fate of selected emerging micropollutants during mesophilic, thermophilic and temperature co-phased...

8
Fate of selected emerging micropollutants during mesophilic, thermophilic and temperature co-phased anaerobic digestion of sewage sludge Vasilios G. Samaras a , Athanasios S. Stasinakis a,, Nikolaos S. Thomaidis b , Daniel Mamais c , Themistokles D. Lekkas a a Water and Air Quality Laboratory, Department of Environment, University of the Aegean, University Hill, 81100 Mytilene, Greece b Laboratory of Analytical Chemistry, Department of Chemistry, University of Athens, Panepistimioupolis Zografou, 15771 Athens, Greece c Department of Water Resources, Faculty of Civil Engineering, National Technical University of Athens, Zografou, Athens 15773, Greece highlights Two single-stage and a two-stage thermophilic/mesophilic AD systems were used. Removal efficiency of synthetic EDCs and NSAIDs was investigated. NSAIDs were highly removed during sludge AD, EDCs removal was moderate. The use of thermophilic/mesophilic system slightly enhanced removal of EDCs. Biotransformation of NP 1 EO and NP was affected by digesters’ temperature. article info Article history: Received 31 January 2014 Received in revised form 24 March 2014 Accepted 28 March 2014 Available online 5 April 2014 Keywords: Anaerobic digestion Sludge Endocrine disrupting compounds Pharmaceuticals Removal abstract The removal of endocrine disrupting compounds (EDCs) and non-steroidal anti-inflammatory drugs (NSAIDs) was studied in three lab-scale anaerobic digestion (AD) systems; a single-stage mesophilic, a single-stage thermophilic and a two-stage thermophilic/mesophilic. All micropollutants underwent microbial degradation. High removal efficiency (>80%) was calculated for diclofenac, ibuprofen, naproxen and ketoprofen; whereas triclosan, bisphenol A and the sum of nonylphenol (NP), nonylphenol monoeth- oxylate (NP 1 EO) and nonylphenol diethoxylate were moderately removed (40–80%). NSAIDs removal was not affected by the type of AD system used; whereas slightly higher EDCs removal was observed in two- stage system. In this system, most microcontaminants were removed in thermophilic digester. Biotrans- formation of NP 1 EO and NP was affected by the temperature applied to bioreactors. Under mesophilic conditions, higher removal of NP 1 EO and accumulation of NP was noticed; whereas the opposite was observed under thermophilic conditions. For most analytes, higher specific removal rates were calculated under thermophilic conditions and 20 days SRT. Ó 2014 Elsevier Ltd. All rights reserved. 1. Introduction Anaerobic digestion (AD) is currently one of the most widely adopted treatment processes for sludge produced in wastewater treatment plants (WWTPs). The application of single stage meso- philic AD results to efficient sludge stabilization and biogas pro- duction (Appels et al., 2008; Cao and Pawlowski, 2012). Single stage thermophilic AD and two-stage thermophilic/mesophilic AD are also used worldwide to assure safer sludge disposal and better operational stability, respectively (Chen et al., 2008; Ge et al., 2010; Rubio-Loza and Noyola, 2010). Nowadays, almost 10 million tons of dry sludge are produced in EU27 and more than half of this amount is spread to the land for agricultural purposes (Kelessidis and Stasinakis, 2012). Recent studies have indicated the occurrence of several emerging organic micropollutants such as endocrine disrupting compounds (EDCs) and non-steroidal anti-inflammatory drugs (NSAIDs) in sludge at concentrations ranging between few lg kg 1 to some mg kg 1 DS (Stasinakis, 2012; Narumiya et al., 2013; Stasinakis et al., 2013). Among the EDCs commonly detected in sludge, the non-polar and highly lipophilic nonylphenol (NP), nonylphenol monoethoxylate http://dx.doi.org/10.1016/j.biortech.2014.03.154 0960-8524/Ó 2014 Elsevier Ltd. All rights reserved. Corresponding author. Tel.: +30 22510 36257; fax: +30 22510 36209. E-mail address: [email protected] (A.S. Stasinakis). Bioresource Technology 162 (2014) 365–372 Contents lists available at ScienceDirect Bioresource Technology journal homepage: www.elsevier.com/locate/biortech

Upload: themistokles-d

Post on 30-Dec-2016

217 views

Category:

Documents


0 download

TRANSCRIPT

Bioresource Technology 162 (2014) 365–372

Contents lists available at ScienceDirect

Bioresource Technology

journal homepage: www.elsevier .com/locate /bior tech

Fate of selected emerging micropollutants during mesophilic,thermophilic and temperature co-phased anaerobic digestionof sewage sludge

http://dx.doi.org/10.1016/j.biortech.2014.03.1540960-8524/� 2014 Elsevier Ltd. All rights reserved.

⇑ Corresponding author. Tel.: +30 22510 36257; fax: +30 22510 36209.E-mail address: [email protected] (A.S. Stasinakis).

