Study on Thermal Environment and Energy Consumption of Typical Residential Courtyards in Beijing’s Old City

Abstract

Old city courtyards are crucial elements of Beijing’s ancient capital. However, existing ones face heating problems. This study focuses on renovated and original-style courtyards. By employing ENVI-met and DeST software, we comprehensively analyzed the courtyard’s thermal environment, ventilation, indoor conditions, and energy consumption. Findings reveal that both types have thermal discomfort. Original courtyards are colder in winter and hotter in summer due to wind and radiation. They possess better ventilation but a higher winter heating load. Both require winter heating, with the original ones having a larger unit area load because of envelope heat loss and ventilation differences. Their direct electric heating consumptions, 187.6 kWh/m² and 229.6 kWh/m², respectively, surpass ordinary residences. This study defines issues for future green and low-carbon courtyard work.

1. Introduction

Courtyard-style residential blocks, centered around traditional Siheyuan (quadrangle courtyards), are a defining feature of Beijing’s historic urban landscape and serve as iconic cultural landmarks. Primarily designed for residential use, these courtyard clusters have faced significant challenges due to population growth and urban development, gradually eroding their traditional historic character. Furthermore, a substantial portion of these areas consists of multi-family “big courtyard” dwellings, which exacerbate conflicts between outdated living conditions and modern lifestyle needs, creating an urgent need for improvement [1,2]. In recent years, Beijing has launched comprehensive initiatives to preserve and renew these historic districts, focusing on protecting traditional urban aesthetics, population relocation to reduce density, and improving residential environments. These efforts have achieved initial success and are expected to continue advancing in the coming years [3].

With the introduction of the national dual-carbon development strategy, green and low-carbon development has become a key focus in the renewal of Beijing’s old city courtyards [4]. From the perspective of residential thermal environments and energy supply, the buildings in these courtyards are predominantly single-story structures with significant shape coefficients. Issues such as widespread neglect and lack of maintenance of building envelopes are common, resulting in poor thermal comfort. Additionally, the absence of centralized heating and other municipal infrastructure leads residents to rely on decentralized, individual heating systems. Electricity consumption for heating represents a significant portion of household energy expenditures, with many residents relying on government subsidies to offset the financial burden [5]. Therefore, studying how to improve residential thermal environments and achieve energy efficiency and carbon reduction within the constraints of historical preservation is of significant importance. Such research not only enhances the quality of life in the old city but also alleviates financial burdens on residents and public finances, promoting low-carbon development. A critical step in this process is to evaluate the existing thermal environments and energy consumption patterns in old city courtyards, providing a comprehensive understanding of thermal comfort and energy usage across different courtyard typologies.

Although traditional residential buildings’ thermal performance and energy efficiency, such as Siheyuan, have been studied extensively in China Siheyuan [6], research specifically addressing the courtyards of Beijing’s old city remains scarce due to various limitations. Current studies primarily rely on computer simulations, supplemented by limited on-site investigations and tests. Considering the interrelationships among physical environmental factors such as wind, light, and thermal conditions, this paper briefly reviews representative literature in this field. Lin et al. [7] were among the first to apply CFD (PHOENICS 3.4) simulation technology to analyze Beijing’s traditional Siheyuan courtyard wind environment. When the outdoor space of the courtyard doubles, the wind speed at a height of 2.5 m increases from 1 m/s to 5 m/s. Shi et al. [8] simulated the wind and thermal environments of traditional Siheyuan. The research has concluded that the courtyard form with the aspect ratio of the inner courtyard close to 1 and the height ratio of the principal house to the opposite house ranging from 1.2 to 1.4 can simultaneously meet the ventilation requirements in summer and windproofing warmth preservation in winter. Zhao et al. [9] used CFD (PHOENICS 4.4) and DeST (3.0) software to analyze the Siheyuan traditional and renovated Siheyuan courtyard wind environment and indoor temperatures. The green construction design strategy was put forward, which suggests that the optimal width–length–height ratio of the internal courtyard is 1:1.4:1. If the usable area needs to be increased and a two-story structure is required for the quadrangle courtyard in some parts, it is recommended to add an additional opposite house. Zeng et al. [10] utilized CFD (PHOENICS 4.4) technology to study the impact of veranda and loft designs on courtyard ventilation in traditional Siheyuan. The results show that the average wind speed in the courtyard is 0.19 m/s when there is a front porch and 0.9 m/s when there is no front porch. Wu et al. [11] used Computational Fluid Dynamics (PHOENICS 4.4) to conduct a simulation analysis on the outdoor wind environment of brick–wood folk houses during the heating period. The wind speed at the pedestrian height was within the range of 0.31–3.12 m/s. However, the wind pressure difference on the horizontal plane at 5.3 m was 5.1 Pa, and that on the horizontal plane at 6.5 m reached 6.88 Pa. It can be seen that the cold air infiltration was serious. Therefore, it was proposed that the airtightness of the enclosure structure on the second floor should be enhanced during the renovation. Wang et al. [12] conducted a comparative CFD (PHOENICS 4.4) analysis on the effects of different gatehouse layouts on the wind environment of Siheyuan. The wind speeds of various schemes are within the range of 0–2.9 m/s. Through comparison, setting doorways in the windward position during the season significantly enhances the wind comfort of quadrangle dwelling courtyards. Li et al. [13] investigated and redesigned the lighting environment of a courtyard in the old city, using Radiance software to simulate and compare natural lighting conditions before and after renovation. It was concluded that after the additional buildings in the surrounding area were demolished, the average illuminance value of the courtyard increased by more than one. Yao et al. [14] measured and simulated the indoor lighting effects of using paper windows in Siheyuan dwellings. The use of window paper can improve the uniformity of indoor illuminance. It can increase the illuminance uniformity of the principal house by 97%, that of the western-side house by 109%, and that of the eastern-side house by 146%. He et al. [15] employed on-site testing and ENVI-met (4.4.2) software simulations to study thermal comfort conditions and improvement strategies for historical street districts in Beijing’s old city. It was concluded that the Universal Thermal Climate Index (UTCI) thermal neutral temperature is 20.6 °C, and increasing greenery is the optimal strategy. Su et al. [16] focused on the Dashilar area of Beijing, using testing and ENVI-met (5.0.1) simulations to investigate outdoor thermal environments and propose natural cooling strategies to mitigate urban heat island effects. It was concluded that by increasing green infrastructure and street trees, 1 °C could reduce the average temperature, and the proportion of high-temperature areas could be decreased by 13.87%. Xu et al. [17] used Design Builder (V4) to conduct a simulation analysis on the annual natural room temperature of the quadrangle courtyards in Yixing City. It was found that when a long corridor was added outside the quadrangle courtyards, the average natural room temperature in the rooms decreased by 1.8 °C in summer and increased by 1.1 °C in winter, with the cooling load reduced by 19.6% and the heating load reduced by 22.3%. Piao et al. [18] used the CFD (PHOENICS 3.5) software to conduct simulations. They explored the impact of the position of the main gate of a three-courtyard quadrangle courtyard on the outdoor thermal environment in Shenyang City under the dominant wind speeds and directions in summer and winter. It was concluded that when the position of the main gate was 20° south by west in winter and 10° south by east in summer, the average temperatures were −8.8 °C and 22.5 °C, respectively, and the thermal environment was at its best. Cai et al. [19] used Ecotect software to calculate and analyze traditional Siheyuan wind, light, and thermal environments under different configurations and structural forms. It was concluded that the light and heat conditions of the side rooms in the differentiated model with a courtyard scale of 1:1 were the best. Li et al. [20] applied Fluent (Airpak 3.0), Ecotect (5.6), and DeST (3.0) software to simulate the courtyard wind field, sunlight exposure, indoor lighting, and winter heating loads of traditional Siheyuan, evaluating the effectiveness of various energy-saving measures. It was concluded that the model with rooms with a depth of 7 m and a south-facing window-to-wall ratio of 0.45 had the highest daylighting comfort and energy consumption of the exterior walls. The energy consumption can be reduced by increasing the thermal insulation of the exterior walls. Yang et al. [21] conducted on-site measurements and simulation analyses of the wind, light, and thermal environments of a traditional Siheyuan in Beijing, proposing climate-adaptive design strategies related to site planning, courtyard dimensions, functional layouts, roof slopes, and building colors. The results show that under the optimal strategy, the average wind speed is 1.8 m/s, the average value of ground solar radiation increases by 67.1 Wh, and the energy consumption is reduced. Shang et al. [22] simulated energy consumption trends during the morphological evolution of Beijing’s traditional Siheyuan, analyzing the impacts of layout and dimensions, building shape and coefficients, and window sizes on heating energy demands, compared with the earliest form, the energy consumption has been reduced by 79%. Zhang et al. [23] proposed a simplified method for calculating the winter heating and domestic hot water energy consumption of houses in Beijing’s old city, analyzing the energy-saving effects of thermal insulation upgrades for buildings from different eras. After the renovation, the energy-saving rates of the buildings completed between 1978 and 1985, between 1986 and 1995, and between 1996 and 2005 can reach 65.53%, 55.04%, and 32.79%, respectively.

