PSI - Issue 78

Livio Pedone et al. / Procedia Structural Integrity 78 (2026) 1991–1998

1994

skills, which lead to a higher effort/investment. Finally, the highest knowledge level (i.e., “Level 2”) employs advanced data , including the materials’ mechanical properties and construction details. The collection of these data typically requires tests on material samplings and in-situ inspection, significantly increasing the effort/investment cost. A key feature of the proposed framework relies on the possibility of “mixing” the knowledge levels for different components. Therefore, a more refined level of analysis can be adopted only for strategic structures, while accepting higher uncertainties for non-strategic ones. Moreover, the models can be dynamically updated once more documentation becomes available. More details on the framework can also be found in Matteoni et al. (2024b). 3. Illustrative application 3.1. Description of the case-study urban area and simulated earthquake scenario A case- study urban area is selected for illustrative purposes. For the sake of clarity and simplicity, a “virtual” urban area is defined rather than using a real (i.e., part of the construction environment) demonstrator. The selected urban area consists of 2 main layers: a water distribution system (WDS) serving 12 residential buildings (Fig. 2a). The considered WDS is characterized by an overall extension of 9 km and a global dimension equal to approximately 1x1 km. The water network is composed of 12 nodes and 19 branches, with pipelines of different construction materials and diameters. Differently, the case-study buildings are selected from a neighborhood in Udine, north Italy. The building footprints have been obtained from the OpenStreetMap Database, as in Matteoni et al. (2024a). The structural skeleton consists of RC moment-resisting frames, and the number of stories ranges between 1 and 4. Moreover, available data suggest that the neighborhood was constructed before the 1970s.

(a)

(b)

Fig. 2. (a) Schematic representation of the case study urban area: geometry of the water distribution system and the building stock; (b) simulated earthquake scenario: seismic intensity of nodes in Peak Ground Acceleration (PGA).

To implement the study, a magnitude Mw = 6.15 earthquake scenario is assumed; the epicentral distance from the nodes of the WDS is shown in Fig. 2a. For each node, the seismic intensity is estimated in terms of both peak ground acceleration (PGA) and peak ground velocity (PGV) through the ground-motion prediction equations (GMPEs) proposed by Kawashima et al. (1984) and Yu and Jin (2008). PGA values range between 0.26g and 0.28g (Fig. 2b), while PGV values range between 39 and 43 cm/s. It is worth highlighting that seismic intensities at each node are deemed available only for intermediate-to- refined levels of analysis (i.e., “Level 1” and “Level 2”). Differently, for a basic knowledge level (“Level 0”), the seismic classification in Italy is considered (OPCM 3519, 2006): the entire Italian territory is classified into four “seismic zones”, i.e. from “Zone 1” to “Zone 4”, where a lower number corresponds to higher seismicity. This classification is based on expected PGA values corresponding to an earthquake with 10% probability of exceedance in 50 years, according to the Italian seismic hazard model. Different values of PGA are thus considered for each seismic zone, as representative of the expected intensity of a possible design earthquake and the related dispersion (upper and lower bounds). Since the urban area is assumed to be in a high- seismicity zone (i.e., “Zone 1”), expected PGA values range between 0.25g and 0.35g.