- Email: [email protected]
Performance of a combined system of microbial fuel cell and membrane bioreactor: Wastewater treatment, sludge reduction, energy recovery and membrane fouling
Biosensors and Bioelectronics 49 (2013) 92–98
Contents lists available at SciVerse ScienceDirect
Biosensors and Bioelectronics journal homepage: www.elsevier.com/locate/bios
Performance of a combined system of microbial fuel cell and membrane bioreactor: Wastewater treatment, sludge reduction, energy recovery and membrane fouling Xinying Su b, Yu Tian a,b,n, Zhicai Sun c, Yaobin Lu b, Zhipeng Li b a
State Key Laboratory of Urban Water Resource and Environment, Harbin Institute of Technology (SKLUWRE, HIT), Harbin 150090, China School of Municipal and Environmental Engineering, Harbin Institute of Technology, Harbin 150090, China c Daqing Water Group Company Limited, Daqing 163311, China b
art ic l e i nf o
a b s t r a c t
Article history: Received 25 February 2013 Received in revised form 2 April 2013 Accepted 4 April 2013 Available online 18 April 2013
A novel combined system of sludge microbial fuel cell (S-MFC) stack and membrane bioreactor (MBR) was proposed in this study. The non-consumed sludge in the MBR sludge-fed S-MFC was recycled to the MBR. In the combined system, the COD and ammonia treatment efficiencies were more than 90% and the sludge reduction was 5.1% higher than that of the conventional MBR. It's worth noting that the energy recovery and fouling mitigation were observed in the combined system. In the single S-MFC, about 75 mg L−1 COD could be translated to electricity during one cycle. The average voltage and maximum power production of the single S-MFC were 430 mV and 51 mW m−2, respectively. Additionally, the combined system was able to mitigate membrane fouling by the sludge modification. Except for the content decrease (22%), S-MFC destroyed simple aromatic proteins and tryptophan protein-like substances in loosely bound extracellular polymeric substances (LB-EPS). These results indicated that effective wastewater treatment, sludge reduction, energy recovery and membrane fouling mitigation could be obtained in the combined system. & 2013 Elsevier B.V. All rights reserved.
Keywords: Membrane bioreactor (MBR) Microbial fuel cell (MFC) Membrane fouling Loosely bound EPS (LB-EPS) Energy recovery
1. Introduction As efficient technology for wastewater treatment, membrane bioreactors (MBRs) have experienced unprecedented growth in recent years (Fan and Zhou, 2007). MBRs offer several advantages over conventional activated sludge system, including stable and high effluent quality, ease of operation, small footprint, and absolute removal of bacteria (Wang et al., 2009b). However, membrane fouling remains as a major obstacle for wider application of MBRs (Kimura et al., 2012). The most common approaches to reduce membrane fouling are supplying an excessive amount of air (high energy con-sumption), cleaning the membrane module (reduction the membrane life-span), optimizing the operating parameters and improving activated sludge characteristics in the reactor (Akamatsu et al., 2010; Bani-Melhem and Elektorowicz, 2010). Microbial fuel cells (MFCs) have been developed as a promising technology to recover energy from wastewater, marine sediments, sludge etc. (Min et al., 2005). Several systems combining MBR and MFC have been reported previously. A bioelectrochemical membrane reactor, which takes advantage of both MBR and MFC processes, was recently developed (Wang et al., 2011). In this reactor, stainless steel n Corresponding author at: State Key Laboratory of Urban Water Resource and Environment, Harbin Institute of Technology (SKLUWRE, HIT), Harbin 150090, China. Tel./fax: +86 451 8628 3077. E-mail address: [email protected] (Y. Tian).
0956-5663/$ - see front matter & 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.bios.2013.04.005
mesh was designed as both the cathode and the membrane. Wang et al., 2012 developed a more practical MFC–MBR integrated process, which used the aeration MBR tank as the cathode chamber and the carbon felt as the cathode. Both systems took MFCs as wastewater treatment units for simultaneously enhancing wastewater treatment and achieving energy recovery. However, for MBR process, stable and high effluent quality is one of the advantages (98% total organic carbon removal and 99% ammonia removal), while membrane fouling is a major obstacle (Pan et al., 2010; Visvanathan et al., 2000). It would be exciting if a combined MBR–MFC system offers the option of membrane fouling mitigation. It has been reported that sludge extracellular polymeric substances (EPS), which have been considered as the major cause of membrane fouling in MBRs (Chang et al., 2001; Drews et al., 2006), could be removed by 36.8% in the form of dissolved organic carbon (DOC) after MFC treatment (Jiang et al., 2010). Therefore, taking MFC as a sludge treatment unit might have benefits for membrane fouling mitigation due to EPS reduction and modification of sludge characteristics. A novel combined system of MBR and MFC was established in this study. The MFC was operated with a direct feed of activated sludge from the MBR. Then the non-consumed sludge in the MFC was returned to the MBR. MFC could convert the chemical energy in organic matter directly into useful electrical energy by the catalytic reaction of microorganisms; on the other hand, the sludge is hydrolyzed, converted and reduced during electricity production of MFCs (Xiao et al., 2011). 25.1% reduction in total suspended solids
X. Su et al. / Biosensors and Bioelectronics 49 (2013) 92–98
(TSS) and 22.8% reduction in volatile suspended solids (VSS) were observed in the MFC which used sewage sludge as fuel (Xiao et al., 2011). Consequently, the combined system may also have the potential for both energy recovery and sludge reduction. Therefore, the objective of this research was undertaken to evaluate the performances of the combination of MBR and MFC: (1) assess the efficiencies of wastewater treatment and sludge reduction; (2) calculate the power generation and COD transformation; and (3) investigate the membrane fouling mitigation and potential mechanisms. On this basis, a new combination approach for effective wastewater treatment, sludge reduction, energy recovery and membrane fouling mitigation is proposed.
2. Materials and methods 2.1. Microbial fuel cell (MFC) Single-chamber air cathode MFC with a cube anodic chamber (7.5 cm 7.5 cm 4 cm) was used in this study. The MFC was fitted with carbon cloth anode (projected surface area¼ 40 cm2, without wet proofing; E-TEK, USA) and cathode (projected surface area¼ 40 cm2, contained 10% platinum as catalyst and four-coating poly tetra fluoro ethylene (PTFE) diffusion layers; E-TEK, USA). The distance between two electrodes was 4 cm. An external resistance of 1000 Ω was connected across the anode and cathode electrodes. The working volume of the anodic chamber was 200 mL. The anodic compartment of the MFC directly used MBR sludge as anodic inoculum and substrate. The sludge in MFC was retained for 5 days. The MFC was inoculated with MBR sludge mentioned above repeatedly until the constant electric power was produced. After approximately 30 days culture (6 cycles), constant electrical power was obtained with MBR sludge added as fuels. Experiments were conducted in batch mode at room temperature (22 73 1C). Sludge-MFC (S-MFC) contained anaerobic digestion as well as electricity generation process. Simultaneously, a control MFC with no applied load (C-MFC) was adopted in this study. 2.2. MBR Two 8 L MBRs used in this study were operated in parallel. Each MBR was installed with a submerged hollow fiber microfiltration (MF) membrane module. The membrane modules were made of polyvinylidene fluoride (PVDF) with a nominal pore size of 0.1 μm and an effective surface area of 0.1 m2 (Motian, China). The MBR was fed with synthetic wastewater (glucose 227 mg L−1; starch 227 mg L−1; NaHCO3 254 mg L−1; urea 33 mg L−1; (NH4)2SO4 121 mg L−1; KH2PO4 15.4 mg L−1; K2HPO4 19.6 mg L−1; MgSO4 7H2O 51 mg L−1; CaCl2 12 mg L−1; ZnCl2 0.13 mg L−1; FeSO4 7H2O 17.48 mg L−1; Pb(NO3)2 0.27 mg L−1 and MnSO4 4H2O 0.13 mg L−1) from a wastewater tank. All the membrane modules were operated at the constant flux of 10 L m−2 h−1 with an intermittent suction of 8-min on and 2-min off. The mixed liquid suspended solids (MLSS) concentrations of these MBRs were maintained at almost the same level. During the experiments, chemical cleaning (soaking for 2–8 h in 0.5% sodium hypochlorite solution) was provided when the transmembrane pressure (TMP) reached 30 kPa.
