Floatation of granular sludge and its mechanism: A key approach for high-rate denitrifying reactor

Bioresource Technology 152 (2014) 414–419

Contents lists available at ScienceDirect

Bioresource Technology journal homepage: www.elsevier.com/locate/biortech

Floatation of granular sludge and its mechanism: A key approach for high-rate denitrifying reactor Wei Li a, Ping Zheng a,⇑, Junyuan Ji b, Meng Zhang a, Jun Guo c, Jiqiang Zhang a, Ghulam Abbas a,d a

Department of Environmental Engineering, Zhejiang University, Hangzhou 310058, China College of Environmental Science and Engineering, Ocean University of China, Qingdao 266100, China c College of Environmental Science and Engineering, Tongji University, Shanghai 200000, China d Department of Chemical Engineering, University of Gujrat, Gujrat, Pakistan b

h i g h l i g h t s  A high-rate denitrifying automatic circulation (DAC) reactor was developed.  Granule floatation was a limiting factor for performance of high-rate DAC reactor.  The floating mechanism of denitrifying granule at high load was revealed.  An evaluation model for physical characteristics of floated granules was proposed.

a r t i c l e

i n f o

Article history: Received 6 October 2013 Received in revised form 14 November 2013 Accepted 20 November 2013 Available online 28 November 2013 Keywords: High-rate DAC reactor Working performance Granular sludge floatation Floatation mechanism

a b s t r a c t A high-rate denitrifying automatic circulate (DAC) reactor has been developed recently, and it is promising to become an alternative in nitrogen removal from wastewaters. However, the performance of DAC reactor was disturbed by the floatation of granular sludge at high-loads. The results showed that: the floatation of granular sludge led to a serious biomass washout and a sharp decrease of biomass concentration. The floatation of granular sludge was ascribed to a low sludge density originated from the holdup of gaseous products. The average density and average gas holdup ratio of floated granular sludge were 913 kg m3 and 11.8% (by volume), respectively. The floatation of granular sludge could disappear by releasing gas when sludge was in the state of elastic expansion, but it would become worse by holding gas when it entered the plastic expansion state. The plastic expansion of granules was significantly correlated with the less content of extracellular polymeric substances. Ó 2013 Elsevier Ltd. All rights reserved.

1. Introduction Nitrogen and carbon-containing compounds are the major contaminants in wastewaters which cause serious environmental problems (Guo, 2007; Smith et al., 1999; Lettinga et al.,1980). Simultaneous elimination of nitrogen and carbon by biological denitrification (Cuervo-Lopez et al., 1999; Lew et al., 2012; Nancharaiah and Venugopalan, 2011) has become increasingly important in wastewater treatment (Gupta and Gupta, 2001; Reyes-Avila et al., 2004). Denitrifying reactor is a promising technology for denitrification and development of high-rate denitrifying reactor can improve denitrification technology. The denitrifying granular sludge reactor was reported in 1975 for the first time (Miyaji and Kato, 1975) and after that it became a focus of wastewater treatment due to its good performance and low cost. So far, the reported maximum nitrogen removal rate (NRR) and ⇑ Corresponding author. Tel./fax: +86 57188982819. E-mail address: [email protected] (P. Zheng). 0960-8524/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.biortech.2013.11.056

COD removal rate (ORR) in literature are 25 kgN m3 d1 (Bode et al., 1987) and 67.5 kg COD m3 d1 (Franco et al., 2006), respectively. However, floatation of denitrifying activated sludge has been reported both in literatures and engineering applications (Cuervo-Lopez et al., 1999; Liu and Sun, 2011). And the floating sludge is more likely to be washed out from the reactor resulting in the deterioration of denitrification process, especially in highrate reactor. Some researchers have reported that addition of enough Ca2+ to influent was an effective way to control the floatation of sludge (Jin et al., 2012b; Liu and Sun, 2011). However, this can cause sludge calcification and sludge activity decline in the end (Chen et al., 2010a). So far, most of the reported loads for denitrifying reactor are lower than 15 kgN m3 d1 (Franco et al., 2006; Isaka et al., 2012; Rabah and Dahab, 2004) and the key factors causing floatation of denitrifying granular sludge remain unclear. Until now, rare literature is available on the mechanism of floating denitrifying granular sludge and its effect on on the performance of reactor at nitrogen loading rate (NLR) greater than 25 kg m3 d1.

