-
9CoagulationandFlocculation9-1 Role of Coagulation and
Flocculation Processes in Water
TreatmentCoagulation ProcessFlocculation ProcessPractical Design
Issues
9-2 Stability of Particles in WaterParticle–Solvent
InteractionsElectrical Properties of ParticlesParticle
StabilityCompression of the Electrical Double Layer
9-3 Coagulation TheoryAdsorption and Charge
NeutralizationAdsorption and Interparticle BridgingPrecipitation
and Enmeshment
9-4 Coagulation PracticeInorganic Metallic
CoagulantsPrehydrolyzed Metal SaltsOrganic PolymersCoagulant and
Flocculant AidsJar Testing for Coagulant EvaluationAlternative
Techniques to Reduce Coagulant Dose
9-5 Coagulation of Dissolved ConstituentsEffects of NOM on
Coagulation for Turbidity RemovalEnhanced CoagulationDetermination
of Coagulant Dose for DOC RemovalRemoval of Dissolved
Inorganics
9-6 Flocculation TheoryMechanisms of FlocculationParticle
CollisionsFlocculation of Spherical ParticlesFractal Flocculation
ModelsFloc BreakupUse of Spherical Particle Models for Reactor
Design
541MWH’s Water Treatment: Principles and Design, Third Edition
John C. Crittenden, R. Rhodes Trussell, David W. Hand, Kerry J.
Howe and George TchobanoglousCopyright © 2012 John Wiley &
Sons, Inc.
-
542 9 Coagulation and Flocculation
9-7 Flocculation PracticeAlternative Methods of
FlocculationVertical Turbine FlocculatorsHorizontal Paddle Wheel
FlocculatorsHydraulic FlocculationImportant Design Features in
Flocculation
Problems and Discussion TopicsReferences
Terminology for Coagulation and Flocculation
Term Definition
Coagulation Addition of a chemical to water with the objective
ofdestabilizing particles so they aggregate or forminga precipitate
that will sweep particles from solutionor adsorb dissolved
constituents.
Coagulant aid Chemicals (typically synthentic polymers) added
towater to enhance the coagulation process.
Counterions Ions of opposite charge to the surface charge
ofparticles.
Critical coagulationconcentration(CCC)
Concentration of coagulant that reduces the electricdouble layer
to the point where flocculation canoccur.
Destabilization Process of eliminating the surface charge on
aparticle so that flocculation can occur.
Electric double layer(EDL)
Electrostatic potential surrounding a charged particlein
solution, consisting of a layer of counterionsadsorbed directly to
the surface and a diffuse layerof ions forming a cloud of charge
around theparticle.
Enhancedcoagulation
Coagulation process with the objective of removingnatural
organic matter, typically for minimizing theformation of
disinfection by-products (see Sec 9-5).
Enmeshment orsweep floc
Entrapment or capture of particles by amorphousprecipitates that
form when a coagulant is added towater.
Flocculation Aggregation of destabilized particles into
largermasses that are easier to remove from water thanthe original
particles.
Flocculant aid Organic polymers used to enhance settleability
andfilterability of floc particles.
-
9 Coagulation and Flocculation 543
Term Definition
Inorganic metalcoagulant
Metal salts such as aluminum sulfate and ferric chloridethat
will hydrolyze, forming mononuclear andpolynuclear species of
varying charge. When added inexcess, metal coagulants form chemical
precipitates.
Jar test Procedure to study effect of coagulant addition
towater; used to determine required doses andoperating conditions
for effective coagulation andflocculation.
Stable particlesuspension
Suspension of particles that will stay in solutionindefinitely;
stable particles have a surface chargethat causes them to repel
each other and preventaggregation into larger particles that would
settle ontheir own.
Synthetic organiccoagulant
High-molecular-weight (typically 104 to 107 g/mol)organic
molecules that can carry positive (cationic),negative (anionic), or
neutral (nonionic) charge.
Zeta potential Measurement of the charge at the shear plane
ofparticles, used as a relative measure of particlesurface
charge.
Natural surface waters contain inorganic and organic particles.
Inorganicparticulate constituents, including clay, silt, and
mineral oxides, typicallyenter surface water by natural erosion
processes. Organic particles mayinclude viruses, bacteria, algae,
protozoan cysts and oocysts, as well asdetritus litter that have
fallen into the water source. In addition, surfacewaters will
contain very fine colloidal and dissolved organic constituentssuch
as humic acids, a product of decay and leaching of organic
debris.Particulate and dissolved organic matter is often identified
as naturalorganic matter (NOM).
Removal of particles is required because they can (1) reduce the
clarityof water to unacceptable levels (i.e., cause turbidity) as
well as impart colorto water (aesthetic reasons), (2) be infectious
agents (e.g., viruses, bacteria,and protozoa), and (3) have toxic
compounds adsorbed to their externalsurfaces. The removal of
dissolved NOM is of importance because manyof the constituents that
comprise dissolved NOM are precursors to theformation of
disinfection by-products (see Chap. 19) when chlorine is usedfor
disinfection. NOM can also impart color to the water.
The most common method used to remove particulate matter and a
por-tion of the dissolved NOM from surface waters is by
sedimentation and/orfiltration following the conditioning of the
water by coagulation and floc-culation, the subject of this
chapter. Thus, the purpose of this chapter is topresent the
chemical and physical basis for the phenomena occurring in
IBMHighlight
IBMHighlight
-
544 9 Coagulation and Flocculation
the coagulation and flocculation processes. Specific topics
include (1) therole of coagulation and flocculation processes in
water treatment, (2) sta-bility of particles in water, (3)
coagulation theory, (4) coagulation practice,(5) coagulation of
dissolved and organic constituents, (6) flocculationtheory, and (7)
flocculation practice.
9-1 Role of Coagulation and Flocculation Processes in Water
Treatment
The importance of the coagulation and flocculation processes in
watertreatment can be appreciated by reviewing the process flow
diagramsillustrated on Fig. 9-1. As used in this book, coagulation
involves the additionof a chemical coagulant or coagulants for the
purpose of conditioningthe suspended, colloidal, and dissolved
matter for subsequent processingby flocculation or to create
conditions that will allow for the subsequentremoval of particulate
and dissolved matter. Flocculation is the aggregationof
destabilized particles (particles from which the electrical surface
chargehas been reduced) and precipitation products formed by the
additionof coagulants into larger particles known as flocculant
particles or, morecommonly, ‘‘floc.’’ The aggregated floc can then
be removed by gravitysedimentation and/or filtration. Coagulation
and flocculation can also bedifferentiated on the basis of the time
required for each of the processes.Coagulation typically occurs in
less than 10 s, whereas flocculation occursover a period of 20 to
45 min. An overview of the coagulation and floc-culation processes
is provided below.
CoagulationProcess
The objective of the coagulation process depends on the source
of thewater and the nature of the suspended, colloidal, and
dissolved organic
Settledsolids
SedimentationInfluent fromsurface water
Effluent todistributionsystem
Granularfiltration
Oxidant/disinfectant
Oxidant/disinfectantPolymer
Flocculation
Bypass flocculationfor contact (in-line) filtration
Bypasssedimentation
for direct filtration
Liquid processing
Residuals processingand management Waste
washwater
CoagulantFlashmix
Figure 9-1Typical water treatment process flow diagram employing
coagulation (chemical mixing) with conventional treatment,
directfiltration, or contact filtration.
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
-
9-1 Role of Coagulation and Flocculation Processes in Water
Treatment 545
constituents. Coagulation by the addition of the hydrolyzing
chemicals suchas alum and iron salts and/or organic polymers can
involve
1. Destabilization of small suspended and colloidal particulate
matter
2. Adsorption and/or reaction of portions of the colloidal and
dissolvedNOM to particles
3. Creation of flocculant particles that will sweep through the
waterto be treated, enmeshing small suspended, colloidal, and
dissolvedmaterial as they settle
Coagulants such as alum, ferric chloride, and ferric sulfate
hydrolyzerapidly when mixed with the water to be treated. As these
chemicalshydrolyze, they form insoluble precipitates that
destabilize particles byadsorbing to the surface of the particles
and neutralizing the charge(thus reducing the repulsive forces)
and/or forming bridges betweenthem. Natural or synthetic organic
polyelectrolytes (polymers with multiplecharge-conferring
functional groups) are also used for particle destabi-lization.
Because of the many competing reactions, the theory of
chemicalcoagulation is complex. Thus, the simplified reactions
presented in this andother textbooks to describe the various
coagulation processes can only beconsidered approximations, as the
reactions may not necessarily proceedas indicated (Letterman et
al., 1999).
FlocculationProcess
The purpose of flocculation is to produce particles, by means of
aggrega-tion, that can be removed by subsequent particle separation
proceduressuch as gravity sedimentation and/or filtration. Two
general types of floc-culation can be identified: (1)
microflocculation (also known as perikineticflocculation) in which
particle aggregation is brought about by the ran-dom thermal motion
of fluid molecules (known as Brownian motion) and(2)
macroflocculation (also known as orthokinetic flocculation) in
whichparticle aggregation is brought about by inducing velocity
gradients andmixing in the fluid containing the particles to be
flocculated. Another formof macroflocculation is brought about by
differential settling in which largeparticles overtake small
particles to form larger particles.
Practical DesignIssues
When it comes to the practical design of coagulation and
flocculationfacilities, designers must consider four process
issues: (1) the type andconcentration of coagulants and flocculant
aids, (2) the mixing intensityand the method used to disperse
chemicals into the water for destabilization,(3) the mixing
intensity and time for flocculation, and (4) the selectionof the
liquid–solid separation process (e.g., sedimentation, flotation,
andgranular filtration). With the exception of sedimentation and
flotation(considered in Chap. 10) and filtration (considered in
Chaps. 11 and 12),these subjects are addressed in the subsequent
sections of this chapter.
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
-
546 9 Coagulation and Flocculation
9-2 Stability of Particles in Water
The particles in water may, for practical purposes, be
classified as suspendedand colloidal, according to particle size.
Because small suspended andcolloidal particles and dissolved
constituents will not settle in a reasonableperiod of time,
chemicals must be used to help remove these particles. Thephysical
characteristics of particles found in water including particle
size,number, distribution, and shape have been discussed previously
in Chap. 2,Sec 2-3.
