Colloid Destabilization With

Cationic Polyelectrolytes as
Affected by Velocity Gradients
water
technology John J. Morrow and Emerson G. Rausch
Cationic poly electrolytes can effectively destabilize colloidal and particu-
late matter without the use of inorganic salts by one's applying increased
mixing energy to rapid-mix basins. A comparison of the sludges produced
by poly electrolytes and inorganic coagulants indicates that polymer sludge
has a greater (jensity, more rapid settling, improved dewatering charac
teristics, and significantly lower volume. Because of this and the increased
ease of handling, cationic polyelectrolytes provide the advantage of lower
overall treatment costs.

A paper presented at the Annual Con tention time with intense mixing and to
ference on May 15, 1973, by John J. yield velocity gradients in order of mag
Morrow (Active Member, AWWA), nitude of 300 fps/ft (s-1 ) or more. The
chemist; and Emerson G. Rausch (Active purpose of rapid mixing is to provide
Member, AWWA), group leader R&D, complete and uniform distribution of
Cafgon Corp., Pittsburgh, Pa. the coagulant and the energy necessary
to drive colloidal particles together by
chemical or electrostatic forces. How
ever, in some instances a longer deten
tion time with less intense mixing is
Increased demands for greater vol preferred, and flow through baffled ba
umes of high-quality water are prompt sins provides an alternative to mechani
ing municipalities and industries to ex cal agitation. Flocculation basins, in
amine new products and application which the coagulated destabilized parti
techniques in an effort to improve exist cles coalesce and compact have a veloc
ing water-treatment facilities.1'5 One ity-gradient range of 5-100 s~l . The
method that has been successful on a lesser velocity gradients are particularly
pilot scale incorporates the use of syn important in instances of weak floe for
thetic cationic polyelectrolytes as pri mation.
mary coagulants;. The results from field use of a pilot-
This article discusses the use of water- clarification unit when employing con
soluble synthetic cationic polyelectro ventional mixing have indicated that at a
lytes as primary coagulants and the ap velocity gradient of 300 s"1 , cationic
plication of high-energy rapid mixing for polyelectrolytes are not always effective
colloid destabilization. The effect that in producing complete particle desta
velocity gradients have on performance bilization. This may be attributed to
and the removal of natural-water tur nonuniform distribution of the polyelec-
bidity is compared with treatments that trolyte in conventional rapid-mix basins.
employ organic coagulants, synthetic ca Also, a short contact time limits the in
tionic polyelectrolytes applied in addi teraction between colloidal particles in
tion to inorganic coagulants, and syn the water and the polyelectrolyte that is
thetic polyelectrolytes alone. used.
Pilot studies were conducted at three Experience has shown that cationic
sites: (1) the ,'S. S. Sahli water plant in polyelectrolytes are effective when ap
New Brighton, Pa. located on the Beaver plied at the head of a sufficiently long
River; (2) the central plant of the St. raw-water intake line. The benefits of
Louis County Water Co., Chesterfield, extended mixing include increased tur
Mo. located on the Missouri River; and bulence, sufficient contact time, and
(3) the East St. Louis-Interurban plant, better distribution of the polymer with
East St. Louiii, 111., located along the the suspended particles in the |ine. How
Mississippi River. Velocity gradients ever, this is only possible at a small num
were used to determine the intensity of ber of treatment facilities. With an in
rapid mixing required to coagulate the crease in the power input, high-energy
raw water successfully from each river mixing in the rapid-mix zone can in
evaluated. crease coagulant dispersion and the fre
quency of particle collisions and thus in
Rapid Mixing crease the effectiveness of polyelec
Generally, mechanical rapid-mixing trolytes in obtaining complete particle
units are designed to provide 10-60-s de destabilization.

