Gulation settling of natural organic matter from soft tropical water

International Journal of Advanced Research in Engineering and Technology (IJARET), ISSN
0976 – 6480(Print), ISSN 0976 – 6499(Online) Volume 4, Issue 4, May – June (2013), © IAEME
118
COAGULATION-SETTLING OF NATURAL ORGANIC MATTER
FROM SOFT TROPICAL WATER USING ALUMINIUM AND
IRON(III) SULPHATE
J.M. Sieliechi1,
*, R. Kamga1
, G. Joseph Kayem2
1
. Department of Applied Chemistry, ENSAI, P.O. Box 455, University of Ngaoundere,
Cameroon
2
.Department of Process Engineering, Water Treatment and Filtration Group, ENSAI, P.O.
Box 455, University of Ngaoundere, Cameroon
ABSTRACT
High natural organic matter (NOM) containing tropical forest river waters are very
difficult to treat for drinking because of their low values of alkalinity, mineralization and
turbidity. In addition, owing to cost of chemicals, coagulation of these waters is performed
under variable pH conditions. We report studies on such high NOM water, using synthetic
water containing NOM extract from a tropical forest river and natural raw water from the
same river, by Jar test and column settling experiments. Two coagulants were tested,
aluminium and iron (III) sulphate, in order to determine, firstly, the pH range at which the
initial water pH should be set for best coagulation and minimisation of residual NOM, and
secondly, the influence of each coagulant on NOM removal and floc sedimentation. Jar test
results showed that minimum coagulant consumption for maximum NOM removal attended
by minimum NOM residuals obtained for initial pH in the range 5.5 – 6.5 for both
coagulants. The amount of Al sulphate required was at least 2 times that of Fe (III) sulphate.
In column settling of metal-NOM flocs, Al was generally better than Fe but both gave
comparable results at hydraulic loading rates ≤ 1.0 m/hr.
Keywords: Coagulation, humic substances, sedimentation, soft water, water treatment.
1. INTRODUCTION
In the hot humid tropical regions, surface waters, especially river waters constitute the
principal source of raw water for drinking water production for cities. The river waters are
soft with slightly acidic to neutral pH, have low levels of alkalinity and mineralization. In the
INTERNATIONAL JOURNAL OF ADVANCED RESEARCH IN
ENGINEERING AND TECHNOLOGY (IJARET)
ISSN 0976 - 6480 (Print)
ISSN 0976 - 6499 (Online)
Volume 4, Issue 4, May – June 2013, pp. 118-133
© IAEME: www.iaeme.com/ijaret.asp
Journal Impact Factor (2013): 5.8376 (Calculated by GISI)
www.jifactor.com
IJARET
© I A E M E

119
dense luxuriant rainforest zones, river waters have low turbidity and medium to high values
of dissolved natural organic matter (NOM), generally called humic substances [1, 2, 3].
Aquatic humic substances represent at least 50% of the dissolved organic on matter
(DOM) in natural surface waters, and in the tropical rainforest, they are essentially, the water
extractable fraction of soil humus that derives from biogeochemical processes such as the
decay of plant residues [1,4]. The yellow to brown colour of medium to high NOM
containing waters can at least in part be attributed to the presence of dissolved coloured
humic substances. At typical natural surface water pH, the humic substances present are a
mixture of fulvic and humic acids, which account for about 80 % of DOM; with fulvic acids
predominating in most waters especially non-coloured waters and humic acids can be of
greater proportion in highly coloured waters [4,5]. Fulvic and humic acids possess phenolic
and carboxylic groups whose dissociation increase with pH so that in the pH conditions of
natural surface waters, they have negative charges making them more soluble and stable, and
enabling them to contribute to the stabilisation of aquatic inorganic colloids [2].
Several water quality problems are posed by the presence of humics, necessitating
their complete removal or at least minimization. First, aesthetically, colour, taste and odour
make these waters unfit for drinking. The presence of humic substances in natural waters
renders these waters difficult to treat, causing higher coagulant demand [6, 7]. Indeed,
membrane fouling, trihalomethanes formation during chlorine desinfection, or biological
regrowth in the distribution network, have all been linked to the presence of residual humic
substances in clariﬁed water [8,9,10]. Humics substances form complex with hazardous
inorganic and organic contaminants such as trace (heavy) metals and pesticides, thus
probably facilitating their transport through treatment systems [11,12].
Chemical coagulation by hydrolysing metal salts is the major technique used around
the world for removal of NOM from water, with Al salts most widely used and iron (III) salts
to a lesser extent [13,14, 15, 16]. For the purpose of improving dissolved NOM removal,
considerable research efforts have been made in recent years towards investigating the
mechanism of destabilisation of aquatic NOM by Al and Fe salts [17, 18, 19, 20]. Three main
mechanisms are generally invoked to explain the removal of humic substances by
coagulation: charge neutralization/complexation preferentially applies at acidic pH and ﬁnds
experimental support from stoichiometric relationships between coagulant demand and
dissolved organic matter concentration, and from suspension restabilization upon overdosing
[21, 22, 23]. On the other hand, under conditions favoring metal hydroxide precipitation,
physical ensmeshment and/or adsorption onto the freshly formed precipitate are assumed to
play a major role in humic substances elimination [24].
The studies undertaken so far on NOM removal from natural waters have been
essentially concerned with hard, high alkalinity waters. However, Eikebrokk [25, 26], has
investigated NOM removal by Al salts from low alkalinity soft Norwegian waters at constant
pH and found that minimisation of NOM residuals occurred at pH 5 – 6 and Al metal
residuals at pH 5 - 7. The case of tropical soft humic waters with low alkalinity,
mineralization and turbidity, has hardly received any attention. Removal of NOM from these
waters by water suppliers notably in tropical Africa is based on turbidity and colour reduction
with Al sulphate. Insufficient importance is accorded to process improvement with respect to
minimisation of DOM residuals, despite the risks associated with the presence of humics in
water as indicated earlier above.
In addition, although it has been established that pH is one of the most crucial factors
regulating NOM removal and metal ion speciation [19, 27, 28, 29, 30], coagulation of

