Evaluation of simplified building energy models for urban-scale energy analysis of buildings

Simplification of building energy models is one of the most common approaches for efficiently estimating the energy performance of buildings over a whole city. In city-scale models, the abstraction of a building into an information model and the division of the model into representative thermal zones cannot be building-specific but must be generic and applicable to many buildings. Considering the limited research on the performance of such methods, in this study, a comprehensive evaluation of the most relevant assumptions on zoning configurations and levels of detail is conducted in three building energy simulation tools IDA ICE, TRNSYS, and EnergyPlus. The findings from the evaluation of zoning configuration on building level and its comparison with the measured energy performance of buildings suggest that a single-zone model of a residential building gives a very similar result to a multi-zone model with one core zone and perimeter zones for every floor of the building. For the single-zone model, IDA ICE overestimates and EnergyPlus underestimates the energy demand compared to the more complex models by approximately the same amount, but EnergyPlus is preferred due to the shorter simulation time. It is also proven that higher levels of detail in building models can increase the accuracy of the results by approximately 6% annually. When extending the scope of the study from building level to district level analysis where a somewhat lesser degree of accuracy can be allowed on the individual building, the simplified models give acceptable results. •Zoning configurations and LoDs of UBEMs were analyzed using common BEM tools.•Results were validated using measured energy performance of buildings.•The BEM tools gave different responses to model simplifications.•Simplified single-zone models deviated negligibly from detailed multi-zone models.