This paper aims to solve three issues frequently present in the optimal placement of water quality sensors for protecting water distribution systems (WDSs) from both accidental and intentional contamination, namely i) computational intractability of the optimization problem as the size of the WDS increases, ii) unrealistic assumption that sensors are positioned at nodes, rather than on system pipes, and iii) neglection of site-specific practical conditions. The three drawbacks were tackled by i) restraining the optimization to the hydraulic/topological-wise most important pipes, ii) introducing dummy nodes in the middle of these pipes as potential sensor locations, iii) applying a multi-criteria decision-making tool incorporating urbanistic and economic factors for selecting the most effective sensor locations. The method is tested on the WDS of the town of Parete (Italy), showing the manyfold benefits of the solution obtained.