w a t e r r e s e a r c h 7 1 ( 2 0 1 5 ) 9 7 e1 0 6
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.elsevier.com/locate/watres
Membrane distillation combined with an anaerobic moving bed biofilm reactor for treating municipal wastewater Hyun-Chul Kim a, Jaewon Shin b,c, Seyeon Won d, Jung-Yeol Lee c, Sung Kyu Maeng e, Kyung Guen Song b,* a
Water Resources Research Institute, Sejong University, 98 Gunja-dong, Gwangjin-gu, Seoul 143-747, Republic of Korea b Center for Water Resource Cycle Research, Korea Institute of Science and Technology, Hwarangno 14-gil 5, Seongbuk-gu, Seoul 136-791, Republic of Korea c School of Civil, Environmental & Architectural Engineering, Korea University, 145 Anam-ro, Seongbuk-gu, Seoul 136-701, Republic of Korea d Han-River Environment Research Center, 627 Yangsu-ri, Yangseo-myeon, Yangpyeong-kun, Kyounggi-do 476-823, Republic of Korea e Department of Civil and Environmental Engineering, Sejong University, 98 Gunja-dong, Gwangjin-gu, Seoul 143747, Republic of Korea
article info
abstract
Article history:
A fermentative strategy with an anaerobic moving bed biofilm reactor (AMBBR) was used
Received 3 July 2014
for the treatment of domestic wastewater. The feasibility of using a membrane separation
Received in revised form
technique for post-treatment of anaerobic bio-effluent was evaluated with emphasis on
17 November 2014
employing a membrane distillation (MD). Three different hydrophobic 0.2 mm membranes
Accepted 29 December 2014
made of polytetrafluoroethylene (PTFE), polyvinylidene fluoride (PVDF), and polypropylene
Available online 7 January 2015
(PP) were examined in this study. The initial permeate flux of the membranes ranged from 2.5 to 6.3 L m2 h1 when treating AMBBR effluent at a temperature difference between the
Keywords:
feed and permeate streams of 20 C, with the permeate flux increasing in the order
Membrane distillation
PP < PVDF < PTFE. The permeate flux of the PTFE membrane gradually decreased to 84% of
Anaerobic moving bed biofilm
the initial flux after the 45 h run for distillation, while a flux decline in MD with either the
reactor
PVDF or PP membrane was not found under the identical distillation conditions. During
Phosphorus removal
long-term distillation with the PVDF membrane, total phosphorus was completely rejected
Effluent organic matter
and >98% rejection of dissolved organic carbon was also achieved. The characterization of
Organic characterization
wastewater effluent organic matter (EfOM) using an innovative suite of analytical tools verified that almost all of the EfOM was rejected via the PVDF MD treatment. © 2015 Elsevier Ltd. All rights reserved.
* Corresponding author. Tel.: þ82 2 958 5842; fax: þ82 2 958 6854. E-mail address: [email protected] (K.G. Song). http://dx.doi.org/10.1016/j.watres.2014.12.048 0043-1354/© 2015 Elsevier Ltd. All rights reserved.
98
1.
w a t e r r e s e a r c h 7 1 ( 2 0 1 5 ) 9 7 e1 0 6
Introduction
Membrane distillation (MD) is a thermally-driven separation process in which two aqueous solutions at different temperatures are separated by a porous hydrophobic membrane and thus the water vapor flux is formed across the membrane from the feed liquid (hot saline water) to the distillate due to the evolved partial vapor pressure (Curcio and Drioli, 2005; Lawson and Lloyd, 1997; Song et al., 2007). Macro-porous polymeric or inorganic membrane disposed between the feed and permeate streams in the MD process acts as a physical barrier, providing interfaces on which heat and mass are simultaneously exchanged (Cerneaux et al., 2009; Gryta and Tomaszewska, 1998; Hendren et al., 2009; Khemakhem and Amar, 2011; Lawson and Lloyd, 1997; Schofield et al., 1987). The most common materials employed for MD membranes are polytetrafluoroethylene (PTFE), polyvinylidene fluoride (PVDF), and polypropylene (PP) (Jiao et al., 2004; Song et al., 2007; Zhang et al., 2010a). Among these materials, PTFE has the highest hydrophobicity, good chemical and thermal stability, and oxidation resistance (Alklaibi and Lior, 2005; Mulder, 1996). PVDF and PP also exhibit good hydrophobicity and thermal/chemical resistance and can be easily developed into membranes with versatile pore structures (Curcio and Drioli, 2005). Recent explorations have been extended to include new membrane materials (such as carbon nanotubes, fluorinated copolymer materials, and surface
modified polyethersulfone (PES)) that achieve good mechanical strength and high porosity (Dumee et al., 2010; Suk et al., 2006, 2010; Zhang et al., 2010b). Such hydro-repellent membranes enable the complete rejection of non-volatile solutes (e.g., macromolecules, colloidal fraction, ionic species and so forth). Typical feed temperature ranges from 30 C to 60 C (Bandini and Sarti, 2002; Hsu et al., 2002; Khayet et al., 2000; Martinez and Florido-Diaz, 2001) and lower temperatures than those usually applied to conventional distillation are desirable since the heat loss through thermal conduction is also linear to the temperature difference across the membrane. MD has primarily been applied for desalination of sea water and is increasingly studied for advanced treatment of wastewater in the context of water reuse (Gryta et al., 2006). Some investigators have also explored the feasibility of using MD for brackish water desalination, process water treatment, and resource concentration for industrial uses (Alkhudhiri et al., 2012). The MD process can readily utilize the advantages of anaerobic processes, while the mesophilic (or thermophilic) operating conditions that are commonly needed to effectively run fermentation processes can result in no or less heating requirement for the subsequent MD treatment. While comparing with aerobic treatment strategies, anaerobic organic stabilization has many advantages, such as no aeration requirement, lower sludge production, and energy-rich biogas recovery. Therefore, the main objective of this research was to evaluate the applicability of the
Fig. 1 e Experimental set-up for fermentative wastewater treatment with AMBBR and subsequent distillation using hydrophobic microfilters. The biogas collected was analyzed in a gas chromatograph.