W. Li et al. / Bioresource Technology 152 (2014) 414–419

415

Nomenclature EPS GPR NLR NRR OLR ORR PN PS QG Qin R SR

extracellular polymeric substances gas production rate nitrogen loading rate nitrogen removal rate COD loading rate COD removal rate extracellular proteins extracellular polysaccharide gas flow rate influent flow rate recycling ratio cross-sectional area of reactor

SDA SGPR Uup VLR eB eg

ql qg qgc qs qG x0

sludge denitrifying activity specific gas production rate two-phase upflow velocity volumetric loading rate bed porosity granular porosity liquid density, (1000 kg m3) density of granular sludge critical density for settling solid density of granular sludge density of gas production critical gas holdup for floating

Recently, a high-rate denitrifying automatic circulation (DAC) reactor has been successfully developed in laboratory. However, the performance of the DAC reactor decline at high NLR caused by the floatation of granular sludge. Taking DAC reactor and granular sludge as models, the mechanism of floating granular sludge has been investigated in order to carry forward the development and application of high-rate granular sludge bed reactor.

diagram of the reactor system is shown in Fig. 1. The lower part of DAC reactor is made of a 0.45 m high plexiglass-made tube with an inner diameter of 0.06 m and effective volume of 1.25 L. The top of DAC reactor was closed with a gas-collecting hood. The reactor was fed by a peristaltic pump (Lead Fluid, China).

2. Methods

The DAC reactor was initially inoculated with 1 L anaerobic granular sludge with VSS/SS content of about 70%. The reactor was set at NLR of 2 kg m3 d1 with a fixed effluent recycling ratio (recycling flow to inflow ratio) of 2.0. Both the synthetic wastewater and the recycling liquid were mixed in the manifold at the bottom of the reactor. The loading rate was increased by shortening hydraulic retention time (HRT) (Tang et al., 2011). Data was recorded after NLR reached up to 25 kg m3 d1. The temperature was set at 30 ± 1 °C.

2.1. Synthetic wastewater The concentrations of sodium nitrate and methanol were 1 g NO3–N L1 and 5 g COD L1, respectively. Nitrogen to methanol ratio was 1:3.33 to keep the nitrogen as the limiting substrate. The other constituents of the mineral medium was (g L1): KH2PO3 0.05, CaCl2 0.4, MgSO47H2O 0.1 and 1 ml L1 of trace elements solution. The trace elements solution contained (g L1): 5 EDTA, 5 MnCl24H2O, 3 FeSO47H2O, 0.05 CoCl.6H2O, 0.04 NiCl26H2O, 0.02 H3BO3, 0.02 (NH4)6Mo7O24H2O, 0.01 CuSO45H2O and 0.003 ZnSO4. The pH of synthetic wastewater was in the range of 6.6–6.9 (Li et al., 2013). 2.2. Laboratory-scale DAC reactor The experimental work was carried out in a plexiglass-made denitrifying automatic circulation (DAC) reactor. The schematic

