International Journal of Architectural Heritage

Conservation, Analysis, and Restoration

ISSN: 1558-3058 (Print) 1558-3066 (Online) Journal homepage: http://www.tandfonline.com/loi/uarc20

Seismic risk assessment of the historical urban

areas of Annaba city, Algeria

Allaeddine Athmani, Tiago Miguel Ferreira & Romeu Vicente

To cite this article: Allaeddine Athmani, Tiago Miguel Ferreira & Romeu Vicente (2018) Seismic
risk assessment of the historical urban areas of Annaba city, Algeria, International Journal of
Architectural Heritage, 12:1, 47-62, DOI: 10.1080/15583058.2017.1370508

To link to this article: https://doi.org/10.1080/15583058.2017.1370508

Accepted author version posted online: 03

Oct 2017.
Published online: 03 Oct 2017.

Submit your article to this journal

Article views: 106

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

http://www.tandfonline.com/action/journalInformation?journalCode=uarc20

Download by: [CERIST] Date: 10 December 2017, At: 05:19

 INTERNATIONAL JOURNAL OF ARCHITECTURAL HERITAGE
2018, VOL. 12, NO. 1, 47–62
https://doi.org/10.1080/15583058.2017.1370508

Seismic risk assessment of the historical urban areas of Annaba city, Algeria
Allaeddine Athmania, Tiago Miguel Ferreira b
, and Romeu Vicentec
a
Laboratory of Civil Engineering, Department of Civil Engineering, University of Badji Mokhtar – Annaba, Annaba, Algeria; bISISE - Institute
for Sustainability and Innovation in Structural Engineering, Department of Civil Engineering, University of Minho, Guimarães, Portugal;
c
RISCO – Department of Civil Engineering, Campus Universitário de Santiago, University of Aveiro, Aveiro, Portugal

ABSTRACT ARTICLE HISTORY

To improve the resilience of old building stocks to seismic events, it is absolutely essential to Received 8 March 2017
understand the seismic risk associated with these built environments in a reliable and accurate way. Accepted 12 August 2017
Three variables, among others, that compose the mathematical description of the seismic risk KEYWORDS
include: vulnerability, hazard and exposure. Vulnerability is of great importance for technicians loss estimation; old buildings;
and decision makers, because, of the three variables, it is the only one where engineering research old city center of Annaba; risk
can mitigate problems. An example includes evaluating the seismic vulnerability of existing con- mitigation; seismic scenarios;
structions to determine the need for retrofitting solutions. From this basis, this article addresses and urban area
discusses the seismic vulnerability associated to the old urban areas of the Algerian city of Annaba.
To perform this task, different seismic scenarios were developed using urban-scale analysis, where
Downloaded by [CERIST] at 05:19 10 December 2017

the vulnerability index method employed was derived from the Italian GNDT approach. About 380
of the historical masonry buildings were assessed using the selected methodology, where the aim is
to evaluate the expected physical damage as well as the relevant economic and human losses for
different macroseismic intensities. For depicting and analyzing the obtained results, a Geographic
Information System (GIS) tool is used, where the ultimate goal is to support the city council of
Annaba in implementing risk mitigation and emergency planning strategies.

1. Introduction potential of a buildings to survive a certain damage

level when subjected to a seismic event of defined
Assessing the impact of an earthquake upon an urban
intensity (Lang and Bachmann 2003). According to
area is of fundamental importance for several reasons,
Ferreira et al. (2013), this parameter indirectly mea-
but perhaps the most important of them is that disaster
sures the reduction in a building’s structural efficiency,
impact scenarios can be subsequently used by decision
or a building’s residual ability to function under speci-
makers to define preventive measures aiming at miti-
fied seismic conditions. Finally, Exposure reflects the
gating the social and economic consequences of a seis-
value of the exposed elements at risk (Ferreira et al.
mic event. Consequently, bolstering the resiliency of
2013).
communities to such hazardous events should begin
This article is divided into three sections, where each
with a comprehensive risk assessment. The concept of
one is related to the three variables that define the
seismic risk, as defined in the past by several research-
seismic risk. The first part consists of identifying and
ers, namely by Caicedo et al. (1994), Cardona (2001),
characterizing the earthquake events that occurred in
Coburn and Spence (2002), McGuire (2004), Barbat
Annaba city and in the surrounding regions, some of
et al. (2010), or Maio et al. (2016), can be defined as a
which demonstrated significant impact, and especially
mathematical convolution between hazard, vulnerabil-
on the old building stock in Annaba city (Benouar
ity and exposure (Equation (1)):
1994, Harbi and Maouche 2009, Harbi, Meghraoui,
Risk ¼ Hazard  Vulnerability  Exposure: (1) and Maouche 2011). Due to degraded state of these
buildings, they are highly vulnerable to seismic actions,
The Hazard refers to the latent danger or the exter-
even those of moderate intensity. To estimate the vul-
nal risk factor for an exposed element, which can be
nerability of the building stock, the use of a reliable but
expressed as the probability of an event occurring with
expedite assessment method, which accounts for build-
a certain intensity at a specific geographical location.
ings features that most influence the seismic
The Vulnerability can be defined as the intrinsic

CONTACT Allaeddine Athmani [email protected] Laboratory of Civil Engineering, Department of Civil Engineering, University of
Badji Mokhtar – Annaba, P.O. Box 12, 23000 Annaba, Algeria.
Color versions of one or more of the figures in the article can be found online at www.tandfonline.com/UARC.
© 2017 Taylor & Francis
 48 A. ATHMANI ET AL.

vulnerability, is an essential step in a meaningful seis- had a maximum intensity of VIII and a surface-wave
mic risk analysis, where the latter constitutes the sec- magnitude (Ms) of 5.2, which violently shook the build-
ond part of this article. The third section estimates the ings and frightened their inhabitants (Benouar 1994).
expected damage levels in the historical masonry build- The cracking of tiles, ceilings, walls and wooden floors
ings located in the old city center of Annaba. This is were heard in the area, and large fissures (of about
achieved using seismic scenarios with different macro- 10 cm) were reported in the military hospital’s struc-
seismic intensities (Grünthal 1998). Economical and ture (Benouar 1994). Moreover, earthquakes from
human losses are also presented and discussed. The Constantine, such as the one that occurred on August
different outputs obtained were then mapped using a 4, 1908 (Ms = 5.2, I0 = VIII), August 6, 1947 (Ms = 5.0,
GIS tool connected to a relational database, which I0 = VIII), and October 27, 1985 (Ms = 6.0, I0 = IX),
enables the storage and the spatial analysis of the results have also devastated Annaba city (Benouar 1994;
into an open, georeferenced and fully upgradable Bounif, Haessler, and Meghraoui 1987; Harbi,
environment. Peresan, and Panza 2010). Other cities were also
Finally, it is important to underline that the main affected. Despite the distance from Annaba city,
objective behind this research is to create a framework destructive events, such as those occurring at Djidjelli
for decision makers, particularly the ones on the city (now called Jijel) on 21 (I0 = VIII) and 22 (I0 = IX)
council in Annaba city, to support the evaluation of August 1856, were strong enough to cause significant
and the potential adoption of risk reduction measures, cracks and collapse buildings (see Harbi, Meghraoui,
emergency response planning and post-disaster and Maouche (2011) for additional details).
Downloaded by [CERIST] at 05:19 10 December 2017

