The Backtracking Uncertainty Bounding Algorithm for Chlorine Sensor Fault Detection

Due to parameter uncertainties, such as errors in water flow estimation and unknown chlorine decay rate, there is typically significant uncertainty in the estimated chlorine concentration in a Drinking Water Distribution Network (DWDN). For certain applications, such as control, monitoring and event detection, it is crucial to have available not only the estimated chlorine concentration but also the possible interval of concentration values. This paper presents a novel methodology, the Backtracking Uncertainty Bounding Algorithm (BUBA), for calculating bounds on chlorine concentration at specific locations in DWDNs. The utility of this algorithm is demonstrated in a chlorine sensor fault detection scheme.