Integrated Evaluation Concept to Assess the Efficacy of Advanced

Article pubs.acs.org/est

Integrated Evaluation Concept to Assess the Efficacy of Advanced Wastewater Treatment Processes for the Elimination of Micropollutants and Pathogens Thomas A. Ternes,*,† Carsten Prasse,†,‡ Christian Lütke Eversloh,† Gregor Knopp,§ Peter Cornel,§ Ulrike Schulte-Oehlmann,∥ Thomas Schwartz,⊥ Johannes Alexander,⊥ Wolfram Seitz,# Anja Coors,g and Jörg Oehlmann∥ †

Federal Institute of Hydrology (BfG), Am Mainzer Tor 1, D-56068 Koblenz, Germany Department of Civil & Environmental Engineering, University of California, Berkeley, 406 O’Brien Hall, Berkeley, California 94720, United States § Institute IWAR, Department Wastewater Technology and Water Reuse, Technische Universität Darmstadt, Franziska-Braun-Straße 7, D-64287 Darmstadt, Germany ∥ Department of Aquatic Ecotoxicology, Goethe University Frankfurt, 60438 Frankfurt am Main, Germany ⊥ Karlsruhe Institute of Technology (KIT)−Campus North, Institute of Functional Interfaces (IFG), Bioengineering and Biosystems Department, 76344 Eggenstein-Leopoldshafen, Germany # Zweckverband Landeswasserversorgung, 89129 Langenau, Germany g ECT Oekotoxikologie GmbH, 65439 Flörsheim, Germany ‡

S Supporting Information *

ABSTRACT: A multidisciplinary concept has been developed to compare advanced wastewater treatment processes for their efficacy of eliminating micropollutants and pathogens. The concept is based on (i) the removal/formation of selected indicator substances and their transformation products (TPs), (ii) the assessment of ecotoxicity via in vitro tests, and (iii) the removal of pathogens and antibiotic resistant bacteria. It includes substances passing biological wastewater treatment plants regulated or proposed to be regulated in the European Water Framework Directive, TPs formed in biological processes or during ozonation, agonistic/antagonistic endocrine activities, mutagenic/genotoxic activities, cytotoxic activities, further activities like neurotoxicity as well as antibiotics resistance genes, and taxonomic gene markers for pathogens. At a pilot plant, ozonation of conventionally treated wastewater resulted in the removal of micropollutants and pathogens and the reduction of estrogenic effects, whereas the in vitro mutagenicity increased. Subsequent post-treatment of the ozonated water by granular activated carbon (GAC) significantly reduced the mutagenic effects as well as the concentrations of remaining micropollutants, whereas this was not the case for biofiltration. The results demonstrate the suitability of the evaluation concept to assess processes of advanced wastewater treatment including ozonation and GAC by considering chemical, ecotoxicological, and microbiological parameters.

■

INTRODUCTION

Consequently, municipal WWTPs are one of the most crucial point sources for the contamination of receiving waters by micropollutants.4,5 Because of the large number of chemicals discharged into rivers and streams6 and ecotoxicological effects observed from WWTP effluents,7−10 there is concern regarding the impact on the aquatic environment. Furthermore, WWTPs have also been identified as important point sources for the emission of bacteria carrying clinically relevant antibiotic resistance genes into receiving waters.11−13 The WHO strongly recommends surveillance and control of the dissemination of

In the E.U., the assessment of municipal wastewater treatment plants (WWTPs) is currently based on compliance with five general wastewater parameters: biochemical oxygen demand in 5 days (BOD5), chemical oxygen demand (COD), total suspended solids (TSS), the sum of inorganic nitrogen, and total phosphorus.1,2 In the U.S., municipal criteria of WWTP secondary effluent quality additionally includes pH and total coliforms as regulated by the EPA under the Clean Water Act.3 The emissions of organic micropollutants, antibiotic-resistant bacteria and pathogens, as well as the resulting effects in receiving waters are not regulated at all. However, chemical monitoring clearly indicates that conventional wastewater treatment, e.g., activated sludge with nutrient removal, is unable to remove most organic micropollutants. © XXXX American Chemical Society

Received: October 10, 2016 Revised: October 28, 2016 Accepted: November 22, 2016

A

DOI: 10.1021/acs.est.6b04855 Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Article

Environmental Science & Technology

Figure 1. Schematic diagram of the pilot plant. Gray dots indicate sampling points.

of the implementation of advanced wastewater treatment processes (either ozone or activated carbon) is determined via the elimination of at least six indicator compounds.35−37 However, this does not take into account the formation of TPs, which can in some instances be even more toxic than the parent compounds.15 Because of these shortcomings of current approaches, the objective of the current study was to develop a multidisciplinary concept that allows for a comprehensive comparable assessment of different advanced wastewater treatment processes based on (i) the elimination of indicator substances as well as their transformation products, (ii) the reduction of in vitro toxicity, and (iii) the removal of pathogens and clinically relevant antibiotic resistance genes.

