Seventeen pharmaceutically active compounds and 22 other trace organic pollutants were analysed regularly in the influent and permeate from a semi-real plant treating municipal wastewater. The plant was operated during 29 months with different configurations which basically differed in the type of biomass present in the system. These processes were the integrated fixed-film activated sludge membrane bioreactor (IFAS-MBR), which combined suspended and attached biomass, the moving bed membrane bioreactor (MBMBR) (only attached biomass) and the MBR (only suspended biomass). Moreover, removal rates were compared to those of the wastewater treatment plant (WWTP) operating nearby with conventional activated sludge treatment. Reverse osmosis (RO) was used after the pilot plant to improve removal rates. The highest elimination was found for the IFAS-MBR, especially for hormones (100% removal); this was attributed to the presence of biofilm, which may lead to different conditions (aerobic–anoxic–anaerobic) along its profile, which increases the degradation possibilities, and also to a higher sludge age of the biofilm, which allows complete acclimation to the contaminants. Operating conditions played an important role, high mixed liquor suspended solids (MLSS) and sludge retention time (SRT) being necessary to achieve these high removal rates. Although pharmaceuticals and linear alkylbenzene sulfonates showed high removal rates (65–100%), nonylphenols and phthalate could only be removed to 10–30%. RO significantly increased removal rates to 88% mean removal rate.

INTRODUCTION

Concern about the so-called emerging pollutants has increased in the last decade as it is generally known that these compounds may affect negatively aquatic life (Santos et al. 2007) and also human health if these compounds reach the water supply sources at significant levels. One important gateway for these compounds is wastewater treatment plants (WWTPs) (Zhang et al. 2008). Emerging micropollutants are slowly or non-biodegradable compounds and are able to pass the biological treatment steps and come out in the effluent in relevant concentrations. There exist several studies dealing with the elimination of recalcitrant compounds using different WWTP schemes, some of them comparing the removal rates found in a membrane bioreactor (MBR) to those of a conventional activated sludge (CAS) system (Sahar et al. 2011; De Wever et al. 2007). The conclusions obtained were that, working at the same operating conditions, the MBR improved neither the elimination of those compounds which are highly biodegradable nor that of the recalcitrant ones (Bouju et al. 2009), but an improvement can be observed for those compounds showing an intermediate elimination rate. Reasons for that may be the absence of suspended solids in the effluent (Clara et al. 2004) and also a larger fraction of planktonic microorganisms and smaller flocs in the MBR (10–100 mm in the MBR compared to 100–500 mm for CAS), which facilitates mass transfer (Cicek et al. 1999). For a higher elimination of recalcitrant compounds (80–100%), reverse osmosis (RO), adsorption (for instance activated carbon) and advanced oxidation technologies belong to the most popular choices. The bibliography on micropollutant removal using other treatment processes such as biofilm systems is scarce. In biofilm systems, the biomass is attached to the carrier and is thus better protected against toxic events (Butler & Boltz 2014; Kriklavova et al. 2012). Moreover, nitrification is favoured even at low temperature, as the residence time of the attached bacteria in the system is longer than in the suspended form (Regmi et al. 2011). In MBR systems, the settling tank is substituted by a membrane so that the effluent quality is increased in terms of pathogen and solids removal. The combination of these two technologies (biofilm systems and MBR) was already reported by Liu et al. (2010), who introduced carriers in an MBR and operated the system for 400 d. When the biofilm had formed, the permeability of the system improved substantially as well as the nitrification efficiency. Leiknes & Ødegaard (2007) performed numerous studies combining moving bed bioreactor (MBBR) and MBR, but none of them analysed organic micropollutants in their systems.

In this study, three processes were compared in terms of organic micropollutant removal: the MBR, the integrated fixed-film activated sludge membrane bioreactor (IFAS-MBR) and the moving bed membrane bioreactor (MBMBR). The main differences between the processes is the activated sludge present in the system, which in the case of the MBR is suspended in the bioreactor and in the case of the MBMBR is attached to a carrier forming a biofilm. The IFAS-MBR would be an intermediate case, part of the biomass being suspended in the reactor and part of it attached to the carriers. The comparison was performed in a semi-real pilot plant treating 2.3 m3/h of municipal wastewater during 29 months and included comparison with the removal rates of the posterior RO treatment and the conventional activated sludge full-scale.

METHODS

The pilot plant

The three treatment schemes evaluated are presented in Figure 1, where AX stands for the anoxic and AE for the aerobic chamber respectively.

