooms, while the probability of air-conditioning use is significantly higher than that in other rooms. The probability of air-conditioning use in the bedrooms of the elderly and children is very low and occurs mostly in the afternoon and evening.

Households A1, C1, and D2 have the highest probability of active behavior, but their behavioral composition is completely different. Household D2, with a probability of air-conditioning use close to 60%, is a typical high-energy household. Households A1, B1, and C3 have a probability of air-conditioning use of less than 20% and are low-energy users. This is consistent with the conclusions obtained from previous years’ electricity bills and illustrates the reasonableness of the monitoring data. The frequency of air-conditioning use during the air-conditioning period for different household structures, in descending order, is one-person households, young couple households, standard nuclear households, three-generation lineal households, and elderly couple households.

By summarizing the data, it was found that the sample homes had a tolerance temperature of 29 °C, a set temperature of 27 °C, and an average temperature of 29 °C when the air-conditioning unit was not turned on. This is consistent with the data obtained from the questionnaire. The tolerance temperature of the master bedroom was 1 °C lower than the other rooms, and the set temperature was also lower. The setting temperature of the elderly bedroom was high and the same as the tolerance temperature, indicating that the elderly are more resistant to heat. Tolerance temperatures were lowest in one-person households and highest in an elderly couple household C3. The high energy-consuming one-person household D2 had a low set temperature of 25 °C, resulting in a large amount of energy consumption and a high potential for energy savings.

4.2.2. Analysis of Personnel Behavior in the Winter

The measured average heating period for households is from 7 December to February of the following year. This matches the 1 December to 22 February interval of the questionnaire study. As shown in Figure 8, the trend of room presence during the heating period of the sample households is similar to that of the air-conditioning period. However, the overall room presence rate is higher than that of the air-conditioning period, with the average rate exceeding 60%. This indicates that households prefer to stay at home during the heating period.

The average probability of active behavior for the sample households is about 40%, with about 20% opening windows and 20% using air conditioning. It is significantly lower than the air-conditioning period, where the probability of opening windows is only about half that of the air-conditioning period and the probability of air-conditioning use is slightly lower than that of the air-conditioning period. Residents of the Jiangsu–Zhejiang–Shanghai Area prefer to protect themselves from the cold by adding more clothing, and their ability to resist the cold has also increased. Therefore, the energy consumption of heating equipment in winter is much lower than that in northern heating areas and developed countries.

The active behavior of the sample households during the heating period shows the lowest probability of adopting active behaviors after the night break in terms of the time period. Window-opening and air-conditioning behaviors complement each other during the active hours (9:00 to 24:00), maintaining the active behavior relatively stable at around 40% with no significant peaks (Figure 9).

In terms of window-opening behavior, it remained stable at around 20% throughout the day, dropping slightly to around 15% between 20:00 and 24:00, when the probability of room presence and air-conditioning use was highest, suggesting that nighttime is a time for family gatherings. Residents in the area prefer using heating equipment. The pattern of window opening during the heating period can be summarized as “constant throughout the day, slightly lower at night”. This is consistent with the results of the questionnaire survey. In terms of air-conditioning use behavior, the probability of air-conditioning use is significantly higher during activity time than after resting at night and continues to increase with the time of day, reaching a peak between 20:00 and 24:00 at night. The pattern of air-conditioning use during the heating period can be summarized as “lowest after bedtime, slightly higher at night” (Figure 7).

The probability of active behavior in each room in descending order is living room, master bedroom, children’s bedroom, and elderly bedroom. It indicates that the elderly have the least willingness to make active changes to the indoor environment during the heating period. The living room, on the other hand, has more active behaviors towards the indoor environment due to the intensive activities of the people. In terms of energy use behavior, the probability of window-opening and air-conditioning use in the living room is the highest, and it occurs mostly in the evening when the family gathers (18:00 to 24:00). The probability of air-conditioning use is lowest in the elderly’s bedroom, and the probability of window opening is lowest in the children’s bedroom. One-person households are more likely to use air conditioning than three-generation lineal households, although they are all below 20% and are not considered high-energy users during the heating period. This is consistent with the conclusions drawn from previous years’ electricity bills and illustrates the reasonableness of the monitoring data.

