Evaluating the Impact of Road Layout Patterns on Pedestrian-Level Ventilation Using Computational Fluid Dynamics (CFD)

Abstract

The development of new energy vehicles and road dust removal technologies presents opportunities for constructing urban ventilation systems based on road patterns. However, the impact of road system layouts on pedestrian-level wind environments remains insufficiently understood. This study utilizes the general-purpose CFD software Phoenics to analyze the effects of road orientation, width, density, and intersection configurations on block ventilation. The standard k-ε model and three-dimensional steady-state RANS equations are employed to calculate pedestrian-level mean air age as an indicator of ventilation efficiency. Grid convergence analysis and validation against previous wind tunnel measurements were conducted. Results show that road layouts influence overall ventilation efficiency by affecting airflow volume, direction, and velocity. Optimal ventilation occurs when road orientation aligns with the prevailing wind at 0° or exceeds 70°. Recommended widths for trunk, secondary, and local roads are 46 m, 30 m, and 18 m, respectively. Lower densities of local road systems enhance ventilation, while higher densities of trunk and secondary roads are beneficial. Intersection configurations impact airflow distribution, with windward segments aiding lateral ventilation of side roads. Finally, ventilation design strategies for road systems are proposed, offering potential for leveraging urban road networks to construct efficient ventilation systems.

1. Introduction

Over the past three decades, China’s rapid urbanization has driven significant population concentration in cities, fostering the swift expansion of high-density urban building zones. This urban densification, however, has introduced pronounced environmental health challenges, including severe air pollution and intensified urban heat island effects. According to research by the World Health Organization (WHO), approximately 7 million people worldwide die each year due to air pollution [1]. The Lancet Countdown on Health and Climate Change: China report highlights a fourfold increase in mortality associated with high-temperature heatwaves in China since 1990 [2]. As densely populated areas, cities face heightened health risks linked to air pollution and urban heat island effects. Urban ventilation has been extensively acknowledged as a key strategy for alleviating urban heat island effects and reducing air pollutant concentrations [3,4].

To improve urban ventilation, researchers often focus on developing urban ventilation corridors. These corridors use natural wind patterns to enhance airflow, creating efficient pathways that allow wind to circulate smoothly through cityscapes. German scholars suggest classifying urban ventilation systems into zones based on the climate functions of the underlying surface: action spaces, compensation spaces, and air-guiding corridors. Ventilation corridors, as a type of air-guiding channel, have clear implementation standards, such as requirements for roughness, width, length, and obstacle control [5,6]. To address the urban heat island effect in the Tokyo Bay area, Bas et al. [7] conducted a study on the impact of four morphological parameters—street layout, canyon form, building arrangement, and tower height—on pedestrian-level wind environments. The study proposed methods to enhance urban ventilation efficiency. Similarly, in 2006, Hong Kong formally integrated the concept of ventilation corridors into its urban planning framework [8], advocating for airflow improvement strategies targeting green space configurations and building morphology at the neighborhood and block scales. These strategies were subsequently codified in the revised Hong Kong Urban Design Guidelines by the Hong Kong Planning Department. In China, cities such as Beijing, Wuhan, and Xi’an have similarly undertaken research on urban ventilation corridors [9,10]. This concept has proven to be both feasible and effective, with notable benefits in promoting air exchange between urban cores and peripheral areas, isolating heat island zones, and enhancing ecological climate regulation [11].

Urban surface spaces not only accommodate the majority of city functions but also serve as the principal activity zones for residents. At an approximate height of 1.5 m, corresponding to the average human breathing height, the ecological environment exerts a substantial influence on human health. This pedestrian-level layer, however, is highly vulnerable to environmental obstructions, contributing to a notably intricate wind environment. Ding et al. [12] employed an efficient numerical method to investigate the effects of building height and width on airflow and pollutant dispersion. They found that the environment around low-rise buildings may pose greater risks, with wider buildings creating larger lateral risk zones, while narrower buildings result in longer central wake regions. Zhong et al. studied 20 residential modules from 4 typical residential cluster types to investigate wind conditions at the 1.5-m pedestrian level. Their findings highlight how ventilation efficiency is closely tied to the coupling of building layout patterns across different module arrangements [13].