Figure 1

Studied MBR treatment schemes: MBR (left), IFAS-MBR (middle) and MBMBR (right).

Figure 1

Studied MBR treatment schemes: MBR (left), IFAS-MBR (middle) and MBMBR (right).

The influent was taken from the wastewater treatment plant of Almuñécar (Granada), which is operating with CAS technology. After sand and grease removal, a constant flow of 2.3 m3/h wastewater was biologically treated in the pilot plant and separated by a hollow fiber membrane module from Koch Membrane Systems (Puron, Germany) with a pore size of 0.01 μm. Before entering the RO system, an antiscalant was dosed to the hollow fibre effluent (Osmotec 1,261, BKG Water Solutions, Germany) and the pH was adjusted to 7.1 using sulfuric acid. The recovery of the RO was restricted to 40% due to the mechanical limitations of the pilot plant. The RO membrane used was a low fouling membrane made of aromatic polyamide–urea from TRISEP (USA). The pilot plant was operated as an IFAS-MBR during 9 months, as an MBMBR during 5 months and during 15 months as a conventional MBR (12 months at the beginning and 3 months in the end). The reverse osmosis system was in operation during 16 months in between.

For the IFAS-MBR and the MBMBR configurations, 50% of the volume of the aerobic reactor of this MBR was filled with high density polypropylene plastic carriers from Christian Stöhr (Germany) with high surface density. As was previously mentioned, nitrification is favoured in biofilm systems and therefore the carriers were added in the aerobic chamber to improve nitrification, which is a common application of MBBR and IFAS systems. Screens were located in the outlet and inlet of the aerobic tank in order to avoid losses of carriers. Especially important is to avoid the entrance of carriers into the membrane tank, as the impact of the carriers could potentially damage the membrane. As the MBR was already in operation, no inoculum was needed for the start-up of the IFAS-MBR phase. For the start-up of the MBMBR phase, the tanks were emptied in order to drain the suspended activated sludge.

After 7 months operating as an IFAS-MBR at an SRT of 10 d, the SRT of the plant was increased to 20 d by changing the sludge wastage and thus the mixed liquor suspended solids (MLSS) in order to study the system at different operational conditions.

Analytical methods

Dissolved oxygen (DO), temperature and pH were regularly monitored in the plant. 24-h composite influent and effluent samples from each experimental installation were collected on a daily basis using a time controller (4 h). Chemical oxygen demand (COD) was determined by colorimetric methods (Hach-Lange, Germany). MLSS were measured by gravimetric determination. All analytical methods were applied according to Standard Methods (APHA 1998).

Several micropollutants were monitored during the study (Table 2). Pharmaceutical compounds (PhACs), linear alkylbenzene sulfonates (LASs), nonylphenols (NPEs), monoethoxylate (NP1EO) and diethoxylate (NP2EO), and di-(2-ethylhexyl)phthalate (DEHP) were extracted by solid-phase extraction (SPE) and determined with high performance liquid chromatography (HPLC) and detected with mass spectrometry. Recovery (R), limit of detection (LOD) and limit of quantification (LOQ) are shown in Table 2 for samples analysed after 2011. Samples analysed before that were analysed with high performance liquid chromatography with diode array and fluorescence detectors sited on line and the R, LOD and LOQ of these methods are described in Camacho-Muñoz et al. (2012). Polycyclic aromatic hydrocarbons (PAHs) were extracted by SPE and determined by HPLC and detected with UV and fluorescence simultaneous detection. The studied PAHs were: Naphthalene (Naf), Acenaphthene (Ace), Fluorene (Flu), Phenanthrene (Fen), Anthracene (Ant), Fluoranthene (Fluo), Pyrene (Pir), 1,2-Benzanthracene (BaA), Chrysene (Cri), Benzo(b)fluoranthene, (BbF), Benzo(k)fluoranthene (BkF), Benzo(a)pyrene (BaP), 1,2,5,6-Dibenzanthracene (DahA), Indeno(1,2,3-C.D) pyrene (I123cdP) and 1,12-Benzopyrelene (BghiP).

RESULTS AND DISCUSSION

Operation and effluent COD results

In Table 1, the operating conditions of the different biological treatment schemes evaluated (hydraulic retention time (HRT), sludge retention time (SRT), MLSS, temperature and COD) and the COD values of their effluents are presented. As can be seen, as the MBMBR operates without recirculation, the MLSS values of the aerobic tank are much lower than for the rest of the configurations. However, the COD values in the effluent showed that the attached biomass was enough to achieve a high-quality effluent.

