Encyclopedia of Earthquake Engineering

Living Edition
| Editors: Michael Beer, Ioannis A. Kougioumtzoglou, Edoardo Patelli, Ivan Siu-Kui Au

Seismic Vulnerability Assessment: Lifelines

  • Kyriazis PitilakisEmail author
  • Sotiris Argyroudis
Living reference work entry
DOI: https://doi.org/10.1007/978-3-642-36197-5_255-1



Main Features of Lifelines

Lifelines refer to the complex system and network assets of connected components, usually interacting with other components and systems, which are performing vital functions that are essential to sustain the life and the growth of a community, such as producing, transporting, and distributing goods or services. Their global value for the society and economy is permanently increased in our modern, technologically advanced, highly demanding, and fragile world. They constitute by themselves a set of critical facilities for the proper and safe functioning of the societies. In case of a strong earthquake motion, their physical damages and the consequent disruption of the services they provide may contribute seriously to the global economic loss. At the same time, their reparation cost may be very high, reaching in several cases 10 % or 15 % of the initial construction cost of the whole system to which they belong.

The term lifeline is ambiguous. Lifelines may be distinguished in two major categories: (i) Utility systems including potable water, natural gas, oil, electric power distribution, wastewater, or communication systems and (ii) transportation systems comprising roadways, railways, airport, and port facilities. Sometimes the terms infrastructures or critical facilities are used instead, at least for some of them.

Compared to buildings, lifeline systems present three distinctive features: (a) spatial variability and topology and exposure to different geological and geotechnical hazards, (b) wide variety of component typologies and material used, and (c) specific functionality requirements, which make them highly hierarchical networks. For practical reasons they are generally categorized in links (e.g., pipelines, roads) and nodes (e.g., tanks, power substations). Moreover, for their vulnerability assessment, an important parameter is the presence of synergies between components within the same system (intra-dependencies) or between different systems (interdependencies). Due to these synergies, the physical damage of a component of a system interacting with another one may affect seriously the second system’s performance and functionality. Therefore, it is essential to define the taxonomy of each interacting lifeline system that describes the individual components of each system and their role in the network as well as the way other systems affect its performance.

Due to their spatial extent, lifeline systems are exposed to variable seismic ground motions, often presenting important incoherency (shaking effects), and to geotechnical hazards expressed in terms of permanent ground deformations, resulted from fault crossing, landslides, and liquefaction (e.g., lateral spreading, settlements, buoyancy effects). Various parameters may be used to describe and characterize the strong ground motion and the induced phenomena. The selection of appropriate parameter (i.e., intensity measure), which is efficiently correlated with the response and the expected damages of each exposed elements, depends on the technical and response characteristics of each element.

As an example, a motorway network presents the following features: It includes multiple elements at risk like bridges, overpasses, interchanges, tunnels, culverts, embankments, cuts, slopes, retaining walls, signage and markings, electrical systems, rest areas, and buildings (e.g., tolls, maintenance, traffic management). These structures are usually built with different design methods, construction techniques, and materials. The seismic design requirements, and therefore their vulnerability, depend on the type and the location of the structure, the seismicity, and the tectonic features of the area. In particular, roadways are crossing different geological and topographical conditions such as valleys, rivers, canals, gulfs, or mountainous regions. Obviously, the network components can be exposed to various seismic ground motions in terms of amplitude, frequency, or duration, as well as to different geotechnical hazards (subsidence, liquefaction, landslides, faults, etc.). In this context, the design of a long bridge crossing an active seismic fault is much more demanding than a single span bridge in the same area. On the other hand, nowadays, the design level (and subsequently the resistance to earthquakes) is higher than of 30 years ago, making old structures more vulnerable, introducing a new factor, the aging or time effects. Furthermore, the hierarchy (importance) of roads depends on their traffic capacity, i.e., highways and major arterials and secondary and local urban roads. The importance of the road segments, serving different part of the urban fabric, depends also on the land use and the density of the population. The functional role of roads is related to the specific served areas such as industrial or residential areas, hospital facilities, and transport hubs (e.g., airports and major harbors). Therefore, the importance of each individual infrastructure and part of the network is variable and at the same time an essential factor in the seismic risk management. Finally, the synergies that are presented in a road network and can affect its overall functionality in case of earthquake include the supply of electric power for the signaling and lighting and the proximity to buildings and utility networks (e.g., water or gas pipes) that may induce the interruption of road traffic.

Experience from Past Earthquakes

Several devastating earthquakes occurred during the last century, causing extensive losses in terms of fatalities, injuries, damages, and disruptions. Knowledge, experience, and lessons learned from past earthquakes have significantly contributed to disaster risk reduction efforts across the globe.

The 1971 San Fernando earthquake in USA (ML6.6), an event that caused catastrophic damage to almost all types of lifelines, motivated the evolution of the lifeline earthquake engineering as well as many new efforts to better understand the causes of these failures and identify ways to mitigate future earthquake damage and disruption (e.g., the establishment of the Technical Council of Lifeline Earthquake Engineering/TCLEE, in ASCE). Other strong and damaging earthquakes that contributed to the progress of lifeline earthquake engineering are the 1978 Miyagiken-oki (Japan, Ms7.7), the 1985 Michoacan (Mexico, Mw8.3), the 1989 Loma Prieta (USA, Mw6.9), the 1990 Luzon (the Philippines, Ms7.8), the 1993 Kushiro-Οki (Japan, Mw7.6), the 1993 Hokkaido–Nansei-Oki (Japan, Mw7.7), the 1994 Northridge (USA, Mw6.7), the 1995 Hyogoken–Nanbu (Kobe) (Japan, Mw6.8), the 1999 Chi-Chi (Taiwan, Mw7.6), the 1999 Kocaeli (Turkey, Mw7.4), and the 2001 El Salvador (Mw7.6) events. More recent earthquakes which caused serious damages and losses to utility systems and transportation infrastructures, even in developed countries, include the 2004 Niigata ken Chuetsu (Mw6.8) and 2007 Niigata-Chuetsu Oki (Mw6.6) earthquakes in Japan, the 2008 Wenchuan (China, Mw7.9), the 2009 L’Aquila (Italy, Mw6.3), the 2010 Maule (Chile, Mw8.8), and recently in 2011 the Christchurch (New Zealand, Mw6.3) and the mega 2011 Tohoku (Japan, Mw9.0) earthquake and the tsunami associated with this earthquake.

The experience from major earthquakes proved that most lifeline elements are vulnerable, and the mid- and long-term effects, mainly in terms of financial losses, may be very important. A typical example of long-term effects is the port of Kobe in Japan, one of the largest container cargo ports in the world, which suffered major damages, mostly due to liquefaction-induced phenomena, as a result of the 1995 Mw6.8 Hyogoken–Nanbu earthquake. A year after almost all port infrastructures have been reconstructed, but it is remarkable that in 1998, 3 years after the disaster, cargo traffic remained at roughly half of pre-disaster levels. For the same earthquake, the direct repair cost for the Hanshin Expressway was $4.6 billion (NCEER 1995), while the total cost was actually much higher if the $3.4 million of daily income from tolls and the losses from business interruption and traffic delays was taken into account. In the 1994 Mw6.7 Northridge earthquake, the repair cost for damaged bridges was evaluated at $190 million, almost the 10 % of the total cost of the whole transportation system ($1.8 billion, Basöz and Kiremidjian 1998).

Interactions between systems may have a significant impact exceeding direct consequences. Loss of power or communication can severely impact the emergency response, especially when the disruptions concern critical facilities, such as hospitals, command centers, etc. Loss of power may have severe indirect effects due to the synergies between the lifelines and the dependence of all networks on the power supply system. In 1989 Mw6.9 Loma Prieta earthquake, the damages that occurred in the water network greatly affected the fire-fighting capacity in the area of San Francisco, Oakland, and Berkeley. Representative cases of interactions have been also reported after the 1994 Mw6.7 Northridge and 1995 Mw6.8 Hyogoken–Nanbu earthquakes, when the reparation of the water and gas networks have been delayed due to traffic congestion, street blocking, damaged buildings, and water that flowed into gas pipelines. A large area was also burned down in Kobe due to the disruption and the inefficiency of the fire-fighting network. Considering that modern societies and built environment are relying on the good performance of their interconnected infrastructures, it is clear that an efficient rescue policy and an optimum recovery strategy after a strong earthquake require the evaluation of the interactions between different lifeline systems.

