Article

Assessment of Building Air Quality in Respect of Eight

Different Urban Block Designs Based on CFD Simulations
Ana Perišić, Marko Lazić * and Ines Perišić

Faculty of Technical Sciences, University of Novi Sad, 21000 Novi Sad, Serbia; [email protected] (A.P.);
[email protected] (I.P.)
* Correspondence: [email protected]

Abstract: Different urban block morphologies can greatly influence the air quality inside the
buildings of the block. The model presented in this paper determines the correlation between block
morphology and air quality, and outputs the indoor air quality via computational fluid dynamics
(CFD) simulations. In this study, stagnant air was assumed to have a velocity lower than 0.15 m/s
and considered to be low-quality air in the context of human health. The geometry of the urban
blocks was simplified based on real-life buildings. Doors and windows were not 3D-modeled, and
all the vertical surfaces of the buildings were considered as potential locations for them. Eight of the
highest-frequency wind directions out of sixteen main directions per block were used. Wind
directions and velocities were determined based on the weather data for one location chosen for
testing. The simulation used the Reynolds-averaged Navier–Stokes (RANS) equations with the k-ε
turbulence model. The results were then interpreted through the specific algorithm using 3D
graphic software. The surface of the building envelope was divided into smaller meshes. For each
mesh, the average velocity was calculated and meshes were marked for values below the stagnant
air threshold. The eight results, one from each wind direction, were synthesized into one final result.
The model was tested on eight different urban block morphologies based on real-life blocks, i.e.,
blocks in Novi Sad, Serbia. The pressure on the building surfaces determined via CFD analyses is
presented alongside results from the method described in this paper. The results show that urban
block morphologies with clustered buildings inside the urban block, which are typical for the most
Citation: Perišić, A.; Lazić, M.;
newly built structures in Novi Sad, have areas on the facades where windows cannot provide
Perišić, I. Assessment of Building Air elemental natural ventilation throughout the year. To interpret the results obtained in this research,
Quality in Respect of Eight Different graphs and 3D color-coding models were used. The best results show a 1-tower urban block
Urban Block Designs Based on CFD morphology with only 0.7% of all vertical faces of the model registering a wind velocity lower than
Simulations. Appl. Sci. 2023, 13, 7408. the set minimum. The worst results were measured for a traditional urban block typical in old city
https://doi.org/10.3390/app13137408 centers. A total of 54.5% of all the vertical surfaces show no problems with air stagnation in close
Academic Editor:
proximity to them.
Juan García Rodríguez
Keywords: 3D model visualization; air quality; computational fluid dynamics (CFD); natural
Received: 14 May 2023
ventilation; urban microclimate; urban morphology; urban wind flow
Revised: 20 June 2023
Accepted: 21 June 2023
Published: 22 June 2023

1. Introduction
The process of rapid urbanization in the last few decades has increased the urban
Copyright: © 2023 by the authors.
Licensee MDPI, Basel, Switzerland.
density. In most cases, new buildings and urban blocks are not built with consideration
This article is an open access article for effective ventilation inside them. Air pollutants cannot be diluted quickly enough and
distributed under the terms and this can cause problems for the populations living in these areas. Failing to provide high-
conditions of the Creative Commons quality air can put people at risk of respiratory disease and other health issues [1,2]. Air
Attribution (CC BY) license pollution is worsening in most cities worldwide. More than half of the urban population
(https://creativecommons.org/license lives in places that have more than six-times higher levels of air contamination compared
s/by/4.0/). to the levels recommended in the WHO air quality guidelines [3].

Appl. Sci. 2023, 13, 7408. https://doi.org/10.3390/app13137408 www.mdpi.com/journal/applsci

Appl. Sci. 2023, 13, 7408 2 of 20

Evaluation of the air conditions in an urban environment is paramount for air quality
management. Outdoor air quality directly impacts indoor air quality through natural or
mechanical ventilation. Stagnant outdoor airflow near windows can influence the natural
ventilation capacity inside the building and increase the density of air pollutants. The
airflow that is observed as stagnant does not have to have a velocity value of 0 m/s. Very
slow velocities can have similar effects, and those are common in outdoor urban
environments.
An urban city block’s density and its morphological characteristics are crucial for the
local ventilation performance and pollutant dispersion [4]. Airflow simulations in an
urban environment are usually limited to representing only how wind intensity and
direction influence the outdoor space of the buildings. More research needs to be done in
order to highlight the problematic areas in terms of air quality around individual houses
or apartments. The geometrical characteristics of the block can change, but most parts are
immutable. It is possible to change the type of ventilation in critical areas from natural to
mechanical, even in cases where there is no possibility of removing structures inside the
block. Developing a method for low air-quality area analysis is very important. Methods
used for the detection of critical stagnant air areas have complexity regarding wind
velocities and directions. The issue cannot be solved by simulation of one wind-tunnel
scenario. Multiple wind direction simulations on one urban block or building are well
studied in the literature; however, to our knowledge, there are no published papers with
the goal of presenting the results synthesized from multiple simulations that account for
conditions throughout the year. The method described in our paper addresses this issue.
Assessment of air quality should address all possible wind directions and every velocity
for every day throughout the year. This could be interpreted as a task requiring hundreds
of simulations for each variation. Because this is a time-consuming task, some
simplification is required regarding the number of analyses. To overcome the issues
related to the previously stated problems, we propose a novel method that can identify
areas with the most stagnant air in an urban block environment directly on the 3D mesh.
This method presents the results in a form where stagnant air areas are highlighted from
data that represent the wind conditions from most of the year.
The paper is structured as follows: Section 2 presents related work and the literature
review on the CFD simulations used in the context of urban airflow. In Section 3, the
overview of the framework for the proposed method is outlined. In the same section, the
procedures of the method are described in detail. Section 4 presents the results of the
method described in this paper in respect of 8 blocks with different urban morphologies.
The discussion of the results is included in the same section. Section 5 describes the main
findings, suggestions for further research, and some limitations of the study.

