A study of the impact of entrance space on indoor air quality in vernacular dwellings in desert areas during sandstorms

Abstract

Alxa League, located in the western part of Inner Mongolia, China, frequently suffers from sandstorms. Such dusty conditions result in sharp increases in respirable particulate matter (PM10) concentration in outdoor and indoor environments, posing a significant health risk to local inhabitants. Vernacular dwellings in this area feature a specific floor plan that includes an entrance space to safeguard indoor air quality during sandstorms while being compatible with the local climate. This study utilises CONTAM, a multizone indoor air quality and ventilation analysis computer program, and field measurements to quantitatively evaluate the effectiveness of the entrance space in protecting indoor air quality against ambient PM10 pollution under both dusty and normal conditions. The simulation results reveal that the entrance space can effectively mitigate PM10 pollution in the middle room, lowering the average concentration from 47.0 μg/m³ to 37.5 μg/m³ during dust periods. However, PM10 pollution may increase in the entrance space, reaching an average concentration of up to 70.0 μg/m³. Experimental outcomes align with the simulated results. Given that construction activities exacerbate desertification and frequent sandstorms, the above findings help identify the optimal design strategies for energy-efficient green vernacular dwellings in the targeted desert area, promoting environmental harmony and addressing climate change challenges.

1 Introduction

A sandstorm is a meteorological phenomenon characterised by a significant amount of sand and dust lifted from the ground by strong winds or upper air currents, causing reduced visibility to less than 1 km (Goudie 2009). These occurrences typically occur in arid and semi-arid regions (Soleimani et al. 2020), particularly in desert areas, where loose, dry soil and high wind speeds provide optimal conditions for developing and transporting dust and sand particles. In recent years, worldwide energy consumption has exponentially risen, with buildings being the major consuming sector (Ahmad and Zhang 2020). Substantial research has proved a strong correlation between climate change and extensive energy consumption (Akhmat et al. 2014). Accordingly, land degradation brought about by climate change, such as global warming and the El Niño phenomenon, has exacerbated desertification and increased the frequency of dust and sandy weather. According to China Climate Bulletin 2021¹, the number of dust events in China has remained high since 2000, with an annual average of 10.7 times.

Sandstorms are often associated with high levels of respirable particulate matter (PM10, particulate matter with an aerodynamic diameter of 10 μm or less), which poses a significant threat to public health. The outdoor concentration of PM10 during sandstorms can increase substantially and has been observed to exceed far the 24-h concentration limit (i.e., 45 μg/m³) recommended by the World Health Organization (WHO 2021) and the 24-h limit value (i.e., 150 μg/m³) given by the Ambient Air Quality Standards GB3095-2012 of China (Ministry of Ecology and Environment 2012). Respiratory diseases are the main health impact of sandstorms (Fussell and Kelly 2021; Oberdorster et al. 2004). Moreover, epidemiological data suggests that sandstorms can also increase the incidence and fatality of cardiovascular diseases (Hefflin et al. 1994; Ashley et al. 2015; Baddock et al. 2013). Ambient PM10 can penetrate indoor spaces through leaks and cracks of the building envelope even with closed doors and windows, leading to indoor PM10 levels that may approach outdoor concentrations in some cases (Wu et al. 2021; Yu et al. 2009; Liu and Nazaroff 2001; Wallace 1996). Accordingly, there is an urgent need for a low-energy approach to mitigating the impact of outdoor PM10 pollution on the indoor environment, especially during dusty weather.

Building design plays a pivotal role in shaping indoor environmental quality, including indoor air quality. In particular, the infiltration of outdoor particulate matter into indoor spaces depends on various building-specific factors, including air exchange rates, indoor particle removal rates, and building volume (Mitchell et al. 2007). The architectural layout and design of the building inherently influence these building-specific parameters. Jung et al. (2020) have highlighted the substantial impact of building type on indoor levels of PM2.5. Their findings revealed higher indoor PM2.5 concentrations in apartments compared with townhouses, which can be attributed to the larger cumulative effect of PM2.5 within the relatively smaller volume of apartments. Furthermore, our prior studies have emphasised the significant influence of diverse indoor floorplans in rural residences on the distribution of indoor PM concentrations across different rooms (Shi et al. 2023). This underscores the need to consider building design and layout carefully to manage and improve indoor air quality. Vernacular architecture, shaped by thousands of years of experience by local inhabitants, provides excellent examples.