In addition to ground shaking, ground failures (i.e., landslides, liquefaction, later spreading, fault ruptures), if they occur, may produce even more severe damages and losses to lifelines and infrastructures. As an example, the 2008 Mw7.9 Wenchuan earthquake in China triggered more than 15,000 landslides of various types within an area of 50,000 km2 and causing more than 20,000 fatalities representing one-quarter of the total fatalities. The landslides produced extensive damage to housing settlements, irrigation channels, highways, bridges, and other infrastructures. The city of Wenchuan and many other towns were isolated from the rest of the country, and the rescue and relief efforts were greatly affected due to blocking and damages in road network (Tang et al. 2011). During the 2010 Mw7.1 followed by the 2011 Mw6.3 Christchurch earthquake sequences in New Zealand, the ground and slope failures, mainly due to liquefaction, produced extensive damages to the water, electricity, and road networks rendering many of them inoperable (O’Rourke et al. 2012). On the contrary the gas system performed rather well. A large number of slope failures, landslides, and rock falls occurred during the 2004 Mw6.8 Niigata ken Chuetsu earthquake on steep natural slopes and artificial cuts in the area, resulting in major damage to road and railway infrastructures (cuts, embankments, pavements, tunnels, abutments, retaining walls), as well as to other lifeline components (water pipes, electric power poles). Similar damages and disruptions, related to geotechnical failures, were recorded after the subsequent 2007 Mw6.6 Niigata-Chuetsu Oki earthquake in Japan and in several recent events in Europe (e.g., 2003 Mw6.4 Lefkas earthquake in Greece, the 2009 Mw6.3 L’Aquila earthquake in Italy, the 1999 Mw7.4 Kocaeli earthquake in Turkey) and worldwide. In Figs. 1, 2, 3, 4, 5, 6, 7, and 8, some representative damages for utility and transportation networks are shown.
Fig. 1

Damages to road, telecommunication, and gas system in 1994 Northridge (Mw6.7) earthquake (Photo Credit: US Geological Survey)

Fig. 2

Damaged equipment in Adapazari substation in 1999 Kocaeli (Mw7.4) earthquake (Photo Credit: Kandilli Observatory and Earthquake Research Institute/KOERI)

Fig. 3

Lateral deformation of train tracks at the Arahama station due to surface waves in 2007 Niigata-Chuetsu Oki (Mw6.6) earthquake (Photo Credit: US Geological Survey)

Fig. 4

Damages to road embankments in 2007 Niigata-Chuetsu Oki (Mw6.6) earthquake (Photo Credit: US Geological Survey)

Fig. 5

Collapse of a section of the Nishinomiya in 1995 (Mw6.8) Hyogoken–Nanbu earthquake (Photo by K. Pitilakis)

Fig. 6

Damage to Trans-European Motorway due to fault rupture in 1999 Kocaeli (Mw7.4) earthquake (Photo Credit: Kandilli Observatory and Earthquake Research Institute/Photo by D. Kalafat)

Fig. 7

Blocking of Pan-American Highway due to landslide in 2001 El Salvador (Mw7.6) earthquake (Photo Credit: US Geological Survey/Photo by E.L. Harp)

Fig. 8

Failure of port infrastructures in 1995 (Mw6.8) Hyogoken–Nanbu earthquake (Photo by K. Pitilakis)

Several utility systems and networks are now outdated and lack proper seismic design and protection. For example, there are parts of the water and wastewater systems in several cities built more than a century ago. Hence, their material properties and strength are seriously affected from aging effects and are consequently more vulnerable and exposed to earthquake risk. Currently, most of urban infrastructures are not designed to sustain severe seismic hazard and complex systemic threats, and it is difficult to anticipate what may be the complex socioeconomic losses due to potential strong earthquakes. The poor performance of elements at risk in a network is likely to make the whole system much more vulnerable and susceptible to strong earthquakes.

Finally, the short- and long-term economic impact due to earthquake damages of lifelines can be significant. In a recent study by CATDAT (Daniell and Vervaeck 2013), the disaggregation of direct economic losses of 61 selected major earthquakes from 1900 to 2012 showed that around 30 % of the direct losses came from infrastructure losses in transport, communications, pipelines, energy supply systems, etc.

Lifeline Earthquake Engineering Research

In the last two decades, several important projects and research efforts related to the vulnerability and risk assessment of different lifelines at regional or urban scale have been undertaken. They resulted in the development of relevant methods and guidelines. A short reference to the most important contributions is following.

United States

In 1992, the Multidisciplinary Center for Earthquake Engineering Research (MCEER) initiated the Highway Project, funded by the Federal Highway Administration (FHWA). The project uniquely examined the impact of earthquakes on the highway system as an integrated network, rather than a collection of individual road segments, bridges, embankments, tunnels, etc. The purpose of the project was to ensure the usability of highways following earthquakes, by improving performance of all interconnected components. Overall goals were to deepen the understanding of the seismic hazard impact to highways and to improve and develop analysis methods, screening procedures and additional tools, retrofit technologies, design criteria, and other approaches to reduce seismic vulnerability of existing and future highway infrastructure. REDARS (Risks due to Earthquake DAmage to Roadway Systems) is a public-domain software package developed under this MCEER project that accounts for how earthquake damage affects post-event traffic flows and travel times and estimates losses from these travel-time and traffic-flow impacts (http://mceer.buffalo.edu/research/redars/).

In 1996, the PEER Lifelines Program (http://peer.berkeley.edu/lifelines) was initiated, under the coordination of the Pacific Earthquake Engineering Research Center, funded by the California Department of Transportation and the Pacific Gas and Electric Company. The aim was to improve the seismic safety and reliability of lifeline systems with particular focus on the characterization of ground motions, local soil and site response, and the performance of bridge structures and electric substation equipment.

In 1997, the Federal Emergency Management Agency (FEMA) released the first edition of HAZUS methodology and software that estimates potential building and infrastructure losses from earthquakes, built using GIS technology. The current version is HAZUS-MH V2.0, including also losses from floods and hurricanes. It was released in 2004 (www.fema.gov/hazus). HAZUS-MH loss estimates reflect the state-of-the-art scientific and engineering knowledge and can be used to inform decision makers at all levels of government by providing a reasonable basis for developing mitigation, emergency preparedness, response, and recovery plans and policies. The main merit of the HAZUS platform is that it provided for the first time an unparalleled set of fragility models for all the main components and systems of the built environment. However, it must be recognized that many of these models have been derived solely on expert judgment and the level of knowledge at that time. One effect of the sheer size of the HAZUS framework and its set of tools is that it established itself very soon as the reference for all studies in the sector worldwide.

In 1998, FEMA and the American Society of Civil Engineers entered into a cooperative agreement to establish the American Lifelines Alliance (ALA). ALA’s objectives were to facilitate the creation, adoption, and implementation of design and retrofit guidelines and other national consensus documents to improve the performance of utility and transportation systems (electric power, telecommunication, water, wastewater, oil, natural gas, rail, and shipping ports) in natural hazard events, including earthquakes. ALA has not attempted to focus efforts on all lifeline systems and components but has instead chosen to place priority on specific topics relevant to selected lifeline systems where a need for improved hazard mitigation practices is identified. The primary function was to bridge the gap between hazard mitigation practices for buildings and lifeline systems; therefore a close collaboration with lifeline stakeholders, owners/operators, and recognizant regulatory bodies was established to facilitate use of developed guidelines (http://www.americanlifelinesalliance.com/).

ERGO (the new name for the Multi-Hazard Assessment, Response, and Planning software previously known as mHARP and MAEviz) started somewhat later than HAZUS and was the product of the research efforts carried out at the Mid-America Earthquake Center in collaboration with the National Center for Supercomputing Applications (NCSA). It is an open-source software and incorporates many of the design concepts and capabilities motivated by NCSA efforts to develop “cyberenvironments” that span scientific disciplines and that can rapidly evolve to incorporate new research results (Elnashai et al. 2008). An important aspect of ERGO is its extensibility, both in terms of analysis/features modules and of visualization/representation (GIS) modules. The framework has been designed to implement the consequence-based risk management (CRM) paradigm supported by the NCSA and MAE center.