Table 1

Operating parameters of the plant during the different periods

  SRT MLSS aerobic tank MLSS membrane tank HRT Total COD influent COD filtered influent pH COD effluent 
 g/L g/L °C mg/L mg/L – mg/L 
IFAS-MBR 10 10 2.5 7.3 13 18–27 3,138 168 7.6 13.7 
IFAS-MBR 20 20 5.3 8.0 13 16–20 1,022 56 7.4 10.7 
MBMBR – 0.3 9.7 13–23 1,147 66 7.4 14.6 
MBR 20 7.4 10.6 13 12–30 1,233 176 7.2 17.6 
  SRT MLSS aerobic tank MLSS membrane tank HRT Total COD influent COD filtered influent pH COD effluent 
 g/L g/L °C mg/L mg/L – mg/L 
IFAS-MBR 10 10 2.5 7.3 13 18–27 3,138 168 7.6 13.7 
IFAS-MBR 20 20 5.3 8.0 13 16–20 1,022 56 7.4 10.7 
MBMBR – 0.3 9.7 13–23 1,147 66 7.4 14.6 
MBR 20 7.4 10.6 13 12–30 1,233 176 7.2 17.6 

Trace organic pollutants in the influent

In the 104 composite samples analysed throughout the experimentation period, 16 of the 17 pharmaceutical compounds analysed (all except for estrone) were detected. The concentration of the compounds showed high variability, as can be seen from the standard deviation values (Table 2). The compounds with the highest concentrations were the anti-inflammatory drugs, reaching average concentrations of 59.7 μg/L for acetaminophen and 24.5 μg/L for ibuprofen. Although hormone concentrations are relatively low (0.7–2.1 μg/L), it is known that hormones may cause adverse effects in aquatic life at levels of only few ng/L (Purdom et al. 1994). The concentration of LASs were extremely high compared to PhACs, reaching concentrations up to 1,967 μg/L for C11, probably because LASs are common products present in domestic detergents. Nevertheless, they are highly biodegradable compounds and their concentrations along the biological treatment decreased significantly. Only some of the polycyclic aromatic hydrocarbons were detected and most of the samples were below LOD.

Table 2

Influent concentration

 Influent (μg/L)
 
Analysis method
 
Group Compound Average STD Max Min R (%) LOD (ng/L) LOQ (ng/L) 
PhAC Acetaminophen 59.7 62.8 284.0 <LOD 95.6 15 44 
Diclofenac 1.1 4.3 38.0 <LOD 89.3 11 43 
Ibuprofen 24.5 25.7 111.0 <LOD 82.3 
Ketoprofen 1.5 6.0 44.3 <LOD 96.2 21 
Naproxen 3.1 4.4 34.9 <LOD 68.7 16 48 
Salicylic acid 13.9 13.1 57.0 <LOD 94.0 10 21 
Sulfamethoxazole 0.3 1.8 17.5 <LOD 66.2 0.4 
Trimethoprim 0.2 1.1 9.8 <LOD 70.6 0.2 
Carbamazepine 0.9 2.5 16.9 <LOD 75.0 0.1 
Propanolol 0.4 1.2 13.4 <LOD 91.0 0.1 
Caffeine 1.8 2.8 22.5 <LOD 114.5 25 
17α-Ethinylestradiol 2.1 5.0 29.2 <LOD 85.1 20 59 
17β-Estradiol 0.7 2.5 21.2 <LOD 86.9 14 
Estriol 1.3 4.1 36.6 <LOD 85.0 22 75 
Estrone <LOD <LOD <LOD <LOD 79.6 25 
Clofibric acid 5.2 19.4 138.3 <LOD 79.3 22 
Gemfibrozil 3.9 18.9 217.0 <LOD 95.6 0.02 0.1 
LAS C10 108.3 195.7 1,125.0 <LOD 80 8.1 27 
C11 304.0 321.8 1,967.0 <LOD 75 4.31 14.4 
C12 247.4 316.2 1,759.0 <LOD 73 375 1,249 
C13 151.3 166.0 806.0 <LOD 70 0.33 1.11 
NPE NP2EO 2.4 5.7 59.7 <LOD 79 0.17 0.58 
NP1EO 5.2 13.2 145.8 <LOD 82 0.29 0.97 
NP 1.5 4.6 28.7 <LOD 80 0.64 2.12 
DEHP DEHP 27.1 60.3 379.3 <LOD 72 0.08 0.25 
PAH Naf 0.06 0.27 3.1 <LOD 90 29 92 
Ace + Flu <LOD 0.09 0.79 <LOD 72 10 
Fen <LOD 0.12 1.5 <LOD 99 10 
Ant <LOD <LOD 0.0 <LOD 70 
Fluo <LOD <LOD 0.41 <LOD 68 12 
Pir <LOD <LOD 0.36 <LOD 75 78 29 
BaA <LOD 0.07 0.82 <LOD 93 21 
Cri <LOD <LOD 0.28 <LOD 91 0.2 
BbF <LOD <LOD <LOD <LOD 84 
BkF <LOD <LOD <LOD <LOD 55 
BaP <LOD <LOD <LOD <LOD 50 24 40 
DahA <LOD <LOD <LOD <LOD 82 11 36 
BghiP <LOD <LOD 0.43 <LOD 87 45 150 
I123cdP <LOD <LOD <LOD <LOD 51 29 92 
 Influent (μg/L)
 