To appreciate the role of chemical coagulants and flocculant
aids, itis important to understand particle solvent interactions
and the electricalproperties of the colloidal particles found in
water. These subjects alongwith the nature of particle stability
and the compression of the electricaldouble layer are considered in
this section.
Particle–SolventInteractions
Particles in natural water can be classified as hydrophobic
(water repelling)and hydrophilic (water attracting). Hydrophobic
particulates have a well-defined interface between the water and
solid phases and have a low affinityfor water molecules. In
addition, hydrophobic particles are thermodynam-ically unstable and
will aggregate irreversibly over time.
Hydrophilic particles such as clays, metal oxides, proteins, or
humicacids have polar or ionized surface functional groups. Many
inorganicparticulates in natural waters, including hydrated metal
oxides (iron or alu-minum oxides), silica (SiO2), and asbestos
fibers, are hydrophilic becausewater molecules will bind to the
polar or ionized surface functional groups(Stumm and Morgan, 1996).
Many organic particulates are also hydrophilicand include a wide
diversity of biocolloids (humic acids, viruses) and sus-pended
living or dead microorganisms (bacteria, protozoa, algae).
Becausebiocolloids can adsorb on the surfaces of inorganic
particulates, the par-ticles in natural waters often exhibit
heterogeneous surface properties.Some particulate suspensions such
as humic or fulvic acids can be reversiblyaggregated because of
their hydrogen bonding tendencies.
ElectricalPropertiesof Particles
The principal electrical property of fine particulate matter
suspended inwater is surface charge, which contributes to relative
stability, causingparticles to remain in suspension without
aggregating for long periodsof time. The particulate suspensions
are thermodynamically unstable and,given sufficient time, colloids
and fine particles will flocculate and settle.However, this process
is not economically feasible because it is very slow.A review of
the causes of particulate stability will provide an understandingof
the techniques that can be used to destabilize particulates, which
arediscussed in the following section.
-
9-2 Stability of Particles in Water 547
HO
HO
Si
Si4+
O
O
Al3+
O
O
Si
O
O
Si
OH
OH
Silicon atom displacedby aluminum atom
Figure 9-2Charge acquisition through isomorphous substitution of
Al for Si.
ORIGIN OF PARTICLE SURFACE CHARGE
Most particulates have complex surface chemistry and surface
charges mayarise from several sources. Surface charge arises in
four principal ways, asdiscussed below (Stumm and Morgan,
1996).
Isomorphous replacement (crystal imperfections)Under geological
conditions, metals in metal oxide minerals can bereplaced by metal
atoms with lower valence, and this will impart a neg-ative charge
to the crystal material. An example where an aluminum atomreplaced
a silicon atom in a clay particle is shown on Fig. 9-2. This
process,known as isomorphous replacement, produces negative charges
on thesurface of clay particles (van Olphen, 1963).
Structural imperfectionsIn clay and similar mineral particles,
imperfections that occur in theformation of the crystal and broken
bonds on the crystal edge can lead tothe development of surface
charges.
Preferential adsorption of specific ionsParticles adsorb NOM
(e.g., fulvic acid), and these large macromoleculestypically have a
negative charge because they contain carboxylic acidgroups:
R − COOH � R − COO− + H+ (pKa = 4 to 5) (9-1)Consequently,
particle surfaces that have adsorbed NOM will be negativelycharged
for pH values greater than ∼5.
Ionization of inorganic groups on particulate surfacesMany
mineral surfaces contain surface functional groups (e.g.,
hydroxyl)and their charge depends on pH. For example, silica has
hydroxyl groupson its exterior surface, and these can accept or
donate protons as shownhere:
Si − OH+2 � Si − OH + H+ � Si − O− + 2H+
pH � 2 pH = 2 pH � 2 (9-2)
-
548 9 Coagulation and Flocculation
pH
0
2 4 6 8 10 12
Par
ticle
surf
ace
char
ge
Silica
Alumina+ψ0
−ψ0
Figure 9-3Variation in particle charge with pH.
The zero point of charge, as shown on Fig. 9-3, forsilica is at
pH 2, whereas the zero point of chargefor alumina is about pH 9.
The pH correspondingto a surface charge of zero is defined as the
zeropoint of charge (ZPC). Above the ZPC the surfacecharge will be
negative (anionic), and below theZPC the charge will be positive
(cationic). TheZPC for other particles that commonly occur inwater
are listed in Table 9-1. When examiningTable 9-1, it is important
to realize that many ofthe measurements that are reported are in
low-ionic-strength waters (i.e., distilled water); conse-quently,
the reported pHzpc values are higher thanis observed in natural
waters.
Table 9-1Surface characteristics of inorganic and organic
particulates commonlyfound in natural waters
Zero Point of Charge,Type of Particle pHzpc
InorganicAl(OH)3 (amorphous) 7.5–8.5Al2O3 9.1CuO3 9.5Fe(OH)3
(amorphous) 8.5MgO 12.4MnO2 2–4.5SiO2 2–3.5Clays
Kaolinite 3.3–4.6Montmorillonite 2.5
AsbestosChrysotile 10–12Crocidolite 5–6
CaCO3 8–9Ca5(PO4)3OH 6–7FePO4 3AlPO4 4
OrganicAlgae 3–5Bacteria 2–4Humic acid 3Oil droplets 2–5
Source: From Parks (1967) and Stumm and Morgan (1981).
-
9-2 Stability of Particles in Water 549
ELECTRICAL DOUBLE LAYER
In natural waters, negatively charged particulates accumulate
positive coun-terions on and near the particle’s surface to satisfy
electroneutrality. Asshown on Fig. 9-4, a layer of cations will
bind tightly to the surface of anegatively charged particle to form
a fixed adsorption layer. This adsorbedlayer of cations, bound to
the particle surface by electrostatic and adsorp-tion forces, is
about 5 Å thick and is known as the Helmholtz layer (alsoknown as
the Stern layer after Stern, who proposed the model shown onFig.
9-4). Beyond the Helmholtz layer, a net negative charge and
electricfield is present that attracts an excess of cations (over
the bulk solutionconcentration) and repels anions, neither of which
are in a fixed position.These cations and anions move about under
the influence of diffusion(caused by collisions with solvent
molecules), and the excess concentrationof cations extends out into
solution until all the surface charge and electricpotential is
eliminated and electroneutrality is satisfied.
Approximate shearlayer measured byelectrophoresis
Fixed charge(Stern) layer
Negatively chargedparticle surface
Diffuseion layer
Ions in equilibriumwith bulk solution
Negative ion
Positivecounterion
Ele
ctro
stat
ic p
oten
tial,
mV
Double layer
Distance from particle surface, A
0
−ψ0
−ψm
−ψζ
κ−1
Nernstpotential
Zeta(Helmholtz)
potential
Zetameasuredpotential
Figure 9-4Structure of the electricaldouble layer. The
potentialmeasured at the shear planeis known as the zeta
potential.The shear plane typicallyoccurs in diffuse layer.
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
-
550 9 Coagulation and Flocculation
The layer of cations and anions that extends from the Helmholtz
layer tothe bulk solution where the charge is zero and
electroneutrality is satisfiedis known as the diffuse layer. Taken
together the adsorbed (Helmholtz)and diffuse layer are known as the
electric double layer (EDL). Dependingon the solution
characteristics, the EDL can extend up to 300 Å into thesolution
(Kruyt, 1952). It is interesting to note that the double-layer
modelproposed by Stern (see Fig. 9-4) is a combination of the
earlier modelsproposed by Helmholtz–Perrin and Gouy–Chapman. In
fact, the diffuselayer is often identified as the Gouy–Chapman
diffuse layer (Voyutsky, 1978).
MEASUREMENT OF SURFACE CHARGE
The electrical properties of highly dispersed particle systems
having a soliddispersed phase and a liquid dispersion medium can be
defined in terms offour phenomena: (1) electrophoresis, (2)
electroosmosis, (3) sedimentation poten-tial (also known as the
Dorn effect), and (4) streaming potential . Collectivelythese four
phenomena, described in Table 9-2, are known as
electrokineticphenomena because they involve the movement of
particles (or a liquid)when a potential gradient is applied or the
formation of the potential
Table 9-2Description and application of electrochemical
phenomena
Phenomena Description Application in Water Treatment
Electrophoresis,discovered by R. Reuss,circa 1808
Refers to the movement of charged particlesrelative to a
stationary liquid subject to anapplied electrical field. The
particles movealong the lines of the electrical field.
Used to assess the destabilizationof particles subject to the
additionof coagulant chemicals. Also usedin laboratory studies to
isolate newproteins and other organicmolecules.
Electroosmosis,discovered by R. Reuss,circa 1808
Refers to the movement of liquid relative toa stationary charged
surface (e.g., a porousplug) subject to an applied electrical
field.
Streaming potential,discovered by G.Quincke, circa 1859
Refers to the creation of a potential gradientwhen liquid is
made to flow along astationary charged surface (e.g., whenforced
through a porous plug). The chargesfrom the particles are carried
along with thefluid.
Used to assess the destabilizationof particles subject to the
additionof coagulant chemicals. Onlineinstruments are now available
thatcan be used to control chemicaladdition in water treatment.
Sedimentation potential,discovered by Dorn,circa 1878
Refers to the creation of a potential gradientwhen charged
particles move (e.g., settling)relative to a stationary liquid
medium.Sedimentation potential is the opposite
ofelectrophoresis.
Source: Adapted from Voyutsky (1978) and Shaw (1966).
IBMHighlight
-
9-2 Stability of Particles in Water 551
gradient when particles (or liquid) move. It should be noted
that theseaforementioned electrical phenomena are caused by the
opposite chargeof the particle (solid) and liquid. Although there
is no direct measure ofthe electrical field surrounding a particle
or method to determine whenparticles have been destabilized from
the addition of coagulants, the sur-face charge on a particle can
be measured indirectly using one of the fourelectrokinetic
phenomena (Voyutsky, 1978).
ZETA POTENTIAL
When a charged particle is subjected to an electric field
between twoelectrodes, a negatively charged particle will migrate
toward the positiveelectrode, as shown on Fig. 9-5, and vice versa.
