Natural ventilation in a traditional city: an exploratory computational study of Bushehr-Iran

ABSTRACT This exploratory research discusses natural ventilation in a traditional city through employing a computational method. The mechanism of airflow in traditional cities has not yet been sufficiently studied through experimental and computational methods. This paper reports work on simulating natural ventilation in traditional Bushehr City by means of computational models of airflow in the study zone under three design scenarios. Simulations show that the unstructured plan of the city is less likely to improve air circulation and velocity ratios. But the main corridor with a relatively wider width and straight direction is more likely to contribute to urban ventilation. Further use of evidence-based methods to test climate-responsive strategies of traditional cities may be helpful.


Simulation of natural ventilation in traditional cities
This study applies a computerised computational approach to simulate and explore the mechanism of natural ventilation in a traditional city. A digital simulation of natural ventilation in traditional cities could clarify the efficiency and shortcomings of the vernacular and bioclimatic strategies applied in architectural and urban design. Traditional urban forms have been regarded as successful vernacular responses by their design to the need to adapt to climate [1][2][3][4]. This paper examines the traditional city of Bushehr in Iran in regard to the local hot and humid climate. Bushehr City is located in the southwest of Iran [28.9° N, 50.8° E] along the Persian Gulf [5]. The main bioclimatic vernacular strategies in Bushehr City are listed as the following [6][7][8][9][10][11][12][13].
(1) Compact urban fabrics with high wind permeability and several breezeways and air paths, (2) narrow pathways, i.e. the height-to-width ratio of about 5 or above, surrounding multiple sides of buildings and directed towards the Persian Gulf (3) porous building facades having several openings on the exterior and interior skins on the first floor and above (4) courtyards with height-to-width ratios of about 2 and above surrounded most often by interior balconies projected from the first floor and above, i.e. also known as Shenashir (5) some buildings with exterior balconies (Shenashir) made of the wood awning and louvres projected most often from the first floor and above Further experiments and simulations are, however, required to provide comprehensive, evidence-based evaluations of such bioclimatic vernacular strategies characterised in traditional cities, particularly in Iran. There appears to be little discussion of deficiencies in traditional urban design strategies. In regard to Bushehr City, the literature presents several accounts of effective bioclimatic vernacular strategies based on abstract, descriptive diagrams, sketches, and writing [6][7][8][9][10][11][12][13]. Few studies have, thus far, employed numerical and computational methods to evaluate the performance and efficiency of bioclimatic vernacular strategies, such as ventilation in the traditional city of Bushehr.
Urban forms and building interior configurations act as an integrated system in modifying ventilation and air circulation inside public and private spaces, such as streets, alleys, courtyards, and indoors. More specifically, urban forms affect the airflow condition and circulation in the canopy and boundary layers [14]. Building forms and interior configurations could also change the air regime and circulation. Improving natural ventilation, air corridors, and wind speed can promote convective heat exchange, disperse air pollution, and reduce air temperature [15][16][17]. Such effects can reduce the negative impact of urban heat island (UHI) phenomena, i.e. the higher temperature of metropolitan and dense urban areas compared to their suburbs [18][19][20]. Air velocity could compensate for the negative effect of high temperature and humidity on people living in a hot climate [21][22][23]. The air velocity of about 1 m=s ð Þ to 2 m=s ð Þ at an ambient temperature of about 24 � to 32 � C could provide an acceptable thermal comfort zone [24]. Modelling urban ventilation, especially integrated with indoors, is complex [25][26][27]. Previous research on Bushehr City ventilation has mainly focused on the urban form with no regard to interior building configurations and courtyards. Integrated impacts of urban form, courtyards, and building interior spaces on ventilation and air circulation in Bushehr City have not yet been comprehensively simulated and evaluated.

General objective and structure of the paper
The aim is to promote an evidence-based understanding of natural ventilation mechanisms in traditional cities through using computational methods and digital simulation. To this end, three design scenarios are hypothesised to explore the impact of urban forms, courtyards, and interior building configurations on ventilation and air circulation in traditional Bushehr City as a case study in Iran. Section 2 describes the methodological framework including the characteristics of the studied scenarios. This exploratory architectural study uses simplified two-dimensional (2D) computational fluid dynamics (CFD) modelling to evaluate the overall validity and importance of hypotheses related to each design scenario. The supplementary online material provides further details. Section 3 reports the results of the simulation in terms of ventilation and air circulation in urban pathways, courtyards, and interiors under the three design scenarios. Section 4 discusses the potential and shortcoming of using computational methods to evaluate natural ventilation as well as vernacular strategies in traditional cities. The Conclusion reports major outcomes such as developing a robust three-dimensional (3D) CFD modelling as well as simulating all environmental factors affecting thermal, lighting, and acoustic comforts.

