:-:t

FUNDAMENTALS OF DEWATERINGFINE PARTICLE SLURRIES

P. SanasundaranProfessor

Columbia UniversityNew York, N.Y.

INTRODUCTION

Sludges and slimes ~o not often cEwater 'at the cEsired rates al:td as a re-

The prooleJ!! issuit their handling is a'serious proolem for many industries.

particularly acute whenever certain clay minerals are present in these waste

(' A prime example of this is the phosphatic slime that is generatedproducts.

Other examples ofduring the mining and processing of the phosphate rocks.

slow settling suspensions include red mud, acid mine drainage sludge, and coal

Large amounts of these slimes are generated .annually (Table l' and inslimes.

view of the loss of mineral matter in it and the environmental hazards created

by them, it is imperative that methods for dewatering slow settlinq suspensions

It would prove fruitful in this reqard to have a proper understan-be devised.

ding of the fundamentals of the dewatering or the subsidence behavior of t.'rtese

However, whi Ie sedimentation theories have been adequately deve lopedslurries.

to cover the two ext~mes of slurry concentrations, very little is known on the

basic aspects of sedimentation of slurries of intermediate concentration ranae

Thus we have theories for very dilute suspensionsthat is of interest he re.

based on Stoke's law and for consolidated beds based on models devel~ed for

sedimentation or dewatering behavior of concentratedflow through porous medii:

Close ex=ina~iC:1suspensions or slurries has not been adequately modelled.

2.[,

of the actual behavior of such a slurry during dewatering has enabled us to develop a

A systematic atte~t was also madephencmenological model which is described here.

as a part of this study to identify the characteristics of the minerals that are

responsible for the slow subsidence. behavior of such slurries. This has been

reviewed by us (12, 13) and is briefly summarized below

SEDIMENTATION MODELS

Sedimentation models can be developed on the basis of different properties of

While in some operations such as thickeni~g,the dewatering system (See Table 2).

others such as ef~the so~ds concentration of the pulp is the criterion, in

fluent treatment, it is the clarity of the supernatant that is of major cOncern.

Sedimentation of slurries have been modelled in t"-he past mostly on the basis of the

changes in the solids concentration of the slurry as a function of time in bat~,

sedimentation processes and is represented usu~ly as height of the slurry/super-

The si~lest sed~mentationnatant interface vs sedimentation time (See Figure 1).

curve is produced by a suspension of uniform size and is characterized by a linear

region indicative of constant settling rate followed immediately by a region of

Settling stops abruptly since the bed, if settled unhindere~ iszero settlinq.

On the other hand, if the slurry is made oftotally incompressible at this point.

compressible material such as clay, settling can continue for ever at a decreasing

Sedimentation curve of such systems will not shQl any constant settlinC]rate.

rate region (14) and is characterized by a reverse S-shape.

Number of investigators have studied the sedimentation of clays, even though

The peculiarnone has developed a model to accurately describe it (15-18).

*Sedimentation of network slurries are adequately called as subsidence in this

paper.

3.

sedimentation behavior of the clays is ~~e result of the tendencj for clays to

Michaels and Bolger (19) were among the first two to take the floc-floccula.te.

culation properties of the sedimenting material into account in the formulation

They considered the basic unit to be a floc ratherof a sedimentation model.

shear rates, the flocs group into largethan the individual particle. At:: low

clusters called aggregates, and the aggregates in turn form an extended three

Three different types of settling curves are obtained indimensional network.

this case. one for each solid concentration region. Fiqure 1 illustrates these

three types also.

