See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/327867568

Impacts of urban morphology on reducing cooling load and increasing

ventilation potential in hot-arid climate

Article in Applied Energy · December 2018

DOI: 10.1016/j.apenergy.2018.09.116

CITATIONS READS

64 2,976

3 authors:

Kavan Javanroodi Mohammadjavad Mahdavinejad

École Polytechnique Fédérale de Lausanne Tarbiat Modares University
25 PUBLICATIONS 226 CITATIONS 196 PUBLICATIONS 1,186 CITATIONS

SEE PROFILE SEE PROFILE

Vahid Nik
Lund University
76 PUBLICATIONS 1,641 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Social sustainability in Housing Design View project

analytic Hierarchy Process (AHP) View project

All content following this page was uploaded by Vahid Nik on 25 September 2018.

The user has requested enhancement of the downloaded file.

 1
2
3
4
5 Impacts of urban morphology on reducing cooling load and increasing
6 ventilation potential in hot-arid climate
7
8
9
10
11
Kavan Javanroodi1,2, Mohammadjavad Mahdavinejad 1* , Vahid M. Nik2,3,4
1
12 Department of Architecture, Art and Architecture Faculty, Tarbiat Modares University, Tehran, Iran
13 (kavan.javanrudi@modares.ac.ir; Mahdavinejad@modares.ac.ir)
14
2
15 Division of Building Physics, Department of Building and Environmental Technology, Lund University, SE- 223
16 63, Lund, Sweden (kavan.javanroodi@byggtek.lth.se, vahid.nik@byggtek.lth.se, nik.vahid.m@gmail.com)
17
3
18 Division of Building Technology, Department of Civil and Environmental Engineering, Chalmers University of
19 Technology, Gothenburg, Sweden (vahid.nik@chalmers.se).
20 4
21 Institute for Future Environments, Queensland University of Technology, Garden Point Campus, 2 George Street,
22 Brisbane, QLD, 4000, Australia (vahid.nik@qut.edu.au).
23
24
25
Abstract
26
27 Cooling buildings in urban areas with hot-arid climate put huge loads on the energy system. There is an
28 increasing trend in urban energy studies to recognize the urban design variables and parameters
29 associated with the energy performance of buildings. In this work, a novel approach is introduced to
30
investigate the impacts of urban morphology on cooling load reduction and enhancing ventilation
31
32 potential by studying a high-rise building (target building), surrounded by different urban configurations,
33 during six warm months of the year in Tehran at four major sections including: (1) generating 1600
34 urban case studies considering three parameters (Urban Density, Urban Building Form, and Urban
35 Pattern) and modelling the urban morphology of Tehran based on a technique namely “Building
36 Modular Cells” , (2) validation study of CFD simulation of the wind flow around buildings, (3)
37 calculating the average cooling load and wind flow at the rooftop of the target building, and (4)
38 investigating sixteen best urban configurations with the lowest cooling load and highest ventilation
39 potential. Results indicate that urban morphology has a notable impact on the energy consumption of
40
41
buildings, decreasing cooling load and increasing ventilation potential more than 10% and 15%
42 respectively, compared to the typical cases. This work also proposes design solutions for architects and
43 urban designers, based on Top 100 configurations (out of 1600), for improved energy performance and
44 better ventilation of buildings in urban areas.
45
46
Keywords: Urban morphology; urban energy; cooling load; ventilation; high-rise buildings; hot-arid
47 climate;
48
49
50 1-Introduction
51
52 The world population in urban areas has increased rapidly during the last century [1, 2]; from 30% in
53 1950 to 54% in 2014 and it is expected to increase to 66% in 2050 [3, 4]. This rapid urbanization and the
54 consequent changes in urban density have increased the air pollution and energy consumption in cities,
55
especially due to the demands from building sector [5]. Heating, cooling and lighting are accounted for
56
57 more than 70% of energy consumption in the commercial and residential buildings in cities [6]. In
58 countries with hot-arid and tropical climate a huge amount of energy, particularlyly electricity, is
59
60
61
62
1
63
64
65
 1
2
3
4 consumed for cooling buildings [7, 8]. Conditions may worsen by climate change which induces warmer
5 weather on average as well as stronger and more frequent extreme conditions [9, 10] increasing the
6
7
average and peak cooling load as well as thermal discomfort in urban areas with hot summers [11, 12].
8 The energy performance of buildings in urban areas is highly affected by urban microclimate and urban
9 heat island (UHI). A large amount of solar radiation is reflected from and stored in surfaces of buildings
10 and urban components, increasing the surface and air temperature. Kalnay and Cai [13] have shown that
11 rapid urbanization and high urban density have caused 0.27 °C mean air temperature increase during the
12 last decade in the United States. This results in higher demands for air-conditioning and cooling in longer
13 time periods, compared to rural areas [14, 15].
14
15 The urban population in Iran has grown from 50.6% of the whole country population in 1987 to 74% in
16 2016, which is significantly higher than the global average growth rate [16]. Energy consumption per
17 capita in Iran – with 1.06% of the world population – is five times higher than the world average [17].
18 The average electricity consumption growth in Iran (7.9%) is more than twice of the global average
19 (3.3%) [18] and more than 33% of the annual electricity production is consumed for air-conditioning
20
systems in cities [19]. The situation is more critical for megacities of Iran with high urban population
21
22 growth and long hot-arid summers such as Tehran, Isfahan, Tabriz, and Mashhad; cooling demand during
23 hot seasons (April to September) introduces very large base and peak loads on the electricity (and water)
24 network. The current situation in induced by several factors such as climatic conditions, poor design and
25 inefficient construction of buildings and HVAC systems (e.g. using typical “Evaporative Air-Cooler” in
26 more than 70% of buildings in cities with very poor efficiency) as well as the lack of practical building
27 codes and regulations for having low-energy and sustainable buildings. Moreover, air pollution is a big
28 problem in the mega-cities (and the southern part of the country), especially in Tehran – the capital city of
29 Iran – as one of the most polluted cities in the world [20, 21]. This results in the need for air purification
30
and makes it difficult to use typical passive methods such as natural ventilation or regular air conditioning
31
32 systems as a part of a comprehensive plan to reduce cooling energy consumption and preserve thermal
33 comfort simultaneously.
34 Improving the living conditions and energy performance of urban areas depends on several technical,
35
environmental and economic factors. Although several studies have addressed the role of urban
36
37 morphology on energy demand [22, 23] and wind potential for ventilation [24, 25], the impacts of urban
38 morphology needs to be further explored considering larger variations and deeper analysis [26]. It is
39 shown that wind flow plays an important role in passive or active ventilation systems, helping to reduce
40 the cooling load of buildings [28, 29] and urban heat island [30, 31] and to enhance the thermal comfort
41 [32] and thermal circulation around buildings [33]. Furthermore, research works have pointed to an
42 indirect relation between higher wind flow rate in urban canopies and average surface temperature in
43 urban areas [34, 35], which can directly or indirectly affect the heat gain through external walls [36] and
44 consequently the cooling load of buildings [37, 38]. Thus, both thermal and wind flow characteristics
45
46 should be taken into account to evaluate cooling load and ventilation potential in an urban area.
47 Selecting the influencing physical parameters to model an urban morphology is an inevitable challenge
48 considering numerous variables and inputs such as urban form, type of buildings, microclimate, urban
49 density, streets and canopies [39, 40]. This becomes more complicated for simulating the cooling demand
50
51
and ventilation potential in urban areas due to their dependence on numerous unsteady climatic inputs and
52 complex boundary conditions, which are influenced by the heat transfer mechanisms, rates and
53 coefficients. Assessing such a complex system requires solving the equations for conservation of energy,
54 mass and momentum [41]. Thus, an integrated method is needed to combine these influencing parameters
55 and indicators to model an urban area in order to simulate energy demand and ventilation potential.
56 Although there have been some attempts to develop integrated methods (e.g. [42, 43]), it has not been
57 investigated in the modern architecture of Iran, despite of serious issues with the energy supply and
58 consumptions in urban areas. Interestingly, the traditional Iranian architecture has implemented several
59
effective and sustainable solutions to provide the highest possible thermal comfort in buildings and cities
60
61
62
2
63
64
65
 1
2
3
4 in the microclimate scale with the minimum energy consumption, such as wind catchers [44, 45], domed
5 and vaulted roof shape [46, 47], courtyard houses [48, 49] and using efficient materials [50, 51]. Some
6
7
attempts have been recently made in Iran to establish a national low-energy construction regulation (the
8 most recent one is National construction regulation No.19) [52] based on modern concepts similar to
9 LEED in the US, CGBL in China, DGNB in Germany, HQE in France and BREEAM in UK [7, 53].
10 However, results are not stratifying in practice and unfortunately not enough attention has been paid to
11 include the traditional methods during the urbanization boom. In fact, the national low-energy and green
12 construction regulations have been developed more based on engineering calculations and mathematical
13 language, while urban designers and architects – as one of the most influencing groups in designing
14 buildings and cities – are neglected. Regulations and building codes should be more practical with design-
15
based low-energy recommendations, easy to be applied by architects and urban designers at the early
16
17 design stage in both building and urban scale. To improve the energy performance of buildings in urban
18 areas and to take into account the important urban physical characteristics in urban planning, it is required
19 to develop general and easy-to-use regulations and/or recommendations – based on integrated methods
20 that model and evaluate numerous urban morphologies in cities – for architects and urban designers.
21
22 The literature in modeling urban morphology can be divided into three main approaches, based on
23 studying hypothetical configurations (e.g. [54, 55]), real-site configurations (e.g. [56, 57]) or both
24 configurations [58]. Through applying any of these approaches, several studies have investigated the
25 relation between urban morphology and the energy demand and supply of buildings in urban areas, each
26 considering some influencing parameters of urban morphology in terms of energy consumption [3, 59]
27 and wind assessment for ventilation [60, 61]. Four mostly applied methods for simulating thermal
28 behavior and ventilation potential of buildings in urban areas can be recognized including: observations in
29 real-sites (e.g. [62]), real-site measurement (e.g. [63]), experiments in the laboratory scale (e.g. [64]) and
30
numerical methods (e.g. [5]). Each method has its practical limitations; observational method is used in
31
32 short-term measurement studies [62], real-site measurement is usually adopted in studies with real-site
33 models in a long-term period [63], experiments are suitable for small scale studies such as a distinct part
34 or section of an urban system usually in the building laboratories [64], while numerical methods are
35 flexible for different scale objectives under validation and verification process [65]. Numerical simulation
36 has been adopted as a popular approach to study impacts of urban morphologies on energy consumption
37 and ventilation through simulating the air flow and heat (and moisture) transfer by means of
38 computational fluid dynamics (CFD), using more sophisticated software tools (such as ANSYS Fluent [66,
39
67] and OpenFOAM [68, 69]) or more user-friendly programs (such as Air Pack [70], ENVI-met [71, 72],
40
41 Grasshopper plugins [73] or EnergyPlus -for thermal simulation- [74, 75]) applying different numerical
42 methods such as RNG Realizable k-ɛ turbulence standard model which is one of the largely accepted
43 numerical methods for simulating wind flow in an urban context with various influencing parameters [76,
44 77].
45
46 Based on the literature review, the most influencing parameters of an urban morphology that should be
47 taken into account to save energy for cooling of urban areas are (1) urban density or compactness (e.g.
48 [78, 79]), (2) building form or geometry in neighborhoods (e.g. [80, 81]), (3) streets or canopies geometry
49 (e.g. [82, 83]), (4) H/W ratio (building height divided by street width)_(e.g. [84, 85]), (5) shading
50 obstacles (trees or adjacent buildings) or devices (e.g. [86, 87]) and (6) buildings orientation (e.g. [88]).
51 Moreover, there are other variables and indicators which have been addressed in different studies (e.g.
52 [89, 90]). However, to the best of our knowledge, a comprehensive research work on the possible design-
53 based parameters which form an urban morphology is still not well-addressed. Current studies on the
54
urban-scale, have focused on one or two physical aspects of urban areas; where in the most cases only
55
56 some cube forms are representing a complex urban neighborhood (hypothetical configurations). In more
57 detailed studies, a part of a neighborhood is simplified and modeled as a whole city (real-site
58 configurations); however the shape and geometry of buildings and urban areas are not considered in the
59 modeling process. Therefore, results out of such studies are not valid for the whole city or cannot be
60 generalized for other cities with similar climate.
61
62
3
63
64
65
 1
2
3
4 As a step beyond the latest developed approaches in urban-scale studies, this work aims at investigating
5 the impacts of urban morphology (considering numerous compositions) on the energy demand and
6
7
ventilation potential of high-rise buildings within urban areas by means of a novel approach emphasizing
8 on geometry, form and design-based parameters and by introducing a new technique, namely ‘Building
9 Modular Cells’ (BMC), for accurate modeling of an urban morphology. The considered case study if the
10 city of Tehran with high urban population, various areas with different urban densities and long lasting
11 hot seasons [16]. Moreover, through considering three comprehensive parameters – urban density, urban
12 building form and urban pattern – this work attempts to reveal the unexplored aspects of urban
13 morphology’s impacts on the building performance by adopting an integrated method using EnergyPlus
14 for thermal simulations and RNG Realizable k-ɛ turbulence standard model as the CFD approach for
15
ventilation potential. Furthermore, this work formulates a platform to consider urban morphology in the
16
17 urban design process, developed based on the introduced approach and techniques with the potential of
18 being implemented in newly built urban areas with similar climate conditions.
19 To facilitate the readability of the proposed approach, the workflow (visualized in Figure 1) is presented
20
in four main sections and twelve steps. In “methodology” (section 2), Tehran’s urban characteristics and
21
22 morphology is recognized and defined by several statistical indicators (step 1). Then in the second step,
23 three main parameters of the study (urban density, urban pattern and urban building form) are defined by
24 introducing “Building Modular Cell” (BMC) technique in detail; this results in generating 1600 urban
25 configuration case studies. Then, governing equations (step 03), weather data and boundary conditions
26 (step 04) and simulation tools and procedure (step 05) are discussed respectively. In section 3, a
27 “validation study” is performed for the introduced CFD approach considering the governing equations,
28 boundary conditions and input data in two steps (6 and 7). The validation cases are a simple cube and one
29 of the case studies with cuboid forms with pressure coefficient and wind velocity according to
30
experimental and real-scale measured data. In section 4, the impacts of urban morphology on cooling
31
32 demand and ventilation potential are thoroughly discussed for each introduced parameter (steps 8 and 9).
33 In step 10, the best sixteen urban configurations are presented and investigated in detail; exploring the 3D
34 forms, layouts, cooling demand and CFD wind speed results for each case. In section 5, after
35 summarizing the main findings of the study, Top 35 and Top 100 cases with the lowest cooling demand
36 and highest ventilation potential are discussed extensively to derive design-based results and suggestions
37 out of the findings (step 11). In the final step (12), the results of the study, considered parameters and
38 their impacts are compared to the other relevant studies (state-of-the-art) to find the probable similarities
39
and new aspects of this work, implying the novelty of the work. In the conclusions (section 6), design-
40
41 based suggestions are presented with respect to the architectural design language and future investigations
42 on the introduced approach and techniques are addressed.
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
4
63
64
65
 1
2
3
4 Figure 1. The structured workflow of the research
5
6 2. Methodology
7
8 2.1 Step 1: The city of Tehran
9
10 Tehran, a megacity which is the capital of Tehran province and Iran, is located at the latitude and
11 longitude of 35.69°N and 51.42°E with the average elevation of 1200 m above the sea level [94]. The
12 climate conditions are what is known as the dominant semi-arid climate [95], with cold winters (a
13
14
minimum of -10°C in January), hot-arid summers (a maximum of 40°C in July) [96], annual relative
15 humidity of 41% and annual rainfall of 233 mm [97]. The population of the city has increased during the
16 last century (from 1.8 million in 1956 to 8.37 million in 2016). Consequently, the land coverage has
17 increased rapidly from 100 km2 in 1956 to about 730 km2 in 2016 [98]. Tehran is divided into twenty-two
18 municipal districts with different demographic and geographic characteristic (Figure 2). Districts 4, 5, 15
19 and 2 are the most populated districts (respectively) and 14, 17, 7 and 12 are the densest ones. Since each
20 district of Tehran is consisting of dozens of neighborhoods with distinct physical characteristics, to
21 recognize its urban morphology, an extensive study has been conducted by authors [99] with the aid of
22
the Comprehensive Master Plan [100], official reports, building codes and published studies [101, 102],
23
24 noticing a clear tendency to build high-rise buildings due to rapid urbanization, high land cost and lack of
