An inquiry into the reliability of window operation models in building performance simulation

Given the impact of inhabitants' control actions on indoor environment and the complex nature of such interactions, sophisticated models of occupants' presence and behavior are increasingly deployed to enhance the reliability of building performance simulations. However, the use of occupant behavior models in building simulation efforts and their predictive performance in different contexts involves potentially detrimental uncertainties. To address this issue, the present study deploys long-term monitored data from an office area and its calibrated simulation model to conduct an external evaluation of a number of stochastic and non-stochastic window operation models in view of their a) potential in predicting occupants' operation of windows, and b) effectiveness to enhance the reliability of building performance simulation efforts. The results suggest that, while stochastic models can emulate the seemingly random character of occupant behavior and provide probabilistic distributions of performance indicators, their use does not guarantee more reliable predictions. Leaving aside the large errors resulted from using such models without the necessary adjustments, stochastic window operation models overestimated the occupants' operation of windows in heating season and thus the annual and peak heating demands. However, as compared with rule-based models, the stochastic models displayed a better performance in predicting window operations and thermal comfort assessment in the free-running season.