Flocculation Research Paper

8/11/2019 Flocculation Research Paper

1/5

Settling velocities of particulate systems 18: Solid ux density determination by

ultra-occulation

F. Concha a,, N.N. Rulyovb, J.S. Laskowski c

a Department of Metallurgical Engineering, University of Concepcin, Chileb Institute of Biocolloid Chemistry, National Academy of Sciences of Ukraine, Kiev, Ukrainec Department of Mining Engineering, University of British Columbia, Vancouver, Canada

a b s t r a c ta r t i c l e i n f o

Article history:Received 3 August 2011

Received in revised form 21 December 2011

Accepted 27 December 2011

Available online 5 January 2012

Keywords:

Thickening

Settling

Flocculation

Ultra-occulation

Flocculant dosage

The scale-up of thickening parameters from laboratory to industrial plant is still an open problem. Severaltechniques have been in use but none is free of criticism. In order to clarify these issue, in this work otation

tailings from one of the major Chilean copper mines was subjected to occulation-settling tests with Orioc-

2010 polycarylamide in a Couette type reactor. By varying the shear rate from 100 to 2000 [s1] the solid

concentration from 1 to 15 [% by volume] and the occulant dosage from 0 to 20 [g/ton] it was shown that an

important interaction exists between these variables. At the optimal occulant dosage, the optimal suspension

concentration and the optimal occulation time, an increase by 50% in the solid ux density function is possible

when the shear rate of _ 100 s1

is changed to the optimum value of around _400 s1

.

2012 Elsevier B.V. All rights reserved.

1. Introduction

In many countries water has become a scarce commodity. The

necessity of recovering and recycling water has turned thickening into

an important process, requiring the best operational practice. This

objective requires improvement in the control system that maximizes

thickener underow concentration, minimize occulant consumption

and, in general, permits the stabilization of the whole operation. Several

feedback control systems have been proposed for thickening opera-

tions, unfortunately, due to the non-linearity of the process and to the

slow response to perturbations, thickening is difcult to control and

classical control systems fail to operate adequately. Even more impor-

tantly, current methods of thickener control are based on feedback to

change operational variables that do not take into consideration the

properties of the feed material. Two such parameters are the solidux-densitythat expresses, for any concentration of a suspension, the

product of the settling velocity and of the solid concentration (and, for

a material of constant concentration, represents the momentum of the

motion), and the solid effective stress that represents the compressibility

of the sediment produced by settling. These parameters permit a

phenomenological description of the thickening process. Models

using these two parameters permit improved operation and optimal

control strategies. Determination of these parameters is still an openproblem (Brger et al., 1999). In this paper we will discuss the

determination of one of these parameters, the solid ux-density

function.

Several techniques have been proposed to determinethe settling ve-

locity in laboratory experiments, thejar tests being the most common

(Coe and Clevenger, 1916; Richardson and Zaki, 1954; Michael and

Bolgers, 1962). The Jar test involves homogenization of suspensions

varying solid concentrations in settling cylinders, introduction of the

occulant and mixing by moving a plunger up and down in the cylin-

ders, or by inverting the cylinders several times. This procedure is

claimed not to be satisfactory because of the local over-dosing that

can occur when the relatively concentrated occulant solution meets

the slurry (Kitchener, 1978); but more important is that the agitation

obtained by this method does not produce the optimum occulation.

Farrow and Swift (1996)show that jar test has three main inconve-

nient: (1) the occulation efciency, measured by the sedimentation

results, depend strongly on the mixing method (numbers of inver-

sions), (2) that the diameter of the cylinders have a signicant effect

on the settling rate, diminishing as the diameter of the cylinder

increases and (3) the experimental reproducibility associated with the

process is low, with standard errors of the mean from 8 to 12% for the

settling velocity.

Flocculation is used to aggregate ne particle to maximize the solid

ux-density, which is directly related to thickener capacity. Hogg et al.

(1993) showedthatthe choiceof a occulant is determined by chemical

factors such as the mineral composition and solution chemistry. But

International Journal of Mineral Processing 104-105 (2012) 5357

Corresponding author.

E-mail addresses:[email protected](F. Concha),[email protected](N.N. Rulyov),

[email protected](J.S. Laskowski).

