A Quantitative Investigation of the Impact of Climate-Responsive Indoor Clothing Adaptation on Energy Use

Abstract

Clothing adjustment by building occupants is a highly effective and prevalent thermal adaptation behavior aimed at achieving thermal comfort. This paper aims to quantify the impact of climate-responsive indoor clothing adaptation on heating/cooling energy consumption. A climate-responsive indoor temperature control strategy based on rural residents’ indoor clothing adaptation was proposed and integrated into building energy simulations. Indoor clothing insulations were obtained using a predictive model from the author’s prior research. These values were used to calculate indoor setpoint temperatures in terms of the PMV model, which were then input into the building energy simulations. The simulations were conducted using “Ladybug Tools” in Grasshopper. Four simulation scenarios were proposed for winter and summer, respectively, to compare heating/cooling energy use with different indoor clothing strategies (constant and dynamic) and thermal comfort requirements (neutral and 80% acceptable). The results indicated that indoor clothing adaptation significantly reduced indoor setpoint temperatures by 5.0–6.7 °C in winter. In contrast, the impacts on summer indoor setpoint temperatures were not significant. The impacts of indoor clothing adaptation on energy use were evident in both seasons and more pronounced in winter. With a neutral thermal comfort requirement (PMV = 0), total heating and cooling energy use decreased by 35.6% and 20.2%, respectively. The influence was further enhanced with lower indoor thermal comfort requirements. With an 80% acceptable thermal comfort requirement ($\mathrm{P}\mathrm{M}\mathrm{V}=\pm 0.85$), total heating and cooling energy use decreased by 63.1% and 34.4%, respectively. The climate-responsive indoor temperature control strategy based on indoor clothing adaptation and its impact on heating/cooling energy consumption suggested a viable approach for improving building energy efficiency in China’s rural area and similar cost-sensitive and fuel-poverty contexts.

1. Introduction

The majority proportion of building energy consumption is used for conditioning the living and working environments of buildings and providing comfortable indoor thermal conditions for occupants. According to a study presented by the International Energy Agency (IEA), air conditioning electricity consumption has become one of the main driving forces for the growth of global electricity demand [1,2,3]. The setpoint temperature (i.e., the temperature at which a building system is set to activate or deactivate) influences the energy consumption of air conditioning and heating systems. The adjustment of setpoint temperatures has been widely used and studied as an energy-saving strategy [2,4,5]. Li et al. [2] developed demand response dynamic room temperature setpoints based on heat balance equations with a thermal comfort model as a constraint. They found that the dynamic temperature setpoint scenario may reduce electricity consumption by 2.8% and operational expenses by 3.73%. Sánchez-García et al. [5] explored the use of adaptive setpoint temperatures and obtained energy savings in air conditioning ranging from 52 to 58%. Another study by Wang et al. [6] indicated that a 4 °C decrease in the heating setpoint temperature resulted in a 43.3% reduction in heating energy consumption in residential buildings in China.

Clothing adaptation is one of the most efficient thermal adaptation behaviors for occupants to achieve thermal comfort [7,8,9,10]. People can efficiently adapt themselves to a wide range of indoor temperatures and achieve thermal comfort by increasing or reducing their clothing level [11,12]. According to the heat balance thermal comfort model, the dynamic adjustment of indoor clothing insulation results in dynamic setpoint temperatures, denoting higher indoor setpoint temperatures in summer and lower indoor setpoint temperatures in winter, thereby reducing heating and cooling energy consumption. Humphreys’ study of office buildings in northwest Pakistan [13] concluded that one-half of the seasonal changes in the comfort temperature could be attributed to clothing adaptation. Indraganti and Boussaa [14] proposed that the indoor neutral temperature decreased by 0.7 °C for every 0.1 clo increase in clothing insulation, according to the field surveys conducted in nine air-conditioned office buildings in Qatar during the five summer months. Another study by Liu et al. [15] in two typical naturally ventilated offices in Changsha, located in the hot summer cold winter climate zone of China, indicated that a 0.1 clo increase in clothing insulation resulted in a 0.4 °C and 0.7 °C decrease in winter and summer neutral temperatures, respectively.

