International Journal of Strategic Property Management

ISSN: 1648-715X (Print) 1648-9179 (Online) Journal homepage: http://www.tandfonline.com/loi/tspm20

Earthquake risk mitigation: the impact of seismic

retrofitting strategies on urban resilience

Tiago Miguel Ferreira, Rui Maio, Romeu Vicente & Anibal Costa

To cite this article: Tiago Miguel Ferreira, Rui Maio, Romeu Vicente & Anibal Costa
(2016) Earthquake risk mitigation: the impact of seismic retrofitting strategies on urban
resilience, International Journal of Strategic Property Management, 20:3, 291-304, DOI:
10.3846/1648715X.2016.1187682

To link to this article: http://dx.doi.org/10.3846/1648715X.2016.1187682

Published online: 19 Jul 2016.

Submit your article to this journal

Article views: 5

View related articles

View Crossmark data

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

http://www.tandfonline.com/action/journalInformation?journalCode=tspm20

Download by: [University of California, San Diego] Date: 25 July 2016, At: 15:10
 International Journal of Strategic Property Management
ISSN 1648-715X / eISSN 1648-9179

2016 Volume 20(3): 291–304

doi:10.3846/1648715X.2016.1187682

earthquake risk mitigation: the impact of seismic

retrofitting strategies on urban resilience
Tiago Miguel FErreira a,*, Rui maio a, Romeu Vicente a, Aníbal Costa a

a RISCO, Department of Civil Engineering, University of Aveiro, Campus Universitário de Santiago,

3810-193, Aveiro, Portugal

Received 27 March 2015; accepted 29 September 2015

Abstract. It is recognized that both community and urban resilience depends on the capacity
Downloaded by [University of California, San Diego] at 15:10 25 July 2016

of built environment to maintain acceptable structural safety levels during and after unforeseeable
events, such as earthquakes, as well as to recover their original functionality. Investing in disaster risk
mitigation policies is a great step towards promoting urban resilience and community preparedness,
implicitly limiting damage to the built environment and subsequently reducing human, economic and
environmental losses. Portugal is a moderate to high seismic prone area as the latest estimates indicate
that within the next 50 years the country could be severely hit by a strong quake similar to the histori-
cal 1755 event, which left a trail of destruction and death, mainly in densely populated areas, such as
Lisbon. This manuscript aims to mitigate the impact of earthquake damage scenarios on social and
economical terms, as well as evaluating the outcome of implementing traditional retrofitting strategies
to existing masonry building stock located in urban areas of high seismic risk, using the old city centre
of Horta, Faial Island (Azores, Portugal), as a case study.

Keywords: Built environment; Resilience; Risk mitigation; Seismic vulnerability; Retrofitting solu-
tions; Loss estimation

1. INTRODUCTION: URBAN RESILIENCE

AND EARTHQUAKE RISK MITIGATION effects. Several organisations such as GFDRR
(Global Facility for Disaster Reduction and Recov-
The concept of urban resilience is often defined ery) are committed on helping governments and
as the ability of exposed communities to sustain- communities in vulnerable disaster-prone areas by
ably resist, support and recover from the effects increasing the perception, awareness and access
of hazards and is directly connected to mitigation, to comprehensive information about physical and
preparedness, disaster, response, recovery and societal exposure to disaster risk (Arshad, Athar
reconstruction disaster risk management phases 2013). This way, governments, communities, and
(Coaffee 2008). With this work, the authors aim private stakeholders are able to better quantify
at contributing for infrastructural resilience and and predict potential impacts of natural hazards
seismic risk mitigation of historical city centres’ on both society and economy, and also to carry out
built environment, since they are frequently vul- risk-sensitive decision-making. Moreover, these
nerable areas, with safeguarding importance. Be- global knowledge-sharing partnerships usually
fore embracing our particular focus, one should work together along with governments, civil soci-
acknowledge the consequences and impact of natu- ety and the private sector to create and improve
ral disasters at a global scale from previous learn- the policies and legislation needed for better land
ing experiences, as well as be aware of successful use planning, to drive investment aimed at risk
strategies and practices carried out recently all mitigation and acting as a moderator over the
over the world and at a large assessment scale, often-difficult dialogue between stakeholders (Ar-
in terms of communities’ resilience. In this sense, shad, Athar 2013).
risk identification from natural hazards is consid- As seismic hazard is still ruled by its unpredict-
ered the first step towards reducing their adverse ability and insusceptibility to be completely elimi-
* Corresponding author. E-mail: tmferreira@ua.pt nated, preparedness and awareness by means of

Copyright © 2016 Vilnius Gediminas Technical University (VGTU) Press

http://www.tandfonline.com/tspm
 292 T. M. Ferreira et al.

developing and optimising contingency and emer- In many countries, including Portugal, civil pro-
gency response plans, are therefore mandatory tection bodies are the agencies responsible for the
strategies to save lives and protect communities. general emergency response plan, which covers all
Even so, as communities can still be highly ex- types of hazards (natural and man-made), both be-
posed to financial shocks, it is necessary the de- fore and after an event. So that they can fulfil this
velopment and implementation of adequate dis- task, the agents involved in planning earthquake
aster risk financial protection strategies, allowing risk mitigation strategies should be able to define
a quicker and balanced response, improving the which zones are physically more vulnerable and to
resilience of all stakeholders. Education plays a prepare logistic and field exercises to simulate sit-
crucial role on this process, as several educational uations that may arise in a real earthquake situa-
programs have been conducted to promote coopera- tion (Goula et al. 2006). However, as discussed by
tion and innovation among Higher Education In- Ferreira et al. (2013), risk management of urban
stitutions to increase society’s resilience to disaster areas is frequently undertaken without the use of
of both human and natural origin, as the case of a general planning tool. A primary consequence
the ANDROID academic network. Created in the of this situation is that technicians and decision
Downloaded by [University of California, San Diego] at 15:10 25 July 2016

framework of the ANDROID Working Package 7, makers (city councils or regional authorities) do
this working group has focused their research on not have a global view of the area under analysis,
the vulnerability and resilience in multi-hazard which can seriously commit the effectiveness of fu-
scenarios for the Venice lagoon case study (Kaluar- ture rehabilitation strategies and risk mitigation
achchi et al. 2014; Indirli et al. 2014; Knezic et al. measures. Several authors have been taking the
2014; Borg et al. 2014). The complexity of this advantage of these multi-purpose tools on hazard
case study concerns the probability of occurrence and vulnerability related projects, as those car-
of cross-border multi-hazard disastrous events ried out by Indirli (2009) for the historical centre
that possibility may involve other surrounding of San Giuliano di Puglia (in Italy) and the city
countries such as Slovenia and Croatia. Moreover, of Valparaiso (in Chile), or for the Vesuvian re-
there are multiple variables involved: population; gion in Italy (Mazzolani et al. 2009). Hence, with
heritage; environment; Industrial facilities; tour- this paper, the authors intend to bring attention
ism; the lagoon itself, the surrounding islands and to this matter, through better understanding the
the mainland territory. cost-to-benefit balance of seismic retrofitting of old
There are several successful examples of the masonry buildings located in historical centres.
positive influence of disaster risk management and Furthermore, it is expected that the outputs re-
planning on communities’ resilient capacity, as the sulting from this work can contribute to clarify the
case of the RHRP (Rural Housing Reconstruction common ideas that there is no need for seismic ret-
Program) in the aftermath of the October 2005 rofitting in Portugal and that these interventions
earthquake in northern Pakistan, which roughly are too expensive
caused 73,000 deaths and more than 2.8 million
homeless people (Arshad, Athar 2013), the Febru-
2. SEISMIC VULNERABILITY ASSESSMENT
ary 2010 Chile (Astroza et al. 2012) or the Febru- AND RETROFITTING SOLUTIONS
ary 2011 Christchurch earthquakes (Mitchelson
2011). Moreover, the March 2011 GEJE (Great
2.1. The case study of the old city centre of
East Japan Earthquake), the first ever recorded Horta, Portugal
mega-disaster comprehending earthquake, tsu-
nami, nuclear power plant accident, power supply The present case study concerns the seismic vul-
failure and large-scale disruption of supply chains, nerability assessment of the city centre of Horta,
caused 20,000 casualties, over than 130,000 col- in Faial Island, Azores. This island was severely
lapsed buildings and 270,000 severely damaged, hit by the July 9, 1998 Azores earthquake, leav-
with a direct economic cost estimated in $210 bil- ing a trail of destruction (roughly 70% of the built
lion (Ranghieri, Ishiwatari 2014). Although Ja- environment), affecting directly more than 5000
pan’s community preparedness is internationally people and causing 8 deaths, 150 injured and
acclaimed and disaster risk management strate- 1500 homeless. All data and information collected
gies had been developed and implemented for dec- during the 10-year reconstruction process of Faial
ades, no one could ever have foreseen this com- Island (hereinafter designated by Faial database),
plexity derived from the 9.0 magnitude earthquake conducted by the Society of Promotion for Housing
cascading effects. and Infrastructures Rehabilitation (SPRHI), was
 Earthquake risk mitigation: the impact of seismic retrofitting strategies on urban resilience 293

