PSI - Issue 78

Carpanese Pietro et al. / Procedia Structural Integrity 78 (2026) 536–543

542

Then, adding up the individual losses of the 33 components results in the estimated economic loss for the entire building. The INSYDE model calculates the economic losses related to the structural/non-structural components of the building, but does not take into account the losses associated with content damage. The contents economic loss is calculated by PERIL multiplying the value of the contents (150% of the value of the building, as previously presented) by the probability of damage given on the basis of a loss function defined on the water height. As for the seismic economic loss, the results can be expressed in absolute terms [€], as a loss per square meter of building [€/m 2 ], or as a percentage of loss to the total value of the building [%]. Although it is possible for flood events with particular characteristics to lead to structural damage and thus a consequent unusability of the building, more generally floods cause the interruption of commercial and industrial activities because of the time required to restore damage to the building's nonstructural components, the time required for cleanup, and the time needed to complete administrative activities such as inspections and permits. Therefore, PERIL also calculates the business interruption recovery time on the basis of the water height, as proposed by FEMA (2022c). 5. Insurance products The last module of PERIL is built to generate insurance products for insurance companies. This module automates the creation of insurance policies based on risk assessment processes and calculates the capital requirement according to the EU Solvency II directive (European Union 2009). For insurance policies generation, PERIL computes the Average Annual Loss (AAL), which is an indicator used in risk analysis to quantify the average economic loss expected in a year due to catastrophic events. AAL is obtained by adding up all possible economic losses from events with different intensities, weighted by their probability of occurrence over a period of one year. For both seismic and flood hazards, the Average Annual Loss (AAL) is computed through an unconditional analysis, as detailed in §4, which considers all possible events with varying return periods and their associated annual probabilities of occurrence. In the insurance sector, AAL estimates support the determination of appropriate insurance premiums and the assessment of the financial reserves required to cover potential claims. The PERIL insurance module allows for the calculation of the Average Annual Loss (AAL) in both ground up and top down terms, considering in the former all the economic losses, and in the latter the financial parameters (e.g., deductibles), taking into account the "terms and conditions" included in the policy. The AAL can be expressed in absolute terms or as a percentage of the total value of the sum insured. In addition to the AAL calculation, another fundamental parameter to be evaluated is the capital necessary to guarantee solvency (Solvency Capital Requirement, SCR) as defined by the European Union's Solvency II (European Union 2009) regulatory framework. The capital requirement that an insurance company must be available for is calculated considering seismic events with a return time of Tr = 201 years and floods with Medium Probability Hazard (MPH). The conditional risk analysis is carried out for all buildings in the project: by adding up the economic losses of all assets within the project, an estimate of the total capital requirement can be obtained. As with the AAL, the capital requirement parameter can also be calculated in ground up terms (starting from the total economic loss of the buildings) and in top down terms (including the terms and conditions of any policies). 6. Conclusions This work presents PERIL, an innovative insurtech platform designed to address the specific needs of risk assessment and insurance product generation for natural hazards in the Italian context. By integrating updated hazard, vulnerability, and exposure models for seismic and flood events, and leveraging AI technologies such as CNN, PERIL allows for detailed and scalable multi- risk analyses. Unlike traditional “black box” tools, the platform provides transparent, user-friendly workflows, enabling insurance companies and policymakers to perform reliable loss estimations, determine Average Annual Loss (AAL), and comply with capital requirement standards. The integration of AI components, automated input enrichment, and interoperable modules makes PERIL a flexible tool adaptable to different users and territorial contexts. The possibility of customizing vulnerability and damage models further strengthens its interoperability. Overall, PERIL contributes to bridging the gap between