In summary, current research on Beijing’s old city courtyards focuses more on traditional Siheyuan and their design optimization, with little attention paid to the more prevalent “big courtyard” dwellings, which lack representativeness of the courtyard forms in Beijing’s old city. At the same time, existing evaluations of courtyard thermal environments and energy consumption often focus on isolated aspects, such as the courtyard thermal environment, indoor thermal environment, or heating and cooling energy consumption, lacking comprehensiveness and relevance. In short, existing research fails to fully support a comprehensive understanding and analysis of the thermal environment and energy consumption levels of Beijing’s old city courtyards.

To address existing research’s shortcomings and better understand the thermal environments and energy consumption of various residential courtyards in the old city, this paper examines two typical courtyard forms in Beijing’s old city: the traditional Siheyuan and the multi-household “big courtyard”. Using simulation methods, it calculates various performance indicators such as courtyard thermal environment, indoor thermal environment, indoor and outdoor natural ventilation, heating and cooling energy consumption, and associated costs. By comparing the two typical courtyards, this paper quantitatively evaluates and analyzes the thermal environment and energy consumption levels of residential courtyards in Beijing’s old city, identifies existing problems, and provides a reference for the green and low-carbon renewal of courtyards in the old city.

Since the significant courtyard forms in Beijing’s old city are large in scale and complex in terms of housing conditions, population composition, and property ownership (2), it is difficult to conduct direct on-site investigations and testing, leading to a general lack of foundational test data related to their thermal environment and energy consumption. Therefore, this study mainly adopts simulation methods. For comparison, a real-world big courtyard (referred to in this paper as the “extended courtyard”) and its historical original layout (referred to in this paper as the “original courtyard”) were selected as two typical courtyard forms for simulation analysis. The research process is shown in Figure 1. First, a multi-household big courtyard in Beijing’s Dongcheng District, which has undergone resident relocation, was selected. Field investigations and relevant regulations determined the parameters required for courtyard modeling. Next, ENVI-met and DeST software were used to construct models for both the traditional Siheyuan and the hefty courtyard. Simulations were conducted to calculate the performance indicators for the two courtyards. Finally, the results for the two courtyard forms’ thermal environments and heating and cooling energy consumption were summarized and compared. These results were validated using partial field investigation data. The main factors affecting the thermal environment and heating and cooling energy consumption of the two courtyard forms were analyzed, and noteworthy issues and key points were identified.

This article will be elaborated according to the following structure. In Section 2, field measurements were carried out on some typical courtyards’ modeling parameters, temperature, and electricity consumption. In Section 3, performance evaluation indicators were selected to provide comparative parameters for subsequent results. In Section 4, the simulation methods and models were described. In Section 5, the similarities and differences of different indicators of the two types of courtyards were presented and discussed based on the results. In Section 6, comprehensive research conclusions were drawn, and targeted suggestions were put forward.

2. Case Study Investigation

In the old city districts of Beijing, some courtyards have been well-preserved and passed down, retaining or restoring their historical appearance from the Ming and Qing dynasties. These are referred to in this paper as “original courtyards”. However, due to limited living space, the majority of courtyards have undergone unauthorized expansions by residents. These expansions often encroach on the original courtyard space, transforming the properties into “big courtyard” dwellings. In this paper, the resident-constructed additions are referred to as “extension spaces”, such courtyards are categorized as “extended courtyards”. The existing courtyard typologies in Beijing’s old city are predominantly these two forms, with extended courtyards being the most prevalent.

2.1. Survey of Modeling Parameters for Typical Courtyards

From 2022 to 2023, the author surveyed 54 courtyards in the Sanyanjing area of Beijing. These courtyards are predominantly state-owned housing compounds, with property ownership belonging to work units or collectives, and are characterized by multi-household shared living arrangements. Under the guidance of relevant municipal policies, more than half of these courtyards have undergone resident relocation and are planned for renovation while retaining their residential function.

This study selects two representative courtyards from the Sanyanjing area—Nos. 70 and 72 Sanyanjing Hutong—as case examples. Originally constructed as a complete Siheyuan in 1956, this courtyard underwent incremental spatial additions beginning in 1972, which resulted in its subdivision into two separate courtyards. By the end of 2021, the site underwent relocation of residents, and all unauthorized extensions were demolished, restoring the original layout. This courtyard, in its two states—before and after the removal of the extensions—serves as a proxy for the two typical courtyard forms found in Beijing’s old city: the “original courtyard” and the “extended courtyard”. The floor plans of these two courtyard states are illustrated in Figure 2. The items marked in red are length and width, with the unit being a meter. In the diagram, dashed lines indicate the partitions created by the extensions. Light gray areas represent the state-owned housing, with a total floor area of 628.6 m², while dark gray areas denote the extensions, with a total floor area of 299.0 m². The total courtyard area is 927.6 m², with the extensions accounting for 32.2% of the total area. Numbers ①–⑧ correspond to the façade photographs taken during the field survey after removing the extensions, and the detailed façade images are provided in

Table 1. Facade Photos Corresponding to Specific Locations.

Point Location	Site Photos	Point Location	Site Photos
①		②
③		④
⑤		⑥
⑧		⑧

. The items marked in red are length and height, with the unit being a meter.

During the field survey, we collected critical geometric parameters and material properties required for building energy simulations alongside data on room functionalities, occupant behavior, and appliance usage profiles. Geometric dimensions of the courtyard, buildings, walls, doors, and windows were obtained using laser distance meters and 360-degree cameras. The “original courtyard” features overhanging eaves on both the front (entrance side) and rear sides. The principal rooms have a depth of 12 m with an eave height of 3.5 m, while the reverse oriented-room, back enclosed-rooms, and east and west wing rooms each have depths of 5.5 m and eave heights of 3.2 m. The side rooms have depths ranging from 4.2 to 4.5 m with eave heights of 2.8 m. The specific dimensions are shown in the aforementioned chart. The primary windows in all rooms face inward toward the courtyard, while walls facing outward have only a few small windows with areas ranging from 0.1 m² to 6.0 m². Wall materials primarily consist of gray and red bricks, with no external wall insulation. Roofs are typically covered with small gray tiles, and doors and windows are made of single-pane glass with plastic frames.

Prior to relocation, this large courtyard housed 32 households across 41 rooms. Due to limited living space, the state-owned rooms served as multifunctional residential units combining functions such as bedrooms, living rooms, and kitchens. The extended spaces, on the other hand, were typically used as living rooms, kitchens, or storage areas. Residents typically go out for work or study during the day and return home at night. Under the climatic conditions of Beijing, people’s heating and cooling habits are as follows: for winter heating, the majority of residents used storage-type electric heaters, which were generally operated continuously throughout the day. In summer, split air-conditioning units were employed for cooling, usually running at night and turned off during the day.