Until now, no coordinated action for the development of a comprehensive methodology and software tool at European level is made. However, important efforts have been undertaken through research projects mainly funded by European Union. Brief description of the most important is following.

RISK-UE (2001–2004) project involved the assessment of earthquake scenarios based on the analysis of the global impact of one or more plausible earthquakes at city scale, within a European context. The primary aim of these scenarios was to increase awareness within the decision-making centers of a city. The developed methodology for creating earthquake scenarios focused on the distinctive features of European cities with regard to current and historical buildings and lifelines, as well as on their functional and social organization, in order to identify weak points within the urban system. The approach was applied to 7 European cities: Barcelona, Bitola, Bucharest, Catania, Nice, Sofia, and Thessaloniki (Mouroux and Le Brun 2006).

LESS-LOSS (2003–2007) was a European Integrated Project focusing on risk mitigation for earthquakes and landslides that relied on the active participation of 46 partners from both academia and industry. The project addressed research issues on seismic engineering, earthquake risk and impact assessment, landslides monitoring, mapping and management strategies, improved disaster preparedness and mitigation of geotechnical hazards, development of advanced methods for risk assessment, methods of appraising environmental quality, and relevant pre-normative research. Separate sections were devoted on earthquake disaster scenario predictions and loss modeling for urban areas and infrastructures with emphasis on water and gas systems (Faccioli 2007).

SYNER-G (2009–2013, www.syner-g.eu) is the most recent European funded research project that developed an integrated, general methodology and a comprehensive framework for the systemic vulnerability and risk analysis of buildings, lifelines, and infrastructures. The proposed methodology and tools encompass in an integrated fashion all aspects in the chain, from hazard to the vulnerability assessment of components and systems and to the socioeconomic impacts of an earthquake, accounting for most relevant uncertainties within an efficient quantitative simulation scheme and modeling interactions between the multiple component systems. It systematically integrates the most advanced fragility or vulnerability functions to assess the vulnerability of physical assets. The increasing impact due to interdependencies and intra-dependencies between different components and among different interacting systems is treated in a comprehensive way, providing specifications for each network and infrastructure. The proposed socioeconomic model integrates social vulnerability into the physical systems modeling approaches providing to decision makers with a dynamic platform to capture post disaster emergency issues like shelter demand and health impact decisions (Pitilakis et al. 2014a, b).

Other Efforts Worldwide

The Central American Probabilistic Risk Assessment (CAPRA) platform was developed in partnership with Central American governments, the support of the Central American Coordination Centre for Disaster Prevention (CEPREDENAC), the Inter-American Development Bank (IDB), the International Strategy of United Nations for Disaster Reduction (UN-ISDR), and the World Bank. CAPRA is an information platform to enhance decision-making in risk management across various sectors, such as emergency management, territorial planning, public investment, and the financial sectors. Through the application of probabilistic risk assessment principles to threats like hurricanes, earthquakes, volcanic activity, floods, tsunamis, and landslides, CAPRA allows to measure and compare different types of risks and to develop sector-specific applications for risk management (Cardona et al. 2012; http://www.ecapra.org/). It is not directly designed for lifeline vulnerability and risk assessment, but it can be used for this purpose as well.

Key Elements of Seismic Risk Assessment

Risk assessment is the process used to determine risk management priorities by evaluating and comparing the level of risk to specific standards, either defined by a code or a set of target risk levels and criteria. The estimation of losses, both material and immaterial, is the ultimate target of the risk assessment. Then, the risk management procedure should define pre-seismic (retrofitting and strengthening actions), coseismic, and post-seismic strategies and policies. The risk assessment of lifelines and infrastructures follows the general scheme of seismic risk assessment:
$$ \left[\mathrm{seismic}\ \mathrm{risk}\right]=\left[\mathrm{seismic}\ \mathrm{hazard}\right]\times \left[\mathrm{vulnerability}\right]\times \left[\mathrm{exposure}-\mathrm{elements}\ \mathrm{at}\ \mathrm{risk}\right] $$
The complexity of elements at risk, their variability from one place and one country to another, and, till recently, the lack of well-validated damage and loss data from strong earthquakes make the vulnerability assessment of each particular component and of the network as a whole a challenging task. Adding to that the spatial extent of lifelines, the synergies between different systems, and the several sources of uncertainties that are inherent in the various aspects (e.g., typology, damage states), models (e.g., seismic hazard, spatial correlation, fragility curves), and finally tools and methods for estimation of losses, it is obvious that the risk assessment of lifelines is indeed a very complex and challenging issue. Therefore, multidisciplinary task and combined efforts are really needed to reach reliable and comprehensive estimates.

This difficulty is amplified due to lack of well-documented data and the partial or poor knowledge of the geometric and other features of the systems; in some cases it is also due to the confidentiality of data (i.e., communication or oil networks). The situation however has been improved the last decade. Important strong earthquakes provided valuable good quality data, while the public awareness and the reported huge direct and indirect losses associated to lifeline damages drew the attention of the scientific community, the governmental authorities, and the insurance sector. Moreover, the development of geographical information systems (GIS) as well as remote sensing technologies offered an excellent platform for the implementation of efficient and innovative techniques to screen, capture, store, manipulate, analyze, manage, and present geographical data.

Few terms employed in this chapter are described in the following. Inventory and taxonomy/typology definitions of all elements at risk and systems are the necessary initial steps to describe the elements and networks exposed to seismic risk; at the same time, the development of detailed inventories is the most expensive task in terms of financial cost and time. The estimation of seismic hazard and site-specific seismic ground motion characteristics provides appropriate seismic intensity measures (IM), which will be used later in the vulnerability and loss assessment modeling. Vulnerability is the propensity of damage of an element (building, bridge, pipeline, roadway, oil tank, etc.) or a network of specific elements (e.g., water or gas system) to a given seismic intensity defined in the seismic hazard analysis. It is commonly assessed based on fragility models, which estimate the probability of damage for a given seismic intensity. Fragility functions are constructed with respect to the taxonomy and typological characteristics of each element at risk. The expected damages and losses of the entire network are estimated through a systemic analysis, which takes into account the physical damages of the various parts and segments of the system, and in some cases, the intra-dependencies between different components within the system and the interdependencies between systems, in order to estimate the serviceability or flow capacity of the damaged network. The results are commonly provided in terms of performance indicators (PI), which are describing the network functional or the expected economic or socioeconomic losses. Efficient risk management is based on the evaluation of the resulting performance indicator values.

Seismic Hazard for Distributed Networks

Lifelines are in most cases spatially distributed systems. Therefore, seismic hazard should meet the specific features of each system considering also its spatial variability. The vulnerability analysis and risk assessment must be evaluated according to the precise typological characteristics of the components and networks, taking also into account the models used to describe vulnerability, usually in terms of fragility curves and relationships. Moreover, due to the spatial extent of utility and transportation networks, the seismic hazard should describe the spatial variability of ground motion considering local soil conditions, topographic effects, and geotechnical hazards; site-specific seismic hazard analyses are always necessary. Traditional seismic hazard analysis, while effective in translating the hazard into a probabilistic formulation, is limited in the extent to which it can incorporate spatial coherency of the form needed for estimation of loss to spatially distributed portfolios. The extension of seismic risk analysis to multiple systems of spatially distributed infrastructures presents new challenges in the characterization of the seismic hazard input, particularly with respect to the spatial correlation structure of the ground motion residuals (i.e., residuals from empirical intensity models) that form the basis for the systemic risk analysis.

A general procedure entitled “Shakefield” has been recently established in the frame of SYNER-G project (Pitilakis et al. 2014b), which allows for the generation of samples of ground motion fields both for single-scenario-type events and for stochastically generated sets of events needed for probabilistic seismic risk analysis (Weatherill et al. 2014). For a spatially distributed infrastructure of vulnerable elements, the spatial correlation of the ground motion fields for different measures of the ground motion intensity is incorporated into the simulation procedure. This is extended further to consider spatial cross-correlation between different measures of ground motion intensity.