25 suitable lands. Moreover, our study has indicated that a typical neighborhood with at least one high-rise
26 building in Tehran consists of three types of surrounding buildings including Short-rise Buildings (1 to 4
27 floors), Mid-rise Buildings (5 to 8 floors) and High-rise Building (9-12 and higher floors) and in most
28 cases together with an open green urban space, besides streets, pathways and canopies with different
29 widths. Table A.1 in Appendix A shows the main demographic and regional characteristics of Tehran.
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59 Figure 2. Twenty-Two districts of Tehran with different demographic and geographic characteristic [103]
60
61
62
5
63
64
65
 1
2
3
4 2.2 Step 2: Modeling Urban morphology based on “BMC”
5
6
To model the urban morphology of a city, all influencing physical parameters should be considered. In
7
8 order to consider the influencing parameters, similarities in different part of the city should be recognized
9 and categorized as distinguished neighborhoods with similar morphological properties such as
10 compactness, geometry or form [104].. In fact, accurate defining and modeling of several indicators and
11 variables of compactness parameter in an urban area is a key to categorize different neighborhoods [105].
12 In this study, the main demographic and regional characteristics of Tehran are categorized based on
13 several parameters to enhance the accuracy of the urban models. First, a hypothetical site with the total
14 area of 24000 m2 was defined as the basic neighborhood template. Since the modeling process in this
15 study is based on a parametric algorithm, a new technique namely “building modular cell” (BMC) is
16
17
introduced to generate the urban buildings; which is based on 8×8×X m cuboid module (X is the variable
18 height according to the following introduced form generation rules) on the defined site as the main
19 platform for urban configurations. These dimensions are selected according to the typical reinforced
20 concrete structures in Tehran, in which the height of each floor is 4m (considering practical limitations).
21 In the second stage, considering the tendency to build high-rise buildings in Tehran, a high-rise building
22 was selected as the target building, which is defined by official building codes of Iran as a twelve-story or
23 taller structure [107]. The size of the twelve-story target building is 40×40×48 m with a simplified typical
24 open plan office, having twenty 2×2m windows in each floor (236 windows in overall) and two 4×3 m
25
entrance glass doors in west and east of the building. A stair box, two elevators and a 22.8×22.8 m
26
27 internal zone with four 4×3 m entrance are also considered and modeled (Figure 3). Moreover, physical
28 characteristics of the building such as materials (glass, metal, bricks, concrete wood etc.), structure,
29 ceiling, walls, insulations, glazing ratio and some other influencing variables designed based on national
30 building codes of Iran and the results of a sensitivity analysis with adopted tools (Table A.3 in Appendix
31 A). Furthermore, typical cooling demand of such a building in Tehran was calculated and verified based
32 on introduced methods of national building code No.19 (entitled “System Performance Method and
33 Prescriptive Method”) , where average cooling load of six-warm months in an efficient twelve-story
34 building within different neighborhoods (designed and built based on the building codes restrictions) of
35
Tehran can approximately range between 1930 to 2410×103 kWh based on different variations such as
36
37 insulation type, glazing etc. [52]. However, cooling demand can be much higher – as it is in the real
38 conditions of Tehran – where numerous buildings are constructed regardless of these building codes.
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58 Figure 3. the target building, a twelve-story office building:(a): S/E view of the 3D model of the building. (b):
59 Architectural typical plan of the target building (2 nd floor to 12 floor)
60
61
62
6
63
64
65
 1
2
3
4 In the third stage, the hypothetical site was fragmented into nine equal square blocks (40×40 m) including
5 seven building blocks, one open green space and the target high-rise building, while each block is divided
6
7
into twenty-five 8×8m cells. In other words, the template site is a matrix with three rows and columns
8 where the target building is placed in the center of the site, surrounded by seven buildings and one open
9 green space and two highways between blocks. Figure 4 shows modeling process of the template
10 hypothetical site considering the BMC technique. According to BMC, three main parameters of the study
11 are urban density, urban building forms and urban pattern, introduced to model the urban morphology of
12 Tehran as it follows.
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44 Figure 4. Modeling process of the hypothetical template site based on BMC; (a) basic cuboid module (8×8×X
45 (variable height) m), (b) each block is divided into twenty-five 8×8m cells based on the module, (c) seven blocks and
46 one open green space placed around the target building as default template site, (d) rectangle template site with
47 24000m2 area and two highways between blocks based on the module, (e) 3D template site with the target building,
48 highways, surrounding blocks and open green space based on the module.
49
50 2.2.1 Urban Density
51
52 Urban density is the most comprehensive parameter of the study, which determines layout, plan and the
53 heights of the case studies. Tehran’s neighborhoods have a wide range of built density, population and
54 number of buildings. In order to recognize, identify and model all possible urban configuration with a
55 comparative approach, five density ranges were defined according to Table A.1 (Appendix A) and
56
57
categorized based on the number of selected cells out of twenty-five:
58 a) Proposed density by the building codes of Tehran Municipality (Norms): 60% (selecting 15 cells)
59 b) Maximum Density: 85.6%, Normalized density (Max): 84% (selecting 21 cells)
60
61
62
7
63
64
65
 1
2
3
4 c) Minimum Density: 35.8%, Normalized density (Min): 36% (selecting 9 cells)
5 d) Mean Density: 70.24%, Normalized density (Mean): 68% (selecting 17 cells)
6
7
e) Middle Density: 50%, Normalized density (Middle): 48% (selecting 12 cells)
8 As stated, in order to increase the accuracy of the urban model, six other indicators are used in this study
9 in a way to cover different built densities in Tehran (Table 1) as well as the main indicators in modeling
10 urban morphology (considered in this study), which are Building Density (BD), Volume-Area Ratio
11
12
(VAR), Plot Area Ratio (PAR) and Site Coverage (SC). Descriptions and calculation method of the six
13 indicators used in the study are presented in Figure 5.
14
15
16
17
Table 1: Urban density ranges characteristics
18
19 TBEA TAUO TAUB
Ranges SC BD PAR VAR
20 km2 km2 km2
21 Max 139 3271 3923 85.6 0.0014 2.91 11.64
22 Middle 20.3 531 1349 50 0.0010 1.28 5.12
23 Min 8.07 1186 609 35.8 0.0006 0.72 2.88
24 Mean 32.17 950.27 1938.59 70.24 0.0011 2.08 8.34
25
Norms - - - 60 0.0008 1.76 7.08
26
TBEA: Total Built Environment Area
27 TAUO: Total Area of Urban open spaces
28 TAUB: Total Area of Urban Buildings
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46 Figure 5. urban morphology parameters used in this study to define urban density; a) volume area ratio (VAR):
47 total volume of surrounding buildings divided by the total site area of site (24000 m 2), b) site coverage (SC) the total
48 area of the ground floor of the building divided by 1600 (the area of each sub-site: 40×40m), c) plot area ratio
49 (PAR): total floor area of the surrounding buildings divided by the total site area, d) building density (BD): number
50 of selected cells divided by the total site area, e) urban plan area density (λp): the built area projected onto the
51 ground surface divided by the site area in a horizontal section (A s), f) frontal area density (λf): is the area of frontal
52 surface of façade to the As
53
54
The next challenge is modeling the height of the surrounding buildings. As stated before, there are three
55
56 types of buildings (in term of height) in Tehran based on the building codes and authors’ study: low-rise,
57 mid-rise, and high-rise. To simplify the simulation process and decrease the calculation load, two constant
58 building heights was considered for each type of building height:
59
60
 Low-rise buildings (BL): two-story (h=8m) and four-story (h=16m),
61
62
8
63
64
65
 1
2
3
4  Mid-rise buildings (BM): six-story (h=24m) and eight-story (h=32m),
5
6
 High-rise building (BH): ten-story (h=40m) and twelve-story (h=48m).
7 Distributions of the height of the surrounding buildings were also based on the five urban density ranges,
8 in which total building floor considered for Max urban density with 52, Mean urban density with 46,
9
Norms urban density is 42, Middle urban density with 40 and Min urban density with 30 floors (Figure
10
11
6). The differences between these height distributions are based on the morphology indicators and total
12 urban density ranges. In fact, after generating the layouts of the urban cases, each case was extruded
13 according to the introduced height distribution.
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47 Figure 6. Urban Density height, PAR, VAR and λf morphology in the context of urban pattern one (P01):
48 (a): “Max” density morphology, (b): “Middle” density morphology, (c): “Min” density morphology,
49 (d): “Norms” density morphology, (e): “Mean” density morphology
50 (f): Basic Cubic Module and forty-cell diagram plan of each urban building
51
52 2.2.2 Urban building forms
53
54 After defining the template sites, target building and urban density parameter, case studies were generated
55 according to a series of form generation rules based on the basic module. The generating rules were
56 developed based on the building codes, architectural design logic and Tehran Municipality regulations.
57 As stated, five defined urban density ranges were normalized according to the basic module. Two initial
58 rules were defined at the first step: 1) selected cells must be connected from their sides and not through
59
60
their corners (based on structural and architectural logics), and 2) in all selections, only one integrated
61
62
9
63
64
65
 1
2
3
4 form must be chosen (configurations with two separate forms are not allowed). Figure 7 is showing the
5 first initial rule in (a), (b), (c) and (d) and the second rule in (e). Moreover, one true and false sample
6
7
example for each urban density range is graphically illustrated considering these initial rules. For
8 instance; in the case of maximum density range, the false pattern is having two disconnected forms; in the
9 case of Mean density range, the false pattern is having two disconnected forms and a corner joint in cell
10 number 25; in the case of Min density range, connections between cells 17-23, 23-19 and 19-15 are from
11 their corners which is unacceptable. These two initial rules in addition to a series of other rational
12 limitations were defined in the form generation algorithm in Grasshopper.
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44 Figure 7. Top: initial rules: (a): the 8×8 module cell, (b): selected cells must be connected from their sides, (c):
45 connection from edges are not acceptable, (d): unacceptable cell connection from edges, (e) detached selected cell
46 and edge connection
47 Bottom: examples of acceptable and unacceptable form generation based on the initial rules for five density ranges
48
49 As one might expect, there would be millions of combinations based on the defined rules. However, a
50 large number of these combinations are not eligible from the architectural point of view. Moreover, many
51 other generated layouts have negligible differences in cooling load and wind flow characteristics. To deal
52 with this situation, authors defined another parameter entitled “Urban Building Form” to select the best
53
eligible layouts out of generated combinations. Thus, the most frequent building forms in Tehran were
54
55 recognized according to the cuboid module, and four main forms were selected as indicators including:
56 “L” forms, “U” forms, cube forms “C” and courtyard forms “CY”. Furthermore, using a novel
57 combination algorithm based on the initial rules in Grasshopper, Top 20 generations nearest to these
58 urban building forms in each urban density ranges were selected, resulting in 100 cases for each form. All
59 in all, 400 configurations were generated and selected as the most familiar possible plans which are used
60 by architects and urban designers in Tehran to include complexities in urban areas into the models. Figure
61
62
10
63
64
65
 1
2
3
4 8 shows all 400 forms according to density ranges, building forms and influencing variables and
5 indicators.
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39 Figure 8. 400 case studies categorization based on building forms and influenced urban density indicators
40 according to BMC technique
41
42 2.2.3 Urban pattern
43
44 As discussed before, another influencing parameter on urban energy demand and the behavior of wind
45 flow around buildings is urban streets or canopies [82, 83]. In order to assess the role of this parameter,
46 two types of streets were considered in the template site (acquired from different urban planning codes of
47 Iran) including highways and low-width streets (only in northern and southern sides of the target
48 building) between buildings in the context of four patterns in the second phase (Figure 9). It should be
49 noted that all patterns have the same area (24000m2) and street to site area ratio in all four patterns is 40%
50
(patterns with wider streets have a lower width highway). The other indicator used in this categorization
51
52 is the ratio of building height to street width (H/W). These four urban patterns are:
53 a) P01: no street between buildings and 40m highways
54 b) P02: 4m width street (H/W: 12) and 33.75m highways
55
c) P03: 6m width street (H/W: 8) and 30.9m highways
56
57 d) P04: 8m width street (H/W: 6) and 28.23 m highways
58
59
60
61
62
11
63
64
65
 1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37 Figure 9. Four urban patterns in the study: All patterns have the same total area (24000m 2) and street to site area
38 ratio (40%) (the dash-line rectangle shows the margin of the P01 in other patterns)
39
40 The last phase of step 2 is placing these 400 cases in the context of four patterns. It means that every 400
41 cases are individually placed in patterns P01, P02, P03 and P04 and consequently 1600 urban
42 morphologies are generated based on BMC (Figure 10). To avoid complexity in recognizing the urban
43 morphologies, authors used an abbreviation logic based on these main parameters: urban pattern, form
44
and density range. For example, P01 C Max41 is the 41st case study generated at pattern one (P01) with
45
46 cube form (C) and maximum density range (Max). Table A.2 in the appendix shows abbreviations used to
47 define all 1600 case studies based on the main parameters.
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
12
63
64
65
 1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22 Figure 10. categorization of 1600 generated case studies based on three main parameters of the study,
23 (The number in the blanket is the total number of generated cases in each category)
24 2.3 Step 3: Governing equations
25
26 Analytical methods were applied to validate the numerical simulation of cooling load and wind flow.
27 Cooling loads are calculated considering both sensible and latent loads [108]. The calculated cooling load
28 is affected by the heat transfer through building external surfaces (walls, roof, floors, windows, ceilings
29 etc.) – known as building envelope – as well as internal sources (people, appliances, equipment, lighting
30
etc.) and infiltration through the building envelope and openings [109]. According to ASHREA [110],
31
32 cooling load for a typical building is calculated according to equation (1):
33 Eq (1)
34
35 Where QW, QR, and QWd are heat transfer through windows, walls and roof (considering conduction,
36 convection and radiation) all in Watt. Heat transfer through windows is calculated as the following:
37
38 Eq (2)
39
Here, QCon, QDir, and QDif are conductive, direct solar and diffuse solar heat gain from windows:
40
41 = .A ( ) Eq (3)
42
43 Eq (4)
44
Eq (5)
45
46 2 2
Here UWd is heat transfer coefficient (W/m .K), A is windows area (m ), To and Ti (K) are respectively the
47 average indoor and outdoor air temperature, EDir is direct irradiance (W/m2), SHGC is direct solar heat
48 gain coefficient as a function of incident angle θ [degree], which can be obtained from tables, EDif is
49
50 diffuse irradiance (W/m2), Er is ground-reflected irradiance (W/m2). IAC (inside shading attenuation
51 coefficient) is assumed to be 1.0 in this study since no inside shading device is considered in the target
52 building. Heat transfer through the external walls and roof is calculated according to equations (6) and
53 (7):
54
55 Eq (6)
56 Eq (7)
57
58 Where Uw and UR, U-values of the walls and roof (W/m2/K), were calculated according to the thickness
59 and material of their constructing layers (check Table A.4). The internal heat gains from lighting (office
60 space) are calculated using equation (8) according to EnergyPlus default configurations
61
62
13
63
64
65
 1
2
3
4 Eq (8)
5
6 Here Lwa is the total light wattage (W), Ful and Fsa are respectively the lighting use factor (assumed 1.0 for
7 commercial applications) and special allowance factor (for fluorescent fixtures assumed 12.0 for quad 13
8 W lamps in this study). Infiltration rate was calculated using equation (9) based on Sherman and
9 Grimsrud method [111]:
10
11 Eq (9)
12
13 In this equation AL is effective air leakage area (cm2) and its value was assumed based on ASHRAE for
14 the office buildings [110], Cs is stack-induced infiltration coefficient ((L/S)2/(cm4.K)), Cw is wind-induced
15 infiltration coefficient ((L/S)2/(cm4.(m/s)2)), ʋ is the local average wind speed [m/s] and is the absolute
16
17
temperature difference between indoor and outdoor zones (K). Surrounding buildings can create shadows
18 on the target building and surfaces around it, which can cause significant changes in air temperature and
19 radiations, it is important to include these temperature differences in calculations [80, 112]. Since
20 EnergyPlus engine is able to provide an accurate representation of different terms in simulating heat flux,
21 air temperature and radiation [113], these values obtained from the tool and used in the presented cooling
22 load equations to include surrounding building impact. Finally, calculated cooling load through these
23 equations is compared to the cooling load from EnergyPlus (check Section 2.5).
24
25 For the CFD simulations, the standard k-ɛ turbulence model was used for simulating turbulent flow
26 conditions. The standard k- ɛ model is a mature turbulence model that has been used and validated
27 extensively by other researchers; and it is used as the default option [30] where k is turbulent kinetic