Vasilios G. Samaras a, Athanasios S. Stasinakis a,⇑, Nikolaos S. Thomaidis b, Daniel Mamais c,Themistokles D. Lekkas a

a Water and Air Quality Laboratory, Department of Environment, University of the Aegean, University Hill, 81100 Mytilene, Greeceb Laboratory of Analytical Chemistry, Department of Chemistry, University of Athens, Panepistimioupolis Zografou, 15771 Athens, Greecec Department of Water Resources, Faculty of Civil Engineering, National Technical University of Athens, Zografou, Athens 15773, Greece

h i g h l i g h t s

� Two single-stage and a two-stage thermophilic/mesophilic AD systems were used.� Removal efficiency of synthetic EDCs and NSAIDs was investigated.� NSAIDs were highly removed during sludge AD, EDCs removal was moderate.� The use of thermophilic/mesophilic system slightly enhanced removal of EDCs.� Biotransformation of NP1EO and NP was affected by digesters’ temperature.

a r t i c l e i n f o

Article history:Received 31 January 2014Received in revised form 24 March 2014Accepted 28 March 2014Available online 5 April 2014

Keywords:Anaerobic digestionSludgeEndocrine disrupting compoundsPharmaceuticalsRemoval

a b s t r a c t

The removal of endocrine disrupting compounds (EDCs) and non-steroidal anti-inflammatory drugs(NSAIDs) was studied in three lab-scale anaerobic digestion (AD) systems; a single-stage mesophilic, asingle-stage thermophilic and a two-stage thermophilic/mesophilic. All micropollutants underwentmicrobial degradation. High removal efficiency (>80%) was calculated for diclofenac, ibuprofen, naproxenand ketoprofen; whereas triclosan, bisphenol A and the sum of nonylphenol (NP), nonylphenol monoeth-oxylate (NP1EO) and nonylphenol diethoxylate were moderately removed (40–80%). NSAIDs removal wasnot affected by the type of AD system used; whereas slightly higher EDCs removal was observed in two-stage system. In this system, most microcontaminants were removed in thermophilic digester. Biotrans-formation of NP1EO and NP was affected by the temperature applied to bioreactors. Under mesophilicconditions, higher removal of NP1EO and accumulation of NP was noticed; whereas the opposite wasobserved under thermophilic conditions. For most analytes, higher specific removal rates were calculatedunder thermophilic conditions and 20 days SRT.

� 2014 Elsevier Ltd. All rights reserved.

1. Introduction

Anaerobic digestion (AD) is currently one of the most widelyadopted treatment processes for sludge produced in wastewatertreatment plants (WWTPs). The application of single stage meso-philic AD results to efficient sludge stabilization and biogas pro-duction (Appels et al., 2008; Cao and Pawlowski, 2012). Singlestage thermophilic AD and two-stage thermophilic/mesophilicAD are also used worldwide to assure safer sludge disposal and

better operational stability, respectively (Chen et al., 2008; Geet al., 2010; Rubio-Loza and Noyola, 2010).

Nowadays, almost 10 million tons of dry sludge are produced inEU27 and more than half of this amount is spread to the land foragricultural purposes (Kelessidis and Stasinakis, 2012). Recentstudies have indicated the occurrence of several emerging organicmicropollutants such as endocrine disrupting compounds (EDCs)and non-steroidal anti-inflammatory drugs (NSAIDs) in sludge atconcentrations ranging between few lg kg�1 to some mg kg�1 DS(Stasinakis, 2012; Narumiya et al., 2013; Stasinakis et al., 2013).Among the EDCs commonly detected in sludge, the non-polar andhighly lipophilic nonylphenol (NP), nonylphenol monoethoxylate

366 V.G. Samaras et al. / Bioresource Technology 162 (2014) 365–372

(NP1EO) and nonylphenol diethoxylate (NP2EO) as well as triclosan(TCS) and bisphenol A (BPA) are of critical environmental concerndue to their high concentrations and toxicological properties. Onthe other hand, diclofenac (DCF), naproxen (NPX), ketoprofen(KFN) and ibuprofen (IBF) are a group of NSAIDs with worldwidehigh consumption due to their analgesic and antipyretic effects.Despite the often detection of the aforementioned compounds insludge samples, so far, little work has been conducted to systemat-ically track their removal in sludge AD. Relevant information usu-ally originates from the monitoring of full-scale anaerobicdigesters (Narumiya et al., 2013; Samaras et al., 2013), while thereis no study comparing emerging contaminants’ removal in parallelmesophilic, thermophilic and two-stage AD systems.

Specifically, according to monitoring studies, NP1EO and NP2EOare partially removed during sludge AD, whereas in many casesconcentrations of NP in digested sludge are higher comparing toraw sludge (González et al., 2010; Samaras et al., 2013). NP1EOand NP2EO seem to be biodegraded to some extent during AD,while contradictory results have been presented for NP by a num-ber of authors (Hernandez-Raquet et al., 2007; Chang et al., 2005;Paterakis et al., 2012). Regarding BPA and TCS, in a recent monitor-ing study a mean removal rate lower than 40%, was observed forboth compounds in a mesophilic full-scale anaerobic digester(Samaras et al., 2013). To the best of our knowledge, no lab-scaleexperiments have been performed to investigate the fate of thesemicropollutants during sludge AD. Regarding the removal of targetNSAIDs in full-scale anaerobic digesters, Samaras et al. (2013)reported that IBF and NPX were removed at a rate higher than80% and Narumiya et al. (2013) reported a removal of KFN andDCF between 30% to 40%. Carballa et al. (2006), using mesophiliclab-scale anaerobic digesters, reported removals ranging between40% to 87% for IBF, DCF and NPX.