For very dilute suspensions, flocs can be considered to settle individually

on the basis of Stoke's law and the following expression based on Richardson and

Zaki equation for the group settling of uniform, spherical p~rticles describe$ this

4-65adequately. (1)v = V £:

i a

is the Stoke'sWhere Vi is the settling rate of slurry/supernatant interface, Va

velocity of single aggregates and £ is the void fraction. Assu.~ng the ave rage

to be independent of the clay concentration and in-diameter of the aggregate da

variant during settling, equation(l) can be rewritten as-2

q(p -o)dVi(O) - s rw a (l-C)4-65~18 Cas' s

~w asWhere V (0) is the initial settling rate,g is the gravitational acceleration, Ps

i

(2)

and ~ are densities of solids, and water respectively, ~ is the viscosity ofw w

is the ratio of vol-the water and ~ is the solids

volume concentration, and Cas

The above mo:ie]. has been tested withume concentration of aggreqate to solid.

some success for kaolin (19) and Ti °2 (20) but it proved inadequate for the case

of alum It'.ud (20).

SIur~ies of intermediate concentrations often ~ppear like a smal~ bulky aggre-

The settlinggate and they can be considered to settle as a coherent network.

will be determined in this case essentially by the balance between the qravity

4.

forces (fg>' support force (fu> exerted 'by t:-.~ underlying material, the wall

support force (fw> and frictional force (fi> for the flow of fluid through

spaces between the aggregate. At equilibriu.-n:

f + f + f + f f - 0U w 9 (3)

(4)

= 1fD 6Hac y

f ( 5)w

=~D4

f (6)q c

T281. 7')

the underlying compressed zone, a is the yield stress of the slurry, < is they

Kozeny's shape constant in the Kozeny - Ca~,an equation, ~is the tortuosity fac-

tor, and S is the specific surface area. F~om equaticns 3 to?, the following

expression results relating settling of the slurry to the container diameter,

v. = v' (0) (1- ~ - ~)1. i Dc Hi 8)

where

(9)

D =(10)y

4 0-Y

9(PS-Pw) 4>s

ac H.1-

~H - -q(.p - p; ) Its w. s (1 I}y

The above expression has been tested for kaolin suspensions and the results suc-

gested that the container diameter b...s .a :'Ie-::ligible effect on the settling of. floc-

culated slurries provided that it is not as small as D-Uur results (21) fo'ry

phosphatic slime, however, showed a definite effect of the container diameter on the

settling rate (See Figure 2). This effect and other features that are Char-

acteristic of the slow settling sludges can however be explained well using our

phenomenological model.

Phenomenological Model

The observed behavior of the slurry during dewatering is schematically shown

There are four stages that are distinctly exhibited by the slurryin Figure 3.

during this time:

(1) Almost immediately after mixing is stopped, all rotacional movements cre-

ated during the suspension terminate

(II) In a few minutes, coarser size particles and small air bubbles trapped

Water seeps up throughin the solidified slurry n\ove through it creating tears.

these tears and forms lenses of water.

leading to channels connecting various(III} Additional tears develop

lenses, and finally opening up at the slurrj - water interface to permit water

The slurry/supernatant interface now subsides rapidlY and water along withto exit.

entrained particles can be seen sprouting through in the fo~ of microvolcanoes

The effect of the containerLiquid also seeps up along the container walls.

diameter that was discussed earlier can in fact be explained by taking into

account wall area/unit mass of slurry available for water seepage.

(IV) Continuous compression of the slurry during this dewatering process

finally causes the contraction of the channels which in turn reduces the rate of

additional dewatering-

Accordinq to this mod~l, dewaterinq is dependent essentially upon the

paths for the water and in the present system suchavailability of seepage

paths are proposed to be essentially the result of the heteroyeneity of the sys-

This hypothesis was tested by determining the effect of the addition oftern.

eoarser particles to the slurry and the generation of microbubbJ.es in it. If

"""" ,.;o'~,,~~,~q~,.,., '._"~ -'-_'.-y---:;;-_c;:~~ '~~~,:,"~~~~~..L-= ~,., -,,-- ~~.- '-'-'",,',"~'-":"'-'-'"

the hypothesis is valid, such procedures shoULdenhance the dewatering of the

slurry.