2. Related Work
There are multiple studies on natural ventilation, including theoretical, analytical,
and experimental approaches, and using computational fluid dynamics (CFD). The CFD
method represents a computer simulation method that can analyze air flows using
numerical analysis and data structures. CFD simulation accounts for about 70% of the
scientific research on natural ventilation [5]. CFD analysis is fast and relatively low-cost
compared to other methods, but the results are less accurate. Comparative studies of
different methods describe this method as having a high level of accuracy [6,7] and being
suitable for scientific model experiments. In recent decades, this has been achieved by
developing more accurate turbulence models, increasing the power of computer
hardware, and producing guidelines for best practice [8,9].
One of the aspects of CFD simulation represents the quality of 3D models of urban
environments. The studies in a number of cases simplify the urban block geometry to
cubical form [10–13]. These studies are mostly conducted to obtain results that are general
in nature in order to observe the influence of external conditions on the temperature,
Appl. Sci. 2023, 13, 7408 3 of 20

pressure, turbulence, or velocity in a generic urban environment. Simulations on real-life

urban blocks [14–16], in contrast, are used in order to study urban environments in detail.
It is necessary to test scenarios with different wind directions for natural ventilation
analysis. Teittel et al. [17] experimented with the airflow from four clustered directions.
The building in this study was an open-terrain greenhouse, and the researchers found a
significant effect of the wind direction on the flow patterns inside the house. Norton et al.
[18] performed CFD simulation in similar conditions in an open field. They studied
various situations by changing the wind direction from 0 to 90 degrees in intervals of 10
degrees. The results show a correlation between the air exchange rate from the outdoor to
the indoor space and the wind direction. Horan and Finn [19] studied natural ventilation
using CFD simulation with a model of a simple cubic building. Four different wind
directions were studied, and they concluded that wind direction has a high impact on the
natural ventilation rate. All three examples used only one building for the experiment.
Hooff et al. [15] performed CFD simulations on a stadium building with and without
surrounding buildings. The eight wind directions, represented as all cardinal and ordinal
directions, were used as boundary conditions with one velocity. The study showed that
surrounding structures can influence the airflow results by up to 96% and concluded that
natural ventilation performance should be studied in a larger urban area. This study also
showed that 2 wind directions can have a difference in airflow up to 152%, so the choice of
directions inside the wind tunnel has a high impact on the final results. Research by Wiren
[20] also included surrounding buildings in the CFD simulation. Wiren concluded that wind
pressure distributions and ventilative heat losses were influenced considerably if more than
one building was included in the research. Costola et al. [21], in their review paper, stated
that very little data exist on urban surroundings and the related sheltering effect. This
observation was further supported by Hooff et al. [15], who stated that detailed studies on
the effect of the wind direction and surrounding buildings on indoor natural ventilation in
an urban environment are not commonly performed. In studies of natural ventilation, there
are two main approaches in regard to indoor and outdoor air simulations. The coupled
outdoor and indoor air flows are mixed in CFD simulations. There are multiple studies
using this approach [10,15,22]. The decoupled approach observes indoor and outdoor
spaces separately and performs simulations on different geometries. This approach has an
advantage for larger systems where the potential of natural ventilation is calculated.
Weerasuriya et al. [23] used building information modeling (BIM) to estimate the potential
of natural ventilation of high-rise structures using the decoupled approach. Gan et al. [24]
also used this approach in their research. They conducted CFD simulations on five buildings
with complex shapes to calculate the wind pressure on the exterior surfaces and used data
as input for the deep neural network model. The results predicted air change rates for
different apartments inside the structures. Although the coupled approach can give higher
accuracy results, this approach is harder to implement when multiple windows and doors
can be opened or closed in simulations.

3. Model for Urban Area Air Quality Analysis Based on CFD Simulations
The main focus of this research was to generate a complex analysis containing crucial
data that can easily be interpreted by urban designers. Frameworks such as those
presented by Perisić et al. [25] propose systems in which multiple simulations from
different fields of study, such as wind analysis, illumination, or energy efficiency, can be
performed on an urban block environment. The models of buildings would be uniformly
represented, and the results of the simulations would be read in a way that can identify
critical areas which are not suitable for windows or living spaces. In this paper, buildings
inside the analyzed block are represented without windows and doors. Every vertical
surface of the analyzed geometry is a potential location for a window. This assumption is
made in order to assess the domain that cannot be easily changed by urban developers or
inhabitants. If the area with stagnant air is detected near windows used for the natural
ventilation of living spaces, the new opening can be placed in a more suitable location.
Appl. Sci. 2023, 13, 7408 4 of 20

Excluding windows and doors can change the air flow, but mostly when the circulation is
provided inside of the building. In order to reduce the complexity of the simulations, we
excluded natural ventilation potential from inside the buildings, although we
acknowledge the recent adoption of BIM technologies in the construction industry and
the potential that most of the interior space of urban block buildings could be fully
represented and available in the future. Another problem for urban block natural
ventilation assessment is that simulations can give different results if only certain doors
and windows are opened. The results can be different in scenarios where all the doors and
windows are opened, all are closed, or when combinations of dozens or hundreds of doors
and windows inside the block are opened or closed. For each combination, a new model
must be prepared. Although the results will not reflect the complexity of real conditions
in every aspect, the model presented in this paper reduces the complexity of certain
aspects of geometry. This is done in order to significantly reduce the computational time.
Several studies [26–28] have made a connection between stagnant air outside of
buildings with higher air pollution and human health issues for the inhabitants of an
urban area. In this regard, in this study, we consider air stagnation in the urban block
environment in close proximity to windows as the air of low quality. In this research, air
quality is a parameter that can have either a high or a low value. A low value is considered
for air velocities that do not meet the minimum requirement for air circulation, which is
0.15 m/s. This threshold was chosen based on several studies [29–31]. The number is a
variable that can be changed based on research preferences. The results are a graphical
representation of the areas with stagnant air on surfaces of the 3D model of the analyzed
urban block. The problem is presented in respect of not just one, but eight wind directions
in order to provide results that can illustrate to users more accurate data in respect of
natural ventilation throughout the year. The surfaces representing the walls of buildings
do not have velocity parameters, so in order to represent the results, the assumption is
made that if the area that is 1 m away from a part of a wall has an average velocity lower
than the threshold, that part of the wall is labeled as a stagnant air area.
Our framework consists of the following procedures:
First, the geometry of the 3D models representing the urban environment are
optimized (Section 3.1). Second, the main wind directions are chosen for the simulation
(Section 3.2). Third, simulations in CFD software are performed (Section 3.3). Fourth, the
algorithm is used to interpret the results (Section 3.4). Fifth, visualization of the results are
conducted (Section 3.5). Figure 1 shows the process of the method used to perform the air
quality analysis of an urban block based on CFD simulations.
Appl. Sci. 2023, 13, 7408 5 of 20

Figure 1. Schematic workflow of the study. Step 1: Location data interpretation; Step2: CFD
simulation; Step 3: Synthesis of the results; Step 4: Visualization of the results as color-coded 3D
mesh and a graph.

