| Title |
Impact Analysis in the Residential Energy Benchmark Data by Net Usable Area and Household Size |
| Authors |
박수환(Park, Soo-Hwan) ; 김혜진(Kim, Hye-Jin) ; 서동현(Seo, Dong-Hyun) |
| DOI |
https://doi.org/10.5659/JAIK.2025.41.7.343 |
| Keywords |
Building energy consumption; Prototypical residential building; Household energy panel survey (HEPS); Net Usable Area (NUA);Household size; Energy benchmark; Energy use intensity (EUI); Linear regression |
| Abstract |
In South Korea, the energy performance of residential buildings is primarily evaluated through the Energy Efficiency Rating System.
However, the current system is designed mainly for new buildings and relies on design-based simulation data, making it difficult to apply to
existing buildings where actual total energy consumption data must be used. Furthermore, the system does not adequately account for detailed
consumption characteristics such as housing type, climate zone, floor area, and household size. To address these limitations, previous studies
have proposed group-average-based benchmark data, but these approaches also fail to capture the continuous variations in energy use resulting
from differences in floor area and household size. This study aims to overcome these limitations by utilizing microdata from Korea’s
Household Energy Panel Survey (HEPS) to quantitatively analyze how Net Usable Area (NUA) and the number of occupants affect residential
energy consumption. Based on this analysis, a regression-based benchmark adjustment model is developed. Among three evaluated regression
models, the model combining linear regression for NUA with an independent adjustment coefficient for household size was selected as the
most appropriate in terms of accuracy, interpretability, and usability. The results show that as NUA increases, the annual Energy Use
Intensity (EUI) decreases, leading to a variation range of approximately ±30% compared to the conventional group-average benchmarks.
Additionally, each additional household member increases electricity consumption by approximately 5?7% and fuel consumption by 8?10%.
The final benchmark range defined by the proposed model includes 49.7% of the total sample, while the remaining 50.3% are identified as
households with atypical energy consumption, characterized by either overconsumption or energy-saving behavior. The findings of this study
offer a practical alternative that complements existing benchmarks and supports more precise assessments of residential energy performance. |