New Perspectives on the Interactions between Adsorption and Degradation of Organic Micropollutants in Granular Activated Carbon Filters.

The removal of organic micropollutants in granular activated carbon (GAC) filters can be attributed to adsorption and biological degradation. These two processes can interact with each other or proceed independently. To illustrate the differences in their interaction, three 14C-labeled organic micropollutants with varying potentials for adsorption and biodegradation were selected to study their adsorption and biodegradation in columns with adsorbing (GAC) and non-adsorbing (sand) filter media. Using 14CO2 formation as a marker for biodegradation, we demonstrated that the biodegradation of poorly adsorbing N-nitrosodimethylamine (NDMA) was more sensitive to changes in the empty bed contact time (EBCT) compared with that of moderately adsorbing diclofenac. Further, diclofenac that had adsorbed under anoxic conditions could be degraded when molecular oxygen became available, and substantial biodegradation (≥60%) of diclofenac could be achieved with a 15 min EBCT in the GAC filter. These findings suggest that the retention of micropollutants in GAC filters, by prolonging the micropollutant residence time through adsorption, can enable longer time periods for degradations than what the hydraulic retention time would allow for. For the biologically recalcitrant compound carbamazepine, differences in breakthrough between the 14C-labeled and nonradiolabeled compounds revealed a substantial retention via successive adsorption-desorption, which could pose a potential challenge in the interpretation of GAC filter performance.

Biological degradation of organic micropollutants in GAC filters-temporal development and spatial variations.

The capacity for organic micropollutant removal in granular activated carbon (GAC) filters for wastewater treatment changes over time. These changes are in general attributed to changes in adsorption, but may in some cases also be affected by biological degradation. Knowledge on the degradation of organic micropollutants, however, is scarce. In this work, the degradation of micropollutants in several full-scale GAC and sand filters was investigated through incubation experiments over a period of three years, using 14C-labeled organic micropollutants with different susceptibilities to biological degradation (ibuprofen, diclofenac, and carbamazepine), with parallel 16S rRNA gene sequencing. The results showed that the degradation of diclofenac and ibuprofen in GAC filters increased with increasing numbers of bed volumes when free oxygen was available in the filter, while variations over filter depth were limited. Despite relatively large differences in bacterial composition between filters, a degradation of diclofenac was consistently observed for the GAC filters that had been operated with high influent oxygen concentration (DO >8 mg/L). The results of this comprehensive experimental work provide an increased understanding of the interactions between microbial composition, filter material, and oxygen availability in the biological degradation of organic micropollutants in GAC filters.

Primary filtration of municipal wastewater with sludge fermentation - Impacts on biological nutrient removal.

Primary filtration is a compact pre-treatment process for municipal wastewater, which can lead to high removal of total suspended solids (TSS) if polymer is added prior to filtration. Extensive carbon removal with rotating belt filter (RBF) can be combined with filter primary sludge fermentation at ambient temperature, in order to produce volatile fatty acids (VFAs) as carbon source for biological nutrient removal (BNR). This process was implemented at large pilot-scale and operated for more than a year. The results showed that the RBF efficiently removed particles >10 µm, and that the TSS removal had a strong linear correlation to the influent TSS concentration. Fermentation of the sludge at ambient temperature and five days retention time and addition of the fermentate to the wastewater could nearly double the VFA concentration in the wastewater by adding 31 ± 9 mg VFA-COD/L. Meanwhile, an increase of 2 mg/L of ammonium nitrogen, and 0.7 mg /L of phosphate phosphorus would be added to the wastewater with the fermentate. Adding the fermented sludge to the wastewater stream and removing the particles with RBF makes it possible to utilize nearly all the produced VFAs for BNR, and the feasibility of this configuration was shown at pilot-scale. According to simulations of subsequent BNR, the pre-treatment would lead to lower effluent total nitrogen concentrations. Alternatively, the required BNR volume could be reduced by 11-18 %. The estimated total biogas production was similar for pre-treatment with primary settler and RBF with fermentation. RBF without fermentation gave the most favourable energy balance, but did not reach the same low effluent value for total nitrogen as RBF with fermentation.

Assessing the potential of a membrane bioreactor and granular activated carbon process for wastewater reuse - A full-scale WWTP operated over one year in Scania, Sweden.