The impact of clothing adaptation on building energy consumption has been demonstrated in a limited number of studies. Newsham [16] examined the effect on energy consumption of clothing adjustment in Toronto and indicated heating energy reductions from 1241.9 to 656.9 kWh and cooling energy reductions from 274.8 to 217.7 kWh. However, this research only assumed four values for clothing adaptation, which were 0.5 clo, 0.75 clo, 1.0 clo, and 1.25 clo. Xu et al. [12] investigated the energy-saving potential of clothing adaptation in an urban office building in Changsha, China, and indicated that up to 65.5% of energy consumption could be saved.

The indoor clothing behaviors of rural residents were found to have distinct characteristics and differences from those of urban residents. Studies have presented obvious discrepancies in clothing insulation between rural and urban residents in Harbin [17,18,19,20,21], Beijing [17,22,23], Yinchuan [24,25], Nanyang [26,27], and Xi’an [21,28]. Rural residents wore significantly heavier clothing in winter and lower clothing levels in summer than urban residents. Consequently, the findings on the impact on energy use of clothing adaptations in urban contexts may not apply to rural contexts in China. The clothing adaptation patterns of rural residents may have a higher impact on energy consumption, especially in heating seasons. The extent to which rural residents’ clothing adaptation affects building energy consumption is largely unclear.

This study aims to quantitatively investigate the impact on heating/cooling energy use of indoor clothing adaptation by rural residents. A climate-responsive indoor temperature control strategy was proposed based on the indoor clothing adaptation and integrated into building energy simulations. The objectives of this research were as follows:

• To propose a climate-responsive indoor temperature control strategy based on the climate-responsive indoor clothing adaptation.
• To devise a methodology for integrating the control strategy into building energy simulations.
• To quantitatively assess the impact on energy use of indoor clothing adaptation among rural residents across different seasons and varying thermal comfort requirements.

2. Methodology

2.1. Climate-Responsive Indoor Temperature Control Strategy

This study introduced a climate-responsive indoor temperature control strategy based on the clothing adaptation of residents. Its objective was to create an acceptable climate-responsive indoor thermal environment whilst reducing air conditioning energy consumption.

The control strategy calculated the indoor setpoint temperature based on the climate-responsive indoor clothing of rural residents (Figure 1). The indoor clothing level was determined in terms of the quantitative correlation between outdoor climate and indoor clothing insulation, identified in the author’s prior studies [7]. The control strategy comprises two parts: (1) indoor clothing prediction (Section 2.1.1) and (2) indoor comfort temperature calculation (Section 2.1.2).

The control strategy offers several advantages: Firstly, it widens the setpoint temperature range, lowering it in winter and raising it in summer, thereby eliminating the need for heating or cooling energy. Secondly, it reduces the intensity of occupants’ clothing adjustments when transitioning between indoor and outdoor environments.

2.1.1. Indoor Clothing Prediction

An indoor clothing predictive model was proposed in the author’s prior study, where detailed information on its development, validation, and assumptions can be found [7]. The research revealed that the 7-day running mean outdoor temperature ($T_{rm,7}$) was the primary determinant of indoor clothing levels among rural residents. An asymmetric five-parameter logistic mathematical model was found to best fit with the relationship between indoor clothing insulation and 7-day running outdoor temperature (Equation (1)). The equation applies in a $T_{rm,7}$ range between −4 °C and 5 °C for winter indoor clothing insulation prediction and in a $T_{rm,7}$ range between 24 °C and 31 °C for summer indoor clothing insulation prediction.

$$I_{cl}=0.381+{\displaystyle \frac{2.044-0.381}{{(1+e^{(0.158·T_{rm,7}-3.168)})}^{2.977}}}(R^2=0.982)$$

(1)