Study zone and locations
A study zone with three design scenarios is specified to simulate the mechanism of ventilation and explore the impact of urban form and indoor configurations in the traditional city of Bushehr. Figure 1(a) shows that the study zone is adjusted to represent the original traditional urban fabrics based on the Urban Renewal and Rehabilitation Action Plan released by Bushehr Municipality [28]. The overall characteristic of the study zone is almost similar to the original traditional configuration. The main corridor, however, includes some major interventions implemented in recent decades which shifted the original traditional configuration by creating a wider width and straight direction compared to pathways within the zone. Table 1 displays the permeability ratio of each scenario, representing the area of open spaces to the total building footprints or solid surface in the zone. Further characteristics of the exploratory design scenarios are as follows.
(1) Scenario One aims at exploring the impact of urban forms on airflow and circulation without considering courtyards and building interior configurations ( Figure 1(b)). The paths and alleys are considered wind corridors which are completely discounted from courtyards and indoors. Air can flow through the passages enclosed by building façades and walls. Buildings are assumed to be a solid mass. All building openings and windows outdoors are also assumed to be closed. Scenario one has a permeability ratio of 17.9% which represents the percentage of open space areas available for airflow (Table 1).
(2) Scenario Two aims at exploring the impact of the exterior courtyards integrated with pathways and alleys on urban ventilation and air circulation (Figure 1(c)). Buildings are still assumed to be a solid mass and have no openings and penetration to the outside. Exterior courtyards refer to open spaces surrounded by exterior building facades and walls. Interior courtyards are enclosed by interior facades. Interior courtyards are considered to be part of the building mass. Exterior balconies (Shenashirs) are also not considered projected from buildings. The permeability ratio of scenario two is 25.4% representing the potential area for air circulation (Table 1).
(3) Scenario Three aims at exploring the integrated impact of urban forms, exterior/ interior courtyards, and building interior design on city ventilation and air circulation ( Figure 1(d)). Because of the limited computational power and computational capacity, only buildings 1 and 2 are modelled for integrated urban ventilation evaluations. All windows and openings of the buildings are considered to be open towards exteriors or interiors. The interior spaces of the buildings as well as the interior courtyards are hypothetically included in urban ventilation and air circulation through exterior courtyards and pathways. Interior and exterior Shenashirs are also assumed not to be projected from façades. Figure 1(d) shows that urban fabrics and interior building configuration, hence, offer an integrated layout with possibilities for natural ventilation and air circulation. The permeability ratio of scenario three is about 30% corresponding to the potential area for natural ventilation.
Wind velocity ratios in the main corridor and six pathways are calculated to evaluate the impacts of different design scenarios on air circulation in the study zone. In contrast to the main corridor, pathways 1-6 have narrower widths and non-straight overall directions, i.e. related to several random changes along their paths. Figure 1 shows that the main corridor and pathway 1 have no direct connections with buildings 1, 2, and 3 in locations A and B marked in the study zone. Pathways 2-4 surround building 1 in location A. Pathways 3-5 surround buildings 1 and 2 in location A of the study zone. Pathway 3 starts from location A and passes through location B. Airflow in pathways 2-4 is presumed affected by the interiors and ventilation of building 1. Air circulation in pathways 3-6 is hypothetically impacted by interior configurations and ventilation of buildings 1 and 2. Pathway 3 is, hence, assumed to be affected by buildings in both locations A and B. The wind velocity ratio represents the ratio of air velocity in a spot in the study zone to the upstream inlet air velocity, i.e. 3.8 m/s.