2.3. Reactor operation

2.4. Analytical methods The determination of pH, nitrate, nitrite, ammonium, suspended solids (SS) and volatile suspended solids (VSS) concentrations were carried out according to the Standard Methods (APHA, 2005). 2.4.1. Granular morphology The digital macro photography was performed by SteREO Discovery stereomicroscopes (Carl Zeiss, Germany) or Digital singlelens reflex camera (Nikon, Japan). The digital macro photos were analyzed by Image Pro Plus 6.0 (Media Cybernetics, USA). The size of granular sludge was determined by QICPIC system (Sympatec, Germany). 2.4.2. Granular density (qg) Sucrose was used to make a series of solutions with densities of 1400, 1390, 1380. . .1020, 1010 g L1. Ethanol was used to make a series of solutions with densities of 990, 980, 970. . .920 gL1, 910 g L1. The granular samples were added into each of the 50 ml graduated cylinders having sucrose or ethanol solutions of different densities (Lu et al., 2012). Under a quiescent condition, the granules moved up or down in the graduated cylinder depending on the density of the solution. Thus, the specific wet gravity of the granule was measured.

Fig. 1. DAC reactor system. (1, 2) Peristaltic pump, (3) DAC reactor, (4) Wet gas meter.

2.4.3. Granular porosity (eg) The granular porosity was determined by size exclusion chromatography (Adav et al., 2008; Zheng and Yu, 2007). The experiment was performed at 4 °C to inhibit the sludge activity without affecting the granule characteristics (Alphenaar et al.,

416

W. Li et al. / Bioresource Technology 152 (2014) 414–419

1992). A glass column with a diameter of 2.0 cm and a height of 40 cm was packed with the granule samples (Zheng and Yu, 2007). 1.0 ml acetate solution with a concentration of 25 mg ml1 was used for all the tests as a tracer. The eluant was pumped with a peristaltic pump (Lead Fluid, China) at a constant rate of 0.5 ml min1. The eluate was continuously collected and the concentration of acetate was determined by gas chromatography (Agilent 6890, United States). Then, granular porosity was calculated according to the average residence time of the acetate as solutes. 2.4.4. Extracellular polymeric substances (EPS) EPS were extracted from the granular sludge by EDTA (Sheng et al., 2005), then the extracellular proteins (PN) were determined by the Lowry method taking egg albumin as standard and the polysaccharide (PS) content was analyzed by the anthrone method taking glucose as standard (Wu et al., 2009). 2.4.5. Bed porosity (eB) Bed porosity (eB) was determined using the method developed by Jimenez et al. (1988). At first, residence time of the tracer (sodium fluoride) in the reactor was determined. NaF was selected as a tracer because it was non-reactive and not absorbed by granular sludge (Asraf-Snir and Gitis, 2011). The measurement of fluoride ion concentration was carried out through fluoride ion-selective electrode (APHA, 2005). Then, the bed porosity was calculated. 2.5. Calculation of critical density (qgc) As the NLR exceeds the 25 kg m3 d1, the gas production rate (GPR) is too high to be elided (Tang et al., 2009). The two-phase upflow velocity (Uup) was used to characterize the average velocity of gas and liquid flow. Assuming the ratio of project area of gas, liquid and sludge to the cross section of reactor as a, b, c, respectively, they satisfy Eq. (1).

aþbþc¼1

ð1Þ

up to 25 kg m3 d1, the DAC reactor was kept in operation with increasing NLRs by shortening HRT and fixing the influent substrate concentration. The volumetric capacity of the DAC reactor at high-load is depicted in Fig. 2A and B. The operation period can be divided into two phases, namely stable phase and decline phase. In the stable phase (P1), the higher the volumetric loading rate (VLR), the larger the volumetric removal rate and gas production rate (GPR). On 59th day, the NLR reached up to 50.76 kg m3 d1 and the NRR and the GPR increased to 49.49 and 42.24 L L1 d1, respectively. The observed NRR was far more than the reported top value (25 kg N m3 d1) (Bode et al., 1987). At the same time, the COD loading rate (OLR) reached up to 273.24 kg m3 d1, and the COD removal rate (ORR) of 187.68 kg m3 d1 was also far more than the reported top value (67.5 kg m3 d1) (Franco et al., 2006). In the decline phase (P2), the volumetric removal rate and the GPR decreased gradually, although VLR was constant with the NLR of 55 kg m3 d1 and the OLR of 275 kg m3 d1. The NLR, the ORR and the GPR decreased from 50.80 kg m3 d1, 198.72 kg m3 d1 and 43.20 L L1 d1 (64th day) to 42.34 kg m3 d1, 142.60 kg m3 d1 and 17.28 L L1 d1 (86th day), respectively.