This movement is termedelectrophoresis. It should be noted that
when a particle moves in an electricalfield some portion of the
water near the surface of the particle moves withit, which gives
rise to the shear plane, as shown on Fig. 9-4. Typically, asshown
on Fig. 9-4, the actual shear plane lies in the diffuse layer to
the rightof the theoretical fixed shear plane defined by the
Helmholtz layer. Theelectrical potential between the actual shear
plane and the bulk solution iswhat is measured by electrophoretic
measurements. This potential is calledthe zeta potential or the
electrical potential and is given by the expression
Z = v0kzμεε0
(9-3)
where Z = zeta potential, Vv0 = electrophoretic mobility,
(μm/s)/(V/cm)
= νE/EνE = electrophoretic velocity of migrating particle, μm/s
(also
reported as nm/s and mm/s)
Diffuse ion cloudtravels with particle
Negativelycharged ion
Negativepole
Positivepole
Positively chargedcounterions attracted
to negative pole
Particle with high negativesurface charge moves toward
positive pole
Figure 9-5Schematic illustration ofelectrophoresis in which a
chargedparticle moves in an electrical field,dragging with it a
cloud of ions.
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
-
552 9 Coagulation and Flocculation
E = electrical field at particle, V/cmkz = constant that is 4π
or 6πμ = dynamic viscosity of water, N · s/m2ε = permitivity
relative to a vacuum (ε for water is 78.54, unitless)
ε0 = permitivity in a vacuum, 8.854188 × 10−12 C2/J · m
(notethat C2/J · m is equivalent to N/V2)
Typical values for the electrophoretic mobility for particles in
naturalwaters vary from about −2 to +2 (μm/s)/(V/cm). The constant
kz is usedto account for the shape of the particle. The value of 4π
appears in thederivation put forth by Smoluchowski and applies if
the extent of the diffuselayer is small relative to the curvature
of the particle. The value of 6π isused when the particle is much
smaller than the thickness of the doublelayer (Kissa, 1999).
For example, if the value of the constant is 4π and the
electrical mobilityis 0.5 (μm/s)/(V/cm), the value of the zeta
potential at 25◦C is 80.4 mV asgiven below:
Z = (0.5 μm·cm/s·V)(4π)(0.890×10−3N·s/m2)(1 m/106μm)(1
m/102cm)(
78.54)(
8.854188 × 10−12 C2/J · m)= 80.4 mV
Empirically, when the absolute value of the zeta potential is
reduced belowapproximately 20 mV, rapid flocculation occurs (Kruyt,
1952). The zetapotential will vary with the size and shape of the
particle, with the numberof charges on the particle, with the
strength of the electric field, and withthe nature of the ions in
the diffuse layer.
Particle Stability The stability of particles in natural waters
depends on a balance between(1) the repulsive electrostatic force
and (2) the attractive force known asthe van der Waals force.
REPULSIVE ELECTROSTATIC FORCES
The principal mechanism controlling the stability of hydrophobic
andhydrophilic particles is electrostatic repulsion. Electrostatic
repulsionoccurs, as discussed above, because particles in water
have a net negativesurface charge. The magnitude of the
electrostatic force will depend onthe charge of the particle and
the solution characteristics.
VAN DER WAALS ATTRACTIVE FORCE
Van der Waals forces originate from magnetic and electronic
resonancethat occurs when two particles approach one another. This
resonance iscaused by electrons in atoms on the particle surface,
which develop astrong attractive force between the particles when
these electrons orientthemselves in such a way as to induce
synergistic electric and magnetic
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
-
9-2 Stability of Particles in Water 553
fields. Van der Waals forces are proportional to the
polarizability of theparticle surfaces. Van der Waals attractive
forces (
-
554 9 Coagulation and Flocculation
condition 1 contains a repulsive maximum that must be overcome
if theparticles are to be held together by the van der Waals force
of attraction.Although floc particles can form at long distances as
shown by the netenergy curve for case 1, the net force holding
these particles together isweak and the floc particles that are
formed can be ruptured easily. In case2, there is no energy barrier
to overcome. Clearly, if colloidal particles areto be flocculated
by microflocculation, the repulsive force must be reducedas shown
in case 2. With the addition of a coagulant, which reduces
theextent of the electrical double layer, rapid flocculation can
occur.
Compressionof the ElectricalDouble Layer
It has been observed that, if the electrical double layer is
compressed,particles in water will come together as a result of
Brownian motion andremain attached due to van der Waals forces of
attraction, as discussedabove. As the ionic strength of a solution
is increased, the extent ofthe double layer decreases, which in
turn reduces the zeta potential. Thethickness of the double layer
and the effects of ionic strength and electrolyteaddition on the
compression of the double layer are described below.
DOUBLE-LAYER THICKNESS
The thickness of the electrical diffuse layer as a function of
the ionicstrength and electrolyte is given in Table 9-3. The
thickness of the diffuselayer may be calculated using the following
equation (Gouy, 1910):
κ−1 = 1010[(2) (1000) e2NAI
εε0 kT
]−1/2(9-4)
where κ−1 = double-layer thickness, Å1010 = length conversion,
Å /m1000 = volume conversion, L/m3
Table 9-3Thickness of electrical double layer (EDL) as function
of ionic strength andvalence at 25◦C
Molarity z+ : z− I, mol/L κ, cm−1 1/κ, Å
0.001 1:1 0.001 1.04 × 106 96.22:2 0.004 2.08 × 106 48.13:3
0.009 3.12 × 106 32.1
0.01 1:1 0.01 3.29 × 106 30.42:2 0.04 6.57 × 106 15.23:3 0.09
9.86 × 106 10.1
0.1 1:1 0.1 1.04 × 107 9.62:2 0.4 2.08 × 107 4.83:3 0.9 3.12 ×
107 3.2
IBMHighlight
IBMHighlight
IBMHighlight
-
9-2 Stability of Particles in Water 555
e = electron charge, 1.60219 × 10−19 CNA = Avagadro’s number,
6.02205 × 1023/mol
I = ionic strength, 12∑
z2M , mol/Lz = magnitude of positive or negative charge on
ion
M = molar concentration of cationic or anionic species, mol/Lε =
permittivity relative to a vacuum (ε for water is 78.54,
unitless)ε0 = permittivity in a vacuum, 8.854188 × 10−12 C2/J ·
mk = Boltzmann constant, 1.38066 × 10−23 J/K
T = absolute temperature, K (273 + ◦C)The relationship given in
Eq. 9-4 is not actually the double-layer thicknessbut is related to
how far out into the solution the repulsive force will reach.It is
approximately equal to the distance at which the electrical
potential is37 percent of the value at the particle surface.
However, it is still importantto know the EDL thickness because it
provides insight into the particlestability and the coagulation
process.
EFFECT OF IONIC STRENGTH
Of the many factors that affect double-layer thickness, ionic
strength isperhaps the most important. As reported in Table 9-3,
the EDL thicknessshrinks dramatically with increasing ionic
strength and valance. Accordingto the DLVO theory, van der Waals
forces extend out into solution about10 Å; consequently, if the
double layer is smaller than this, a rapidly floc-culating
suspension is formed. While it is possible to reduce the
thicknessof the EDL by increasing the ionic strength, this is not a
practical methodfor destabilizing particles in drinking water
treatment because the requiredionic strengths are greater than are
considered acceptable in potable water.It is interesting to note
that ionic strength can be used to explain whyparticles are stable
in freshwater (low ionic strength but high electricalrepulsive
forces) and flocculate rapidly in salt water (high ionic
strengthbut low electrical repulsive forces). Determination of the
thickness of thedouble layer as function of the ionic strength is
illustrated in Example 9-1.
EFFECT OF COUNTERIONS
If the charge on the counterions in solution is altered, the
thickness of theEDL will be reduced, as illustrated in Table 9-3.
The ionic concentrationthat results in the reduction of the EDL to
the point where flocculationoccurs is defined as the critical
coagulation concentration (CCC) and willdepend on the type of
particulate as well as the dissolved ions. Accordingto the DLVO
theory, the CCC is inversely proportional to the sixth powerof the
charge on the ion. Thus, the CCC values for mono-, di-, and
trivalentions are in the ratio of 1: 12
6: 13
6, or 100: 1.6: 0.14 percent, assuming that the
electrolytes do not adsorb or precipitate. The above
relationship is knownas the Schultz–Hardy rule, which was
originally observed in the 1880s
IBMHighlight
IBMHighlight
-
556 9 Coagulation and Flocculation
Example 9-1 Determination of thickness of electricaldouble
layer
Verify that the values in Table 9-3 are correct for 0.001 M
solutions ofmonovalent and divalent ions using Eq. 9-4.
Solution1. Determine the ionic strength I for a molarity of
0.001 for chemical
constituents with a charge of 1 and 2.a. Determine the ionic
strength for Z = +1 and −1:
1 = 12
∑Z2M = 1
2
∑ (+1)2 (0.001) + (−1)2 (0.001) = 0.001 mol/Lb. Determine the
ionic strength for Z = +2 and −2:
1 = 12
∑Z2M = 1
2
∑ (+2)2 (0.001) + (−2)2 (0.001) = 0.004 mol/L2. Substitute known
terms in Eq. 9-4 and solve for 1/κ:
a. For M = 0.001, Z = +1, −1, and I = 0.001 mol/L,κ−1 =
(1010 Å/m
)
×
⎡⎢⎣ (2)
(1000 L/m3
)(1.60219 × 10−19 C
)2(6.02205 × 1023 mol−1
)(0.001 mol/L
)(78.54)
(8.854188 × 10−12 C2/J · m
)(1.38066 × 10−23 J/K)(273 + 25 K)
⎤⎥⎦
−1/2
= 96.2 Åb. For M = 0.001, Z = +2, −2, and I = 0.004 mol/L,
κ−1 =(1010 Å/m
)
×
⎡⎢⎣ (2)
(1000 L/m3
)(1.60219 × 10−19 C
)2(6.02205 × 1023 mol−1
)(0.004 mol/L
)(78.54)
(8.854188 × 10−12 C2/J · m
)(1.38066 × 10−23 J/K)(273 + 25 K)
⎤⎥⎦
−1/2
= 48.1 Å
Comment
The above computation illustrates the importance of the charge
of the ionicspecies, as reported in Table 9-3.
