Energy and Buildings 91 (2015) 131138

Contents lists available at ScienceDirect

Energy and Buildings

journal homepage: www.elsevier.com/locate/enbuild

Effect of housing density on energy efciency of buildings located in

hot climates
Omar S. Asfour , Ebtesam S. Alshawaf
Faculty of Engineering, Islamic University of Gaza, PO Box 108, Gaza City, Palestinian Territories

a r t i c l e i n f o a b s t r a c t

Article history: Residential buildings are considered among the most important elements in the city, as they form most
Received 1 April 2014 of our built environment. Among the factors that affect urban planning and energy consumption of
Received in revised form residential neighbourhoods, the role of housing density is essential. Thus, this study aims to examine
20 November 2014
the effect of housing density on energy efciency of buildings considering the hot climatic conditions.
Accepted 20 January 2015
To achieve this aim, a numerical analysis based on computerized simulation has been carried out to
Available online 28 January 2015
compare different congurations of housing densities in terms of their energy consumption. The study
concluded that energy efciency in residential buildings is highly dependent on their density. This is true
Keywords:
Housing
when housing type is changed, or even within the same housing category. In addition, the study found
Density that compact horizontal housing congurations can perform better in terms of energy efciency when
Energy compared to the vertical ones. In this context, the examined row houses conguration offers a reduction
Hot climate in average energy consumption that reaches 28% compared to the rest of examined residential buildings
Thermal modelling types located in urban situation.
2015 Elsevier B.V. All rights reserved.

1. Introduction if planned properly, help reducing construction and energy costs of

Energy is one of the most fundamental needs in our life. As
population of the world increases, reliance on energy produced
from non-renewable resources should be gradually reduced and
replaced by renewable energies. This is in addition to the envi-
ronmental benets of renewable energy sources in facing the
problems of pollution, global warming, and climate change. The
role of architects and engineers is essential here. They have large
share of responsibility for designing buildings in a green way that
ensures healthy and environmental friendly built environment. In
this regard, the role of housing is essential as it represents most
of our built environment. Housing is the largest or at least one of
the largest sectors in energy demand [1]. Thus, every little helps
when it comes to energy savings in this sector. In this context,
housing modernization and utilization of technology can lead to
a signicant reduction in energy consumption and CO2 emissions
[2]. There are many sustainability design guidelines related to the
design of individual residential buildings and the planning process
of green neighbourhoods. This includes housing density, which can,
housing [3].
Housing density is usually measured in dwelling unit per land
area unit. This could be acre (du/acre), or hectare (du/ha). Hous-
ing density is an essential planning indicator as it helps estimating
how many land for housing is needed, and this would affect other
land uses. If the prescribed limit of density is exceeded, then
housing units are considered overcrowded. In general, increas-
ing density properly has the advantage of reducing housing cost.
This is because higher density means [4]: less land to be devel-
oped, shorter roads and infrastructure lines, and smaller housing
units which are more affordable for low-income people. On the
other hand, high density may reduce solar potential of building as
a result of increasing shading of neighbouring buildings [5]. Thus, it
is required to achieve a balance between advantages and disadvan-
tages of compact housing environment. Increasing housing density
can be achieved through implementing several housing patterns.
This includes walk-up and mid-rise apartments, courtyard and row
houses, single family homes on small plots, and high-rise apart-
ments.
There are different methods to measure housing density in res-
idential neighbourhoods. These methods are as follows [6]:

Corresponding author. Tel.: +970 8 2644400; fax: +970 8 2644800. - Gross housing density: This refers to number of housing units
E-mail addresses: oasfour@iugaza.edu.ps, o.asfour@hotmail.com (O.S. Asfour). (or inhabitants) in a given area including all land uses in the

http://dx.doi.org/10.1016/j.enbuild.2015.01.030
0378-7788/ 2015 Elsevier B.V. All rights reserved.
132 O.S. Asfour, E.S. Alshawaf / Energy and Buildings 91 (2015) 131138

neighbourhood. This includes roads and parking, commercial ser- investigation of additional housing congurations with reference to
vices, schools, and public open spaces. This method is useful for the hot climatic conditions.
planning new neighbourhoods.
- Net housing density: As above, but land considered here is the
one used for residential purposes only. This method is useful for 2. Methodology
planning housing clusters within existing residential sites, where
services and roads already exist. This research is carried out using parametric numerical analysis.
- Site coverage ratio: This is simply the percentage of the built area This has been implemented using Ecotect program as a main tool to
to the site area. Although this method does not consider building estimate energy consumption in several hypothetical cases. Ecotect
height, it is a useful indicator in the case of housing projects of is a widely used program in environmental performance simulation
uniform building height. studies, where its three-dimensional CAD interface makes it more
- Plot ratio or oor area ratio (FAR): This density is the most accurate practical in building simulation. Ecotect is known to give a reason-
method when it comes to design issues like natural ventilation able estimate of energy loads in buildings [13], and has been solely
and lighting as it takes building height into account. It simply used in several recent energy modelling studies [1418]. In addi-
means the ratio of building oor area to the total site area multi- tion, its thermal modelling of buildings has been compared to other
plied by number of oors. tools [19] and eld measurements [20] where good agreement has
been observed. Ecotect is based on the Chartered Institute of Build-
ing Services Engineers (CIBSE) admittance method. This method
The interacting connection between climate, urban form, and uses a exible algorithm which has no restrictions on building
energy consumption of buildings is a demanding subject of research geometry or number of analysed thermal zones. Several outputs
that has been discussed in several studies [79]. More specically, can be produced by the software including external and internal
several studies have tackled the effect of housing density on energy temperature proles, thermal comfort, heating and cooling loads,
consumption. For example, Hui [10] investigated how urban den- and heat gain and losses proles.
sity affects building energy design strategies in highly populated Design Builder program has been used as a secondary tool. This
cities, considering Hong Kong as a case study. The paper evaluated has been done at the beginning of the study to ensure proper use
the major factors that affect low energy building design in this con- of Ecotect modelling code. Design Builder is a powerful software
text. The study concluded that densication of cities could have tool for assessing thermal performance of buildings. It can be used
both positive and negative effects on total energy consumption for several applications including energy consumption, shading
related to transport, infrastructure, thermal performance, and nat- design, and lighting control systems. It uses Energy Plus simulation
ural lighting and ventilation. However, with conscious urban and engine to simulate thermal performance of buildings, but with a
building design, the overall environmental quality can be improved. more user-friendly interface. Several graphical or tabulated outputs
Heinonen and Junnila [1] carried out a holistic analysis of energy can be obtained including energy consumption, mean radiant and
consumption patterns in housing urban and rural environment in operative temperatures and humidity, heat transmission through
Finland. This has been done based on a eld survey and data collec- building fabric, and CO2 generation.
tion for selected cases. These cases included apartment buildings, As mentioned above, results of Ecotect and Design Builder
row houses, and detached houses. The study found that the differ- are compared to ensure proper use of Ecotect modelling code.
ent cases appear to be less energy-consuming in rural areas. For Table 1 shows the implemented modelling settings in both pro-
example, the overall average energy use per housing unit increases grams. Several residential blocks with different types and urban
from 12,600 kW h/a in apartment buildings located in rural areas congurations have been simulated. In all simulation cases, energy
to 15,500 kW h/a in apartment buildings located in urban ones. consumption was the dependent variable, presented as heating,
Hachem et al. [11] investigated the inuence of geometry, density cooling, and total loads as an indicator of energy consumption. If
and site layouts on solar energy utilization potential of two-story these loads are counted by a factor 1 to 1, then refrigeration kW h
housing unit. This has been done using numerical analysis imple- equals electricity kW h. Thus, it was possible to observe the effect of
mented by EhergyPlus simulation package. Several congurations
of attached and detached houses have been investigated, where
spacing between housing rows and the resulting shading was a Table 1
The implemented settings in the comparison of Ecotect and Design Builder.
key determinant of solar potential of the examined housing unit.
The study found that at higher densities achieved by congurations Ecotect Design Builder
such as row houses, shading effect is fundamental on irradiation. Location and site
This effect is strongly dependent on the distance between the rows Location
of units, and can offer a reduction in insolation that reaches 55% in Weather data Al-Arish city Al-Arish city
some congurations. Terrain types Urban Urban
Site orientation NS NS
Strmann-Andersen and Sattrup [12] investigated the effect of Thermal condition
urban density on daylight and passive solar gains in ofce and HVAC system Full air conditioning Full air conditioning
domestic buildings. Density is examined by changing the aspect Thermostat range ( C) 18.026.0 18.026.0
ratio of urban canyon in several orientations. The study found that Heating set point 18.0 18.0
Cooling set point 26.0 26.0
there is in general an increase in energy consumption as den-
Design condition
sity increases. With reference to the cold climatic conditions, the Occupancy (occupants/unit) 6 6
observed increase in energy loads is mainly related to heating loads Occupancy pattern Continuous Continuous
increase due to the reduction in solar gains. On the contrary, cool- Internal sensible gain (W/m2 ) 15 15
ing demand decreases with density due to overshadowing. These Internal latent gain (W/m2 )
Inltration rate (air change/hr.) 1 1
studies show that the use of parametric numerical analysis to U-value (W/m2 K)
investigate the relationship between urban density and energy con- Exterior walls 2.30 2.55
sumption is a practical approach. This is facilitated by the use of Roof 2.60 2.51
computer simulation using a variety of tools. However, most studies Floor 2.56 2.24
Window 5.50 6.12
focus on the cold climatic conditions. Thus, this study offers further
 O.S. Asfour, E.S. Alshawaf / Energy and Buildings 91 (2015) 131138 133