93
degradation and energy recovery. At the same time, about 200 mL mixed sludge liquor was discharged from the S-MFC and recycled to the combined MBR. A parallelly operated conventional MBR system was running simultaneously as a control experiment. Both systems were performed schematically as shown in Fig. S1 lasting for 6 months. The physicochemical properties of the mixed liquid in the combined MBR, conventional MBR and MFC were monitored every two days. 2.4. Analytical methods 2.4.1. Electrical parameters Cell voltages (V) were monitored at 1 min intervals using a data acquisition system (VX5100R/C2/U, Hangzhou, China) connected to a personal computer. The maximum power densities were determined from polarization curves by varying the external resistor ranged 5000–50 Ω. Current (I), power (P ¼IV) and coulomb efficiency (CE) were calculated as previously described (Liu et al., 2004), and normalized by the cathode surface area. 2.4.2. Other analysis The concentrations of TSS, VSS, ammonia and Chemical Oxygen Demand (COD) were determined according to Standard Methods (APHA, 1995). The preparation of the supernatant was made according to the following procedures. The mixed liquor samples were firstly centrifuged at 4000 rpm for 5 min, and then the extracted supernatant was filtered through a 0.45 μm membrane filter. The filtrate was regarded as soluble microbial products (SMP). The loosely bound EPS (LB-EPS) and tightly bound EPS (TB-EPS) were extracted by twostep heating methods and analyzed for the contents of proteins and carbohydrates (Li and Yang, 2007). The calculation method of sludge yield is listed in supporting information. Excitation-emission matrix (EEM) spectra (FP 6500, JASCO, Japan) were collected with corresponding scanning emission spectra from 220 nm to 550 nm at 2 nm increments by varying the excitation wavelength from 220 nm to 400 nm at 5 nm sampling intervals. The molecular weight (MW) distribution of the organic matters in SMP was determined using a GPC (Gel Permeation Chromatography) (Agilent 1100, Agilent, USA). The number-average molecular weight (Mn) and weight-average molecular weight (Mw) were calculated. The morphological properties of the mixed liquid were described by the floc size distribution and distribution spread index (DSI) of sludge flocs. At the end of the continuous experiments, membrane resistance analyses of the fouling layer were also investigated by resistance-in-series model J ¼ ΔP T =ðμRt Þ
ð1Þ
Rt ¼ Rm þ Rc þ Rf
ð2Þ
where ΔP T is the transmembrane pressure (Pa), μ is the viscosity of the permeate (Pa s), Rc is the cake resistance formed by the cake layer deposited over the membrane surface, (m−1); Rf is the resistance caused by pore plugging and/or solute adsorption onto the membrane surface and pores (m−1); Rm is the intrinsic membrane resistance (m−1) and Rt is the total resistance (m−1). 3. Results and discussions 3.1. The performance of S-MFC, combined system and conventional MBR
2.3. Combined system In this study, a combined system of a MBR (combined MBR) and a MFC stack was set up. The stacked MFC was obtained by connecting five single S-MFCs in parallel. Every day, 200 mL raw sludge discharged from the combined MBR was pumped into the settle pool to remove the residual dissolved oxygen (DO). Then, the sludge in the settle pool was pumped into S-MFC for sludge
For new wastewater treatment systems, it is necessary to evaluate their nutrient removal efficiencies. After 5 days retention, the COD and NH4+-N concentrations of sludge supernatant within S-MFC were increased from 57 mg L−1 and 2.27 mg L−1 to 616 mg L−1 and 8.22 mg L−1 respectively (Table 1). When the treated sludge was recycled to the combined MBR, the COD and NH4+-N loads were increased by 1.49% and 0.3%, respectively. The averages of permeate
94
X. Su et al. / Biosensors and Bioelectronics 49 (2013) 92–98
COD and NH4+-N of the combined MBR were 31 mg L−1 and 2.25 mg L−1, while the corresponding values for the conventional MBR were 30 mg L−1 and 1.45 mg L−1, respectively. The COD and NH4+-N treatment efficiencies in the combined and conventional MBRs achieved more than 90%, meeting the water quality standards (50 mg L−1 for COD and 5 mg L−1 for NH4+-N), although the former were slight lower than the latter. Also listed in Table 1, a remarkable difference of sludge yield between the combined system (0.195 kgVSS kgCODremoved−1) and conventional MBR (0.221 kgVSS kgCODremoved−1) was observed. Compared with the conventional activated sludge (CAS) process (0.5 kgVSS kgCODremoved−1) (Tchobanoglous and Burton, 1991), the sludge production in the conventional MBR was reduced by 55.9%. The sludge reduction in the conventional MBR was mainly attributed to the maintenance metabolism, which was known as a process with relatively high decay rate due to the high sludge concentration in the reactor (Wen et al., 2004). It was worth noting that compared with the CAS process, the sludge production in the combined system was reduced by 61.0%, which was 5.1% higher than that in the conventional MBR. The additional sludge reduction (5.1%) obtained in the combined system might be due to the S-MFC process, which was equivalent to about 328 mgCOD d−1 consumed in the S-MFC. This value can be evaluated by the following formula: T t ¼ T s ½ðC ICi −CECi ÞY CONMBR −ðC ISi −CESi ÞY COMsys
ð3Þ
where Tt is the total COD (TCOD) of the reduced sludge, mgCOD d−1; Ts is the TCOD of sludge in the combined MBR, mgCOD gVSS−1. The other parameters are listed in supporting information. 3.2. Mass balances of COD in the combined system and the conventional MBR
analyzed. S-MFC contained anaerobic digestion as well as electricity generation process. Therefore, a control MFC with no applied load (C-MFC) was adopted in this study. The TCOD removal in the C-MFC can be attributed to the anaerobic digestion. The difference in the TCOD removal efficiency between the two MFCs might be induced by the electricity-generation process in the S-MFC. Sewage sludge with 12,741 mg L−1 TCOD was fed into S-MFC and C-MFC. The MFC produced stable voltage outputs of 430730 mV over a 400 h test (Fig. 1a) and the maximum power density obtained from the polarization experiments (Fig. 1b) was 5172 mW m−2. After 5 days of operation, the average TCOD removal efficiency for the S-MFC was 21%, i.e. the average TCOD removal capacity was about 300 mg d−1 approximately equivalent to the additional sludge reduction (5.1%) mentioned above. The corresponding TCOD removal efficiency for the C-MFC was 11%. The 10% difference (1274 mg L−1) between the two reactors might be correlated with the electricity load in the S-MFC. The COD consumption related to electricity process might include COD recovery in the form of electricity, COD metabolized by electricity-generating microbes and probable COD metabolized by common microbes stimulated by electricity etc. The theoretical amount of coulombs (Cth ) available based on the COD removed in the S-MFC (ΔCOD) could be evaluated by the formula as follows: Cth ¼ Fbes V An ΔCOD
ð4Þ
where F is Faraday constant, bes is the number of electrons exchanged per mole of oxygen, 4 mole– mol–1, and V An is volume (V) in anode. The total coulombs calculated (C R ) by integrating the current over time could be evaluated by formula as follows: Z tb Idt ð5Þ C R ¼ Ms 0
To further insight into the reason for the increase of sludge reduction in the combined system, the mass balances of COD were
where M s is the molecular weight of oxygen, 32 g mol−1, t b is the operation time (s), I is the current (A) at time t.
Table 1 Characteristics of the mixed liquors and effluent. Parameter (units) −1
MLSS (mg L ) MLVSS (mg L−1) Influent COD (mg L−1) Soluble COD in supernatant (mg L−1) COD in permeate (mg L−1) Influent NH4+-N (mg L−1) NH4+-N in supernatant (mg L−1) NH4+-N in permeate (mg L−1) Average sludge yield of the whole system (kgMLSS kgCODremoved−1)
S-MFC
Combined MBR
8350 7361
91607 472 82167 431 436 7 30 57 74 317 5 32.1 75 2.27 71.4 2.25 7 1.5 0.214
577 4 6167 24 2.277 1.5 8.22 72.1
Combined system (Combined-MBR+S-MFC)
Conventional MBR
430 7 30 577 4 31 74 327 5 2.277 1.4 2.25 7 1.5 0.195
9490 7532 8588 7 388 430 7 30 557 3 307 3 327 5 1.61 70.8 1.45 7 1.2 0.221
Values in parenthesis are standard deviations (n¼ 18).
Fig. 1. Variation of voltage output with time (a) and power density (b) for single S-MFC.