2.1.2. Indoor Comfort Temperature Calculations

The PMV model was proposed by Fanger in 1970 and has become the most commonly used index in thermal comfort research and practice [29]. PMV is a function of indoor air temperature ($T_a$), mean radiant temperature ($T_r$), relative humidity (RH), air velocity ($V_a$), activity level (Met), and clothing insulation ($I_{cl}$), as shown in Equation (2). The thermal environment is considered thermally neutral if the PMV equals zero and is considered comfortable for 80% of occupants if the PMV is between −0.85 and 0.85.

$$PMV=f(T_a,T_r,RH,V_a,Met,I_{cl})$$

(2)

A comfortable thermal environment should have no noticeable cold or hot radiant sources that may cause discomfort and should have low air velocity and medium relative humidity [12]. It is commonly assumed that the mean radiant temperature equals the indoor air temperature in a comfortable thermal environment [16]. Based on this assumption, the comfort temperature ($T_{comf}$) can be expressed as a function of relative humidity (RH), air velocity ($V_a$), activity level (Met), clothing insulation ($I_{cl}$), and PMV, shown as Equation (3).

$$T_{comf}=T_a=T_r=f(RH,V_a,Met,I_{cl},PMV)$$

(3)

By specifying values for relative humidity, air velocity, and activity level within rational ranges, Equation (3) can be converted into Equation (4). In this equation, comfort temperature becomes a function of indoor clothing insulation and PMV. The comfort temperature can be calculated by letting PMV equal specific values (0 and ±0.85), thus becoming a function of indoor clothing insulation.

$$T_{comf}=f(I_{cl},PMV)$$

(4)

The ISO 7730 [30] PMV model is applicable within an indoor air temperature range of 10.0 $\mathrm{℃}$ and 30.0 $\mathrm{℃}$. Consequently, Equation (4) produces comfort temperatures between 10.0 $\mathrm{℃}$ and 30.0 $\mathrm{℃}$ with sufficient levels of clothing insulation. If the inputs yield comfort temperatures beyond this range, the values at the extremes are used. For example, if the inputs result in comfort temperatures lower than 10.0 $\mathrm{℃}$, the indoor comfort temperatures are set at 10.0 $\mathrm{℃}$.

2.2. Building Energy Simulations

This research simulated a prototype house rather than a specific rural house. The prototype house was assumed to be well insulated, airtight, and equipped with an automated HVAC system implementing the aforementioned climate-responsive indoor temperature control strategy. This was because a well-insulated and airtight prototype was more reasonable as these features align better with an automated HVAC system. This could represent the prototype for future rural houses. Simulations were conducted using Grasshopper (GH). Two Python-based GH components were developed to integrate the climate-responsive indoor temperature control strategy into the building energy simulation. The simulations encompassed four scenarios for each season, generating daily heating and cooling energy use over the simulation periods. Comparative analysis of energy use in different scenarios demonstrated the impact on energy use across various circumstances. Figure 2 presents the flowchart for this research.

2.2.1. Simulation Software

This study utilized the building simulation tools “Ladybug Tools” for building energy simulation within Grasshopper (GH). Ladybug tools, built upon the validated energy simulation engine “EnergyPlus V.8.2.0”, provided building energy simulation capabilities. Ladybug tools have been widely used in previous studies for building energy simulation, demonstrating exceptional predictive accuracy [31,32,33,34].

Two Python-based grasshopper components were developed to implement the climate-responsive control strategy and integrate it into building energy simulations (Figure 3). The two components were “Indoor Clothing Prediction (ICP)” and “Find Indoor Comfort Temperature (FICT)”.

The author programmed Equation (1) into the ICP component, which utilized outdoor temperature as input to predict the indoor clothing insulation of rural residents. Equation (4) was programmed into the FICT component. The FICT component took season, clothing insulation, and PMV as inputs to compute indoor comfort temperature. The Python package “pythermalcomfort”, developed by the Lawrence Berkeley National Laboratory for thermal comfort research, was imported for the development of FICT [35].