A full-scale membrane bioreactor (MBR) with ultrafiltration, followed by granular activated carbon (GAC), was examined to determine the potential of reusing treated water as a source of drinking water or for irrigation. The major part of the bacteria removal took place in the MBR, whereas the GAC removed substantial amounts of organic micropollutants. Annual variations in inflow and infiltration resulted in a concentrated influent during summer and a diluted influent in the winter. The removal of E. coli was high throughout the process (average log removal 5.8), with effluent concentrations meeting the threshold for class B water standards for irrigation (EU 2020/741) but exceeding those for drinking water in Sweden. The total bacterial concentration increased over the GAC, indicating the growth and release of bacteria; however, E. coli concentrations declined. The effluent concentrations of metals met the Swedish criteria for drinking water. The removal of organic micropollutants decreased during the initial operation of the treatment plant, but after 1 year and 3 months, corresponding to 15,000 bed volumes, the removal increased. Maturation of the biofilm in the GAC filters might have resulted in biodegradation of certain organic micropollutants, in combination with bioregeneration. Although there is no legislation in Scandinavia with regard to many organic micropollutants in drinking water and water for irrigation, the effluent concentrations were generally in the same order of magnitude as to those in Swedish source waters that are used for drinking water production.

Seasonal variations in acidogenic fermentation of filter primary sludge.

Primary treatment of municipal wastewater by rotating belt filtration followed by hydrolysis and acidogenic fermentation of the filter primary sludge (FPS) at ambient temperature was studied at pilot-scale during one year. The seasonal variations of volatile fatty acids (VFAs), nutrient release and soluble COD production as well as microbial community assembly were assessed, leading to novel findings for fermentation at ambient temperature. The reproducibility of VFA production performance was first established by operating the two fermentation reactors under the same conditions, showing similar results regarding VFA production and microbial community structure. One year of operation at 5 d retention time (RT) and 16-29 °C resulted in an average VFA yield of 180±35 mg COD/g VSin and soluble COD yield of 242±40 mg COD/g VSin. The VFA formation was temperature-dependent, with Ï´=1.033±0.005 ( [Formula: see text] . The seasonal variations of the acetic and propionic acid productions were pronounced, whereas the productions of VFAs with longer chains were more stable regardless of temperature. The community structure of the reactor microbiomes was also clearly affected by season and temperature and linked with the production spectrum of VFAs. The ammonium and phosphate releases were stable during the year, leading to a decrease in ratios of soluble COD to NH4+-N and PO43--P during winter. The soluble COD yield was 11% and 27% higher at 5 d RT compared to 3 and 2 d RT respectively, but the corresponding volumetric productivities were lower. The dissimilarities between microbiomes in influent FPS and fermenters were significant even at a short RT of 2 d, and increased with longer RT of 3 and 5 d, primarily caused by selection of bacteria within Bacteroidota in the fermentation reactors.

Chemically enhanced primary treatment, microsieving, direct membrane filtration and GAC filtration of municipal wastewater: a pilot-scale study.

Chemically enhanced primary treatment (CEPT) followed by microsieving and direct membrane filtration (DMF) as ultrafiltration, was evaluated on pilot scale at a municipal wastewater treatment plant. In addition, a granular activated carbon (GAC) filter downstream of DMF was evaluated for the removal of organic micropollutants. Up to 80% of the total organic carbon (TOC) and 96% of the total phosphorus were removed by CEPT with microsieving. The additional contribution of subsequent DMF was minor, and only five days of downstream GAC filtration was possible due to fouling of the membrane. Of the 21 organic micropollutants analysed, all were removed (≥ 98%) by the GAC filter until 440 bed volumes, while CEPT with microsieving and DMF removed only a few compounds. Measurements of the oxygen uptake rate indicated that the required aeration for supplementary biological treatment downstream of CEPT with microsieving, both with and without subsequent DMF, was 20-25% of that in the influent wastewater. This study demonstrated the potential of using compact physicochemical processes to treat municipal wastewater, including the removal of organic micropollutants.

Ozonation of 14C-labeled micropollutants - mineralization of labeled moieties and adsorption of transformation products to activated carbon.

Ozonation transformation products (OTPs) are largely unknown compounds that are formed during the ozonation of micropollutants, and it is uncertain to which extent these compounds can be removed by subsequent adsorption to activated carbon. Thus, 14C-labeled micropollutants were ozonated to generate 14C-labeled OTPs, for which the adsorption of the sum of all 14C-labeled OTPs to activated carbon could be determined, based on the adsorption of the labeled carbon. Further, 14CO2 traps were used to examine the mineralization of 14C-labeled moieties during ozonation. 14CO2-formation revealed a partial mineralization of the 14C-labeled moieties in all compounds except for propyl-labeled bisphenol A and O-methyl-labeled naproxen. A similar degree of mineralization was noted for different compounds labeled at the same moiety, including the carboxylic carbon in diclofenac and ibuprofen (∼40% at 1 g O3/g DOC) and the aniline ring in sulfamethoxazole and sulfadiazine (∼30% at 1 g O3/g DOC). Aromatic ring cleavage was also confirmed for bisphenol A, sulfamethoxazole, and sulfadiazine through the formation of 14CO2. The adsorption experiments demonstrated increased adsorption of micropollutants to powdered activated carbon after ozonation, which was connected to a decreased adsorption of dissolved organic matter (DOM). Conversely, the OTPs showed a substantial and successive decline in adsorption at increased ozone doses for all compounds, likely due to decreased hydrophobicity and aromaticity of the OTPs. These findings indicate that adsorption to activated carbon alone is not a viable removal method for a wide range of ozonation transformation products.