resistant microorganisms to address the growing problem of infections by multiresistant pathogens.14 Another major challenge is that the overall load and composition of micropollutants entering WWTPs are sitespecific and vary considerably depending on the discharger within the catchment area.15 Therefore, treatment processes have to be tailored toward individual demands. Today, advanced wastewater processes such as ozonation, dosing of powdered-activated carbon (PAC), or granular-activated carbon filtration (GAC) are the most promising options to tackle new challenges related to micropollutants, their transformation products (TPs), pathogens, and antibiotic resistance genes.16,17 However, most assessment approaches that evaluate wastewater treatment processes still do not consider transformation products and ecotoxicological in vitro assays in combination with microbiological criteria. Approaches that have combined chemical and ecotoxicological assessment have so far primarily focused on the analysis of endocrine-disrupting chemicals and their correlation with overall activities via in vitro assays. These include estrogens, antiestrogens, androgens, antiandrogens, glucocorticoids, and progestagens.18−24 For glucocorticoids, most of the activity observed in in vitro assays could be explained by the presence of known chemicals,25 whereas for other end points, such as estrogenic activity, the gap between in vitro activities and detected concentrations of known estrogens indicates the presence of other to date unknown bioactive compounds.26 Advances in analytical chemistry and in particular “omic” technologies such as metabolomics have led to the development of methodologies that study the effects of micropollutants on a whole organism level in vivo.27−29 Even though these approaches are still very time and labor intensive, future improvements might result in more widespread usage for the assessment of wastewater quality. Efforts to use standard wastewater parameters such as BOD5, COD, and TSS as indicators for ecotoxicological effects indicate that these are not suitable to assess overall water quality.30,31 As an alternative, the analysis of indicator compounds, which are used as representatives of compounds with environmental relevance due to their incomplete removal in conventional biological treatment, is currently one of the main measures being used to evaluate advanced wastewater treatment technologies.16,32−34 In Switzerland, for example, the efficacy

■

MATERIALS AND METHODS A pilot plant operated for 618 days was fed by the secondary effluent of a municipal WWTP connected to approximately 42,000 person equivalents.38 The WWTP effluent was filtered by a microscreen (10 μm mesh size) and subsequently directed into an ozonation system coupled to two parallel operating granular activated carbon (GAC) filters or two biological filters (BF). The ozonation system consisted of two bubble columns connected in series, an equalization tank, and an ozone generator (WEDECO GSO 10, Xylem Water Solutions Herford, Herford Germany). GAC filters and the BF were identically designed and were top-down fed and bottom-up back washed. The filters contained either a 2 m GAC layer (tapped state) or an expanded clay layer. GAC filter 1 and BF 1 were aerated (ae) by compressed ambient air (air velocity of 2 m/h). GAC filter 2 and BF 2 were nonaerated (nae). The efficacy of the ozonation and post-treatment steps were optimized prior to application, leading to a specific ozone consumption of 0.98 ± 0.24 gO3/gDOC and a hydraulic retention time in the bubble columns of 17 ± 3 min (average flow rate: 0.8 m3/h). The filter systems were operated similarly with a filter velocity of approximately 3.9−4.8 m/h and an empty bed contact time of 28−34 min, achieving a net specific throughput of approximately 24,000−27,000 m3/m3 bed volumes. Further details are described in Knopp et al. (2016).38 All process parameters were selected according to previous studies,39−42 resulting in the appropriate removal of many micropollutants. B


