Development, testing and application of a real-time sensing network for monitoring chemical contamination events in a drinking water distribution system
Date
2007
Journal Title
Journal ISSN
Volume Title
Abstract
Routine on-line monitoring which detects the change of water quality by both chemicals and biofilm of drinking water distribution systems offers potential in eliminating the danger of purposeful contamination events.
One of the objectives of this research was to evaluate on-line UV254 as a surrogate for TOC analysis to detect contaminants in the distribution system. UV254 has a lower cost than on-line TOC analysis or grab sampling techniques. Seven contaminants were monitored by measuring common water quality parameters such as conductivity, pH, chlorine residual, turbidity, TOC, and UV254. Results indicate that the seven chemical contaminants can be detected at relatively low concentrations with routine on-line monitoring. Some (aldicarb, sodium fluoroacetate, sodium cyanide) of the seven chemical contaminants can be detected below a concentration that will cause significant health impacts.
Because of the insensitivity of the turbidimeter, the indigenous biofilm in the presence of toxic chemicals may provide an effective, indirect surrogate response with either turbidity or UV254. The hypothesis is that, if toxic chemicals are added to the distribution system, the biofilm would die and slough off to an extent that would change the UV254 absorbance and light scattering of the water so that relatively inexpensive monitors could detect the event. This hypothesis was proven to be correct and turbidity was found to be an excellent indirect monitor for intentional contamination events.
The real-time data collected by the cost effective MSU (Micro Sensor Unit) network can be used to predict the behavior and spread of substances throughout the water system. H2O Map software that can model the hydraulic and chemical characteristics of a distribution system was used to study the CSU facility. Several intentional contamination scenarios were simulated at different points to determine the optimum location and number of MSUs. The results of this modeling showed that there was a probability of detection of greater than 60% at 2 hours and 80% at 4 hours with 5 MSUs in the CSU system.
The results of this research indicate that low cost, reliable, bulk water quality monitors can provide significant chemical detection in a public water distribution system.
One of the objectives of this research was to evaluate on-line UV254 as a surrogate for TOC analysis to detect contaminants in the distribution system. UV254 has a lower cost than on-line TOC analysis or grab sampling techniques. Seven contaminants were monitored by measuring common water quality parameters such as conductivity, pH, chlorine residual, turbidity, TOC, and UV254. Results indicate that the seven chemical contaminants can be detected at relatively low concentrations with routine on-line monitoring. Some (aldicarb, sodium fluoroacetate, sodium cyanide) of the seven chemical contaminants can be detected below a concentration that will cause significant health impacts.
Because of the insensitivity of the turbidimeter, the indigenous biofilm in the presence of toxic chemicals may provide an effective, indirect surrogate response with either turbidity or UV254. The hypothesis is that, if toxic chemicals are added to the distribution system, the biofilm would die and slough off to an extent that would change the UV254 absorbance and light scattering of the water so that relatively inexpensive monitors could detect the event. This hypothesis was proven to be correct and turbidity was found to be an excellent indirect monitor for intentional contamination events.
The real-time data collected by the cost effective MSU (Micro Sensor Unit) network can be used to predict the behavior and spread of substances throughout the water system. H2O Map software that can model the hydraulic and chemical characteristics of a distribution system was used to study the CSU facility. Several intentional contamination scenarios were simulated at different points to determine the optimum location and number of MSUs. The results of this modeling showed that there was a probability of detection of greater than 60% at 2 hours and 80% at 4 hours with 5 MSUs in the CSU system.
The results of this research indicate that low cost, reliable, bulk water quality monitors can provide significant chemical detection in a public water distribution system.
Description
Rights Access
Subject
chemical contamination
distribution system
drinking water
real-time sensing network
water monitoring
civil engineering
sanitation
environmental engineering