recovery. Recently, three segments of a thrust fault were seis-

mically active, where all three possess similar lengths,
and are denoted in Figure 1 as S1 (30 km), S2 (30 km),
2. Seismic risk at Annaba city and S3 (20 km). In 2005, the Maradja2 company used
their framework (Kherroubi et al. 2009) to identify and
2.1 Characterization of the seismic hazard locate these thrust faults under the Annaba sea. A
Annaba city is close to two active seismogenic zones, seismic event with moment-magnitude Mw = 7.5 is
Guelma and Constantine, which could have severely implied when assuming, based upon the scaling rela-
damaging effects on Annaba city (Harbi, Maouche, tions between the fault surface displacements and
and Benhallou 2003; Mourabit et al. 2014). For lengths and if the faults rupture in a single event. This
instance, the strongest-known earthquake is from the actual value may be even higher if segment S4 is also
Guelma region, and occurred on February 10, 1937. It considered in this analysis (Kherroubi et al. 2009).

Figure 1. Annaba sea’s active faults and the spatial distribution of north-eastern, Algerian earthquakes from 267 to 2010. The map
was designed after that from Athmani et al. (2015) (squares denote historical earthquakes with I0 ≥ IV, while circles represent
instrumental earthquake with M ≥ 4.0).
 INTERNATIONAL JOURNAL OF ARCHITECTURAL HERITAGE 49

In terms of seismic intensity, Annaba city is classi- Table 1. The applied methodology.
fied as a moderate to low seismic hazard area (CGS Vulnerability Class,
Cvi weight
2011), where intensities are generally a VI and do not
PARAMETERS A B C D Pi
exceed VII (Yelles-Chaouche et al. 2006). The hazard
1. Structural building system
from peak ground acceleration, however, is significant. P1 Typology of resisting system 0 5 25 45 2.50
A recent study of the probabilistic seismic hazard P2 Organization of the resisting system 0 5 25 45 1.00
P3 Conventional strength 0 5 25 45 1.50
assessment, which was performed by the Algerian P4 Maximum distance between walls 0 5 25 45 0.25
National Earthquake Engineering Research Center P5 Number of floors 0 5 25 45 0.75
2. Irregularities and interactions
(CGS) for the Annaba region, indicated an appreciable P6 Location and soil conditions 0 5 25 45 0.75
0.2–0.32 g for a 500-year return period (CGS 2011). P7 Aggregate position and interaction 0 5 25 45 0.75
P8 Plan regularity 0 5 25 45 0.50
Although these numbers are debated, the results are of P9 Vertical regularity 0 5 25 45 0.50
a sufficiently high hazard level to compel us to access 3. Floor and Roof systems
P10 Horizontal diaphragms 0 5 25 45 1.00
the probable level of damages and the construction for P11 Roof system 0 5 25 45 0.25
different seismic scenarios within Annaba city, where 4. Conservation status and modifications
P12 Interventions 0 5 25 45 0.50
the historical urban areas are of interest. P13 General state of preservation 0 5 25 45 1.00
P14 Non-structural elements 0 5 25 45 0.25

2.2 Description of the applied seismic vulnerability

assessment method includes parameters that characterize a building’s
Downloaded by [CERIST] at 05:19 10 December 2017

resisting system, which determines its structural beha-

To assess the seismic vulnerability of the old masonry
viour. These parameters, for instance, represent the
building stock in Annaba city, a methodology based on
masonry type and quality through the material size,
the Italian GNDT II level approach (GNDT-SSN 1994)
shape, and stone type, the masonry fabric, and the
(GNDT 1993) was applied to the city’s urban areas. The
arrangement of walls and the level of connections
GNDT approach (GNDT 1993) is based on post-seis-
between walls are represented with Parameters P1
mic damage observation and survey data covering a
and P2, respectively. The building’s shear strength
vast number of elements, and focuses on the most
capacity is estimated in Parameter P3, while
important aspects and features that define building
Parameters P4 and P5, respectively evaluates the
damage. From this data set, a hybrid method was
potential out-of-plane collapse risk and a building’s
derived and adapted for the Algerian masonry build-
height. The second group of parameters focuses on
ings by Athmani et al. (2015), which used very detailed
the soil conditions under the building’s foundation in
building stock information to discuss and redefine the
Parameter P6, their relative location in the urban
criteria behind some of the most important parameters,
mesh and the soil’s interaction with other buildings
and to introduce new parameters that account for typi-
in Parameter P7, and its plan and height irregularities
cally overlooked building features.
are, respectively, represented through Parameters P8
As already mentioned, this seismic vulnerability
and P9. The third group of parameters evaluate the
method was adopted to fit the number of buildings
horizontal structural systems, namely the quality and
populating the old city center of Annaba. To estimate
type of the floor system are reflected in Parameter
their capacity to withstand seismic events, the current
P10, while the impulsive nature of the pitched roof
method uses a vulnerability index score obtained as the
system is included in Parameter P11:
weighted sum of 14 parameters, Equation (2). The
index scores were distributed into the following 4 vul- X
14
nerability classes Cvi that denote increasing vulnerabil- Iv ¼ Cvi  Pi (2)
i¼1
ity: A, B, C, and D. Each of the 14 parameters evaluate
one aspect of a building’s seismic response. Each para- Finally, the fourth group is composed of the follow-
meter is weighted, pi, with a numerical value ranging ing: (i) Parameter P12 assesses the construction anoma-
from 0.25, for the less important parameters from a lies that affect the structural behaviour of buildings; (ii)
structural vulnerability perspective, to 2.5 for the most Parameter P13 evaluates structural fragilities and the
influential parameters (see Table 1). The final vulner- conservation states of buildings; and (iii) Parameter
ability index, I*v is then normalized to fall within the P14 evaluates the negative influences of non-structural
range of 0–100 for ease of use. elements with poor connections to the main structural
As can be seen in Table 1, the 14 parameters are system.
arranged into 4 groups to emphasise their differences The most common parameters used to compute the
and their relative importance. The first group final vulnerability index, using the selected method,
 50 A. ATHMANI ET AL.