2.2. Measurement of Indoor Temperature and Electricity Consumption in Typical Rooms

Due to the relocation of most residents in Sanyanjing Hutong, the majority of rooms are unoccupied, making it impossible to measure and record indoor thermal environments and electricity consumption. To validate the subsequent simulation results, 11 rooms with occupants were selected from residential areas in Qianmen Xiheyan Street and Xicaoshi East Street in Beijing. These rooms were chosen for their representativeness regarding building envelope characteristics and heating and cooling methods used during winter and summer, respectively (Room 11 was an unoccupied room).

Table 2. Room Information.

Location	Courtyard No.	Site Photos	Room Number	Room Function	Floor Area (m2)	Heating/Cooling Method and Operation Duration	Testing Duration	Testing Point Layout
Qianmen Xiheyan Street	A		A1	Bedroom (State-owned housing)	16.0	Electric Heating All day	1.25~3.8
			A2	Front Hall (Extensions)	20.0	Electric Heating All day
	B		B1	Bedroom (State-owned housing)	22.1	Electric Heating All day	1.25~3.8
			B2	Kitchen (Extensions)	4.0	None
	C		C1	Bedroom 1 (State-owned housing)	11.5	Electric Heating All day	1.25~3.8
			C2	Bedroom 2 (State-owned housing)	14.8	Electric Heating All day
Xicaoshi East Street	D		D1	Bedroom (State-owned housing)	10.3	Split Air Conditioner Part-time	7.5~8.31
			D2	Bathroom (State-owned housing)	3.8	None
	E		E1	Bedroom (State-owned housing)	39.9	None	7.5~8.31 1.25~3.8

shows on-site investigations, including testing and electricity consumption statistics. Exterior walls marked with red in the test point diagram indicate shared walls with adjacent rooms, the dimension unit in the figure is a meter. Due to the varying distances of test points from the electric heaters, slight discrepancies may exist between the measured temperatures and the actual indoor temperatures.

The Xicao Hongmiao district, an old city area constructed during the early years of the People’s Republic of China, has not undergone significant large-scale renovation to date. Most residents have extended their original dwellings inward into the courtyards, constructing self-built additions. Among the most common modifications by residents are upgrades to doors and windows, while external walls and roofs have seen minimal alterations. In the surveyed rooms, all doors and windows have been replaced with aluminum-framed windows with thermal barriers and uPVC doors. The roof assembly consists of a 20 mm wooden plank base, 50 mm straw–clay insulation, and 10 mm small gray tiles. The wall composition includes a 20 mm cement mortar layer, a 120 mm clay brick core, and a 20 mm cement mortar finish. Notably, Room C features an additional 50 mm polyurethane insulation layer applied to the eave walls for enhanced thermal performance. The thermal transmittance coefficients of enclosure structures are shown in

Table 3. Thermal Transmittance Coefficients of Enclosure Structures.

Courtyard No.	Thermal Transmittance (W/m2K)
	Measured Value of Exterior Walls	Estimated Value of Doors	Estimated Value of Windows	Estimated Value of Roof
A	2.5	2.5	3	2.5
B	2.8	2.5	3	2.5
C	1.5	2.5	3	2.5
D	3.8	2.5	3	2.5
E	3.5	2.5	3	2.5

.

In Beijing’s Siheyuan areas, winter indoor heating is provided through the use of storage-type electric heaters. These heaters utilize off-peak electricity rates during nighttime hours to store heat, which is then released during the day to meet residents’ heating needs.

2.2.1. Indoor Temperature Measurement and Analysis

Indoor temperature measurements were conducted in 11 rooms using Aqara temperature and humidity sensors (see 50 for specifications). These sensors recorded data whenever temperature changes occurred, with measurement parameters detailed in

Table 4. Aqara Temperature and Humidity Sensor Information.

Measurement Range	Resolution	Sampling and Recording Time	Image
−20 to 50 °C	0.1 °C	Recorded When Temperature Changes

.

Based on the field tests, the summarized data are shown in Figure 3, which illustrates the winter indoor temperatures of rooms A, B, C, and E and the summer indoor temperatures of rooms D and E. According to the Code for Design of Heating, Ventilation, and Air Conditioning in Civil Buildings (GB50736-2012) [24], the design indoor temperature standards for main rooms in cold regions are 18 °C for heating and 26 °C for cooling. These values were used as benchmarks to assess whether the rooms met indoor temperature requirements.

For heated rooms, the average indoor temperatures for rooms A1, A2, B1, B2, C1, C2, C3, C4, and E1 were 17.1 °C, 13.6 °C, 20.1 °C, 12.3 °C, 22.0 °C, 23.7 °C, 19.0 °C, 9.8 °C, and 2.6 °C, respectively. Among these, only rooms C1, C2, and C3 met the required temperature standard, while other rooms fell short. Room 11, an unoccupied space without heat transfer or radiation from adjacent rooms, exhibited extremely low temperatures, highlighting the inadequacy of indoor heating performance in Siheyuan.

For cooled rooms, the average indoor temperatures for rooms D1, D2, and E1 were 26.7 °C, 28.7 °C, and 30.1 °C, respectively. The cooling performance showed a pattern where occupied cooled rooms < adjacent cooled rooms < unoccupied rooms.

To further analyze temperature patterns, data were collected during six extreme cold days and six extreme hot days, and the hourly indoor and outdoor temperatures for typical rooms were plotted in Figure 4. In Figure 4a, from 15 December to 20 December, outdoor temperatures remained below 0 °C throughout the day due to the arrival of a cold wave. The temperature trends for rooms A1, B1, and C1 closely followed outdoor temperature variations, with notable differences between rooms. All three rooms relied on electric heaters during cold weather, but the hours with indoor temperatures above 18 °C were 19, 15, and 103 for A1, B1, and C1, respectively. While C1 generally met the temperature standard, A1 and B1 showed inadequate heating performance.

In Figure 4b, from 15 July to 20 July, outdoor temperatures exceeded 30 °C for 88 h, with daily peaks above 35 °C. Room E1 showed minimal correlation between indoor and outdoor temperature changes, with indoor temperatures fluctuating between 24.3 °C and 31.2 °C. The hours with indoor temperatures below 26 °C totaled 61. Moreover, these periods with relatively low temperatures are mainly concentrated at night. As can be seen from Figure 3, the number of hours during which the indoor temperature of air-conditioned room D1 is lower than 26 °C is even greater. From this, it can be reasonably inferred that using summer air conditioners can meet indoor temperature requirements.

2.2.2. Actual Electricity Consumption Statistics and HVAC Energy Usage Calculation

Electricity consumption data for Users A, B, and C were collected and analyzed via the ’State Grid Online’ mobile application over a 12-month period (March 2023–February 2024). Data for User D were unavailable. The results are presented in Figure 5. The data clearly show that electricity consumption during the heating season significantly exceeds the cooling season’s, with transitional seasons recording the lowest usage. This trend is attributable to Beijing’s cold climate, where winter heating demands a higher thermal load. Additionally, storage-type electric heaters, which are less energy-efficient and have higher power consumption compared to split air-conditioning systems, contribute to elevated electricity usage during the heating season.

