An Empirically-based Assessment of CFD Utility in Urban-level Air Flow Studies

Application of CFD (Computational Fluid Dynamics) toward performance-based building design has a long tradition. Increasingly, CFD tools are also being applied to address large-scale urban climatic questions. For instance, CFD-based analyses have the potential to provide a more detailed representation of the micro-climatic boundary conditions needed for building energy simulation. Likewise, CFD-based methods can be deployed to pursue queries pertaining to the contribution factors to the so-called Urban Heat Island effect. In this context, the reliability, usability, and the predictive performance of such tool represent a number of challenges, implying the need for additional studies of the effectiveness of CFD utilization concerning urban-level queries. In the present paper, we discuss the results of a systematic effort to generate urban segment CFD models, conduct simulations to compute key variables of the air flow field, and compare the simulation results with measurements. Thereby, the influence of a number of factors on the nature and magnitude of the modelling errors (deviation of simulation and measurement results) are explored. These factors include assumptions pertaining to geometry and surface properties, boundary ("inlet") conditions, mesh resolution and selection of the reference weather station.