Article


5% B, and 6.5 min for equilibration at 5% B. The flow rate was constant at 0.6 mL/min, and the column temperature was 25 °C. Electrospray ionization was operated in positive and negative mode. The following parameters were set: curtain gas, 30 psi; temperature, 600 °C; ion source gas 1, 60 psi; ion source gas 2, 60 psi; declustering potential, 100 V/−100 V; entrance potential, 10 V/−10 V; and ion source, 5000 V/− 4200 V. Analysis was performed by direct injection of 100 μL of water sample. The analytical parameters (e.g., precursor ions, product ions, and collision energy) of the analyzed organic micropollutants as well as their limit of quantification (LOQs) are listed in Table S2. Selection of Ecotoxicological Parameters. For the effect-based evaluation, the results of in vitro tests were used exclusively, although chronic in vivo tests with primary producers and representatives of the most important invertebrate groups in aquatic ecosystems were also conducted at the pilot plant with continuous flow-through exposures (see chapter 2.3 of the Supporting Information and Table S3). However, it turned out that some of the selected test species used for in vivo testing sensitively reacted to nutrients (nitrogen compounds, phosphorus), elevated salt contents, as well as to suspended matter. These reactions may modulate and even mask the impact of micropollutants. Without additional tests, it is in this case impossible to clearly elucidate whether the impact on the tested parameters (e.g., biomass, growth, and reproduction) is caused by the elimination of toxic substances/ formation of potentially toxic TPs or by other properties of the wastewater (e.g., nutrient content, salt content, temperature, pH). Hence, the in vivo tests could not be considered in the multidisciplinary concept as their outcome was not exclusively linked to the efficiency of the treatment processes. The main advantages of in vitro tests for the assessment of wastewater are that the tests are sensitive and can be applied easily and at low cost and for a specific mechanism of action. They are therefore suitable for routine testing. On the other hand, the ecological relevance of in vitro tests is limited because positive results indicate a toxic potential of the analyzed sample but do not provide sufficient evidence for a toxic effect in intact organisms per se.15 However, the main strength can be seen in the combination with the analysis of micropollutants. The response of a bioassay and the occurrence of micropollutants known to be covered by those bioassays will be an ideal supplement. For the assessment concept, primarily standardized in vitro testing methods were applied. In case no standardized method was available, tests protocols with good methodological documentation were used. The test procedures used in this study have proven successful due to their robustness, their high degree of standardization and, as a consequence, the reproducibility of the results. Therefore, they are generally recommended for the assessment of wastewater treatment plants. For projects with different objectives, and depending on the wastewater samples to be analyzed as well as the methodological expertise of the laboratories involved, other test procedures can also be used to investigate the same mechanisms of action. However, these optional procedures should exhibit a comparable sensitivity, robustness, and standardization level as the in vitro tests that have been used in this study. The assessment was based on the results of SPEenriched samples to improve limits of detection. Thus, the in vitro tests are targeted to those substances that are enriched via

Sampling devices were installed before and after the WWTP, the ozone system, and the filtration systems (Figure 1). Composite samples (24 h) were taken constantly in 5 L glass bottles with dosing pumps and a hose clamp throttled sample (free of DEHP). Prior to sampling, the glass bottles were purified by rinsing with a basic and acidic solution, drying (105 °C), rerinsing with acetone (picograde), and heating (200 °C) for 6−8 h. During sampling, glass bottles were stored in cooling boxes cooled with thermal packages. For chemical and in vitro analysis all samples were stored in the dark at 3 °C until analysis (within 1 week) after sampling. For microbiological analyses 24h-composite samples were taken, filtrated, and subsequent DNA extractions were performed, immediately. However, the selected HRTs and EBCTs of the pilot system are sufficiently short with respect to the sampling mode; therefore, there should be no concern that the sampling scheme does not account for the residence time in the treatment units, leading to a mismatch of data between the influent and effluent, as discussed by Ort et al.43 Selection of Chemical Parameters. The chemical parameters have been selected according to the following criteria: (1) Substances that are present in secondary effluents with appreciable concentrations (higher ng/L to μg/L range) at least ten times the limit of quantification over the whole year. (2) Persistent substances that are not or are hardly biodegradable during biological wastewater treatment. (3) Substances with ecotoxicological relevance and/or regulated in the Water Framework Directive (WFD) or proposed for regulation. (4) Substances with a high potential to contaminate groundwater and drinking water. (5) TPs formed during biological wastewater treatment. (6) TPs formed during ozonation. (7) Availability of robust analytical methods. Analytical Methods to Analyze the Selected Indicator Substances. Detection and quantification of indicator substances and their TPs was conducted by LC−MS/MS according to Rühmland et al. (2015), Kaiser et al. (2014), Funke et al. (2015), and Bollmann et al. (2016).44−47 After chromatographic separation, the targets where analyzed by scheduled multiple reaction monitoring (sMRM) using electrospray ionization (ESI) in both positive and negative polarization. At least two mass transitions were measured for quantification and confirmation. Quantification was performed using at minimum a 10 point calibration curve of the target substances and their surrogate standards (deuterated, 13C-, and 15 N-labeled compounds) where available. Analyses of chemical compounds followed the German norms DIN 38407-36:201409 and DIN 38407-47:2015-07. The high-performance liquid chromatography system UltiMate RSLC 3000 (Thermo, Dreieich, Germany), and the mass spectrometer Qtrap 5500 (SCIEX, Darmstadt, Germany) were used. The HPLC column Kinetex C18, 2.6 μm, 100 A, 100 mm × 4.6 mm (Phenomenex, Aschaffenburg Germany) was used. Eluent A was an aqueous solution of 0.1% (v/v) formic acid and 2 mmol/L ammonium formate, and eluent B was acetonitrile with the addition of 5% water, 0.1% (v/v) formic acid, and 2 mmol/L ammonium formate. A multistep gradient with the following parameters was applied: 5 min at 5% B, within 5 min to 20% B, within 5 min to 80% B, hold for 3 min at 80% B, within 0.5 min back to C