Ecotect Design Builder Heang loads Cooling loads Total loads

140 140.0

Energy Load (kWh/m2)

Total Load (kWh/m2)

120 120.0
100 100.0
80 80.0
60 60.0
40 40.0
20 20.0
0 0.0
Case 1 Case 2 Case 3 Case 4 D.1(1) D.2(1) D.3(1) D.4(1)

Modelling Case Modelling Case (NS orientaon)

Fig. 1. A comparison between Ecotect and Design Builders calculations of annual

energy loads (kW h/m2 ). Heang loads Cooling loads Total loads

140.0

Energy Load (kWh/m2)

changing urban density on energy consumption behaviour in these
different building congurations.
Ecotect imports climatic data from weather data les available
in the program directory. In fact, limited cities are available for the
hot climate. Among these cities, Al-Arish, Egypt has been selected.
This city is located on latitude 31N, and can be used to repre-
sent mild hot climatic conditions, where prevailing Mediterranean
winds moderate its temperatures. Daily mean temperature in this
city ranges from 13 C in Jan. to 26 C in Aug. [21]. Table 2 shows
the monthly climatic data of Al-Arish city. In modelling settings, the
used sensible heat gain reects gains from appliances and lighting
based on their number, heat emission, and operation time. Sensible
gains from appliances have been estimated at 4 W/m2 .
As for lighting, it is estimated at 11 W/m2 [22], which means a
total value of 15 W/m2 . As for HVAC settings, it is assumed full air
conditioning to help estimating heating and cooling loads. It is sized
for each single zone, where heating and cooling system run to main-
tain building air temperature within the assumed comfort zone
(1826 C). System efciency is assumed 95%, which means that 5%
is lost in the conversion from electrical energy to thermal heating or
cooling. Finally, buildings have been assumed in the urban terrain
to consider wind-shelter effect. The site is oriented towards north-
south, which means shorter southern building elevations and less
heat gains in summer.
Several cases have been modelled using both programs based on
the settings explained above. These cases include single detached,
row, and multi-story housing units. Table 3 presents geometrical
characteristics of these cases. As shown in Fig. 1, results obtained
from Ecotect and Design Builder show similar behaviour with slight
discrepancy in some values. This is acceptable considering com-
plexity of thermal modelling calculations. Maximum amount of
discrepancy is 13% observed in Case 2. This may be related to the
slight differences in thermal settings, especially U-value, and the
underlying assumptions used in calculation method in each pro-
gram.
3. Simulation study
This study examines the relation between energy consumption
and housing type in the urban context. The simulated 15 cases rep-
resent different housing types including single detached, row, and
apartment housing blocks in both NS and EW orientations, where
orientation represents the long axis of the urban conguration.
The study used abbreviation codes to distinguish these different
types. The method gives each modelling case a special code which
contains three parts: house type, number of oors, and number of
housing unit in each oor. The symbols D, R, Al, Ah have been used
for single detached house, row house, low-rise apartment build-
ing, and high-rise apartment building, respectively. For example,
120.0
100.0
80.0
60.0
40.0
20.0
0.0
D.1(1) D.2(1) D.3(1) D.4(1)
Modelling Case (EW orientaon)
Fig. 2. Estimated annual loads in the single detached house congurations
(kW h/m2 ).
the code D.2(1) refers to two-oor single detached house, with one