3.2. Biomass flotation and wash-out In the DAC reactor, the denitrifying microorganisms existed in the form of granular sludge so the performance of the reactor is closely related to the amount and activity of granular sludge (Chen et al., 2010b). Fig. 3A shows the profile of biomass concentration in the reactor and in the effluent at different NLRs. In P1 (the NLR increased from 25 to 45 kg m3 d1), the effluent biomass concentrations increased slightly from 0.009 ± 0.001 to 0.027 ± 0.005 g VSS L1, while the biomass concentrations in the reactor significantly increased from 60.54 ± 0.01 to 82.23 ± 0.01 g VSS L1. In P2, the effluent biomass concentration sharply increased to 0.133 ± 0.09 g VSS L1, and the biomass concentration in reactor

The Uup can be calculated according to Eq. (2).

U up ¼ ¼

QG a Q ð1 þ RÞ a Q þ Q in ð1 þ RÞ  þ in  ¼ G aþb SR b SR ða þ bÞ SR a a þ b Q G þ Q in ð1 þ RÞ SR ð1  cÞ

ð2Þ

where QG is the gas flow rate, L d1; Qin is the influent flow rate, L d1; R is the recycling ratio; SR is the cross-sectional area of reactor, m2. c can be calculated by Eq. (3).

c¼

total volume of granular sludge ¼ 1  eB total volume of packed bed

ð3Þ

where eB is the bed porosity (m3 of voids/m3 of bed).According to d2 ðqg ql Þg Stocks equation U ¼ 18l (Liu et al., 2005), the critical density of the granular sludge can be calculated by Eq. (4).

qgc ¼

18lU up 2

gd

þ ql

ð4Þ

where l is liquid viscosity, Pa s (0.0008); ql is liquid density, kg m3 (1000); d is the average diameter of the granular sludge, m; g is gravitational acceleration, m s2 (9.8). 3. Results and discussion 3.1. Performance of the high-rate DAC reactor Taking the reported top NLR (25 kg m3 d1) as a standard, the NLR was divided into two groups: normal-load (NLR < 25 kg m3 d1) and high-load (NLR P 25 kg m3 d1). After the NLR reached

Fig. 2. Performance (A, B) of the high-rate DAC reactor.

W. Li et al. / Bioresource Technology 152 (2014) 414–419

417

Fig. 4. Variations of granular sludge density qg and critical granular sludge density qgc under different NLRs. (Relative density difference = qg  qgc.

tor leading to deterioration in settling performance of denitrifying granular sludge was the sharp decrease of qg.

Fig. 3. Variation of biomass concentration in the reactor and the effluent under different NLRs.

sharply decreased to 15.64 ± 0.01 g VSS L1. As shown in Fig. 3A, the sludge denitrifying activity (SDA) of 2.83 ± 0.19 g N g1 VSS d1 in P2 was significantly higher than the counterpart in P1. But the biomass washout rate of 7.30 ± 0.52 g VSS L1 d1 in P2 was higher than the growth rate of 4.32 ± 0.70 g VSS L1 d1 (Fig. 3B) resulting in the negative growth of biomass in the reactor. These results suggest that the direct factor leading to performance decline at high VLR is the less biomass concentration caused by the negative growth of the biomass in the DAC reactor. So the sludge concentration of 15.64 ± 0.01 g VSS L1 in the DAC reactor, can only support NLR of 41.26–47.26 kg m3 d1 and cannot support NLR of 55 kg m3 d1, which needs at least 18.21–20.83 g VSS L1, even with the highest SDA of 2.83 ± 0.19 g N g1VSS d1.

