Utilizing routine water quality instruments for monitoring distribution system security

Abstract

Drinking water system security concerns have been a considerable issue in the United States in recent years, but in the last two years this issue has risen to new levels of urgency. The tragic events of September 11th highlighted America's vulnerability to terrorism and spurred a domestic security response unprecedented since World War II. Currently, significant purposeful contamination of a water system won't be properly characterized until post-symptomatic epidemiological events are manifested in the affected community. One approach to mitigating this potential disaster includes on-line monitoring of drinking water distribution systems. Four credible threat chemical drinking water contaminants (aldicarb, sodium arsenate, sodium cyanide, and sodium fluoroacetate) were analyzed at different concentrations to determine their detectability in a drinking water distribution system using commonly measured parameters. On-line monitoring that measured conductivity, pH, chlorine residual, turbidity, and total organic carbon was completed to determine baseline water quality indicators. The contaminants were then mixed with tap water and analyzed in beakers to determine their contaminant instrument response at various concentrations with bench top instruments. This data was then used to determine dosing into a bench scale distribution system. The contaminants were then added at the specified concentrations to determine detectability using the online monitoring equipment. Results indicate that the four chemical contaminants can be detected at relatively low concentrations with routine monitoring. Three of the four chemical contaminants can be detected below a concentration that will cause significant health impacts. When these same contaminants were introduced into an acclimated bioreactor, it was determined that toxicity induced sloughing off of biomass increased the turbidity response significantly, effectively lowering the limit of detection for the chemical contaminants. To ensure that the increase in turbidity was a result of toxicity induced sloughing off of biomass, cell counts were conducted on the biofilm containing PVC coupons using fluorescence microscopy. It was shown that indeed cell counts were significantly reduced after introduction of the contaminants. Multivariate cluster analysis of the data also demonstrated a potential to reduce the time to detect a contamination event versus univariate analysis.