The sample households had a tolerance temperature of 18 °C during the heating period and a set temperature of 20 °C. The tolerance temperature during the heating period was the same for all rooms, and the set temperature varied considerably. The living room and master bedroom have the highest set temperatures, and the children’s bedroom has the lowest set temperature. The tolerance temperature and setting temperature during the heating period of the three-generation lineal households are significantly lower than those of other families, and they are typically low energy consumption families during the heating period.

The average air-conditioning period in summer and the average heating period in winter measured by the residents were consistent with the interval obtained from the questionnaire survey. In the case of the living room with a high occupancy rate, the window-opening pattern of “always on in the morning and evening” during the air-conditioning period and “constant throughout the day and slightly lower at night” during the heating period is also consistent with the findings of the questionnaire survey. The measured tolerance and setting temperatures during the air-conditioning period are in line with the questionnaire findings of 29–30 °C and 26 °C, respectively. The probability of air-conditioning use is also consistent with the conclusions drawn from previous years’ electricity bills, which also verifies the reasonableness of the monitoring data.

4.3. Simulation and Analysis of Residential Energy Consumption

We selected the typical residence A1, which covers the widest age range, as the standard sample based on the actual energy use patterns of residents in the Jiangsu–Zhejiang–Shanghai Area in the winter and summer seasons obtained from data monitoring. The software DeST-h conducted the simulation in conjunction with its energy use pattern to examine the impact of each energy use behavior factor on energy consumption and comfort. Residence A1 is located in Shanghai, with the main building orientation facing south and a floor area of 90 m² (Figure 10).

4.3.1. Simulation Experiment 1: Impact of Room Presence on Energy Consumption and Comfort Levels

The DeST input parameters are given in

Table 7. Experiment 1 input conditions.

Project		Calculation Mode A	Calculation Mode B
Meteorological parameters		Typical meteorological year
Indoor calorific value		Indoor lighting: 0.0141 kWh/m2·d Indoor personnel and equipment: Personnel thermal resistance mode 1	Indoor lighting: 0.0141 kWh/m2·d Indoor personnel and equipment: Personnel thermal resistance mode 2
Indoor and outdoor ventilation mode		Ventilation mode 1
Air-conditioning equipment use mode	Air-conditioning and heating calculation period	Air-conditioning period: 3 July to 15 September Heating period: 1 December to 28 February of the following year
	Air-conditioning tolerance temperature	Tolerance temperature mode 1
	Air-conditioning setting temperature	Set temperature mode 1
	Air-conditioning operation mode	Air-conditioning operation mode 1
	Air-conditioning and heating energy efficiency ratio	Air-conditioning energy efficiency ratio: 3.1 Heating energy efficiency ratio: 2.5
Others		Kitchen and bathroom without temperature control

. Calculation mode A is the standard calculation mode and uses the standard group. The differences between the two occupant thermal resistance modes are that in “personnel thermal resistance mode 1”, the living room has a higher presence rate, and family members are more likely to congregate in the living room. In “personal thermal resistance mode 2”, households prefer to stay in their bedrooms. The total thermal resistance is the same in both modes.

Calculation mode A shows an increase in comfort level and less energy consumption compared to calculation mode B, which indicates that “gathering in the living room” is more favorable than “staying in one’s own room” both in terms of comfort and energy savings (

Table 8. Experiment 1 calculation results.

Mode	Cooling Energy Consumption (kWh/m2)	Heating Energy Consumption (kWh/m2)	Annual Energy Consumption (kWh/m2)	Comfort Compliance Rate (%)	Comfort Evaluation	Energy Saving Evaluation
Calculation mode A	4.8	1.5	6.3	50%	★★★★	★★★
Calculation mode B	4.0	2.4	6.4	49%	★★★	★★

★ are used here as a representation of the comfort and energy saving evaluation grade, one ★ to five ★, with evaluation from one to five gradually rising.

).

4.3.2. Simulation Experiment 2: Influence of Ventilation Mode on Energy Consumption and Comfort Levels

The input parameters of the DeST software are given in

Table 9. Experiment 2 input conditions.