Effect of Additi9~f Coarse Particles

The effect of the presence of coarse particles was tested by mixing the slime

The results given in Figure 4 show that the addition doeswith sand tailings.

mecha-The proposed seepageindeed enhance the settling rate significantly.

nism itself was tested by conducting experi~ents using particles with a wide

range of properties, and the results are presented in Figure 5. Evidently, in-

crease in the specific gravity of the additive produced no measurable effect o~

subsidence suggesting that the observed effect is not due to any increase that

It is noted that cassi-could have occured in the weight of the sliI:\e network.

terite, the heaviest mineral, was not as ef=ective as the other lighter miner-

This was oonfirmed to be due to the fact that the heavy cassiterite par-

ticles had broken through the slurry even before it gelled and were not thus

Interestingly, siliconeavailable for creating the tears for water seepage

coated glass beads also broke through the slurry similarl~ producing no effect on

Since glass beads without silicone coating did enhancethe subsidence

the settling, the above effect must be attributed to the hydrophobic nature of

Evidently polar interactions between the particlesthe silicone coated surface.

and the aqueous medium are helpful in causing en trapment of the particles in t~e

Indeed irregular quartz particles a~e more effective than even the un-slurry.

hydrO!>hilicity of thecoated qlass beads suggesting that in addition to the

surface, the morphology of the particles also plays a governing role in deter-

The effect of graphite and molybdenite might havemining the subsidence rate.

resulted from the presence of some hydro?hilic sites on the surface as well as

Indeed it is possible to alter the interactions be-irregular morphology.

tween the particles and the slurry by means of chemical additive9~ particularly

Polymers can affect the aggregation and the~efore dewateringpolymers (22-24).

7.

by modifying the structure of flocs and aggregates as well as by e~~~~cing

bridging of particles to clay and thereby trapping of particles in the slurry.

Principles of polymer adsorption and flocculation have been reviewed in detail

Addition of polymers in proper dosages can indeed causerecently (25, 26).

flocculation. the extent of adsorption and flocculation being depencent upon

polymer properties such as molecular weight, functional group distribution, and

charge density, solution properties such as pH, ionic strength and temperature

While polymers inand mineral properties such as surface charge and porosity.

proper dosages can be expected to enhance the initial settling, the final solid

content can often be low due to entrapment of large amounts of water inside bul-

ky flocs. In such cases, restructuring of flocs and aggregates by ~ech~~ical

me~~s becomes a necessary part of the process to achieve the desired solid con-

Effect of Air Bu!>bles

Settling rates of slurries which were subjected to suction, are compared in

Figure 6 with that obtained for control sQm?les. It can be seen tha~ qeneration

The bubblesof bubbles due to suction does enhance the subsidence considerably.

were found to alter the physical features of the slurry such that ~ater seepage

It appears that any means for the generation ofwas possible at an enhanced rate.

tiny bubbles might prove useful for enhancing the subsidence particularly whe~

in combination with other techniques These dbservations do closely support

the proposed mechanism of enhanced subsidence

IDEALIZED PHENOMENOLOGICAL MODEL

Enhanced settling of materials that has a tendency to form a network.stu~-

ture is a complex process because it is governed by a combination 0= severat ~~-

The cot:IPlexchanisms, ea~h predominating in a different concentration range.

Majornature of this sedimentation is illustrated in Figures 7 and 8 (27).

8.

Thestages and their relative durations have been identified in these Figures.

phen~noloqical equatio:n is derived considering the fact that a slurry with a:-.

internal structure subsides according to the action of two interrelated processes

namely qravitational expulsion of water fro~ the slurrl and internal resistive

Wnile the rate of expulsion of wate~forces that oppose the movement of water.

itself depends upon the amount of water contained belo'A the interface in the

suspension, the resistance depends upon t~e amount of water which has already

This can be expressed as:passed through the interface.