99
w a t e r r e s e a r c h 7 1 ( 2 0 1 5 ) 9 7 e1 0 6
MD process for its combination with an anaerobic biological treatment that is capable of not only producing energy-rich biogas, but also reducing energy consumption which is mainly related to aeration for sustaining aerobes in wastewater treatments. In this research, an anaerobic moving bed biofilm reactor (AMBBR), in which the high concentrations of biomass could be retained, was employed for the anaerobic sanitary treatment of domestic wastewater. The effectiveness and applicability of the MD process for the reclamation of the sanitary effluent were determined by monitoring the water quality and the permeate flux. We also employed wellestablished characterization techniques to better trace the fate of wastewater effluent organic matter (EfOM) components through fermentative treatment and the subsequent MD process.
2.
Materials and methods
2.1.
Fermentative wastewater treatment
The synthetic wastewater was prepared on a daily basis using a recipe for artificial domestic wastewater reported in an earlier study (Chu et al., 2005; Hu and Stuckey, 2006; Huang et al., 2008). The anaerobic wastewater treatment employed a down-flow hanging sponge reactor in which the amount of floatable polyurethane cubes (5 mm) were packed up to 50% (v/v). The anaerobic moving bed biofilm reactor (AMBBR) with a working volume of 5 L was located and operated at a flow rate of 40 or 60 L d1 in a walk-in-chamber controlled at 30 C. The effluent from the columnar AMBBR (10 cm 85 cm) was recirculated at 0.7 L min1 using a peristaltic pump. The theoretical hydraulic retention time (HRT) of AMBBR was in the range of 2e3 h and the organic loading rate (as COD) ranged from 0.9 to 1.3 g L1 d1. The initial content of biosolids was 5500 mg L1 and the biogas produced from AMBBR was collected in DuPont Tedlar bags.
2.2.
Membrane distillation of AMBBR effluent
Direct contact membrane distillation (DCMD) was conducted using a closed-loop test unit (Fig. 1) comprising a membrane
module, a heater for a feed stream, a chiller for a permeate stream, and gear circulating pumps. Three different hydrophobic 0.2 mm membranes made of PTFE, PVDF, and PP were examined in this study and their specifications are described in Table 1. The contact angle of water on the surface of the flat sheet membranes (PTFE and PVDF) was measured using a contact angle measurement instrument (DSA 100, Kruss, Germany) which was not available for determining the liquid-membrane contact angle for the capillary PP membrane. The morphology of the membrane surface was observed using a JSM-6390 Scanning Electron Microscope (Jeol, USA). Feed water was circulated at a flow rate of 0.4 L min1 through the retentate side of the module using a gear pump on the recirculation line. Likewise, the distillate was circulated at 0.3 L min1 from a double-layer acrylate reservoir through the membrane module and back to the reservoir under constant temperature (20 C). Permeate flux was controlled by adjusting the transmembrane temperature between the feed and permeate streams. Tertiary-treated carbon-free water (18 MU cm1) was used as the initial condensing fluid. Deionized water was used as the feed water for pure water flux tests, and thereafter AMBBR effluent was distillated with three different membranes to select the most promising membrane among them. The temperature of the feed stream was stepped to a higher temperature up to 60 C and the permeate flux (as mass) was continuously recorded using an electronic balance (GF-6100, A&D Company Ltd.) for 3 h at each feed temperature after the temperature was stabilized. The best membrane was selected for further long-term MD experiments. MD application was examined by introducing secondary effluent from AMBBR into the module where a flatsheet PVDF membrane with an effective distillation area of 32 cm2 was located. PVDF has been comprehensively used as a porous membrane material in many public and industrial fields due to its excellent resistance to harsh environments containing acids, alkaline, oxidants, and halogens (Curcio and Drioli, 2005). The laboratory scale MD system was operated at an initial constant flux of 3.4 L m2 h1 (LMH) without any interruptions of the filtration for backwash or flushing. A peristaltic pump was used to continuously transfer AMBBR effluent to the feed reservoir which had a working volume of
Table 1 e Specification of hydrophobic microfilters used in the membrane distillation. Parameter Supplier Material (membrane/supporter) Type Nominal pore size (mm) Porosity (%) Membrane thickness (mm)b Effective permeate area (cm2) Contact angle ( ) Liquid entry pressure of water (kPa)b a
PTFE
PVDF
PPa
Millipore Polytetrafluoroethylene/polyethylene Flat sheet 0.2 70 130 32 142 ± 10 280
Millipore Polyvinylidene fluoride/none Flat sheet 0.22 75 110 32 94 ± 2 204
Membrana GmbH Polypropylene/none Capillary 0.2 70 450 23 134c 140
The feed water was delivered inside the capillary membrane. Khayet and Matsuura, 2011. c Bojarska et al. (2014) measured the contact angle of the capillary PP membrane using the dynamic Wilhelmy method which was different from our method used for the flat sheet membranes. b
100
w a t e r r e s e a r c h 7 1 ( 2 0 1 5 ) 9 7 e1 0 6
6 L during the distillation. Excess distillate was periodically decanted from the permeate reservoir and collected for analysis. The amount of permeate volume at the end of the test accounted for 0.5% of the feed. Intermittent filtration runs with hydrophilic 0.2 mm polyethersulfone (PES) membrane were also conducted for comparative examination simultaneously with the MD process for post-treatment of anaerobically treated wastewater.
