A set of software tools has been developed to enable performance testing ofcandidate on-line sensors in a utility-specific contaminant warning system designusing simulation methods. Specific sensor performance characteristics such as noise,drift, detection level, and sampling rate can be input into the simulator andcontamination scenarios can be run with a contaminant transport model such asEPANET using utility-specific pipe layouts. The sensor simulator is coupled withevent detection software and together the two modules can provide an indication ofhow well the sensor will perform in a specific utility setting with user-definedcontamination events. Water quality baseline data from utility measurements usingon-line sensors are also incorporated into the analysis for a more realistic simulation.The software enables a determination of the rate of true positive, false positive, truenegative, and false negative events during the simulation. Results from this analysisreveal that on-line residual chlorine and total organic carbon sensors combined withseveral different event detection algorithms can provide reliable detection for avariety of chemical contamination events. Study results further indicate thissimulation tool concept is useful for pre-installation evaluation of contaminantwarning system design in a specific utility setting.
展开▼