- ~ = c + kcll(W) III (1-E;(W» (12)dt

Where W is the fraction of recoverable water remaining in the slurry at

c istime t; k is general rate coefficient and :1:,"l, and ~'are general functions.

the value of constant settling rate and t~<es into account the fact that there

preceeding the development of resista:,ce.may be initial incompressible settling,

The form of this equation that is applicable for various stages can be obtainec. by

examining the nature of subsidence and the type of water (interaggregate, inte~-

floc and intrafloc) that is involved in de.~atering during each stage

In the first stage, called lenticular stage (H(O) to H(tl) , gel structure

c indevelopes in the suspension, but there is negligible interface settling-

Coarse particles and air Qubbles move through theequation (12) is zero.

Water seeps up through the channels and forms le~.-suspension forming channels.

Thus, at the end of this stage, the suspension consists of solids gro~pe~ into ases.

loose network of aggregates made up of flocs ~~th lenticular water filled fissu=as

interspersing throughout the network structure

During the next stage, called reticular stage «~(tl to H(t2»' lenses be-

of water inte:--come interconnected such that continuous filaments or reticules

As channels open up at thesperse the loosely conected floccular aggr~gates.

interface, interaggreqate water is expelled from the sedimenting mass slowly a~

first, then more rapidly and gradually sl~~ing as aggregates make contact with

'J.

each other.

(13.)\

and

(13b)

w~ being the fraction of interaggregate water at time t and r and s being rate

The equation is based on t:--': =act that the rate of expulsion, willcoefficients.

depend upon the amount of water present i~ t~e suspension and on the degree of

bridging which provides a growing resistive force which ~tself is proportional

to the water that has already escaped.

In the vermicular s~age CHCt.2) to HCt3) the aggregates establish closer

Thecontacts reducing the water filaments into a narrow vermiform stucture.

bridging allows forces to be transmitted ~hrough the mass and subsidence proceeds

at a decreasing rate as the water in the structure is expelled and continues until

the channels close. Equation 13 wilJ. descri::>e this s.tage, but the exponential expres-

( 14a)

( 14b)

This equation represents a moderating rate of water ex-a and b being constants.

pulsion as compared with the fast flow rate during the latter part of the reti-

cular stage.

When all the water in the macrochannels between aq~regate has been expellec

the end of the vermicular stage is reached a~d further dewate~ing takes place

HCt) to H(~». Waterdue to the consolidation of flocs (floccula~ stage:

is now removed from the microchannels bet'Neen flocs and at a slower rate from

within the flocs. These processes deterni~ed by gravitational and viscous forces

follow first-order kinetics

(15a)-dt

10.

and:

(lSb)H(t) + m exp (-P(t-t3) +n exp (-q(t-t3»+H(~) q < p

Where Wf is in~rand intra flocs water at time t, m and n are respectively

coefficients referring to fast and slow parts of the floccular reqio~, and

p and q are corresponding rate coefficients.

At ~e end of intrafloc processes, the system approaches an equilibrium be-

tween gravitational, frictional, and electrical forces with viscous forces

Durinq this staqe the plot of suspension height versus timeapproaching zero.

asymptotes to a constant value

The phenomenological model was tested by first determining various ratecoefficients

.0\5 can be seen fromthen comparing the resultant curves with experimental curves.

To developFigure 9, the model fits the experimental results satisfactorily.

a full understanding of this type of subsidence, it is however essential to iden-

the major. physical and chemical factors that determine the slo'~ settling

behavior and then to arrive at a fundamental interpretation for all the para-

meters including the rate coefficients

MOtEL SLOW SETTLl~:G SW RRi

Slow settling behavior of the s~imes is aenerally attributed to clays and

, .

clay type minerals, but not all clays produce slimes of the phosphat1.c type.the

To ideqtify characteristics of clay responsible for/problematic features experiments. . .

