Simulating Intentional Contamination Events in Water Distribution Systems: A Report on the Sensitivity of Estimated Impacts to Major Simulation Parameters

Abstract

We report on results of recent EPA/Argonne studies that examine potential public health impacts associated with hypothetical intentional contamination events in 12 real, diverse water distribution systems. Hydraulic and water quality simulations were performed using each water system’s network model. Impacts are defined as the number of people who receive a contaminant dose above a certain level. We evaluated the equivalent of a wide range of contaminants and determined the sensitivity of impacts to contaminant injection duration, time of injection, injection mass, population distribution, and the model for tap water ingestion. Impacts are examined and ranked (percentile) by their injection location. We also examined the consistency of high percentile injection locations for the different contamination events. The hardening of water systems or the design of contamination warning systems (CWS) requires an understanding of how impacts from contamination events are influenced by the major factors that define the contamination event. We found that contaminant injection time, duration, and mass can have a significant influence on the magnitude of public health impacts. We also found that uncertainties in the population distribution and the model used to estimate dose can be important. For highly toxic contaminants, we found impacts are most sensitive to injection duration; for less toxic contaminants, impacts are most sensitive to injection mass. The overlap of high percentile injection locations decreases with the decreasing toxicity of the contaminant and increasing impact, thereby complicating the hardening of water systems or the design of CWS.