120
tropical soft humic waters is carried out under conditions of variable pH. Coagulant is added
to the water at its natural pH or after raising pH to a desired value by liming, depending on
in-plant experience. The pH is then allowed to drift downwards, towards, it is expected, the
optimum pH for NOM destabilisation. Unfortunately, the coagulation process often fails in
this case, causing coloured water to arrive at the tap. Even though the high cost of treatment
chemicals is at the root of this practice in developing tropical Africa of allowing pH drift
during coagulation. The operative mechanisms have been inferred from the discussions by
Duan and Gregory[31].There are no studies available to indicate the optimum initial pH range
for coagulation of these soft waters under conditions of variable pH, where pH is not adjusted
after adding coagulant.
Al is preferred over Fe III sulphate as coagulant for treating humic surface waters in
the tropics because of the corrosiveness of Fe III salts and the risk of staining of clothes by
these salts. However, research is giving greater consideration to the use of Fe III salts because
the aluminium is non essential metal. Also, it is claimed that Fe III forms stronger and denser
flocs with NOM than does Al and it should thus be expected that Fe-NOM flocs should settle
faster than Al-NOM flocs. Yet, no pilot scale study has been performed to test the settling
properties of the two types of metal-NOM flocs. Deductions are made based on small scale
Jar tests [32].
The settling characteristics of flocs can be evaluated by column settling pilot scale
tests in the laboratory [33,34]. Zanoni et al. [35] have shown that for columns of height equal
to or greater than 2.0 meters and internal diameter of 100 mm or above, the results of the
column settling tests were similar. Most studies by column settling of aqueous flocculent
suspensions have been performed on inorganic colloids and waste water systems.
This study was therefore directed at the investigation of the removal of NOM with
freshly prepared Al and Fe III sulphate from soft, low alkalinity and mineralization tropical
river water in a treatment situation involving pH drop from a preset initial value. It had two
aims, namely, determining the initial pH range for optimum coagulation with minimum NOM
residuals and evaluating on pilot scale the settling of metal-NOM flocs formed by the two
coagulants and hence appreciate their relative effects on clarification. The study was
performed by Jar test and column settling tests and involved synthetic waters and natural
river water.
2. MATERIALS AND METHODS
2.1. Water sources
Feedstock waters used in this study were from two sources, distilled water and raw
river water. Distilled water from a borosilicate glass still (JENCONS, UK), was used for
extraction of natural organic matter (NOM) and preparation of synthetic waters containing
NOM for Jar test experiments and column settling tests. Raw river water was from River
Nyong in the Cameroon rainforest south of Yaounde (Latitude 4°N), on which is located the
water treatment plant supplying Yaounde, Cameroon’s capital. The river water was used for
Jar test and column settling experiments.
2.2. Chemicals
All chemicals were analytical grade reagents. Aluminium sulphate (Al2(SO4)3.18H2O)
coagulant, sulphuric acid (H2SO4) and sodium hydroxide (NaOH) used for pH adjustment,
were obtained from PROLABO France, now MERCK-Eurolab. Ferric sulphate, Fe2 (SO4)3
xH2O coagulant was from RIEDEL DEN HAEN, Germany. Since the number of molecules of

121
water in the ferric sulphate was unknown, this chemical was dried to constant weight at 105°C
and cooled in a desiccators before being used to prepare solutions based on the formula weight of
399.88 g/mol.
2.3. Glassware
All glassware used was of borosilicate glass.
2.4. Apparatus
The pH of solutions and suspensions was measured by means of a HANNA HI 931000
pH-meter (Hanna Instruments, Portugal) with a precision of 0.05 pH units.
A SPECTRONIC GENESYS 2PC model 336003 single beam UV-Visible
spectrophotometer (Spectronic Instruments Co, USA) equipped with a quartz cuvette (10 mm
optical path) was used to determine the absorption peak of dissolved NOM and also for
establishing the calibration graph for determining NOM concentration in waters tested. The
absorption peak was confirmed on a double beam UV-visible spectrophometer.
Coagulation - flocculation of NOM in water was performed on a VITTADINI Jar test
apparatus (PROLABO, France) equipped with 6 paddle stirrers.
Column settling experiments to determine removal rates of coagulated NOM were carried
out with the system shown in Figure 1. It consists of a rapid mixing tank, a pump and transparent
acrylic polymer (Polymethylmethacrylate) Column of height 2m and internal diameter 100mm.
Both the mixing tank and the column are equipped with paddle stirrers as shown. Sampling ports
on the column were fitted with short silicone tubing tightly sealed with screw valves at 0.5m,
1.0m and 1.5m depths. A syringe fitted with a small hypodermic needle was used to withdraw
samples the column were fitted with short silicone tubing tightly sealed with screw valves at
0.5m, 1.0m, and at each port.
Figure 1 Schematic diagram of gravity sedimentation apparatus
0
500 mm
1000 mm
1500 mm
2000 mm
Motor
Sedimentation
columnPump
Mixing tank
Motor
Valve

