Linear Mixed Model of Virus Disinfection by Free Chlorine to Harmonize Data Collected across Broad Environmental Conditions.

Despite the critical importance of virus disinfection by chlorine, our fundamental understanding of the relative susceptibility of different viruses to chlorine and robust quantitative relationships between virus disinfection rate constants and environmental parameters remains limited. We conducted a systematic review of virus inactivation by free chlorine and used the resulting data set to develop a linear mixed model that estimates chlorine inactivation rate constants for viruses based on experimental conditions. 570 data points were collected in our systematic review, representing 82 viruses over a broad range of environmental conditions. The harmonized inactivation rate constants under reference conditions (pH = 7.53, T = 20 °C, [Cl-] < 50 mM) spanned 5 orders of magnitude, ranging from 0.0196 to 1150 L mg-1 min-1, and uncovered important trends between viruses. Whereas common surrogate bacteriophage MS2 does not serve as a conservative chlorine disinfection surrogate for many human viruses, CVB5 was one of the most resistant viruses in the data set. The model quantifies the role of pH, temperature, and chloride levels across viruses, and an online tool allows users to estimate rate constants for viruses and conditions of interest. Results from the model identified potential shortcomings in current U.S. EPA drinking water disinfection requirements.

Critical Role of Trichloramine Interaction with Dichloramine for N-Nitrosamine Formation during Breakpoint Chlorination.

Breakpoint chlorination is prevalent in drinking water and potable reuse water treatment. Breakpoint chlorination enhances the formation of N-nitrosamines through reactions that form nitrosating agents. The most recent study suggests that nitroxyl (HNO) can react with free chlorine (HOCl) to form the nitrosyl chloride (ClNO) nitrosating agent but has not experimentally verified its importance in breakpoint chlorination. This study first assessed the formation of N-nitrosamines from model N-chloro-alkylamine precursors when they were added to a mixture of HOCl and HNO-derived nitrosating agents generated by chlorinating hydroxyurea. Results demonstrated negligible N-nitrosamine formation. Instead, we observed that the interaction of NCl3 with NHCl2 (total Cl2/total N molar ratio = 2.4-3:1) produced an intermediate capable of nitrosating N-chloro-alkylamines to N-nitrosamines at yields 8-fold higher to those observed in NHCl2 treatment alone, within a very short timescale (<3 min). We examined the stoichiometry of the reaction of NCl3 with NHCl2 using a UV-spectrum-based approach. Nitrosyl chloride was proposed as the key intermediate, likely formed alongside the reformation of NHCl2. Further isotopic experiments, byproduct measurements, and kinetic modeling supported the hypotheses. Modeling indicated that the reaction of NCl3 with NHCl2 explained ∼75% of NDMA formation during breakpoint chlorination. Because NCl3 is mainly derived from the reaction of HOCl with NHCl2, controlling NHCl2 (e.g., with additional treatment) is critical for minimizing nitrosamine formation in waters where breakpoint chlorination occurs.

Ferryl Ion in the Photo-Fenton Process at Acidic pH: Occurrence, Fate, and Implications.

Fenton processes produce reactive species that can oxidize organic compounds in natural and engineered systems. While it is well-documented that Fenton reactions produce hydroxyl radical (HO•) under acidic conditions, we demonstrated the generation of ferryl ion (FeIVO2+) in the UV/Fe(III) and UV/Fe(III)/H2O2 systems at pH 2.8 using methyl phenyl sulfoxide (PMSO) as the probe compound. Moreover, we clarified that FeIVO2+ is parallelly formed via the oxidation of Fe(III) by HO• and the O-O homolysis of [FeIII-OOH]2+ in the photo-Fenton process. The rate constant for the reaction between HO• and Fe3+ measured by laser flash photolysis was 4.41 × 107 M-1 s-1. The rate constant and quantum yield for thermal and photo O-O homolysis of [FeIII-OOH]2+ complex were 1.4 × 10-2 s-1 and 0.3, respectively, which were determined by fitting PMSO2 formation. While FeIVO2+ forms predominantly through the reaction between HO• and Fe3+ in the absence of H2O2, the relative contribution of [FeIII-OOH]2+ O-O homolysis to FeIVO2+ formation highly depends on the molar ratio of [H2O2]0/[Fe(III)]0, the level of HO• scavenging, and incident irradiance in the UV/Fe(III)/H2O2 system. Accordingly, an optimized kinetic model was developed by incorporating FeIVO2+-involved reactions into the conventional photo-Fenton model, which can accurately predict Fe(II) formation and contaminant decay in the UV/Fe(III) and UV/Fe(III)/H2O2 systems. Our study illuminated the underlying formation mechanism of reactive oxidative species in the photo-Fenton process and highlighted the role of FeIVO2+ evolution in modulating the iron cycle and pollutant abatement therein.