Roads, acting as linear open spaces in urban areas, are critical for pedestrian-level ventilation. Studies highlight that road traffic is a leading source of air pollution in cities, with vehicles generating heat, exhaust, and dust [14,15]. However, with the rise of new energy vehicles and effective dust control measures, roads are increasingly seen as potential elements of urban ventilation systems. In densely built urban areas, limited land resources make the ventilation role of road systems essential. Researchers are increasingly recognizing the significant contribution of roads and street networks to improving urban airflow. Feng et al. [16] studied traditional neighborhoods in Xi’an, focusing on factors like floor area ratio, building density, height, enclosure, elevation differences, aspect ratios, and views. They analyzed 12 neighborhood layouts from various eras, exploring how these parameters correlate with average pedestrian wind speeds. Liu et al. [17] conducted a numerical study on the dominant summer wind directions in the Sancha Street Canyon of Tianjin. The results showed that the ventilation performance was optimal when the street formed a 30° angle with the prevailing wind direction, while it was least effective at a 60° angle. When the street was parallel to the prevailing wind direction, the airflow penetration was the strongest. Jon et al. [18] examined the effects of street morphology (flat canyon, ascending canyon, and descending canyon) and wind orientation (90°, 60°, 45°, 30°, 0°) on internal ventilation and pollutant dispersion within streets. The study highlighted that, in terms of ventilation within the pedestrian breathing zone, descending canyons had higher ventilation rates in the leeward pedestrian zone compared to ascending and flat canyons. Sin et al. [19] evaluated the impacts of street configuration and wind direction on ventilation and pollutant exposure. Their findings revealed that compared to conventional street canyons, those with ventilated ground-floor designs exhibited higher air exchange rates. Furthermore, wind directions closer to the road axis provided better ventilation efficiency. Juan et al. [20] explored the impact of urban density, building corner designs, and wind direction on ventilation outcomes. They found that compact layouts promote good airflow infiltration, while sparse layouts increase wind speeds around building mid-levels and rooftops. Rounded corners, compared to sharp ones, lowered pedestrian-level wind speeds by 39% but increased wind speeds above mid-levels by 20%. Li et al. [21] investigated the impact of aspect ratios on ventilation potential by comparing idealized canyons with real-world canyons. Their study revealed that under longitudinal inflow conditions, both the height-to-width ratio and the length of the street canyon are closely related to the acceleration of flow within the canyon and its exchange with the surrounding atmosphere. Furthermore, airflow within urban street canyons is affected by a range of factors, including building frontal area density [22], height variations among buildings [23], street irregularities [24], and the continuity of street networks [25].

To summarize, most existing urban-scale studies emphasize constructing ventilation corridors based on climate factors, surface temperatures, and terrain analyses. However, they seldom explore the spatial patterns and interactions of urban pollutants, wind speeds, and air temperatures [26]. While some studies have examined the ventilation performance of urban street canyons, they typically focus on small areas or the airflow within roads themselves, leaving the development of an integrated urban ventilation network largely unexplored. This research employs the Phoenics (2019 v1.1) software program to perform numerical simulations of the 1.5-m pedestrian-level urban space, examining how variations in urban road system density, width, orientation, and intersection configurations influence regional ventilation performance. Based on these findings, it proposes strategies for establishing urban ventilation systems leveraging road networks. The conclusions aim to offer valuable insights for enhancing ventilation conditions in high-density urban areas.

2. Materials and Methods

2.1. Road Form Model

2.1.1. Road Orientation Model

The simulation of road orientation and its influence on ventilation conditions centers on evaluating the effect of the angle between road alignment and prevailing wind direction on regional ventilation efficiency. This study examines a block-scale area, maintaining consistent road density and width, while varying wind inflow angles to simulate ventilation performance. The inflow angle spans from 0° to 90°, with 10° increments, resulting in a total of 10 scenarios. Previous studies, such as those by Jon et al. [18] and Li et al. [27], have demonstrated the significant impact of road orientation on airflow characteristics and urban ventilation efficiency. These studies provide the theoretical foundation for the road orientation simulations conducted in this research.