122
2.5. Procedures
NOM extract for coagulation-flocculation studies was obtained by alkaline extraction
from river sediment (soil) knowing that aquatic NOM and soil humus have the same chemical
behaviour (Stevenson 1994). The extraction was done by the Jackson [36] method as follows.
Fresh sediment was dug out during the rainy season from Nyong river bed upstream from the
water treatment plant. The sediment was allowed to dry in air at room temperature (24±2°C)
inside a vacuum cupboard. The dried sediment, hereafter called soil, was then ground to fine
powder in an agate mortar. Aliquots (100g) of powder soil were placed in glass beakers (2L)
to each of which was then added 1L of NaOH solution (0.25M). The mixture in each beaker
was agitated for 20 minutes with a HB502 magnetic stirrer (Bibby Sterilin, UK). After
stopping agitation, the beakers were covered with parafilm and the suspensions formed were
allowed to settle for 12 hours. The supernatant (pH 12) was carefully poured into other 2L
glass beakers, pH adjusted to 7 by adding drops of H2SO4 (0.5M) and the resultant
suspension allowed to stand from 12 h. The supernatant solutions containing NOM extract
were then carefully transferred to 1L screw capped glass bottles.
The aqueous NOM extract so obtained was filtered by means of a CARBOSEP (TECH-SEP,
France) tubular membrane module having a membrane pore size of 0.14 µm. The clear
filtrate, also called permeate, containing dissolved NOM was stored in 1L screw capped glass
bottles at 4°C until required for experiments.
Before use, the concentration of the stock NOM extract was determined
gravimetrically after drying the extract solution. A sample of the NOM extract solution was
placed in a Spectra/Por membrane dialysis tube (Spectrum Medical Industries, USA) and
dialysed exhaustively against distilled water (conductivity 3 µS/cm) under agitation in a glass
tank. Water was changed every 12 hours until the conductivity of the dialysing water was
constant and close to that of distilled water. Conductivity was measured with a TACUSSEL
CD 60 Resistimeter (TACUSSEL Electronique, France). The concentration of the dialysed
stock NOM extract solution was then determined gravimetrically by drying a sample (200
mL) to constant weight at 105°C. Triplicate determinations were made and the average taken
to calculate the concentration of the stock NOM solution.
The residual concentration of NOM after Jar test coagulation-flocculation and settling
experiments was determined as follows. The UV-Visible absorption peak of dialysed stock
NOM solution at several pH values was found by measuring absorbance versus wavelength
after diluting the NOM solution to give an absorbance value in the range 0.1 to 0.9 [37]. The
absorption peak was at 280 nm was quite independent of pH as can be seen in Figure 2
Korshin et al. [38] in their work show that the presence of coagulant species doesn’t modify
the spectrum of NOM. Knowing the concentration of the dialysed stock NOM solution and
the dilution factor used, a calibration graph of NOM concentration, C (mg/L) versus
absorbance at 280 nm was established and shows a linear graph. This graph of equation,
Absorbance = 0.0230*C + 0.0364, has a correlation coefficient of 0.994.

123
Figure 2 UV-Visible absorption spectra of diluted dialysed NOM extract solution at various
values pH 4 (●), 5 (○), 6 (▼), 7 (∆)
Coagulation-flocculation experiments were performed on synthetic waters and Nyong
River water by means of a Jar test bench with 6 glass beakers of 1L each filled with 800 mL
of water. Nyong River water was tested as such, the average dissolved NOM concentration
being 20±0.6 mg/L and pH 5.9±0.2 during the period of investigation. The synthetic waters
were obtained by diluting stock NOM extract solutions with distilled water down to a desired
concentration (10, 25 or 40 mg/L), and the pH was then adjusted under agitation to a
predetermined value with sulphuric acid or NaOH (0.05M). The Jar test procedure involved
adding coagulant from a freshly prepared stock solution of 10g/L not more than 6 hours old
to the NOM containing water under rapid agitation at 260 rpm, after which, stirring was
continued for 2 minutes at 260 rpm followed by 10 minutes at 50 rpm. Agitation was then
stopped and the resultant suspension was allowed to settle for 20 minutes. The final pH was
measured, the ccc ascertained and residual NOM at ccc determined by absorbance at 280 nm
on a small sample prefiltered on a 0.45 µm porosity disc membrane. The sample analysed for
residual NOM was withdrawn each time at the same level (5cm) below the water surface in
the beaker using a syringe fitted with a small hypodermic needle. The values of ccc, final pH,
as well as residual NOM concentrations, were established on the basis of triplicate
coagulation-flocculation tests.
Column settling studies were carried out with the set-up shown in Figure 1 on
synthetic waters (25mg/L NOM) at initial pH 5.0±0.05 and 6.0±0.05 and on Nyong river
Wavelength (nm)
0 150 200 250 300 350 400 450 500 550 600 650 700 750
Absorbance
0
1
2
3

124
water (20.0±0.6mg/L NOM) at pH 5.9±0.2. For this purpose, the predetermined volume of
coagulant solution to attain the ccc was added to water in the mixing tank under high speed
agitation and stirring continued for 2 minutes at the same speed then stopped. The pump was
then used to rapidly transfer enough suspension (15.7L) to fill the settling column to the zero
mark. The suspension in the column was agitated for 15 minutes at slow speed (velocity
gradient 40/s). Agitation was stopped and samples were withdrawn from the indicated ports
(0.5m, 1.0m, 1.5m) at predetermined times. Sample withdrawal was staggered and alternated
between column sides in order to avoid channelling effects. Removal rate of flocculent NOM
particles was determined by measuring turbidity of withdrawn samples with a HACH
RatioXR turbidimeter (HACH Co., Loveland, USA). The column settling tests were
conducted in duplicate in each case studied. At the end of each run, the percentage removal at
the three selected depths for each settling time was calculated [39]. Then, following the
method of Krishnan [33] the average percentage removal in the column and the
corresponding settling velocity also called hydraulic loading rate (HLR) was evaluated for
each of the chosen settling times. HLR was given by the height of the column divided by the
settling time. In literature, sometimes surface overflow rate (SOR) is also used but that is
simply obtained by dividing the HLR values by the cross-sectional area of the settling tank.
3. RESULTS AND DISCUSSION
3.1. Coagulant Demand and NOM Residuals
The influence of initial pH of water (pHi) on the critical coagulation concentration
(ccc) and thus on coagulant demand on the one hand and on NOM residuals on the other hand
after coagulation and settling in Jar tests are shown in Figure 3 for synthetic water containing
25mg/L NOM. Table 1 presents the results for experiments carried out on Nyong River water
containing 20mg/L NOM.
pH
0,04,0 4,5 5,0 5,5 6,0 6,5 7,0 7,5 8,0
CCC(mg/L)
0
50
100
150
200
250
300
350
pH
0.04.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0
0
2
4
6
8
10
12
ResidualNOM(mg/L)
Figure 3. Variation of the ccc of coagulant (a) and residual NOM (b) with initial solution pH
for coagulation (25±2 °C) of 25 mg/L NOM water with Al sulphate (●); Fe (III) sulphate (○).
(Added coagulant concentration is that at CCC)
(b)(a)