Due to the limitations of the author’s survey methods, it was not possible to directly obtain the electricity consumption of the electric heating systems. Instead, the method proposed by Li Z [25] for processing energy consumption survey data was adopted. In this methodology, the average monthly electricity consumption during the transitional seasons is approximated by subtracting the baseline consumption from non-heating and non-cooling appliances. The energy consumption for air conditioning or heating is then calculated by subtracting the electricity usage of other appliances from the total electricity consumption during the air conditioning or heating season. The calculations are expressed in Equations (1) and (2).

$$W_{C\backslash H}=M_{C\backslash H}·\left(W_{Cs\backslash Hs}-{\displaystyle \frac{W_t}{5}}\right)$$

(1)

$$E_{C\backslash H}={\displaystyle \frac{W_{C\backslash H}}{CO{P}_{C\backslash H}}}$$

(2)

$W_{C\backslash H}$ is total electricity consumption during the cooling or heating season (kWh). The cooling season spans June to August (3 months), and the heating season spans Nov 15 to Mar 15 of the following year (4 months). $M_{C\backslash H}$ is the number of months in the cooling or heating season. $W_{Cs\backslash Hs}$ is the total electricity consumption during the cooling or heating season (kWh). $W_t$ is the total electricity consumption during the transitional season (April, May, September, October, and November) over 5 months (kWh). $E_{C\backslash H}$ is the energy consumption of air conditioning or heating systems (kWh). $CO{P}_{C\backslash H}$ is the coefficient of performance for air conditioning or heating systems.

The electricity consumption statistics for the heating and cooling seasons of the three selected households (A, B, and C) in Xihe are shown in

Table 5. Electricity Consumption Statistics.

Courtyard No.	A	B	C
Annual Total Electricity Consumption (kWh)	6076.0	8544.0	9857.0
Electricity Consumption During Heating Season (kWh)	4325.0	5767.5	7163.0
Electricity Consumption During Cooling Season (kWh)	560.0	980.0	731.0
Electricity Consumption of Other Appliances (kWh)	454.0	491.0	695.0
Heating Electricity Consumption (kWh)	3871.0	5767.5	6468.0
Cooling Electricity Consumption (kWh)	219.5	611.8	209.8
Proportion of Heating Electricity Consumption	63.7%	67.5%	65.6%
Proportion of Cooling Electricity Consumption	3.6%	7.2%	2.1%
Heating and Cooling Area (m2)	36.0	22.1	37.8
Heating Electricity Consumption per Unit Area (kWh/m2)	107.5	261.0	171.1
Cooling Electricity Consumption per Unit Area (kWh/m2)	6.0	27.6	5.6

. The total annual electricity consumption of A, B, and C was 6076 kWh, 8544 kWh, and 9857 kWh, respectively. Heating electricity consumption accounted for 63.7%, 67.5%, and 65.6% of the total annual electricity consumption for A, B, and C, while cooling electricity consumption accounted for 3.6%, 7.2%, and 2.1%, respectively. The heating electricity consumption per unit area was 107.5 kWh/m², 261.0 kWh/m², and 171.1 kWh/m² for A, B, and C, respectively. The cooling electricity consumption per unit area was 6.0 kWh/m², 27.6 kWh/m², and 5.6 kWh/m² for A, B, and C, respectively. These results clearly show that heating electricity consumption is significantly higher than cooling electricity consumption, with heating dominating total energy use.

Energy consumption varied among the three households. The thermal transmittance of exterior walls was 2.5 W/m²K for A, 2.8 W/m²K for B, and 1.5 W/m²K for C. The conditioned floor area was 36.0 m² for A, 22.1 m² for B, and 37.8 m² for C. The average indoor winter temperatures were 15.4 °C for A, 16.2 °C for B, and 18.6 °C for C. Household C, compared to A and B, had the lowest thermal transmittance for exterior walls, the largest conditioned area, and the highest average winter indoor temperature. Combined with C’s specific heating behavior, these factors resulted in unit-area heating electricity consumption higher than A but lower than B, and the lowest unit-area cooling electricity consumption. Table 5 presents detailed electricity consumption statistics for the three households.

3. Selection of Performance Evaluation Indicators

The performance evaluation indicators selected in this study include the following five aspects:

(1)

Thermal Environment and Comfort of Courtyard Spaces

The factors evaluated for the thermal environment of courtyard spaces include air temperature, wind speed, and Mean Radiant Temperature (MRT) [26,27]. These factors influence human thermal sensation differently and may counteract one another; therefore, a comprehensive evaluation of thermal comfort is required. This study adopts the Universal Thermal Climate Index (UTCI) [28,29], a widely recognized metric for outdoor thermal comfort, to assess the thermal conditions within the courtyard spaces. UTCI is defined as the equivalent temperature under actual conditions that elicit the same thermal sensation, and it is generally expressed in the form shown in Equation (3) [30]:

$$\mathrm{UTCI}\left(T_a,T_r,v_a,p_a\right)=T_a+\mathrm{Offset}\left(T_a,T_r,v_a,p_a\right)$$

(3)

In the equation, Offset represents the deviation between UTCI and air temperature, which depends on factors such as air temperature ($T_a$), mean radiant temperature ($T_r$), wind speed ($v_a$), and air humidity (expressed as water vapor pressure in pascals). Based on the UTCI values, outdoor thermal comfort can be classified into ten levels [30], as shown in

Table 6. UTCI comfort classification.

Stress Category	extreme cold stress	very strong cold stress	strong cold stress	moderate cold stress	slight cold stress
UTCI (°C) range	<−40	−40~−27	−27~−13	−13~0	0~9
Stress Category	no thermal stress	moderate heat stress	strong heat stress	very strong heat stress	extreme heat stress
UTCI (°C) range	9~26	26~32	32~38	38~46	>46

.

Due to the three-dimensional distribution of thermal environment factors in courtyard spaces, this study adopts the arithmetic mean of various environmental factors measured at a height of 1.5 m above the ground in the open space of the Siheyuan [31] as the representative value for the overall thermal environment level. The calculation formula is provided in Equation (4):

$$X=({\displaystyle \sum }_{i=1}^n{X}_i)/n$$

(4)

In the equation, $X$ represents the variables listed in

Table 7. Thermal environment and thermal comfort in courtyard spaces.

Variable	X	Units
Air Temperature	$T_a$	°C
Wind Speed	$v_a$	m/s
Mean Radiant Temperature (MRT)	$T_r$	°C
Universal Thermal Climate Index (UTCI)	UTCI	°C

; $n$ is the number of grids at a height of 1.5 m; and $X_i$ is the value of the variable in the i-th grid.

(2)

Indoor–Outdoor Natural Ventilation Performance

This study evaluates the natural ventilation performance of buildings based on the wind pressure difference across building facades [32]. To distinguish between windward and leeward facades, the outdoor wind direction is divided into eight primary directions—north, northeast, east, southeast, south, southwest, west, and northwest—each spanning a 45° interval, as shown in Figure 6. Here, a northward wind is designated as 0° (or 360°), eastward as 90°, southward as 180°, and westward as 270°. The distinction between windward and leeward facades is illustrated in Figure 7. The average facade wind pressure difference for courtyard buildings is the mean wind pressure difference between all windward and leeward facades, as shown in Equation (5).

$$P_m=\left({\displaystyle \sum }_{i=1}^p{S}_i{P}_i\right)/{\displaystyle \sum }_{i=1}^p{S}_i-\left({\displaystyle \sum }_{k=1}^q{S}_k{P}_k)\right)/{\displaystyle \sum }_{k=1}^q{S}_k$$

(5)

The scenario considered by this formula is that at a certain moment when the external air flow blows across the courtyard, each building in the courtyard will correspondingly form a windward side and a leeward side. In the equations, $P_m$ represents the average facade wind pressure difference (Pa); $S_i \mathrm{and} P_i$ are the area and wind pressure of the iii-th windward surface, respectively; $S_k \mathrm{and} P_k$ are the area and wind pressure of the k-th leeward surface, respectively, and $p$ and $q$ are the numbers of windward and leeward surfaces, respectively.