IBMHighlight
-
9-3 Coagulation Theory 557
Kruyt, 1952). Thus, if 3000 mg/L of NaCl will produce rapid
flocculation ofhydrophobic particulates, then 47 mg/L of CaCl2 will
achieve similar results.It should also be noted that if multivalent
ions comprise the fixed layernext to a negatively charged particle,
the EDL will be reduced significantlyand the CCC value would be
much lower than predicted by the theory (forthe Schultz–Hardy
rule).
9-3 Coagulation Theory
The electrical properties of particles were considered in the
previoussection. Coagulation, as described in Sec. 9-1, is the
process used to destabi-lize the particles found in waters so that
they may be removed by subsequentseparation processes. The purpose
of this section is to introduce the prin-cipal coagulation
mechanisms responsible for particle destabilization andremoval.
Coagulation practice including the principal chemicals used
forcoagulation in water treatment and jar testing is presented and
discussedin Sec. 9-4.
Mechanisms that can be exploited to achieve particulate
destabilizationinclude (1) compression of the electrical double
layer, (2) adsorptionand charge neutralization, (3) adsorption and
interparticle bridging, and(4) enmeshment in a precipitate, or
‘‘sweep floc.’’ While these mechanismsare discussed separately
here, it will become apparent that each onehas certain pitfalls,
and this is the reason that destabilization strategiesexploit
several mechanisms simultaneously. It should also be noted
thatcompression of the electrical double layer, discussed in the
previous section,is also considered a coagulation mechanism but is
not discussed herebecause increasing the ionic strength is not
practiced in water treatment.
Adsorptionand Charge
Neutralization
Particulates can be destabilized by adsorption of oppositely
charged ions orpolymers. Most particulates in natural waters are
negatively charged (clays,humic acids, bacteria) in the neutral pH
range (pH 6 to 8); consequently,hydrolyzed metal salts,
prehydrolyzed metal salts, and cationic organicpolymers can be used
to destabilize particles through charge neutralization.Cationic
organic polymers can be used as primary coagulants, but they
aremost often used in conjunction with inorganic coagulants to form
particlebridges, as discussed below. Generally, the optimum
coagulant dose occurswhen the particle surface is only partially
covered (less than 50 percent).Polymers of high charge density and
low to moderate molecular weights(10,000 to 100,000) are believed
to be adsorbed on negatively chargedparticles as a patch on the
surface and do not extend much from thesurface. The optimum dose
appears to increase in proportion to thesurface area concentration
of the particulates.
When the proper amount of polymer has adsorbed, the charge
isneutralized and the particle will flocculate. When too much
polymer has
IBMHighlight
-
558 9 Coagulation and Flocculation
Figure 9-7Destabilization of a kaolinite claysuspension with
cationic polymer No. 4.Initial clay concentration = 73.2
mg/L.(Adapted from Black et al., 1966.)
0 0.2 0.4 0.6 0.8 1.0
Polymer dosage, mg/L
Res
idua
ltu
rbid
ity, N
TU
Ele
ctro
phor
etic
mob
ility
,(μ
m/s
)/(V
/cm
)
+2
+1
0
−2
−1
0
0.2
0.1
0.3
0.4
0.5
120
100
80
60
40
200
Pol
ymer
adso
rbed
, mg/
L
been added, the particles will attain a positive charge and
become stableonce again. This phenomenon is demonstrated by the
classical experimentsof Black et al. (1966), which are shown on
Fig. 9-7. For polymer dosagesup to 0.7 mg/L, the electrophoretic
mobility becomes more positive andthe amount adsorbed increases.
Higher dosages cause charge reversal,particle stability, and a
higher residual turbidity. Cationic polymers andpolyaluminum
chloride (PACl) are said to exhibit stoichiometry becausea certain
amount of charge exists on the particle suspension surface, andwhen
the precise amount of coagulant is added, a rapidly
flocculatingsuspension is created.
Adsorptionand InterparticleBridging
Polymer bridging is complex and has not been adequately
describedanalytically. Schematically, polymer chains adsorb on
particle surfaces atone or more sites along the polymer chain as a
result of (1) coulom-bic (charge–charge) interactions, (2) dipole
interaction, (3) hydrogenbonding, and (4) van der Waals forces of
attraction (Hunter, 2001). Therest of the polymer may remain
extended into the solution and adsorbon available surface sites of
other particles, thus creating a ‘‘bridge’’between particle
surfaces. If the extended polymer cannot find vacant siteson the
surface of other particulates, no bridging will occur. Thus,
there
-
9-3 Coagulation Theory 559
is an optimum degree of coverage or extent of polymer adsorption
atwhich the rate of aggregation will be a maximum. Polymer bridging
isan adsorption phenomenon; consequently, the optimum coagulant
dosewill generally be proportional to the concentration of
particulates present.Adsorption and interparticle bridging occur
with nonionic polymers andhigh-molecular-weight (MW 105 to 107),
low-surface-charge polymers. High-molecular-weight cationic
polymers have a high charge density to neutralizesurface
charge.
REACTION MECHANISMS FOR POLYMERS
A schematic of the reaction mechanisms for polymers is shown on
Fig. 9-8.At the optimum dosage of polymer shown in reaction (a),
the particles aredestabilized and can subsequently flocculate, as
shown in reaction (b). Ifthe particle concentration is very low or
if adequate mixing does not allowflocculation, then nonadsorbed
ends of the polymers will eventually adsorbon the destabilized
particle, causing it to restabilize, as shown in reaction(c). If
too much polymer is added, all adsorption sites will be taken up
andthe particle will not flocculate, as shown in reaction (d). If
the particlesare mixed for too long or too intensively, they will
break up, as shown inreaction (e).
POLYMER SELECTION
Because polymer–solution interactions are complex, polymer
selection isbased on empirical testing. In general, though, anionic
polymers havebeen shown to be effective flocculant aids, while
nonionic polymers havebeen effective as filter aids. Polymer
selection for sludge conditioningis dependent on sludge properties,
polymer properties, and the mixingenvironment (O’Brien and Novak,
1977). Polymer bridging is the dominantmechanism in sludge
conditioning, and thus polymer molecular weight isthe dominant
property of interest. For each system, the optimum polymerdose,
mixing conditions, and pH must be determined empirically.
Precipitationand Enmeshment
When high enough dosages are used, aluminum and iron form
insolubleprecipitates and particulate matter becomes entrapped in
the amorphousprecipitates. This type of destabilization has been
described as precipitationand enmeshment or sweep floc (Packham,
1965; Stumm and O’Melia, 1968).Although the molecular events
leading to sweep floc have not been definedclearly, the steps for
iron and aluminum salts at lower coagulant dosages areas follows:
(1) hydrolysis and polymerization of metal ions, (2) adsorption
ofhydrolysis products at the interface, and (3) charge
neutralization. At highdosages, it is likely that nucleation of the
precipitate occurs on the surfaceof particulates, leading to the
growth of an amorphous precipitate withthe entrapment of particles
in this amorphous structure. This mechanismpredominates in water
treatment applications where pH values are generallymaintained
between pH 6 and 8, and aluminum or iron salts are used at
-
560 9 Coagulation and Flocculation
Particle destabilizationresults from polymer
bonding
Particles and polymerflocculate due to perikinetic
and orthokinetic forces
Polymer added toparticulate suspension
at correct dosage
StableparticlesPolymer
Excessive dosageof polymer added
Insufficient mixing conditionsresults in particle
restabilization
and poor floc formation
Correctmixing
Excessivemixing
Destabilizedparticles Floc
particle
Particles and polymerflocculate due to perikinetic
and orthokinetic forces
Floc ruptures dueto high or prolonged
mixing conditions
Flocparticle
(a) (b)
(c) (d)
(e)
Secondaryadsorption
Inactive particles
Flocfragments
Particlesenmeshedin polymer
matrix
Figure 9-8Schematic representation of the bridging model for the
destabilization of particles by polymers. (Adapted from O’Melia,
1972.)
concentrations exceeding saturation with respect to the
amorphous metalhydroxide solid that is formed.
One interesting finding regarding sweep floc is that, in
general, thesweep floc mechanism does not depend on the type of
particle, and,thus, the same dosage of coagulant is required for
sweep floc formationregardless of the type of particles that may be
present (in the absence of
-
9-4 Coagulation Practice 561
54
Al 2
(SO
4)3
.14H
2O, m
g/L
6 7 8 9 10
pH
0
50
100
150
200
Various claysOrganics
Legend
Figure 9-9Coagulation of various types of clays and
organics,which supports hypothesis that sweep floc is notinfluenced
by type of particles present: (©) clays and(�) organics. (Adapted
from Packman, 1962.)
NOM). The dosage of alum required to reduce the turbidity of a
variety ofparticles is displayed on Fig. 9-9. Although the dosage
does not depend onthe type of particles, it does depend on the pH,
as expected. However, acaveat that should be mentioned is that the
coagulant demand exerted byNOM is not reflected on Fig. 9-9. Thus,
the concentration of hydrolyzingmetal salts that is required for
sweep floc will depend on the type andconcentration of NOM, which
unfortunately is site specific. The effects ofNOM on coagulation
practice are considered in Sec. 9-5.
9-4 Coagulation Practice
Selection of the type and dose of coagulant depends on the
characteristicsof the coagulant, the concentration and type of
particulates, concentrationand characteristics of NOM, water
temperature, and water quality. Atpresent, the interdependence of
these five elements is only understoodqualitatively, and prediction
of the optimum coagulant combination fromcharacteristics of the
particulates and the water quality is not yet possible.The purpose
of this section is to introduce coagulation practice, includingthe
types of inorganic and organic coagulants and coagulant aids used,
andalternative techniques used to reduce coagulant dosages.