Dissolved Methane Recovery and Trace Contaminant Fate Following Mainstream Anaerobic Treatment of Municipal Wastewater.

Anaerobic treatment of municipal wastewater with the staged anaerobic fluidized bed membrane bioreactor (SAF-MBR) shows promise to transform secondary wastewater treatment into an energy-positive process. However, the dissolved methane in SAF-MBR effluent needs to be recovered to reach net energy positive. To recover this methane for energy generation, an air stripping system was constructed downstream of a pilot-scale SAF-MBR facility and operated for over 80 days. The process removed 98% of effluent dissolved methane, and with the addition of intermittent disinfection recovered an average of 90% of the dissolved methane. The exit gas from air-stripping comprised 1.5-2.5% methane and could be utilized by blending with biogas produced from primary solids digestion and the SAF-MBR in an on-site combustion process. The direct energy costs for air stripping methane are <1% of the energy recoverable from the dissolved methane, not accounting for siloxane or sulfide scrubbing. Only siloxanes were observed at levels impacting combustion in this study, with 1.6 mg Si/m3 present in the blended biogas and air stripping mixture. The fate of a subset of trace organic contaminants was examined across the air stripping unit to check for aerobic degradation by methanotrophs or other opportunistic aerobes. Only 1,4-dioxane and benzotriazole showed statistically significant removal among 17 compounds screened, with 0.53 ± 0.13 and 0.34 ± 0.15 fraction removal, respectively. Our results indicate that air stripping is an energy efficient and robust technology for dissolved methane removal and onsite utilization for heat and electricity generation from anaerobic treatment of municipal wastewater.

Chloride Enhances DNA Reactivity with Chlorine under Conditions Relevant to Water Treatment.

Free available chlorine (FAC) is widely used to inactivate viruses by oxidizing viral components, including genomes. It is commonly assumed that hypochlorous acid (HOCl) is the chlorinating agent responsible for virus inactivation; however, recent studies have underscored that minor constituents of FAC existing in equilibrium with HOCl, such as molecular chlorine (Cl2), can influence FAC reactivity toward select organic compounds. This study measures the FAC reaction kinetics with dsDNA and ssDNA extracted from representative bacteriophages (T3 and ϕX174) in samples augmented with chloride. Herein, chloride enhances FAC reactivity toward dsDNA and, to a lesser extent, toward ssDNA, especially at pH < 7.5. The enhanced reactivity can be attributed to the formation of Cl2. Second-order rate constants were determined for reactions of ssDNA and dsDNA with HOCl and Cl2. DNA chlorination kinetics followed the reactivity-selectivity principle, where the more-reactive nucleophilic species (ssDNA, ∼100× more reactive than dsDNA) reacted less selectively with electrophilic FAC species. The addition of chloride was also shown to enhance the inactivation of bacteriophage T3 (dsDNA genome) by FAC but did not enhance the inactivation of bacteriophage ϕX174 (ssDNA genome). Overall, the results suggest that Cl2 is an important chlorinating agent of nucleic acids and viruses.

Recovery of Clean Water and Ammonia from Domestic Wastewater: Impacts on Embodied Energy and Greenhouse Gas Emissions.