interval and integration time were maintained at 1 nm and 0.1 s, respectively. Right-angle geometry was used for liquid samples in a 10 mm fused-quartz cuvette. Three-dimensional spectra were obtained by repeatedly measuring the emission spectra within the range of 200e600 nm, with excitation wavelengths from 200 to 400 nm spaced at 5 nm intervals in the excitation domain. Spectra were then concatenated into an excitation-emission matrix (EEM).
2.3.
2.4.
Characterization of dissolved organic matter
The molecular weight (MW) distribution of dissolved organic matter was determined using liquid chromatography-organic carbon detection (LC-OCD) (DOC-Labor, Germany), for which size-exclusion chromatography was coupled with on-line equipment for the real-time measurement of both UV absorbance at 254 nm (UV254) and organic carbon concentration of effluent passing through the exclusion column. More detailed information of the LC-OCD system can be found in an earlier study (Huber et al., 2011). Fluorescence spectra were collected using a PerkineElmer LS-50B luminescence spectrometer which uses a 450 W xenon lamp source. All liquid samples were diluted with carbon-free electrolyte solution under the ambient pH conditions. The spectroscopic analysis was carried out at a concentration of 1 mg C L1 to minimize the inner-filter effect. The acquisition
Fig. 2 e COD removal (a) and methane yield (b) during fermentative wastewater treatment using AMBBR.
Analytical methods
The Salicylate, Acid Persulfate Digestion, and Dichromate Methods were employed to measure the NH4eN, total phosphorus (TP), and chemical oxygen demand (COD) in the water samples, equivalent to Standard Method 4500-NH3 G, 4500 P. B. 5, and 5220 C for water and wastewater, respectively (APHA, 1998). The total organic carbon (TOC) and the total nitrogen (TN) were determined using a Shimadzu TOC-V/CPN analyzer
Fig. 3 e Pure water flux of three different membranes as a function of transmembrane temperature (a) and the performance of these membranes for the distillation of AMBBR effluent (b). During the pure water flux tests the feed temperature was stepped to a higher temperature to control the transmembrane temperature ranging from 10 to 40 C, while the distillation of AMBBR effluent was conducted at a constant transmembrane temperature of 20 C. The feed flow rate was 0.4 L min¡1, and the permeate temperature was consistently maintained at 20 C during any experiments.
w a t e r r e s e a r c h 7 1 ( 2 0 1 5 ) 9 7 e1 0 6
101
with a TN module (TNM-1). Non-purgeable organic carbon is generally referred to as TOC for whole samples (or dissolved organic carbon (DOC) for a sample filtered with 0.45 mm membranes) according to the operational definition. A DR/ 5000 spectrophotometer (Hach, USA) was used to measure the UV254 after filtration using 0.45 mm membrane filters. The specific UV absorbance (SUVA) value is calculated from the UV254 divided by the DOC of the water sample, where the UV254 is mainly due to electron-rich sites such as aromatic functional groups and double-bonded carbon groups in organic molecules (Kim and Lee, 2006). The amount of dissolved oxygen (DO) was determined using an YSI 58 DO meter. The pH was measured using a pH meter (Orion 3 STAR, USA). Biogas was analyzed to quantify the methane content using an HP 6890 gas chromatograph (HewlettePackard, USA) in which a Carbonex 1004 stainless steel column was installed. Each measurement was carried out in triplicate and average values were reported.
3.
Results and discussion
3.1.
Fermentative wastewater treatment with AMBBR
The AMBBR run achieved from 79% to 92% removal of COD at an HRT of 3 h, which was similar to that registered even after HRT decreased to 2 h (Fig. 2). The fermentative wastewater treatment also produced a small volume of valuable biogas via the bioconversion of carbonaceous organic matter. The content of methane produced ranged from 58% to 72% during the experimental course. The methane yield per removed COD was independent of the HRT employed in this study, and was comparable to that previously reported (Hu and Stuckey, 2006; Kim et al., 2011). In the context of bioenergy recovery, an industrial wastewater with high organic strength could be partially combined with domestic sewage to be treated with anaerobic biological stabilization processes producing biogas (e.g., CH4), which is easily convertible to heat energy (especially for maintaining the moderate temperature of AMBBR effluent at the feed side of MD process); thus, this strategy is encouraged for hybridization with the MD treatment.
