- Email: [email protected]
Removal of emerging organic contaminants with a pilot-scale biofilter packed with natural manganese oxides
Accepted Manuscript Removal of emerging organic contaminants with a pilot-scale biofilter packed with natural manganese oxides Yongjun Zhang, Hong Zhu, Ulrich Szewzyk, Sven Lübbecke, Sven Uwe Geissen PII: DOI: Reference:
S1385-8947(17)30258-9 http://dx.doi.org/10.1016/j.cej.2017.02.095 CEJ 16534
To appear in:
Chemical Engineering Journal
Received Date: Revised Date: Accepted Date:
5 October 2016 24 December 2016 17 February 2017
Please cite this article as: Y. Zhang, H. Zhu, U. Szewzyk, S. Lübbecke, S. Uwe Geissen, Removal of emerging organic contaminants with a pilot-scale biofilter packed with natural manganese oxides, Chemical Engineering Journal (2017), doi: http://dx.doi.org/10.1016/j.cej.2017.02.095
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Removal of emerging organic contaminants with a pilot-scale biofilter packed with natural manganese oxides
Yongjun Zhanga,b,*, Hong Zhuc, Ulrich Szewzykc, Sven Lübbecked, Sven Uwe Geissenb
a, School of Environmental Sciences and Engineering, Nanjing Tech University, Nanjing 211800, PR China b, Technische Universität Berlin, Chair of Environmental Process Engineering, Sekr. KF2, Strasse des 17.Juni 135, 10623 Berlin, Germany c, Technische Universität Berlin, Chair of Environmental Microbiology, Sekr. BH 6-1, Ernst-Reuter-Platz 1, 10587 Berlin, Germany d, Waterleau Deutschland GmbH, Hemelinger Hafendamm 18, 28309 Bremen, Germany
*, Corresponding author Tel.: +86 (0)25 58139656 Fax: +86 (0)25 58139960 E-mail: [email protected]
1
Abstract Emerging
organic
contaminants
(EOCs),
consisting
of
pharmaceuticals,
industrial/household additives, and their transformation products, have been widely detected in surface water bodies and may lead to potential ecological risks. Installing a tertiary treatment step in sewage treatment plants (STP) is an effective method to control their contamination. In this study, a pilot-scale biofilter was set up with natural manganese oxides as carrier materials at a local STP to treat the real secondary effluent. This innovative and simple reactor was on-site operated for approximately 500 days.
Some
EOCs
were
effectively
removed
after
adaptation,
including
10,11-dihydro-10,11-dihydroxycarbamazepine (98%), gabapentin (97%), tramadol (93%), carbamazepine (91%), benzotriazole (91%), sulfamethoxazole (88%), erythromycin (86%). By contrast, the removal of others can be obtained without adaptation, including diclofenac (91%),
carboxy-acyclovir
(91%),
iomeprol
(89%),
1-hydroxybenzotriazol
(87%),
4'-hydroxydiclofenac (86%), acyclovir (73%), tolyltriazole (70%). Overall, 80% of the total mass of 15 detected EOCs was eliminated from the secondary effluent. In addition, 53% of UV254 was removed from wastewater, indicating the aromatic content was damaged to a certain extent.
Keywords:
pharmaceuticals;
transformation
biodegradation; biofiltration.
2
products;
micropollutants;
1 Introduction In past 20 years, various emerging organic contaminants (EOCs) have been detected in water bodies, including pharmaceuticals, personal care products, additives and their transformation products [1]. Although they usually occur at trace concentrations, a potential impact on the aquatic ecological system can still be projected [2]. In the environmental transporting route of EOCs, sewage treatment plants (STPs) play an important role since some EOCs can be effectively removed there but some not. Consequently, STPs are a point source of discharging recalcitrant EOCs into the aquatic bodies. Accordingly, for the effective management of the environmental problems of EOCs, upgrading the current STPs with a tertiary treatment step is an important strategy [3].
Researchers have investigated many technologies to further treat the secondary effluent of STPs, including various chemical oxidation processes, membrane separation, sorption, etc. [4]. Among them, ozonation and activated carbon adsorption are most promising and have been investigated at full scale [5, 6]. Nevertheless, both of them are faced with the challenges of comparably high treatment costs and complex operation. Biotechnological processes are highly attractive for the treatment of wastewater, especially when the contaminants are easily biodegradable. However, EOCs in the secondary effluent are usually recalcitrant to biodegradation.
The biodegradation of recalcitrant pollutants with manganese oxidizing bacteria (MOB) offers a potential technology and however, is not yet extensively studied. The 3
bacteria can oxidize Fe2+ and Mn2+ to their high-valence states and thus play an important role in the natural cycle of iron and manganese [7]. They widely exist in natural water bodies and technical water systems [8]. Recently, researchers found that MOB are capable of degrading recalcitrant pollutants, very likely via the synergetic effect of the biological metabolism of manganese and the oxidation of biogenic manganese oxides (bioMnOx) which is formed when Mn2+ is biologically oxidized to its high valence state [9-12]. For the application of this innovative biological process, Forrez et al. studied lab-scale biofilters packed with manganese oxides and plastic polyethylene rings, separately [13]. They found that the former packing material worked better than the latter in terms of removal and stability. Later, the team collected bioMnOx and then added it into a membrane module (160 mg Mn·l-1) to remove EOCs from a STP secondary effluent [14]. Several EOCs were effectively removed, including naproxen (>95%), morphine (60%), N-acetyl-sulfamethoxazole (92%), clarithromycin (60%) and diuron (90%) at a HRT of 24 h. The results are promising but the approach is possibly hindered by the complexity and costs of installation as well as the long-term stability of the membrane module.
Our previous study compared two lab-scale biofilters packed with natural manganese oxides and zeolites, separately, where two MOB strains were inoculated to treat synthetic and real secondary effluents [15]. Both biofilters can effectively remove diclofenac (DFC) and sulfamethoxazole (SMX). The biofilter with manganese oxides was more efficient than the other with zeolites because of the short adaptation time. It
4
was very interesting to notice that an adaptation time was required to remove DFC in the zeolite biofilter, even with the inoculum of MOB strains whose capacity of degrading DFC had been demonstrated in pure culture previously [16]. It is likely that the degradation in biofilters could be attributed to different strains of MOB.
In the current study, a pilot-scale of biofilter packed with manganese oxides was installed at a local STP to treat real secondary effluent. The objective was to test its applicability and capability of removing EOCs at real conditions.
2 Experimental 2.1 The pilot plant and its operation The pilot plant was installed indoor at an STP (population equivalent 15,000) in Langenau, Germany to treat the secondary effluent. It mainly consisted of a metal sieve (pore size 300-500 µm) to remove particles, a 40-l reservoir, and a fixed-bed biofilter as shown in Fig. 1. The biofilter was constructed with acrylic glass and had a total volume of 250 L (diameter 400 mm, height 2000 mm) and an effective volume of 140 l (packed height 1500 mm). It was packed with natural manganese oxides (Aqua-mandix®, size 3-5 mm) supplied by Aqua-Techniek B.V., Netherlands. Its chemical composition can be found elsewhere [15]. A pump was used to recycle the biofilter upwards at a constant rate of 8 l min-1 which resulted in a superficial velocity of approximately 4 m h-1. The secondary effluent was pumped from a sampling tank to the reservoir via the metal sieve and finally was fed into the biofilter through the recycling pipe. The biofilter effluent was discharged from the top. Aeration was not 5
applied but the dissolved oxygen in the biofilter was daily fluctuating between 1-3 mg l-1. The pH value was around 7.5 without control.