were conducted with components of ~hosphatic slimes, and mixtures of these cor,~onents(:

Three types of minerals were used to arrive at the following model systems:

These three mineralsclay minerals - montmorillonite, attapulgite, kaolin.

are major components of phosphatic slime

These fibrous minerals were in-asbes tos minerals - amphibole, chrysoti Ie..

cluded to determine the role of such a mineral feature

sand - q~rtz, This is also a component of the phosphatic slime. but it is

neither fibrous nor clay type

1L

Single, binary, ternary, and quarternarj mixtures of the above minerals .~'ere

tested for the following eight basic characteristics of the phosphatic sli~e:

1. an initial period of gelling

slow subsidence with continuously varyinq settlinq rate2.

minimum segregation of mineral cons~ituents during subsidence3.

4. a bulky sediment

presence of fissures and channels during subsidences.water exitinq as microvolcanoes6.

a sharp slurry/supernatant interface7.

8. a clear supernatant

Subsidence behavior of the typical single, binary and ternary systems are s~o~~

It was found that none ofin Figure 10 along with that for phosphatic slime.

the constituent minerals themselves or their binaries resembled the industrial

slime with respect to the above eight characteristics. From the various co=-

of montmorillonite, kaolinite, attapulgite, quartz, chrysotile a~dbinatiOns

studied only the montmorillonite-attapulgite-kaolinite te~a~.amphibole

and the montmorillonite-attapulqite-kaolinits-quartz quarternary showed t~e

major slow settling features of the phosphatic slime. The role of the mo~~~r-

illonite and the fibrous attapulqite particles was particularly evident si~~e ~he

absence of either mineral made the behavior of the slurry most different from

It is evident that the morphology of the min~ralsthat of the phosphatic slime.

does playa major role in determining the de'.~atering rate.

The experiments conducted as a functio~ of pH shoTNed the supernatant to be

clear only when the solution pH was such ~hGt the min~rals were oppositely cha~ged.

Apparently fava~able electrostatic interactions that can induce attachment of ~he

colloidal mineral matter to the subsiding mass are required to produce a clear

Thus while the settlinq rate and the solid content is depe~dentsupernatant.

essentially upon the size and morphology of the mineral components, the clarity of

the supernatant is governed by the surface charge properties which can in tu~ be

12.

and salinitycontrolled either by adjusting the solution properties such as pH

or by addition of polymers or other chemicals that can adsorb on various compo-

nent minerals.

CONCLUDING R~:": ~KS

MuCh work has been done on the dewatering of sludges such as phosphatic

slime, acid sludge, coal slime and red mud, and yet very little is established

Also settling of theseon the basic reasons for their slow settling behavior.

slimes have not been adequately modelled.

In our study, the mechanism by which en~anced dewatering of phosphatic

slime occurs has been identified based on the actual Qbservation of the phos-

phatic slime during subsidence. Also a phenomenological model developed to

d~scribe during various stages is successfully tes~ed for this

To conduct any fundamental studies on these slurries, it is most usefulsystem.

to have a reproducible known model slurry and a mineral mixture that closely

resembles the phosphatic slime in its subsidence behavior has been formulated

towards this Clearly, there exists now a need to systematically stu-purpose.

the effect of relevant physical and che~ical factors on the dewaterin9 ofdy

the model slurry and problem slurries. and from the results to identify the

reasons for thc pertinaceous behavior of these slurries inorder that adequate

technology can be developed for their dewatering.

ACKNOWLE IJ;E:.tEl;':'

The support of the Particulate and }bltiphase processing program of the

Contribu-National Science Foudation <CfE-80-ll0l3) is gratefully ackno~11edged.

G.C. Sresty, and Mr. E.L. Smith,tions of Prof. C.C. Harris, n:-. D.R. Na~ra;, Mr

Jr. with the author (Ref12,21,27 & 28) used in this review are also acknowledged

"

'~-'.:_~«~

REFERENCES

1. Imhoff, K., et a1., "Disposal of Sewage and Other ~later-borne

Wastes" , Ann Arbor Sci. Pub., Inc., Ann Arbor, Mich., 1971.