Based on the above, the objective of this study was to investi-gate the removal efficiency of five synthetic EDCs (NP, NP1EO,NP2EO, TCS, BPA) and four NSAIDs (DCF, NPX, IBF, KFN) in lab-scaleanaerobic digesters, operating under single-stage and two-stagethermophilic and mesophilic anaerobic conditions. For this reason,three systems were used; a single-stage mesophilic, a single-stagethermophilic and a two-stage thermophilic/mesophilic. All sys-tems operated at the same total sludge residence time (SRT:20 days), their operation was divided in two phases and lasted466 days. AD performance stability was monitored on a frequentbasis by measuring total suspended solids (TSS), volatile sus-pended solids (VSS), total COD (CODt), VFAs concentration, as wellas biogas production and alkalinity. The removal efficiency andspecific removal rates of target analytes were estimated in allbioreactors.

2. Methods

2.1. Analytical standards and reagents

Analytical standards of IBF, NPX, DFC, KFN, TCS, NP, NP1EO andNP2EO were supplied by Dr. Ehrenstorfer (Germany). BPA was pur-chased by Fluka (Switzerland); whereas the deuterated [2H16]bisphenol A (BPA d-16) and meclofenamic acid were purchasedfrom Sigma–Aldrich (USA). Stock and working solutions of individ-ual compounds were prepared in methanol at 1000 mg L�1 andkept at �18 �C. Methanol (MeOH) and ethyl acetate were of HPLCgrade (Merck, Germany) and they were used as received. Bis(tri-methylsilyl)trifluoroacetamide (BSTFA), BSTFA + 1% trimethyl chlo-rosilane (TMCS) solution and pyridine, used for silylation, werepurchased by Supelco (USA) and Carlo Erba-SDS (France), respec-tively. The solid phase extraction (SPE) cartridges used for samples’clean-up were silica-based bonded C18 (Sep-Pak, 6 ml, 500 mg)

and they were supplied by Waters (Ireland). HPLC grade waterwas prepared in the laboratory using a MilliQ/MilliRO Milliporesystem (USA). Ultra-pure HCl (32%), used for samples’ acidification,was purchased by Merck (Germany). All the abbreviations used inthis study can be found in Table S1.

2.2. Lab-scale anaerobic digesters

Four laboratory-scale continuously stirred anaerobic digesters(CSTRs) were used in this study (Fig. 1). Two of them were oper-ated at the mesophilic temperature range (M1 and M2) and twoat the thermophilic temperature range (T1 and T2). Each reactorwas made of 5 mm thick steel cylinders, had an internal diameterof 100 mm, a total volume of 4 L and an operating liquid volume of3 L. Sealing of the reactors was achieved by bolted O-ring mountedplexiglass stoppers. Three ports were installed in each digester forfeeding, sample collection and biogas venting. Mixing was pro-vided by steel stirrers, while feeding and decanting were carriedout by calibrated peristaltic pumps. In order to provide optimumconditions for each process, the mesophilic anaerobic digesterswere operated at 37 ± 0.5 �C, while the thermophilic at55 ± 0.5 �C. In both cases, temperatures were kept constant usingappropriate water baths.

2.3. Inoculation and experimental set-up

All reactors were initially seeded with an anaerobic inoculumobtained from a full-scale mesophilic anaerobic digester operatedin Athens WWTP. Then, they were flushed with N2 in order toremove oxygen and achieve anaerobic conditions. The substrateused for reactors’ feeding was a mixture of primary sludge and sec-ondary sludge (50:50 volume ratio), obtained from UniversityCampus WWTP and municipal WWTP, respectively. During feedevents, approximately a volume of 150 ml of mixed sludge waspumped to the system using peristaltic pumps. The operationalsequence was identical in all reactors and included feeding for10 min at a flow rate of 15 ml min�1 on a daily basis. The charac-teristics of raw sludge are presented in Table 1.

The operation of anaerobic digesters was divided into twophases. In the first phase (Phase I, 333 days), all four digestersoperated continuously as single-stage systems, at SRT of 20 daysand without spiking of the target analytes. In the second phase(Phase II, 133 days), the digesters T2 and M2 were connected in ser-ies and operated as a two-stage (dual) system (thermophilic/mes-ophilic) at SRT of 8 and 12 days, respectively. The remaining twodigesters (T1 and M1) continued to operate in a single mode at aSRT of 20 days in order to compare and control both processes(Fig. 1). During Phase II and after an acclimation period of fourSRT, target compounds were added into all systems. This wasaccomplished by adding daily a methanolic aliquot of micropollu-tants to the feeding substrate, in order to achieve an environmentalrelevant concentration of target compounds in raw sludge(2–4 lg g�1).

2.4. Analytical methodology

To monitor the performance of the anaerobic processes, rawand digested sludge samples were periodically taken from all bio-reactors during both experimental phases. Routine analyses, suchas CODt, soluble COD (CODs), total solids (TS), volatile solids (VS),TSS, VSS, pH, VFAs and alkalinity, were carried out in accordanceto Standard Methods (APHA, 2005) on a weekly basis. Moreover,daily biogas production was quantified by saline water displace-ment, while biogas composition was determined by a gas analyzer(Model GA 94A).

Fig. 1. Schematic diagram of the lab-scale anaerobic digesters used in this study.