2.1.2. Road Width Model

The Urban Comprehensive Transportation System Planning Standard (GB/T 51328-2018) [28] categorizes urban roads by type, prescribing red line widths: trunk roads range from 40 to 50 m, secondary roads from 20 to 35 m, and branch roads from 14 to 20 m. In this study, simulation scenarios were developed based on these classifications, with road widths adjusted at 2-m intervals within a consistent land area (Figure 1). The influence of road width on urban ventilation has been highlighted in prior research, including studies by He et al. [29], which emphasize the role of road width in determining airflow patterns and pollutant dispersion.

2.1.3. Road Density Model

With China’s current promotion of the “small block, dense road network” urban layout model, previous design standards for road density have been officially phased out. In response, this study examines the effect of road density on ventilation performance by varying road spacing. Building on the findings of Niu et al. [30] and Yin et al. [24], which demonstrated the significant influence of road density on urban airflow patterns and pollutant dispersion, the simulation scenarios were designed to evaluate ventilation efficiency under varying road spacing conditions. Roads are classified into three categories: trunk, secondary, and branch roads. Given the study’s focus on high-density urban environments, the research area is designated as a large metropolitan region. Design standards specify that trunk roads typically have a spacing of 800–1200 m, secondary roads 350–500 m, and branch roads 150–250 m. Since these road types serve different urban functions and cover varying service areas, the study ensures consistent land use for roads of the same type. Separate analyses are conducted (Figure 2), with five scenarios for trunk roads (100-m intervals), four for secondary roads (50-m intervals), and five for branch roads (25-m intervals).

2.1.4. Road Intersection Forms

Given that four-way intersections are the predominant form of urban road crossings, with T-shaped, Y-shaped, and other irregular intersections being relatively uncommon, this study focuses exclusively on variations of four-way intersections. Four-way intersections in urban areas are classified into four categories: orthogonal, single-directional rotation, opposite bidirectional rotation, and adjacent bidirectional rotation [31]. The analysis considers intersection angles of 45°, 60°, and 75°, excluding extreme angles, ultimately deriving 13 road intersection scenarios (Figure 3).

2.2. Building Layout Model

Given the inherent complexity of real-world building block forms and layouts, isolating the effects of individual factors on block ventilation becomes challenging. To address this, the study simplifies the analyzed building blocks into idealized three-dimensional physical models. This approach aligns with previous studies, such as those by Guo et al. [32] and Zeng et al. [33], which emphasize the utility of idealized models in capturing essential aerodynamic characteristics of urban forms while minimizing unnecessary complexity. To represent compact urban environments and adhere to model simplification requirements, high-rise buildings (40 stories) are used for arterial road block scales, mid-rise buildings (15 stories) for secondary road block scales, and low-rise buildings (6 stories) for branch road block scales (Figure 4). In the simulation models addressing road width, density, and orientation, all buildings are standardized as slab-style structures arranged in a row-column layout, assumed to be parallel to adjacent roads. To ensure precision in evaluating the road system, actual building setbacks are excluded from the models. In accordance with Wijesooriya et al.’s recommendations, two upstream rows of buildings and at least one downstream row are incorporated for all wind directions to mitigate the impact of wake recirculation on the leeward side of urban configurations [34]. These design choices are supported by studies such as Wang et al. [35], which underline the importance of including surrounding building rows to account for wake effects in urban ventilation simulations.

2.3. Ventilation Efficiency Evaluation Index

The ventilation performance of urban blocks is assessed using the concept of mean air age. According to the air concentration transport equation, the governing equation for steady-state air age is given by [36]:

$$S_{\phi i}=\nabla \left(\rho \overline{v}{\phi }_i-{\mathsf{\Gamma }}_i\nabla {\phi }_i\right)$$

where ${\phi }_i$ is the air age (s). $\rho$ is the fluid density (kg/m³). $v$ is the fluid velocity (m/s). ${\mathsf{\Gamma }}_i$ is the diffusion coefficient, ${\mathsf{\Gamma }}_{\tau i}=2.88\times {10}^{-5}\rho +{\displaystyle \frac{u_{eff}}{0.7}}$, determined by the effective viscosity $u_{eff}$ of the fluid. $S_{\phi i}$ is the source term, usually set to 1.