3.2. Distillation of AMBBR effluent using hydrophobic microfilters Three different membrane materials were examined for the pure water flux tests (Table 1). The transmembrane temperature ranged from 10 C to 40 C, which was adjusted by increasing the temperature of the feed stream. The permeate flux for all membranes was increased with the increasing temperature difference across the membrane (Fig. 3a), driven by the exponential increase of vapor pressure with the temperature. The permeate flux for the PVDF membrane ranged between 1.8 and 14.4 LMH under the given temperature, which was comparable to that achieved by the PTFE membrane. The PP membrane showed the lowest permeability among the examined membranes, regardless of the temperature applied, which was attributed to a higher membrane thickness of PP when compared to PTFE or PVDF. An insignificant difference
Fig. 4 e Permeate flux profile (a) and the variations in the feed and permeate concentrations (bed) during long-term distillation of AMBBR effluent using 0.22 mm PVDF membrane. Insets show the surface images of the PVDF membrane used in the experiment: fresh PVDF membrane before the distillation (left) and the membrane fouled after the 150 h distillation (right). The temperature of feed stream was maintained at 40 C, while the permeate temperature was controlled to be 20 C using a chiller during the distillation.
102
w a t e r r e s e a r c h 7 1 ( 2 0 1 5 ) 9 7 e1 0 6
was found in the porosity of the three candidates. The porosity increases both the thermal resistance and the permeability of the membrane. Membrane permeability has an increasing tendency with decreasing membrane thickness (Curcio and Drioli, 2005). MD operates successfully if membrane wetting does not occur (Chiam and Sarbatly, 2013). The wetting of the membrane pores leads to reduced product quality, and thus it is advantageous to use membranes with cz liquid entry pressure (LEP) value as high as possible (Ra et al., 2014). The LEP of the membrane decreases with the pore size but increases with the contact angle of water on the surface of the membrane. The relationship between the LEP and these two parameters can be expressed by the Laplace equation (Lawson and Lloyd, 1997). It is possible to relatively evaluate the membrane wetting for different membrane materials by comparing their contact angles only when the identical analytical method was used to determine the liquidmembrane contact angle. Bojarska et al. (2014) measured the contact angle of the capillary PP membrane using the dynamic Wilhelmy method. They reported that the capillary PP membrane had a contact angle of 134 , which was greater than that observed for the PVDF membrane examined in this study. However, it has been reported that the contact angle values of water on some materials such as PTFE, PVDF, and Teflon are generally higher than that observed for PP (Khayet and Matsuura, 2011; Sigurdsson and Shishoo, 1997). When treating AMBBR effluent with the three candidates for 45 h, the permeate flux was always higher for the PTFE membrane than that observed for the other two membranes (Fig. 3b). The initial permeate flux for the PTFE membrane was 6.3 LMH at a transmembrane temperature of 20 C, but gradually decreased to 5.3 LMH as a function of distillation time. The initial flux for the PVDF and PP membranes accounted for 81% and 41% of that registered for PTFE. However, the permeability was consistent for both PVDF and PP during the experimental course. Based on the results from the permeability measurement using the three different membranes, the PVDF membrane was selected for a further longer term distillation test because of the stable permeability for a given period of distillation time.
The PVDF MD treatment was conducted using AMBBR effluent for 150 h. Fig. 4a shows the history of the permeate flux measured during the distillation with the PVDF membrane at the transmembrane temperature of 20 C. The variations of the feed and permeate concentrations as a function of time were also illustrated in Fig. 4 with respect to conductivity, soluble COD (SCOD), and TP. A noticeable decline in the permeate flux occurred after the first 52 h run and it was accelerated as the permeate volume increased, likely due to membrane fouling associated with a gradual increase in the feed concentration. The permeability decreased by 15% at the end of the distillation test, which was strongly associated with an accumulation of non-volatile components in the influent water. The photos inserted in Fig. 4 clearly show the particulates accumulated on the top of the PVDF membrane after the distillation test. Although membrane distillation is more resistant to fouling than conventional thermal processes, the foulants depositing on the membrane surface can significantly reduce the hydrophobicity of the membrane, which causes (or accelerates) membrane wetting resulting in decreased rejection of the non-volatile components.
