Groundwater is one of the major sources for drinking water supply worldwide. Conventional iron removal via aeration-filtration produces about 72,802 t of iron sludge annually in the Netherlands alone. Iron sludge comprises low-density flocs of little to no commercial value. The current study explored a novel concept for iron removal, namely anoxic iron sulfides formation in a fixed bed continuous flow reactor. Iron sulfides usually form dense structures and offer a wider range of re-use applications. A packed bed up-flow column reactor filled with pyrite granules was fed iron and sulfide containing solutions. Produced solids were analyzed applying X-ray diffraction analysis, Raman spectroscopy, digital microscopy, scanning electron microscopy and energy dispersive X-ray spectroscopy. Rapid iron sulfides formation was observed after < 10 min. The formed minerals were partially retained by the pyrite granules. The molar ratio of removed Fe(II) to removed S(-II) equaled up to 0.76 ± 0.16 mol Fe(II)_rem/(mol S(-II)_rem). Our results show that iron sulfides formation can present an interesting alternative to iron removal via aeration-filtration due to its compact particle sizes and fast formation rates.

To meet the increasing drinking water demand, membrane technologies are used to treat previously unavailable water sources. A byproduct of membrane technologies is the concentrate stream, containing valuable resources in higher concentrations. We studied the recovery of iron from different groundwater matrices and anaerobic reverse osmosis (RO) concentrates via precipitation of vivianite and the co-removal of other common groundwater divalent cations Mn²⁺, Mg²⁺ and Ca²⁺ during vivianite precipitation. The formed precipitates were characterized using X-Ray Diffraction and Scanning Electron Microscopy. Vivianite precipitation removed a maximum of 89 % of Fe²⁺ in raw groundwater and 52 % Fe²⁺ from RO concentrate. Substantial co-removal of Mn²⁺ (max 91 %) and limited co-removal of Mg²⁺ (max 7 %) were found, without hindering Fe removal efficiencies or altering morphological changes of the vivianite crystal. In contrast, co-removal of Ca²⁺ occurred at the expense of iron removal, forming amorphous calcium phosphate precipitates. This study shows the potential of vivianite precipitation for iron recovery across a wide range of groundwater matrices and highlights the need for further research to optimize this novel method to treat concentrate streams that are challenging to dispose of.

Methane removal is an essential step in drinking water production from methane-rich groundwaters. Conventional aeration-based stripping results in significant direct methane emissions, contributing up to one-third of a treatment plant's total carbon footprint. To address this, a full-scale trickling filter was operated for biological methane oxidation upstream of a submerged sand filter, and its performance was compared to a conventional aeration–submerged sand filtration set-up. Full-scale data were combined with ex-situ batch assays and metagenome-resolved metaproteomics to quantify the individual contribution of the main (a)biotic processes and characterize the enriched microbial communities. Both treatment setups fully removed methane, iron, ammonium, and manganese, yet the underlying mechanisms differed significantly. Methane was completely removed from the effluent after trickling filtration, with stripping and biological oxidation each accounting for half of the removal, thereby halving overall methane emissions. Methane-oxidizing bacteria not only outcompeted nitrifiers in the trickling filter, but also likely contributed directly to ammonia oxidation. In contrast to the submerged filter preceded by methane stripping, signatures of biological iron oxidation were almost completely absent in the trickling filter, suggesting that the presence of methane directly or indirectly promotes chemical iron oxidation. All systems had similar ex-situ manganese oxidation capacities, yet removal occurred only in the submerged filters but not the trickling filter. Ultimately, our results demonstrate that trickling filtration is effective in promoting biological methane oxidation at comparable produced drinking water quality, highlighting its potential for advancing sustainable drinking water production.