As climate change becomes more of a problem, vernacular architectures have been more often revisited and studied as tangible examples of architectural sustainability in contemporary architectural practices. Alxa League, situated on the Alxa Plateau, is characterised by a fragile ecological environment, with a desert landscape being the predominant geographical feature (Da et al. 2012). The vernacular houses in the Alxa League have transformed from Mongolian yurts to three-roomed settled vernacular houses. According to the field interview materials, residents have expressed the intent to include entrance spaces in their dwellings to mitigate the adverse impact of the harsh, sandy and windy climate on indoor air quality (Liang and Zhou 2020). This adaptive design observed in the vernacular dwellings can be a passive measure for cost-effectively optimising indoor air quality. However, a quantitative analysis of the efficacy of such architectural design in safeguarding indoor air quality from sandstorms still needs to be done.

The primary aim of this research is to assess the efficacy of entrance spaces in safeguarding indoor air quality within vernacular dwellings situated in the desert settlement of Alxa Right Banner, Inner Mongolia, against sandstorms. To achieve this objective, this study begins by identifying representative floorplans of local residences derived from the evolution of local vernacular dwellings. Then, the effectiveness of the entrance space in mitigating the impact of outdoor PM10 pollution for the selected residences during both sandy and non-sandy periods was simulated and evaluated utilising the multizone indoor air quality and ventilation analysis computer program (CONTAM). Furthermore, the effectiveness of active measures was evaluated, specifically using air purifiers to preserve indoor air quality during ambient sandy and non-sandy weather conditions. This is a comparative reference for the passive measures by incorporating the entrance space in the middle room as a foyer. Finally, field measurements were conducted within residences featuring a three-room structure, with the primary objective of observing the influence of entrance spaces on indoor air quality under authentic, real-world conditions.

2 Studied tangible heritage of vernacular dwellings

2.1 Study area: Alxa League, Badain Jaran Gacha

The study area, Alxa League, is located in the west of Inner Mongolia Autonomous Regions, China, 37°24' ~ 42°47'N, 97°10' ~ 106°53'E, and is bordered by Ningxia Hui Autonomous Region to the south and Gansu Province to the west. The region is distinguished by its arid climate and desert landscape, which provide pasture for animal husbandry, such as the desert steppe. According to the summary of historical meteorological data of Alxa Right Banner (Wang 2019; Li 2016), gales tend to be most frequent during the spring months (March to May), while they are less common in autumn and winter. Spring seasons in this area are characterised by elevated wind speeds and minimal rainfall, with a prevailing southeasterly wind direction, creating conducive conditions for dust events. Notably, Alxa Right Banner hosts the Badain Jaran Gacha, a traditional pastoralist desert settlement (Fig. 1), a typical representation of the region’s local built environment.

Fig. 1

Location map of the pastoralist desert settlement (Source: Liang 2019, Page 37)

2.2 Evolution of vernacular dwellings

Since the establishment of the banner in the 17th century, the Alxa League has formed a unique residential culture of multi-ethnic intermingling under the influence of the migration movement (Li 2016). In the past, the local herdsmen lived in the Mongolian yurts, a product of nomadic civilisation. The yurt is easy to erect and dismantle, mostly consisting of wooden supports and felt carpet. Due to the migration movement, the residential cultures of different ethnic groups began to converge, and the forms of dwelling gradually transformed from nomadic to settled architecture influenced by the Han Chinese. Although the settled vernacular dwelling still maintains a nomadic identity, the construction structure and layout have changed considerably.

During the nomadic society, the Mongolians used wood, felt and rope to build yurts. Wood was used to build the hana² and ugni³ as the main support structure. After the Han Chinese moved in, earthen construction techniques were introduced and popularised in Alxa. Mongolians began to learn how to use local raw earth for construction. Wood is used as the roof structure, with raw earth and brick as the enclosure and main load-bearing structure.

The layout of the sedentary settled dwelling in Alxa evolved under the mutual influence of the official-style main house of the upper class and the three-room vernacular dwelling of the lower class in north China (Liang and Zhou 2020). Figure 2 shows the evolution of the main house plan. During the Qing dynasty (1644– 1911), while herdsmen lived in yurts, the upper class of Mongolians, including upper lamas and nobles, lived in the settled official residences, having a top-down influence that shifted the society from nomadic to sedentary. The long migration movement brought Han construction techniques to Alxa, promoting the transformation of dwellings from yurts to adobe houses with the simple ‘one family, one room’ layout (A-2). Influenced by the official-style residences (Fig. 3), a three-room layout (C-1) was developed, which was very similar to Han farmhouses. The three-room plan is symmetrical, with the middle room as a traffic space, connecting the left and right spaces.

