w a t e r r e s e a r c h 5 0 ( 2 0 1 4 ) 1 7 1 e1 7 8

Available online at www.sciencedirect.com

ScienceDirect journal homepage: www.elsevier.com/locate/watres

A rheological approach to analyze aerobic granular sludge Yun-Jie Ma a, Cheng-Wang Xia a, Hai-Yang Yang b, Raymond J. Zeng a,* a

Department of Chemistry, University of Science and Technology of China, Hefei, Anhui 230026, People’s Republic of China b Department of Polymer Science and Engineering, University of Science and Technology of China, Hefei, Anhui 230026, People’s Republic of China

article info

abstract

Article history:

Aerobic granular sludge is one promising biotechnology in wastewater treatment. Despite

Received 20 August 2013

intensive researches on granular architecture and strategies to improve treatment effi-

Received in revised form

ciency, there are still some elusive material parameters needed to stimulate the granula-

28 November 2013

tion process. The main aim of this study was to evaluate aerobic granular sludge

Accepted 30 November 2013

innovatively using the universal rheology methodology, in terms of processability or

Available online 13 December 2013

quality and texture. Steady shear and oscillatory measurements were performed. Basic rheological characterization showed that aerobic granular sludge was a shear-thinning

Keywords:

HerscheleBulkley fluid with yield pseudoplasticity. Meanwhile, granular sludge pre-

In-situ rheological characterization

sented characterized viscoelastic behaviors in dynamic sweeps highlighting its superiority

Steady shear flow

to flocculent sludge. Furthermore, a Wagner-type constitutive model incorporating a

Viscoelastic properties

relaxation and damping function was introduced and able to describe the time-dependent

Non-linear viscoelasticity

and non-linear viscoelastic behaviors. This study could make a further step on predicting

Wagner-type model

rheological properties, helping improve the actual sludge treatment process and the operation of sludge dewatering. ª 2013 Elsevier Ltd. All rights reserved.

1.

Introduction

Compared with conventional activated sludge system, granular sludge exhibits a more robust aggregate structure, better solideliquid separation, higher biomass concentration and greater endurance to withstand shock loadings (Liu et al., 2009). Based on these advantages mentioned, granular sludge has been regarded as a promising technology in wastewater treatment. With decades of researches, predecessors have made plentiful micro- and macro-scale observations of the aerobic granular sludge, in terms of both

structure and function, including settleability, morphology, permeability, porosity, thermodynamics, mechanical stability, surface hydrophobicity, and its practical applications in removal of phenol or heavy metals (Liu and Tay, 2004; Liao et al., 2001). The knowledge about its physicochemical characteristics and technological parameters has steadily improved to stimulate this technology. Nevertheless, there still exist some unanswered puzzles in the granulation process, for example, it is hard to give a precise qualitative or quantitative delimitation when granular sludge system comes to deterioration (Adav et al., 2008). To get a better understanding about granulation and achieve higher treatment

* Corresponding author. Tel.: þ86 551 63600203; fax: þ86 551 63601592. E-mail address: [email protected] (R.J. Zeng). 0043-1354/$ e see front matter ª 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.watres.2013.11.049