646 WATER TECHNOLOGY/QUALITY JOURNAL AWWA

 By one's applying a high degree of measuring the amperage of the running Design of Pilot Unit
mixing turbulence to rapid-mix basins, motor across the electric circuit and
cationic salts completely as primary calculated into horsepower. The water The pilot clarification unit is designed
coagulants. The ability of cationic poly- horsepower is the difference between as a conventional rapid-mix clarification
electrolytes to provide efficient removal running the mixing motor with an at plant scaled down to deliver a raw-water
of particulate matter from water sources tached blade in air and in water. flow rate of 0.5 gpm. The pilot unit is
of both low (1-15 Jtu) and high Velocity gradient determinations for fabricated from stainless steel and con
(50-2000 Jtu) turbidity has been dem existing plants that use mechanical flash sists of three interconnected rapid-mix
onstrated by using velocity gradient mixers can be calculated easily by using basins sized to contain 5 gal of water
values of 400 s"1 or greater. It has been the previous formula. The water horse each with several inches of freeboard.
discovered that coagulant dosage de power, P, can be obtained by calculating Each basin contains a variable-speed
creases as the velocity gradient increases the product of the amperage drawn by laboratory mixer. The effluent from the
and that the time necessary to provide the mixer, the voltage of the power line, third rapid-mix basin flows by rein
complete distribution of the polymer and the power factor if alternating cur forced rubber hose to a two-channel
and colloid destabilization also decreases rent is used, divided by 746. Neglecting slow-mix flocculation tank containing
as the velocity gradient is increased. frictional drag, this value can be used revolving variable-speed mechanical
The completeness of the coagulation directly in the formula to provide an ap paddles.
process is dependent upon the number proximation of the velocity gradient. The flocculation tank effluent flows
of particle collisions. This is directly pro into a settling tank, the bottom of which
portional to the velocity gradient G times 1 hp = 746 watts consists of two V-shaped basins with
the detention time T, and the rate at amperage x voltage motor-operated stainless-steel ball
which these collisions occur. The value x power factor valves for sludge collection. Exiting
of GT is ordinarily in the range of hp of the mixer from the settling tank by an overflow
3 000-50 000 for most rapid-mixing ap
746 weir, the water proceeds to a dual-media
plications. However, as the value of G is pilot filter where the effluent is filtered
For baffled basins the following for and the finished water is obtained. The
increased, the destabilized particles be mula is used: 7 flow scheme is shown in Fig. 1.
come subject to shearing forces. /62.4/7
It is apparent that the shear resistance Three rapid-mix tanks provide the
of the formed floe varies with different in which benefit of evaluating the high-energy
^T mix and polymer feed points at different
structured types of polymers. Retention locations as well as providing for tapered
times from several seconds to less than H = head loss due to friction in feet
two minutes during rapid mixing would T = detention time in seconds mixing rates and flocculation. The de
be the optimum range for coagulation
M = viscosity sign of the pilot unit at 0.5 gpm provides
since shearing of the polyelectrolyte and Preliminary Laboratory and Pilot 15 min of detention time in each floc
adsorbed particles occurs at longer re culation channel and approximately 4 hr
tention times. Camp6 discovered that,
Unit Studies settling time in the sedimentation basin.
when using ferric sulfate and employing High-energy laboratory batch studies The rapid-mix basins, because of short-
G values as high as 12 500 s"1 , essen were conducted prior to the installation circuiting, provide a detention time of 3
tially all coagulation and flocculation was and the operation of the pilot unit. Using min at conventional-plant mixing speeds
complete within a 2-min rapid-mix 20-1 batch samples of Ohio River water to 30 s for high-energy mixing, depend
period at G values of 500 and 1 000 s~>. and velocity gradient (G) values of a ing upon the velocity gradient and mix
A higher G value in the initial rapid mix magnitude greater than 400 s~', re ing speed chosen.
could delay floe formation during the searchers discovered that optimum The pilot filter used during the
slow-mix period. coagulation occurred in less than 2 min Missouri River study at the central plant
when cationic polyelectrolytes were of the St. Louis County Water Co. was
Velocity Gradient Determination used. Floe shearing occurred if G values constructed of two 6-in-ID circular 5-ft
Determination of the velocity gra equal to 400-450 SH were continued for sections of acrylic plastic, joined togeth
dients used in the pilot work was ob any length of time over 2 min with a er by a rubber seal and bolted steel
tained from the following formula:7 resultant small floe that did not reform plates. The sections are mounted on a
during flocculation. At velocity gradients support stand, which contains three out
of 300-1 400 s"1 , the time necessary to lets used individually to obtain the fin
achieve optimum destabilization de- ished water and backwash water and to
creased so that at a G value of 1 000 s~l monitor the filter head loss.
in which only several seconds of rapid mixing was The media used for this downflow
G = velocity gradient in feet per second necessary. filter duplicated the central plant's full-
per foot (s"1) Floe shearing also was found to occur scale filters and consisted of 22 in. an
H = viscosity (0.273 x lO^4 Ib/s/sq ft at when rapid-mix speeds were suddenly thracite coal, 6 in. fine sand, 2 in. grade 3
50F) lowered to flocculation speeds where G sand, 2 in. grade 2 sand, and 2 in. gravel
V = volume of basin in cubic feet values of 5-100 s~' could be expected. A supported by a wire-mesh screen with a
P = water horsepower in foot pounds small floe formed, which neither grew hollow cavity below. During this evalua
per second nor settled during the flocculation tion the flow rate to the filter was main
period. However, floe shear did not oc tained at 2.5 gpm/sq ft. During the
A velocity gradient represents the cur if tapered flocculation, obtained by Mississippi River evaluation a 5-in.-ID
mean square ""space rate of change of use of sequential rapid mixing, was used pilot filter was employed having the
velocity in a direction normal to the to lower the mixing speed gradually. The same design features; however, a rated
direction of the velocity."8 In a fixed design of two-compartment rapid-mix flow of 3.7 gpm/sq ft. was evaluated.
basin with constant volume, the velocity facilities or a series of variable-speed Refer to Fig 2.
gradient is a direct function of the effi motors turning at different speeds could
Pilot-Plant Results
ciency of the mixing device. For the provide a maximum input of energy,
pilot clarification unit, horsepower of the thereby enhancing the floe growth while Inorganic coagulants. New Brighton study.
mechanical mixers was determined by preventing the floe from shearing. The Beaver River is a low turbidity

