Transport of viruses during bank filtration

Quantitative approaches for optimising the multi-compartment concept and for model-based risk management for bank filtration in hydraulically and hydrogeochemically heterogeneous aquifer systems

image of illustration: Concept of a bank filtration well — Figure 1: Concept of a bank filtration well

Drinking water production using bank filtration is an important method for the sustainable usage of water resources, especially in urbanized areas. For example, bank filtration is used for 50% of the drinking water production of Berlin, and for the city of Düsseldorf it is even 100%.

Bank filtration can be threatened by polluted surface waters. From wastewater treatment plants and surface runoff of agricultural areas, pollutants including pathogens like viruses and bacteria can enter the surface waters.

image of concept for the creation of the tool box for the estimation of virus transport during bank filtration — Figure 2: Concept for the creation of the tool box for the estimation of virus transport during bank filtration

Goal

The goal of this project is to develop a toolbox for waterworks managers for a better prediction of pathogen transport in bank filtration, taking into account the natural attenuation of the subsurface and naturally occurring variable hydraulic and geochemical boundary conditions (Figure 2).

Environmental Relevance

Adenovirus_symbol — Figure 3. Structure of an Adenovirus. Copyright Martin (2012)

While research in the past was mostly focused on pathogenic bacteria, knowledge about viruses in aquatic systems is still limited. From excretions of humans and animals they can get into the wastewater or directly into the environment. Results of previous studies for the natural attenuation of viruses in groundwater are ambiguous (Krauss & Griebler, 2011). Viruses can stay infectious in aquatic environments for more than a hundred days. Additionally, the infectious dose for viruses is generally lower than for pathogenic bacteria (Krauss & Griebler, 2011).

Processes

konzept_borderless — Figure 4: Transport of virus particles (not to scale) along 4 selected streamlines. (a) straining (physical filtration at small pore throats). (b) Sorption after collision with grain surfaces at locally favorable conditions for attachment (e.g. positively charged grain surface). (c) No sorption due to no collision. (d) No sorption after collision under locally unfavorable conditions for attachment (e.g. negatively charged grain surface).

In the environment, different processes lead to the inactivation and elimination of viruses, for example UV-light or increased temperatures (Schijven et al., 1999). An important process for the retention of viruses in sediments is the sorption at charged surfaces (Figure 4). Various studies have shown that sorption processes can depend on water chemistry, water saturation, biological activity, sediment mineral composition or grain size distribution (Xagoraraki et al., 2014). Hydrological events can have a significant influence on the transport of viruses in groundwater, for example flooding or strong precipitation events can change the groundwater chemistry and/or lead to an increased entry of viruses into the groundwater.

Quantitatively, transport of virus can be described using the colloid-filtration-theory (CFT). But CFT has only been completely validated for sediments with homogeneous grains and without significant repulsion between colloids and grains, which is the typical case under natural conditions for pathogens (Hunt & Johnson, 2017; Tufenkji & Elimelech, 2004).

Results

grafik_ufer_rhein — Figure 5. Change of the oxygen concentration through redox reactions (left) along a 1D transect at the waterworks Flehe (red in top right). (bottom right) temperature in the Rhein (black line) and in one of the observation wells (measurements = red/orange lines., model = blue line).

Analysis of samples from a 1-year monitoring at the waterworks Flehe in Düsseldorf showed, that the river Rhein contains significant concentrations of pathogen (indicators), for example up to 10000 MPN/100mL Coliforms and 100 PFU/100mL for Coliphages. But pathogen retention appears to be especially strong in the first part of the bank filtration.

Therefore, in the untreated water of the extraction wells concentrations are in most cases below the detection limit, for example not higher than 10 MPN/100 mL for Coliforms and viruses were never detected. Even though transport times from the river to the extraction wells for conservative species like chloride are comparably low with around 30 days.

Furthermore, a seasonal especially temperature-driven variability was found in the aquifer, which results in nitrate-reducing conditions in the late summer (Figure 5). Modelling of the hydrochemical data has shown, that hotspots for redox reactions (oxygen depletion are found in or close to the colmation layer at the interface between aquifer and river (Figure 5, depletion of oxygen after entry into the subsurface).

Project Coordination

TU Berlin (Dep. Hydrogeology)

Project leader: Prof. Dr. Irina Engelhardt
Project worker: MSc. Dustin Knabe

Project partners:

Stadtwerke Düsseldorf, Abteilung Qualitätsüberwachung Wasser (en.: Dep. Water Quality Control)
Universität Wien, Department of Limnology and Bio-Oceanography (formerly Helmholtz-Zentrum München, Institute for Groundwater Ecology)
VisDat Geodatentechnologie GmbH

Funded by the Deutsche Bundesstiftung Umwelt (DBU)
Project period: 2017 - 2021