172

w a t e r r e s e a r c h 5 0 ( 2 0 1 4 ) 1 7 1 e1 7 8

efficiency, a deeper exploration of mechanism and a clearer definition of technological parameters are needed urgently. Rheology is a subject describing the deformation of a body under influence of mechanical stress, and is a valuable tool for characterizing non-Newtonian and viscoelastic properties of the materials, e.g. sludge suspension (Chen et al., 2010; Wloka et al., 2004). As not only can it reflect the internal architectural features (Lin et al., 2013), but also can quantify flow behavior in actual process or sludge cake behavior during mechanical expression (Chang and Lee, 1998; Sozanski et al., 1997). These research fields about rheology applied in sludge system have already been proved successfully. Baudez et al. (2012) discussed these similarities of the viscoelastic behaviors of raw and anaerobic digested sludge with soft-glassy materials, and differentiated these two kinds of sludge in micro-rheology to a certain extent. To describe rheological parameters in sludge flow, Eshtiaghi et al. (2012) tried to use certain model fluids to emulate thickened digested sludge. And Stoodley et al. (1999) gave an in-situ investigation of bacterial bio-film rheology reflecting structural deformation caused by short-term fluctuations in fluid shear and concluded that bio-film viscoelasticity could increase fluid energy losses in pipelines. In brief, predecessors have given a fundamental inter-linkage between rheology and flocculent sludge, and that innovation makes it feasible to apply rheological analysis in a granular sludge system, as similarly implied by Mu and Yu (2006). However, the understanding of rheological dynamics properties of granular sludge is rather limited because of fragmented studies on the whole. Most subtle researches have been concentrated on the study of extracellular polymeric substance (EPS) or its individual components (Ying et al., 2010; Seviour et al., 2009a; Lin et al., 2010). To our knowledge, the insitu rheological characterization of aerobic granular sludge has barely been reported as a complete system, although important parameters, such as limiting viscosity and intrinsic viscoelasticity, play big roles in differentiating liquid-like from solid-like regime in granules (Seviour et al., 2009b). What’s more, it is a challenge to match the viscoelastic response with steady-state flow behavior in sludge flow system, and limited attempts have been made. Mackley et al. (1994) extended an integral form Wagner equation originally developed for the description of polymer melts to test its applicability on an associative thickener, a kaolinite slurry, an oil-based paint and so on. Also Liang and Mackley (1994) adopted the same model, incorporating with a relaxation and damping function, to perform non-linear viscoelasticity and predict stress growth and steady-shear behaviors for polyisobutylene solutions. So Wagner-type model seems to be a promising solution. In this paper, the aim was to elaborate the fundamental understanding of intrinsic rheological properties of aerobic granular sludge by performing basic rheological characterization and viscoelasticity measurement. Firstly, in steady shear measurement, shear deformation under the simple shearing flow was measured and shear stress was recorded to obtain its basic flow behaviors. Secondly, in oscillatory measurement, strain, frequency and temperature sweeps were conducted to obtain its viscoelastical response to external forces, and its granular configuration was highlighted in comparison with flocculent sludge reported by Baudez et al.

(2012). Finally, to match the time-dependent viscoelasticity with the steady-state flow, a Wagner-type model, was established to predict the sludge flow preliminarily.

2.

Materials and methods

2.1.

Aerobic granules for investigation

The aerobic granules were sampled from a lab-scale granular enhanced biological phosphorus removal reactor treating synthetic wastewater with COD: N: P as 22: 1.6: 1. The reactor had a working volume of 8 L, operated with a cycle time of 6 h, and temperature was kept between 22 and 25  C. Each cycle was consisted of a 130 min-anaerobic stage, a 190 min-aerobic stage, a 30 min-settling period and a 10 min-decant period. The suspended solid concentration (SS) was about 4 g/L, and the reactor was operated for over 3 months staying at a pseudo stable level. SS was measured following the Standard Methods (APHA, 1998). Granules for rheological tests were collected at the end of aerobic period. Before tests, granules were gently concentrated to every objective concentration, and stored at 4  C overnight in order to reduce temporal variability.

2.2.

Extraction of extracellular polymeric substances

Granules with EPS extracted were obtained using a heat extraction method (Morgan et al., 1990). The sludge suspension was first dewatered by centrifugation in a 50-mL tube at 6000 rpm for 5 min. The pellet in the tube was re-suspended into 0.1 M NaCl solution and washed twice. Then sludge was diluted with NaCl solution to its original volume, and was heated to 60  C in a water bath for 20 min. The supernatant collected after a centrifugation was regarded as EPS extraction, and the retained granules were used for following experiments to give a qualitative comparison.

2.3.

Rheological tests

Rheological tests were carried out using a rotational ARES-G2 rheometer, a coaxial cylindrical measurement device, connected to a temperature controlled water bath. The rheometer was equipped with a cup and bob geometry. A 20-mL aliquot of granule sample in a certain concentration was poured into the cup and the bob was lowered until the space between bob and cup was fully wetted. The temperature was kept at 25  C through water bath. Before each rheological measurement was performed, granular sludge was pre-sheared for 5 min at a shear rate of 500 s1, which was to erase material memory and have reproducible results. Then granular sludge was left at rest for 10 min for different tests performed: In the steady shear mode, shear stress (s) was measured as _ increased from 1 to 800 s1 and then vice versa. shear rate (g) The temperature was kept at 25  C. In the oscillatory shear mode, granular sludge was subjected to small and reversible periodic oscillations of frequency (u) with shear strain (g) and shear stress (s) recorded. Strain, frequency and temperature sweeps followed the same

w a t e r r e s e a r c h 5 0 ( 2 0 1 4 ) 1 7 1 e1 7 8

173

rheology methodology with many other systems (Baudez et al., 2012). Storage modulus (G0 ) and loss modulus (G00 ), namely the real and imaginary parts of dynamic complex modulus, are determined according to G0 ¼

s00 g

(1)

G00 ¼

s0 g

(2)

tan d ¼

s00 s0

(3)

where s0 , s00 , d are in-phase stress, out-of-phase stress and the phase shift, respectively.

2.4.

Fig. 1 e Typical rheogram of aerobic granular sludge at SS of 35 g/L and 25  C.

