This paper gives an overview of some main achievements of the NARSIS European project, which aimed at proposing some possible improvements related to the Probabilistic Safety Assessment of nuclear sites, in case of combined and/or cascade natural disasters. The project proposed a multi-hazard as well as a Bayesian Networks-based multi-risk framework, allowing to better: (i) characterize single and combined natural hazards for a given site; (ii) estimate the vulnerability of technical and human systems and related consequences; (iii) evaluate and constrain uncertainties. As part of the effort, many datasets were collected, states of the art produced and methodologies tested for nuclear safety assessment of virtual/real pressurized water reactors, leading to the release of various software tools and recommendations for use. This paper presents a rapid tour illustrated by one of the project applications related to the fragility assessment of a protection dyke under a multi-hazard scenario.

The objective of this study is to establish a method for evaluating the risk of damage to critical infrastructures due to the impact action of falling rocks and rockslides caused by earthquakes. First, based on the results of the existing rolling and free fall experiments on rock models, the time history of the impact force is modelled as the impact action. The characteristic values of the model were quantitatively evaluated based on the uncertainty of the impact force against the momentum of the rock at collision to the structure. Next, the response of the side wall of the critical facility in a nuclear power plant was evaluated by dynamic response analysis using the impact action model, and the fragility curve corresponding to the damage mode was evaluated using the momentum as a strength index. As a result, it is found that the fragility curve can be used to evaluate the influence based on the shape and location information of the target rock because the magnitude of the velocity that affects to the structure is determined from the size of rock.

Retaining walls are crucial elements in infrastructure road and railway networks, as they provide essential support for earth structures, such as embankments and bridge abutments. A proper evaluation of their seismic performance is therefore a fundamental step for assessing the functionality of a transportation network after the occurrence of an earthquake. This study investigates the seismic performance of retaining walls supporting roads in a probabilistic framework. Hundreds of nonlinear dynamic analyses were executed by considering recordings from different soil classes, as well as different wall layouts and critical accelerations. The results were interpreted to develop fragility curves, based on damage scales available in literature. The latter were then exploited to derive new functionality loss functions referring to different road types. The proposed fragility and loss of serviceability curves provide an effective tool for evaluating the probability of failure for different types of retaining walls. Therefore, these curves can be used in risk assessment and related decision-making when the unsatisfactory seismic performance of geotechnical systems may result in a compromised functionality of the overall road network.

Soil liquefaction occurring in urban areas may result in road subsidence, settlement and tilt of buildings, and manhole uplifting displacement after a heavy earthquake. The limitation of using the soil liquefaction potential map shows that it cannot provide quantitative information about the above-mentioned disasters. Using an appropriate risk assessment of failure modes may serve as a better reference for decision-making agencies, instead of relying solely on a soil liquefaction hazard map. This paper aims to focus on the types of damage that can occur on road infrastructure with the goal of developing a risk assessment strategy for emergency routes. Based on different levels of road damage, we determine the vulnerability level by considering the degree of impact on rescue operations when a disaster occurs. Furthermore, the hazard level was evaluated by taking into account both the liquefaction potential index and the groundwater level. Vulnerability and hazard levels were combined into a risk matrix to determine risk level. As an example, we will use a segment of emergency routes in order to explain how the road damage risk level affects the rescue response in the event of a disaster. The proposed risk assessment demonstrates promising potential as an intermediate tool for emergency response in practical cases.

Tunnels are critical infrastructure for the sustainable development of urban areas worldwide, especially for modern metropolises. It is important to evaluate the seismic performance of tunnel under a wide range of earthquake intensities. For this purpose, this study firstly illustrates a Bayesian-based framework for evaluation seismic risk of tunnels, in which the probability values of different fragility curves models are treated as random variables, characterized by their probability density functions was proposed in order to propagate inter-model variability in seismic risk assessments. Then the nominal values of seismic risk for the three main damage states, including minor, moderate and extensive were calculated. Moreover, the effects of the weighting of different fragility functions on the expected seismic risk assessments are discussed. The results show that Bayesian-based framework efficiently projects the whole variability of input models into risk estimations. The findings of this study can serve as the basis for decision-making, seismic risk, and resilience management of city infrastructure.

On February 6th, 2023, at 4:17 a.m. local time, a devastating earthquake struck the territories of southeastern Türkiye and northwestern Syria. The moment magnitude recorded for the first strong quake was 7.7. Subsequently, numerous quakes of moment magnitude greater than 3 were recorded in the already affected areas. A further strong quake of magnitude 7.6 at 13:24 local time resulted in further extensive damage in the eleven provinces of Türkiye already hit. This paper focuses on the assessment of the earthquake-induced damage to the built environment observed in Gölbaşı city, located in the Adıyaman province of Türkiye close to the Gölbaşı Lake. The damage data have been collected through an in-situ reconnaissance performed by the Turkish author 10 days after the main seismic events. Analysis of the strong ground motions recorded in Gölbaşı city has been cross-checked with data on geological and subsoil conditions, also integrated with data reported in other studies, and the observed damage mechanisms. The features of the detected damage highlighted the presence of both inertia-induced effects, due mainly to the building vulnerability, and typical damage patterns related to the earthquake-induced liquefaction of the foundation soils, such as punching failure and consolidation settlements. This preliminary assessment shows the significance of this case for the advancement of the knowledge on earthquake-induced liquefaction and traces the route for deeper future studies.

Surface faulting along active faults poses a threat to the facilities that might be affected. It is especially important for lifelines, trunk pipelines in particular, that, unlike other structures that can be just replaced, often cannot be retraced to avoid crossing with active faults able to produce surface faulting associated with large earthquakes. Pipelines' damage may produce significant economic, social and ecological losses. Pipelines, however, can sustain significant offsets that was demonstrated successfully by the Trans-Alaska oil pipeline during the 2002 Denali Fault earthquake. The critical point is that, estimating the design displacement that might occur during anticipated earthquake, researchers and engineers have to deal with significant uncertainties produced by combination of several factors: (1) significant and hardly predictable slip variability along fault strike; (2) very large, up to 1-2 orders of magnitude, scatter of maximal displacement of surface ruptures with the specified rupture length and during earthquakes with certain magnitude; (3) lack of data on the displacement values consistency at a particular fault section during successive rupturing events. Amount of fieldwork required to get such information is, usually, too large to be performed during site investigations for a particular lifeline project. Moreover, traces of older events are poorly recognizable and can be hardly characterized quantitively in the same way as those of modern surface ruptures. The pipeline design solutions, however, are "discrete" – i.e. they are appropriate for some range of the design displacement values. Stress-strain state of the pipeline strongly depends not only on fault displacement value, but also on fault kinematics, on relative orientation of a pipeline and fault at their crossing point, on the shape of a trench, backfill properties, pipe diameter and material. The preliminary analysis of all these factors and the numerical modeling of the pipeline-fault crossing can provide threshold displacement values which excess requires change of the crossing design. Thus, the main task of site investigation is to find if the anticipated displacement will exceed such threshold or not. If not – there is no need to waste time and efforts trying to specify the exact offset value and work can be concentrated on specifying fault kinematics, exact position of the anticipated fault plane, etc. Close cooperation of geologists and engineers at the earliest stages of pipeline projects implementation will help selection of optimal design solutions and optimization of the site investigations program.