Method to establish time-series building energy data inventory based on frequency for energy-sharing community planning

Abstract

This study aims to establish a foundation for estimating energy demand according to building types by proposing a method to establish a time-series energy data inventory solely based on energy consumption in countries where typical models or benchmarking data by building types have not been developed. First, we presented the characteristics of energy consumption patterns that can be defined from the energy consumption itself based on frequency characteristics. Then, we defined a method to normalize and classify types of daily energy consumption patterns using this method and validated it with measured data. Finally, based on the results, a generalized method for establishing building time series energy data inventory was proposed. This method has a simple procedure and its result can be interpreted intuitively. It does not require data to be converted into stationary time-series data, unlike statistical methods, and it can avoid data distortion due to building operational errors. Moreover, it is easy to find the reasons for the results because the processes are easy to understand, unlike artificial intelligence methods. This study proposes a method to establish a time-series building energy data inventory by classifying types based solely on energy consumption. Furthermore, it suggests the possibility of its application in the early stages of planning an energy-sharing community.