NOVEMBER 1974 J. J. MORROW AND E. G. RAUSCH 647

 Finished Water

Fig. 1. Pilot-Clarification-Unit Flow Scheme

High-energy mix I and II; polymer feed I, II, or III

range of 40-2 500 Jtu. The tremendous

fluctuations in flow and suspended
solids of the lower Missouri River make
it one of the most difficult midwestern
2.5
rivers to treat because it requires cons
3
I tant monitoring of plant and river condi
£
i° tions.
I 2.0 An evaluation of several inorganic 0 From Pressure
Line
treatments was conducted to duplicate
the results obtained by the lime soften
ing, ferric-sulfate-clarification process Fig. 2. Pilot Filter
used by the central plant. Results from 1 —filter effluent line; 2—support stand;
250 360 450 550 650 750
the pilot unit with ferric sulfate as the 3 wire-mesh screen; 4~grave/: 2 in.;
coagulant indicated taat turbidity was 5—grade No. 2 sand: 2 in.; 6—grade No.
Velocity Gradient—s '
only marginally reduced and did not ap 3 sand: 2 in.; 7—fine sand: 6 in.; 8—
Fig. 3. Results of Application of Velocity proach the results obtained by the plant. anthracite: 22 in.; 9 —steel plating
Gradients Greater Than 400 sec,, Prior to filtration at the central plant, the (bolted); 10—backwash drain; 11—over
(O) = 2.0 mg/l; (A) = 1.0 mg/l clarified water had a turbidity range of flow drain; 12 —two circular columns;
1.0-5.0 Jtu with a median value of 2.4 13—head-loss gage; 14—steel plating
(bolted); 15—backwash valve
Jtu. Ferric sulfate at 25.0 mg/l produced
(5-15 Jtu), low-colored industrial river. a settled water turbidity with a median
The raw river water is prechlorinated, value of 30.0 Jtu.
coagulated with alum, and adjusted for Alum, at a dosage of 45 mg/l, was the
pH with lime by the water plant. The ini second treatment evalaated. Turbidity duced a finished water containing 0.3 Jtu
tial treatment investigated used alum to was reduced to an average of 12.6 Jtu of turbidity.
duplicate the results and conditions but was still not comparable to the plant East St. Louis interurban study. The
found in the plant. While maintaining effluent. A combined alum-lime treat raw-water source for this plant is the
the New Brighton plant's alum treat ment of 30 mg/l alum and 25 mg/l lime, Mississippi River, which has a normal
ment (24.0 mg/l) and mixing condition s, however, produced a turbidity of 2.6 Jtu, turbidity range of 75-600 Jtu. Alum was
an effluent was obtained from the equal to the values obtained by the plant. the coagulant evaluated under mixing
clarification unit that provided equiva The installation of the pilot filter at this conditions of both a high and low rapid-
lent results to the settled water quality time added to the capability of compar mix velocity gradient. At a dosage of
obtained by the plant. A turbidity range ing filtrability of the pilot unit effluent 45.0 ppm and a velocity gradient of 600
of 0.5-3.1 Jtu was obtained from the set with the various treatments. s"1 , the settled water from the pilot unit
tled water in the pilot unit, which com Overall, the pilot clarification unit averaged 5.0 Jtu and the finished water
pared favorably to the plant's range of achieved results comparable to the showed little improvement at an average
1.5-3.8 Jtu. Direct comparisons between plant's settled water, but the finished- of 3.5 Jtu. The highly turbulent mixing
the plant's finished water and the pilot water turbidity averaged higher than conditions were readily shearing the
unit effluent could not be made because that obtained by the plant because a alum floe, leaving behind colloidal mate
a pilot filter was not in service during the filter aid was not in use. The central rial that passed through the filter easily.
New Brighton study. plant produces a finished water contain A velocity gradient of 300 s~' and the
St. Louis County study. The central ing an average of 0.1 Jtu of turbidity same alum dosage produced settled-
plant treats water from the Missouri using a nonionic polyacrylamide as an water averages of 1.5 Jtu and finished-
River, which has an annual turbidity aid, whereas the pilot unit and filter pro- water average values of 0.3 Jtu.
JOURNAL AWWA
648 WATER TECHNOLOGY/QUALITY
Cationic Polyelectrolytes and TABLE 1
Inorganic Salts Beaver R iver Study
New Brighton study. Alum and a ca- Polyelectrolyte and Inorganic Coagulant Results
tionic quaternary ammonium polymer,*
Average Settled Water Turbidity
used as a coagulant aid with conven Velocity Coagulant Raw Water
tional-plant mixing conditions, were Coagulant Gradient
*-) Dosage Turbidity Pilot Unit Plant
Treatment mg/1 Jtu Jtu Jtu
evaluated. The plant uses one-step rapid
mix with a velocity gradient of approxi Alum 250 24.0 6-14 1.5 2.5
Alum/polymer* A 250 12.0/0.2 4-8 1.9 1.6
mately 250 s"1 . The alum treatment used Polymer A 425 1.0 5-7 1.85 1.8
in the plant was reduced by 50 per cent Polymer A 425 2.0 4-6 1.4 1.25
to 12 mg/1 with the addition of 0.2 mg/1 Polymer A 600 1.0 5-6 1.3 1.25
Polymer A 600 2.0 4-7 1.4 1.65
of the polymer. Both coagulants were Polymer A 730 1.0 4-6 1.25 1.45
fed simultaneously to the third rapid- Polymer A 730 2.0 4-6
4-6
2.2 1.45
Polymer A 250 2.0 3.0 2.3
mix basin of the pilot unit. The effluent
quality from the pilot unit was equal to *Cat-Floc B, polyquaternarz ammonium compound
the quality of the plant's settled water.
Turbidity values for the settled water in
the plant range from 1.2 to 2.5 Jtu. In the TABLE 2
pilot unit a range of 1.4-2.2 Jtu was ob Missouri River Study
tained. Higher velocity gradients were Pilot Unit Inorganic Applications
not evaluated because of the successful Average Finished
coagulation and treatment of the raw- Water Turbidity
water turbidity using conventional treat Velocity Coafiuliint Raw Water Pilot-Unit
Plant
Coagulation Gradient Dosage Turbidity Average Pilot Unit
ment methods. Treatment
j-l
mg/1 Jtu Settled Turbidity Jtu Jlu
St. Louis County study. Cationic poly-
300 80-380 31.0 0.10
electrolyte applications, first with Ferric sulfate
Alum 300
25.0
45.0 60-450 12.6 0.10
alum and then with lime, were investi Alum lime 300 30/17 45-110 10.3 1.30 0.10
30/25 25-160 2.6 0.35 0.16
gated. The initial concentrations of alum Alum/lime
A!um/polymer*B
300
300 25/1.0 90-100 8.0 0.45 0.10
with polyDMDAACt formed a large floe Alum/polymer B 300 25/0.5 130-600 12.5 1.10 0.17
that settled rapidly but allowed an ex Lime/polymer B 550 77/2.0 800-2 400 12.0 0.57 0.05
Lime/polymer B 550 68/2.0 200-800 10.4 0.50 0.10
cessive amount of dispersed material to
carry over into the pilot filter causing *Cat-Floc, Poly DMDAAC
severe plugging and backwash problems.
Reducing the dosage resulted in a more TABLE 3
uniform floe but poorer clarification that Missouri River Study Using Cationics Fluctuations
still caused poor filtrability and severe
head loss on the filter. This treatment Average Finished Water Turbidity
failed to work because of the high solids Velocity Coagulant Raw Water Pilot-Unit
Average Plant
Coagulant Gradient Dosage Turbidity Pilot Unit
in the raw water and a poorly formed Treatment
s~' mg/1 Jlu Settled Turbidity Jtu Jlu
floe that would not settle.
PolyDMDAAC 550 4.0 210-550 1.70 0.66 0.07
The second treatment evaluated with PolyDMDAAC 550 2.0 120-140 1.20 0.50 0.10
the pilot unit utilized high-energy mix PolyDMDAAC 1000 2.0 120-180 3.50 0.60 0.10
ing of polyDMDAAC followed by lime PolyDMDAAC 1450 2.0 550-1 200 4.0 1.10 0.10
PolyDMDAAC 300 3.0 200-1 000 20.0 14.30 0.20
coagulation for the removal of hardness Modified poly 700 2.0 160-700 17.40 0.30 0.10
and alkalinity. With average raw-water DMDAAC
0.38 0.10
Modified poly 700 3.0 130-560 2.90
turbidities ranging from 200-2400 Jtu, DMDAAC
the pilot unit produced settled water that Condensation 700 2.0 110-180 2.96 0.37 0.13
ranged from 3.1 to 27.0 Jtu. The finished polyamine
Epichlorohydrin 700 2.0 200-700 4.60 0.44 0.10
water from the pilot filter ranged from quat.
0.14 to 1.7 Jtu. Alkalinity was greatly re
duced by the lime addition, and hard
ness values were reduced 30-40 per TABLE 4
cent. Color was unaffected by the poly Mississippi River Study
mer-lime treatment, but some removals
were recorded through the filter. Average Finished Water Turbidity