3.3. Removal and characterization of dissolved organic fractions Separate MD treatment continued for 24 h to trace the fate of EfOM components through fermentative treatment and the subsequent distillation. The anaerobic wastewater treatment using AMBBR achieved >94% removal of carbonaceous components (determined as DOC) and also decreased UV254 from 0.58 to 0.20 cm1, which resulted in increased SUVA (Table 2). SUVA has been used as an indicator of the humic content in water environmental systems (EPA, 1999), and a substantial increase in the SUVA value upon the anaerobic stabilization indicates the accumulation of highly-condensed organic compounds that could not be easily utilized by anaerobes in the fermentative treatment processes. The subsequent MD treatment removed almost all of the organic components that remained after the fermentative treatment, and the organic
Table 2 e Overall performance of membrane distillation following fermentative treatment using AMBBR for domestic wastewater treatment. Parameter pH TCOD (mg L1) DOC (mg L1) UV254 (cm1) SUVA (mg m1 L1) TN (mg L1) NH4eN (mg L1) TP (mg L1)
AMBBR influent 7.4 326 109 0.582 0.53 42 30 5.4
± 0.1 ± 89 ± 27 ± 0.070 ± 0.21 ±5 ±9 ± 0.3
AMBBR effluent
MD permeatea
7.6 ± 0.1 55 ± 1 5.7 ± 0.2 0.202 ± 0.007 3.54 ± 0.02 45 ± 2 47 ± 2 6.0 ± 0.5c
8.5 ± 0.1 (7.3 ± 0.1) 1.3 ± 0.1 (15 ± 1) 0.03 ± 0.01 (4.3 ± 0.1)
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.elsevier.com/locate/watres
Membrane distillation combined with an anaerobic moving bed biofilm reactor for treating municipal wastewater Hyun-Chul Kim a, Jaewon Shin b,c, Seyeon Won d, Jung-Yeol Lee c, Sung Kyu Maeng e, Kyung Guen Song b,* a
Water Resources Research Institute, Sejong University, 98 Gunja-dong, Gwangjin-gu, Seoul 143-747, Republic of Korea b Center for Water Resource Cycle Research, Korea Institute of Science and Technology, Hwarangno 14-gil 5, Seongbuk-gu, Seoul 136-791, Republic of Korea c School of Civil, Environmental & Architectural Engineering, Korea University, 145 Anam-ro, Seongbuk-gu, Seoul 136-701, Republic of Korea d Han-River Environment Research Center, 627 Yangsu-ri, Yangseo-myeon, Yangpyeong-kun, Kyounggi-do 476-823, Republic of Korea e Department of Civil and Environmental Engineering, Sejong University, 98 Gunja-dong, Gwangjin-gu, Seoul 143747, Republic of Korea
article info
abstract
Article history:
A fermentative strategy with an anaerobic moving bed biofilm reactor (AMBBR) was used
Received 3 July 2014
for the treatment of domestic wastewater. The feasibility of using a membrane separation
Received in revised form
technique for post-treatment of anaerobic bio-effluent was evaluated with emphasis on
17 November 2014
employing a membrane distillation (MD). Three different hydrophobic 0.2 mm membranes
Accepted 29 December 2014
made of polytetrafluoroethylene (PTFE), polyvinylidene fluoride (PVDF), and polypropylene
Available online 7 January 2015
(PP) were examined in this study. The initial permeate flux of the membranes ranged from 2.5 to 6.3 L m2 h1 when treating AMBBR effluent at a temperature difference between the
Keywords:
feed and permeate streams of 20 C, with the permeate flux increasing in the order
Membrane distillation
PP < PVDF < PTFE. The permeate flux of the PTFE membrane gradually decreased to 84% of
Anaerobic moving bed biofilm
the initial flux after the 45 h run for distillation, while a flux decline in MD with either the
reactor
PVDF or PP membrane was not found under the identical distillation conditions. During
Phosphorus removal
long-term distillation with the PVDF membrane, total phosphorus was completely rejected
Effluent organic matter
and >98% rejection of dissolved organic carbon was also achieved. The characterization of
Organic characterization
wastewater effluent organic matter (EfOM) using an innovative suite of analytical tools verified that almost all of the EfOM was rejected via the PVDF MD treatment. © 2015 Elsevier Ltd. All rights reserved.
* Corresponding author. Tel.: þ82 2 958 5842; fax: þ82 2 958 6854. E-mail address: [email protected] (K.G. Song). http://dx.doi.org/10.1016/j.watres.2014.12.048 0043-1354/© 2015 Elsevier Ltd. All rights reserved.
98
1.
w a t e r r e s e a r c h 7 1 ( 2 0 1 5 ) 9 7 e1 0 6
Introduction
Membrane distillation (MD) is a thermally-driven separation process in which two aqueous solutions at different temperatures are separated by a porous hydrophobic membrane and thus the water vapor flux is formed across the membrane from the feed liquid (hot saline water) to the distillate due to the evolved partial vapor pressure (Curcio and Drioli, 2005; Lawson and Lloyd, 1997; Song et al., 2007). Macro-porous polymeric or inorganic membrane disposed between the feed and permeate streams in the MD process acts as a physical barrier, providing interfaces on which heat and mass are simultaneously exchanged (Cerneaux et al., 2009; Gryta and Tomaszewska, 1998; Hendren et al., 2009; Khemakhem and Amar, 2011; Lawson and Lloyd, 1997; Schofield et al., 1987). The most common materials employed for MD membranes are polytetrafluoroethylene (PTFE), polyvinylidene fluoride (PVDF), and polypropylene (PP) (Jiao et al., 2004; Song et al., 2007; Zhang et al., 2010a). Among these materials, PTFE has the highest hydrophobicity, good chemical and thermal stability, and oxidation resistance (Alklaibi and Lior, 2005; Mulder, 1996). PVDF and PP also exhibit good hydrophobicity and thermal/chemical resistance and can be easily developed into membranes with versatile pore structures (Curcio and Drioli, 2005). Recent explorations have been extended to include new membrane materials (such as carbon nanotubes, fluorinated copolymer materials, and surface
modified polyethersulfone (PES)) that achieve good mechanical strength and high porosity (Dumee et al., 2010; Suk et al., 2006, 2010; Zhang et al., 2010b). Such hydro-repellent membranes enable the complete rejection of non-volatile solutes (e.g., macromolecules, colloidal fraction, ionic species and so forth). Typical feed temperature ranges from 30 C to 60 C (Bandini and Sarti, 2002; Hsu et al., 2002; Khayet et al., 2000; Martinez and Florido-Diaz, 2001) and lower temperatures than those usually applied to conventional distillation are desirable since the heat loss through thermal conduction is also linear to the temperature difference across the membrane. MD has primarily been applied for desalination of sea water and is increasingly studied for advanced treatment of wastewater in the context of water reuse (Gryta et al., 2006). Some investigators have also explored the feasibility of using MD for brackish water desalination, process water treatment, and resource concentration for industrial uses (Alkhudhiri et al., 2012). The MD process can readily utilize the advantages of anaerobic processes, while the mesophilic (or thermophilic) operating conditions that are commonly needed to effectively run fermentation processes can result in no or less heating requirement for the subsequent MD treatment. While comparing with aerobic treatment strategies, anaerobic organic stabilization has many advantages, such as no aeration requirement, lower sludge production, and energy-rich biogas recovery. Therefore, the main objective of this research was to evaluate the applicability of the
Fig. 1 e Experimental set-up for fermentative wastewater treatment with AMBBR and subsequent distillation using hydrophobic microfilters. The biogas collected was analyzed in a gas chromatograph.