Table 1Characteristics of the raw sludge (mixture of primary and secondary sludge, volume ratio 50:50) used for the feeding of lab-scale anaerobic digesters during the experiment.

pH TSin (g L�1) VSin (g L�1) TSSin (g L�1) VSSin (g L�1) CODt,in (g L�1)

Average 6.56 32.19 25.30 28.97 24.55 32.65Standard deviation 0.19 7.31 6.72 3.91 3.56 2.18

V.G. Samaras et al. / Bioresource Technology 162 (2014) 365–372 367

During Phase II, the concentration of the target compounds inraw and treated sludge was determined at regular intervals asdescribed below. Due to the high TS concentration in the liquidphase of raw sludge and the presence of fats and lipids that madedifficult the separation of dissolved and particulate phase, determi-nation of the total mass of target compounds in sludge samples wasmade after drying of 50 ml sludge. For this reason, sludge sampleswere oven dried in 60 �C until constant weight and stored at�18 �C before sample preparation. Thermal stability of the targetcompounds at 60 �C has been previously reported by Saleh et al.(2011) and Samaras et al. (2011).

The analytical method for the determination of EDCs and NSAIDsin sewage sludge samples was developed and optimized by theauthors (Samaras et al., 2011). In brief, the target analytes wereextracted by ultrasound sonication. The sonication was carriedout once at 50 �C for 30 min using a mixture of methanol andMilli-Q grade water as the extraction solvent. The supernatantwas collected after centrifuging and diluted to a final volume of100 ml using acidified Milli-Q grade water. Afterwards, it wasextracted using C18 cartridges. Cartridges were conditioned by3 � 2 ml of ethyl acetate, 3 � 2 ml of methanol and 3 � 2 ml Milli-Q water. The samples were passed through the cartridge with a flowrate of 5 ml min�1. Finally, cartridges were washed with 2 ml Milli-Q water and then they dried under vacuum for 60 min. The com-pounds were eluted with 3 � 2 ml of ethyl acetate at a flow rateof 1 ml min�1. The eluates were evaporated to dryness, under a gen-tle stream of nitrogen. Finally, the dried residues were subjected toderivatization reaction by adding a volume of 50 ll of BSTFA plus10 ll of pyridine. The analysis of silylated derivatives of the ana-lytes was performed by a Hewlett Packard Gas Chromatograph

5890 Series II connected to a Hewlett Packard Mass SpectrometerHP5971 MSD (USA). The separation of EDCs and NSAIDs wasachieved using a DB5MS capillary column (60 m) with a film thick-ness of 0.25 mm and an internal diameter of 0.32 mm (Supelco,USA). The analytical method presented satisfactory precision, withrelative standard deviations less than 12% for all the tested com-pounds. Satisfactory recoveries were obtained for target micropol-lutants, ranging from 78% to 104%. Limits of detection (LODs) variedbetween 15 ng g�1 (TCS) to 32.9 ng g�1 (BPA).

2.5. Calculations

The removal efficiencies of each target compound in AD sys-tems were calculated taking into account the analyte masses inraw (Min) and treated sludge (Mout), according to the followingequation:

Removalð%Þ ¼ ðMinfÞ � ðMeff ÞðMinfÞ

� 100

¼ ðQ in � CinÞ � ðQ out � CoutÞðQ in � CinÞ

� 100 ð1:1Þ

where Qin and Qout are the mass rates of raw sludge and treatedsludge, respectively (g TSS d�1) and Cin and Cout the concentrationsof target compounds in raw and treated sludge, respectively(lg g�1 TSS).

Target compounds are considered thermodynamically stableand relatively non-volatile, while photo degradation was prohib-ited by the experimental conditions used. Based on these, it could

368 V.G. Samaras et al. / Bioresource Technology 162 (2014) 365–372

be supposed that removal of these compounds was mainly due tobiodegradation/biotransformation.

The calculation of specific removal rate of each analyte in lab-scale anaerobic digesters was performed using Eq. (1.2):

ES ¼Min �Mout

XVSS � Vð1:2Þ

where ES is the analyte specific removal rate (lg g VSS�1 d�1), XVSS

is the concentration of VSS in the reactor (g L�1), and V is the reactorvolume (L).

3. Results and discussion

3.1. Operation of anaerobic digesters

The operational conditions applied to the reactors in Phase I andII, as well as their performance are summarized in Table 2.

During Phase I, all bioreactors were operated for a long periodas single-stage systems to ensure stable and favorable conditionsat a SRT of 20 days. According to the results, the average pH valuesin digesters remained in the neutral range, no accumulation ofVFAs was observed; whereas the mean ratio of VFAs to alkalinitywas lower than 0.5, indicating a relatively balanced process forall systems. The average daily biogas production was rangedbetween 0.84 and 0.96 L day�1; mean methane level was between56% and 64%, VSS removal was ranged from 31% (T1) to 46% (M2),while mean CODt removal was higher than 40% in all bioreactors(Table 2).

During Phase II, no obstruction was caused in bioreactors M2

and T2 after changing operating conditions and combining theminto a two-stage system. Moreover, in all bioreactors the processwas not affected by the addition of target compounds in rawsludge. Regarding the role of reactors’ configuration and tempera-ture on AD performance, a slight increase in CODt and VSS elimina-tion was observed in the two-stage system (T2–M2) comparing tosingle-stage digesters, M1 and T1 (Table 2, Fig. S1). Higher CODs,out

and VFAs concentrations (significant different using t-test, p < 0.05)were observed in the single-stage thermophilic system (T2) com-paring to the other systems, demonstrating higher hydrolysis ofthe organic material at high temperatures (Table 2). Regardingthe effect of SRT on AD performance, the elimination efficiency ofVSS and CODt was higher at higher SRT values (significant different

Table 2Operational conditions of lab-scale anaerobic digesters and characteristics of digested sludPhase II (19 weeks, n = 3 measurements per week).