gathered over an 8-month-length period in 2007 by city centre of Horta was evaluated in this work us-
the Regional Secretariat for Housing and Equip- ing the vulnerability index methodology developed
ment (SRHE) of the Faial Island, which funded by Vicente (2008) on the basis of the Italian GNDT
this initiative (Neves et al. 2012a). The quality II level approach (GNDT 1994). It is worth not-
and uniqueness of this database in both national ing that this methodology has been used in recent
and international context have encouraged the de- years for the seismic vulnerability assessment of
velopment of several advanced studies throughout several historical urban centres in Portugal (see
the years. Even though the fully open access pro- Vicente et al. 2011; Ferreira et al. 2013; Maio et al.
vided to the mentioned database collected in the 2015).
aftermath of the 1998 earthquake, the authors car- Similarly to above mentioned past surveying
ried out in field surveying work, prospected in the and assessment case studies, the difficulties en-
scope of the FCT URBSIS project (Assessing Vul- countered in accessing the interior of all the build-
nerability and Managing Earthquake Risk at the ings and time constraint related issues, led the au-
Urban Scale), collecting appraisal data in order to thors to distinguishing two different assessment
understand the evolution and diachronic process levels. Thus, a total of 313 buildings were divided
Downloaded by [University of California, San Diego] at 15:10 25 July 2016

resulting from rehabilitation interventions imple- into four different groups based on the detail level
mented since 1998. Although the damage levels of the available information. The first group (De-
observed in the built environment of Horta were tailed assessment), composed of the 50 buildings
not as severe as in the remaining parishes of the for which a detailed inspection was undertaken,
Faial Island, classified with a macroseismic inten- was evaluated resorting to detailed specific infor-
sity IEMS-98 = V/VI (Zonno et al. 2010), this urban mation such as architectural and structural plans,
area was still selected as case study due to its im- photographic and post-earthquake damage reports,
portance in both cultural and architectural herit- gathered in the Faial database, allowing for the
age context. Moreover, taking into consideration full evaluation of the required input parameters
the goal of the present work, the authors based of the vulnerability index methodology used. The
their assessment on the existing building condition second group was composed of 142 buildings for
at the time of the earthquake that hit Faial Island which only a non-detailed exterior inspection was
in 1998, before likely late comprehensive retrofit- available (Non-detailed assessment). A third group
ting actions ever occurred, hereinafter designated composed of 93 reinforced concrete buildings was
as BR (before retrofitting). not included in this study as the used vulnerabil-
When performing vulnerability assessment ity index methodology is only suitable for masonry
of a large number of buildings and over an ur- building typologies. Finally, a fourth group was
ban centre or region, the resources and quantity created to include other 28 non-assessable build-
of information to collect and deal with can be ings related to religious or governmental use, and
enormous and thus the use of more expedite ap- also buildings in pre-ruin or buildings undergoing
proaches results more adequate and reasonable. a retrofitting process. Thus, the outputs of mean
Methodologies for vulnerability assessment either vulnerability index values presented further on,
at the national and urban scale should be based were obtained through assembling both detailed
on few parameters, defined through the knowledge and non-detailed assessment groups, hereinafter
of the effects of past earthquakes, which can then designated as overall assessment.
be treated statistically (Neves et al. 2012b). The The seismic vulnerability index methodology
definition and nature of such approach (qualitative herein applied, classified by Calvi et al. (2006) as
and quantitative) naturally limits the formulation an hybrid technique suitable for large-scale as-
of the methodologies and the level at which the sessment of masonry buildings, comprehends the
evaluation is conducted, from the expedite evalu- calculation of a vulnerability index score, I v* , for
ation of buildings based on visual observation to each building as the weighted sum of 14 param-
the most complex numerical modeling of single eters (in Eq. 1), each one of them evaluating one
structures. aspect related to the building’s seismic response,
Despite several different methodologies have distributed into four vulnerability classes (Cvi) of
been developed and validated during the last dec- growing vulnerability, from A to D (Vicente et al.
ades, such as the FAMIVE method (D’Ayala, Sper- 2011).
anza 2002) or the MEDEA procedure (MEDEA 14
2013) – which the results were compared by Indirli I v* = ∑Cvi × pi . (1)
et al. (2013) – the seismic vulnerability of the old i =1
 294 T. M. Ferreira et al.

Table 1. Vulnerability index methodology (Vicente 2008)

Parameters by group Class Cvi Weight Relative weight
pi over I v*
A B C D
1. Structural building system
P1 Type of resisting system 0 5 20 50 0.75 46/100
P2 Quality of resisting system 0 5 20 50 1.00
P3 Conventional strength 0 5 20 50 1.50
P4 Maximum distance between walls 0 5 20 50 0.50
P5 Number of floors 0 5 20 50 1.50
P6 Location and soil conditions 0 5 20 50 0.75
2. Irregularities and interactions
P7 Aggregate position and interaction 0 5 20 50 1.50 27/100
P8 Plan configuration 0 5 20 50 0.75
P9 Height regularity 0 5 20 50 0.75
Downloaded by [University of California, San Diego] at 15:10 25 July 2016

3. Floor slabs and roofs

P10 Wall façade openings and alignments 0 5 20 50 0.50 15/100
P11 Horizontal diaphragms 0 5 20 50 1.00
P12 Roofing system 0 5 20 50 1.00
4. Conservation status and other elements
P13 Fragilities and conservation status 0 5 20 50 1.00 12/100
P14 Non-structural elements 0 5 20 50 0.50