The natural ventilation rate of a room can directly measure the indoor–outdoor natural ventilation performance, expressed in terms of air exchange rate, and it influences the indoor thermal environment as well as HVAC energy consumption. For simplification, this study converts the courtyard’s average facade wind pressure difference into the air exchange rate for each room based on Equations (6) and (7).

$$\mathrm{Q}=\mathsf{\mu }\mathrm{F}\sqrt{2\mathsf{\Delta }P/\rho }$$

(6)

$$\mathrm{n}=Q/V$$

(7)

In the equations, $Q$ represents the natural ventilation rate of each room $\left({\mathrm{m}}^3/\mathrm{s}\right);\mu$ is the discharge coefficient of window openings, generally less than 1 [33];$F$ is the operable window area of the room $\left({\mathrm{m}}^2\right);\mathsf{\Delta }P$ is the facade wind pressure difference $\left(\mathrm{Pa}\right);\rho$ is the air density ($\left.1.2 \mathrm{kg}/{\mathrm{m}}^3\right);n$ is the air exchange rate $\left({\mathrm{h}}^{-1}\right)$; and $V$ is the room volume $\left({\mathrm{m}}^3\right)$. This study accounts for residents’ window-opening behavior, with a discharge coefficient of 0.83 for fully open windows and 0.11 for closed windows [34].

(3)

Indoor Thermal Environment

This study uses the natural room temperature (i.e., the indoor air temperature without air conditioning or heating systems) [35] as the indicator for the indoor thermal environment. The study also incorporates the degree-hours ($DH$) metric, as defined in the literature [36,37], to quantify the building’s heating and cooling demands. The calculation formulas for degree-hours are provided in Equations (8)–(11), defined as the difference between indoor heating or cooling design and natural room temperatures.

$$H_{DH, i}={\displaystyle \sum }_{\tau =1}^{\mathrm{HHs}}\max \left(t_{h,set}-t_{ni}\left(\tau \right), 0\right)$$

(8)

$$C_{DH,i}={\displaystyle \sum }_{\tau =1}^{\mathrm{CHs}}\max \left(t_{ni}\left(\tau \right)-t_{c,set}, 0\right)$$

(9)

$${\mathrm{H}}_{DH}={\displaystyle \sum }_{i=1}^n{H}_{DH,i}/{\displaystyle \sum }_{i=1}^n{A}_i$$

(10)

$${\mathrm{C}}_{DH}={\displaystyle \sum }_{i=1}^n{C}_{DH,i}/{\displaystyle \sum }_{i=1}^n{A}_i$$

(11)

In the equations, $H_{\mathrm{DH},i}$ and $C_{\mathrm{DH},i}$ represent the heating and cooling degree-hours for the $i$-th room (°C $\cdot \mathrm{h}$); $t_{n,i}\left(\tau \right)$ is the hourly natural room temperature for the $i$-th room (°C); $t_{\mathrm{h},\mathrm{set} }$ and $t_{\mathrm{c}, \mathrm{set} }$ are the indoor heating and cooling design temperatures set to ${18}^{ \circ }\mathrm{C}$ and 26 °C, respectively, in this study; HHs and CHs are the total hours of the winter heating period and summer cooling period, respectively; $H_{\mathrm{DH}}$ and $C_{\mathrm{DH}}$ are the heating and cooling degree-hours per unit area (°C$\left.\cdot \mathrm{h}/{\mathrm{m}}^2\right);A_i$ is the area of the $i$-th room $\left({\mathrm{m}}^2\right)$; and represents the total building area.

(4)

HVAC Energy Consumption

HVAC energy consumption for the courtyard is quantified using the annual cumulative heating and cooling loads, as well as the corresponding electricity consumption for these processes. The calculation formulas for these energy consumption indicators are provided in Equation (12):

$$E_s=E_{yr}/{\displaystyle \sum }_{i=1}^n{A}_i$$

(12)

In the equations, $E_{yr}$ represents the annual heating and cooling loads, as well as the electricity consumption for heating and cooling in the courtyard, measured in $\mathrm{kWh}$. ${\mathrm{A}}_i$ and ${\displaystyle \sum }_{i=1}^n{A}_i$ have the same meanings as defined above. $E_s$ denotes the energy intensity per unit area, measured in $\mathrm{kWh}/{\mathrm{m}}^2$

4. Simulation Calculations

The aforementioned performance evaluation indicators were obtained through simulations using ENVI-met (5.0.2) and DeST (20230713) software. The ENVI-met microclimate simulation software, developed by Bruse and Fleer in 1998 [29], can calculate indicators such as courtyard air temperature, wind speed, MRT, and UTCI. The DeST building energy simulation software [38], developed by Tsinghua University, can compute annual natural room temperatures and HVAC energy consumption for building spaces. The accuracy of both software tools has been extensively validated in previous studies [39,40,41,42,43,44,45,46].

4.1. Establishment of the Courtyard Geometric Model

To simplify the courtyard geometric model, the pitched roofs of all buildings within the courtyard were approximated as flat roofs based on their average roof heights. Following this adjustment, the average heights of the state-owned buildings and extended structures were set to 3 m or 4 m.

4.1.1. ENVI-Met Model Establishment

Based on the geometric information obtained from the field survey, the ENVI-met modeling diagram of the courtyard is shown in

Table 8. Model diagrams of ENVI-met for two types of courtyard patterns.

Courtyard	Extended Courtyards	Original Courtyards
Plan view
Isometric drawing

.

4.1.2. DeST Model Establishment

Similarly, the DeST model was established based on the geometric information of the courtyard, with the modeling diagram shown in

Table 9. Model diagrams of DeST for two courtyard patterns.

Courtyard	Extended Courtyards	Original Courtyards
Plan view
Isometric drawing

. Four typical rooms—the back enclosed-room, reverse oriented-room, east wing, and west wing—were selected from the two courtyard models for ease of result presentation and analysis.

4.2. Parameter Settings

4.2.1. Outdoor Meteorological Parameters

The outdoor meteorological data used in this study are derived from the Meteorological Dataset for Building Thermal Environment Analysis in China, jointly developed by Tsinghua University and the China Meteorological Administration, which provides typical meteorological year data for Beijing [47].

As ENVI-met simulations can model a maximum of four consecutive days, typical days are generally used for calculations [48]. This study divides the year into four seasons—spring (March to May), summer (June to August), autumn (September to November), and winter (December to February)—and selects 18 January, 16 April, 15 July, and 16 October, representing days with relatively abundant solar radiation, as the typical days for winter, spring, summer, and autumn, respectively. The key meteorological parameters for these four typical days are shown in Figure 8.

As some performance evaluation indicators have varying impacts during daytime and nighttime, it is not appropriate to use daily averages. This study divides the daytime and nighttime durations of each day based on the presence or absence of solar radiation. The division results for the four typical days are shown in Figure 9.

4.2.2. Thermal Performance Parameters of the Building Envelope

The materials and construction of the state-owned housing’s roofs, walls, doors, and windows were determined based on field survey results. For the extended buildings, as the building envelope materials were sourced from diverse origins and have since been dismantled, it was assumed that their building envelope materials were consistent with those of the state-owned housing.

The thermal performance parameters of the building envelope for the state-owned housing were estimated based on the physical properties of materials commonly used in historic buildings, as shown in

Table 10. Parameters of non-transparent building envelopes.

Building Envelopes	Materials	Thickness mm	Heat Transfer Coefficient W/m2·K
Wall	Cement mortar + Red brick + Cement mortar	10 + 240 + 10	1.356
Courtyard	Cement mortar + Red brick + Cement mortar	10 + 240 + 10	1.356
Roof	Chinese-style black tiles + Cement mortar + Wood roof boards	10 + 10 + 20	3.409
Door	Polyvinyl chloride + Glass	40	3.749

and

Table 11. Parameters of transparent building envelopes.

Building Envelopes	Materials	Thickness mm	Heat Transfer Coefficient W/m2·K	SHGC
window	Single-layer common glass	6	5.7	0.548

. According to the Residential Building Energy Efficiency Design Standard (DB11/891-2020) [49] for Beijing, the maximum allowable thermal transmittance for external walls, roofs, external doors, and windows are 0.35, 0.21, 1.5, and 1.1 $\mathrm{W}/{\mathrm{m}}^2\mathrm{K},$ respectively. Compared to these standards, the thermal performance of the building envelope in the state-owned housing is significantly inferior.

