JAUE2024-068: Refining Residential Energy Baselines through Predictive Degree Day Modeling

Abstract

As global urban energy consumption and carbon emissions continue to rise, the building sector accounts for approximately 34% of total energy use. In response, South Korea's Ministry of Land, Infrastructure, and Transport established a standard baseline for calculating greenhouse gas emissions and reductions in energy efficiency projects. This study focuses on improving the accuracy of this baseline by incorporating heating and cooling degree days (HCDD) as key variables. We developed a model using data from 1999 to 2022 to predict average HCDD for 2023 to 2027, enabling a more precise standard baseline for residential buildings across various climate zones. The study reclassifies the existing four climate zones into five, enhancing the model's ability to predict energy consumption per unit area. Using the Prophet model, which excels in handling time series data, we achieved a high predictive accuracy, as indicated by an average R2value of 0.87. The updated baseline reflects climate variability, providing essential data for sustainable urban development and energy policy formulation. These results will assist in setting realistic energy efficiency improvement targets, ultimately supporting efforts to reduce greenhouse gas emissions in the building sector.