28 energy equation (10) and ɛ as dissipation of kinetic energy in equation (11):
29
30 Eq (10)
31
32
33 Eq (11)
34
35 Where Eij (-) is the rate of deformation and ui (m/s) represents velocity component in the corresponding
36 direction, and σk, σɛ, C1ɛ, C2ɛ are adjustable dimensionless constant values which are set to 1, 1.30, 1.44,
37 1.92 respectively and is the empirical dimensionless constant that is assumed 0.09 in this study.
38 (kg/m/s) is eddy viscosity which is calculate from equation (12) using Boussinesq approximation:
39
40 Eq (12)
41
42 Wall and energy Prandtl number were assumed as an average of 0.85 for turbulent thermal diffusivity in
43 this study. Table A.2 (check Appendix A) shows other specifications applied in CFD numerical
44
simulation (air density, viscosity, emissivity, conductivity etc.)
45
46 2.4 Step 4: Weather data, inputs and boundary conditions
47
48 To facilitate the simulation process and limit the number of simulations, inputs such as boundary
49 conditions, internal loads, heat gains through windows, natural ventilation, infiltration etc. were adjusted
50
with constant values and schedules to simplify the calculations process both in cooling load and wind
51
52 flow simulations for all case studies based on a sensitivity analysis (check Table A.3 in the Appendix A).
53 In addition, some physical influencing parameters on cooling load such as glazing ratio of the target
54 building, materials, insulations and wall thickness assumed constant for all the case studies.
55
The weather data of Tehran (epw file from Mehrabad Airport) with the hourly temporal resolution for the
56
57 six warm months (April, May, June, July, August and September) was used for simulating average cooling
58 loads with hourly time-step, resulting in 9600 sets of simulations (1600 case studies in six months). In
59 order to reduce the number of graphs and figures, the final results are the monthly average cooling load of
60 each month for each case study out of the hourly time-step simulations. In a parallel process, average
61
62
14
63
64
65
 1
2
3
4 wind velocity in all directions, prevailing wind direction and air temperature of the six warm months were
5 calculated (from a 50-year weather data of 1964-2014) and adjusted with Tehran weather file data, to
6
7
obtain monthly average values. In order to consider the impact of urban surrounding building and
8 obstacles on wind flow patterns, inflow velocity for each month was retrofitted to the height of 10 to 90 m
9 to define wind profile using equation (13):
10
11 Eq (13)
12
13 Here (m/s) is the mean wind speed at the height of 10 m in an urban area, Vref (m/s) is the average
14 wind velocity of 50 years of weather data at Mehrabad Airport, α is a coefficient based on surface
15 smoothness in urban area and is set to 0.36 in this study. Average wind velocity obtained from this
16
process at the height of 10 m in urban simulation used for CFD wind velocity simulation in 1600 urban
17
18 configurations for six warm months (Table 2). It should be noted that in the boundary conditions only
19 average prevailing wind direction has been applied to the CFD computational domain. Moreover, wind
20 angle assumed θ=0 in all simulations and Tehran solar heating data (radiation and heat flux) in Autodesk
21 CFD included for the 20th day of each month at 12.00 noon (peak loads) in 2016.
22
Table 2: Average weather data used in CFD simulation
23
24 Tmean Tmin Tmax Vmean Dominant Wind-direction
Months
25 C C C m.s-1 Degree
26 April 21.1 4 26.2 3.54 270
27 May 25.9 9.2 31.2 3.65 270
28 June 28.7 10.6 34.7 3.34 260
29 July 31.1 16.6 36.4 2.88 270
30 August 31 20.4 36.4 2.52 270
31
September 27.2 18 32.5 2.41 270
32
33
34 Computational domain of CFD simulation was determined based on standards like AIJ and COST [114,
35 115], in which distances from inlet and sides of the domain to model should not be less than 5H and at
36 least 10H from behind the model, where H is the tallest model height. In this study, lateral and top of the
37 domain considered 10H and rear of the domain is 20H. “H” is the height of target building which is 48m
38 in all case studies (Figure 11). The characteristics of computational domain in the both CFD tools (ANSYS
39 CFD and Autodesk CFD) is illustrated in Figure 10. Table A.4 shows boundary conditions of
40 computational domain planes in Appendix A.
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
15
63
64
65
 1
2
3
4 Figure 11. (a): Computational domain and boundary conditions used for CFD simulation (H= 48 m), (b): Section
5 computational domain and boundary condition, (c): Dimensions of computational domain used for urban patterns
6 P01-P04 W (width of the site), L (length of the site), Domain W (domain width), Domain L (domain length),
7
8 2.5 Step 5: Simulation tools and procedure
9
10 This study has adopted several tools such as RhinocerosTM and GrasshopperTM, ArchsimTM (a
11 Grasshopper-based plug-in as a part of Diva for Rhino 4.0) [116], Autodesk Inventor TM, Autodesk CFDTM
12 and ANSYS Fluent TM. These tools were linked together to simulate cooling load and wind flow patterns in
13 all 1600 case studies separately. For cooling load, Rhinoceros, Grasshopper, Archsim and EnergyPlus
14
15
were coupled together. Grasshopper is an open source plugin in Rhinoceros for parametric modeling and
16 Archsim is a plugin in Grasshopper that can convert geometries into the EnergyPlus files parametrically.
17 At the first step, 1600 urban morphology case studies were modeled in Rhinoceros and Grasshopper. For
18 each case study, all parameters and zoning were adjusted and defined using twelve calculation points on
19 each floor of the twelve-story target building (Q1 to Q12) in Archsim to be convertible into EnergyPlus
20 engine using Tehran weather data, then cooling load was simulated from April to September separately.
21 Finally, for each case study, average cooling load amounts in each month and for all six months (Qall (1-
22
1600)) were calculated (115200 cooling load values of twelve floors of each 1600 case studies were
23
gathered and analyzed in Excel).
24
25 In an independent process, the 1600 case studies were modeled in Autodesk Inventor to be convertible to
26 Autodesk CFD, where calculation zones, materials, boundary conditions, meshing etc. were defined based
27 on assumed values and average wind speed at all directions was simulated using ten calculation points (ʋ1
28
to ʋ10) at 200 cm above the roof (50 m height from the ground) from April to September using normalized
29
30 values (96000 values of 10 wind speed calculation points in six months for 1600 case studies were
31 collected and analyzed in Excel). Autodesk CFD selected as the main CFD tool in order to conduct the
32 wind flow simulations faster (compared to ANSYS Fluent).
33 Due to air pollution in Tehran, natural ventilation through openings is not the best practice and it is
34
35 recommended to purify the inlet air entering buildings. Thus, a low-cost active/passive system applied to
36 the target building based on the concepts of traditional wind catchers, in which, due to dusts and mists in
37 many central regions of Iran rising from surrounding deserts, inlet air was purified by placing different
38 natural fibers on their openings [17]. Moreover, the concept of wind catchers can be embodied in modern
39 buildings by implement of mechanical ventilators and evaporative chillers in the mechanism of the device
40 [117]. Therefore, one typical commercial air cleaner device – affordable for office buildings – was
41 considered as a combined part of the ventilation system. Hatches of this system are considered to be on a
42 plane at 2m above the target building’s roof (50 meters from ground). Ten calculation points placed on
43
this plane to assess average wind speed on the openings of the hypothetical hatches in order to obtain a
44
45 better accuracy in results (average wind speed of all ten points in all directions calculated for each case).
46 Figure 12 shows the cooling load and wind speed calculation points and their positions for one case study
47 (P04 C Middle 46).
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
16
63
64
65
 1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41 Figure 12. (a): wind velocity calculation points and dominant wind flow on one of the case studies, (b): Sample plan
42 of the target building for 2nd to 12th floors, (c): dimensional coordinates of wind velocity magnitude calculation
43 points in four urban patterns, (d): An elevation of the target building, locations of cooling load (Q) and wind
44 velocity (ʋ) calculation points (The case study is P04 C Middle 46)
45 3. Validation study
46
47 3.1 Step 6: Validation of the CFD model: a cube case
48
49 Validation is an indispensable part of the studies based on CFD simulations [118]. Several researchers
50 have reported the possibility of using single simple forms like a cube (one of most studied wind flow
51 problems) to validate settings, parameters and codes of a more complex model for virtual/typical case
52
studies [119, 120, 121]. In addition, results of different in-site measurements and wind tunnel tests for
53
54 these simple forms are widely available in published research works [122, 123]. Thus, considering the
55 large number of virtual cases with urban configurations and computational complexity of the CFD
56 simulation, at the first step, pressure coefficient of a single 3D cube (6×6×6 m3) – without urban
57 configuration – was calculated using the two CFD tools used in this study (Autodesk CFD and ANSYS
58 Fluent) and compared to in-site measurement and wind tunnel results of the same form. In the next step,
59 an urban morphology case study of the current research (P02 C Min 41) with cube form buildings were
60 selected to analyze the pressure coefficients using both the CFD tools. For the single (6×6×6 m3) cube, in-
61
62
17
63
64
65
 1
2
3
4 site measurements of pressure coefficient obtained from the results of a real scale cube study (Silsoe cube)
5 conducted by Richards et al [124] and wind tunnel test results obtained from Murakami and Mochida
6
7
[125]. In these experimental models, there are several calculation points on a vertical and on a horizontal
8 centerline section with additional tapping points on the roof, which calculated pressure coefficient from
9 points on the vertical centerline of both models were gathered and compared to CFD results of the current
10 study.
11
12
Furthermore, the validated CFD results of the same cube obtained from Abohela et al [123], used for a
13 more accurate comparison where 3D modeling and simulation (boundary conditions, weather data etc.) of
14 the cube were performed exactly according to Abohela et al [123]. The computational domain is 3H
15 backward, 1.5 H for lateral and 1.5 H windward and above the cube. Applying a high-resolution
16 hexahedron grid, the computational domains were discretized (Figure 13). In Autodesk CFD, the
17 minimum number of grids on edges was 10 with the growth rate of 1.35 in all directions for meshing. A
18 grid sensitivity analysis was conducted with different grid sizes in Autodesk CFD, compared to ANSYS
19 Fluent, in-site measurements and wind tunnels pressure coefficient results to select the finest meshing
20
size with three different mesh resolutions (0.1, 0.3 and 0.5 m). Based on the results of the validation
21
22 study, mesh resolutions of 0.1 m and 0.5m were applied for the cube and computational domain
23 respectively. The resolution for the urban case study (P02 C Min 41) was 0.3m around the target building
24 and 0.8m for the rest of the domain (see Table A.3 in Appendix A).
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
Figure 13. (a) Computational domain applied for all 1600 case studies simulation in Autodesk CFD,
43
(b): Computational domain applied for cube validation study from Abohela et al [123]
44
45
46 This geometry, domain and boundary conditions were modeled and simulated in Autodesk CFD (with
47 higher grid size and lower mesh cells due to limitations of the tool) and ANSYS Fluent (with the same grid
48 size and mesh cells as discussed above). Then, pressure coefficient results on the roof and windward and
49 leeward sides of the cube were compared to the results of the reference in-site measurement, wind tunnel
50 test and validated CFD (in twenty-one computational points, each surface seven points). Furthermore,
51 wind velocity and pressure coefficient contours with both CFD tools (Figure 14) were compared to
52 Abohela et al [123] and Murakami and Mochida [125], showing similar wind flow patterns in both
53
velocity and pressure coefficient contours.
54
55
56
57
58
59
60
61
62
18
63
64
65
 1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50 Figure 14. (a) ANSYS Fluent wind velocity magnitude at Y section contours, (b) ANSYS Fluent wind velocity
51 magnitude Plan contours, (c) ANSYS Fluent pressure coefficient at Y section contours, (d) ANSYS Fluent pressure
52 coefficient Plan contours
53 (e) Autodesk CFD wind velocity magnitude at Y section contours, (f) Autodesk CFD wind velocity
54 magnitude Plan contours, (g) Autodesk CFD pressure coefficient at Y section contours, (h) Autodesk CFD
55 pressure coefficient Plan contours
56
57 According to figure 14, four main streams are observed when wind approaches the cube; two deviate over
58
59
and bottom of windward façade and the other two streams approach the cube sides. Maximum pressure
60 coefficient is observed at the deviation point (stagnation point) where the wind flow changes its direction
61
62
19
63
64
65
 1
2
3
4 into lower parts of the windward surface of the cube. In pressure coefficient counters, the stagnation point
5 at the windward surface in this study is 0.85H for Autodesk CFD and 0.83H for ANSYS Fluent, while for
6
7
Silsoe cube is 0.8-82H, in wind tunnel test is about 0.8-0.83H and for validated CFD is 0.81H, where H is
8 the height of Cube. Comparison of pressure coefficient results of the study (ANSYS Fluent and Autodesk
9 CFD), wind tunnel and Silsoe cube in twenty-one points (seven points at each façade) is illustrated in
10 Figure 15. Thus, it can be claimed that results of both CFD tools have a good agreement with published
11 results and CFD settings, boundary conditions and meshing is accurate enough compared to the site-
12 measurements and wind tunnel results.
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Figure 15. Simulated Pressure coefficient in twenty-one calculation point at the centerline of windward, roof and
34
leeward surfaces of a sample cube in comparison with two wind tunnel test and a validated CFD results
35
36 3.2 Step 7: Validation of the CFD model: the case studies
37
38 To validate the effects of urban morphologies on the wind flow around the target building, one of the
39
generated case studies with nearly cubic forms, “P02 C Min 41” (Figure 16) was selected in Step 2,
40
41 modeled and simulated using Autodesk CFD and ANSYS Fluent with the same boundary conditions,
42 meshing, grid size and weather data as discussed in sections 2.4 and 2.5. In Autodesk CFD, meshing
43 tetrahedral cells and triangular interior faces were set to “37170”and “1098759”respectively, while these
44 numbers in ANSYS Fluent was “1290537”and “2527950”. According to retrofitted values of Table 3,
45 wind velocity considered 3.65 m/s for May in both tools and other data was set according to Table 4.
46
47
48
49
50
51
52
53
54
55
56
57 Figure 16. P02 C Min 41, case study selected for the validation study
58
Wind velocity magnitude and pressure coefficient results out of both tools are similar to the cube
59
60 simulation, where Autodesk CFD results showed a good agreement with ANSYS CFD values. Three main
61
62
20
63
64
65
 1
2
3
4 streams at the top and sides of the windward 10-story building are observed in the wind velocity contours
5 (Figure 17) where the mainstream is deviated above the building and approaches the roof of the target
6
7
building. In both CFD tools, a similar flow pattern is noticed and a large amount of flow has covered the
8 roof surface.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55 Figure 17. (a) ANSYS Fluent wind velocity magnitude at Y section contours, (b) ANSYS Fluent wind velocity
56 magnitude Plan contours, (c) ANSYS Fluent pressure coefficient at Y section contours, (d) ANSYS Fluent pressure
57 coefficient Plan contours
58 (e) Autodesk CFD wind velocity magnitude at Y section contours, (f) Autodesk CFD wind velocity
59 magnitude Plan contours, (g) Autodesk CFD pressure coefficient at Y section contours, (h) Autodesk CFD
60 pressure coefficient Plan contours
61
62
21
63
64
65
 1
2
3
4
5 At the windward and leeward facades of the target building, wind velocity is lower than the applied
6 average value (3.65 m/s). However, wind velocity reaches the maximum amount at the lateral 4m width
7
8
streets. Maximum pressure coefficient is observed at the windward façade of the target building (0.71H at
9 both tools: 0.65 in Autodesk CFD and 0.47 in ANSYS CFD). The stagnation point at the windward surface
10 of the target building in both tools is 0.86H with similar values. Figure 18 shows pressure coefficient
11 comparison in both CFD tools in twenty-one points (seven points at each façade) where a good agreement
12 is also observed for both results. By comparing all the simulation results between Autodesk CFD and
13 ANSYS Fluent, the accuracy of Autodesk CFD simulations for this study was confirmed (as some previous
14 studies [107]), therefore this tool was used to perform the rest of CFD simulations, helping to decrease
15 the calculation load and runtime.
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33 Figure 18. Simulated Pressure coefficient in twenty-one calculation point at the centerline of windward, roof and
34 leeward surfaces of a P02 C Min 01 case study in both CFD tools used in the study
35
36 4. Results
37
38 Results of this study are discussed in three sub-sections, by assessing the impacts of urban morphology on
39 cooling demand (Section 4.1) and ventilation potential (Section 4.2) as well as investigating sixteen best
40 cases (Section 4.3). In the first two sub-sections, impacts of urban density (Density and Height: Max,
41
42
Min, Middle, Mean, Norms), urban building forms (L, U, C, and CY) and urban patterns (P01-P04) on
43 cooling load and average wind speed are assessed. In the third section, sixteen energy efficient urban
44 configurations with the lowest cooling load and highest average wind speed for all directions at the
45 rooftop of the target building are presented as the “best urban configurations”. The selection criteria of
46 these sixteen best cases are based on the urban pattern and urban building forms parameters where for
47 each pattern and surrounding building forms, the most efficient case is selected.
48
49 4.1 Step 8: Impact of urban morphology on cooling load
50
51 The monthly average cooling loads during six warm months for all 1600 case studies are assessed
52 considering three introduced urban morphology parameters: urban density (Section 4.1.1), urban building
53 form (Section 4.1.2) and urban pattern (Section 4.1.3). In order to facilitate the comparison between
54 parameters and objectives and enhance the readability of the figures, the “highest cooling demand”
55 (HCD) in each month, is marked and its difference compared to the other cases are shown (in percentage)