Wagner-type model

The Wagner model, a generalized, multiple Maxwell, linear viscoelastic model, derives from the more general KBKZ model (Larson, 1988). Considering only the shear stress component s (t) and time-strain reparability gðt; t0 Þ, a form of the equation is given Zt

dGðt  t0 Þ hðgÞgðt; t0 Þdt dt0

sðtÞ ¼  N

(4)

where Gðt  t0 Þ is the linear relaxation modulus and h (g) is a damping function. According to Wagner (1976) hðgÞ ¼

Gðg; t  t0 Þ 0 ¼ ekjgðt;t Þj Gðt  t0 Þ

(5)

where Gðg; t  t0 Þ is the non-linear relaxation modulus, and k, the damping factor, is an empirical parameter which quantifies the level of non-linearity. The relaxation modulus can be described in terms of a discrete set of Maxwell elements as Bower et al. (1999) Gðg0 ; tÞ ¼

X

gi et=li

(6)

i

where (gi, li) are discrete spectra of modulus and relaxation _ in steady time, respectively. And the apparent viscosity hðgÞ shear can be calculated with Eq. (7) as reported by Liang and Mackley (1994) _ ¼ hðgÞ

3.

X

gi li

i

_ ð1 þ kli gÞ

2

(7)

Results

3.1. The basic rheological characterization of aerobic granules in the steady shear flow Fig.1 shows a typical rheogram of aerobic granular sludge in the steady-shear flow at SS of 35 g/L and room temperature. _ increased from Shear stress (s) was measured as shear rate (g) 1 to 800 s1, and then vice versa. Compared to descending shear rate path, stress s on ascending path was higher. It showed a high hysteresis surface, dependent on the data

sampling (Baudez, 2006). The apparent viscosity (happ) decreased rapidly with increasing shear rate when g_ was smaller than 100 s1, and kept relatively unchanged at about 1 Pa s for higher rates ðg_ > 200 s1 Þ. To describe the basic non-Newtonian rheology of aerobic granular sludge, Herschel-Bulkley equation, with an additional yield stress (sy), was adopted. s ¼ sy þ K$g_ n

(8)

where K is the fluid consistency index, and n is the flow behavior index. Similar pattern was also found in the SS of 10, 15, 20 g/L (data not shown). Applied to Eq. (8), it gave n value between 0.36 and 0.53, all smaller than 1, and showed that sy inclined to increase as SS went up, equal to 0.14 Pa, 0.35 Pa, 0.65 Pa, 23.5 Pa, respectively. It can be concluded that aerobic granular sludge was a shear-thinning fluid and had yield pseudoplasticity. The basic rheological behavior can reflect the internal structure in dense aggregated suspensions (Chen et al., 2010). Fig. 2A shows there was an exponential relationship between SS and limiting viscosity (hN) in the aerobic granular sludge. This relation follows the scaling behavior model, proposed by Mu and Yu (2006), and is able to describe the intra-floc structure in granules. The exponential dependence is described as follows: hN wTSS1=ð3Df Þ

(9)

where Df is the fractal dimension. Fig. 2B shows the doublelogarithmic plots of hN versus SS, and Df was estimated from the slope. It equaled 2.74, greater than 2.51 of flocculent sludge in the same reactor. Based on the basic rheological characterization, it demonstrates that aerobic granules belong to the weak-link regime where links in flocs have higher elasticity than those between neighboring flocs.

3.2. The viscoelastic properties of aerobic granules in the oscillatory shear flow Fig.3 shows the strain dependence of G0 , G00 and d (hysteretic phase angle) of aerobic granular sludge. G0 kept constant at 87

174

w a t e r r e s e a r c h 5 0 ( 2 0 1 4 ) 1 7 1 e1 7 8

Fig. 3 e Evolution of storage modulus, loss modulus and delta during stress sweep at: SS [ 19.4 g/L, T [ 25  C, pH [ 7e8 and [ 5 rad/s. The insert shows the stressestrain relationships for aerobic granules with and without EPS.

Fig. 2 e Correspondence between SS and h N at T of 25  C. (A) h N as a function of SS. (B) Double-logarithmic plots of h N versus SS.