3.3. The mechanism of granular sludge flotation 3.3.1. Relative density difference of granular sludge According to Stocks equation, the critical density of granular sludge (qgc) for settlement can be calculated (Eq. (4)) using Uup as the critical setting velocity. Then the relative density difference can be obtained based on the density of granular sludge (qgc) (Fig. 4). Afterwards, the settling potential of granular sludge can be estimated by the relative density difference on specific conditions: the larger the relative density difference, the greater the settling performance of the granular sludge. In P1, the qg increased gradually from 1227 ± 31 kg m3 (25 kg m3 d1) to 1321 ± 21 kg m3 (45 kg m3 d1) with the increase of NLR, while the qgc changed less than 1 kg m3 resulting in an increase of the relative density difference from 226.0 to 320.2 kg m3. The results indicated that the settling performance of denitrifying granular sludge was becoming better in P1. However, the qg sharply decreased to 1031 ± 15 kg m3, and the qgc almost remained constant, resulting in decrease of the relative density difference to 30.6 kg m3 in P2. In summary, the direct fac-

3.3.2. Gas holdup of granular sludge In P2, some of the denitrifying granular sludge with larger diameter of 4.93 mm and less density of 913 ± 0.042 kg m3 floated (Fig. 5A). Stereoscopic observation (Carl Zeiss, Germany) of the floating granules demonstrated that there were a large amount of voids inside the granular sludge with a lot of gases (Fig. 5B). Interestingly, a slight press with a small spoon made all the floating granular sludge (Fig. 5C) sink (Fig. 5D) and increased the density to 1.025 ± 0.004 kg m3. This phenomenon demonstrated that the floating granular sludge was in the expansion state due to inner gas pressure, and it could return to the equilibrium state and increase its density by releasing gases. In other words, the floating granular sludge could recover its settleability by releasing gases when it was in the state of elastic expansion. Critical gas holdup (x0) was defined as the critical gas percentage of granular sludge for floating at specific VLR. The value of x0 is relevant to the densities and the volume distributions of gas, liquid and solid in granular sludge. Since the liquid density is equal to the solution density, x0 mostly depends on the density difference between the gas phase and solid phase in granule: the larger the density difference, the higher the x0 and the better the settling performance. Comparison of the x0 values at different NLRs (Table 1) indicates that: the x0 of 3.1% at the NLR of 55 kg m3 d1 (in P2) is significantly lower than the critical x0 of 22.6% in P1 at the NLR of 25 kg m3 d1. But the specific gas production rate (SGPR)

Fig. 5. Flotation and settlement of denitrifying granular sludge.

418

W. Li et al. / Bioresource Technology 152 (2014) 414–419

Table 1 Physical characteristics of high-load denitrifying granular sludge at NLRs. NLR (kg m3 d1) SGPR(L g1 VSS d1) Granular porosity (eg) (%) Solid density (qs)a (kg m3) Critical gas holdup (x0)b (%) a b

25 0.47 73.5 1857 22.6

35 0.48 77.1 2166 25.5

45 0.48 80.7 2663 32.1

55 1.23 88.5 1270 3.1

Solid density of granular sludge (qs):qs = (qg  eg  ql)/(1  eg). Critical gas holdup (x0) related equation: qgc = qs(1  eg) + ql(eg  x0) + qGx0.

