Understanding the evolution of stratified extracellular polymeric substances in full-scale activated sludges in relation to dewaterability

Abstract

Activated sludge is a highly changeable colloidal system. In this study, the dynamic variation in physicochemical characteristics, especially distribution and abundance of extracellular polymeric substances (EPS) of activated sludges from different WWTPs was investigated in order to establish the relationships between floc properties and the sludge dewatering property. Firstly, it was observed that the total EPS content of the activated sludge was significantly decreased with the rise in temperature. Three-dimensional fluorescence excitation–emission matrix (3D-EEM) spectroscopy analysis indicated that each sludge fraction (soluble EPS, loosely-bound EPS (LB-EPS), tightly bound EPS (TB-EPS) and pellet) from different WWTPs had a similar fluorescence fingerprint in the same time period. In addition, protein-like substances were found to be the dominant components in TB-EPS and pellets regardless of operating time for each WWTP sludge. At low temperatures, soluble EPS and LB-EPS also mainly contained protein-like compounds, while the amount of humic acids of them was increased significantly in the summer. According to Pearson's correlation analysis, normalized CST correlated well with the composition and content of soluble EPS, indicating that the change in soluble EPS properties caused fluctuation of sludge dewatering behavior. Finally, we proposed some operating strategies for improving the dewatering performance of activated sludge in full-scale WWTPs by regulating the soluble EPS properties.

---

1. Introduction

The management of wastewater sludge, now often referred to as biosolids, accounts for a major portion of the cost of the wastewater treatment process and represents significant technical challenges. Dewatering has been proven to be an efficient method to reduce sludge volume, cutting transportation and disposal cost.^1,2

Extracellular polymeric substances (EPSs) are the most crucial constituents of activated sludge and account for 60–80% of the total biomass.³ Consisting of high molecular-weight (MW) polysaccharides, proteins, glycoproteins, nucleic acids, phospholipids, and humic acids, EPS affects flocculating, settling and dewatering properties of activated sludge.³ Getting insight into the role of EPS on the sludge dewatering property became a very popular topic in last few decades. The history on this study has generally experienced three stages: (1) EPS content;⁴ (2) EPS composition and abundance;^5,6 (3) fractionation of EPS⁷ and stratification of sludge floc.^8,9 Initially, Knocke et al. suggested that biopolymer content had a major effect on dewatering performance of biological sludge, and sludge floc density determined the ultimate solid concentration after filtration dewatering.⁴ Houghton et al. also demonstrated that sludge dewaterability was mainly dependent on EPS content and there appeared to be an optimum level of EPS for each type of sludge (activated sludge or digested sludge) at which the sludge should show the best dewatering performance.¹⁰ Besides, EPS content also affected the surface charge, zeta potential, moisture content and floc stability.⁶ Afterward, it was found that EPS composition and abundance had more significant effects on sludge properties. Many studies suggested that mass ratio of protein and polysaccharide in EPS determined the sludge dewaterability.³ Especially, protein played a more important role that high content of bound protein was beneficial to sludge dewatering.¹¹ On the contrary, Murthy and Novak reported that sludge dewaterability deteriorated with increasing protein concentration of EPS.¹² EPS of activated sludge particles are not uniform and are likely to have a dynamic double-layered structure of loosely bound EPS (LB-EPS) diffused from the tightly bound EPS (TB-EPS).¹³ According to the strength of their binding to floc, Li et al. used a two step heat extraction method to separate the LB-EPS and TB-EPS from sludge floc. The sludge dewatering property was found to be much more strongly correlated with the concentration of LB-EPS than with the concentration of TB-EPS, and excessive LB-EPS could weaken the floc structure resulting in poor bioflocculation and sludge–water separation performance.⁷ Additionally, they found that the sludge flocculating behavior, settleability, compressibility and dewaterability correlated well with LB-EPS content rather than TB-EPS content.¹⁴ Yu et al. fractionated sludge floc into five layers (supernatant, slime, LB-EPS, TB EPS and pellet) with combined ultrasound and centrifugation treatment. A strong correlation was found between the parameters for sludge dewatering characteristics and protein concentration in the supernatant, slime, and LB-EPS, while no correlation showed between organic composition in the pellet and the sludge flocs as a whole.⁹ Furthermore, three-dimensional excitation emission matrix combined with parallel factor analysis was proven to be an effective tool to characterize the biopolymers. Yu et al. also found that sludge dewaterability was mainly determined by protein, humic acid and fulvic acid in slime and LB-EPS layers.⁸ In addition, it was noted that divalent and multivalent cations could bind the EPS together with electrostatic force, thus affected EPS distribution and composition and dewatering behavior of activated sludge.^2,15

As mentioned above, great efforts have been made to understand the effect of physicochemical properties of sludge floc on sludge–liquid separation in biological system. However, the dynamic variation in physicochemical properties, especially EPS distribution and composition of activated sludge from different full-scale WWTPs with operating time and its influence on sludge dewatering property is still not fully understood. Therefore, the aims of this study were to: (1) understand the physicochemical properties of activated sludge floc collected from five different WWTPs in Beijing; (2) investigate the dynamic variation of distribution, abundance and composition of EPS with operating time (from April to July, 2013) by using 3D-EEM spectroscopy; (3) get an insight into relationships between EPS properties and other physicochemical properties of floc and sludge dewatering behavior.