Air age serves as a key indicator for explaining pollutant concentration levels and temporal dynamics within a space. For homogeneous pollution sources with consistent emission timescales, air age can help determine the concentration levels at the emission source. Areas with poor ventilation exhibit higher air ages, indicating prolonged times for fresh air to reach specific regions, slower pollutant removal rates, and increased pollutant accumulation. The simulations in this study were conducted using PHOENICS, a computational fluid dynamics (CFD) software package widely used for urban ventilation studies. Phoenics was selected due to its robust implementation of the Reynolds-averaged Navier–Stokes (RANS) equation and its capability to handle complex geometries and urban morphologies. Its suitability for air age calculations and ventilation analysis has been validated in previous studies [37]. To evaluate the average air age and spatial frequency distribution across the simulation domain, 1500 sampling points were strategically distributed within the region of interest.

2.4. Validation Experiment Data

The validation experiments in this study utilized wind tunnel simulation data from He et al. (2019) [38], conducted in the boundary layer wind tunnel at the National University of Singapore. The urban morphology was scaled to 1:500 and positioned on a 3.4 m diameter turntable. The windward section was covered with 60 rows of square roughness elements, measuring 0.04 m in width, 0.07 m in length, and 0.09 m in height, achieving a scaled aerodynamic roughness length Z₀ = 0.065 m. At a full-scale height of 10 m, the mean wind speed was 4 m/s. The turbulence intensity at the entrance to the measuring section was approximately 10%, which is consistent with typical urban boundary layer conditions. Roughness length and friction velocity were derived from a linear fit to the measured U-profile, ensuring fully developed turbulence and minimizing viscous stress effects in the test section. Experimental data included wind speed coefficients measured along the road centerline at 0.008 m height (scaled to 4 m in full size). Wind speeds were captured at 21 test points across four road segments, with pressure readings converted into average wind speeds. Validation based on the available velocity field was considered appropriate, as previous studies show that 3D steady RANS models with linear k-ε schemes can accurately estimate mean wind speeds in urban environments.

2.5. Model Parameter Settings

2.5.1. Computational Domain and Grid

Phoenics was employed to simulate the wind environment in this study. The downstream length of the computational domain was defined following best practice guidelines, while the upstream length was set to three times the maximum building height to prevent the unintended development of streamwise gradients. To minimize blockage effects, all road configuration models were proportionally scaled based on their wind tunnel counterparts. The computational domain for the road orientation model was set at 4600 × 4600 × 350 m³, whereas the domains for road width and intersection models measured 1600 × 1600 × 350 m³.

To ensure optimal control over grid topology and quality, the computational domain was divided into two subdomains: a central subdomain and an outer subdomain. The two subdomains employed distinct grid generation approaches, with the outer subdomain using coarser grid sizes and the central subdomain utilizing finer grid resolutions for greater accuracy. A grid sensitivity analysis was performed by refining and coarsening the grid by a factor of approximately 2 to validate grid accuracy. Based on the sensitivity test results, a suitable grid density was determined and applied as the standard for other cases.

2.5.2. Boundary Conditions and Solver Settings

For the inlet boundary conditions, U∗ is calculated from the reference velocity (Uref) at the building height (zref) using Equation (1). The inlet velocity profile, fitted using Equation (1) and calculated in Equation (2), is shown in Figure 5. Similarly, the turbulence kinetic energy (k) and turbulence dissipation rate (ε) distributions are estimated using Equation (1) in conjunction with Equations (3) and (4).

$$U_{\mathrm{*}}={\displaystyle \frac{k{U}_{ref}}{\ln \left(\left(z_{ref}+z_0\right)/z_0\right)}},$$

(1)

$$U={\displaystyle \frac{U_{\mathrm{*}}}{k}}\ln \left({\displaystyle \frac{z+z_0}{z_0}}\right),$$