The mechanisms and by-product formation of electrochemical oxidation (EO) for As(III) oxidation in drinking water treatment using groundwater was investigated. Experiments were carried out using a flowthrough system, with an RuO ₂/IrO ₂ MMO Ti anode electrode, fed with synthetic and natural groundwater containing As(III) concentrations in a range of around 75 and 2 µg/L, respectively. Oxidation was dependent on charge dosage (CD) [C/L] and current density [A/m ²], with the latter showing plateau behaviour for increasing intensity. As(III) concentrations of <0.3 µg/L were obtained, indicating oxidation of 99.9 % of influent As(III). Achieving this required a higher charge dosage for the natural groundwater (>40 C/L) compared to the oxidation in the synthetic water matrix (20 C/L), indicating reaction with natural organic matter or other compounds. As(III) oxidation in groundwater required an energy consumption of 0.09 and 0.21 kWh/m ³, for current densities of 20 and 60 A/m ², respectively. At EO settings relevant for As(III) oxidation, in the 30–100 C/L CD range, the formation of anodic by-products, as trihalomethanes (THMs) (0.11–0.75 µg/L) and bromate (<0.2 µg/L) was investigated. Interestingly, concentrations of the formed by-products did not exceed strictest regulatory standards of 1 µg/L, applicable to Dutch tap water. This study showed the promising perspective of EO as electrochemical advanced oxidation process (eAOP) in drinking water treatment as alternative for the conventional use of strong oxidizing chemicals.

Rapid sand filters (RSF) are an established and widely applied technology for the removal of dissolved iron (Fe²⁺) and ammonium (NH₄⁺) among other contaminants in groundwater treatment. Most often, biological NH₄⁺oxidation is spatially delayed and starts only upon complete Fe²⁺ depletion. However, the mechanism(s) responsible for the inhibition of NH₄⁺oxidation by Fe²⁺ or its oxidation (by)products remains elusive, hindering further process control and optimization. We used batch assays, lab-scale columns, and full-scale filter characterizations to resolve the individual impact of the main Fe²⁺ oxidizing mechanisms and the resulting products on biological NH₄⁺ oxidation. modeling of the obtained datasets allowed to quantitatively assess the hydraulic implications of Fe²⁺ oxidation. Dissolved Fe²⁺ and the reactive oxygen species formed as byproducts during Fe²⁺ oxidation had no direct effect on ammonia oxidation. The Fe³⁺ oxides on the sand grain coating, commonly assumed to be the main cause for inhibited ammonia oxidation, seemed instead to enhance it. modeling allowed to exclude mass transfer limitations induced by accumulation of iron flocs and consequent filter clogging as the cause for delayed ammonia oxidation. We unequivocally identify the inhibition of NH₄⁺oxidizing organisms by the Fe³⁺ flocs generated during Fe²⁺ oxidation as the main cause for the commonly observed spatial delay in ammonia oxidation. The addition of Fe³⁺ flocs inhibited NH₄⁺oxidation both in batch and column tests, and the removal of Fe³⁺ flocs by backwashing completely re-established the NH₄⁺removal capacity, suggesting that the inhibition is reversible. In conclusion, our findings not only identify the iron form that causes the inhibition, albeit the biological mechanism remains to be identified, but also highlight the ecological importance of iron cycling in nitrifying environments.

Iron-based adsorbents are commonly used to remove arsenic (As) from water for drinking water purposes. Here, we study the role of biological As(III) oxidation on iron-based adsorbents in filters and its effect on overall As uptake. A lab-scale filter with iron oxide coated sand (IOCS), a commonly used adsorbent, was operated with water containing As(III) and As(V), while water samples were taken periodically over its height. As(III) oxidation initiated after approximately 10 days and increased to a first order rate constant of 0.09 s⁻¹ after 57 days resulting in full oxidation of As(III) in <50 s. Consequently, the filter shifted from an As(III) to an As(V) adsorbing filter. Oxidation was not observed after inhibiting the microbial activity using sodium azide confirming its biogenic nature. This implies that As(III) oxidizing biomass can grow on iron-based adsorbents in water filters without requiring inoculation. As the experimental conditions were similar to full-scale As treatment plants, we believe that biological As(III) oxidation is widely overlooked in these systems. Occurrence of biological oxidation is, however, beneficial for removal, as at pH <8 the adsorption capacity for As(V) can be up to 10-fold higher than for As(III). With these new insights, arsenic treatment using iron-based adsorbents can be further optimized. We suggest a more robust new design with a biological active As(III) oxidizing top layer and an As(V) adsorbing bottom layer.