2. Materials and methods

2.1. Sludge samples

24 wastewater sludge samples were collected from sludge recycle pumping lines of the 5 full-scale WWTPs in Beijing from April to July, 2013, each mainly treating domestic wastewater (Table 1). The locations of WWTPs investigated in this study were shown in Fig. S1 of ESI.† The operating status and sampling date of activated sludges from different wastewater treatment plants (WWTPs) can be found in Table 2. The collected samples were transported to a laboratory within 6 h after sampling and then stored at 4 °C prior to analysis. The characteristics of activated sludge obtained from different WWTPs were given in Table 3.

Table 1 Process descriptions of the wastewater treatment plants

Treatment plant	Source of wastewater	Treatment capacity (ton per day)	Biological process	Chemical dosage	SRT^a (days)	SVI^b (mL g^−1)
a Solids retention time (SRT).b Sludge volume index (SVI); data obtained from the treatment plants.
Qinghe MBR (QHM)	Mainly domestic	200 000	Membrane bioreactor (MBR)	—	10–12	96–197
Qinghe A/A/O (QHA)	Mainly domestic	200 000	Anaerobic/Anoxic/Oxic (A/A/O)	—	20–30	45–101
Jiuxianqiao (JXQ)	Mainly domestic	350 000	Oxidation ditch	Aluminum sulfate	10–12	140–308
Gaobeidian (GBD)	Mainly domestic	1 200 000	A/O (primary settling tank)	Aluminum sulfate	10–12	87–116
Xiaohongmen (XHM)	Mainly domestic	600 000	A/A/O (primary settling tank)	—	10–12	103–152

---

Table 2 Operating status of different wastewater treatment plants

Sample	Treatment plant	Operating conditions					Sampling date
		Temperature (°C)	COD (mg L^−1)	NH4^+–N (mg L^−1)	TP (mg L^−1)	COD/nitrogen ratio
1	QHAO	15.0	249	48.9	5.5	5.1	April 16, 2013
2	QHM	15.0	696	53.4	5.7	13.0	April 16, 2013
3	JXQ	14.4	656	60.3	6.5	10.9	April 16, 2013
4	GBD	14.9	390	41.2	4.3	9.5	April 16, 2013
5	XHM	15.5	613	55.5	6.0	11.0	April 16, 2013
6	QHAO	20.3	539	51.2	6.2	10.5	May 06, 2013
7	QHM	19.6	850	54.9	6.1	15.5	May 06, 2013
8	JXQ	19.9	698	59.8	5.8	11.7	May 06, 2013
9	GBD	19.6	440	39.7	4.3	11.1	May 06, 2013
10	XHM	19.5	619	57.3	7.61	10.8	May 06, 2013
11	QHAO	22.3	508	50.2	6.0	10.1	June 03, 2013
12	QHM	22.8	837	50.9	6.8	16.4	June 03, 2013
13	JXQ	22.8	700	53.8	6.5	13.0	June 03, 2013
14	GBD	22.5	548	41.3	5.2	13.3	June 03, 2013
15	XHM	23.0	616	55.2	5.9	11.2	June 03, 2013
16	QHAO	24.1	441	45.1	5.5	9.8	June 25, 2013
17	QHM	24.0	744	41.6	6.6	17.9	June 25, 2013
18	JXQ	23.8	578	48.6	5.7	11.9	June 25, 2013
19	GBD	23.7	430	36.9	4.6	11.7	June 25, 2013
20	QHAO	25.1	441	40.2	6.1	11.0	July 10, 2013
21	QHM	24.5	744	42.5	4.7	17.5	July 10, 2013
22	JXQ	25.1	578	42.4	5.0	13.6	July 10, 2013
23	GBD	25.0	430	32.5	3.9	13.2	July 10, 2013
24	XHM	25.3	441	42.2	5.0	10.5	July 10, 2013

---

Table 3 Physicochemical properties of all wastewater sludges

Sample	CST (s g^−1 L^−1)	d0.1 (μm)	d0.5 (μm)	d0.9 (μm)	Zeta (mV)	pH	VSS/TSS (%)
1	17.36	44.08	124.66	263.70	−17.90	6.81	72.56
2	13.79	23.47	68.36	151.85	−12.10	7.04	63.85
3	15.78	26.08	84.85	234.12	−14.30	7.15	69.63
4	4.74	40.52	106.93	220.16	−17.40	7.08	67.34
5	30.93	11.93	42.33	114.86	−16.30	7.08	64.63
6	6.26	36.77	105.25	215.94	−10.50	6.53	74.00
7	4.05	22.34	65.07	144.61	−11.10	6.76	63.91
8	11.75	22.48	72.23	195.00	−13.00	6.71	68.81
9	3.57	41.19	108.78	222.91	−18.20	6.69	66.56
10	7.18	17.77	56.43	125.16	−25.80	6.81	67.43
11	4.36	31.94	91.06	190.85	−8.70	6.64	70.35
12	3.83	30.23	79.41	166.94	−12.40	6.92	62.83
13	8.78	17.14	54.80	133.97	−15.80	7.04	63.20
14	4.65	47.93	117.80	231.23	−17.00	6.84	59.58
15	4.14	17.90	54.80	112.82	−8.50	6.98	60.26
16	4.14	26.37	78.61	163.37	−12.00	6.78	72.80
17	3.55	33.82	89.38	189.68	−12.10	6.90	59.92
18	5.39	17.62	51.87	129.87	−14.00	7.17	65.47
19	3.59	41.61	120.73	244.04	−15.20	6.82	62.17
20	3.72	23.44	70.93	150.26	−10.30	6.81	65.67
21	5.30	34.59	85.79	170.09	−12.80	7.06	56.45
22	4.11	19.02	53.04	123.33	−12.10	7.07	64.22
23	6.76	32.36	106.56	231.14	−12.30	6.91	58.31
24	7.66	15.63	44.14	93.73	−10.40	6.96	61.36

