Predicting adaptive responses - simulating occupied environments

Simulation of building performance is increasingly being used in design practice to predict comfort of occupants in finished buildings. This is an area of great uncertainty: what actions does a person take when too warm or suffering from glare; how is comfort measured; how do groups of people interact to control environmental conditions, etc? An increasing attention to model these issues is evident in current research. Two issues are covered in this paper: how comfort can be assessed and what actions occupants are likely to make to achieve and maintain a comfortable status. The former issue describes the implementation of existing codes within a computational framework. This is non-trivial as information on local air velocities, radiant temperature and air temperature and relative humidity have to be predicted as they evolve over time in response to changing environmental conditions. This paper also presents a nascent algorithm for modelling occupant behaviour with respect to operable windows. The algorithm is based on results of several field studies which show the influence of internal and external temperatures on decision making in this respect. The derivation and implementation of the algorithm is discussed, highlighting areas where further effort could be of benefit