The plant was operated with an effective contact time (ECT) of five hours for initial 150 days and later with ECT 10 h until the end of operation (500th day). Grab samples were taken every three weeks from the reservoir (biofilter inlet) and the discharge pipe (biofilter outlet) and used to analyze EOCs, DOC, TN, TP, Mn, and UV254. The last three samples were taken once a month to further validate the results (see Section 3.1). In total, 29 EOCs were analyzed as target compounds but only 15 of them were constantly detected, including: 1-hydroxybenzotriazol (1-OH-BTZ), 4'-hydroxydiclofenac (4-OH-DFC), 10.11-dihydro-10.11-dihydroxycarbamazepine (CBZ-diol), acyclovir (ACV), benzotriazole (BTA), carbamazepine (CBZ), carboxy-acyclovir (ACV-COOH), diclofenac (DFC), diatrizoic acid (DTA), erythromycin (ETC), gabapentin (GPT), iomeprol (IMP), sulfamethoxazole (SMX), tramadol (TMD), tolyltriazole (TTZ).
2.2 Analysis The water samples were filtrated with 0.45 µm membrane and frozen-stored for the collective analysis. For the analysis of EOCs, 100 µl of water samples was injected to a HPLC-MS system (Thermo Dionex UltiMate 3000 RSLC and Sciex Qtrap 5500) without enrichment. The separation was conducted with a C18 column (Phenomenex Kinetex C18 100 x 4,6 mm, 2,6 µ m). The mobile phase consisted of water and acetonitrile (ACN) with 5 % (v/v) water in addition, both of which were dosed with 0.1 % (v/v) formic acid and 2 mM ammonium formate. A flow rate of 0.6 ml min-1 was used to elute the sample with the following gradient: 6
5% ACN at 0 min, 5% ACN at 5 min, 20% ACN at 10 min, 80% ACN at 15 min, 80% ACN at 18 min, 5% ACN at 18.5 min, 5% ACN at 25 min. Positive/negative polarity switching (5000 / -4200 V) was applied for the detection of different target compounds. The limit of -1
qualification (LOQ) was 25 ng l , calculated according to the German norm DIN 32645 (P=95 %, k=2).
The analysis of DOC, TN, TP, Mn, and UV254 was described elsewhere [15].
3. Results and discussion 3.1 Removal of EOCs The removal efficiency of EOCs was defined as: ܥ − ܥ௨௧ ܥ -1
where, Cout is the concentration of a substance at the biofilter outlet (µg l ); Cin its concentration at the biofilter inlet (µg l-1). When the outlet concentration of a substance was below its LOQ, the LOQ value was used instead. By summating the concentrations of individual EOCs, the total mass concentrations of all EOCs can be obtained. The inlet and outlet concentrations and the removal efficiencies of all EOCs are listed in Supplementary data (Table S1, S2). Fig. 2 depicts the total mass concentrations in the biofilter inlet and outlet and the corresponding removal efficiencies of total EOCs along the operation time. The removal was gradually increased from 30% at the beginning to 80% after approx. 350 days of operation. It is also likely that the removal enhancement was not only due to the increase of ECT from 5 h to 10 h since in the initial 50 days of 10 h ECT the removal did not change significantly. The adaptation could have played an important role. When inspecting the 7
removal of individual EOCs, one can find a clear increasing removal of a group of EOCs, independently from ECTs, including BTA, CBZ, CBZ-diol, GPT, SMX, ETC, and TMD. As shown in Fig. 3, the removal of those EOCs was very low at the beginning and sometimes a negative removal was obtained. Nevertheless, after about 350-day operation, their removal was dramatically ascended up to ~90% and at the last sampling point (499th day), their removal ranked as: CBZ-diol (98%) > GPT (97%) > TMD (93%) > CBZ (91%) = BTA (91%) > SMX (88%) > ETC (86%). In addition, no removal of DTA was observed.
Among this group of EOCs, some are well known as being recalcitrant to biodegradation and their removal in activated sludge processes is mostly <10%, including CBZ, CBZ-diol, GPT, TMD [17-22]. The increasing removal of those EOCs did not fit with a typical adsorption phenomenon but did well with the adaptive or acclimating biological removal. Nevertheless, CBZ was not removed either in our previous lab-scale biofilter which was inoculated with two MOB strains and operated for 1.5 years [15]. A possible reason might be that the lab-scale biofilter started with a synthetic wastewater which contained more easily biodegradable substances than the real secondary effluent used in this pilot biofilter and thus led to the excessive growth of other microorganisms than MOB. Furthermore, the pilot biofilter also took longer adaptation time than the previous lab-scale biofilter (350 d vs 90 d), indicating a different microorganism population. Therefore, a population analysis should be conducted in a further study.
On the other hand, some EOCs achieved good removal immediately after the start of the biofilter. The removal of this group of EOCs was clearly influenced by the ECT: 10-h led to a 8
higher removal than 5-h ECT. Fig. 4 shows their inlet levels, their outlet levels along the whole operation time. One can find that in the inlet of the pilot plant, i.e. the secondary effluent of the local STP, DFC and TTA occurred at the highest concentration (median -1
-1
value >1.5 µg l ) while others mostly did at the levels of hundreds of ng l . Fig. 4 also presents the removal at two ECT settings: 5 h and 10 h, separately. The removal enhancement by the higher ECT can be clearly identified. At ECT 10 h, the removal ranked as DFC (91%) = ACV-COOH (91%) > IMP (89%) > 1-OH-BTA (87%) > 4-OH-DFC (86%) > ACV (73%) > TTA (70%).
The immediately effective removal should be related to the oxidation induced by the
manganese oxides (MnOx). Previous studies have demonstrated that MnOx can oxidize DFC [23] and can also decolorize industrial wastewater [24]. It was also found that the oxidation by MnOx cannot be sustained since the released Mn2+ will cover the active site and inhibit the further reaction [23, 24]. No inhibition effect was observed in this study, indicating that Mn2+ might be re-oxidized by MOB.
Assuming that the removal of those EOCs in the biofilter follows the pseudo-first-order kinetics, the following equation can be used to calculated the kinetic constants: ln
ܥ௨௧ = −݇ ∙ ݐ ܥ
where, Cout is the outlet concentration (µg l-1); Cin the inlet concentration (µg l-1); k, the first-order kinetic constant (h-1); t, reaction time, i.e., ECT (h). 18-20 sampling points were used to calculate the kinetic constants for the non-adaptive removal while only last 4 were applied for the adaptive removal where the optimal removal was established. The kinetic 9
constants of two groups of EOCs can be found in Table 2.
3.3 Removal of conventional parameters Besides EOCs, DOC, TN, TP and UV254 were also monitored for the biofilter at 5 h and 10 h ECT. Their values along the operation time are listed in Supplementary data (Fig. S1). The inlet concentrations of those parameters, i.e., their concentrations in the secondary effluent, -1
-1
-1
were in a typical range: DOC 8.8±2.6 mg l , TN 17.6±3.0 mg l , TP 0.12±0.03 mg l , UV254 0.099±0.015 AU. However, the difference of their removal at two ECT settings was not noticeable. Therefore, the inlet, outlet and removal data of those parameters are presented in Fig. 5 regardless of ECT. DOC, TN, and TP all obtained low removal efficiencies with the following rank (median values): TP (30%) > DOC (26%) >TN (10%). In contrast, the biofilter resulted in a good removal of UV254 (53%, median value), an indicator of aromatic contents in wastewater.