---

2.2. Analytical methods

2.2.1. Determination of sludge dewaterability. The capillary suction time (CST) was a rapid indication of the filterability of sludge, with results of good reproducibility. Meanwhile, it had good correlation with specific resistance filtration (SRF).⁹ The dewaterability of the sludge flocs was measured with a CST instrument (Model 319, Triton, UK) equipped with an 18 mm diameter funnel and Whatman no. 17 chromatography-grade paper. The CST values were normalized by dividing them by the initial total suspended solid (TSS) concentration and then expressed in units of seconds per liter per gram TSS.

2.2.2. Characteristics of activated sludge floc. A laser diffraction instrument (Malvern Mastersizer 2000, Malvern, UK) was used to measure sludge floc size. The d(0.1), d(0.5) and d(0.9) values mean that 10%, 50% and 90% of the particles measured were less than or equal to the size stated. pH was measured by pH meter (pHS-3C, Shanghai, China). Zeta potential measurements were performed on a Zetasizer (Zetasizer 2000, Malvern, UK). Other sludge parameters, including TSS, volatile suspended solids (VSS), chemical oxygen demand (COD), ammonium nitrogen (NH₄⁺) and total phosphorus (TP) were analyzed following (APHA-AWWA-WEF, 1998).¹⁶ TOC was determined with a TOC analyzer (Teledyne Tekmar, USA).

2.3. Extraction and analysis of EPS

2.3.1. EPS extraction. EPS fractionation process of sludge samples was modified based on the procedures described by Yu et al.⁹ Firstly, raw sludge sample was settled down at 2000 g for 15 min, and the supernatant was collected as soluble EPS. The sediment was resuspended to its initial volume with phosphate buffer solution consisting of Na₃PO₄, NaH₂PO₄, NaCl and KCl. And then the suspension was transferred and centrifuged at 5000 g for 15 min, and the supernatant and sediments were collected separately. This supernatant was used to determine LB-EPS later. The collected sediment was re-suspended with the phosphate buffer solution to the original volume following by treatment with ultrasound at 20 kHz and 480 W for 10 min. The extracted solution was centrifuged at 20 000g for 20 min and separated as TB-EPS. Finally, the solid residues were again re-suspended with the buffer to the original volume. This fraction was the pellet. The particulates present in the supernatant, soluble EPS, LB-EPS, and TB-EPS solutions were removed with polytetrafluoroethylene membranes with a pore size of 0.45 μm prior to fluorescence EEMs and dissolved organic carbon (DOC) analysis.

2.3.2. EPS analysis. All chemical analyses were carried out in duplicate using chemicals of analytical grade. The protein and carbohydrate were determined according to procedure described by Wang et al.¹⁷ Polysaccharide was measured using the anthrone method with a glucose standard (Sinopharm). Protein was determined with the Lowry procedure using bovine serum albumin (BSA) (Sigma) as standard respectively. In addition, 3D-EEM spectra were also used to characterize the organic composition different sludge fractions. Before EEM analysis, the samples were diluted with Milli Q water until concentration of DOC was below 10 mg L⁻¹ 3D-EEM spectra were measured by a Hitachi F-7000 fluorescence spectrophotometer (Japan) with an excitation range from 200 to 400 nm at 10 nm sampling intervals and an emission range from 280 to 500 nm at 10 nm sampling interval. The spectra were recorded at a scan rate of 12 000 nm min⁻¹, using excitation and emission slit bandwidths of 10 nm. Each scan had 37 emission and 27 excitation wavelengths.

2.4. Statistical analysis

Correlation analysis was carried out using the software SPSS version 16.0 for Windows (SPSS). Pearson's correlation coefficient (R) was used to evaluate the linear correlation between physicochemical properties and normalized CST. According to Dancey and Reidy's suggestions, this R from −1 to +1 A correlation of 1, whether it's positive or negative, is a perfect correlation. A correlation of 0 denotes no relationship between two variables. A correlation of 0.7 to 0.9 (positive or negative) is regarded as a strong correlation. R of 0.4 to 0.6 means moderate correlation. R of 0.1 to 0.3 regarded as weak correlations. The correlations were considered statistically significant at a 95% confidence interval (p < 0.05).¹⁸

3. Results and discussion

3.1. Physicochemical properties of sludge floc

As depicted in Table 3, CST for activated sludge from various WWTPs showed different changing trends with operating time. Sludge from JXQ WWTP exhibited the poorest dewatering property while no significant variation in sludge dewaterability for other sludges. It is worthy to note that the CST for most of sludges reached minimum in the summer. The average floc size (d_0.5) of all sludge samples were in the range from 54.80 μm to 126.33 μm, no obvious variation was observable for each WWTP with operating time. The average floc size of GBD, QHA, JXQ and QHM were 95 μm, 83 μm, 62 μm and 71 μm respectively. In addition, the all sludge particles carried negative charge with a zeta potential of −3.8 to −19 mV. This is due to the ionization of the anionic functional groups, such as carboxylic and phosphate. It should be pointed out that the aluminum sulfate was dosed in influent of secondary settling tank for enhancing phosphorus removal in GBD WWTP. The trivalent cations (Al³⁺, Fe³⁺) can serve to bind the negatively charged EPS and improve bioflocculation through electrical neutralization and bridging.¹⁹ Therefore, the activated floc size in GBD WWTP was larger than that in other WWTPs.

