Big-data Analysis on Energy Consumption of Office Buildings in Seoul, Korea

Abstract

It is important to understand the characteristics of actual energy consumption of existing buildings, e.g. energy use intensity (EUI, kwh/m2.yr) vs. building age. This paper reports a big-data analysis on the actual energy consumption of 4,625 office buildings in Seoul, South Korea. The data, a total of 226,625 data points, were obtained from the national building energy database released by Korean government in 2015 and include the following information of each individual building: location, age, usage, total floor area, the number of floors, the number of elevators, and monthly energy consumption. In order to deliver an intuitive understanding of the thermal characteristics on the large building stock, this paper presents a big-data analysis as follows: (1) energy consumption of buildings vs. their location, (2) energy consumption of buildings vs. building age, and (3) correlations between energy consumption of buildings and other building data. The analyses were visualized with scatterplot matrix and density plot. In the paper, it is shown that building energy consumption, represented as EUI (kwh.m2.yr), was irrelevant to the building age as well as strict prescriptive building energy codes.