Article


Wells with an optical density >0.6 were considered as revertant wells. Selection of Microbiological Parameters. The selection of microbiological parameters was based on health relevance and the likelihood of abundances of microbiological targets: • Enterococci and vancomycin resistance (vanA)53−55 • Imipenem resistance in Pseudomonas aeuginosa (blaVIM)56 • Erythromycin resistance (ermB)57 • Enterobacteria and ampicillin resistance (ampC)58−60 • Pseudomonas aeruginosa61,62 • Staphylococcus aureus and CNS63,64 Methodology to Determine Microbiology Parameters. For separating the surviving bacterial population from inactivated or injured bacteria after ozone treatment, the propidium monoazid method was used to block DNA from dead or injured bacteria and free DNA for qPCR amplification. For that, the PhAST Blue Photo-Activation System (GenIUL, Barcelona, Spain) was used. The protocol is according to Varela-Villarreal et al.65 and Nocker et al.,66−68 where the method was also optimized for natural mixed populations (see Supporting Information for further details). The filtration and DNA extraction of the wastewater samples as well as the quantification of ARGs and opportunistic microorganisms via SYBR Green qPCR were performed according to Alexander et al.12 Data analysis of qPCR amplification was performed using Bio-Rad CFX Manager software. For calculating the gene abundance of the respective antibiotic resistance gene and taxon-specific gene marker, reference strains carrying the genetic targets of interest were used. Serial dilutions of reference strains were made to determine the correlation between plate count experiments and Ct values. The obtained data were used for calculation of the calibration curves. Using these curves, the measured Ct values of antibiotic resistance genes or taxon-specific gene markers from water samples can be converted into cell equivalents.12,69 The coefficient of determination of the standard curves was above 0.979 in all experiments, indicating minimal variability within the linear data range. The numbers of cell equivalents in each wastewater sample were derived using the corresponding calibration curve and normalized per 100 ng of total isolated DNA (cell equivalents per 100 ng of DNA). Hence, we calculated the relative abundance with respect to the population size. Additionally, the absolute abundances of the respective cell equivalents in each wastewater sample were derived using the corresponding calibration curve and normalized to 100 mL sample volumes (cell equivalents per 100 mL of wastewater sample). Further details of the microbial methodologies are provided in chapter 3.1 of the Supporting Information, such as the list of primers used to detect pathogens, antibiotic resistance genes, the amplicon sizes, accuracies of the calibration curves, as well as a comparison of TaqMan and SybrGreen analyses.

the used SPE procedure, while results are not influenced by other water properties such as pH, temperature, nutrients or salt contents like in the in vivo experiments. Only if the analysis of a mechanism of action shows that the causative substances cannot or are insufficiently enriched on SPE cartridges should the assessment instead be carried out based on the analytical results of native samples. Methodology of Testing via in Vitro Bioassays. The in vitro testing was conducted according to Magdeburg et al. (2014) and Stalter et al. (2011).48,49 For in vitro bioassays, 100 mL of the wastewater samples were acidified to pH 2 with HCl and extracted using Telos C18_ENV+ cartridges (Abimed GmbH, Langenfeld, Germany). Loaded cartridges were dried under a stream of nitrogen. For analysis, cartridges were eluted with 6 mL of methanol (MeOH) and 6 mL of methyl-tertbutyl-ether (MTBE). The MTBE and MeOH eluates were merged, and 500 mL of dimethyl sulfoxide (DMSO) was added for solvent exchange. The eluates were concentrated to a final volume of 500 μL under a stream of nitrogen to provide 1000fold concentrated WW extracts solved in DMSO. DMSO solvent exchange was applied to minimize the loss of volatile substances during the test procedure.49 All yeast screens in this study (YAAS, yeast antiandrogen screen; YAES, yeast antiestrogen screen; YAS, yeast androgen screen; YES, yeast estrogen screen; and YDS, yeast dioxin screen) were conducted in 96-well microtiter plates loaded with DMSO extracts of the sample providing a 2-fold sample concentration per well. The assay procedure and data analysis were conducted as described previously.49 The Microtox assay or bioluminescence inhibition test with the bacterium Aliivibrio f ischeri (former Vibrio fischeri) was conducted to assess cytotoxicity. The assay was performed according to the standard operating procedure of the International Organization for Standardization50 modified to a 96-well plate format as previously described.51,52 In brief, controls (negative/solvent), SPE-Blank, reference compound (3,5-dichlorophenol), and SPE extracts were serially diluted (1:2) in saline buffer. Then, 100 μL of each sample was added to 50 μL of A. f ischeri solution (not exceeding 1% DMSO in the final medium volume). For inhibition to be detected, luminescence was measured prior to sample addition and after 30 min incubation using a microplate reader (Spark10M, Tecan, Crailsheim, Germany). Mutagenicity was assessed in the same sample extracts using the Ames fluctuation test without metabolic activation with the Salmonella typhimurium strains TA98 and YG7108 for detecting frameshift mutations and base pair transitions according to Magdeburg et al.48 The strain YG7108 is sensitive to mutagenic alkylating agents and nitrosamines that can be formed during ozonation. Overnight cultures were diluted to 180 (TA98) and 250 (YG7108) formazine attenuation units in the assay medium. Salmonella strains were exposed for 100 min to extracted water samples (as 2% DMSO) in triplicates in 24-well microplates. Propylene oxide (0.2%, YG7108) and 4-nitro-ophenylenediamine (10 mg/mL, TA98) served as positive controls. After incubation, the exposure media were diluted by 6-fold with histidine-deficient minimal medium and indicator dye bromocresol purple. Cultures were transferred to 384-well plates (48 wells per replicate) and incubated for 48 h at 37 °C. Growth of revertants led to pH reduction in the respective wells causing a color change in the indicator dye, which was detected with a multiwell plate reader at 414 nm.

