Incomplete mixing models have recently been shown to better represent solute transport at junctions of pressurized water systems, compared to a complete mixing assumption. The present work incorporated an incomplete solute mixing model into a methodology for sensor network design. Water quality simulations conducted using both mixing models were carried out to generate pollution matrices that provided the input data for the set covering optimization formulation. Multiple contamination and detection scenarios were simulated by considering both a minimum hazard level of the contaminant and a maximum volume of contaminated water consumed. Examination and comparison of outcomes demonstrated that the water quality solver used may impact sensor network designs in three ways by altering: (i) the minimum number of monitoring stations required for full detection coverage, (ii) the optimal layout of stations over the water network and (iii) the detection domain of some stations.
Keywords: contamination, optimization, sensor, solute mixing, water system