125
Table 1: Coagulation of Nyong River water, Cameroon
Raw water
characteristics
(in rainy season)
pH
initial
Al sulphate coagulation Fe sulphate coagulation
CCC
(mg/L)
Residual NOM
(mg/L)
CCC
(mg/L)
Residual NOM
(mg/L)
pH 5.9±0.2
Alkalinity 9.1±0.5
(mg/L)
NOM 20.0±0.6
(mg/L)
Turbidity 3.8±0.4
(NTU)
Colour 180±20
Pt/Co (mg/L)
Temperature
24±0.5°C
5.0±0.1
5.5±0.1
6.0±0.1
6.5±0.1
7.0±0.1
44.59
59.31
59.82
57.80
89.19
4.82
2.69
2.68
2.74
2.36
29.98
38.73
37.67
39.05
59.90
3.40
2.10
2.20
1.95
1.70
It can be seen that the graph of figure 3a, ccc vs pHi, has three segments defined by
pHi for both Al and Fe(III) sulphate coagulation. First, when pHi is in the range 4.5 – 5.5,
coagulant demand rises gently with increase in pH, 54.33mg/L per pH unit for Al and
29.63mg/L per pH unit for Fe(III). Then for pHi between 5.5 and 6.5, coagulant demand
roughly stabilizes, with a difference per pH unit of only 11.38 mg/L for Al and 5.88 mg/L for
Fe (III). Finally, for pHi in the range 6.5 – 7.5, coagulant demand rises rapidly with increase
in pH, being per pH unit, 128.00 mg/L for Al and 78.45mg/L for Fe (III). The trends seen in
figure 3a were also observed in coagulation experiments on other synthetic waters containing
NOM at concentrations of 15mg/L and 40mg/L except that the values of ccc were lower for
15mg/L NOM and higher for 40mg/L NOM. The results shown in table 1 for raw river water
confirm the stabilization of ccc for pHi in the range 5.5 – 6.5. Throughout the pHi range
investigated in figure 3a, coagulant demand is much higher for Al than for Fe (III), and on
average 100% higher; the ccc value for Al sulphate is at least two times that of Fe (III)
sulphate. The higher coagulant demand involved in using Al sulphate is confirmed in table 1
even though the difference between the two coagulants is around 50% only for the
coagulation of raw Nyong River water. Figure 3b also presents the variation of NOM
residuals against pHi for synthetic water containing 25mg/L NOM. The curves shows three
segments for Al sulphate coagulation whereas Fe (III) sulphate coagulation has only two
segments for NOM residuals vs pHi. Both coagulants show the same trend for NOM residuals
vs pHi for pHi below 5.5. Above this value of pHi, the curve for Al has two segments, one
with almost constant (4.0 + 0.5 mg/L) but slightly increasing NOM residuals for pHi between
5.5 and 6.5, and one with falling NOM residuals at pHi above 6.5; whereas for Fe (III), at pHi
above 5.5, there is only one segment, with roughly constant NOM residuals generally less
than 1mg/L. In general, Al sulphate coagulation leaves higher NOM residuals than does Fe
(III) sulphate coagulation. This is corroborated by the results shown in table 1 on raw river
water but the difference in NOM residuals arising from the use of the two coagulants is much
smaller. Furthermore, table 1 also confirms the stabilization of NOM residuals for the two
coagulants for the condition, 5.5 < pHi > 6.5.
The trends shown by the results displayed in Figure 3 for synthetic water as well as
those in table 1 for raw Nyong river water, when considered together, enable establishment of

126
an optimum pH region for minimization of both coagulant demand and NOM residuals
during coagulation of soft, low alkalinity, low mineralization and low turbidity water under
variable pH conditions where, pH is preset but allowed to vary by dropping after coagulant
addition. The optimum value of pHi is between 5.5 and 6.5. Both coagulant demand (ccc) and
NOM residuals are roughly constant in this pHi range. This corresponds to a final pH in
solution (pHf) of 4.5 – 5.0 for Al sulphate and 4.5 – 5.5 for Fe (III) sulphate. For pHi below
5.5, NOM residuals are high and for pHi above 6.5, NOM residuals are minimized but at the
expense of steeply increasing coagulant demand. The optimum pH range is not affected by
the presence of particulate turbidity as shown by the results on raw river water (table 1). This
is in agreement with Edwards and Amirtharajah [40], who studied removal of colour caused
by humic acids using alum and found that added turbidity had no effect on the optimum
coagulation domain for medium to high humic acid water. Eikebrokk[25], Cheng et al, [41]
studied the removal of humic substances from soft, low alkalinity and low turbidity waters
by Al coagulants and stated that Al was minimized at pH 5 – 7 while NOM residuals were
minimized at pH 5 – 6.
The results of ccc vs pH and NOM residuals vs pH as seen above show that Fe (III) is
a better coagulant than Al both in terms of coagulant demand and NOM residuals. This result
is in accord with Black et al. [42] who pointed out that Fe (III) sulphate was more than two
times as effective as Al sulphate in the coagulation of colour causing organic compounds in
water. In the optimum pHi domain (5.5 – 6.5) where ccc is stabilized in our studies,
consideration of coagulant demand in terms of metal ion to NOM mass ratio (mg metal ion /
mg NOM) gives for Al, 0.27± 0.01, and for Fe (III), 0.46 ± 0.02 for synthetic water; but for
raw river water, 0.24 mg Al / mg NOM and 0.50 mg Fe (III) / mg NOM. Outside the
optimum pH domain for ccc, it is found that NOM residuals increase below 0.22 mg Al/ mg
NOM and 0.4 mg Fe (III) / mg NOM; also, at pHi = 6.5 – 7.5, mg Al / mg NOM rises from
0.27 to 0.49 and mg Fe (III) / mg NOM rises from 0.46 to 0.91 but NOM is still minimized.
Thus, NOM residuals are minimized at mg Al /mg NOM values of 0.27 – 0.49 and at mg Fe
(III) / mg NOM values of 0.46 – 0.91. These results are in general agreement with those in
the literature. In the case of Al, Eikebrokk [25] for soft, low alkalinity Norwegian waters
reported dosage requirements of 0.29 – 0.56 mg Al / mg NOM to minimize NOM and metal
residuals and Jekel [43] suggest a minimum dosage 0.4 – 0.5 mg Al/mg NOM for low Al
residuals and best removal of NOM. As regards Fe (III), the results of Cheng [44],
recalculated along these lines give mg Fe / mg NOM values of 0.34 – 0.57 for good NOM
removal with minimization of residuals.
In fact in the optimum range of initial pH of water, pHi 5.5 – 6.5, determination of
NOM residuals after Jar Test coagulation experiments show that NOM removal was as high
as 97.58% for Fe (III) as against 86.69% for Al. The difference in behaviour of Al and Fe
(III) sulphate with respect to pH depression during coagulation, coagulant demand and NOM
residuals, points to the possibility that in at least part of the pH range explored in these
studies, some of the types of species involved in NOM coagulation by the two metals are
different.
3.2. Column Settling of Metal-NOM Flocs
Figure 4 represent the column settling results obtained on the humic synthetic water
tested and Figure 5 shows those obtained on raw natural humic water (Nyong River). The
values of pHi at which experiments were performed on synthetic waters were selected on the