Project		Calculation Mode A	Calculation Mode C1	Calculation Mode C2
Meteorological parameters		Typical meteorological year
Indoor calorific value		Indoor lighting: 0.0141 kWh/m2·d Indoor personnel and equipment: Personnel thermal resistance mode 1
Indoor and outdoor ventilation mode		Ventilation mode 1	Ventilation mode 2	Ventilation mode 3
Air-conditioning equipment use mode	Air-conditioning and heating calculation period	Air-conditioning period: 3 July to 15 September Heating period: 1 December to 28 February of the following year
	Air-conditioning tolerance temperature	Tolerance temperature mode 1
	Air-conditioning setting Temperature	Set temperature mode 1
	Air-conditioning operation mode	Air-conditioning operation mode 1
	Air-conditioning and heating energy efficiency ratio	Air-conditioning energy efficiency ratio: 3.1 Heating energy efficiency ratio: 2.5
Others		Kitchen and bathroom without temperature control

. The differences between the three ventilation modes are as follows: Ventilation mode 1 is the actual window-opening situation, which means that there is a certain probability of opening the window when the air conditioner is not in use. Ventilation mode 2 is to open the window only without turning on the air conditioner, and ventilation mode 3 is basically not opening windows.

The comfort evaluations were the same for all three modes, indicating that window opening does not affect comfort (

Table 10. Experiment 2 calculation results.

Mode	Cooling Energy Consumption (kWh/m2)	Heating Energy Consumption (kWh/m2)	Annual Energy Consumption (kWh/m2)	Comfort Compliance Rate (%)	Comfort Evaluation	Energy Saving Evaluation
Calculation mode A	4.8	1.5	6.3	50%	★★★★	★★★
Calculation mode C1	4.0	1.8	5.8	50%	★★★★	★★★★★
Calculation mode C2	6.1	1.8	7.9	50%	★★★★	★

★ are used here as a representation of the comfort and energy saving evaluation grade, one ★ to five ★, with evaluation from one to five gradually rising.

). However, the energy consumption is very different; the “no air conditioning, only opening windows” mode is the most energy efficient among the three modes for the whole year. The energy consumption of this mode is higher than that of mode A during the heating period, which suggests that windows should be opened when the heating equipment is not in use. Comparing the temperature in the living room of household A1 with that of the outdoors, it is found that the outdoor temperature is cooler than the indoor temperature in the morning and evening of the summer, so it is better to use natural ventilation (Figure 11). In winter, the outdoor temperature in the middle of the day is not too low compared to the indoor temperature, so the windows can be opened appropriately for ventilation. Therefore, the optimal ventilation pattern is to “open the windows as much as possible in summer and open them appropriately in winter”.

4.3.3. Simulation Experiment 3: Impact of the Air-Conditioning (Heating) Calculation Period on Energy Consumption and Comfort Levels

As given in

Table 11. Experiment 3 input conditions.

Project		Calculation Mode A	Calculation Mode D1	Calculation Mode D2
Meteorological parameters		Typical meteorological year
Indoor calorific value		Indoor lighting: 0.0141 kWh/m2·d Indoor personnel and equipment: Personnel thermal resistance mode 1
Indoor and outdoor ventilation mode		Ventilation mode 1
Air-conditioning equipment use mode	Air-conditioning and heating calculation period	Air-conditioning period: 3 July to 15 September Heating period: 1 December to 28 February of the following year	Air-conditioning period: 15 July to 31 August Heating period: 15 December to 15 February of the following year	Air-conditioning period: 15 June to 30 September Heating period: 15 November to 15 March of the following year
	Air-conditioning tolerance temperature	Tolerance temperature mode 1
	Air-conditioning setting temperature	Set temperature mode 1
	Air-conditioning operation mode	Air-conditioning operation mode 1
	Air-conditioning and heating energy efficiency ratio	Air-conditioning energy efficiency ratio: 3.1 Heating energy efficiency ratio: 2.5
Others		Kitchen and bathroom without temperature control

, the air-conditioning and heating periods in calculation mode D1 are shorter than those in calculation mode A by 1 month each, while calculation mode D2 is longer by 1 month each.

The results of Experiment 3 are given in

Table 12. Experiment 3 calculation results.