As mentioned above, MnOx may release Mn2+ by oxidizing organics and lead to the
inhibition effect and also to formation of secondary contamination by Mn2+. In this study, the content of Mn was analyzed for the biofilter inlet and outlet. As depicted in Fig. 6, -1
Mn was removed with the biofilter and its outlet concentration was below 40 µg l which is -1
below the value of 50 .µg l recommended by USA and Europe for drinking water. It also indicates that the Mn2+ content on MnOx should be limited as well.
3.4 Comparison with BATs As mentioned before, the best available technologies (BATs) for the reduction of EOCs from the secondary effluents of nutrient removing STP (secondary effluent) are powdered or 10
granular activated carbon (AC) and ozonation. AC has to be replaced regularly as well as further processed after loading and transformation products of ozonation have to be evaluated for every application. However, both processes are reducing EOCs to a certain extend and are applied in full-scale [5, 6].
Table 2 compares the reduction of EOCs by AC, ozonation, and MOB biofilter investigated in this study. The biofilter reduced most of the analyzed substances by more than 80%. However, any process does not reduce DTA significantly, which can be expected due to its designed molecular stability as contrast agent. The biofilter also reduced Gabapentin, an antiepileptic drug, by more than 80% whereas ozone (40-80%) and activated carbon (<40%) are not so efficient. The DOC reduction of 27% is in-between both BATs whereas UV254 is reduced significantly by 53% which is comparable to AC. In summary, the MOB biofilter achieved results comparable to both BATs.
The studied biofilter requires a pressurized filter similar to the granular activated carbon adsorption. However, the residence time based on the currently achieved results is more than 5 times higher in comparison to the granular activated carbon processes [29]. The hydrodynamics of the biofilter have not been optimized so far. This offers the opportunity to decrease the residence time by improving the mass transfer and adapting the residence time distribution to the reaction kinetics.
The costs of the MnOx are about 3 times higher than for granular AC. Nahrstedt et al. (2014) found that the replacement of granular AC is necessary every 11 to 20 months, which indicates that less than 3 years of operation is compensating the higher cost of the carrier 11
material.
4. Conclusions and outlook
After nearly one-year adaption, the biofilter removed 80% of the total mass of 15 detected EOCs. Diatrizoic acid was the only detected EOC with no removal.
Some EOCs can be removed only after the adaption and their removal efficiencies at ECT 10 h were ranked as: CBZ-diol (98%) > GPT (97%) > TMD (93%) > CBZ (91%) = BTA (91%) > SMX (88%) > ETC (86%).
Other EOCs can effectively remove without adaption: DFC (91%) = ACV-COOH (91%) > IMP (89%) > 1-OH-BTA (87%) > 4-OH-DFC (86%) > ACV (73%) > TTA (70%), at ECT 10 h.
Conventional parameters like TN, DOC, and TP were not significantly removed (10%, 26%, and 30%, respectively) while a good removal of UV254 (53%) was obtained.
Further studies should be conducted to understand the adaption process, such as the development of microorganism population and the influence of wastewater matrix. The reactor hydrodynamics can also be optimized to shorten the ECT.
Acknowledgement This study is part of the project TransRisk (grant No. 02WRS1276J) financed by the German BMBF within the program of RiskWa. We appreciate the analytical support of Wolfram Seitz from Landeswasserversorgung, Langenau, Germany. We also thank the colleagues in the
sewage treatment plant of Langenau for taking care of the pilot plant. 12
Reference [1] B. Petrie, R. Barden, B. Kasprzyk-Hordern, A review on emerging contaminants in wastewaters and the environment: Current knowledge, understudied areas and recommendations for future monitoring, Water Res., 72 (2015) 3-27.
[2] J. Brausch, K. Connors, B. Brooks, G. Rand, Human Pharmaceuticals in the Aquatic Environment: A Review of Recent Toxicological Studies and Considerations for Toxicity Testing, in: D.M. Whitacre (Ed.) Reviews of Environmental Contamination and Toxicology Volume 218, Springer US, 2012, pp. 1-99.
[3] Y. Zhang, S.-U. Geißen, Prediction of carbamazepine in sewage treatment plant effluents and its implications for control strategies of pharmaceutical aquatic contamination, Chemosphere, 80 (2010) 1345-1352.
[4] T. Basile, A. Petrella, M. Petrella, G. Boghetich, V. Petruzzelli, S. Colasuonno, D. Petruzzelli, Review of Endocrine-Disrupting-Compound Removal Technologies in Water and Wastewater Treatment Plants: An EU Perspective, Ind. Eng. Chem. Res., 50 (2011) 8389-8401.
[5] J. Reungoat, M. Macova, B.I. Escher, S. Carswell, J.F. Mueller, J. Keller, Removal of micropollutants and reduction of biological activity in a full scale reclamation plant using ozonation and activated carbon filtration, Water Res., 44 (2010) 625-637.
[6] S.G. Zimmermann, M. Wittenwiler, J. Hollender, M. Krauss, C. Ort, H. Siegrist, U. von Gunten, Kinetic assessment and modeling of an ozonation step for full-scale 13
municipal wastewater treatment: Micropollutant oxidation, by-product formation and disinfection, Water Res., 45 (2011) 605-617.
[7] D. Emerson, E.J. Fleming, J.M. McBeth, Iron-Oxidizing Bacteria: An Environmental and Genomic Perspective, Annu. Rev. Microbiol., 64 (2010) 561-583.
[8] J.M. Cerrato, J.O. Falkinham Iii, A.M. Dietrich, W.R. Knocke, C.W. McKinney, A. Pruden, Manganese-oxidizing and -reducing microorganisms isolated from biofilms in chlorinated drinking water systems, Water Res., 44 (2010) 3935-3945.
[9] J.S. Sabirova, L.F.F. Cloetens, L. Vanhaecke, I. Forrez, W. Verstraete, N. Boon, Manganese-oxidizing bacteria mediate the degradation of 17-ethinylestradiol, Microb. Biotechnol., 1 (2008) 507-512.
[10] I. Forrez, M. Carballa, K. Verbeken, L. Vanhaecke, M. Schlüsener, T. Ternes, N. Boon, W. Verstraete, Diclofenac Oxidation by Biogenic Manganese Oxides, Environ. Sci. Technol., 44 (2010) 3449-3454.
[11] F. Meerburg, T. Hennebel, L. Vanhaecke, W. Verstraete, N. Boon, Diclofenac and 2-anilinophenylacetate degradation by combined activity of biogenic manganese oxides and silver, Microb. Biotechnol., 5 (2012) 388-395.
[12] H. Zhu, W. Sun, Y. Zhang, U. Szewzyk, S.-U. Geissen, Removal of pharmaceutical compound diclofenac by iron bacteria, Proceddings of Annual Conference of the Association for General and Applied Microbiology (VAAM), Tübingen, 2012. 14
[13] I. Forrez, M. Carballa, H. Noppe, H. De Brabander, N. Boon, W. Verstraete, Influence
of
manganese
and
ammonium
oxidation
on
the
removal
of
17α-ethinylestradiol (EE2), Water Res., 43 (2009) 77-86.
[14] I. Forrez, M. Carballa, G. Fink, A. Wick, T. Hennebel, L. Vanhaecke, T. Ternes, N. Boon, W. Verstraete, Biogenic metals for the oxidative and reductive removal of pharmaceuticals, biocides and iodinated contrast media in a polishing membrane bioreactor, Water Res., 45 (2011) 1763-1773.