Treatment of domestic wastewater can recover valuable resources, including clean water, energy, and ammonia. Important metrics for these systems are greenhouse gas (GHG) emissions and embodied energy, both of which are location- and technology-dependent. Here, we determine the embodied energy and GHG emissions resulting from a conventional process train, and we compare them to a nonconventional process train. The conventional train assumes freshwater conveyance from a pristine source that requires energy for pumping (US average of 0.29 kWh/m3), aerobic secondary treatment with N removal as N2, and Haber-Bosch synthesis of ammonia. Overall, we find that this process train has an embodied energy of 1.02 kWh/m3 and a GHG emission of 0.77 kg-CO2eq/m3. We compare these metrics to those of a nonconventional process train that features anaerobic secondary treatment technology followed by further purification of the effluent by reverse osmosis and air stripping for ammonia recovery. This "short-cut" process train reduces embodied energy to 0.88 kWh/m3 and GHG emissions to 0.42 kg-CO2eq/m3, while offsetting demand for ammonia from the Haber-Bosch process and decreasing reliance upon water transported over long distances. Finally, to assess the potential impacts of nonconventional nitrogen removal technology, we compared the embodied energy and GHG emissions resulting from partial nitritation/anammox coupled to anaerobic secondary treatment. The resulting process train enabled a lower embodied energy but increased GHG emissions, largely due to emissions of N2O, a potent greenhouse gas.

Reactivity of Viral Nucleic Acids with Chlorine and the Impact of Virus Encapsidation.

Free chlorine disinfection is widely applied to inactivate viruses by reacting with their biomolecules, which include nucleic acids, proteins, and lipids. Knowing the reactivities of viral genomes with free chlorine and the protection that encapsidation provides would ultimately help predict virus susceptibility to the disinfectant. The relative reactivities of different viral genome types and the impact of viral higher order structure with free chlorine are poorly characterized. Here, we studied the reactivity of viral genomes representing four genome types from virus particles with diverse structures, namely, (+)ssRNA (MS2), dsRNA (φ6), ssDNA (φX174), and dsDNA (T3) with free chlorine. We compared the reactivities of these viral nucleic acids when they were suspended in phosphate buffer solutions (naked forms) and when they were in the native virus particles (encapsidated forms). The reactivities of nucleic acids were tracked by polymerase chain reaction (PCR)-based assays. The naked dsDNA of T3 was the least reactive with free chlorine, with an average second order rate constant normalized by the number of bases in the measured regions (in M-1 s-1 b-1) that was 34×, 65×, and 189× lower than those of the dsRNA of φ6, ssRNA of MS2, and ssDNA of φX174, respectively. Moreover, different regions in the ssRNA genome of MS2 and the dsRNA genome of φ6 exhibited statistically different reaction kinetics. The genomes within virus particles reacted slower than the naked genomes overall, but the extent of these differences varied among the four viruses. The results on viral nucleic acid reactivity help explain different susceptibilities of viruses to inactivation by free chlorine and also provide a valuable comparison of the susceptibilities of different nucleic acids to oxidants.

Control of sulfides and coliphage MS2 using hydrogen peroxide and UV disinfection for non-potable reuse of pilot-scale anaerobic membrane bioreactor effluent.

Anaerobic membrane bioreactors reduce the energy cost of wastewater treatment and meet filtration requirements for non-potable reuse. However, sulfides (H2S/HS-) formed during anaerobic treatment exert a high chlorine demand and inhibit UV disinfection by photon shielding at 254 nm. This study evaluated the feasibility of hydrogen peroxide (H2O2) for sulfide oxidation, UV disinfection for inactivation of MS2 bacteriophage, and chlorine to provide a residual for distribution. H2O2 treatment at pH ≥ 8 favored sulfide oxidation to sulfate in 30 min at a 4:1 H2O2:sulfide stoichiometry. Compared to a 6:1 H2O2:sulfide molar ratio, treatment of anaerobic effluent with 0.5 mM sulfides with a 4:1 H2O2:sulfide molar ratio would increase the applied UV fluence needed for 5-log MS2 inactivation from 180 mJ cm-2 to 225 mJ cm-2. However, the lower H2O2 dose reduced the dose of chlorine needed to quench residual H2O2 and provide a residual for distribution. Treatment at the 4:1 H2O2:sulfide molar ratio was favored, because the cost savings in H2O2 and chlorine reagents outweighed the energy savings associated with UV treatment. However, H2O2/UV/chlorine treatment of anaerobic effluent was cost-competitive with conventional treatment of aerobic effluent for non-potable reuse only for < 285 µM sulfides.

