Understanding the Impact of Parametric Uncertainty in Validating Building Thermal Behaviour in Computational Fluid Dynamics (CFD)

Understanding the uncertainty in modelling procedures is important to provide confidence in model predictions, but is rarely carried out. Uncertainty in CFD model parameters resulting from empirical investigation and applying accepted standards are assessed to validate model behaviour. The results show that a basic structured approach to uncertainty considerations (i.e. sensitivity analysis) in a steady-state CFD model result in different conclusions of model behaviour than otherwise achieved by a standard deterministic evaluation.