come directly from the CTC database. For some items, structures, has become an important issue in Annaba
however, the data is available in other sources, such as city. Indeed, as it was reported by the technical organ-
in the documents provided by the Direction of ism in control of the Annaba city council (CTC 2010),
Urbanism and Construction. Whereas, for a few other the first two months of 2005 experienced a series of
parameters, the information does not exist in the tech- tragic events in the historical city center of Annaba
nical check list, it cannot be inferred by the CTC that resulted in 49 collapse episodes. The most tragic
screener during the inspection process or it was collapse killed an entire family when an old building
checked and collected from the field (Athmani et al. (CTC 2010) in a historic district slumped.
2014). The method proposed here is considered robust Consequently, in response to the frequent complaints
and considers that the majority of building inspections of the inhabitants, the Direction of Urban Planning
provided sufficiently detailed and accurate geometrical and Construction of Habitation (DUCH) launched a
information for this analysis. Therefore, uncertainty in general program aimed at evaluating the vulnerability
the assignment of vulnerability classes to each para- of the old buildings in 12 of the 29 districts in the
meter can be considered low (Athmani et al. 2015). Annaba municipality, which have been declared as
Finally, the associated expected mean damage grade, historical and heritage areas. Because of this program,
µD, given by Equation (3), is computed as a function of a detailed field survey was conducted and provides
the macroseismic intensity, IEMS-98, which is described data for selected regions in Annaba city (Figure 2).
according by the EMS-98 scale (Grünthal 1998), the The survey was conducted by expert structural engi-
vulnerability index used in the Macroseismic Method neers from Annaba city’s Technical Control of
Downloaded by [CERIST] at 05:19 10 December 2017

(Milutinovic and Trendafiloski 2003), as expressed Construction organism (CTC 2010). They used check
through Equation (4), and a ductility factor, Q, that lists pertaining to survey items both outside and inside
describes the ductility of a specific construction typol- the envelope of each individual building. The main
ogy (using values that range between 1 and 4). A mean scope of this survey was the evaluation of each build-
value of 2.3 was adopted in this work and was chosen ing’s state of degradation based upon the experts’
based upon this parameter’s calibration by Vicente judgment; the purpose of this work was to assess
et al. (2011). It is important to highlight that the and support different types of intervention strategies.
approaches applied herein assume that, due to their Although the data was not originally developed for
constructive and structural characteristics, the seismic purposes, the special engineering expertise
Algerian buildings can be considered typologically provided on structural vulnerability is valuable and
similar to Euro-Mediterranean construction technology of great importance for obtaining valid risk outputs
used in the original framework (further discussion can (Athmani et al. 2015).
be found in Athmani et al. (2015)). This assumption is An example of the CTC’s datasheets is presented
also supported by the great compatibility and similarity in Figure 2, which outlines the type of information
found between the existing information contained collected during the field survey. The figure is sub-
within the study area (e.g., the CTC data) and the divided in three parts: (i) description; (ii) diagnosis;
main parameters used in the equations: and (iii) conclusions and recommendations. As
   observed in Figure 2, the CTC data was used to
IEMS98 þ 6:25  13:1
μD ¼ 2:5 1 þ tanh (3) classify buildings into one of the four conservation
Q
classes: a good state, slightly degraded, moderately
Thus, the vulnerability index, Iv, can be related to the degraded, and highly degraded. These four groups
vulnerability index, V, through Equation (4), which are arranged in order of severity to define interven-
enables calculation of the mean damage grade, µD, tion priorities. The first group consists of buildings
and estimation of the physical, human and economic that do not require any structural intervention,
losses (in Section 4): except for some possible minor repairs, to ensure
V ¼ 0:56 þ 0:0064  Iv (4) the safety of their inhabitants (CTC 2010). The sec-
ond group consists of buildings that need repair and/
or slight strengthening interventions, while the third
group consists of buildings that require repair and/or
2.3 Characterization of the old city center of
serious rehabilitation intervention (e.g., moderate to
Annaba
heavy) (CTC 2010). Finally, the fourth group consists
2.3.1 Building inspection and appraisal of buildings that need urgent strengthening interven-
The necessity for assessing the seismic risk of existing tion or that should be demolished and reconstructed
urban areas, in terms of the damage potential to the (CTC 2010).
 INTERNATIONAL JOURNAL OF ARCHITECTURAL HERITAGE 51
Downloaded by [CERIST] at 05:19 10 December 2017

Figure 2. Example of the parameters that compose the CTC data sheets (CTC 2010).

2.3.2 Context of the study area most interesting of them, called “Place d’arme,” shelters
Annaba city is one of the most urbanized and densely about 12,000 people (DUC 2006).
populated Algerian cities, and is currently the largest Place d’arme is one of the first settlement areas in
industrial and commercial hub of the north-eastern Annaba, which started in the Arab-Turkish period in
region of Algeria (see Figure 3). The population is the 18th century. Later, during the colonial French era
mostly concentrated within the old urban areas; the between 1830 and 1962, the town experienced a large
 52 A. ATHMANI ET AL.

results of the inspections have shown that the buildings are

being used for a variety of purposes (Figure 4a)—such as
administrative, scholarly, cultural or other uses that include
military and production centers. In addition, the built-up
area is dominated by the multi-family residential buildings,
which are usually fully occupied (Figure 4b). The ground
story of a considerable number of those buildings is used for
commercial ends. It is worth noting that this study does not
account for the monumental, scholarly or the cultural
buildings, which must be subjected to a more specific
analysis.
Figure 5a shows that two typologies dominate the
materials constituting the resisting system: rubble, stone,
and adobe. Some of these buildings present a regular
structural layout with thick walls that usually range from
Figure 3. Aerial view of the old city center of Annaba (DUC 2006).
40–50 cm in thickness. As for the floor system, steel
beams are used to support brick masonry arches, while
expansion and modifications over the existing building timber floor construction includes timber beams covered
blocks (DUC 2006). Some of old buildings were with timber planks, ballast fill, and tile flooring. In the
Downloaded by [CERIST] at 05:19 10 December 2017

replaced by colonial ones. Due to this fact, a large CTC data, building floors with two or more material types
portion of the buildings in this urban area are a mixture are marked as a mixed structure. Moreover, reinforced
of traditional and colonial structures (DUC 2006). It is concrete and stone or brick vaults floors are also observed
important to stress that this frame in Figure 3 is repre- in the old city center of Annaba (Figure 5b).
sentative of most historical urban areas in Algeria. The statistical data presented in Figure 6a shows that
The area selected in this work is composed only of old the majority of the assessed buildings are in very poor
masonry structures, whereas the existing CTC data covers condition, and reveals the presence of high levels of
380 building surveys (out of 602 entries) (CTC 2010). The deterioration in their interior and the exterior

(a) Usage (b) Occupation

Figure 4. Distribution of buildings stock according to their usage type and occupation level (Athmani et al. 2014).

