Passive sampling for monitoring the removal of organic micropollutants by different wastewater treatment processes

• Passive Probenahme zur Überwachung der Entfernung von organischen Mikroschadstoffen durch unterschiedliche Abwasserreinigungsverfahren

Kämpfer, David; Schäffer, Andreas (Thesis advisor); Hollert, Henner (Thesis advisor)

Aachen : RWTH Aachen University (2022)
Dissertation / PhD Thesis

Dissertation, RWTH Aachen University, 2022

Abstract

Even though water can be found nearly everywhere, clean freshwater is a scarce and valuable resource. Through the increased use of synthetic organic chemicals in our everyday lives, for industrial processes, for agricultural purposes etc., freshwater resources are often polluted with these substances. To prevent their entry into the environment, suitable wastewater treatment technologies are necessary. Furthermore, the efficiency of the different treatment steps needs to be monitored which can be challenging due to the variability in the organic micropollutant (OMP) concentrations and their low levels. In this thesis, analytical methods for measuring the concentrations of 49 OMPs in water samples were established. These were then used in conjunction with grab and passive sampling to monitor the removal of these substances by different technologies in a modular wastewater treatment plant (WWTP). Passive sampling was selected to provide time-integrative information about the OMP concentrations, while the grab samples were used as a general verification of the trends in the water concentrations. Despite the semi-quantitative nature of the polar organic chemical integrative sampler (POCIS) applied for deriving dissolved concentrations due to the problem of defining accurate field sampling rates, POCIS was an effective tool to assess the relative reduction of OMPs between the inflow and outflow of different treatment steps since the hydrodynamic conditions were similar. Over the duration of about one and a half years, from the initial phase through to the routine operation, the performance of the pilot-scale modular WWTP for removing OMPs was investigated. The modular WWTP comprised a primary treatment step with an anaerobic baffled reactor, a moving bed biofilm reactor, and a shaking bed filter, while the secondary treatment step included a retention soil filter (RSF), a UV-light treatment, and ultrafiltration (UF) which were operated in parallel. The OMP removal efficiencies of the complete primary treatment step, as well as of the different secondary treatment steps, were investigated. Additionally, further treatment of the eluent coming from the RSF or UF using low-pressure reverse osmosis (RO) was studied. Out of the 49 target OMPs, up to 21 were routinely detected during four monitoring campaigns using grab and passive sampling. More easily degradable substances such as acetaminophen and metformin showed consistently high removal efficiencies (>70%) during the primary treatment. In addition, the typically problematic X-ray contrast agent iomeprol, and the antibiotic trimethoprim were well removed during the later campaigns by the RSF. In general, the primary treatment and the RSF were able to effectively remove some OMPs from the wastewater showing the potential of particularly the RSF as a low-tech barrier against the entrance of OMPs in the environment. In contrast to these biological treatment steps, the UF showed a varying removal of a broader set of OMPs over the study treatment. Furthermore, directly after chemical cleaning of the UF membrane, higher removal rates were observed for many OMPs. This could be due to its increased sorption capacity directly after the cleaning. In the third and fourth campaign, however, the observed retentions by UF were mostly in a low to medium range for the OMPs and a high variability between compounds and between these campaigns was observed. This shows that the UF module was not suitable to reliably remove OMPs under standard operating conditions. In contrast, the RO treatment resulted in a good removal (>80%) for almost all of the measured OMPs. Only 1H-benzotriazole, a small hydrophilic molecule, showed a higher passage over the RO membrane with removal rates of less than 50% during normal operation. In conclusion, only the RO treatment showed a consistent and high removal of almost all measured OMPs over the complete study period. The POCIS applied for the monitoring above has also become increasingly popular for monitoring a wide range of OMPs in aquatic environments. To reduce artefacts due to the polyether sulfone (PES) membranes typically used in current POCIS configurations and simplify the mechanistic description of the OMP uptake in order to derive quantitative dissolved concentrations, stainless steel mesh was tested as an inert membrane barrier (m-POCIS). Sampling rates (RS) through the water boundary layer (WBL) of the metal membrane were determined for a range of OMPs. The uptake kinetics of the m-POCIS were first measured in a laboratory calibration experiment. They were then deployed in the field in the effluent of the RSF. Under laboratory conditions, the m-POCIS showed measured RS in the range of around 9.502 ± 1.250 mL/(d*cm2) which are at the upper range of values reported for POCIS using PES membranes. Similar results were achieved by applying a mechanistic approach to describe the uptake process using a single reference compound to calibrate the WBL thickness. Additionally, the RS during the field deployment could be successfully estimated for several OMPs via the same approach. Overall, passive sampling could be successfully applied to evaluate the performance of the different treatment steps. This highlighted in particular the RO for the OMP removal but also showing the potential of the RSF. Additionally, an adapted POCIS using an inert metal membrane was successfully tested which avoids the negative effects of PES membranes.