Operational conditions M1 (Phase I) M1 (Phase II) M2 (Phase I)

T (�C) 37 ± 0.5 37 ± 0.5 37 ± 0.5SRT (d) 20 20 20pH 7.07 ± 0.05 7.09 ± 0.15 7.08 ± 0.05TSout (g L�1) 19.0 ± 4.7 18.4 ± 3.2 19.1 ± 3.1VSout (g L�1) 15.1 ± 3.0 14.4 ± 2.9 14.4 ± 3.9TSSout (g L�1) 15.9 ± 3.7 15.8 ± 1.2 15.7 ± 4.1VSSout (g L�1) 13.7 ± 5.7 12.1 ± 1.3 13.1 ± 4.2CODt,out (g L�1) 21.8 ± 6.0 18.3 ± 2.3 20.7 ± 5.0CODs,out (g L�1) 0.40 ± 0.06 0.39 ± 0.11 0.38 ± 0.05Alkalinity (g L�1 CaCO3) 2.1 ± 0.3 1.6 ± 0.3 4.2 ± 0.5VFAs (g/L�1) 1.3 ± 0.6 1.2 ± 0.1 1.6 ± 0.2VFAs/alkalinity 0.47 ± 0.18 0.46 ± 0.06 0.36 ± 0.02Daily biogas production (L day�1) 0.96 ± 0.19 0.92 ± 0.20 0.92 ± 0.22Biogas composition CH4 (%) 64 ± 4 68 ± 1 63 ± 7Biogas composition CO2 (%) 30 ± 7 35 ± 1 30 ± 11Removal TS (%) 41 ± 12 45 ± 15 41 ± 8Removal VS (%) 40 ± 12 44 ± 15 44 ± 10Removal TSS (%) 38 ± 10 45 ± 6 44 ± 12Removal VSS (%) 40 ± 9 51 ± 7 46 ± 10Removal CODt (%) 40 ± 14 44 ± 8 42 ± 17

using t-test, p < 0.05), while system stability was enhanced withSRT increase (Fig. S2).

3.2. Removal of target EDCs and NSAIDs during sludge AD

Fig. 2 shows the average concentrations (lg g�1) of target EDCsand NSAIDs in untreated and anaerobically digested sludge sam-ples, originated for the three AD systems used in Phase II. Accord-ing to the results, mean concentrations of target compounds inuntreated sludge ranged between 1.97 lg g�1 (DCF) and4.25 lg g�1 (BPA). Target NSAIDs were detected in treated sludgeat mean concentrations lower than 0.69 lg g�1 (NPX, T1); whereasfor some samples their concentrations were below method’s LODs.On the other hand, higher mean concentrations were determinedfor target EDCs, ranging up to 4.18 lg g�1 (NP, M1).

Calculation of micropollutants removal rates’ showed that alltarget compounds were significantly removed from the first dayof their addition and no acclimatization period was needed.According to Fig. 3, the highest removal rates were calculated forDCF, NPX, IBF and KFN (>80%), while mean removal rangingbetween 40% and 80% were observed for TCS, BPA and the sumof NP, NP1EO, NP2EO (NPE). Comparison with the literature showedthat NPX removal rate was consistent with the results obtained byCarballa et al. (2007) who reported removal efficiency between80% and 90%, under mesophilic and thermophilic conditions, whilein the current study higher removal efficiency was observed forDCF and IBF. The higher removal of these compounds could bedue to the lower concentrations of TS in reactors M1, T1 and T2–M2 (16.9–18.4 g L�1), comparing to 36.3–37.3 g L�1 reported inthe article of Carballa et al. (2007). Up to date, several authors havereported lower removal efficiency of micropollutants in the pres-ence of higher inert solid fraction (Chang et al., 2005; Carballaet al., 2007). To the best of our knowledge, so far, there is no datareporting the removal of KFN, BPA and TCS in lab-scale sludgeanaerobic digesters.

In the present study, the two-stage AD system (T2–M2) pre-sented similar removal rates with the single-stage mesophilic(M1) and thermophilic (T1) system for the target NSAIDs (Fig. 3).In contrast, the combination of thermophilic and mesophilic anaer-obic digesters slightly favored the removal of EDCs comparing tothe conventional mesophilic and thermophilic system (no statisti-cal significant difference using t-test). It should be noted thataccording to the results presented in Paragraph 3.1, this system

ge and produced biogas during Phase I (48 weeks, n = 3 measurements per week) and

T2–M2 (Phase II) T1 (Phase I) T1 (Phase II) T2 (Phase I) T2 (Phase II)