Analysis method
 
Group Compound Average STD Max Min R (%) LOD (ng/L) LOQ (ng/L) 
PhAC Acetaminophen 59.7 62.8 284.0 <LOD 95.6 15 44 
Diclofenac 1.1 4.3 38.0 <LOD 89.3 11 43 
Ibuprofen 24.5 25.7 111.0 <LOD 82.3 
Ketoprofen 1.5 6.0 44.3 <LOD 96.2 21 
Naproxen 3.1 4.4 34.9 <LOD 68.7 16 48 
Salicylic acid 13.9 13.1 57.0 <LOD 94.0 10 21 
Sulfamethoxazole 0.3 1.8 17.5 <LOD 66.2 0.4 
Trimethoprim 0.2 1.1 9.8 <LOD 70.6 0.2 
Carbamazepine 0.9 2.5 16.9 <LOD 75.0 0.1 
Propanolol 0.4 1.2 13.4 <LOD 91.0 0.1 
Caffeine 1.8 2.8 22.5 <LOD 114.5 25 
17α-Ethinylestradiol 2.1 5.0 29.2 <LOD 85.1 20 59 
17β-Estradiol 0.7 2.5 21.2 <LOD 86.9 14 
Estriol 1.3 4.1 36.6 <LOD 85.0 22 75 
Estrone <LOD <LOD <LOD <LOD 79.6 25 
Clofibric acid 5.2 19.4 138.3 <LOD 79.3 22 
Gemfibrozil 3.9 18.9 217.0 <LOD 95.6 0.02 0.1 
LAS C10 108.3 195.7 1,125.0 <LOD 80 8.1 27 
C11 304.0 321.8 1,967.0 <LOD 75 4.31 14.4 
C12 247.4 316.2 1,759.0 <LOD 73 375 1,249 
C13 151.3 166.0 806.0 <LOD 70 0.33 1.11 
NPE NP2EO 2.4 5.7 59.7 <LOD 79 0.17 0.58 
NP1EO 5.2 13.2 145.8 <LOD 82 0.29 0.97 
NP 1.5 4.6 28.7 <LOD 80 0.64 2.12 
DEHP DEHP 27.1 60.3 379.3 <LOD 72 0.08 0.25 
PAH Naf 0.06 0.27 3.1 <LOD 90 29 92 
Ace + Flu <LOD 0.09 0.79 <LOD 72 10 
Fen <LOD 0.12 1.5 <LOD 99 10 
Ant <LOD <LOD 0.0 <LOD 70 
Fluo <LOD <LOD 0.41 <LOD 68 12 
Pir <LOD <LOD 0.36 <LOD 75 78 29 
BaA <LOD 0.07 0.82 <LOD 93 21 
Cri <LOD <LOD 0.28 <LOD 91 0.2 
BbF <LOD <LOD <LOD <LOD 84 
BkF <LOD <LOD <LOD <LOD 55 
BaP <LOD <LOD <LOD <LOD 50 24 40 
DahA <LOD <LOD <LOD <LOD 82 11 36 
BghiP <LOD <LOD 0.43 <LOD 87 45 150 
I123cdP <LOD <LOD <LOD <LOD 51 29 92 

