Development of a window opening algorithm based on adaptive thermal comfort to predict occupant behavior in Japanese dwellings

Window opening behavior and thermal comfort were monitored in relation to the thermal environment over a 4 year period in the living rooms and the bedrooms of dwellings in the Kanto region of Japan. 36 144 sets of physical and subjective data were collected from 243 residents of 120 dwellings. This paper explores relationships between the different variables in the data. The likelihood of windows being open depended on the three modes of operation of the dwelling, free running (FR), heating (HT) or cooling (CL). In the FR mode, the likelihood was much higher than in either the CL or the HT modes. The likelihood that a window is open correlated well with both indoor temperature and outdoor air temperature in the FR Mode. The indoor comfort temperature correlated well with the running mean of the outdoor temperature. Window opening behavior as predicted by logistic regression analysis is in agreement with the measured data. The deadband of window opening was narrower, and the constraint on window opening was smaller than had previously been found in studies in office buildings. Equations are given to quantify these relations and to enable window opening and comfort temperature to be predicted from outdoor temperature.