Colloidal silica removal in coagulation processes for

Water Science & Technology Vol 55 No 1–2 pp 187–195 Q IWA Publishing 2006
wastewater reuse in a high-tech industrial park
S.H. Chuang*, T.C. Chang**, C.F. Ouyang*** and J.M. Leu*
*Department of Environmental Engineering and Management, Chaoyang University of Technology, Wufeng,
Taichung County 413, Chinese Taiwan (E-mail: [email protected])
**Institute of Environmental Planning and Management, National Taipei University of Technology, Taipei 106,
Chinese Taiwan
***Institute of Environmental Engineering, National Central University, Chungli 320, Chinese Taiwan

Abstract Four experiments of coagulation and flocculation were conducted to investigate the characteristics
of colloidal silica removal in a high-tech industrial wastewater treatment plant for reclamation and reuse of
the effluent. Experimental results illustrated that poly-aluminium chloride (PACl) showed higher performances
on colloidal silica removal than alum. Interestingly, the two coagulants demonstrated the same capacity on
silica removal. The specific silica removal capacity was approximately 0.135 mg SiO2/mg Al2O3 when the
dosage of coagulants was in the range 30 –150 mg/L Al2O3. In addition, the silica was reduced significantly
at the condition of pH above 8. Experimental data implied that precipitation of aluminium flocs was the major
mechanism for colloid silica removal in PACl and alum coagulation, besides, charge adsorption was also
important for improving removal efficiency. Moreover, the addition of polyacrylic acid (PAA) as a flocculant
could slightly advance silica removal in the PACl coagulation. The combined PACl/PAA/flocs coagulation
was effective for the removal of colloidal silica, soluble COD, and turbidity and also suitable as a
pretreatment unit in wastewater reclamation and reuse processes.
Keywords Alum; coagulation; colloidal silica; poly-aluminium chloride; wastewater reuse

Introduction
The occurrence of high concentrations of colloidal silica in wastewater is a critical con-
straint for reclamation and reuse of the effluent in a wastewater treatment plant or in
waste cooling, especially in a high-tech industrial park. In wastewater, amorphous silica
can usually be classified into soluble, colloidal and particulate forms depending on its
reactivity and suspension (Sheikholeslami and Tan, 1999). In our previous study at a
high-tech industrial park, the concentration of silica in the effluent of the wastewater
treatment plant was in the range 23.1 –42.9 mg/L; moreover, over 95% of silica was in
the colloidal category (Chuang et al., 2005). Based on silica chemistry, the solubility of
colloidal silica in pure water is limited to the range of 100–140 mg/L at 25 8C and at pH
below 8 (Roque, 1996), the soluble species being in the form of Si(OH)4. Owing to the
limited solubility of silica in water, the standard industrial guideline has been to limit the
maximum silica concentration in the retentate to about 120 mg/L at 25 8C in order to
reduce the scaling and fouling problems in reverse osmosis (RO) (Freeman and Majerle,
1995). Therefore, this limitation becomes an indispensability of reclamation and reuse of
silica-rich wastewater or waste cooling.
The chemistry of silica fouling is very complex; amorphous silica can foul the
membranes by monomeric precipitation, deposition of colloidal particles or as biogenic
amorphous silica (Iler, 1979; Sheikholeslami and Tan, 1999). Clearly, pretreatment is
indeed required to reduce silica content and to increase the recovery of RO membranes in
reclamation and reuse processes for silica-rich wastewater or in waste cooling. Various
doi: 10.2166/wst.2007.054 187
 treatment processes can be used for the removal of silica from water, such as a lime
softener, hot process softener, anion exchange, membrane filtration, etc. Of these, chemi-
cal clarification is one of the most practical units employed as pretreatment for membrane
desalination (Masarwa et al., 1997; Sheikholeslami et al., 2001, 2002; Al-Rehaili, 2003).
Chemicals such as lime, soda ash and iron salts are used extensively for silica removal in
softening processes. The research of Al-Rehaili (2003) showed that the addition of
precipitation aids such as alum, sodium aluminate or ferric chloride could improve
S.H. Chuang et al.

performance of the lime-soda ash process and reduce the content of silica in groundwater.
Also, anionic polymer was found helpful for silica reduction at a low dosage. However,
Sheikholeslami et al. (2002) reported that sodium aluminate alone was not able to reduce
the silica concentration significantly; moreover, the addition of soda ash only slightly
improved the efficiency. In addition, the study of Duan and Gregory (1996) concluded
that coagulation by aluminium salts can be greatly affected by quite low levels of soluble
silica; also, the effect is highly dependent on pH. However, studies dealing with silica
removal in coagulation with aluminium salts in wastewater reported in the literature
are limited. Therefore, the behaviour of silica removal by aluminium salts needs more
comprehensive studies in the laboratory and field.
The study described in this article was part of extensive research aimed at wastewater
reclamation and reuse in a high-tech industrial park. The objectives of this article were to
investigate the characteristics of chemical coagulation processes when using aluminium
salts in silica removal in the effluent of a wastewater treatment plant, and moreover, to
establish the optimum conditions of coagulation for further application.