X. Su et al. / Biosensors and Bioelectronics 49 (2013) 92–98
Based on the eqs. (4) and (5) mentioned above, the current over time in the form of ΔCOD could be calculated as 75 mg L−1, accounting for 0.59% of the feed COD in the S-MFC. Consequently, the mass balances of COD in S-MFC are shown in Fig. 2. After the S-MFC treatment, 11% COD was metabolized by general anaerobic process, 0.59% COD was translated into electricity, 9.41% COD was metabolized by electricity-generating microbes and common microbes stimulated by electricity etc. The coulombic efficiency in this study was calculated as 5.9%, similar with the values obtained by other sludge MFC (More and Ghangrekar, 2010). These results indicated that combination of an S-MFC and MBR process had the advantage of decomposing sludge and recovering energy.
3.3. Membrane fouling of the combined MBR and conventional MBR TMP changes in the combined MBR and conventional MBR are shown in Fig. 3. After the operation reaching a steady state, the TMP variations of the combined MBR and the conventional MBR were taken for comparison. The TMP in the conventional MBR took about 22 days to reach 31.5 kPa while it reached 30.5 kPa after 44 days in the combined MBR, nearly twice as long as that in the conventional MBR. The rise of TMP indicated the increase of membrane fouling (Le-Clech et al., 2006). In other words, the membrane fouling of the combined MBR was dramatically mitigated. As illustrated in Fig. 3, the periods of TMP for both MBRs showed a two-phase evolution (Zhang et al., 2006): Phase 1 is a prolonged period of slow TMP rise, which might be ascribed to the accumulation of organic macromolecules either deposited from the bulk liquor or produced by biofilms on the membrane surface. Phase 2 is a sudden rise in TMP, which was mainly driven by suspended flocs. In this phase, the overall permeate productivity redistributes to the less fouled membrane areas or pores, leading to local flux increases exceeding a critical flux. The gradual fouling rate and rapid fouling rate could be analyzed to compare the fouling layer rate of development (Wu and Huang, 2010). As calculated in this study, the rates of gradual fouling and rapid fouling in the combined MBR were 0.28 kPa d−1 and 1.95 kPa d−1, respectively, which were 61% and 27% lower than those in the conventional MBR (0.72 kPa d−1 and 2.68 kPa d−1). The combined MBR was therefore effective at mitigating sudden fouling and gradual fouling. The membrane filtration resistances were calculated to explore the reasons for the slow increase in the total resistance of the combined
95
MBR (Table S1). For the combined MBR, the values of Rt, Rm, Rc and Rf were 10.90 1012 m−1, 0.98 1012 m−1, 9.43 1012 m−1, and 0.49 1012 m−1, respectively. And the corresponding values for the conventional MBR were 11.26 1012 m−1, 0.99 1012 m−1, 8.65 1012 m−1 and 1.62 1012 m−1, . The effects of both MBR systems on cake layer formation can be assessed by the average increase rates of cake resistance, which were calculated as 0.39 1012 m−1 d−1 for the conventional MBR and 0.21 1012 m−1 d−1 for the combined MBR. The increase rate of Rc in the combined MBR showed a 46% decrease compared to that in conventional MBR. Correspondingly, the average increased rates of Rf in the combined MBR and the conventional MBR were calculated as 0.011 1012 and 0.074 1012 m−1 d−1 respectively, indicating a 85% decrease of Rf increase rate for combined MBR. The results of average increased rates of Rc and Rf, agreed with the analysis of TMP profiles, confirmed that the combined MBR could effectively limit the extents of pore blocking and cake layer formation. 3.4. Mechanisms of membrane fouling mitigation As mentioned above, the combined MBR could dramatically mitigate membrane fouling by effectively limiting the rate of pore blocking and cake layer formation. It was well accepted that the membrane fouling was mainly attributed to the SMP and EPS (Meng et al., 2009). Aimed to clarify the mechanisms of fouling mitigation in the combined MBR, the effects of S-MFC on SMP and EPS in the combined MBR were investigated, respectively. 3.4.1. Effect of S-MFC on SMP Generally, the SMP present in MBR originated from the feed substrate, the release from cell lysis and the excretion of microorganism (Barker and Stuckey, 1999; Ji et al., 2008). In this study, the average SMP concentrations before and after the S-MFC treatment were 11.5 mg L−1 and 67.2 mg L−1, respectively. The increase of SMP were about 55.7 mg L−1, including 22.5 mg L−1 carbohydrates and 33.2 mg L−1 proteins. The increase of SMP occurred in S-MFC process might be due to the nutrient (such as proteins and carbohydrates) release induced by the integration of anaerobic digestion and electricity generation. Three-dimensional EEM fluorescence spectra of the SMP before and after S-MFC treatment are illustrated in Fig. 4. For the SMP in feed sludge, four peaks were identified at Ex/Em 275–285/310–340 nm, 220–225/338–342 nm, 290–320/410–420 nm and 255–265/430–440
Fig. 2. The mass balances of COD in the conventional MBR and combined system.
X. Su et al. / Biosensors and Bioelectronics 49 (2013) 92–98
Ex. Wavelength [nm]
96
400 380 360 340 320 300 280 260 240 220 220 260 300 340 380 420 460 500 550
nm, which were related to tryptophan protein-like substances (Peak A), aromatic protein-like substances (Peak B), humic-like substances (Peak C) and fulvic-acid like substances (Peak D), respectively (Baker, 2001; Chen et al., 2003). After S-MFC treatment, the fluorescent intensities of aromatic protein-like substances (Peak B) and tryptophan protein-like substances (Peak A) in SMP showed significant increase, ascribing to the accumulation of those protein-like substances. Moreover, after S-MFC treatment, the locations of the Peak A in the SMP demonstrated a blue-shift (20–30 nm) compared to those of the feed sludge. The Peak A was believed to represent the fraction of the biodegradable compounds, which were easily used for the energy requirements of microorganisms (He et al., 2011). Aromatic proteinlike substances (Peak B) were thought to be easier for biodegradation by microorganisms (Huang et al., 2011). The blue-shift of peak A is associated with the break-up of the large molecules into smaller fragments (Świetlik et al., 2004). From the results mentioned above, it could be inferred that the proteins and carbohydrates released by SMFC were degraded when they were returned to the combined MBR. The molecular weight distribution of SMP in both MBRs and MFC were compared (Table S2) to further confirm the easy degradation of SMP released by the S-MFC. In the conventional MBR, the proportion of large fraction (Mw420 kDa) was 17.63% while this proportion increased to 29.55% after S-MFC treatment. However, when the digested sludge was recycled to the combined MBR, the proportion of large fraction (Mw420 kDa) in the combined MBR (17.50%) did not show obvious difference compared with that in the conventional MBR. The substances with large molecular weight, which played a marked role in fouling, could be easily retained by the membrane. These results clearly showed that the S-MFC treated sludge could be degraded by the existing sludge in the MBR, resulting in no more large-molecular accumulated in the combined MBR.
3.4.2. Effect of S-MFC on EPS EPS were usually reported as a controlling factor of membrane fouling in MBRs (Zhang et al., 2009). The latest studies demonstrated that LB-EPS played a more significant role in membrane fouling during MBR operation (Hwang et al., 2009; Wang et al., 2009b). Li and Yang(2007) indicated that excessive EPS in the form of LB-EPS could weaken cell attachment and the floc structure, resulting in the poor bioflocculation, greater cell erosion and retarded sludge-water separation. The concentrations of EPS and LB-EPS before and after the S-MFC treatment were extracted and compared (Fig. 5). After S-MFC treatment, the EPS and LB-EPS concentrations were decreased from 94.2 mg gSS−1 to 75.63 mg gSS−1, and from 16.14 mg gSS−1 to
Ex. Wavelength [nm]
Fig. 3. Profile of TMP increases in the combined MBR and conventional MBR.
Ex. Wavelength [nm]
Em. Wavelength [nm] 400 380 360 340 320 300 280 260 240 220 220 260 300 340 380 420 460 500 550 Em. Wavelength [nm] 400 380 360 340 320 300 280 260 240 220 220 260 300 340 380 420 460 500 550 Em. Wavelength [nm]
1000
-100 1000
-100 1000
0
Fig. 4. EEM spectra of SMP in (a) combined MBR, (b) conventional MBR and (c) S-MFC treated sludge (25 times dilution).