2.2.2. Modeling the Prototype

The simulations assumed a well-insulated and airtight building equipped with an automated HVAC system utilizing the aforementioned climate-responsive indoor temperature control strategy.

The prototype house was a south-oriented, single-storey, detached house with a courtyard at the front. The floor plan of the house is illustrated in Figure 4. The dimensions of the house model were as follows: width 13.2 m, depth 6.0 m, and height 3.0 m, resulting in a total area of 79.2 ${\mathrm{m}}^2$. The central room served as the living room. One of the two terminal rooms functioned as the primary bedroom and the other was used as a multifunctional room.

The thermal performance of building envelopes, presented in

Table 1. U-values of building envelopes.

Envelope	U-Value	U-Value Limit
Exterior wall	0.52	0.65
Interior wall	1.95	\
Roof	0.30	0.50
Ground floor	0.34	\
Unit: W/(m2·K)

, referred to the U-value limits and building envelope prototypes suggested by the Chinese standard GB/T 50,824 2013 [36]. The model’s airtightness was assigned the value specified in the Passivhaus criteria (0.6 air changes per hour at 50 Pascals pressure) [37].

As there is a lack of information on rural house airtightness, the power consumption of home appliances, and residents’ schedules, several assumptions were made in the simulation. Heat loads from cooking and home appliances were assumed to be negligible to prevent potential interference with heating/cooling energy simulation results caused by thermal emissions from these appliances. The house was assumed to accommodate two occupants and to be occupied 24 h a day. The HVAC system was set to the “Ideal Air Loads” provided by the software. The “Ideal Air Loads” object was modeled as an ideal VAV terminal unit with variable temperature and humidity. The supply airflow rate was varied between zero and the maximum in order to satisfy the zone heating or cooling loads, zone humidity controls, outdoor air requirements, and other constraints, if specified. More specific explanations of the system can be found in the “Engineering reference” of EnergyPlus V.8.2.0 [38].

2.2.3. Simulation Periods

The proposed indoor clothing adaptation model was based on field survey data from Kaifeng City, which is located in the cold climate zone of China. Monthly outdoor air temperatures in Kaifeng are presented in Figure 5. The typical meteorological year’s weather data of Kaifeng City were employed for the simulation [39].

The summer simulation ran from 17 June to 26 August. The simulation period was defined in terms of the 7-day running mean outdoor temperature. It comprised consecutive days on which the $T_{rm,7}$ fell within the summer application range of the indoor clothing predictive model (Figure 6). The period represented 71 days.

The winter simulation took place from 6th December to 14th February of the following year. It was determined based on a comprehensive consideration of the daily $T_{rm,7}$ and the number days in the summer simulation period. Firstly, a period of consecutive days was identified during which the $T_{rm,7}$ remained within the winter application range of the indoor clothing predictive model (Figure 7). Then, the period was restricted to 71 days to align with the length of the summer simulation period.

Diurnal time-step simulations were performed because Equation (1) predicted daily mean indoor clothing insulations.

2.2.4. Simulation Scenarios

For each season, four scenarios were proposed to evaluate the impact of dynamic indoor clothing compared to constant indoor clothing under different thermal comfort conditions (

Table 2. Scenario settings of winter and summer simulations.

	Winter Simulation Scenarios				Summer Simulation Scenarios
	WC0	WD0	WC1	WD1	SC0	SD0	SC1	SD1
Relative humidity	60%	60%	60%	60%	60%	60%	60%	60%
Air velocity	0.02 m/s	0.02 m/s	0.02 m/s	0.02 m/s	0.24 m/s	0.24 m/s	0.24 m/s	0.24 m/s
Metabolic rate	1.2 met	1.2 met	1.2 met	1.2 met	1.0 met	1.0 met	1.0 met	1.0 met
PMV	0	0	−0.85	−0.85	0	0	+0.85	+0.85
Clothing insulation	1.0 clo	Dynamic	1.0 clo	Dynamic	0.46 clo	Dynamic	0.46 clo	Dynamic

).