(2) Pa at low shear strain, suggesting a linear viscoelastic regime (LVE). In LVE regime, G00 /G0 was equal to 0.17 and the value had no relevance with strain change. Then moduli became strain dependent as g increased, leading to a nonlinear viscoelastic regime (non-LVE). When g exceeded 6%, the Payne-like effect (Wang, 1999) appeared: G0 decreased monotonically and G00 passed through a peak followed by a decrease as well. Results showed that the peak remained almost constant even though temperature went up to 50  C. Considering the fact that d increased from 8 to 85 over the sweep range, sludge was also supposed to transform from solid-like to liquid-like. In non-LVE regime, namely liqiud-like regime, the strain-dependence of both G0 and G00 followed a power-law equation: G00 Ngn G0 Ng2n

(10)

This peculiar relationship is known to be the hallmark of soft-glassy materials, so aerobic granules can share the universal rheology methodology with many other systems, such as colloidal glasses, gels and raw sludge (Mason and Weitz, 1995). Interestingly, the transition from LVE to non-LVE was not smooth (the insert in Fig. 3): there existed a “fluctuation” in the

applied stress when performing strain sweep. When g increased from 5 to 21%, G0 dropped sharply from 76 to 10 Pa, but the applied stress waved between 3.22 and 4.54 Pa. It was weakened and even disappeared when EPS was partly extracted. Based on our knowledge (Mikkelsen and Keiding, 2002), the fluctuation is attributed to a larger amount of EPS in granules and its non-homogeneous stratification, which increase interaction via supra-molecular entanglement, leading to more prominent relaxation processes. This inference is proved by the direct yield in raw sludge and the smooth transition in digested sludge (Baudez et al., 2012), Nevertheless, it was limited to a qualitative conclusion. The contribution that EPS made to gross dynamic modulus couldn’t be described quantitatively by comparison in Fig. 3, due to partial disruption of the structural integrity of granules in EPS extraction. Based on Fig. 3, fixed strains of 1% and 50% were chosen for frequency dependency sweeps in LVE and non-LVE regimes, respectively. In the linear regime (Fig. 4A), both G0 and G00 illustrated a similar weak power-law dependence across the frequency range, from 0.01 to 10 Hz (power-law index smaller than 0.15). G0 was 5 times larger than G00 , and it showed a solidlike regime. In higher frequency range, G00 had a shallow minimum which wouldn’t appear in granules without EPS. On the whole, G00 /G0 was almost 0.17, in consistent with strain sweep one in Fig. 3. In the non-linear regime (Fig. 4B), G00 was always larger than G0 , so it demonstrated a liquid-like regime. Aerobic granules exhibited plateau for both moduli at low frequencies, especially G0 . At higher frequencies, moduli followed apparent power-law relationships, and all indexes turned out to be nearly 0.5. The increasing trend after plateau was due to an entangled network of disordered polymer coil, indicating the characteristic of the time scale of molecular entanglement in granular sludge. According to the decrease of viscosity in both Figs. 4A and B, it identically demonstated that aerobic granular sludge behaved like a viscoelastic solid and a shearthinning fluid.

175

w a t e r r e s e a r c h 5 0 ( 2 0 1 4 ) 1 7 1 e1 7 8

direct proportionality between water viscosity and shear strain at the cross-points where G0 equals G00 and above (R2 ¼ 0.96). This relation infers that viscoelastic properties of aerobic granules follow an Arrhenius law with temperature, just like water viscosity. Liquid-like granules share the same activation energy as water viscosity, and the same molecular movements are involved in the liquid-like regimes of both granules and flocs (Baudez et al., 2012). However, experimental condition was controlled in a moderate temperature range, because extreme conditions weren’t taken into our plan due to probable EPS detachment in the time-consuming frequency sweeps. That is also a point requiring further study to explore rheology of biological system in extreme conditions.

3.3. Evaluation the Wagner-type constitutive model in sludge system

Fig. 4 e Frequency sweep of aerobic granular sludge at: SS [ 19.4 g/L, T [ 25  C, pH [ 7e8. (A) LVE regime: strain [ 1%; (B) non-LVE regime: strain [ 50%.

As a pervasive influence factor, temperature is always investigated in rheological study (Baroutian et al., 2013). Table 1 tells that the temperature increase didn’t lead to an obvious decrease of rheological characteristics of aerobic granular sludge: G0 almost had no change and the peak of G00 before its decrease did exist along the entire temperature sweep. There is an overall evaluation that granular sludge is more thermally stable. Both the maximum of G0 /G00 and the minimum of loss factor at 25  C here, suggesting the highest structural strength, reveal an optimum condition for cultivating granules, and it fills the knowledge gap about the role of temperature in aerobic granulation (Liu and Tay, 2004). Moreover, it emphasizes a

Table 1 e Water viscosity and strain where G0 [ G00 at 5 rad/s and the average modulus in LVE regime for aerobic granules at 4 temperatures (10  C, 25  C, 30  C, 40  C). T ( C) Water viscosity (mPa.s) Strain (%)

G (G0 > G00 ) G0 (Pa) G00 (Pa)