dwelling unit per oor.
Each type is modelled using different housing densities pre-
sented in number of units per 1000 m2 . In order to examine the
effect of both housing type and housing density, the following
variables are assumed xed: housing unit area (130 m2 ), site area
(3300 m2 ), and thermal settings (see Table 1). As building type and
density changes, several variables change such as building height,
surface area, and internal gains. To allow for fair comparison of
energy consumption in the different congurations, it is essential
to normalize the aforementioned variables. This has been done by
dividing the estimated annual total energy consumption of each
conguration by the gross oor area of all units, which means that
energy consumption is presented in kW h/m2 .
3.1. Effect of different single detached house densities on energy
consumption
This part investigates four cases of single detached housing
located in linear urban conguration. Building height increases
from 1 to 4 oors, which resulted in housing density ranging from 3
to 12 dwelling unit per 1000 m2 (site area is 3300 m2 ). These cases
in addition to the geometrical parameters of the single unit are
illustrated in Table 4.
Fig. 2 shows the results obtained for NS and EW orientations.
It shows that energy consumption increases as building height
increases. However, energy consumption in the NS orientation is
slightly higher by about 3% due to the observed reduction in cool-
ing loads in EW orientation, where the cluster has less exposure
to south in the latter case. As for the NS orientation, the observed
increase in total load compared to D.1(1) is 16%, 23%, and 26.6%
for cases D.2(1), D.3(1), and D.4(1), respectively. It is commonly
believed that increasing housing density leads to a reduction in
energy consumption per housing unit. This is true here for the
heating loads only, where increasing building density means that
a heat loss from one unit in winter will become a heat gain for
134 O.S. Asfour, E.S. Alshawaf / Energy and Buildings 91 (2015) 131138

Table 2
Al-Arish city climatic data [21].

Annual Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

Average temperature ( C)
20 13 14 16 18 21 24 25 26 25 22 20 16

Average precipitation (mm)

110 30 10 10 10 20

Average relative humidity (%)

72 71 74 70 71 68 73 74 73 70 69 74 78

Table 3
Geometrical parameters of the comparative modelling study.

Case no. 1 2 3 4

Building type Single detached Semi-detached Low-rise High-rise

Perspective

Plan

Height (m) 3.3 (1 oors) 6.6 (2 oors) 9.9 (3 oors) 19.8 (6 oors)
Unit area (m2 ) 130 130 130 130
Units/oor 1 2 4 4

the other one. As shown in Fig. 2, the increase in total loads is
mainly caused by cooling loads, which are much higher since the
study is carried out with reference to hot climatic conditions. This
is because increasing site density reduces building exposure to
wind, which in turn reduces heat loss through building fabric,
and increases the thermal island effect. This is in agreement with
Strmann-Andersen and Sattrup [12] ndings in cold climates, but
in a reversed manner, as the total load increased as a result of
heating load instead.
3.2. Effect of different row house densities on energy consumption
The second group in this modelling study investigates three
cases of row housing located in linear urban conguration.
Building height increases from 1 to 3 oors, which resulted
in housing density ranging from 4 to 12 dwelling unit per
1000 m2 (site area is 3300 m2 ). These cases in addition to the
geometrical parameters of the single unit are illustrated in
Table 5.

