Novel cake characteristics of waste-activated sludge

Printed in Great Britain 0043-1354/01/$ - see front matter

PII: S0043-1354(00)00513-3

RAPID COMMUNICATION

NOVEL CAKE CHARACTERISTICS OF WASTE-ACTIVATED

SLUDGE

R. M. WU

_{, D. J. LEE}

_{*, C. H. WANG}

_{, J. P. CHEN}

_{and R. B. H. TAN}

1_{Department of Chemical Engineering, National Taiwan University, Taipei 10617, Taiwan ROC and} 2

Department of Chemical and Environmental Engineering, National University of Singapore, 4, Engineering Drive 4, Singapore 117576, Singapore

(First received 29 May 2000; accepted in revised form 18 October 2000)

Abstract}Breaking down the time limit constraints for conventional compression–permeation (C–P) cell test, this work has, for the first time, experimentally evaluated the cake characteristics of viable waste-activated sludge subject to polyelectrolyte flocculation and to freeze/thaw treatment under a pressure range of 25–200 kPa. There exists a threshold pressure exceeding which the cake structure would significantly deteriorate. Also, the present biological sludge is a ‘‘super-compactible’’ sludge, whose compactibility is greater than most data ever reported in open literature. The information presented herein has implications to filter design/operation and can be used as a reference data set for examining the existing filtration theories. # 2001 Elsevier Science Ltd. All rights reserved

Key words}compression-permeability cell, porosity, specific resistance, yield stress

INTRODUCTION

The design and operation of mechanical dewatering

apparatus for sludge, such as the belt filter press or

the screw press, requires the knowledge of cake

characteristics changes when subject to pressure

compression and liquid permeation (Chang et al.,

1997; Chang and Lee, 1998). The local cake porosity

(e) and the specific filtration resistance (a) are the

most essential parameters in the applications of

filtration theory to filter design and operation. Ruth

(1946) introduced the compression–permeability cell

(the C–P cell) for measuring the porosity and the

specific resistance as functions of applied pressure (p).

Many studies employed the C–P cell for analyzing

the local properties in the filter cake, as briefly

summarized in Lu et al. (1998a, b) and in He et al.

(1997a). Despite the other technical drawbacks

claimed for the C–P cell tests, the major difficulties

to adopt the C–P cell tester in practice include the

long testing time, say, up to 2–4 weeks for obtaining a

complete set of data. Therefore, most C–P cell studies

employed inorganic substance as their testing

ma-terials, like clay or calcium carbonate, to prevent

possible quality changes during the relatively long

testing time.

Waste-activated sludge (WAS) is a mixture

con-taining bacteria and water, whose filter cake

com-monly exhibits high cake compressibility and a vast

amount of bound water (Wu et al., 1998). The

increase in the applied pressure difference to filter a

highly compactible sludge cake would not yield an

increasing filtrate rate (Tiller et al., 1999; Lee et al.,

2000). Mechanical dewatering is widely employed in

WAS dewatering practice. However, the C–P cell

data for the WAS are still largely lacking. Such a

drawback is mainly attributed to the serious time

limit for any biological sludge testing, for not

exceeding 5–7 days before quality deterioration

occurring (Sanin et al., 1994). With the same

reasoning, Kwon (1995) used formalin to ‘‘fix’’ his

biological sludge from degradation and completed

his C–P cell tests in 2 months. The effects of such a

chemical treatment on the sludge cake characteristics

are not clear.

He et al. (1997b) proposed a multifunctional test

cell that could conduct filtration and C–P cell tests in

the same apparatus. Since most of the sampling and

consolidation procedures were automated with the

assistance of a computer, one could perform each test

in a relatively short period time (6–12 h). Hence,

using this newly proposed C–P cell tester, the

collection of C–P cell data for biological sludge

becomes feasible. This work for the first time

reported the cake characteristics of WAS in both

compression and permeation stages. Effects of adding

polyelectrolyte and conditioned by freezing and

thawing method were investigated. The fitting

para-meters of certain constitutive equations were

eval-uated and tabulated. The information presented

herein could be used as a reference set of data in

*Author to whom all correspondence should be addressed. Fax: +886-2-2362-3040; e-mail: [email protected]

engineering design/operation or to check up with the

existing guidelines for sludge dewatering

manage-ment.