This method has two main input parameters. The first parameter is the geometry of the
buildings and the second is location. The geometry of the urban block is extracted from the
geospatial data from the chosen location. The graphical outputs of this method visualize
highlighted areas on the 3D mesh representing the problematic area and graph results.

3.1. Geometry Optimization

The geometry of objects in CFD simulations is usually optimized, especially on
models representing urban blocks. It is possible to construct models automatically or
semi-automatically using geospatial data [32,33]. The manual 3D modeling approach is
also possible with consideration of the real object represented. In this study, the manual
approach is used, and it has several constraints:
(a) buildings are modeled individually from a reliable source;
(b) the ground shape of the buildings is extruded;
(c) roof slopes are not modeled, because of complexity;
(d) the smallest edge in the model is 30 cm;
Appl. Sci. 2023, 13, 7408 6 of 20

(e) vegetation is not modeled;

(f) only areas that exceed 150 cm on the facade are modeled;
(g) buildings are merged into one solid object for simulation optimization.
In Figure 2, the real-world block and the block model are presented.

Figure 2. Example of real-world urban block modeling.

The geometrical model was loaded into the CFD software for the analysis. It was also
triangulated for the further analysis described in Section 3.4. The surface of the model was
divided into sub surfaces no more than 3 m by 3 m, and then triangulated.

3.2. Wind Directions

Before the CFD simulations were run, the main wind directions and velocities were
calculated. There are 16 cardinal, ordinal, and intercardinal wind directions, and most
measurements represent their velocities and frequencies. Using all of the wind directions
is possible, but it is very time-consuming. We excluded eight directions from the analysis
to simplify the model and lower the computing time by half. Winds were excluded based
on frequency criteria, which represent the following process: (1) finding relative
frequencies of all wind directions during the year for a given location; (2) sorting the data
based on the highest number of relative frequencies; (3) including the first eight wind
directions, which will provide data for most of the time throughout the year.
In Table 1, an example is shown for the city of Novi Sad, Serbia. The 8 most frequent
directions represent data for 57% of all situations per year. The data also show that, on
average, for 87% of the year no wind is detected at the location. In windless conditions,
the results will show velocities of 0 m/s inside the urban block and poor ventilation
conditions. From the perspective of our research, 65.8% of all wind conditions are
included in this example. For the cities Belgrade, Smederevska Palanka, and Sjenica, the
values would be 76%, 76%, and 80%, respectively. The percentage of relative frequencies
of the excluded wind directions is not insignificant, but it can be a valid way of considering
the setup and computational time for the analyses.

Table 1. Wind direction frequencies and velocities for the city of Novi Sad, Serbia, for the period
1981–2010 [34]. Directions are sorted from lowest to highest relative frequency.

Wind Windles WN
NE ENE ENE NNE SW SSW S WSW SE ESE WSW SSE NW N NNW
Directions s W
relative
frequencies 87 28 35 35 39 47 51 55 55 55 57 59 59 63 66 98 109
(‰)
average
velocities / 2.1 2.2 2.3 2.1 2.1 1.7 2 2.3 2.6 2.7 2.7 2.6 2 3 3.1 3
(m/s)
Appl. Sci. 2023, 13, 7408 7 of 20

The directions and their corresponding velocities are the basic conditions for wind
tunnel simulation in CFD software.

3.3. CFD Simulations

The software used in this study was Autodesk CFD [35], but the method is not limited
to specific software. The Reynolds-averaged Navier–Stokes (RANS) model and the Chen–
Kim k–ε turbulence model [36] were adopted for the computation. These models are used
in most research on fluid dynamics in built environments [4,37,38] for their relative
precision and speed. The finite element method was used for numerically solving
differential equations, and the modified Petrov–Galerkin method was used as the upwind
method. The mesh resolution was set to 0.5 m and the mesh was divided into hexahedral
elements. The material used for the fluid was standard air material and the analysis was
set as steady state. The building and ground material was set as generic concrete material.
The computational domain and boundary conditions were based on the Architectural
Institute of Japan standards and the COST group guidelines [39,40]. In Figure 3, the
domain configuration is illustrated. Parameter H represents the height of the tallest
building of the urban block. The vertical and lateral extent is 10H, except for the side
toward the outlet. The wind speed boundary condition is defined at the inlet. The outlet
is set to uniform zero-gradient conditions. The geometry of the urban block and ground
have a no-slip condition assigned. Possible wind angles were set from 0 to 337.5 degrees
with an increment of 22.5 degrees. Each wind direction corresponds to a specific air
velocity as an input parameter. The turbulence kinetic energy (TKE) tolerance was set to
10−4. Eight RANS simulations were performed for each urban block based on wind
direction and velocity. On average, the mesh was divided into 150 k hexahedral elements,
and the value for parameter H ranged from 10 to 109 m.

Figure 3. Dimensions of the computational domain.