37 ± 0.5 55 ± 0.5 55 ± 0.5 55 ± 0.5 55 ± 0.512 20 20 20 87.18 ± 0.12 7.37 ± 0.04 7.18 ± 0.09 7.37 ± 0.07 7.13 ± 0.1116.9 ± 1.4 21.7 ± 4.2 17.2 ± 1.1 23.1 ± 3.0 20.3 ± 5.311.7 ± 2.9 15.5 ± 4.1 13.5 ± 2.2 17.7 ± 4.5 16.4 ± 3.512.3 ± 2.7 18.3 ± 3.7 13.5 ± 1.1 21.6 ± 4.8 18.1 ± 9.09.6 ± 2.0 14.8 ± 4.7 10.5 ± 1.1 17.2 ± 4.5 15.1 ± 5.415.2 ± 1.8 24.5 ± 8.0 16.8 ± 1.5 25.3 ± 7.4 21.3 ± 4.90.46 ± 0.27 1.45 ± 0.03 1.55 ± 0.01 1.25 ± 0.04 1.44 ± 0.193.8 ± 0.3 4.3 ± 0.5 3.7 ± 0.6 4.4 ± 0.3 3.1 ± 1.11.1 ± 0.2 1.2 ± 0.0 1.5 ± 0.6 1.5 ± 0.0 1.6 ± 0.20.32 ± 0.05 0.27 ± 0.04 0.47 ± 0.18 0.33 ± 0.02 0.25 ± 0.360.87 ± 0.23 0.84 ± 0.29 0.97 ± 0.16 0.85 ± 0.28 0.95 ± 0.2871 ± 1 61 ± 10 56 ± 1 56 ± 8 61 ± 138 ± 1 31 ± 4 29 ± 1 31 ± 11 27 ± 150 ± 10 38 ± 9 49 ± 7 32 ± 10 41 ± 1352 ± 15 35 ± 12 49 ± 10 33 ± 6 42 ± 1957 ± 8 36 ± 4 53 ± 5 27 ± 10 40 ± 2758 ± 10 40 ± 4 57 ± 5 31 ± 5 46 ± 1353 ± 17 48 ± 8 48 ± 8 43 ± 2 43 ± 2

Fig. 2. Concentrations of target EDCs and NSAIDs in raw and treated sludge of (a) mesophilic digester, M1, (b) thermophilic digester, T1 and (c) thermophilic/mesophilicsystem, T2–M2 (n = 11).

V.G. Samaras et al. / Bioresource Technology 162 (2014) 365–372 369

achieved slightly better removal of TSS and CODt. Regarding therole of each digester in T2–M2 system, excepting NPE, the majorpart of target compounds was removed in the thermophilic reactor(Fig. 4). On the other hand, the removal of NPE was almost equalbetween two reactors, indicating the important role of mesophilicreactor in two-stage system for these compounds.

To investigate the role of SRT on micropollutants removal, theremoval efficiency of target compounds was compared in T1

(SRT: 20 days) and T2 (SRT (8 days). According to the results pre-sented in Fig. S3, the mean removal of most micropollutants wasslightly increased by SRT increase (no statistical significantdifference using t-test). So far, contradictory results have been

0 20 40 60 80 100 120

Sum NPE

TCS

BPA

KFN

DCF

NPX

IBF

T2-M2 T1 M1

Fig. 3. Removal of EDCs and NSAIDs in single-stage mesophilic (M1), single-stagethermophilic (T1) and two-stage thermophilic–mesophilic (T2–M2) AD system.

370 V.G. Samaras et al. / Bioresource Technology 162 (2014) 365–372

reported for the role of SRT on micropollutants removal duringsludge AD (Benabdallah El-Hadj et al., 2006; Carballa et al., 2007).

3.3. Fate of NPE under different AD treatment systems

In this study, NP, NP1EO and NP2EO were the only compoundsfor which negative removal rates were calculated on certain daysunder thermophilic and mesophilic conditions. These negativeremoval rates resulted to the high standard deviations presentedin Table 3. Among the three NPE, NP2EO showed the highest aver-age removal rate in all anaerobic systems (67–91%). The removal

Fig. 4. Contribution of each reactor in the removal of EDCs and NSAID

Table 3Average removal rates and standard deviations of the target NPE in the three anaerobic d

Compound M1 (SRT: 20d) T1 (SRT: 20d)

Removal (%) Standard deviation Removal (%) Standard deviat

NP 33 67 79 13NP1EO 60 40 11 60NP2EO 67 20 75 18

efficiency of the other two compounds was significantly affectedby the anaerobic system used. Specifically, higher removal ratesfor NP1EO were calculated in the mesophilic single-stage and thetwo-stage system (60% and 54%, respectively), while NP was elim-inated at a greater extent in the thermophilic single-stage and two-stage system (79%).

The increase of NPE concentration during anaerobic digestionhas also been reported in some previous studies (Hernandez-Raquet et al. 2007). The presence of several precursors in sewagesludge (e.g. nonylphenol polyethoxylates) complicates the discus-sion for NPE fate during AD and as a result contradictory resultshave been presented in the literature. Up to date, most authorsagree that NP1EO and NP2EO are biodegraded to some extent dur-ing sludge AD and contribute to the formation of NP (Minamiyamaet al., 2006; Patureau et al., 2008; McNamara et al., 2012; Bozkurtand Sanin, 2013). According to Hernandez-Raquet et al. (2007) andJanex-Habibi et al. (2009), NP does not undergo further transfor-mation and consequently accumulates in biosolids. On the con-trary, other researchers have reported the biological degradationof NP during sludge AD (Chang et al., 2005; Patureau et al., 2008;Paterakis et al., 2012). In the literature, several factors have beenreported to affect the removal of NPE. Recently, Paterakis et al.(2012) reported higher removal rate of NP1–2EO in a mixture ofprimary and secondary sludge comparing to primary sludge. More-over, it is generally accepted that the degradation of lipophiliccompounds such as NPE takes place at a slower rate in the pres-ence of natural organic matter which reduces bioavailability anddiffusion of the target compounds. However, the characteristicsand composition of the feed sludge vary considerably in most stud-ies and as a result; it is difficult to draw conclusions about theireffect (Barret et al., 2012).