(2)

$$k={\displaystyle \frac{U_{\mathrm{*}}^2}{\sqrt{C_{\mu }}}},$$

(3)

$$\epsilon ={\displaystyle \frac{U_{\mathrm{*}}^3}{k\left(z+z_0\right)}},$$

(4)

where Z₀ = 1.5 m, the von Karman constant (k) is 0.4, and the model constant (Cμ) is 0.09. The remaining boundary conditions, to replicate the wind tunnel test section velocity profile gradient, involve setting a rough surface with Z₀ = 0.065 m on the ground. Zero static pressure is applied at the domain outlet, and free-slip conditions are set on the top and side walls of the computational domain. To solve the 3D steady-state RANS equations for incompressible isothermal flow, both the standard k-ϵ turbulence model and the RNG k-ϵ turbulence model are employed. The standard k-ϵ model is widely recognized for its computational efficiency and robustness in simulating general urban flow conditions, particularly in regions with relatively uniform turbulence distributions. However, the RNG k-ϵ model incorporates an additional term in its dissipation rate equation, improving its ability to capture flow separation and recirculation in areas with high turbulence intensity, such as building wakes. The use of both models allows for a comparative analysis of their performance under complex urban geometries and varying turbulence conditions. This approach ensures that the chosen model for subsequent parametric studies provides reliable and accurate results for the intended applications.

3. Results and Discussion

3.1. Grid Sensitivity Study

To validate the model, three grid densities were generated by refining and coarsening the standard grid, applying an adjustment factor of 2. The coarse grid contained 846,544 cells, the standard grid 1,596,048 cells, and the fine grid 3,290,148 cells. The grid convergence index (GCI) was utilized to quantify the discretization errors associated with the grid resolutions. Figure 6a presents the horizontal wind speed profiles along the road centerline derived from three different grid resolutions, and Figure 6b illustrates the error bounds resulting from GCI calculations based on the standard grid solution. As shown in Figure 6a, variations in grid resolution exert minimal influence on the horizontal wind speed along the road centerline. The overall trends remain consistent across all grid resolutions, with only slight discrepancies observed in specific regions. The error band generated by the standard grid validates the high grid sensitivity of horizontal wind speed simulation results along the road centerline. Downwind road segments (e, g) demonstrated narrower error bands compared to crosswind segments (f, h). Despite these observations, the relatively minor discrepancies justified the use of the standard grid for analyzing other cases.

3.2. Model Validation Study

Case VS-3 was used to evaluate the accuracy of different turbulence models (Figure 5). Two widely adopted turbulence models, standard k-ε (STD) and RNG k-ε(RNG), were validated against wind tunnel experimental data provided by He et al. (2019) [38]. Figure 7 presents a comparison of the average wind speeds at test points along four road directions with the corresponding wind tunnel results. Overall, the simulation outcomes aligned well with the experimental data, with the STD model demonstrating the highest consistency.

Figure 8 further illustrates the wind speed distribution at each test point. The smallest deviation between simulation and measurement was found in the flow direction sections (Figure 8a,f). Additionally, smaller deviations were observed at most test points in the oblique inflow sections (10c,d). However, significant differences were seen in the building-shadowed areas of the transverse sections (Figure 9).

These discrepancies can be attributed to two reasons. First, the turbulence model overestimated the mean wind speed in the building wake, where turbulence intensity is high. Second, it underestimated k in the wake and ultimately overestimated the size of the cavity region [39]. In the current results, the STD outperformed the RNG, consistent with previous validation findings. For high wind areas, the performance of STD was similar to RNG, except for when β = 45° in S2 (Figure 8d). It is known that STD overestimates k near the building corners, thus underestimating flow separation and recirculation sizes. However, the adequate performance of STD in the current results suggests limited impact on pedestrian-level flows in the canyon. Considering STD’s better overall estimation of canyon flow, which is the focus of this study, it was chosen for the parametric study. It should be noted that STD has also been used in other numerical studies assessing street canyon wind conditions [40].

3.3. CFD Simulation Results

3.3.1. Models with Different Angles Between Roads and Prevailing Wind Direction

The relationship between the prevailing wind direction and block orientation reveals that as the angle β rotates from 0° to 90°, the block’s average air age initially increases and then decreases (

Table 1. Average air age of blocks with different wind directions and road intersections.