Microbial bromate reduction following ozonation of bromide-rich wastewater in coastal areas.

Ozonation of wastewater can reduce the release of organic micropollutants, but may result in the formation of undesirable by-products, such as bromate from bromide. Bromide is one of the most abundant ions in seawater, the primary precursor of bromate during ozonation, and the end product in microbial bromate reduction. Investigations were carried out to compare the concentration of bromide in wastewater in coastal and non-coastal catchment areas, to monitor bromate formation during ozonation, and to assess the potential for subsequent bromate reduction with denitrifying carriers. Higher bromide concentrations were systematically observed in wastewater from coastal catchment areas (0.2-2 mg Br-/L) than in wastewater from non-coastal areas (0.06-0.2 mg Br-/L), resulting in elevated formation of bromate during ozonation. Subsequent investigations of bromate reduction in contact with denitrifying carriers from two full-scale moving bed biofilm reactors (MBBRs) showed that 80 % of the bromate formed during ozonation could be reduced to bromide in 60 min with first-order rate constants of 0.3-0.8 L/(gbiomass·h). Flow-through experiments with denitrifying carriers also showed that combined reduction of bromate and nitrate could be achieved below a concentration of 2 mg NOx--N/L. These findings indicate that bromide-rich wastewater is more likely to be of concern when using ozonation in coastal than in non-coastal areas, and that bromate and nitrate reduction can be combined in a single biofilm reactor.

Digging deep into a GAC filter - Temporal and spatial profiling of adsorbed organic micropollutants.

A large pilot-scale granular activated carbon (GAC) filter was operated downstream in a full-scale wastewater treatment plant to remove organic micropollutants. To describe the spatial and temporal developments of micropollutant adsorption profiles in the GAC filter, micropollutants were extracted from GAC media taken at various filter depths and number of treated bed volumes. At a low number of treated bed volumes (2600 BVs), most micropollutants were adsorbed in the top layers of the filter. At increasing number of treated bed volumes (7300-15,500 BVs), the adsorption front for micropollutants progressed through the filter bed at varying rates, with sulfamethoxazole, fluconazole, and PFOS reaching the bottom layer before carbamazepine and other well-adsorbing micropollutants, such as propranolol and citalopram. Higher amounts of adsorbed micropollutants in the bottom layer of the filter bed resulted in decreased removal efficiencies in the treated wastewater. Mass estimations indicated biodegradation for certain micropollutants, such as naproxen, diclofenac, and sulfamethoxazole. A temporary increase in the concentration of the insecticide imidacloprid could be detected in the filter indicating that extraction of adsorbed micropollutants could provide an opportunity for backtracking of loading patterns.

A comparison of adsorption of organic micropollutants onto activated carbon following chemically enhanced primary treatment with microsieving, direct membrane filtration and tertiary treatment of municipal wastewater.

The adsorption of organic micropollutants onto powdered activated carbon (PAC) was investigated in laboratory scale based on samples from four wastewater process streams (matrices); three from a pilot-scale plant with different degrees of physicochemical treatment of municipal wastewater and one from a full-scale activated sludge plant with post-precipitation. The pilot-scale treatment consisted of chemically enhanced primary treatment with microsieving followed by direct membrane filtration as microfiltration or ultrafiltration. The results showed highest adsorption of micropollutants in the tertiary (biologically and chemically) treated wastewater and lowest adsorption in the microsieve filtrate. Adsorption of micropollutants in the direct membrane microfiltration (200 nm) permeate was generally similar to that in the direct membrane ultrafiltration (3 nm) permeate. The higher adsorption of micropollutants in the tertiary treated wastewater could be related to a lower concentration of dissolved organic carbon (DOC) and lower affinity of DOC for PAC at low dosage (<15 mg PAC/L) in this matrix. At a PAC dose of 10 mg/L, sulfamethoxazole was removed by 33% in the tertiary treated wastewater and 7% in the direct membrane microfiltration permeate. In addition to the PAC experiments, a pilot scale sand filter and a proceeding GAC filter was operated on tertiary treated wastewater from the full-scale treatment plant. Similar removal trends in the PAC and GAC experiments were observed when studying a weighted average micropollutant removal in the GAC filter and a similar dose of activated carbon for both PAC and GAC. Positively charged micropollutants were removed to a higher extent than negatively charged ones by both PAC and GAC.

