Subfloor and house ventilation rates: comparing measured and predicted values.

This paper reports on the use of BRE's domestic ventilation model, BREVENT, to predict subfloor and whole house ventilation rates in a BRE/DoE test house. Before the model could be used though some minor adjustments were necessary because one of its underlying assumptions was that the subfloor temperature was equal to the external temperature. Temperatures measurements over a number of months showed this assumption to be false and so an extra stack term was introduced into the model. However, the overall difference this makes is still quite small, only a few percent at most. The predicted subfloor ventilation rate matched the calculated value well, particularly when it was stack dominated. When wind played a significant part though the level of agreement deteriorated, particularly when subfloor air bricks were located on unsheltered walls. However, both the subfloor and whole house ventilation rate of the test house appears to be heavily influenced by the stack effect because the suspended floor and ceiling are leaky in comparison to the walls. As a result, subfloor ventilation will be stack dominated about 61% of the time. To improve the prediction of wind affected subfloor ventilation better pressure coefficient data is required. In a similar vein BREVENT can predict whole house ventilation rates best when the flow is stack dominated. Analysis of the separate stack and wind effects show that the ventilation in the test house will be stack dominated for about 86% of the time. When wind speed does influence ventilation then wind direction also has an effect: winds blowing from the East and West generally give ventilation rates 25% lower than those blowing from the North and South.