Mode	Cooling Energy Consumption (kWh/m2)	Heating Energy Consumption (kWh/m2)	Annual Energy Consumption (kWh/m2)	Comfort Compliance Rate (%)	Comfort Evaluation	Energy Saving Evaluation
Calculation mode A	4.8	1.5	6.3	50%	★★★★	★★★
Calculation mode D1	4.0	1.6	5.6	46%	★	★★★★★
Calculation mode D2	5.7	2.1	7.8	51%	★★★★★	★

★ are used here as a representation of the comfort and energy saving evaluation grade, one ★ to five ★, with evaluation from one to five gradually rising.

. As the air-conditioning (heating) period is shortened, the energy consumption is reduced accordingly, but the comfort level is seriously reduced. When the air-conditioning (heating) period is extended, the comfort level is slightly improved, but the energy consumption increases significantly. Both are unbalanced energy use patterns, so it is not recommended to shorten or lengthen the air-conditioning (heating) period excessively.

4.3.4. Simulation Experiment 4: Impact of Air-Conditioning Tolerance Temperature on Energy Consumption and Comfort Levels

As given in

Table 13. Experiment 4 input conditions.

Project		Calculation Mode A	Calculation Mode E1	Calculation Mode E2	Calculation Mode E3	Calculation Mode E4
Meteorological parameters		Typical meteorological year
Indoor calorific value		Indoor lighting: 0.0141 kWh/m2·d Indoor personnel and equipment: Personnel thermal resistance mode 1
Indoor and outdoor ventilation mode		Ventilation mode 1
Air-conditioning equipment use mode	Air-conditioning and heating calculation period	Air-conditioning period: 3 July to 15 September Heating period: 1 December to 28 February of the following year
	Air-conditioning tolerance temperature	Tolerance temperature mode 1	Air-conditioning period: 27 °C Heating period: 19 °C	Air-conditioning period: 28 °C Heating period: 18 °C	Air-conditioning period: 30 °C Heating period: 16 °C	Air-conditioning period: 31 °C Heating period: 15 °C
	Air-conditioning setting temperature	Set temperature mode 1
	Air-conditioning operation mode	Air-conditioning operation mode 1
	Air-conditioning and heating energy efficiency ratio	Air-conditioning energy efficiency ratio: 3.1 Heating energy efficiency ratio: 2.5
Others		Kitchen and bathroom without temperature control

, the average tolerance temperature range of calculation mode A is 17–29 °C, and the tolerance temperature ranges of calculation modes E1–E4 increase sequentially.

The relationship between tolerance temperature and energy consumption or comfort during the air-conditioning period is shown in Figure 12. Based on Equations (2) and (3), taking the calculation mode A as the benchmark, which is about 29 °C, it can be seen that with the rise in tolerance temperature, the comfort degree basically shows a linear decreasing trend, while the energy consumption shows a stepwise decreasing trend. When the tolerance temperature rises to 30 °C or falls to 28 °C, there is a sharp rise in energy consumption or a sharp drop in comfort, so it is not recommended. In summary, the optimal tolerance temperature is 29 °C in terms of energy consumption and comfort during the air-conditioning period.

The relationship between tolerance temperature and energy consumption or comfort during the heating period is shown in Figure 13. Taking calculation mode A as a benchmark, which is about 17 °C, it can be seen that as the tolerance temperature rises, comfort and energy consumption basically have a linearly increasing trend, and the growth rate is faster than that of the air-conditioning period. When the tolerance temperature rises to 18 °C or falls to 16 °C, there is a sharp rise in energy consumption or a sharp drop in comfort, so it is not recommended. In summary, the optimal tolerance temperature during the heating period is 17 °C.

The calculation results are given in

Table 14. Experiment 4 calculation results.

Mode	Cooling Energy Consumption (kWh/m2)	Heating Energy Consumption (kWh/m2)	Annual Energy Consumption (kWh/m2)	Comfort Compliance Rate (%)	Comfort Evaluation	Energy Saving Evaluation
Calculation mode A	4.8	1.5	6.3	50%	★★★★	★★★
Calculation mode E1	6.6	2.2	8.9	56%	★★★★★	★
Calculation mode E2	6.3	2.1	8.4	56%	★★★★★	★
Calculation mode E3	4.7	1.2	6.0	48%	★★	★★★★
Calculation mode E4	2.7	0.9	3.6	43%	★	★★★★★

★ are used here as a representation of the comfort and energy saving evaluation grade, one ★ to five ★, with evaluation from one to five gradually rising.