N-Nitrosodimethylamine Formation during UV/Hydrogen Peroxide and UV/Chlorine Advanced Oxidation Process Treatment Following Reverse Osmosis for Potable Reuse.

Chloramines applied to control microfiltration and reverse osmosis (RO) membrane biofouling in potable reuse trains form the potent carcinogen, N-nitrosodimethylamine (NDMA). In addition to degrading other contaminants, UV-based advanced oxidation processes (AOPs) strive to degrade NDMA by direct photolysis. The UV/chlorine AOP is gaining attention because of its potential to degrade other contaminants at lower UV fluence than the UV/hydrogen peroxide AOP, although previous pilot studies have observed that the UV/chlorine AOP was less effective for NDMA control. Using dimethylamine (DMA) as a model precursor and secondary municipal wastewater effluent, this study evaluated NDMA formation during the AOP treatment via two pathways. First, NDMA formation by UV treatment of monochloramine (NH2Cl) and chlorinated DMA (Cl-DMA) passing through RO membranes was maximized at 350 mJ/cm2 UV fluence, declining at higher fluence, where NDMA photolysis outweighed NDMA formation. Second, this study demonstrated that chlorine addition to the chloramine-containing RO permeate during the UV/chlorine AOP treatment initiated rapid NDMA formation by dark breakpoint reactions associated with reactive intermediates from the hydrolysis of dichloramine. At pH 5.7, this formation was maximized at a chlorine/ammonia molar ratio of 3 (out of 0-10), conditions typical for UV/chlorine AOPs. At 700 mJ/cm2 UV fluence, which is applicable to current practice, NDMA photolysis degraded a portion of the NDMA formed by breakpoint reactions. Lowering UV fluence to ∼350 mJ/cm2 when switching to the UV/chlorine AOP exacerbates effluent NDMA concentrations because of concurrent NDMA formation via the UV/NH2Cl/Cl-DMA and breakpoint chlorination pathways. Fluence >700 mJ/cm2 or chlorine doses greater than the 3:1 chlorine/ammonia molar ratios under consideration for the UV/HOCl AOP treatment are needed to achieve NDMA control.

Efficacy of ozone for removal of pesticides, metals and indicator virus from reverse osmosis concentrates generated during potable reuse of municipal wastewaters.

This study evaluated ozone treatment to address concerns regarding the discharge to marine waters of chemical contaminants and pathogens in reverse osmosis (RO) concentrates generated during the potable reuse of municipal wastewaters. Previous studies indicated that contaminants can be sorted into five groups based on their reaction rate constants with ozone and hydroxyl radical to predict degradation of chemical contaminants during ozonation of municipal effluents. Spiking representatives of each group into five RO concentrate samples, this study demonstrated that the same contaminant grouping scheme could be used to predict contaminant degradation during ozonation of RO concentrates, despite the higher concentrations of ozone and hydroxyl radical scavengers. The predictive capability of the contaminant grouping scheme was further validated for four contaminants of concern in RO concentrates, including the pesticides fipronil and imidacloprid, and the metal chelates Ni-EDTA and Cu-EDTA. After measuring their ozone and hydroxyl radical reaction rate constants, these compounds were assigned to contaminant groups, and their degradation during ozonation matched predictions. Addition of 300 mg/L CaO at pH 11 achieved partial removal of the native nickel and copper by precipitation. Ozone pretreatment further enhanced precipitation of nickel, but not copper. Ozonation achieved 5-log inactivation of MS2 in all five concentrate samples at 1.18 mg O3/mg DOC. Ozonation at 0.9 mg O3/mg DOC formed 139-451 µg/L bromate. Pretreatment of RO concentrates with chlorine and ammonia reduced bromate formation by a maximum of 48% but increased total halogenated DBP concentrations from 20 µg/L to 36 µg/L. Regardless, neither bromate nor trihalomethane concentrations exceeded threshold concentrations of concern for discharge to marine waters.