■

RESULTS AND DISCUSSION To address the overall quality of treated wastewater by advanced treatment processes, a multidisciplinary approach was developed including chemical, ecotoxicological, and microbiological measurements. The approach was tested for ozonation and the use of activated carbon. However, the concept is highly flexible and can therefore also be used for D


Article

Environmental Science & Technology other processes such as membrane filtration or AOPs. However, in contrast to ozonation, no current oxidation products of micropollutants are known for AOPs. Therefore, category D (see section ‘Chemical Assessment’) cannot be used. Furthermore, the concept allows for integrating new micropollutants, pathogens, and/or further in vitro bioassays. Chemical Assessment. Exclusion Criteria. In receiving rivers and streams, the exceedance of environmental quality standards (EQS) from the Water Framework Directive (WFD) for micropollutants and the exceedance of the thresholds for bromate and NDMA from the drinking water directive have to be avoided. Thus, in the E.U. WWTP effluents should not exceed the EQS and drinking water thresholds unless this is permitted by water authorities to a defined factor due to elevated dilution in the receiving waters and the limited discharge of those compounds by other sources. If that cannot be guaranteed, the technical concept of the respective WWTP has to be improved with regard to the removal of those substances that exceed discharge regulations. If these exclusion criteria are not fulfilled, indicator substances are suggested to assess the process efficacy for the removal of organic micropollutants in comparison to that of conventionally treated wastewater. Indicator substances (Table S1) are grouped in four categories (A−D), and the suggested list of indicator substances should be regularly updated based on new scientific findings, revised regulations, and regional contaminations. The chemical evaluation is based on the removal of individual micropollutants/TPs (eq 1) and the formation of TPs in ozonation (eq 2). For at least six substances of each category A−C, an average removal Rc (X) (eq 3), and for at least four TPs of category D, an average formation Fc (D) (eq 4), is calculated. The removal refers to the concentration c0 present in the conventionally treated wastewater. It should be noted that the number of available indicator substances for category D is lower than for the other categories due to limited knowledge about ozonation products that are permanently formed. As the fifth category, the DOC removal Rc(DOC) is included. Finally, a chemical assessment index (CAI) summarizes the outcomes of all categories (eqs 5 and 6). Higher CAI results in higher average removal of the indicator substances/TP and lower potential to form ozonation TPs. The following indicator substances were proposed for categories A−D due to (i) available analytical methods, (ii) the potential to contaminate groundwater and drinking water, and (iii) their potential (eco)toxicological relevance, which is generally given for antibiotics due to the potential development of resistance13,70 as well as for the drinking water carcinogens bromate and NDMA71−74 (≥6 from each cat. A−C and ≥4 from cat. D; a: potential to be present in drinking water; b: potentially of (eco)toxicological relevance; see Table S1): A. Substances Not or Hardly Biodegradable during Biological Wastewater Treatment. Carbamazepinea,b, tramadola, venlafaxinea,b, primidona, diatrizoatea, sotalola,b, metoprolol, benzotriazolea,b, tolyltriazolea,b, sulpiride, amisulpride, acesulfamea, lamotrigine, sucralose, citalopram, candesartanb, irbesartanb. B. Substances Regulated in the WFD or Proposed for Regulation. Sulfamethoxazolea,b, diclofenacb, trimethoprimb, irgarolb, terbutrynb, clarithromycinb, roxithromycinb, azithromycinb, PFOSa,b, 17α-ethinylestradiolb, 17β-estradiolb. C. TPs Formed in Biological Wastewater Treatment. Carboxy-acyclovira,b, carboxy-emtricitabinea, carboxy-abacavira, carboxy-lamivudinea, oxypurinola, carboxy-acridinea,b, valsartan

acida, iopromide-TP643a, iopromide-TP701Aa, iopromideTP701Ba, diclofenac-lactam, carboxy-diclofenac. D. TPs Formed during Ozonation. Sulpiride-N-oxidea, amisulpride-N-oxidea, lamotrigine-N-oxidea, COFAa,b, tramadol-N-oxide, venlafaxine-N-oxide, BQDa, BQMa, bromatea,b, NDMAa,b. For normalization, the precursor compound Si needs to be measured as well. Bromate and NDMA are normalized by the thresholds for drinking water with 10 μg/L and 10 ng/L, respectively. Removal (Ri) of the individual indicator substance (Si) and the formation (Fi) of the transformation product (TPi) are calculated according to eqs 1 and 2 ⎛ cS (advanced treatment) ⎞ ⎟ × 100 [%] R i = ⎜⎜1 − i cSi(WWTPeffluent ) ⎟⎠ ⎝

