Automatic generation of building stock databases with hourly heat demand profiles for district analysis using large-scale available data

Kurzfassung

The building sector is a major contributor to global energy consumption, accounting for nearly one-third of greenhouse gas emissions. In response, Germany's new energy laws mandate municipalities to develop heating plans for climate neutrality by 2045. However, this process is not just costly, time-consuming, but it also requires both skilled personnel and reliable data on building stock and heat demand. Reliable data is therefore essential for making informed investment decisions in sustainable heating solutions. This dissertation addresses these challenges by developing a data-driven methodology for the generation of high-resolution building stock databases with hourly heat demand profiles, while at the same time shifting from individual building analysis to a broader district-level perspective, enhancing urban planning and integration of renewable energy systems.The primary objective is to automate the determination of heat load profiles with high temporal and spatial resolution using large-scale available data for district analysis. This is achieved through three key components: first, an automated workflow to collect, standardize, and enrich urban data for a consistent building stock inventory; second, a method to generate dynamic heat load profiles based on weather conditions, refurbishment scenarios, and building characteristics; and third, a methodology to describe typical settlement patterns and energy demands in urban districts.The model developed in this dissertation automatically calls and integrates diverse datasets at large regional scales using geospatial, statistical, engineering, and machine learning techniques to enhance the consistency and completeness of the results. These datasets include but are not limited to 3D building models, census data, weather data, building archetypes, official cadastral data and OpenStreetMaps data. This approach enables the creation of standardized building stock databases with heat demand profiles for any region in Germany from 2010 to 2045. Furthermore, a novel methodology for segmenting urban areas into 16 distinct morphological units, known as Urban Energy Units (UEUs) is proposed. The UEU-methodology facilitates the identification of high- and low-heat-demand areas supporting strategic decisions on integrating technologies such as heat pumps and district heating networks.The methodologies developed in this thesis, are designed to efficiently collect and analyze building data across large areas while optimizing computational and financial resources. The model is intended for small municipalities and cities with limited technical resources. The workflow presented in this dissertation was successfully applied in pilot projects across Germany and the results demonstrate its effectiveness in supporting municipal heat plans and large-scale energy analyses. By integrating automated data processing, heat load profiling, and urban classification, this study presents a comprehensive framework for managing the composition of the building stock and forecasting its energy requirements, and therefore, accelerating the heating transformation plans of the country.