Wind Direction and Road Intersection Angle	Mean Air Age(s)
0°	245.8378
10°	333.0408
20°	336.4999
30°	330.8685
40°	319.0469
50°	302.9407
60°	284.5542
70°	269.4638
80°	263.3725
90°	262.0616

and Figure 9). At β = 0°, the average air age is at its minimum (245.8378 s), while the maximum air age of 336.4999 s is observed at a wind direction of 20° relative to vertical roads, reflecting the weakest ventilation performance. Beyond 20°, the average air age decreases progressively. Between β = 10° and 40°, localized regions of high air age develop on the leeward left side of the block, with the size of these regions diminishing as the wind angle increases. This suggests that lateral airflow is weak and vortex zones form on the leeward side of buildings, thereby impairing ventilation.

In conclusion, the block demonstrates superior ventilation performance when β = 0° or exceeds 70°, with minimal impact on surrounding blocks. This can be attributed to an increase in the number of inflow channels as the block aligns at specific angles to the wind direction. At angles greater than 70°, the width of the inflow channels reaches its maximum, while at β = 0°, the vortex zones on the leeward sides of buildings are significantly reduced. These observations align with the findings of Sari et al. [41].

3.3.2. Impact of Different Road Widths on Ventilation Capacity

As depicted in Figure 10, the introduction of wider urban branch roads leads to a fluctuating pattern in block ventilation capacity (

Table 2. Average air age of blocks with different road widths.

Type of Road	Width (m)	Mean Air Age (s)
Spue road	14	733.1439
	16	691.5106
	18	684.7663
	20	687.8025
Secondary road	22	652.8138
	24	651.9089
	26	654.2791
	28	627.6394
	30	650.1060
	32	623.0651
	34	538.3546
Trunk road	40	535.7709
	42	538.3824
	44	477.5441
	46	457.2494
	48	464.8015
	50	460.1007

). At W = 14 m, the block exhibits the highest average air age of 733.1439 s. Similarly, simulations of secondary road widths reveal that block ventilation performance, indicated by average air age, shows a degree of variability as road width increases. The maximum average air age of 654.2791 s is observed when W = 26 m. When W ≥ 34 m, a notable shift in airflow patterns occurs within the block, resulting in a significant reduction in average air age. At W = 34 m, the secondary road block achieves the lowest average air age of 538.3546 s. It is worth noting that the average air age at W = 30 m shows a slight increase compared to W = 28 m and W = 32 m. This phenomenon can be attributed to localized airflow changes near the leeward sides of buildings, where partial recirculation or turbulence temporarily reduces ventilation efficiency [42]. Additionally, as the reported values are derived from multiple simulation runs, minor fluctuations in average air age may occur due to inherent variability in computational fluid dynamics simulations. Despite this localized variation, the overall trend indicates that increasing road width generally improves ventilation capacity. The relationship between arterial road width and block ventilation capacity demonstrates a fluctuating trend. For W = 44/46/48/50 m, the airflow pattern is consistent, with localized high-air-age zones predominantly located on the leeward sides of buildings in the downwind half of the block. The highest average air age of 538.3824 s is observed at W = 42 m, whereas the lowest, 457.2494 s, occurs at W = 46 m.

The findings indicate that block average air age follows the order: arterial road blocks > secondary road blocks > branch road blocks. Wider streets facilitate stronger ventilation flows, enhancing their effectiveness in promoting airflow. Specifically, for roads narrower than 14 m, block ventilation is weakest, with an average air age exceeding 700 s. As road width increases to 14–32 m, ventilation capacity improves, resulting in an average air age of 600–700 s. For road widths between 32 m and 42 m, ventilation capacity further increases, with an average air age of 500–600 s. When road widths exceed 42 m, block ventilation is at its peak, with an average air age below 500 s. Simulation results indicate that with increasing road width, the extent of localized high-air-age regions within the block diminishes, accompanied by a reduction in the area of high-air-age zones on the leeward sides of upstream buildings. This phenomenon is likely attributed to the restricted airflow caused by narrow urban roads, which impede air circulation in the upstream spaces of confined street configurations [43].

3.3.3. Impact of Different Road Densities on Ventilation Capacity

Simulation results for urban road density reveal a non-linear relationship between branch road density and block ventilation (

Table 3. Average air age of blocks with different road densities.