4.2.3. Occupant Activities and Equipment Schedules

Based on the survey, all state-owned housing units were designated as multifunctional rooms, while the extended buildings were assigned as auxiliary room types such as living rooms, kitchens, storage rooms, and entry halls, according to the actual survey findings. The densities of occupants, lighting, and household appliances for each room type are shown in

Table 12. The density of indoor occupants, lighting, and equipment.

Functional Areas	Room Area (m2)	Maximum Number of People	Personnel (W/Person)	Illumination (W/m2)	Household Appliances (W/m2)
Functionally composite rooms	4.2~39.2	2.0	53.0	5.0	15.0
Living room	4.2~27.6	4.0	53.0	5.0	9.3
Kitchen	8.9~12	2.0	53.0	5.0	48.2
Storeroom, Lobby	1.4~3.1	0.0	0.0	0.0	0.0

[50].

The schedules for occupants, lighting, household appliances, window operations, and HVAC usage were simplified based on actual conditions, as shown in Figure 10. It is assumed that courtyard residents leave home during the day and return at night, cooking before leaving and after returning, with lighting and appliances turned off during sleeping hours. In winter, residents open windows for ventilation for one hour before leaving and after returning home, while in spring, summer, and autumn, windows are kept open during the day and closed when residents return home. All room types except storage rooms and entry halls are equipped with heating and air-conditioning systems. The conditioned floor areas for extended courtyards and original courtyards are 917.2 m² and 628.6 m², respectively. Based on survey results and the Energy Efficiency Design Standard for Residential Buildings in Severe Cold and Cold Zones (JGJ26-2018) [51], the winter heating period is set from 15 November to 15 March, with heating systems operating continuously throughout the day, maintaining an indoor temperature of 18 °C and a relative humidity of 30%. The summer cooling period is set from 1 June to 31 August, with air-conditioning systems operating intermittently, off during the day and on at night, maintaining an indoor temperature of 26 °C, relative humidity of 60%, and considering a startup temperature of 29 °C [52].

4.3. Calculation of Performance Indicators

The performance evaluation indicators selected in this study were primarily obtained directly through simulations and processing using ENVI-met and DeST software, while also considering the impact of courtyard natural ventilation on indoor thermal environments and HVAC energy consumption. The specific calculation process is as follows:

(1)

ENVI-met was used to simulate the courtyard microclimate for typical days in each of the four seasons, generating NetCDF data files. Data processing was conducted using Python to compute the average values of key thermal environment indicators across the courtyard, including thermal comfort and facade wind pressure differences.

(2)

Based on the ENVI-met results for the typical days, a mathematical relationship was established to fit the average facade wind pressure difference with incoming wind direction and speed. Using this relationship and hourly wind speed and direction data from meteorological files, the annual hourly average facade wind pressure difference was calculated, which was then converted into hourly air exchange rates. These rates were used as ventilation schedules and input into DeST.

(3)

DeST was utilized to simulate and calculate the annual hourly natural room temperatures and HVAC loads for the courtyard buildings, obtaining indicators such as degree-hours, heating loads, and cooling loads.

(4)

Based on the hourly heating and cooling loads of the courtyard buildings and a fixed COP, the hourly HVAC electricity consumption was calculated. According to GB21455-2019 [53]: Minimum Allowable Values of Energy Efficiency and Energy Efficiency Grades for Room Air Conditioners, the COP for cooling was set at 3.7 for room air conditioners, while the COP for storage-type electric heaters was set at 1.

5. Results and Analysis

5.1. Thermal Environment of Courtyard Spaces

5.1.1. Thermal Environment Factors of Courtyard Spaces

Table 13. Distribution of thermal environment parameters at a height of 1.5 m at 12 o’clock on typical winter days in the courtyard.

Index	Extended Courtyards	Original Courtyards	Legend
Air Temperature
Wind speed
M R T

presents the distribution of environmental parameters, including air temperature, wind speed, and mean radiant temperature (MRT), at a height of 1.5 m in the courtyard space at noon on a typical winter day. It is apparent that the thermal environment parameters in courtyard spaces are significantly influenced by the extended structures, resulting in notable differences between the two courtyard configurations. Generally, compared to the original courtyard, the extended courtyard exhibits slightly higher air temperatures, significantly lower wind speeds, and reduced mean radiant temperatures.

Averaging these distributed data points provides a clearer understanding of the overall conditions. Figure 11 summarizes the hourly average air temperature, wind speed, MRT, and corresponding meteorological data for typical days across the four seasons in the two courtyard types. Figure 11a shows that the overall difference in air temperature between the original and extended courtyards is minimal, with both closely aligning with the outdoor meteorological temperature. Figure 11b illustrates that courtyard wind speeds in both types of courtyards are significantly lower than the incoming meteorological wind speeds, primarily due to the courtyard layout and structural shielding. The average wind speed throughout the day in the original courtyards was 0.5 m/s, 0.5 m/s, 0.3 m/s, and 0.4 m/s higher than in the extended courtyards on the four typical winter, spring, summer, and autumn days, respectively. The extended courtyard experiences a reduction in wind speed of approximately 50% compared to the original courtyard, primarily due to the additional built structures. While reduced wind speeds can improve thermal comfort by alleviating cold sensations in winter, they can hinder pollutant dispersion and heat dissipation during other seasons.

Figure 11c demonstrates that MRT values in the courtyards mainly fluctuate with solar radiation intensity, being higher during the day and lower at night. MRT values exhibit the lowest levels in winter and the highest in summer. The original courtyard, characterized by lower building density and greater solar exposure, exhibits higher daytime MRT values than the extended courtyard. On typical days in winter, spring, summer, and autumn, the maximum MRT values in the original courtyard exceed those of the extended courtyard by approximately 8 °C, 4 °C, 3 °C, and 6 °C, respectively. At night, the differences in MRT between the two courtyard types are negligible. The average nighttime MRT in the original courtyards was 0.9 °C, 0.1 °C, 0.1 °C, and 0.1 °C lower than that in the extended courtyards on the four typical days in winter, spring, summer, and autumn, respectively. Higher MRT values in winter improve thermal comfort by mitigating cold sensations in courtyard spaces, whereas in summer, they have the opposite effect, intensifying heat discomfort and reducing thermal comfort.

5.1.2. Thermal Comfort in Courtyard Spaces

The UTCI index was used to analyze the combined effects of the aforementioned factors on thermal comfort in courtyard spaces. Figure 12 presents the hourly average UTCI values and the daytime and nighttime statistical averages for typical days in each season. Figure 12a shows that the hourly average UTCI values for both courtyard types vary with outdoor meteorological conditions, following a pattern of “higher during the day and lower at night”. The minimum UTCI values occur at sunrise, while the maximum values are observed around noon or in the afternoon, specifically at 13:00 in winter and spring, 16:00 in summer, and 12:00 in autumn. The UTCI peaks do not align precisely with the times of maximum outdoor air temperature or solar radiation due to the combined effects of outdoor air temperature, wind speed, and solar radiation. For the two most thermally extreme seasons, winter and summer, the maximum daytime UTCI values for the extended and original courtyards are −1.9 °C/−2.7 °C and 45.3 °C/46.6 °C, respectively. This indicates that the original courtyard is 0.8 °C colder than the extended courtyard in winter and 1.3 °C warmer in summer. These differences are primarily due to wind speed variations dominating under winter’s cold conditions and MRT differences playing a more significant role under summer’s high-temperature conditions.

Figure 12b compares the arithmetic mean UTCI values for both courtyard types during the day and night. According to the UTCI comfort classification standards in Table 2, both courtyard types fall within the same thermal comfort ranges during winter, spring daytime, summertime, and autumn. Additionally, during the entire winter, the entire summer, and the daytime in autumn, both courtyards fall into non-comfort zones. Notably, both courtyards experience “strong cold” conditions during winter nights, though the extended courtyard has a significantly higher nighttime UTCI average than the original courtyard (by 3.1 °C). Furthermore, the thermal comfort in the original courtyard changes from “comfortable” to “slightly cold” during spring nights and from “strong heat” to “very hot” during summer days, compared to the extended courtyard. This highlights that both courtyard types experience varying degrees of thermal discomfort in their spaces.