99
w a t e r r e s e a r c h 7 1 ( 2 0 1 5 ) 9 7 e1 0 6
MD process for its combination with an anaerobic biological treatment that is capable of not only producing energy-rich biogas, but also reducing energy consumption which is mainly related to aeration for sustaining aerobes in wastewater treatments. In this research, an anaerobic moving bed biofilm reactor (AMBBR), in which the high concentrations of biomass could be retained, was employed for the anaerobic sanitary treatment of domestic wastewater. The effectiveness and applicability of the MD process for the reclamation of the sanitary effluent were determined by monitoring the water quality and the permeate flux. We also employed wellestablished characterization techniques to better trace the fate of wastewater effluent organic matter (EfOM) components through fermentative treatment and the subsequent MD process.
2.
Materials and methods
2.1.
Fermentative wastewater treatment
The synthetic wastewater was prepared on a daily basis using a recipe for artificial domestic wastewater reported in an earlier study (Chu et al., 2005; Hu and Stuckey, 2006; Huang et al., 2008). The anaerobic wastewater treatment employed a down-flow hanging sponge reactor in which the amount of floatable polyurethane cubes (5 mm) were packed up to 50% (v/v). The anaerobic moving bed biofilm reactor (AMBBR) with a working volume of 5 L was located and operated at a flow rate of 40 or 60 L d1 in a walk-in-chamber controlled at 30 C. The effluent from the columnar AMBBR (10 cm 85 cm) was recirculated at 0.7 L min1 using a peristaltic pump. The theoretical hydraulic retention time (HRT) of AMBBR was in the range of 2e3 h and the organic loading rate (as COD) ranged from 0.9 to 1.3 g L1 d1. The initial content of biosolids was 5500 mg L1 and the biogas produced from AMBBR was collected in DuPont Tedlar bags.
2.2.
Membrane distillation of AMBBR effluent
Direct contact membrane distillation (DCMD) was conducted using a closed-loop test unit (Fig. 1) comprising a membrane
module, a heater for a feed stream, a chiller for a permeate stream, and gear circulating pumps. Three different hydrophobic 0.2 mm membranes made of PTFE, PVDF, and PP were examined in this study and their specifications are described in Table 1. The contact angle of water on the surface of the flat sheet membranes (PTFE and PVDF) was measured using a contact angle measurement instrument (DSA 100, Kruss, Germany) which was not available for determining the liquid-membrane contact angle for the capillary PP membrane. The morphology of the membrane surface was observed using a JSM-6390 Scanning Electron Microscope (Jeol, USA). Feed water was circulated at a flow rate of 0.4 L min1 through the retentate side of the module using a gear pump on the recirculation line. Likewise, the distillate was circulated at 0.3 L min1 from a double-layer acrylate reservoir through the membrane module and back to the reservoir under constant temperature (20 C). Permeate flux was controlled by adjusting the transmembrane temperature between the feed and permeate streams. Tertiary-treated carbon-free water (18 MU cm1) was used as the initial condensing fluid. Deionized water was used as the feed water for pure water flux tests, and thereafter AMBBR effluent was distillated with three different membranes to select the most promising membrane among them. The temperature of the feed stream was stepped to a higher temperature up to 60 C and the permeate flux (as mass) was continuously recorded using an electronic balance (GF-6100, A&D Company Ltd.) for 3 h at each feed temperature after the temperature was stabilized. The best membrane was selected for further long-term MD experiments. MD application was examined by introducing secondary effluent from AMBBR into the module where a flatsheet PVDF membrane with an effective distillation area of 32 cm2 was located. PVDF has been comprehensively used as a porous membrane material in many public and industrial fields due to its excellent resistance to harsh environments containing acids, alkaline, oxidants, and halogens (Curcio and Drioli, 2005). The laboratory scale MD system was operated at an initial constant flux of 3.4 L m2 h1 (LMH) without any interruptions of the filtration for backwash or flushing. A peristaltic pump was used to continuously transfer AMBBR effluent to the feed reservoir which had a working volume of
Table 1 e Specification of hydrophobic microfilters used in the membrane distillation. Parameter Supplier Material (membrane/supporter) Type Nominal pore size (mm) Porosity (%) Membrane thickness (mm)b Effective permeate area (cm2) Contact angle ( ) Liquid entry pressure of water (kPa)b a
PTFE
PVDF
PPa
Millipore Polytetrafluoroethylene/polyethylene Flat sheet 0.2 70 130 32 142 ± 10 280
Millipore Polyvinylidene fluoride/none Flat sheet 0.22 75 110 32 94 ± 2 204
Membrana GmbH Polypropylene/none Capillary 0.2 70 450 23 134c 140
The feed water was delivered inside the capillary membrane. Khayet and Matsuura, 2011. c Bojarska et al. (2014) measured the contact angle of the capillary PP membrane using the dynamic Wilhelmy method which was different from our method used for the flat sheet membranes. b
100
w a t e r r e s e a r c h 7 1 ( 2 0 1 5 ) 9 7 e1 0 6
6 L during the distillation. Excess distillate was periodically decanted from the permeate reservoir and collected for analysis. The amount of permeate volume at the end of the test accounted for 0.5% of the feed. Intermittent filtration runs with hydrophilic 0.2 mm polyethersulfone (PES) membrane were also conducted for comparative examination simultaneously with the MD process for post-treatment of anaerobically treated wastewater.