Materials and methods

Experimental methods

Coagulation and flocculation. Four different experiments, namely (1) single

coagulation, (2) single flocculation, (3) combined coagulation and flocculation, and (4)
combined coagulation and flocs flocculation, were designed and carried out in a modified
jar test apparatus for investigating the characteristics of colloidal silica removal. Jar tests
were conducted in the laboratory at room temperature (25–28 8C) using 1 L plastic jars.
Coagulation experiments were performed with rapid mixing at 150 rpm for 5 min with a
dosage of 0–180 mg/L Al2O3 of coagulants. Flocculation experiments were operated with
slow mixing at 20 rpm for 20 min with a dose of 0 –8 mg/L of polymer. During
coagulation and flocculation processes, the pH was adjusted to an appointed value using
diluted sulphuric acid or sodium hydroxyl. Experiments of flocculation with addition of
flocs was designed to investigate the effectiveness of flocs on silica and COD removal;
the flocs was produced from previous coagulation; moreover, it was concentrated
gravitationally and kept in a non-disturbed condition for use. Samples were collected and
filtered after 30 min settling through 0.45 mm glass fibre filter paper for total silica and
colloid silica analysis.

Analysis methods. The total silica, colloidal silica, soluble COD, turbidity, total solid
and silt density index (SDI) were used to evaluate the performances in this study. The
silica concentration was determined immediately using a spectrophotometer (Hitachi
U2001), according to the molybdosilicate method (Standard Method, 1995). The SDI
measurement conformed to the ASTM D4189-95 standards at fixed 30 psi with a time
188 interval of 5 min.
Materials
Raw water. Table 1 shows the average characteristics of the effluent of the secondary
wastewater treatment plant in a high-tech industrial park. The effluent was collected
monthly for a series of jar tests during the experimental period. Table 1 shows that the
effluent was neutral with a pH of 6.8 ^ 0.1. The silica concentration varied from
23.1 ^ 15.5 to 27.9 ^ 10.7 mg/L; moreover, over 95% of silica was in the form of
colloid. In addition, the average of the COD was lower than 25.0 mg/L; chloride and

S.H. Chuang et al.

sulphate were the major components of anions in the effluent.

Reagents. Two commercial coagulants were used for coagulation experiments,

including PACl (polyaluminium chloride, Al2(OH)nCl62n, 30% Al2O3) and alum
(aluminium sulphate, Al2(SO4)3†18H2O, 98% Al2O3). Two reagent grade flocculants
were selected for flocculation tests, namely PAA (polyacrylic acid, (C3H5NOC3H4O2)n,
molecular weight 1,500–5,000) and PDADMA (polydiallyl dimethyammonium chloride,
(C8H16NCl)n, molecular weight 200,000–350,000). Plastic reactors and other reserved
containers were used to prevent possible leaching of silica from glass.

Results and discussion

Effects of coagulants, flocculants and pH on colloidal silica removal
All single coagulation and flocculation experiments were carried out at a pH of 7.5 in jar
tests. Table 2 summarises the performances of coagulation by PACl and alum. It shows
that the concentration of colloid silica was significantly reduced to 14.0 mg/L at a PACl
dosage of 150 mg/L Al2O3; moreover, the removal ratios were increased with the dosage
of coagulants. Compared with alum, the coagulant PACl demonstrated a higher achieve-
ment on colloid silica removal in the dosage range of 0–150 mg/L Al2O3. A previous
study on coagulation of aluminium salts showed that alum and PACl could precipitate to
form different solid phases of flocs for colloidal removal (Van Benschoten and Edzwald,
1990a, b). Apparently, the polymeric structure of PACl could provide a more positively
charged surface within the PACl precipitates and colloid, thus producing a lower silica
concentration than the alum flocs in this study.
From another point of view, Figure 1 illustrates that a positive linear relationship
between coagulants dosage and removed silica was established with a good correlation.
This demonstrates a very interesting occurrence of coagulation of the effluent. Although
PACl showed higher performance on overall colloidal silica removal than alum, the
specific silica removal capacity of the two coagulants was almost the same. It was
approximately 0.135 mg SiO2/mg Al2O3 when the dosage of coagulants was in the range
30–150 mg/L Al2O3. This phenomenon implied that the precipitation of aluminium flocs