Subsequently, a weight pi, is assigned to each of these modifiers influenced the final vulnerabil-
parameter, ranging from 0.50 for the less impor- ity index as a sum of the scores for all modifiers.
tant parameters (in terms of structural vulnerabil- Thus, the vulnerability index of each Non-detailed
ity) up to 1.50 for the most important ones. The in- building, I v , is defined in the following Eq. 2:
itial value of I v* , ranging in between 0 and 650, is
then normalised to vary in between 0 and 100, des- Iv = Iv + ∑ ∆ Iv , (2)
ignated as I v to further ease of use, namely when where: I v is the mean vulnerability index result-
estimating the building’s damage condition based ing from the Detailed assessment, and ∑ ∆ I v is
on different macroseismic intensities (Grünthal the sum of the modifier scores for the attributed
1998) and also in both human and economic loss class. It is important to note that this strategy
estimation. According to the Table 1, these 14 pa- is valid only if a reliable Detailed assessment of
rameters are arranged into four groups to empha- a large number of buildings in the study area is
size their differences and relative importance on initially obtained and the strategy is applied to a
the global seismic response of the building. single building typology (Santos et al. 2013).
According to the previous definition of the De- The following Eq. (3) shows how the scores of
tailed assessment group (first inspection phase), each modifier parameter class was estimated:
the evaluation of the vulnerability index was made pi
× (Cvi - Cvi ) , (3)
for those buildings to which detailed information 6
was available (50 buildings out of 313). Accord- ∑ i =1pi
ingly, a more expeditious approach for the Non- where: pi is the weight assigned to parameter i;
6
detailed assessment group of the remaining 142
buildings was conducted (second inspection phase), ∑ pi is the sum of parameter’s weights; Cvi is the
i =1
using the mean values obtained from the Detailed modifier factor for a determined vulnerability class
analysis of the first group of buildings, assuming and Cvi is the mean vulnerability class of param-
the masonry building characteristics homogeneous eter i, defined by the Detailed assessment.
in this area. Starting from this principle, the mean From the application of the vulnerability index
vulnerability index value obtained in the first De- methodology described above to the 50 buildings
tailed evaluation was used as a typological vulner- assessed in a Detailed manner, corresponding to
ability index (mean value) that could be affected by the first inspection phase, a mean value of the seis-
modifiers of the mean vulnerability index for each mic vulnerability index, I v , of 26.32 was obtained,
building (Ferreira et al. 2013). The classification to which was associated a standard deviation
 Earthquake risk mitigation: the impact of seismic retrofitting strategies on urban resilience 295

terials; quality of construction; building and site

value, σ I of 9.73. Assembling the complementary
v conditions; intensity of damage sustained by the
approach, used in the Non-detailed assessment of
building in past earthquakes; and the expected
the remaining 142 buildings to which the avail-
ground shaking in the area. The following strat-
able information was somehow incomplete (sec-
egies have been point out by Bothara and Brzev
ond phase of assessment), an overall assessment
(2011) as presenting the highest cost-to-benefit
seismic vulnerability index mean value, I v OA , of
ratio in terms of improving the seismic safety of
26.55 with a corresponding standard deviation val-
stone masonry buildings: i) enhancing integrity of
ue, σ I , of 5.45, was estimated. The maximum
v OA the entire building by ensuring the box-like seis-
and minimum I v index values obtained from the mic behaviour; ii) enhancing the wall strength for
detailed assessment were evaluated in 55.00 and in-plane and out-of-plane effects of seismic loads;
10.96, respectively. It is important to note that the improving floor and roof diaphragm action; and fi-
results obtained for the BR building condition are nally, iii) the strengthening of the existing founda-
well adjusted to the building characteristics and tion, which strategy is not considered practical and
fragilities the assessed buildings, an evidence of economically feasible in most cases.
the method’s robust nature.
Downloaded by [University of California, San Diego] at 15:10 25 July 2016

As massive demolition and replacement of this

vulnerable building typology seems neither afford-
2.2. Application of traditional retrofitting able nor feasible due to historical, cultural, social
solutions and economic constraints, this section presents
The protection of the lives of building occupants the retrofitting strategies adopted in this study,
in the event of an earthquake is the main goal which are based on the reconstruction method-
of the retrofit, referred as “life safety” perfor- ology defined shortly after the 1998 earthquake
mance in building codes. In many cases the life by the Regional government of Azores, aiming to
safety level prescribed by building codes cannot enhance the seismic performance, and thus the
be achieved without major structural intervention vulnerability of existing stone masonry buildings,
and investment. Thus, retrofitting is considered through retrofitting solutions that comply with
to be unfeasible when the required capital invest- the mentioned constraints. Bearing in mind the
ment exceeds the initial building cost, or when a exposed, six retrofitting solutions of increasing in-
building is in an advanced degradation state or in trusiveness and cost (from S1 to S6), grouped into
pre-ruin. Costs associated with demolition, debris three cumulative retrofitting packages (from RP1
disposal and reconstruction determine the feasibil- to RP3), were herein addressed (see Table 2).
ity of each retrofit project. Moreover, legal issues These solutions were adopted in the aftermath
arise when the safety of a building is dependent of the 1980 and 1998 Azores earthquakes by dif-
on adjacent housing units, like in the case of row ferent design offices based on the design recom-
buildings where several owners share a building, mendations specially prepared for the Faial reha-
and housing units with different owners have a bilitation process (Cansado et al. 1998), developed
common wall. Retrofitting a single house in row by the Civil Engineering Regional Laboratory of
housing has low benefit when adjacent units are Azores (LREC) in partnership with several ex-
seismically deficient (Bothara, Brzev 2011). perienced technicians in this field (Oliveira et al.
Although the establishment of retrofit strate- 1990; Costa et al. 2008). The costs associated to
gies for a specific building depends on socio-eco- each package (presented in Table 2) were defined
nomic constraints, a number of technical issues by Costa et al. (2008) from a sample of 40 struc-
arise, such as: structural system; building ma- tural design projects that undergone a thorough

Table 2. Seismic retrofitting solutions adopted

Retrofitting Retrofitting Description Estimated cost
packages solution in €/m2
RP1 = S1 + S2 + S1 Wall-to-wall connection improvement through tie-rods 35
S3 + S4 S2 Floors stiffening with diagonal bracing and new timber planks
S3 Wall-to-floor connection improvement
S4 Wall-to-roof connection improvement through tie-rods
RP2 = RP1 + S5 S5 Wall-to-roof connection improvement through concrete strapping 100
beam
RP3 = RP2 + S6 S6 Stone masonry consolidation through reinforced plasters 230
 296 T. M. Ferreira et al.

analysis process. As the adopted strategy pursues visions: the installation of 75 mm thick diagonal
the cumulative implementation of retrofitting so- wood braces at the floor level between principal
lutions and the authors have considered this set timber beams, anchored with φ10 galvanized steel
of retrofitting solutions effective on enhancing the threaded rods of and 3 mm thick galvanized steel
box-like behaviour of stone masonry buildings, re- angle brackets, and a new layer of timber planks,
sorting to low-to-moderate intrusiveness and esti- laid perpendicular to the existing planks and ade-
mated costs, the following solutions S1 to S4 were quately nailed to the floor as shown in Figure 1 (b).
grouped in RP1 package (shown in Figs 1 and 2). The retrofitting of wall-to-floor connections so-
The retrofitting of wall-to-wall connections by lution (S3) was enhanced by introducing 3 mm
means of effectively tying walls together with steel thick full-length steel angle brackets adequately
tie-rods, addressed in the retrofitting solution S1, anchored to walls through steel connectors and
is an ancient provision to enhance the building anchor plates, as depicted in Figure 2 (a), comple-
integrity, seen as a crucial requirement for sur- menting the previous solution S2. Finally, Figure
vival during an earthquake, which has been used 2 (b) illustrates the retrofitting of wall-to-roof con-
for many centuries in Mediterranean European nections solution (S4), ensured by applying the
Downloaded by [University of California, San Diego] at 15:10 25 July 2016

