A Utility-Aware Knowledge Graph to Support Multi-Scale Residential Energy Modelling

Abstract: Building operations account for major global energy use, yet decarbonisation is hindered by fragmented data silos. This paper presents a utility-aware knowledge graph developed for multi-scale residential energy modelling. By leveraging Linked Data principles and the DDIM server, the framework integrates building-level performance data with national electricity distribution networks. Using Ireland as a case study, the methodology semantically associates 2.3 million residences with over 46,000 low-voltage transformers through a novel enrichment process. This scalable approach enables planners to perform complex queries for grid capacity analysis, bridging the gap between individual building physics and urban-scale energy simulation.