Using response surface regression method to evaluate the influence of window types on ventilation performance of Hong Kong residential buildings

Abstract Natural ventilation and the type of window used are closely related. Many works have already been done on identifying the window types that can enhance ventilation performance in residential buildings. However, window performance has been investigated either independently, without considering the interactive effects of other apertures (relative positions of windows and window orientations) or it has been examined based on very limited wind data sets. To fill this research gap, this study aims to evaluate the influence of window types on indoor ventilation performance of residential units in Hong Kong, taking into account the interactive effect and based on representative wind data sets. Air change per hour (ACH) was used to quantify natural ventilation. In this study, site measurement was conducted at a carefully selected residential unit to provide data for validation of the Computational Fluid Dynamics (CFD) settings. The validated settings were used for further CFD simulations. The Central Composite Design method and the Squeeze Theorem were used to determine the representative wind data sets. The Response Surface Regression method was used to develop a mathematical model for quantifying the interactive influence of different apertures on indoor ventilation and to enable quick estimation of ACH. The results lead to the conclusion that for all apertures, side hung (SH) window is the most effective design, followed in descending order by top hung (TH) window and sliding (SLD) window. The maximum achievable ACH for SH and TH windows were 124% and 97% higher than the SLD window. It was also found that if windows can be located only on the same side of a residential unit, SH windows and south facing TH windows should be preferred.