56
in graphs. The cooling load variations in each month are discussed with regard to the introduced ranges of
57
58 urban density, urban building form, and pattern. Furthermore, for better understanding and comparison,
59 the six-month average cooling load (Average) is presented in each graph and compared to the HCD in the
60 other six months.
61
62
22
63
64
65
 1
2
3
4 4.1.1 Urban Density and cooling load
5
6 To investigate the impacts of urban density on cooling demand, monthly average cooling load for the five
7
introduced urban density ranges (Min, Middle, Norms, Mean and Max) is plotted in Figure 19-left for the
8
9 six warm months. The highest cooling demand (HCD) in all months and six-month average occur in cases
10 with Min density surrounding buildings; a lower urban density range results in a higher cooling demand.
11 The cooling load reduces from Min to Max urban density, where “the lowest cooling demand” in all
12 months occurs in cases with Max urban density surrounding buildings. According to Figure 19, in August
13 (with the highest cooling demand in all six months), cases with Max urban density showed 2.4% cooling
14 reduction (11.21×103 kWh) compared to HCD in Min urban density cases (467.29×103 kWh). The largest
15 cooling load reduction occurs in June, where Max urban density cases have 4.5% reduction (16.66×103
16 kWh) against HCD in Min urban density (370.39×103 kWh). In the other months, difference between
17
18
cases with Max urban densities (the lowest cooling demand) and Min (HCD) is: 4.3% in September
19 (3.98×103 kWh), 3.4% in July (14.87×103 kWh), 1.4% in April (1.39×103 kWh) and 1.3% in May
20 (2.39×103 kWh). Cases with Mean urban density also showed a remarkable cooling load reduction
21 compared to the HCD in each month; for example 4% in June (9.88×103 kWh), 2.7% in September
22 (3.63×103 kWh) and 2.1% in August (0.74×103 kWh).
23
24 To get an overall picture of the impacts of urban densities on the cooling demand, the Average HCD,
25 which occurs in cases with Min urban density (311.4×103 kWh) is presented. The annual cooling demand
26 reduces for 0.2%, 0.3%, 1.8% and 3.9% respectively in cases with Middle, Norms, Mean and Max urban
27 density (1% is equivalent to 3.11×103 kWh). It is clear that increasing the urban density reduces the
28 average cooling load. One important reason is that urban density affects both sub-site coverage and height
29 of surrounding buildings. While higher sub-site coverage can increase UHI in a neighborhood, increase in
30 the height of the surrounding buildings can enhance shading over the target buildings and streets around
31
it. Another benefit of denser neighborhoods, with the impact of reducing cooling load, is the less exposure
32
33 of the façade surface area and openings of the target building which decreases heat gains through
34 irradiation (blocks 03 and 04 can cover external walls and openings); this can considerably reduce the
35 direct irradiation and heat gain through external surfaces.
36 4.1.2 Building forms and cooling load
37
38
39
Impacts of the forms of the surrounding buildings on cooling demand are assessed by studying the
40 average cooling load for the four introduced building forms (L, U, CY, and C) as is shown in Figure 19-
41 right. According to Figure 19-right, except for August and July where cases with CY form have the HCD,
42 in the other months, HCD occurs in cases with L form. The monthly cooling load of the target building is
43 the lowest in cases with U form surrounding buildings. The relative reduction for U form cases (compared
44 to the HCD in each month) are 8.2% in April (13.01×103 kWh reduction against L form with
45 HCD=158.59×103 kWh), 6% in May (15.19×103 kWh reduction against L form with HCD=253.32×103
46 kWh), 7.6% in September (26.68×103 kWh reduction against L form with HCD=351.06×103 kWh), 2.9%
47
in June (11.51×103 kWh reduction against L form with HCD=397.24×103 kWh), 2.3% in July (10.24×103
48
49 kWh reduction against CY form with HCD=445.37×103 kWh), and 4.9% in August (22.96×103 kWh
50 reduction against CY form with HCD=468.70×103 kWh). The cases with C urban building form also
51 show a notable cooling load reduction against the HCD in all six warm months (L and CY urban building
52 forms). For instance, cooling load reduces around 4.8% in April (7.57×103 kWh) and May (12.15×103
53 kWh), and 5.4% in September (18.88×103 kWh), compared to HCD in L urban building form in the same
54 months. To obtain an overall picture about the impacts of urban building form on cooling load, reduction
55 in the annual cooling load (Average) is compared to HCD in cases with L urban building form (346.3×103
56 kWh for six warm months). The reduction is 1.2%, 2.3% and 3.9% for cases with CY, C and U
57
58
surrounding building form (1% is equivalent to 3.46×103 kWh).
59 It can be noted that adjacent building around the target building (blocks 03 and 04) have a major impact
60 on cooling load reduction, where the flow rate around U forms is notably lower than L and CY forms
61
62
23
63
64
65
 1
2
3
4 which can reduce thermal circulation and heat gain through infiltration and scheduled natural ventilation
5 in lower floors of the building.
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22 Figure 19. Left: average cooling load simulation of different Urban Density in six months
23
24 Right: average cooling load simulation of different Urban Building Form in six months
25 4.1.3 Urban pattern and cooling load
26
27
Results of the monthly cooling demand simulation for four urban patterns (P01, P02, P03 and P04) are
28
29 plotted in Figure 20-left in all six warm months. As stated earlier, the main variable in the urban pattern is
30 H/W ratio of streets and highways (height of the target building to the width of the north-south street).
31 The monthly and annual HCD occur in P03, except for May where P02 cases have the highest cooling
32 demand, while P01 results in the lowest annual cooling demand. According to Figure 20-left, cooling
33 reduction for P01 is 9.8% in April (23.35×103 kWh), 10.2% in June (39.530×103 kWh), 3.5% in July
34 (15.10×103 kWh), 3.9% in August (17.98×103 kWh) and 6.5% in September (21.47×103 kWh) compared
35 to P03 as the HCD in these months, and 7.8% in May (18.49×103 kWh) against P02. On the other hand,
36 P04 shows a notable reduction compared to HCD cases in all months, where the highest cooling reduction
37
occurs in June with 5.6% (21.75×103 kWh reduction against P03 cases with HCD=388.53×103 kWh) and
38
39 in May 4.1% (9.75×103 kWh reduction against P02 cases with HCD=237.98×103 kWh). Furthermore, the
40 average HCD occurs in P03 case (331.44×103 kWh), where P01, P02 and P04 cases showed 6.8%, 2.3%
41 and 3.7% cooling load reduction respectively (1% is equivalent to 3.31×103 kWh). A reason for higher
42 cooling demand in P03 cases, is the less exposure of the façade surface area and openings of the target
43 building which decreases heat gains through irradiation. In most of the case studies in P01, the openings
44 and external walls of the target building has been fully covered by surrounding buildings (blocks 03 and
45 04), particularly between first and 8th floors of the target building. Moreover, when the streets become
46 wider (4 to 6m), wind flow rates with high-temperature increases around the target building, inducing
47
48
warm air through openings and infiltration through external walls. On the other hand, in P04 (with 8m
49 wide streets), wind flow rates decrease considerably and interestingly the cooling load reduces compared
50 to the HCDs in P02 and P03 cases (see Section 4.2).
51 To further understand the impact of urban density and urban pattern on cooling load, Figure 20-right
52
53
shows the relation between these two parameters in term of cooling load reduction compared to HCD for
54 Min density cases in P03 (HCD=2043×103 kWh). The lowest cooling demand occurs for P01Max cases
55 with 16.9% reduction compared to the HCD. In fact, cooling load for P01 and P04 in all urban density
56 ranges are considerably lower than P02 and P03. Moreover, Max density cases for all patterns showed a
57 notable reduction in comparison with other density ranges (1% is equivalent to 20.43×103 kWh).
58
59
60
61
62
24
63
64
65
 1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Figure 20. Left: average cooling load simulation of different Urban Pattern in six months
20
21 Right: total average cooling load in term of urban density and urban patterns for all six months
22
23 4.2 Step 9: Impact of urban morphology on ventilation potential
24
25 To investigate the ventilation potential of the considered urban morphologies, CFD simulation results of
26 1600 case studies, in six warm months of the year, are categorized in three main urban morphology
27 parameters: urban density (Section 4.2.1), urban building form (Section 4.2.2), and urban pattern (Section
28 4.2.3). In order to facilitate the comparison between parameters and objectives, the Lowest Ventilation
29 Potential (LVP) in each month – as the cases with the lowest wind speed in all directions at the rooftop of
30 the target building – is marked and the difference between the value of LVP and the other investigated
31
parameters (in percentage) are illustrated. By means of 3D CFD simulation, average wind speed at all
32
33 directions in each month is discussed individually in term of introduced ranges of urban density, building
34 form and pattern. Furthermore, for better understanding and comparison, the “six-month average wind
35 speed” (Average) is plotted in each graph.
36 4.2.1 Urban density and ventilation potential
37
38
39 To evaluate the ventilation potential, the average wind speed (out of CFD simulations) in all directions is
40 plotted in Figure 21-left for the five introduced urban density ranges (Min, Middle, Norms, Mean and
41 Max). The lowest wind speed at the rooftop of the target building is marked as the “lowest ventilation
42 potential” (LVP), and the difference between other ranges is shown in percentage for each month. The
43 LVP in all months and the six-month average occurs in cases with Min density surrounding buildings,
44 except for June (13.1% equivalent to 0.44 m/s increase against LVP= 3.01 m/s) and September (15.3%
45 equivalent to 0.32 m/s increase against LVP= 2.16 m/s) where cases with Mean density range have the
46 highest wind speed at the rooftop of the target building. For all the other months, the highest wind speed
47
is observed for Max urban density cases, with the average higher speed of 15.2% in August (0.34 m/s
48
49 against LVP= 2.27 m/s), 14.7% in July (0.37 m/s against LVP= 2.56 m/s), 13.9% in April (0.44 m/s
50 against LVP= 3.21 m/s) and 14.3% in May (0.46 m/s against LVP= 3.25 m/s) compared to the LVP. It can
51 be noted that a lower urban density range results in a lower wind speed at the rooftop of the target
52 building. To get an overall picture about the impacts of urban densities, the six-month average wind speed
53 increase is compared to LVP in cases with Min urban density (2.74 m/s). The annual average values
54 (considering only six warm months of the year) are increased 10.6%, 10.9%, 13.9%, and 14.1%
55 respectively for cases with Middle, Norms, Mean and Max urban density compared to Min urban density
56 cases (1% is equivalent to 0.02 m/s).
57
58 One reason for these variations in wind velocity is the generation of different flow patterns across block
59 01 and block 02 buildings when wind approaches denser case studies (check Figure 25). This does not
60 happen for cases with low density or those in the P02 to P04 since the wind flow enters streets and vacant
61
62
25
63
64
65
 1
2
3
4 spaces around buildings. Denser surrounding buildings block the wind flow and separate the mainstream
5 by one or more stagnation points into two different flow fields; an intense one above the surrounding
6
7
buildings and a weaker one at windward façade of block 01 and block 02 (check Figure B.1 to 4 in
8 Appendix B).
9 4.2.2 Building forms and wind flow potential
10
11 The averaged wind speed at all directions for four introduced building forms (L, U, CY, and C) is plotted
12
13 in Figure 21-right to investigate the impacts of urban building form of the surrounding building on
14 ventilation potential during warm months. Results showed that the LVP in all months occurs in cases with
15 U form surrounding buildings. In comparison with the LVP, except the C urban building form in April
16 with 4.6% higher wind speed (0.15 m/s against LVP= 3.38 m/s), the CY form increases wind speed the
17 most for 7.8% in May (0.24 m/s compared to LVP= 3.33 m/s), 10.2% in June (0.28 m/s compared to
18 LVP= 3.0 m/s), 10.6% in July (0.23 m/s), 8.6% in August (0.17 m/s) and 10.8% in September (0.19 m/s).
19
20 To get an overall picture about the impact of urban building form, the Average is compared to LVP
21 (which occurs in U form cases). The annual average wind speed increases 6.8%, 8.3% and 7.4% for cases
22 with L, CY, and C surrounding building form compared to LVP (1% is equivalent to 0.028 m/s).
23 According to the results, the adjacent buildings around the target building (blocks 03 and 04) have a
24 major impact on wind velocity and wind speed in all directions at the rooftop level. A closer look into
25 cases with U urban building form indicated that flow fields are notably thinner – with lower wind speed in
26
27
all directions – than those of L and CY form cases which can directly affect cooling load calculations
28 through infiltration and openings (check figure 25). Moreover, windward buildings (blocks 01 and 02) in
29 U urban building form cases have a major role in reducing wind speed both on the rooftop and around the
30 target building by directing the flow into their open spaces (check Figure B.1 to 4 in Appendix B)
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46 Figure 21. Left: average wind velocity magnitude, out of CFD simulations, in term of Urban Density in six months
47
48 Right: average wind velocity magnitude, out of CFD simulations, in term of Urban Building Form in six months
49 4.2.3 Urban pattern and wind flow potential
50
51
To assess the impacts of the urban pattern on ventilation potential, average wind speed at all directions
52
53 for the four urban patterns (P01, P02, P03 and P04) is plotted in Figure 22-left for all the six warm
54 months. The LVP in all months occurs in P04 and the highest wind (at the rooftop of target building) in
55 P02. The difference between P04 and P02 cases is 4.5% in all months with different LVPs: 0.16 m/s in
56 April (LVP=3.40 m/s), 0.15 m/s in May (LVP=3.44 m/s), 0.14 m/s in June (LVP=3.15 m/s), 0.12% in July
57 (LVP=2.73 m/s), 0.11 m/s in August (LVP=2.39 m/s) and 0.11 m/s in September (LVP=3.40 m/s).
58 Moreover, cases in P04 and P01 have a similar wind speed in all directions with less than 1% difference
59 in each month. Wind speed at the rooftop of the target building in P03 showed 3% increase during all six
60
61
62
26
63
64
65
 1
2
3
4 months compared to cases in P04 (between 0.07 m/s and 0.10 m/s). The annual average wind speed in
5 P04 (in all directions and compared to LVP) is equal to 2.90 m/s which shows an increase for 0.09%,
6
7
4.5%, and 3% compared to P01, P02, and P03, respectively (1% is equivalent to 0.029m/s).
8 A deeper analysis of CFD contours reveals that narrow streets notably accelerate the airflow through the
9 urban fabric, in particular when wind velocity is higher than 2.88 m/s (check June, August and September
10 contours in Figure B.1 to 4 in Appendix B). Moreover, the intersect of the accelerated flows with
11
12
surrounding buildings and street causes eddies with many bending angles in northern and southern streets
13 around the target building and it can increase the wind speed at the rooftop of the target building. On the
14 other hand, when the street width increases to 6m in P03 (H/W=8), a considerable reduction occurs in
15 wind velocity in the fabric and rooftop of the target building. This reduction is more remarkable when the
16 street width increases to 8m in P04 (H/W=6) as the LVP in all months which is quite similar to the P01
17 case with no streets at north and south of the target building (check Figure B.1 to 4 in Appendix B).
18
19 For a better understanding of the relationship between the urban morphology and ventilation potential at
20 the rooftop of the target building, wind speed variations compared to the LVP are illustrated in Figure 22-
21 right. As it is clear, cases in Min urban density in P04 (P04 Min) have the lowest wind speed at the
22 rooftop of the target building (LVP=2.64 m/s). On the other hand, the highest wind speed at the rooftop of
23 the target building occurs in cases with Max urban density for P02 (P02 Max) with the average wind
24 speed of around 3.34 m/s (23.7% higher than the LVP). For cases in P02, all urban densities have the
25
highest wind speed compared to those of three other patterns. Cases with Mean and Max urban density
26
27 have the highest wind speed at the rooftop of the target building for all urban patterns (1% is equivalent to
28 0.026 m/s). According to the Figure 22, cases in P02 and P03 have the highest possible wind speed, while
29 cooling loads of cases in P01 and P04 are considerably lower compared to P02 and P03 (see Figure 20).
30 Thus, finding cases with lowest cooling load and highest wind speed out of 1600 cases studies in these
31 patterns is a difficult challenge; which is discussed in the next section.
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47 Figure 22. Left: average wind velocity magnitude, out of CFD simulations, in term of Urban Pattern in six months
48 Right: total average wind CFD velocity magnitude in term of Urban Density and Urban Patterns in all six months
49
50 4.3 Step 10: Best urban configurations
51
52 Impacts of urban morphology on reducing cooling demand and increasing ventilation potential are further
53 investigated for sixteen case studies with the lowest cooling demand (as the main priority) and highest
54 ventilation potential (highest average wind speed for all directions at the rooftop of the target building).
55
These sixteen best urban configurations are considered as the most efficient cases for each urban pattern
56
57 and urban building form category. These cases for each urban building form category (U, L, CY, and C)
58 in P01 to P04 with different urban density are illustrated in Figure 23 (check Figure 10 for main
59 parameters categorization). In addition to the cooling load and wind speed results (Figure 24), CFD
60
61
62
27
63
64
65
 1
2
3
4 contours of these best cases in April are also discussed in Figure 25 (check Figure B.1 to 4 in Appendix B
5 for CFD contours of best urban configurations in the six warm months).
6
7 Results of the best cases provides a deeper insight about the relation between urban morphology and the
8 objectives of the study (in practice) for engineers, designers and stakeholders. In Figure 24, cooling load
9 and wind speed of the best cases are compared to the “average cooling load” (ACL) and “average wind
10 speed” (AWS) of all 1600 cases and the differences are shown in percentages in both graphs (1% of
11
12
cooling load is 18.60×103kWh and 1% of wind speed is 0.03 m/s). In the first urban pattern (P01 cases),