Fig. 2

Evolution of the main house plan from yurt to settled dwelling (Source: Liang and Zhou 2020, Page 38)

Fig. 3

Official residences in Dingyuan Camp (Source: Liang and Shan 2020, Page 90)

Among the 150 examined residences situated in the Alxa region, the floor plans of 16 were surveyed and visually depicted in Fig. 4, which notably reveals that S02, S03, S06, S07, S10, S13, S14, S15, and S16 adhere to the conventional three-room layout design and incorporate an entrance space. In the cases of S02, S04, S06, S07, S14, and S16, deviations from the original three-room layout are evident, primarily prompted by the spatial demands arising from the tourism development within the area. Based on the 150 sampled residences, the three-room layout was identified as a prevalent and archetypal architectural configuration of the local dwellings. Besides, entrance space was also considered a dominant form of entry for the local residence.

Fig. 4

Field survey planes of 16 samples (Source: the authors)

The wind rose diagram in 2021 and a satellite map of local residences in the Alxa Right Banner are shown in Fig. 5. A notable feature of the local climatic conditions is the consistent predominance of northwesterly winds throughout the year (Fig. 5a), with a significant increase in the intensity of southeast winds during the spring season (Fig. 5b), causing the dispersal of substantial amounts of sand and dust, the local dwellings are generally oriented to the southeast, as shown in Figs. 4 and 5c. To mitigate the influence of the annual prevailing north winds on indoor environmental conditions, most local dwellings deliberately position their openings (i.e., windows and doors) to face south. Concurrently, an entrance space is customary to be placed on the southern side of the dwellings, facing the prevailing wind direction during sandstorm seasons. According to the field interview with the residents, the design of the entrance space is intended to serve as a protective barrier, shielding the indoor air quality of the leeward area from the heightened risk of sandstorms. Concurrently, it also obstructs the sunlight from the south, influencing the middle room’s lighting and radiation heat gain in winter. As a result, the middle room frequently incorporates a window on the north side to fulfil the light requirement.

Fig. 5

a Wind rose diagram in 2021 in Alxa Right Banner (Source: http://www.aqistudy.cn/). b Wind rose diagram in March 2021 in Alxa Right Banner (Source: http://www.aqistudy.cn/). c Partial satellite map of desert settlements in Alxa Right Banner (Source: Google Map)

Therefore, two representative vernacular dwelling designs were identified according to the analysis of the architectural layouts of local dwellings within the Alxa Right Banner region. These selected designs were subsequently employed in CONTAM simulations to assess the protective effect of the entrance space from the sandstorms. Detailed floorplans of these selected dwellings are visually presented in Fig. 6; both were built in 2002 and are of earthen construction. In particular, Case A epitomises the prevalent three-room layout devoid of an entrance space, while Case B is a prototypical example of the three-room layout incorporating an entrance space. These chosen cases facilitate a comparative analysis of the protective attributes of the entrance space in safeguarding the indoor environment against the adverse effects of sandstorms.

Fig. 6

Floorplans of the representative vernacular dwellings (Source: the authors)

3 Materials and method

This section presents the methodological framework employed in this study to evaluate the performance of the interventions in safeguarding indoor air against the detrimental effects of sandstorms.

3.1 CONTAM simulation

The CONTAM software is a widely recognised and validated tool for multizone airflow and contaminant transport modelling, relying on the multizone airflow network model (Emmerich et al. 2004, Lee et al. 2017). The modelling of transient inter-zone airflows within CONTAM is achieved by applying the mass conservation law to each zone within the residence, as expressed in the following equation (Dols and Polidoro 2015):

$$\begin{array}{c}{F}_{j,i}=f\left({P}_{j}-{P}_{i}\right)\\ \frac{\partial {m}_{i}}{\partial t}={\rho }_{i}\frac{\partial {V}_{i}}{\partial t}+{V}_{i}\frac{\partial {\rho }_{i}}{\partial t}=\sum\limits_{j}{F}_{j,i}+{F}_{i}\end{array}$$