Trace organic pollutant removal rates

Removal rates for all processes were depicted as box-and-whisker plots in Figure 2 for the PhACs, NPEs and DEHP and in Figure 3 for the PAHs. Error bars reflect min and max values. The lower side of the box represents the 5th percentile, the upper side of the boxes is the 95th percentile and the intersection of the boxes represents the median of the data. The points represent the mean values. All removal rates are referred to the concentration in the influent. Nonylphenols and DEHP showed high concentrations both in the influent (Table 2) and the effluent for all process configurations studied, showing for all configurations severe difficulties to remove these compounds; even after RO treatment, the removal rates are only 70 to 80% (Figure 2). Nonylphenol and DEHP belong to the priority substances listed in the EU Waterframework Directive (WFD) (Directive 2008/105/EC), which means that their concentration shall be controlled in the near future in surface water bodies. Recently, diclofenac, 17α-ethinylestradiol and 17β-estradiol were proposed to be included in the WFD in a ‘watch list’ and may be added to this priority substances list in the future. For these priority substances and depending on the receiving media, biological treatment may not be enough to comply with regulation and further treatment such as RO, adsorption or advanced oxidation could be needed. Although PAHs belong also to the list of priority substances, their concentration constantly remained under the limits reflected in the WFD, as can be seen from Table 2 and Figure 3.

Figure 2

Removal rates for the different treatment processes for PhACs, LASs, NPEs and DEHP.

Figure 2

Removal rates for the different treatment processes for PhACs, LASs, NPEs and DEHP.

Figure 3

Removal rates for the different treatment processes for PAHs.

Figure 3

Removal rates for the different treatment processes for PAHs.

As can be seen in Figures 2 and 3, the process showing the best performance was the IFAS-MBR, operating at 20 d SRT. Its removal rates are not only higher but also less scattered compared to the rest of the processes and the boxes are smaller, indicating a more robust process. Operating conditions seemed to be critical for this process; when the system was operated at lower SRT and MLSS the removal rates achieved by the IFAS-MBR were much lower. An increase in removal for longer SRT and higher MLSS has been shown in other studies (Clara et al. 2005) and it is related to a better acclimation of the biomass to the contaminants due to a longer residence time of the sludge and also to the presence of slower growing species, which may play an important role in degradation of some compounds. Moreover, higher MLSS is related to lower F/M ratios, which may induce microorganisms to metabolise poorly degradable compounds due to food shortage (Verlicchi et al. 2012). Conversely, biofilms need long maturation time and in the first months of operation the low removal rates achieved may also have been a consequence of an immature biofilm, as these first months coincided with the IFAS-MBR 10 d phase. Especially surprising is the high removal rate for the IFAS-MBR20 for carbamazepine, where 88% removal was achieved with this process. The removal rates found in the literature for this compound normally range between 0 and 13%, as it is highly recalcitrant and not biodegradable (Tadkaew et al. 2011). However, carbamazepine-degrading bacterium have been effectively isolated by Li et al. (2013), which were able to biodegrade carbamazepine to 46.6% at low temperature (10 °C). It has been demonstrated that for carbamazepine, a higher sludge age is not related to a higher removal (Kreuzinger et al. 2004; Clara et al. 2005), which means that the SRT change from 10 to 20 d was not a crucial factor here. The reason for the higher removal of the IFAS-MBR may be then attributed to the introduction of a new type of biomass (namely the attached biomass of the carriers) which means a broader variety of bacteria present in the system in the case of the IFAS-MBR. Moreover, the existence in the biofilm of different profiles (aerobic, anoxic and anaerobic) might have led to higher elimination of recalcitrant compounds as they have a wider choice of conditions to achieve degradation and/or sorption. This would be in agreement with Hai et al. (2011), who found exceptionally high removal (68%) for carbamazepine working in anoxic conditions. For hormones, an increase in their removal by an IFAS process was reported in the literature by Koh et al. (2008), who associated it with the high SRT of the biofilm, which allows the biomass to completely adapt to the influent. This was in agreement with this study where, although only estriol was detected in the IFAS-MBR 20 d phase, its removal was 100%, higher than for the MBR and the CAS. The rest of studied hormones (17α-ethinylestradiol and 17β-estradiol) were detected during the IFAS-MBR 10 d phase and they were also completely removed by the system, which was not the case for the MBR and the CAS. An increase in the removal rate of hormones by the IFAS-MBR is thus demonstrated here and it seems to be independent of operating conditions (SRT and MLSS).