[15] Y. Zhang, H. Zhu, U. Szewzyk, S.U. Geissen, Removal of pharmaceuticals in aerated biofilters with manganese feeding, Water Res., 72 (2015) 218–226.
[16] H. Zhu, Y. Zhang, S.-U. Geißen, U. Szewzyk, Elimination of Diclofenac by Pure Cultures of Iron Bacteria, Proceddings of FEMS, 2013.
[17] P. Falås, A. Wick, S. Castronovo, J. Habermacher, T.A. Ternes, A. Joss, Tracing the limits of organic micropollutant removal in biological wastewater treatment, Water Res., 95 (2016) 240-249.
[18] A. Joss, S. Zabaczynski, A. Göbel, B. Hoffmann, D. Löffler, C.S. McArdell, T.A. Ternes, A. Thomsen, H. Siegrist, Biological degradation of pharmaceuticals in municipal wastewater treatment: Proposing a classification scheme, Water Res., 40 (2006) 1686-1696.
[19] R. Gurke, M. Rößler, C. Marx, S. Diamond, S. Schubert, R. Oertel, J. Fauler, Occurrence and removal of frequently prescribed pharmaceuticals and corresponding 15
metabolites in wastewater of a sewage treatment plant, Sci. Total Environ., 532 (2015) 762-770.
[20] Y. Zhang, S.-U. Geißen, C. Gal, Carbamazepine and diclofenac: removal in wastewater treatment plants and occurrence in water bodies, Chemosphere, 73 (2008) 1151-1161.
[21] B. Kasprzyk-Hordern, R.M. Dinsdale,
A.J. Guwy, The removal of
pharmaceuticals, personal care products, endocrine disruptors and illicit drugs during wastewater treatment and its impact on the quality of receiving waters, Water Res., 43 (2009) 363-380.
[22] M. Pomies, J.M. Choubert, C. Wisniewski, M. Coquery, Modelling of micropollutant removal in biological wastewater treatments: A review, Sci. Total Environ., 443 (2013) 733-748.
[23] M. Huguet, M. Deborde, S. Papot, H. Gallard, Oxidative decarboxylation of diclofenac by manganese oxide bed filter, Water Res., 47 (2013) 5400-5408.
[24] M.M. Arimi, Y. Zhang, S.-U. Geißen, Color removal of melanoidin-rich industrial effluent by natural manganese oxides, Sep. Purif. Technol., 150 (2015) 286-291.
[25] J. Altmann, A. Sperlich, M. Jekel, Integrating organic micropollutant removal into tertiary filtration: Combining PAC adsorption with advanced phosphorus removal, Water Res., 84 (2015) 58-65. 16
[26] J. Altmann, D. Rehfeld, K. Trader, A. Sperlich, M. Jekel, Combination of granular activated carbon adsorption and deep-bed filtration as a single advanced wastewater treatment step for organic micropollutant and phosphorus removal, Water Res., 92 (2016) 131-139.
[27] L. Kovalova, H. Siegrist, U. von Gunten, J. Eugster, M. Hagenbuch, A. Wittmer, R. Moser, C.S. McArdell, Elimination of Micropollutants during Post-Treatment of Hospital Wastewater with Powdered Activated Carbon, Ozone, and UV, Environ. Sci. Technol., 47 (2013) 7899-7908.
[28] G. Knopp, C. Prasse, T.A. Ternes, P. Comel, Elimination of micropollutants and transformation products from a wastewater treatment plant effluent through pilot scale ozonation followed by various activated carbon and biological filters, Water Res., 100 (2016) 580-592.
[29] A. Nahrstedt, H. Burbaum, C. Mauer, K. Alt, T. Sürder, J. Fritzsche, Einsatz granulierter Aktivkohle auf dem Verbandsklärwerk „Obere Lutter“, KA Korrespondenz Abwasser, Abfall, 61 (2014) 408-526.
17
Table 1. First-order kinetic constants of emerging organic contaminants in the biofilter. Pseudo-first-order Emerging organic kinetic constants, contaminants
h-1
Adaptive
CBZ-diol
0.34 ± 0.03
removal
Carbamazepine
0.23 ± 0.01
n=4
Benzotriazole
0.22 ± 0.01
Erythromycin
0.19 ± 0.05
Gabapentin
0.43 ± 0.04
Sulfamethoxazole
0.20 ± 0.01
Tramadol
0.23 ± 0.04
Non-adaptive
1-Hydroxybenzotriazole
0.23 ± 0.07
removal
4'-Hydroxydiclofenac
0.20 ± 0.08
n = 18-20
Acyclovir
0.14 ± 0.06
Carboxyacyclovir
0.29 ± 0.14
Diclofenac
0.33 ± 0.11
Iomeprol
0.24 ± 0.12
Tolytriazole
0.14 ± 0.06
18
Table 2. Removal efficiencies of emerging contaminants and UV254 by activated carbon, ozonation in literature and by the biofilter in this study. Ozonationd
MOB biofilter
>92% >99%
73% 91%
27%
0%
91%
91%
87%
86%
1-Hydroxybenzotriazole
>92%
87%
Tolyltriazole
95%
70%
89%
91%
89%
98%
>99%
91%
>98%
86%
78%
97%
56%
89%
98%
88%
Emerging contaminants and UV254
Activated carbon
Acyclovir Carboxy-acyclovir 23% by PACa Diatrizoic acid
0 by GAC after 5,000 BVb 90% by PACa ~50% by
Benzotriazole
GACb 88% by
Erythromycin
PACc
94% by PACa ~40% by
Carbamazepine
GACb 10,11-Dihydro-10,11-Dihydroxycarbamazepine 90% by PACa
Diclofenac
~50% by GACb
4-Hydroxydiclofenac 21% by PACa ~40% by
Gabapentin
GACb 66% by PACa ~60% by GACb
Iomeprol
Sulfamethoxazole
70% by 19
PACa 0 by GAC after 8,000 BVb Tramadol
>99% by PACc
>97%
93%
UV254
50% by PACa
61%
53%
a: PAC, powder activated carbon, 50 mg/L [25]; b: GAC, granular activated carbon, after 25,000 BV (bed volumes) if not specified [26]; c: 43 mg/L PAC [27]; d: ozone dosage 0.87 gO3/gDOC [28].
20
List of figure captions Fig. 1 The process scheme (a) and a photo (b) of the pilot plant.
Fig. 2. The inlet and outlet overall concentrations of detected emerging organic contaminants and the corresponding overall removal efficiencies at different effective contact time (ECT 5 h, 10 h).
Fig. 3. The adaptive removal of some EOCs at different effective contact time (ECT 5 h, 10 h).
Fig. 4. The inlet concentration levels along the whole operation time, outlet levels and removal of EOCs immediately removed without adaptation at different ECT time (ECT 5 h, 10 h). The whiskers show max. and min. value and the times signs stand for the mean value.
Fig. 5. The inlet value, outlet value, and removal efficiency (boxplots from right to left, n=15) of DOC, TN, TP, UV254. The whiskers show max. and min. value and the times signs stand for the mean value.
Fig. 6. The inlet and outlet concentration of manganese.
21
Fig. 1
22
Fig. 2
23
Fig. 3
24
Fig. 4
25
Fig. 5
26
Fig. 6
27
Graphic Abstract
28
Highlights A pilot-scale biofilter packed with MnOx was operated for approx. 500 d. 80% of the total mass of 15 detected EOCs was effectively removed. 53% of UV254 was removed. Adaptation was required for the removal of some contaminants.