3.4. Algorithm-Based Interpretation of the CFD Analysis

For the interpretation of the results, different 3D software is used. After solving eight
analyses, the results are exported as a point cloud, where each point is represented with
various properties such as temperature, velocity, pressure, and more. Out of these
properties, our method only considers velocity values calculated from x, y, and z vectors.
These point clouds are then imported into McNeel Rhinoceros 3D [41]. The software has
a built-in plug-in named Grasshopper, further enhancing its capacity for analyzing and
representing the results.
Appl. Sci. 2023, 13, 7408 8 of 20

The surface envelope of the urban block is divided into smaller parts. Parts of 3D
mesh marked as suitable for windows of the object are divided into rectangular polygons
and from each of the polygons, a box is created. The base of the box is the polygon itself
and the height of the box is 1 m in front of the polygon. This box represents the potential
for natural ventilation for windows hosted by the observed wall. The results of operations
inside the box are interpreted as results that are assigned to the contact segment of the
mesh. The algorithm for the process is presented in Figure 4.

Figure 4. Algorithm for interpretation of CFD simulation results. The output is a set of binary
numbers correlated to wind directions.

The next step involves the observation of points with velocity information that we
acquired earlier. For each wind direction, collisions are checked for each box and the point
cloud. Points inside the box are considered for further calculation. From their velocities,
the average speed is calculated for each box and for each wind direction. If the value is a
smaller value than 0.15 m/s, then it is marked as 0, and it is marked with 1 in other cases.
It is represented as 8 binary numbers for each mesh segment; these are formulated as a set
Si = [n1, n2, n3, n4, n5, n6, n7, n8] for segment Mk.
If all numbers are 0, the overall performance is 0. If all numbers are 1, the overall
performance is 1. In other cases, the performance is calculated from the wind direction
using the following equations.
εi
φi = 8 (1)
∑i = 1 εi
Appl. Sci. 2023, 13, 7408 9 of 20

R k = ∑ ni × φi (2)
i=1

In Equation (1), φi represents the relative percentage of the wind direction frequency.
It is calculated once per location. A variable represents relative frequencies of wind
directions that are obtained through weather data such as in Table 1. R k is a percentage of
the stagnant air distribution throughout the year for mesh segment M k. A percentage
value of 100 can be translated as perfect airflow through most of the year near the observed
mesh. This result is then assigned to the part of the surface mesh segment connected to
the box.

3.5. Representation of the Results

The results are presented as graphs directly on the 3D models of the urban blocks.
On the models, surfaces representing possible natural ventilation problems are color-
coded in grayscale. A white color on the surface represents which windows in this area
can provide good natural ventilation throughout most of the year, and a black color means
a complete lack of natural ventilation according to the standards discussed in Section 3.4.
The color for mesh segment Mk is calculated using the following equation:
Ck = R k × 255 (3)
where Ck represents the RGB values of a color. The mesh is then presented from two
opposite axonometric views. For the graph view, we divide the scale for the air quality
distribution throughout the year from 0 to 100 by an increment of 5. Results for each mesh
segment are then placed on the scale into groups and the area sums for all groups are
calculated.

4. Results and Discussion

The results for eight individual blocks are presented for the city of Novi Sad. The
same wind conditions were applied to all of the examples shown in this section. Wind
directions and intensities used in this study are from reliable sources [34]. Urban blocks
were carefully chosen to represent different urban morphologies that are typical for the
area of Novi Sad. All urban blocks have a total area of around 10,000 m2 to represent
similar location potential. Some of them were slightly modified to represent similar
external conditions for all the examples. Graphical results for corresponding blocks are
also presented. Ideal graph results for the urban block would have only values of 100 in
respect of the stagnant air area of the building envelope throughout the year. More values
to the right side of the graph represent problems in geometry.
The first morphology represents an urban block comprising one-family residential
buildings. In Figure 5, the results of the method presented in this paper have been applied.
Vertical areas with stagnant air are shown graphically. East- and west-oriented faces of
the structures have results that are worse than north and south faces. This problem is most
visible in the structures in the 3D model that are in close proximity to each other. A
graphical representation of the analysis (Figure 6) shows that this type of block is
moderately ventilated. The curve is set to a y-axis value of 0 near 85% of the x-axis. This
should be interpreted that the most critical areas of this block have 15% air stagnation
situations throughout most of the year.
Appl. Sci. 2023, 13, 7408 10 of 20

Figure 5. Axonometric result of the method described in this paper on one-family residential
building block type. N points to the north relative to the 3D model.

Figure 6. Graphical result of the quality of air in the context of air velocities higher than 0.15 m/s
near surfaces of one-family residential building urban block type.

The second morphology analyzed in this paper is the traditional urban block. This
type of block is typically located in old city centers. Most of the areas inside the block are
not properly ventilated (Figure 7). The close proximity of buildings in the inner courtyard
is the main reason for this. This can result in the deterioration of health conditions in the
housing area. The graphical analysis (Figure 8) shows lower levels of ventilation than for
previously analyzed urban blocks. From all vertical surfaces, 19% of their calculated area
has less than 85% air quality distribution throughout the year. The gradual shape of the
curve suggests a high percentage of the vertical surfaces are not suitable for windows in
regard to natural ventilation. The 0 value on the x-axis is reached at 45%, and if those
surfaces host windows, stagnant air is present for most of the year.

Figure 7. Axonometric result of the method described in this paper on traditional urban block type.
N points to the north relative to the 3D model.
Appl. Sci. 2023, 13, 7408 11 of 20

Figure 8. Graphical result of the quality of air in the context of air velocities higher than 0.15 m/s
near surfaces of traditional urban block type.

The third urban morphology that was analyzed is an urban block with multiple
towers, typically built in the second half of the 20th century in Europe. From the results of
the analysis in Figure 9, only small parts of the surfaces have problems with ventilation.
A low build density has an influence on wind intensity in the whole area. There are
surfaces with lower ventilation, but they are created because of the turbulence effect. A
graphical representation of the results (Figure 10) shows that the air quality distribution
throughout the year is very good. Only 13% of all vertical surfaces have values of air
quality distribution that are below 95% throughout the year. The highest value on the y-
axis suggests that the ideal area for window placement is much higher than that in
previous urban blocks.

Figure 9. Axonometric result of the method described in this paper on multiple-tower urban block
type. N points to the north relative to the 3D model.

