Buildings represent a significant share of global energy consumption, positioning them at the forefront of worldwide decarbonization efforts. While Building Energy Modelling (BEM) enables highly detailed and accurate simulations for individual buildings, it becomes computationally prohibitive and data-intensive at urban scales. Urban Building Energy Modelling (UBEM) enables scalable city- and district-scale energy analyses through model simplifications, yet the myriad of methodological simplifications inevitably introduce prediction errors. As UBEM is increasingly adopted for building decarbonization policies and interventions, it is critical to evaluate the accuracy loss during the transition from detailed BEM to more abstract UBEM approaches.
To address this gap, this study systematically quantifies the accuracy loss arising from two key simplification strategies (viz. thermal zoning and template assignment) through a series of simulation experiments with measured validation data. Four model variants are developed for individual buildings, reflecting graduated levels of internal zoning detail: detailed room-level zones, HVAC/non-HVAC partitioning, core/perimeter modelling according to solar exposure (based on ASHRAE standards), and a single-zone-per-floor approach (utilized in most UBEM studies). For each zoning variant, the impact of template assignment granularity is evaluated by comparing differentiated templates assigned per zone (e.g., per room, per functional zone, or per floor) to unified templates assigned per building. Simulated results are validated against measured energy data from diverse buildings covering various functional use types.
The study further examines how prediction errors accumulate as the number of buildings increases, providing a critical estimate of accuracy loss when scaling from individual buildings to urban districts. These insights enable researchers and practitioners to select appropriate zoning and template assignment strategies tailored to the scope and objectives of UBEM applications, enhancing UBEM reliability in building retrofitting strategies and urban decarbonization policies. This analysis not only clarifies the complexity-accuracy trade-offs but also lays the foundation for developing robust, physics-informed UBEM methodologies.