Slow sand filters (SSFs) are widely used in drinking water production to improve microbial safety and biological stability of water. Full-scale SSFs are maintained by scraping the biomass-rich top layers of sand. The period of downtime required for filter recovery after scraping is a major challenge due to limited knowledge of the re-stabilisation of purification processes. This study examined the recovery of microbial biomass, and removal of dissolved organic carbon (DOC) and ammonium (NH4+) in water phase and/or on sand along the depth of a scraped full-scale SSF. Scraping reduced microbial biomass on sand in the top layers, while the main prokaryotic taxa remained unaltered. Cellular ATP (cATP) and intact cell counts (ICC) in water sampled from the top layers increased, indicating a temporary disruption in functionality for 37 days. However, stable concentrations of cATP and ICC and similar microbial community composition in the effluent after scraping revealed that deeper layer biofilms offset any scraping effect. Consistent DOC and NH4+ removal after scraping showed that deeper layers effectively performed the role of the top layer. These findings highlight the resilience and robustness of microbial communities in mature full-scale SSFs and their contribution to water treatment efficiency after disturbances caused by scraping.

Gravity-driven sand filters are the dominant groundwater treatment technology for drinking water production. In the past, physicochemical reactions were often assumed to play the main role in the removal of contaminants, but recent breakthroughs showcase the vital role of microorganisms. In this Current Opinion, we thoroughly assess the current understanding of biology in sand filters and explore the potential benefits of shifting toward designs aimed at promoting biological reactions. We highlight the main bottlenecks and propose key areas to be explored toward the next generation of sustainable, resource-efficient groundwater biofilters.

Nitrate leaching from agricultural soils is increasingly found in groundwater, a primary source of drinking water worldwide. This nitrate influx can potentially stimulate the biological oxidation of iron in anoxic groundwater reservoirs. Nitrate-dependent iron-oxidizing (NDFO) bacteria have been extensively studied in laboratory settings, yet their ecophysiology in natural environments remains largely unknown. To this end, we established a pilot-scale filter on nitrate-rich groundwater to elucidate the structure and metabolism of nitrate-reducing iron-oxidizing microbiomes under oligotrophic conditions mimicking natural groundwaters. The enriched community stoichiometrically removed iron and nitrate consistently with the NDFO metabolism. Genome-resolved metagenomics revealed the underlying metabolic network between the dominant iron-dependent denitrifying autotrophs and the less abundant organoheterotrophs. The most abundant genome belonged to a new Candidate order, named Siderophiliales. This new species, “Candidatus Siderophilus nitratireducens,” carries genes central genes to iron oxidation (cytochrome c cyc2), carbon fixation (rbc), and for the sole periplasmic nitrate reductase (nap). Using thermodynamics, we demonstrate that iron oxidation coupled to nap based dissimilatory reduction of nitrate to nitrite is energetically favorable under realistic Fe3+/Fe2+ and NO3−/NO2− concentration ratios. Ultimately, by bridging the gap between laboratory investigations and nitrate real-world conditions, this study provides insights into the intricate interplay between nitrate and iron in groundwater ecosystems, and expands our understanding of NDFOs taxonomic diversity and ecological role.