Cationic Polyelectrolytes Coagulant Gradient

s~' Dosage Turbidity Average Pilot Unit
Jlu
Plant
Jiu
Treatment mgtl Jtu Settled Turbidity
New Brighton study. The use of
450 2.0 100-300 1.71 0.50 0.47
modified polyDMDAAC in the absence Poly me
Polyme
A*
A 450 1.0 135-270 0.43 0.14 0.26
of inorganic salts for effecting particle Polyme A 450 0.5 115-180 5.40 2.10 0.22
destabilization was evaluated. The poly- Polyme A 600 0.5 120-200 5.20 1.60 0.29
Polyme A 600 2.0 150-300 0.72 0.48 0.67
electrolyte was tested at dosages of 1.0
and 2.0 mg/1 at various mixing inten 'Modified polyDMDAAC, Cat-Floe B, a product of Calgon Corp.
sities. Each dosage was evaluated at G
values of 730,600,425, and 250 s"1 .
With the application of velocity gra results, as shown ,in Fig. 3, indicate that
dients greater than 400 s"1 , the poly- cationic polyelectrolytes work most ef
Cat-Floe B, modified polydiallyldimethylammonium
chloride, a product of Calgon Corp. electrolyte produced an effluent that fectively when high G values are used.
TCat-Floc, polydiallyldimethylammonium chloride, a equalled the quality of the water ob The most successful treatment was with
product of Calgon Corp. tained by treatment with alum. The 1.0 mg/1 modified polyDMDAAC and a
NOVEMBER 1974 J. J. MORROW AND E. G. RAUSCH 649
 TABLE 5 mixing produced an effluent that was
Mississippi River Study superior in quality to the settled water
obtained by the plant. A treatment of 2.0
Average Finished Water Turbidity mg/l poiyDMDAAC at a G of 550 s"1
Velocity Coagulant Raw-Water Pilot-Unit produced a settled water of 0.8-1.6 Jtu.
Coagulant Gradient Dosage Turbidity Average Pilot Unit Plant
Treatment s~l mg/l Jtu SettledTurbidity Jtu Jtu Without the benefit of a filter aid, the
Poly- 150-235 0.34 0.16 0.26
pilot-unit finished water averaged 0.5 Jtu
600 2.0
DMDAAC in comparison to the plant's finished
Poly. 300 2.0 75-110 2.60 0.72 0.27 water of 0.1 Jtu. The water quality was
DMDAAC
Poly- 600 1.0 50-100 1.54 0.42 0.36 enhanced by the addition of 10-20 ppb
DMDAAC of a nonionic polymer filter aid that
Alum 300 45.0 80-200 1.5 0.3 0.28
Alum 600 45.0 100-260 5.0 3.5 0.25
reduced the finished water turbidity to
levels of 0.15 Jtu. Tables 2 and 3 indicate
the effectiveness of each treatment both
with and without the use of high-energy
TABLE 6 mixing.
Buchner Funnel Test Results: The results of high-energy mixing at
Pilot Clarification Unit Sludge levels approaching 1 000 s^show
noticeable difficulties in maintaining a
Sludge Sample Polymer Op'.imum Net Filtrate large and settleable floe as the higher
Solids Size Conditioning Ebsage After 2 min
Sludge Type mgll ml Agent mg/l ml velocity gradients are attained. The diffi
164
culty in obtaining good filtration data for
Inorganic 14000 200 Cationic 75
Inorganic 14000 200 Non ionic 50 63 the treatment using a G value of 1 450
Inorganic 14000 200 Anionic 25 87 s""1 indicated that a greater share of the
Inorganic 14000 200 0 38
Inorganic/polymer 18000 200 Cationic 50 176 colloidal material was not being captured
Inorganic/polymer 18000 200 Nonionic 5 86 even though the settled-water results
Inorganic/polymer 18000 200 Anionic 50 125 appeared acceptable. It is possible that
Inorganic/polymer 18000 200 0 82
Polymer 20000 200 Cationic 25 185 shearing of the polymer and floe had oc
Polymer 20000 200 Nonionic 5 73 curred with resultant nonreformation of
Polymer 20000 200 Anionic 5 115
Polymer 20000 . 200 0 109 the sheared particles.
From the results of the Missouri River
study, it was discovered that cationic
polyelectrolytes function most effective
TABLE 7 ly as primary coagulants within a G
Comparison of Pilot Clarification Unit Sludges range of 500-1 000 s"1 . Comparing this
to the previous study indicates that for
Sludge
Sludge Solids Bulk Density Sludge
both high- and low-turbidity waters, the
Treatment mg/l g/cm3 Iblmil gal H^O treated effective G range and polyelectrolyte
1. Inorganic 14000 2.0 77
dosage requirements are nearly the
2. Inorganic't 61000 3.4 same. Because the polyelectrolyte
3. Inorganic/polymer 18000 2.8 60 dosage was not optimized for each
4. PoiyDMDAAC' 92000 3.5
5. Modified poiyDMDAAC 20000 4.4 48 change in river conditions, the accepta
(Beaver River) ble results obtained over a wide turbidity
6. Modified poiyDMDAAC 22000 4.2 range indicate that with high G values
(Mississippi River)
polyelectrolytes have a much wider
'Sludges were gravity thickened for 48 hr. range of effectiveness at a single dosage.
tBlend of alum, lime, and ferric sulfate sludge.
East St. Louis interurban study. The use
of modified poiyDMDAAC as the pri
G of 730 s""1 , which gave an average tur less polymer is necessary to effect mary coagulant at a velocity gradient of
bidity of 1.25 Jtu compared with 1.45 Jtu efficient turbidity removal since more 450 s~l was evaluated first. The polymer
obtained in the plant. effective distribution and utilization of was fed ahead of the high velocity rapid-
By using 2.0 mg/l of the polymer, the polyelectrolyte occurs. By using high mix zone instead of directly into the
researchers obtained efficient turbidity velocity gradients, researchers found rapid-mix tank as previously done in the
removals at G values between 400-600 that synthetic cationic polyelectrolytes first two plants. This treatment showed
s~', but at a higher G value, overcoagula- can successfully function as primary the benefits of premixing the polyelec
tion and charge repulsion appeared to coagulants and coagulate low-turbidity trolyte before subjecting it to a high-
take place. Treatment 8 in Table 1 shows surface waters. No effect on clarification energy mix.
this effect. efficiency due to hardness, alkalinity, As shown in Table 4, the most effec
It was discovered also that the poly pH, or temperature during any of the tive treatment was with 1.0 mg/l poly
mer alone cannot completely destabilize polyelectrolyte treatments was observed. mer and a G value of 450 s"1 . An average
colloidal particulate matter in surface None of these parameters was affected settled-water turbidity of 0.43 Jtu was
waters when applied with conventional by the addition of cationic polymer. obtained with finished-water turbidity
rapid-mix G values. Plant trials with St. Louis County study. The use of averaging 0.14 Jtu from the pilot filter.