Tracking 14C-Labeled Organic Micropollutants to Differentiate between Adsorption and Degradation in GAC and Biofilm Processes.

Granular activated carbon (GAC) filters can be used to reduce emissions of organic micropollutants via municipal wastewater, but it is still uncertain to which extent biological degradation contributes to their removal in GAC filters. 14C-labeled organic micropollutants were therefore used to distinguish degradation from adsorption in a GAC-filter media with associated biofilm. The rates and extents of biological degradation and adsorption were investigated and compared with other biofilm systems, including a moving bed biofilm reactor (MBBR) and a sand filter, by monitoring 14C activities in the liquid and gas phases. The microbial cleavage of ibuprofen, naproxen, diclofenac, and mecoprop was confirmed for all biofilms, based on the formation of 14CO2, whereas the degradation of 14C-labeled moieties of sulfamethoxazole and carbamazepine was undetected. Higher degradation rates for diclofenac were observed for the GAC-filter media than for the other biofilms. Degradation of previously adsorbed diclofenac onto GAC could be confirmed by the anaerobic adsorption and subsequent aerobic degradation by the GAC-bound biofilm. This study demonstrates the potential use of 14C-labeled micropollutants to study interactions and determine the relative contributions of adsorption and degradation in GAC-based treatment systems.

Integrating dissolved and particulate matter into a prediction tool for ozonation of organic micropollutants in wastewater.

Ozonation is an established technique used to reduce the discharge of organic micropollutants into the aquatic environment, but the possibility of predicting the ozone demand for different wastewater matrices is still limited, especially in the presence of suspended solids (SS). A new tool for the prediction of the removal of organic micropollutants with ozone, based on dissolved and particulate matter in activated sludge effluents, was therefore developed. The removal of 25 organic micropollutants was determined on laboratory scale in the presence and absence of suspended solids. The linear trajectories of the dose-response curves enabled the determination of a new set of removal constants, based on dissolved chemical oxygen demand (COD) and SS. The presence of SS had a more negative effect on the removal of slow-reacting micropollutants (removal constant <3.5 mg CODCr,diss·mg O3-1) with ozone than on the fast-reacting micropollutants (removal constant >3.5 mg CODCr,diss·mg O3-1). However, the decreased removal of the organic micropollutants was generally small, <10%, at typical SS concentrations, <25 mg SS·L-1. Integration of the new removal constants based on COD and SS enabled the removal in an ozone pilot plant to be modelled with an average deviation of <10% for several organic micropollutants. The use of the frequently measured parameters, COD and SS, as input parameters could facilitate the future use of the tool to predict the removal of micropollutants during ozonation.

Promoting the degradation of organic micropollutants in tertiary moving bed biofilm reactors by controlling growth and redox conditions.

A novel process configuration was designed to increase biofilm growth in tertiary moving bed biofilm reactors (MBBRs) by providing additional substrate from primary treated wastewater in a sidestream reactor under different redox conditions in order to improve micropollutant removal in MBBRs with low substrate availability. This novel recirculating MBBR was operated on pilot scale for 13 months, and a systematic increase was seen in the biomass concentration and the micropollutant degradation rates, compared to a tertiary MBBR without additional substrate. The degradation rates per unit carrier surface area increased in the order of ten times, and for certain micropollutants, such as atenolol, metoprolol, trimethoprim and roxithromycin, the degradation rates increased 20-60 times. Aerobic conditions were critical for maintaining high micropollutant degradation rates. With innovative MBBR configurations it may be possible to improve the biological degradation of organic micropollutants in wastewater. It is suggested that degradation rates be normalized to the carrier surface area, in favor of the biomass concentration, as this reflects the diffusion limitations of oxygen, and will facilitate the comparison of different biofilm systems.

MBBRs as post-treatment to ozonation: Degradation of transformation products and ozone-resistant micropollutants.