0301-7516/$ see front matter 2012 Elsevier B.V. All rights reserved.

doi:10.1016/j.minpro.2011.12.007

Contents lists available at SciVerse ScienceDirect

International Journal of Mineral Processing

j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / i j m i n p r o


2/5

occulant adsorption and bridging occulation take place simulta-

neously and dynamic variables, such as the intensity of mixing and

the mixing time, affect particle-particle linkages (Keys and Hogg;,

1979; Hogg, 1999). Molecular weight of typical occulants fall in the

range of 1020 million Daltons. The transfer of the occulant from the

stock solution into thesuspension and distribution within the dispersed

systemtakes some time. An improved procedureto assessthe efciency

ofocculation is via the use of a shear vessel, which is similar to a rota-

tional Couette viscometers and has the advantage of quantifying the

mixing quality through the shear rate. Several investigators have used

the shear vessel in the past in coagulation experiments (Ives and

Bhole, 1977; Smith and Kitchener, 1978; Stein et al., 1986) and in oc-

culation studies (Muhle and Domasch, 1991; Farrow and Swift, 1996;

Rulyov, 1999, 2004; Rulyov et al., 2005a, 2005b; Rulyov et al., 2009).

Farrow and Swift (1996)designed their shear vessel with concen-

tric cylinders of 200 and 210 [mm] in diameter and 120 [mm] in

length (Fig. 1). At the bottom of the vessel a glass tube 14 mm indiameter and 220 [mm] in length was used to measure the settling

velocity. The experiments were carried out at a constant rotational

velocity of 200 [rpm]. The outow of the shear vessel was introduced

immediately in the settling column, depicted as A and B in the gure.

The authors conclude that the combination of shear vessel and set-

tling column overcame most of the problems associated with the jar

test, in particular the strong dependence of batch settling test on

mixing rate and cylinder diameter.

It was shown byRulyov (1999)and Rulyov et al. (2000) that the

mixing time in occulation can be reduced down from minutes to

56 s by the appropriate hydrodynamic treatment of the suspension

in a shear vessel at high shear rate _. This treatment, termed ultra-oc-

culation(Rulyov, 2004; Rulyov et al., 2005a, 2005b), ensures that, not

only occulant macro molecules are quickly and evenly distributed

within the suspension and adsorb onto the surface of the particles, but

also provides for the formation of large and dense ocs. Depending on

the size distribution and density of the solid particles in the dispersion,

as well as on their volume concentration, the optimum values of the

mean shear rate _ may vary in a wide range such as 300b _b5000

[s

1

]. The signi

cant advantage of ultra-

occulation is that it ensuresa good mix of small and large particles in ocs before they get into the

settler, thus providing for fast sedimentation and high degree superna-

tant clarication (Rulyov et al., 2009). It was recently shown that under

certain conditions intense agitation for short times may even change

the nature ofocculation, from total occulation to a selective occula-

tion of only some mineral constituents (Ding and Laskowski, 2007).

In this work an instrument called UltraocTester has been used,

which combines a shear vesselwith variable shear rateand an optoelec-

tronic device that measures the uctuation via intensity of the a light

beampassing normally through the transparent tube, while the formed

ocs pass through this tube, to (1) analyze the relationship ofoccula-

tion efciency (or mean ocs size) with solid concentration, occulant

dosage and shear rate, (2) the effect of the concentration and shear

rate on the settling velocity and solid ux density and (3) the effect of

the solid concentration on the optimum shear rate for occulation.

2. Material, experimental set-up and method

Flotation tailings from one of the major copper otation plants in

Chile were used in all the experiments. Solid concentration was varied

over the range from 1.8 to 15 [%] by volume] (4.7 and 32.3[%] by

weight); material density was 2700 [kg/m3]. An average particle size

of ,x50 20 m with size distribution characterized by x80 40 m

and x20 0:5 m , was determined using a Sympatec Helos-Rhodos

laser dispersion instrument. Orioc-2020 polyacrylamide was used as

a occulant with a molecular weight of 9.64106[g/mol].