10 25 30 40

1.3077 0.8937 0.8007 0.656

49.6 27.0 25.0 22.8

4.69 7.084 5.201 6.641

1.349 1.804 1.696 1.792

Accurate description and simulation of the rheological behaviors in sludge process depends upon the selection of a constitutive equation, which can reflect flow behaviors where material, namely sludge suspension, undergoes fast changes with strong time and strain dependence (Liang and Mackley, 1994; Johnson and Cook, 1983). Here the Wagner-type model, incorporated with the relaxation and damping function, was used to predict fluid viscosity of aerobic granular sludge in steady-shear flow, and match the viscoelastic responses with steady-state flow behaviors. The self consistency was identified by comparing the predicted and measured data. Starting with the sweeps above, a discrete spectrum of relaxation time was employed to fit the frequency sweeps (Table 2). In order to obtain the discrete spectrum, the 1stOpt general global optimization method was used (1stOpt 5.0 program, 7D-Soft High Technology lnc., Beijing, China), and a time domain equivalent to the experimental frequency domain was chosen from 0.0001 to 1000 s. Because of the poly-disperse relaxation processes, an eight-element Maxwell model instead of a single element model was built to describe the frequency dependence of G0 and G00 (Liang and Mackley, 1994). To check the rationality of the calculation, the dynamic modulus between experiment and prediction was compared, and Fig. 5 shows a good coherence between them (R2 > 0.90), so did the distribution trend of the values of giet/li (Eq. (6)). The Wagner damping factor h (g), characterizing the degree of non-linearity (Eq. (5)), was determined from linear and non-linear step strain data (Table 2). An estimated value of 2.5 was adopted to predict the viscosity response of aerobic

Table 2 e Discrete relaxation spectrum of aerobic granules at T of 25  C. i

li (s)

1 2 3 4 5 6 7 8

0.0001 0.001 0.01 0.1 1 10 100 1000

gi (Pa) LVE

Non-LVE

927.858 9.578 30.374 12.575 11.216 17.007 1.7 38.072

3498.142 1.611 7.247 6.041 1.941 5.632 8.466 2.659

176

w a t e r r e s e a r c h 5 0 ( 2 0 1 4 ) 1 7 1 e1 7 8

4.

Discussion

4.1. Rheology in micro-structure analysis of aerobic granular sludge

Fig. 5 e Frequency sweep (shear strain [ 1%) with discrete linear relaxation spectrum fit.

granular sludge when subjected to steady shear rate from 1 to 300 s1 (Eq. (7)). Fig.6 illustrates the predicted viscosity response based on the Wagner-type constitutive model, using the discrete spectrum of relaxation time. Compared to experimental data, the predicted viscosity declared an excellent agreement at shear rates from 100 to 200 s1, and had a similar tendency to keep relatively stable when shear rate surpassed 200 s1. To check the validity, a semi-empirical calculation method proposed by Schwarzl and Staverman, was introduced without the damping factor. The dynamic relaxation modulus is obtained directly from approximate value.   2 HðlÞ ¼ G00 ðuÞ1=u ¼ l p

(10)

Fig. 6 shows there was a big error at either low or high shear rates with this semi-empirical fitting method (p ¼ 0.228 in statistics). To put it another way, it supports to select the Wagner-type model (p ¼ 0.769 > 0.228), which matches the viscoelastic response with steady-state flow behavior successfully.

Fig. 6 e Experimental data and Wagner model based prediction in steady-state flow.

Microbial granules growing in aqueous environment generally have heterogeneous structures consisting of cell clusters (aggregates of bacterial cells held together in an extracellular slime matrix) separated by interstitial voids and channels. Physical properties, including rheological parameters, play a big role in determining its shape and mechanical stability (Seviour et al., 2009a). That is why we try to elaborate the reflection of rheology on internal microstructure. In this research, basic rheological characterization is extended in aerobic granular sludge. Some basic rheological parameters, e.g. limiting viscosity, reduced hysteresis area and yield stress, are right to reflect structure characteristics of aerobic granules, and show different sludge qualities, just as Tixier et al. (2003) did in activated sludge. Moreover, viscoelasticity has made more differences in analyzing material structure. As known, even one of the simplest systems (i.e. a suspension of hard-sphere particles) shows a complex viscoelastic behavior that is tightly linked to changes in microstructure (Wyss et al., 2005). Therefore, it is significant to apply viscoelastic study, which has only been discussed in flocculent sludge system before, for elucidating these associations composing the fine continuous network in granules. Firstly, the appearance of Payne-like effect (Fig. 3), attributed to the breakage and recovery of weak physical bonds linking adjacent filler clusters, proves the fact that physical or non-covalent associations are more common with biopolymer gels than covalent links (Seviour et al., 2009b). Secondly, comparing sweeps of granules with and without EPS (the insert in Fig. 3), it is concluded that entanglements between high-polymer molecules cause mechanical interference, and it indeed has a vital role in matrix linking, besides hydrogen bonds and Coulomb interactions. And a bold assumption is even made: the so-called “double network technique” (Haque et al., 2012), realizes the more stable network for EPS with addition of calcium ions. The first brittle network in flocs serves as sacrificial bonds, while the second ductile polymer chains in granules act as hidden length, extending to sustain large deformation. Finally, due to the significant contribution of EPS in rheological properties for the intactness and structural integrity/strength of granular sludge, it is concluded that more EPS the sludge produces, more robust the structure is, which ensures superiorities of granules over flocculent sludge. At last, results obtained in frequency sweep of aerobic granules (Fig. 4) are compared with literature data of digested sludge or raw sludge (Baudez et al., 2012). A particular average relaxation value of the order of 10 s, even larger is obtained, suggesting a more durable relaxation process. Then, why? As well-known, average relaxation time (l ¼ 1/uint) is characterized as average lifetime of “junction points”. Applied stress gets to relax through these points under certain conditions (Wloka et al., 2004). Therefore, reasons for the more durable relaxation process are as follows: Firstly, when experiment is performed at higher frequencies, there is not sufficient