29
S1385-8947(17)30258-9 http://dx.doi.org/10.1016/j.cej.2017.02.095 CEJ 16534
To appear in:
Chemical Engineering Journal
Received Date: Revised Date: Accepted Date:
5 October 2016 24 December 2016 17 February 2017
Please cite this article as: Y. Zhang, H. Zhu, U. Szewzyk, S. Lübbecke, S. Uwe Geissen, Removal of emerging organic contaminants with a pilot-scale biofilter packed with natural manganese oxides, Chemical Engineering Journal (2017), doi: http://dx.doi.org/10.1016/j.cej.2017.02.095
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Removal of emerging organic contaminants with a pilot-scale biofilter packed with natural manganese oxides
Yongjun Zhanga,b,*, Hong Zhuc, Ulrich Szewzykc, Sven Lübbecked, Sven Uwe Geissenb
a, School of Environmental Sciences and Engineering, Nanjing Tech University, Nanjing 211800, PR China b, Technische Universität Berlin, Chair of Environmental Process Engineering, Sekr. KF2, Strasse des 17.Juni 135, 10623 Berlin, Germany c, Technische Universität Berlin, Chair of Environmental Microbiology, Sekr. BH 6-1, Ernst-Reuter-Platz 1, 10587 Berlin, Germany d, Waterleau Deutschland GmbH, Hemelinger Hafendamm 18, 28309 Bremen, Germany
*, Corresponding author Tel.: +86 (0)25 58139656 Fax: +86 (0)25 58139960 E-mail: [email protected]
1
Abstract Emerging
organic
contaminants
(EOCs),
consisting
of
pharmaceuticals,
industrial/household additives, and their transformation products, have been widely detected in surface water bodies and may lead to potential ecological risks. Installing a tertiary treatment step in sewage treatment plants (STP) is an effective method to control their contamination. In this study, a pilot-scale biofilter was set up with natural manganese oxides as carrier materials at a local STP to treat the real secondary effluent. This innovative and simple reactor was on-site operated for approximately 500 days.
Some
EOCs
were
effectively
removed
after
adaptation,
including
10,11-dihydro-10,11-dihydroxycarbamazepine (98%), gabapentin (97%), tramadol (93%), carbamazepine (91%), benzotriazole (91%), sulfamethoxazole (88%), erythromycin (86%). By contrast, the removal of others can be obtained without adaptation, including diclofenac (91%),
carboxy-acyclovir
(91%),
iomeprol
(89%),
1-hydroxybenzotriazol
(87%),
4'-hydroxydiclofenac (86%), acyclovir (73%), tolyltriazole (70%). Overall, 80% of the total mass of 15 detected EOCs was eliminated from the secondary effluent. In addition, 53% of UV254 was removed from wastewater, indicating the aromatic content was damaged to a certain extent.
Keywords:
pharmaceuticals;
transformation
biodegradation; biofiltration.
2
products;
micropollutants;
1 Introduction In past 20 years, various emerging organic contaminants (EOCs) have been detected in water bodies, including pharmaceuticals, personal care products, additives and their transformation products [1]. Although they usually occur at trace concentrations, a potential impact on the aquatic ecological system can still be projected [2]. In the environmental transporting route of EOCs, sewage treatment plants (STPs) play an important role since some EOCs can be effectively removed there but some not. Consequently, STPs are a point source of discharging recalcitrant EOCs into the aquatic bodies. Accordingly, for the effective management of the environmental problems of EOCs, upgrading the current STPs with a tertiary treatment step is an important strategy [3].
Researchers have investigated many technologies to further treat the secondary effluent of STPs, including various chemical oxidation processes, membrane separation, sorption, etc. [4]. Among them, ozonation and activated carbon adsorption are most promising and have been investigated at full scale [5, 6]. Nevertheless, both of them are faced with the challenges of comparably high treatment costs and complex operation. Biotechnological processes are highly attractive for the treatment of wastewater, especially when the contaminants are easily biodegradable. However, EOCs in the secondary effluent are usually recalcitrant to biodegradation.
The biodegradation of recalcitrant pollutants with manganese oxidizing bacteria (MOB) offers a potential technology and however, is not yet extensively studied. The 3
bacteria can oxidize Fe2+ and Mn2+ to their high-valence states and thus play an important role in the natural cycle of iron and manganese [7]. They widely exist in natural water bodies and technical water systems [8]. Recently, researchers found that MOB are capable of degrading recalcitrant pollutants, very likely via the synergetic effect of the biological metabolism of manganese and the oxidation of biogenic manganese oxides (bioMnOx) which is formed when Mn2+ is biologically oxidized to its high valence state [9-12]. For the application of this innovative biological process, Forrez et al. studied lab-scale biofilters packed with manganese oxides and plastic polyethylene rings, separately [13]. They found that the former packing material worked better than the latter in terms of removal and stability. Later, the team collected bioMnOx and then added it into a membrane module (160 mg Mn·l-1) to remove EOCs from a STP secondary effluent [14]. Several EOCs were effectively removed, including naproxen (>95%), morphine (60%), N-acetyl-sulfamethoxazole (92%), clarithromycin (60%) and diuron (90%) at a HRT of 24 h. The results are promising but the approach is possibly hindered by the complexity and costs of installation as well as the long-term stability of the membrane module.
Our previous study compared two lab-scale biofilters packed with natural manganese oxides and zeolites, separately, where two MOB strains were inoculated to treat synthetic and real secondary effluents [15]. Both biofilters can effectively remove diclofenac (DFC) and sulfamethoxazole (SMX). The biofilter with manganese oxides was more efficient than the other with zeolites because of the short adaptation time. It
4
was very interesting to notice that an adaptation time was required to remove DFC in the zeolite biofilter, even with the inoculum of MOB strains whose capacity of degrading DFC had been demonstrated in pure culture previously [16]. It is likely that the degradation in biofilters could be attributed to different strains of MOB.
In the current study, a pilot-scale of biofilter packed with manganese oxides was installed at a local STP to treat real secondary effluent. The objective was to test its applicability and capability of removing EOCs at real conditions.
2 Experimental 2.1 The pilot plant and its operation The pilot plant was installed indoor at an STP (population equivalent 15,000) in Langenau, Germany to treat the secondary effluent. It mainly consisted of a metal sieve (pore size 300-500 µm) to remove particles, a 40-l reservoir, and a fixed-bed biofilter as shown in Fig. 1. The biofilter was constructed with acrylic glass and had a total volume of 250 L (diameter 400 mm, height 2000 mm) and an effective volume of 140 l (packed height 1500 mm). It was packed with natural manganese oxides (Aqua-mandix®, size 3-5 mm) supplied by Aqua-Techniek B.V., Netherlands. Its chemical composition can be found elsewhere [15]. A pump was used to recycle the biofilter upwards at a constant rate of 8 l min-1 which resulted in a superficial velocity of approximately 4 m h-1. The secondary effluent was pumped from a sampling tank to the reservoir via the metal sieve and finally was fed into the biofilter through the recycling pipe. The biofilter effluent was discharged from the top. Aeration was not 5
applied but the dissolved oxygen in the biofilter was daily fluctuating between 1-3 mg l-1. The pH value was around 7.5 without control.
