ECO62: STREAM – Substance Testing for River-bank filtration And Mobility

Description

River-bank Filtration (RBF) plays an important role for drinking water supply as water treatment technique, which exploits the natural filtration processes of subsurface materials to remove pollution and improve the quality of the extracted water. RBF is a simple, cheap and effective technology that has been widely applied for access to safe drinking water. According to Murshed et al. (2020), 75 % of the drinking water of Berlin depends on RBF. In the Slovak Republic, around 50% of drinking water was supplied by RBF, 45% in Hungary, 25% in Switzerland, 16% in Germany and 5% in Netherlands, and also other countries out of Europe started RBF for water treatment. 

However, certain contaminants, resistant to degradation and highly mobile, may be transported (directly /or indirectly) across the river-bank affecting the drinking water quality. In order to identify such mobile contaminants to protect drinking water sources, PMT (persistent, mobile, toxic) and vPvM (very persistent, very mobile) criteria are under the EU REACH Regulation ((EC) No 1907/2006) (Neumann and Schliebner, 2019). The trigger values for the P and T criterion are already laid down in Annex XIII of the REACH guidance document.  

The proposed measure for the M criterion is based on the organic carbon normalised adsorption coefficient (Koc). 

For the regulatory and modelling requirements usually only Koc is necessary and widely accepted as screening criterion for lipophilic, non-polar substances. However, the use of the Koc approach is too simplistic and does not reflect the complexity of sorption behaviour to soil and sediments (ECETOC TR-139). For example, the Koc does neither account for sorption to clay and mineral surfaces nor for non-linear sorption (Freundlich isotherms) nor for aging processes, slow adsorption to the inner particles, entrapment and irreversible sorption (non-extractable residues NER). Hawthorne et. al (2006) show a high variability of Koc values for polyaromatic hydrocarbons depending on type of sediment and aging, The Koc approach needs to be applied with caution especially for polar (mobile) compounds (Franco et al 2010). 

Another issue with the use of Koc for RBF assessment is the fact, that sediments important for RBF in most cases are very poor of organic carbon and other properties/parameters become more important for mobility considerations. This could be for example electrical interactions with solid surfaces. 

Current status of sorption determination and mobility assessment 

For the experimental determination of the Kd/Koc values, standard guidelines such as OPPTS 835.1110 guidance; OECD TG 106, OECD 312 are used within the given limits. 

For chemicals registered under REACH, only a small fraction report experimentally determined Koc values. This is due to the fact, that experimental Koc determination according to, e.g., OECD 106 works in a certain Koc window only, up to a log Koc of about 4. Where Koc cannot be measured any more, the Koc value is estimated based on retention on HPLC-column material (OECD 121), the Kow, read-across or by QSAR.  

A common lack of all standard tests listed above is their limited use for ionisable substances. Ions are more hydrophilic than neutral chemicals and therefore particular suspect for transport with the water phase. About half of all registered chemicals for REACH are partly or fully ionized (Franco et al. 2010), and an even higher fraction of pharmaceuticals (Manallack 2009). A widely used regression for the Koc of ionizable substances (188 citations) has been provided by Franco and Trapp (2008), but later studies (Franco et al. 2009) could only confirm its applicability at varying pH for weak acids. For cationic substances, Droge and Goss (2013) proposed a cation-exchange model that defines the sorption of organic cations to soil as a summed contribution of sorption to organic matter (OM) and sorption to clay minerals.  

 “Mobility” assessment in the context of RBF 

For the mobility assessment in RBF, the description of adsorption by Koc alone is rather problematic for a number of reasons listed in ECETOC TR-139. Among these reasons are the low organic carbon of aquifer material, different conditions of the sediment (e.g., lower anion and cation content), and the prolonged time for adsorption during bank passage (weeks to months). The mobility assessment for RBF should thus not be described by Koc-values derived for soils, but requires a dedicated newly-designed flow-through adsorption test. 

During RBF, due to the long underground passage, chemicals may simultaneously undergo advection, adsorption and degradation. This is similar to the vadose zone passage, and in silico tools established for leaching from soil and vadose zone transport, such as Pearl or PELMO, might be adapted to the RBF to simulate possible break-through to the drinking water wells.