several polymer products confirm the several cationic polyelectrolytes for Values as low as 0.04 Jtu were achieved.
difficulty found in completely replacing effecting complete colloid destabilization These values correspond to the plant's
inorganic coagulants with polyelec- at high G values wai; investigated. Those average of 0.26 Jtu on finished water.
trolytes when conventional rapid mixing polyelectrolytes were discovered to be No filter aid applications were neces
is used. All comparative data for the capable of completely replacing inor sary to treat the Mississippi raw water. A
treatments evaluated are shown in Table ganic coagulant for high turbity waters. comparison between treatments C and
1. ' D in Table 4, in which 0.5 mg/l polymer
Application of a synthetic cationic
Figure 3 shows that by increasing G, polyelectrolytes with high-energy rapid and G values of 450 and 600 s~l were
650 WATER TECHNOLOGY/QUALITY JOURNAL AWWA
used, demonstrates that additional mix TABLE 8
ing alone cannot compensate for an in Estimated Costs
adequate polymer dosage. Average fin High Energy Rapid-Mixing
ished water for the two treatments only Water Clarification Plants
improved from 2.1 to 1.6 Jtu by increas
ing G. Plant Size
PolyDMDAAC was evaluated under Costs
\mgil 10 mgd 100 mgd
the following conditions established in
the pilot unit. These conditions are the Basin requirement (2-min retention) 186 cu ft 1 860 cu ft 18600 cu ft
Basin dimensions ( 2-ft freeboard) 5 x 5 x 10. ft 13 x 13 x 13 // 24 x 48 x 18.5 /(
same as those used in the initial treat Capital Expense
ment. Estimated cost of concrete basin $ 4 000.00 14 100.00 55000.00
Estimated cost of mixing equipment
Treatment 2.0 mg/1 for a velocity gradient*t of
400s"1 -$ 2 000.00 3 400.00
poly- 700 s~'
27 950.00
2 650.00 5 200.00 44 850.00
DMDAAC 1 000 s-1 2 750.00 8 700.00 84 300.00
Flow rate 0.5 gal/min Total cost of basin and equipment
Rapid mix G value 600 s-1 for a velocity gradient of
400 s"1 - $ 6 000.00 1 7 500.00 82 950.00
Tapered mix G value 300 s-1 700 s"1 6650.00 19300.00 99 850.00
45s-' 1000s"1
Slow mix G value Operating Expense
6 750.00 22 800.00 139300.00
Flocculation speed 0.3 ft/s Water horsepower required at 41F
Flocculation for a velocity gradient of
400s"1 -hp 2.02 20.2 202.0
detention time 30 min 700 s-> 5.42 54.2 542.0
Sedimentation time 4hr 1 000 s" 1 10.80 108.0 1 080.0
Power cost at 11/kWhr for
Filtration rate 3.7 gal/sq a velocity gradient* of
ft/mi n 400s~'-$ 0.36/mil gal 3.60/day 36.00/day
700 s"1 0.98/mil gal 9.80/day 98.00/day
The polymer was again fed in advance 1 000 s~' 1.94/milgal 19.40/day 194.00/day
of the rapid-mix chamber. A high floc-
culation speed was used to maintain 'Courtesy Philadelphia Mixer Co., King of prussia. Pa.
gear «...-.
tSingle-speed g^». drives, walkways, and railings not included
suspension of the floe since it had a tWater horsepower only
tendency to settle out in the flocculation
channels.
The best treatment obtained was with
2.0 mg/1 of the polyelectrolyte and a G TABLE 9
value of 600 s"1 . Settled water from the Cost Comparison of
unit produced an average turbidity of Coagulants for Clarification
0.34 Jtu, and finished-water values were
measured as low as 0.03 Jtu. An addi Dosage Cost
Plant and Characteristic Turbidity Treatment mg/1 Stmgd
tional treatment using a conventional
rapid-mix G value of 300 s"1 with a 10- Plant A low turbidity
river water alum 24.0 5.04
min mix time produced results with alum with
polymer alone that did not match plant polymer 12.0/0.2 3.24
performance; however, acceptable re Plant B high turbidity
polymer* 1.0 2.52
sults of 0.72 Jtu on the finished water river water lime and 94.0 8.97
were obtained. ferric sulfate 5.1
polymer with 1.0 8.19
Settled-water quality had an average of alum 25.0
2.6 Jtu, which is a 98 per cent removal of polymer* ' 2.0 5.88
Plant C high turbidity
raw turbidity using the polyelectrolyte river water alum 45.0 9.45
alone under plant conditions. As seen in polymer with 2.1 6.13
Table 5, using 1.0 mg/1 of the poly ferrous sulfate
polymer*
17.1
1.0 2.10
electrolyte at a G value of 600 s"1 , pilot
results were equivalent in quality to the Treatments employing high-energy rapid mixing
plant's treatment of 1.2 mg/1 polymer
and 17.0 mg/1 ferrous sulfate.
Color Removal have while acting as an oxidant. The use trolyte addition was equal to floe formed
of pre- or post-chlorination of both with inorganic flocculants. Coagulation,
At no time could color be adequately could significantly alter the results ob flocculation, and sedimentation of the
removed with any of the treatments. tained since none of the waters tested much more variable Missouri River raw
Polyelectrolytes on their own do not was treated with chlorine. The added water produced settled water of much
capture natural color bodies satisfac benefits of prechlorination could pro higher turbidity than that obtained from
torily. Removal of the large colloidal vide for the removal of the oxidizable Mississippi River water and, conse
fractions was very good using 1 mg/1 of organic color bodies, which could signifi quently, the best finished-water results
the polymer with a G range of 400-600 cantly improve the finished water color. obtained by filtering the treated Mis
s~' being the most effective. Because cat- However, complete color removal by souri River water were values near 0.3
ionic polyelectrolytes do not affect pH, high-energy mixing and polyelectrolyte Jtu. With the use of a nonionic synthetic
color must be removed at the pH of addition alone appears unobtainable at polymer filter aid,* however, finished-
natural waters instead of the preferred this time. water turbidities were reduced to levels
acidic range of 4.0-6.09, which is ob .near 0.15 Jtu, or those obtained by the
tained by the use of iron and aluminum Filtration Data lime-softening treatment employed by
inorganic salts. The operation of the pilot filters at the St. Louis County Water Co.
One integral parameter not investi both the 2.5 and 3.7 gpm/sq ft rates pro
gated during the operation of the pilot duced comparable results. Filtrability of
unit was the effect that chlorine would the floe formed by cationic polyelec *C.A. 233, a product of Calgon Corp.

