A combined workflow to generate citywide building energy demand profiles from low-level datasets

Abstract

Urban building energy models provide an efficient way to estimate the district's energy demand and have been vigorously developed over the years. However, due to the lack of crucial information about building's physical descriptions, existing datasets cannot always meet the significant data demand for modeling. In addition, it is challenging to balance simulation time while ensuring simulation accuracy. Therefore, in order to solve the above difficulties, this paper makes some improvements and optimizations based on reduced-order building modeling and experience-based archetypes and presents a novel two-step approach combining bottom-up and top-down methods. A case study in a suburban district in Germany was chosen as a sample test case to demonstrate these workflow automation tools. The proposed workflow achieved a mean absolute percentage error of less than 2% for citywide energy demand prediction using low-level public data, compared to an error of 15% in the scenario without public data. The results indicate that although there was a lack of specific input data, the presented workflow was capable of obtaining accurate results with reduced computation time. The presented methodology offers an efficient way to establish urban building energy models using low-level datasets and provide relatively accurate results on a district level.