increased from 0.47 to 1.23 L g1VSS d1 (Table 1) at the same time. This result indicated that the granular sludge was more easily to float in P2. In denitrification processes, the reactants (nitrate and methanol) were dissolved in liquid phase, while the products (nitrogen gas and carbon dioxide) were in gas phase. The gaseous products will form bubbles once their concentrations exceed the saturated values. As the denitrification goes on, the bubbles will create a gas pressure inside the granular sludge, which can lead to bubbles movement towards outside and gas release (Zhang et al., 2012). During this process, if the gas holdup ratio exceeds x0, the granular sludge floats. At the NLR of 55 kg m3 d1, both the SDA and the SGPR were as high as 2.83 ± 0.19 g N g1 VSS d1 and 1.23 L g1 VSS d1, respectively. Therefore, the granular sludge at this NLR would float easily for its internal gas holdup ratio exceeding the x0 (only 3.1%). In the DAC reactor, there is a diaphragm covering the upper reaction zone with two pipes (Fig. 1). At high VLRs, the compression produced by the sharp reduction of cross-sectional area (Barnes et al., 1989) (the cross section ratio is more than 36) at the top of reaction zone can promote extrusion of internal gases which helps the floating granular sludge sink to some degree. Nevertheless, because of the low value of x0 (3.1%), some amount of granular sludge still floated. Our experimental results showed that the ratios of gas holdup inside floating granular sludge ranged from 3.1% to 22.6%, with an average value of 11.8%, and elastic expansion ones can decrease the gas holdup ratio to 0.61% (less than x0 of 3.1%) by compression and then settle (Fig. 5). These results also suggest that DAC reactor can recover the settleability of floating granular sludge with x0 more than 11.8% (see Fig. 6). 3.3.3. Solid density of granular sludge The calculations (Table 1) show that large decrease in solid density of granular sludge plays a key role in the reduction of x0. Determination of the VSS/SS, EPS and PS/PN at the NLR of 55 and 45 kg m3 d1 showed that the VSS/SS at the NLR of 55 kg m3 d1 (in P2) was 0.56 more than the counterpart of 0.46 at the NLR of 45 kg m3 d1 (in P1). While the EPS of 9.10 ± 1.98 mg g1 VSS and the PS/PN of 0.72 ± 0.01 were less than the counterpart values of 23.77 ± 4.92 mg g1 VSS and 3.21 ± 0.02, respectively. These suggest that at high VLR, organic components in granular sludge increased but the EPS content decreased. Generally speaking, the density of organics is less than the density of inorganic matter, so, the more the organic contents, the less the solid density of granular sludge. In addition, the organic components mainly consisted of EPS and functional bacteria in the granular sludge. The EPS, especially the PS, were assumed to be binder in the sludge (Liu et al., 2004; Puñal et al., 2000; De Beer et al., 1996). Therefore, decrease of both the EPS content and the PS/PN weakened the viscous forces in solid of the granular sludge making it loose. The loose granular sludge with less EPS content was more likely to expand plastically under high internal gas pressure and hydrostatic pressure in high-rate reactor. Naturally, loose granular sludge was low in density and floated easily at high VLR. Moreover, Puñal et al. (2000) reported that the condition of COD/N more than 300/4 was not conducive to the synthesis of EPS for anaerobic granular sludge. In our experiment, we found that the COD/N has been less

Fig. 6. Mechanism promoting gas release from denitrifying granular sludge.