Organic micropollutants (OMPs) enter the aquatic environment via municipal wastewater treatment plants (WWTPs). As conventional WWTPs have limited capacity for the removal of OMPs, additional processes are required, like ozone - granular activated carbon (GAC) filtration. A specific lay-out of this process is the O3-STEP® process, in which the removal of suspended solids, OMPs, phosphate and nitrate is combined. However, ozonation may result in formation of bromate, a compound with a strict water quality standard of 1 μg/L for surface waters in The Netherlands. This limits the applicability of ozonation in wastewater treatment. This study examined biological bromate removal associated with denitrification processes in the GAC filter of the O3-STEP® process. In this GAC filter methanol is dosed for nitrate removal by biological denitrification. In column experiments, bromate and nitrate were removed simultaneously under both anoxic and oxic conditions. Depletion of oxygen within the biofilm surrounding the GAC granules most probably is the reason for denitrification under oxic bulk conditions, although aerobic denitrification cannot be excluded. In batch experiments, the presence of nitrate did not affect bromate removal, whereas the presence of dissolved oxygen had a slight inhibitory effect on bromate removal and nitrate removal. Addition of methanol increased both nitrate and bromate removal, which is hypothesized to occur through an increased availability of electron donors in the water. The results show that a denitrifying GAC filter in the ozone - GAC filtration process mitigates the bromate formation, which broadens the applicability of this process for OMP removal from wastewater.

Iron (Fe²⁺), manganese (Mn²⁺), and ammonium (NH₄⁺) oxidation processes were studied in three single media and three dual media full-scale rapid sand filters (RSFs) using reactive transport modelling (RTM) in PHREEQC and parameter estimation using PEST. Here, we present the insights gained into the spatial distribution of Fe and Mn mineral coatings in RSFs and its influence on the oxidation sequence and rates. Fe²⁺ and Mn²⁺ oxidation predominantly occurred simultaneously in the RSFs, contrary to the expected sequential oxidation based on Gibbs free energy calculations. During backwashing, RSF grains become fully mixed, which initiates heterogeneous Mn²⁺ oxidation on Mn-coated grains that end up in the top layer. The resulting grains have a mixed Fe/Mn mineral coating, which is limiting heterogeneous Mn²⁺ oxidation due to the limited Mn mineral surface available. Mixed coatings did not seem to affect Fe²⁺ oxidation rates, instead oxidation rates were increasing at lower pH. We found that RSFs can be designed to spatially separate Fe²⁺ and Mn²⁺ oxidation, which results in optimal conditions for Mn²⁺ oxidation. The RSF needs to consist of two layers with varying density to inhibit mixing and complete Fe²⁺ oxidation should occur in the top layer. The developed RTM can be used to estimate the depth at which Fe²⁺ oxidation is complete, and thus the ideal intersection depth of the two layers. A novel perspective is provided on how mineral coating distribution in single and dual media filters influence removal rates and the sequence of oxidation, which contributes to the design of more efficient groundwater filters.

Long-term consumption of groundwater containing elevated levels of arsenic (As) can have severe health consequences, including cancer. To effectively remove As, conventional treatment technologies require expensive chemical oxidants to oxidise neutral arsenite (As(III)) in groundwater to negatively charged arsenate (As(V)), which is more easily removed. Rapid sand filter beds used in conventional aeration-filtration to treat anaerobic groundwater can naturally oxidise As(III) through biological processes but require an additional step to remove the generated As(V), adding complexity and cost. This study introduces a novel approach where As(V), produced through biological As(III) oxidation in a sand filter, is effectively removed within the same filter by embedding and operating an iron electrocoagulation (FeEC) system inside the filter. Operating FeEC within the biological filter achieved higher As(III) removal (81 %) compared to operating FeEC in the filter supernatant (67 %). This performance was similar to an analogous embedded-FeEC system treating As(V)-contaminated water (85 %), confirming the benefits of incorporating FeEC in a biological bed for comparable As(III) and As(V) removal. However, operating FeEC in the sand matrix consumed more energy (14 Wh/m³) compared to FeEC operated in a water matrix (7 Wh/m³). The efficiency of As removal increased and energy requirements decreased in such embedded-FeEC systems by deep-bed infiltration of Fe(III)-precipitates, which can be controlled by adjusting flow rate and pH. This study is one of the first to demonstrate the feasibility of embedding FeEC systems in sand filters for groundwater arsenic removal. Such systems capitalise on biological As(III) oxidation in aeration-filtration, effectively eliminating As(V) within the same setup without the need for chemicals or major modifications.

