PFAS in Managed Aquifer Recharge - Solution or hidden risk in drinking water production?

Abstract

Managed Aquifer Recharge (MAR) is a globally implemented, cost-effective method to mitigate groundwater depletion while simultaneously meeting the need for higher quantities of groundwater during drinking water production with increasing populations and climate change effects. Through MAR, surface waters from rivers and lakes artificially infiltrate into groundwater systems before being pumped up for further drinking water treatment. However, surface waters contain variable concentrations of the harmful per- and polyfluoroalkyl substances (PFAS). The amphiphilic, persistent, and highly mobile nature of PFAS distinguishes their behavior from that of many traditional persistent organic contaminants, raising concerns that PFAS end up in ground- and drinking water systems during MAR, thus posing a risk for carcinogenicity and infertility to humans. The aim of this thesis was therefore to assess whether MAR acts as a source or sink for PFAS into groundwater systems and eventually drinking water supplies by (1) identifying periods of increased risk of PFAS infiltration by monitoring surface water PFAS variations, (2) understanding PFAS fate and transport in the subsurface at low ambient concentrations, as well as (3) assessing possible optimization strategies to meet upcoming stricter PFAS drinking water guidelines. Therefore, a two-year monitoring campaign at two operating Swedish MAR sites, saturated column experiments, as well as a pilot study of foam fractionation as potential pre-treatment to MAR, were conducted. The key finding of this thesis is that MAR can act both as a source and temporary sink for PFAS. This thesis further shows that other environmental proxies in surface waters cannot predict PFAS variations that are instead highly site-specific and depend on local conditions and PFAS point sources. PFAS transport and retardation within MAR systems are governed by complex interactions between sorption processes, aquifer heterogeneity, and groundwater level fluctuations, which can promote both temporary retardation and later remobilization of PFAS. Here, long-term monitoring of stable water isotope (δ18O) signals infiltrated via MAR provide valuable insight into water flow velocities, mixing processes, and PFAS transport mechanisms. The findings highlight the need for integrated management approaches that combine appropriate (pre-)treatment technologies with strategic MAR design and operation to produce drinking water that meets increasingly stringent current and upcoming PFAS guidelines. The results therefore support the design, optimization, and expansion of MAR systems while minimizing the input of PFAS into artificially recharged groundwater systems and drinking water supplies.