Table 4
Geometrical parameters of the detached house modelling cases.

Case code D.1(1) D.2(1) D.3(1) D.4(1)

Density (du./1000 m2 ) 3 6 9 12

Geometrical parameters of single block

Plan

No. of oors 1 2 3 4
Area of unit (m2 ) 130 130 130 130
Units/oor 1 2 3 4
No. of people 6 2 6 = 12 3 6 = 18 4 6 = 24
 O.S. Asfour, E.S. Alshawaf / Energy and Buildings 91 (2015) 131138 135

Table 5
Geometrical parameters of row house modelling cases.

Case code R.1(1) R.2(1) R.3(1)

Density (du/1000 m2 ) 4 8 12

Geometrical parameters of single block

Plan

No. of oors 1 2 3
Area of unit (m2 ) 130 130 130
Units/oor 1 2 3
No. of people 6 2 6 = 12 3 6 = 18

Load relation to density at NS and EW orientations has a similar
behaviour to the one observed in the previous modelling group
(Fig. 3). As density increases, heating loads cease while cooling
loads increase. However, energy consumption in the NS orienta-
tion is slightly higher by about 5% due to the observed reduction
in cooling loads. At the NS orientation, the observed increase in
total load compared to R.1(1) is 16% and 21.7% for cases R.2(1)
and R.3(1), respectively. However, recorder loads here are less than
the detached house case in general. For example, total load in case
R.1(1) is 78.5 kW h/m2 , compared to 87.6 in case D.1(1). This means
Heang loads Cooling loads
140.0
Energy Load (kWh/m2)
a reduction of 10.4%. This is related to that row houses have less
exposed surface area compared to the single detached housing con-
guration, which help reducing heat gains through building fabric.
In this regard, the average daily heat gains through fabric in summer
is 320 and 269 W h/m2 for cases D.1(1) and R.1(1), respectively.
3.3. Effect of different low-rise apartment building densities on
energy consumption
The third group investigates four cases of low-rise apartment
housing blocks located in linear urban conguration. Building
Total loads height is either 3 to 4 oors, assuming two or four units per oor.
This resulted in housing density ranging from 9 to 14 dwelling unit
per 1000 m2 (site area is 3300 m2 ). These cases in addition to the

120.0
geometrical parameters of the single unit are illustrated in Table 6.
100.0
80.0 Cases have been modelled at both NS and EW orientations. As
60.0 observed in the previous types, results at both orientations show
40.0 similar behaviour but with less cooling loads in the EW orientation.
20.0 Thus, results will be presented here only for the NS orientation as
0.0 shown in Fig. 4. In general, the estimated energy loads are higher
R.1(1) R.2(1) R.3(1) than the previous congurations, i.e. row houses. For example, at
density of 12 unit per 1000 m2 , case Al.3(4) recorded higher loads by
Modelling Case (NS orientaon) 14.2% compared to R.3(1). When comparing the cases modelled in
this group, it is possible to note that increasing number of units per
oor from two to four resulted in an increase in the total load. Unlike
Heang loads Cooling loads Total loads
the previous groups, this increase is attributed to both heating
140.0 and cooling loads, which indicates the effect of increasing build-
Energy Load (kWh/m2)

120.0 ing height on solar access to the external exposed surfaces in both
100.0 winter and summer. To quantify this effect, heat gains and losses
80.0 for cases Al.3(4) and Al.4(4) have been investigated. Fabric and solar
60.0 heat gains and losses have been estimated for the average hottest
40.0 day in the year to represent summer, and for the average coldest day
20.0 in the year to represent winter. Values obtained have been normal-
0.0 ized per square meter to allow for comparison. As shown in Table 7,
R.1(1) R.2(1) R.3(1) when comparing case Al.3(4), which has less height, to case Al.4(4),
it is possible to notice that case Al.3(4) has less heat gains in sum-
Modelling Case (EW orientaon)
mer and less heat losses in winter. This in turns reduces its need
Fig. 3. Estimated annual loads in the row house congurations (kW h/m2 ). for cooling in summer and for heating in winter.
136 O.S. Asfour, E.S. Alshawaf / Energy and Buildings 91 (2015) 131138

Table 6
Geometrical parameters of low-rise apartment modelling cases.