interval and integration time were maintained at 1 nm and 0.1 s, respectively. Right-angle geometry was used for liquid samples in a 10 mm fused-quartz cuvette. Three-dimensional spectra were obtained by repeatedly measuring the emission spectra within the range of 200e600 nm, with excitation wavelengths from 200 to 400 nm spaced at 5 nm intervals in the excitation domain. Spectra were then concatenated into an excitation-emission matrix (EEM).
2.3.
2.4.
Characterization of dissolved organic matter
The molecular weight (MW) distribution of dissolved organic matter was determined using liquid chromatography-organic carbon detection (LC-OCD) (DOC-Labor, Germany), for which size-exclusion chromatography was coupled with on-line equipment for the real-time measurement of both UV absorbance at 254 nm (UV254) and organic carbon concentration of effluent passing through the exclusion column. More detailed information of the LC-OCD system can be found in an earlier study (Huber et al., 2011). Fluorescence spectra were collected using a PerkineElmer LS-50B luminescence spectrometer which uses a 450 W xenon lamp source. All liquid samples were diluted with carbon-free electrolyte solution under the ambient pH conditions. The spectroscopic analysis was carried out at a concentration of 1 mg C L1 to minimize the inner-filter effect. The acquisition
Fig. 2 e COD removal (a) and methane yield (b) during fermentative wastewater treatment using AMBBR.
Analytical methods
The Salicylate, Acid Persulfate Digestion, and Dichromate Methods were employed to measure the NH4eN, total phosphorus (TP), and chemical oxygen demand (COD) in the water samples, equivalent to Standard Method 4500-NH3 G, 4500 P. B. 5, and 5220 C for water and wastewater, respectively (APHA, 1998). The total organic carbon (TOC) and the total nitrogen (TN) were determined using a Shimadzu TOC-V/CPN analyzer
Fig. 3 e Pure water flux of three different membranes as a function of transmembrane temperature (a) and the performance of these membranes for the distillation of AMBBR effluent (b). During the pure water flux tests the feed temperature was stepped to a higher temperature to control the transmembrane temperature ranging from 10 to 40 C, while the distillation of AMBBR effluent was conducted at a constant transmembrane temperature of 20 C. The feed flow rate was 0.4 L min¡1, and the permeate temperature was consistently maintained at 20 C during any experiments.
w a t e r r e s e a r c h 7 1 ( 2 0 1 5 ) 9 7 e1 0 6
101
with a TN module (TNM-1). Non-purgeable organic carbon is generally referred to as TOC for whole samples (or dissolved organic carbon (DOC) for a sample filtered with 0.45 mm membranes) according to the operational definition. A DR/ 5000 spectrophotometer (Hach, USA) was used to measure the UV254 after filtration using 0.45 mm membrane filters. The specific UV absorbance (SUVA) value is calculated from the UV254 divided by the DOC of the water sample, where the UV254 is mainly due to electron-rich sites such as aromatic functional groups and double-bonded carbon groups in organic molecules (Kim and Lee, 2006). The amount of dissolved oxygen (DO) was determined using an YSI 58 DO meter. The pH was measured using a pH meter (Orion 3 STAR, USA). Biogas was analyzed to quantify the methane content using an HP 6890 gas chromatograph (HewlettePackard, USA) in which a Carbonex 1004 stainless steel column was installed. Each measurement was carried out in triplicate and average values were reported.
3.
Results and discussion
3.1.
Fermentative wastewater treatment with AMBBR
The AMBBR run achieved from 79% to 92% removal of COD at an HRT of 3 h, which was similar to that registered even after HRT decreased to 2 h (Fig. 2). The fermentative wastewater treatment also produced a small volume of valuable biogas via the bioconversion of carbonaceous organic matter. The content of methane produced ranged from 58% to 72% during the experimental course. The methane yield per removed COD was independent of the HRT employed in this study, and was comparable to that previously reported (Hu and Stuckey, 2006; Kim et al., 2011). In the context of bioenergy recovery, an industrial wastewater with high organic strength could be partially combined with domestic sewage to be treated with anaerobic biological stabilization processes producing biogas (e.g., CH4), which is easily convertible to heat energy (especially for maintaining the moderate temperature of AMBBR effluent at the feed side of MD process); thus, this strategy is encouraged for hybridization with the MD treatment.