13 P01LMax03 is the most efficient case study with the lowest cooling load with 29.2% reduction compared
14 to the ACL and highest average wind speed at the rooftop the target building with 0.08% increase against
15 AWS. Cooling load reduction for other cases in P01 are: 17.8% in P01CMax17 and P01UMax03 and
16 17.9% in P01CYMax02 compared to the ACL. On the other hand, except for P01UMax03 where wind
17 speed decreases notably (14.1% compared to the AWS), it increases for 2.9% in P01CMax17 and 3.4% in
18 P01CYMax02. A deeper investigation of P01 cases reveals that fifteen cases from U Max category are
19 among the Top 20 list with lowest cooling load and highest possible wind speed at the rooftop of the
20
target building.
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52 Figure 23. best urban configuration chosen in four urban patterns and urban building form
53 CFD simulation results (Figure 25) showed that during April, there is a stagnation point at the edge of
54
55 block 01 and 02 as the main obstacles in front of the prevailing flow (270 degrees) in P01LMax02. The
56 stagnated flow on the windward buildings is diverged into two kinds of flow; an upward flow with a
57 strong layer of turbulence that blows toward rooftop of the target building, and a downward flow towards
58 the front of the obstacle buildings. The upstream flow at the rooftop of the target building causes
59 maximum possible wind speed at all ten calculation points at 1.04H (H is the height of the target
60 building). As a result, a large amount of air is carried into the roof level and consequently into the placed
61
62
28
63
64
65
 1
2
3
4 hatches. The flow fields in P01LMax02, are generally similar for the other five months with lower wind
5 speeds. Moreover, in P01CYMax02 and P01UMax03 the windward buildings act as obstacles and
6
7
redirect the flow above the buildings to a higher level than the rooftop of the target buildings. In
8 P01CMax17, the approaching flow stocks in the front and lateral spaces of the windward buildings
9 (blocks 03 and 04) and only a portion of the flow field is directly disturbed on the rooftop of the target
10 building in comparison with L case.
11
12
In the second urban pattern, P02UMean04 with 23.3% cooling load reduction, compared to ACL, and
13 7.3% decreased wind speed on the rooftop of the target building, compared to AWS, is the most efficient
14 case. The cooling load of P02LMax02, P02CYMax02, and P02CMiddle05 shows 20.4%, 26.2% and
15 26.2% increase compared to the ACL. The wind speed at the rooftop of the target building in these three
16 cases is 7%, 0.8% and 7.3% (respectively) higher than AWS.
17
18 According to the CFD simulation results for the wind flow in P02CMiddle05 and P02CYMax02, an
19 intense flow field (with high mass flow rate and speed) is observed above the target building, particularly
20 in April, May, and June with higher wind velocity rates in X and Y directions (Figure 25). Although the
21 flow rates above the target building in P02LMax02 reach the maximum at 1.08H (the calculation points
22 are placed at 1.04H), during April, May, June, and July, wind speed is still 7% higher than average wind
23 speed. Furthermore, the flow fields in P02UMean04 clearly show the impact of U form surrounding
24 buildings on changing the direction of the flow over 1.2H and decreasing wind speed of the calculation
25
points at the rooftop of the building. It is also observed that wind flow rate and speed at lower parts of the
26
27 target buildings is considerably higher in L form case, which can negatively affect cooling load through
28 heat gain.
29 For the third urban pattern, P03UMax03 is the most efficient case with 17.8% cooling load reduction,
30
compared to ACL, and 0.08% decrease in wind speed at the rooftop of the target building, compared to
31
32 AWS. Except for P03CYMean01 with 2.3% cooling load reduction, the other two cases have a remarkably
33 higher cooling load, 26.2% higher than the ACL in P03CMiddle05 and 20.4% in P03LMax02. On the
34 other hand, wind speed for P03CYMean01, P03CMiddle05, and P03LMax02 is 5.1%, 4.9% and 1.8%
35 higher than AWS. This reveals the great impact of the urban pattern on cooling demand of the target
36 building.
37
38 According to the wind velocity contours in Figure 25, the considered points at 1.04H height reach the
39 highest possible wind speed in P03CYMean01 and P03CMiddle05. Although the upward flow toward the
40 target building in P03 has a thin layer of turbulence compared to P02 cases, wind speed is notably higher
41 than P01 cases. In these two cases, the approaching wind intersects with windward buildings (blocks 01
42 and 02) and a stagnation point occurs at the edge of the buildings, where the upward flow reaches the
43 rooftop of the target building. Moreover, the lateral flow from the stagnation point is directed towards the
44 entrance of streets through the endpoint of the site, while a downward flow at the front of windward
45
buildings is also observed. In P03LMax02 the main portion of flow field is distributed into the urban
46
47 fabric and around the target building at lower floors, while in P03UMax03 the main flow blows over the
48 calculation points (1.09H), which results in a considerable decrease in wind speed at the rooftop.
49 In the fourth urban pattern, P04UNorms19 and P04CMax01 with 17.6% and 17.7% cooling load
50
reduction, compared to ACL, and 20.8% and 0.5% lower wind speed, compared to AWS, are the most
51
52 efficient cases. On the other hand, the cooling load increases for 20.4% in P04CYMean01 and
53 P04LNorms13, showing that CY and L are not efficient options for P04. For these two cases, wind speed
54 increases for 8.1% in P04CYMean01 and decreases for 8.4% in P04LNorms13 (compared to AWS).
55
In the wind speed contours (Figure 25), for P04CYMean01 the stagnation points locate exactly at the
56
57 edges of the block 01 and block 02. The downward flow at the front of the windward buildings and the
58 lateral flow into the entry of streets from the stagnation region are higher than those of P01, P02, and P04
59 cases. Although a similar flow field is observed at P04CMax01 in April, May, and June, during July,
60 August and September, the simulated wind speed is considerably lower than the average wind speed.
61
62
29
63
64
65
 1
2
3
4 Another interesting point is the flow fields around L and CY forms at the lower floors and around the
5 target building which is significantly higher than those of U form cases which can affect cooling load as
6
7
discussed before.
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22 Figure 24. Left: comparison of ACL and cooling load of the best urban configuration,
23 Right: comparison of AWS and wind speed of the best urban configuration
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51 Figure 25. Wind speed contours for the best urban configuration in April
52 5. Discussion
53
54
According to the results of this comprehensive study, variations in the three considered parameters (urban
55
56 density, building form and pattern) can induce substantional decrease in cooling demand and increase in
57 ventilation potential. For urban density (including variables such as VAR, PAR, SC, λp and λf), a clear
58 trend was observed, in which the higher density resulted in lower monthly cooling load and higher
59 average wind speed in all directions at the rooftop of the target building, where up to 14.87×10 3 kWh
60 cooling load reduction was observed in July (compared to HCD), and 15.3% higher wind speed in
61
62
30
63
64
65
 1
2
3
4 September (compared to LVP). Moreover, the target building in a neighborhood with U form building has
5 the lowest possible cooling load compared to C, L, and CY forms, where up to 26.49 kWh reduction was
6
7
observed compared to HCD in September. This implies that constructing U form surrounding buildings
8 with high densities (Max, Mean or Norms) is the best option to reduce cooling load of buildings within a
9 neighborhood. On the other hand, up to 10.2% higher wind speed was observed in CY and C form cases
10 compared to LVP in U form cases during all the six warm months.
11
12
Furthermore, urban pattern was the most influencing parameter in this study, inducing up to 39.53×10 3
13 kWh (10.2%) cooling reduction compared to HCD in one month (June). It can imply that dense
14 neighborhoods with the lowest possible H/W ratio have the best opportunity to reduce cooling load;
15 however, the highest wind speed was observed at the P02 cases with lower density, thus finding efficient
16 cases with the lowest cooling load and highest wind speed at the rooftop was quite challenging. By
17 having wider streets (4m) in P02, the flow pattern changes notably due to variation of both urban pattern
18 and urban density. In P02 cases, the channeling effect between buildings and inflow distribution in urban
19 fabric was observed. Nevertheless, by adopting the introduced framework, a large number of best cases
20
have been generated in P01 and P02 cases.
21
22 5.1 Step 11: Top 35 and Top 100 urban configurations
23
24 To further understand the results for real application and practice for designers, Figure 26 illustrates the
25 statistics of Top 100 urban configurations with the lowest cooling demand and highest ventilation
26 potential in terms of three main parameters of the study, and the plan configuration of thirty-five
27 suggested plans in Top 35 (in addition to sixteen best cases), which are categorized based on urban
28 building forms. As it is clear, among Top 100 cases, 61% were from Max and Norms density, 15% from
29 each Min and Middle density category and only 9% from Mean urban density. Among the first-twenty top
30 cases, twelve cases were from Max, three cases from Mean and five case studies from Norms density.
31
Moreover, all thirty-one Max cases (with different urban forms on P01) were among the first top-sixty
32
33 cases. About 76% of Top 100 cases had U or C urban building forms, 20% for L forms and only 5% had
34 CY forms. According to the results, 21 U form cases with Max density where between Top 100. It can be
35 concluded that case studies with Max and Norms urban density and U or C urban forms have high
36 efficiency potential in term of cooling load reduction and ventilation potential. Furthermore, 58% of Top
37 100 case studies were on P01, 27% on P02, and thirty-four cases in Top 35 cases had P01 urban pattern.
38 In the other words, case studies with Max density which were placed on P01 with U or C urban building
39 forms have the highest potential to be efficient both in cooling demand and ventilation potential.
40
41 These configurations can be a resource for urban designers and architects to design efficient architectural
42 plans. An overview of Figure 26 shows that cases with coherent cells and compact shapes have the
43 highest number on Top 35 urban configurations. Moreover, it is clear that for cases with non-compact
44 forms, the empty cells are placed on the windward side. In other words, the lowest possible mass and
45 voids in each block is suggested in the process of architectural plan designing. Another interesting point is
46
47
that cases with lower area of external surfaces and façade, compared to their volumes, show more
48 efficient results in terms of the objectives of this study. By deeper investigation of Top 100 urban
49 configurations, it is suggested that designers try to reduce push and pull in the architectural plans as much
50 as possible, but if it is required to consider these variations and fractions in line with architect’s idea,
51 cases with fractions and empty cells on the leeward side of urban buildings are more energy efficient.
52
53
54
55
56
57
58
59
60
61
62
31
63
64
65
 1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37 Figure 26. Left: Statistics of Top 100 cases in three parameters of the study
38 Right: Some of efficient cases among Top 35 cases in each urban building form category
39
5.2 Step 12: Verifying results
40
41
42 As an attempt to verify our results, Table 3 illustrates a number of relevant published studies (including
43 their methods, approaches, tools and main parameters) which are in good agreement with our findings,
44 considering energy demand, thermal behavior, ventilation and wind flow around buildings in urban areas.
45 Most of the studies have focused only on one or two parameters for evaluating an objective based on
46 some generic configurations regardless of real and design-based geometry. Moreiver, most of published
47 literature studies in urban-scale building simulation have not addressed air pollution issue. However, the
48 current study has considered all the influencing parameters in the literature, in addition to some new
49
features, trying to model the morphology of Tehran with the highest possible accuracy and providing an
50
51 extensive database for real applications.
52 Table 3: A number of studies with good agreement as compared to the results of this paper (Approaches are
53 hypothetical and real-site configurations, Methods are NS: numerical simulation, EX: experimental study, RM: Real-
54 site measurement, O-R: Observations on real data)
55
56 Description
Studies Year Considering Air
57 Ref Approach methods Tools Main parameters
pollution
58 Huang and Li, 2017 [14] Hypothetical NS ENVI-ment, EnergyPlus H/W, Density -
59 Salvati et al, 2017 [42]
Real-site\
NS EnergyPlus
Compactness,
-
60 Hypothetical Geometry
61
62
32
63
64
65
 1
2
3
4 Perera et al, 2018 [43] Hypothetical NS CitySim, CIM Density, Height -
5 Zinzi et al, 2018 [59] Real-site RM
Data loggers and climate
Density -
6 stations
Real-site\ ENVI-ment, Ecotect,
7 Taleghani et al, 2014 [72]
Hypothetical
NS/ RM
Design Builder
Geometry -
8 Real-site\ NS Density, Geometry,
9 Samuelson et al, 2016 [73] EnergyPlus, Archsim -
Hypothetical H/W
10 Yang et al, 2016 [77] Real-site NS/ RM ANSYS CFD
Density, Height,
-
11 Geometry
Chatzidimitrioua Real-site\
12 2017 [88] NS ENVI-ment, RayMan H/W -
and Yannas Hypothetical
13 Compactness,
14 Mohajeri et al, 2016 [89] Real-site NS/ RM CitySim, RADIANCE -
Density
15 NS
Screening Tool for Estate
Density, Height,
16 Wong et al, 2011 [90] Generic Environment Evaluation -
Green spaces
(STEVE)
17
NS/E Density, Surface
18 Montavon et al, 2004 [126] Real-site Ray tracing technique -
area
19 NS Density, Shading,
Li et al, 2015 [127] Hypothetical CitySim, RADIANCE -
20 Cluster
21 NS Façade, Roof,
Sarralde et al, 2015 [128] Real-site CitySim, SUNtool -
Density
22
Ramponi et al, 2015 [129] Hypothetical NS/E CFD, LES, RANS Density, H/W -
23 Density, Geometry,
24 Kontokosta and Tull, 2017 [130] Real-site O-R MatLab -
Floor area
25 Urban form,
26 Silva et al, 2017 [131] Real-site O-R Arc GIS Diversity of -
activities
27
Compactness,
28 Eicker et al, 2015 [132] Hypothetical NS EnergyPlus -
Geometry
29 Urban Energy Index for
Rodrigues, 2016 [133] Real-site NS Geometry -
30 Buildings (UEIB)
31 Hong and Lin, 2015 [134] Hypothetical NS CFD Geometry, Shading -
32 Real-site\ EnergyPlus, RADIANCE,
Nikkho et al, 2017 [135] NS/ RM H/W, Density -
Hypothetical OpenFOAM, Open Studio
33 Weather Research and
34 Rosado et al, 2017 [136] Hypothetical NS H/W, Geometry -
Forecasting (WRF)
35 Wen et al, 2007 [137]
Real-site\
NS/ RM CFD H/W, Geometry -
36 Hypothetical
37
38
39 6. Conclusions
40
41 This study thoroughly investigated the impacts of urban morphology on reducing cooling load and
42 increasing ventilation potential in hot-arid climate of Tehran in four main steps: modeling 1600 urban
43 morphology case studies; validation study; cooling load and ventilation potential simulations; and
44 analysis of the best sixteen urban configurations. In this respect, a twelve-story building (40×40×48 m)
45 with a simplified open-plan office was set up as the target building based on the building codes of Iran. A
46 novel approach, namely “Building Modular Cell” (BMC), introduced to model 1600 urban morphology
47 case studies considering three major parameters: urban density (Min, Middle. Norms, Mean and Max);
48
49
urban building form (“U”, “L”, “C” or cubic and “CY” or courtyard forms); and urban pattern one to four
50 (P01-P04). In addition, several variables and influencing indicators considered in the process of modeling
51 the target and surrounding buildings. Reducing the cooling demand and increasing the ventilation
52 potential of the target buildings were the main objectives of the study.
53
54
The cooling load of the target building and the average wind speed at its rooftop (for all directions) were
55 simulated individually for all 1600 case studies, by means of several validated tools, for six warm months
56 of the year (April, May, June, July, August, and September). All case studies were assessed in term of
57 cooing demand and ventilation potential, where the “highest cooling demand” (HCD) and “lowest
58 ventilation potential” (LVP) were marked in each month and other cases were compared to these
59 indicators. Moreover, sixteen best urban configurations were selected in terms of the lowest cooling
60 demand and highest ventilation potential in each urban density and building form category. To obtain a
61
62
33
63
64
65
 1
2
3
4 clear overview, the cooling load and wind speed at rooftop of these best cases were compared to “average
5 cooling load” and “average wind speed” of all 1600 cases. In the discussion section, characteristics of Top
6
7
100 efficient cases were investigated individually and Top 35 urban configurations were presented in the
8 layouts (in addition to sixteen best urban configurations) in order to derive the best design-based
9 suggestions. Results clearly demonstrated the important effects of urban morphology in improving the
10 cooling and ventilation conditions for buildings within urban areas. According to the results, the proposed
11 method is in good agreement with the major studies, while shortens the process time and enhances the
12 accuracy of the assessment. Moreover, several unique findings were obtained with potential applications
13 for researchers, engineers, designers and stakeholders. Some important conclusions and recommendation
14 out of results of this study and analysis of best urban configuration geometry, with the potential of being
15
applied during the early stage design, are as the following:
16
17  Urban Density
18 o
Findings
19  HCD occurred in urban configurations with Min urban density in all six warm months and the
20
21
average cooling demand.
22  Highest HCD reduction occurred in June, where Max urban density cases showed 4.5% reduction
23 (16.66×103 kWh) compared to Min urban density cases (370.39×103 kWh).
24  Average HCD reduction in cases with Middle, Norms, Mean and Max urban density are 0.2%,
25 0.3%, 1.8% and 3.9% respectively compared to Min urban density cases (1% is equivalent to
26 3.11×103 kWh).
27  LVP occurred in cases with Min urban density in all six warm months and the average ventilation
28 potential.
29
 Max urban density cases showed 15.2% wind speed increase (0.34 m/s) compared to LVP in Min
30
31 urban density cases (2.27 m/s).
32  Average wind speed in cases with Middle, Norms, Mean and Max urban density increased 10.6%,
33 10.9%, 13.9% and 14.1% respectively compared to LVP in Min urban density cases (2.74 m/s).
34 o Recommended
35  To design newly-built neighborhoods and buildings based on higher urban density (Norms and