Inorganic MetallicCoagulants
Inorganic coagulants, coagulant aids, and the chemicals used for
alkalinityand pH adjustment are summarized in Table 9-4. Of the
chemicals listedin Table 9-4, the principal inorganic coagulants
used in water treatmentare salts of aluminum and ferric ions and
prehydrolyzed salts of thesemetals. These hydrolyzable metal
cations are readily available as sulfate or
-
562 9 Coagulation and Flocculation
Table 9-4Common inorganic coagulants, coagulant aids, and pH and
alkalinity adjusting chemicals used inwater treatment
MolecularClassification Chemical Formula Weight, g/mol
Application
CoagulantsAluminum sulfate Al2(SO4)3 · 14H2O 594.4 Primary
coagulantSodium aluminate Na2Al2O4 163.9 Used with alum;
provides
alkalinity and pH controlAluminum chloride AlCl3 160.3 Used in
blends with organic
polymersPolyaluminum chloride(PACl)a
Ala(OH)b(Cl)c(SO4)d Variable Primary coagulant
Polyaluminum sulfate (PAS)b Ala(OH)b(Cl)c(SO4)d Variable Primary
coagulant, producedonsite
Polyiron chloridec Fea(OH)b(Cl)c(SO4)d Variable Primary
coagulant, producedonsite
Ferric chloride FeCl3 162.2 Primary coagulantFerric sulfate
Fe2(SO4)3 400.0 Primary coagulant
Coagulant aidsActivated silica SiO2 60.0 Coagulant aid used with
alum
during cold winter monthsSodium silicate Na2O(SiO2)3−25 242–1562
Coagulant aid, produced onsiteBentonite Al2Si2O5(OH)4 258 Used to
provide nucleation
sites for enhanced coagulation
Alkalinity and pH adjustmentCalcium hydroxide Ca(OH)2 56.1 as
CaO Used to provide alkalinity and
adjust pHSodium hydroxide NaOH 40.0 Used to provide alkalinity
and
adjust pHSoda ash Na2CO3 106.0 Used to provide alkalinity
and
adjust pH
aPrehydrolyzed metal salts made from aluminum
chloride.bPrehydrolyzed metal salts made from aluminum
sulfate.cPrehydrolyzed metal salts made from iron chloride.
chloride salts in both liquid and solid (dry) forms. In the
United States, thepredominant water treatment coagulant is aluminum
sulfate, or ‘‘alum,’’sold in a hydrated form as Al2(SO4)3 · xH2O,
where x is usually 14 becauseit is the least expensive coagulant.
The action, solubility, and application ofthese coagulants are
considered in the following discussion.
ACTION OF ALUM AND IRON SALTS
When ferric or aluminum ions are added to water, a number of
paralleland sequential reactions occur. Initially, when a salt of
Al(III) and Fe(III)
IBMHighlight
IBMHighlight
IBMHighlight
-
9-4 Coagulation Practice 563
is added to water, it will dissociate to yield trivalent Al3+
and Fe3+ ions, asgiven below:
Al2 (SO4)3 � 2Al3+ + 3SO42− (9-5)FeCl3 � Fe3+ + 3Cl− (9-6)
The trivalent ions of Al3+ and Fe3+ then hydrate to form the
aquometalcomplexes Al(H2O)63+ and Fe(H2O)63+, as shown on the
left-hand side ofEq. 9-7. As shown, the metal ion (aluminum in this
case) has a coordinationnumber of 6 and six water molecules orient
themselves around themetal ion:⎡
⎣H2O OH2H2O − Al − OH2H2O OH2
⎤⎦
3+
�
⎡⎣H2O OHH2O − Al − OH2
H2O OH2
⎤⎦
2+
+ H+ (9-7)
These aquometal complexes then pass through a series of
hydrolytic reac-tions, as illustrated on the right-hand side of Eq.
9-8, which give riseto the formation of a variety of soluble
mononuclear (one aluminumion) and polynuclear (several aluminum
ions) species, as illustrated onFig. 9-10. The mononuclear
species—Al(H2O)5(OH)2+ [or just Al(OH)2+]and Al(H2O)4(OH)2+ [or
just Al(OH)2+]—are among the many speciesformed. Similarly, iron
forms a variety of soluble species, including mononu-clear species
(one iron ion) such as Fe(H2O)5(OH)2+ [or just Fe(OH)2+]and
Fe(H2O)4(OH)2+ [or just Fe(OH)2+].
Al(H2O)63+
Al(OH)(H2O)52+
Al(OH)3(s)
Al(OH)4−
Hydrogen ion
Hydrogen ion
Hydrogen ion
Hydrogen ion
Aquo Al ion
Mononuclear species
Polynuclear species
Precipitate
Aluminate ion
Al13O4(OH)247+
Figure 9-10Aluminum hydrolysis products. The dashed lines are
used todenote an unknown sequence of reactions. (Adapted
fromLetterman, 1981)
-
564 9 Coagulation and Flocculation
Polynuclear species such as Al18(OH)204+ form via hydroxyl
bridges. Forexample, a hydroxyl bridge for two aluminum atoms is
shown below:
2 Al(H2O)5(OH)2+ [(H2O)4
OH
Al Al
OH
(H2O)4]4+ + 2H2O (9-8)
It should be noted that all of these mononuclear and polynuclear
speciescan interact with the particles in water, depending on the
characteristics ofthe water and the number of particles.
Unfortunately, it is difficult to controland know which mononuclear
and polynuclear species are operative. Aswill be discussed later,
this uncertainty gave rise to the development ofprehydrolyzed metal
salt coagulants.
SOLUBILITY OF METAL SALTS
The solubility of the various alum [Al(III)] and iron [Fe(III)]
species areillustrated on Figs. 9-11a and 9-11b, respectively, in
which the log molar con-centrations have been plotted versus pH.
The equilibrium diagrams shownon Figs. 9-11a and 9-11b were created
using equilibrium constants for themajor hydrolysis reactions that
have been estimated after approximately 1 hof reaction time (upper
limit of coagulation/flocculation detention times).Accordingly, the
composition of aluminum and iron species in contact withthe freshly
precipitated hydroxide (amorphous) is illustrated on Figs. 9-11aand
9-11b. In preparing these diagrams, only the mononuclear species
for
−8
−7
−6
−5
−4
−3
−2
1
10
100300
30
3
0.3
0 2 4 6 8 10 12 14
log[
Al],
mol
/L
pH of mixed solution pH of mixed solution
−12
−10
−8
−6
−4
−2
0
0.27
2.727270
0 2 4 6 8 10 12 14
log[
Fe]
, mol
/L
(a) (b)
Fer
ric a
s F
eCl 3
. 6 H
2O, m
g/L
Alu
m a
s A
l 2(S
O4)
3. 1
4.3
H2O
, mg/
L
Adsorptiondestabilization
Sweepcoagulation
Fe3+
Fetotal
Fe(OH)2
+
Fe(OH)4−
(AM) Fe(OH) (s)3
Fe(OH)2+
Restabilization zone(changes with colloidor surface area)
Charge neutralizationto zero zeta potentialwith Al(OH)3 (s)
Charge neutralization tozero zeta potential withAlx(OH)y
n+/Al(OH)3 (s)
Sweepcoagulation Al(OH)4
-
Al(OH)2+
Altotal
Adsorptiondestabilization
Restabil-ization zone
(boundarychanges
with colloid)
Al3+
Combination(sweep andadsorption)
Optimal sweep
(AM) Al(OH)3 (s)
Figure 9-11Solubility diagram for (a) Al(III) and (b) Fe(III) at
25◦C. Only the mononuclear species have been plotted. The metal
species areassumed to be in equilibrium with the amorphous
precipitated solid phase. Typical operating ranges for coagulants:
(a) alumand (b) iron. (Adapted from Amirtharajah and Mills,
1982)
-
9-4 Coagulation Practice 565
Table 9-5Reactions and associated equilibrium constants for
aluminum and iron species in equilibriumwith amorphous aluminum
hydroxide and ferric hydroxide
Acid Equilibrium Constants
Equilibrium Range, Used forReaction Constant log K Fig. 9-11
Aluminum, Al(III)Al(OH)3(s) + 3H+ → Al3+ + 3H2O Ks0 9.0–10.8
10.8Al(OH)3(s) + 2H+ → Al(OH)2+ + 2H2O Ks1 4.0–5.8 5.8Al(OH)3(s) +
H+ → Al(OH)2+ + H2O Ks2 0.7–1.5 0.7Al(OH)3(s) → Al(OH)03 Ks3 −4.2
to −6.1 −6.1Al(OH)3(s) + H2O → Al(OH)4− + H+ Ks4 −7.7 to −12.5
−11.9Species not considered: Al2(OH)24+ ,
Al8(OH)204+, Al13O4(OH)247+ , Al14(OH)3210+
Iron, Fe(III)Fe(OH)3(s) + 3H+ → Fe3+ + 3H2O Ks0 3.2–4.891
3.2Fe(OH)3(s) + 2H+ → Fe(OH)2+ + 2H2O Ks1 0.91–2.701 1.0Fe(OH)3(s)
+ H+ → Fe(OH)2+ + H2O Ks2 −0.779 to −2.5 −2.5Fe(OH)3(s) → Fe(OH)30
Ks3 −8.709 to −12.0 −12.0Fe(OH)3(s) + H2O → Fe(OH)4− + H+ Ks4
−16.709 to −19 −18.4Species not considered: Fe2(OH)24+ ,
Fe3(OH)45+
Source: Benefield et al. (1982), McMurry and Fay (2003), Morel
and Hering (1993), Nordstrom and May (1989a, b), Pankow(1991),
Snoeyink and Jenkins (1980), Sawyer et al. (2002), and Stumm and
Morgan (1981).
alum and iron have been plotted. The various mononuclear species
foralum and iron are given in Table 9-5, along with the
corresponding rangeof acid solubility products reported in the
literature and the values usedto prepare Figs. 9-11a and 9-11b. The
approximate total concentration ofresidual soluble alum (see Fig.
9-11a) and iron (see Fig. 9-11b) in solutionafter precipitation is
identified by the solid line. Aluminum hydroxide andferric
hydroxide are precipitated within the shaded areas, and
polynuclearand polymeric species are formed outside of the shaded
areas at higher andlower pH values. It should also be noted that
the structure of the precip-itated iron is far more compact and
inert as compared to the amorphousnature of precipitated
aluminum.