MATERIALS AND METHODS

The sample

Activated sludge samples (pH 6.8–7.4) were taken from the reflux stream of Ulu Pandan Sewage Treatment Works, Singapore. The design of the Works is based on conven-tional activated sludge process using diffused air aeration system, having a total treatment capacity of 286,000 m3_per day. The dry solid content of sludge, determined by weighing and drying at 1028C, was 0.24% w/w. The chemical oxygen demand for the supernatant (COD) and that for the total sludge (TCOD) are 103 and 8500 mg/L, respectively. After mixing and prior to settling, a small quantity of sludge-polymer aggregates in the vessel was transferred carefully into the fresh electrolyte at the same pH and electrolyte concentration as the original electrolyte. Zeta potentials of aggregates were then mea-sured by the zeta meter (Meter System 3.0, Zeter-Meter Inc., USA). The result for original sludge flocs read ÿ19.3 mV. The particle size distribution was deter-mined by Sedigraph 5100C (Micromeritics) as a

mono-dispersed distribution with a mean diameter of

approximately 145.7 mm. The true solid density was measured by Accupyc Pycometer 1330 (Micromeritics), giving a measure of 1378 kg/m3_{with a relative deviation of} less than 0.5%. Capillary suction apparatus as described in Lee and Hsu (1992, 1993) was employed to estimate the sludge filterability. The capillary suction time (CST) for original sludge is 62 s.

The sludge was subject to chemical or physical con-ditioning. For chemical conditioning, the cationic polyelec-trolyte indicated as polymer T-3051 was obtained from Kai-Guan Inc., Taiwan. The polymer T-3051 is a polyacryla-mide with an average molecular weight of 107_{, and a charge} density of 20%. The mixing unit was a magnetic stirrer. The weighed powder was first suspended in distilled water. Solution of the polymer solution was then gradually poured into the mixing vessel with 500 rpm of stirring for 25 min. The settleability of the chemical-conditioned sludge was determined using hindered settling tests performed in tubes of diameter 2.5 cm and height 18.5 cm. The zone settling velocity (ZSV) could be obtained by linear regression of the interface height versus time data for the constant-rate period with a regression coefficient higher than 0.98. Other experimental details could be found in Chen et al. (1996). The ZSVs for original sludge, and for those conditioned at 50, 100 and 200 ppm of polyelectrolyte are 800, 930, 4880, and 2810 mm/s, respectively. According to the settling test, the application of polyelectrolyte could markedly enhance the zone settling velocity, while the so-called ‘‘optimal dose’’ for the present activated sludge could be identified as around 100 ppm of polyelectrolyte.

Freeze/thaw conditioning is an efficient method of changing floc structure and reducing the bound water content in sludge (Lee and Hsu, 1994). The sample is placed in a stainless-steel vessel measuring 25 cm in diameter, 0.15 cm thick and 25 cm high. The vessel is immersed for 48 h in a freezing pool at a temperature ofÿ158C. After freezing, the sample was thawed at room temperature for another 12 h. Such an experimental condition was chosen for providing sufficiently low freezing speed for sufficient conditioning of the sludge (Hung et al., 1997).

C–P cell and test

Figure 1 illustrates the schematic diagram of the C–P cell and other supportive apparatus. The load at the top (p) and

the transmitted pressure to the bottom surface (pT) are

measured together with the cake height during each test. The cell has a cylinder made of stainless steel of inner diameter 75.3 mm. A complete C–P cell test comprises two stages: the compression stage and the permeation stage. The cell could track both stages since all data from pressure transducers as well as from the displacement measurement were automatically sent to a computer for storage and processing.

Prior to the C–P cell test the septum was first filled with filtrate. The slurry was carefully poured into the cylinder and drained to form a saturated, wet cake. The piston was positioned at the top of the formed cake, through which the mechanical force was applied. During the compression the valve-A (permeation valve) was close and valve-B open, which allows the drainage of the filtrate. Before and after having reached mechanical equilibrium with the applied load, the thickness of compressed cake was continuously measured and recorded. Then the permeation test was conducted by allowing the filtrate to flow from a constant-head reservoir through the valve-A and the cake. An electronic balance measured the filtrate weight. With the flow rate and the pressure drop data, the specific resistance of filtration of cake could be determined. The temperatures during testing were used to correct the viscosity of filtrate.