The plant was operated with an effective contact time (ECT) of five hours for initial 150 days and later with ECT 10 h until the end of operation (500th day). Grab samples were taken every three weeks from the reservoir (biofilter inlet) and the discharge pipe (biofilter outlet) and used to analyze EOCs, DOC, TN, TP, Mn, and UV254. The last three samples were taken once a month to further validate the results (see Section 3.1). In total, 29 EOCs were analyzed as target compounds but only 15 of them were constantly detected, including: 1-hydroxybenzotriazol (1-OH-BTZ), 4'-hydroxydiclofenac (4-OH-DFC), 10.11-dihydro-10.11-dihydroxycarbamazepine (CBZ-diol), acyclovir (ACV), benzotriazole (BTA), carbamazepine (CBZ), carboxy-acyclovir (ACV-COOH), diclofenac (DFC), diatrizoic acid (DTA), erythromycin (ETC), gabapentin (GPT), iomeprol (IMP), sulfamethoxazole (SMX), tramadol (TMD), tolyltriazole (TTZ).
2.2 Analysis The water samples were filtrated with 0.45 µm membrane and frozen-stored for the collective analysis. For the analysis of EOCs, 100 µl of water samples was injected to a HPLC-MS system (Thermo Dionex UltiMate 3000 RSLC and Sciex Qtrap 5500) without enrichment. The separation was conducted with a C18 column (Phenomenex Kinetex C18 100 x 4,6 mm, 2,6 µ m). The mobile phase consisted of water and acetonitrile (ACN) with 5 % (v/v) water in addition, both of which were dosed with 0.1 % (v/v) formic acid and 2 mM ammonium formate. A flow rate of 0.6 ml min-1 was used to elute the sample with the following gradient: 6
5% ACN at 0 min, 5% ACN at 5 min, 20% ACN at 10 min, 80% ACN at 15 min, 80% ACN at 18 min, 5% ACN at 18.5 min, 5% ACN at 25 min. Positive/negative polarity switching (5000 / -4200 V) was applied for the detection of different target compounds. The limit of -1
qualification (LOQ) was 25 ng l , calculated according to the German norm DIN 32645 (P=95 %, k=2).
The analysis of DOC, TN, TP, Mn, and UV254 was described elsewhere [15].
3. Results and discussion 3.1 Removal of EOCs The removal efficiency of EOCs was defined as: ܥ − ܥ௨௧ ܥ -1
where, Cout is the concentration of a substance at the biofilter outlet (µg l ); Cin its concentration at the biofilter inlet (µg l-1). When the outlet concentration of a substance was below its LOQ, the LOQ value was used instead. By summating the concentrations of individual EOCs, the total mass concentrations of all EOCs can be obtained. The inlet and outlet concentrations and the removal efficiencies of all EOCs are listed in Supplementary data (Table S1, S2). Fig. 2 depicts the total mass concentrations in the biofilter inlet and outlet and the corresponding removal efficiencies of total EOCs along the operation time. The removal was gradually increased from 30% at the beginning to 80% after approx. 350 days of operation. It is also likely that the removal enhancement was not only due to the increase of ECT from 5 h to 10 h since in the initial 50 days of 10 h ECT the removal did not change significantly. The adaptation could have played an important role. When inspecting the 7
removal of individual EOCs, one can find a clear increasing removal of a group of EOCs, independently from ECTs, including BTA, CBZ, CBZ-diol, GPT, SMX, ETC, and TMD. As shown in Fig. 3, the removal of those EOCs was very low at the beginning and sometimes a negative removal was obtained. Nevertheless, after about 350-day operation, their removal was dramatically ascended up to ~90% and at the last sampling point (499th day), their removal ranked as: CBZ-diol (98%) > GPT (97%) > TMD (93%) > CBZ (91%) = BTA (91%) > SMX (88%) > ETC (86%). In addition, no removal of DTA was observed.
Among this group of EOCs, some are well known as being recalcitrant to biodegradation and their removal in activated sludge processes is mostly <10%, including CBZ, CBZ-diol, GPT, TMD [17-22]. The increasing removal of those EOCs did not fit with a typical adsorption phenomenon but did well with the adaptive or acclimating biological removal. Nevertheless, CBZ was not removed either in our previous lab-scale biofilter which was inoculated with two MOB strains and operated for 1.5 years [15]. A possible reason might be that the lab-scale biofilter started with a synthetic wastewater which contained more easily biodegradable substances than the real secondary effluent used in this pilot biofilter and thus led to the excessive growth of other microorganisms than MOB. Furthermore, the pilot biofilter also took longer adaptation time than the previous lab-scale biofilter (350 d vs 90 d), indicating a different microorganism population. Therefore, a population analysis should be conducted in a further study.
On the other hand, some EOCs achieved good removal immediately after the start of the biofilter. The removal of this group of EOCs was clearly influenced by the ECT: 10-h led to a 8
higher removal than 5-h ECT. Fig. 4 shows their inlet levels, their outlet levels along the whole operation time. One can find that in the inlet of the pilot plant, i.e. the secondary effluent of the local STP, DFC and TTA occurred at the highest concentration (median -1
-1
value >1.5 µg l ) while others mostly did at the levels of hundreds of ng l . Fig. 4 also presents the removal at two ECT settings: 5 h and 10 h, separately. The removal enhancement by the higher ECT can be clearly identified. At ECT 10 h, the removal ranked as DFC (91%) = ACV-COOH (91%) > IMP (89%) > 1-OH-BTA (87%) > 4-OH-DFC (86%) > ACV (73%) > TTA (70%).
The immediately effective removal should be related to the oxidation induced by the
manganese oxides (MnOx). Previous studies have demonstrated that MnOx can oxidize DFC [23] and can also decolorize industrial wastewater [24]. It was also found that the oxidation by MnOx cannot be sustained since the released Mn2+ will cover the active site and inhibit the further reaction [23, 24]. No inhibition effect was observed in this study, indicating that Mn2+ might be re-oxidized by MOB.
Assuming that the removal of those EOCs in the biofilter follows the pseudo-first-order kinetics, the following equation can be used to calculated the kinetic constants: ln
ܥ௨௧ = −݇ ∙ ݐ ܥ
where, Cout is the outlet concentration (µg l-1); Cin the inlet concentration (µg l-1); k, the first-order kinetic constant (h-1); t, reaction time, i.e., ECT (h). 18-20 sampling points were used to calculate the kinetic constants for the non-adaptive removal while only last 4 were applied for the adaptive removal where the optimal removal was established. The kinetic 9
constants of two groups of EOCs can be found in Table 2.
3.3 Removal of conventional parameters Besides EOCs, DOC, TN, TP and UV254 were also monitored for the biofilter at 5 h and 10 h ECT. Their values along the operation time are listed in Supplementary data (Fig. S1). The inlet concentrations of those parameters, i.e., their concentrations in the secondary effluent, -1
-1
-1
were in a typical range: DOC 8.8±2.6 mg l , TN 17.6±3.0 mg l , TP 0.12±0.03 mg l , UV254 0.099±0.015 AU. However, the difference of their removal at two ECT settings was not noticeable. Therefore, the inlet, outlet and removal data of those parameters are presented in Fig. 5 regardless of ECT. DOC, TN, and TP all obtained low removal efficiencies with the following rank (median values): TP (30%) > DOC (26%) >TN (10%). In contrast, the biofilter resulted in a good removal of UV254 (53%, median value), an indicator of aromatic contents in wastewater.