(1)

Summer simulation scenarios

In summer, scenarios SC0 and SC1 adopted a constant indoor clothing insulation of 0.46 clo, the minimum recommended value in ASHRAE Standard 55 [40], with PMVs of 0 and +0.85, respectively. Scenarios SD0 and SD1 utilized dynamic indoor clothing insulation predicted by Equation (1), with PMVs of 0 and +0.85, respectively. A PMV of 0 indicated a neutral thermal condition, while a PMV of −0.85 corresponded to an 80% acceptability thermal comfort condition.

All four summer scenarios set the relative humidity as 60%, the air velocity as 0.24 m/s, and the metabolic rate as 1.0 met. The assumptions originated from field survey results in the author’s prior research. Nearly 80% of the respondents remained in the “sitting, relaxed” activity during the summer field survey, corresponding to the 1.0 met. The mean value of measured air velocity during summer field surveys was 0.24 m/s. Mean relative humidity during prior field surveys was 63% for summer and 58% for winter. The distribution was also similar. Consequently, a relative humidity of 60% was adopted for both winter and summer simulations.

(2)

Winter simulation scenarios

In winter, scenarios WC0 and WC1 adopted a constant indoor clothing insulation of 1.0 clo, the maximum recommended value in ASHRAE Standard 55 [40], with PMVs of 0 and −0.85, respectively. Scenarios WD0 and WD1 adopted dynamic winter indoor clothing insulation predicted by Equation (1), with PMVs of 0 and −0.85, respectively.

The relative humidity, air velocity, and metabolic rate were assigned specific values of 60%, 0.02 m/s, and 1.2 met, respectively. The winter metabolic rate was found to be slightly higher as rural residents intensified their activities to resist coldness in winter. Referring to “Metabolic rate for typical activities” in ASHRAE Standard 55, the winter metabolic rate was set at 1.2 met. Air velocity for the winter simulations also referred to the field survey results in the author’s prior studies [41].

3. Results

3.1. Predicted Daily Indoor Clothing Insulation

Figure 8 illustrates the daily indoor clothing insulation throughout the winter and summer simulations. The dynamic indoor clothing insulation, predicted by Equation (1), is depicted by the dashed blue line, while the constant indoor clothing insulation is represented by the red line.

In winter, as presented in

Table 3. Description of daily clothing insulation in two scenarios in winter and summer simulations.

	Winter					Summer
	Min	Max	Mean	S.D.	CV (RMSE)	Min	Max	Mean	S.D.	CV (RMSE)
Constant clothing	1.00	1.00	1.00	0.00	81.6%	0.46	0.46	0.46	0.00	11.5%
Dynamic clothing	1.72	1.93	1.82	0.05		0.39	0.45	0.41	0.01
Unit: clo.

, the dynamic indoor clothing insulations ranged from 1.72 clo to 1.93 clo, with a mean value of 1.82 clo. The CV(RMSE) of 81.6% indicated a significant discrepancy from the winter constant indoor clothing insulation. The climate-responsive indoor clothing adaptation resulted in much higher indoor clothing insulation in winter.

For summer (Table 3), the dynamic indoor clothing insulations ranged from 0.39 clo to 0.45 clo, with a mean value of 0.41 clo. The CV(RMSE) was only 11.5%. Although the climate-responsive indoor clothing adaptation resulted in less dynamic indoor clothing insulation in summer, the discrepancies were much less significant compared to winter.

3.2. Daily Indoor Comfort Temperature

(1)

Winter simulation scenarios

Figure 9a illustrates the daily indoor comfort temperature for the four winter simulation scenarios. In scenario WC0 (1.0 clo, PMV = 0) and WC1 (1.0 clo, PMV = −0.85), constant indoor comfort temperatures of 21.2 $\mathrm{℃}$ and 17.4 $\mathrm{℃}$ were observed, respectively.