Table 1 Average characteristics of the effluent of the secondary wastewater treatment plant

Parameter Feb –April May– Nov

Mean Std. dev. Mean Std. dev.

pH – 6.8 0.1 6.8 0.1

SS mg/L 13.9 6.5 27.5 23.4
Turbidity NTU 2.01 1.59 5.14 1.90
TDS mg/L 1,773 200 1,789 218
Conductivity mS/cm 1,709 586 1,934 30
Hardness mg/L as CaCO3 184.1 113.6 119.0 23.0
SiO2 mg/L 23.1 15.5 27.9 10.7
COD mg/L 21.1 3.3 23.2 3.7
Cl2 mg/L 706 476 621 93
SO22
4 mg/L 360 34 406 99
189
 Table 2 Performances of coagulants on silica removal

Dosage of coagulant PACl Alum

(mg/L Al2O3)
T. SiO2* C. SiO2** Removal ratio T. SiO2 C. SiO2 Removal ratio
(mg/L) (mg/L) (%) (mg/L) (mg/L) (%)

0 42.3 41.0 0.0 42.3 41.0 0.0

30 30.6 29.9 27.2 38.2 37.2 9.2
S.H. Chuang et al.

60 26.5 25.4 38.1 36.4 35.8 12.6

90 22.6 21.9 46.5 32.0 31.4 23.3
120 17.5 16.6 59.6 25.6 24.1 41.2
150 14.3 14.0 65.9 23.3 23.0 43.9
*T. SiO2 represents the total silica; **C. SiO2 represents the colloid silica

should be the major mechanism of coagulation for colloid silica removal at a neutral
environment of pH 7.5; however, the higher charged precipitates of PACl can improve
silica removal with charge neutralisation. Figure 2 illustrates the characteristics of single
flocculation with cation polymer (PDADMA) and anion polymer (PAA). Neither
PDADMA nor PAA could reduce colloid silica content, regardless of charge type and
molecular weight.
The pH is one of the important parameters for colloidal silica control. Figure 3 shows
the results of coagulation at different pH levels with the initial silica concentration of raw
water at 36.2 mg/L. In experiments without adding coagulant, the variation of silica was
slight at pH below 8; however, when pH increased to 9, the silica concentration dropped
significantly to 21.8 mg/L. Coagulation experiments with adding PACl 60 mg/L Al2O3
and PAA 0.5 mg/L also showed similar behaviour; moreover, the silica concentration was
reduced to about 10.4 mg/L at a pH of 9. Effects of pH on silica solubility indicate that
the dissociation constants of amorphous silica increases significantly when pH is above 8;
it will ionise and form more silicate ions (Roque, 1996). Therefore, with the pH below 8,
the precipitation of PACl precipitates with amorphous silica performed an essential mech-
anism for colloidal silica removal; otherwise, the ionised silicate ions could be adsorbed
on the high positively charged PACl precipitates enhancing silica removal when pH was
controlled above 8.
Conductivity was an important parameter for selecting wastewater reclamation and
reuse processes. Figure 4 displays a linear relationship between coagulant dosage and
increased conductivity. The coagulant PACl contributed more electrolytes than alum in
solution. The conductivity of the effluent used for this study was in the range 2,750–
3,080 mS/cm, which is a typical quality of industrial wastewater at a treatment plant.

190 Figure 1 Relationship between coagulants dosage and removed colloid silica
 S.H. Chuang et al.
Figure 2 Characteristics of colloid silica removal in flocculation

Figure 4 illustrates that the addition of 100 mg/L Al2O3 of coagulants PACl and alum
would increase conductivity of 174 and 96 mS/cm, respectively; it was acceptable for
wastewater reclamation.