(1)

⎛ c TP(ozonation) ⎞ i ⎟⎟ × 100 [%] Fi = ⎜⎜ c (WWTP ⎝ Si effluent ) ⎠

(2)

where cSi (WWTP effluent) the measured concentration of substance S i in the conventional treated wastewater, c S i (advanced treatment) the measured concentration of substance Si after advanced treatment, and cTPi (ozonation) the measured concentration of TPi after ozonation. Average removal Rc (X) of selected indicator substances/TPs (cat. A−C) and formation Fc (D) of TPs (cat. D) relative to the effluent without advanced treatment are calculated according to eqs 3 and 4 n

Rc(X ) =

∑i R i(X )

Fc(D) = −

(3)

n n ∑i Fi(D)

(4)

n

where X is cat. A−C, and n is the number of indicator substances used per category. The CAI is calculated according to eq 5 that is foreseen for advanced treatment without ozonation (e.g., only GAC) CAI =

Rc(A) + Rc(B) + Rc(C) + Rc(DOC) 4

(5)

and eq 6 that is foreseen for advanced treatment without ozonation. CAI =

Rc(A) + Rc(B) + Rc(C) + Fc(D) + Rc(DOC) 5 (6)

which is advanced treatment including ozonation. Ecotoxicological (Effect-Based) Assessment Using in Vitro Tests. In vitro assays according to standardized ISO guidelines or Standard Operating Procedures (SOP) are applied in the ecotoxicological assessment scheme. If no appropriate procedure is available for a relevant mechanism of action, tests with good methodological documentation can also be used. The assessment is based on SPE-enriched samples to ensure sufficient sensitivity. However, small polar compounds may be lost during SPE. Five categories (E−I) of specific mode-of-action groups are considered in the assessment concept: E,F. Agonistic (E) and antagonistic (F) endocrine activities. Activities at the estrogen receptor (ER) α and the androgen E


Article

Environmental Science & Technology receptor (AR) can be analyzed with recombinant yeast reporter gene assays with proliferation assays (e.g., E-screen) or cell linebased reporter gene assays (e.g., ER-Calux, AR-Calux). The agonistic and antagonistic activities contribute to the ecotoxicological assessment with a share of 15% each (30% total) regardless of the number of tests. G. Mutagenic/Genotoxic Activities. Mutagenic/genotoxic activities are detected using the Ames fluctuation test (ISO 11350, 2012). In addition or alternatively, other genotoxicity assays (e.g., umu test, Comet assay, micronucleus test) can be used. The mutagenic/genotoxic activity constitutes 40% of the ecotoxicological assessment regardless of the number of tests. H. Cytotoxic Activities. Cytotoxic activities are detected using mammalian (e.g., GH3) or other vertebrate cell lines (e.g., RTL-W1) or bioluminescence inhibition in luminescent bacteria.51,52 Cytotoxicity constitutes 15% of the ecotoxicological assessment. I. Additional Activities. Additional activities can be taken into account, e.g., dioxin-like (e.g., yeast dioxin screen), neurotoxic (e.g., inhibition of acetylcholinesterase), and/or phytotoxic effects (e.g., inhibition of photosystem II), and contribute to 15% of the ecotoxicological assessment. If one or more categories are not considered, the proportions of the remaining categories to the total assessment rise proportionally. As a minimum requirement, endocrine and mutagenic/genotoxic activities (categories E−G) have to be covered. The in vitro test results are reported as relative change of activity Ai [%] in the advanced treated wastewater compared to conventional effluent according to eq 7 ⎛ a (advanced treatment) ⎞ A i (X ) = ⎜1 − i ⎟ × 100 [%] a i(WWTPeffluent ) ⎠ ⎝