countries, such as Italy. With diameters ranging same technique as in solution S1 but at the roof
normally from 16 to 20 mm (Bothara, Brzev 2011), level, by introducing steel tie-rods underneath the
these threaded steel tie-rods are installed horizon- ceiling joists, to sustain horizontal thrusts in the
tally beneath floors (S1) or roofs (S4) on both sides event of an earthquake.
of the wall, and are restrained at the ends by steel The retrofitting package RP2 adds the retro-
anchor plates, as depicted in Figure 1 (a), adapted fitting solution S5 to the previous package RP1,
from Cansado et al. (1998), D’Ayala and Speranza comprehending the introduction of a reinforced
(2002). This solution is not only effective in in- concrete strapping beam (with 4 φ10 longitudinal
creasing the stiffness of flexible floor diaphragms steel bars and φ6//.20 stirrups) at the top of stone
but also in enhancing the connections with exte- masonry walls, executed along the whole perimeter
rior load-bearing walls and frontal walls. of the building, enhancing the connection between
Moreover, through the assessment of the stone roof and load-bearing stone masonry, see Figure
masonry building stock under study, the authors 3 (a). In the same figure it is also illustrated a
have confirmed the predominance of flexible tim- strengthening detail of the connection between the
ber floors, many of which besides aging have been roof and gable masonry walls.
deteriorated over time. Furthermore replacing all Retrofitting package RP3 comprises retrofit-
deteriorated structural timber elements of floors ting solution S6, presented in Figure 3 (b), which
diaphragms by new parts adequately connected, involves the shear strengthening and confinement
restoring their original resistant capacity, the solu- of masonry structural walls by the implementa-
tion adopted in this study for retrofitting of floor tion of reinforced render, as specified in Costa
connections (S2), joins two different stiffening pro- (2002). Thus, after application of a first layer of

(a) Retrofitting solution S1 (b) Retrofitting solution S2

Fig. 1. Details of retrofitting solutions S1 (a) and S2 (b) of package RP1, adapted from Costa (2006)
 Earthquake risk mitigation: the impact of seismic retrofitting strategies on urban resilience 297

(a) Retrofitting solution S3 (b) Retrofitting solution S4

Downloaded by [University of California, San Diego] at 15:10 25 July 2016

Fig. 2. Details of retrofitting solutions S3 (a) and S4 (b) of package RP1, adapted from Costa (2006)

(a) Retrofitting solution S5 (b) Retrofitting solution S6

Fig. 3. Details of retrofitting solutions S5 of package RP2 (a) and S6 of RP3 (b), adapted from Costa (2006)

filling mortar in the proportion of 1:3 (local sand used. According to the following Table 3, each ret-
extracted from Fajã Beach: Portland cement: wa- rofitting solution has directly contributed to the
ter) for voids and surface regularisation, a 0.5 mm gradual enhancement of the vulnerability index
thick welded steel mesh made of Fe430 steel and value I v , by upgrading the vulnerability classes
10 cm spaced ribs, is then fixed on both sides of the Cvi, of parameters P1, P2, P3, P11, P12 and P13.
masonry wall through a system composed of M20 It is important to refer that these solutions were
galvanized screws, φ20 galvanized steel threaded cumulatively implemented, from S1 to S6. With
rods and 4 mm thick anchor plates (20×20 mm), the exception of retrofitting solutions S4 and S6
spaced of 150 cm. Finally, a 3 cm thick second lay- that do not define directly the vulnerability class
er of fine sand-blasted finishing mortar is applied upgrade, the remaining vulnerability classes pre-
for finishing (Costa 2002). sented in Table 3 were directly attributed to all the
evaluated buildings.
3. comparative analysis While retrofitting solution S4 the enhancement
was simply guaranteed by improving in one class
3.1. Seismic vulnerability assessment the original vulnerability class Cvi of parameter
Following the previous section wherein the retro- P12, the explanation concerning retrofitting solu-
fitting solutions and packages were presented, the tion S6, influencing over parameter P3, requires
current section begins by explaining how exactly deeper attention. Acknowledging the fact that in
these retrofitting solutions were accounted on the an urban context, as in the present case study,
seismic vulnerability index methodology herein the observed masonry typology and fabric is quite
 298 T. M. Ferreira et al.

Table 3. Influence of each retrofitting solution over the vulnerability index value, I v
Retrofitting Retrofitting Description Parameter Vulnerability
packages solution class, Cvi
RP1 = S1 Wall-to-wall connection improvement through P1 B
S1 + S2 + S3 + S4 tie-rods
S2 Floors stiffening with diagonal bracing and new P11 A
timber planks
S3 Wall-to-floor connection improvement P1 A
S4 Wall-to-roof connection improvement through P12 +1
tie-rods
RP2 = P1 + S5 S5 Wall-to-roof connection improvement through P12 A
concrete strapping beam
RP3 = P2 + S6 S6 Stone masonry consolidation through reinforced P2 A
plasters P3 τ0
P13 A
Downloaded by [University of California, San Diego] at 15:10 25 July 2016

distinct than in rural areas and in order to esti- within the current urban environment, namely
mate reliable values for the stone masonry build- features regarding building’s geometry and global
ings within the city centre of Horta, the authors alignments, as in P4, P5, P7, P8, P9 and P10.
have matched the masonry typological classifica- The application of the same vulnerability index
tion argued by Costa (2002) with the Italian Seis- methodology considering the retrofitting package
mic Code (DM 2008) classification and correspond- RP1, led to a reduction of the seismic vulnerability
ing mechanical properties of those masonry typolo- index mean value, I v OA , of roughly 18.9%. Moreo-
gies. Through observing the available information ver, introducing the retrofitting package RP2, this
collected in the aftermath of the 1998 earthquake, reduction slightly increased to 23.1%, again when
it was possible to clearly differentiate two mason- compared to the BR vulnerability index mean val-
ry typologies: i) Type M1, masonry walls of good ue. Finally, by considering the retrofitting package
quality, built with regular size stones, which is RP3 (which includes the previous packages RP1
described in NTC 2008 (DM 2008) as “Masonry in and RP2), the seismic vulnerability index mean
squared stony blocks” and is commonly observed in value was reduced in 51.7%. The results in terms
the noblest and most magnificent Azorean build- of seismic vulnerability index values are sum-
ings; and ii) Type M3, masonry walls of irregular marised in the following Table 4, as well as the
stones interconnected using smaller fragments of attained reduction of I v OA for each retrofitting
stone or clay to fill in the small voids and to en- package (RPi) with respect to the original building
sure adequate strength, described in NTC 2008 condition (BR).
(DM 2008) as “Masonry in disorganized (irregular)
Table 4. Vulnerability index values and reduction values
stones typology”.
Accordingly, for masonry Type M1, the values Build- Detailed assess- Overall assess- Reduc-
ing con- ment ment tion
adopted for the mechanical properties of the ma- dition
terial are based on the corresponding lower limit (%)
σ σ
Iv Iv I v OA Iv OA
of NTC 2008 guidelines (DM 2008), which are in
line with those adopted in past research works car- BR 26.32 9.73 26.55 5.45 –
ried out within the same urban area of the city of RP1 21.30 7.64 21.52 4.52 18.9
Horta (Neves et al. 2012b; Cunha 2013). For ma- RP2 20.19 7.18 20.41 4.32 23.1
sonry Type M3, the value considered was obtained RP3 12.61 4.72 12.84 3.34 51.7
from the experimental work developed by Costa
et al. (2012). Thus, by introducing retrofitting solu-
3.2. Damage scenarios and loss estimation
tion S6 (of RP3), ultimate shear strength value τ 0
, required to estimate the conventional strength in This section presents loss estimation obtained for
parameter P3, were enhanced in about 60% and different damage scenarios computed for several
110% for the masonry typology M1 and M3, re- macroseismic intensities, IEMS-98. According to
spectively. As mentioned in Section 2.1, the seismic Nunes (2008), ever since the second half of the 20th
vulnerability index mean value, I v OA , of 26.55 was century, maximum intensities of IEMS-98 = VII and
obtained for the BR building stock condition, re- VIII were observed in the Azorean archipelago,
flecting the good general quality of these buildings during the earthquakes of 1952 (São Miguel), 1964
 Earthquake risk mitigation: the impact of seismic retrofitting strategies on urban resilience 299

