Development of a CSPF Prediction Model for ≤3RT Air Conditioners Based on Public Data

As global warming intensifies, rising summer temperatures and extended cooling seasons have significantly increased the demand for cooling energy in residential buildings, particularly in multi-family housing. In response, South Korea has promoted Zero Energy Buildings (ZEBs) by enhancing regulations on insulation and window performance. However, the performance of cooling systems is only recommended, rather than strictly regulated, resulting in variations in efficiency and difficulties in quantifying energy savings. Heat pump-type air conditioners, commonly found in ZEBs, exhibit varying performance based on their manufacturing year and capacity. This study utilizes public datasets, including a high-efficiency equipment registry, to analyze the performance trends of electric air conditioners in Korea, focusing on year and capacity. A multiple linear regression model was developed, using Cooling Seasonal Performance Factor (CSPF) and power consumption as key variables. The data were cleaned and categorized into standard and network-connected types. The proposed model exhibited high predictive accuracy, effectively capturing CSPF improvements and efficiency changes based on capacity. This model is anticipated to aid in forecasting the energy-saving potential of high-efficiency cooling systems in ZEBs.