NOVEMBER 1974 J. J. MORROW AND E. G. RAUSCH 651

 The filtration results of Mississippi samples evaluated showed polymer dosage necessary to effect colloid desta-
River water indicated that some color sludges to be twice as dense as alum and bilization.
was removable in the filter, but settled- ferric sludges. A significant reduction in Floe shearing occurs if high-velocity
water turbidities and finished-water volume of sludge produced is realized gradients are continued beyond the
values were much better than the results when polyelectrolytes are used as the period of optimum colloid destabiliza-
obtained from Missouri River water. No primary coagulants in water-treatment tion and if high (/values are reduced too
filter aid was necessary during this study plants. rapidly.
since water quality of less than 0.1 Jtu A 40 per cent reduction in pounds of The floe produced by polyelectrolytes
was obtained. The use of polyelectro- sludge solids produced per million gal with high-energy mixing is filterable,
lytes as primary coagulants indicated lons of treated raw rater was obtained does not contribute to shortening of fil
that they would produce a readily fil during the Beaver River study. Because ter runs, and is often improved by the
terable water that did not shorten filter of the reductions in sludge volume, the addition of a nonionic polymeric filter
runs since several treatments produced use of cationic polyelectrolytes as pri aid.
filter runs over 100 hr. mary coagulants significantly eases the A significant reduction in volume of
problem of sludge handling and disposal. sludge produced is realized when poly
Sludge Handling electrolytes are used, which contributes
As a result of tests performed on Plant-Scale Costs to increased ease of sludge handling and
sludge samples collected from the pilot To provide a basis for comparison and disposal.
clarification unit during the three stud design of a suitable facility employing Sludge produced by polyelectrolyte
ies, several observations were noted. Of high-velocity gradients for polymer co treatment was heavier and denser than
the synthetic polymers evaluated, ca- agulation, a table of estimated costs list any inorganic sludge evaluated and had
tionic polymers for dewatering sludges ing both capital and operating expenses better sludge compaction and settling
proved to have the most potential. Using appears in Table 8. Prices are based on characteristics.
the Buchner funnel test to determine using standard motor-driven, single- None of the polymers evaluated was
dewatering rates and filtrability, it was speed gear drives with impellers de able to remove the natural color satisfac
discovered that anionic and nonionic signed to draw a minimum water horse torily.
polymers produced excellent coagula power. All in-tank parts are assumed to Hardness, alkalinity, pH, and temper
tion of the sludge solids, but blinding of be carbon steel. ature had no effect on the performance
the filter pad caused low dewatering Water-horsepower figures indicated of the polyelectrolytes nor were these
rates. that the 100 mgd plan 1: would require use parameters affected by polymer addition.
As demonstrated both with and with of multiple units. For purposes of this When high G values are used, cationic
out polymer additions, marked improve estimate, two mixers were used for polyelectrolytes are effective at a single
ment in dewatering rates was shown velocity gradients of 400 and 700 s~' and dosage over a much wider range of raw-
with polyelectrolyte sludges over in three mixers for 1 OOC s~'. Use of multi water turbidities.
organic sludges. The resultant dosage re ple mixers would require a redesign of
quirements for dewatering polyelectro- tank shape from square to rectangular, Acknowledgment
lytes sludges were found to be more which could also accommodate a tapered The authors of this article would like
than 50 per cent lower than those re mix. Tank dimensions were arbitrarily to express their thanks and appreciation
quired for dewatering an alum sludge. A selected at 24 ft wide by 48 ft long by 18 to the following people, and the person
polyelectrolyte sludge produced a much 1/2 ft deep. Prices quoted for the mixing nel in their respective plants, for their
drier cake and a better dewatering rate units would not include shipping costs assistance in collecting the data pre
than the inorganic alum sludge. (Table from the manufacturer. sented. Their support and cooperation
6). Table 9 compares the cost of treating provided an invaluable contribution to
Comparative densities of each sludge low-turbidity river water and high-tur the research presented in this article.
sample evaluated were obtained by bidity river water. Mr. Frank A. Richter, supt, S. S.
weighing a 1-1 volumetric flask, filling it In reviewing Table 8, one must Sahli water plant, New Brighton,
with the sludge sample, and then taking remember that from the formula sup Pa.
a weight measurement and adjusting for plied for velocity gradient determina Mr. J. L. Teupker, mgr. of pro
the weight of the flask. By determining tions, a twofold increase in G requires a duction, St. Louis County Water
the percentage of solids in the sludge fourfold increase in the water horse Co., St. Louis, Mo.
sample and then obtaining the dif power requirement. An alternative to Mr. John Hamilton, dist. engr.,
ference between the dry and liquid building expensive mixing basins would and Mr. Clarence Blanck, asst.
weight of the sludge, the weight of the be the use of in-line blenders that can dist. engr., American Water
liquid-sludge fraction is obtained. Com effectively produce a high velocity gra Works Service Co., East St.
pared by weight with a litre of water, the dient for a few second?. A series of mix Louis Interurban Water Co.,