Scenario WD0 (dynamic clothing, PMV = 0) showed dynamic indoor comfort temperatures ranging between 15.6 $\mathrm{℃}$ and 16.8 $\mathrm{℃}$, with an average of 16.2 $\mathrm{℃}$. The implementation of the climate-responsive control strategy resulted in an average reduction of 5.0$\mathrm{℃}$ in indoor comfort temperatures under the neutral thermal comfort requirement.

Scenario WD1 (dynamic clothing, PMV = −0.85) showed dynamic indoor comfort temperatures between 10.0 $\mathrm{℃}$ and 11.5 $\mathrm{℃}$, with an average value of 10.7 $\mathrm{℃}$. Compared to scenario WC1, indoor comfort temperatures were, on average, reduced by 6.7 $\mathrm{℃}$ under the 80% acceptability thermal comfort requirement. The indoor comfort temperature decreased significantly as a result of the climate-responsive control strategy under both thermal comfort conditions in winter.

(2)

Summer simulation scenarios

Similarly, in the summer simulation (Figure 9b) scenarios SC0 (0.46 clo, PMV = 0) and SC1 (0.46 clo, PMV = +0.85), constant indoor comfort temperatures of 26.7 $\mathrm{℃}$ and 28.8 $\mathrm{℃}$ were observed, respectively. Scenario SD0 (dynamic clothing, PMV = 0) showed dynamic indoor comfort temperatures between 26.8 $\mathrm{℃}$ and 27.1 $\mathrm{℃}$, with an average value of 27.0 $\mathrm{℃}$. Scenario SD1 (dynamic clothing, PMV = +0.85) showed dynamic indoor comfort temperatures ranging from 28.8 $\mathrm{℃}$ to 29.1 $\mathrm{℃}$, with an average of 29.0 $\mathrm{℃}$. The mean indoor comfort temperature slightly increased by 0.3 $\mathrm{℃}$ and 0.2 $\mathrm{℃}$ due to the climate-responsive control strategy under neutral and 80% acceptability thermal comfort conditions, respectively. Notably, the alterations in indoor comfort temperature due to dynamic indoor clothing adaptation among rural residents were less pronounced during summer compared to winter.

3.3. Heating Loads

Table 4. Statistics of daily heating loads (in kWh) over winter simulations.

PMV	Scenario	Min	Max	Mean	Total	Energy Use Reduction
0	WC0	25.5	41.1	34.4	2442.1	868.6
	WD0	14.2	28.0	22.5	1573.5
−0.85	WC1	16.6	31.7	25.2	1778.3	1129.0
	WD1	2.0	14.2	9.3	659.2

presents daily heating load statistics, indicating reductions in peak and mean heating loads, evidenced by notable decreases in both maximum and mean values attributed to the climate-responsive control strategy.

Figure 10 illustrates the total heating loads across the winter simulations. The total heating loads decreased from 2442.1 kWh to 1573.5 kWh, representing a reduction of 35.6%, under the neutral condition and decreased from 17,789.3 kWh to 659.2 kWh, resulting in a 63.1% reduction, under the 80% acceptability condition. This suggests that indoor clothing adaptation significantly reduces heating energy consumption. Additionally, the higher rate of reduction observed under the 80% acceptability thermal comfort condition highlights the enhanced impacts on heating energy use of indoor clothing adaptation in environments with lower thermal comfort requirements.

3.4. Cooling Loads

The statistics in

Table 5. Statistic descriptions of daily cooling loads (in kWh) over summer simulations.

PMV	Scenario	Min	Max	Mean	Total	Energy Use Reduction
0	SC0	0.0	47.8	12.1	858.4	173.8
	SD0	0.0	44.4	9.6	684.6
+0.85	SC1	0.0	19.6	1.6	108.2	37.2
	SD1	0.0	16.3	1.0	71.0

show 7% and 16.8% reductions in maximum values under the neutral and 80% acceptability thermal comfort conditions. This demonstrates the limited effects of the climate-responsive control strategy on peak load in summer compared to winter.

The climate-responsive control strategy also extended the number of no-cooling days over the summer simulations. The number of no-cooling days increased from 9 to 12 under the neutral thermal comfort condition and rose from 28 to 33 under the 80% acceptability thermal comfort condition.