than 200/4 in reaction solution at the NLR of 55 kg m3 d1 with relatively low content of EPS. These results corroborated the findings of Puñal. 4. Conclusions In high-rate DAC reactor, floatation of granular sludge could lead to a serious biomass washout and a sharp deterioration of reactor performance. The sludge floatation was ascribed to the decrease in density, which originated from the holdup of gaseous products. The average density and gas holdup ratio of floated granular sludge were 913 kg m3 and 11.8% (by volume), respectively. The floating granular sludge could recover its settleability by releasing gases during elastic expansion, but it would lose its settleability when it entered plastic expansion state. The plastic expansion of granular sludge was significantly correlated with the less content of EPS. Acknowledgements This research was supported by the Natural Science Foundation of China (31070110), the Natural Science Foundation of China (51278457), the Specialized Research Fund for the Doctoral Program of Higher Education of China (20110101110078) and the Natural Science Foundation of Zhejiang Province (Z5110094). References Adav, S.S., Chang, C.H., Lee, D.J., 2008. Hydraulic characteristics of aerobic granules using size exclusion chromatography. Biotechnol. Bioeng. 99 (4), 791–799. Alphenaar, P.A., Perez, M.C., Berkel, W.J.H., Lettinga, G., 1992. Determination of the permeability and porosity of anaerobic sludge granules by size exclusion chromatography. Appl. Microbiol. Biotechnol. 36 (6), 795–799. APHA, 2005. Standard Methods for the Examination of Water and Wastewater, 21st ed. American Public Health Association, Washington, DC, USA. Asraf-Snir, M., Gitis, V., 2011. Tracer studies with fluorescent-dyed microorganisms– a new method for determination of residence time in chlorination reactors. Chem. Eng. J. 166 (2), 579–585. Barnes, H.A., Hutton, J.F., Walters, K., 1989. An Introduction to Rheology. Elsevier, 66–68. Bode, H., Seyfried, C., Kraft, A., 1987. High-rate denitrification of concentrated nitrate wastewater. Water Sci. Technol. 19 (1–2), 163–174. Chen, J., Ji, Q., Zheng, P., Chen, T., Wang, C., Mahmood, Q., 2010a. Floatation and control of granular sludge in a high-rate anammox reactor. Water Res. 44 (11), 3321–3328. Chen, J., Zheng, P., Yu, Y., Tang, C., Mahmood, Q., 2010b. Promoting sludge quantity and activity results in high loading rates in Anammox UBF. Bioresour. Technol. 101 (8), 2700–2705. Cuervo-Lopez, F.M., Martinez, F., Gutierrez-Rojas, M., Noyola, R.A., Gomez, J., 1999. Effect of nitrogen loading rate and carbon source on denitrification and sludge settleability in upflow anaerobic sludge blanket (UASB) reactors. Water Sci. Technol. 40 (8), 123–130. De Beer, D., O’Flaharty, V., Thaveesri, J., Lens, P., Verstraete, W., 1996. Distribution of extracellular polysaccharides and flotation of anaerobic sludge. Appl. Microbiol. Biotechnol. 46 (2), 197–201. Franco, A., Roca, E., Lema, J.M., 2006. Granulation in high-load denitrifying upflow sludge bed (USB) pulsed reactors. Water Res. 40 (5), 871–880. Guo, L., 2007. Doing battle with the green monster of Taihu Lake. Science 317 (5842), 1166.

W. Li et al. / Bioresource Technology 152 (2014) 414–419 Gupta, A.B., Gupta, S.K., 2001. Simultaneous carbon and nitrogen removal from high strength domestic wastewater in an aerobic RBC biofilm. Water Res. 35 (7), 1714–1722. Isaka, K., Kimura, Y., Osaka, T., Tsuneda, S., 2012. High-rate denitrification using polyethylene glycol gel carriers entrapping heterotrophic denitrifying bacteria. Water Res. 46 (16), 4941–4948. Jimenez, B., Noyola, A., Capdeville, B., Roustan, M., Faup, G., 1988. Dextran blue colorant as a reliable tracer in submerged filter. Water Res. 22 (10), 1253–1257. Jin, X.B., Wang, F., Liu, G.H., Yan, N., 2012. A key cultivation technology for denitrifying granular sludge. Process Biochem. 47 (7), 1122–1128. Lettinga, G., van Velsen, A.F.M., Hobma, S.W., de Zeeuw, W., Klapwijk, A., 1980. Use of the upflow sludge blanket (USB) reactor concept for biological wastewater treatment, especially for anaerobic treatment. Biotechnol. Bioeng. 22 (4), 699– 734. Lew, B., Stief, P., Beliavski, M., Ashkenazi, A., Svitlica, O., Khan, A., Tarre, S., de Beer, D., Green, M., 2012. Characterization of denitrifying granular sludge with and without the addition of external carbon source. Bioresour. Technol. 124, 413– 420. Li, W., Zheng, P., Wang, L., Zhang, M., Lu, H., Xing, Y., Zhang, J., Wang, R., Song, J., Ghulam, A., 2013. Physical characteristics and formation mechanism of denitrifying granular sludge in high-load reactor. Bioresour. Technol. 142, 683–687. Liu, Y.J., Sun, D.D., 2011. Calcium augmentation for enhanced denitrifying granulation in sequencing batch reactors. Process Biochem. 46 (4), 987–992. Liu, Y.Q., Liu, Y., Tay, J.H., 2004. The effects of extracellular polymeric substances on the formation and stability of biogranules. Appl. Microbiol. Biotechnol. 65 (2), 143–148. Liu, Y., Wang, Z.W., Liu, Y.Q., Qin, L., Tay, J.H., 2005. A generalized model for settling velocity of aerobic granular sludge. Biotechnol. Progr. 21 (2), 621–626. Lu, H.F., Zheng, P., Ji, Q.X., Zhang, H.T., Ji, J.Y., Wang, L., Ding, S., Chen, T.T., Zhang, J.Q., Tang, C.J., Chen, J.W., 2012. The structure, density and settlability of anammox granular sludge in high-rate reactors. Bioresour. Technol. 123, 312–317.