The results of Pearson correlation analysis was given in Table S1.† It can be seen from Fig. 1(a) that moderate correlation was found between sludge dewaterability (measured with CST) and d_0.1 (R = −0.66, p < 0.05) rather than d_0.5 and d_0.9, indicating that the large amounts of fine colloidal particles were detrimental to sludge dewatering process. Interestingly, it was found that pH negatively correlated with d_0.5 (R = −0.83, p < 0.05) and d_0.9 (R = −0.82, p < 0.05) (Fig. 1(b)). This result revealed that sludge particle size distribution was affected by solution pH. Liao et al. suggested that the protons (H⁺) present within the bulk liquid tend to neutralize the floc negative charges, more especially the negative functional groups located on the EPS surface, improving the particle flocculating property.²⁰ Meanwhile, Zhu et al. demonstrated that the activated sludge growing in weak acid conditions were larger than that growing in mild alkali medium.²¹ Karr and Keinath observed that the supracolloidal solids of particle size in the range from 1–100 μm were reduced in the presence of acids.²² It was generally accepted that the filter medium and cake layer were more likely to be blocked by fine bio-colloidal particles. Furthermore, it was interesting to note that VSS/TSS was negatively correlated to the solution pH level (R = −0.75, p < 0.01). The sludge with higher organic content was more likely to undergo hydrolysis acidification process. Overall, no strong correlation was found between normalized CST and properties of sludge floc.

Fig. 1 Correlations between particle size and (a) CST, (b) pH.

3.2. EPS distribution and composition in sludge flocs

3.2.1. Change in organic content of four sludge fractions. The variation in organic content of different sludge fraction can be found in Fig. 2. The soluble EPS was the dominant fraction of sludge EPS. The soluble EPS, LB-EPS and TB-EPS accounted for 53.1–72.3%, 9–17% and 15–27% of total EPS. The organic content of soluble EPS was much higher than that of others. In addition, the total extractable EPS content of samples collected in April was much higher than that in June and July. This observation was in agreement with the report of Wilen et al., who also found that the sludge flocs were much more open and irregularly shaped during the winter months compared to that in the summer. Through long-term monitoring of activated sludge floc properties, they also demonstrated that bound EPS content reached maximum in the winter while minimum in the summer, and low EPS concentration was always related to good filterability.²³

Fig. 2 Change in DOC concentrations of four sludge fractions with operating time.

3.2.2. Dynamic variation in EPS distribution and composition with operating time. Fig. 3(a) showed that, for all WWTPs studied, the protein concentration in EPS fractions (soluble EPS, LB-EPS and TB-EPS) was significantly lower while that in pellet was greatly increased at higher temperatures (May–July). This observation revealed that the binding strength of protein-like compounds to microbial cells was much stronger at higher temperatures. Proteins were found to be more dominant than polysaccharides in most samples. In addition, it was noted that extracellular proteins secreted by bacteria were actually enzymes, the total content of which was higher in summer, and this was corresponding to the enhancement of microbial activity.

Fig. 3 Variation in protein (a) and polysaccharide (b) concentration of four EPS fractions in activated sludge from different WWTPs.

Most microbial exopolysaccharides are highly soluble in water, and capsule-forming polysaccharides are attached to the cells surface through covalent bonds to other surface polymers.²⁴ Fig. 3(b) showed that a similar changing trend was observed for polysaccharide in EPS fractions from different WWTPs. The extractable PS in all sludge layers showed very similar pattern, it reached the minimum in summer (22 May to 10 July 2004). More than 80% of PS was located in pellet. Since the polysaccharides are highly biodegradable polymers, they are very likely to be more easily degraded in the presence of high content of exoenzymes in summer.

3.2.3. EEM analysis of EPS properties. EEM spectroscopy was proven to be an appropriate and effective method to characterize the biopolymers from various origins in wastewater treatment systems. Each EEM gives spectral information about the chemical compositions of EPS samples. As showed in Fig. 4 and S2–S5 of ESI,† the change in fluorescence characteristic of DOM from different WWTPs showed a similar pattern with time. However, it was observed that the chemical composition of TB-EPS and pellet was similar and relatively stable with time. Two dominant peaks at excitation/emission (Ex/Em) of 230/330 (aromatic protein II) and 280/330 (tryptophan-protein like substances) were detected in the TB-EPS and pellet fractions.^25,26 Only a weak fluorescent signal of humic acids was observed after June, 2013. However, it was found that the chemical components in soluble EPS and LB-EPS changed considerably with time. As stated by Sheng and Yu, the intensity of fluorescent peak had a linear relationship with the EPS concentration at a low concentration (e.g., <10 mg C/L), so it can be used to quantify the EPS.²⁷ Additionally, the Fig. 4 and S2–S5 of ESI† also clearly showed that the peak located at Ex/Em of 330/400 related to the humic-like substances began to appear and become a dominant fraction in soluble EPS on July. This may be attributed to the fluctuation of temperature and influent quality (COD, N, P etc.). Humic substances in sludge EPS primarily came from two sources: adsorption from wastewater²³ and hydrolytic conversion from other biopolymers such as proteins.²⁸ Since the microbial activity was enhanced at higher temperature, the increase of degradation of the organic matter in the sewer system caused the accumulation of humic-like substances.²³ As stated by Xi et al. the microorganisms produced more biomass associated products (BAP) mainly composed by humic substances under substrate-limited condition, and BAP primarily originated from hydrolysis of LB-EPS.²⁸ This study confirmed that the soluble EPS and LB-EPS had very similar chemical constituents regardless of operating conditions and influent quality of WWTPs.

