Analyzing energy signature of building retrofits using public data

Buildings account for a large share of global energy consumption and greenhouse gas emissions, necessitating energy-efficiency retrofits. However, in-depth quantitative and systematic evaluations of the effectiveness of retrofits are required, and robust evidence of actual energy savings and emission reductions is lacking. We thus developed data-driven framework for assessing the performance of the retrofit of 92 public buildings in South Korea. First, we applied a change-point model using monthly energy consumption data from before and after retrofitting to derive energy signatures. Second, we used hierarchical clustering based on these signatures to classify the buildings according to the changes in their energy patterns and to associate clusters with specific retrofit technologies. Finally, we found that retrofitting resulted in an average 23 % reduction in primary energy consumption: 59 buildings achieved a mean energy savings of 34 %, but the energy consumption of 33 buildings increased by 14 %. On-site thermal imaging inspections of the underperforming buildings revealed issues such as envelope heat leakage and insulation defects, highlighting the importance of rigorous quality control during retrofitting. Our methodology provides a precise, scientific basis for evaluating retrofit interventions by integrating statistical modeling, clustering analysis, and field verification. Our method produces results that offer policymakers and practitioners reliable evidence for formulating strategies that maximize energy savings and mitigate greenhouse gas emissions across large building portfolios.