12.56 mg gSS−1, respectively. The carbohydrates and proteins concentrations in the LB-EPS were decreased by 25.6% and 21.2%, respectively, compared to those of the feed concentrations. Thus, S-MFC sludge recycling was effective in delaying the membrane fouling by decreasing LB-EPS content. Recently, to deeply understand the relationship between EPS and fouling propensity, the EPS characteristics other than concentration were taken into account. Fig. S3 shows the EEM fluorescence spectra of the LB-EPS extracted from the sludge in the conventional MBR, combined MBR and that after S-MFC treatment. Two main peaks (Peak A and B) were readily identified at the Ex/Em of 275–280/338– 342 nm and 220–225/338–342 nm for all the samples, which were associated with the tryptophan protein-like substances and simple aromatic proteins, respectively. Compared with the conventional MBR, fluorescence intensities of the peak A and B in the LB-EPS after S-MFC treatment decreased substantially, indicating effective hydrolysis of these protein-like substances occurred in the S-MFC.
X. Su et al. / Biosensors and Bioelectronics 49 (2013) 92–98
97
Effective wastewater treatment and sludge reduction were
obtained in the combined system. The results showed that the effluent COD and NH4+-N concentrations of the combined system were always less than 40 mg L−1 and 5 mg L−1, respectively. Compared with the conventional MBR, additional sludge reduction (5.1%) was obtained in the combined system. Membrane fouling was mitigated by the sludge modification in the combined system. The operation cycle in the combined system was found nearly twice as long as that in the conventional MBR. The increased rates of Rc and Rf in the combined system were 46% and 85%, respectively, lower than those in the conventional MBR. Further investigation indicated that the MFC could effectively reduce the LB-EPS content by 22%. Additionally, the intensities of the simple aromatic proteins and tryptophan protein-like substances in LB-EPS decreased by 10% and 8% respectively, compared with the conventional MBR. The electrical energy was recovered directly from biodegradable compounds of sewage sludge. About 75 mg d−1 COD in MFC stack was translated to 5768C electricity energy. Continuous and stable electricity voltage output of over 400 mV was achieved in the single S-MFC during the operation. Additionally, 11% TCOD removal in the single S-MFC originated from electricity process.
Acknowledgments
Fig. 5. The variety of (a) bound EPS and (b) LB-EPS in conventional MBR, combined MBR and S-MFC treated sludge.
This study was supported by the Major Science and Technology Program for Water Pollution Control and Management (No. 2013Z X07201007), the State Key Laboratory of Urban Water Resource and Environment, Harbin Institute of Technology (No. 2011DX01) and the Specialized Research Fund for the Doctoral Program of Higher Education (No. 20112302110060). The authors also appreciate the Funds for Creative Research Groups of China (No. 51121062).
Appendix A. Supplementary Information When the sludge was recycled to the combined MBR, the intensities of the Peak A and peak B were 8% and 10% lower than those of the conventional MBR, respectively. Wang et al. (2009a) demonstrated that the two main peaks (Peaks A and B) appeared in the EEM fluorescence spectra of membrane foulants. Liu et al. (2011) observed that the reduction of protein-like substances represented by Peak B in EPS resulted in alleviating the membrane fouling. Thus, it was logical to suspect that the reduction of content, the simple aromatic proteins and tryptophan protein-like substances in the LB-EPS after S-MFC treatment may be favorable for membrane fouling mitigation. Besides SMP and EPS, the capillary suction time (CST), specific resistance to filtration (SRF) (Table S3) and floc size distribution (Fig. S2) were also analyzed. The detailed results were presented in the Supporting Information. Compared to the sludge in the conventional MBR, the sludge in the combined MBR had less proportion of small particles, lower DSI of sludge flocs and better dewaterability and filterability. These analyses further confirmed that the combination of MFC and MBR could limit the fouling propensity by modifying the sludge characteristics. Thus the combined system can achieve wastewater treatment, sludge reduction, energy recovery and fouling mitigation simultaneously.
4. Conclusions The effective performances were obtained in a combined system of MBR and MFC. A list of the specific results was achieved in the long-term observation:
Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.bios.2013.04.005.
References Akamatsu, K., Lu, W., Sugawara, T., Nakao, S., 2010. Water Research 44 (3), 825–830. APHA, 1995. American Public Health Association Washington, DC. Baker, A., 2001. Environmental Science and Technology 35 (5), 948–953. Bani-Melhem, K., Elektorowicz, M., 2010. Environmental Science and Technology 44 (9), 3298–3304. Barker, D., Stuckey, D., 1999. Water Research 33 (14), 3063–3082. Chang, I.S., Bag, S.O., Lee, C.H., 2001. Process Biochemistry 36 (8–9), 855–860. Chen, W., Westerhoff, P., Leenheer, J.A., Booksh, K., 2003. Environmental Science and Technology 37 (24), 5701–5710. Drews, A., Vocks, M., Iversen, V., Lesjean, B., Kraume, M., 2006. Desalination 192 (1–3), 1–9. Fan, F.S., Zhou, H.D., 2007. Environmental Science and Technology 41 (7), 2523–2528. He, X.-S., Xi, B.-D., Wei, Z.M., Jiang, Y.-H., Yang, Y., An, D., Cao, J.L., Liu, H.L., 2011. Journal of Hazardous Materials 190 (1–3), 293–299. Huang, G.C., Meng, F.G., Zheng, X., Wang, Y., Wang, Z.G., Liu, H.J., Jekel, M., 2011. Applied Microbiology and Biotechnology 90 (5), 1795–1803. Hwang, B.K., Kim, J.H., Ahn, C.H., Lee, C.H., Song, J.Y., Ra, Y.H., 2009. Water Research 44 (6), 1833–1840. Ji, J., Qiu, J.P., Wong, F.S., Li, Y.Z., 2008. Water Research 42 (14), 3611–3622. Jiang, J., Zhao, Q., Wei, L., Wang, K., 2010. Water Research 44 (7), 2163–2170. Kimura, K., Tanaka, I., Nishimura, S.-I., Miyoshi, R., Miyoshi, T., Watanabe, Y., 2012. Water Research 46 (17), 5725–5734. Le-Clech, P., Chen, V., Fane, T.A.G., 2006. Journal of Membrane Science 284 (1–2), 17–53. Li, X.Y., Yang, S.F., 2007. Water Research 41 (5), 1022–1030. Liu, H., Ramnarayanan, R., Logan, B.E., 2004. Environmental Science and Technology 38 (7), 2281–2285. Liu, T., Chen, Z., Yu, W., You, S., 2011. Water Research 45 (5), 2111–2121.
98
X. Su et al. / Biosensors and Bioelectronics 49 (2013) 92–98
Meng, F.G., Chae, S.R., Drews, A., Kraume, M., Shin, H.S., Yang, F.L., 2009. Water Research 43 (6), 1489–1512. Min, B., Kim, J., Oh, S., Regan, J.M., Logan, B.E., 2005. Water Research 39 (20), 4961–4968. More, T.T., Ghangrekar, M.M., 2010. Bioresource Technology 101 (2), 562–567. Pan, J.R., Su, Y.-C., Huang, C., Lee, H.-C., 2010. Journal of Membrane Science 349 (1–2), 287–294. Świetlik, J., Dabrowska, A., Raczyk-Stanislawiak, U., Nawrocki, J., 2004. Water Research 38 (3), 547–558. Tchobanoglous, G., Burton, F.L., 1991. McGraw-Hill, Inc. Visvanathan, C., Ben Aim, R., Parameshwaran, K., 2000. Critical Reviews in Environmental Science and Technology 30 (1), 1–48. Wang, Y.-K., Sheng, G.-P., Li, W.-W., Huang, Y.-X., Yu, Y.-Y., Zeng, R.J., Yu, H.-Q., 2011. Environmental Science and Technology 45 (21), 9256–9261.
Wang, Y.-P., Liu, X.-W., Li, W.-W., Li, F., Wang, Y.-K., Sheng, G.-P., Zeng, R.J., Yu, H.-Q., 2012. Applied Energy 98 (0), 230–235. Wang, Z.W., Wu, Z.C., Tang, S.J., 2009a. Water Research 43 (6), 1533–1540. Wang, Z.W., Wu, Z.C., Tang, S.J., 2009b. Water Research 43 (9), 2504–2512. Wen, X.H., Ding, H.J., Huang, X., Liu, R.P., 2004. Process Biochemistry 39 (11), 1427–1431. Wu, J.L., Huang, X., 2010. Bioresource Technology 101 (15), 6019–6027. Xiao, B.Y., Yang, F., Liu, J.X., 2011. Journal of Hazardous Materials 189 (1–2), 444–449. Zhang, H.M., Xia, J., Yang, Y., Wang, Z.X., Yang, F.L., 2009. Journal of Environmental Sciences 21 (8), 1066–1073. Zhang, J., Chua, H.C., Zhou, J., Fane, A.G., 2006. Journal of Membrane Science 284 (1–2), 54–66.