Fig. 4 Change in EEM profiles of sludge from QHM WWTP with time (a) 2013-04-16 (b) 2013-05-06 (c) 2013-05-23 (d) 2013-06-03 and (e) 2013-07-10.

In order to better understand the influence of chemical composition of DOMs in four floc layers on sludge dewaterability, the fluorescence regional integration (FRI) method was employed to analyze the five excitation–emission regions as described by Chen.²⁶ As shown in Fig. S6,† peaks at shorter wavelengths (<250 nm) and shorter emission wavelengths (<350 nm) are associated with simple aromatic proteins such as tyrosine and tryptophan (Regions I and II). Peaks at intermediate excitation wavelengths (250–280 nm) and shorter emission wavelengths (<380 nm) also are related to protein-like substances (region IV) while peaks located at the excitation wavelengths (200–250 nm) and the emission wavelengths (>380 nm) represent fulvic acid-like substances (region III). Peaks at longer excitation wavelengths (>280 nm) and longer emission wavelengths (>380 nm) are related to humic substances (region V). It can be seen from Fig. 5 that relative proportion of humic-like substances in soluble EPS and LB-EPS was increased by around 20–30% while that of aromatic protein-like substances reduced by about 20% from April to July, 2013. Furthermore, it is interesting to note that tryptophan-like substances increased significantly in the summer, because they were related to microbial activity which could be enhanced with increase in temperature.

Fig. 5 Change in FRI distribution with operating time (a) 2013-04-16 (b) 2013-05-06 (c) 2013-05-23 (d) 2013-06-03 and (e) 2013-07-10.

3.3. Correlations of influent quality and operating condition and EPS properties of AS

Table 4 showed that the temperature was negatively correlated with content of SEPS (R = −0.86, p < 0.01), LB-EPS (R = −0.89, p < 0.01), TB-EPS (R = −0.77, p < 0.01) and pellet (R = −0.89, p < 0.01). PS content in SEPS, LB-EPS, TB-EPS and PN in SEPS were also negatively correlated to water temperature. There was no correlation between C/N ratios of the wastewater and EPS properties in the sludge. These results indicated that temperature had significant effect on EPS production of activated sludges. Table 2 showed that the operating temperature of biochemical tank rose from 15 to 25 °C from April to July while the influent COD concentration decreased. This result agreed with the finding of Wang et al., who observed that content of bound EPS reached maximum in the winter while minimum in the summer.²⁹ The high EPS production at low temperature might be attributed to a shift in microbial populations due to the stress on microorganisms at low temperatures and/or a reduced degradation of these substances due to kinetic considerations.²⁹

Table 4 Coefficients of correlation (R) for linear regression between operating conditions and EPS properties of activated sludges (n = 24)

Operating conditions		Sludge fractions				PS				PN
		SEPS	LB	TB	Pellet	SEPS	LB	TB	Pellet	SEPS	LB	TB	Pellet
a Correlation is significant at the 0.01 level (2-tailed).b Correlation is significant at the 0.05 level (2-tailed); SEPS-soluble EPS; LB-loosely bound EPS; TB-Tightly bound EPS; OC-organic content; ZP-zeta potential; PS-polysaccharide; PN-protein.
Organic loading	Pearson correlation	−0.49	−0.66	−0.51	−0.33	−0.48	−0.26	−0.22	−0.47	−0.16	−0.60	−0.44	0.58
	Significance	0.32	0.05	0.27	0.65	0.29	0.73	0.82	0.31	0.90	0.08	0.36	0.11
Carbon/nitrogen (C/N)	Pearson correlation	−0.57	−0.64	−0.52	−0.49	−0.55	−0.40	−0.26	−0.36	−0.27	−0.58	−0.49	0.47
	Significance	0.13	0.05	0.21	0.27	0.12	0.42	0.72	0.50	0.71	0.08	0.21	0.26
Temperature (°C)	Pearson correlation	−0.86^a	−0.89^a	−0.77^a	−0.89^a	−0.68^b	−0.71^b	−0.69^b	0.30	−0.72^b	−0.51	−0.41	0.20
	Significance	0.00	0.00	0.00	0.00	0.02	0.01	0.01	0.66	0.19	0.01	0.37	0.83

---

3.4. Relationship between characteristics of different sludge fractions and dewaterability

Table 5 presented the relationships between CST and distribution, abundance, composition of four sludge fractions. A strong positive correlation was found between CST and total content of soluble EPS (R = 0.79, p < 0.01) rather than organic content of other fractions (Fig. 6(a)). It is generally believed that the increase in content of total EPS are positive correlated to dynamic viscosity of mixed liquor, resulting in a higher degree of accumulation of polymers and sludge particles on membrane surface and furthermore deterioration of sludge filterability and dewaterability.³⁰ However, there was no significant correlation between LB-EPS content and CST. This was contradictory to the finding of Li and Yang, who found that sludge properties (settleability, filterability and dewaterability) were mainly influenced by the characteristics of LB-EPS.⁷

Table 5 Correlation between distribution, abundance, composition of organic compounds in different sludge layers and sludge dewaterability (n = 24)^c

