Effect of parameter uncertainty on water quality predictions in distribution systems-case study

The effect of parameter uncertainty on water quality in a distribution system under steady and unsteady conditions is analyzed using Monte Carlo simulation (MCS). Sources of uncertainties for water quality include decay coefficients, pipe diameter and roughness, and nodal spatial and temporal demands. Results from the system analyzed suggest that water quality estimates are robust to individual parameter estimates but the total effect of multiple parameters can be important. The largest uncertainties occur when flow patterns are altered. The study also provides guidance on difficulties in model calibration. For example, the wall decay had the largest influence on model prediction for the system that was reviewed and is one of the most difficult to measure given its variability between pipes.