(a) Load bearing wall (b) Floor

Figure 5. Distribution of buildings stock according to their structural compound systems (Athmani et al. 2014).
 INTERNATIONAL JOURNAL OF ARCHITECTURAL HERITAGE 53

(a) Degradation state (b) Number of floors

Figure 6. Building distribution based upon their degradation state and the number of storeys (Athmani et al. 2014).

envelopes. In addition, the majority of these buildings values obtained were 23.67 and 78.98, respectively.
were built with poor quality mortars and low resistance Finally, it is worth highlighting that about 90% of the
materials. Despite most of them having 2 or 3 stories, it buildings have a vulnerability index over 45, which is
is also possible to find buildings with more than four equivalent to a class A vulnerability (Giovinazzi and
stories (see Figure 6b) that were built during the French Lagomarsino 2004). Figure 7 shows the spatial distribu-
Downloaded by [CERIST] at 05:19 10 December 2017

colonial period. tion of building stock as a function of their vulnerabil-

The geological characteristics of Annaba city reveal ity index. According to Vicente et al. (2011), it is
that the underlying ground consists of three main geolo- recommended that masonry buildings with Iv values
gical rock units: the upper part of the town is character- over 45, i.e., buildings where severe damage and poten-
ized by “Gneiss,” whereas the majority of the area from tial local collapse are expected, should be subjected to a
the middle to the northwest part is above “Micaschistes” further assessment involving a more detailed approach.
and “Cipolins” layers. Throughout the entire study area, It is worth stressing that such seismic vulnerability
the soil conditions can be considered stiff to very stiff, maps enable the identification of the more vulnerable
which yield an average capacity of 200 kPa (DUC 2006). buildings, which, as previously mentioned, represents a
As a result, it was assumed in the present study that the valuable output for preliminary organization of urban
spatial distribution of damage should not be significantly management and protection strategies. Moreover, they
affected by site conditions, but rather, that the distribu- contribute to the development of a more supported and
tion mainly reflects building vulnerabilities. effective risk management environment, such as
through the identification of weakness in the exposed
elements and their evaluation in time, and the analysis
3. Mapping and scenarios
of global and individual impacts when implementing
The preparation of damage scenarios is the compre- different strengthening strategies (as discussed by
hensive final step that, upon combining building vul- Ferreira, Maio, and Vicente (2017)).
nerability and earthquake intensity data, returns an Once the vulnerability is defined by means of
estimation of the building damage and loss scenarios Equation (3), the mean damage grade, μD, can be eval-
(e.g., human casualties or economic losses). These esti- uated for different macroseismic intensities using
mations are of fundamental importance for seismic risk Equation (4). Thus, the vulnerability curves for the
mitigation and emergency planning and management historical masonry buildings of the built-up area
(Chiauzzi et al. 2012). under analysis were computed for the mean value of
the vulnerability index, Iv,mean, and were affected by its
standard deviation value when calculating the upper
3.1 Seismic vulnerability assessment and lower bounds (Iv,mean-2σIv; Iv,mean-σIv; Iv,mean; Iv,
As previously mentioned, the masonry building stock mean+σIv; Iv,mean+2σIv) for events with different macro-
of the old city centre of Annaba was assessed herein. A seismic intensities defined by the EMS-98 scale
vulnerability index was assigned to each building (Grünthal 1998). These vulnerability curves provide
according to the applied method (see Equation (2)). global appreciation of the estimated damage grade,
The mean vulnerability index value of the overall sam- μD, for the entire buildings stock (Figure 8). For
ple, Iv,mean, is 57.86, and the standard of deviation is instance, when considering an expected earthquake of
9.39. In addition, the minimum and maximum index a relatively moderate intensity of VII, 56% of the
 54 A. ATHMANI ET AL.
Downloaded by [CERIST] at 05:19 10 December 2017

Figure 7. Spatial distributions of the vulnerability index according to the applied method (Athmani et al. 2015).

buildings that possess a vulnerability index equal to or probability mass function and the beta probability density
greater than the mean value, Iv,mean, would probably function given by Equation (5):
incur significant damage (2 < μD ≤ 3). For the high
ΓðtÞ ðxaÞr1 ðbxÞtr1
macroseismic intensity of IX (IEMS-98 = IX), the same PDF : Pβ ðxÞ ¼ ΓðrÞΓðtrÞ ðbaÞt1
a  x < b ; (5)
number of buildings are expected to suffer a near
collapse to complete collapse (4 < μD ≤ 5). where a, b, t, and r are geometric parameters associated
As will be addressed below, the comprehensive seis- with the damage distribution and Γ is the gamma
mic vulnerability assessment of an old urban center, function. Research carried out by Giovinazzi (2005)
with the purpose of mitigating earthquake risks,
has shown that the beta distribution is the most versa-
assumes great importance in preventing physical,
tile of these parameters, where the shape of the distri-
human, economical and patrimonial losses in these
bution is controlled via the parameters t and r. This
areas.
function enables the fitting of both very narrow and
broad damage distributions. The mean value, μx , is
3.2 Assessment of the damage distribution defined in Equation (6) as a function of the same
probability parameters as follows if x is a continuous variable that
ranges between a and b with variance, σ 2x , Equation (7):
To assess the damage rates for different macroseismic
intensities, seismic damage needs to be quantified and r
μx ¼ a þ ðb  aÞ: (6)
measured in a standard manner. The most common t
approach to quantifying seismic damage rates is to per- r ðt  r Þ
form fragility analyses (Askan, Erberik, and Bilal 2014). σ 2x ¼ ðb  aÞ2 þ (7)
t 2 ðtþ 1Þ:
Indeed, once the expected mean damage grade is known,
then different damage distribution histograms can be If t = 8, a = 0, and b = 6, the probability associated
defined using a probabilistic approach. The most-com- with damage grade k is calculated in Equation (8),
monly applied methods are based on the binomial where k 2 ½0; 5:
 INTERNATIONAL JOURNAL OF ARCHITECTURAL HERITAGE 55

pk ¼ Pβ ðk þ 1Þ  pβ ðkÞ: (8) accordance with the vulnerability curves, Figure 8,

obtained from the entire building stock. As for two
addressed seismic intensity scenarios, IEMS-98 = VII and
Following this definition, the mean damage grade,
IEMS-98 = IX, respectively, represent moderate to signifi-
μD , and the mean value of the beta distribution, μx , can
cant damages (D2–D4) from an earthquake of moderate
be correlated through the third-degree polynomial in
intensity (IEMS-98 = VII), and severe damages to total
Equation (9): collapse (mainly D4–D5) for events of high intensity
μx ¼ 0:042μ3D  0:31μ2D þ 1:72μD : (9) (IEMS-98 = IX). Another way to represent damage using
the damage distribution histograms involves the use of
Following Giovinazzi (2005), Equations (6) and (9) fragility curves. Fragility curves are widely used to esti-
are used to correlate parameters t and r with the mean mated damage and to define the probability of reaching or
damage grade, μD in Equation (10), which are then exceeding a certain damage grade or state, Dk ðk 2 ½0; 5Þ.
represented as a function of the vulnerability index, Iv, Fragility curves can be obtained directly from the physical
and the macroseismic intensity, IEMS-98, through building damage distributions derived from the beta
Equation (3), while the variance of the distributions is probability function as follows (Milutinovic and
defined using the previously assumed value of 8 for Trendafiloski 2003):
parameter t (Vicente et al. 2011): PðD  Dk Þ ¼ 1  pβ ðkÞ: (11)
r ¼ tð0:007μ3D  0:052μ2D þ 0:2875μD Þ (10) As with the vulnerability curves, fragility curves define
Downloaded by [CERIST] at 05:19 10 December 2017

the relationship between earthquake intensity and

Figure 9 presents the results of the damage distribution damage in terms of the conditional cumulative probabil-
for the mean value of the vulnerability index for the entire ity of reaching a certain damage state. Figure 10 presents
built-up area under study. The results in Figure 9 are in both probability curves for different damage grades and
fragility curves obtained for the mean vulnerability index
of the buildings stock, Iv,mean = 57.86.
As discussed by Ferreira, Vicentea, and Varum
(2014), these results can be considered a primary tool
for developing and implementing risk mitigation and/
or seismic retrofitting strategies at the urban scale.