In most water treatment applications for removal of turbidity,
disinfec-tion by-product precursors (NOM), and color, the pH during
coagulationranges between 6 and 8. The lower limit is imposed to
prevent acceleratedcorrosion rates that occur at pH values below pH
6. The regions shown onFigs. 9-11a and 9-11b correspond to the
operating pH and dosage rangesthat are normally used in water
treatment when alum and iron are used
-
566 9 Coagulation and Flocculation
in the sweep floc mode of operation. The operating region for
aluminumhydroxide precipitation is in a pH range of 5.5 to about
7.7, with minimumsolubility occurring at a pH of about 6.2 at 25◦C,
and from about 5 to 8.5for iron precipitation, with minimum
solubility occurring at a pH of 8.0.The importance of pH in
controlling the concentration of soluble metalspecies that will
pass through the treatment process is illustrated on Figs.9-11a and
9-11b. The effect of temperature on the solubility products
foraluminum is also illustrated on Fig. 9-11a. As shown, the point
of minimumsolubility for alum shifts with temperature, which has a
significant impacton the operation of water treatment plants where
alum is used as thecoagulant. Comparing the solubility of alum and
ferric species, the ferricspecies are more insoluble than aluminum
species and are also insolubleover a wider pH range. Thus, ferric
ion is often the coagulant of choiceto aid destabilization in the
lime-softening process, which is carried out athigher pH values (pH
9).
STOICHIOMETRY OF METAL ION COAGULANTS
The overall stoichiometric reactions for aluminum and ferric ion
in theformation of hydroxide precipitates are given by Eqs. 9-9 and
9-10. As shown,each mole of trivalent ion will produce 1 mole of
the metal hydroxide and3 moles of hydrogen ions:
Al3+ + 3H2O � Al (OH)3, am↓ + 3H+ (9-9)Fe3+ + 3H2O � Fe
(OH)3,am↓ + 3H+ (9-10)
The ‘‘am’’ subscripts in Eqs. 9-9 and 9-10 refer to amorphous
solids(hours old), which have a much higher solubility product than
crystallineprecipitates.
When alum is added to water and aluminum hydroxide precipitates,
theoverall reaction is
Al2(SO4)3 · 14H2O → 2Al(OH)3↓ + 6H+ + 3SO42− + 8H2O
(9-11)Similarly, the overall reactions for ferric chloride and
ferric sulfate are asfollows:
Ferric chloride:
FeCl3 + 3H2O → Fe(OH)3↓ + 3H+ + 3Cl− (9-12)Ferric sulfate:
Fe2(SO4)3 · 9H2O → 2Fe(OH)3↓ + 6H+ + 3SO42− + 3H2O (9-13)After
Al(OH)3 or Fe(OH)3 precipitates, the species remaining in water
arethe same as if H2SO4 or HCl had been added to the water. Thus,
addingalum or ferric is like adding a strong acid. A strong acid
will lower the pHand consume alkalinity. Alkalinity is the
acid-neutralizing capacity of waterand is consumed on an equivalent
basis; that is, 1 meq/L of alum or ferric
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
-
9-4 Coagulation Practice 567
will consume 1 meq/L of alkalinity. Since alkalinity buffers
water againstchanges in pH, the change in pH following coagulant
addition depends onthe initial alkalinity. If the natural
alkalinity of the water is not sufficient tobuffer the pH, it may
be necessary to add alkalinity to the water to keepthe pH from
dropping too low. Alkalinity can be added in the form ofcaustic
soda (NaOH), lime [Ca(OH)2], or soda ash (Na2CO3). In manywater
plants, caustic soda is often used because it is easy to handle
andthe required dosage is relatively small. The reaction for alum
with causticsoda is
Al2 (SO4)3 · 14H2O + 6NaOH → 2Al (OH)3,am↓ + 3Na2SO4 +
14H2O(9-14)
The corresponding reaction for lime is given by the
expression
Al2 (SO4)3 · 14H2O + 3Ca (OH)2 → 2Al (OH)3,am↓ + 3CaSO4 +
14H2O(9-15)
Coagulants are typically purchased in a concentrated liquid
form. Cal-culating coagulant doses can be confusing because the
stock chemicalconcentration is often reported in percent by weight
and the density ofthe stock solution will be significantly heavier
than water. In addition, theextent of hydration of the alum or
ferric will vary or be unknown in the stocksolution, which affects
the formula weight of the chemical. To get aroundthis issue,
chemical manufacturers will sometimes report the concentrationof
the coagulant as a different formula entirely, for example, stock
alumconcentration is often reported as percent as Al2O3, even
though the chem-ical present is Al2(SO4)3 · xH2O. Ferric chloride
may be reported with orwithout waters of hydration (i.e., FeCl3 ·
6H2O or FeCl3) or as soluble iron(Fe3+). To calculate doses
accurately, the density and chemical formulaused by the chemical
manufacturer to report the concentration must beknown. The
application of these principles and the above equations is
illus-trated in Example 9-2. Note that the sludge produced during
coagulationconsists of both the precipitate formed in the reactions
shown above andthe solids that were present in the source water.
Example 21-2 in Chap. 21demonstrates the procedure for calculating
the amount of sludge producedconsidering both components.
Example 9-2 Calculation of coagulant doses,
alkalinityconsumption, and precipitate formation
A chemical supplier reports the concentration of stock alum
chemical as8.37 percent as Al2O3 with a specific gravity of 1.32.
For the stock chemical,calculate (a) the molarity of Al3+ and (b)
the alum concentration if reportedas g/L Al2(SO4)3 · 14H2O. Also,
for a 30-mg/L alum dose applied to a
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
-
568 9 Coagulation and Flocculation
treatment plant with a capacity of 43,200 m3/d (0.5 m3/s),
calculate (c) thechemical feed rate in L/min, (d) the alkalinity
consumed (expressed as mg/Las CaCO3), (e) the amount of precipitate
produced in mg/L and kg/day,and (f) the amount of NaOH that would
need to be added to counteract theconsumption of alkalinity by
alum.
Solution1. Calculate the formula weights (FW) for Al2O3,
Al2(SO4)3 · 14H2O,
Al(OH)3, and NaOH, given molecular weights: Al = 27, O = 16, H =
1,S = 32, and Na = 23 g/mol.FW: Al2O3 = 2(27) + 3(16) = 102
g/molFW: Al2(SO4)3·14H2O = 2(27) + 3(32) + 26(16) + 28(1) = 594
g/molFW: Al(OH)3 = 27 + 3(16) + 3(1) = 78 g/molFW: NaOH = 23 + 16 +
1 = 40 g/mol
2. Calculate the molar concentration of Al3+ in the stock alum
chemical.a. Calculate the density of stock chemical:
ρstock = 1.32(1 kg/L
) = 1.32 kg/Lb. Calculate the concentration of alum in the stock
chemical as mg/L
Al2O3:
Cstock = 0.0837(1.32 kg/L
) (103 g/kg
)= 110.5 g/L Al2O3
c. Calculate the molar concentration of Al3+ in the stock
alumchemical:[
Al3+]
= 110.5 g/L Al2O3(
mol Al2O3102 g Al2O3
)(2 mol Al3+
mol Al2O3
)= 2.17 mol/L
3. Calculate the stock alum concentration if reported as g/L
Al2(SO4)3 ·14H2O.
Cstock = 110.5 g/L Al2O3(
594 g/mol alum102 g/mol Al2O3
)= 643.5 g/L alum
4. Calculate the chemical feed rate.By mass balance:
CstockQfeed = CprocessQprocess
Qfeed =CprocessQprocess
Cstock=
(30 mg/L
) (43,200 m3/d
) (103 L/m3
)643.5 g/L
(103 mg/g
) (1440 min/d
) = 1.40 L/min
-
9-4 Coagulation Practice 569
5. Calculate the alkalinity consumed using Eq. 9-11:
Alk = [30 mg/L alum] ( 1 mmol alum594 mg alum
) (3 mmol SO42−
mmol alum
)(2 meq SO42−
mmol SO42−
)
×(
1 meq alkmeq SO42−
) (50 mg CaCO3
meq alk
)= 15 mg/L as CaCO3
6. Calculate the precipitate formed using Eq. 9-11:[Al(OH)3
] = [30 mg/L alum] ( 1 mmol alum594 mg alum
) [2 mmol Al(OH)3
mmol alum
] [78 mg Al(OH)3mmol Al(OH)3
]
= 7.88 mg/L Al(OH)3
[Al(OH)3
] =(7.88 mg/L
) (43,200 m3/d
) (103 L/m3
)(106 mg/kg
) = 340 kg/d7. Calculate the NaOH dose required to counteract
the alkalinity con-
sumption using Eq. 9-14:[NaOH
] = [30 mg/L alum] ( 1 mmol alum594 mg alum
) (6 mmol NaOH
mmol alum
) (40 mg NaOHmmol NaOH
)
= 12.1 mg/L NaOH
CommentThe sludge produced by coagulation has two components the
precipitateformed by the reactions shown above and the particles
from the raw water.Calculation of the total amount of sludge
produced during coagulationconsidering both components is
illustrated in Example 21-2.
APPLICATION OF METAL SALTS IN WATER TREATMENT
Because of the sequence of reactions that follow the addition of
alum oriron salts, as discussed above and illustrated on Fig. 9-10,
it is not possible topredict a priori the performance of a
coagulation process. Consequently,jar testing is typically used for
coagulant/coagulant aid screening, andthese results must be
evaluated in the full-scale operation. Nevertheless, itis useful to
review some general aspects of coagulation practice, including(1)
the operating regions for the alum and iron, (2) interactions with
otherconstituents in water, (3) typical dosages, and (4) the
importance of initialblending when using metal salts. As noted in
Chap. 6, blending is a mixingprocess to combine two liquid streams
to achieve a specified level of unifor-mity. Guidance on the use of
coagulants is provided in Table 9-6. Additionaleffects of NOM on
the coagulation process are considered in Sec. 9-5.
IBMHighlight
-
Tabl
e9-
6Ap
plic
atio
ngu
idan
cefo
rAl
(III)
and
Fe(II
I)as
coag
ulan
tsan
dpr
ehyd
roly
zed
met
alco
agul
ants
used
inw
ater
trea
tmen
t
Wat
erQ
ualit
yC
oagu
lant
Para
met
erAl
um(II
I)Fe
(III)
PAC
I
Turb
idity
For
low
-turb
idity
wat
ers
(i.e.
,low
part
icle
conc
entr
atio
n),s
wee
pflo
cw
illbe
requ
ired.
For
low
-turb
idity
wat
ers
(i.e.
,lo
wpa
rtic
leco
ncen
trat
ion)
,sw
eep
floc
will
bere
quire
d.
For
low
-turb
idity
wat
ers
(i.e.
