Impacts of diversity in commercial building occupancy profiles on district energy demand and supply

Urban building energy models (UBEM) have the potential to become integral planning tools for district energy systems due to the dynamic, interactive and complex nature of temporal building energy demand patterns. Although the demand patterns are related to the occupancy profiles of buildings supplied by district energy systems, occupant behavior in current UBEM approaches does not usually consider diversity in occupancy profiles among buildings of the same use-type. In this work, a novel method to create context-specific, data-driven commercial building occupancy profiles was used to generate diverse and non-diverse urban building occupant presence models (UBOP). Diverse UBOP randomly assigned occupancy profiles to buildings. Non-diverse UBOP assigned the data-driven mean or median profile to all buildings. ASHRAE standard profiles and occupant densities served as a baseline for comparison. The impact of diverse vs. non-diverse UBOP was assessed by comparing UBEM simulations for district energy efficiency benchmarking, renewable energy integration potential, and district energy system design, using a case study in Singapore. The results demonstrate that, because of the relationship between occupant presence and building systems operation, occupancy profiles are highly sensitive parameters for district energy demand predictions. For the case study, the energy demand estimation is significantly influenced by the shape of occupancy profiles. In particular, the choice of UBOP influences the cooling demand to the degree that district cooling system design decisions might be impacted. Therefore, it is advisable to use diverse UBOP and to run probabilistic UBEM simulations for district energy system design. •Data-driven occupancy profiles are used to simulate district energy demand and supply.•Diversity in building occupancy profiles significantly impacts simulation results.•District energy system design decisions might be impacted by occupancy profiles.