The consideration of hazard from permanent ground deformation (PGD) is essential in modeling the seismic risk to lifeline systems. For pipelines and similar systems with linear elements, fragility models are generally given in terms of PGD, as they are mostly vulnerable to permanent displacement of the ground rather than transient shaking. Four primary causes of permanent ground displacements are commonly considered: liquefaction-induced lateral spread, liquefaction-induced settlement, slope displacement, and coseismic fault rupture. In addition to strong shaking, another transient effect that poses a potential risk to lifeline systems is the transient ground strain (PGS). Several models are available for the estimation of PGD; some of them are intended to relate the degree of deformation and the probability of the geotechnical hazard occurring to the intensity of the ground motion (Weatherill et al. 2014). However, it should be noted that most theoretical and empirical models relating PGD to strong shaking require a level of geotechnical detail that may be impractical to obtain for a spatially distributed set of sites. HAZUS methodology (FEMA 2003) provides a rather simple “baseline” model that can be implemented in the widest variety of applications.

When the scenario type or probabilistic site-specific seismic hazard is available, the damage assessment of a lifeline system is evaluated using appropriate fragility or vulnerability functions relating the probability of damages to seismic intensity measures. The data and parameters needed to perform this analysis are shortly described in the following.

Taxonomy, Inventory, and Typology

The key assumption of the vulnerability assessment of lifelines is that the structures and components of the systems, having similar structural characteristics (e.g., a bridge of a given typology), being in similar geotechnical conditions (e.g., a segment of a pipeline crossing the same soil conditions and exposed to the same geotechnical hazards), are expected to perform in the same way for a given seismic excitation. Within this context, expected damages are directly related to the structural properties of the elements at risk. Taxonomy and typology are thus fundamental descriptors of a system that are derived from the inventory of each element and system. Geometry, material properties, morphological features, age, seismic design level, anchorage of the equipment, soil conditions, and foundation details are among usual typology descriptors/parameters. Buildings, bridges, pipelines (gas, fuel, water, wastewater), tunnels, road embankments, power substations, harbor facilities, and road and railway networks have their own specific set of typologies and different taxonomy.

The taxonomy of any lifeline network is thus an essential step for identifying, characterizing, and classifying all types of lifeline elements according to their specific typology and their distinctive geometric, structural, and functional features. It allows the classification of the network elements in an ordered classification system adequate for the seismic risk assessment of each component and the network (or system) as a whole. The taxonomy is specific for each lifeline system and it needs a detailed typology definition of all elements at risk. For that, an adequate inventory process is essential. However, several difficulties arise in the collection and archiving of the data related to the absence of well-organized archives in public and often in private organizations managing the systems, the oldness of the networks (e.g., water network), and the high cost to draw up complete inventories. To this respect remote sensing techniques and GIS offer a useful and indispensable instrument and platform to enhance with relatively low cost the collection of the data and implement any inventory inquires. Different types of satellite data can be used to extract information and parameters needed to compile or update inventories for seismic risk assessment. Examples include the width of roads, the geometry or material of buildings, building aggregates, infrastructures like bridges and channels, land use characteristics, and others.

The inventory of a specific structure in a region and the capability to create classes of structural types (e.g., with respect to material, geometry, and design code level) are among the main challenges when carrying out a general seismic risk assessment, for example, at a city or region scale, where it is practically impossible to perform this assessment separately for each single structure. It is necessary to classify all elements at risk, in “as much as possible” homogeneous classes and subclasses presenting more-or-less similar response characteristics to ground shaking. Thus, the derivation of appropriate fragility curves for any type of structure depends entirely on the creation of a reasonable taxonomy that is able to classify the different kinds of structures and infrastructures in any system exposed to seismic hazard. Several uncertainties are inherent to each taxonomy system, which are inevitably propagated through the generic fragility (vulnerability) curves to the final loss estimation. The most coherent and comprehensive taxonomy presently available in Europe is produced in SYNER-G project (Table 1, Pitilakis et al. 2014a). Detailed classifications are also provided in HAZUS (FEMA 2003) and ALA (2001).
Table 1

SYNER-G infrastructure taxonomy (Pitilakis et al. 2014a)


Component (and subcomponents)

BDG: buildings

Force-resisting mechanism (FRM1): moment-resisting frame structural wall, flat slab, bearing walls, precast, confined masonry

FRM material (FRMM1): concrete, masonry

Plan (P): regular, irregular

Elevation (E): regular/irregular geometry

Cladding (C): regular infill vertically, irregular infill vertically, bare

Detailing (D): ductile, non-ductile, with tie rods/beams, without tie rods/beams

Floor system (FS): rigid, flexible

Roof system (RS): peaked, flat, gable end walls

Height level (HL): Low rise, mid-rise, high rise, tall

Code level (CL): none, low, moderate, high

EPN: electric power network

EPN01: electric power grid

EPN02: generation plant

EPN03: substation

EPN04: distribution circuits

EPN05–09: substation macro-components (autotransformer line; line without transformer; bar-connecting line; bars; cluster)

EPN10–23: substation micro-components (circuit breaker; lightning arrester or discharger; horizontal disconnect switch or horizontal sectionalizing switch; vertical disconnect switch or vertical sectionalizing switch; transformer or autotransformer; current transformer; voltage transformer; box or control house; power supply to protection system; coil support; bar support or pothead; regulator; bus; capacitor tank)

EPN24: transmission or distribution line

GAS: natural gas system

GAS01: production and gathering facility (onshore, offshore)

GAS02: treatment plant

GAS03: storage tank

GAS04: station (compression; metering/pressure reduction; regulator; metering)

GAS05: pipeline


OIL: oil system

OIL01: production and gathering facility (onshore, offshore)

OIL02: refinery

OIL03: storage tank farm

OIL04: pumping plant

OIL05: pipeline


WSS: water-supply network

WSS01: source (springs, rivers, natural lakes, impounding reservoirs, shallow or deep wells)

WSS02: treatment plant

WSS03: pumping station

WSS04: storage tank

WSS05: pipe

WSS06: tunnel

WSS07: canal

WSS08: SCADA system

WWN: wastewater network

WWN01: wastewater treatment plant

WWN02: pumping (lift) station

WWN03: pipe

WWN04: tunnel

WWN05: SCADA system

RDN: road network

RDN01: bridge (material, type of deck, deck structural system, pier to deck connection, type of pier to deck connection; type of section of the pier, spans, type of connection to the abutments, skew, bridge configuration, foundation type, seismic design level)

RDN02: tunnel

RDN03: embankment (road on)

RDN04: trench (road in)

RDN05: unstable slope (road on or running along)

RDN06: road pavement (ground failure)

RDN07: bridge abutment

RWN: railway network

RWN01: bridge

RWN02: tunnel

RWN03: embankment (track on)

RWN04: trench (track in a)

RWN05: unstable slope (track on or running along)

RWN06: track

RDN07: bridge abutment

RWN08: station

HBR: harbor

HBR01: waterfront components (gravity retaining structures; sheet pile wharves; piers; breakwaters mooring and breasting dolphins)

HBR02: earthen embankments (hydraulic fills and native soil material)

HBR03: cargo handling and storage components (cranes, tanks, etc.)

HBR04: buildings (sheds, warehouse, offices, etc.)