As mentioned above, MnOx may release Mn2+ by oxidizing organics and lead to the
inhibition effect and also to formation of secondary contamination by Mn2+. In this study, the content of Mn was analyzed for the biofilter inlet and outlet. As depicted in Fig. 6, -1
Mn was removed with the biofilter and its outlet concentration was below 40 µg l which is -1
below the value of 50 .µg l recommended by USA and Europe for drinking water. It also indicates that the Mn2+ content on MnOx should be limited as well.
3.4 Comparison with BATs As mentioned before, the best available technologies (BATs) for the reduction of EOCs from the secondary effluents of nutrient removing STP (secondary effluent) are powdered or 10
granular activated carbon (AC) and ozonation. AC has to be replaced regularly as well as further processed after loading and transformation products of ozonation have to be evaluated for every application. However, both processes are reducing EOCs to a certain extend and are applied in full-scale [5, 6].
Table 2 compares the reduction of EOCs by AC, ozonation, and MOB biofilter investigated in this study. The biofilter reduced most of the analyzed substances by more than 80%. However, any process does not reduce DTA significantly, which can be expected due to its designed molecular stability as contrast agent. The biofilter also reduced Gabapentin, an antiepileptic drug, by more than 80% whereas ozone (40-80%) and activated carbon (<40%) are not so efficient. The DOC reduction of 27% is in-between both BATs whereas UV254 is reduced significantly by 53% which is comparable to AC. In summary, the MOB biofilter achieved results comparable to both BATs.
The studied biofilter requires a pressurized filter similar to the granular activated carbon adsorption. However, the residence time based on the currently achieved results is more than 5 times higher in comparison to the granular activated carbon processes [29]. The hydrodynamics of the biofilter have not been optimized so far. This offers the opportunity to decrease the residence time by improving the mass transfer and adapting the residence time distribution to the reaction kinetics.
The costs of the MnOx are about 3 times higher than for granular AC. Nahrstedt et al. (2014) found that the replacement of granular AC is necessary every 11 to 20 months, which indicates that less than 3 years of operation is compensating the higher cost of the carrier 11
material.
4. Conclusions and outlook
After nearly one-year adaption, the biofilter removed 80% of the total mass of 15 detected EOCs. Diatrizoic acid was the only detected EOC with no removal.
Some EOCs can be removed only after the adaption and their removal efficiencies at ECT 10 h were ranked as: CBZ-diol (98%) > GPT (97%) > TMD (93%) > CBZ (91%) = BTA (91%) > SMX (88%) > ETC (86%).
Other EOCs can effectively remove without adaption: DFC (91%) = ACV-COOH (91%) > IMP (89%) > 1-OH-BTA (87%) > 4-OH-DFC (86%) > ACV (73%) > TTA (70%), at ECT 10 h.
Conventional parameters like TN, DOC, and TP were not significantly removed (10%, 26%, and 30%, respectively) while a good removal of UV254 (53%) was obtained.
Further studies should be conducted to understand the adaption process, such as the development of microorganism population and the influence of wastewater matrix. The reactor hydrodynamics can also be optimized to shorten the ECT.
Acknowledgement This study is part of the project TransRisk (grant No. 02WRS1276J) financed by the German BMBF within the program of RiskWa. We appreciate the analytical support of Wolfram Seitz from Landeswasserversorgung, Langenau, Germany. We also thank the colleagues in the
sewage treatment plant of Langenau for taking care of the pilot plant. 12
Reference [1] B. Petrie, R. Barden, B. Kasprzyk-Hordern, A review on emerging contaminants in wastewaters and the environment: Current knowledge, understudied areas and recommendations for future monitoring, Water Res., 72 (2015) 3-27.
[2] J. Brausch, K. Connors, B. Brooks, G. Rand, Human Pharmaceuticals in the Aquatic Environment: A Review of Recent Toxicological Studies and Considerations for Toxicity Testing, in: D.M. Whitacre (Ed.) Reviews of Environmental Contamination and Toxicology Volume 218, Springer US, 2012, pp. 1-99.
[3] Y. Zhang, S.-U. Geißen, Prediction of carbamazepine in sewage treatment plant effluents and its implications for control strategies of pharmaceutical aquatic contamination, Chemosphere, 80 (2010) 1345-1352.
[4] T. Basile, A. Petrella, M. Petrella, G. Boghetich, V. Petruzzelli, S. Colasuonno, D. Petruzzelli, Review of Endocrine-Disrupting-Compound Removal Technologies in Water and Wastewater Treatment Plants: An EU Perspective, Ind. Eng. Chem. Res., 50 (2011) 8389-8401.
[5] J. Reungoat, M. Macova, B.I. Escher, S. Carswell, J.F. Mueller, J. Keller, Removal of micropollutants and reduction of biological activity in a full scale reclamation plant using ozonation and activated carbon filtration, Water Res., 44 (2010) 625-637.
[6] S.G. Zimmermann, M. Wittenwiler, J. Hollender, M. Krauss, C. Ort, H. Siegrist, U. von Gunten, Kinetic assessment and modeling of an ozonation step for full-scale 13
municipal wastewater treatment: Micropollutant oxidation, by-product formation and disinfection, Water Res., 45 (2011) 605-617.
[7] D. Emerson, E.J. Fleming, J.M. McBeth, Iron-Oxidizing Bacteria: An Environmental and Genomic Perspective, Annu. Rev. Microbiol., 64 (2010) 561-583.
[8] J.M. Cerrato, J.O. Falkinham Iii, A.M. Dietrich, W.R. Knocke, C.W. McKinney, A. Pruden, Manganese-oxidizing and -reducing microorganisms isolated from biofilms in chlorinated drinking water systems, Water Res., 44 (2010) 3935-3945.
[9] J.S. Sabirova, L.F.F. Cloetens, L. Vanhaecke, I. Forrez, W. Verstraete, N. Boon, Manganese-oxidizing bacteria mediate the degradation of 17-ethinylestradiol, Microb. Biotechnol., 1 (2008) 507-512.
[10] I. Forrez, M. Carballa, K. Verbeken, L. Vanhaecke, M. Schlüsener, T. Ternes, N. Boon, W. Verstraete, Diclofenac Oxidation by Biogenic Manganese Oxides, Environ. Sci. Technol., 44 (2010) 3449-3454.
[11] F. Meerburg, T. Hennebel, L. Vanhaecke, W. Verstraete, N. Boon, Diclofenac and 2-anilinophenylacetate degradation by combined activity of biogenic manganese oxides and silver, Microb. Biotechnol., 5 (2012) 388-395.
[12] H. Zhu, W. Sun, Y. Zhang, U. Szewzyk, S.-U. Geissen, Removal of pharmaceutical compound diclofenac by iron bacteria, Proceddings of Annual Conference of the Association for General and Applied Microbiology (VAAM), Tübingen, 2012. 14
[13] I. Forrez, M. Carballa, H. Noppe, H. De Brabander, N. Boon, W. Verstraete, Influence
of
manganese
and
ammonium
oxidation
on
the
removal
of
17α-ethinylestradiol (EE2), Water Res., 43 (2009) 77-86.
[14] I. Forrez, M. Carballa, G. Fink, A. Wick, T. Hennebel, L. Vanhaecke, T. Ternes, N. Boon, W. Verstraete, Biogenic metals for the oxidative and reductive removal of pharmaceuticals, biocides and iodinated contrast media in a polishing membrane bioreactor, Water Res., 45 (2011) 1763-1773.
