Investigation on Electric Energy Consumption Patterns of Residential Buildings in Four Cities through the Data Mining

Abstract

The government and private sectors in Korea have actively conducted the studies associated with high energy-efficient buildings to enable the zero-energy building (ZEB) targets. Establishing a reference baseline condition is essential to achieve the targets effectively. This study explores electrical energy consumption of residential building types under the city-level condition to identify baseline energy patterns and set up the benchmark dataset. An open-source dataset is used, which is provided by the public housing management information system (i.e., the K-apt website) of the government management office. For this study, five (5) years of electricity usage are reflected for the data-mining process (e.g., removing noise data) and its statistical analysis. This study considers four representative cities for the energy benchmarking analysis, including Seoul, Busan, Gwangju, and Daejeon. Results indicated that the energy distribution patterns of the residential buildings were varied according to the regions, analysis years, and other combined effects such as the range of dataset in the processing. This study concludes that the benchmark dataset will be useful and informative to determine the relative system capacity potential under monthly and yearly trends in the representative locations in Korea.