Contents lists available at SciVerse ScienceDirect
Biosensors and Bioelectronics journal homepage: www.elsevier.com/locate/bios
Performance of a combined system of microbial fuel cell and membrane bioreactor: Wastewater treatment, sludge reduction, energy recovery and membrane fouling Xinying Su b, Yu Tian a,b,n, Zhicai Sun c, Yaobin Lu b, Zhipeng Li b a
State Key Laboratory of Urban Water Resource and Environment, Harbin Institute of Technology (SKLUWRE, HIT), Harbin 150090, China School of Municipal and Environmental Engineering, Harbin Institute of Technology, Harbin 150090, China c Daqing Water Group Company Limited, Daqing 163311, China b
art ic l e i nf o
a b s t r a c t
Article history: Received 25 February 2013 Received in revised form 2 April 2013 Accepted 4 April 2013 Available online 18 April 2013
A novel combined system of sludge microbial fuel cell (S-MFC) stack and membrane bioreactor (MBR) was proposed in this study. The non-consumed sludge in the MBR sludge-fed S-MFC was recycled to the MBR. In the combined system, the COD and ammonia treatment efficiencies were more than 90% and the sludge reduction was 5.1% higher than that of the conventional MBR. It's worth noting that the energy recovery and fouling mitigation were observed in the combined system. In the single S-MFC, about 75 mg L−1 COD could be translated to electricity during one cycle. The average voltage and maximum power production of the single S-MFC were 430 mV and 51 mW m−2, respectively. Additionally, the combined system was able to mitigate membrane fouling by the sludge modification. Except for the content decrease (22%), S-MFC destroyed simple aromatic proteins and tryptophan protein-like substances in loosely bound extracellular polymeric substances (LB-EPS). These results indicated that effective wastewater treatment, sludge reduction, energy recovery and membrane fouling mitigation could be obtained in the combined system. & 2013 Elsevier B.V. All rights reserved.
Keywords: Membrane bioreactor (MBR) Microbial fuel cell (MFC) Membrane fouling Loosely bound EPS (LB-EPS) Energy recovery
1. Introduction As efficient technology for wastewater treatment, membrane bioreactors (MBRs) have experienced unprecedented growth in recent years (Fan and Zhou, 2007). MBRs offer several advantages over conventional activated sludge system, including stable and high effluent quality, ease of operation, small footprint, and absolute removal of bacteria (Wang et al., 2009b). However, membrane fouling remains as a major obstacle for wider application of MBRs (Kimura et al., 2012). The most common approaches to reduce membrane fouling are supplying an excessive amount of air (high energy con-sumption), cleaning the membrane module (reduction the membrane life-span), optimizing the operating parameters and improving activated sludge characteristics in the reactor (Akamatsu et al., 2010; Bani-Melhem and Elektorowicz, 2010). Microbial fuel cells (MFCs) have been developed as a promising technology to recover energy from wastewater, marine sediments, sludge etc. (Min et al., 2005). Several systems combining MBR and MFC have been reported previously. A bioelectrochemical membrane reactor, which takes advantage of both MBR and MFC processes, was recently developed (Wang et al., 2011). In this reactor, stainless steel n Corresponding author at: State Key Laboratory of Urban Water Resource and Environment, Harbin Institute of Technology (SKLUWRE, HIT), Harbin 150090, China. Tel./fax: +86 451 8628 3077. E-mail address: [email protected] (Y. Tian).
0956-5663/$ - see front matter & 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.bios.2013.04.005
mesh was designed as both the cathode and the membrane. Wang et al., 2012 developed a more practical MFC–MBR integrated process, which used the aeration MBR tank as the cathode chamber and the carbon felt as the cathode. Both systems took MFCs as wastewater treatment units for simultaneously enhancing wastewater treatment and achieving energy recovery. However, for MBR process, stable and high effluent quality is one of the advantages (98% total organic carbon removal and 99% ammonia removal), while membrane fouling is a major obstacle (Pan et al., 2010; Visvanathan et al., 2000). It would be exciting if a combined MBR–MFC system offers the option of membrane fouling mitigation. It has been reported that sludge extracellular polymeric substances (EPS), which have been considered as the major cause of membrane fouling in MBRs (Chang et al., 2001; Drews et al., 2006), could be removed by 36.8% in the form of dissolved organic carbon (DOC) after MFC treatment (Jiang et al., 2010). Therefore, taking MFC as a sludge treatment unit might have benefits for membrane fouling mitigation due to EPS reduction and modification of sludge characteristics. A novel combined system of MBR and MFC was established in this study. The MFC was operated with a direct feed of activated sludge from the MBR. Then the non-consumed sludge in the MFC was returned to the MBR. MFC could convert the chemical energy in organic matter directly into useful electrical energy by the catalytic reaction of microorganisms; on the other hand, the sludge is hydrolyzed, converted and reduced during electricity production of MFCs (Xiao et al., 2011). 25.1% reduction in total suspended solids
X. Su et al. / Biosensors and Bioelectronics 49 (2013) 92–98
(TSS) and 22.8% reduction in volatile suspended solids (VSS) were observed in the MFC which used sewage sludge as fuel (Xiao et al., 2011). Consequently, the combined system may also have the potential for both energy recovery and sludge reduction. Therefore, the objective of this research was undertaken to evaluate the performances of the combination of MBR and MFC: (1) assess the efficiencies of wastewater treatment and sludge reduction; (2) calculate the power generation and COD transformation; and (3) investigate the membrane fouling mitigation and potential mechanisms. On this basis, a new combination approach for effective wastewater treatment, sludge reduction, energy recovery and membrane fouling mitigation is proposed.
2. Materials and methods 2.1. Microbial fuel cell (MFC) Single-chamber air cathode MFC with a cube anodic chamber (7.5 cm 7.5 cm 4 cm) was used in this study. The MFC was fitted with carbon cloth anode (projected surface area¼ 40 cm2, without wet proofing; E-TEK, USA) and cathode (projected surface area¼ 40 cm2, contained 10% platinum as catalyst and four-coating poly tetra fluoro ethylene (PTFE) diffusion layers; E-TEK, USA). The distance between two electrodes was 4 cm. An external resistance of 1000 Ω was connected across the anode and cathode electrodes. The working volume of the anodic chamber was 200 mL. The anodic compartment of the MFC directly used MBR sludge as anodic inoculum and substrate. The sludge in MFC was retained for 5 days. The MFC was inoculated with MBR sludge mentioned above repeatedly until the constant electric power was produced. After approximately 30 days culture (6 cycles), constant electrical power was obtained with MBR sludge added as fuels. Experiments were conducted in batch mode at room temperature (22 73 1C). Sludge-MFC (S-MFC) contained anaerobic digestion as well as electricity generation process. Simultaneously, a control MFC with no applied load (C-MFC) was adopted in this study. 2.2. MBR Two 8 L MBRs used in this study were operated in parallel. Each MBR was installed with a submerged hollow fiber microfiltration (MF) membrane module. The membrane modules were made of polyvinylidene fluoride (PVDF) with a nominal pore size of 0.1 μm and an effective surface area of 0.1 m2 (Motian, China). The MBR was fed with synthetic wastewater (glucose 227 mg L−1; starch 227 mg L−1; NaHCO3 254 mg L−1; urea 33 mg L−1; (NH4)2SO4 121 mg L−1; KH2PO4 15.4 mg L−1; K2HPO4 19.6 mg L−1; MgSO4 7H2O 51 mg L−1; CaCl2 12 mg L−1; ZnCl2 0.13 mg L−1; FeSO4 7H2O 17.48 mg L−1; Pb(NO3)2 0.27 mg L−1 and MnSO4 4H2O 0.13 mg L−1) from a wastewater tank. All the membrane modules were operated at the constant flux of 10 L m−2 h−1 with an intermittent suction of 8-min on and 2-min off. The mixed liquid suspended solids (MLSS) concentrations of these MBRs were maintained at almost the same level. During the experiments, chemical cleaning (soaking for 2–8 h in 0.5% sodium hypochlorite solution) was provided when the transmembrane pressure (TMP) reached 30 kPa.