Computational modelling of airflow
The researchers have developed 2D CFD models to simulate natural ventilation throughout the study scenarios. The CFD settings are validated by the experimental study published by Tominaga et al. [29] from the Architectural Institute of Japan (AIJ). ANSYS Fluent was used to develop the CFD simulation. The supplementary online material provides the detail of CFD simulation as well as calibration processes. The scenarios are assumed to be located at the height of 6.5 m above the ground to represent the first floor of traditional buildings in Bushehr City which are commonly designed for bedrooms, and living, dining, and guest rooms, especially for use during summers.
The metrological data of Bushehr City show that the wind speed at the height of 10 m above sea level fluctuates between 3 m/s and 5 m/s throughout the year [30]. As further explained in the supplementary online material, wind velocity at the target height of 6.5 m above the ground is calculated to be about 3.8 m/s using the power-law equation and the mean speed of 4 m/s at 10 m height. Note that, as an exploratory architectural study, this research developed 2D simulations to evaluate the overall validity of the hypotheses and to give a general view of the computational assessment approach to the traditional urban and architectural design. Developing a robust 3D computation simulation requires further studies, computer power, and interdisciplinary collaborations, especially the contributions of wind engineering and CFD specialists.

Air circulation in the studied traditional city
The computational modelling simulates air circulation and velocity representing natural ventilation in the traditional City of Bushehr. Figure 2 shows velocity contours, representing air circulation patterns in the study zone under scenarios 1 to 3. As can be seen, pathways in the direction of the upstream wind allow the air to penetrate the study zone under all scenarios. The main corridor has a relatively wider width and straight direction, because of contemporary interventions. This enables a higher flow rate and penetration compared to narrower pathways within the zone. The air circulation pattern and velocity contours in the main corridor are almost identical for all study scenarios. Pathways 1-6 inside the study zone, however, establish no effective wind corridor with a dominant airflow direction. Multiple air stagnation zones are created inside the study zone and along the edges and several corners of the main corridor and pathways. The exploratory scenarios 2 and 3 show that exterior/interior courtyards and the indoor configuration of buildings modify air flow rates and circulation patterns in the study zone. Air circulation in pathways 1-6 under scenario 1 is noticeably altered in scenarios 2 and 3, creating some spots with flow rates of nearly zero. Figure 3 enlarges Locations A and B encompassing buildings 1 to 3 to clarify the impact of courtyards and indoors on airflow patterns. Spots 1-8 identify the changes that occurred in air circulation in both locations under three design scenarios. In regard to the marked spots, courtyards and indoor configurations considerably shifted air circulation in pathways and around buildings. Courtyards and indoors also created spots with almost zero air velocity. Such impacts of courtyards and indoor configurations on air circulation patterns are observed for buildings located within the study zone (Figure 3-Loc A) or at the edge in the direction of upstream flow (Figure 3-Loc B).
In regard to scenario 3, interior building configurations and indoor ventilation patterns changed air circulation in adjacent pathways and courtyards. The simulations also show that some inner spaces are not ventilated despite being fully exposed to exteriors. Moreover, courtyards and indoors create connections between the airflow in non-adjacent pathways. The air circulation inside interior and exterior courtyards is also connected indoors.
Simulation results visualise air circulation and velocity contours in courtyards. Figure 4 illustrates some examples of courtyards with different adjacent sides, dimensions, and connections with indoors. As illustrated in Figure 4(a), courtyards with a single side adjacent to the outdoors are less likely to be ventilated and involved with the airflow in the pathway compared to courtyards with double sides adjacent to pathways. Figure 4(a,b) show that narrow courtyards with a depth-to-inlet width (d/w) ratio of above 2 are more likely to offer a large air stagnation zone with nearly zero velocity. Such negative impacts could occur for courtyards with either a single side or double sides adjacent to pathways. Yet, the single-sided narrow courtyards show a higher likelihood of creating air stagnation zones. Furthermore, interior building configurations can also completely change air circulation and flow patterns inside adjacent courtyards as depicted in Figure 4(c).
The simulations show that courtyards with relatively regular and irregular polygon shapes have almost similar impacts on air circulation. Figure 4(b) enlarges some examples of regular-and irregular-shaped courtyards with almost similar (d/w) ratios. As can be seen, these courtyards enable almost similar air velocity patterns and stagnation zones. Figures 2 and 3 visualise such behaviours in all regular-and irregular-shaped courtyards throughout the study zone.