Removal rates achieved by the CAS were generally lower than those of the MBR and the IFAS-MBR, which can be again attributed to a lower sludge age, which was approximately 10 d for the CAS, as well as to the absence of biofilm. The process showing the worst performance was the MBMBR. This had been already expected, as this system operated at the lowest HRT and with the lowest kilograms of biomass, as it practically only counted on the biomass attached to the carriers. Nevertheless, this system performed quite well considering its simplicity (only one aerobic chamber and the membrane tank, no recirculation). When considering the RO, removal rates increased, obtaining a mean removal of 88.8% for all compounds and 97.8% for the PhACs.

In Table 3, the studied micropollutants have been grouped in order to evaluate the general performance of the processes. Thus, the IFAS-MBR20 was the best process for hormones (100% removal) and for PhACs (here considering all listed PhACs from Table 2 except the hormones), followed by the MBR, both showing removal rates around 82% for PhACs. This rate improves for the IFAS-MBR20 if we do not take into account sulfamethoxazole; this compound showed 0% removal but its representativeness is restricted to one sample, as its concentration in the influent was generally below the LOD. Excluding this compound, mean PhAC removal for IFAS-MBR20 was 91%. Removal rates achieved by the RO were high for PhACs, hormones and LAS, with removal rates greater than 95%. Although for NP and DEHP removal rates were only medium high (70–80%), if we consider that the removal in the previous step was relatively low (1.2–39.5%), the removal rates achieved by RO for these compounds become relevant.

Table 3

Comparison of removal rates per groups of substances

Compound IFAS-MBR10 IFAS-MBR20 MBMBR MBR CAS RO 
PhACs 72.3 82.8 64.2 82.4 69.9 97.8 
Hormones 100 100 66.7 75.4 93.3 99.0 
LASs 69.3 91.1 78.2 86.4 85.1 95.5 
Nonylphenols 39.5 19.4 13.0 29.8 36.3 78.4 
DEHP 30.0 40.2 1.2 27.8 26.5 71.5 
PAHs 64.9 97.4 82.2 81.0 70.1 90.8 
Global 62.7 71.8 50.9 63.8 63.5 88.8 
Compound IFAS-MBR10 IFAS-MBR20 MBMBR MBR CAS RO 
PhACs 72.3 82.8 64.2 82.4 69.9 97.8 
Hormones 100 100 66.7 75.4 93.3 99.0 
LASs 69.3 91.1 78.2 86.4 85.1 95.5 
Nonylphenols 39.5 19.4 13.0 29.8 36.3 78.4 
DEHP 30.0 40.2 1.2 27.8 26.5 71.5 
PAHs 64.9 97.4 82.2 81.0 70.1 90.8 
Global 62.7 71.8 50.9 63.8 63.5 88.8 

CONCLUSIONS

The trace organic pollutant removal rates of different biological treatment processes were compared using a pilot plant and a full-scale plant treating the same real municipal wastewater. The processes were the MBR, IFAS-MBR and MBMBR at pilot scale and the CAS at full scale. The IFAS-MBR showed the best performance, with mean removal rates of 72%, followed by the MBR, with 64% mean removal rates. The increase in removal rates by the IFAS-MBR was attributed to the presence of both suspended and attached biomass, which may have increased the degradation possibilities by offering different conditions (aerobic–anoxic–anaerobic) along its profile and also to a high sludge age of the biofilm which allowed conditioning of the biology to the contaminants. The operating conditions for this system proved to be of great importance; the IFAS-MBR operating at low SRT and low MLSS concentrations showed only medium removal rates. The MBMBR showed the worst performance of all membrane processes, which was attributed to the lower content of biomass in the system and the lower operating HRT. All compounds were effectively removed except for the nonylphenols and the DEHP, showing poor removal rates (10–30%). For these compounds, further treatment such as RO, activated carbon or advanced oxidation process is needed in order to achieve efficient elimination. The RO step improved removal rates substantially up to 100%, being very effective for PhACs, hormones and LASs (>95% removal) but lower for nonylphenols and DEHP (70–80% removal). IFAS-MBR is an interesting process for trace organic pollutant removal in wastewater.

ACKNOWLEDGEMENTS

The authors would like to acknowledge Corporación Tecnológica de Andalucía for providing funding and Aguas y Servicios de la Costa Tropical, Silvia Ruiz, Maria de Mar González, Jorge Ignacio Pérez and Jordi Bacardit for their help.