Rapid sand filters are established and widely applied technologies for groundwater treatment. In these filters, main groundwater contaminants such as iron, manganese, and ammonium are oxidized and removed. Conventionally, intensive aeration is employed to provide oxygen for these redox reactions. While effective, intensive aeration promotes flocculent iron removal, which results in iron oxide flocs that rapidly clog the filter. In this study, we operated two parallel full-scale sand filters at different aeration intensities to resolve the relative contribution of homogeneous, heterogeneous and biological iron removal pathways, and identify their operational controls. Our results show that mild aeration in the LOW filter (5 mg/L O2, pH 6.9) promoted biological iron removal and enabled iron oxidation at twice the rate compared to the intensively aerated HIGH filter (>10 mg/L O2, pH 7.4). Microscopy images showed distinctive twisted stalk-like iron solids, the biosignatures of Gallionella ferruginea, both in the LOW filter sand coatings as well as in its backwash solids. In accordance, 10 times higher DNA copy numbers of G. ferruginea were found in the LOW filter effluent. Clogging by biogenic iron solids was slower than by chemical iron flocs, resulting in lower backwash frequencies and yielding four times more water per run. Ultimately, our results reveal that biological iron oxidation can be actively controlled and favoured over competing physico-chemical routes. The production of more compact and practically valuable iron oxide solids is of outmost interest. We conclude that, although counterintuitive, slowing down iron oxidation in the water before filtration enables rapid iron removal in the biofilter.

Iron (Fe)-based treatment methods are widely applied to remove carcinogenic arsenic (As) from drinking water, but generate toxic As-laden Fe (oxyhydr)oxide waste that has traditionally been ignored for resource recovery by the water sector. However, the European Commission recently classified As as a Critical Raw Material (CRM), thus providing new incentives to re-think As-laden groundwater treatment sludge. Before As recovery techniques can be developed for groundwater treatment waste, detailed information on its structure and composition is essential. To this end, we comprehensively characterized sludge generated from a variety of As-rich groundwater treatment plants in different geographic regions by combining a suite of macroscopic measurements, such as total digestions, leaching tests and BET surface area with molecular-scale solid-phase analysis by Fe and As K-edge X-ray absorption spectroscopy (XAS). We found that the As mass fraction of all samples ranged from ∼200–1200 mg As/kg (dry weight) and the phosphorous (P) content reached ∼0.5–2 mass%. Notably, our results indicated that the influent As level was a poor predictor of the As sludge content, with the highest As mass fractions (940–1200 mg As/kg) measured in sludge generated from treating low groundwater As levels (1.1–22 µg/L). The Fe K-edge XAS data revealed that all samples consisted of nanoscale Fe(III) precipitates with less structural order than ferrihydrite, which is consistent with their high BET surface area (up to >250 m²/g) and large As and P mass fractions. The As K-edge XAS data indicated As was present in all samples predominantly as As(V) bound to Fe(III) precipitates in the binuclear-corner sharing (²C) geometry. Overall, the similar structure and composition of all samples implies that As recovery methods optimized for one type of Fe-based treatment sludge can be applied to many groundwater treatment sludges. Our work provides a critical foundation for further research to develop resource recovery methods for As-rich waste.

