Water quality sensors are one of the most effective ways to minimize the catastrophic consequences of pollution in water distribution networks (WDNs). The main challenge is arranging sensors properly in the network. In this study, the NSGA-III algorithm is developed to improve the optimal locations of sensors by balancing four conflicting objectives. 1. Detection likelihood, 2. Expected detection time, 3. Detection redundancy, and 4. The affected nodes before detection. The proposed procedure is based on chlorine concentration variation between defined upper and lower limits. The upper and lower bounds of chlorine concentration were determined utilizing the Monte Carlo simulator. To deal with the problem of a large size matrix of possible contaminants a heuristic method was utilized for selecting a representative collection of contaminations with the same characteristics and effects. Importance coefficients were introduced to avoid the same importance of contamination events and network nodes. The proposed simulation-optimization approach was tested on the benchmark and real water networks, then the optimal Pareto fronts were computed for each of the two sets of conflicting objectives.Moreover, the sensitivity analysis related to the number of sensors installed in the networks was conducted for the results obtained from different objective functions. According to the sensitivity analysis, the Pareto fronts became more efficient when the number of sensors increased. Also increasing the number of sensors to more than 10 and 15 in the benchmark and real systems, respectively, will provide little additional detection likelihood.