127
basis of the results obtained for NOM residuals, with pH 5 (Fig.3b) being in the zone of high
NOM residuals and pH 6 (Fig. 3b) in the zone of minimum NOM residuals.
It can be seen from Figure 4 for synthetic waters that, for both coagulants, the graphs
show the same trends generally. Metal - NOM floc removal is low (< 20%) and independent
of HLR at HLR higher than 5 m/hr. For HLR less than 2 m/hr, NOM removal increases
rapidly with decreasing HLR. Comparison of Figure 4a and b shows Metal – NOM floc
removal is generally higher for Al coagulated NOM than for Fe (III) coagulated NOM. This
implies that Al – NOM flocs have higher settling rates compared to Fe (III) – NOM flocs. At
all values of HLR investigated, Fe (III) – NOM flocs seem to have the same settling rates at
pH 5 and pH 6, as indicated by results of NOM removal. The absence of a pH effect on floc
settling points to the possibility that in the conditions prevailing in these studies, the
structuring of aggregates, Fe (III) – NOM flocs, deriving from Fe (III) coagulation of NOM
in these waters occurs by the same mechanism at both pH values. This situation is similar to
that applicable in the mechanism of coagulation which is complexation – charge
neutralization at both pH values, as suggested earlier. However, Vilgé – Ritter et al. [45]
stated that floc structuring in Fe (III) coagulation of NOM in natural waters (Lake Ribou and
Seine River) was pH dependent as indicated by the fractal dimensions of aggregates and was
controlled by the nature of the organic matter. It is suggested that the nature of the tropical
river organic matter may be responsible for the absence of a pH effect in our case. Regarding
Al –NOM flocs, settling rates are similar at both pH values for HLR less than 2 m/hr, but at
higher HLR, pH has a small but weak effect with pH 6 giving slightly higher floc settling
(removal) rates than pH 5. This could be due to a small difference in coagulating species
present, in view of the suggestion made earlier that at pH 5, Al coagulates NOM purely by
complexation – charge neutralization while at pH 6, although complexation – charge
neutralization is predominant. At higher HLR shearing would tend to reduce floc aggregate
sizes and hence cause slower settling rates.
The results presented in Figure 5, pertaining to Nyong River water (pH 5.9), confirm
the metal cation effect on settling rates. At all HLR values greater than 2m/hr, the percent
removal of Al-NOM flocs is higher than that of Fe(III)-NOM flocs, indicating that Al-NOM
flocs have a higher settling rate compared to Fe(III)-NOM flocs for HLR above 2m/hr. But
for HLR below 2m/hr, the difference between the two coagulants reduces rapidly so that
percent removal and hence settling rates are comparable at HLR below 1m/hr. As in Figure 4,
the percent removal of NOM, for both coagulants and for HLR greater than 5m/hr, is quite
low (< 30%). Interestingly, for greater than 50% NOM removal, it is necessary to operate at
HLR values less than 2m/hr, that is, at quite low HLR as observed in practice at the Nyong
River water treatment plant. NOM removal values of 80% or more can be achieved at HLR
values of 1m/hr or less. The need for very low HLR values to achieve high percent removal
of NOM points to low density of metal – NOM flocs in the absence of significant turbidity in
these waters. It has been observed from the results presented above that, Al – NOM flocs
generally settle faster than Fe (III) – NOM flocs, the difference being more evident for Nyong
River water.

128
Figure 4 Effect of hydraulic loading on the removal of NOM from 25 mg/L NOM water
coagulated at pH 5 (a) and pH (6.5) (b) with Al sulphate (●) and Fe (III) sulphate (○)
Figure 5 Effect of hydraulic loading on NOM removal from Nyong river raw water (20
mg/L NOM) coagulated at pH 5.9 with Al sulphate (●) and Fe (III) sulphate (○)
(b)
Hydraulic loading rate (m/hr)
0 5 10 15 20 25 30
%NOMremoval
0
20
40
60
80
100
(b)
Hydraulic Loading rate (m/hr)
0 5 10 15 20 25 30
%NOMremoval
0
20
40
60
80
100
(a)
Hydraulic loading rate (m/hr)
0 2 4 6 8 10 12 14
%NOMRemoval
0
20
40
60
80
100