Drinking water treatment plants (DWTPs) are designed to remove physical, chemical, and biological contaminants. However, until recently, the role of DWTPs in minimizing the cycling of antibiotic resistance determinants has got limited attention. In particular, the risk of selecting antibiotic-resistant bacteria (ARB) is largely overlooked in chlorine-free DWTPs where biological processes are applied. Here, we combined high-throughput quantitative PCR and metagenomics to analyze the abundance and dynamics of microbial communities, antibiotic resistance genes (ARGs), and mobile genetic elements (MGEs) across the treatment trains of two chlorine-free DWTPs involving dune-based and reservoir-based systems. The microbial diversity of the water increased after all biological unit operations, namely rapid and slow sand filtration (SSF), and granular activated carbon filtration. Both DWTPs reduced the concentration of ARGs and MGEs in the water by circa 2.5 log gene copies mL−1, despite their relative increase in the disinfection sub-units (SSF in dune-based and UV treatment in reservoir-based DWTPs). The total microbial concentration was also reduced (2.5 log units), and none of the DWTPs enriched for bacteria containing genes linked to antibiotic resistance. Our findings highlight the effectiveness of chlorine-free DWTPs in supplying safe drinking water while reducing the concentration of antibiotic resistance determinants. To the best of our knowledge, this is the first study that monitors the presence and dynamics of antibiotic resistance determinants in chlorine-free DWTPs.

Natural organic matter (NOM) is present in water matrix that serves as a drinking water source. This study examined the effect of low and high NOM concentrations on inactivation kinetics of a model RNA virus (MS2) and a model DNA virus (PhiX 174) by copper (Cu²⁺) and/or silver (Ag⁺) ions. Cu and Ag are increasingly applied in household water treatment (HHWT) systems. However, the impact of NOM on their inactivation kinetics remains elusive despite its importance for their application. The presence of NOM in water led to faster virus inactivation by Cu²⁺ but slower by Ag⁺. The fastest inactivation kinetics of MS2 (K_obs = 4.8 h⁻¹) were observed by Cu in water containing high NOM (20 mg C/L). Meanwhile, for PhiX 174, the fastest inactivation kinetics (av. K_obs = 3.5 h⁻¹) were observed by Cu and Ag synergism in water containing high NOM. Altogether, it can be concluded that the combination of Cu and Ag is promising as a virus disinfectant in treatment options allowing for multiple hours of residence time such as safe water storage tanks.

Rapid sand filters (RSF) are an established and widely applied technology for groundwater treatment. Yet, the underlying interwoven biological and physical-chemical reactions controlling the sequential removal of iron, ammonia and manganese remain poorly understood. To resolve the contribution and interactions between the individual reactions, we studied two full-scale drinking water treatment plant configurations, namely (i) one dual-media (anthracite and quartz sand) filter and (ii) two single-media (quartz sand) filters in series. In situ and ex situ activity tests were combined with mineral coating characterization and metagenome-guided metaproteomics along the depth of each filter. Both plants exhibited comparable performances and process compartmentalization, with most of ammonium and manganese removal occurring only after complete iron depletion. The homogeneity of the media coating and genome-based microbial composition within each compartment highlighted the effect of backwashing, namely the complete vertical mixing of the filter media. In stark contrast to this homogeneity, the removal of the contaminants was strongly stratified within each compartment, and decreased along the filter height. This apparent and longstanding conflict was resolved by quantifying the expressed proteome at different filter heights, revealing a consistent stratification of proteins catalysing ammonia oxidation and protein-based relative abundances of nitrifying genera (up to 2 orders of magnitude difference between top and bottom samples). This implies that microorganisms adapt their protein pool to the available nutrient load at a faster rate than the backwash mixing frequency. Ultimately, these results show the unique and complementary potential of metaproteomics to understand metabolic adaptations and interactions in highly dynamic ecosystems.