5.2. Natural Ventilation Effect of Indoor-Outdoor Courtyards

5.2.1. Average Facade Wind Pressure Difference in Courtyards

Figure 13a presents the hourly variation in the average facade wind pressure difference for typical days across the four seasons. The average wind pressure for both courtyard types ranges from 0.1 Pa to 13 Pa, with the smallest values typically occurring in the summer. The hourly meteorological wind speed and direction influence the facade wind pressure difference. Under the same wind direction, higher wind speeds result in a more tremendous average wind pressure difference. As a result, the difference in the average wind pressure between the extended and original courtyards is significant, similar to the differences in courtyard wind speeds shown in Figure 11b. Additionally, due to the layout and building density of the courtyards, wind pressure differences vary even under the same wind speed but in different wind directions. Figure 13b shows that the daily average wind pressures for the extended and original courtyards range from 0.5 Pa to 2.5 Pa and from 1.0 Pa to 5.5 Pa, respectively. The original courtyard’s wind pressure is approximately twice as large as that of the extended courtyard. This suggests that the natural ventilation effect of the original courtyard is better than that of the extended courtyard. Effective natural ventilation helps with indoor air exchange and the dispersion of pollutants. In the summer, it aids in cooling through ventilation, but it may adversely affect thermal insulation in the winter. According to the wind pressure difference scoring criteria in GB/T50378-2019 “Green Building Evaluation Standard” [54], the original courtyard, with a facade wind pressure exceeding 5 Pa during the winter, would experience significant air infiltration, increasing the heating load. In contrast, both courtyards’ average summer facade wind pressure is below 1 Pa, indicating potentially poor indoor–outdoor natural ventilation.

5.2.2. Ventilation Air Changes

To convert the courtyard façade’s average wind pressure difference into ventilation air changes per hour (ACH), polynomial relationships are derived for the eight primary wind directions, correlating typical hourly facade wind pressure differences with corresponding meteorological wind speeds. The general form of the relationship is shown in Equation (13). In this equation, P represents the average wind pressure difference at each moment of the typical day, and s is the corresponding meteorological wind speed at that time. The coefficients a, b, c, and d, along with the goodness of fit $R^2$, for each wind direction, are provided in

Table 14. Extended courtyards fitting coefficient and goodness of fit.

Wind Direction	a	b	c	d	${\mathit{R}}^2$
N	3.1	−2.5	5.9	−3.7	0.8
NE	0.0	−0.2	0.6	−0.3	0.9
E	−1.9	1.8	−5.1	4.6	0.9
SE	0.1	−0.5	1.3	−0.6	0.8
S	0.1	−0.5	1.0	−0.3	1.0
SW	0.0	0.3	−0.7	0.5	0.8
W	0.1	−0.4	0.9	−0.3	0.8
NW	0.0	0.6	−1.7	1.5	0.8

and

Table 15. Original courtyards fitting coefficient and goodness of fit.

Wind Direction	a	b	c	d	R2
N	0.8	−0.7	1.6	1.0	0.9
NE	0.1	−0.2	0.3	0.3	1.0
E	−1.0	0.9	−2.7	2.4	0.9
SE	−0.1	0.6	−1.3	1.2	0.9
S	0.2	−1.2	2.6	−1.2	1.0
SW	0.0	0.0	−0.3	0.7	0.9
W	0.0	−0.2	0.2	0.1	0.9
NW	0.0	0.0	−0.1	0.5	0.9

.

Then, the hourly wind pressure difference for the entire year is obtained by applying Equation (13) along with the fitted coefficients from Table 14 and Table 15 and the typical hourly wind direction and speed data from the meteorological year. Finally, using Equations (6) and (7), the facade’s average wind pressure difference is converted into the air changes per hour (ACH) for each room. These air change values are then input into the DeST software to calculate the room’s natural indoor temperature and HVAC energy consumption.

$$P=a{s}^4+b{s}^3+c{s}^2+ds$$

(13)

Based on the conversion results from the above formulas, the ventilation air changes per hour (ACH) for typical rooms such as the back enclosed-room, reverse oriented-room, east wing, and west wing in the extended courtyard and original courtyard range from 0.003 h⁻¹ to 2.5 h⁻¹ and 0.04 h⁻¹ to 3.3 h⁻¹, respectively, on typical days. Variations in ACH across rooms are attributed to room volume and window area differences.

5.3. Indoor Thermal Environment

When analyzing the results of natural room temperature simulation by DeST, we intercepted the temperature data that exactly matched the measured time period and then calculated the absolute error between the simulated temperatures of the two courtyards and the measured temperature of Room 11 at corresponding moments. Specifically, in the comparison between the added courtyard and Room 11, 75% of the temperature error values were less than 4 °C; in the comparison between the original courtyard and Room 11, 75% of the temperature error values were less than 3 °C. Room 11 has no added space, and compared with the original courtyard, the environmental factors are more similar, so the error range between them is even smaller. These series of comparison results powerfully and effectively verify that the simulation results in this paper have a high degree of accuracy. Figure 14 presents the annual heating and cooling degree-hours for each room in the state-owned buildings of the two courtyards, arranged in descending order of heating degree-hours for the original courtyard rooms. For state-owned rooms in the extended courtyard, heating degree-hours ranged from 31,977 °C·h to 46,465 °C·h, while cooling degree-hours ranged from 1071 °C·h to 3534 °C·h. In the original courtyard, heating degree-hours ranged from 32,570 °C·h to 46,939 °C·h, and cooling degree-hours ranged from 1302 °C·h to 5143 °C·h. In both courtyard types, the heating degree-hours for most rooms exceeded the cooling degree-hours by more than tenfold, indicating that heating demand during winter is overwhelmingly dominant. Compared to the extended courtyard, approximately 84% of the state-owned rooms in the original courtyard exhibited higher heating degree-hours, and approximately 95% showed higher cooling degree-hours. This is due to the removal of extended rooms in the original courtyard, which increased the number of external walls in the state-owned rooms (previously, internal walls became external walls), leading to higher total wall heat transfer.

Table 16. Heating and cooling degree-hours.

	$\mathbf{Total} \mathbf{Degree}-\mathbf{Hours} (°\mathbf{C}\mathbf{·}\mathbf{h}$)		$\mathbf{Degree}-\mathbf{Hours} \mathbf{per} \mathbf{Unit} \mathbf{Area} (°\mathbf{C}\mathbf{·}\mathbf{h}/{\mathbf{m}}^2)$
	Extended Courtyards	Original Courtyards	Extended Courtyards	Original Courtyards
Heating	1,880,523.5	1,605,804.3	2027.3	2554.6
Air-conditioning	158,990.6	122,481.4	171.4	194.8

summarizes the total and per-unit-area heating and cooling degree-hours for all rooms (including state-owned rooms and extended spaces in the extended courtyard) under the two courtyard configurations. In terms of total degree-hours, the extended courtyard exhibited 17.1% higher heating and 29.8% higher cooling degree-hours than the original courtyard. Conversely, on a per-unit-area basis, the original courtyard showed 26.0% higher heating and 13.6% higher cooling degree-hours than the extended courtyard. This indicates that the extended courtyard has greater overall thermal and cooling demands, while the original courtyard requires more heating and cooling per unit area. Consequently, it can be inferred that the HVAC energy consumption relationship between the two courtyard types should follow a similar pattern.

5.4. HVAC Energy Consumption

5.4.1. Heating and Cooling Loads

Table 17. Statistics of total replacement heat consumption and cold consumption.