The set-up to perform the ultra-occulation tests is shown inFigs. 2

and 3. It consists of a small shear vessel, referred to as ultra-occulator

inFig. 2. This Couette reactor, with a rotating cylinder of 28 [mm] indiameter and a gap of 1.5 [mm] was fed continuously with the suspen-

sion of tailings by a measuring peristaltic pump. Before entering the

Couette reactor the pulp receives continuouslya diluted occulant solu-

tion, at a ow-rate to give a pre-determined dosage. After 6 s treatment

at a pre-determined shear rate in the Couette reactor, the occulated

suspension is discharged from the ultra-occulator through a 3 [mm]

inner diameter transparent tube equipped with an opto-electronic

sensor which registers the uctuation of intensity of the light beam

passing normally through the tube [in accordance with techniques

proposed by Gregory and Nelson (1984)]. The electronic signal is

processed and displayed in a three digital format, thus showing, in

Fig. 1.Illustration ofFarrow and Swift shear vessel (1996) .

Fig. 2.Illustration of the UltraocTester layout.

54 F. Concha et al. / International Journal of Mineral Processing 104-105 (2012) 5357


3/5

relative units, the values ofocculation efciency (or mean ocs size)

and the mean shear rate _.

The operational conditions consisted in changing theocculant feed

rate and the shear rate while maintaining a constant treatment time of

the suspension in the Couette reactor (6 s). When the feed suspension

concentration exceeded the threshold of 6[%by volume]of its optical

analysis capacity, it was diluted by introducing clean water between

the shear reactor and the optoelectronic sensor (shown by a dash line

inFig. 1). In the tests designed to measure settling rate of the treated

suspension, dilution was not used. In this case the suspension from

the outlet of the tester was continuously fed to a 14 mm diameter

settling cylinder and, as soon as the suspension lled the cylinder, it

was one time inverted and the suspension allowed settling and the

initial settling velocity was measured.

3. Results and discussion

Table 1shows the operational conditions of the experiments and

the output of the instrument.(In Table 1, vopt and v100 are the initial settling velocity of the

suspension after treatment, at optimal shear rate _optand at shear rate

equal to _100 s1

, fbk(opt) and fbk100are the corresponding solid-

ux densities.)

3.1. Effect of the occulant dose on the efciency ofocculation

The occulation was carried out over 6 s at optimal values of the

mean shear rate _. (for the respective suspension concentrations see

Table 1).Fig. 4demonstrates that the occulation efciency increases

monotonically with occulant dosage, reaching the relative value of

90 with a dosage of 10 [g/ton] for the low range of particle concentra-

tion, and 20 [g/ton] for the higher range. The observed increase in theocculant dosage with the increase in the suspension concentration

can most likely be attributed to the slowdown of theprocess of theoc-

culant macro molecules distribution within the volume of the suspen-

sion with the increase in solid concentration.

3.2. Effect of the shear rate on the efciency ofocculation

Fig. 5shows the effect of shear rate _on theocculation efciency

(or mean ocs size) and clearly demonstrate that a maximum exist

between 400 and 600 [s1] withhigher values for increasing concentra-

tion. The shift of the maximum occulation efciency to higher shear

values for higher occulant dosages may be due to the increased

strength of the bridges bonding particles within a oc, as shown by

Rulyov et al. (2005a, 2005b). (The operational principle of the ultra-occulator is based on the optoelectronic measuring system proposed and

described by J. Gregory and D.W. Nelson (J. Gregory and D.W. Nelson

A new method for occulation monitoring in Solidliquid Separation

(J. Gregory / Ed)Ellis Horwood, Chichester, 1984, pp.172182). The reading

of the device is proportional to the average size ofocs and, since these

readings are not in length units, they were called occulation efciency).

3.3. Effect of the shear rate on the settling velocity

Since the shear rate inuences the occulation efciency (or mean

ocs size) in the way expressed in the previous sections, one would

expect a similar inuence on the settling velocity. This was conrmed

as shown inFig. 6. The results provided in this gure indicate that the

optimum shear rate corresponding to the maximum occulation ef-

ciency also corresponds to the maximum initial settling velocity of the

occulated suspension, thus the ultra-occulation test is an effective

method for identication of the optimalocculation conditions.

