PSI - Issue 78

Gianrocco Mucedero et al. / Procedia Structural Integrity 78 (2026) 1959–1966

1962

employing the ground motion model by Kotha et al. (2020). Ten return periods were considered: 30 (SLO), 50 (SLD), 475 (SLV), 975 (SLC), 2475, 4950, 9900, 19800, 39600, and 79200 years. For each return period, 20 pairs of ground motion records were selected and scaled using the conditional spectrum (CS) approach (Baker 2011), conditioned on AvgSa, and including the ESHM20 correlation models. The MSA results were used to derive collapse fragility curves for all the case-study buildings, through the maximum likelihood method outlined by Baker (2015). As regards the collapse criteria, an Interstorey Drift Ratio (IDR) threshold of 3.5% was set. The collapse fragility curves are illustrated in Fig. 1(a). An increase in the number of storeys is associated with a corresponding rise in ϑ ; specifically, when increasing from four to eight stories, ϑ increases by approximately 51%. It is noted that, for residential buildings, NTC-18 requires design verification solely for the SLD and SLV limit states, rendering safety checks at the SLC level non-compulsory. The probability of collapse for the 975-year return period (SLC) ranges from 0.42 to 0.11. Following the derivation of the collapse fragility curves, the seismic loss estimation was carried out following the detailed component-based loss assessment methodology suggested in FEMA-P58 (FEMA 2018) and using the PELICUN tool (Zsarnoczay et al. 2024)). A detailed summary of the building’s component inventory, including all the structural/non-structural elements and contents, can be found in (Mucedero et al. 2025b). For each return period, 200 realisations were used. The obtained loss ratio, i.e., the repair costs normalised to the replacement cost, is presented in Fig. 1(b) for the case- study buildings. The annual rates (λ) associated with the return periods foreseen by the Italian code are also indicated. The loss ratios obtained for the two- and four-storey buildings are almost similar, regardless of the return period; the same happens for the six- and eight-storey buildings. The loss ratio values obtained at SLD (50 years), SLV (475 years) and SLC (975 years) are in the range of [2.2-4.7] %, [44.6-56.9] % and [66.6-92.7] %, respectively, with higher values obtained for the four-storey building. The loss ratios obtained for both SLV and SLC are aligned with the those proposed in the Italian guidelines for risk classification of constructions. In particular, for the 475- and 975-year return periods, the suggested loss ratio values are 50% and 80%, respectively. In terms of Expected Annual Losses (EAL), the highest value (EAL=0.92%) is obtained for the four-storey building and, overall, the EALs are in the range of [0.7-0.92] %. It is worth noting that, according to the Sismabonus guidelines (Cosenza et al. 2018), Italian buildings compliant with modern seismic code provisions belong to class (0.50% < ≤ 1.0%) or + ( ≤ 0.50%). As such, the outcomes obtained herein , through a detailed loss assessment methodology, are aligned with those foreseen in Sismabonus, although the latter uses a simplified approach for estimating EAL.

(a) (b) Fig. 1. (a) Collapse fragility curves and (b) expected loss ratio curves.

3.2. Downtime assessment results Fig. 2 illustrates the median building recovery time (RT) to achieve functional recovery (FR), re-occupancy (RO), and shelter-in-place (SiP) states for the considered return periods, as a function of the building height. The return periods corresponding to SLO (30 years), SLD (50 years), SLV (475 years) and SLC (975 years) limit states are also highlighted. For low-intensity earthquakes with return periods of 30 and 50 years, all building configurations show