Abstract

Meeting climate targets requires sustainable energy systems and proactive municipal heat planning. This paper aims to investigate how different spatial aggregation levels impact the accuracy of heat demand prediction. For this purpose, the energy demand calculated by a state-provided and publicly available urban building energy model is compared with a comprehensive high-resolution consumption dataset from the city of Krefeld in Germany. The latter includes detailed consumption data for district heating, natural gas, heat pumps, and night storage heaters. Box plot diagrams and statistical performance indicators are applied to evaluate the precision of various spatial aggregation levels. The results demonstrate that aggregations at the level of postcode areas, statistical districts and cadastral sectors can provide a reliable foundation for planning purposes. Aggregation at the level of building blocks and heat lines provides an improvement compared to individual parcels, but it should be applied in planning practice with due consideration of the remaining uncertainty. This paper also examines the model deviation between simulated and measured data based on building age class, building type and heat carrier. Furthermore, this paper identifies an underestimation by the state-provided Urban Building Energy Model of around 25 %.