2. 18S1,Stokes, G.G., Trans. Cambridge Phil. Soc.. .2.. 11

3. Happel, J. and Brenner, H., Low Reynolds Number Hydrodynamics

with Spo'cial Ref. Engelwoodto Particulate Media. Prentice-Hall.

Cliffs, 1965

4. Cambridge University Press, London, 1932.Lamb, H., Hydrodynamics.

5. Trans. lust. ofCarman, P.C., Fluid Flow Through Granular Beds,

Chem. Engrs., !i, 1937, 150.

6. P.'C., Determination of the Specific SuTface of Powders.Carman,

.1. Soc. Chem. Ind., 11., 1938, 225.

7. Kangelsdorf, P.C., Convenient Plot for Exponential Functions

30, 1959,- 442with Unknown Asymptotes, J. Appl. Phys.,

8. 11, 1962. 89Ergan, S., them. Eng. Progr.,

9. Wyllie, "Fluid Flow Through Unconsoli-M.R.J. and Gregory, A.R.,

dated Porous Aggregates, Effect of Porosity and Particle Shape

1379.1955,on Kozeny-Carman Constants," Ind. Eng. Chem., £,

"The Flow of Fluids Through Granular Beds: Effect10. Coulson, J.M.,

of Particle Shape and Voids in Streamline Flow." Trans. Inst.

Chem. Engrs., rr. 1949, 237.

"Determination of Specific11. Edmundsort, I.C. and Toottll, J.P.R.,

!.§..Surface Area by Air Permeation at Low Porosities," Analyst,

1963, 128.

12. "Dewatering of Slow-sett.li:-.gG.C. .Somasundaran, P. and Sresty,

Sludges," Second Handbook Volume on Particle-Fluid Separation tech-

nology, 1977.Chicago,private publication by lIT Research Institute,

Somasundaran» P.13. "Th.icken:lng or Dewatering of Slow settling min-

"eral suspensions, XIII the International Mineral Process. COGar.

Proceedings, Elsevier (in Press)

"Effect of14. Somasundaran, C.C. tP., Smith, E.L., Jr. and Harris,

Coarser Particles on the Settling Characteristics of Phosphatic

Slimes," Proc. 1st Int. Chicago,Conf. on Particle Technology,

August 21-24, 144-150.1973, I.I.T. Research Institute,

15. "Flocculation of Mineral Sus-Wadsworth, H.E. and Cutler, I.B..pensions with Coprecipitated Polyelectrolytes," Mining Eng.. ~,

1956, 830.

Subsidence,16. La Mer, V.K. and Smellie, Jr., R.H., "Flocculation,

g, 1957,and Filtration of Phosphate Slimes," J. of Collo~d Sci..230.

17. 1.b1d., 566.

"Measuring Sedimentation-Flocculation18. Bramer, H.C. and Hoak, R.D.,

Chem. Process Design Development~ ~~Efficiencies," Ind. & Engng.

1966. 316.

Hichacls, A.S., Bolger, J.C., Settling Rates and Sediment Volumes19.

of Flocculated Kaolin Suspensions, Ind. Eng. Chern. Fundam't It 1962

(Feb.) 24.

20. Bodman, S.W., Shah, Y.T., and Skriba, M.C., Settling of Flocculated

Ind.Suspensions of Titanium Dioxide and Alum Mud in Water, Eng.

Chern. Process Design Develop., ll. 1972. 46.

21. Somasundaran, P., Smith, .E.L., Jr., and Harris, C.C., "Dewatering

"XI th Int. Miner.of Phosphate Slimes Using Coarse Additives,

Proc. Ist1tuto de Arte Minerari e PreparazioneCongr. Proceedings,

dei Minerali, Universita de Cagliari, 1975, 1301.