Wind velocity ratio in the studied traditional city
Velocity ratios further reveal the impact of design scenarios on airflow and natural ventilation in the main corridors and pathways. As shown in Figures 5 and 6, the wind velocity fluctuates in the main corridor and pathways under all study scenarios. Figure 5 shows that the velocity ratio inside the main corridor is higher than pathway 1 under all design scenarios. The airflow in the main corridor has up to 40% higher velocity ratios in some spots under scenario 1 than in scenarios 2 and 3 ( Figure 5(a)). The ratio of wind speed throughout the main corridor is relatively similar for scenarios 2 and 3. On the contrary, the wind velocity ratio in pathway 1 is most likely to be up to 20% higher under scenarios 2 and 3 than in scenario 1 ( Figure 5(b)). Velocity ratios are almost similar for scenarios 2 and 3 of pathway 1. Overall, the main corridor and pathway 1 are mainly affected by courtyards as they have no direct relationships with case study buildings. Courtyards also reduce air velocity ratios in the main corridor. Such design scenarios with courtyards, however, slightly increase the wind speed in pathway 1 which is narrower than the main corridor.
Wind velocity ratios in pathways 2 to 6 are changed with different patterns under scenarios 1, 2, and 3. Figure 6(a,b), pathways 2 and 3, i.e. relating to building 1 in location A, show inverse velocity ratio patterns comparing scenarios 1 to 3. The wind velocity ratio is generally reduced in pathway 2 under scenario 3 compared to the corresponding patterns under scenarios 1 and 2. The velocity ratio under scenario 3 can be nearly 60% lower than the values computed for scenarios 1 and 2. Yet, the inlet velocity ratio of pathway 1, i.e. west side, is higher by about 20% under scenario 3 than in scenarios 1 and 2. On the contrary, the velocity ratio of scenario 3 for pathway 3 is higher than scenarios 1 and 2, i.e. up to about 20%. The inlet velocity ratio of pathway 3 is reduced to about 60% under scenario 3 compared to scenarios 1 and 2. Furthermore, the velocity ratio patterns for pathways 2 and 3 under scenario 2, i.e. including only courtyards, fluctuate between the corresponding patterns for scenarios 1 and 3.
As for pathways 5 and 6 surrounding buildings 1 and 2 in location B, velocity ratio patterns behave similarly to pathways 2 and 3. As illustrated in Figure 6(d), pathway 5 has lower velocity ratios under scenarios 1 and 2 compared to scenario 3. The wind velocity ratio in pathway 5 under scenario 1 increases by about 40% compared to scenarios 2 and 3. Another bump of a higher velocity ratio with a nearly 20% increase is also observed for both scenarios 1 and 2. The velocity ratio of pathway 5 has fewer fluctuations and bumps under scenario 3. In regard to pathway 6 in Figure 6(e), velocity ratios increase by up to 20% under scenario 3 compared to the relatively identical patterns for scenarios 1 and 2.
The velocity ratio patterns of pathway 4 represent the impacts of design scenarios in locations A and B. About the first half of pathway 3 passes through location A. The second half passes from location B. The velocity ratio of pathway 4 is generally reduced by about 20% to 40% under scenario 3 compared to scenarios 1 and 2. Pathway 4 has a few spots with about 20% increases under scenario 3 which are most likely outside locations A and B (Figures 1 and 2). Scenarios 1 and 2 enable fairly similar velocity ratio patterns throughout pathway 4.