,low
part
icle
conc
entr
atio
n),s
wee
pflo
cw
illbe
requ
ired.
Med
ium
-bas
icity
PACl
s(4
0–5
0%)a
resu
itabl
efo
rco
ldw
ater
sw
ithlo
wtu
rbid
ity.
Alka
linity
High
alka
linity
valu
esm
ake
pHad
just
men
tfor
optim
umco
agul
atio
nm
ore
diffi
cult.
Ifsu
ffici
enta
lkal
inity
isno
tpre
sent
,sol
uble
alum
inum
isfo
rmed
,whi
chca
nre
sult
inpo
stflo
ccul
atio
nin
dow
nstr
eam
proc
esse
s.Su
pple
men
tala
lkal
inity
shou
ldbe
adde
dbe
fore
coag
ulan
t.
Alth
ough
high
alka
linity
valu
esm
ake
pHad
just
men
tfor
optim
umco
agul
atio
nm
ore
diffi
cult,
itsim
pact
onco
agul
atio
nus
ing
Feis
less
than
Al.
pHTh
eop
timum
pHra
nge
isbe
twee
n5.
5an
d7.
7bu
twill
fluct
uate
seas
onal
ly(s
eeFi
g.9-
11).
Typi
cally
,the
optim
umpH
will
bene
arer
6in
the
sum
mer
and
7in
the
cold
erw
inte
rm
onth
s.Hi
gher
pHle
vels
ofte
nco
rres
pond
tope
riods
ofal
galg
row
th,w
hich
intu
rnw
illaf
fect
the
coag
ulan
tdos
e.
The
optim
umpH
rang
eis
from
5to
8.5
orm
ore
(see
Fig.
9-11
).
PACl
sar
ele
ssse
nsiti
veto
pH.C
anbe
used
over
the
pHra
nge
of4.
5–9
.5.
570
IBMHighlight
IBMHighlight
-
NO
MTh
ere
mov
alof
NO
Mw
illno
rmal
lyco
ntro
lthe
coag
ulan
tdos
e.Re
mov
alof
NO
Mte
nds
toin
crea
seas
pHis
redu
ced.
Rem
oval
ofup
to70
%ha
sbe
enac
hiev
ed.
The
rem
oval
ofN
OM
will
norm
ally
cont
rolt
heco
agul
ant
dose
.Rem
oval
ofN
OM
tend
sto
incr
ease
aspH
isre
duce
d.Re
mov
alof
upto
80%
has
been
achi
eved
.
The
rem
oval
ofN
OM
will
norm
ally
cont
rolt
heco
agul
antd
ose.
Rem
oval
ofN
OM
tend
sto
incr
ease
asth
epH
isre
duce
d.Re
mov
als
ofup
to70
%ha
vebe
enac
hiev
ed.L
ow-b
asic
ityPA
Cls
(up
to20
%)a
resu
itabl
efo
rw
ater
shi
ghin
colo
ran
dto
talo
rgan
icca
rbon
.
Tem
pera
ture
Tem
pera
ture
affe
cts
solu
bilit
ypr
oduc
ts.F
loc
form
edin
cold
erw
ater
tend
sto
bew
eake
r.
Floc
form
edin
cold
erw
ater
tend
sto
bew
eake
r.
Mix
ing
Hydr
olys
isre
actio
nsar
eve
ryfa
st.
Mix
ing
times
shou
ldbe
less
than
1s
and
pref
erab
lyle
ssth
an0.
5s.
Hydr
olys
isre
actio
nsar
eve
ryfa
st.M
ixin
gtim
essh
ould
bele
ssth
an1
san
dpr
efer
ably
less
than
0.5
s.
Beca
use
the
PACl
ispr
ehyd
roly
zed,
the
initi
albl
endi
ngtim
eis
som
ewha
tle
sscr
itica
l.
571
-
572 9 Coagulation and Flocculation
OPERATING REGIONS FOR METAL SALTS
Because the chemistry of the various reactions discussed above
is so complex,there is no complete theory to explain the action of
hydrolyzed metal ions.To qualitatively describe the application of
alum as a function of pH, takinginto account the action of alum as
described above, Amirtharajah and Mills(1982) developed the
diagrams shown on Fig. 9-11. It is important to notethat the
generalized information represented on Fig. 9-11 does not
reflectthe effects of NOM on the dosages of coagulant required. The
approximateregions in which the different phenomena associated with
particle removalin conventional sedimentation and filtration
applications are plotted as afunction of the alum dose and the pH
of the treated effluent after alumhas been added. For example,
optimum particle removal by sweep flococcurs in the pH range of 7
to 8 with an alum dose of 20 to 60 mg/L.With proper pH control it
is possible to operate with extremely low alumdosages.
Interactions with other constituents in waterAs with all cations
in water, hydrolysis products of aluminum and iron reactwith
various ligands (e.g., SO42−, NOM, F−, PO43−) forming both
solubleand insoluble products that will influence the quantity or
dose of thecoagulant required to achieve a desired level of
particle destabilization.Thus, the optimum dose of a coagulant
depends strongly on the particularwater chemistry and the types of
particles.
Typical dosagesA typical dosage of alum ranges from 10 to 150
mg/L, depending on raw-water quality and turbidity. Typical dosages
of ferric sulfate [Fe2(SO4)3 ·9H2O] and ferric chloride (FeCl3 ·
6H2O) range from 10 to 250 mg/L and5 to 150 mg/L, respectively,
depending on raw-water quality and turbidity.Ferric chloride is
more commonly used than ferric sulfate and comes as aliquid.
Importance of initial mixing with metal saltsThe rapid initial
mixing (known as blending) of the metal salts in watertreatment is
extremely important. The sequence of reactions shown onFig. 9-10
occurs rather rapidly (Rubin and Kovac, 1974). For example, ata pH
of 4, half of the Al3+ hydrolyzes to Al(OH)2+ within 10−5 s
(Baseand Mesmer, 1976). Hudson and Wolfner (1967) noted that
‘‘coagulantshydrolyze and begin to polymerize in a fraction of
second after beingadded to water.’’ Hahn and Stumm (1968), studying
the coagulation ofsilica dispersions with Al(III), reported that
the time required for theformation of mono- and polynuclear
hydroxide species was on the order of10−3 s, and the time of
formation for the polymer species was on the orderof 10−2 s. The
importance of initial and rapid mixing is also discussed
byAmirtharajah and Mills (1982) and Vrale and Jorden (1971).
-
9-4 Coagulation Practice 573
Clearly, based on the literature and actual field evaluations,
the instan-taneous rapid and intense mixing of metal salts is of
critical importance,especially where the metal salts are to be used
as coagulants to lower thesurface charge of the colloidal
particles. It should be noted that, althoughachieving extremely low
blending times in large treatment plants is oftendifficult, low
blending times can be achieved by using multiple mixers.Typical
blending times for various chemicals are reported in Table 6-10
inSec. 6-10, where the subject of mixing is considered in
detail.
PrehydrolyzedMetal Salts
From the previous discussion of the use of alum and iron salts,
it is clearthat it is difficult to control the metal species
formed, especially at lowdosages. The unpredictability associated
with alum and iron salts led to thedevelopment of prehydrolyzed
metal salts. Prehydrolyzed metal salts areprepared by reacting alum
or ferric with various salts (e.g., chloride, sulfate)and water and
hydroxide under controlled mixing conditions. Severaladvantages of
preformed aluminum metal salts include the following:(1) lower
dosages may be required for effective coagulation (on the basisof
Al3+) for cases where NOM does not dictate the coagulant dosage
atneutral or slightly acidic conditions, (2) flocs tend to be
tougher and denser(although flocculation aids are still necessary
in many cases), and (3) theperformance of prehydrolyzed alum salts
is less temperature dependent ascompared to unmodified alum salts.
General guidance on the applicationof prehydrolyzed metal salts is
given in Table 9-6.
CHEMICAL COMPOSITION
The commercial prehydrolyzed alum salts, commonly known as PACl,
havethe following overall formula: Ala(OH)b(Cl)c(SO4)d . Although
many for-mulations do not contain any sulfate; the presence of
sulfate ions helps tostabilize the aluminum polymers and keep them
from precipitating. Thesepolymers can be more effective than those
formed by simply adding alu-minum salts to solution because the
larger cationic polymers can be formedby increasing the
hydroxide-to-aluminum ratio (R = OH/Al, see followingbasicity
discussion), which can lead to enhanced charge
neutralization.Another benefit is that, as the polymer becomes
larger, it becomes morecrystalline, compact, and dense. However, as
the value of R increases, thepolymers become less stable and may
begin to precipitate, which can causea problem in the storage of
PACl.
BASICITY
As given by Eqs. 9-9 and 9-10, when metal salts such as alum and
ironhydrolyze, hydrogen ions are released, which will react with
the alkalinityof the water. In the formulation of PACl coagulants,
some of the acidthat would have been released is neutralized with
base (OH−) when thecoagulant is manufactured. The degree to which
the hydrogen ions thatwould be released by hydrolysis are
preneutralized is known as the basicity
-
574 9 Coagulation and Flocculation
of the product and is given by the following relationship for
prehydrolyzedmetal salts that do not contain oxygen:
Basicity, % = [OH][M] ZM
× 100 (9-16)
where [OH]/[M] = molar ratio of hydroxide bound to metal ionZM =
charge on metal species
For example, the basicity of the PACl Al2(OH)4Cl2 is 66.7
percent{[4/(3 × 2)] × 100}. It should be noted that, if oxygen is
included inthe formulation, the basicity of the compound will
increase by the effect ofthe oxygen. For example, the basicity for
the compound Al13O4(OH)24 is82.1 percent {[24 + (4 × 2)]/(13 × 3)]
× 100}. In effect, each mole of oxy-gen will neutralize 2 moles of
hydrogen. Most prehydrolyzed alum productshave an OH/Al ratio of
0.45 to 2.5, which corresponds to basicity values of15 [(0.45/3) ×
100] and 83.3 [(2.5/3) × 100] percent.
