Designing and validating a groundwater sampling campaign in an unmonitored aquifer: Patiño aquifer case

Groundwater quality sampling campaigns are crucial for the characterization and monitoring of aquifers, especially in urban settings where contamination have an anthropogenic origin due to common practices developed in urban areas. In Paraguay, the Patiño aquifer is located below the largest and most densely populated urban area, and supplies water to 31% of the country's population (approximately 2.1 million people), it has more than 8,000 potential pollution sources and over 2800 deep wells. Water quality campaigns are performed irregularly, and there is no characterization of water type or water quality for the aquifer. This study presents a novel well selection protocol for the purposes of a water quality campaign and the results of its application. This protocol is based on a multi-objective evolutionary algorithm that uses known risk of contamination data, well location, and accessibility to wells (i.e. public vs. privately owned wells) to maximize the chances of finding groundwater contamination. In total, twenty-one water quality parameters were evaluated in 66 wells that were selected based on our sampling protocol. Of these wells, 83% were found to have values outside the permissible limit, according to the regulations considered. In addition, the presence of nitrate concentrations was found to be above the permissible limits in 42% of the wells. Our well selection protocol had a 50% success rate at finding samples outside the permissible limits, while two previous campaigns with no optimized selection protocol showed a 21 and 37% success rate.