Figure 11 illustrates the total cooling loads throughout the summer simulations. The reductions observed were minimal, with only 173.8 kWh and 37.2 kWh under the neutral and 80% acceptability thermal comfort conditions, respectively. Despite these modest decreases, the decrease rates reached 20.2% and 34.4% under the respective thermal conditions. This could be attributed to the relatively low total cooling loads. The climate-responsive control strategy could reduce the cooling load, although its effectiveness was not as pronounced as in winter. Additionally, the higher reduction rate under the 80% acceptability thermal comfort condition also demonstrated its more noticeable impacts on energy use under lower thermal comfort requirements.

4. Discussion

4.1. Impact of Outdoor Temperature on Energy Use Reduction

Figure 12 plots daily heating/cooling loads with daily outdoor temperatures. The regression lines of the daily heating/cooling loads with daily outdoor temperature were formulated. Equation WC0 had a steeper slope than WD0, indicating a faster increase in daily heating loads of WC0 than WD0 and expanding gaps between WC0 and WD0 with decreasing outdoor temperature. The 80% acceptability thermal comfort condition also presented a steeper gradient in WC1 than in WD1 and an increasing gap between the two equations with decreasing outdoor temperatures. It could be inferred that the energy use reductions were amplified in colder outdoor climates in winter. Similarly, it can be deduced from Figure 12b that the energy use reductions were amplified in the hotter outdoor climate of summer.

4.2. Seasonal Differences in Energy Reduction

As discussed in Section 3, with the influences of clothing adjustment, the decreases in indoor comfort temperatures in winter were much higher than the increases in indoor comfort temperature in summer. Moreover, the decreases in the maximum daily heating loads in winter exceeded those in summer. Most importantly, the total energy use reduction rate was more significant in winter than in summer. It can be deduced that the indoor clothing adaptation of rural residents presented a more significant impact on energy use in winter than in summer.

This phenomenon can partly be attributed to the following two reasons. Firstly, rural residents in China tolerate heavier clothing in winter. This was demonstrated by the upper limit of Equation (1) and the mean daily indoor clothing insulation presented in Section 3.1. Both values significantly exceeded the 1.0 clo limitation suggested in ASHRAE Standard 55. Higher indoor clothing insulation levels denote lower indoor comfort temperatures and, correspondingly, less heating energy use. In contrast, the summer clothing of rural residents in China approximates the suggested lower limitation in the standard in terms of the function’s lower limit and mean daily indoor clothing insulation over the summer simulation. Secondly, the indoor clothing adaptation of rural residents was more active in winter than in summer. Figure 13 presents the indoor clothing adaptation function and predicted indoor clothing insulation distributions over the simulation periods. The adaptation function in the red region has a steeper slope than in the blue region, implying swift-rising indoor clothing insulation with dropping outdoor temperatures in winter.

4.3. Rethinking Setpoint Temperatures

Indoor temperature significantly affects occupants’ health, especially in extreme conditions. According to the “Fuel for Life” report, low indoor temperatures can lead to cold and damp environments, increasing the risk of respiratory infections, hypothermia, and cardiovascular issues. High indoor temperatures can cause heat exhaustion and heat stroke and exacerbate chronic conditions such as heart disease and diabetes. Prolonged exposure to extreme temperatures induces thermal stress, posing potential health risks, particularly for vulnerable populations such as the elderly, children, and individuals with pre-existing health conditions [42].

In this research, a very low setpoint temperature range between 10.0 and 11.5 $\mathrm{℃}$ was observed during the winter simulation under the 80% acceptable thermal comfort requirement. These temperatures may pose significant health threats to occupants. Similarly, setpoint temperature ranges between 28.8 $\mathrm{℃}$ and 29.1 $\mathrm{℃}$ were obtained during summer simulations. High setpoint temperatures may also cause health issues among residents. The achievement of energy efficiency in residential buildings, even for those experiencing fuel poverty, should not compromise the health of occupants.