129
3.3. Coagulation Mechanisms
Reaction of NOM with Al and Fe (III) depends on the metal hydrolysis species
present and the degree of dissociation of carboxyl and phenolic groups of the NOM. The
hydrolysis of Al and Fe (III) in the presence of humic acids and NOM is limited by the
presence of the latter but still dependent on solution pH. Masion et al. [46] studied the
hydrolysis of Al in the presence of small organic acid ligands using small angle X –ray
scattering and found that speciation of Al in the pH range of 3 – 8 was limited to monomers,
dimers and small oligomers (trimers). Further, Vilgé-Ritter et al. [45] found that Fe (III)
hydrolysis in the presence of lake and river NOM at pH 5.5 and 7.5 was limited to small
oligomers (trimers) due to hindrance of the hydrolysis of Fe (III) by NOM. Rose et al. [47]
who studied Fe speciation in NSIMI river water from the same tropical forest zone as Nyong
River in Cameroon, stated that NOM hindered Fe hydrolysis even at pH 6, Fe was poorly
polymerized due to complexation with NOM which blocks the Fe growth sites. Hence the
presence of NOM prevents formation of polymers and oxyhydroxides of Fe at pH below 7.5.
In view of the foregoing, it can be inferred that the mechanisms of coagulation of Al and Fe
(III) sulphate with NOM as suggested by the trends in Fig. 3 are as follows.
In the case of Al sulphate, there are three segments. For pHi 5.5 – 4.5 corresponding
to final pH (pHf) 4.3 – 4.5, reaction between positively charge Al ionic species (monomers,
dimers and trimers) and ionized carboxyl of NOM occurs by electrostatic attraction followed
by complexation –charge neutralization. The effective pH (4.3 – 4.5) is too low for the
formation of a significant amount of Al hydroxide and it is also known that at least 80 % of
the carboxyl groups are ionized in this pHf range [19,48]. The high level of residuals in this
pH range is probably due to soluble metal NOM complexes and non-complexed NOM. In the
range of pHi 5.5 – 6.5 (pHf 4.5 – 5.0), the amount of NOM residuals roughly stabilizes as
does the ccc. This could be due to the fact that most of the NOM carboxyl groups are ionized.
However, in this pHf range, dimer and trimer, aluminum species with different numbers of
sulfate ions also forms [48-50] and so there is formation of Al – NOM complexes by reaction
of NOM with various Al ionic species. Coagulation probably occurs by complexation charge
neutralization. At pHi 6.5 – 7.7 (pHf 5.0 – 5.4) all the carboxyl groups of NOM and some of
the phenolic groups have ionized and dimeric and trimeric aluminium species has become
predominant. In this region, coagulation also occurs principally by complexation-charge
neutralization. The fall in NOM residuals as pHi increases above 6.5 in Figure 3 is consistent
with the total ionization of carboxyl and partial ionization of phenolic groups of NOM. As
regards removal of NOM by Fe (III) sulphate, Figure 3b shows only two segments. NOM
residuals increase as pH falls (pHi 5.5 – 4.5, pHf 4.5 – 3.9). In this pH zone, coagulation
occurs by complexation – charge neutralization, similar to the case of Al sulphate. Increase in
NOM residuals as pH falls in this zone is probably due to the formation of soluble Fe (III) –
NOM species. Indeed, from studies of coagulation of humic acid by Fe (III) salts using
fluorescence quenching, Cheng [44] stated that under low pH conditions, dissolved
complexes of Fe (III) – organic matter were found in solution. For pHi 5.5 – 7.5, that is at pHf
4.5 – 6.3, the amount of NOM residuals is constant and independent of pH, thus suggesting
strongly that in this pH range, the mechanism of NOM removal by Fe is independent of pH.
Since, as stated above, there is no evidence of Fe (III) hydroxide in the presence of NOM
[45, 47], coagulation occurs by complexation – charge neutralization. It is clear that
throughout the pH range (pHi 4.5 – 7.5, pHf 3.9 – 6.3) explored in our studies, removal of
NOM by Fe (III) takes place by complexation – charge neutralization. The size of humic
substances is also known to depend on pH: a stretched conﬁguration occurs at alkaline pH

130
due to intramolecular electrostatic repulsions, whereas small humic aggregates can be formed
below pH 5 [51, 52].Thus the aggregation of humic acid with hydrolyzed-Fe species can be
ascribed as the restructuration of flexible humic network during association with coagulant
species. As shown in recent electron energy loss spectroscopy and pyrene fluorescence
experiments [29, 22], conformational rearrangements can also be evidenced during the
coagulation of negatively charged humic colloids with cationic hydrolyzed metal species.
4. CONCLUSIONS
Studies of NOM coagulation by Al and Fe (III) sulphate in soft, low alkalinity, low
turbidity high NOM containing tropical water using Jar test show that coagulant demand is
much higher (twice or more) for Al compared to Fe (III) salt. Whereas, column settling
experiments show that Al gives better floc settling rates than Fe (III) except at very low
hydraulic loading rates. The difference between Al and Fe (III) sulphate in terms of coagulant
demand (ccc) and removal of NOM from soft, low alkalinity and low turbidity humic water
observed is due to differences in metal hydrolysis species present in the pHi range concerned.
In the pHi zone where NOM residuals are minimized (pHi > 5.5) Al and Fe (III) coagulates
NOM by complexation – charge neutralization.
REFERENCE
[1] Thurman. E. M., (1985), ‘‘Organic geochemistry of natural waters’’, Martinus
Nijhoff and W. Junk Publishers, Dordrecht, p. 497.
[2] Zumstein. J., Buffle. J., (1989), ‘‘Circulation of pedogenic and aquagenic organic
matterin an eutrophic lake’’, Water Res., Vol. 23, pp. 229–239.
[3] Peuravuori. J., (2005), ‘‘NMR spectroscopy study of freshwater humic material in
light of supramolecular assembly’’, Environ. Sci.Technol. Vol. 39, pp. 5541–5549.
[4] Malcolm. R. L. (1985), Geochemistry of stream fulvic and humic substances, in: G.
R. Aiken, D. McKnight. M., Wershaw. R. MacCarthy. L., P. (Eds.), Humic substances
in soil, sediment and water. Geochemistry, isolation and characterisation, Wiley-
Interscience, New York, pp. 181-210.
[5] Steelink. (1983), ‘‘Humic substances in soil, sediment, and water’’, John Wiley and
Sons Inc., New York, pp. 457-476,
[6] Jekel. M. R., and Heinzmann. B., (1989), ‘‘Residual aluminium in drinking water
treatment’’, J. Water SRT-Aqua., Vol. 38, pp. 281 – 288.
[7] Bernhardt. S. H., (1993), ‘‘Control of flocculants by use of streaming current detector,
J. Water SRT-Aqua, Vol. 42, p. 239.
[8] Owen. D.M., et. al. (1995), ‘‘NOM characterization and treatability’’, J. AWWA, Vol.
87, pp. 46–63.
[9] Lin. C.F., Lin. T.Y., Hao. O.J., (2000), ‘‘Effects of humic substance characteristics on
UF performance’’, Water Res. Vol.34, pp.1097–1106.
[10] Shon. H.K., et. al.(2009), ‘‘A study on the influence of ionic strength on the elution
behaviour of membrane organic foulant using advanced separation tools’’,
Desalination and Water Treatment, Vol.11, pp. 38–45.
[11] Suffet. I. H., MacCarthy P., (1989), ‘‘Aquatic humic substances influence on fate and
treatment of pollutants’’, American Chemical Society, Washington DC,.