REFERENCES

REFERENCES
APHA, AWWA & WEF
1998
Standard Methods for the Examination of Water and Wastewater
.
18th
edn,
American Public Health Association
,
Washington, DC
.
Bouju
H.
Buttiglieri
G.
Malpei
F.
2009
Are MBRs really more efficient in removing pharmaceutical substances? Comparison of a full scale conventional activated sludge process and a MBR pilot plant. 6th IWA MTC
,
Beijing
,
China
.
Butler
C. S.
Boltz
J. P.
2014
Biofilm Processes and Control in Water and Wastewater Treatment
.
Earth Systems and Environmental Sciences
3
,
90
107
.
Camacho-Muñoz
D.
Martín
J.
Santos
J. L.
Alonso
E.
Aparicio
I.
de la Torre
T.
Rodríguez
C.
Malfeito
J. J.
2012
Effectiveness of three configurations of membrane bioreactors on the removal of priority and emergent organic compounds from wastewater: comparison with conventional wastewater treatments
.
Journal of Environmental Monitoring
14
(
5
),
1428
1436
.
Clara
M.
Strenn
B.
Ausserleitner
M.
Kreuzinger
N.
2004
Comparison of the behaviour of selected micropollutants in a membrane bioreactor and a conventional wastewater treatment plant
.
Water Science and Technology
50
(
5
),
29
36
.
De Wever
H.
Weiss
S.
Reemtsma
T.
Vereecken
J.
Müller
J.
Knepper
T.
Rördend
O.
Gonzalez
S.
Barceló
D.
Hernando
M. D.
2007
Comparison of sulfonated and other micropollutants removal in membrane bioreactor and conventional wastewater treatment
.
Water Research
41
,
935
945
.
Hai
F. I.
Li
X.
Price
W.
Nghiem
L. D.
2011
Removal of carbamazepine and sulfamethoxazole by MBR under anoxic and aerobic conditions
.
Bioresource Technology
102
(
22
),
10386
10390
.
Koh
Y. K.
Chiu
T. Y.
Boobis
A.
Cartmell
E.
Scrimshaw
M. D.
Lester
J. N
.
2008
Treatment and removal strategies for estrogens from wastewater
.
Environmental Technology
29
(
3
),
245
267
.
Kriklavova
L.
Lederer
T.
Jirků
V.
2012
The use of composite fibers for production of biomass carriers
. In:
Microbes in Applied Research
(
Mendez-Vilas
A.
, ed.),
World Scientific Publishing
,
Singapore
, pp.
187
191
.
Kreuzinger
N.
Clara
M.
Strenn
B.
Kroiss
H.
2004
Relevance of the sludge retention time (SRT) as design criteria for wastewater treatment plants for the removal of endocrine disruptors and pharmaceuticals from wastewater
.
Water Science and Technology
50
,
149
156
.
Leiknes
T.
Ødegaard
H.
2007
The development of a biofilm membrane bioreactor
.
Desalination
202
,
135
143
.
Li
A.
Cai
R.
Cui
D.
Qiu
T.
Panga
C.
Yang
J.
Maa
F.
Ren
N.
2013
Characterization and biodegradation kinetics of a new cold-adapted carbamazepine-degrading bacterium, Pseudomonas sp. CBZ-4
.
Journal of Environmental Sciences
25
(
11
),
2281
2290
.
Purdom
C. E.
Hardiman
P. A.
Bye
V. V. J.
Eno
N. C.
Tyler
C. R.
Sumpter
J. P.
1994
Estrogenic effects of effluents from sewage treatment works
.
Chemistry and Ecology
,
8
(
4
),
275
285
.
Sahar
E.
Ernst
M.
Godehardt
M.
Hein
A.
Herr
J.
Kazner
C.
Melin
T.
Cikurel
H.
Aharoni
A.
Messalem
R.
Brenner
A.
Jekel
M.
2011
Comparison of two treatments for the removal of selected organic micropollutants and bulk organic matter: conventional activated sludge followed by ultrafiltration versus membrane bioreactor
.
Water Science and Technology
63
(
4
),
733
740
.
Tadkaew
N.
Hai
F. I.
McDonald
J. A.
Khan
S. J.
Nghiem
L. D.
2011
Removal of trace organics by MBR treatment: The role of molecular properties
.
Water Research
45
,
2439
2451
.