Type of Road	Distance (m)	Mean Air Age (s)
Spue road	150	274.3000
	175	262.0653
	200	250.1670
	225	249.3367
	250	268.6297
Secondary road	350	441.3983
	400	458.7060
	450	452.6969
	500	449.3969
Trunk road	800	870.3474
	900	846.9341
	1000	881.5802
	1100	891.8147
	1200	908.8112

and Figure 11). As branch road spacing increases from 150 m to 225 m, the average air age within the block initially decreases, reaching its lowest localized air age of 249.3367 s at 225 m spacing. Conversely, at 150 m spacing, the average air age peaks at 274.3 s. Similarly, simulations for secondary road density indicate that as road spacing increases, block average air age first rises and then falls. The lowest average air age of 441.3983 s occurs at a spacing of 350 m, while the highest value of 458.7060 s is observed at 400 m spacing. Analysis of air age distribution within the block indicates that reduced secondary road density leads to an increase in localized high-air-age regions. Simulations of arterial road density reveal a fluctuating upward trend in average air age as arterial road density decreases. Similar to the secondary road results, localized high-air-age regions within the block become more pronounced. The lowest average air age, 846.9341 s, is observed at an arterial road spacing of 900 m, while the highest average air age, 908.8112 s, occurs at 1200 m spacing.

The findings indicate that urban road density significantly influences overall block ventilation efficiency. Higher densities of arterial and secondary roads are conducive to enhanced block ventilation; however, this trend does not extend to branch roads, where increased density adversely affects ventilation performance. These results are inconsistent with the conclusion by Xie and Maing [44] that “small blocks with dense road networks” facilitate ventilation. Conversely, other research demonstrates a weak negative correlation between arterial and secondary road densities and air pollution, while branch road density exhibits a moderately strong positive correlation with air pollution, aligning with our study’s results [45,46].

3.3.4. Impact of Road Intersection Forms on Ventilation Capacity

In orthogonal road intersection configurations (α = 90°), the block exhibits optimal ventilation performance, characterized by an average air age of 254.4644 s (

Table 4. Average air age of blocks in different road intersection forms.

Type of Road	Angle of Rotation	Age of Air (s)
orthogonal	0°	254.4644
One-way forward rotation	45°	266.5024
	60°	259.6972
	75°	257.2023
One-way reverse rotation	45°	280.5455
	60°	271.5965
	75°	275.2031
Opposite two-way rotation	45°	284.6735
	60°	280.0662
	75°	279.2981
Adjacent two-way rotation	45°	262.7773
	60°	269.3913
	75°	267.9079

and Figure 12). As the downwind segment S3 rotates from 75° to 45° (single-directional rotation), the block’s average air age steadily increases, reflecting a gradual decline in ventilation efficiency. The highest average air age, 266.5024 s, occurs at α = 45°. The single-directional rotation intersection simulation reveals a notable rise in localized air age in the downstream a4 block, accompanied by a rightward shift of the high-air-age region in the a3 block. Angled road configurations, compared to orthogonal layouts, facilitate greater horizontal flow across a larger number of road segments [47]. However, this configuration causes the flow from S1 to be partially diverted into S2, limiting airflow reaching S3 and thereby reducing its ventilation capacity. Additionally, the rotation of S3 alters wind direction and expands the area of the a4 block, enhancing ventilation in the a3 block while decreasing it in the a4 block. These outcomes align with the observations reported by Guo et al. [31].

The simulation results indicate that when the downwind segment S1 rotates from 75°to 45°(single-direction counter-rotation), the block’s average air age initially decreases but subsequently increases, peaking at 280.5455 s at α = 45°. Similarly, when both S1 and S3 rotate from 75° to 45°(opposite bidirectional rotation), the average air age gradually rises, reflecting a decline in overall ventilation performance. At α = 45°, the average air age reaches its maximum of 284.6735 s. Both scenarios highlight reduced ventilation efficiency, primarily attributed to the rotation of S1, which reduces inflow volume in the upstream S1 segment. This is partly due to decreased external wind shear acting on the airflow in S1 [48]. The simulation results reveal that the average air age in single-direction counter-rotation scenarios is marginally higher than in opposite bidirectional rotation scenarios. This difference can be attributed to the rotation of the S3 segment, which causes a significant portion of airflow from S1 to be diverted into S2. When S1 and S3 are aligned with the same rotation angle, airflow from S1 transitions into S3 more seamlessly, resulting in the slightly lower average air age observed in opposite bidirectional rotation scenarios. Simulation results for adjacent bidirectional rotation indicate that as S1 and S2 segments rotate from 75° to 45°, the block’s average air age initially rises but subsequently declines. At α = 45°, the lowest average air age of 262.7773 s is achieved. When S1 and S2 rotate to 60°, a noticeable increase in average air age occurs, indicating a reduction in ventilation efficiency due to diminished inflow to S1 caused by its rotation. When both S1 and S2 rotate to 45°, they act as upwind inflow routes, significantly improving overall block ventilation and reducing the average air age to its lowest value.