[15] Y. Zhang, H. Zhu, U. Szewzyk, S.U. Geissen, Removal of pharmaceuticals in aerated biofilters with manganese feeding, Water Res., 72 (2015) 218–226.
[16] H. Zhu, Y. Zhang, S.-U. Geißen, U. Szewzyk, Elimination of Diclofenac by Pure Cultures of Iron Bacteria, Proceddings of FEMS, 2013.
[17] P. Falås, A. Wick, S. Castronovo, J. Habermacher, T.A. Ternes, A. Joss, Tracing the limits of organic micropollutant removal in biological wastewater treatment, Water Res., 95 (2016) 240-249.
[18] A. Joss, S. Zabaczynski, A. Göbel, B. Hoffmann, D. Löffler, C.S. McArdell, T.A. Ternes, A. Thomsen, H. Siegrist, Biological degradation of pharmaceuticals in municipal wastewater treatment: Proposing a classification scheme, Water Res., 40 (2006) 1686-1696.
[19] R. Gurke, M. Rößler, C. Marx, S. Diamond, S. Schubert, R. Oertel, J. Fauler, Occurrence and removal of frequently prescribed pharmaceuticals and corresponding 15
metabolites in wastewater of a sewage treatment plant, Sci. Total Environ., 532 (2015) 762-770.
[20] Y. Zhang, S.-U. Geißen, C. Gal, Carbamazepine and diclofenac: removal in wastewater treatment plants and occurrence in water bodies, Chemosphere, 73 (2008) 1151-1161.
[21] B. Kasprzyk-Hordern, R.M. Dinsdale,
A.J. Guwy, The removal of
pharmaceuticals, personal care products, endocrine disruptors and illicit drugs during wastewater treatment and its impact on the quality of receiving waters, Water Res., 43 (2009) 363-380.
[22] M. Pomies, J.M. Choubert, C. Wisniewski, M. Coquery, Modelling of micropollutant removal in biological wastewater treatments: A review, Sci. Total Environ., 443 (2013) 733-748.
[23] M. Huguet, M. Deborde, S. Papot, H. Gallard, Oxidative decarboxylation of diclofenac by manganese oxide bed filter, Water Res., 47 (2013) 5400-5408.
[24] M.M. Arimi, Y. Zhang, S.-U. Geißen, Color removal of melanoidin-rich industrial effluent by natural manganese oxides, Sep. Purif. Technol., 150 (2015) 286-291.
[25] J. Altmann, A. Sperlich, M. Jekel, Integrating organic micropollutant removal into tertiary filtration: Combining PAC adsorption with advanced phosphorus removal, Water Res., 84 (2015) 58-65. 16
[26] J. Altmann, D. Rehfeld, K. Trader, A. Sperlich, M. Jekel, Combination of granular activated carbon adsorption and deep-bed filtration as a single advanced wastewater treatment step for organic micropollutant and phosphorus removal, Water Res., 92 (2016) 131-139.
[27] L. Kovalova, H. Siegrist, U. von Gunten, J. Eugster, M. Hagenbuch, A. Wittmer, R. Moser, C.S. McArdell, Elimination of Micropollutants during Post-Treatment of Hospital Wastewater with Powdered Activated Carbon, Ozone, and UV, Environ. Sci. Technol., 47 (2013) 7899-7908.
[28] G. Knopp, C. Prasse, T.A. Ternes, P. Comel, Elimination of micropollutants and transformation products from a wastewater treatment plant effluent through pilot scale ozonation followed by various activated carbon and biological filters, Water Res., 100 (2016) 580-592.
[29] A. Nahrstedt, H. Burbaum, C. Mauer, K. Alt, T. Sürder, J. Fritzsche, Einsatz granulierter Aktivkohle auf dem Verbandsklärwerk „Obere Lutter“, KA Korrespondenz Abwasser, Abfall, 61 (2014) 408-526.
17
Table 1. First-order kinetic constants of emerging organic contaminants in the biofilter. Pseudo-first-order Emerging organic kinetic constants, contaminants
h-1
Adaptive
CBZ-diol
0.34 ± 0.03
removal
Carbamazepine
0.23 ± 0.01
n=4
Benzotriazole
0.22 ± 0.01
Erythromycin
0.19 ± 0.05
Gabapentin
0.43 ± 0.04
Sulfamethoxazole
0.20 ± 0.01
Tramadol
0.23 ± 0.04
Non-adaptive
1-Hydroxybenzotriazole
0.23 ± 0.07
removal
4'-Hydroxydiclofenac
0.20 ± 0.08
n = 18-20
Acyclovir
0.14 ± 0.06
Carboxyacyclovir
0.29 ± 0.14
Diclofenac
0.33 ± 0.11
Iomeprol
0.24 ± 0.12
Tolytriazole
0.14 ± 0.06
18
Table 2. Removal efficiencies of emerging contaminants and UV254 by activated carbon, ozonation in literature and by the biofilter in this study. Ozonationd
MOB biofilter
>92% >99%
73% 91%
27%
0%
91%
91%
87%
86%
1-Hydroxybenzotriazole
>92%
87%
Tolyltriazole
95%
70%
89%
91%
89%
98%
>99%
91%
>98%
86%
78%
97%
56%
89%
98%
88%
Emerging contaminants and UV254
Activated carbon
Acyclovir Carboxy-acyclovir 23% by PACa Diatrizoic acid
0 by GAC after 5,000 BVb 90% by PACa ~50% by
Benzotriazole
GACb 88% by
Erythromycin
PACc
94% by PACa ~40% by
Carbamazepine
GACb 10,11-Dihydro-10,11-Dihydroxycarbamazepine 90% by PACa
Diclofenac
~50% by GACb
4-Hydroxydiclofenac 21% by PACa ~40% by
Gabapentin
GACb 66% by PACa ~60% by GACb
Iomeprol
Sulfamethoxazole
70% by 19
PACa 0 by GAC after 8,000 BVb Tramadol
>99% by PACc
>97%
93%
UV254
50% by PACa
61%
53%
a: PAC, powder activated carbon, 50 mg/L [25]; b: GAC, granular activated carbon, after 25,000 BV (bed volumes) if not specified [26]; c: 43 mg/L PAC [27]; d: ozone dosage 0.87 gO3/gDOC [28].
20
List of figure captions Fig. 1 The process scheme (a) and a photo (b) of the pilot plant.
Fig. 2. The inlet and outlet overall concentrations of detected emerging organic contaminants and the corresponding overall removal efficiencies at different effective contact time (ECT 5 h, 10 h).
Fig. 3. The adaptive removal of some EOCs at different effective contact time (ECT 5 h, 10 h).
Fig. 4. The inlet concentration levels along the whole operation time, outlet levels and removal of EOCs immediately removed without adaptation at different ECT time (ECT 5 h, 10 h). The whiskers show max. and min. value and the times signs stand for the mean value.
Fig. 5. The inlet value, outlet value, and removal efficiency (boxplots from right to left, n=15) of DOC, TN, TP, UV254. The whiskers show max. and min. value and the times signs stand for the mean value.
Fig. 6. The inlet and outlet concentration of manganese.
21
Fig. 1
22
Fig. 2
23
Fig. 3
24
Fig. 4
25
Fig. 5
26
Fig. 6
27
Graphic Abstract
28
Highlights A pilot-scale biofilter packed with MnOx was operated for approx. 500 d. 80% of the total mass of 15 detected EOCs was effectively removed. 53% of UV254 was removed. Adaptation was required for the removal of some contaminants.
29