PSI - Issue 78

Rita Couto et al. / Procedia Structural Integrity 78 (2026) 1951–1958 Rita Couto / Structural Integrity Procedia 00 (2025) 000–000

1954

4

mainland Portugal. The updated GEM exposure model for Portugal was used to characterise the exposure of the country. This exposure model combines 2011 and 2021 census data to characterise the residential building stock, including attributes such as material, structural system, floor type, number of storeys, occupancy, and replacement costs, all classified using the GEM taxonomy. Buildings were also assigned seismic design levels and lateral force coefficients (β) based on their construction year and seismic zoning regulations. The exposure data were mapped at the parish level and aggregated by district and NUTS-2 regions for regional risk prioritisation. Several fragility curves, specifically tailored to the Portuguese building stock, have been selected to assess the seismic vulnerability. In this study, the fragility curves developed by Silva et al. (2015a, 2015b), Martins and Silva (2021), and Romão et al. (2021) were selected due to their relevance, coverage of key building attributes, and compatibility with seismic hazard models. These models differ in analytical methods and building representations, so a logic tree approach with equal weighting was adopted to incorporate epistemic uncertainty and ensure a more comprehensive representation of building vulnerability. These fragility models were then converted into vulnerability curves using damage-to-loss relationships, enabling the estimation of the average annual economic losses (AAEL). A fatality vulnerability model (Romão et al. 2021) was employed to quantify the average annual loss of life (AALL). Finally, the seismic risk metrics (AAEL and AALL) were calculated by combining hazard curves for each parish with assigned vulnerability and fatality functions for building classes, averaging results across models with equal weights. These metrics were aggregated to calculate the seismic risk indicator (I S ), which is presented in Fig. 2(a). The results reveal higher losses in southern Portugal and the Lower Tagus Valley, with Lisbon, Faro, Setúbal, and Santarém districts showing the greatest AAEL due to high seismic hazard and building density. 3.2. Energy performance metrics and indicator (I E ) The energy performance assessment is based on heating (HDD) and cooling (CDD) degree-days, as well as annual electricity and gas consumption. HDD and CDD represent the energy required to maintain indoor comfort relative to an outdoor baseline temperature (typically between 18°C and 25°C). The mean HDD and CDD patterns for the period 1979-2023 were sourced from (Eurostat 2024). The annual electricity and gas consumption data by municipality and domestic sector in Portugal were retrieved from DGEG, (2022), while annual CO 2 emissions were estimated by applying a fixed greenhouse gas emission factor (APA 2024) and carbon emission cost (Rennert et al. 2022), resulting in emission patterns that closely mirror the energy consumption distribution across districts. The energy-performance indicator (I E ) was obtained by combining these data and using the equations listed in Table 1. The results, depicted in Fig. 2(b), reveal marked regional disparities: Lisbon emerges as the most critical district, followed by Porto, Setúbal, Faro, Braga, and Aveiro. In contrast, Guarda ranks lowest, preceded by Beja, Viseu, Castelo Branco, and Vila Real. Despite these variations, the I E values show a relatively narrow distribution within the [0,1] range, with scores across districts being generally low and more homogeneous. This is primarily due to Lisbon’s exceptionally high score, which skews the distribution under the min-max normalisation method. Overall, districts in the eastern part of Portugal tend to exhibit lower priority in terms of energy performance needs, while western coastal regions present greater challenges. 3.3. Socioeconomic vulnerability metrics and indicator (I G,SVI ) Socioeconomic vulnerability significantly influences how communities experience and recover from disasters, prompting the inclusion of equity-focused metrics. For this reason, three metrics to perform the regional socioeconomic assessment of mainland Portugal were adopted: the Socioeconomic Vulnerability Index (SVI), energy poverty and earthquake risk awareness (ERA). SVI combines two key indicators of regional development and social well-being: the Human Development Index (HDI) and the EU Regional Social Progress Index (EU-SPI). The HDI captures core aspects of human development (health, education, and standard of living) while the EU-SPI offers a broader perspective on social and environmental progress, including access to basic services, energy, education, and sustainability. Energy poverty refers to the inability of households to access essential energy services, which affects quality of life, health, and social equity. This issue is pronounced in Portugal, where approximately 20% of the population is