Comparing zonal and CFD models of air flows in large indoor spaces to experimental data

Algebraic multi-zone infiltration models (e.g. COMIS, CONTAM) have been developed to predict air flows and contaminant transport in complex buildings. However these models assume that each building zone is a well-mixed volume. This assumption is not appropriate to model large indoor spaces. We describe two simplified approaches, called zonal methods, to describe air flows in large indoor spaces, intended to provide an improvement over the well-mixed assumption. We compare velocity predictions from different formulations of zonal methods and coarsegrid k—e CFD models, to measurements, in a 2D mechanically-ventilated isothermal room. Our results suggest that coarse-grid CFD is a better simplified method to predict air flows in large indoor spaces coupled to complex multizone buildings, than are the zonal methods when airflow details are required. Based on the comparison of pressure predictions from the different models, we propose a way of coupling a model of detailed airflow in large spaces to an algebraic multi-zone infiltration model.