The rejection of emerging trace organics by a variety of commercial reverse osmosis (RO), nanofiltration (NF), and ultra‐low‐pressure RO (ULPRO) membranes was investigated using TFC‐HR, NF‐90, NF‐200, TFC‐SR2, and XLE spiral membrane elements (Koch Membrane Systems, Wilmington, Massachusetts) to simulate operational conditions for drinking‐water treatment and wastewater reclamation. In general, the presence of effluent organic matter (EfOM) improved the rejection of ionic organics by tight NF and RO membranes, as compared to a type‐II water matrix (adjusted by ionic strength and hardness), likely as a result of a decreased negatively charged membrane surface. Rejection of ionic pharmaceutical residues and pesticides exceeded 95% by NF‐90, XLE, and TFC‐HR membranes and was above 89% for the NF‐200 membrane. Hydrophobic nonionic compounds, such as bromoform and chloroform, exhibited a high initial rejection, as a result of both hydrophobic–hydrophobic solute‐membrane interactions and steric exclusion, but rejection decreased significantly after 10 hours of operation because of partitioning of solutes through the membranes. This resulted in a partial removal of disinfection byproducts by the RO membrane TFC‐HR. In a type‐II water matrix, the effect of increasing feed water recoveries on rejection of hydrophilic ionic and nonionic compounds was compound‐dependent and not consistent for different membranes. The presence of EfOM, however, could neutralize the effect of hydrodynamic operating condition on rejection performance. The ULPRO and tight NF membranes were operated at lower feed pressure, as compared to the TFC‐HR, and provided a product water quality similar to a conventional RO membrane, regarding trace organics of interest.

Environmental Context.The contamination of public water supplies by drug residues is an issue of importance for public health. While soil may act as a natural filter to remove some contaminants from groundwater, there is a lack of information on the fate and transport of pharmaceutical residues during groundwater recharge. In this study, the fate and the transport of eight drug residues during groundwater recharge of contaminated surface water which was investigated at an artificial groundwater replenishment plant in Berlin, Germany, as part of the international collaboration NASRI (Natural and Artificial Systems for Recharge and Infiltration). The study shows that some of the contaminants would indeed end up in the water supply if the water only underwent normal groundwater recharge. Abstract.Recently, several new types of organic contaminants such as pharmaceuticals and their metabolites have been found in sewage or surface waters. Some of the polar pharmaceuticals have also been detected in samples of ground and drinking water, especially when water from induced recharge is used for drinking water production. The fate and the transport of eight drug residues during groundwater recharge (GWR) of contaminated surface water was investigated at an artificial groundwater replenishment plant in Berlin, Germany. After a recharge distance of only a few meters, bezafibrate, a blood lipid regulator, and indomethacin, an analgesic, were removed below their detection limits. Clofibric acid, a metabolite of blood lipid lowering agents, and the analgesic drugs diclofenac and propyphenazone were also attenuated during GWR. However, they were still detectable in the receiving water supply wells at low concentration levels (≤ 40 ng L–1). The anti-epileptic drugs carbamazepine and primidone and the drug metabolite AMDOPH (1-acetyl-1-methyl-2-dimethyl-oxamoyl-2-phenylhydrazide) were not significantly affected by GWR occurring in the water supply wells at mean individual concentrations between 100 and 1570 ng L–1.

No abstract