,1957,

La Mer, V.K. and Smellie, R.H.,22. Jr. ~ "Theory of Flocculat ion ~

Subsidence and Filtration Rates of Colloidal Dispersions

"Clays and Clay Minerals, !'Flocculated by Polyelectrolytes,

1962.295.

23. Smelley, A.G. and Feld, I.L.. "Flocculation Dewatering of Florida

Phosphatic Clay Wastes, ..U.s. Bureau of Hines Report of

Investigations, No. 8349, 1979

24. "Dewatering ofSmelley, A.G., Scheiner, B.3., and Zatko. 3.R.,

Industrial Clay Wastes", u.s. Bureau of Mines Report of Invest-

igations, No. 8498, 1980.

25. Sresty, G.C., Raja, A., and Somasundaran, P., "Selective

:Flocculation of Mineral Slimes Using Polymers" , in Recent

Developments in Separation Science, Vol. 4. CRC Press, West

Palm Beach, Fl., 1978, p' 93-105; Hollander, A., Somasundaran, P.,

"Adsorption of polyacrylamide and sulfooatedc.c. Gr~te,

polyacrylamide on kaolinite", in Adso!:ption from solution, Plenum,

(in press)

26. "Principles of Flocculation, Dis.persio..n, andSomasundaran, P.,

Selective FloccUc.1ation", in Fin-e Particles Processing, P.

947 - 976Somasundaran, ed., !, AIME, New York, 1980, p.

"Sedimentation of21. Harris, C.C., Somasundaran, P.. and Jenson, R.R.,

Compressible Materials: Analysis of Batch Sedimentation Curves,"

Powder Tech., li, 1975, 75.

28. "SubsidenceNagaraj, D.R., McAllist~r, L., and Somasundaran, P. t

of Suspension of Phosphatic Slime and its Major Constituents,

"Int. J. Mnl. Proc., ,i, 1977, Ill.

Table 1

Million Tons per YearWaste

phosphatic slime 40-50

8-10Mud

Acid Sludge 0.5

Slimes 10.

Others: Potash, Clay fine-s, Uranium taili.ngs, drilling mud

waste, paper and pulp effluent, TiO2

-

. ""'-;~,~.i!;="- -c-c-~,~,~,L~ " ~'.~'

Table 2. Kajor Criteria in Dewatering

Solid concentration of sediment

Volume of sediment

Settling rate

Supernatant clarity

Cake yield strength

Cake moisture

LIST OF FIGURES

Typical batch sedimentation curves for pulps of various solid1.concentrations (19).

Diagram illustrating the effect of container dia~eter on the slurry!2.supernatant interface height, after 8 hours of subsidence;

-37pm slimes; Initial height of the slime column. 17..8 cm

Schematic representation of processes that a 2.6% phos~~atic slime3.containing 0.5g c'oarse particles underwent during subsidence

Observations: 0 min., suspension gels immediately after

mixing is stopped; ~lO ~inutes, moving particles and air b~~les

to ~3S gin., water concentrates into len-create tears; "'13 min.

channels and microvolcanoes permitses around the tears; ~2 hours,

channelswater filaments depleted,"'4 hours,enhanced dewatering;

(27j.begin to close and dewatering rate becomes very s~all

Diagram showing the effect of quartz flotation tailings on the4.vertical bars indicatesubsidence of 2.6% ~-37pm phosphatic slime,

(21).range when larger than symbol

Effect of addition of various types of coarse particles on the5.(2.6% -37~m phosphatic sliQe)! super-.height obtained for the slurry

natant interface at various subsidence times (21).

Diagram illustrating the effect of air bubbles generated b}" suc-6.(28).tion on the subsidence of phospha~ic slime

Batch subsidence curve, H(t} vs. plot showing pricipal7 108 t re-

glons. The duration of the stages depend upon the characteristics

of the solids and the experimental conditions (27).