In contrast, under the neutral thermal comfort requirement (PMV = 0), these setpoint temperature ranges were between 15.6 $\mathrm{℃}$ and 16.8 $\mathrm{℃}$ for winter and 26.8 $\mathrm{℃}$ and 27.1 $\mathrm{℃}$ for summer simulations. These temperature ranges are more rational settings for practice.

4.4. Research Limitations

This research emphasized the impact of indoor clothing adaptation on building energy use. It is possible that the simulations in the research may not have accurately mirrored the energy-reducing benefits of indoor clothing adaptations in real-world settings. Simulations were carried out with a well-insulated and airtight modern rural house prototype that did not represent any actual rural houses. Chinese rural houses exhibit diverse typologies, with varying parameters affecting energy use. Additionally, the indoor clothing predictive model was based on field survey results from current Chinese rural houses lacking energy-efficient strategies and smart HVAC control systems. Clothing adaptation behaviors may change when people acclimate to modern energy-efficient buildings. Furthermore, as income among rural residents rises, their clothing preferences at home may also change. Further observations and research on the indoor clothing behaviors of rural residents in rapidly developing rural regions in China are necessary for understanding the impact of clothing adaptation on energy use in rural houses.

5. Conclusions

This study quantitatively evaluated the impact of indoor clothing adaptation on building energy use among rural residents in China’s cold climate zone. A climate-responsive indoor temperature control strategy was proposed based on the rural residents’ climate-responsive indoor clothing adaptation and integrated into building energy simulations. The conclusions drawn from this investigation are as follows:

• The influence of indoor clothing adaptations on indoor comfort temperature varied between winter and summer. During winter, indoor comfort temperature was on average reduced by 5.0 $\mathrm{℃}$ ($\mathrm{P}\mathrm{M}\mathrm{V}=0$) and 6.7 $\mathrm{℃}$ ($\mathrm{P}\mathrm{M}\mathrm{V}=-0.85$). Conversely, in summer, mean indoor comfort temperatures increased marginally by only 0.3 $\mathrm{℃}$ ($\mathrm{P}\mathrm{M}\mathrm{V}=0$) and 0.2 $\mathrm{℃}$ ($\mathrm{P}\mathrm{M}\mathrm{V}=0.85$).
• The impact of indoor clothing adaptation on energy use was significant in both seasons. Peak loads were significantly reduced during the winter simulations, while the number of no-cooling days increased during the summer simulations. The total heating and cooling energy consumption could be reduced by 35.6% and 20.2%, respectively, under optimal thermal comfort conditions ($\mathrm{P}\mathrm{M}\mathrm{V}=0$). The energy use reduction was more significant with lower levels of thermal comfort requirements. The proportion of the reduction in total heating and cooling energy use increased to 63.1% and 34.4%, respectively, under the 80% acceptable thermal comfort conditions ($\mathrm{P}\mathrm{M}\mathrm{V}=\pm 0.85$).
• The climate-responsive indoor temperature control strategy based on indoor clothing adaptation and its significant impact on energy consumption suggested a viable approach for improving building energy efficiency in rural China and similar cost-sensitive contexts where economic factors, such as income and fuel costs, take precedence in decision-making regarding air conditioning. This approach leverages rural residents’ climate-responsive indoor clothing adaptation capability, providing acceptable indoor thermal environments while achieving significant energy reductions.

This research highlights the promising role of indoor clothing adaptation coupled with climate-responsive control strategies in mitigating energy consumption in rural contexts. The consideration of occupants’ indoor clothing adaptation in building energy modeling may be one mitigating factor in closing the performance gaps between simulated and actual energy uses, thereby improving the accuracy of simulation results. The findings of this research serve as a valuable reference for architects and building system engineers seeking to optimize building and system designs. Moreover, this study underscores the importance of occupant behavior in building energy consumption. Energy use behavior and its impacts on building energy consumption deserve further investigation in future research for the benefits of building energy modeling, building design, and policy making.