Bench-scale column evaluation of factors associated with changes in N-nitrosodimethylamine (NDMA) precursor concentrations during drinking water biofiltration.

Biofiltration has been observed to increase or decrease the concentrations of N-nitrosodimethylamine (NDMA) precursors in the effluents of full-scale drinking water facilities, but these changes have been inconsistent over time. Bench-scale tests comparing biofiltration columns side-by-side exposed to different conditions were employed to characterize factors associated with changes in NDMA precursor concentrations, as measured by application of chloramines under uniform formation conditions (UFC). Side-by-side comparisons of biofiltration media from different facilities fed with water from each of these facilities demonstrated that differences in source water quality were far more important than any original differences in the microbial communities on the biofiltration media for determining whether NDMA precursor concentrations increased, decreased or remained constant across biofilters. Additional tests involving spiking of specific constituents hypothesized to promote increases in NDMA precursor concentrations demonstrated that inorganic nitrogen species associated with nitrification, including ammonia, hydroxylamine and chloramines, and biotransformation of known precursors (i.e., municipal wastewater and the cationic polymer, polyDADMAC) to more potent forms were not important. Biotransformation of uncharacterized components of source waters determined whether NDMA precursor concentrations increased or decreased across biofilters. These uncharacterized source water component concentrations varied temporally and across locations. Where biotransformation of source water precursors increased NDMA precursor concentrations, ∼30-60% of the levels observed in column effluents fed with biofiltration influent water remained associated with the media and could be rinsed therefrom in either the dissolved or particulate form. Ozone pre-treatment significantly reduced NDMA precursor concentrations at one facility, suggesting that pre-oxidation could be an effective technique to mitigate the increase in NDMA precursor concentrations during biofiltration. Biofiltration decreased the concentrations of halogenated disinfection byproduct precursors.

Pilot-scale evaluation of oxidant speciation, 1,4-dioxane degradation and disinfection byproduct formation during UV/hydrogen peroxide, UV/free chlorine and UV/chloramines advanced oxidation process treatment for potable reuse.

Advanced oxidation using UV/free chlorine and UV/chloramines are being considered as alternatives to UV/H2O2 for treatment of reverse osmosis (RO) permeate in treatment trains for the potable reuse of municipal wastewater. This pilot-scale comparison of the three advanced oxidation processes (AOPs) evaluated three factors important for selecting among these alternatives. First, the study characterized the speciation of oxidants serving as the source of radicals within the AOPs to facilitate process modeling. Kinetic modeling that included consideration of the chloramines occurring in RO permeate accurately predicted oxidant speciation. Modeling of the UV/free chlorine AOP indicated that free chlorine is scavenged by reactions with ammonia and monochloramine in RO permeate, such that oxidant speciation can shift in favor of dichloramine over the short (∼30 s) timescale of AOP treatment. Second, the order of efficacy for degrading the target contaminant, 1,4-dioxane, in terms of minimizing UV fluence was UV/free chlorine > UV/H2O2 ≫ UV/chloramines. However, estimates indicated that the UV/chloramines and UV/H2O2 AOPs could be similar on a cost-effectiveness basis due to savings in reagent costs by the UV/chloramines AOP, provided the RO permeate featured >3 mg/L as Cl2 chloramines. Third, the study evaluated whether the use of chlorine-based oxidants within the UV/free chlorine and UV/chloramines AOPs enhanced disinfection byproduct (DBP) formation. Even after AOP treatment and chloramination, total halogenated DBP formation remained low at <15 µg/L for all three AOPs. DBP formation was similar between the AOPs, except that the UV/free chlorine AOP promoted haloacetaldehyde formation, while the UV/H2O2 and UV/chloramines AOPs followed by chloramination increased chloropicrin formation. However, total DBP formation on a toxic potency-weighted basis was similar among the AOPs, since haloacetonitriles and haloacetamides were the dominant contributors and did not differ significantly among the AOPs.

Transcriptome analyses of cells carrying the Type II Csp231I restriction-modification system reveal cross-talk between two unrelated transcription factors: C protein and the Rac prophage repressor.