419

Miyaji, Y., Kato, K., 1975. Biological treatment of industrial wastes water by using nitrate as an oxygen source. Water Res. 9 (1), 95–101. Nancharaiah, Y.V., Venugopalan, V.P., 2011. Denitrification of synthetic concentrated nitrate wastes by aerobic granular sludge under anoxic conditions. Chemosphere 85 (4), 683–688. Puñal, A., Trevisan, M., Rozzi, A., Lema, J.M., 2000. Influence of C:N ratio on the startup of up-flow anaerobic filter reactors. Water Res. 34 (9), 2614–2619. Rabah, F.K., Dahab, M.F., 2004. Nitrate removal characteristics of high performance fluidized-bed biofilm reactors. Water Res. 38 (17), 3719–3728. Reyes-Avila, J., Razo-Flores, E.A., Gomez, J., 2004. Simultaneous biological removal of nitrogen, carbon and sulfur by denitrification. Water Res. 38 (14–15), 3313– 3321. Sheng, G.P., Yu, H.Q., Yu, Z., 2005. Extraction of extracellular polymeric substances from the photosynthetic bacterium Rhodopseudomonas acidophila. Appl. Microbiol. Biotechnol. 67 (1), 125–130. Smith, V.H., Tilman, G.D., Nekola, J.C., 1999. Eutrophication: impacts of excess nutrient inputs on freshwater, marine, and terrestrial ecosystems. Environ. Pollut. 100 (1), 179–196. Tang, C.J., Zheng, P., Mahmood, Q., 2009. The shear force amendments on the slugging behavior of upflow Anammox granular sludge bed reactor. Sep. Purif. Technol. 69 (3), 262–268. Tang, C.J., Zheng, P., Wang, C.H., Mahmood, Q., Zhang, J.Q., Chen, X.G., Zhang, L., Chen, J.W., 2011. Performance of high-loaded ANAMMOX UASB reactors containing granular sludge. Water Res. 45 (1), 135–144. Wu, J., Zhou, H.M., Li, H.Z., Zhang, P.C., Jiang, J., 2009. Impacts of hydrodynamic shear force on nucleation of flocculent sludge in anaerobic reactor. Water Res. 43 (12), 3029–3036. Zhang, J.B., Wu, J., Poncin, S., Hamelin, M., Li, H.Z., 2012. Microscale investigation of anaerobic biogas production under various hydrodynamic conditions. Environ. Sci. Technol. 46 (16), 8698–8704. Zheng, Y.M., Yu, H.Q., 2007. Determination of the pore size distribution and porosity of aerobic granules using size-exclusion chromatography. Water Res. 41 (1), 39–46.