w a t e r r e s e a r c h 5 0 ( 2 0 1 4 ) 1 7 1 e1 7 8

fluctuation of “junction points”. A permanently cross-linked network is observed and it has insufficient time to come part. Secondly, granules behave more like ideal elastomer with a longer relaxation time than flocs. Modulus attains plateau at frequencies about 0.1 Hz in non-LVE regime, which is lower than that for flocculent sludge. It is demonstrated as a rule that with increased concentration of EPS, the crossover inclines to shift to lower frequencies. Eventually, the intermediate frequency range of aerobic granules is characterized by ambivalent viscoelastic behaviors. Based on these rheological properties, granular sludge obviously performs a more robust structure and greater endurance to withstand shock loadings, having advantages over flocculent sludge, as confirmed in other ways identically (Adav et al., 2008). The results differentiate aerobic granular sludge from other sludge systems, and highlight the utility of rheological techniques for studying the particular three-dimensional microstructure and quantifying the granulation process.

4.2.

Rheology in macro-scope view of sludge flow

Generally, microscopic study in rheology means molecular rheology, where microstructure (molecular motion) is correlated with rheological property. Comparatively speaking, macroscopic study means a typical phenomenological research. Starting with a viscoelastic model, researchers conduct a strain-stress analysis, describing its response to external stimulation, just as Larson and Monroe (1984) adopted the BKZ and the Wagner model to fit shear and elongational flow data of an LDPE (low-density polyethylene) melt. Rheological method is a useful tool for evaluating sludge, e.g. aerobic granular sludge, in terms of processability or quality and texture. But it is a challenge to describe the timedependent and non-linear behavior in sludge flow. Most researches are focused on two well known limits of shear rheology: linear viscoelasticity and steady-state shear measurements (Bengoechea et al., 2008). Fewer articles are devoted to study non-linear viscoelasticity, which are required to fill the gap between both limits and to provide valuable information on the effect of shear on microstructure. Therefore, it is one of new trials in sludge system to introduce a non-linear viscoelastical constitutive model, particularly in granular sludge systems. This work demonstrates that the non-linear Wagner-type model has a higher accuracy than empirical ones which describe non-linearity from only one single test (Figoni and Shoemaker, 1983). The good self consistency between oscillatory and step-strain data proves its practical applicability, which has also been confirmed through predicting flow properties for three emulsions with timestrain separability (Bengoechea et al., 2008). This achievement promotes the establishment of an appropriate constitutive model in the sludge flow. Considering the inherent opacity of sludge, it is impossible to visualize mix, flow and stress patterns inside wastewater treatment system. This also makes it meaningful to introduce a constitutive model in the sludge flow. Eshtiaghi et al. (2012) has tried to use three clear model fluids to emulate rheological properties of thickened digested sludge: Carboxymethyl cellulose fluid modeling the flow behavior at high shear rates,

177

Carbopol gel fluid in short-time flow process, and Laponite clay fluid where time dependence is dominant. Nevertheless, what makes it more complex is to compare these three model fluids to mimic uniform behaviors of actual sludge flow. But it turns to be explicit if a model reflecting textural properties is successfully used. These rheological parameters in the model are very important for sewage sludge management, e.g. designing parameters in transporting, storing, spreading operations or in determining the design requirements for a pump scheme, and will improve the actual sludge flow process. To illustrate, high EPS content of AGS, which plays a big role in determining structural integrity/strength of granules, makes it harder and problematic to dewater. Also these parameters can give a semi-quantitative or qualitative delimitation when flocculent sludge granulates itself or granular sludge comes to deterioration. What’s more, as sludge disposal costs constitute a large component of operating costs in wastewater treatment plants, and time factors, governing the sludge consolidation process, is essential to describe the relation between void ratio and compressive pressure of the sludge cake during mechanical expression, a viscoelastic model has practical significance in sludge dewatering, compression and deformation relaxation (Raynaud et al., 2010). A constitutive model indeed brings improved techniques and principles, supporting the development in municipal and industrial wastewater treatment.