36 Max urban density have 61 cases among top 100), because higher urban density parameter (both
37 site coverage and height) results in lower cooling load for buildings within a neighborhood placed
38 in hot-arid climate.
39
40
 To consider normal built density (50-60% site coverage) in a situation that all the neighborhood
41 buildings should have low-rise buildings, regarding the relation between site coverage and height
42 in urban density parameter.
43 o Not-recommended
44  To design neighborhoods with low urban density and since the cooling demand is significantly
45 higher in cases with Min and Middle urban density ranges.
46  To design large private open spaces in the hot-arid climate to fulfill requirements of standard green
47 space per capita (according to regulations for a neighborhood), it is better to consider an open
48 green space in each neighborhood as a public place and reduce private open spaces area.
49
50  Urban Building Form
51 o Findings
52  HCD occurred in cases with CY urban building form in August and July, and cases with L urban
53 building form in April, May, September, June and average cooling demand.
54  Highest HCD reduction occurred in April, where cases with U urban building form showed 8.2%
55 reduction (13.01×103 kWh) compared to L urban building form cases (158.59×103 kWh).
56  Average HCD reduction in cases with CY, C and U urban building form are 0.4%, 2.3% and 3.9%
57
compared to configurations with L urban building form (1% is equivalent to 3.46×103 kWh).
58
59  LVP occurred in cases with U urban building form in all six warm months and the average
60 ventilation potential.
61
62
34
63
64
65
 1
2
3
4  Urban configurations with CY form buildings showed 10.8% wind speed increase (0.23 m/s) in
5 September compared to LVP in cases with U urban building form (2.21 m/s).
6
7
 Average wind speed in cases with L, CY and U urban building form increased 6.8%, 8.3% and
8 7.4% respectively compared to lowest ventilation potential in configurations with U form buildings
9 (2.81 m/s).
10 o Recommended
11  To use efficient L form layouts presented in the best urban configurations (sixteen and Top 35), or
12 at least similar forms for designing the overall architectural plans at the early stage (Although the
13 monthly average cooling load of L form buildings is higher than to be called efficient, with some
14 consideration it might be used as a basic form for building design (20 L form cases observed
15
among Top 100)).
16
17  To consider push and pull in the architectural plan ( if it is required in line with the architect’s
18 ideas) on the eastern side of the buildings as results showed a significant cooling load reduction in
19 cases with empty cells in their eastern side.
20 o Not-recommended
21  To use CY form buildings as the only form in neighborhoods; there were only five pure courtyard
22 cases among Top 100. Also, average cooling load of CY form buildings are higher than other
23 analyzed forms. However, averaged wind speed at the rooftop in these cases were much higher
24 than other cases (in the case of designing courtyard buildings, deeper investigation for each case is
25
26
required).
27  To design forms with high area of external surfaces and façade because of the potential for high
28 heat transfer through surfaces and consequently higher cooling load.
29  Urban pattern
30 o Findings
31  HCD occurred in cases in P03 in all months, except for May where HCD occurred in P02 cases.
32  Highest HCD reduction occurred in June, where cases in P01 showed 10.2% reduction
33 (39.530×103 kWh) compared to urban configurations in P03.
34
35
 Average HCD reduction occurred in urban configuration in P04, P02 and P01 was 3.7%, 2.3% and
36 6.8% compared to cases in P03 (1% is equivalent to 3.31×103 kWh).
37  LVP occurred in cases in urban pattern four in all six warm months and the average LVP.
38  Cases placed on P02 showed 4.6% wind speed increase in June (0.14 m/s) compared to urban
39 configurations in P04 (3.15 m/s).
40  LVP in P03, P02 and P01 increased 3%, 4.5% and 0.09% compared to lowest ventilation potential
41 in cases placed on P04 (2.90 m/s).
42 o Recommended
43
 To design denser neighborhood with lower width streets around buildings (H/W (height to width
44
45 ratio) of 12 or higher for design newly-built neighborhoods.
46  To consider urban streets and canopies with H/W of 6 to 8, which has the highest potential for
47 natural ventilation through openings in the cities with no air pollution issues (in this study main
48 opening were assumed to be closed considering severe air pollution of Tehran).
49 o Not-recommended
50  To design detached buildings in a neighborhood with Hot-arid climate considering air pollution
51 (while cases with H/W lower than 12, showed a significant decrease in ventilation potential).
52  To consider the building as a non-coherent form and to place vacant cells on the western side next
53
to the possible surrounding streets (lowest possible mass and voids in each block is suggested).
54
55  General
56  It is recommended to use the “building modular cells” as a technique for architects and urban
57 designers to control cooling demand and ventilation potential in different design problems.
58 However, the database of the results (layouts and graphs) is also can be used as a reference for
59 designers by adapting their designed urban forms in the early stages to the best configurations.
60
61
62
35
63
64
65
 1
2
3
4  Applying suggested fenestration by the national building code in Iran was effective in cooling load
5 reduction, however, further research should be conducted to optimize openings in high-rise
6
7
buildings.
8  The hatches of inlet air in this system, should be placed on the centerline of the building 0.1w to
9 0.6w (w is the width of the target building). However, further investigation is required to design
10 such an active-passive ventilation system in order to maximize the efficiency of the process.
11
12
Compared to other studies on urban morphologies, this work focused on comprehensive design-related
13 parameters based on a parametric modeling process by a novel approach and technique. The proposed
14 method and results are in good agreement with the major studies, while shortens the process time and
15 enhances the accuracy of the assessment. Several unique findings were obtained with potential
16 applications for researchers, engineers, designers and stakeholders considering air pollution issue at whole
17 process. In the real practice, designers tend to focus on their concepts as their main priority and design
18 urban and building forms regardless of its thermal and aerodynamic behavior. The results of this study are
19 a technical and graphical reference for designers as a guide through the design process. The introduced
20
method and workflow can be adopted by researchers, engineers and designers for study cases and
21
22 developing urban plans, codes and regulations and enables to bridge the gap between engineers and urban
23 development beneficiaries (architects, urban designers and city authorities) and to initiate further
24 investigations on the topic in the global-scale. The derived design-based suggestions out of this extensive
25 database can shape the urban and building form in the newly-built neighborhoods in Tehran or cities with
26 similar climate.
27
28
29 References
30
31 [1] "World urbanization prospects: The 2014 revision," United Nations, New York, 2014.
32 [2] N. Grimm, S. Faeth, N. Golubiewski, C. Redman, J. Wu, X. Bai and et al, "Global change and the ecology of cities,"
33 Science, vol. 319, no. 5864, pp. 756-60, 2008.
34
[3] L. Frayssinet, L. Merlier, F. Kuznik, J. Hubert and M. Milliez, "Modeling the heating and cooling energy demand of urban
35 buildings at city scale," Renewable and Sustainable Energy Reviews, vol. in press, no. in press, p. in press, 2017.
36
[4] B. Cohen, "Urbanization in developing countries: current trends, future projections and key challenges for sustainability,"
37
Technology in society, vol. 28, no. 1, pp. 83-80, 2006.
38
39 [5] J. Srebric, M. Heidarinejad and J. Liu, "Building neighborhood emerging properties and their impacts on multi-scale
modeling of building energy and airflows," Building and Environment, vol. 91, pp. 246-262, 2015.
40
41 [6] Y. Wu, P. Krishnan, L. E.Yu and M. Zhang, "Using lightweight cement composite and photocatalytic coating to reduce
42 cooling energy consumption of buildings," Construction and Building Materials, vol. 145, pp. 555-564, 2017.
43 [7] S. Gou, V. Nik, J. Scartezzini, Q. Zhao and Z. Li, "Passive design optimization of newly-built residential buildings in
44 Shanghai for improving indoor thermal comfort while reducing building energy demand," Energy and Building, vol. in
45 press, p. In press, 2018.
46 [8] D. Dominković, K. Abdul Rashid, A. Romagnoli, A. Pedersen, K. Leong, G. Krajačić and N. Duić, "Potential of district
47 cooling in hot and humid climates," Applied Energy, vol. 208, pp. 49-61, 2017.
48 [9] V. M. Nik and A. Sasic Kalagasidis, "Impact Study of the Climate Change on the Energy Performance of the Building
49 Stock in Stockholm Considering Four Climate Uncertainties," Building and Environment, vol. 60, pp. 291-304, 2013.
50 [10] V. M. Nik, Hygrothermal Simulations of Buildings Concerning Uncertainties of the Future Climate, Gothenburg, Sweden:
51 Chalmers University of Technology, 2012.
52 [11] M. Santamouris, "On the energy impact of urban heat island and global warming on," Energy and Buildings, vol. 82, pp.
53 100-113, 2014.
54 [12] V. Nik, "Making energy simulation easier for future climate- Synthesizing typical and extreme weather data sets out for
55 regional climate models," Applied Energy, vol. 177, pp. 204-226, 2016.
56 [13] E. Kalnay and M. Cai, "Impact of urbanization and land-use change on climate," Nature, vol. 423, no. 6939, pp. 528-31,
57 2003.
58 [14] K. Huang and Y. Li, "Impact of street canyon typology on building’s peak cooling energydemand: A parametric analysis
59 using orthogonal experiment," Energy and Buildings, vol. 154, pp. 448-464, 2017.
60
61
62
36
63
64
65
 1
2
3
4 [15] C. Hsieh, J. Li, L. Zhang and B. Schwegler, "Effects of tree shading and transpiration on building cooling energy use,"
5 Energy and Buildings, vol. 159, pp. 382-397, 2018.
6 [16] P. a. H. Censuses, "CENSUS," Statistical Center of Iran, Tehran, 2016.
7
[17] M. Mahdavinejad and K. Javanroodi, "Natural ventilation performance of ancient wind catchers, an experimental and
8
analytical study case studies: one-sided, two-sided and four-sided wind catchers," International Journal of Energy
9 Technology and Policy, vol. 10, no. 1, pp. 36-60, 2014.
10
[18] E. Pourazam and A. Cooray, "Estimating and forecasting residential electricity demand in Iran," Economic Modelling, vol.
11
35, pp. 546-558, 2015.
12
13 [19] I. E. E. Organization, "Annual Energy Fact Sheet," IEEO, Tehran, 2016.
14 [20] H. Vafa-Arani, S. Jahani, H. Dashti, J. Heydari and S. Moazen, "A system dynamics modeling for urban air pollution: A
15 case study of Tehran, Iran," Transportation Research Part D, vol. 31, pp. 21-36, 2014.
16 [21] P. Pahlavani, H. Sheikhian and B. Bigdeli, "Assessment of an air pollution monitoring network to generate urban air
17 pollution maps using Shannon information index, fuzzy overlay,and Dempster-Shafer theory, A case study: Tehran, Iran,"
18 Atmospheric Environment, vol. 167, pp. 254-269, 2017.
19 [22] E. Krüger, "Impact of site-specific morphology on outdoor thermal perception: A case-study in a subtropical location,"
20 Urban Climate, vol. 21, pp. 123-35, 2017.
21 [23] J. An, D. Yan, T. Hong and K. Sun, "A novel stochastic modeling method to simulate cooling loads in residential districts,"
22 Applied Energy, vol. 206, pp. 134-149, 2017.
23 [24] J. Hang, M. Sandberg and Y. Li, "Effect of urban morphology on wind condition in idealized city models," Atmospheric
24 Environment, vol. 43, pp. 869-87, 2009.
25 [25] S. M. W. U. C. L. A. Skote M, "Numerical and experimental studies of wind environment in an urban morphology,"
26 Atmospheric Environment, vol. 39, pp. 6147-58, 2005.
27 [26] H. Campaniço, P. Hollmuller and P. Soares, "Assessing energy savings in cooling demand of buildings using passive
28 cooling systems based on ventilation," Applied Energy, vol. 134, pp. 426-438, 2014.
29 [27] S. Khoshdel Nikkho, M. Heidarinejad, J. Liu and J. Srebric, "Quantifying the impact of urban wind sheltering on the
30 building energy consumption," Applied Thermal Engineering , vol. 116, pp. 850-65, 2017.
31
[28] A. Razak, A. Hagishima, N. lkegaya and J. Tanimoto, "Analysis of airflow over building arrays for assessment of urban
32 wind environment," Building and Environment, vol. 59, pp. 56-65, 2013.
33
[29] H. Taleb and M. Musleh, "Applying urban parametric design optimisation processes to a hot climate: Case study of the
34
UAE," Sustainable Cities and Society, vol. 14, pp. 236-253, 2015.
35
36 [30] K. Al-Sallal and L. Al-Rais, "Outdoor airflow analysis and potential for passive cooling in the traditional urban context of
Dubai," Renewable Energy, vol. 36, pp. 2494-2501, 2011.
37
38 [31] T. Kim, K. Kim and B. Kim, "A wind tunnel experiment and CFD analysis on airflow performance of enclosed-arcade
39 markets in Korea," Building and Environment, vol. 45, pp. 1329-38, 2010.
40 [32] A. Tablada, F. Troyera, B. Blocken, J. Carmeliet and H. Verschure, "On natural ventilation and thermal comfort in
41 compact urban environments in the Old Havana case," Building and Environment, vol. 44, pp. 1943-58, 2009.
42 [33] A. Dallman, S. Magnusson, R. Britter, L. Norford, D. Entekhabi and H. Fernando, "Conditions for thermal circulation in
43 urban street canyons," Building and Environment, vol. 80, pp. 184-191, 2014.
44 [34] "Wind tunnel simulation studies on dispersion at urban street canyons and intersections-a review," Journal of Wind
45 Engineering and Industrial Aerodynamics, vol. 93, pp. 697-717, 2005.
46 [35] R. Memon, D. Leung and D. Liu, "Effects of building aspect ratio and wind speed on air temperatures in urban-like street
47 canyons," Building and Environment, vol. 45, pp. 176-88, 2010.
48 [36] T. Ihara, Y. Kikegawa, K. Asahi, Y. Genchi and H. Kondo, "Changes in year-round air temperature and annual energy
49 consumption in office building areas by urban heat-island countermeasures and energy-saving measures," Applied Energy,
50 vol. 85, no. 1, pp. 12-25, 2008.
51 [37] M. Mourshed, "The impact of the projected changes in temperature on heating and cooling requirements in buildings in
52 Dhaka, Bangladesh," Applied Energy, vol. 88, p. 3737–3746, 2011.
53 [38] W. Mohd Nazi, M. Royapoor, Y. Wang and A. P. Roskilly, "Office building cooling load reduction using thermal analysis
54 method – A case study," Applied Energy, vol. 85, pp. 1574-1585, 2017.
55 [39] J. Keirstead, M. Jennings and A. Sivakumar, "A review of urban energy system models: approaches, challenges and
56 opportunities," Renewable and Sustainable Energy Reviews , vol. 16, no. 6, pp. 3847-66, 2012.
57
[40] M. Kavgic, A. Mavrogianni, D. Mumovic, A. Summerfield, Z. Stevanovic and M. Djurovic-Petrovic, "A review of bottom-
58 up building stock models for energy consumption in the residential sector," Building and Environment, vol. 45, no. 7, pp.
59 1683-97, 2010.
60
61
62
37
63
64
65
 1
2
3
4 [41] F. de la Flor and S. Dominguez, "Modelling microclimate in urban environments and assessing its influence on the
5 performance of surrounding buildings," Energy and Building, vol. 36, no. 5, pp. 403-13, 2004.
6 [42] A. Salvati, H. Coch and M. Morganti, "Effects of Urban compactness on the building energy performance in
7 Mediterranean climate," in CISBAT, Lausanne, 2017.
8
[43] A. Perera, S. Coccolo, J. Scartezzini and D. Mauree, "Quantifying the impact of urban climate by extending the boundaries
9 of urban energy system modeling," Applied Energy, vol. 222, pp. 847-860, 2018.
10
[44] H. Montazeri and F. Montazeri, "CFD simulation of cross-ventilation in buildings using rooftop wind-catchers: Impact of
11
outlet openings," Renewable Energy, vol. 118, pp. 502-20, 2018.
12
13 [45] S. Hosseini, E. Shokry, A. Ahmadian Hosseini, G. Ahmadi and J. Calautit, "Evaluation of airflow and thermal comfort in
buildings ventilated with wind catchers: Simulation of conditions in Yazd City, Iran," Energy for Sustainable
14
Development, vol. 35, pp. 7-24, 2016.
15
16 [46] M. Mahdavinejad and K. Javanroodi, "Impact of roof shape on air pressure, wind flow and indoor temperature of
residential buildings," International Journal of Sustainable Building Technology and Urban Development, vol. 7, no. 2, pp.
17
87-103, 2016.
18
19 [47] A. Faghih and M. Bahadori, "Thermal performance evaluation of domed roofs," Energy and Buildings, vol. 43, no. 6, pp.
1254-1263, 2011.
20
21 [48] N. Nasrollahi, M. Hatami, S. RazieyehKhastar and M. Taleghani, "Numerical evaluation of thermal comfort in traditional
22 courtyards to develop new microclimate design in a hot and dry climate," Sustainable Cities and Society, vol. 35, pp. 449-
467, 2017.
23
24 [49] F. Soflaei, M. Shokouhian, H. Abraveshdar and A. Alipour, "The impact of courtyard design variants on shading
25 performance in hot- arid climates of Iran," Energy and Buildings, vol. 143, pp. 71-83, 2017.
26 [50] A. Mohammadi, M. RezaSaghafi, M. Tahbaz and F. Nasrollahi, "The study of climate-responsive solutions in traditional
27 dwellings of Bushehr City in Southern Iran," Journal of Building Engineering, vol. 16, pp. 169-183, 2018.
28 [51] M. Khalili and S. Amindeldar, "Traditional solutions in low energy buildings of hot-arid regions of Iran," Sustainable
29 Cities and Society, vol. 13, pp. 171-181, 2014.
30 [52] Construction Engineering Organization, National Building Codes of Iran (19), Tehran: Ministry of Roads & Urban
31 Development, 2009.
32 [53] S. Verbeke and A. Audenaert, "Thermal inertia in buildings: A review of impacts across climate and building use,"
33 Renewable and Sustainable Energy Reviews, p. In press, 2018.
34 [54] D. V. C. J. Moonen P, "Evaluation of the ventilation potential of courtyards and urban street canyons using RANS and
35 LES," Journal of Wind Engineering , vol. 99, pp. 414-23, 2011.
36 [55] X. Li, C. Liu and D. Leung, "Development of a k-e model for the determination of air exchange rates for street canyons,"
37 Atmospheric Environment, vol. 39, pp. 7285-96, 2005.
38 [56] Y. Chen, T. Hong and M. A. Piette, "Automatic generation and simulation of urban building energy models based on city
39 datasets for city-scale building retrofit analysis," Applied Energy, vol. 205, pp. 323-335, 2017.
40
[57] J. Allegrini and J. Carmeliet, "Simulations of local heat islands in Zurich with coupled CFD and building energy models,"
41 Urban Climate, vol. in press, p. in press, 2017.
42
[58] Y. Tominaga, "Visualization of city breathability based on CFD technique: case study for ur-ban blocks in Niigata City,"
43
Journal Visualization, vol. 15, pp. 239-276, 2012.
44
45 [59] M. Zinzi, E. Carnielo and B. Mattoni, "On the relation between urban climate and energy performance of buildings. A
three-years experience in Rome, Italy," Applied Energy, vol. 221, pp. 148-160, 2018.
46
47 [60] T. Van Hooff and B. Blocken, "Coupled urban wind flow and indoor natural ventilation modelling on a high-resolution
48 grid: a case study for the Amsterdam ArenA stadium," Environmental Modeling and Software, vol. 25, pp. 51-65, 2010.