disparity in the results represents the ers with progressively lower intensities East St. Louis, 111.
volume in which the solid-sludge frac could find application where space is un
tion is concentrated. Total weight of the available or equipment is already in use. References
dry solids over the displaced volume 1. COHEN, J. M.; ROURKE, G. A.; & WOOD
produces the bulk density of the sludge. Conclusions WARD, R. L. Natural and Synthetic
Comparative values of the sludges an Cationic polyelectrolytes are able to Polyelectrolytes as Coagulant Aids. Jour.
alyzed are shown in Table 7. replace inorganic coagulants completely AWWA, 50:4:463 (Apr. 1958).
It was discovered that a polyelec as primary coagulants for both low- and 2. CROOK, E. H. & POLLIO, F. X. Removal of
trolyte sludge yields a higher percentage high-turbidity surface waters when G Soluble and Insoluble Organic and In
of solids, higher bulk density, and better values of greater than 400 s"1 are ap organic Material by Flocculation. Proc.
compactability than sludge produced by plied. 26th Annual Meeting of the Intl. Wtr.
inorganic coagulants. In both the Beaver Optimum coagulation with cationic 3. HEMENWAY,Conf., Pittsburgh, Pa. (Oct. 1962).
D. R. & KESHAVAN, K.
and Missouri River studies the poly- polyelectrolytes using high-energy mix Determination of Optimum Velocity
DMDAAC sludge was concentrated 25 ing occurs in less than 2 min and an in Gradients for Waters Coagulated with
per cent more than the inorganic crease in the velocity gradient decreases Polyelectrolytes. Wtr. Sew. Works,
sludges. Comparative densities of the the retention time ami polyelectrolyte 115:12:554 (Dec. 1968).
652 WATER TECHNOLOGY/QUALITY JOURNAL AWWA
4. KLEBER, J. P. Field Use of a Cationic Treatment (Harold B. Crawford and Dilute Colloidal Suspensions by Poly
Polymer for Clarification. Wtr. Wastes Daniel N. Fischel, editors), McGraw-Hill electrolytes. PhD thesis, Univ. of Illinois
Engrg., 6:6:42 (Jun. 1969). Book Co., New York (3rd ed., 1971). (1968).
5. TEKIPPE, R. J. & HAM, R. K. Velocity COHEN, J. M.; ROURKE, G. A.; & WOODWARD, KLEBER, J. P. Field Use of A Cationic Polymer
Gradient Paths in Coagulation. Jour. R. L. Natural and Synthetic polyelectro- for Clarification. Wtr. Wastes Engrg.,
AWWA, 63:7:439 (Jul. 1971). lytes as Coagulant Aids. Jour. AWWA, 6:6:42 (Jun. 1969).
6. CAMP, T. R. Floe Volume Concentration. 50:4:463 (Apr. 1958). MUIA, R. A. Clarification with Cationic
Jour. AWWA, 60:6:656 (Jun. 1968). CROOK, E. H. & POLLIO, F. X. Removal of Sol Polymers. 22nd Okla. Waste and Pollu
7. Water Treatment Plant Design. ASCE and uble and Insoluble Organic and In tion Cont. Conf. and Advanced Wtr.
AWWA, New York (1969). organic Materials by Flocculation. Proc. Conf., Okla. State Univ. (Mar. 1971).
8. CAMP, T. R. & STEIN, P. C. Velocity Gra 26th Annual Meeting of Intl. Wtr. Conf., PRESSMAN, M. Cationic Polyelectrolytes as
dients and Internal Work in Fluid Mo Pittsburgh, Pa. (Oct. 1962). Prime Coagulants in Natural Water. Jour.
tion. Jour. Boston Soc. Engrs., 30:219:4 FAIR, G. M. & GEYER, J. C. Water Supply and AWWA, 59:12:169 (Feb. 1967).
(Oct. 1943). Waste Water Disposal. John Wiley & PUGH, T. & HELLER W. Coagulation and
9. COHEN, J. M. & HANNAH, S. A. Coagula Sons, New York (1954). Stabilization of Colloid Solutions with
tion and Flocculation, Water Quality and FRISCH, H. L. Polymer Chain Configeration Polyelectrolytes. Jour. Poly. Sci., 47:219
Treatment (Harold B. Crawford & Daniel Near a Boundary Exerting Forces. Jour. (Nov. 1960).
N. Fischel, editors), McGraw-Hill Book Phys. Chem., 59:7:633 (Jul. 1955). SAMPLES, W. R. Removal of ABS from Waste
Co., New York (3rd ed., 1971). Fuoss, R. M. & GATHERS, G. I. Polyelectro- Water Effluent. Jour. WPCF, 34:1070
lytes. I Picrates of 4-Vinyl-Pyridine- (Oct. 1962).
Bibliography Styrene Copolymers. Jour. Poly. Sci., 2:12 SIMHA, R.; FRISCH, H. L.; & EIRICH, F. R. The
BIRKNER, F. B. & MORGAN, J. J. Polymer Floc (Feb. 1947). Adsorption of Flexible Macromolecules.
culation Kinetics of Dilute Colloidal HAWKES, B. E. Polyelectrolytes in Water II. Jour. Phys. Chem., 58:507 (Jun. 1954).
Suspensions. Jour. AWWA, 60:2:175 Treatment. Jour. N. E. Water Works TEKIPPE, R. J. & HAM, R. K. Velocity Gra
(Feb. 1968). Assn., 84:189 (Jun. 1970). dient Paths in Coagulation. Jour. A WWA,
BLACK, A. P.; BIRKNER, F. B.; & MORGAN, HEMENWAY, D. R. & KESHAVAN, K. Deter 63:7:439 (Jul. 1971).
J. J. Destabilization of Dilute Clay Sus mination of Optimum Velocity Gra TENNEY, M. W. & STUMM, W. Chemical Floc
pensions with Labeled Polymers. Jour. dients for Waters Coagulated with culation of Microorganisms in Biological
AWWA, 57:2:1547 (Dec. 1965). Polyelectrolytes. Wtr. Sew. Works, Waste Treatment. Jour. WPCF, 37:1370
CAMP, T. R. Flocculation and Flocculation 115:12:554 (Dec. 1968). (Oct. 1965).
Basins. Trans. ASCE, 120:1 (Sep. 1955). HUDSON, H. E. JR. <<K WOLFNER, J. P. Design of THOMAS, A. W. Colloid Chemistry. McGraw-
CAMP, T. R. Floe Volume Concentration. Jour. Mixing and Flocculation Basins. Jour. Hill Book Co., New York (1934).
AWWA, 60:6:656 (Jun. 1968). AWWA, 59:1257 (Dec. 1967). Water Treatment Plant Design. ASCE and
CAMP, T. R. & STEIN, P. C. Velocity Gradients JENCKEL, E. & RUMBACH, B. The Adsorption AWWA, New York, N. Y. (1969).
and Internal Work in Fluid Motion. Jour. of High Molecular Weight Materials WILDE, P. F. & DEXTER, R. W., Parameter Af
Boston Soc. Engrs., 30:219:4 (Oct. 1943). from Solution. Ger. Z. Electrochem, 55:612 fecting the Performance of Polyelectro
COHEN, J. M. & HANNAH, S. A. Coagulation (Oct. 1951). lytes as Aids to Water Clarification. Brit.
and Flocculation. In Water Quality and JORDEN, R. M. Destabilization-Aggregation of Polymer Jour., 4:239 (May 1972).