Restriction-modification (R-M) systems represent an effective mechanism of defence against invading bacteriophages, and are widely spread among bacteria and archaea. In acquiring a Type II R-M system via horizontal gene transfer, the new hosts become more resistant to phage infection, through the action of a restriction endonuclease (REase), which recognizes and cleaves specific target DNAs. To protect the host cell's DNA, there is also a methyltransferase (MTase), which prevents DNA cleavage by the cognate REase. In some R-M systems, the host also accepts a cis-acting transcription factor (C protein), which regulates the counteracting activities of REase and MTase to avoid host self-restriction. Our study characterized the unexpected phenotype of Escherichia coli cells, which manifested as extensive cell filamentation triggered by acquiring the Csp231I R-M system from Citrobacter sp. Surprisingly, we found that the cell morphology defect was solely dependent on the C regulator. Our transcriptome analysis supported by in vivo and in vitro assays showed that C protein directly silenced the expression of the RacR repressor to affect the Rac prophage-related genes. The rac locus ydaST genes, when derepressed, exerted a toxicity indicated by cell filamentation through an unknown mechanism. These results provide an apparent example of transcription factor cross-talk, which can have significant consequences for the host, and may represent a constraint on lateral gene transfer.

Comparison of Toxicity-Weighted Disinfection Byproduct Concentrations in Potable Reuse Waters and Conventional Drinking Waters as a New Approach to Assessing the Quality of Advanced Treatment Train Waters.

Advanced treatment trains based on oxidation, biofiltration, and/or granular activated carbon (Ox/BAF/GAC) are an attractive alternative to those based on microfiltration, reverse osmosis, and advanced oxidation (MF/RO/AOP) for the potable reuse of municipal wastewater effluents, but their effluent quality is difficult to validate with respect to chemical contaminants. This study evaluated the sum of the concentrations of 46 disinfection byproducts (DBPs) after treatment by chlorine or chloramines weighted by metrics of toxic potency in 10 full- or pilot-scale reuse trains to estimate the DBP-associated toxicity of their effluents. These total toxicity-weighted DBP concentrations were compared to those measured in their local, conventional drinking waters as a benchmark for water quality receiving regulatory and widespread public acceptance. The results indicated that while the DBP-associated quality of MF/RO/AOP-based reuse waters can readily exceed that of drinking waters, that of Ox/BAF/GAC-based reuse waters can approach or exceed that of drinking waters, particularly when they are chloraminated. Unregulated, halogenated DBPs were the dominant contributors to the estimated DBP-associated toxicity. While RO/AOP treatment preferentially reduced the concentrations of the more toxic brominated DBP species, BAC and GAC treatment favored brominated DBP species by removing DOC but not bromide. Comparing the total toxicity-weighted DBP concentration between reuse and drinking waters provides drinking water as a rational benchmark for water quality comparison, explicitly recognizes that contaminants occur as mixtures, provides utilities flexibility in selecting the most efficient treatment trains to reduce estimated toxicity, and can be expanded to encompass new contaminants as toxic potency data become available.

Evaluation of a Pilot Anaerobic Secondary Effluent for Potable Reuse: Impact of Different Disinfection Schemes on Organic Fouling of RO Membranes and DBP Formation.

Anaerobic biological secondary treatment has the potential to substantially reduce the energy cost and footprint of wastewater treatment. However, for utilities seeking to meet future water demand through potable reuse, the compatibility of anaerobically treated secondary effluent with potable reuse trains has not been evaluated. This study characterized the effects of different combinations of chloramines, ozone, and biological activated carbon (BAC), applied as pretreatments to mitigate organic chemical fouling of reverse osmosis (RO) membranes, and the production of 43 disinfection byproducts (DBPs). The study employed effluent from a pilot-scale anaerobic reactor and soluble microbial products (SMPs) generated from a synthetic wastewater. Ozonation alone minimized RO flux decline by rendering the dissolved organic carbon (DOC) more hydrophilic. When combined with chloramination, ozone addition after chloramines maintained a higher RO flux. BAC treatment was ineffective for reducing the pressure and energy requirements for a set permeate flux. Regardless of pretreatment method prior to RO, the total DBP concentrations were <14 µg/L upstream of RO. After treatment by RO, the UV/hydrogen peroxide advanced oxidation process, and chloramination, the total DBP concentrations were ≤5 µg/L. When DBP concentrations were weighted by metrics of toxic potency, the total DBP calculated toxicity was 4-fold lower than observed previously in full-scale potable reuse facilities receiving aerobically treated secondary effluent. The RO fouling and DBP formation behavior of anaerobic SMPs were similar to that of the pilot-scale anaerobic effluent. The results of this study are promising, but more research is needed to evaluate whether anaerobic effluent is suitable as an influent to potable reuse trains.