HBR05: liquid fuel system (as per the OIL system)

FFS: fire-fighting system

FFS01: fire-fighter station

FFS02: pumping station

FFS03: storage tank

FFS04: fire hydrant

FFS05: pipe

Each component is further classified according to its specific features regarding its seismic behavior. For instance, the main typological features of gas pipelines include the material type, material strength, diameter, wall thickness and smoothness of coating, type of connection and joints, as well as the nominal design and actual flow. Among them, material and the connection types (joints) together with the soil type are the most critical parameters for their seismic response. Table 2 outlines common types of pipelines based on the above classification criteria.
Table 2

Common types of materials and connections for buried pipelines

Material type

Connection type

Asbestos cement (AC)

Arc welded

Cast iron (CI)

Bell and spigot

Ductile iron (DI)


Concrete (C)


Polyvinyl chloride (PVC)

Rubber gasket

Welded steel (WS)

Gas welded

Medium-density polyethylene (MDPE)


High-density polyethylene (HDPE)

Another example is storage tanks, which are usually categorized according to their material type (steel or reinforced concrete), construction type (at grade or elevated), anchorage (anchored or unanchored), roof type and capacity, shape factor (height-on-diameter ratio), and amount of content in the tank (full, half full, empty). Among them, material type, construction type, and anchorage are considered as the most important. Table 3 presents the classification of tanks in USA according to ALA (2001).
Table 3

Typology of tanks (ALA 2001)

Unanchored redwood tank (5 × 104–5 × 105 gal)

Unanchored post-tensioned circular concrete tank (>1 × 106 gal)

Unanchored steel tank with integral shell roof (1 × 105–2 × 106 gal)

Unanchored steel tank with wood roof (1 × 105–2 × 106 gal)

Anchored steel tank with integral steel roof (1 × 105–2 × 106 gal)

Unanchored steel tank with integral steel roof (>2 × 106 gal)

Anchored steel tank with wood roof (>2 × 106 gal)

Anchored reinforced (or prestressed) concrete tank (5 × 104–1 × 106 gal)

Elevated steel tank with no seismic design

Elevated steel tank with nominal seismic design

Open-cut reservoir

Fiberglass tanks

For a systematic and detailed description of all classes and subclasses of all elements at risk in most lifeline systems and networks in the European context, the reader should refer to Pitilakis et al. (2014a).

Damage Assessment

The damage patterns and mechanisms are different for each lifeline system and element (component), and they are strongly depending on the typology of each structure (i.e., materials, geometry, structural types). For example, the damage mechanisms of pipelines are generally classified in the form of breaks or leaks; material type (i.e., brittle or ductile) and joints (i.e., flexible or rigid) are also factors affecting the seismic response and damages of pipelines. In case of tunnels, damage patterns include lining cracks (longitudinal or transverse) and spalling, wall deformation, bending and buckling of reinforcing bars, obstruction of the opening, or pavement cracks. In subway stations damages to columns may also occur. Damages to quay walls are related to excessive lateral pressures or decrease of shear strength of the foundation soil that can cause sliding, deformation, and tilting of the walls. Settlements and lateral movement of the backfill materials and cracking of apron pavements can be also induced. Damage to electric power substations includes failures of the various subcomponents (e.g., disconnect switches, circuit breakers, transformers) or damage of the building. Earth structures such as highway and railway embankments can spread laterally and settle, resulting in opening of cracks in the road pavement or displacement of the railway tracks. The list of possible damage patterns is unlimited.

Therefore, classification of damage and the subsequent definition of specific damage states are important in the vulnerability assessment as the seismic intensity is correlated to the expected damage level through the fragility or vulnerability functions. Again, the form of the fragility functions depends on the typology of the element at risk. For common structures (e.g., buildings, bridges) and other not extended elements (e.g., cranes, tanks, substations), the fragility curves describe the response and damage level of particular subcomponents (e.g., columns, transformers) or of the entire structure. For linear elements of extended networks such as gas pipelines, the fragility functions describe the number of expected damages along a certain length (i.e., per km). Examples and further details are given in the next sections.

Damage States

In seismic risk assessment, the performance levels of a structure, for example, a reinforced concrete building, belonging to a specific class (Pitilakis et al. 2014a) can be defined through damage thresholds called limit states. A limit state defines a boundary between two different damage conditions often referred to as damage states. The thresholds are related to functionality and serviceability levels while they are usually defined based on engineering judgment and common sense. They are also strongly depending on the model applied for the analysis and derivation of the fragility functions. Uncertainties related to this stage of the analysis are normally referred as epistemic uncertainties (see Pinto 2014). Different damage criteria have been proposed depending on the typologies of elements at risk and the approach used for the derivation of fragility curves. The most common way to define earthquake consequences is a classification in terms of the following damage states: no damage, slight/minor, moderate, extensive, and complete. This qualitative approach requires an agreement on the meaning and the content of each damage state. The number of damage states is variable and is related with the functionality of the components and/or the repair duration and cost. In this way the total losses of the system (economic and functional) can be estimated. In particular, physical damages are related to the expected serviceability level of the component (i.e., fully or partial operational or inoperative) and the corresponding functionality (e.g., power availability for electric power substations, number of available traffic lanes for roads, flow or pressure level for water system). These correlations provide quantitative measures of the component’s performance and can be applied for the definition of specific performance indicators (PIs), which are introduced in the systemic analysis of each network. Therefore, the comparison of a demand with a capacity quantity, the consequence of a mitigation action, or the accumulated consequences of all damages (usually referred as the “impact”) can be evaluated.

Methods for deriving fragility curves generally describe damages on a discrete damage scale. In the empirical procedures, the scale is used in survey efforts to produce post-earthquake damage statistics, and sometimes it is rather subjective (i.e., there is often a discrepancy between the damage levels that any two different inspectors would assign for the same incident). In analytical procedures the scale is related to limit state of selected mechanical properties that are described by appropriate indices, such as the displacement capacity or the storey drift in the case of buildings or simple drift in pier bridges. For other elements at risk, the definition of the performance levels or the limit states may be more vague and follows other criteria related, for example, in the case of pipelines, to the limit strength characteristics of the material used in each typology. The definition and consequently the selection of the damage thresholds, i.e., limit states, are among the main, yet, unavoidable sources of uncertainties.

Intensity Measures

An important issue related to the fragility curve construction and implicitly to the risk assessment is the selection of an appropriate earthquake intensity measure (IM) that characterizes the strong ground motion that best correlates with the response of each element, for example, building, pipeline, or harbor facilities like cranes. Several measures of the intensity of ground motion (IMs) have been developed. Each intensity measure may describe different characteristics of the motion, some of which may be more adverse for the structure or the system under consideration. The use of a particular IM in seismic risk analysis should be guided by the extent to which the measure corresponds to damage to the components of a system. Optimum intensity measures are defined in terms of practicality, effectiveness, efficiency, sufficiency, robustness, and computability (Mackie and Stojadinovic 2003).

In general, IMs are grouped in two general classes: empirical intensity measures and instrumental intensity measures. With regard to the empirical IMs, different macroseismic intensity scales could be used to identify the observed effects of ground shaking over a limited area. Instrumental IMs are, by far, more accurate and representative of the seismic intensity characteristics and the severity of ground shaking. For example, for bridges the best descriptor is a spectral response value at a specific period (i.e., T = 1.0 s). For other lifeline components, it may be the peak ground acceleration (e.g., buildings, tanks, electric power substations), peak ground velocity (e.g., pipelines), or even the permanent ground deformations (e.g., pipes, embankments, roadways, railways). The correlation between damages of specific elements at risk and intensity measures is not simple and never unique. Several other descriptors like peak ground strain, Arias intensity, cumulative absolute velocity, and other parameters of the ground motion have been also used for different structures composing a lifeline system. More recently, it is proposed to use two descriptors instead of one, and hence the fragility curve is transformed in fragility surfaces (Seyedi et al. 2010; Douglas et al. 2014).

The selection of the adequate intensity parameter is also related to the approach that is followed for the derivation of fragility curves and the typology of elements at risk. The identification of the proper IM is determined from different constraints, which are first of all related to the adopted hazard model, but also to the element at risk under consideration and the availability of data and fragility functions for all different exposed assets. Empirical fragility functions are usually expressed in terms of the macroseismic intensity defined according to different macroseismic scales (EMS, MCS, and MM). Analytical or hybrid fragility functions are, on the contrary, related to instrumental IMs, which are related to parameters of the ground motion (PGA, PGV, PGD) or of the structural response (spectral acceleration Sa or spectral displacement Sd, for a given value of the period of vibration T). When the vulnerability of elements due to ground failure is examined (i.e., liquefaction, fault rupture, landslides), permanent ground deformation (PGD) is the most appropriate IM.