4. Loss scenarios
Loss estimation plays an important role in the imple-
mentation of urban planning and retrofitting strategies,
and enables costs to be placed alongside various ben-
eficial measures, such as repair costs and life safety
(D’Ayala et al. 1997). The loss estimation models are
Figure 8. Seismic vulnerability curves for the buildings stock inevitably dependent on the physical damage grades,
according to the applied method. including the definition of correlations between both

(a) IEMS-98=VII (b) IEMS-98=IX

Figure 9. Distribution of the damage probability for seismic scenarios with two different intensities.
 56 A. ATHMANI ET AL.

Figure 10. Damage to the buildings stock represented by (a) probability and (b) fragility curves.

the probability of a certain damage’s occurrence and of buildings deemed unusable after minor and moder-
the probability of different loss phenomena (Ferreira ate earthquakes. Although such events produce lower
2010). Consequently, loss assessments either in terms of levels of structural and non-structural damage, higher
collapsed or unusable buildings that stem from the lack mean damage grade values are associated with a higher
of structural safety conditions, as well as related death probability of building collapse. The probabilities asso-
and severe injuries, are in direct relation with the ciated with the occurrence of a certain damage grade
Downloaded by [CERIST] at 05:19 10 December 2017

damage scenarios calculated above. In this work, the are used in the loss estimation model and are affected
loss estimation results are organized and discussed by multiplier factors, which range from 0–1. Equations
through the construction of damage scenarios based (12) and (13), respectively, were used for the analysis of
on the global probabilistic distributions obtained for collapsed and unusable buildings:
the 380 buildings evaluated.
The development of GIS tools in recent years has Pcollapse ¼ p5 (12)
highlighted their potential for the management of data
(Vicente et al. 2011). Building vulnerability assessment Punusablebuildings ¼ p3  wub;3 þ p4  wub;4
and loss evaluation algorithms (mathematical and þ p5  wub;5 (13)
probability functions) were also integrated into a GIS
tool to enhance the entire analytical process. As already where p3, p4, and p5 are the probability of occurrence
noted in the introductory section, the example tool for damage D3, D4, and D5, whereas wub,3, wub,4, and
developed in this work represents the implementation wub,5 are the weight factors for this probability func-
of a workable and progressive tool that integrates all tion, and are thought to provide the percentage of
seismic risk evaluation from building characteristics to uninhabitable dwellings for each damage grade.
estimation of economic loss. This procedure enabled Following destructive seismic events in Italy,
rapid data layering and editing of building information, Bramerini et al. (1995) indicated that, based upon sur-
inter-correlation of building feature vulnerability, phy- veyed data, all buildings with a damage level, μD, equal
sical risk, and loss estimation results and supports to or greater than 4 (D4 and D5) and a portion of the
responsible risk-management actions and decisions. buildings with damage level, μD, equal to 3 (40%) are
This GIS tool not only enables scenario modeling and assumed to be unusable. Under this assumption, the
simulation, it also enables the impact forecasting of unusable buildings rate is calculated herein, highlight-
retrofitting strategies in the reduction of vulnerability ing its value at each seismic intensity, without account-
and potential economic losses. ing for the rate of collapsed buildings, which are
emphatically unusable and uninhabitable. Therefore,
by eliminating the probability rate of collapse from
4.1 Collapsed and unusable buildings
Equation (13), which was already evaluated in Pcollapse
The most frequently employed approaches in the Euro- for the built-up area under study, the value assigned to
Mediterranean region are based on observed damage each factor was wub,3 = 0.4, wub,4 = 1.0, and wub,5 = 0.
data after destructive earthquake events. Accordingly, Figure 11 shows the results of building collapse and
the loss estimation model adopted in this work was unusable building estimations for the mean value of the
proposed by Servizio Sísmico Nazionale (SSN) and vulnerability index, Iv,mean, as well as for other represen-
based on the work of Bramerini et al. (1995), who tative values of vulnerability, namely Iv,mean-2σIv, Iv,mean-
approached the data analysis through the probability σIv, Iv,mean, Iv,mean+σIv, and Iv,mean+2σIv.
 INTERNATIONAL JOURNAL OF ARCHITECTURAL HERITAGE 57

Figure 11. Estimates of the probability of (a) collapsed and (b) unusable buildings rates for different seismic scenarios.

Table 2. Results of the collapsed and unusable buildings for different macroseismic scenarios.
Intensity, I (EMS-98)
Total number of the assessed buildings: 380 VII VIII IX X
Collapsed buildings 3 (0.84%) 37 (9.67%) 161 (42.45%) 298 (78.45%)
Unusable buildings 82 (21.53%) 179 (47.21%) 168 (44.31%) 72 (18.94%)
Total of affected buildings 85 (22.37%) 216 (56.88%) 329 (86.76%) 370 (97.39%)
Downloaded by [CERIST] at 05:19 10 December 2017

It is worth highlighting that, as the number of buildings Using the developed GIS tool, an evaluation of the
that suffer collapse increases with the increase of the seis- collapsed and unusable buildings for seismic scenarios
mic intensity, the number of unusable buildings tends to with intensities of VII and IX can be presented for each
decrease (see Table 2). More representative and explicative building, where a single ID code is associated with each
results are listed in Table 2 for seismic events with moder- polygon on the map. Since the information is linked to
ate to strong intensities (e.g., VII, VIII, IX. and X) all building data, it is also possible to combine different
(Grünthal 1998) and for the mean value of the estimated data layers, as shown in Figures 12 and 13, where the
vulnerability obtained for the 380 buildings evaluated in probability of collapse and the number of inhabitants
the old city center of Annaba. As can be seen in Table 2, are plotted together. Note that the results combining
22.37% of the analyzed constructions would be affected by high occupancy rates and building collapse probabil-
a seismic event with an intensity of VII, and more than the ities higher than 80% are especially worrisome.
half of the assessed buildings (56.88%) would be severely
affected by a seismic event with an intensity of VIII.
Moreover, for intensities of IX and X, this percentage 4.2 Human casualties and homeless
would increase to approximately 86.76% and 97.39% A very serious consequence of an earthquake is the loss
respectively, which represents a scenario wherein the of human life and thus one of the major goals of all risk
region would be completely destroyed. mitigation strategies is ensuring human safety.

Figure 12. Estimates of the probability of (a) collapse and (b) unusable buildings rates for a seismic scenario of IEMS-98 = VII.
 58 A. ATHMANI ET AL.

Figure 13. Estimates of the probability of (a) collapse and (b) unusable building rates for a seismic scenario of IEMS-98 = IX.