, and the comfort evaluation decreases with the expansion of the tolerance temperature range while the energy-saving evaluation increases correspondingly.

4.3.5. Simulation Experiment 5: Impact of Air-Conditioning Setting Temperature on Energy Consumption and Comfort Levels

As displayed in

Table 15. Experiment 5 input conditions.

Project		Calculation Mode A	Calculation Mode F1	Calculation Mode F2	Calculation Mode F3	Calculation Mode F4
Meteorological parameters		Typical meteorological year
Indoor calorific value		Indoor lighting: 0.0141 kWh/m2·d Indoor personnel and equipment: Personnel thermal resistance mode 1
Indoor and outdoor ventilation mode		Ventilation mode 1
Air-conditioning equipment use mode	Air-conditioning and heating calculation period	Air-conditioning period: 3 July to 15 September Heating period: 1 December to 28 February of the following year
	Air-conditioning tolerance temperature	Tolerance temperature mode 1
	Air-conditioning setting temperature	Set temperature mode 1	Air-conditioning period: 25 °C Heating period: 20 °C	Air-conditioning period: 26 °C Heating period: 19 °C	Air-conditioning period: 28 °C Heating period: 17 °C	Air-conditioning period: 29 °C Heating period: 16 °C
	Air-conditioning operation mode	Air-conditioning operation mode 1
	Air-conditioning and heating energy efficiency ratio	Air-conditioning energy efficiency ratio: 3.1 Heating energy efficiency ratio: 2.5
Others		Kitchen and bathroom without temperature control

, the average set temperature range of calculation mode A is 18–27 °C, and the set temperature ranges of calculation modes F1–F4 increase sequentially.

The relationship between the set temperature and energy consumption or comfort during the air-conditioning period is shown in Figure 14. Based on Equations (4) and (5), taking calculation mode A as a benchmark, which is about 27 °C, it can be seen that as the set temperature decreases, the energy consumption basically increases linearly, while the comfort level peaks at 27 °C. When the set temperature is lower than 27 °C or higher than 27 °C, the energy consumption rises sharply or the comfort level drops sharply. The setting temperature of the air conditioner can be adjusted appropriately. In summary, the optimal setting temperature for the air-conditioning period is 27 °C, and it can be increased appropriately for people who are sensitive to low temperatures.

The relationship between the set temperature and energy consumption or comfort during the heating period is shown in Figure 15. Taking the calculation mode A as the benchmark, which is about 18 °C, it can be seen that the energy consumption basically has a linear growth trend as the setting temperature rises, while the comfort level stabilizes after 18 °C and no longer grows. In summary, the optimal set temperature for the heating period is 18 °C.

The results of the experimental calculations are given in

Table 16. Experiment 5 calculation results.

Mode	Cooling Energy Consumption (kWh/m2)	Heating Energy Consumption (kWh/m2)	Annual Energy Consumption (kWh/m2)	Comfort Compliance Rate (%)	Comfort Evaluation	Energy Saving Evaluation
Calculation mode A	4.8	1.5	6.3	50%	★★★★	★★★
Calculation mode F1	5.8	2.3	8.1	45%	★	★
Calculation mode F2	5.3	1.9	7.2	46%	★	★★
Calculation mode F3	3.5	1.1	4.5	46%	★	★★★★★
Calculation mode F4	2.1	0.8	2.9	46%	★	★★★★★

★ are used here as a representation of the comfort and energy saving evaluation grade, one ★ to five ★, with evaluation from one to five gradually rising.

, where the energy savings evaluation increases with the expansion of the setting temperature range, but the comfort level does not increase correspondingly.

4.3.6. Simulation Experiment 6: Impact of Air-Conditioning Operation Modes on Energy Consumption and Comfort Levels

As displayed in

Table 17. Experiment 6 input conditions.