Vulnerability Assessment and Fragility Curves

The fundamental tool in seismic risk assessment of lifeline components is the fragility curves which describe the probability that a structure will reach or exceed a certain damage state for a given ground motion intensity. An extensive review of available fragility functions and state-of-the-art methods for vulnerability assessment of buildings and lifeline components can be found in Pitilakis et al. (2014a). Fragility curves are usually represented by two-parameter (median and log-standard deviation) cumulative lognormal distributions. Several approaches are used to establish fragility functions. They are grouped in the following five categories:
  • Empirical fragility curves, based on post-earthquake surveys and observations of actual damage. They are specific to particular sites and seismotectonic, geological, and geotechnical conditions, as well as the properties of the damaged structures. Consequently, the use of these functions in different regions is always questionable (Figs. 9 and 10).

Fig. 9

Empirical fragility curves for power grids made up of substations of different voltages based on data from US west coast earthquakes (Dueñas-Osorio et al. 2007)

Fig. 10

Empirical fragility curves for on-ground steel tanks subjected to ground shaking (ALA 2001)

In case of pipelines, the empirical fragility functions relate the repair rates (RR) expressed as repairs/km with the peak ground velocity (PGV) or permanent ground deformation (PGD) (Fig. 11). The curves may be further adapted to the material properties and geometry of the pipelines and soil conditions. In the last two decades, the increased density of high-quality strong ground motion records in different soil conditions, in combination with new technologies such as geographical information systems (GIS) and remote sensing technologies (e.g., LiDAR) capable to measure more accurately ground movements, contributed significantly to the development and verification of such relationships (O’Rourke et al. 2012).
Fig. 11

Empirical fragility functions for common pipeline typologies provided by ALA (2001): (a) pipelines subjected to wave propagation (PGV), (b) pipelines subjected to permanent ground deformation (PGD)

  • Analytical fragility curves, based on numerical simulations of structural models under increasing earthquake loads (Figs. 12 and 13). Analytical methods, validated with large-scale experimental data and observations from recent strong earthquakes, have become more popular in recent years. The main reason is the considerable improvement of computational tools, methods, and skills, which allows comprehensive parametric studies covering most common typologies to be undertaken. Moreover, several of the associated uncertainties, e.g., material properties, are better controlled.

  • Judgmental or expert elicitation fragility curves, using questionnaires by which the experts are queried on the probability of a component being in a certain damage state for a given intensity (Fig. 14). They are versatile and relatively fast to establish, but their reliability is questionable because of their dependence on the experiences of the experts consulted.

  • Hybrid fragility curves, which combine any of the above-mentioned techniques in order to compensate for their respective drawbacks.

  • Fragility curves based on a fault-tree analysis, where complex components (e.g., substations, pumping plants, hospitals) are broken down into subcomponents and the global fragility is obtained based on the relationships between the subcomponents and their individual fragilities (Fig. 15).

Fig. 12

Analytical fragility curves for alluvial bored tunnels due to ground shaking, classified to ground type B, C, and D according to Eurocode 8 (Argyroudis and Pitilakis 2012)

Fig. 13

Analytical fragility curves for gravity waterfront structures due to ground shaking, classified according to the wall height (H) and the soil foundation conditions (Vs values) (Kakderi and Pitilakis 2010)

Fig. 14

Expert judgment fragility curves for railway tracks subjected to permanent ground deformation (Argyroudis and Kaynia 2014)

Fig. 15

Fault-tree based fragility curves for water treatment plant with anchored components subjected to ground shaking (Pitilakis et al. 2014a)

An example of damage estimation for “on-ground unanchored steel tanks” due to ground shaking and “pipelines” due to ground failure is given in Table 4. In the first case, the exceedance probabilities of each damage state are estimated based on the fragility curves in Fig. 10, and then the punctual probabilities of each damage state are obtained. In the second case, the repairs per kilometer are estimated for a given value of PGD based on the curves in Fig. 11. The estimated repairs should be modified according to the length of the pipe segment under study.
Table 4

Example of damage estimation for steel tanks and pipelines


On-ground unanchored steel tanks subjected to ground shaking (PGA = 0.4 g)

Damage state

No damage





Probability of exceedance





Probability of occurrence











Pipelines subjected to ground failure (PGD = 40 cm)

Material diameter

Welded steel – large

Welded steel – small

PVC – small





Systemic Analysis

The majority of the methodologies that have been developed worldwide for the seismic risk assessment of lifelines and infrastructures refer to single systems, without considering interactions, cascading failures, and complex impacts. The study of interdependent infrastructures is challenging due to heterogeneous quality and insufficient data availability and the need to account for their spatial and temporal aspects of complex supply–demand operation (Satumtira and Duenas-Osorio 2010). The various approaches available in the literature depend on the simulation method, modeling objectives, scale of analysis, availability of input data, and end-user type or needs. A comprehensive categorization scheme describing different levels is provided in Modaressi et al. (2014):
  • Vulnerability analysis: This level considers only the potential physical damages of the components of the systems, with no consideration of functionality of either the elements or the whole system.

  • Connectivity analysis: This level describes the probability of the demand nodes to be connected to functioning supply nodes through undamaged paths. In this approach the damaged components are removed from the network and the adjacency matrix is updated accordingly, thus pointing out the nodes or areas that are disconnected from the rest of the system. This qualitative approach is used for all utility networks (water, electricity, gas) and the road transportation system.

  • Capacitive analysis: This level describes the ability of the system to provide to the users the required functionality through a quantitative approach. For utility networks, graph algorithms and flow equations can be used to estimate capacitive flows from sources (e.g., generators, reservoirs) to sinks (i.e., distribution nodes), based on the damages sustained by the network components (from total destruction to slight damages reducing the capacity).

  • Fault-tree analysis: This level of analysis concerns critical infrastructures, where multiple conditions are necessary for the systems to ensure its task. This type of approach aims to evaluate the remaining operating capacity (residual operation capacity) of objects such as health-care facilities. The system is broken down into structural, nonstructural, or human components, each one of them being connected with logic operators.

The performance of each network (e.g., utility or accessibility losses) is commonly measured through appropriate performance indicators (PIs), which, if combined with direct losses from physical damages, can yield a first partial estimate of the overall socioeconomic impact of an earthquake. Performance indicators, at the component or the system level, depend on the type of analysis that is performed. Connectivity analysis gives access to indices such as the connectivity loss (measure of the reduction of the number of possible paths from sources to sinks). Capacitive modeling yields more elaborate performance indicators at the distribution nodes (e.g., head ratio for water system, voltage ratio for electric buses) or for the whole system (e.g., system serviceability index comparing the customer demand satisfaction before and after the seismic event). The fault-tree analysis method is generally used for the derivation of fragility curves for specific components that comprise a set of subcomponents (e.g., health-care facilities, water treatment plants).

The importance of the interconnection between different systems is a more recent acquisition that targets two or, rarely, more systems (Pitilakis et al. 2014b). Several classifications have been proposed to categorize the types of interactions. The most common are physical, demand, and geographic interactions (Rinaldi et al. 2001). Physical interaction describes physical reliance on material flow from one infrastructural system to another as, for example, the supply of power to various network facilities by electric generators. Demand interactions correspond to a supply–demand from a given component to another system. An example is the number of casualties that should be treated by health-care facilities after an earthquake. Finally, geographic interactions describe the way that a local environmental event affects components across multiple infrastructural systems due to physical proximity. For instance, the collapse of buildings in city centers can induce the blockage of adjacent roads due the debris accumulation.

A comprehensive methodological framework for the assessment of physical as well as socioeconomic seismic vulnerability and risk of buildings, lifelines, and infrastructures at urban and regional level considering inter-element and intra-systems dependencies has been developed in SYNER-G project. The reader is referred to Pitilakis et al. (2014b) for more details and applications of the proposed framework.

Seismic Risk Assessment of Lifelines: Examples

Representative examples of seismic risk assessment studies for lifelines are given in the following, for different scales of analysis.