(1)

where F_j.i is the airflow rate from zone j to zone i (kg/s), P_j is the pressure of zone j (Pa), P_i is the pressure of zone i (Pa), m_i is the mass of air in zone i (kg), t is the time (h), ρ_i is the density of air in zone i (kg/m³), V_i is the volume in zone i (m³). The specific type of airflow path determines the relationship between interzone airflows and interzone pressure differences. To simulate the zonal concentrations of PM10, the zonal mass conservation of PM10 based on the resultant inter-zone airflows can be expressed as follows (Dols and Polidoro 2015):

$$\frac{d{m}_{i}}{dt}=\sum_{j}{F}_{j,i}\left(1-{\eta }_{j}\right){C}_{j}-\sum_{j}{F}_{i,j}{C}_{i}-{K}_{i}{V}_{i}{C}_{i}$$

(2)

In this equation, C_i and C_j represent indoor PM10 concentration in zone i and zone j (μg/m³). η represents the filter efficiency of the path, equaling the particle penetration coefficient for the path connected to the ambient environment and being unity for the other paths. K represents the indoor particle deposition rate.

In this study, the zonal PM10 concentration within the two identified representative vernacular dwellings, with and without the entrance space, was simulated utilising CONTAM. This simulation was conducted to assess the impact of the entrance space on the preservation of indoor air quality against ambient PM10 pollution. The representative dwellings were represented graphically in CONTAM’s sketchpad, as illustrated in Fig. 7, with a floor height of 3 m (Liang and Shan 2019).

Fig. 7

Model plans of the representative vernacular dwellings in CONTAM (Source: the authors)

The simulation assumes that interior doors are open while the exterior doors and windows remain closed, a scenario representative of typical conditions observed during severe ambient PM pollution events. The determination of the effective unit leakage area, which characterises the average leakage of the exterior wall, was determined using a multiple linear regression equation based on building characteristics, including the building type, the construction year, and the floor area, as documented in a prior study (Chan et al. 2005). This calculation yielded an effective unit leakage area of 2.59 cm²/m² for the representative dwellings. The wind pressure coefficients for the flow paths along the exterior wall and roof were adopted from the research findings of Swami and Chandra (1988). The parameters of the remaining flow paths were configured in accordance with the built-in database of the CONTAM.

The effectiveness of air purifiers in maintaining indoor air quality within the representative dwelling lacking an entrance space was also assessed, utilising the CONTAM simulation. This simulation was designed to establish a comparative analysis between the passive protection offered by an entrance space and the active air purification measures targeting ambient PM10 pollution. One air purifier was positioned in the middle room of the representative dwelling without the entrance space (i.e., Case A in Fig. 6), which serves as an active measure for preserving indoor air quality in the face of sandstorms. The clean air delivery rate (CADR) is a metric used to evaluate the performance of air purifiers in removing airborne particles from indoor air, as established by the American Association of Home Appliance Manufacturers. In this study, the CADR index of the air purifier was determined to be 275 m³/h, a value derived from the performance parameters of top-selling brands of portable air purifiers in China, as documented in a previous study (Shi et al. 2018). Within the CONTAM simulation, a sink element was incorporated into the middle room to simulate the removal of indoor PM10 by the air purifier. The removal rate of this sink term was set as the ratio between the CADR index and the room volume. Other CONTAM settings for the air purifier simulation were configured in accordance with those utilised in the passive measure simulation, including the effective unit leakage area and the wind pressure coefficient, maintaining consistency in key parameters.

In both the passive and active measure simulations, the orientation of the two representative dwellings was set to face southeast, aligning with the prevalent dwelling orientation observed in the field survey (Fig. 4) and satellite map (Fig. 5c). The simulation duration spanned from March 1st to March 31st, 2021, to encompass varying weather conditions, including both non-sandstorm and sandstorm events. The measured indoor temperature using a temperature and humidity data logger during the study period was incorporated into the CONTAM model by creating a continuous value file. Local meteorological parameters for the study period were acquired from local meteorological stations and subsequently integrated into CONTAM using a weather file. As illustrated in Fig. 5b, the predominant wind during the study period is southeast, which is consistent with the typical meteorological characteristics of Alxa Right Banner (Wang 2019). Ambient PM10 concentrations acquired from the online air quality monitoring and analysis platform were integrated into the CONTAM model by creating a contaminant file. Particle-related parameters, including the penetration coefficient (P) and the particle deposition rate (K) of PM10, were set at 0.3 and 4 h⁻¹, respectively, based on previous studies (Li and Chen 2003; Shi and Zhao 2014). Subsequently, transient CONTAM simulations were conducted for both the passive and active measures throughout the study period, enabling an exploration of the dynamic characteristics of indoor PM10 levels. These simulations employed a time step of 5 min, facilitating the capture of short-term fluctuations in PM10 concentrations.