House code Al.3(2) Al.4(2) Al.3(4) Al.4(4)

2
Density (du/1000 m ) 9 12 11 14

Geometrical parameters of single block

Plan

No. of oors 3 4 3 4
Area of unit (m2 ) 130 130 130 130
Units/oor 6 8 12 16
No. of people 6 6 = 36 8 6 = 48 12 6 = 72 16 6 = 96

Heang loads Cooling loads Total loads 3.4. Effect of different high-rise apartment building densities on
energy consumption
140.0
Energy Load (kWh/m2)

120.0
Finally, the fourth group investigates four cases of high-rise
100.0
80.0 apartment housing blocks located in linear urban conguration.
60.0 Building height increases from 6 to 12 oors, assuming four units
40.0 per oor. This resulted in housing density ranging from 22 to 43
20.0 dwelling unit per 1000 m2 (site area is 3300 m2 ). These cases in
0.0 addition to the geometrical parameters of the single unit are illus-
Al.3(2) Al.4(2) Al.3(4) Al.4(4) trated in Table 8.
Cases have been modelled at both NS and EW orientations. As
Modelling Case (NS orientaon) observed in the previous types, results at both orientations show
similar behaviour. Thus, results will be presented here only for the
NS orientation. Fig. 4 shows the results obtained from the simu-
Heang loads Cooling loads Total loads
lation of high-rise housing congurations. It can be noticed that
140.0 the effect of increasing density on increasing energy consumption
Energy Load (kWh/m2)

120.0 is recorded here as well. The observed increase in total load com-
100.0 pared to Ah.6(4) is 4%, 9%, and 7% for cases Ah.8(4), Ah.10(4), and
80.0 Ah.12(4), respectively. It can be noticed from Fig. 4 as well that
60.0 the estimated energy loads here are higher than the previous con-
40.0 guration, i.e. low-rise buildings, by 10% in average. This shows
20.0 the restrictions caused by increasing building height on solar and
0.0 wind exposure, which affects passive heating in winter and passive
Ah.6(4) Ah.8(4) Ah.10(4) Ah.12(4) cooling in summer.
Finally, Table 9 summarizes the effect of different housing types
Modelling Case (NS orientaon)
and densities on energy consumption considering NS orientation as
Fig. 4. Estimated annual loads in the low-rise (above) and the high-rise (bottom) the worst case. It shows the different modelled cases, their densi-
house congurations (kW h/m2 ). ties, the normalized total energy load, and the averaged one for each
category. It can be noticed that the lower energy consumption is
observed in the case of row houses. The average normalized energy
load for this housing conguration is 90.7 kW h/m2 , which is lower
Table 7
by 12%, 19.5%, and 28% compared to the single detached, low-rise,
Heat gains and losses for cases Al.3(4) and Al.4(4).
and high-rise congurations, respectively. This means that com-
Case Gain/loss Fabric Solar Total pact horizontal housing congurations can perform better in terms
(W h/m2 ) (W h/m2 ) (W h/m2 ) of energy efciency, when compared to the vertical ones. Despite
Summer 373.0 74.8 447.8
the fact that vertical congurations provide higher densities, this
Al.3(4)
Winter 680.7 22.5 658.2 occurs on the account of passive heating in winter and passive
Al.4(4) Summer 389.2 61.7 451.0 cooling in summer, which increases energy consumption.
Winter 713.7 17.4 696.3 Even for the same housing density, some interesting results
can be observed. For example, Table 7 shows that it is possible to
 O.S. Asfour, E.S. Alshawaf / Energy and Buildings 91 (2015) 131138 137

Table 8
Geometrical parameters of high-rise apartment modeling cases.