	Heat Consumption and Cold Consumption (MWh)		Heat Consumption and Cold Consumption per Unit Area (kWh/m2)
	Extended Courtyards	Original Courtyards	Extended Courtyards	Original Courtyards
Heating	174.0	144.3	187.6	229.6
Air-conditioning	7.8	5.9	8.4	9.3

presents the two courtyard types’ annual cumulative heating and cooling load indicators. Similar to the heating and cooling degree-hours, both the original and extended courtyards are dominated by heating demand, with annual heating loads significantly exceeding annual cooling loads. Compared to the extended courtyard, the original courtyard shows a 17.2% reduction in total annual heating load but an 18.3% increase in per-unit-area heating load. Similarly, the total annual cooling load decreases by 25.2%, while the per-unit-area cooling load increases by 10.4%. The reduction in total heating and cooling loads for the original courtyard is primarily attributed to its smaller total building area, which is 32.2% less than that of the extended courtyard.

Gui Chenxi et al. [55], combining actual energy-use data with simulation results based on the 2010 Energy Efficiency Design Standards (with roof, external wall, and external window thermal transmittance values of 0.35, 0.82, and 2.5 W/(m²·K), respectively), found that the annual heating and cooling load intensities of typical low-rise residential buildings (≤3 stories) in Beijing were 82.9 kWh/m² and 10.4 kWh/m², respectively. Compared to these benchmarks, the unit-area heating load of Siheyuan buildings is significantly higher, while the unit-area cooling load is slightly lower. This disparity arises from differences in building shape coefficients and the thermal performance of the building envelope.

To analyze the differences in heating load intensity between the two courtyard types, this study considers the indoor air as the object of heat balance, neglecting the heat storage of air, with heat gain considered positive and heat loss negative. The hourly heat transfer components for typical winter days were calculated and plotted, as shown in Figure 15. The results indicate that the unit-area heat loss through the building envelope in both extended and original courtyards is approximately equivalent, with a maximum value of around 130 W/m². This heat loss accounts for 77~97% and 53~86% of total heat loss in the extended and original courtyards, respectively, making it the largest source of heat loss and the primary contributor to heating loads. The differences in heating loads between the two courtyards are primarily attributed to variations in heat loss due to ventilation and air infiltration. The original courtyard exhibits greater facade wind pressure differences, resulting in higher air infiltration rates and increased heat loss during winter. Consequently, its unit-area heating load is larger than the extended courtyard’s.

5.4.2. HVAC Electricity Consumption and Costs

Table 18. Statistics on annual electricity consumption for air conditioning and heating.

	Cumulative Electricity (MWh)		Power Consumption Intensity (kWh/m2)
	Extended Courtyards	Original Courtyards	Extended Courtyards	Original Courtyards
Direct electric	174.0	144.3	187.6	229.6
Air-conditioner cooling	2.1	1.6	2.3	2.5

presents the annual HVAC electricity consumption metrics for both courtyard configurations. The annual heating electricity consumption for the extended courtyard is 187.6 kWh/m², and the cooling electricity consumption is 2.3 kWh/m². The original courtyard’s annual heating electricity consumption is 229.6 kWh/m², and the cooling electricity consumption is 2.5 kWh/m². In both courtyard types, heating electricity consumption accounts for over 98% of the total HVAC electricity consumption, making it the dominant factor. The annual total electricity consumption for HVAC systems is lower in the original courtyard than in the extended courtyard, but the unit-area HVAC electricity consumption is higher. This is consistent with the earlier analysis of heating and cooling loads.

According to Beijing’s “Coal-to-Electricity” heating subsidy policy [8], residents in old city areas do not follow the tiered electricity pricing structure. During the heating season (15 November to 15 March), a time-of-use pricing system is applied: peak hours (8:00–20:00) are charged at 0.4883 RMB/kWh, while off-peak hours (20:00–8:00) are charged at 0.1 RMB/kWh. Using this pricing structure, the annual HVAC electricity costs for the two courtyard types were calculated and are shown in

Table 19. Statistics of air-conditioning and heating electricity bills.

	Electricity Bill (RMB)		Electricity Bill per Unit Area (RMB/m2)
	Extended Courtyards	Original Courtyards	Extended Courtyards	Original Courtyards
Direct electric	50,747.1	42,085.1	55.2	67.4
Air-conditioner cooling	1074.3	830.1	1.2	1.3

. Based on field surveys of electricity usage by courtyard residents in the nearby Xihe Street area, direct electric heating consumption ranged from 110 kWh/m² to 260 kWh/m², with heating costs between 50 RMB/m² and 77 RMB/m². The simulation results for electricity consumption and costs in Table 19 fall within this range, indicating that the study’s simulation results effectively reflect the actual heating electricity usage levels in old city courtyards.

6. Conclusions and Recommendations

Taking Courtyard Nos. 72 and 70 in Sanyanjing Hutong, Dongcheng District, Beijing, as sample cases, this research employed ENVI-met and DeST software to calculate the indoor and outdoor thermal environment indices and HVAC energy consumption of the extended and original courtyards in Beijing’s old city. A comprehensive evaluation of their thermal and energy consumption features was made, with field-measured data used to verify the simulation results. Based on the findings, the similarities and differences in the courtyard and indoor thermal environments and HVAC energy consumption between the two types of courtyards were identified:

• Thermal Environment of Courtyard Spaces: (a) The wind speeds in both courtyards were notably lower than the incoming meteorological wind speeds, confirming the wind-blocking effect. However, the wind speed in extended courtyards was about 50% lower than in the original ones due to added structures. Lower speeds enhance winter comfort but hinder heat and pollutant dispersion in other seasons. (b) Original courtyards, with lower building densities, allow more solar radiation, leading to higher daytime MRT values. This benefits winter comfort but worsens summer heat. (c) On typical days, the UTCI values in both courtyards’ spaces were outside the comfort range in winter and summer, indicating thermal discomfort.
• Indoor–Outdoor Natural Ventilation: (a) The average facade wind pressure of original courtyards was over twice that of extended ones, matching their wind speed differences. (b) In winter, the facade wind pressures of the original courtyards exceeded 5 Pa, increasing air infiltration and heating loads. In summer, both had pressures below 1 Pa, suggesting poor ventilation.
• Indoor Thermal Environment: (a) Current Siheyuan rooms show inadequate heating performance. (b) The heating degree hours of rooms in both courtyards were over ten times their cooling degree hours, highlighting the dominant winter heating demand. (c) Extended courtyards had higher total heating and cooling degree-hours but lower unit-area values than original ones due to added structures.
• HVAC Energy Consumption: (a) Both courtyard types were mainly driven by heating needs, with annual heating loads far exceeding cooling loads. (b) The heating load intensity of Siheyuan buildings was significantly higher than that of typical low-rise residences in Beijing, while the cooling load intensity was slightly lower due to shape and envelope differences. (c) Heat loss through the building envelope was the major contributor to heating loads. Ventilation heat loss differences mainly caused the variation in heating load intensity. (d) Field measurements showed significant differences in HVAC electricity consumption among households, with heating accounting for about 65% of total energy use. (e) Extended courtyards consumed 187.6 kWh/m² for heating and 2.3 kWh/m² for cooling annually; original courtyards, 229.6 kWh/m² for heating and 2.5 kWh/m² for cooling. The high heating costs demand energy savings in old city courtyards.

Compared to previous studies, this paper analyzes the thermal environment and HVAC energy consumption more thoroughly, analyzing their differences. It helps understand key issues and guides future research and improvements. For the preservation and renewal of old city courtyards in Beijing, focusing on energy-efficient heating, it is recommended to enhance the insulation and airtightness of exposed walls and windows after removing extended structures. Partly, retention as thermal buffers could control heating energy consumption for non-disruptive extended spaces. Nevertheless, this study has limitations due to current simulation and data constraints: (1) Simulations could not be performed with one type of software, preventing yearly integrated hourly calculations. Simplified assumptions may affect accuracy; (2) The lack of direct measurement data for courtyard and room thermal environments restricts actual thermal comfort analysis. Future work should enhance simulation capabilities and strengthen field data collection for in-depth research on energy-efficient, low-carbon renewal.