3.2 Field measurement

While local dwellings with a typical three-room plan, including an entrance space identical to the simulated floor plans, do exist in the vicinity, these vernacular dwellings were unoccupied and lacked electricity during the measurement duration. Consequently, a north–south oriented residence located in the Badain Jaran Desert of Alxa League, which is measurable, was recruited for the measurement to investigate the protective efficacy of the entrance space against sand storms under real conditions. The floor plan of the measured vernacular dwelling featured a layout that had evolved from the typical three-room plan with an entrance space, as depicted in Fig. 8.

Fig. 8

Floor plan of the measured vernacular dwelling (Source: the authors)

It can be observed from the floor plan that room AS01 serves a similar role to the entrance space, functioning as the windward room shielding the leeward room, AS02, from sandstorms. As a result, the monitoring campaign involved monitoring indoor PM10 concentrations within two rooms of the measured dwelling: room AS01 and room AS02. The field measurements were conducted from January 19, 2022, to January 21, 2022. Indoor PM10 concentrations were monitored using the PM-Model-II particle counter equipped with the low-cost sensor (Plantower PMS3003, Beijing, China), which has been extensively used in previous research on indoor air pollutants (Men et al. 2021; Shen et al. 2020), at intervals of one second. The sensor measures the particle concentrations in three size ranges, i.e., 0.3–1.0 μm, 1.0–2.5 μm, and 2.5–10 μm, with an upper limit of 500 μg/m³. The sensor was placed at a height of approximately 0.7 m, away from direct ventilation in each monitored room. Before the field measurement, all instruments were calibrated for one week at the same location and height. The calibration factors were obtained by regressing the measured PM10 concentrations against those reported by the nearest meteorological station.

In the monitored residence, the oven for heating was placed in AS01, whereas cooking activities were conducted in room AS02. Throughout the experiment, the external windows and doors remained predominantly closed, while the internal doors of both room AS01 and room AS02 were left open. The hourly meteorological data was obtained from the sandstorm network monitoring station in Badain Jaran Town, Alxa Right Banner. During the experimental period, outdoor temperatures ranged from -9.0℃ to 4.8℃. The prevailing wind direction was east, and the average wind speed was 3.4 m/s, with a maximum of 6.7 m/s.

4 Results

In the data analysis, the dust periods were identified to be between 14:00 on Mar 13th to 12:00 on Mar 20th and 14:00 on Mar 27th to 23:00 on Mar 30th, while the rest of time as non-dust periods based on the hourly PM10 concentrations recorded by the meteorological station and the sandstorm time points recorded in China Climate Bulletin 2021.⁴ The non-dust periods were further categorised into two distinct control scenarios, guided by the 24-h concentration limit of PM10 (i.e., 150 μg/m³) recommended by GB3095-2012 of China (MEE 2012). Therefore, the period between 19:00 on Mar 1st to 13:00 on Mar 13th was designated as control period 1, characterised by hourly PM10 concentrations generally lower than 150 μg/m³. In contrast, the period between 4:00 on Mar 23rd to 13:00 on Mar 26th was designated as control period 2, marked by hourly PM10 concentrations frequently exceeding 150 μg/m³ (see Fig. 9 for further details).

Fig. 9

Ambient hourly PM10 concentrations in March 2021 (Source: Meteorological station of Alxa Right Banner)

