A Multidimensional Analysis of Energy Vulnerability and Gendered Patterns in Seville’s Residential Building Stock

Energy poverty is a multidimensional phenomenon resulting from the interaction of different factors. This study investigated energy vulnerability in Seville by integrating a gender perspective with residential building performance. A total of 530 census tracts and 58,819 energy certificates were analyzed to evaluate indicators such as building efficiency ratings (E-F-G), energy burden, and gendered household structures. Unsupervised machine learning algorithms (K-means++) and geographic information systems (GISs) were used to identify patterns and map the obtained results, respectively. The findings revealed three distinct vulnerability profiles. Cluster 3 demonstrates severe vulnerability, linking low building efficiency with high unemployment and significant energy costs. Conversely, Cluster 1 highlights a specific risk among older women living alone, despite higher average income levels. The results conclude that energy poverty is deeply influenced by the intersection of building characteristics and gender roles. This GIS-enabled classification provides a robust framework for prioritizing residential retrofitting interventions, ensuring that mitigation strategies address both technical deficiencies in the building stock and the social realities of gender-driven vulnerability.