Figure 10. Graphical result of the quality of air in the context of air velocities higher than 0.15 m/s
near surfaces of multiple-tower urban block type.
Appl. Sci. 2023, 13, 7408 12 of 20

The fourth urban morphology is a slab city block type. This is similar to the previous
morphology, but with a lower building height and increased land cover (Figure 11). The
buildings are freestanding, and their disposition influences the airflow direction inside
the urban block. Through strategic urban planning, it is possible to create better air quality
distribution in the entire area of the block. The most surfaces that correspond to air
stagnation problems are at the ground floor level. Figure 12 shows good results for air
quality distribution throughout the year, similar to the results from the previous example.
The area between 70% and 80% on the x-axis is more prominent, which can be seen in
Figure 11 where darker shades of gray are present. The main reason for this result is the
disposition of the blocks and their correlation to the most frequent wind directions for the
location.

Figure 11. Axonometric result of the method described in this paper on slab city block type. N points
to the north relative to the 3D model.

Figure 12. Graphical result of the quality of air in the context of air velocities higher than 0.15 m/s
near surfaces of slab city block type.

The fifth example is a U-shaped city block. The position of the open side of the block
in the context of important wind directions can give different results. The real-life urban
block that was used as an input for the geometry is orientated towards two opposite
directions. In this example (Figure 13), the wind flow is well balanced from all directions
and the results of the analysis are good. Areas with less than 75% air distribution quality
are not detected (Figure 14). The curve of the graph Is relatively smooth in contrast to our
expectations. The highest number on the y-axis is around 8000; this is significantly lower
than that for the multiple tower block, which is similar in volume to the U-shaped block.
Appl. Sci. 2023, 13, 7408 13 of 20

Figure 13. Axonometric result of the method described in this paper on U-shaped block type. N
points to the north relative to the 3D model.

Figure 14. Graphical result of the quality of air in the context of air velocities higher than 0.15 m/s
near surfaces of U-shaped urban block type.

The sixth urban morphology is the densely populated perimeter city block. This
block is typical of the urban blocks that developed from single-family housing to collective
buildings in recent decades in Novi Sad. The results (Figures 15 and 16) show that most
of the problems detected are inside the urban block. The graph shows that area from 70%
to 80% air quality distribution is prominent, compared to most other examples. These
values are representations of the area that is not suitable for window placement for natural
ventilation of the building. The graph curve gradually reaches 0 between 40% and 50%,
suggesting that inside of the block there are surfaces near which there is stagnant air
present for most of the year.

Figure 15. Axonometric result of the method described in this paper on densely populated perimeter
city block type. N points to the north relative to the 3D model.
Appl. Sci. 2023, 13, 7408 14 of 20

Figure 16. Graphical result of the quality of air in the context of air velocities higher than 0.15 m/s
near surfaces of densely populated perimeter city block type.

A one-tower urban block is the seventh example of urban block geometry. The height
of the building is 109 m, and the area around the object is flat. This type of block is
historically usually built so natural ventilation and daylight illumination are provided for
all of its inhabitants. Less than 0.7% of all surfaces have values below 95%. The results are
presented in Figures 17 and 18. The graph curve is steepest for all the examples, presenting
almost perfect results in regard to natural ventilation potential. This result is predictable
considering the surrounding of the one-block building.

Figure 17. Axonometric result of the method described in this paper on one-tower urban block type.
N points to the north relative to the 3D model.

Figure 18. Graphical result of the quality of air in the context of air velocities higher than 0.15 m/s
near surfaces of one-tower urban block type.
Appl. Sci. 2023, 13, 7408 15 of 20

The eighth example represents a hybrid urban block. This urban block type does not
have a specific form and can have higher variety compared to previously examined
blocks. The wind flow is slower in the east–west direction, as shown in Figure 19. North-
facing surfaces and surfaces in the west are locations that have lower values of air velocity.
The values are similar to results of the U-shaped block, but with more problematic areas
for values under 75% air quality distribution. For those areas, the hybrid block has 1.1%
of all surfaces compared to 0% for the U-shaped block. The graphical results are presented
in Figure 20. The graph curve shows variation around 50% and 65% on the x-axis. This
shows small areas with stagnant air in close proximity to the building surface.

Figure 19. Axonometric result of the method described in this paper on hybrid urban block type. N
points to the north relative to the 3D model.

Figure 20. Graphical result of the quality of air in the context of air velocities higher than 0.15 m/s
near surfaces of hybrid urban block type.

The results can be seen comparatively between different urban morphologies. In

Table 2, the results of all the examples are presented. It can be concluded that different
morphologies produce different measurable conditions in regard to air quality with the
method described in this paper. Urban blocks are modeled without surrounding blocks
and trees, and thus, the maximal potential for urban blocks can be observed through this
study. In cases that represent a more detailed approach, less favorable results can be
expected.
Appl. Sci. 2023, 13, 7408 16 of 20

Table 2. Air quality distribution throughout the year for eight different urban block types based on
the method described in this study.

Block Vertical Air Quality Area on Building

Site Maximum
Type Volume Surface Envelope throughout the Year (%)
Coverage Height (m)
(m3) Area (m2) >95 85–95 75–85 65–75 55–65 <55
One-family
residential 28,374 33.76% 11 10,075 Area (m2) 7850 1904 230 76 15 0
building block
Percent
77.92% 18.90% 2.28% 0.75% 0.15% 0.00%
of total
Traditional
block for city 55,169 71.04% 10.5 5736 Area (m2) 3123 1516 632 227 151 87
center
Percent
54.45% 26.43% 11.02% 3.96% 2.63% 1.52%
of total
Multiple-tower
55,330 10.35% 52 15,955 Area (m2) 13,886 1620 302 141 6 0
block
Percent
87.03% 10.15% 1.89% 0.88% 0.04% 0.00%
of total
Slab city block 56,350 20.26% 31 12,588 Area (m2) 10,247 1355 669 223 94 0
Percent
81.40% 10.76% 5.31% 1.77% 0.75% 0.00%
of total
U-shaped city
54,256 29.28% 19 11,890 Area (m2) 10,296 1339 255 0 0 0
block
Percent
86.59% 11.26% 2.14% 0.00% 0% 0%
of total
Perimeter dense
103,372 56.96% 23 17,921 Area (m2) 12,417 2799 1666 429 310 300
block
Percent
69.29% 15.62% 9.30% 2.39% 1.73% 1.67%
of total
One-tower
105,522 9.33% 109 14,408 Area (m2) 14,307 101 0 0 0 0
block
Percent
99.30% 0.70% 0% 0% 0% 0%
of total
Hybrid block 107,015 43.81% 42 15,143 Area (m2) 13,285 1361 331 120 46 0
Percent
87.73% 8.99% 2.19% 0.79% 0.30% 0.00%
of total