4.1 Simulated air change rate

The simulated average air change rates of the two simulated dwellings are represented in Table 1. As it shows, the simulated average air change rates of the two simulated dwellings under the same weather conditions exhibited tiny differences, indicating that the entrance space had a negligible effect on the overall ventilation of the whole residence. The average air change rates for rooms A2 and B2-2 are 0.45 h⁻¹ and 0.59 h⁻¹, respectively, which implies that the presence of an entrance space has a diminishing effect on the air change rate in the room situated leeward to it. The dynamic features of the air change rate can be discerned from Fig. 10a and b, which reveal a consistent fluctuation pattern in air change rates across all rooms within the two simulated dwellings. Furthermore, the middle rooms (A2 and B2-2) exhibit lower air change rates than the side rooms (A1, A3, B1, B3), further demonstrating the lower air infiltration of the middle room due to the relatively smaller surface area exposed to the external environment. Figure 10a and c illustrate a similarity in the fluctuating patterns between the air change rate and wind speed, particularly during episodes of strong winds. This observation underscores the prominent role of wind pressure as the primary driving force influencing the air change rate throughout the simulation period. Generally, an increased air change rate means introducing a larger volume of ambient air, which can dilute indoor air pollution and result in lower indoor contaminant concentrations if the incoming air is uncontaminated (Kraus 2016). However, it is essential to note that during dusty weather conditions, a higher inflow of ambient air may result in increased infiltration of PM10 particles. Our simulation results indicate that the air change rates of the simulated dwellings during dusty periods are significantly elevated compared to those observed during control periods. For instance, an exceptional peak in outdoor wind speed, reaching 30.6 m/s with a westward wind direction at 17:00 on Mar 27th, led to substantial air change rates of 26.16 h⁻¹ for rooms A1 and B1. These greater air exchange rates, combined with the elevated ambient PM10 concentrations during dusty and polluted periods, can substantially harm indoor air quality.

Table 1 Average air change rate of the whole dwelling (h⁻¹)

Fig. 10

a Hourly air change rate of each room in dwelling A. b Hourly air change rate of each room in dwelling B. c Variation of wind speed. (Source: http://www.aqistudy.cn/) d Variation of temperature. (Source: http://www.aqistudy.cn/)

4.2 Effectiveness of the entrance space

The simulated dynamic zonal PM10 concentrations of the two simulated dwellings during different periods were depicted in Fig. 11, while the statistical analysis of the hourly zonal PM10 concentrations was summarized in Table 2. Overall, the concentration differences observed in side rooms were not significantly different. However, in the middle rooms, there were considerable discrepancies in PM10 concentration between the two modelled residences. The time-averaged PM10 concentration in all rooms of the simulated dwellings except for B2-2 exceeded the WHO’s concentration limit of 45 μg/m³ due to the concurrent high level of outdoor PM10 concentration. For both simulated dwellings, the left and the right rooms of the three-room structure exhibited similar zonal PM10 concentration levels, with a relative difference of 6.1% and 5.5% between the two rooms for residences A and B, respectively. The middle rooms in both simulated dwellings showed lower time-averaged PM10 concentrations compared to the side rooms, which may be attributed to the smaller specific surface area of the exterior walls of the middle rooms. Among the examined rooms, B2-2, the protected middle room with the entrance space, displayed the lowest time-averaged PM10 concentration of 37.5 μg/m³, while A2, the middle room without the entrance space, had a larger time-averaged PM10 concentration of 47.0 μg/m³. Notably, during the dusty period, the prevailing wind direction is southeast. The modelled results suggest that an entrance space facing the prevailing wind direction mitigated the time-averaged PM10 concentration in the middle room by 20.2%. However, the time-averaged PM10 concentration in the entrance space (B2-1) was 70.0 μg/m³, greater than those in the middle rooms in both residences.

Fig. 11

Zonal hourly PM10 concentration of the two dwellings during different periods (Source: the authors)

Table 2 Statistical analysis of simulated hourly zonal PM10 concentrations (μg/m³)

Figure 11a illustrates the variation in simulated indoor PM10 concentrations in the modeled residences during the dusty period. The PM10 concentrations in the right rooms of the two residences reached a peak of 901.0 μg/m³ and 891.7 μg/m³, respectively, at 8:00 on 15th March. The highest peak values in the right rooms are likely attributed to the east wind direction during this period. Meanwhile, lower peaks in PM10 concentrations were observed in the remaining rooms with a lag time of less than 1 h. Generally speaking, the transient PM10 concentrations in the protective B2-2 were considerably lower than those in A2, which lacked an entrance space. However, B2-2 exhibited an outlier of 407.7 μg/m³ at 7:00 on 19th March, exceeding the maximum value observed in A2. This was attributed to the corresponding north-westerly wind direction and high wind speed of up to 9.4 m/s, indicating that B2-2 served as the entrance space for B2-1 during this period. The results suggest that the protective effect was closely related to the wind direction, with the windward room experiencing reduced inlet airflows from the ambient environments to the leeward side. Notably, the PM10 concentrations in B2-2 returned to normal levels (< 150 μg/m³) within 2 h after the first peak, significantly faster than the remaining rooms, which remained dusty for approximately 25 h. Furthermore, the results indicate that the entrance space did not affect the PM10 concentrations in side rooms. Therefore, it can be concluded that the entrance space facing the predominant wind direction provided protection against outdoor PM pollution for the middle room during the dusty climate while sacrificing its air quality.