House code Ah.6(4) Ah.8(4) Ah.10(4) Ah.12(4)

Density (du/1000 m2 ) 22 29 36 43

Plan

No. of oors 6 8 10 12
Area of unit (m2 ) 130 130 130 130
Units/oor 4 4 4 4
No. of people 24 6 = 144 32 6 = 192 40 6 = 240 48 6 = 288

achieve housing density of 12 dwelling per 1000 m2 using three

Table 9
Total energy consumption by different house types in different urban densities
congurations oriented towards NS: four-oor single detached,
considering NS orientation. D.4(1), three-oor row house, R.3(1), and four-oor apartment
house, Al.4(2). Amongst these three congurations, the row house
Housing Density Total Average
conguration has the lowest height, and the lowest energy con-
conguration (du/1000 m2 ) normalized normalized
load (kW h/m2 ) load (kW h/m2 ) sumption too. The average normalized energy load for this housing
conguration is 100.2, compared to 110.9, and 110.4 kW h/m2 in
Single D.1(1) 3 87.6 102.2
detached D.2(1) 6 102.1 the single detached and low-rise congurations, respectively. This
D.3(1) 9 108.3 is further illustrated in Fig. 5.
D.4(1) 12 110.9

Row R.1(1) 4 78.5 91.4 4. Conclusion

house R.2(1) 8 93.4
R.3(1) 12 100.2
The global concern about depletion of non-renewable energy
Low- Al.3(2) 9 110.3 113.6 resources is increasing. As a result, the current global research
rise Al.4(2) 12 111.8 efforts pay signicant attention to the application of strategies that
Al.3(4) 11 113.7 improves sustainability in our built environment. This includes sus-
Al.4(4) 14 118.5
tainable urban planning, and impact of urban pattern on energy
High- Ah.6(4) 22 119.5 127.1 consumption. Housing density is an important parameter in this
rise Ah.8(4) 29 120.4 regard, which signicantly affects urban planning of residential
Ah.10(4) 36 122.9 neighbourhoods. Thus, the main aim of this study is to highlight
Ah.12(4) 43 121.4
the impact of housing density on energy consumption considering
different urban congurations.
This has been carried out using computerized simulation to
compare different parametric hypothetical cases of different hous-
ing congurations. This included single detached housing, row
housing, and low and high-rise apartment buildings. Examined
densities ranged from 3 to 43 housing unit per 1000 m2 . In general,
it has been found that energy consumption in residential buildings
located in hot climates is highly dependent on residential density. In
general, energy consumption in the NS orientation is slightly higher
than the EW one due to the observed reduction in cooling loads.
This is a direct result of changing the orientation, where the site has
less exposure to south. The observed relationship between housing
density and energy consumption is true in two cases: different den-
sities within the same housing type; and the same density achieved
by different housing types.
In the rst case, results showed that energy loads are propor-
tional to housing density increase. This is mainly caused by the
Fig. 5. A comparison between annual energy consumption in three housing cong- observed increase in cooling loads, where increasing density is
urations at a housing density of 12 du/1000 m2 .
restricts passive heat loss from buildings. This increase is more
138 O.S. Asfour, E.S. Alshawaf / Energy and Buildings 91 (2015) 131138
signicant in relatively low densities presented in single detached
and row houses (312 dwelling unit per 1000 m2 ). In the second
case, it has been found that the examined row houses congu-
ration offers a reduction in the average energy consumption by
12%, 19.5%, and 28% compared to the single detached, low-rise,
and high-rise congurations, respectively. Thus, compact horizon-
tal housing congurations can perform better in terms of energy
efciency when compared to the vertical ones. This shows that
despite the fact that vertical congurations provide higher den-