49 [61] L. Z. S. M. G. j. Hang J, "Natural ventilation assessment in typical open and semi-open urban environments under various
50 wind conditions," Building and Environment, vol. 70, pp. 318-33, 2013.
51 [62] H. Chen, R. Ooka, H. Huang and T. Tsuchiya, "Study on mitigation measures for outdoor thermal environment on present
52 urban blocks in Tokyo using coupled simulation," Building and Environment, vol. 44, no. 11, pp. 2290-9, 2009.
53 [63] Y. Shimazaki, A. Yoshida, R. Suzuki, T. Kawabata, D. Imai and S. Kinoshita, "Application of human thermal load into
54 unsteady condition for improvement of outdoor thermal comfort," Building and Environment, vol. 46, no. 8, pp. 1716-
55 1724, 2011.
56 [64] T. Farea, D. Ossen, S. Alkaff and H. Kotani, "CFD modeling for natural ventilation in a lightwell connected to outdoor
57 through horizontal voids," Energy and Building, vol. 86, pp. 502-513, 2015.
58 [65] B. Hong and B. Lin, "Numerical studies of the outdoor wind environment and thermal comfort at pedestrian level in
59 housing blocks with different building layout patterns and trees arrangement," Renewable Energy, vol. 73, pp. 18-27, 2015.
60 [66] C. Gromke, B. Blocken, B. Janssen, W. Merema, B. van Hooff and H. Timmermans, "CFD analysis of transpirational
61
62
38
63
64
65
 1
2
3
4 cooling by vegetation: case study for specific meteorological conditions during a heat wave in Arnhem, Netherlands.,"
5 Building and Environment, vol. 83, pp. 11-26, 2015.
6 [67] Y. Toparlar, B. Blocken, P. Vos, G. Van Heijst, W. Janssen, T. van Hooff, H. Montazeri and H. Timmermans, "CFD
7 simulation and validation of urban microclimate: a case study for Bergpolder Zuid, Rotterdam," Building and Environment,
8 vol. 83, pp. 79-90, 2015.
9 [68] J. Allegrini, V. Dorer and J. Carmeliet, "Influence of morphologies on the microclimate in urban neighbourhoods," Journal
10 of Wind Engineering and Industrial Aerodynamics, vol. 144, pp. 108-117, 2015.
11
[69] J. Allegrini and J. Carmeliet, "Coupled CFD and building energy simulations for studying the impacts of building height
12 topology and buoyancy on local urban microclimates," Urban Climate, vol. 21, pp. 278-305, 2017.
13
[70] E. Wurtz, L. Mora and C. Inard, "An equation-based simulation environment to investigate fast building simulation,"
14
Building and Environment, vol. 41, no. 11, pp. 1571-1583, 2006.
15
16 [71] K. Perini and A. Magliocco, "Effects of vegetation, urban density, building height, and atmospheric conditions on local
temperatures and thermal comfort," Urban Forestry & Urban Greening, vol. 13, no. 3, pp. 495-506, 2014.
17
18 [72] M. Taleghani, L. Kleerekoper, M. Tenpierik and A. Van den Dobbelsteen, "Outdoor thermal comfort within five different
19 urban forms in the Netherlands," Building and Environment, vol. 83, pp. 65-78, 2014.
20 [73] H. Samuelson, S. Claussnitzer, A. Goyal, Y. Chen and a. Romo-Castillo, "Parametric energy simulation in early design:
21 High-rise residential buildings in urban contexts," Building and Environment, vol. 101, pp. 19-31, 2016.
22 [74] A. Vartholomaios, "A parametric sensitivity analysis of the influence of urban form ondomestic energy consumption for
23 heating and cooling in aMediterranean city," Sustainable Cities and Society, vol. 28, pp. 135-145, 2017.
24 [75] E. Elbeltagi, H. Wefki, S. Abdrabou, M. Dawood and A. Ramzy, "Visualized strategy for predicting buildings energy
25 consumption during early design stage using parametric analysis," Journal of Building Engineering , vol. 13, pp. 127-136,
26 2017.
27 [76] S. Vardoulakis, B. FIsher, K. Pericleous and N. Gonzalez-Flesca, "Modelling air quality in street canyons: a review,"
28 Atomespheric Environment, vol. 37, no. 2, pp. 155-182, 2003.
29 [77] A. Yang, Y. Su, C. Wen, Y. Juan, W. Wang and C. Cheng, "Estimation of wind power generation in dense urban area,"
30 Applied Energy, vol. 171, pp. 213-230, 2016.
31 [78] M. Taleghani, M. Tenpierik, A. Van Den Dobbelsteen and R. De Dear, "Energyuse impact of and thermal comfort in
32 different urban block types in the Netherlands," Energy and Buildings, vol. 67, pp. 166-175, 2013.
33 [79] J. Strømann-Andersen and P. Sattrup, "The urban canyon and buildingenergy use: Urban density versus daylight and
34 passive solar gains," Energy andBuildings, vol. 43, pp. 2011-2020, 2011.
35 [80] S. Gracik, M. Heidarinejad, J. Liu and J. Srebric, "Effect of urban neighborhoods on the performance of building cooling
36 systems," Building and Environment, vol. 90, pp. 15-29, 2015.
37
[81] Y. Chen and T. Hong, "Impacts of building geometry modeling methods on the simulation results of urban building energy
38 models," Applied Energy, vol. 215, pp. 717-735, 2018.
39
[82] J. Allegrini, V. Dorer and J. Carmeliet, "Influence of the urban microclimate in street canyons on the energy demand for
40
space cooling," Building and Energy, vol. 55, pp. 823-832, 2012.
41
42 [83] K. Niachou, S. Hassid, M. Santamouris and I. Livada, "Experimental performance investigation of natural, mechanical and
hybrid ventilation in urban environment," Building and Environment, vol. 43, no. 8, pp. 1373-1382, 2008.
43
44 [84] K. Perez, W. Cole, J. Rhodes, A. Ondeck, M. Webber and M. Baldea, "Nonintrusive disaggregation of residential air-
45 conditioning loads from sub-hourly smart meter data," Energy and Building, vol. 81, pp. 316-25, 2014.
46 [85] R. Evins, V. Dorer and J. Carmeliet, "Simulating external longwave radiation exchange for buildings," Energy and
47 Buildings, vol. 75, pp. 472-82, 2014.
48 [86] T. Ichinose, L. Lei and Y. Lin, "Impacts of shading effect from nearby buildings on heating and cooling energy
49 consumption in hot summer and cold winter zone of China," Energy and Buildings, vol. 136, pp. 199-210, 2017.
50 [87] Z. H. Wang, X. Zhao, J. Yang and J. Song, "Cooling and energy saving potentials of shade trees and urban lawns in a
51 desert city," Applied Energy, vol. 161, pp. 437-444, 2016.
52 [88] A. Chatzidimitrioua and S. Yannas, "Street canyon design and improvement potential for urban open spaces, the influence
53 of canyon aspect ratio and orientation on microclimate and outdoor comfort," Sustainable Cities and Society, vol. 33, pp.
54 85-101, 2017.
55 [89] N. Mohajeri, G. Upadhyay, A. Gudmundsson, D. Assouline, J. Kampf and J. Scartezzini, "Effects of urban compactness on
56 solar energy potential," Renewable Energy, vol. 93, pp. 469-82, 2016.
57 [90] N. Wong, S. Jusuf, N. Syafii, Y. Chen, N. Hajadi, H. Sathyanarayanan and Y. Manickavasagam, "Evaluation of the impact
58 of the surrounding urban morphology on building energy consumption," Solar Energy, vol. 85, pp. 57-71, 2011.
59 [91] A. Hargreaves, V. Cheng, S. Deshmukh, M. Leach and K. Steemers, "Forecasting how residential urban form affects the
60 regional carbon savings and costs of retrofitting and decentralized energy supply," Applied Energy, vol. 186, pp. 549-561,
61
62
39
63
64
65
 1
2
3
4 2017.
5
[92] J. He, A. Hoyano and T. Asawa, "A numerical simulation tool for predicting the impact of outdoor thermal environment on
6 building energy performance," Aplied Energy, vol. 86, no. 9, pp. 1596-605, 2009.
7
[93] K. Steemers, "Energy and the city: Density, buildings and transport," Energyand Buildings, vol. 35, no. 1, pp. 3-14, 2003.
8
9 [94] I. Emadodin, A. Taravat and M. Rajaei, "Effects of urban sprawl on local climate: A case study, north central Iran," Urban
10 Climate, vol. 17, pp. 230-247, 2016.
11 [95] M. Peel, B. Finlayson and T. McMahon, "Updated world map of the Köppen-Geiger climate classification," Hydrology and
12 earth system sciences, vol. 4, no. 2, pp. 439-473, 2007.
13 [96] T. M. S. d. t. i. C, "I.R. of Iran Meteorological Organizaition (IRIMO)," 20 July 2015. [Online]. Available:
14 http://reports.irimo.ir/jasperserver/login.html. [Accessed 10 July 2015].
15 [97] G. Roshan, I. Rousta and M. Ramesh, "Studying the effects of urban sprawl of metropolis on tourism-climate index
16 oscillation: a case study of Tehran city," Journal of Geography and Regional Planning, vol. 2, no. 12, pp. 310-21, 2009.
17 [98] A. Bashiri and H. Alizadeh, "The analysis of demographics, environmental and knowledge factors affecting prospective
18 residential PV system adoption: A study in Tehran," Renewable and Sustainable Energy Reviews, vol. 81, no. 2, pp. 3131-
19 3139, 2018.
20 [99] M. Mahdavinejad, M. Khabari and R. Moghadam, "Revitalization and the Iranian Contemporary Architecture after the
21 Victory of Islamic Revolution," Iranian Islamic City, vol. 2, pp. 95-119, 2011.
22 [100] Deputy of Urbanism and Architecture , "Comprehensive Master Plan," Tehran Municipality, Tehran, 2016.
23 [101] Zista Consultant and Tadbir Baft Consultant, Design Criteria and locating of high-rise buildings, Tehran: Tehran
24 Municipality, 1998.
25
[102] T. U. R. a. P. Center, "Terms and conditions of zoning in Tehran," Tehran Municipality, Tehran , 2010.
26
27 [103] T. Municipality, "Tehran Statasitcal Report 2004-2014," Tehran Municipality Statasitcal Center, Tehran, 2014.
28 [104] A. Rasheed, D. Robinson, A. Clappier, C. Narayanan and D. Lakehal, "Representing complex urban geometries in
29 mesoscale modeling," INternational Journal of Climatology, vol. 31, no. 2, pp. 289-301, 2011.
30 [105] J. Liu, J. Srebric and N. Yu, "Numerical simulation of convective heat transfer coefficients at the external surfaces of
31 building arrays immersed in a turbulent boundary layer," INternational Journal of Heat Transfer, vol. 61, pp. 209-25,
32 2013.
33 [106] A. Rasheed, D. Robinson, A. Clappier, C. Narayanan and D. Lakehal, "Representing complex urban geometries in
34 mesoscale modeling," International Journal of Climatology, vol. 31, no. 2, pp. 289-301, 2011.
35 [107] E. a. T. m. organization, "Executive code of building Conservation from Fire, Code 112," State Management and Planning
36 Organization, Tehran, 1992.
37 [108] A. Bhatia, "HVAC cooling load calculations and principles," Continuing Education and Development, Inc, NewYork,
38 2014.
39 [109] A. Burdick, "Strategy Guideline: Accurate Heating and Cooling Load Calculations," US Department of Energy, Building
40 Technologies Program, Pennsylvania, 2011.
41 [110] ASHRAE, 2001 ASHRAE Handbook of Fundamentals, Atlanta: American Society of Heating, Refrigerating and Air-
42 Conditioning Engineers, 2001.
43 [111] M. Sherman and D. Grimsurd, "Infiltration-pressurization correlation: Simplified physical modeling," ASHRAE
44 Transactions, vol. 86, no. 2, p. 778, 1980.
45
[112] J. Liu, M. Heiderinejad, S. Gracik and J. Srebric, "The impact of exterior surface convective heat transfer coefficients on
46 the building energy consumption in urban neighborhoods with different plan area densities," Energy and Buildings, vol. 83,
47 pp. 449-63, 2015.
48
[113] E. TM, "EnergyPlus engineering reference," 2013.
49
50 [114] A. Mochida, Y. Tominaga, S. Murakami, R. Yoshie, T. Ishihara and R. Ooka, "Comparison of various k-e models and
DSM applied to flow around a high-rise building-Report on AU cooperative project for CFD prediction of wind
51
environment," Wind and Structures, vol. 5, pp. 227-244, 2002.
52
53 [115] Y. Tominaga, A. Mochida, R. Yoshie, H. Kataoka, T. Nozu, M. Yoshikawa and T. Shirasawa, "AIJ guidelines for practical
applications of CFD to pedestrian wind environment around buildings," Journal of Wind Engineering and Industrial
54
Aerodynamics, vol. 96, pp. 1749-1761, 2008.
55
56 [116] D. f. Rhino, "Solemma," Harvard’s Graduate School of Design – Sustainable Design (G(SD)2) research group, [Online].
Available: www.Solemma.net.
57
58 [117] J. K. Calautit, B. R. Hughes and D. SNM Nasir, "Climatic analysis of a passive cooling technology for the built
59 environment in hot countries," Applied Energy, vol. 186, pp. 321-335, 2017.
60 [118] B. Blocken, T. Sathopoulos, J. Carmeliet and J. Hensen, "Application of computational fluid dynamics in building
61
62
40
63
64
65
 1
2
3
4 performance simulation for the outdoor environment: an overview," Journal of Building Performance Simulation, vol. 4,
5 pp. 157-84, 2011.
6 [119] S. Huang, X. Hu and N. Zeng, "Impact of wedge-shaped roofs on airflow and pollutant dispersion inside urban street
7 canyons," Building and Environment, vol. 44, no. 12, pp. 2335-2347, 2009.
8
[120] L. Ledo, P. Kosasih and P. Cooper , "Roof mounting site analysis for micro-wind turbines," Renewable Energy, vol. 36,
9 no. 5, pp. 1379-1391, 2011.
10
[121] S. Vardoulakis, R. Dimitrova, K. Richards, D. Hamlyn, G. Camilleri and M. Weeks, "Numerical model inter-comparison
11
for wind flow and turbulence around single-block buildings," Environmental Modeling and Assessment, vol. 16, pp. 169-
12 81, 2011.
13
[122] P. Richards and R. Hoxey, "Quasi-steady theory and point pressures on a cubic building," Journal of Wind Engineering
14
and Industrial Aerodynamics, vol. 92, no. 14, pp. 1173-1190, 2004.
15
16 [123] I. Abohela, N. Hamza and S. Dudek, "Effect of roof shape, wind direction, building height and urban configuration on the
energy yield and positioning of roof mounted wind turbines," Renewable Energy, vol. 50, pp. 1106-1118, 2013.
17
18 [124] P. Richards, R. Hoxey and L. Short, "Wind pressures on a 6 m cube," Journal of Wind Engineering and Industrial
19 Aerodynamics, vol. 89, pp. 1553-1564, 2001.
20 [125] S. Murakami and A. Mochida, "3-D numerical simulation of airflow around a cubic model by means of the k-[epsilon]
21 model," Journal of Wind Engineering and Industrial Aerodynamics, vol. 31, pp. 283-303, 1988.
22 [126] M. Montavon, J. Scartezzini and R. Compagnon, "Comparison of the solar energy utilisation potential of different urban
23 environments,," in 21-st Conf. on passive and low energy architecture, Eindhoven, 2004.
24 [127] D. Li, G. Liu and S. Liao, "Solar potential in urban residential buildings," Solar Energy, vol. 111, pp. 225-235, 2015.
25 [128] J. Sarralde, D. Quinn, D. Wiesmann and K. Steemers, "Solar energy and urban morphology: scenarios for increasing the
26 renewable energy potential of neighbourhoods in London," Renewable Energy, vol. 73, pp. 10-17, 2015.
27 [129] R. Ramponi, B. Blocken, L. de Coo and W. Janssen, "CFD simulation of outdoor ventilation of generic urban
28 configurations with different urban densities and equal and unequal street widths," Building and Environment, vol. 92, pp.
29 152-166, 2015.
30 [130] C. Kontokosta and C. Tull, "A data-driven predictive model of city-scale energy use in buildings," Applied Energy, vol.
31 197, pp. 303-17, 2017.
32 [131] M. Silva, I. M. Horta, V. Leal and V. Oliveira, "A spatially-explicit methodological framework based on neural networks
33 to assess the effect of urban form on energy demand," Applied Energy, vol. 202, pp. 386-398, 2017.
34
[132] U. Eicker, D. Monien, E. Duminil and R. Nouvel, "Energy performance assessment in urban planning competitions,"
35 Applied Energy, vol. 155, pp. 323-333, 2015.
36
[133] J. Rodrigues_Alvarez, "Urban energy index for building (UEIB): A new method to evaluate the effect of urban form on
37
buildings energy demand," Landscape and Urban Planning, vol. 148, pp. 170-187, 2016.
38
39 [134] B. Hong and B. Lin, "Numerical studies of the outdoor wind environment and thermal comfort at pedestrian level in
housing blocks with different building layout patterns and trees arrangement," Renewable Energy, vol. 73, pp. 18-27, 2015.
40
41 [135] S. Nikkho, M. Heidarnejad, J. Liu and J. Srebric, "Quantifying the impact of urban wind sheltering on the building energy
42 consumption," Applied Thermal Engineering , vol. 116, pp. 850-865, 2017.
43 [136] P. Rosado, G. Ban-Weiss, A. Mohegh and R. Levinson, "Influence of street setbacks on solar reflection and air cooling by
44 reflective streets in urban canyons," Solar Energy, vol. 144, pp. 144-157, 2017.
45 [137] C. Wen, Y. Juan and A. Yang, "Enhancement of city breathability with half open spaces in ideal urban street canyons,"
46 Building and Environment , vol. 113, pp. 322-336, 2007.
47 [138] M. Wong, J. Nichol, P. TO and J. Wang, "A simple method for designation of urban ventilation corridors and its
48 application to urban heat island analysis," Building and Environment , vol. 45, pp. 1880-1889, 2010.
49
50
51
52
53 Appendix A
54
55 Table A.1: Demographic and regional description of 22 districts of Tehran city (from [103])
56
District Population TBEA TAUO TAUB NB District Population TBEA TAUO TAUB NB
57
km2 km2 km2 km2 km2 km2
58
1 439467 45.74 11.86 33.88 112159 12 240720 16.01 2.88 13.13 66729
59 632917 47.61 14.2 33.41 181295 276027 17 3.51 13.49 71782
2 13
60
61
62
41
63
64
65
 1
2
3
4 3 314112 29.39 8.44 20.95 91364 14 484333 24.12 3.48 20.64 138222
5 4 861280 61.61 22.77 38.84 232108 15 638740 31.3 9.48 21.82 168192
6 5 793750 53.22 22.04 31.18 200727 16 287803 16.53 3.88 12.65 77294
7 6 229980 21.42 5.31 16.11 69796 17 248589 8.22 1.33 6.89 63721
8 7 309745 15.36 2.52 12.84 98527 18 391368 38.08 10.64 27.44 81553
9 8 378118 13.22 2.41 10.81 115684 19 244350 20.53 7.19 13.34 60257
10 9 158516 19.51 12.53 6.98 44949 20 340861 22.55 10.47 12.08 89213
11 10 302852 8.07 1.98 6.09 97035 21 162681 55.5 16.27 39.23 42957
12 11 288884 12.05 3.16 8.89 82171 22 128958 58.51 32.71 25.8 28763
13 TBEA :Total Built Environment Area TAUO : Total Area of Urban open spaces
14 TAUB: Total Area of Urban Buildings NB: Number of Buildings
15
16 Table A.2:1600 generated forms abbreviations
17
18 Forms Cases Patterns Max Mean Middle Min Norms
19 P01 P1 L Max01-20 P1 L Mean01-20 P1 L Middle01-20 P1 L Min01-20 P1 L Norm01-20
P02 P2 L Max01-20 P2 L Mean01-20 P2 L Middle01-20 P2 L Min01-20 P2 L Norm01-20
20 L 400
P03 P3 L Max01-20 P3 L Mean01-20 P3 L Middle01-20 P3 L Min01-20 P3 L Norm01-20
21 P04 P4 L Max01-20 P4 L Mean01-20 P4 L Middle01-20 P4 L Min01-20 P4 L Norm01-20
22 P01 P1 U Max21-40 P1 U Mean21-40 P1 U Middle21-40 P1 U Min21-40 P1 U Norm21-40
23 U 400
P02 P2 U Max21-40 P2 U Mean21-40 P2 U Middle21-40 P2 U Min21-40 P2 U Norm21-40
24 P03 P3 U Max21-40 P3 U Mean21-40 P3 U Middle21-40 P3 U Min21-40 P3 U Norm21-40
P04 P4 U Max21-40 P4 U Mean21-40 P4 U Middle21-40 P4 U Min21-40 P4 U Norm21-40
25
P01 P1C Max41-60 P1 C Mean41-60 P1 C Middle41-60 P1 C Min41-60 P1 C Norm41-60
26 P02 P2C Max41-60 P2 C Mean41-60 P2 C Middle41-60 P2 C Min41-60 P2 C Norm41-60
27 C 400
P03 P3 C Max41-60 P3 C Mean41-60 P3 C Middle41-60 P3 C Min41-60 P3 C Norm41-60
28 P04 P4 C Max41-60 P4 C Mean41-60 P4 C Middle41-60 P4 C Min41-60 P4 C Norm41-60
29 P01 P1CY Max61-80 P1CY Mean61-80 P1CY Middle61-80 P1CY Min61-80 P1CY Norm61-80
30 P02 P2CY Max61-80 P2CY Mean61-80 P2CY Middle61-80 P2CY Min61-80 P2CY Norm61-80
CY 400
P03 P3CY Max61-80 P3CY Mean61-80 P3CY Middle61-80 P3CY Min61-80 P3CY Norm61-80
31 P04 P4CY Max61-80 P4CY Mean61-80 P4CY Middle61-80 P4CY Min61-80 P4CY Norm61-80
32 Total Cases 1600 320 320 320 320 320
33
34
35 Table A.3: boundary conditions of the validation study of the cube
36 boundary conditions descriptions
37 Inlet Retrofitted average wind velocity
38 Outlet Zero-pressure
39 Laterals Symmetric
40 Top Symmetric
41 Ground Standard WF
42 Grid size around cube 0.1m
43 Grid size domain 0.3m
44
Meshing (ANSYS Fluent): tetrahedral cells 1870991
45
46 Meshing (ANSYS Fluent): triangular interior faces 3701969
47 Meshing (Autodesk CFD): tetrahedral cells 27170
48 Meshing (Autodesk CFD): triangular interior faces 72689
49
50 Table A.4: applied variables and parameters in simulations
51 variables Sections Description Value Unit
52
53 People Residency 0.2 p/m2
54 Schedule behavior As a simple office
55 Equipment Typical office 12 w/m2
Loads