	SEPS	LB	TB	Pellet	S1	S2	S3	S4	S5	LB1	LB2	LB3	LB4	LB5	TB1	TB2	TB3	TB4	TB5	P1	P2	P3	P4	P5
a Correlation is significant at the 0.05 level (2-tailed).b Correlation is significant at the 0.01 level (2-tailed).c SEPS-soluble EPS, LB-LB-EPS, TB-TB-EPS, S1-region I of soluble EPS, S2-region II of soluble EPS, S3-region III of soluble EPS, S4-region IV of soluble EPS,…
CST	0.79^b	0.61	0.36	0.64	0.74^a	0.76^a	0.62	0.75^a	0.41	0.57	0.60	0.60	0.60	0.48	0.42	0.43	0.41	0.22	−0.12	0.63	0.55	0.63	0.59	0.61
SEPS	1	0.89^b	0.84^b	0.85^b	0.94^b	0.95^b	0.93^b	0.95^b	0.51	0.88^b	0.85^b	0.90^b	0.89^b	0.65^b	0.88^b	0.88^b	0.87^b	0.64	0.56	0.86^b	0.84^b	0.86^b	0.81^b	0.78^b
LB	—	1	0.72^a	0.93^b	0.91^b	0.89^b	0.89^b	0.78^b	−0.11	0.99^b	0.99^b	0.99^b	0.98^b	0.63	0.83^b	0.82^b	0.81^b	0.40	0.24	0.94^b	0.93^b	0.95^b	0.87^b	0.82^b
TB	—	—	1.00	0.57	0.71^a	0.75^a	0.82^b	0.85^b	0.59	0.71^a	0.64	0.74^a	0.77^a	0.65	0.95^b	0.97^b	0.96^b	0.92^b	0.85^b	0.57	0.54	0.57	0.59	0.55
Pellet	—	—	—	1	0.91^b	0.89^b	0.82^b	0.70^a	−0.40	0.94^b	0.95^b	0.93^b	0.86^b	0.32	0.72^a	0.70^a	0.67	0.18	−0.40	0.99^b	1.00^b	1.00^b	0.97^b	0.91^b
S1	—	—	—	—	1	0.99^b	0.90^b	0.78^b	−0.46	0.93^b	0.90^b	0.93^b	0.85^b	0.41	0.86^b	0.84^b	0.83^b	0.29	−0.24	0.94^b	0.92^b	0.92^b	0.81^b	0.72^a
S2	—	—	—	—	—	1	0.92^b	0.81^b	0.38	0.90^b	0.88^b	0.91^b	0.85^b	0.52	0.86^b	0.85^b	0.84^b	0.44	0.26	0.92^b	0.90^b	0.90^b	0.80^b	0.72^a
S3	—	—	—	—	—	—	1	0.82^b	0.33	0.87^b	0.86^b	0.91^b	0.90^b	0.74^b	0.84^b	0.85^b	0.86^b	0.65	0.62	0.83^b	0.82^b	0.84^b	0.78^b	0.76^b
S4	—	—	—	—	—	—	—	1	0.77^b	0.75^b	0.72^b	0.76^b	0.83^b	0.68^a	0.83^b	0.83^b	0.81^b	0.77^b	0.73^a	0.69^a	0.66	0.70^a	0.71^a	0.74^a
S5	—	—	—	—	—	—	—	—	1	−0.33	−0.33	−0.21	0.41	0.67	0.31	0.39	0.38	0.75^a	0.80^b	−0.49	−0.50	−0.43	0.19	0.46
LB1	—	—	—	—	—	—	—	—	—	1	0.99^b	0.99^b	0.95^b	0.47	0.86^b	0.84^b	0.82^b	0.32	−0.21	0.95^b	0.94^b	0.96^b	0.87^b	0.81^b
LB2	—	—	—	—	—	—	—	—	—	—	1	0.98^b	0.96^b	0.52	0.79^b	0.77^b	0.75^b	0.20	−0.29	0.96^b	0.95^b	0.97^b	0.89^b	0.83^b
LB3	—	—	—	—	—	—	—	—	—	—	—	1	0.96^b	0.61	0.86^b	0.85^b	0.84^b	0.42	0.20	0.94^b	0.94^b	0.95^b	0.87^b	0.81^b
LB4	—	—	—	—	—	—	—	—	—	—	—	—	1	0.77^b	0.83^b	0.82^b	0.82^b	0.56	0.51	0.86^b	0.85^b	0.88^b	0.82^b	0.78^b
LB5	—	—	—	—	—	—	—	—	—	—	—	—	—	1	0.48	0.54	0.60	0.71^a	0.81^b	0.28	0.28	0.36	0.35	0.44
TB1	—	—	—	—	—	—	—	—	—	—	—	—	—	—	1	0.99^b	0.98^b	0.74^a	0.62	0.75^a	0.72^a	0.73^a	0.62	0.57
TB2	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	1	0.99^b	0.79^b	0.67	0.73^a	0.7^a	0.71^a	0.65	0.57
TB3	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	1	0.78^b	0.70^a	0.70^a	0.66	0.68^a	0.60	0.51
TB4	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	1	0.96^b	−0.16	−0.2	−0.2	0.43	0.48
TB5	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	1	−0.46	−0.48	−0.42	−0.14	0.31
P1	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	1	1^b	0.1^b	0.94^b	0.87^b
P2	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	1	0.1^b	0.95^b	0.87^b
P3	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—		1	1^b	0.94^b
P4	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—			1	0.95^b
P5	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—				1

---

Fig. 6 Correlations of CST and (a) DOM content in four sludge layers; (b) biopolymers located in different EEM regions.