Medium- to High-Voltage Electric Power Network of the Sicily Region, Southern Italy (Regional Scale)

A power flow analysis is performed for the electric power network or Sicily, which is composed of 181 nodes and 220 transmission lines (Cavalieri et al. 2014). The nodes, i.e., the buses, are subdivided into 175 demand or load nodes and six supply nodes, five of which are power plants and one is the balance node (or slack bus), which is coinciding with the generation node providing the highest power. The load nodes (two for transmission/distribution and one for distribution substations) deliver power to users. In total, 390 municipalities are served by the network. All transmission lines are overhead lines and considered as non-vulnerable elements. They are classified into high- (HV), medium- (MV), and low-voltage (LV) lines (Fig. 16a). The vulnerable elements are the components within substations, called micro-components. A probabilistic evaluation of the performance of network is carried out by means of Monte Carlo simulation by sampling seismic events for 18 faults taken from the Italian DISS database employing the truncated Gutenberg and Richter recurrence model for the source activity.
Fig. 16

Seismic risk assessment of electric power network in Sicily, Italy (Source: Cavalieri et al. 2013). (a) Transmission lines, classified by voltage. (b) Contour map of expected values of voltage ratio (VR). (c) Mean annual frequency (λ) curve for electric power system serviceability index (SSI)

The distribution of performance losses is shown in Fig. 16c as the mean annual frequency (MAF) of exceedance for the system serviceability index (SSI). SSI provides a global scalar measure of the system performance. It is defined as the ratio of the sum of the real power delivered from load buses after an earthquake to that before the earthquake. For each bus inside the substations, voltage ratio (VR) is defined as the ratio of the voltage magnitude in the seismically damaged network to the reference value for non-seismic, normal conditions. Figure 16b displays a contour map of the expected values of VR, averaged on the whole simulation for each demand node. It can be seen that the reduction in voltage due to induced damage is less than the tolerated threshold of 10 %, allowing the power demand delivery everywhere in the island.

Water Network of Thessaloniki, Greece (City Level)

A connectivity analysis is performed for the main water system (WSS) of Thessaloniki in Greece considering the interaction of electric power network (EPN) with pumping stations (Pitilakis et al. 2014b). If a pump serving a source node is not fed by the reference EPN node due to damage in EPN substation, then the pump itself is considered out of service and the relative WSS node is removed from the system. A probabilistic evaluation of the performance of networks is carried out by means of Monte Carlo simulation by sampling seismic events for five seismic zones with Mmin = 5.5 and Mmax = 7.5. Pipeline damage is evaluated for each simulation considering both wave propagation and ground failure due to liquefaction. The network is analyzed for each sampled event and the results are aggregated all over the sampled events, in order to numerically obtain the marginal distribution of performance losses (Fig. 17). The interaction can be important; as an example the water connectivity loss is increased from 1 to 1.8 % for λ = 0.001 (corresponding to mean return period T = 1,000 years) when the connections of water pumping stations to EPN are included in the analysis.
Fig. 17

Mean annual frequency (λ) curve for water connectivity loss with and without interaction with electric power network (EPN)

Figure 18 shows the level of correlation between the water connectivity loss and damages in pipes as well as the nonfunctional EPN substations supplying the water pumping stations. The most correlated pipes are concentrated along the coastline where liquefaction susceptibility is high and therefore damages due to permanent ground displacement are expected. A higher level of correlation is obtained for the EPN transmission substations. The highest value of 80 % is attributed to a component in the southeast part of the city, where several pumping stations (connected to EPN) are located.
Fig. 18

Correlation of damaged pipes and nonfunctional EPN transmission stations to water network connectivity

Road Network of Thessaloniki, Greece (City Level)

The main network of the urban area is considered in this case study, together with the ring road and the main exits of the city where the majority of bridges and overpasses are located. In particular, 594 nodes and 674 edges are included in the simulation. The nodes are subdivided into 15 external nodes, 127 traffic analysis zone (TAZ) centroids, and 452 simple intersections. Edges are assumed to be the only vulnerable components in the network. They are classified into road pavements and bridges, with fragility models expressed in terms of permanent ground deformation (PGD) due to liquefaction and peak ground acceleration (PGA) for ground shaking, respectively (Pitilakis et al. 2014b). Road closures are estimated due to soil liquefaction and bridge damages. Moreover, the interaction with collapsed buildings that can induce road blockages is considered (Pitilakis et al. 2014b). A probabilistic evaluation of the network’s performance is carried similarly to the one described in the previous example.

The interaction with building collapses can be important especially for return periods higher than 500 years (λ = 0.002). As an example the connectivity loss is increased from 20 % to 33 % for λ = 0.001 (T = 1,000 years) when the building collapses are included in the analysis (Fig. 19). Figures 20 and 21 show the level of correlation between the connectivity loss and the distribution of blockages due to building collapses and damage in bridges and road pavements, respectively. Relatively higher correlation factors are found for edges blocked by building collapse, demonstrating the importance of this failure mechanism in the analysis. A few road segments near the coastline which are subjected to ground failure due to liquefaction are also highly correlated to the network connectivity. The high risk of failure for bridges is attributed to their typology characteristics (old, simple span bridges) and the high values of PGA.
Fig. 19

Mean annual frequency (λ) curve for road network connectivity loss with and without interaction with collapsed buildings

Fig. 20

Correlation of edges blocked by buildings’ collapse to road network connectivity

Fig. 21

Correlation of broken edges (bridges due to ground shaking or road segments due to liquefaction) to road network connectivity

Port System of Thessaloniki, Greece (Infrastructure Level)

The port covers an area of 1,550,000 m2 and trades approximately 16,000,000 t of cargo annually, having a capacity of 370,000 containers and six piers with 6,500 m length. In this case study, waterfront structures, cargo handling equipment, power supply system, roadway system, and buildings are examined. In particular, waterfront structures of a total 6.5 km length, 48 crane nodes, and two terminals (one container and one bulk cargo) are considered. The interactions accounted for in the analysis are the supply of EPN to cranes and the road closures due to building collapses. A probabilistic evaluation of the performance of networks is carried out by means of Monte Carlo simulation by sampling seismic events for five seismic zones affecting the city of Thessaloniki and the harbor with Mmin = 5.5 and Mmax = 7.5. The performance of the port is described through the total cargo or containers handled in a predefined time frame per terminal and for the whole port system (Pitilakis et al. 2014b).

Figures 22 and 23 show the level of correlation between the total cargo handled per day (TCaH) and the distribution of damages in cranes and non-functionality of electric power distribution substations, respectively. In this way the most critical components can be defined in relation with their contribution to the performance loss of the system. All cranes have medium (40–70 %) to high (over 70 %) levels of correlation, indicating their great importance to the functionality of the overall port system. A higher level of correlation is estimated for the EPN distribution substations, with 40 % of the components having values greater than 70 %.
Fig. 22

Correlation of damaged cranes to port performance (PI = TCaH)

Fig. 23

Correlation of nonfunctional electric power distribution substations to port performance (PI = TCaH)


Lifelines are spatially distributed systems that provide essential services to any modern society. They play also an important role for emergency response and restoration in the aftermath of disastrous earthquakes; in general they are vital for the resilience of society. They are often grouped to transportation and utility systems and comprise multiple components, which are exposed to different ground shaking effects and geotechnical hazards. Research efforts and studies undertaken the last 20 years, and after numerous devastating earthquakes that caused extensive losses and disruptions to lifelines and infrastructures, contributed to important improvement of knowledge and expertise in the seismic vulnerability and risk assessment of lifelines. The main objective of these efforts has been the development of methodologies and tools for the estimation of probable direct and indirect losses (physical, performance, economic, social) due to future earthquakes in order to develop efficient emergency response and mitigation strategies. A general layout for the seismic risk assessment of lifelines is outlined in Fig. 24. Inventory and taxonomy/typology definitions of all elements at risk and systems are the necessary initial steps to describe the elements and networks exposed to seismic risk. The estimation of site-specific seismic ground motion and other geotechnical hazards and the selection–estimation of appropriate seismic intensity measures (IM) is the basis of the seismic hazard analysis. Vulnerability is the expected response of an element or a network to a given seismic intensity. It is commonly assessed based on fragility models, which estimate the probability of exceeding certain damage states for given seismic intensity. Fragility functions are adapted with respect to the typological characteristics of each element at risk. The expected damage and loss of the entire network are estimated through appropriate systemic analysis, which takes into account the physical damage and the relations and interactions between the different components and systems. The interdependencies among different systems may considerably increase the overall impact. The results, which are commonly provided in terms of performance indicators that describe the network functional losses or the expected economic or socioeconomic losses, provide the means for an efficient mitigation or recovery planning. Obviously, several sources of uncertainties are inherent in the various definitions, models, tools, and methods for estimation of losses, which make the risk assessment of lifelines a complex and challenging topic.
Fig. 24