8. Idealized sedimentation model. A: Reticular stage commences,

Aggregates come in contact, water lenses form; B: Reticular stage

completes, Aggregates are increasing the number of ~oints of mu-

tual contact; C: Vermicular stage commences, Aggregates are brid-

sed strongly, water filaments interconnected but less intensely

Vermicular stage partiall}" ccm-than during the pr~vious reticular stage: D:

plete; E: Floccular stage commencing, water pr~sent only as i~ter

and intrafloc water (27).

(21).9. Diagram illustrating the fit of the phenoClenological equation

16. a. Reight of the slurry/supernatant interface vs. settling time for

(2.5%); in parantheses indicate thethe numberssingle mineral systems

pH values (28).

b. Height of the slurry/supernatant interface vs. settling time for

binary systems containing kaolin (K) and chrysot ile (Ch), at-:.ap".:.l-

site (At). amphibole (Am), and quartz (Q). Curve 1 for K: Co"} - 1:1

(pH 6.7 and 8.5),(pH 8.8), 2 for K:At-l:4 (pH 8-3); 3 for K:Am=l:l

(pH 7-8). 5 for K: Qal:~ (pH 6.4). 6 for phosphatic4 for K:At-l:l

slimes (pH 8-2), (28)

c. Reight of the slurry/supernatant interface versus settling time

for binary systems at pH~5 containing M~ntmorillonite(:1) and A~~~

Curves 1 and 4 for At:M-4:1,pulgite(At) and Chrysotile(Ch).

and 6: 1 respect1.ve-ly, 2, 3 and 5 fo!" Am: M=4:1, 521 and 0:1 respectiva:l?, 6

for Ch: M=4:l, 7 and 8 for phosphatic slimes at pH 6-2 and 8-2 respectivel? (28)-

d. Height of the slurry/supernatant interface versus settling

time of 'Montmor111onite(M) , Attapulgite(At), Kaolin(K), and

(pH 4-8), 1:1:3Quartz(Q). Curves 1,2 and 4 for M:K:At.l:2:3

3 and 5 for M:Q:At-2:2:2(pH"-6) and 1:1:6 (pH "-6) respectively.

(pH 3-8) and 1:1:4 (pH~4) respectively (28)

.

0r()~

w~~(/)

E:l-

t'--f't')I

<..?

U').

0

~

qv

0C\J.0 <J t>c

U')Q:J:

w~-J-

0-

-i

000

0m

0a>

0V

0 0 0I"- W In

'3~'V'.:IH3.lNI :JO. J.H~13H%..

'3""~~ '~':~~'!'~:'l'!I'~~~~-~O::~ '~-:~v-.,--~~-,~ -~."7:-r ~~~~.;i"'~~~'

(.!)

0 .<!J

0.(\J

~

"00-%

c 3::>

-v.:fH3J.N

I 3H

.1 .:f0 l.HS

I3H

2:0-..-(,)::>U

>

02:0

0 0

0 ~

I cr.w~(!)z

--O(/)

J:::>eE

~a(/)(\J::>

a:zO

O-::>

~zz'«o-cr.-'--1-20..uou>::>

u'«(/)0

(!)z-(/):).~

-~a:

000:I:N

'-E

>-<

tE

a:~Q

)wQ

.->

(/)Lw

<t

Z

.a:o~

w

~~

~uo~:)->

'(/)0

1,

<!)

z(/)~.

z~coft>

a..

:I:>-~

Ea::J

Ew

a..

O>

~~

W:J

-.ZU

:JcW

O-U

>~

"-&n

U

.-

~(\J"",

(/)<J

-L0

w~-I-

.z-~

CD

(,0

~

" ~

I

.

.~~Ic),.~,.~."

%

.z-~

. 3~-V

-.:IH3l.N

I 3Hl..:I0

.lH~

13H