5.

Conclusions

In this study, rheology was applied to analyze aerobic granular sludge via steady-shear and oscillatory measurements. The main outcomes were:  Rheology provided insight into these intrinsic architectural properties of aerobic granular sludge, and showed that rheological characteristics reflected internal microstructure.  Aerobic granular sludge was a shear-thinning fluid with yield pseudoplasticity, similar to HerscheleBulkley fluid, a typical non-Newtonian class of fluids.  When oscillatory sweeps were conducted, aerobic granules behaved like elastic and viscoelastic solid or viscoelastic liquid. Its peculiar viscoelastic response highlighted the necessity of EPS in granulation, and made a distinction between granular sludge and other sludge systems.  Wagner-type model was able to characterize time-strain separability, and match viscoelastic responses with steadystate flow, which makes a promising step on predicting flow behaviors for sludge flow with a viscoelastic model.

Acknowledgments The authors would like to acknowledge the financial support of the Hundred-Talent Program of Chinese Academy of Sciences and the Program for Changjiang Scholars and Innovative Research Team in University

178

w a t e r r e s e a r c h 5 0 ( 2 0 1 4 ) 1 7 1 e1 7 8

references

Adav, S.S., Lee, D.J., Show, K.Y., Tay, J.H., 2008. Aerobic granular sludge: recent advances. Biotechnol. Adv. 26 (5), 411e423. APHA, A, 1998. WEF, Standard Methods for the Examination of Water and Wastewater. 4500-NO3-D nitrate Electrode Method, twentieth ed. American Public Health Association, Washington, DC. Baroutian, S., Eshtiaghi, N., Gapes, D.J., 2013. Rheology of a primary and secondary sewage sludge mixture: dependency on temperature and solid concentration. Bioresour. Technol. 140, 227e233. Baudez, J.-C., 2006. About peak and loop in sludge rheograms. J. Environ. Manag. 78 (3), 232e239. Baudez, J.-C., Gupta, R.K., Eshtiaghi, N., Slatter, P., 2012. The viscoelastic behaviour of raw and anaerobic digested sludge: strong similarities with soft-glassy materials. Water Res. 47 (1), 173e180. Bengoechea, C., Puppo, M.C., Romero, A., Cordobes, F., Guerrero, A., 2008. Linear and non-linear viscoelasticity of emulsions containing carob protein as emulsifier. J. Food Eng. 87 (1), 124e135. Bower, C., Gallegos, C., Mackley, M., Madiedo, J., 1999. The rheological and microstructural characterisation of the nonlinear flow behaviour of concentrated oil-in-water emulsions. Rheol. Acta 38 (2), 145e159. Chang, I., Lee, D., 1998. Ternary expression stage in biological sludge dewatering. Water Res. 32 (3), 905e914. Chen, D.T., Wen, Q., Janmey, P.A., Crocker, J.C., Yodh, A.G., 2010. Rheology of soft materials. Cond. Mat. Phys. 1. Eshtiaghi, N., Yap, S.D., Markis, F., Baudez, J.-C., Slatter, P., 2012. Clear model fluids to emulate the rheological properties of thickened digested sludge. Water Res. 46 (9), 3014e3022. Figoni, P.I., Shoemaker, C.F., 1983. Characterization of time dependent flow properties of mayonnaise under steady shear. J. Texture Stud. 14 (4), 431e442. Haque, M.A., Kurokawa, T., Gong, J.P., 2012. Super tough double network hydrogels and their application as biomaterials. Polymer 53 (9), 1805e1822. Johnson, G.R., Cook, W.H., 1983. A Constitutive Model and Data for Metals Subjected to Large Strains, High Strain Rates and High Temperatures. International Ballistics Committee, The Hague, Netherlands, pp. 541e547. Larson, R.G., 1988. Constitutive Equations for Polymer Melts and Solutions. Butterworths, Boston. Larson, R., Monroe, K., 1984. The BKZ as an alternative to the Wagner model for fitting shear and elongational flow data of an LDPE melt. Rheol. Acta 23 (1), 10e13. Liang, R., Mackley, M., 1994. Rheological characterization of the time and strain dependence for polyisobutylene solutions. J. Non-newton. Fluid Mech. 52 (3), 387e405. Liao, B., Allen, D., Droppo, I., Leppard, G., Liss, S., 2001. Surface properties of sludge and their role in bioflocculation and settleability. Water Res. 35 (2), 339e350. Liu, Y., Tay, J.-H., 2004. State of the art of biogranulation technology for wastewater treatment. Biotechnol. Adv. 22 (7), 533e563. Liu, X.-W., Sheng, G.-P., Yu, H.-Q., 2009. Physicochemical characteristics of microbial granules. Biotechnol. Adv. 27 (6), 1061e1070.