93
degradation and energy recovery. At the same time, about 200 mL mixed sludge liquor was discharged from the S-MFC and recycled to the combined MBR. A parallelly operated conventional MBR system was running simultaneously as a control experiment. Both systems were performed schematically as shown in Fig. S1 lasting for 6 months. The physicochemical properties of the mixed liquid in the combined MBR, conventional MBR and MFC were monitored every two days. 2.4. Analytical methods 2.4.1. Electrical parameters Cell voltages (V) were monitored at 1 min intervals using a data acquisition system (VX5100R/C2/U, Hangzhou, China) connected to a personal computer. The maximum power densities were determined from polarization curves by varying the external resistor ranged 5000–50 Ω. Current (I), power (P ¼IV) and coulomb efficiency (CE) were calculated as previously described (Liu et al., 2004), and normalized by the cathode surface area. 2.4.2. Other analysis The concentrations of TSS, VSS, ammonia and Chemical Oxygen Demand (COD) were determined according to Standard Methods (APHA, 1995). The preparation of the supernatant was made according to the following procedures. The mixed liquor samples were firstly centrifuged at 4000 rpm for 5 min, and then the extracted supernatant was filtered through a 0.45 μm membrane filter. The filtrate was regarded as soluble microbial products (SMP). The loosely bound EPS (LB-EPS) and tightly bound EPS (TB-EPS) were extracted by twostep heating methods and analyzed for the contents of proteins and carbohydrates (Li and Yang, 2007). The calculation method of sludge yield is listed in supporting information. Excitation-emission matrix (EEM) spectra (FP 6500, JASCO, Japan) were collected with corresponding scanning emission spectra from 220 nm to 550 nm at 2 nm increments by varying the excitation wavelength from 220 nm to 400 nm at 5 nm sampling intervals. The molecular weight (MW) distribution of the organic matters in SMP was determined using a GPC (Gel Permeation Chromatography) (Agilent 1100, Agilent, USA). The number-average molecular weight (Mn) and weight-average molecular weight (Mw) were calculated. The morphological properties of the mixed liquid were described by the floc size distribution and distribution spread index (DSI) of sludge flocs. At the end of the continuous experiments, membrane resistance analyses of the fouling layer were also investigated by resistance-in-series model J ¼ ΔP T =ðμRt Þ
ð1Þ
Rt ¼ Rm þ Rc þ Rf
ð2Þ
where ΔP T is the transmembrane pressure (Pa), μ is the viscosity of the permeate (Pa s), Rc is the cake resistance formed by the cake layer deposited over the membrane surface, (m−1); Rf is the resistance caused by pore plugging and/or solute adsorption onto the membrane surface and pores (m−1); Rm is the intrinsic membrane resistance (m−1) and Rt is the total resistance (m−1). 3. Results and discussions 3.1. The performance of S-MFC, combined system and conventional MBR
2.3. Combined system In this study, a combined system of a MBR (combined MBR) and a MFC stack was set up. The stacked MFC was obtained by connecting five single S-MFCs in parallel. Every day, 200 mL raw sludge discharged from the combined MBR was pumped into the settle pool to remove the residual dissolved oxygen (DO). Then, the sludge in the settle pool was pumped into S-MFC for sludge
For new wastewater treatment systems, it is necessary to evaluate their nutrient removal efficiencies. After 5 days retention, the COD and NH4+-N concentrations of sludge supernatant within S-MFC were increased from 57 mg L−1 and 2.27 mg L−1 to 616 mg L−1 and 8.22 mg L−1 respectively (Table 1). When the treated sludge was recycled to the combined MBR, the COD and NH4+-N loads were increased by 1.49% and 0.3%, respectively. The averages of permeate
94
X. Su et al. / Biosensors and Bioelectronics 49 (2013) 92–98
COD and NH4+-N of the combined MBR were 31 mg L−1 and 2.25 mg L−1, while the corresponding values for the conventional MBR were 30 mg L−1 and 1.45 mg L−1, respectively. The COD and NH4+-N treatment efficiencies in the combined and conventional MBRs achieved more than 90%, meeting the water quality standards (50 mg L−1 for COD and 5 mg L−1 for NH4+-N), although the former were slight lower than the latter. Also listed in Table 1, a remarkable difference of sludge yield between the combined system (0.195 kgVSS kgCODremoved−1) and conventional MBR (0.221 kgVSS kgCODremoved−1) was observed. Compared with the conventional activated sludge (CAS) process (0.5 kgVSS kgCODremoved−1) (Tchobanoglous and Burton, 1991), the sludge production in the conventional MBR was reduced by 55.9%. The sludge reduction in the conventional MBR was mainly attributed to the maintenance metabolism, which was known as a process with relatively high decay rate due to the high sludge concentration in the reactor (Wen et al., 2004). It was worth noting that compared with the CAS process, the sludge production in the combined system was reduced by 61.0%, which was 5.1% higher than that in the conventional MBR. The additional sludge reduction (5.1%) obtained in the combined system might be due to the S-MFC process, which was equivalent to about 328 mgCOD d−1 consumed in the S-MFC. This value can be evaluated by the following formula: T t ¼ T s ½ðC ICi −CECi ÞY CONMBR −ðC ISi −CESi ÞY COMsys
ð3Þ
where Tt is the total COD (TCOD) of the reduced sludge, mgCOD d−1; Ts is the TCOD of sludge in the combined MBR, mgCOD gVSS−1. The other parameters are listed in supporting information. 3.2. Mass balances of COD in the combined system and the conventional MBR
analyzed. S-MFC contained anaerobic digestion as well as electricity generation process. Therefore, a control MFC with no applied load (C-MFC) was adopted in this study. The TCOD removal in the C-MFC can be attributed to the anaerobic digestion. The difference in the TCOD removal efficiency between the two MFCs might be induced by the electricity-generation process in the S-MFC. Sewage sludge with 12,741 mg L−1 TCOD was fed into S-MFC and C-MFC. The MFC produced stable voltage outputs of 430730 mV over a 400 h test (Fig. 1a) and the maximum power density obtained from the polarization experiments (Fig. 1b) was 5172 mW m−2. After 5 days of operation, the average TCOD removal efficiency for the S-MFC was 21%, i.e. the average TCOD removal capacity was about 300 mg d−1 approximately equivalent to the additional sludge reduction (5.1%) mentioned above. The corresponding TCOD removal efficiency for the C-MFC was 11%. The 10% difference (1274 mg L−1) between the two reactors might be correlated with the electricity load in the S-MFC. The COD consumption related to electricity process might include COD recovery in the form of electricity, COD metabolized by electricity-generating microbes and probable COD metabolized by common microbes stimulated by electricity etc. The theoretical amount of coulombs (Cth ) available based on the COD removed in the S-MFC (ΔCOD) could be evaluated by the formula as follows: Cth ¼ Fbes V An ΔCOD
ð4Þ
where F is Faraday constant, bes is the number of electrons exchanged per mole of oxygen, 4 mole– mol–1, and V An is volume (V) in anode. The total coulombs calculated (C R ) by integrating the current over time could be evaluated by formula as follows: Z tb Idt ð5Þ C R ¼ Ms 0
To further insight into the reason for the increase of sludge reduction in the combined system, the mass balances of COD were
where M s is the molecular weight of oxygen, 32 g mol−1, t b is the operation time (s), I is the current (A) at time t.
Table 1 Characteristics of the mixed liquors and effluent. Parameter (units) −1
MLSS (mg L ) MLVSS (mg L−1) Influent COD (mg L−1) Soluble COD in supernatant (mg L−1) COD in permeate (mg L−1) Influent NH4+-N (mg L−1) NH4+-N in supernatant (mg L−1) NH4+-N in permeate (mg L−1) Average sludge yield of the whole system (kgMLSS kgCODremoved−1)
S-MFC
Combined MBR
8350 7361
91607 472 82167 431 436 7 30 57 74 317 5 32.1 75 2.27 71.4 2.25 7 1.5 0.214
577 4 6167 24 2.277 1.5 8.22 72.1
Combined system (Combined-MBR+S-MFC)
Conventional MBR
430 7 30 577 4 31 74 327 5 2.277 1.4 2.25 7 1.5 0.195
9490 7532 8588 7 388 430 7 30 557 3 307 3 327 5 1.61 70.8 1.45 7 1.2 0.221
Values in parenthesis are standard deviations (n¼ 18).
Fig. 1. Variation of voltage output with time (a) and power density (b) for single S-MFC.