Additionally, modeling earthquake casualties (dead and Following the same logic used to present and discuss the
severely injured) is fundamental not only for emer- buildings loss, the two rates are discussed next: the prob-
gency response management and for mitigation strat- ability of casualties, which is computed through Equation
Downloaded by [CERIST] at 05:19 10 December 2017

egy planning, but also for health preparedness planning (14), and the probability of homeless people that relies
(Giovinazzi 2005). Among the various casualty rate upon Equation (15). For this purpose, the value of each
analyses and correlation laws found in the literature, multiplier factor was wub,3 = 0.4, wub,4 = 1.0, and
those developed by Coburn et al. (1992), Tiedemann wub,5 = 0.7:
(1989), Hazus (1999) and Bramerini et al. (1995) are
Pdead and severely injured ¼ p5  0:3 (14)
the most frequently cited. Once again, the Servizio
Sismico Nazionalle proposal (Bramerini et al. 1995) Phomeless ¼ p3  wub;3 þ p4  wub;4 þ p5
was used here to guarantee the typological consistency
 wub;5 (15)
of the loss assessment procedure. Within this approach,
estimating the rate of dead and severely injured refers Figure 14 shows the estimation of the number of
only to collapsed buildings with a percentage of 30%, deaths, severe injuries and homeless associated with the
where survivors are assumed to seek short-term shel- mean value of the vulnerability index, Iv,mean, and the
ters. In this context, shelter need is another important previously presented vulnerability values (Iv,mean-2σIv,
parameter following an earthquake. Iv,mean-σIv, Iv,mean, Iv,mean+σIv and Iv,mean+2σIv).
This study also aims to estimate the homeless popula- From the explanation above, it is important to note
tion, which is associated with the uninhabitable building that the number of homeless decreased with the growth
rate computed previously. The rate of unusable buildings of the seismic intensity, as the dead and injured popu-
can be therefore combined with demographic data to lations increased. Further results are summarized in
quantify the number and composition of population Table 3 for seismic events with moderate to strong
requiring short-term shelter (Giovinazzi 2005). intensities (e.g., VII, VIII, IX, and X) (Grünthal 1998),

Figure 14. Estimates of the probability of (a) casualty and (b) homeless rates for different seismic scenarios.
 INTERNATIONAL JOURNAL OF ARCHITECTURAL HERITAGE 59

Table 3. Results of the dead and severely injured, and homelessness inhabitants for different macroseismic scenarios.
Intensity, I (EMS-98)
Total population of the assessed buildings: 8255 VII VIII IX X
Dead and severely injured 21 (0.25%) 240 (2.90%) 1051 (12.73%) 1943 (23.54%)
Homeless 1826 (22.12%) 4456 (53.98%) 6111 (74.02%) 6097 (73.85%)
Total of affected inhabitants 1847 (22.37%) 4696 (56.88%) 7162 (86.75%) 8040 (97.39%)

and for the mean value of the estimated vulnerability Using the developed GIS tool, the evaluated loss in
obtained for the 380 buildings evaluated in the old city terms of dead and severely injured and homeless are
center of Annaba. In Table 3, the intensities VII–X evaluated. The probability of homelessness may be
show that 22.37% and 97.39% of inhabitants were combined with the number of inhabitants of each
affected, and stresses the importance of efficient logis- building, where Figures 15 and 16 show the loss esti-
tical preparation by competent authorities for the tem- mation results for macroseismic scenarios with inten-
porary relocation of a potentially high number of sities of VII and IX, respectively.
residents. It is critical that Civil Protection bodies and
responsible regional authorities be aware of these out-
puts so that they can perform simulation exercises and 4.3 Economic loss and repair cost estimation
prioritize all logistical and financial resources that The correlation between damage grades and the repair
would inevitably be necessary in such a scenario.
Downloaded by [CERIST] at 05:19 10 December 2017

costs adopted in the present work, is obtained from the

Figure 15. Estimates of the probability of (a) dead and severly injured and (b) shelter requirements or homelessness rates for a
seismic scenario of IEMS-98 = VII.

Figure 16. Estimates of the probability of (a) dead and severly injured and (b) shelter requirements or homelessness rates for a
seismic scenario of IEMS-98 = IX.
 60 A. ATHMANI ET AL.

analysis of the data collected, using the GNDT-SSN pro- IX confirm the rather high vulnerability of the building
cedure established by Dolce et al. (2006). The repair cost stock. It should also be noted that the total repair cost
probabilities P[R|I] for a certain seismic event character- ratios, in terms of building area and the cost of replacing
ized by an intensity I, can be obtained from the product of the entire building stock, are 23.55%, 49.69%, and 78.20%
the conditional probability of the repair cost for each for seismic intensities of IEMS-98 = VII, VIII and IX
damage level, P[R|Dk] (Dolce et al. 2006), with the condi- respectively.
tional probability of the damage condition for each level
of building vulnerability and seismic intensity, P[Dk|Iv,I],
given by Equation (16) (Vicente et al. 2011): 5. Conclusions
5 X
X 100
The analysis of seismic risk in Annaba city provided an
P½PjI ¼ P½RjDk   P½Dk j Iv ; I (16)
Dk ¼1 Iv ¼0
opportunity to develop a tool for constructing a scenario
with proper characteristics for the urban areas. The results
Despite being created in April 2006, the Unit of Work obtained in this study are closely correlated to the
and Maintenance (UWM) of the OPRM (Office of observed building construction features and general fra-
Promotion and Real Estate Management) of Annaba city gilities of the built environment. It is important to note
was responsible by now for very few rehabilitation and that some of the usual interventions on historical build-
strengthening processes in the historical buildings of the ings are responsible for the increase in these buildings’
old city center of Annaba (about 20 operations). In fact, seismic vulnerability, namely the interruption or suppres-
Downloaded by [CERIST] at 05:19 10 December 2017

there is no reference to cost planned yet for this kind of sion of the resistant systems on the wall bases, the increase
operations. Consequently, to estimate the repair costs asso- in the number of floors and the replacement of original
ciated with the different vulnerability values used in the loss roof structures with heavier structures that are normally
evaluation (Iv,mean-2σIv; Iv,mean-σIv; Iv,mean; Iv,mean+σIv; Iv, constructed from reinforced concrete.
mean+2σIv) an average cost value per unit area of 100,000 The actual overall conservation state of the evaluated
DZD/m2 was considered in this work (equivalent to the buildings is poor, which presents fragilities that could
750 €/m2 that are assumed by Ferreira, Maio, and Vicente compromise their seismic behavior, even for low to
(2017) for Portugal). Figure 17 shows the expected global moderate intensities. Indeed, even though the city of
costs of repair estimated for the entire study area of 380 Annaba is located within a moderate seismic hazard
buildings considering different seismic scenarios. region, the moderate to high values of damage and loss
Based on observations from Figure 17, it should be obtained for different earthquake scenarios developed
stressed that, for intensities within the range of V to IX, on the basis of a building-by-building survey revealed a
the difference between the minimum and maximum considerable global seismic risk to the building stock
repair costs estimated for the vulnerability scenarios and to the historic area. The loss estimation model
under consideration is quite significant. This difference yielded results related to collapsed and unusable build-
is much smaller for higher earthquake intensities due to ing rates, such as for moderate macroseismic intensities
the high damage levels caused by severe seismic events. initially allocated to Annaba city. It is worth under-
Accounting for hazards in the Annaba city region, the lining the high number of buildings affected. With
repair cost estimations for earthquakes with intensities of respect to the human loss estimation, several factors
VII, VIII and IX are the most representative because these have contributed to the expressivity of the numbers
are the maximum historical intensities felt in the region. attained, namely the high population density and the
The high repair costs for events with intensities of VII to high seismic vulnerability of Annaba’s building stock.