where AF is the assessment factor for cat. E−I, and EAI is the effect-based assessment index. Microbiological Assessment. The microbiological assessment of the treated water monitors the alteration of different microbial parameters that are known to be crucial for human health and thus cause substantial financial expenses for healthcare facilities.75−77 It contains two main criteria that are determined via DNA-based qPCR: removal/increase of (i) pathogens and (ii) antibiotic resistance genes with clinical relevance. For both, the absolute (100 mL sample volume) and relative (referred to 100 ng of DNA of the microbial population) abundances are considered to calculate the process efficacy. Currently, there are no regulations based on the abundance of microbial parameters to monitor WWTP effluent. However, the combination of volume- and population-based abundances is a novel approach and can help to identify the microbial selection processes as well as to evaluate the regrowth potential of crucial microbial parameters in downstream aquatic systems. The microbial assessment index (MAI) is derived from microbial parameters (on the basis of absolute as well as relative abundances; see also the Supporting Information) grouped in categories J and K. Higher MAI results in higher average removal of pathogens and antibiotic resistance genes. The selected microbiological parameters targeting clinically relevant antibiotic resistance genes and marker genes for pathogens can be extended (e.g., ESBL-forming pathogens, quinoloneresistant bacteria, Klebsiella spp., Acinetobacter spp.). Because of their clinical relevance (see Supporting Information), the removal of the following selected healthrelated microbial parameters were grouped into two categories: J. Removal of Antibiotic Resistance Genes. 1. vanA (vancomycin resistance in enterococci) 2. blaVIM (imipenem resistance in Pseudomonas aeruginosa) 3. ampC (ampicillin resistance in Enterobacteriaceae) 4. ermB (erythromycin resistance in Streptococcus spp.) K. Removal of Taxonomic Gene Markers Referring to Pathogens (Also Called Opportunistic Bacteria). 5. enterococci 6. Pseudomonas aeruginosa 7. staphylococci 8. enterobacteria Separately for absolute as well as for relative abundances,78 these eight parameters are transferred to assessment factors (AF) ranging from −1 (strong increase) to 1 (high removal). • removal ≥ 99% = +1.0 AF • 40% < removal < 99% = +0.5 AF • change ± 40% = 0.0 AF • 40% < increase ≤ 5-fold = −0.5 AF • increase > 5-fold = −1.0 AF The individual removal (Ri(X)) of microbial parameters is calculated according to

(7)

where ai is the measured activity in the in vitro assay I, and X is cat. E−I. These activity changes are converted to assessment factors (AF): • removal > 80% = +1 AF • 20% < removal ≤ 80% = +0.5 AF • change < ±20% = 0 AF • 20% < increase ≤ 100% = −0.5 AF • increase >100% = −1 AF If several tests are used in one category in parallel, the assessment is performed on a worst-case basis, i.e., the test with the most negative result (lowest activity removal or highest increase) sets the AF. As long as the assays reflect different receptor types in endocrine categories E and F, the result is calculated as an average value for the agonistic (E) and antagonistic activity (F). The ecotoxicological assessment of treated wastewater samples considers the AFs of all categories after offsetting against the corresponding assessment factor (15% each for cat. E, F, H, and I; 40% for cat. G). Category I contributes with a higher share of 40% to the overall assessment because mutagenic/genotoxic activities represent an undesirable trait of treated wastewater and pose a specifically high risk for the aquatic environment (G). The results are transformed into an effect-based assessment index (EAI) with a value of 100 showing strong reduction of activities for all activity groups in comparison to a conventional treatment, and a value of −100 showing a strong increase (eq 8)

R i(X ) =

AFrel + AFabs 2

(9)

where AFrel is the relative abundance of pathogens/antibiotic resistance bacteria, and AFabs is the absolute abundance of pathogens/antibiotic resistance genes. The average removal in categories J and K is calculated according to eq 10

EAI = (0.15AFE + 0.15AFF + 0.4AFG + 0.15AFH + 0.15AFI) × 100

(8) F


G

0.04 ± 0.01 0.020 ± 0.003 0.08 ± 0.02 0.84 ± 0.26 4.2 ± 0.8 0.01b 0.09 ± 0.02 0.02b 0.02b

0.01b 0.01b 0.01b 0.01b 0.01b 0.01b 0.02b

0.02b 0.01b 1.6 ± 0.2 0.01b 0.01b 0.05 ± 0.02

0.02b 0.05 ± 0.02 0.02b 3.3 ± 0.7 0.02b 0.10 ± 0.02 0.57 ± 0.11 0.04 ± 0.008

5.6 ± 0.6 1.0 ± 0.1 3.3 ± 0.6 2.2 ± 0.7 4.4 ± 0.9 1.0 ± 0.1 0.65 ± 0.13 0.57 ± 0.06 0.36 ± 0.09

0.040 ± 0.006 0.060 ± 0.015 0.070 ± 0.015 0.12 ± 0.02 0.050 ± 0.012 0.030 ± 0.004 1.4 ± 0.2