General layout for the seismic risk assessment of lifelines



  1. American Lifelines Alliance [ALA] (2001) Seismic fragility formulations for water systems. Part 1 – guideline. ASCE-FEMA, Washington, DC, 104 ppGoogle Scholar
  2. Argyroudis S, Kaynia AM (2014) Fragility functions of highway and railway infrastructure. In: Pitilakis K, Crowley H, Kaynia AM (eds) SYNER-G: typology definition and fragility functions for physical elements at seismic risk, vol 27, Geotechnical, geological and earthquake engineering. Springer, Dordrecht. doi:10.1007/978-94-007-7872-6_10Google Scholar
  3. Argyroudis S, Pitilakis K (2012) Seismic fragility curves of shallow tunnels in alluvial deposits. Soil Dyn Earthq Eng 35:1–12CrossRefGoogle Scholar
  4. Basöz N, Kiremidjian AS (1998) Evaluation of bridge damage data from the Loma Prieta and Northridge, California earthquake. Technical report MCEER-98-0004. State University of New York, BuffaloGoogle Scholar
  5. Cardona OD, Ordaz MG, Reinoso E, Yamín LE, Barbar AH (2012) CAPRA – comprehensive approach to probabilistic risk assessment: international initiative for risk management effectiveness. In: Proceedings of the 15th world conference of earthquake engineering, Lisbon, 24–28 Sept 2012Google Scholar
  6. Cavalieri F et al (2013) Application and validation study to an electric power network in Italy. In: Pitilakis K, Argyroudis S (eds) Systemic seismic vulnerability and loss assessment: validation studies SYNER-G. Reference report 6, Publications Office of the European Union, doi: 10.2788/16706Google Scholar
  7. Cavalieri F, Franchin P, Pinto PE (2014) Application to selected transportation and electric networks in Italy. In: Pitilakis K et al (eds) SYNER-G: systemic seismic vulnerability and risk assessment of complex urban, utility, lifeline systems and critical facilities. Methodology and applications. Springer, Dordrecht. ISBN 978-94-017-8834-2Google Scholar
  8. Daniell J, Vervaeck A (2013) CATDAT Damaging earthquakes in 2012 – the year in review. http://earthquake-report.com/2013/01/07/damaging-earthquakes-2012-database-report-the-year-in-review/
  9. Douglas J, Seyedi DM, Ulrich T, Modaressi H, Foerster E, Pitilakis K, Pitilakis D, Karatzetzou A, Gazetas G, Garini E, Loli M (2014) Evaluation of seismic hazard for the assessment of historical elements at risk: description of input and selection of intensity measures. Bull Earthquake Eng. doi:10.1007/s10518-014-9606-0Google Scholar
  10. Dueñas-Osorio L, Craig JI, Goodno BJ (2007) Seismic response of critical interdependent net-works. Earthq Eng Struct 36(2):285–306CrossRefGoogle Scholar
  11. Elnashai A, Hampton S, Lee JS, McLaren T, Myers JD, Navarro C, Spencer B, Tolbert N (2008) Architectural overview of MAEviz-HAZTURK. J Earthq Eng 12(S2):92–99. doi:10.1080/13632460802013610CrossRefGoogle Scholar
  12. Faccioli E (ed) (2007) Prediction of ground motion and loss scenarios for selected infrastructure systems in European urban environments: LESSLOSS report No. 2007/08, ISBN: 978-88-6198-012-9. IUSS Press, PaviaGoogle Scholar
  13. Federal Emergency Management Agency [FEMA] (2003) Multi hazard loss estimation methodology: earthquake model – HAZUS-MH MR3 technical manual. FEMA, Washington, DCGoogle Scholar
  14. Kakderi K, Pitilakis K (2010) Seismic analysis and fragility curves of gravity waterfront structures. In: Fifth international conference on recent advances in geotechnical. Earthquak Engineering and Soil Dynamics and Symposium in Honour of Prof. I. M. Idriss, 6.04a, San Diego, CA, May 24–29Google Scholar
  15. Mackie K, Stojadinovic B (2003) Seismic demands for performance-based design of bridges. PEER Report 2003/16. Pacific Earthquake Engineering Research Center, University of California, BerkeleyGoogle Scholar
  16. Modaressi H, Desramaut N, Gehl P (2014) Specification of the vulnerability of physical systems. In: Pitilakis K et al (eds) SYNER-G: systemic seismic vulnerability and risk assessment of complex urban, utility, lifeline systems and critical facilities. Methodology and applications. Springer, Dordrecht. ISBN 978-94-017-8834-2Google Scholar
  17. Mouroux P, Le Brun B (2006) Risk-UE project: an advanced approach to earthquake risk scenarios with application to different European towns. In: Oliveira CS, Roca A, Goula X (eds) Assessing and managing earthquake risk. Springer, Netherlands, pp 479–508. doi:10.1007/978-1-4020-3608-8_23Google Scholar
  18. NCEER (1995) The Hanshin-Awaji earthquake of January 17, 1995: performance of lifelines, technical report NCEER-95-0015 (ed: Shinozuka M), State University of New York, BuffaloGoogle Scholar
  19. O’Rourke TD, Jeon SS, Toprak S, Cubrinovski M, Jung JK (2012) Underground lifeline system performance during the Canterbury earthquake sequence. In: Proceedings of the 15th world conference on earthquake engineering, LisbonGoogle Scholar
  20. Pinto P (2014) Modeling and propagation of uncertainties. In: Pitilakis K, Crowley H, Kaynia AM (eds) SYNER-G: typology definition and fragility functions for physical elements at seismic risk, vol 27, Geotechnical, geological and earthquake engineering. Springer, Dordrecht. doi:10.1007/978-94-007-7872-6_2Google Scholar
  21. Pitilakis K, Crowley E, Kaynia A (eds) (2014a) SYNER-G: typology definition and fragility functions for physical elements at seismic risk, vol 27, Geotechnical, geological and earthquake engineering. Springer, Heidelberg. ISBN 978-94-007-7872-6Google Scholar
  22. Pitilakis K, Franchin P, Khazai B, Wenzel H (eds) (2014b) SYNER-G: systemic seismic vulnerability and risk assessment of complex urban, utility, lifeline systems and critical facilities. Methodology and applications, Geotechnical, geological and earthquake engineering. Springer, Heidelberg. ISBN 978-94-017-8834-2Google Scholar
  23. Rinaldi SM, Peerenboom JP, Kelly TK (2001) Identifying, understanding, and analyzing critical infrastructure interdependencies. IEEE Contr Syst Mag 21(6):11–25CrossRefGoogle Scholar
  24. Satumtira G, Duenas-Osorio L (2010) Synthesis of modeling and simulation methods on critical infrastructure interdependencies research. In: Gopalakrishnan K, Peeta S (eds) Sustainable and resilient critical infrastructure systems. Springer, Berlin/Heidelberg. doi:10.1007/978-3-642-11405-2Google Scholar
  25. Seyedi DM, Gehl P, Douglas J, Davenne L, Mezher N, Ghavamian S (2010) Development of seismic fragility surfaces for reinforced concrete buildings by means of nonlinear time-history analysis. Earthq Eng Struct 39:91–108Google Scholar
  26. Tang C, Zhu J, Qi X (2011) Landslide hazard assessment of the 2008 Wenchuan earthquake: a case study in Beichuan. Can Geotech J 48:128–145CrossRefGoogle Scholar
  27. Weatherill G, Esposito S, Iervolino I, Franchin P, Cavalieri F (2014) Framework for seismic hazard analysis of spatially distributed systems. In: Pitilakis K et al (eds) SYNER-G: systemic seismic vulnerability and risk assessment of complex urban, utility, lifeline systems and critical facilities. Methodology and applications. Springer, Dordrecht. ISBN 978-94-017-8834-2Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2014

Authors and Affiliations

  1. 1.Department of Civil EngineeringAristotle UniversityThessalonikiGreece