Figure 17. Estimation of repair costs.

 INTERNATIONAL JOURNAL OF ARCHITECTURAL HERITAGE 61

Finally, it is worth highlighting that very important ORCID

values, in terms of economic losses, were found, but
Tiago Miguel Ferreira http://orcid.org/0000-0001-6454-
particularly for macroseismic intensities larger than VI. 7927
Based upon these scenarios, the results obtained and
the conclusions gathered in this study offer a great
opportunity to guide the action and decision-making References
in the field of seismic risk in Annaba city’s urban areas
in two regards. First, this research could enable the Askan, A., M. A. Erberik, and M. Bilal. 2014. Estimation of
development of a framework for a comprehensive data- potential seismic damage in urban areas. In Encyclopedia of
Earthquake Engineering, edited by M. Beer, I. A.
base and guidance tool for local authorities responsible Kougioumtzoglou, E. Patelli, and I. S.-K. Au, 1–9. Berlin,
for rehabilitation and renewal of the historical buildings Germany: Springer.
in Annaba city. This avenue should consider the seis- Athmani, A., A. Gouasmia, F. M. Tiago, R. Vicente, and A.
mic strengthening measures in the intervention pro- Khemis. 2015. Seismic vulnerability assessment of historical
cess. Secondly, this work can be an important element masonry buildings located in Annaba city (Algeria) using
non ad-hoc data survey. Bulletin of Earthquake Engineering
in disaster recovery and hazard mitigation strategies
13 (8):2283–2307. doi:10.1007/s10518-014-9717-7.
and prevention and emergency responses for urban Athmani, A. E., T. Ferreira, R. Vicente, A. Gouassmia, and A.
areas in the short, medium and long term. The authors Khemis. 2014. Processing of the CTC data survey for seismic
believe that considering such suggestions can produce vulnerability assessment of the masonry buildings located in
major benefits for the Annaba council after eventual Annaba city, Algeria. Aveiro, Portugal: University of Aveiro.
Downloaded by [CERIST] at 05:19 10 December 2017

seismic events; the expected losses can be considerably Barbat, A. H., M. L. Carreño, L. G. Pujades, N. Lantada, O. D.
Cardona, and M. C. Marulanda. 2010. Seismic vulnerabil-
reduced and consequently the number of expected ity and risk evaluation methods for urban areas. A review
casualties and the economic cost could fall with application to a pilot area. Structure and
dramatically. Infrastructure Engineering 6 (1–2):17–38. doi:10.1080/
As a recommendation, it is worth noting that the 15732470802663763.
development of more reliable vulnerability assessment Benouar, D. 1994. Material for the investigation of the seis-
micity of Algeria and adjacent region during the twentieth
models that combine statistical and mechanical methods
century. Annali Di Geofisica XXXVII 37 (4):511–528.
to post-seismic data collection for the Algerian events. Bounif, A., H. Haessler, and M. Meghraoui. 1987. The
This is still an issue due to uncertainties associated with Constantine (northeast Algeria) earthquake of October
statistical approaches, and must be further studied in the 27, 1985: Surface ruptures and aftershock study. Earth
case of Annaba city. Moreover, including more input and Planetary Science Letters 85 (4):451–460. doi:10.1016/
data, such as the definition of trigger-induced phenomena 0012-821X(87)90140-3.
Bramerini, F., G. Di Pasquale, A. Orsini, A. Pugliese, R.
(i.e., accounting for site effects), it is certainly possible to Romeo, and F. Sabetta 1995. Rischio sismico del territorio
achieve more accurate results. Additionally, the vulner- italiano. Proposta per una metodologia e risultati prelimi-
ability analyses of lifelines and essential facilities, as well as nari. Roma: Rapporto tecnico del Servizio Sismico
a detailed analysis of high-stakes, structural vulnerabilities Nazionale, SSN/RT/95/01.
considered fragile in the first assessment, should help Caicedo, C., A. H. Barbat, J. Á. Canas, and R. Aguiar. 1994.
Vulneradilidad sísmica de edifícios. Barcelona: Monografía
provide comforting measures and reduce the exposure
CIMNE IS-6.
of people and property to seismic risk. Cardona, O. D. 2001. Estimación holística del riesgo sísmico
utilizando sistemas dinámicos complejos. Ph.D. thesis,
Universidade Politécnica de Cataluña, Barcelona, Spain.
Acknowledgments CGS. 2011. Etude de l’aléa sismique de la Région de Annaba –
Skikda – Guelma. Alger, Algérie: Centre national de
The authors are very grateful to Dr. Mezouar Nourredine and recherche appliqueé en génie parasismique.
Dr. Boukri Mehdi, researchers at the Algerian National Chiauzzi, L., A. Masi, M. Mucciarelli, M. Vona, F. Pacor, G.
Earthquake Engineering Research Centre (CGS), and Dr. Cultrera, F. Gallovič, and A. Emolo. 2012. Building
Assia Harbi, Research Director of the Algerian Centre for damage scenarios based on exploitation of Housner inten-
Research in Astronomy, Astrophysics and Geophysics sity derived from finite faults ground motion simulations.
(CRAAG) for their assistance in obtaining valuable data per- Bulletin of Earthquake Engineering 10 (2):517–545.
taining to the seismicity of the study area. The authors thank doi:10.1007/s10518-011-9309-8.
both the technical organism in charge of the Technical Control Coburn, A. W., and R. Spence. 2002. Earthquake protection.
of Construction in Annaba city (CTC) and the Direction of Chicheter: Wiley.
Urban planning and Construction of Habitation (DUCH) for Coburn, A. W., R. Spence, and A. Pomodis. 1992. Factors
providing the technical files on the buildings diagnoses. determining human casualty levels in earthquakes:
Finally, the authors would like to thank the two anonymous Mortality prediction in building collapse. In Proceedings
reviewers for their valuable comments and suggestions. of the First International Forum on Earthquake-Related
 62 A. ATHMANI ET AL.

