Evaluation of the LBL and AIM-2 air infiltration models on large single zones: Three historical churches

Air infiltration in ancient churches and other historical and monumental buildings is of great importance considering moisture transfer, energy consumption, thermal comfort and air pollutants that induce surface soiling. Two of the most established models for predicting air infiltration rate in buildings are the Lawrence Berkeley Laboratory (LBL) model and the Alberta air Infiltration Model (AIM-2). Being originally developed mainly for dwellings, their applicability to large single zone buildings is evaluated in this study by comparing model predictions with field measurements in three historical stone churches that are naturally ventilated only through infiltration. The somewhat more developed AIM-2 model yielded slightly better predictions than the LBL model. However, an LBL version that allows inclusion of the Neutral Pressure Level (NPL) of the building envelope produced even better predictions and also proved less sensitive to assumptions on air leakage distribution at the building envelopes. All models yielded however significant overpredictions of the air infiltration rate. Since NPL may be difficult to attain in practice, the AIM-2 model was chosen for model modification to improve predictions. Tuning of this model by varying its original coefficients yielded however unrealistic model behaviors and the eventually suggested modification implied introducing a correction factor of 0.8. This reduced the median absolute prediction error from 25% to 11%. Thus, especially when the NPL is not at hand, this modification of the AIM-2 model may suit better for air infiltration assessment of churches and other buildings similar to the tested kind.