Rapid prediction of the transient effect of the initial contaminant condition using a limited number of sensors

Abstract

Rapid prediction of a transient rise in a contaminant concentration attributed to a sudden release of a contaminant is crucial for decision-making in relation to deployment of emergency ventilation. The superposition theorem for a fixed flow field is an alternative approach that has been used to accomplish fast prediction. In this study, the accuracy of a superposition method was investigated based on a novel method to obtain the initial condition using a limited number of sensors. Sensors were placed at the centre of each divided zone to record the initial zonal concentration, and the prediction was performed in accordance to an algebraic expression that uses the measured concentrations and the index of transient accessibility of the sub-initial condition (TASIC) as inputs. The analysis of these numerical cases has led to the following conclusions: (1) the prediction accuracy is improved when the prediction time is longer, and when there are more zones and less nonuniformity for the sub-initial conditions, (2) the relative prediction deviations are larger when the initial condition is measured using sensors rather than those elicited based on the predicted initial conditions, and (3) the use of eight sensors is appropriate for the studied case in accordance to sensor cost and accuracy. The method is helpful for fast decision-making when the initial conditions are unpredictable.