The degradation potential of micropollutants and transformation products in biological post-treatment after ozonation is partly unknown. A pilot plant with ozonation and subsequent biological treatment in a moving bed biofilm reactor (MBBR) was thus operated over 16 months to investigate the removal of micropollutants and the formation and removal of N-oxide transformation products. Lab-scale kinetic experiments were performed in parallel. At a moderate ozone dose of 0.5 g O3 g-1 DOC, further degradation of gabapentin and 3 iodinated contrast media (iomeprol, iopamidol, and iohexol) could be induced by the biofilm at prolonged exposure times. To facilitate comparison of feeding regimens in biofilm systems a new surface-related degradation rate constant was introduced. The availability of substrates in the pilot MBBR influenced the micropollutant degradation kinetics with increasing and decreasing degradation rates. N-oxides from erythromycin, clarithromycin, tramadol, and venlafaxine were formed during ozonation and could not be degraded by the biofilm.

Removal of pharmaceutical metabolites in wastewater ozonation including their fate in different post-treatments.

Advanced treatment technologies for the removal of pharmaceuticals and other organic micropollutants in WWTPs primarily target the removal of parent compounds. Nevertheless, the removal of metabolites originating from human- or microbial metabolism during biological treatment needs comparable consideration, as some of them might be present in high concentrations and contribute to toxicity. This study was conducted to elucidate the removal of human and microbial metabolites of pharmaceuticals as a function of the specific ozone dose. Ozonation was performed on four sites with two pilot- and two full-scale plants operated downstream of conventional activated sludge plants. The ozone reactivity of all metabolites (expressed as the ozone dose to remove 90% of the compound/decadic ozone dose) was lower than those of their parent compounds. The decadic ozone dose was 1.0, 1.3 and 1.1 mg O3/mg DOC for Epoxy-carbamazepine, Di-OH-carbamazepine and N-Desmethyl tramadol, respectively. 20-40% of the remaining metabolites were removed in a polishing sand/BAC-filter (biological activated carbon). Similar removal was observed for Epoxy-carbamazepine, Di-OH-carbamazepine and Hydroxy-diclofenac in a constructed wetland. However, the sand/anthracite filter had no effect. All four metabolites were removed in a GAC (granulated activated carbon) filter.

Ozone dose dependent formation and removal of ozonation products of pharmaceuticals in pilot and full-scale municipal wastewater treatment plants.

The removal of micropollutants from municipal wastewater is challenged by the number of compounds with diverse physico-chemical properties. Ozonation is increasingly used to remove micropollutants from wastewater. However, ozonation does not necessarily result in complete mineralization of the organic micropollutants but rather transforms them into new compounds which could be persistent or have adverse environmental effects. To explore ozone dose dependency of the formation and successive removal of ozonation products, two pilot-scale and one full-scale ozonation plants were operated subsequent to a conventional activated sludge treatment. The results from these trials indicated that the concentrations of several N-oxides, such as Erythromycin N-oxide, Venlafaxine N-oxide and Tramadol N-oxide, increased up to an ozone dose of 0.56-0.61 mg O3/mg DOC while they decreased at elevated doses of 0.7-1.0 mg O3/mg DOC. Similar results were also obtained for two transformation products of Diclofenac (Diclofenac 2,5-quinone imine and 1-(2,6-dichlorophenyl)indolin-2,3-dione) and one transformation product of Carbamazepine (1-(2-benzoic acid)-(1H,3H)-quinazoline-2,4-dione), where the highest concentrations appeared around 0.27-0.31 mg O3/mg DOC. The formation maximum of a given compound occurred at a specific ozone dose that is characteristic for each compound, but seemed to be independent of the wastewater used for the experiments at the two pilots and the full-scale plant.

PAC dosing to an MBBR - Effects on adsorption of micropollutants, nitrification and microbial community.

Two nitrifying MBBR reactors were operated in parallel, one with PAC dosing and one without, to determine the effects of PAC dosing on nitrification and micropollutant adsorption in municipal wastewater. The removal of micropollutants was evaluated for several doses of PAC and batch experiments were performed to measure adsorption kinetics and nitrification rates. The influence of PAC on the nitrifying microbial community was examined by high-throughput amplicon sequencing. Long-term operation of the pilot reactors showed that nitrification could be maintained while supplying PAC at increasing doses, as confirmed by high nitrification rates and significant abundance of nitrifying bacteria. The adsorption of organic micropollutants could be controlled by the PAC dose, and increased dosing resulted in corresponding improvements in removal efficiency. Biomass, suspended or attached to carriers, did not interfere with the adsorption of organic micropollutants. Freundlich isotherms obtained from the batch experiments were used to predict removal of organic micropollutants in the pilot reactors, suggesting that batch adsorption experiments can be used to predict micropollutant removal on a full scale. Collectively, the results show that nitrification and adsorption of organic micropollutants can be performed simultaneously in an MBBR.