Project		Calculation Mode A	Calculation Mode G1	Calculation Mode G2
Meteorological parameters		Typical meteorological year
Indoor calorific value		Indoor lighting: 0.0141 kWh/m2·d Indoor personnel and equipment: Personnel thermal resistance mode 1
Indoor and outdoor ventilation mode		Ventilation mode 1
Air-conditioning equipment use mode	Air-conditioning and heating calculation period	Air-conditioning period: 3 July to 15 September Heating period: 1 December to 28 February of the following year
	Air-conditioning tolerance temperature	Tolerance temperature mode 1
	Air-conditioning setting temperature	Set temperature mode 1
	Air-conditioning operation mode	Air-conditioning operation mode 1	Air-conditioning operation mode 2	Air-conditioning operation mode 3
	Air-conditioning and heating energy efficiency ratio	Air-conditioning energy efficiency ratio: 3.1 Heating energy efficiency ratio: 2.5
Others		Kitchen and bathroom without temperature control

, the differences between the three air-conditioning operation modes are that mode 1 is an assumed operation mode based on the actual use probabilities of air-conditioning in households, and differentiation is made between different bedrooms. Mode 2 is an air-conditioning operation mode that ignores the differences in energy users and is set according to a conventional understanding. Mode 3 is an “all-time, all-space” energy use mode.

As displayed in

Table 18. Experiment 6 calculation results.

Mode	Cooling Energy Consumption (kWh/m2)	Heating Energy Consumption (kWh/m2)	Annual Energy Consumption (kWh/m2)	Comfort Compliance Rate (%)	Comfort Evaluation	Energy Saving Evaluation
Calculation mode A	4.8	1.5	6.3	50%	★★★★	★★★
Calculation mode G1	4.6	1.8	6.4	48%	★★	★★
Calculation mode G2	9.0	4.3	13.3	78%	★★★★★	★

★ are used here as a representation of the comfort and energy saving evaluation grade, one ★ to five ★, with evaluation from one to five gradually rising.

, the energy consumption of calculation mode G1 is slightly increased compared with calculation mode A, but the comfort level is decreased. Calculation mode G2 significantly improves comfort, but energy consumption also increases significantly, which is not conducive to the promotion of energy savings. In conclusion, the air-conditioning operation mode proposed in calculation mode A is the best.

It can be seen from the results of the six experiments that the window-opening situation does not have an effect on the comfort level. As the air-conditioning (heating) period is shortened, energy consumption correspondingly decreases, but the comfort level is severely degraded. When the air-conditioning (heating) period is lengthened, the comfort level gets a small improvement, but the energy consumption increases significantly. As the tolerance temperature increased, the comfort level basically showed a linear decline, while the energy consumption showed a stepwise decline. The optimal tolerance temperatures were 29 °C for the air-conditioning period and 17 °C for the heating period. The optimum setting temperature was 27 °C during the air-conditioning period and 18 °C during the heating period.

4.4. Comprehensive Evaluation System for Energy Consumption and Comfort

In this section, we try to establish a comprehensive evaluation system for energy consumption and comfort. This is based on the statistical results of the “energy use mentality” in the questionnaire survey and the simulation results. The weighting of how important “comfort” and “saving energy” are to the respondents is about 6:4, and by applying this weighting to the dual-indicator evaluation in the simulation analysis above, we can generate

Table 19. Comprehensive evaluation system.

Factors	Calculation Mode		Comfort Evaluation	Energy Saving Evaluation	Comprehensive Evaluation
The standard group	A		★★★★	★★★	★★★★
Room presence	B		★★★	★★	★★★
Ventilation mode	C	C1	★★★★	★★★★★	★★★★
		C2	★★★★	★	★★★
Air-conditioning (heating) calculation period	D	D1	★	★★★★★	★★★
		D2	★★★★★	★	★★★
Air-conditioning tolerance temperature	E	E1	★★★★★	★	★★★
		E2	★★★★★	★	★★★
		E3	★★	★★★★	★★★
		E4	★	★★★★★	★★★
Air-conditioning set temperature	F	F1	★	★	★
		F2	★	★★	★
		F3	★	★★★★★	★★★
		F4	★	★★★★★	★★★
Air-conditioning operation modes	G	G1	★★	★★	★★
		G2	★★★★★	★	★★★

★ are used here as a representation of the comfort and energy saving evaluation grade, one ★ to five ★, with evaluation from one to five gradually rising.

.

5. Conclusions and Prospects for Further Research

In recent years, more and more scholars have realized the important influence of personnel behavior on energy consumption and comfort. They have launched relevant research, but there is still room for improvement in data monitoring and evaluation indexes. Based on previous studies, this paper mainly makes the following innovations:

(1) This study subdivided the energy-using people by family structure and analyzed their energy-using patterns through questionnaire research, data monitoring, and software simulation. This provides a new idea for the typological study of the impact of personnel behavior on energy consumption and comfort. It also improves the precision of the research conclusions.

