PSI - Issue 78

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

1995

3.2. Direct loss assessment for individual layers According to the proposed framework, direct economic losses for the building stock and the water distribution network are evaluated considering alternative refinement levels of knowledge and analysis. At this stage of the research, only “Level 0” and “Level 1” are presented and discussed. The “Level 0” approach is typically based on a rapid tool for a first estimation (in the form of a range/domain) of economic losses. For the building stock, in the case of basic knowledge, a building-level seismic-risk assessment is performed, considering readily available data for hazard, vulnerability, and exposure models and in line with current practice for large-scale applications. To define the exposure-vulnerability model, a database of fragility relationships has been developed, considering the fragility models adopted in the last National Risk Assessment (Dolce et al., 2021). Then, different building classes are defined based on the construction period (i.e., pre-1970, 1970-2003, post 2003) and the seismic design level (i.e., gravity load design, GLD; earthquake-resistant design, ERD), according to the basic building knowledge at this level of analysis. A set of fragility relationships is thus assigned to each building class, and vulnerability analysis is conducted using the damage-to-loss model adopted by Dolce et al. (2021). Since several fragility models are assigned to each building class, different vulnerability models are also obtained. Vulnerability relationships are finally used to estimate the expected direct losses in scenario-based seismic risk assessment, considering the range of PGA values adopted for the seismic zonation in Italy. As a final step, the effective loss is obtained from the loss ratios, assuming a replacement cost of 1 ’ 350 EUR/sqm (Dolce et al., 2021). A similar approach is also adopted for the WDN. When little information is available (i.e., “Level 0”), the utility network topology is determined according to the most relevant material used for the pipelines, which can be identified by referring to the urbanization year. Two types of material are considered: “brittle” and “ductile” material. Thus, the WDN can be characterized either by the predominance of one specific material or equally distributed materials (Fig. 3a). This way, an innovative methodology/tool for a typology-based seismic vulnerability assessment of WDNs can be developed. Given the material behavior, the repair rate (RR; i.e., the number of repairs needed per unit of length of pipe; O’Rourke et al., 1993) is evaluated according to available formulations in the literature (Fig. 3a). Since these relationships typically depend on PGV values, assumptions are needed to correlate the PGA values to the corresponding PGV ones . Finally, by assuming an average value of intervention (between € 3’450 and € 5’240 for replacement or repair of the pipe, respectively; e.g. Mazumder et al., 2020), the cost of intervention for the WDN can be estimated with the related uncertainties.

(a)

(b)

Fig. 3. (a) Refinement Level 0: simplified methodology for the classification of water distribution network and the assessment of expected repair rate (RR); (b) Refinement Level 1: methodology for building-specific seismic performance, vulnerability, and risk assessment.

Differently, in “Level 1” analysis, it is assumed that the building height and building footprint are collected, thanks to open-source projects (e.g., OpenStreetMaps). These data are deemed sufficient to move to a more detailed