The pressure range under investigation is 25–200 kPa. The choice of such a pressure range for test is attributed to the following two reasons. Firstly, in preliminary tests at the applied pressure of 50 kPa, the cake structure of activated sludge reveals a significant collapse, which had not occurred at the test of the less pressure. Hence the lower limit for applied pressure under investigation is set at 25 kPa. Moreover, since the total testing time has to be limited within 7 days, the upper pressure limit was taken as 200 kPa. Although the practical range for sludge dewatering could be up to 500–700 kPa, however, as the present experimental data illustrated, the basic characteristics for cake properties would remain unchanged at pressures exceeding 200 kPa.

RESULTS AND DISCUSSION

Compression test

Figure 2 illustrates the time evolutions for the cake

thickness. Apparently, at p¼ 25 kPa, the cake could

be only mildly compressed. For the original sludge,

the cake reaches equilibrium at around 30,000 s. For

flocculated sludges, on the other hand, the

mechan-ical equilibrium has been established rather rapidly.

Also, the cake thickness reduces for less than 10%.

At a higher load of 50 kPa, the cake structure starts

to collapse markedly. The reduction in cake thickness

could be as high as 80–90%. Although the

con-solidation times needed for flocculated sludge are still

much less than the original sludge, they are 3–5 times

longer than those at 25 kPa. Further increase in the

applied pressure would continuously compress the

cake. However, no qualitative cake characteristic

changes had been observed at the elevated pressures.

Such an observation indicates that there exists

a threshold pressure exceeding which the cake

structure would significantly deteriorate, attributed

to the network structure built over the entire sludge.

The compaction characteristics for the present

WAS under low applied load (like in sedimentation)

and for that under higher load (like filtration,

centrifugation, and consolidation) would be very

different.

The trend for the consolidation curves less than the

threshold pressure is opposite to that for greater than

the threshold pressure. At 20 and 50 kPa, the cake

becomes the stiffest at the dosage of 100 ppm.

However, exceeding the threshold pressure, the trend

reverses. The optimal dosage would yield the fastest

dewatering, which corresponds to the previous

studies with clay slurries (Chu and Lee, 2000). The

cake compaction at applied load less than the

threshold pressure would behave very differently

from those at elevated pressures.

Permeation test

Figures 3(a) and (b) illustrate, respectively, the

measured e

ÿ p

and the a

ÿ p

data, where p

is

the log-mean pressure difference defined as

p

¼

p

ÿ p

ln p=p

ð

Þ

Firstly, the cake porosity (e) decreases with

increas-ing solid pressure, which correlates with the common

knowledge that a higher pressure leads to a more

compacted cake (Fig. 3(a)), whence a greater specific

resistance of the filter cake (a) (Fig. 3(b)). Secondly,

at a prescribed pressure, although with some data

scattering, the cake porosity first decreases with

polyelectrolyte dose, after reaching minimum at the

optimal dose (100 ppm), and then increases in the

overdosing regime. Restated, the presence of

poly-electrolyte would yield a more compact structure but

easier to permeate sludge cake. Such an observation

should be attributed to the existence of a vast amount

of bound water that changes the ‘‘effective’’ porosity

for the filter cake (Lee and Hsu, 1995). For original

sludge, for example, the effective porosity of cake

would be much less than demonstrated if the bound

water has been taken into account. Such an

information is not available for the present C–P cell

test. Finally, the freeze/thaw treatment would yield a

more compact and easier to permeate sludge cake

when compared with the original sludge. The effects

are not as significant as that for chemical

condition-ing for the present WAS.