The highest-ranking results obtained from the tower urban block morphology,
measuring 99.3% of all of its vertical surfaces, have the uppermost values for air quality
distribution throughout the year. U-shaped blocks have no surfaces below 70% air quality.
The lowest ranking is for the traditional urban block; the distances between buildings are
short and very densely distributed. Only 54.45% of surfaces have adequate wind
conditions for natural ventilation. Most of them are on the outer facades of the block. The
recommendation for most inner areas is to install mechanical ventilation systems. Results
for the perimeter block also show only 69% of surfaces with the highest rating. Traditional
and perimeter blocks are the only ones that contain surfaces with a rating of less than 55%.
Considering that the perimeter block has double the volume of the traditional block, more
people living in it can be affected.
For comparison, the results of the static pressure from CFD simulations were
exported. Low absolute values of pressure under 0.3 Pa are colored in black, and higher
values, up to 11 Pa in some cases, are colored in white. The value of 0.3 Pa does not
Appl. Sci. 2023, 13, 7408 17 of 20

represent the threshold for experiments, but the measure for which low pressure results
can easily be seen. The results are shaded according to the wind frequencies, similar to
Equations (1) and (2). The results are presented with the ground plane and can be seen in
Figure 21. Several similarities between the results of velocities and pressure levels can be
observed. Bad air flow is detected in the courtyards of the first and second urban blocks.
The perimeter dense block and hybrid block show results that suggest slower movement
of air around them. The best results are show for the one-building block, as in Figure 17.

Figure 21. Static pressure measured for absolute values below 0.3 Pa. The illustrations show yearly
averages similar to the principle used for velocities described in Sections 3.2 and 3.4. Urban blocks
labeled with letters: (a) one-family residential buildings; (b) traditional; (c) multiple towers; (d) slab
city block; (e) U-shaped; (f) perimeter dense block; (g) one tower; (h) hybrid block.
Appl. Sci. 2023, 13, 7408 18 of 20

5. Conclusions
In this paper, a novel method for natural ventilation capacity assessment in urban
areas was presented using average yearly data for multiple wind directions. This enabled
automation and color-coding of the surfaces areas of urban blocks that have problems
with ventilation and visualization on a single 3D model for most of the annual data. To
our knowledge, this problem has not been considered by other researchers. The proposed
method consists of a four-step process: location data interpretation, CFD simulation,
synthesis of the results, and visualization. Location data were used for building creation
and wind direction and velocity data interpretation. Buildings were 3D modeled without
a roof. Simulations were conducted for eight main wind directions for the location. These
simulations were synthesized by an algorithm that labeled stagnant air areas near the
surfaces of the buildings. Results were represented in 3D form and as graphs.
The method was used for the analysis of eight different urban building morphologies
from the city of Novi Sad, Serbia. The weather data were obtained from reliable sources.
The results show that natural ventilation can be influenced significantly by urban
morphology. Urban block types such as the one-tower block type showed the best results,
and the worst results were obtained for traditional urban blocks from old city centers. The
best results show that all of the vertical surfaces of the block are suitable for windows if
the object is naturally ventilated.
The findings from this study make an important contribution to the field of CFD
simulation of urban environments, providing solutions for natural ventilation analysis of
urban blocks by analyzing multiple wind direction conditions. The results are represented
in an easy-to-read format for urban block evaluation.
Our procedure can be augmented in a number of ways. First, the surrounding blocks
can have a great impact on the results of CFD simulations. Our tests were conducted in
order to compare the results. For studying real-world urban blocks, the surrounding
geometry is necessary for better results. Second, trees were not part of the 3D model, and
their influence on wind flow is well-established in the literature. Third, different
turbulence models could be used for the analysis. Large-eddy simulation (LES) is
considered a higher accuracy model than RANS, which was used in this study. Further,
the system can be directly implemented in software solutions where most of the processes
would be automated. This could help experts from the field of urban design in making
better decisions.

Author Contributions: Conceptualization, A.P. and M.L.; methodology, A.P. and I.P.; software,
M.L. and A.P.; validation, M.L.; investigation, resources, data curation, and visualization, M.L. and
A.P.; writing—original draft preparation, M.L.; writing—review and editing, A.P. and I.P. All
authors have read and agreed to the published version of the manuscript.
Funding: This research received no external funding.
Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Data Availability Statement: Our algorithm and data can be publicly accessed [42].
Conflicts of Interest: The authors declare no conflicts of interest.

References
1. Brunekreef, B.; Holgate, S.T. Air Pollution and Health. Lancet 2002, 360, 1233–1242. https://doi.org/10.1016/S0140-6736(02)11274-
8.
2. Raaschou-Nielsen, O.; Andersen, Z.J.; Beelen, R.; Samoli, E.; Stafoggia, M.; Weinmayr, G.; Hoffmann, B.; Fischer, P.;
Nieuwenhuijsen, M.J.; Brunekreef, B.; et al. Air Pollution and Lung Cancer Incidence in 17 European Cohorts: Prospective
Analyses from the European Study of Cohorts for Air Pollution Effects (ESCAPE). Lancet Oncol. 2013, 14, 813–822.
https://doi.org/10.1016/S1470-2045(13)70279-1.
3. World Health Organization. WHO Global Air Quality Guidelines: Particulate Matter (PM2.5 and PM10), Ozone, Nitrogen Dioxide,
Sulfur Dioxide and Carbon Monoxide; World Health Organization: Bonn, Germany, 2021; ISBN 978-92-4-003422-8.
Appl. Sci. 2023, 13, 7408 19 of 20