Fig. 5 shows the contribution of NP, NP1EO and NP2EO in thetotal mass of NPE in raw and treated sludge for the AD systems usedin this study. According to the results, the reduction of NPE in thesingle-stage mesophilic system (M1) is mainly due to the significantelimination of NP1EO, NP2EO and to a lesser extent to the reductionof NP. In contrast, the reduction of NPE in the thermophilic single-stage system (T1) seems to be mainly due to the degradation ofNP2EO and NP, whereas NP1EO is accumulated. In the two-stage

s for the two-stage thermophilic–mesophilic (T2–M2) AD system.

igestion systems.

T2–M2 (SRTs: 8d-12d) T2 (SRT: 8d)

ion Removal (%) Standard deviation Removal (%) Standard deviation

79 18 67 4054 60 �1 6191 8 62 33

0

5

10

15

20

25

30

35

40

45

Inlet 1 M1 T1 Inlet 2 T2 T2-M2

Mas

s(g

d-1)

NP2EONP1EONP

Fig. 5. Masses (g d�1) of NP, NP1EO and NP2EO in raw and treated sludge of single-stage mesophilic (M1), single-stage thermophilic (T1) and two-stage thermophilic–mesophilic (M2–T2) AD system.

V.G. Samaras et al. / Bioresource Technology 162 (2014) 365–372 371

AD system (T2–M2), there is a significant elimination of all threeNPE, while the mesophilic reactor (M2) appears to have a significantcontribution on the removal of NP1EO. The above results indicatethat all three NPE can be biotransformed during sludge AD, whileit is clear that the temperature applied to bioreactors affects differ-ently each of them. Higher removal of NP1EO and accumulation ofNP is observed under mesophilic conditions; whereas accumulationof NP1EO and higher removal of NP is achieved under thermophilicconditions. Further experiments should be performed to investigatewhether the accumulation of NP1EO in thermophilic conditions isdue to the greater conversion of precursor compounds found insludge comparing to mesophilic conditions or due to the interrup-tion/decrease of its transformation to NP.

3.4. Removal rates of target EDCs and NSAIDs during sludge AD

Specific removal rates (ES) for each analyte were calculatedaccording to Eq. (1.2). Excepting NP2EO, higher mean removal ratesfor all target compounds were calculated in bioreactor T1 underthermophilic conditions and SRT of 20 days (Fig. 6). For NP1EO, ES

calculated in this bioreactor was significantly different compared

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.80

0.90

NP2EO NP1EO NP TCS BPA KFN DCF NPX

Spec

ific

Rem

oval

Rat

e(μ

g g

VSS

-1 d

ays-1

)

M1 T1 T2

Fig. 6. Specific removal rates (lg g VSS�1 d�1) of the target micropollutants in thethree AD systems.

to those calculated in other AD systems; whereas no statisticallysignificant differences were observed for the other compounds.

4. Conclusions

The performance of all anaerobic digesters was efficientthroughout the entire period of the study, while their operationat higher SRT resulted to better systems’ performance and stability.In all systems, it was observed mean removal of DCF, KFN, NPX andIBF higher than 80% and mean removal of TCS, BPA and NPEbetween 40% and 80%. The use of the two-stage AD system slightlyfavored the removal of EDCs; whereas did not affect the removal ofNSAIDs. NP2EO, NP1EO and NP can be biodegraded during sludgeAD; however the temperature applied to bioreactors affects differ-ently their removal.

Acknowledgements

This research has been co-financed by the European Union(European Social Fund - ESF) and Greek national funds throughthe Operational Program "Education and Lifelong Learning" of theNational Strategic Reference Framework (NSRF) - ResearchFunding Program: THALES. Investing in knowledge throughthe European Social Fund. WATERMICROPOL (www.aegean.gr/environment/watermicropol). The authors would like to thankMrs Alexandra Mathiopoulou, Mrs Ioanna Sirigou, Mrs KorinaBertoli and Mrs Niki Spirou for their valuable help during theexperiments.

Appendix A. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at http://dx.doi.org/10.1016/j.biortech.2014.03.154.

References

APHA, 2005. Standards Methods for the Examination of Water and Wastewater,twenty first ed. American Public Health Association, USA.

Appels, L., Baeyens, J., Degrève, J., Dewil, R., 2008. Principles and potential of theanaerobic digestion of waste-activated sludge. Prog. Energy Combust. Sci. 34,755–781.

372 V.G. Samaras et al. / Bioresource Technology 162 (2014) 365–372

Barret, M., Delgadillo-Mirquez, L., Trably, E., Delgenes, N., Braun, F., Cea-Barcia,G., Steyer, J.P., Patureau, D., 2012. Anaerobic removal of trace organiccontaminants in sewage sludge: 15 years of experience. Pedosphere 22,508–517.

Benabdallah El-Hadj, T., Dosta, J., Mata-Alvarez, J., 2006. Biodegradation of PAH andDEHP micro-pollutants in mesophilic and thermophilic anaerobic sewagesludge digestion. Water Sci. Technol. 53, 99–107.

Bozkurt, H., Sanin, F.D., 2013. Toxicity of nonylphenol diethoxylate in lab-scaleanaerobic digesters. Chemosphere 104, 69–75.