Water-Treatment Sludge
Filtration Studies
FredC. Hawkins, J. F. Judkins Jr., andJ. M. Morgan
Water-treatment processes remove impurities from ground water and sur in 1881. In 1885 Austin and Wilbur of
face water. The result of treatment, however, is a form of sludge that Rutgers Univ. investigated alum and
must be disposed of properly. Disposal of this sludge constitutes a problem several other coagulants and found alum
for plant operators. But if the volume of sludge can be reduced prior to to possess the greatest potential. 1 The
sludge produced from alum coagulation
treatment, then the problem of disposal can be accomplished more readily. of turbidity consists mainly of alumi
nium hydroxide and is bulky and
gelatinous in nature.2 As a result, it set
primarily aluminum sulfate (alum), are tles to a low-solids concentration and is
A paper presented at the Annual Con
ference on May 17, 1973, by Fred C. added to remove impurities. Ground so compressible that it resists passage of
Hawkins, instructor; J. F. Judkins Jr., water is usually in essense devoid of water from itself.2' 3
assoc. prof.; and J. M. Morgan, asst. clay, silt, and bacteria but may contain Disposal of alum sludges has com
prof., Dept. of Civ. Engrg., Univ. of larger quantities of hardness-causing monly involved direct discharge into re
Auburn, Auburn, Ala. (All are Active minerals. Hardness is removed primarily ceiving streams, but the adverse effect of
Members, AWWA). by lime-soda softening. In both of these this practice has led to strict pollution-
water-treatment processes, impurities control laws. As a result, more accepta
are removed in the form of sludge. The ble methods of disposal must be dis
There are two basic sources of water disposal of sludge, whether alum sludge covered. Disposal can be accomplished
supply surface water and ground or lime-softening sludge, is a very seri more readily if the volume of the sludge
water. Surface water contains clay, silt, ous problem in water treatment. can be reduced prior to treatment. Vac
and bacteria, which must be removed The earliest recorded use of alumi uum filtration,4' 5 sand-drying beds,5 cen-
prior to domestic use. In conventional num sulfate as a coagulant for a munici trifugation,5' 6 and alum recovery6 have
treatment of surface water, metal salts, pal water supply was in Bolton, England, been used for this purpose.

NOVEMBER 1974 F: C. HAWKINS ET AL 653