3.4. Effect of the solid concentration on the optimal shear rate

It is important to establish the optimum solid concentration for

occulation in a commercial thickener. In the majority of industrial

thickeners occulation is perform in the feedwell where the feed is

diluted with up-coming and circulating water. Knowing the solid

concentration that gives the bestocculation should permit calculation

of the water dilution ow rate.

Fig. 7shows the effect of the suspension volume concentration onthe optimum shear rate for a given occulation. Interestingly, at a

certain solid concentration value, the dependence of optimal shear

rate on solid volume concentration has a minimum, which shows that

for much diluted and very concentrated suspensions higher shear rate

must be used. This relationship between shear rate and solids concen-

tration can be explained using Smoluchowski theory because, at a

given suspension concentration, the oc size increases to a maximum

within a short time interval. On the other hand, with the increase in

suspensionconcentration the distributionofocculant macromolecules

within the volume of suspension slows down. In particular, this is

conrmed by the increased consumption of the occulant with the

increased suspension concentration at a constant time interval.

However with increasing shear rate, due to convective diffusion, the

rate ofocculant molecules dissemination in the suspension signicantlyincreases, leading to the growth of the dependence of theoptimum shear

rate on concentration in the region of large concentration values. This

may also lead to some decrease in the required occulant dosage as

shown byRulyov et al. (2005a, 2005b).

Fig. 3. Photograph of the UltraocTester: (Model: UFT-TFS-029 Turbootservice

Company).

Table 1

General data.

Solid concentration

[g/l]

Concentration [%] solid

by volume

Settl. veloc. vopt/v100[mm/s]

Shear rate _

[s1]

Flocculant dose

[g/ton]

Csvopt 104

[g/cm2s]

fbkopt/fbk100104

[m/s]

50 1.8 20/14 600 10 1000 3.72/2.60

100 3.7 13.6/9.0 500 8 1360 5.03/3.51

200 7.4 2.26/1.50 350 16 452 1.67/1.11

300 11.1 0.50/0.24 300 10 150 0.51/0.27

405 15.0 0.15/0.07 600 20 60 0.22/0.10

55F. Concha et al. / International Journal of Mineral Processing 104-105 (2012) 5357


4/5

3.5. Effect of the shear rate on the ux-density function

The best way to show the effect of shear rate on sedimentation is

to plot theux-density function versus the suspension concentration

for several shear rates at the optimum occulant dosage. This is

plotted inFig. 8.

If one takes into account that occulation in a commercial thickener

is generally performed at shear values lower than _ 100 s1

, the

results of this work reveals an interesting improvement that could be

obtained by ultra-occulating the feed before entering the thickener.

3.6. Effect of the shear rate on the supernatant clarity

It is well established and known that in the process of ultra-

occulation treatment supernatant clarity is directly related to

occulation efciency, so in the context of the paper this information

has been regarded as redundant.

4. Application to industrial thickeners

Solidliquid separation of tailings in thickeners relies on the ef-

cient mixing of slurry and dilute occulant solution in the tube

conveying the pulp to the thickener or within the feedwell (Owen

et al., 1999). In both cases the mixing, measured in shear rates around

10 to 20[s1], is too low for efcient occulation. One of the major

resent improvements is thickener feedwell design, for example,

Outotec company on its webpage claims that in the Vane Feedwell

the upper zone, into which feed, dilution water and occulant are

added, provides enhanced mixing and energy dissipation. This maximizesocculant adsorption, eliminates the possibility of coarse-nes segrega-

tion and ensures all particles are aggregated together by the occulant.

Efcient operation is maintained in this upper zone over varying feed

rates. The lower zone promotes gentle mixing for continued aggregate

growth, with the option for secondaryocculant dosing. This zone also

enables aggregates to uniformly discharge under low shear conditions.

Unfortunately there is no indication of the shear rate in the feedwell.

Based on its Laboratory Ultra-Flocculator,Rulyov et al. (2009)de-

veloped industrial sized Ultra-occulators, such as the one depicted

in Fig. 9. Fig. 10 shows the occulation efciency versus the shear

rate of a 200 [m3/h] for the quartz suspension treated by this utra-

occulation.