Slow Sand Filtration is popular in drinking water treatment for the removal of a wide range of contaminants (e.g., particles, organic matter, and microorganisms). The Schmutzdecke in slow sand filters (SSFs) is known to be essential for pathogen removal, however, this layer is also responsible for increased head loss. Since the role of deeper layers in bacteria and virus removal is poorly understood, this research investigated the removal of E.coli WR1 and PhiX 174 at different depths of a full-scale SSF. Filter material from top (0–5 cm), middle (5–20 cm) and deep (20–35 cm) layers of an established filter was used in an innovative experimental set-up to differentiate physical-chemical and biological removal processes. In the analysis, we distinguished between removal by biological activity, biofilm and just sand. In addition, we modelled processes by a one-side kinetic model. The different layers contributed substantially to overall log removal of E.coli WR1 (1.4–1.7 log₁₀) and PhiX 174 (0.4–0.6 log₁₀). For E.coli WR1, biological activity caused major removal, followed by removal within biofilm and sand, whereas, removal of PhiX 174 mainly occurred within sand, followed by biofilm and biological activity. Narrow pore radii in the top layer obtained by micro-computed tomography scanner suggested enhanced retention of bacteria due to constrained transport. The retention rates of E.coli WR1 and PhiX 174 in top layer were four and five times higher than deeper layers, respectively (k_ret 1.09 min⁻¹ vs 0.26 min⁻¹ for E.coli WR1 and k_ret 0.32 min⁻¹ vs of 0.06 min⁻¹ for PhiX 174). While this higher rate was restricted to the Schmutzdecke alone (top 5 cm), the deeper layers extend to around 1 m in full-scale filters. Therefore, the contribution of deeper layers of established SSFs to the overall log removal of bacteria and viruses is much more substantial than the Schmutzdecke.

At wastewater treatment plants (WWTPs), additional steps are introduced for removal of organic micropollutants (OMPs) from the treated effluents, especially pharmaceutical residues. At the same time, a new concern is emerging: antibiotic resistance (AR). This research studied the effect of ozonation, coagulation and granular activated carbon (GAC) filtration applied as tertiary treatment for the removal of OMPs and nutrients, on AR removal. Bacterial culture methods in selective media were used to screen for four different microorganisms: two faecal indicators (Escherichia coli and Enterococci) as antibiotic sensitive bacteria (ASB), and a resistant strain of each of these bacteria, namely Extended-Spectrum Beta-lactamase producing E. coli (ESBL-E.coli) and Vancomycin Resistant Enterococci (VRE) as antibiotic resistant bacteria (ARB). At laboratory scale, ozonation experiments (ozone dose 0.4–0.6 g O ₃/g DOC) and coagulation experiments using Polyaluminum chloride (PAX-214) and FeCl ₃ (coagulant dose 0.004–1 mM/L) were performed using secondary effluent from two municipal WWTPs. In addition in a pilot plant and full-scale plant ozonation (ozone dose 0.4 g O ₃/g DOC) and GAC filtration (empty bed contact time 15 min) were studied for AR removal. No significant differences were found between ARB and ASB removal for coagulation and ozonation which could indicate that ASB can be used as an initial proxy for ARB removal for these technologies. In the laboratory experiments, ozonation and coagulation showed a good removal of both ARB and ASB. However, the doses needed to reach 2–3 log removal were a factor 2.5–4 (ozonation) and 250 (coagulation) higher than applied for OMP removal (by ozonation) and phosphorus (P) removal (by coagulation). In the GAC filters, the risk of ARB enhancement occurred, especially in filters with a matured biology. Although these bacteria are not necessarily directly harmful, they can pass down their resistance to pathogenic bacteria via horizontal gene transfer.

Iron in anaerobic groundwater is commonly removed by oxidation followed by sand filtration. This produces large volumes of iron(III)(hydr)oxide sludge with little value. Our research investigates the novel concept of anaerobic iron(II) recovery from groundwater as the valuable mineral vivianite (Fe₃(PO₄)₂ • 8 H₂O) by the addition of phosphate to the water. We found that vivianite precipitated both in synthetic and natural groundwater when the saturation index (SI) was higher than 4. The SI can be increased by elevating the pH, which allows for iron removal at lower concentrations. Anaerobic iron removal reached 93.7% in natural groundwater, which increased further to 99.9% after a subsequent aeration step. Vivianite precipitation followed second order kinetics with a rate constant of 2.3 M⁻¹s⁻¹ and the sludge volume decreased by two third compared to iron oxidation. We therefore conclude that anaerobic iron removal is a promising new approach towards sustainable groundwater treatment.