Musser A, Yuill G K
Year:
1999
Bibliographic info:
USA, American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc (ASHRAE), 1999, in: the ASHRAE Transactions CD, proceedings of the 1999 ASHRAE Winter Meeting, held Chicago, USA, January 1999

Residential air infiltration rates predicted by a detailed multizone computational model are compared with those predicted by a single-zone model. The multizone model is created using the public domain program CONTAM96, which allows the user to break the house into a number of Zones connected to one another and the outdoors by leakage paths with user-defined characteristics. Actual floor plans for a ranch-style house and typical published leakage characteristics of residential building components are used to construct a very detailed model with roughly 2,000 zones and 7,000 leakage paths. The leakage path configuration of this multizone model is then validated by performing fan pressurization tests on two houses constructed according to the floor plan used to develop the computational model. At pressure differences typical of infiltration conditions, the leakage of the multizone model is in between that of the two identical houses. Infiltration rates computed by the multizone model for representative outdoor temperatures and wind speeds are then compared to those predicted by the single-zone LBL model. Four ventilation systems are modeled: no mechanical ventilation or exhaust, supply fan only, exhaust fan only, and balanced supply and exhaust fans. Comparisons are initially made based on the single-zone model predictions using typical assumptions. The multizone computational model is then used to calculate more precise wind parameters and building leakage characteristics for use in the single-zone model, and the resulting infiltration is again compared with that predicted by the multizone model. These comparisons show that the predictions of both models are sensitive to the choice of wind-related parameters and that the assumption that leakage is evenly distributed throughout the building envelope has little effect on the predictions of the single-zone model. The predictions of the single-zone model most closely match those of the multizone model when flows are added using a quadrature method that takes into account the flow exponent obtained using the multizone model.