Organic Polymers Organic polymers are long-chain molecules
consisting of repeating chem-ical units with a structure designed
to provide distinctive physicochemicalproperties. The chemical
units usually have an ionic functional groupthat imparts an
electrical charge to the polymer chain. Hence, organicpolymers are
often termed polyelectrolytes. Organic polymers have two prin-cipal
uses in water treatment: (1) as a coagulant for the destabilization
ofparticles and (2) as a filter aid to promote the formation of
larger andmore shear-resistant flocs. While destabilization occurs
primarily throughcharge neutralization, nonionic and anionic
polymers can be used to forma bridge between particles. Organic
polymers are not generally used asprimary coagulants and are often
used after the particles have been destabi-lized to some degree
with metal coagulants. Polymers are broadly classifiedas being
natural or synthetic in origin. Because of their greater use in
watertreatment, the synthetic polymers are discussed first.
SYNTHETIC ORGANIC POLYMERS
Generally, synthetic organic polymers are much cheaper than
those madefrom natural sources and consequently are used more often
in the UnitedStates than natural organic polymers. The principal
synthetic organic poly-mers used for water treatment are summarized
in Table 9-7. Syntheticorganic polymers are made either by
homopolymerization of the monomeror by copolymerization of two
monomers. Polymer synthesis can be manip-ulated to produce polymers
of varying size (molecular weight), chargegroups, number of charge
groups per polymer chain (charge density), andvarying structure
(linear or branched). A typical example is the productionof
polyacrylamide in which the monomer, acrylamide,
homopolymerizesunder appropriate conditions to form the polymer.
Polyacrylamide carriesno ionic charge and is referred to as a
nonionic polymer. Subsequent
-
Tabl
e9-
7Ty
pica
lorg
anic
coag
ulan
tsus
edin
wat
ertr
eatm
ent
Mol
ecul
arW
eigh
t,C
omm
onTy
peC
harg
eg/
mol
eAp
plic
atio
nsTy
pica
lExa
mpl
esa
Oth
erEx
ampl
es
Anio
nic
Neg
ativ
e10
4−
107
Coag
ulan
taid
,filte
rai
d,flo
ccul
anta
id,s
ludg
eco
nditi
onin
g
Hydr
olyz
edpo
lyac
ryla
mid
esHy
drol
yzed
poly
acry
lam
ides
,po
lyac
ryla
tes,
poly
acry
licac
id,p
olys
tyre
nesu
lfona
te
CH
2
NH
2
CH
CO
x
CH
2
Na+
CH
CO
Oy
Catio
nic
Posi
tive
104
−10
6Pr
imar
yco
agul
ant,
turb
idity
and
colo
rre
mov
al
Epic
hlor
ohyd
rindi
met
hyla
min
e(e
pi-D
MA)
Amin
omet
hyl
poly
acry
lam
ide,
poly
alky
lene
,pol
yam
ines
,po
lyet
hyle
nim
ine
N+
Cl−
CH
2C
H2
CH
CH
3
CH
3
OH
x
Slud
geco
nditi
onin
gPo
lydi
ally
ldim
ethy
lam
mon
ium
chlo
ride
(pol
y-DA
DMAC
)Po
lydi
met
hyl
amin
omet
hyl
poly
acry
lam
ide,
poly
viny
lben
zyl,
trim
ethy
lam
mon
ium
chlo
ride
N+
Cl−
CH
2
CH
CH
2
CH
3C
H3
CH
x
N+
Cl−
CH
2
CH
CH
CH
2CH
3
CH
2C
H3
y
Non
ioni
cN
eutr
al10
5−
107
Coag
ulan
taid
,filte
rai
d,fil
ter
cond
ition
ing
Poly
acry
lam
ides
Poly
acry
lam
ides
,po
lyet
hyle
neox
ide
CH
2
NH
2
CH
CO
x
Oth
ers
Varia
ble
Varia
ble
—So
dium
algi
nate
Algi
nic
acid
,dex
tran
,gu
argu
m,s
tarc
hde
rivat
ives
a Num
ber
ofm
onom
erm
olec
ules
inpo
lym
erde
sign
ated
byx
and
y.
575
IBMHighlight
-
576 9 Coagulation and Flocculation
hydrolysis of polyacrylamide under basic conditions produces a
polymerwith anionic charges. Thus, the number of anionic groups, in
this case acarboxyl group, can be controlled, providing anionic
polymers of differentmolecular weights and charge density. The
third type of polymer has acationic or positive charge group
incorporated in the polymer chain, usuallyby a copolymerization
process.
APPLICATION OF POLYMERS
Since their introduction in the early 1950s, the use of organic
polyelec-trolytes such as poly diallyl-dimethyl ammonium chloride
(poly-DADMAC)and epichorohydrin dimethylamine (epi-DMA) (see Table
9-7) has gainedwidespread use for water treatment in the United
States. The MW rangesfrom 104 to 105 and the basic polymer units
are shown in Table 9-7.
Cationic polymersIn water treatment applications, cationic
organic polymers are generallydesigned to be water soluble, to
adsorb on or react rapidly with particulates,and to possess a
chemical structure suitable for the intended use. When usedas
primary coagulants, cationic polymers, in contrast to aluminum or
ferricions, do not produce large floc volumes because organic
coagulants canbe effective at much lower coagulant dosages than
inorganic coagulants.However, sludge from organic coagulants is
usually denser and stickierthan sludge from inorganic coagulants.
Consequently, cationic organiccoagulants are not suitable for every
type of separation process.
It should be noted that, because organic coagulants do not
always pro-duce the same water quality as is obtained with metallic
ion coagulants,cationic organic polymers are rarely used alone
except for direct filtration.Furthermore, if cationic organic
polymers are used alone, they are ineffec-tive in removing
dissolved substances (e.g., NOM, As, F). It is common touse
cationic organic polymers and metallic ion coagulants together.
Themain advantage of the combined usage is that the dosage of
metallic ioncoagulants can be reduced by 40 to 80 percent. The
lower metallic ioncoagulant dosage in turn reduces sludge and
alkalinity consumption. Withlower alkalinity consumption, the pH
will not be depressed as much, whichcan improve metallic ion
coagulation.
Polymer dosagesBecause of the complex interactions between
polymers and particulatesand the uncertain influence of water
quality on these interactions, polymerselection is empirical. The
typical dosage rates for sedimentation are onthe order of 1 to 10
mg/L for DADMAC and epi-DMA. Low dosages ofhigh-molecular-weight
nonionic polymers (0.005 to 0.05 mg/L) are oftenadded before
granular filtration and to the backwash water to improve
filterperformance. Incorrect dosing can cause mudball formation in
the filters,which are not usually broken apart during normal
backwashing operations.
-
9-4 Coagulation Practice 577
Impact of solution parametersSolution parameters will also
impact polymer dose. If the polymer chargedensity depends on pH, as
with nonquaternized polyamines (see Table 9-7),then the optimum
polymer dose will vary with pH, generally decreasing asthe pH
decreases [the charge on secondary and tertiary amines dependson pH
because the amine group will tend to protonate at lower pH
values(less than 6) and will remain uncharged at neutral pH]. The
charge densityof quaternized polymers such as poly-DADMAC are only
slightly affectedby pH. Changes in ionic strength and composition
do not appear to affectpolymer dose strongly over typical ranges
encountered in water treatment(TDS between 50 and 500 mg/L).
MIXING OF POLYMERS
Most polymers are available in liquid form and can be used
without apreparation stage, but they must be injected directly
following in-linedilution. Successful use of polymers in water
treatment requires adequatedispersion of the polymer to promote
more uniform polymer adsorption.Jar testing, as described in the
previous section, may be used to assess theeffect of mixing.
NATURAL POLYMERS
Sodium alginate is a natural organic polymer extracted from
brown seaweed.The polymeric structure of sodium alginate is
comprised of mannuronicacid and glucuronic acid and the molecular
weight is on the order of104 to 2 × 105 (Degrémont, 2007). Sodium
alginate is particularly effectiveas a flocculant aid with ferric
salts, and good results have also beenobtained with aluminum salts,
with typical dosages ranging from 0.5 and2 mg/L. Chitosan, another
natural organic, is obtained from chitin shells(crab, lobster,
etc.). Natural starches are also classified as natural
polymers.Starches can be obtained from a number of sources,
including potatoes,tapioca, or plant seed extracts. Starches are
branched and nonlinearglucopyranose polymers, which are sometimes
partially broken down withOH− or derivatized to form
carboxy-ethyl-dextrose. Starches are used inconcentrations of 1 to
10 mg/L, preferably together with aluminum salts.
Coagulantand Flocculant
Aids
A variety of chemicals and additives known as coagulant aids and
flocculantaids, used to enhance coagulation and flocculation
processes, are describedbelow. Some of the commonly used inorganic
coagulant aids are given inTable 9-4. Polymers used as flocculant
aids are given in Table 9-7.
COAGULANT AIDS
Coagulant aids, typically insoluble particulate materials, are
added toenhance the coagulation process. Clay (bentonite,
kaolinite), sodium sili-cate, pure precipitated calcium carbonate,
diatomite, powdered activated
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
IBMHighlight
-
578 9 Coagulation and Flocculation
carbon (used as an adsorbent), and fine sand have all been used
as coag-ulation aids. Coagulant aids are often added to waters that
contain lowconcentrations of particles to form nucleating sites for
the formation oflarger flocs. Coagulant aids are used in
conjunction with inorganic coagu-lants, organic polyelectolytes, or
both. Because the density of these particlesis higher than that of
most floc particles, the settling velocities of
flocculatedparticles is increased.
FLOCCULANT AIDS
Uncharged and negatively charged organic polymers that were
discussedin the previous section are used as flocculant aids as
opposed to primarycoagulants. As previously discussed, the main
advantage of using flocculantaids is that a stronger floc is
formed. Flocculant aids are added after thecoagulants are added and
the particles are already destabilized. The timerequired for
destabilization depends on water temperature and the typeof
particles; consequently, jar tests have to be conducted. The
importantfactors that need to be evaluated in jars and full-scale
implementation arefloc strength, size, and settling rate. It should
be noted that improperdosing of flocculant aids can cause mudballs
to form in gravity filters thatare not easily eliminated by
backwashing.
Activated silica is an important inorganic flocculant aid that
is used incombination with alum and can be effective in cold water.
It is usually storedas sodium silicate, which is soluble at high
pH. Usually, the concentratedsodium silicate is partially
neutralized (usually with sulfuric acid) priorto use and then added
immediately to the water. In some instances,aluminosilicate is used
where alum is the primary