Drinking water can contain low levels of micropollutants, as well as disinfection by-products (DBPs) that form from the reaction of disinfectants with organic and inorganic matter in water. Due to the complex mixture of trace chemicals in drinking water, targeted chemical analysis alone is not sufficient for monitoring. The current study aimed to apply in vitro bioassays indicative of adaptive stress responses to monitor the toxicological profiles and the formation of DBPs in three drinking water distribution systems in France. Bioanalysis was complemented with chemical analysis of forty DBPs. All water samples were active in the oxidative stress response assay, but only after considerable sample enrichment. As both micropollutants in source water and DBPs formed during treatment can contribute to the effect, the bioanalytical equivalent concentration (BEQ) approach was applied for the first time to determine the contribution of DBPs, with DBPs found to contribute between 17 and 58% of the oxidative stress response. Further, the BEQ approach was also used to assess the contribution of volatile DBPs to the observed effect, with detected volatile DBPs found to have only a minor contribution as compared to the measured effects of the non-volatile chemicals enriched by solid-phase extraction. The observed effects in the distribution systems were below any level of concern, quantifiable only at high enrichment and not different from bottled mineral water. Integrating bioanalytical tools and the BEQ mixture model for monitoring drinking water quality is an additional assurance that chemical monitoring is not overlooking any unknown chemicals or transformation products and can help to ensure chemically safe drinking water.

In vitro bioassays indicative of different modes of action demonstrate the high quality of produced drinking water.

Ozonation plays an essential role in water disinfection to inactivate viruses, bacteria and some parasites (Giardia). Ozone treatment rates to attain disinfection goals also result in oxidation reactions of emerging pollutants. Pharmaceuticals–except Ciprofloxacin–are very reactive to ozone: they are removed as early as the transfer compartment outlet even at an ozone treatment rate of less than 1 g/m3. Glyphosate, AMPA, Amitrole and Diuron–the four major pesticides in the Seine, Marne and Oise rivers–are reactive to ozone. Twenty-one pesticides are only partially reactive to ozone and an additional “GAC filtration” is needed to remove them. Further investigations have been planned to study the removal of Phthalates, Nonylphenols and Hormones by combining the “Ozone” and “GAC filtration” process units.

Microplastics (MP) have been detected in almost all matrices, including drinking water, and assessing the contamination of drinking water with this type of pollution is of the utmost sanitary importance. This study aims to evaluate MP contamination of inlet river water and drinking water at three drinking water treatment plants (DWTPs) in the Paris region in France. Each plant performs water treatment processes that are efficient for particulate matter removal such as coagulation-flocculation, sand filtration, and granular activated carbon filtration. One of the plants also has a parallel water treatment file that uses microfiltration and nanofiltration processes. This file was investigated to assess its efficiency compared to the others. To our knowledge, this study is the first to investigate MP contamination in a DWTP using nanofiltration processes. The drinking water distribution network was also investigated, with samples taken at three network points. Microplastics contamination of sizes 25–5,000 μm was characterized using micro-Fourier transform infrared spectroscopy (μ-FTIR) in large volume samples (500 L) with complete mapping of each sample. Concentrations ranging from 7.4 to 45.0 MP/L were found in inlet water while concentrations ranging from blank level (0.003 MP/L) to 0.260 MP/L were found in outlet drinking water (overall removal rate above 99%). Polyethylene, polypropylene, and polyethylene terephthalate were the main polymers found both at the inlet and outlet, but ratios varied significantly at the outlet. No MP were detected in four out of the six samples from the nanofiltration file, and were not found to have significantly different concentrations compared to blank level. Concentrations in the distribution network were higher overall than at the corresponding DWTP outlet, although a high degree of variation between samples was observed. Our results suggest that membrane processes of microfiltration and nanofiltration are more efficient than typical treatment processes, and also that a MP re-contamination within the distribution network itself might occur.