Fig. 2. The time evolutions for cake thickness under axial loads: (a) p=0.25 105

Pa; (b) p=0.5 105 Pa; (c) p=1 105

Pa; (d) p=2 105_{Pa. The initial cake thickness (L}

To correlate the measured cake characteristics

as functions of applied pressure, like porosity and

specific resistance to filtration, various constitutive

equations had been previously proposed in the

literature (Lee and Wang, 2000). Tiller and Leu

(1980) proposed the following constitutive equations

as follows:

a

¼ a

1

þ

p

p





ð1Þ

1

ÿ e ¼ 1 ÿ e

ð

Þ 1 þ

p

p





ð2Þ

k

¼ k

1

þ

p

p





ð3Þ

In Equations (1)–(3), e

, a

and k

are the cake

porosity, specific resistance, and permeability under

null-stress condition, and p

, b, n, and d are the fitting

parameters. Notably, d

¼ b þ n for a specific cake.

Therefore, only two of the three equations in

Equations (1)–(3) are independent. Tiller and Leu

(1980) proposed a graphic method to determining all

parameters in Equations (1) and (2).

Table 1 lists the best-fitting parameters, showing a

d value greater than 3.6. The presence of

polyelec-trolyte would yield an even greater d value (exceeding

4.0). Literature works regarded the sludge cake with

d > 1 as ‘‘highly compactible’’ (Tiller and Kwon,

1998). This work demonstrates that the present WAS

is a ‘‘super-compactible’’ sludge. In reality, the

compactibility noted for this WAS is much higher

than the most sludges ever reported. For example, La

Heij (1994) estimated d

¼ 2:3 and Kwon (1995) gave

d

¼ 1:66 for their activated sludges. Such a

discre-pancy might be arisen from the quality changes for

La Heij and Kwon’s sludges during the C–P cell test.

The present C–P cell could provide the cake

characteristics for viable activated sludge within a

relatively short period of time.

The recognition of the ‘‘super-compactibility’’ of

the WAS suggests that, owing to the formation of a

‘‘skin layer’’ close to the septum, to simply increase

the applied pressure would not yield a greater filtrate

flow rate. The basic equations to design and

operation of filters would hence be very different

from those adopted for those of low-to-medium

compactibility (Tiller and Kwon, 1998, Tiller et al.,

1999; Lee et al., 2000). In filtration chamber, a

low-but-sufficient pressure drop should thereby be

adopted. Simply raising the applied pressure could

not help in enhancing the filter performance.

CONCLUSIONS

All tests regarding biological sludge have to be

completed within 5–7 days for preventing the quality

deterioration. However, a conventional

compres-sion–permeation (C–P) cell test commonly required

a relatively long period of time. With the assistance

Fig. 3. (a) Porosity (e) versus pressure of the filter cake. (b) Specific filtration resistance (a) versus pressure of the filter cake.

Table 1. Model parameters in equations (1)–(3). d¼ bþn

Polymer dose b n d pa e0 a0 (kPa) (1010 m/kg) Original 0.83 2.7 3.5 3.0 0.993 5.0 50 ppm 1.10 2.9 4.0 3.3 0.995 3.0 100 ppm 0.88 3.1 4.0 2.6 0.994 0.1 200 ppm 1.10 >3.1 >4.2 2.6 0.995 NA F/T 0.78 2.8 3.6 2.5 0.993 0.6

of a newly proposed C–P cell by He and co-workers,

this work has, for the first time, experimentally

evaluated the cake characteristics of waste-activated

sludge subject to polyelectrolyte flocculation and to

freeze/thaw treatment under a pressure range of 25–

200 kPa. Compression tests indicate the existence of a

threshold pressure exceeding which the cake structure

would be significantly deteriorated. The compaction

characteristics for the waste activated sludge under

low applied load and for that under higher load

would be very different.

In permeation test, the cake porosity was noted to

decrease with increasing applied load, whence

yield-ing a greater specific resistance of the filter cake.

However, the presence of polyelectrolyte would lead

to a more compact structure but easier to permeate

sludge cake. The effects of freeze/thaw treatment are

milder than the chemical conditioning adopted

here-in. Using the correlation proposed by Tiller and Leu

(1980), the present WAS was identified as a

‘‘super-compactible’’ sludge, whose compactibility was

greater than the most data ever reported in the

literature. The chemical conditioning would further

increase its compatibility. Thus, the cake

character-istics of the present biological sludge would shed out

the benefit for raising the applied pressure in

filtration practice. The information presented herein

could be used as a reference set of data in engineering

design/operation or to check up with the existing

guidelines for sludge dewatering management.