(2) In previous studies [35,36], data monitoring mostly used temperature and humidity recorders, power meters, and infrared body sensors to record indoor temperature and humidity conditions, air-conditioning usage, and movement of people in the room, respectively. This study adopts a new generation of smart home sensors to measure and comprehensively analyze the presence of sample households, indoor temperature and humidity, air-conditioning energy use behavior, and window-opening behavior. It then summarizes their energy use patterns to obtain a standard sample. During the study, the concepts of “comprehensive energy use behavior” and “active behavior” were proposed to more accurately describe the energy use behavior of households.

(3) We distinguish between “activity in the room” and “resting in the room” and more accurately describe the indoor comfort situation.

(4) The dual-evaluation indexes of energy consumption and comfort are proposed for the first time. We also propose a comprehensive evaluation system based on the results of the research, which provides a new way of thinking about the criteria for judging energy use patterns and improves the accuracy and practicability of the evaluation results.

The main outcome of the work and conclusions of this paper are as follows:

(1) The proportion of respondents emphasizing “comfort” and “energy efficiency” in dwellings is about 6:4, and the importance of comfort increases from low-energy to high-energy households. In terms of room presence, “gathering in the living room” is more favorable than “staying in one’s own room” in terms of both comfort and energy savings.

(2) The tolerated temperature during the air-conditioning period was 29 °C, and the set temperature was 27 °C. The tolerated temperature during the heating period was 18 °C, and the set temperature was 20 °C. The tolerance temperatures for air conditioning in different energy consumption groups showed significant differences, with high-energy users being significantly less resistant to heat than low-energy users, which has great potential for energy savings.

(3) The average probability of active behavior for the sample households was about 60% during the air-conditioning period, of which 35% opened windows and 25% used air conditioning. During the heating period, the average probability of active behavior was about 40%, with about 20% opening windows and 20% turning on the air conditioner. The pattern of window opening during the air-conditioning period can be summarized as “often in the morning and evening”, and the pattern of air-conditioning use as “often in the afternoon and evening”. The window-opening pattern during the heating period can be summarized as “constant throughout the day and slightly lower at night”, and the air-conditioning usage pattern as “lowest after bedtime and slightly higher at night”. “Basically not opening windows” will result in a large amount of energy consumption. “Opening windows as much as possible in summer and appropriately in winter” is the most favorable for energy savings. It is recommended that windows be opened in the morning and evening in summer and at noon in winter for ventilation.

The above-suggested methods can also be applied to other hot-summer and cold-winter regions of the world, but due to the different periods of development of building energy consumption in each country as well as the special characteristics of Chinese residents’ energy use habits, corresponding adjustments should be made according to the actual climatic conditions of other countries and regions, residents’ requirements for the comfort of the indoor environment, residents’ behaviors, and their energy use habits. Building energy consumption should be minimized under the premise of ensuring building comfort.

This paper proposes some innovations in monitoring methods and evaluation indexes based on previous studies. However, there are still some deficiencies due to the limitations of research time and scope. These are expected to be improved in subsequent studies. Firstly, this paper selects 10 households based on the current situation of household structure in Shanghai, but due to the small sample size and large chance, an accurate typology cannot be obtained. The accuracy of summarizing the energy use patterns for different household structures needs to be improved, and in the subsequent study, we will obtain a larger sample size of data to monitor and exclude the interference of factors, such as floor area, number of floors, and thermal performance of the envelope. Moreover, smart home sensors are used in this paper to monitor the energy use of the occupants, but the activation of doors in the building, the increase or decrease of dress code, window shading, indoor wind speed, and the energy use behavior of other equipment (electric fans, oil heaters, etc.) are not covered in this paper. It is expected that subsequent studies will have more comprehensive monitoring data, making the analysis of behavioral impacts on energy consumption and comfort more accurate. Finally, this paper proposes for the first time a dual-evaluation index of energy consumption and comfort as well as a comprehensive evaluation system, but the evaluation criteria need to be improved. It is expected that the subsequent research can determine more reasonable evaluation standards through large-sample social research.