4. Peng, Y.; Gao, Z.; Buccolieri, R.; Ding, W. An Investigation of the Quantitative Correlation between Urban Morphology
Parameters and Outdoor Ventilation Efficiency Indices. Atmosphere 2019, 10, 33. https://doi.org/10.3390/atmos10010033.
5. Chen, Q. Ventilation Performance Prediction for Buildings: A Method Overview and Recent Applications. Build. Environ. 2009,
44, 848–858. https://doi.org/10.1016/j.buildenv.2008.05.025.
6. Blocken, B.; Stathopoulos, T.; van Beeck, J.P.A.J. Pedestrian-Level Wind Conditions around Buildings: Review of Wind-Tunnel
and CFD Techniques and Their Accuracy for Wind Comfort Assessment. Build. Environ. 2016, 100, 50–81.
https://doi.org/10.1016/j.buildenv.2016.02.004.
7. van Hooff, T.; Blocken, B.; Tominaga, Y. On the Accuracy of CFD Simulations of Cross-Ventilation Flows for a Generic Isolated
Building: Comparison of RANS, LES and Experiments. Build. Environ. 2017, 114, 148–165.
https://doi.org/10.1016/j.buildenv.2016.12.019.
8. Franke, J.; Hellsten, A.; Schlünzen, K.H.; Carissimo, B. Best Practice Guideline for the CFD Simulation of Flows in the Urban
Environment—a Summary. In Proceedings of the 11th Conference on Harmonisation within Atmospheric Dispersion Modelling
for Regulatory Purposes, Cambridge, UK, 2–5 July 2007.
9. Blocken, B. Computational Fluid Dynamics for Urban Physics: Importance, Scales, Possibilities, Limitations and Ten Tips and
Tricks towards Accurate and Reliable Simulations. Build. Environ. 2015, 91, 219–245.
https://doi.org/10.1016/j.buildenv.2015.02.015.
10. Ramponi, R.; Blocken, B.; de Coo, L.B.; Janssen, W.D. CFD Simulation of Outdoor Ventilation of Generic Urban Configurations
with Different Urban Densities and Equal and Unequal Street Widths. Build. Environ. 2015, 92, 152–166.
https://doi.org/10.1016/j.buildenv.2015.04.018.
11. Sanaieian, H.; Tenpierik, M.; Van Den Linden, K.; Seraj, F.M.; Shemrani, S.M.M. Review of the Impact of Urban Block Form on
Thermal Performance, Solar Access and Ventilation. Renew. Sustain. Energy Rev. 2014, 38, 551–560.
https://doi.org/10.1016/j.rser.2014.06.007.
12. Shirzadi, M.; Naghashzadegan, M.; Mirzaei, P.A. Improving the CFD Modelling of Cross-Ventilation in Highly-Packed Urban
Areas. Sustain. Cities Soc. 2018, 37, 451–465. https://doi.org/10.1016/j.scs.2017.11.020.
13. Hang, J.; Luo, Z.; Sandberg, M.; Gong, J. Natural Ventilation Assessment in Typical Open and Semi-Open Urban Environments
under Various Wind Directions. Build. Environ. 2013, 70, 318–333. https://doi.org/10.1016/j.buildenv.2013.09.002.
14. Antoniou, N.; Montazeri, H.; Wigo, H.; Neophytou, M.K.-A.; Blocken, B.; Sandberg, M. CFD and Wind-Tunnel Analysis of
Outdoor Ventilation in a Real Compact Heterogeneous Urban Area: Evaluation Using “Air Delay”. Build. Environ. 2017, 126,
355–372. https://doi.org/10.1016/j.buildenv.2017.10.013.
15. Van Hooff, T.; Blocken, B. On the Effect of Wind Direction and Urban Surroundings on Natural Ventilation of a Large Semi-
Enclosed Stadium. Comput. Fluids 2010, 39, 1146–1155. https://doi.org/10.1016/j.compfluid.2010.02.004.
16. Hu, Y.; Peng, Y.; Gao, Z.; Xu, F. Application of CFD Plug-Ins Integrated into Urban and Building Design Platforms for
Performance Simulations: A Literature Review. Front. Archit. Res. 2023, 12, 148–174. https://doi.org/10.1016/j.foar.2022.06.005.
17. Teitel, M.; Ziskind, G.; Liran, O.; Dubovsky, V.; Letan, R. Effect of Wind Direction on Greenhouse Ventilation Rate, Airflow
Patterns and Temperature Distributions. Biosyst. Eng. 2008, 101, 351–369. https://doi.org/10.1016/j.biosystemseng.2008.09.004.
18. Norton, T.; Sun, D.-W.; Grant, J.; Fallon, R.; Dodd, V. Applications of Computational Fluid Dynamics (CFD) in the Modelling
and Design of Ventilation Systems in the Agricultural Industry: A Review. Bioresour. Technol. 2007, 98, 2386–2414.
https://doi.org/10.1016/j.biortech.2006.11.025.
19. Horan, J.M.; Finn, D.P. Sensitivity of Air Change Rates in a Naturally Ventilated Atrium Space Subject to Variations in External
Wind Speed and Direction. Energy Build. 2008, 40, 1577–1585. https://doi.org/10.1016/j.enbuild.2008.02.013.
20. Wiren, B.G. Effects of Surrounding Buildings on Wind Pressure Distributions and Ventilative Heat Losses for a Single-Family
House. J. Wind Eng. Ind. Aerodyn. 1983, 15, 15–26. https://doi.org/10.1016/0167-6105(83)90173-3.
21. Costola, D.; Blocken, B.; Hensen, J.L.M. Overview of Pressure Coefficient Data in Building Energy Simulation and Airflow
Network Programs. Build. Environ. 2009, 44, 2027–2036. https://doi.org/10.1016/j.buildenv.2009.02.006.
22. Zhang, X.; Weerasuriya, A.U.; Tse, K.T. CFD Simulation of Natural Ventilation of a Generic Building in Various Incident Wind
Directions: Comparison of Turbulence Modelling, Evaluation Methods, and Ventilation Mechanisms. Energy Build. 2020, 229,
110516. https://doi.org/10.1016/j.enbuild.2020.110516.
23. Weerasuriya, A.U.; Zhang, X.; Gan, V.J.L.; Tan, Y. A Holistic Framework to Utilize Natural Ventilation to Optimize Energy
Performance of Residential High-Rise Buildings. Build. Environ. 2019, 153, 218–232.
https://doi.org/10.1016/j.buildenv.2019.02.027.
24. Gan, V.J.; Wang, B.; Chan, C.M.; Weerasuriya, A.U.; Cheng, J.C. Physics-Based, Data-Driven Approach for Predicting Natural
Ventilation of Residential High-Rise Buildings. In Proceedings of the Building Simulation; Springer: Berlin/Heidelberg, Germany,
2022; Volume 15, pp. 129–148. https://doi.org/10.1007/s12273-021-0784-9.
25. Perisic, A.; Lazic, M.; Perisic, B. The Extensible Orchestration Framework Approach to Collaborative Design in Architectural,
Urban and Construction Engineering. Autom. Constr. 2016, 71, 210–225. https://doi.org/10.1016/j.autcon.2016.08.005.
26. Norris, G.; Larson, T.; Koenig, J.; Claiborn, C.; Sheppard, L.; Finn, D. Asthma Aggravation, Combustion, and Stagnant Air.
Thorax 2000, 55, 466–470. https://doi.org/10.1136/thorax.55.6.466.
27. Kerr, G.H.; Waugh, D.W. Connections between Summer Air Pollution and Stagnation. Environ. Res. Lett. 2018, 13, 084001.
https://doi.org/10.1088/1748-9326/aad2e2.
Appl. Sci. 2023, 13, 7408 20 of 20