sities, this occurs on the account of passive heating in winter and
passive cooling in summer, which increases energy consumption.
Finally, the research recommends further investigation using other
modelling programs and experimental methods to highlight the
effect of housing density on energy consumption considering more
detailed climatic conditions, and to implement the ndings as hous-
ing strategies that promote energy efciency in buildings design
and planning.
References
[1] J. Heinonen, S. Junnila, Residential energy consumption patterns and the overall
housing energy requirements of urban and rural households in Finland, Energy
Build. 76 (2014) 295303.
[2] M. Bell, R. Lowe, Energy efcient modernisation of housing: a UK case study,
Energy Build. 32 (2000) 267280.
[3] B. Edwarwds, P. Hyett, Rough Guide to Sustainability, RIBA, London,
2002.
[4] R. Untermann, R. Small, Site Planning for Cluster Housing, Van Nostrand Rein-
hold, New York, NY, 1977.
[5] K. Steemers, Energy and the city: density, buildings and transport, Energy Build.
35 (1) (2003) 314.
[6] A. Allam, City Planning, Anglo Egyptian Library, Cairo, 1998.
[7] C. Ratti, N. Baker, K. Steemers, Energy consumption and urban texture, Energy
Build. 37 (2005) 762776.
[8] P. Littlefair, Daylight, sunlight and solar gain in the urban environment, Sol.
Energy 70 (2001) 177185.
[9] C. Georgakis, M. Santamouris, Experimental investigation of air ow and tem-
perature distribution in deep urban canyons for natural ventilation purposes,
Energy Build. 38 (2006) 367376.
[10] S.C.M. Hui, Low energy building design in high density urban cities, Renewable
Energy 24 (2001) 627640.
[11] C. Hachem, A. Athienitis, P. Fazio, Investigation of solar potential of housing
units in different neighborhood designs, Energy Build. 43 (2011) 22622273.
[12] J. Strmann-Andersen, P.A. Sattrup, The urban canyon and building energy use:
urban density versus daylight and passive solar gains, Energy Build. 43 (2011)
20112020.
[13] I. Knight, S. Stravoravdis, S. Lasvaux, Assessing the operational energy pro-
les of UK education buildings: ndings from detailed surveys and modeling
compared to measured consumption, in: 2nd PALENC Conference and 28th
AIVC Conference on Building Low Energy Cooling and Advanced Ventilation
Technologies in the 21st Century, Crete Island, Greece, 2007, pp. 531536.
[14] N. Sada, E. Salleh, L.C. Haw, Z. Jaafar, Evaluating thermal effects of internal
courtyard in a tropical terrace house by computational simulation, Energy
Build. 43 (2011) 887893.
[15] F. Faizi, M. Noorani, A. Ghaedi, M. Mahdavinejad, Design an optimum pattern
of orientation in residential complexes by analyzing the level of energy con-
sumption (case study: Maskan Mehr complexes, Tehran, Iran), Procedia Eng.
21 (2011) 11791187.
[16] A. Rashwan, O. Farag, W.S. Moustafa, Energy performance analysis of integrat-
ing building envelopes with nanomaterials, Int. J. Sustainable Built Environ. 2
(2013) 209223.
[17] L. Yang, B. He, M. Ye, Application research of Ecotect in residential estate plan-
ning, Energy Build. 72 (2014) 195202.
[18] J. Goussousa, H. Siama, H. Alzoubib, Prospects of Green Roof Technology for
Energy and Thermal Benets Inbuildings: Case of Jordan, Sustainable Cities
and Society, 2014, http://dx.doi.org/10.1016/j.scs.2014.05.012 (in press).
[19] J.S. Carlos, M.C. Nepomuceno, A simple methodology to predict heating load at
an early design stage of dwellings, Energy Build. 55 (2012) 198207.
[20] A. Oikonomou, F. Bougiatioti, Architectural structure and environmental per-
formance of the traditional buildings in Florina, NW Greece, Build. Environ. 46
(2011) 669689.
[21] Weatherbase.com, Records and Averages for Middle East, 2010, Available at:
www.weatherbase.com [Accessed 31 Mar. 2014] [online].
[22] CIBSE, Energy Efciency in Buildings, CIBSE, London, 1998.