(São Jorge), 1973 (Pico) and 1988 (São Miguel). substantially, ranging between 2.30 and 2.88 and
Moreover, maximum intensities of IEMS-98 = IX from 3.12 to 3.63 (minimum and maximum) for
were recorded in 1958, 1980 and 1998. macroseismic intensities of IEMS-98 = IX and IEMS-
Among the several methods described in the 98 = X, respectively. In this sense it is pertinent
literature for estimating losses in function of the to note that according to some authors (Pagnini
probability of occurrence of a certain damage et al. 2011; Vicente et al. 2011; Ferreira et al.
grade, this task was herein carried out through 2013), buildings with a vulnerability index higher
the construction of damage scenarios based on than 45, i.e. building for which severe damages
global probabilistic distributions, using the seismic ( 3 ≤ µ D ≤ 4 ) and potential local collapse ( µ D > 4 )
vulnerability index values, I v , obtained to the dif- are expected, should be subjected to a further as-
ferent above-mentioned building conditions (BR, sessment resorting to a more detailed approach.
RP1, RP2 and RP3). The damage estimation The loss estimation model adopted in this re-
models are inevitably dependent on the physical search is based on damage grades that relate the
damage grades, including the definition of cor- probability of exceeding a certain damage level
relations between the probability of exceeding a with the probability of collapse and functional loss.
Downloaded by [University of California, San Diego] at 15:10 25 July 2016

certain level of damage and the probability of dif- Supported on observed damage data, the approach
ferent loss phenomena (such as the probability of herein applied has been proposed by Servizio
collapsing or becoming unusable due to the lack Sísmico Nazionale (SSN) based on the work of
of structural safety conditions, the probability of Bramerini et al. (1995). This approach involves
death and severely injured, etc.). Thus, and based the analysis of data associated with the probabil-
on the previously discussed seismic vulnerability ity of buildings to be deemed unusable after minor
assessment outputs, mean damage grades, µ D , and moderate seismic action. As referred in Maio
were estimated and discussed next for different et al. (2015), although such events produce lower
macroseismic intensities, based on each building levels of structural and non-structural damage,
vulnerability index, I v . To this end, Bernardini higher mean damage grade values are associated
et al. (2007) proposed an analytical expression that with a higher probability of building collapse. The
correlates hazard with the mean damage grade ( probabilities of exceeding a certain damage grade
0 ≤ µ D ≤ 5 ) of the damage distribution in terms of are used in the loss estimation and are affected by
vulnerability value, Eq. (4): multiplier factors, which range from 0 to 1 accord-
  I + 6.25 × V -13.1  ing to different proposals. In Italy, data processing
µ D = 2.5 × 1+ tanh    ; (4) undertaken by Bramerini et al. (1995) has enabled
  Q  the establishment of these weighted factors and
0 ≤ µ D ≤ 5, respective expressions for their use in the estima-
where: I is the macroseismic intensity in accord- tion of building losses.
ance to the European Macroseismic Scale (EMS- The following Eq. (6) and (7) were used for the
98); V is the vulnerability index used in the mac- determination of collapsed and unusable buildings:
roseismic methodology, which can be related to the Pcollapse = P( D5 ) , (6)
vulnerability index value I v , through Eq. (5); and
Q, which is a ductility factor that describes the Punusable buildings = P ( D3 ) × Wei ,3 + P ( D4 ) × Wei ,4 , (7)
ductility of the constructive typology under study
(ranging from 1 to 4). In order to provide the best where: P ( Di ) is the probability of the occurrence of
fit between the GNDT curves and the EMS-98 a certain damage grade (from D1 to D5 ) and Wei , j
functions, a ductility factor, Q, of 3.0 was adopted are multiplier factors that indicate the percentage
in this work (see Vicente et al. 2011; Ferreira et al. of buildings associated with the damage grades;
2013). Di , that suffer collapse or are considered unusa-
ble. Following the work of Maio et al. (2015), these
V = 0.592 + 0.0057 × I v (5)
multiplier factors are assumed here as Wei ,3 = 0.4
Globally, the estimated damage for the origi- and Wei ,4 = 0.6. Figure 4 presents the probability
nal building condition (BR) ranged from 2.49 to of building collapse and unusable buildings for the
3.69 and 3.30 to 4.23 for earthquake scenarios four building conditions studied in this work (BR,
corresponding to IEMS-98 = IX and IEMS-98 = X, RP1, RP2 and RP3).
respectively. When the Retrofitting Package RP3 Moreover, Table 5 summarizes the overall re-
is applied to the building stock of Horta, the val- sults in terms of collapsed and unusable build-
ues obtained for the mean damage grade decrease ings, obtained for those building conditions by
 300 T. M. Ferreira et al.

Collapsed Buildings Unusable Buildings

0.8 0.8
BR RP1 RP2 RP3 BR RP1 RP2 RP3

Damage Probability, Pk
0.6 0.6
Damage Probability, Pk

0.4 0.4

0.2 0.2

0.0 0.0
VIII IX X XI XII VIII IX X XI XII
I EMS-98 I EMS-98

(a) (b)

Fig. 4. Probability of collapsed (a) and unusable buildings (b) for the different building conditions analysed
Downloaded by [University of California, San Diego] at 15:10 25 July 2016

Table 5. Estimation of the number of collapsed and unusable buildings, for a total of 192 buildings evaluated
Intensity Collapsed buildings Unusable buildings
IEMS-98
BR RP1 RP2 RP3 BR RP1 RP2 RP3
VIII 0 0 0 0 9 (4.5%) 6 (3.1%) 5 (2.8%) 3 (1.5%)
IX 1 (0.3%) 0 0 0 43 (22.4%) 35 (18.1%) 33 (17.1%) 22 (11.5%)
X 14 (7.2%) 9 (4.7%) 8 (4.2%) 4 (2.0%) 92 (47.7%) 85 (44.1%) 83 (43.2%) 70 (36.4%)
XI 67 (34.9%) 55 (28.6%) 52 (27.3%) 36 (18.8%) 91 (47.3%) 96 (49.8%) 96 (50.2%) 100 (51.8%)
XII 129 (67.3%) 120 (62.4%) 118 (61.2%) 101 (52.7%) 52 (27.1%) 59 (30.7%) 60 (31.5%) 72 (37.3%)

considering macroseismic intensities in the range assumed to require short-term shelters. Casualties
of IEMS-98 = VIII to IEMS-98 = XII. and homeless rates were determined using Eq. (8)
As in the previous case, a proposal of the and (9), respectively.
Servizio Sismico Nazionalle (Bramerini et al. 1995) Pdeath and severely injured = 0.3 × P ( D5 ) , (8)
was used to estimate the casualties (deaths and
severely injured) and homeless rates. Regarding Phom eless = P ( D3 ) × Wei ,3 + P ( D4 ) × Wei ,4 + 0.7 × P ( D5 ) .
the deaths and severely injured rate, it is defined (9)
as being 30% of the inhabitants living in collapsed Using the same presentation scheme, Figure 5
and unusable buildings. In this case, the survivors presents the probability of casualties and homeless