During both control period 1 and control period 2, B2-2 demonstrated the lowest time-averaged PM10 concentration among the examined rooms. Specifically, in control period 1, B2-2 exhibited a time-averaged PM10 concentration of 1.2 μg/m³, while A2 had a time-averaged PM10 concentration of 2.4 μg/m³, suggesting that the presence of the entrance space reduced the time-averaged PM10 concentration in the middle room by 50.0%. During the more polluted control period 2, the time-averaged PM10 concentration was 19.1 μg/m³ in B2-2 and 30.0 μg/m³ in A2, indicating that the presence of the entrance space reduced the time-averaged PM10 concentration in the middle room by 36.3%. Nonetheless, it is worth noting that the protective effect conferred on the middle room comes at the expense of an increased time-averaged concentration in the entrance space during both control periods. Therefore, the entrance space is unsuitable for being arranged as the main living space for everyday human activities.

4.3 Effectiveness of the indoor air purification

The effectiveness of indoor air purification was also investigated to evaluate the effectiveness of passive intervention more accurately and with comparative reference. Table 3 summarises the simulated results of the statistical analysis of hourly zonal PM10 concentrations in A2 in the presence of an air purifier. The findings show that the PM10 concentrations in A2 were significantly lower with the use of an air purifier across all three periods studied. During the dust period, the time-averaged PM10 concentration in A2 declined from 47.0 μg/m³ to 24.9 μg/m³, representing a 47% reduction with the aid of the air purifier. Similarly, the air purifier lowered the time-averaged PM10 concentrations in A2 by 50% and 40% in control period 1 and control period 2, respectively. In contrast, the PM10 concentrations in the side rooms exhibited negligible differences between the air purifier and the no air purifier scenarios. During the dust period, the time-averaged PM10 concentration was 73.2 μg/m³ in A1 and 77.5 μg/m³ in A3, respectively, which only reduced by 0.4% and 0.6%, correspondingly, after the air purifier was deployed. Consistent with the dust period, the differences in time-averaged PM10 concentrations of side rooms between the air purifier and no air purifier scenarios were within 1 μg/m³ in control period 1 and control period 2.

Table 3 Statistical analysis of simulated hourly PM10 concentrations after placing the air purifier (μg/m³)

According to the modelled results of the entrance space scenario and the air purifier scenario, both interventions were found to be effective in safeguarding the indoor air quality of the middle room against the ambient dusty weather, but did not show a protective effect for the side rooms. Using an air purifier reduced 47.0%, 50.0%, and 40.0% in the time-averaged PM10 concentration of the middle room during the dust period, control period 1, and control period 2, respectively. Meanwhile, the entrance space displayed reductions of 20.2%, 50.0%, and 36.3% in time-averaged PM10 concentration of the middle room in dust period, control period 1, and control period 2. While the two methods demonstrated similar pollution reduction impacts based on time-averaged concentration values during the control periods, the active intervention (i.e., air purifier) was more efficient in reducing pollutant levels during the dust period.

4.4 Sensitive analysis

A sensitivity analysis was conducted during the dust period to evaluate the impact of two influential factors, including building orientation and envelope leakage area, on the protective efficacy of the entrance space, safeguarding indoor air quality from sandstorms. The envelope leakage area signifies the air tightness of the building façade. The parameters utilised in the CONTAM simulation, as elucidated in Section 3.1, served as the baseline for this analysis. Then, indoor PM10 concentrations within the two representative dwellings were simulated utilising CONTAM with varied influential factors. Each simulation pertained to the change of one influential factor, while all other input variables were maintained at their respective baseline values. In the sensitivity analysis, building orientation was systematically varied from south to east, with intervals of 30°, encompassing the predominant range of building orientations observed in the local dwellings. The leakage area was set to deviate by ± 50% from the baseline of 2.59cm²/m² in the sensitivity analysis, setting as 1.30cm²/m² and 3.89cm²/m², respectively.

