Stochastic prediction of ventilation system performance.

Traditionally, prediction of ventilation systems performance has been based on deterministic approach, which implies that the spread of the input parameters values is zero. The deterministic approach is valid if the effect of fluctuations in the forcing functions (wind speed and direction, temperature, radiation, occupants' behavior, etc.) is negligible when compared to the mean value. When random fluctuations of the forcing functions are significant, the variance of response can no longer be neglected, as for example, for a hybrid-ventilated building, where a part of or the entire required ventilation load is supplied naturally. Most of the parameters influencing the performance of a hybrid-ventilated building are inherently stochastic in nature. The paper briefly reviews the existing techniques for predicting the airflow rate due to the random nature of forcing functions, e.g. wind speed. The effort is to establish the relationship between the statistics of the output of a system and the statistics of the random input variables and parameters of a model.