AWWA RSP64407

This slide presentation outlines a study on incorporating uncertainty and variability in the design of microbial monitoring programs. There are multiple objectives for this effort relating to regulatory issues, process control, security issues, and special requirements. The steps involved in the analysis include: define purpose and metrics for monitoring system; design alternative monitoring programs; define the model to be used; define variability/uncertainty distributions; and, use Monte Carlo simulation to test the "value" of each alternative. The presentation applies these steps to an example where the high incidence of positive coliform readings indicated that the primary cause was cross connections, and the objective was to design a monitoring system to better detect the positive coliform readings to locate the cross-connections. Observations included: in a full scale application many more iterations should be performed; uncertainty and variability of more parameters should be considered; use more rigorous methods for defining uncertainty and variability distributions such as research-based, data-driven or expert-panel. Includes tables, figures.