5. Conclusions

Proper addition and mixing of a occulants with the pulp is one of

the essential steps in the occulation process. Optimal dosage ofoc-

culant and the correct suspension concentration are not sufcient to

yield a good occulation. The shear rate at which occulation is per-

formed may decide on the quality of the process. Up to recently it

has been believed that mild mixing for a long time was the preferred

mode ofocculation, but work byRulyov (2004)demonstrate that an

intense mixing for a short time gives a better performance.

0

20

40

60

80

100

120

0 5 10 15 20 25

Flocculant dose, g/ton

Flocculationefficiency,relativeunits

1.8%

3.7%

7.4%

11.0%

15.0%

Fig. 4. Flocculation efciency versus occulant dosage with solid concentration as

parameter.

0

20

40

60

80

100

0 400 800 1200 1600 2000

Shear rate, 1/s

Flocculationefficiencyrelativ

,units

1.8%/1 0 g/t

3.7%/1 0 g/t

15.0%/20 g/t

Fig. 5. Effectof shear rate onocculation efciency for concentrations from 1.8 to 15 [%]

solid by volume, with occulant doses between 10 and 20 [g/ton].

0.1

1

10

100

0 400 800 1200 1600 2000

Shear rate, 1/s

Initialse

ttlingvelocity,mm/s

1.8/ 8 g/ton

1.8/ 4 g/ton

3.7/16 g/ton

3.7/10 g/ton

11.0/ 6 g/ton

11.0/16 g/ton

Fig. 6.Initial settling velocity versus shear rate for suspension volume concentrations

from 1.8[%] to 11 [%] by volume and occulant dose from 4 to 16 [g/ton].

200

300

400

500

600

700

0 2 4 6 8 10 12 14 16

Solid volume concentration [%]

Optimalshearrate,

[1/s]

Fig. 7.Optimal average shear rate _versus suspension solid [%] volume concentration.

56 F. Concha et al. / International Journal of Mineral Processing 104-105 (2012) 5357


5/5

In this work copper otation tailings froma major Chilean minewas

occulated with an ultra-occulator yielding the following result:

1. Flocculation at the optimal occulant dosage and optimal time

permitted an increase settling rate by 50% when changing from a

shear rate of _ 100 s1

to the optimum value of around_400 s1

depending on the suspension concentration.

2. The best occulation was obtained at the suspension volume

concentration in the range of 34 [% by volume] (75115 [g/l]).

3. An increase of the suspension concentration above the optimum

value, for example to 15 [% by volume] (400 [g/l]), increased the

necessary occulant dosage from 10 to about 20 [g/ton].

4. The suspension concentration and necessary mixing intensity are

interrelated. For example, an increase in the suspension concentra-

tion from 1.8 to 10[% byvolume] (50 to 280 [g/l])requiresa decrease

in the shear rates from 600 to 280 [s1] to maintain equal occula-

tion efciency, but with further increase in the suspension concen-

tration from 10 to 15 [% by volume] (280 to 400 [g/l]) the optimal

value of the shear rate increases from 280 to 600 [s1].5. The UltraocTester: UFT-TFS-029 used in this work allowed to

dene explicitly theoptimumocculationconditions for maximizing

theux density function for a given thickening operation.

Acknowledgment

This work was nanced by the AMIRA Project P996/CORFO Project

08 MC01-17. This nancial support is gratefully acknowledged.

References

Brger, R., Bustos, M.C., Concha, F., 1999. Settling velocities of particulate systems; 9Phenomenological theory of sedimentation processes ofocculates suspensions inan ideal batch or continuous thickener. Int. J. Miner. Process. 55, 267282.

Coe, K.S., Clevenger, G.H., 1916. Method of determining the capacity of slime settlingtanks. Trans. AIME 55, 203210.

Ding, K.J., Laskowski, J.S., 2007. Effect of conditioning on selective occulation withpolyacrylamide in the coal reverse otation. Trans. IMM 116, 108114.

Farrow, J.B., Swift, J.D., 1996. A new procedure for assessing the performance ofoccu-lants. Int. J. Miner. Process. 46, 263275.

Gregory, J., Nelson, D.W., 1984. A new method for occulation monitoring. In: Gregory,J. (Ed.), Solidliquid Separation. Ellis Horwood, Chichester, pp. 172182.