131
[12] Countway. R. E., Dickhut. R. M., Canuel. E.A., (2003), ‘‘Polycyclic aromatic
hydrocarbon (PAH) distributions and associations with organic matter in surface
waters of the York River’’, VA Estuary. Org. Geochem. Vol. 34 , pp. 209-224.
[13] Edwards. G. A., Amirtharajah. A., (1985), ‘‘Removing color caused by humic acids’’,
J. Am. Water Works Assoc. Vol.77, pp.50-57.
[14] Hundt. T.R. and O’Melia. C.R. (1988), ‘‘Aluminium fulvic acid interactions:
mechanisms and applications’’, J. Am. Water Works Assoc., Vol. 80, p.176.
[15] van Benschoten. J.E., Edzwald. J.K., (1990), ‘‘Chemical aspects of coagulation with
aluminium chloride’’, Water Res. 24, pp.1527 – 1537.
[16] Matilainen. A., Lindqvist. N., Tuhkanen. T., (2005), ‘‘Comparison of the efficiency
of aluminium and ferric sulphate in the removal of natural organic matter during
drinking water treatment process’’, Environmental Technology, Vol. 26, pp. 867-875.
[18] Mazet. M., Angbo. L., Serpaud. B., (1990), Adsorption of humic substances on
preformed aluminium hydroxide flocs, Water Res., Vol. 24, pp.1509-1518.
[19] Edzwald. J.L., (1993), ‘‘Coagulation in drinking water treatment: particles, organics
and coagulants’’, Water Sci. Technol., Vol. 27, pp.21-35.
[20] Masion. A., Vilgé-Ritter. A., Rose, J., Stone, W. E. E., Teppen Brian, J., D.Rybacki,
J.-Y. Bottero, (2000), ‘‘Coagulation-flocculation of natural organic matter with Al
salts: speciation and structure of the aggregates’’, Environ. Sci. Technol., Vol. 34,
pp.3242-3246.
[21] Narkis. N., Rebhun. M., (1983), Inhibition of flocculation processes in systems
containing organic matter, J. WPCF Vol.55, pp. 947-954.
[22] Siéliéchi. J.M., et al. (2008), ‘‘Changes in humic acid conformation during
coagulation with ferric chloride: Implications for drinking water treatment’’, Water
Res., Vol. 42, pp.2111-2123.
[23] Zhang. P., et al. , (2008), ‘‘Coagulation characteristics of polyaluminum chlorides
PAC-Al30 on humic acid removal from water’’, Separation and Puriﬁcation
Technology, Vol. 63, pp. 642–647.
[24] Bose. P. and Reckhow. D.A., (1998), ‘‘Adsorption of natural organic matter on
preformed aluminum hydroxide ﬂocs’’, J. Environ. Eng., Vol. 124, pp. 803–811.
[25] Eikebrokk. B., (1999), ‘‘Coagulation-direct filtration of soft, low alkalinity humic
waters’’, Water Sci. Technol., Vol. 40, pp. 55-62.
[26] Eikebrokk. B., (1996), ‘‘Removal of humic substances by coagulation’’, in H.H.
Hahn, E. Hoffmann, and H. Odegaard (Eds.), Chemical Water and Wastewater
Treatment V, Springer Verlag, Berlin, pp.173 – 187.
[27] Crozes. G., White P., Marshall M., (1995), ‘‘Enhanced coagulation. Its effect on
NOM removal and chemical costs’’, J. Am. Water Works Assoc. Vol. 87, pp.78 – 89.
[28] Jung. A.V., et al., (2005), Coagulation of humic substances and dissolved organic
matter with a ferric salt: An electron energy loss spectroscopy investigation, Water
Res., Vol. 39, pp.3849–3862.
[29] Kazpard. V., et al. (2006), ‘‘Fate of coagulant species and conformational effects
during the aggregation of a model of a humic substance with Al13 polycations’’, Water
Res., Vol. 40, pp.1965 – 1974.
[30] Yang. Z.L., Gao. B.Y., Yue. Q.Y., Wang. Y., (2010), ‘‘Effect of pH on the
coagulation performance of Al-based coagulants and residual aluminum speciation
during the treatment of humic acid–kaolin synthetic water’’, Journal of Hazardous
Materials, Vol. 178, pp.596 – 603.