Additionally, it can be seen from Fig. 6(b) that CST correlated well with the aromatic protein I (R = 0.74, p < 0.05), aromatic protein II (R = 0.75, p < 0.05) and tryptophan protein (R = 0.87, p < 0.05) in soluble EPS, but not with humic acid and fulvic acid. Buntner et al. also suggested that SRF significantly positively correlated with protein content in soluble EPS.³¹ Lyko et al. also suggested the dewaterability of activated sludge was primarily dependent on concentration of dissolved macromolecular compounds,³² and proteins were always high MW substances. Excessive soluble EPS was found to be detrimental to cell cohesion and stability of sludge floc, leading to bad bioflocculation property, cell erosion and poor sludge–water separating performance. As mentioned above, it is interesting to note that the organic compositions in TB-EPS and pellet were relatively stable while that in soluble EPS and LB-EPS varied greatly under non-steady-state conditions.

3.5. Correlation of biopolymers composition located in four sludge layers

Table 5 showed a strong correlation between EEM regions I and II (R > 0.99, p < 0.01) for all sludge samples, indicating that their chemical composition were very similar to each other. Regions I and II were related to simple aromatic proteins such as tyrosine. Again, the fluorescent intensity (FI) of regions I and II could represent biochemical oxygen demand (BOD) of WWTP influent. Furthermore, it is worthy to note that both regions I and II were correlated well to region IV (R > 0.77, p < 0.05). It can be explained by the fact that more soluble EPS was produced at higher BOD concentrations. As mentioned earlier, water temperature negatively correlated well with the EPS content. As depicted in Fig. 5, regions I, II, IV and V were dominant fractions and accounted for more than 90% of total organic content in all floc layers. From April to July, the FRI of regions I and II decreased from 60% to 30% while that of regions IV and V was increased from 30% to 70% respectively. It is well known that the microbial activity was greatly influenced by water temperature. Since the degradation rates of BOD were lower for activated sludge at lower temperature (before April), the aromatic protein could not be rapidly utilised by microorganisms and as a result it will accumulate in sludge floc. After June, the sludge activity was improved with increase in temperature, thereby aromatic protein was more efficiently mineralized or converted into the humic substances through hydrolytic process under anaerobic condition.

4. Change in structural model of activated sludge with operating time

The soluble EPS is weakly attached to cells and dispersed, moving freely among the flocs. By contrast, bound EPS is sludge fraction firmly binding to microbial cells which were only can be removed by physiochemical methods. According to experimental results, we proposed a model of change in structural properties of activated sludge floc with time (Fig. 7). Temperature was found to be the major factor affecting EPS properties. At first, the EPS content reduced with increase in temperature and reached minimum in the summer, consequently the sludge floc strength was improved. Meanwhile, no obvious change in chemical composition pellet and TB-EPS were observed, while soluble EPS was highly changeable. Proteins and polysaccharide were reduced and humic acid greatly increased in the summer. Since the dewaterability was manly influenced by chemical characteristics of soluble EPS, the dynamic variation in soluble EPS resulted in fluctuation of dewatering behavior of activated sludge.

Fig. 7 Model of change in structural properties of activated sludge with operating time.

5. Technological implications

In general, the distribution and composition of EPS were significantly influenced by process flow, influent quality, operating conditions (hydraulic retention time (HRT), sludge retention time (SRT), temperature). The mass load of chemical oxygen demand (COD) to the plant showed a seasonal pattern with higher loading during the winter months. Correspondently, the composition of the activated sludge also showed a seasonal change and higher concentrations of extractable EPS, and the protein content of the total sludge and EPS also increased significantly during the winter.²³ Wang et al. also found that a seasonal variation of bound EPS concentrations with highest values at low temperature periods.³³ Higher COD/N ratios would result in increase of the content and decrease of protein content in EPS.³⁴ As mentioned above, sludge dewaterability negatively correlated with total organic and protein concentration of soluble EPS which were strongly dependent on operating conditions and influent qualities of WWTP. Thus, the operator can control the operating parameters to reduce the production of soluble EPS and provide the sludge with good dewaterability. For instances, Barker and Stuckey suggested that it appeared to decrease to a minimum and then increase again, indicating the existence of an optimal SRT and HRT for minimizing the soluble EPS production.³⁵ Furthermore, the low molecular weight polymers were mainly produced by bacteria in the logarithmic growth phase, which are representative components of the LB-EPS. Conversely, high MW polymers, the components of TB-EPS, are primarily formed in the stationary and decline phases. Therefore, the LB-EPS content decreased, and the TB-EPS content increased with increased SRTs. Biopolymers with high MW in soluble EPS were the determining factors of sludge dewaterability. Multivalent cations exhibited excellent binding capacity for EPS, hence they could absorb the sticky soluble EPS from water phase and enhance the floc strength of activated sludge through electrical neutralization and bridging. Many studies reported that sludge dewatering and filtration performance can be effectively improved by dosing inorganic salt coagulants, such as ferric and aluminum salts.^19,36

6. Conclusion

This study focused on discussing the effects of floc properties, especially dynamic variation of EPS distribution and composition in sludge from five WWTPs on sludge dewatering behavior. The following conclusions can be drawn from the experimental results:

• Overall, sludge dewatering property was not significantly influenced by the charge property, VSS/TSS ratio and pH of sludge system. Furthermore, there was only moderate correlation between CST and the amount of fine particles.