Deaths and severely injured Homeless

0.8 0.8
BR RP1 RP2 RP3 BR RP1 RP2 RP3
Damage Probability, Pk

Damage Probability, Pk

0.6 0.6

0.4 0.4

0.2 0.2

0.0 0.0
VIII IX X XI XII VIII IX X XI XII
I EMS-98 I EMS-98

(a) (b)

Fig. 5. Probability of deaths and severely injured (a) and homeless (b)
for the evaluated building conditions (BR to RP3)
 Earthquake risk mitigation: the impact of seismic retrofitting strategies on urban resilience 301

Table 6. Estimation of the number of deaths and severely injured, and homeless,
for a total number of 1596 inhabitants
Inten- Deaths and severely injured Homeless
sity
IEMS-98 BR RP1 RP2 RP3 BR RP1 RP2 RP3
VIII 0 0 0 0 73 (4.5%) 49 (3.1%) 45 (2.8%) 24 (1.5%)
IX 2 (0.1%) 1 (0.1%) 1 (0.1%) 0 363 (22.7%) 290 (18.2%) 275 (17.2%) 185 (11.6%)
X 34 (2.2%) 22 (1.4%) 20 (1.3%) 9 (0.6%) 841 (52.7%) 756 (47.4%) 737 (46.2%) 603 (37.8%)
XI 167 (10.5%) 137 (8.6%) 131 (8.2%) 90 (5.6%) 1144 (71.7%) 1114 (69.8%) 1106 (69.3%) 1037 (65.0%)
XII 322 (20.2%) 299 (18.7%) 293 (18.4%) 253 (15.8%) 1184 (74.2%) 1186 (74.3%) 1187 (74.4%) 1184 (74.2%)

obtained for each of the building conditions (BR to these authors obtained statistical values based on
RP3) for the previous seismic intensity scenarios the estimated cost of typical replacement actions
(from IEMS-98 = VIII to XII). applied to more than 50,000 buildings.
In addition, the global frequencies computed Thus, and according to Vicente (2008), the re-
Downloaded by [University of California, San Diego] at 15:10 25 July 2016

from these two probabilistic outputs are given pair cost probabilities for a certain seismic event
in Table 6, from which it is worth emphasizing characterized by an intensity I, P[R|I], can be ob-
the significant decrease of the homeless attained tained from the product of the conditional prob-
through the application of the retrofitting package ability of the repair cost for each damage level,
solutions. As an example, a decrease of about 15% P[R|Dk], with the conditional probability of the
was observed in the number of homeless people damage condition for each level of building vul-
(from 841 to 603 out of 1596 inhabitants) for an nerability and seismic intensity, P[Dk|Iv,I], given
intensity IEMS-98 = X. by Eq. (10):
5 100
The earthquake that occurred in 2009 in the
Italian city of L’Aquila caused about 300 deaths
P[ R | I ] = ∑ ∑ P  R | Dk  × P  Dk | I v , I  . (10)
Dk =1I v = 0
and rendered 40,000 homeless. This example
To estimate the replacement costs associated
should provide to other countries an important les-
with the different building conditions, an aver-
son about strategies that disbelieve the community
age cost value of 700 €/m2 was considered for the
engagement in post-disaster decision-making (Liel
building stock within the historical centre of the
et al. 2013). Therefore, it is suggested that both
city of Horta, value that was in line with the one
communities and governments should put more
estimated by Dolce et al. (2006) from the recon-
emphasis on planning for post-disaster, valuing
struction process undertaken in the aftermath of
the community engagement and decision-making,
the Irpinia earthquake. Moreover, to account for
especially planning for emergency response.
built and cultural heritage issues, whereas the
implementation of traditional building techniques
3.3. Economic loss
and materials can be slightly more expensive than
In this section, the behavioral influence of the ret- current solutions, this average cost value per unit
rofitting packages is analyzed, not only over the area was considered 1000 €/m2 for the BR build-
estimation of the seismic vulnerability index as- ing condition. It is worth noting that this value
sociated to each, but also how these actions con- was already suggested in the past by Vicente et al.
tribute to mitigate the earthquake risk in general. (2011) for the old city centre of Coimbra, Portugal.
As addressed by Benedetti and Petrini (1984), the Based on these probabilistic values it is then
mean damage grade, discussed in Section 3.2, can possible to estimate the global replacement costs
be interpreted either economically or by means of for the entire study area (192 buildings) and to ob-
an economic damage index representing the ratio tain the economic balance computed for each one of
between repair and replacement costs (i.e. building the three retrofitting packages, in relation do the
value). The correlation between damage grades and BR buildings condition. This output is presented
these repair and replacement costs was proposed by in Figure 6 for macroseismic intensities ranging
Dolce et al. (2006) by processing and analysing post- between IEMS-98 = V and XII and shown Table 7
earthquake damage data collected after the Um- shows in the form of global savings associated to
bria-Marche (1997) and Pollino (1998) earthquakes such economic balance. Moreover, the mean peri-
(Dolce et al. 2006), using the GNDT-SSN procedure ods of inactivity referred by Nunes (2008) for the
(GNDT 1994). From that extensive amount of data, Archipelago of Azores (i.e. the mean amount of
 302 T. M. Ferreira et al.

of 100 €/m2, i.e. about 14% of the mean replace-

ment cost, reductions of about 0.9%, 6.5%, 3.0%
and 4.5% were obtained respectively in the ratios
of deaths and severely injured, homeless, collapsed
and unusable buildings, for a macroseismic inten-
sity of IEMS-98 = X. Repeating this exercise for the
most expensive and complete retrofitting package,
RP3, its costs represent about 33% of the mean re-
placement cost, but its implementation leads to a
reduction of 51.7% over the mean vulnerability in-
Fig. 6. Economic balance for the three retrofitting dex value, which, in terms of loss estimation for a
packages considered, in relation to the BR building macroseismic intensity of IEMS-98 = X, represents a
condition decrease of 1.6%, 14.9%, 5.0% and 11.5% in terms of
the respective ratios of deaths and severely injured,
time elapsed between two earthquakes of inten- homeless and collapsed and unusable buildings.
Downloaded by [University of California, San Diego] at 15:10 25 July 2016

sities V < IEMS-98 < VIII and IEMS-98 ≥ VIII), are

also presented in Figure 6 so that the economic 4. Conclusions
viability of the retrofitting packages analysed in
this work can be easily analysed. The broad conclusions and recommendations from
From the analysis of Figure 6, it is easy to con- the exposed work converge towards the enhancement
clude that the three retrofitting packages proved to in terms of public awareness and perception, educa-
be cost effective for macroseismic intensities equal tion, training and research concerning the seismic
or higher than IEMS-98 = IX. Notwithstanding this risk mitigation. Ideally, the strategy to mitigate seis-
fact, for intensities of IEMS-98 = VII and VII respec- mic risk should address land use zoning (reducing
tively, which are already within the mean period of exposure), planning and of adequate strengthening
inactivity of 12 years, global savings of about 1.08 campaigns and the implementation of seismic build-
and 5.85 million euros (M€) can be obtained with ing codes suitable both for new and existing struc-
the application of RP1 to the 192 buildings of the tures (reducing the seismic vulnerability of the built
old building stock of Horta (see Table 7). environment). Moreover, with an appropriate and as-
As would be expected, the global savings ob- sertive policies, financial and institutional supports
tained from the application of the herein analysed at both national and local levels it is possible to carry
retrofitting packages are more expressive for the this strategy into a workable action plan.
higher macroseismic intensities, as the reduction on As expected, in terms of damage and loss esti-
the relative level of damage suffered by the build- mation, the results achieved for the case study of
ings is more important for these intensities, and Horta’s historical centre were found, not as impres-
from a strictly economic point of view RP1 proved to sive as in other potential case studies (e.g. histori-
be the most cost effective retrofitting package with cal centre of Faro city), due to the large influence
global saving of around 20 million euros. However, of the vulnerability index over the loss estimation
it is important to note that these outputs must be formulations and to the low range of the obtained
seen and analysed along with the already discussed vulnerability index values of the evaluated build-
damage scenarios and loss estimation results, since ing stock, estimated through the application of the
although for some intensities retrofitting packages seismic vulnerability index methodology. Despite
RP2 and RP3 could lead to lower economic sav- the simplifications inherent to the methodology it-
ings but to a significant decrease in terms of hu- self, these results were well adjusted to the slight
man losses. As an example, considering the retro- damage levels observed in the city of Horta, in the
fitting package RP2, which has an estimated cost aftermath of the 1998 Azores earthquake.

