Volume 22: Sustainable Energy Solutions for a Post-COVID Recovery towards a Better Future: Part V

The nonlinear influence of urban and building factors on residential building energy use: An empirical study with quantile regression in Seoul Sujin Lee, Steven Jige Quan

https://doi.org/10.46855/energy-proceedings-9382

Abstract

The relationship between urban and building factors and building energy use is important for low-carbon city development. However, previous empirical studies often encountered the nonlinearity and non-normality issues common in complex urban datasets. This study examines the nonlinear influence of urban and building factors on building energy use under complex distribution conditions using the quantile regression model. This study focuses on Seoul’s residential building electricity use in August 2017 and compares the quantile regression model and the ordinary least squares (OLS) model. The quantile regression results are generally in line with OLS results. However, considering the energy use distributions, the quantile regression results show that nonlinear influences of the urban and building factors on building energy use are strong at the right tail of the energy use distribution. Specifically, the positive relationship between coverage ratio and building energy seems to change rapidly above quantile 80%. The influence of distance to water body on building energy is insignificant in 25% and 50% quantile models, but it turns to a significantly negative effect at quantile 75% and above. Unlike the OLS results, no significant difference between the older adults ratio is found in all quantile models. This study suggests that the quantile regression reveals the nonlinear relationship between urban and building factors and building energy use, providing more detailed evidence for policymaking.

Keywords building electricity use, quantile analysis, building density, urban factors, conditional distribution

Copyright ©
Energy Proceedings