Table 4 overviews the simulated average zonal PM10 concentrations within the representative dwellings during the dust period. Then, the relative difference in the indoor PM10 concentrations of the middle room between the dwellings with and without the entrance space (i.e., room A2 to B2-2 in Fig. 7) was computed as the decline ratio. This decline ratio serves as an indicator of the protective efficacy of the entrance space across different scenarios. The building orientation has a greater influence on the protective effect of the entrance space. When the building orientation ranges from south to east in 30° increments, the decline ratio is 15.5%, 20.2%, 22.9%, and 22.4%, respectively. When the residence is oriented to the southeast, aligning with the prevailing wind direction during the dust period, the entrance space shows a more effective protective impact on the indoor air quality of the middle room. However, the influence of envelope leakage area on the protective effect of the entrance space, concerning the preservation of indoor air quality against sandstorms, is relatively weaker. The decline ratio is determined to be 19.3%, 20.2%, and 20.8% when the leakage area is 1.30 cm²/m², 2.59 cm²/m and 3.89 cm²/m², respectively.

Table 4 Simulated average zonal PM10 concentrations and decline ratio of the entrance space in different simulated scenarios (μg/m³)

4.5 Field measurement analysis

Figure 12 displays the time series of measured PM10 concentrations from January 19, 2022 to January 21, 2022. Overall, outdoor PM10 concentrations varied from 24.0 μg/m³ to 155.0 μg/m³, while indoor PM10 concentrations varied from 25.0 μg/m³ to 196.0 μg/m³ in AS01 and from 20.1 μg/m³ to 211.3 μg/m³ in AS02. The PM10 concentration in AS02 showed intermittent spikes at around 8:00, 10:30 and 19:00, indicating the significant influence of cooking activities. To evaluate the protective effects of the entrance space against ambient PM10 pollution, the periods without any apparent indoor PM sources were selected for further analysis. The statistical analysis of the corresponding indoor PM10 concentrations is shown in Table 5. During these periods, the time-averaged PM10 concentration was 37.4 μg/m³ in AS01 and 30.7 μg/m³ in AS02, indicating that AS02 was shielded from outdoor PM10 pollution. As the prevailing wind direction during the measurement period was east and the dwelling was oriented southeast, the infiltration of PM10 from the ambient environment to AS02 was reduced due to the obstruction provided by AS01. These findings confirmed the simulation results and validated the protective role of the entrance space.

Fig. 12

a Temporal variations in the measured PM10 concentrations. b Hourly I/O ratio of the measured rooms. (Source: the authors)

Table 5 Statistical analysis of the processed PM10 concentrations (μg/m³)

5 Discussion and conclusions

This study presents a quantitative assessment of the effectiveness of entrance space, traditionally integrated into vernacular dwellings, as a passive measure for safeguarding indoor air quality against sandstorms in Alxa Right Banner. The key findings of this investigation highlight the capacity of entrance spaces to reduce indoor PM10 concentrations in the middle room effectively. The room located on the windward side mitigates the influence of outdoor sand and dust on the leeward side room; however, it comes at the cost of increased PM10 concentration within the windward side room itself. Although the active intervention is more effective in diminishing indoor PM10 concentrations in the middle room when contrasted with passive interventions via the entrance space, the passive intervention has an absolute advantage in reducing energy consumption, serving as a sustainable approach to air optimisation. It is therefore recommended to follow the floorplan with entrance space and to spend less time in it during dusty weather. However, it is important to acknowledge that the incorporation of an entrance space occupies a portion of the living area and can also adversely affect the middle room’s natural lighting and radiation heat gain. Consequently, further analyses, including illumination testing and thermal environment assessments, should be conducted when designing entrance spaces for local dwellings.

In Inner Mongolia, the built heritage is hard to preserve because of its nomadic nature. Compared to the yurt, the settled vernacular dwelling seems inconspicuous and valueless, with simple and rough construction techniques. However, learning from the invisible law of inheritance, its strategy for responding to the local culture and climate conditions can be essential; the specific layout with entrance space should be treated as a tangible or living heritage relating to the environmental adaptation. This quantitative analysis can serve as a theoretical basis for the design of dwellings in the ecologically fragile desert areas of not only Inner Mongolia but also other arid places across the world, which may face a similar challenge in the future, drawing on local construction wisdom to promote habitat development.