28. Lai, L.-W. Public Health Risks of Prolonged Fine Particle Events Associated with Stagnation and Air Quality Index Based on
Fine Particle Matter with a Diameter <2.5 Μm in the Kaoping Region of Taiwan. Int. J. Biometeorol. 2016, 60, 1907–1917.
https://doi.org/10.1007/s00484-016-1177-0.
29. Basford, A.; Spearpoint, M. Defining a Velocity Criterion for Stagnant Air in Buildings. Int. Fire Prof. J. 2021, 36, 32–37; ISSN:
2050-8638.
30. Gong, N.; Tham, K.; Melikov, A.; Wyon, D.; Sekhar, C.; Cheong, K.W.D. The Acceptable Air Velocity Range for Local Air
Movement in The Tropics. Hvacr Res. 2006, 12, 1065–1076. https://doi.org/10.1080/10789669.2006.10391451.
31. Kato, S.; Hiyama, K. Ventilating Cities: Air-Flow Criteria for Healthy and Comfortable Urban Living; Springer Science & Business
Media: Dordrecht, Netherlands, 2012; ISBN 978-94-007-2771-7, doi: 10.1007/978-94-007-2771-7.
32. Deininger, M.E.; von der Grün, M.; Piepereit, R.; Schneider, S.; Santhanavanich, T.; Coors, V.; Voß, U. A Continuous, Semi-
Automated Workflow: From 3D City Models with Geometric Optimization and CFD Simulations to Visualization of Wind in
an Urban Environment. ISPRS Int. J. Geo-Inf. 2020, 9, 657. https://doi.org/10.3390/ijgi9110657.
33. Piepereit, R.; Deininger, M.; Kada, M.; Pries, M.; Voß, U. Sweep-Plane Algorithm for the Simplification of 3D Building Models
in the Application Scenario of Wind SimulationsS. Int. Arch. Photogramm. Remote Sens. Spat. Inf. Sci. 2018, XLII-4-W10, 151–156.
https://doi.org/10.5194/isprs-archives-XLII-4-W10-151-2018.
34. RHMS of Serbia—Republic Hydrometeorological Service of Serbia. Available online:
https://www.hidmet.gov.rs/latin/meteorologija/stanica_sr.php?moss_id=13168 (accessed on 19 April 2023).
35. Autodesk CFD—Computational Fluid Dynamics Simulation Software. Available online:
https://www.autodesk.com/products/cfd/overview (accessed on 19 April 2023).
36. Chen, Y.-S.; Kim, S.-W. Computation of Turbulent Flows Using an Extended K-Epsilon Turbulence Closure Model. NASA
STI/Recon Tech. Rep. N. 1987, 88, 11969.
37. Badach, J.; Wojnowski, W.; Gębicki, J. Spatial Aspects of Urban Air Quality Management: Estimating the Impact of Micro-Scale
Urban Form on Pollution Dispersion. Comput. Environ. Urban Syst. 2023, 99, 101890.
https://doi.org/10.1016/j.compenvurbsys.2022.101890.
38. Chen, L.; Hang, J.; Sandberg, M.; Claesson, L.; Di Sabatino, S.; Wigo, H. The Impacts of Building Height Variations and Building
Packing Densities on Flow Adjustment and City Breathability in Idealized Urban Models. Build. Environ. 2017, 118, 344–361.
https://doi.org/10.1016/j.buildenv.2017.03.042.
39. Tominaga, Y.; Mochida, A.; Yoshie, R.; Kataoka, H.; Nozu, T.; Yoshikawa, M.; Shirasawa, T. AIJ Guidelines for Practical
Applications of CFD to Pedestrian Wind Environment around Buildings. J. Wind Eng. Ind. Aerodyn. 2008, 96, 1749–1761.
https://doi.org/10.1016/j.jweia.2008.02.058.
40. Mochida, A.; Tominaga, Y.; Murakami, S.; Yoshie, R.; Ishihara, T.; Ooka, R. Comparison of Various Ke Models and DSM Applied
to Flow around a Highrise Building-Report on AU Cooperative Project for CFD Prediction of Wind Environment. Wind Struct.
2002, 5, 227–244. https://doi.org/10.12989/was.2002.5.2_3_4.227.
41. McNeel Rhinoceros 3D. Available online: https://www.rhino3d.com/ (accessed on 28 April 2023).
42. AQurbanblock assessment tool. 2023. Available online:
https://drive.google.com/drive/folders/1PaUwuEDHBZYyLJ2uEXDe3lywGiEs7AE6?usp=sharing (accessed on 22 Jun 2023).

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury
to people or property resulting from any ideas, methods, instructions or products referred to in the content.