132
[31] Duan. J. and Gregory J., (2003), ‘‘Coagulation by hydrolysing metal salts’’, Adv.
Colloid Inter. Sci. Vol.100-102, pp. 475-502.
[32] Yan. M., et al. (2006), ‘‘Enhanced coagulation in a typical North-China water
treatment plant’’, Wat. Res., pp.3621 – 3627.
[33] Krishnan. P., ‘‘Column settling tests for flocculant suspension Discussion’’. J.
Environ. Eng. Div. ASCE 102 (1976), pp. 227 – 229.
[34] Yao. K.M., (1979), ‘‘Column settling test and tube settling’’, J. Am. Water Works
Assoc., Vol. 71, pp. 109 – 112.
[35] Zanoni. A.E., Blomquist. M.W, (1975), ‘‘Column settling tests for flocculant
suspensions’’, J. Environ. Eng. Div. ASCE. 6, pp.309 – 318.
[36] Jackson. (1968), ‘‘Soil Chemical Analysis-Advanced Course’’, 4th
edition.
Department of Soil Science, University of Wisconsin, Madison.
[37] Degremont. (1989), Mémento technique de l’eau, Technique et Documentation,
Lavoisier, Paris, pp. 23-38.
[38] Korshin. G. V., Li. C.W., Benjamin. M.M., (1997), ‘‘Monitoring the properties of
natural organic matter through UV spectroscopy: A consistent theory’’, Water Res.
Vol. 31, pp.1787-1795.
[39] Zanoni. A.E., Blomquist M.W., (1975), ‘‘Column settling tests for flocculant
suspensions’’, J. Environ. Eng. Div. ASCE. 6, pp.309 – 318.
[40] Edwards G. A., Amirtharajah A., (1985), Removing color caused by humic acids. J.
Am. Water Works Assoc., Vol. 77, pp.50-57.
[41] Cheng. W.P., Chi. F.H., Li. C.C., Yu. R.F., (2008), ‘‘A study on the removal of
organic substances from low-turbidity and low-alkalinity water with metal-
polysilicate coagulants’’, Colloids Surf. A, Vol.312, pp.238–244.
[42] Black. A.P., Singley. J.E., Whittle, G.P., Maulding, J.S. (1963), ‘‘Stoichiometry of the
coagulation of color-causing organic compounds with ferric sulphate’’, J. Am. Water
Works Assoc., Vol. 10, pp.1347 – 1366.
[43] Jekel. M. R., (1986), ‘‘Interactions of humic acids and aluminum salts in the
flocculation process’’, Water Res., Vol. 20, pp.1535-1542.
[44] Cheng .W. P., (2002), ‘‘Comparison of hydrolysis/coagulation behavior of polymeric
and monomeric iron coagulants in humic acid solution’’, Chemosphere, Vol. 47, pp.
963-969.
[45] Vilgé-Ritter. A., Rose., J., Masion. A., Bottero. J.-Y., Lainé. J.-M., (1999),
‘‘Chemistry and structure of aggregates formed with Fe-salts and natural organic
matter’’, Colloids Surf. A: Physicochem. Eng. Aspects, Vol. 147, pp.297-308.
[46] Masion. A., Bottero. J.Y., Thomas. F., Tchoubar. D., (1994), ‘‘Chemistry and
structure of Al(OH)/organic precipitates. A small-Angle X-ray scattering study. 2.
Speciation and structure of the aggregates’’, Langmuir, Vol. 10, pp. 4349 – 4352.
[47] Rose. J., Vilgé. A., Olivier-Lauquet. G., Masion. A., Frechou. C., Bottero. J. Y.,
(1998), ‘‘Iron speciation in natural organic matter colloids’’, Colloids Surf. A:
Physicochem. Eng. Aspects, Vol. 136, pp.11-19.
[48] Lu. X., Chen. Z., Yang. X., (1999), ‘‘Spectroscopic study of aluminium speciation in
removing humic substances by Al coagulation’’, Water Res., Vol. 33, pp.3271 – 3280.
[49] Urabe. T., Tsugoshi. T., Tanaka. M., (2008), ‘‘Electrospray ionization mass
spectrometry investigation of the blocking effect of sulfate on the formation of
aluminum tridecamer’’, Journal of Molecular Liquids 143, pp.70–74.

133
[50] Wang. Y., et al. (2009), ‘‘Characterization of floc size, strength and structure in
various aluminum coagulants treatment’’, Journal of Colloid and Interface Science,
Vol. 332, pp. 354–359.
[51] Lead. J.R., Wilkinson. K.J., Starchev. K., Canonica. S., Buffle. J., (2000),
‘‘Determination of diffusion coefficients of humic substances by fluorescence
correlation spectroscopy: role of solution conditions’’, Environ. Sci. Technol., Vol.
34, pp.1365–1369.
[52] Plaschke. M., Romer. J., Klenze. R., Kim. J.I.,. (1999), ‘‘In situ AFM study of sorbed
humic acid colloids at different pH’’, Colloids Surf. A Physicochem. Eng. Aspects,
Vol. 160, pp. 269–279.
[53] P.S. Thué, J. M. Siéliéchi, P.P. Ndibewu and R. Kamga, “Physico-Chemical Studies
on The Adsorption of Atrazin on Locally Mined Montmorillonites”, International
Journal of Advanced Research in Engineering & Technology (IJARET), Volume 4,
Issue 1, 2013, pp. 79 - 95, ISSN Print: 0976-6480, ISSN Online: 0976-6499.
[54] R Radhakrishanan and A Praveen, “Sustainability Perceptions on Wastewater
Treatment Operations in Urban Areas of Developing World”, International Journal of
Civil Engineering & Technology (IJCIET), Volume 3, Issue 1, 2012, pp. 45 - 61,
ISSN Print: 0976 – 6308, ISSN Online: 0976 – 6316.

Gulation settling of natural organic matter from soft tropical water

Recommended

Recommended

More Related Content

What's hot

What's hot (20)

Viewers also liked

Viewers also liked (9)

Similar to Gulation settling of natural organic matter from soft tropical water

Similar to Gulation settling of natural organic matter from soft tropical water (20)

More from IAEME Publication

More from IAEME Publication (20)

Recently uploaded

Recently uploaded (20)

Gulation settling of natural organic matter from soft tropical water