Table 7. Global savings obtained for each retrofitting package RPi (in millions of €)
Retrofitting Macroseismic intensity, IEMS-98
package
V VI VII VIII IX X XI XII
RP1 - - 1.08 M€ 5.85 M€ 12.24 M€ 17.15 M€ 19.23 M€ 20.10 M€
RP2 - - - 1.20 M€ 7.89 M€ 12.53 M€ 14.39 M€ 15.13 M€
RP3 - - - - 1.80 M€ 5.86 M€ 6.12 M€ 5.91 M€
 Earthquake risk mitigation: the impact of seismic retrofitting strategies on urban resilience 303

With respect to the considered retrofitting solu- ty of masonry buildings, L’industria delle Costruzioni
tions, their implementation led in general, to a rea- (149): 66–74.
sonable improvement of the building stock global Bernardini, A.; Giovinazzi, S.; Lagomarsino, S.; Parodi,
S. 2007. Vulnerabilità e previsione di danno a scala
seismic behaviour, namely in terms of the vulner-
territoriale secondo una metodologia macrosismica
ability index values, reflecting with great accuracy coerente con la scala EMS-98, in Proceedings of
its known influence over the shear strength ca- the 12th conference of Italian National Association
pacity, ductility and on the improvement of the of Earthquake Engineering – ANIDIS, 10–14 June
so-called box-behaviour. However, from loss esti- 2007, Pisa, Italia.
mation results, the enhancement observed by im- Borg, R. P.; Indirli, M.; Romagnoli, F.; Rochas, C.; Kuzņecova,
T. 2014. The ANDROID case study; Venice and its ter-
plementing these retrofitting packages could have ritory: vulnerability and resilience in multi-hazard sce-
been more expressive if the building stock seismic narios, Procedia Economics and Finance 18: 825–836.
vulnerability was higher, therefore implicating http://dx.doi.org/10.1016/S2212-5671(14)01008-9
higher vulnerability index values. Bothara, J.; Brzev, S. 2011. A TUTORIAL: improving
From the loss estimation results related to col- the seismic performance of stone masonry buildings.
1st ed. EERI, Oakland, California, USA.
lapsed and unusable building rates obtained to
Bramerini, F.; Di Pasquale, G.; Orisini, A.; Pugliese, A.; Ro-
Downloaded by [University of California, San Diego] at 15:10 25 July 2016

the maximum observed macroseismic intensities

meo, R.; Sabetta, F. 1995. Rischio sismico del territorio
in Faial of IEMS-98 = IX and X, it is relevant to Italiano: proposta per una metodologia e risultati pre-
point out, on the one hand, an extremely low rate liminary, Technical report SSN/RT/95/01, Rome, Italy.
of collapsed buildings, and on the other hand, the Calvi, G. M.; Pinho, R.; Magenes, G.; Bommer, J.; Re-
significant rate of unusable buildings expected in strepo-Velez, L. F.; Crowley, H. 2006. Development
the event of an earthquake within this range of of seismic vulnerability assessment methodologies
over the past 30 years, ISET Journal of Earthquake
intensities. With respect to human loss estimation,
Technology 43: 75–104.
several factors have contributed to the low num-
Cansado, E.; Oliveira, C. S.; Fragoso, M.; Miranda, V. 1998.
bers attained when compared to the total num- Regras gerais de reabilitação e reconstrução de edifícios
ber of inhabitants estimated for the case study correntes afectados pela crise sísmica do Faial, Pico e
area, among them the pronounced seasonality, São Jorge, iniciada pelo sismo de 9 de Julho de 1998.
low population density and again the low range Coaffee, J. 2008. Risk, resilience, and environmentally
of vulnerability index values obtained to this par- sustainable cities, Energy Policy 36(12): 4633–4638.
http://dx.doi.org/10.1016/j.enpol.2008.09.048
ticular building stock. Nonetheless, the number of
Costa, A. 2002. Determination of mechanical properties
homeless population has resulted quite expressive,
of traditional masonry walls in dwellings of Faial Is-
naturally following to some extent the trend ob- land, Azores, Earthquake Engineering & Structural
served of the rate of unusable buildings. Finally Dynamics 31(7): 1361–1382.
regarding economic losses, it is worth highlighting Costa, A. 2006. Strengthening and repairing earthquake
here that the three analysed retrofitting packages damaged structures, in Oliveira, C.; Roca, A.; Goula,
proved to be cost-effective for the highest macro- X. (Eds.). Assessing and Managing Earthquake Risk,
Chapter 19, Springer Netherlands, 403–426.
seismic intensities, i.e., from IEMS-98 = IX to XII,
Costa, A.; Oliveira, C. S.; Neves, F. 2008. Reinforcing tech-
presenting maximum global savings of about 20.10 niques more commonly used in Faial reconstruction. Sis-
million euros. Moreover, for moderate to high mac- mo 1998 – Açores. Uma década depois. 1st ed., 531–555.
roseismic intensities (IEMS-98 = VII and VIII), very Costa, A. A.; Arêde, A.; Costa, A.; Guedes, J.; Silva, B.
interesting savings of about 5.85 and 1.20 million 2012. Experimental testing, numerical modelling
euros were obtained with the application of the and seismic strengthening of traditional stone ma-
retrofitting packages RP1 and RP2, respectively. sonry: comprehensive study of a real Azorian pier,
Bulletin of Earthquake Engineering 10(1): 135–159.
Cunha, S. 2013. Retrofitting techniques of masonry
References buildings. MSc Thesis. Department of Civil Engi-
neering. University of Aveiro, Portugal.
Arshad, S.; Athar, S. 2013. Rural housing reconstruc- D’Ayala, D.; Speranza, E. 2002. An integrated procedure
tion program post-2005 earthquake - learning from for the assessment of seismic vulnerability of historic
the Pakistan experience. A Manual for Post-Disaster buildings, in Proceedings of the 12th European Con-
Housing Program Managers. 116. ference on Earthquake Engineering, 2–6 September
Astroza, M.; Ruiz, S.; Astroza, R. 2012. Damage as- 2002, London, UK. London: Elsevier Science Ltd. Pa-
sessment and seismic intensity analysis of the 2010 per Reference 561.
(Mw 8.8) Maule earthquake, Earthquake Spectra 28: DM. 2008. Decreto Ministeriale DM 14/01/2008. Norme
S145–S164. http://dx.doi.org/10.1193/1.4000027 tecniche per le costruzioni, 428.
Benedetti, D.; Petrini, V. 1984. Sulla vulnerabilita sis- Dolce, M.; Kappos, A.; Masi, A.; Penelis, G.; Vona, M.
mica di edifici in muratura: un metodo di valutazi- 2006. Vulnerability assessment and earthquake
one. A method for evaluating the seismic vulnerabili- damage scenarios of the building stock of Potenza
 304 T. M. Ferreira et al.