Evaluation of Enhanced Ozone-Biologically Active Filtration Treatment for the Removal of 1,4-Dioxane and Disinfection Byproduct Precursors from Wastewater Effluent.

Ozonation followed by biologically active filtration (BAF) (O3-BAF) treatment has become an alternative to reverse osmosis in potable wastewater reuse applications because of the ability to produce a high-quality effluent while reducing brine production and disposal. In this study, effluent from a sequencing batch membrane bioreactor (SBMBR) was treated by O3-BAF at three specific ozone doses (0.5, 0.7, and 1.0 mg O3/mg DOC) and different empty bed contact times (EBCTs; 15-45 min). The reaction of O3 with granular activated carbon (GAC) (O3/GAC) to promote the formation of hydroxyl radicals (·OH) was evaluated at 1.0 mg O3/mg DOC followed by BAF at 15-45 min EBCT. The efficacy of these techniques was compared for the removal of O3 refractory 1,4-dioxane and the reduction in the formation of bromate, 35 regulated and unregulated halogenated disinfection byproducts (DBPs), and 8 N-nitrosamines after chloramination. Conventional ozonation (without the presence of GAC during ozonation) removed 6-11% of 1,4-dioxane, while BAF increased the removal to ∼25%. O3/GAC improved the removal of 1,4-dioxane to ∼40%, while BAF increased the removal to ∼50%. No bromate was detected during conventional ozonation. Although O3/GAC formed 12.5 µg/L bromate, this concentration was reduced during BAF treatment to <6.8 µg/L. Even though conventional ozonation was more effective than O3/GAC for the reduction in chloramine-reactive N-nitrosodimethylamine (NDMA) precursors, BAF treatment after either conventional or enhanced ozonation reduced NDMA formation during chloramination to <10 ng/L. O3/GAC was more effective at reducing halogenated DBP formation during postchloramination. Regardless, the reduction in halogenated DBP formation during postchloramination achieved by BAF treatment was ∼90% relative to the formation in the SBMBR effluent after either conventional or enhanced ozonation. The reduction of haloacetic acid (HAA) formation improved moderately with increasing BAF EBCT. Both O3-BAF and (O3/GAC)-BAF met regulatory levels for trihalomethanes, HAAs, NDMA, and bromate.

Pilot-scale comparison of microfiltration/reverse osmosis and ozone/biological activated carbon with UV/hydrogen peroxide or UV/free chlorine AOP treatment for controlling disinfection byproducts during wastewater reuse.

Ozone and biological activated carbon (O3/BAC) is being considered as an alternative advanced treatment process to microfiltration and reverse osmosis (MF/RO) for the potable reuse of municipal wastewater. Similarly, the UV/free chlorine (UV/HOCl) advanced oxidation process (AOP) is being considered as an alternative to the UV/hydrogen peroxide (UV/H2O2) AOP. This study compared the performance of these alternative treatment processes for controlling N-nitrosamines and chloramine-reactive N-nitrosamine and halogenated disinfection byproduct (DBP) precursors during parallel, pilot-scale treatment of tertiary municipal wastewater effluent. O3/BAC outperformed MF/RO for controlling N-nitrosodimethylamine (NDMA), while MF/RO was more effective for controlling N-nitrosomorpholine (NMOR) and chloramine-reactive NDMA precursors. The UV/H2O2 and UV/HOCl AOPs were equally effective for controlling N-nitrosamines in O3/BAC effluent, but UV/HOCl was less effective for controlling NDMA in MF/RO effluent, likely due to the promotion of dichloramine under these conditions. MF/RO was more effective than O3/BAC for controlling chloramine-reactive halogenated DBP precursors on both a mass and toxicity-weighted basis. UV/H2O2 AOP treatment was more effective at controlling the toxicity-weighted chloramine-reactive DBP precursors for most halogenated DBP classes by preferentially degrading the more toxic brominated species. However, the total toxicity-weighted DBP precursor concentrations were similar for treatment by either AOP because UV/H2O2 AOP treatment promoted the formation of iodoacetic acid, which exhibits a very high toxic potency. The combined O3/BAC/MF/RO train was the most effective for controlling N-nitrosamines and the total toxicity-weighted DBP precursor concentrations with or without treatment by either AOP.