3.2. Distillation of AMBBR effluent using hydrophobic microfilters Three different membrane materials were examined for the pure water flux tests (Table 1). The transmembrane temperature ranged from 10 C to 40 C, which was adjusted by increasing the temperature of the feed stream. The permeate flux for all membranes was increased with the increasing temperature difference across the membrane (Fig. 3a), driven by the exponential increase of vapor pressure with the temperature. The permeate flux for the PVDF membrane ranged between 1.8 and 14.4 LMH under the given temperature, which was comparable to that achieved by the PTFE membrane. The PP membrane showed the lowest permeability among the examined membranes, regardless of the temperature applied, which was attributed to a higher membrane thickness of PP when compared to PTFE or PVDF. An insignificant difference
Fig. 4 e Permeate flux profile (a) and the variations in the feed and permeate concentrations (bed) during long-term distillation of AMBBR effluent using 0.22 mm PVDF membrane. Insets show the surface images of the PVDF membrane used in the experiment: fresh PVDF membrane before the distillation (left) and the membrane fouled after the 150 h distillation (right). The temperature of feed stream was maintained at 40 C, while the permeate temperature was controlled to be 20 C using a chiller during the distillation.
102
w a t e r r e s e a r c h 7 1 ( 2 0 1 5 ) 9 7 e1 0 6
was found in the porosity of the three candidates. The porosity increases both the thermal resistance and the permeability of the membrane. Membrane permeability has an increasing tendency with decreasing membrane thickness (Curcio and Drioli, 2005). MD operates successfully if membrane wetting does not occur (Chiam and Sarbatly, 2013). The wetting of the membrane pores leads to reduced product quality, and thus it is advantageous to use membranes with cz liquid entry pressure (LEP) value as high as possible (Ra et al., 2014). The LEP of the membrane decreases with the pore size but increases with the contact angle of water on the surface of the membrane. The relationship between the LEP and these two parameters can be expressed by the Laplace equation (Lawson and Lloyd, 1997). It is possible to relatively evaluate the membrane wetting for different membrane materials by comparing their contact angles only when the identical analytical method was used to determine the liquidmembrane contact angle. Bojarska et al. (2014) measured the contact angle of the capillary PP membrane using the dynamic Wilhelmy method. They reported that the capillary PP membrane had a contact angle of 134 , which was greater than that observed for the PVDF membrane examined in this study. However, it has been reported that the contact angle values of water on some materials such as PTFE, PVDF, and Teflon are generally higher than that observed for PP (Khayet and Matsuura, 2011; Sigurdsson and Shishoo, 1997). When treating AMBBR effluent with the three candidates for 45 h, the permeate flux was always higher for the PTFE membrane than that observed for the other two membranes (Fig. 3b). The initial permeate flux for the PTFE membrane was 6.3 LMH at a transmembrane temperature of 20 C, but gradually decreased to 5.3 LMH as a function of distillation time. The initial flux for the PVDF and PP membranes accounted for 81% and 41% of that registered for PTFE. However, the permeability was consistent for both PVDF and PP during the experimental course. Based on the results from the permeability measurement using the three different membranes, the PVDF membrane was selected for a further longer term distillation test because of the stable permeability for a given period of distillation time.
The PVDF MD treatment was conducted using AMBBR effluent for 150 h. Fig. 4a shows the history of the permeate flux measured during the distillation with the PVDF membrane at the transmembrane temperature of 20 C. The variations of the feed and permeate concentrations as a function of time were also illustrated in Fig. 4 with respect to conductivity, soluble COD (SCOD), and TP. A noticeable decline in the permeate flux occurred after the first 52 h run and it was accelerated as the permeate volume increased, likely due to membrane fouling associated with a gradual increase in the feed concentration. The permeability decreased by 15% at the end of the distillation test, which was strongly associated with an accumulation of non-volatile components in the influent water. The photos inserted in Fig. 4 clearly show the particulates accumulated on the top of the PVDF membrane after the distillation test. Although membrane distillation is more resistant to fouling than conventional thermal processes, the foulants depositing on the membrane surface can significantly reduce the hydrophobicity of the membrane, which causes (or accelerates) membrane wetting resulting in decreased rejection of the non-volatile components.
3.3. Removal and characterization of dissolved organic fractions Separate MD treatment continued for 24 h to trace the fate of EfOM components through fermentative treatment and the subsequent distillation. The anaerobic wastewater treatment using AMBBR achieved >94% removal of carbonaceous components (determined as DOC) and also decreased UV254 from 0.58 to 0.20 cm1, which resulted in increased SUVA (Table 2). SUVA has been used as an indicator of the humic content in water environmental systems (EPA, 1999), and a substantial increase in the SUVA value upon the anaerobic stabilization indicates the accumulation of highly-condensed organic compounds that could not be easily utilized by anaerobes in the fermentative treatment processes. The subsequent MD treatment removed almost all of the organic components that remained after the fermentative treatment, and the organic
Table 2 e Overall performance of membrane distillation following fermentative treatment using AMBBR for domestic wastewater treatment. Parameter pH TCOD (mg L1) DOC (mg L1) UV254 (cm1) SUVA (mg m1 L1) TN (mg L1) NH4eN (mg L1) TP (mg L1)
AMBBR influent 7.4 326 109 0.582 0.53 42 30 5.4
± 0.1 ± 89 ± 27 ± 0.070 ± 0.21 ±5 ±9 ± 0.3
AMBBR effluent
MD permeatea
7.6 ± 0.1 55 ± 1 5.7 ± 0.2 0.202 ± 0.007 3.54 ± 0.02 45 ± 2 47 ± 2 6.0 ± 0.5c
8.5 ± 0.1 (7.3 ± 0.1) 1.3 ± 0.1 (15 ± 1) 0.03 ± 0.01 (4.3 ± 0.1)
