Spatial analysis of public residential housing's electricity consumption in relation to urban landscape and building characteristics: A case study in Singapore

Abstract

In a highly populated country like Singapore, a significant percentage of our gross annual electricity consumption stems from our domestic electricity usage in our residential houses. Analyzing and understanding factors that could influence such patterns is thus essential in order to derive effective measures to reduce usage. In this research, 16 identified variables were calculated and considered in the spatial analyses based on various buffer sizes. Both multilinear regression (MLR) and geographically weighted regression (GWR) based analyses were conducted using each residential housing's Energy Unit Intensity (EUI) as the dependent variable. The analyzed results have shown that building characteristics variables have more significant influences towards energy consumption patterns as compared to urban landscape variables. Although little difference was observed across different buffer sizes, more reliable results were obtained from a smaller buffer size of 50 m, suggesting its suitability in using these obtained values for further prediction model analysis and development. Results obtained from the GWR-based analysis have shown a significant improvement in the goodness-of-fit value compared to the MLR-based analysis, effectively indicating that GWR performs better in this context, apart from its better explanation on the contribution of these identified variables to the EUI in this case study.