Lin, Y., de Kreuk, M., Van Loosdrecht, M., Adin, A., 2010. Characterization of alginate-like exopolysaccharides isolated from aerobic granular sludge in pilot-plant. Water Res. 44 (11), 3355e3364. Lin, Y., Sharma, P., van Loosdrecht, M., 2013. The chemical and mechanical differences between alginate-like exopolysaccharides isolated from aerobic flocculent sludge and aerobic granular sludge. Water Res. 47 (1), 57e65. Mackley, M., Marshall, R., Smeulders, J., Zhao, F., 1994. The rheological characterization of polymeric and colloidal fluids. Chem. Eng. Sci. 49 (16), 2551e2565. Mason, T., Weitz, D., 1995. Linear viscoelasticity of colloidal hard sphere suspensions near the glass transition. Phys. Rev. Lett. 75 (14), 2770. Mikkelsen, L.H., Keiding, K., 2002. Physico-chemical characteristics of full scale sewage sludges with implications to dewatering. Water Res. 36 (10), 2451e2462. Morgan, J., Forster, C., Evison, L., 1990. A comparative study of the nature of biopolymers extracted from anaerobic and activated sludges. Water Res. 24 (6), 743e750. Mu, Y., Yu, H.-Q., 2006. Rheological and fractal characteristics of granular sludge in an upflow anaerobic reactor. Water Res. 40 (19), 3596e3602. Raynaud, M., Heritier, P., Baudez, J.-C., Vaxelaire, J., 2010. Experimental characterisation of activated sludge behaviour during mechanical expression. Proc. Saf. Environ. Prot. 88 (3), 200e206. Seviour, T., Pijuan, M., Nicholson, T., Keller, J., Yuan, Z., 2009a. Gel-forming exopolysaccharides explain basic differences between structures of aerobic sludge granules and floccular sludges. Water Res. 43 (18), 4469e4478. Seviour, T., Pijuan, M., Nicholson, T., Keller, J., Yuan, Z., 2009b. Understanding the properties of aerobic sludge granules as hydrogels. Biotechnol. Bioeng. 102 (5), 1483e1493. Sozanski, M.M., Kempa, E.S., Grocholski, K., Bien, J., 1997. The rheological experiment in sludge properties research. Water Sci. Technol. 36 (11), 69e78. Stoodley, P., Lewandowski, Z., Boyle, J.D., Lappin-Scott, H.M., 1999. Structural deformation of bacterial biofilms caused by short-term fluctuations in fluid shear: an in situ investigation of biofilm rheology. Biotechnol. Bioeng. 65 (1), 83e92. Tixier, N., Guibaud, G., Baudu, M., 2003. Determination of some rheological parameters for the characterization of activated sludge. Bioresour. Technol. 90 (2), 215e220. Wagner, M., 1976. Analysis of time-dependent non-linear stress-growth data for shear and elongational flow of a lowdensity branched polyethylene melt. Rheol. Acta 15 (2), 136e142. Wang, M.-J., 1999. The role of filler networking in dynamic properties of filled rubber. Rubber Chem. Technol. 72 (2), 430e448. Wloka, M., Rehage, H., Flemming, H.-C., Wingender, J., 2004. Rheological properties of viscoelastic biofilm extracellular polymeric substances and comparison to the behavior of calcium alginate gels. Colloid Poly. Sci. 282 (10), 1067e1076. Wyss, H.M., Tervoort, E.V., Gauckler, L.J., 2005. Mechanics and microstructures of concentrated particle gels. J. Am. Ceramic Soc. 88 (9), 2337e2348. Ying, W., Yang, F., Bick, A., Oron, G., Herzberg, M., 2010. Extracellular polymeric substances (EPS) in a hybrid growth membrane bioreactor (HG-MBR): viscoelastic and adherence characteristics. Environ. Sci. Technol. 44 (22), 8636e8643.

A rheological approach to analyze aerobic granular sludge.

Aerobic granular sludge is one promising biotechnology in wastewater treatment. Despite intensive researches on granular architecture and strategies t...
878KB Sizes 1 Downloads 0 Views