On the experimental validation of the infiltration model DOMVENT3D

Buildings represent approximately 40% of global energy demand and heat loss induced by uncontrolled air leakage through the building fabric can represent up to one third of the heating load in a building. This leakage of air at ambient pressure levels, is known as air infiltration and can be measured by tracer gas means, however, the method is disruptive and invasive. Air infiltration models are a non-disruptive way to calculate predictive values for air infiltration in buildings. DOMVENT3D, is an infiltration model that predicts the infiltration rate from air permeability measurements quoted at 50 Pa. The model has been used to calculate the English housing stock infiltration rates, however, it has not been experimentally verified. The objective of this study is to compare DOMVENT3D’s predictions with real measurements. Air permeability measurements using a blower door (50 Pa) and the Pulse method (4 Pa) were carried out in three domestic dwellings in Nottingham, U.K. DOMVENT3D was calibrated to predict infiltration from air permeability measurements taken at 4 Pa. This paper presents a comparison of results for both the blower door and pulse methods predicted values and real infiltration measurements. Initial results show discrepancies (from 34 to 107%) in the predictions between blower door and pulse methods and also with the real infiltration measurements. Nevertheless DOMVENT3D presented results close to the measurements. Further validation is needed before a final conclusion about the model is given.