Effects of PAA on colloidal silica removal

All of the combined coagulation and flocculation experiments were also accomplished at
pH of 7.5. Figures 5 and 6 illustrate the effects of flocculant PAA on silica removal in
PACl and alum coagulation processes, respectively. In PACl coagulation, the addition of
the flocculant PAA slightly increased silica removal ratios when the dosage was in the
range 0.5–3 mg/L. The removal ratio was approximately 80% at the conditions of PACl
dosage of 180 mg/L Al2O3 and PAA of 3.0 mg/L, and it seemed to approach a maximum
of silica removal at a pH of 7.5.
However, compared with PACl, alum coagulation experiments demonstrated a dissim-
ilar behaviour on silica removal. The effectiveness of the addition of flocculant PAA was
not clear; moreover, the removal ratios of silica were increased linearly with the dosage
of alum. The removal ratios were in the range 40 –50% at an alum dosage of 180 mg/L
Al2O3 and PAA of 0.5–3.0 mg/L.
From the results of the coagulation and flocculation experiments, it seems that the
positively charged aluminium precipitates initially contributed a chance for charged neu-
tralisation of amorphous silica on the surface of precipitates. The highly charged PACl
precipitates demonstrated higher performance on silica removal than alum and produced

Figure 3 Characteristics of colloid silica removal in different pH 191

S.H. Chuang et al.

Figure 4 Relationship between coagulants dosage and increased conductivity

lower concentrations of silica. Then, when the complex flocs of aluminium precipitates
and amorphous silica was formed, precipitation of the complex flocs became the major
mechanism of silica removal; meanwhile, the flocs developed from PACl or alum showed
the same specific capacity on silica reduction. Moreover, adjusting pH to above 8 for
increasing silica ionisation was one of the effective operations for improving colloidal
silica removal in the aluminium coagulation processes.

Effects of addition of waste flocs on colloidal silica and COD removal

Experiments with the addition of waste flocs in PACl coagulation were designed to inves-
tigate the extended performances of waste flocs on silica and COD removal at a pH of 8
and PACl of 60 mg/L Al2O3. Figure 7 shows that the addition of waste flocs could mod-
erately increase the efficiency of colloidal silica removal with about 8.8% when the
dosage of flocs was at 90 mg/L; meanwhile, the soluble COD removal ratios were
increased by approximate 22.0%. However, overdosage of waste flocs at 120 mg/L
resulted in the decrease of silica removal; contrarily, soluble COD removal was increased
significantly. It revealed that the precipitation of waste PACl –silica complex flocs
demonstrated more effectiveness on organic matters removal than on colloid silica
reduction. From the view of application, the silica illustrated good potential as a coagu-
lation aid for improving organics reduction. Therefore, addition of an adequate amount of
flocs could be useful for moderately extending silica and COD removal.

192 Figure 5 Performance of PACl coagulation and PAA flocculation on colloid silica removal
 S.H. Chuang et al.
Figure 6 Performance of alum coagulation and PAA flocculation on colloid silica removal

Table 3 summarises the performances of combined PACl/PAA/flocs coagulation at

different pH levels with PACl of 60 mg/L Al2O3, PAA of 0.5 mg/L, and flocs of 90 mg/L.
This demonstrates that single coagulation of PACl was useful for the removal of turbid-
ity, colloid silica and solids, although the fouling index SDI5 was still higher than 20.
The combined PACl/PAA/flocs coagulation showed remarkable improvement on soluble
organics removal and fouling index; nevertheless, the SDI5 was still very high and the
treated water was still inadequate to flow into the membrane process for further recla-
mation. Besides, the addition of waste flocs resulted in a slight increase on turbidity and
total solids.
Figure 8 illustrates the comparisons of performances of combined PACl/PAA/flocs
coagulations under the conditions of pH 7 and 8. It shows that experiments had excellent
results on turbidity reduction in different pH situations. Compared with the experiment at
pH 7, the experiment at pH 8 demonstrated a better performance on colloidal silica and
soluble COD removal; the removal ratios increased approximately 10 –12%. However,
the performance of total solids was reduced in the condition of pH 8. In summary, the
combined PACl/PAA/flocs coagulation was suggested as a pretreatment unit for improv-
ing the removal of turbidity, silica and SCOD in wastewater reclamation processes; the
optimal conditions of operation were determined at PACl of 60 mg/L Al2O3, PAA of
0.5 mg/L, waste flocs of 90 mg/L and pH of 8 in this study.

Figure 7 Effects of addition of waste flocs on colloid silica and SCOD removal 193
 Table 3 Performances of combined PACl/PAA/flocs coagulation

PACl/PAA/flocs pH Cond. Turbi. Colloid SiO2 TS TCOD SCOD SDI5

(mg Al2O3/mg/mg) (mS/cm) (NTU) (mg/L) (mg/L) (mg/L) (mg/L)

Raw water 7.24 2900 7.1 34.0 1980 92.2 73.5 .20
60/0/0 7 2950 0.2 22.0 1088 74.8 56.4 .20
60/0.5/0 2950 0.2 20.1 772 63.7 47.5 17.5
60/0.5/90 2930 0.3 20.3 820 44.2 31.5 15.1
S.H. Chuang et al.