Cao, Y., Pawłowski, A., 2012. Sewage sludge-to-energy approaches based onanaerobic digestion and pyrolysis: brief overview and energy efficiencyassessment. Renew. Sust. Energy Rev. 16, 1657–1665.

Carballa, M., Omil, F., Alder, A.C., Lema, J.M., 2006. Comparison between theconventional anaerobic digestion of sewage sludge and its combination witha chemical or thermal pre-treatment concerning the removal ofpharmaceuticals and personal care products. Water Sci. Technol. 53, 109–117.

Carballa, M., Manterola, G., Larrea, L., Ternes, T., Omil, F., Lema, J.M., 2007. Fate ofpharmaceutical and personal care products (PPCPs) during anaerobic digestionof sewage sludge. Water Res. 41, 2139–2150.

Chang, B.V., Chiang, F., Yuan, S.Y., 2005. Anaerobic degradation of nonylphenol insludge. Chemosphere 59, 1415–1420.

Chen, Y., Cheng, J.J., Creamer, K.S., 2008. Inhibition of anaerobic digestion process: areview. Bioresour. Technol. 99, 4044–4064.

Ge, H., Jensen, P.D., Batstone, D.J., 2010. Pre-treatment mechanisms duringthermophilic–mesophilic temperature phased anaerobic digestion of primarysludge. Water Res. 44, 123–130.

González, M.M., Martín, J., Santos, J.L., Aparicio, I., Alonso, E., 2010. Occurrence andrisk assessment of nonylphenol and nonylphenol ethoxylates in sewage sludgefrom different conventional treatment processes. Sci. Total Environ. 408, 563–570.

Hernandez-Raquet, G., Soef, A., Delgenes, N., Balaguer, P., 2007. Removal of theendocrine disrupter nonylphenol and its estrogenic activity in sludge treatmentprocesses. Water Res. 41, 2643–2651.

Janex-Habibi, M.L., Huyard, A., Esperanza, M., Bruchet, A., 2009. Reduction ofendocrine disruptor emissions in the environment: the benefit of wastewatertreatment. Water Res. 43, 1565–1576.

Kelessidis, A., Stasinakis, A.S., 2012. Comparative study of the methods used fortreatment and final disposal of sewage sludge in European countries. WasteManage. 32, 1186–1195.

McNamara, P.J., Wilson, C.A., Wogen, M.T., Murthy, S.N., Novak, J.T., Novak, P.J.,2012. The effect of thermal hydrolysis pretreatment on the anaerobicdegradation of nonylphenol and short-chain nonylphenol ethoxylates indigested biosolids. Water Res. 46, 2937–2946.

Minamiyama, M., Ochi, S., Suzuki, Y., 2006. Fate of nonylphenol polyethoxylates andnonylphenoxy acetic acids in an anaerobic digestion process for sewage sludgetreatment. Water Sci. Technol. 53, 221–226.

Narumiya, M., Nakada, N., Yamashita, N., Tanaka, H., 2013. Phase distribution andremoval of pharmaceuticals and personal care products during anaerobicsludge digestion. J. Hazard. Mater. 260, 305–312.

Paterakis, N., Chiu, T.Y., Koh, Y.K.K., Lester, J.N., McAdam, E.J., Scrimshaw, M.D.,Soares, A., Cartmell, E., 2012. The effectiveness of anaerobic digestion inremoving estrogens and nonylphenolethoxylates. J. Hazard. Mater. 199–200,88–95.

Patureau, D., Delgenes, N., Delgenes, J.-P., 2008. Impact of sewage sludge treatmentprocesses on the removal of the endocrine disrupters nonylphenol ethoxylates.Chemosphere 72, 586–591.

Rubio-Loza, A., Noyola, A., 2010. Two-phase (acidogenic–methanogenic) anaerobicthermophilic/mesophilic digestion system for producing Class A biosolids frommunicipal sludge. Bioresour. Technol. 101, 576–585.

Saleh, A., Larsson, E., Yamini, Y., Jönsson, J.Å., 2011. Hollow fiber liquid phasemicroextraction as a preconcentration and clean-up step after pressurized hotwater extraction for the determination of non-steroidal anti-inflammatorydrugs in sewage sludge. J. Chromatogr. A 1218, 1331–1339.

Samaras, V.G., Thomaidis, N.S., Stasinakis, A.S., Lekkas, T.D., 2011. An analyticalmethod for the simultaneous trace determination of acidic pharmaceuticalsand phenolic endocrine disrupting chemicals in wastewater and sewagesludge by gas chromatography–mass spectrometry. Anal. Bioanal. Chem.399, 2549–2561.

Samaras, V.G., Stasinakis, A.S., Mamais, D., Thomaidis, N.S., Lekkas, T.D., 2013. Fateof selected pharmaceuticals and synthetic endocrine disrupting compoundsduring wastewater treatment and sludge anaerobic digestion. J. Hazard. Mater.244–245, 259–267.

Stasinakis, A.S., 2012. Review on the fate of emerging contaminants during sludgeanaerobic digestion. Bioresour. Technol. 121, 432–440.

Stasinakis, A.S., Thomaidis, N.S., Arvaniti, O.S., Asimakopoulos, A.G., Samaras, V.G.,Ajibola, A., Mamais, D., Lekkas, T.D., 2013. Contribution of primary andsecondary treatment on the removal of benzothiazoles, benzotriazoles,endocrine disruptors, pharmaceuticals and perfluorinated compounds in asewage treatment plant. Sci. Total Environ. 463–464, 1067–1075.