Hogg, R., Brunnaul, P., Suharyono, H., 1993. Chemical and physical variables inpolymer-induced occulation. Miner. Metall. Process. 10, 8185.

Hogg, R.,1999.Polymeradsorptionandocculation.Polymersin Mineral Processing.: In: Las-kowski,J.S. (Ed.), Proc. 3rdUBC-McGill Int. Symp. CIMMetall. Soc., QuebecCity,pp. 317.

Ives, K.J., Bhole, A.G., 1977. Study of ow-through Couette occulator II. Laboratorystudies ofocculation kinetics. Water Res. 11, 209215.

Keys, R.O., Hogg, R., 1979. Mixing problems in polymer occulation. AIChE Symp. Series,vol. 75, No 75, pp. 6372.

Kitchener, J.A., 1978. Flocculation in mineral suspensions. In: Ives, K.J. (Ed.), The ScienticBasis of Flocculation. Sijthoff & Noordhoff, pp. 283328.

Michael, A.S., Bolger, J.C., 1962. Settling rates and sediment volumes ofocculated kaolinsuspensions. Ind. Eng. Chem. Fundam. 1, 2433.

Muhle, K., Domasch, K., 1991.Stabilityof particle aggregatesin occulation withpolymers.Chem. Eng. Process. 29, 18.

Owen, A.T., Nguyen, T.V., Fawell, P.D., 1999. The effect ofocculant solution transportand addition conditions on feedwell performance in gravity thickeners. Int. J.Miner. Process. 93, 115127.

Richardson, J.F., Zaki, W.N., 1954. Sedimentation and uidization: Part I. Trans. Inst.Chem. Eng. 32, 3553.

Rulyov, N.N., 1999. Application of ultra-occulation and turbulent micro-otation tothe removal ofne contaminants from water. Colloids Surf., A 151, 283291.

Rulyov, N.N., 2004. Utra-occulation: Theory, experiment and applications. Particle SizeEnlargement in Mineral Processing. : In: Laskowski, J.S. (Ed.), Proc. 5th UBC McGill Int.Symp. Fundamentals of Mineral Processing. CIM Metall. Soc., Hamilton, pp. 197214.

Rulyov, N.N., Maes, A., Korolyov, V.J., 2000. Optimization of hydrodynamic treatmentregime in the processes of sorption-occulation water purication from organiccontaminants. Colloids Surf., A 175, 371381.

Rulyov, N.N., Dontsova, T.A., Nebesnova, T.V., 2005a. The pair binding energy of particlesandocs size formed in turbulent ow. Khimiya i Tekhnologtya Vody 27 (1), 117.

Rulyov, N.N.,Dontsova, T.A., Korolyov,V.Ja., 2005b.Ultra-occulation of dilutedne dispersedsuspensions. Miner. Process. Extr. Metall. Rev. 26 (34), 203217 (July-September).

Rulyov, N.N., Korolyov, B.Y., Kovalchuk, N.M., 2009. Ultra-occulation of quartz suspen-sion: effect of shear rate, dispersity and solid concentration. Mineral Processingand Extractive Metallurgy 118 (3), 175181.

Smith, D.K.W., Kitchener, J.A., 1978. The strength of aggregates formed in occulation.Chem. Eng. Sci. 33, 16311636.

Stein, H.N., Logtenberg, E.H., Van Diemen, A.J.G., Peters, P.J., 1986. Coagulation of

suspensions in shear elds of different characters. Colloids Surf. 18, 223

240.

-2.0

-1.0

0.00.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0

Solid volume concentration, %

Flux-den

sityfunction,m/h

Shear rate=opt

Shear rate=100

Fig. 8. Flux-density functions versus solid volume concentration [%] for shear rates of

100 [s1] and the optimum shear rate.

Fig. 9.Ultra-occulator used for the industrial occulation of coal otation tailings in

Ukraine and Russia;Rulyov et al. (2009).

0

20

40

60

80

100

0 2000 4000 6000 8000 10000 12000 14000

Shear rate,1/s

FloculationEfficiency,relativeunits

size

Documents

Flocculation Research Paper