60/0/0 8 2950 0.3 17.5 828 59.2 39.3 .20

60/0.5/0 2950 0.2 16.8 820 47.6 31.3 17.8
60/0.5/90 2930 0.4 16.6 969 33.1 22.9 16.1

Figure 8 Comparisons of coagulation at the conditions of PACl 60 mg/L Al2O3, PAA 0.5 mg/L and waste
flocs 90 mg/L

Conclusion
The objectives of this study were to investigate the characteristics of colloidal silica
removal in chemical coagulation processes by using aluminium salt, and moreover, to
establish the optimum conditions of coagulation for further application on wastewater
reclamation and reuse in a high-tech industrial park. Based on the results of this research,
conclusions were made as follows:
(1) The PACl showed higher performances on colloidal silica removal than alum. Inter-
estingly, the two coagulants demonstrated the same capacity on silica removal. The
specific silica removal capacity was approximately 0.135 mg SiO2/mg Al2O3 when
the dosage of coagulants was in the range 30–150 mg/L Al2O3.
(2) Adjusting pH to above 8 for increasing silica ionisation was one of the effective oper-
ations for improving colloidal silica removal in aluminium coagulation processes.
(3) The precipitation of aluminium flocs was the major mechanism for colloid silica
removal in PACl and alum coagulation; besides, charge adsorption was also import-
ant for improving removal efficiency. Moreover, the addition of polyacrylic acid
(PAA) as a flocculant could slightly advance silica removal in the PACl
coagulation.
(4) Combined PACl/PAA/flocs coagulation was suggested as a pretreatment unit for
improving the removal of turbidity, silica and SCOD in wastewater reclamation pro-
cesses; the optimal conditions of operation were determined at PACl of 60 mg/L
194 Al2O3, PAA of 0.5 mg/L, waste flocs of 90 mg/L and a pH of 8 in this study.
References
Al-Rehaili, A.M. (2003). Comparative chemical clarification for silica removal from RO groundwater feed.
Desalination, 159, 21 –31.
ASTM, D4189-95. Standard Test Methods for Silt Density Index (SDI) of Water.
Chuang, S.H., Chang, T.C., You, S.J. and Ouyang, C.F. (2005). Evaluation of wastewater reclamation
processes in a high-tech industrial park. Desalination, 175, 143 – 152.
Duan, J. and Gregory, J. (1996). Influence of soluble silica on coagulation by aluminium sulphate. Colloid.

S.H. Chuang et al.

Surface. A, 107, 309 – 319.
Freeman, S.D.N. and Majerle, R.J. (1995). Silica fouling revisited. Desalination, 103, 113 – 115.
Iler, R.K. (1979). The Chemistry of Silica, Wiley, New York, Chapter 1, 3.
Masarwa, A., Meyerstein, D., Daltrophe, N. and Kedem, O. (1997). Compact accelerated precipitation
softening (CAPS) as pretreatment for membrane desalination II. Lime softening with concomitant
removal of silica and heavy metals. Desalination, 113, 73 – 84.
Roque, H. (1996). Chemical Water Treatment: Principles and Practice, VHC Publishers, Inc., New York.
Sheikholeslami, R. and Tan, S. (1999). Effects of water quality on silica fouling of desalination plants.
Desalination, 126, 267 –280.
Sheikholeslami, R., Al-Mutaz, I.S., Koo, T. and Young, A. (2001). Pretreatment and the effect of cations and
anions on prevention of silica fouling. Desalination, 139, 83 – 95.
Sheikholeslami, R., Al-Mutaz, I.S., Tan, S. and Tan, S.D. (2002). Some aspects of silica polymerization and
fouling and its pretreatment by sodium aluminate, lime and soda ash. Desalination, 150, 85 – 92.
Standard Methods for the Examination of Water and Wastewater, 19th edn (1995). APHA, AWWA and
WEF. Washington DC, USA.
Van Benschoten, J.E. and Edzwald, J.K. (1990a). Chemical aspects of coagulation using aluminum salts – I.
Hydrolytic reactions of alum and polyaluminum chloride. Wat. Res., 24, 1519– 1526.
Van Benschoten, J.E. and Edzwald, J.K. (1990b). Chemical aspects of coagulation using aluminum salts – II.
Coagulation of fulvic acid using alum and polyaluminum chloride. Wat. Res., 24, 1527– 1535.

195