(Southern Italy) using Italian and Greek methodolo- Mazzolani, F. M.; Indirli, M.; Faggiano, B.; Form-
gies, Engineering Structures 28(3): 357–371. http:// isano, A.; De Gregorio, D.; Zuccaro, G. 2009. Cata-
dx.doi.org/10.1016/j.engstruct.2005.08.009 strophic effects of a Vesuvian eruption on the built
Ferreira, T. M.; Vicente, R.; Mendes da Silva, J. A. R.; environment, in Proceedings of the 2nd international
Varum, H.; Costa, A. 2013. Seismic vulnerability as- workshop on Performance, Protection & Strengthen-
sessment of historical urban centres: case study of ing of Structures under Extreme Loading PROTECT
the old city centre in Seixal, Portugal, Bulletin of 2009, 19–21 August 2009, Hayama, Japan.
Earthquake Engineering 11(5): 1753–1773. MEDEA. 2013. Manuale di Esercitazioni sul Danno
GNDT. 1994. Scheda di esposizione e vulnerabilità e di rile- ed Agibilità per edifici ordinari [User’s handbook
vamento danni di primo livello e secondo livello (mura- to check damage and safety/usability for ordinary
tura e cemento armato). Grupo Nazionale per la Difesa buildings]. Civil Defense, Italy.
dai Terrmoti – Servizio Sismico Nazionale, Rome, Italy. Mitchelson, M. 2011. Haiti and Christchurch earth-
Goula, X.; Roca, A.; Oliveira, C. S. (Eds.). 2006. Assess- quakes viewed through a resilience lens – a compara-
ing and managing earthquake risk. Geo-scientific tive case study. MSc thesis. Auckland University of
and engineering knowledge for earthquake risk miti- Technology, Australia.
gation: developments, tools, techniques. Springer Neves, F.; Costa, A.; Vicente, R.; Oliveria, C. S.; Varum,
Netherlands. H. 2012a. Seismic vulnerability assessment and char-
Grünthal, G. (Ed.). 1998. European Macroseismic Scale acterisation of the buildings on Faial Island, Azores,
Downloaded by [University of California, San Diego] at 15:10 25 July 2016

1998 (EMS-98). European Seismological Commis- Bulletin of Earthquake Engineering 10(1): 27–44.
sion, Subcommission on Engineering Seismology, Neves, N.; Arêde, A.; Costa, A. 2012b. Seismic analysis
Working Group Macroseismic Scales, Luxembourg. of a building block, Bulletin of Earthquake Engineer-
Indirli, M. 2009. Organization of a Geographic Infor- ing 10(1): 235–267.
mation System (GIS) database on natural hazards Nunes, J. 2008. General description of the Azores region
and structural vulnerability for the historical cen- seismicity, in Oliveira, C.; Costa, A.; Nunes, J. C.
tre of San Giuliano di Puglia (Italy) and the City (Eds.). Sismo 1998 – Açores. Uma década depois. 1st
of Valparaiso (Chile), International Journal of Ar- Editio. Horta, Faial Island, Azores, Portugal, 59–72.
chitectural Heritage 3(4): 276–315. http://dx.doi. Oliveira, C. S.; Lucas, A.; Guedes, J. H. C. 1990. 10 Anos
org/10.1080/15583050902803780 após o Sismo dos Açores de 1 de Janeiro de 1980.
Indirli, M.; Kouris, L. A. S.; Formisano, A.; Borg, R. P.; Monografia, Volume I e II. Secretaria Regional da
Mazzolani, F. M. 2013. Seismic damage assessment Habitação e Obras Públicas, Delegação da Ilha Ter-
of unreinforced masonry structures after the Abruzzo ceira, Açores, Laboratório Nacional de Engenharia
2009 earthquake: the case study of the historical cen- Civil, Angra do Heroísmo, Açores, Portugal.
tres of L’Aquila and Castelvecchio Subequo, Interna- Pagnini, L. C.; Vicente, R.; Lagomarsino, S.; Varum, H.
tional Journal of Architectural Heritage 7(5): 536–578. 2011. A mechanical model for the seismic vulner-
http://dx.doi.org/10.1080/15583058.2011.654050 ability assessment of old masonry buildings, Earth-
Indirli, M.; Knezic, S.; Borg, R.; Kaluarachchi, Y.; quakes and Structures 2(1): 25–42. http://dx.doi.
Ranguelov, B.; Romagnoli, F.; Rochas, C. 2014. The org/10.12989/eas.2011.2.1.025
ANDROID case study; Venice and its territory: Ranghieri, F.; Ishiwatari, M. 2014. Learning from
a general overview, Procedia Economics and Fi- Megadisasters: lessons from the Great East Japan
nance 18: 837–848. http://dx.doi.org/10.1016/S2212- Earthquake. International Bank for Reconstruction
5671(14)01009-0 and Development/The World Bank, Washington DC.
Kaluarachchi. Y.; Indirli, M.; Ranguelov, B.; Romagno- http://dx.doi.org/10.1596/978-1-4648-0153-2
li, F. 2014. The ANDROID case study; Venice and Santos, C.; Ferreira, T. M.; Vicente, R.; Mendes da Silva,
its territory: existing mitigation options and chal- J. A. R. 2013. Building typologies identification to
lenges for the future, Procedia Economics and Fi- support risk mitigation at the urban scale – case
nance 18: 815–824. http://dx.doi.org/10.1016/S2212- study of the old city centre of Seixal, Portugal, Jour-
5671(14)01007-7 nal of Cultural Heritage 14(6): 449–463. http://dx.doi.
Knezic, S.; Scudeller, M.; Indirli, M.; Romagnoli, F.; org/10.1016/j.culher.2012.11.001
Kuzņecova, T.; Perdikou, S. 2014. The ANDROID Vicente, R. 2008. Estratégias e metodologias para inter-
case study; Venice and its territory: identification of venções de reabilitação urbana: avaliação da vulner-
hazards and impact of multi-hazard scenarios, Pro- abilidade e do risco sísmico do edificado da Baixa de
cedia Economics and Finance 18: 465–472. http:// Coimbra. Universidade de Aveiro, Aveiro, Portugal.
dx.doi.org/10.1016/S2212-5671(14)00964-2
Vicente, R.; Parodi, S.; Lagomarsino, S.; Varum, H.;
Liel, A. B.; Corotis, R. B.; Camata, G.; Sutton, J.; Holtz- Mendes da Silva, J. A. R. 2011. Seismic vulnerabil-
man, R.; Spacone, E. 2013. Perceptions of decision- ity and risk assessment: case study of the historic
making roles and priorities that affect rebuild- city centre of Coimbra, Portugal, Bulletin of Earth-
ing after disaster: the example of L’Aquila, Italy, quake Engineering 9(4): 1067–1096. http://dx.doi.
Earthquake Spectra 29(3): 843–868. http://dx.doi. org/10.1007/s10518-010-9233-3
org/10.1193/1.4000158
Zonno, G.; Oliveira, C. S.; Ferreira, M. A.; Musacchio, G.;
Maio, R.; Ferreira, T. M.; Vicente, R.; Estêvão, J. 2015. Meroni, F.; Mota-de-Sá, F.; Neves, F. 2010. Assess-
Seismic vulnerability assessment of historical cen- ing seismic damage through stochastic simulation of
tres: case study of the old city centre of Faro, Portu- ground shaking: the case of the 1998 Faial Earth-
gal, Journal of Risk Research http://dx.doi.org/10.10 quake (Azores Islands), Surveys in Geophysics 31(3):
80/13669877.2014.988285 361–381. http://dx.doi.org/10.1007/s10712-009-9091-1