Casualties, Madrid, Spain, July 1992, 5989–5994. Algeria Part II: 20th century seismicity and seismotec-
Rotterdam: A.A. Balkema. tonics analysis. Journal of Seismology 7 (2):221–234.
CTC. 2010. Technical expertise study of the old buildings in doi:10.1023/a:1023571316216.
the city of Annaba. Annaba, Algeria: Technical organism Harbi, A., M. Meghraoui, and S. Maouche. 2011. The
officially in charge of the Technical Control of Djidjelli (Algeria) earthquakes of 21 and 22 August
Construction of Annaba city. 1856 (I0 VIII, IX) and related tsunami effects
D’Ayala, D., R. Spence, C. Oliveira, and A. Pomonis. 1997. Revisited. Journal of Seismology 15 (1):105–129.
Earthquake loss estimation for Europe’s historic town centres. doi:10.1007/s10950-010-9212-9.
Earthquake Spectra 13 (4):773–793. doi:10.1193/1.1585980. Harbi, A., A. Peresan, and G. F. Panza. 2010. Seismicity of
Dolce, M., A. Kappos, A. Masi, G. Penelis, and M. Vona. 2006. Eastern Algeria: A revised and extended earthquake cata-
Vulnerability assessment and earthquake damage scenarios logue. Natural Hazards 54 (3):725–747. doi:10.1007/
of the building stock of Potenza (Southern Italy) using s11069-009-9497-6.
Italian and Greek methodologies. Engineering Structures 28 Hazus, M. H. 1999. Earthquake loss estimation methodology—
(3):357–371. doi:10.1016/j.engstruct.2005.08.009. Technical and user manuals. Washington, D.C: Federal
DUC. 2006. Land use documents of the old town of Annaba city. Emergency Management Agency.
Annaba, Algeria: Direction of urbanism and Construction. Kherroubi, A., J. Déverchère, A. Yelles, B. Mercier De
Ferreira, T. 2010. Seismic vulnerability assessment of old Lépinay, A. Domzig, A. Cattaneo, R. Bracène, V.
urban centres—Case study of the old city centre of Seixal, Gaullier, and D. Graindorge. 2009. Recent and active
Portugal. Porto, Portugal: Post-graduate course in deformation pattern off the easternmost Algerian margin,
Rehabilitation of the Built Heritage, Faculty of Western Mediterranean Sea: New evidence for contrac-
Engineering of the University of Porto, Faculty of tional tectonic reactivation. Marine Geology 261 (1–4):17–
Engineering of the University of Porto. 32. doi:10.1016/j.margeo.2008.05.016.
Downloaded by [CERIST] at 05:19 10 December 2017

Ferreira, T. M., R. Maio, and R. Vicente. 2017. Analysis of the Lang, K., and H. Bachmann. 2003. On the seismic vulner-
impact of large scale seismic retrofitting strategies through ability of existing unreinforced masonry buildings. Journal
the application of a vulnerability-based approach on tradi- of Earthquake Engineering 7 (3):407–426. doi:10.1080/
tional masonry buildings. Earthquake Engineering and 13632460309350456.
Engineering Vibration 16:329–348. doi:10.1007/s11803- Maio, R., T. M. Ferreira, R. Vicente, and J. Estêvão. 2016.
017-0385-x. Seismic vulnerability assessment of historical urban cen-
Ferreira, T. M., R. Vicente, J. A. R. Mendes Da Silva, H. Varum, tres: Case study of the old city centre of Faro, Portugal.
and A. Costa. 2013. Seismic vulnerability assessment of his- Journal of Risk Research 19 (5):551–580. doi:10.1080/
torical urban centres: Case study of the old city centre in 13669877.2014.988285.
Seixal, Portugal. Bulletin of Earthquake Engineering 11 McGuire, R. K. 2004. Seismic hazard and risk analysis EERI
(5):1753–1773. doi:10.1007/s10518-013-9447-2. monography. Oakland, CA: Earthquake Engineering
Ferreira, T. M., R. Vicentea, and R. Varum. 2014. Seismic Research Institute.
vulnerability assessment of masonry facade walls: Milutinovic, Z. V., and G. S. Trendafiloski. 2003. RISK-UE:
Development, application and validation of a new scoring An advanced approach to earthquake risk scenarios with
method. Structural Engineering and Mechanics 50:541–561. applications to different European towns. WP4:
doi:10.12989/sem.2014.50.4.541. Vulnerability of Current Buildings 2003. European Project
Giovinazzi, S. 2005. The vulnerability assessment and the damage EVK4-CT-2000-00014, 110 pp.
scenario in seismic risk analysis. Dissertation, Department of Mourabit, T., K. M. Abou Elenean, A. Ayadi, D. Benouar, A.
Civil Engineering Technical University Carolo-Wilhelmina at Ben Suleman, M. Bezzeghoud, A. Cheddadi, M. Chourak,
Braunschweig and University of Florence. M. N. ElGabry, A. Harbi, M. Hfaiedh, H. M. Hussein, J.
Giovinazzi, S., and S. Lagomarsino 2004. A macroseismic Kacem, A. Ksentini, N. Jabour, A. Magrin, S. Maouche, M.
method for the vulnerability assessment of buildings. Meghraoui, F. Ousadou, G. F. Panza, A. Peresan, N.
Paper read at the 13th World Conference on Earthquake Romdhane, F. Vaccari, and E. Zuccolo. 2014. Neo-deter-
Engineering, No. 896. Vancouver, B.C., Canada. ministic seismic hazard assessment in North Africa.
GNDT. 1993. Rischio sismico di edifici pubblici - Parte I: Journal of Seismology 18 (2):301–318. doi:10.1007/s10950-
Aspetti metodologici. Bologna: Centro Servizi Quasco. 013-9375-2.
Grünthal, G. 1998. L’Echelle Macrosismique Européenne Tiedemann, H. 1989. Casualties as a function of building
(SEM-98). Conseil De l’Europe, Cahiers Du Centre quality and earthquake intensity. Proceedings of the
Européen De Géodynamique Et De Séismologie 15:101. International Workshop on Earthquake Injury
Gruppo Nazionale per la Difesa dai Terremoti (GNDT-SSN). Epidemiology for Mitigation and Response. Baltimore,
1994. Scheda di esposizione e vulnerabilità e di rilevamento MD: Johns Hopkins University.
danni di primo e secondo livello (murata e cemento ar- mato). Vicente, R., S. Parodi, S. Lagomarsino, H. Varum, and J. A. R.
[Exposure sheet and vulnerability assessment of damage of the Mendes Silva. 2011. Seismic vulnerability and risk assess-
first and second level for masonry and reinforced concrete]. ment: Case study of the historic city centre of Coimbra,
Rome: Gruppo Nazionale per la Difesa dai Terremoti. Portugal. Bulletin of Earthquake Engineering 9 (4):1067–
Harbi, A., and S. Maouche. 2009. Les principaux séismes du 1096. doi:10.1007/s10518-010-9233-3.
Nord-Est de l’Algérie. Mémoires du Service Géologique Yelles-Chaouche, A., A. Boudiaf, H. Djellit, and R. Bracene.
National. N° 16, pp. 81. 2006. La tectonique active de la région nord-algérienne.
Harbi, A., S. Maouche, and H. Benhallou. 2003. Re-appraisal Comptes Rendus Geoscience 338 (1–2):126–139.
of seismicity and seismotectonics in the north-eastern doi:10.1016/j.crte.2005.11.002.