A Making method of Time-series energy demand data of non-residential buildings for urban energy analysis

Abstract

Urban energy supply systems are changing to distributed energy supply systems according to the spreading of renewable energy. In order to effectively arrange and operate distributed energy supply systems in the city, it is necessary to (1) predict energy demands for each use in each building, (2) consider operation technology such as interchange or storage, and (3) study on a city scale rather than a building scale. Therefore, we aim to construct a method to calculate the optimal energy supply system with renewable energy based on data from geographical information systems (GIS). This paper describes the development of an energy-demand prediction method for non-residential buildings and a demand analysis in Japanese business areas by using this method. For the demand analysis, we developed a method to predict the demand of electricity and heat (heating, cooling and hot water) of non-residential buildings for one year. This program fluctuations in demand by five-minute intervals depending on the type of buildings (office, hospital, hotel, store, restaurant and school), total floor area, outdoor air temperature and so on. The standard demand amount of each type of buildings is based on statistical data and measurement data about energy consumption of non-residential buildings in Japan. Furthermore the fluctuation method of the demand incorporates random number simulation and probability distribution to reproduce an actual fluctuation. We predicted and accumulated the demand for hundreds of buildings by three districts for the demand analysis in the whole district. One of these analyses showed that there are large fluctuations in the demand of each building, and these fluctuations decrease by grouping the buildings in the block. Moreover, we analyzed the gap of peak demand between aggregated individual buildings and districts. This analysis revealed that some of peak demand in districts are less than 40% of aggregated individual.