• The total EPS content of activated sludge was decreased significantly in the summer. The evolution in fluorescence characteristic of four sludge fractions from different WWTPs showed a very similar pattern with operating time. Additionally, the chemical composition of TB-EPS and pellet was relatively stable and protein-like compounds are the dominant fractions. Soluble EPS and LB-EPS also mainly contained protein-like substances before May while the protein content was decreased and humic-like substances began to appear and increased significantly in the summer.

• Sludge dewaterability negatively correlated significantly with concentration of protein-like substances and/or soluble EPS rather than chemical characteristics of other sludge fractions. Therefore, fluctuation of sludge dewatering performance in full-scale WWTP was primarily dependent on dynamic variation characteristics of soluble EPS during operation.

Acknowledgements

This study was financially supported by the State Water Project for Integrated Water Supply Sludge Quality and Depth of Dewatering Technology (2012ZX07408001-05) and National Natural Science Foundation of China (no. 51025830, 41201498).

References

1. H. Liu, J. Yang, N. Zhu, H. Zhang, Y. Li, S. He, C. Yang and H. Yao, J. Hazard. Mater., 2013, 258, 144.
2. M. Niu, W. Zhang, D. Wang, Y. Chen and R. Chen, Bioresour. Technol., 2013, 144, 337.
3. Y. Liu and H. H. Fang, Crit. Rev. Environ. Sci. Technol., 2003, 33, 237.
4. W. R. Knocke and D. L. Wakeland, Virginia Water Resources Research Center, Virginia Polytechnic Institute and State University, 1982.
5. B. Jin, B. M. Wilén and P. Lant, Chem. Eng. J, 2004, 98, 115.
6. L. H. Mikkelsen and K. Keiding, Water Res., 2002, 36, 2451.
7. X. Li and S. Yang, Water Res., 2007, 41, 1022.
8. G. H. Yu, P. J. He and L. M. Shao, Water Res., 2010, 44, 797.
9. G. H. Yu, P. J. He, L. M. Shao and P. P. He, Environ. Sci. Technol., 2008, 42, 7944.
10. J. I. Houghton, J. Quarmby and T. Stephenson, Water Sci. Technol., 2001, 44, 373.
11. M. J. Higgins and J. T. Novak, Water Environ. Res., 1997, 225–232.
12. S. N. Murthy and J. T. Novak, Water Environ. Res., 1999, 71, 197.
13. T. L. Poxon and J. L. Darby, Water Res., 1997, 31, 749.
14. S. F. Yang and X. Y. Li, Process Biochem., 2009, 44, 91.
15. C. Park and J. T. Novak, Water Res., 2007, 41, 1679.
16. A. Apha, Standard methods for the examination of water and wastewater, American Public Health Association, American Water Work Association, Water Environment federation (Wef), Washington DC, 20th edn, 1998, vol. 252.
17. X. H. Wang, Z. C. Wu, Z. W. Wang, X. Yin and X. X. Du, Bioresour. Technol., 2009, 100, 39–45.
18. C. P. Dancey and J. Reidy, Statistics without Maths for Psychology: using SPSS for Windows, Pearson Education, London, 2004.
19. M. R. Mehrnia, H. Azami and M. H. Sarrafzadeh, Colloids Surf., B, 2013, 109, 90.
20. B. Liao, D. Allen, G. Leppard, I. Droppo and S. Liss, J. Colloid Interface Sci., 2002, 249, 372.
21. Z. Zhu, T. Li, D. Wang and C. Yao, Chin. J. Environ. Eng., 2009, 2, 1599.
22. P. R. Karr and T. M. Keinath, J.-Water Pollut. Control Fed., 1978, 50, 1911.
23. B. M. Wilén, D. Lumley, A. Mattsson and T. Mino, Water Res., 2008, 42, 4404.
24. P. Lembre, C. Lorentz and P. Di Martino, The Complex World of Polysaccharides, InTech, p. 371, 2012.
25. F. Meng, S. R. Chae, A. Drews, M. Kraume, H. S. Shin and F. Yang, Water Res., 2009, 43, 1489.
26. W. Chen, P. Westerhoff, J. A. Leenheer and K. Booksh, Environ. Sci. Technol., 2003, 37, 5701.
27. G. P. Sheng and H. Q. Yu, Water Res., 2006, 40, 1233.
28. L. Xi, N. Lv, H. Zhang and B. Sun, CIESC J., 2013, 64, 3003.
29. Z. Wang, Z. Wu and S. Tang, Water Res., 2009, 43, 2504.
30. F. Meng, H. Zhang, F. Yang, S. Zhang, Y. Li and X. Zhang, Sep. Purif. Technol., 2006, 51, 95.
31. D. Buntner, H. Spanjers and J. van Lier, Water Res., 2014, 51, 284.
32. S. Lyko, T. Wintgens, D. Al-Halbouni, S. Baumgarten, D. Tacke, K. Drensla, A. Janot, W. Dott, J. Pinnekamp and T. Melin, J. Membr. Sci., 2008, 317, 78.
33. X. Wang, Z. Wu, Z. Wang, X. Yin and X. Du, Bioresour. Technol., 2009, 100, 1937.
34. B. Durmaz and F. Sanin, Water Sci. Technol., 2001, 44, 221.
35. D. J. Barker and D. C. Stuckey, Water Res., 1999, 33, 3063.
36. W. Zhang, P. Xiao, Y. Liu, S. Xu, F. Xiao, D. Wang and C. W. Chow, Sep. Purif. Technol., 2014, 132, 430.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c4ra13379j

---