56 Schedule behavior As a simple office

57 Lighting lights 12 w/m2
58 Illumination 500 lux
59 Dimming continuous -
60 Schedule behavior As a simple office -
61
62
42
63
64
65
 1
2
3
4 Heating Limitations No -
5 Cooling Limitations No -
6 Humidity control Limitations No -
7 Mechanical Fresh air 2.5 L/s/person
Conditioning

8 ventilation Fresh air 0.3 L/s/zone area m2

9 Sensible recovery ratio 0.7 -
10 Heat recovery None -
11 Ventilation Scheduled None -
12 Natural Buoyancy driven flow 18-30 C
13 Rel. Humidity 80% -
14 Infiltration ACH 0.2 -
15 Domestic hot Peak flow 0.03 m3/h/m2
water

water Supply Temp 65 C

Hot

16
17 Main Temp 10 C
Schedule behavior As a simple office -
18
External walls 5 layers (500mm) Reinforced U=0.4 W/m2K
19
Adiabatic concrete, polystyrene
20 plaster, insulation
21 insulation, according to
22 mortar, NBC 19 Iran
23 composite
24 facade
25 Internal walls 3 layer (250mm) Bricks, U=0.7 W/m2K
26 Adiabatic plaster, No
27 plaster insulation
28 Roof 4 layers (350mm) Reinforced U=0.30 W/m2K
29 Adiabatic concrete, polystyrene
30 plaster, insulation
Construction

insulation, according to
31 cement NBC 19 Iran
32 mosaic
33 Window Frame Stainless U=0.9 W/m2K
34 steel
35 Glass Low-E U=1.70 W/m2K
36 (0.30)
37 SHGC 0.2 -
38 Shading None -
39 Projection 50% -
40 Factor
41 Glazing North 25% 9* (2*2) win
42 facade
43 South 15% 5* (2*2) win
44 façade
West façade 8% 3* (2*2) win
45
East façade 8% 3* (2*2) win
46
Density Assumed air properties in Equation of state
47 simulations
Viscosity 0.0001817 poise
48 conductivity 002563 w/cm-K
49
air

Specific heat 1.004 J/g-K

50 compressibility Cp/Cv 1.4
51 emissivity 1 -
52 Gas constant 287.05 M2/s2-K
53 Radiation Base on Tehran weather data Months (4-9)/20/2016 -
heating

54
Solar

12:00 Pm
55 Heat flux Base on Tehran weather data Months (4-9)/20/2016
56 12:00 Pm
57
58
59 Table A.5: boundary conditions applied in CFD simulation in the study
60
plane Boundary conditions
61
62
43
63
64
65
 1
2
3
4 Inlet Retrofitted average wind velocity
5 Outlet Zero-pressure
6 Laterals Symmetric
7 Top Symmetric
8
Ground Standard WF
9
10
11
12
13
14
15
16
17 Appendix B
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
44
63
64
65
 1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58 Figure B.1. Wind velocity contours out of CFD simulations of the best urban configuration in Urban Pattern one
59 (P01)
60
61
62
45
63
64
65
 1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59 Figure B.2. Wind velocity contours out of CFD simulations of the best urban configuration in Urban Pattern two
60 (P02)
61
62
46
63
64
65
 1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59 Figure B.3. Wind velocity contours out of CFD simulations of the best urban configuration in Urban Pattern three
60 (P03)
61
62
47
63
64
65
 1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59 Figure B.4. Wind velocity contours out of CFD simulations of the best urban configuration in Urban Pattern four
60 (P04)
61
62
48
63
64
65
Graphical Abstract (for review)
Figures High quality

Figure 1. The structured workflow of the research

Figure 2. Twenty-Two districts of Tehran with different demographic and geographic characteristic [103]
Figure 3. the target building, a twelve-story office building:(a): S/E view of the 3D model of the building. (b):
Architectural typical plan of the target building (2 nd floor to 12 floor)
Figure 4. Modeling process of the hypothetical template site based on BMC; (a) basic cuboid module (8×8×X
(variable height) m), (b) each block is divided into twenty-five 8×8m cells based on the module, (c) seven blocks and
one open green space placed around the target building as default template site, (d) rectangle template site with
24000m2 area and two highways between blocks based on the module, (e) 3D template site with the target building,
highways, surrounding blocks and open green space based on the module.

Figure 5. urban morphology parameters used in this study to define urban density; a) volume area ratio (VAR):
total volume of surrounding buildings divided by the total site area of site (24000 m 2), b) site coverage (SC) the total
area of the ground floor of the building divided by 1600 (the area of each sub-site: 40×40m), c) plot area ratio
(PAR): total floor area of the surrounding buildings divided by the total site area, d) building density (BD): number
of selected cells divided by the total site area, e) urban plan area density (λ p): the built area projected onto the
ground surface divided by the site area in a horizontal section (A s), f) frontal area density (λf): is the area of frontal
surface of façade to the As
Figure 6. Urban Density height, PAR, VAR and λf morphology in the context of urban pattern one (P01):
(a): “Max” density morphology, (b): “Middle” density morphology, (c): “Min” density morphology,
(d): “Norms” density morphology, (e): “Mean” density morphology
(f): Basic Cubic Module and forty-cell diagram plan of each urban building
 Figure 7. Top: initial rules: (a): the 8×8 module cell, (b): selected cells must be connected from their sides, (c):
connection from edges are not acceptable, (d): unacceptable cell connection from edges, (e) detached selected cell
and edge connection
Bottom: examples of acceptable and unacceptable form generation based on the initial rules for five density ranges
Figure 8. 400 case studies categorization based on building forms and influenced urban density indicators according to BMC technique
Figure 9. Four urban patterns in the study: All patterns have the same total area (24000m 2) and street to site area
ratio (40%) (the dash-line rectangle shows the margin of the P01 in other patterns)

Figure 10. categorization of 1600 generated case studies based on three main parameters of the study,
 (The number in the blanket is the total number of generated cases in each category)

Figure 11. (a): Computational domain and boundary conditions used for CFD simulation (H= 48 m), (b): Section
computational domain and boundary condition, (c): Dimensions of computational domain used for urban patterns
P01-P04 W (width of the site), L (length of the site), Domain W (domain width), Domain L (domain length),
Figure 12. (a): wind velocity calculation points and dominant wind flow on one of the case studies, (b): Sample plan
of the target building for 2nd to 12th floors, (c): dimensional coordinates of wind velocity magnitude calculation
points in four urban patterns, (d): An elevation of the target building, locations of cooling load (Q) and wind
velocity (ʋ) calculation points (The case study is P04 C Middle 46)
Figure 13. (a) Computational domain applied for all 1600 case studies simulation in Autodesk CFD,
(b): Computational domain applied for cube validation study from Abohela et al [123]
 Figure 14. (a) ANSYS Fluent wind velocity magnitude at Y section contours, (b) ANSYS Fluent wind velocity
magnitude Plan contours, (c) ANSYS Fluent pressure coefficient at Y section contours, (d) ANSYS Fluent pressure
coefficient Plan contours
(e) Autodesk CFD wind velocity magnitude at Y section contours, (f) Autodesk CFD wind velocity
magnitude Plan contours, (g) Autodesk CFD pressure coefficient at Y section contours, (h) Autodesk CFD
pressure coefficient Plan contours
Figure 15. Simulated Pressure coefficient in twenty-one calculation point at the centerline of windward, roof and
leeward surfaces of a sample cube in comparison with two wind tunnel test and a validated CFD results

Figure 16. P02 C Min 41, case study selected for the validation study
 Figure 17. (a) ANSYS Fluent wind velocity magnitude at Y section contours, (b) ANSYS Fluent wind velocity
magnitude Plan contours, (c) ANSYS Fluent pressure coefficient at Y section contours, (d) ANSYS Fluent pressure
coefficient Plan contours
(e) Autodesk CFD wind velocity magnitude at Y section contours, (f) Autodesk CFD wind velocity
magnitude Plan contours, (g) Autodesk CFD pressure coefficient at Y section contours, (h) Autodesk CFD
pressure coefficient Plan contours
Figure 18. Simulated Pressure coefficient in twenty-one calculation point at the centerline of windward, roof and
leeward surfaces of a P02 C Min 01 case study in both CFD tools used in the study

Figure 19. Left: average cooling load simulation of different Urban Density in six months
Right: average cooling load simulation of different Urban Building Form in six months
 Figure 20. Left: average cooling load simulation of different Urban Pattern in six months
Right: total average cooling load in term of urban density and urban patterns for all six months

Figure 21. Left: average wind velocity magnitude, out of CFD simulations, in term of Urban Density in six months
Right: average wind velocity magnitude, out of CFD simulations, in term of Urban Building Form in six months

Figure 22. Left: average wind velocity magnitude, out of CFD simulations, in term of Urban Pattern in six months
Right: total average wind CFD velocity magnitude in term of Urban Density and Urban Patterns in all six months
Figure 23. best urban configuration chosen in four urban patterns and urban building form

Figure 24. Left: comparison of ACL and cooling load of the best urban configuration,
Right: comparison of AWS and wind speed of the best urban configuration
Figure 25. Wind speed contours for the best urban configuration in April
 Figure 26. Left: Statistics of Top 100 cases in three parameters of the study
Right: Some of efficient cases among Top 35 cases in each urban building form category
Appendix B

Figure B.1. Wind velocity contours out of CFD simulations of the best urban configuration in Urban Pattern one
(P01)
Figure B.2. Wind velocity contours out of CFD simulations of the best urban configuration in Urban Pattern two
(P02)
Figure B.3. Wind velocity contours out of CFD simulations of the best urban configuration in Urban Pattern three
(P03)
 Figure B.4. Wind velocity contours out of CFD simulations of the best urban configuration in Urban Pattern four
(P04)

View publication stats