3.4. Contributions and Limitations of the Study

The parametric study yields several key findings and actionable recommendations, as summarized below. These insights are highly relevant for designing road networks in high-density urban environments. (1) Road orientation and prevailing wind direction: While block buildings impede airflow, the spaces between them act as potential ventilation corridors. Optimal ventilation efficiency is observed when the block orientation aligns with the prevailing wind direction at β = 0° or exceeds 70°. Conversely, ventilation efficiency is at its lowest when the block forms a 20° angle with the prevailing wind. (2) Road width and ventilation efficiency: The analysis reveals that increasing road width does not uniformly enhance ventilation efficiency within blocks, as both airflow volume and velocity play essential roles in determining ventilation performance [49]. Blocks exhibit the weakest ventilation when road width is less than 14 m. Ventilation efficiency improves for road widths between 14 m and 32 m, further increases between 32 m and 42 m, and peaks when road width exceeds 42 m. The optimal road widths for maximizing block ventilation are 18 m for branch roads, 34 m for secondary roads, and 46 m for arterial roads. (3) Road density and ventilation efficiency: The findings indicate that urban road density significantly impacts overall block ventilation performance. Specifically, increased branch road density negatively affects block ventilation efficiency, whereas higher densities of arterial and secondary road networks enhance ventilation performance across the block. (4) Road intersection forms and ventilation efficiency: The configuration of road intersections significantly affects airflow patterns and trajectories within blocks, thereby influencing ventilation efficiency [50]. The findings indicate that adjacent bidirectional rotation at 45° increases the number of inflow channels, enhancing block-wide ventilation efficiency. This highlights the importance of increasing both the number and width of upwind inflow channels to optimize ventilation. Furthermore, rotating downstream segments of downwind roads facilitates lateral airflow on adjacent roads. In urban road planning, this strategy can direct airflow into transverse stagnant zones, improving air circulation in targeted spaces. Roads should be rotated downstream rather than upstream, with downstream segment orientations optimized at α = 75° to balance overall ventilation and lateral airflow. Introducing non-orthogonal road intersections in high-density urban areas can reduce the likelihood of stagnant zones in critical locations and remains effective in regions with multiple prevailing wind directions.

This study acknowledges certain limitations. The arrangement of buildings significantly influences the wind environment; however, the block’s buildings were simplified to row-column slab-style structures, which, while streamlining the computational process, introduces potential discrepancies in result accuracy. Moreover, the effects of road greenery and other urban facilities on the wind environment were excluded from the analysis, possibly contributing to additional errors. Despite these limitations, the findings effectively illustrate the role of road systems in shaping block wind environments and offer valuable insights for urban road system planning and design.

4. Conclusions

This study utilized Phoenics simulations to examine the effects of road layout patterns on block ventilation efficiency in high-density urban areas. The analysis considered four road layout parameters: road orientation, road width, road density, and road intersection forms. The results demonstrate that road layout patterns influence the overall air renewal rate of blocks by modifying inflow volume, airflow direction, and airflow velocity. The findings also highlight the role of wind direction and the relative orientation of adjacent road segments (e.g., longitudinal vs. transverse flow; upstream vs. downstream) in shaping block ventilation efficiency. Practical recommendations for designing road systems in high-density urban environments are proposed, with an emphasis on addressing airflow stagnation zones to improve overall block ventilation. The study is limited to four specific road layout patterns and employs an idealized row-column building arrangement. Future research should investigate additional road and building layouts. Moreover, to overcome the computational limitations of the current simulation software, more advanced simulation techniques should be employed in subsequent studies.