Regulated and unregulated halogenated disinfection byproduct formation from chlorination of saline groundwater.

Coastal utilities exploiting mildly saline groundwater (<150 mg/L chloride) may be challenged by disinfection byproduct (DBP) formation, a concern likely to increase with sea-level rise. Groundwater from North Carolina coastal aquifers is characterized by large variations in concentrations of halides (bromide up to 10,600 µg/L) and dissolved organic carbon (up to 5.7 mg-C/L). Formation of 33 regulated and unregulated halogenated DBPs, including trihalomethanes (THMs), haloacetic acids (HAAs), haloacetonitriles, haloacetamides, and haloacetaldehydes, was measured after simulated chlorination of 24 coastal North Carolina groundwater samples under typical chlorination conditions. Results of chlorination simulation show that THM levels exceeded the Primary Maximum Contaminant Levels in half of the chlorinated samples. Addition of halides to a low salinity groundwater (110 mg/L chloride) indicated that elevated bromide triggered DBP formation, but chloride was not a critical factor for their formation. DBP speciation, but not overall molar formation, was strongly correlated with bromide variations in the groundwater. THMs and HAAs dominated the measured halogenated DBPs on a mass concentration basis. When measured concentrations were weighted by metrics of toxic potency, haloacetonitriles, and to a lesser degree, haloacetaldehydes and HAAs, were the predominant contributors to calculated DBP-associated toxicity. For some samples exhibiting elevated ammonia concentrations, the addition of chlorine to form chloramines in situ significantly reduced halogenated DBP concentrations and calculated toxicity. HAAs dominated the calculated toxicity of chloraminated waters. Reverse osmosis treatment of saline groundwater (chloride >250 mg/L) can reduce DBP formation by removing halides and organic precursors. However, we show that in a case where reverse osmosis permeate is blended with a separate raw groundwater, the residual bromide level in the permeate could still exceed that in the raw groundwater, and thereby induce DBP formation in the blend. DBP-associated calculated toxicity increased for certain blends in this system due to the DBPs resulting from the combination of the elevated bromide concentration in the permeate and the organic precursors from the raw coastal groundwater.

Effect of thiols, zinc, and redox conditions on Hg uptake in Shewanella oneidensis.

Mercury uptake in bacteria represents a key first step in the production and accumulation of methylmercury in biota. Previous experiments with mercury methylating bacteria have shown that Hg uptake is enhanced by some thiols, in particular cysteine, and that it is an energy-dependent process through heavy metal transporters [Schaefer et al. Environ. Sci. Technol. 2014, 48, 3007]. In this study, we examine Hg uptake in the nonmethylating facultative aerobe, Shewanella oneidensis, under both anaerobic and aerobic conditions. Our results demonstrate similar characteristics of the Hg uptake system to those of the Hg methylating strains: (1) uptake is enhanced in the presence of some thiols but not others; (2) uptake is energy dependent as evidenced by inhibition by a protonophore, and (3) uptake is inhibited by high Zn(II) concentrations. Initial cellular uptake rates in S. oneidensis were remarkably similar under aerobic and fumarate-reducing conditions. These data support a similar Hg(II) uptake mechanism within the proteobacteria of accidental Hg(II) transport through heavy metal transporters with similar rates of uptake but differences in the ability to take up Hg bound to different thiols.