3.1 ± 0.5 0.18 ± 0.04 17 ± 3 0.36 ± 0.10 0.12 ± 0.02 0.53 ± 0.12

0.53 ± 0.11 0.05 ± 0.02 3.12 ± 0.37 0.05b 0.33 ± 0.07 0.020 ± 0.004 1.68 ± 0.34 0.020 ± 0.004

benzotriazole tolyltriazole acesulfame sucralose diatrizoate carbamazepine iopamidol sotalol primidone average category A

clarithromycin roxithromycin trimethoprim mecoprop N-acetyl-smx sulfamethoxazole diclofenac average category B

carboxy-acyclovir carboxy-lamivudine oxypurinol carboxy-emtricitabin carboxy-abacavir carboxy-acridine average category C

tramadol tramadol-N-oxide carboxy-acyclovir COFA sulpiride sulpiride-N-oxide lamotrigine lamotrigine-N-oxide average category D 2.4 37

30

106

9.4

99 94 90 97 92 91 94

99 86

75 83 86 92 78

99 98 98 62 5 99 86 96 94 82

elim. /form. [%]

c

d

0.02b 0.02b 0.02b 3.1 ± 0.8 0.02b 0.02b 0.02b 0.02b

0.02b 0.01b 0.025 0.01b 0.01b 0.01b

0.01b 0.01b 0.01b 0.01b 0.01b 0.01b 0.02b

0.02b 0.02b 0.05 ± 0.01 0.26 ± 0.08 2.0 ± 0.4 0.01b 0.08 ± 0.02 0.02b 0.02b

c [μg/L]

c [μg/L]

75 83 86 92 78

1.1 24

6.0

86

3.8

99 94 100 97 92 97 97

99 86

75 83 86 92 78

99 98 98 82 41 99 95 96 94 89

elim. /form. [%]

c

ozone/GACae

Category A 0.02b 0.02b 0.08 ± 0.02 0.4 ± 0.1 2.6 ± 0.5 0.01b 0.03 ± 0.01 0.02b 0.02b

d

Category B 0.01b 0.01b 0.01b 0.01b 0.01b 0.01b 99 0.02b 86 Category C 99 0.02b 94 0.01b 100 0.03 ± 0.01 97 0.01b 92 0.01b 98 0.02 ± 0.01 97 Category D 0.02b 3.8 0.02b 0.02b 99 2.7 ± 0.8 0.02b 6.0 0.02b 0.02b 1.1 0.02b 27

100 98 98 88 55 99 88 96 94 91

elim. /form. [%]

c

ozone/GACnae d

0.02b 0.05 ± 0.01 0.02b 2.6 ± 0.7 0.02b 0.10 ± 0.02 0.63 ± 0.13 0.02b

0.02b 0.01b 1.3 ± 0.2 0.01b 0.01b 0.11 ± 0.03

0.01b 0.01b 0.01b 0.01b 0.01b 0.01b 0.02b

0.030 ± 0.003 0.020 ± 0.003 0.07 ± 0.014 0.86 ± 0.27 3.0 ± 0.6 0.01b 0.05 ± 0.01 0.02b 0.02b

c [μg/L]

c

1.1 31

30

85

9.4

99 94 92 97 92 79 92

99 86

75 83 86 92 78

99 98 98 61 32 99 92 96 94 85

elim. /form. [%]

ozone/BFnae d

0.02b 0.05 ± 0.01 0.02b 2.6 ± 0.5 0.02b 0.10 ± 0.02 0.69 ± 0.14 0.02b

0.02b 0.01b 1.4 ± 0.3 0.01b 0.01b 0.12 ± 0.03

0.01b 0.01b 0.01b 0.01b 0.01b 0.01b 0.02b

0.030 ± 0.001 0.02b 0.06 ± 0.01 1.2 ± 0.4 4.1 ± 0.8 0.01b 0.08 ± 0.02 0.02b 0.02b

c [μg/L]

1.1 31

30

85

9.4

99 94 92 97 92 77 92

99 86

75 75 86 92 78

99 98 98 45 7 99 88 96 94 80

elim.c/form.d [%]

ozone/BFae

Concentration in feed water. bConcentration below LOQ: LOQ was used for calculations. Concentrations are given with expanded uncertainty. cElimination calculated according to eq 1 (cat. A-C). Formation calculated according to eq 2 (cat. A-C).

d

a

c [μg/L]

c0a [μg/L]

ozone

Table 1. Concentrations of Target Substances and Calculation of Elimination Rates

Environmental Science & Technology Article


Article

Environmental Science & Technology Table 2. Activities Measured for the Analyzed Wastewater Samples with the Different in Vitro Assaysa treatment conventional treatment ozone ozone/GACnae ozone/GACae ozone/BFnae ozone/BFae

YES [ng EEQ/L]

YAES [μg OHTEQ/L]

YAS [ng TEQ/L]

8.39

446

92.0

0.705 0.837 0.916 0.836 0.876

1413 475 621 1856 1729

72.5 34.0 54.1 53.9 19.1

YAAS [μg FluEQ/L]

YDS [μg β-NFEQ/L]

microtox assay (% inhibition)

Ames YG7108 (% revertants)

Integrated Evaluation Concept to Assess the Efficacy of Advanced

Recommend Documents