Acknowledgements}DJL wishes to thank National Uni-versity of Singapore for appointing him as a senior fellow during June to September, 1999.

REFERENCES

Chang I. L., Chu C. P., Lee D. J. and Huang C. (1997) Effects of polymer dose on filtration followed by expression characteristics of clay slurries. J. Colloid Interface Sci. 185, 335–342.

Chang I. L. and Lee D. J. (1998) Ternary expression stage in biological sludges dewatering. Water Res. 32, 905–914. Chen G. W., Chang I. L., Hung W. T. and Lee D. J. (1996)

Regimes of zone settling of waste activated sludge. Water Res. 30, 1844–1851.

Chu C. P. and Lee D. J. (2000) Expression characteristics of polyelectrolyte flocculated sludges. J. Chin. Inst. Chem. Engrs. 31, 321–331.

He D. -X., Tan R. B. H. and Tien C. (1997a) An overview of investigations on filter cake characteristics. Adv. Filtr. Sep. Technol. 11, 404–411.

He D.-X., Tan R. B. H. and Tien C. (1997b). An investigation of the filter cake characteristics in a modified compression–permeability cell. 1997 A.I.Ch.E. Annual Meeting, Los Angeles, USA.

Hung W. T., Feng W. H., Tsai I. H., Lee D. J. and Hong S. G. (1997) Unidirectional freezing of waste activated sludge: radial freezing versus vertical freezing. Water Res. 31, 2219–2228.

Kwon J. H. (1995) Effects of compressibility and cake clogging on sludge dewatering characteristics. Ph.D. Dissertation, Seoul National University, Seoul, Korea. La Heij E. J. (1994) An analysis of sludge filtration and

expression. Ph.D. Dissertation, Technische Universiteit Eindhoven, Eindhoven, The Netherlands.

Lee D. J. and Hsu Y. H. (1992) Fluid flow in capillary suction apparatus. Ind. Eng. Chem. Res. 31, 2379–2384. Lee D. J. and Hsu Y. H. (1993) Cake formation in capillary

suction apparatus. Ind. Eng. Chem. Res. 32, 1180–1185. Lee D. J. and Hsu Y. H. (1994) Fast freeze/thaw process on

excess activated sludges: Floc structure and sludge dewaterability. Environ. Sci. Technol. 28, 1444–1449. Lee D. J. and Hsu Y. H. (1995) Measurement of bound

water in sludges: a comparative study. Wat. Environ. Res. 67, 310–317.

Lee D. J., Ju S. P., Kwon J. H. and Tiller F. M. (2000) Filtration of highly compactible filter cake: variable internal flow rate. A.I.Ch.E. J. 46, 110–118.

Lee D. J. and Wang C. H. (2000) Theories of cake filtration and consolidation and implications to sludge dewatering. Water Res. 34, 1–20.

Lu W. M., Huang Y. P. and Hwang K. J. (1998a) Methods to determine the relationship between cake properties and solid compressive pressure. Sep. Purif. Technol. 13, 9–23. Lu W. M., Huang Y. P. and Hwang K. J. (1998b) Stress distribution in a confined wet cake in the compression– permeability cell and its application. Powder Technol. 97, 16–25.

Ruth B. F. (1946) Correlating filtration theory with industrial practice. Ind. Eng. Chem. 38, 564–571. Sanin F. D., Vesilind P. A. and Martel C. J. (1994)

Pathogen reduction capability of freeze/thaw sludge conditioning. Water Res. 28, 2393–2398.

Tiller F. M. and Kwon J. H. (1998) Role of porosity in filtration}XIII}behavior of highly compactible cakes. A.I.Ch.E. J. 44, 2159–2167.

Tiller F. M. and Leu W. F. (1980) Basic data fitting in filtration. J. Chin. Inst. Chem. Engrs. 11, 61–70. Tiller F. M., Lu R., Kwon J. H. and Lee D. J. (1999)

Variable liquid flow rate in compactible filter cakes. Water Res. 33, 15–23.

Wu R. M., Feng W. H., Tsai I. H. and Lee D. J. (1998) An estimate of waste activated sludge floc permeability: a novel hydrodynamic approach. Water Environ. Res. 70, 1258–1264.