X. Su et al. / Biosensors and Bioelectronics 49 (2013) 92–98
Based on the eqs. (4) and (5) mentioned above, the current over time in the form of ΔCOD could be calculated as 75 mg L−1, accounting for 0.59% of the feed COD in the S-MFC. Consequently, the mass balances of COD in S-MFC are shown in Fig. 2. After the S-MFC treatment, 11% COD was metabolized by general anaerobic process, 0.59% COD was translated into electricity, 9.41% COD was metabolized by electricity-generating microbes and common microbes stimulated by electricity etc. The coulombic efficiency in this study was calculated as 5.9%, similar with the values obtained by other sludge MFC (More and Ghangrekar, 2010). These results indicated that combination of an S-MFC and MBR process had the advantage of decomposing sludge and recovering energy.
3.3. Membrane fouling of the combined MBR and conventional MBR TMP changes in the combined MBR and conventional MBR are shown in Fig. 3. After the operation reaching a steady state, the TMP variations of the combined MBR and the conventional MBR were taken for comparison. The TMP in the conventional MBR took about 22 days to reach 31.5 kPa while it reached 30.5 kPa after 44 days in the combined MBR, nearly twice as long as that in the conventional MBR. The rise of TMP indicated the increase of membrane fouling (Le-Clech et al., 2006). In other words, the membrane fouling of the combined MBR was dramatically mitigated. As illustrated in Fig. 3, the periods of TMP for both MBRs showed a two-phase evolution (Zhang et al., 2006): Phase 1 is a prolonged period of slow TMP rise, which might be ascribed to the accumulation of organic macromolecules either deposited from the bulk liquor or produced by biofilms on the membrane surface. Phase 2 is a sudden rise in TMP, which was mainly driven by suspended flocs. In this phase, the overall permeate productivity redistributes to the less fouled membrane areas or pores, leading to local flux increases exceeding a critical flux. The gradual fouling rate and rapid fouling rate could be analyzed to compare the fouling layer rate of development (Wu and Huang, 2010). As calculated in this study, the rates of gradual fouling and rapid fouling in the combined MBR were 0.28 kPa d−1 and 1.95 kPa d−1, respectively, which were 61% and 27% lower than those in the conventional MBR (0.72 kPa d−1 and 2.68 kPa d−1). The combined MBR was therefore effective at mitigating sudden fouling and gradual fouling. The membrane filtration resistances were calculated to explore the reasons for the slow increase in the total resistance of the combined
95
MBR (Table S1). For the combined MBR, the values of Rt, Rm, Rc and Rf were 10.90 1012 m−1, 0.98 1012 m−1, 9.43 1012 m−1, and 0.49 1012 m−1, respectively. And the corresponding values for the conventional MBR were 11.26 1012 m−1, 0.99 1012 m−1, 8.65 1012 m−1 and 1.62 1012 m−1, . The effects of both MBR systems on cake layer formation can be assessed by the average increase rates of cake resistance, which were calculated as 0.39 1012 m−1 d−1 for the conventional MBR and 0.21 1012 m−1 d−1 for the combined MBR. The increase rate of Rc in the combined MBR showed a 46% decrease compared to that in conventional MBR. Correspondingly, the average increased rates of Rf in the combined MBR and the conventional MBR were calculated as 0.011 1012 and 0.074 1012 m−1 d−1 respectively, indicating a 85% decrease of Rf increase rate for combined MBR. The results of average increased rates of Rc and Rf, agreed with the analysis of TMP profiles, confirmed that the combined MBR could effectively limit the extents of pore blocking and cake layer formation. 3.4. Mechanisms of membrane fouling mitigation As mentioned above, the combined MBR could dramatically mitigate membrane fouling by effectively limiting the rate of pore blocking and cake layer formation. It was well accepted that the membrane fouling was mainly attributed to the SMP and EPS (Meng et al., 2009). Aimed to clarify the mechanisms of fouling mitigation in the combined MBR, the effects of S-MFC on SMP and EPS in the combined MBR were investigated, respectively. 3.4.1. Effect of S-MFC on SMP Generally, the SMP present in MBR originated from the feed substrate, the release from cell lysis and the excretion of microorganism (Barker and Stuckey, 1999; Ji et al., 2008). In this study, the average SMP concentrations before and after the S-MFC treatment were 11.5 mg L−1 and 67.2 mg L−1, respectively. The increase of SMP were about 55.7 mg L−1, including 22.5 mg L−1 carbohydrates and 33.2 mg L−1 proteins. The increase of SMP occurred in S-MFC process might be due to the nutrient (such as proteins and carbohydrates) release induced by the integration of anaerobic digestion and electricity generation. Three-dimensional EEM fluorescence spectra of the SMP before and after S-MFC treatment are illustrated in Fig. 4. For the SMP in feed sludge, four peaks were identified at Ex/Em 275–285/310–340 nm, 220–225/338–342 nm, 290–320/410–420 nm and 255–265/430–440
Fig. 2. The mass balances of COD in the conventional MBR and combined system.
X. Su et al. / Biosensors and Bioelectronics 49 (2013) 92–98
Ex. Wavelength [nm]
96
400 380 360 340 320 300 280 260 240 220 220 260 300 340 380 420 460 500 550
nm, which were related to tryptophan protein-like substances (Peak A), aromatic protein-like substances (Peak B), humic-like substances (Peak C) and fulvic-acid like substances (Peak D), respectively (Baker, 2001; Chen et al., 2003). After S-MFC treatment, the fluorescent intensities of aromatic protein-like substances (Peak B) and tryptophan protein-like substances (Peak A) in SMP showed significant increase, ascribing to the accumulation of those protein-like substances. Moreover, after S-MFC treatment, the locations of the Peak A in the SMP demonstrated a blue-shift (20–30 nm) compared to those of the feed sludge. The Peak A was believed to represent the fraction of the biodegradable compounds, which were easily used for the energy requirements of microorganisms (He et al., 2011). Aromatic proteinlike substances (Peak B) were thought to be easier for biodegradation by microorganisms (Huang et al., 2011). The blue-shift of peak A is associated with the break-up of the large molecules into smaller fragments (Świetlik et al., 2004). From the results mentioned above, it could be inferred that the proteins and carbohydrates released by SMFC were degraded when they were returned to the combined MBR. The molecular weight distribution of SMP in both MBRs and MFC were compared (Table S2) to further confirm the easy degradation of SMP released by the S-MFC. In the conventional MBR, the proportion of large fraction (Mw420 kDa) was 17.63% while this proportion increased to 29.55% after S-MFC treatment. However, when the digested sludge was recycled to the combined MBR, the proportion of large fraction (Mw420 kDa) in the combined MBR (17.50%) did not show obvious difference compared with that in the conventional MBR. The substances with large molecular weight, which played a marked role in fouling, could be easily retained by the membrane. These results clearly showed that the S-MFC treated sludge could be degraded by the existing sludge in the MBR, resulting in no more large-molecular accumulated in the combined MBR.
3.4.2. Effect of S-MFC on EPS EPS were usually reported as a controlling factor of membrane fouling in MBRs (Zhang et al., 2009). The latest studies demonstrated that LB-EPS played a more significant role in membrane fouling during MBR operation (Hwang et al., 2009; Wang et al., 2009b). Li and Yang(2007) indicated that excessive EPS in the form of LB-EPS could weaken cell attachment and the floc structure, resulting in the poor bioflocculation, greater cell erosion and retarded sludge-water separation. The concentrations of EPS and LB-EPS before and after the S-MFC treatment were extracted and compared (Fig. 5). After S-MFC treatment, the EPS and LB-EPS concentrations were decreased from 94.2 mg gSS−1 to 75.63 mg gSS−1, and from 16.14 mg gSS−1 to
Ex. Wavelength [nm]
Fig. 3. Profile of TMP increases in the combined MBR and conventional MBR.
Ex. Wavelength [nm]
Em. Wavelength [nm] 400 380 360 340 320 300 280 260 240 220 220 260 300 340 380 420 460 500 550 Em. Wavelength [nm] 400 380 360 340 320 300 280 260 240 220 220 260 300 340 380 420 460 500 550 Em. Wavelength [nm]
1000
-100 1000
-100 1000
0
Fig. 4. EEM spectra of SMP in (a) combined MBR, (b) conventional MBR and (c) S-MFC treated sludge (25 times dilution).
