A scalable graph-theoretical optimization model for predictive maintenance of energy transmission networks integrating georeferencial infrastructure and socioeconomic data

Energy efficiency remains a cornerstone of sustainability; however, regional disparities in energy availability present considerable challenges. Optimizing distribution necessitates the deployment of intelligent networks, particularly in the wake of decades characterized by inefficient exploitation of finite resources. Before the implementation of large-scale infrastructure projects, it is imperative to reassess existing energy matrix strategies. Notably, improving the efficiency of current transmission infrastructure may mitigate the need for high-cost alternatives such as new hydroelectric facilities, which are often inaccurately portrayed as universally clean and economically viable. Preserving the integrity of existing transmission systems is therefore essential, as their progressive degradation imposes significant consequences on national fiscal stability and constitutes a substantial and unnecessary social opportunity cost. In response, this study developed a computational model based on classical network optimization algorithms. The model integrates a multi-criteria weighting system to quantify risk and establish maintenance priorities across a representative sample of the Brazilian grid. The analysis encompasses data from 27 federative units, 79 electricity concessionaires and 623 transmission systems, totaling 105346 km, approximately 68.4% of the georeferenced national grid. The model's predictive capabilities proved effective across the dataset, generating actionable insights for infrastructure management and strategic planning.