Discussion
Simulation results characterise the mechanism of natural ventilation and air circulation in traditional Bushehr City. Scenarios 1-3 reveal that the unstructured, irregular shape of the urban fabric creates narrow pathways with random changes in directions, and is less likely to contribute to natural ventilation in terms of enhancement in air circulation or velocity ratios. The main corridor with a relatively wide width and straight direction, i.e. created by contemporary interventions, presents higher potentials to improve air circulation and ventilation than do multiple narrower pathways within the study zone. The main corridor plays a prominent role in increasing the distribution and penetration of upstream airflow in the study zone. Narrow pathways inside the zone, however, offer lower contributions to urban ventilation represented by air circulation and velocity. More specifically, an unstructured irregular form with random changes in the direction of pathways is more likely to obstruct air circulation and reduce the flow rate. Such urban form configurations produce multiple edges and corners within and around the zone which are likely to create non-ventilated spots with air stagnation and vortices.
Furthermore, the unstructured form of Bushehr City is less likely to enable a consistent and stable air velocity ratio throughout a pathway. As visualised by the simulation results, the wind speed can significantly fluctuate with several high/lowvelocity spots throughout a pathway. Overall, the unstructured urban form creates highly uncertain and unpredictable air circulation and velocity patterns. The irregular, unstructured geometries of building, courtyards, and pathways cause adverse impacts and unplanned changes in the velocity and direction of air in multiple spots through the study zone. Such behaviours appear to contradict the common view that there is a positive impact from the traditional form of Bushehr City on its higher natural ventilation. Future research could study the efficient width and direction of pathways that can improve ventilation in such urban fabrics. Further experimental studies and comprehensive digital simulations are still required to explore the impact of such urban forms on other thermal comfort and natural ventilation variables such as providing shadow and preventing solar radiation absorption by urban surfaces and building façades.
In regard to scenarios 2 and 3, courtyards and interior building configurations could have both positive and negative impacts on urban natural ventilation. Simulation outputs show that courtyards and indoors modify air circulation and velocity ratio patterns in adjacent pathways. As positive contributions, courtyards and interior configurations improve the penetration, circulation, and velocity of airflow in some spots in adjacent pathways. Indoors and courtyards also enhance urban ventilation by connecting air circulation in non-adjacent pathways. Indoors particularly connect air circulation in interior and exterior courtyards to pathways. As negative impacts, courtyards and indoors can shift airflow direction and reduce wind speed in some spots and adjacent pathways. Thus, air circulation and velocity ratios are negatively impacted in some spots of courtyards and interior configurations. Simulation outputs also reveal that the natural ventilation of some inner spaces is not necessarily enhanced by being fully open to exteriors. The natural ventilation performance of traditional Bushehr City should, thus, be evaluated in relation to courtyards and indoors. Research should also develop an integrated simulation of indoor and outdoor natural ventilation for the entire study zone in the cases of traditional cities.
The computerised simulation reveals that air circulation in courtyards is impacted by the number of sides adjacent to pathways, courtyards' dimensions, and connections to indoors. The shape and geometry of courtyards have not, however, shown a significant relation to airflow patterns. Courtyards with double sides adjacent to pathways are more likely to be ventilated than those with a single side adjacent to a pathway. Double sided courtyards also offer a higher likelihood of cross ventilation and air circulation between two adjacent pathways. The dimension of courtyards is likely to affect natural ventilation and air circulation inside courtyards and adjacent pathways. Narrow courtyards with a ratio of courtyard-depth-to-inlet-width of above 2 are more likely to offer a stagnation zone with almost zero airflow rates. Narrow courtyards with a single side adjacent to a pathway, i.e. the inlet for air, are likely to create a stagnant zone, especially at the end farther from the inlet. There are more single sided courtyards in the study zone than double sided courtyards and so the air circulation patterns in the latter should be investigated further.

Conclusion
The digital computational modelling offers ample evidence to characterise the relationship between natural ventilation and design configurations in traditional Bushehr City by simulating air circulation and velocity patterns. As a major conclusion, the unstructured configuration of the city consisting of buildings of regular and non-regular shape and courtyards as well as pathways with unplanned changes in directions could less likely contribute to better ventilation and air circulation. Such urban fabric configurations could have negative impacts on the distribution of airflow with consistent velocity ratios equal to or higher than the upstream inlet. Narrow pathways with low widths and random direction changes could reduce the air velocity and generate stagnation zones in non-desirable, unplanned spots. On the contrary, the main corridor, which has a wider width and relatively straight direction resulting from contemporary interventions, creates dominant positive effects on natural ventilation and air distribution inside the study zone.
Furthermore, this exploratory study also shows that courtyards and indoor configurations could change air circulation and velocity patterns in pathways and around and inside inner spaces. The courtyard's dimension, sides adjacent to pathways, and connections to indoors could impact air circulation and velocity ratios in pathways surrounding a building. Deep, narrow courtyards with a courtyard-depth-to-inlet-width ratio of above 2 are more likely to have a stagnation zone with zero velocity airflow. Results do not, however, reveal any significant relationships between air circulation and the shapes of courtyards.
It is advisable to use robust digital simulation further to examine features of traditional cities, such as shenashirs. Future research could study the overall thermal comfort condition in relation to the fabric configuration and the efficient width of pathways for optimum comfort; and the question of shadows produced by structures to absorb solar radiation.

Disclosure statement
No potential conflict of interest was reported by the authors.