12.56 mg gSS−1, respectively. The carbohydrates and proteins concentrations in the LB-EPS were decreased by 25.6% and 21.2%, respectively, compared to those of the feed concentrations. Thus, S-MFC sludge recycling was effective in delaying the membrane fouling by decreasing LB-EPS content. Recently, to deeply understand the relationship between EPS and fouling propensity, the EPS characteristics other than concentration were taken into account. Fig. S3 shows the EEM fluorescence spectra of the LB-EPS extracted from the sludge in the conventional MBR, combined MBR and that after S-MFC treatment. Two main peaks (Peak A and B) were readily identified at the Ex/Em of 275–280/338– 342 nm and 220–225/338–342 nm for all the samples, which were associated with the tryptophan protein-like substances and simple aromatic proteins, respectively. Compared with the conventional MBR, fluorescence intensities of the peak A and B in the LB-EPS after S-MFC treatment decreased substantially, indicating effective hydrolysis of these protein-like substances occurred in the S-MFC.
X. Su et al. / Biosensors and Bioelectronics 49 (2013) 92–98
97
Effective wastewater treatment and sludge reduction were
obtained in the combined system. The results showed that the effluent COD and NH4+-N concentrations of the combined system were always less than 40 mg L−1 and 5 mg L−1, respectively. Compared with the conventional MBR, additional sludge reduction (5.1%) was obtained in the combined system. Membrane fouling was mitigated by the sludge modification in the combined system. The operation cycle in the combined system was found nearly twice as long as that in the conventional MBR. The increased rates of Rc and Rf in the combined system were 46% and 85%, respectively, lower than those in the conventional MBR. Further investigation indicated that the MFC could effectively reduce the LB-EPS content by 22%. Additionally, the intensities of the simple aromatic proteins and tryptophan protein-like substances in LB-EPS decreased by 10% and 8% respectively, compared with the conventional MBR. The electrical energy was recovered directly from biodegradable compounds of sewage sludge. About 75 mg d−1 COD in MFC stack was translated to 5768C electricity energy. Continuous and stable electricity voltage output of over 400 mV was achieved in the single S-MFC during the operation. Additionally, 11% TCOD removal in the single S-MFC originated from electricity process.
Acknowledgments
Fig. 5. The variety of (a) bound EPS and (b) LB-EPS in conventional MBR, combined MBR and S-MFC treated sludge.
This study was supported by the Major Science and Technology Program for Water Pollution Control and Management (No. 2013Z X07201007), the State Key Laboratory of Urban Water Resource and Environment, Harbin Institute of Technology (No. 2011DX01) and the Specialized Research Fund for the Doctoral Program of Higher Education (No. 20112302110060). The authors also appreciate the Funds for Creative Research Groups of China (No. 51121062).
Appendix A. Supplementary Information When the sludge was recycled to the combined MBR, the intensities of the Peak A and peak B were 8% and 10% lower than those of the conventional MBR, respectively. Wang et al. (2009a) demonstrated that the two main peaks (Peaks A and B) appeared in the EEM fluorescence spectra of membrane foulants. Liu et al. (2011) observed that the reduction of protein-like substances represented by Peak B in EPS resulted in alleviating the membrane fouling. Thus, it was logical to suspect that the reduction of content, the simple aromatic proteins and tryptophan protein-like substances in the LB-EPS after S-MFC treatment may be favorable for membrane fouling mitigation. Besides SMP and EPS, the capillary suction time (CST), specific resistance to filtration (SRF) (Table S3) and floc size distribution (Fig. S2) were also analyzed. The detailed results were presented in the Supporting Information. Compared to the sludge in the conventional MBR, the sludge in the combined MBR had less proportion of small particles, lower DSI of sludge flocs and better dewaterability and filterability. These analyses further confirmed that the combination of MFC and MBR could limit the fouling propensity by modifying the sludge characteristics. Thus the combined system can achieve wastewater treatment, sludge reduction, energy recovery and fouling mitigation simultaneously.
4. Conclusions The effective performances were obtained in a combined system of MBR and MFC. A list of the specific results was achieved in the long-term observation:
Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.bios.2013.04.005.
References Akamatsu, K., Lu, W., Sugawara, T., Nakao, S., 2010. Water Research 44 (3), 825–830. APHA, 1995. American Public Health Association Washington, DC. Baker, A., 2001. Environmental Science and Technology 35 (5), 948–953. Bani-Melhem, K., Elektorowicz, M., 2010. Environmental Science and Technology 44 (9), 3298–3304. Barker, D., Stuckey, D., 1999. Water Research 33 (14), 3063–3082. Chang, I.S., Bag, S.O., Lee, C.H., 2001. Process Biochemistry 36 (8–9), 855–860. Chen, W., Westerhoff, P., Leenheer, J.A., Booksh, K., 2003. Environmental Science and Technology 37 (24), 5701–5710. Drews, A., Vocks, M., Iversen, V., Lesjean, B., Kraume, M., 2006. Desalination 192 (1–3), 1–9. Fan, F.S., Zhou, H.D., 2007. Environmental Science and Technology 41 (7), 2523–2528. He, X.-S., Xi, B.-D., Wei, Z.M., Jiang, Y.-H., Yang, Y., An, D., Cao, J.L., Liu, H.L., 2011. Journal of Hazardous Materials 190 (1–3), 293–299. Huang, G.C., Meng, F.G., Zheng, X., Wang, Y., Wang, Z.G., Liu, H.J., Jekel, M., 2011. Applied Microbiology and Biotechnology 90 (5), 1795–1803. Hwang, B.K., Kim, J.H., Ahn, C.H., Lee, C.H., Song, J.Y., Ra, Y.H., 2009. Water Research 44 (6), 1833–1840. Ji, J., Qiu, J.P., Wong, F.S., Li, Y.Z., 2008. Water Research 42 (14), 3611–3622. Jiang, J., Zhao, Q., Wei, L., Wang, K., 2010. Water Research 44 (7), 2163–2170. Kimura, K., Tanaka, I., Nishimura, S.-I., Miyoshi, R., Miyoshi, T., Watanabe, Y., 2012. Water Research 46 (17), 5725–5734. Le-Clech, P., Chen, V., Fane, T.A.G., 2006. Journal of Membrane Science 284 (1–2), 17–53. Li, X.Y., Yang, S.F., 2007. Water Research 41 (5), 1022–1030. Liu, H., Ramnarayanan, R., Logan, B.E., 2004. Environmental Science and Technology 38 (7), 2281–2285. Liu, T., Chen, Z., Yu, W., You, S., 2011. Water Research 45 (5), 2111–2121.
98
X. Su et al. / Biosensors and Bioelectronics 49 (2013) 92–98
Meng, F.G., Chae, S.R., Drews, A., Kraume, M., Shin, H.S., Yang, F.L., 2009. Water Research 43 (6), 1489–1512. Min, B., Kim, J., Oh, S., Regan, J.M., Logan, B.E., 2005. Water Research 39 (20), 4961–4968. More, T.T., Ghangrekar, M.M., 2010. Bioresource Technology 101 (2), 562–567. Pan, J.R., Su, Y.-C., Huang, C., Lee, H.-C., 2010. Journal of Membrane Science 349 (1–2), 287–294. Świetlik, J., Dabrowska, A., Raczyk-Stanislawiak, U., Nawrocki, J., 2004. Water Research 38 (3), 547–558. Tchobanoglous, G., Burton, F.L., 1991. McGraw-Hill, Inc. Visvanathan, C., Ben Aim, R., Parameshwaran, K., 2000. Critical Reviews in Environmental Science and Technology 30 (1), 1–48. Wang, Y.-K., Sheng, G.-P., Li, W.-W., Huang, Y.-X., Yu, Y.-Y., Zeng, R.J., Yu, H.-Q., 2011. Environmental Science and Technology 45 (21), 9256–9261.
Wang, Y.-P., Liu, X.-W., Li, W.-W., Li, F., Wang, Y.-K., Sheng, G.-P., Zeng, R.J., Yu, H.-Q., 2012. Applied Energy 98 (0), 230–235. Wang, Z.W., Wu, Z.C., Tang, S.J., 2009a. Water Research 43 (6), 1533–1540. Wang, Z.W., Wu, Z.C., Tang, S.J., 2009b. Water Research 43 (9), 2504–2512. Wen, X.H., Ding, H.J., Huang, X., Liu, R.P., 2004. Process Biochemistry 39 (11), 1427–1431. Wu, J.L., Huang, X., 2010. Bioresource Technology 101 (15), 6019–6027. Xiao, B.Y., Yang, F., Liu, J.X., 2011. Journal of Hazardous Materials 189 (1–2), 444–449. Zhang, H.M., Xia, J., Yang, Y., Wang, Z.X., Yang, F.L., 2009. Journal of Environmental Sciences 21 (8), 1066–1073. Zhang, J., Chua, H.C., Zhou, J., Fane, A.G., 2006. Journal of Membrane Science 284 (1–2), 54–66.