Mineral Processing and Extractive Metallurgy Review

An International Journal

ISSN: 0882-7508 (Print) 1547-7401 (Online) Journal homepage: http://www.tandfonline.com/loi/gmpr20

Modeling of flotation process – an overview of

different approaches

Mousumi Gharai & R. Venugopal

To cite this article: Mousumi Gharai & R. Venugopal (2015): Modeling of flotation process – an
overview of different approaches, Mineral Processing and Extractive Metallurgy Review, DOI:
10.1080/08827508.2015.1115991

To link to this article: http://dx.doi.org/10.1080/08827508.2015.1115991

Accepted author version posted online: 20

Nov 2015.

Submit your article to this journal

Article views: 34

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

http://www.tandfonline.com/action/journalInformation?journalCode=gmpr20

Download by: [University of Sussex Library] Date: 04 December 2015, At: 22:08
 Modeling of flotation process – an overview of
different approaches
Mousumi Gharai and R. Venugopal

Department of Fuel and Mineral Engineering, Indian School of Mines, Dhanbad-826004

t
ip
Email id: vrayasam@yahoo.com

cr
Abstract

us
The modelling of flotation has been reviewed with the aim of identifying their usefulness,
Mineral Processing and Extractive Metallurgy Review

significance as well as limitations. From literature it is clear that, various flotation models have

an
been developed based on the processes and sub-processes occurring in flotation. An overview of

the literature indicates that various approaches have been adopted in quantifying the process.
M
Flotation models based on kinetics have prevailed in almost all flotation conditions regardless of

the ore type and ore characteristics as well as flotation cell configurations. It may be concluded
ed

that the classical first order kinetic model is comparatively a better model and can be utilized to

optimize the flotation process as it is applicable to both batch and continuous flotation processes
pt

with high confidence level.

ce

Finally, it is suggested that future work should focus on improving the models which can more

accurately predict the prevailing conditions and enable better optimization of the flotation
Ac

process.

Keywords: flotation, kinetics, modelling approaches, kinetic models, classical models

1
 1. Introduction

Froth flotation is a widely used separation process throughout the mineral processing industries

in the last century because of its ability to treat complex and low grade ores, where the mineral

t
ip
particle size is too small for other separation techniques to be efficient. It can be a highly

efficient process for solid-solid separation of minerals as well as for water and waste water

cr
treatment, and more recently for de-inking of recycled paper, electrolyte cleaning (oil

us
Mineral Processing and Extractive Metallurgy Review

separation), separating plastics from each other, removing radioactive contaminants from soil

and carbon from fly ash.

an
In the case of water and waste water treatment processes, very fine particles of about a few
M
microns in size are removed in very low concentrations, which are collected by small bubbles of

50–100 microns, whereas in mineral processing, particles typically less than 150 microns, in 10–
ed

40% solid suspensions, are selectively collected by bubbles of 0.5–2.5 mm size. The difference is

in the type of flotation used, i.e., dissolved air flotation (DAF) for the water treatment and
pt

dispersed air flotation for minerals separation. Studies on DAF modelling are rare and mainly

related to hydrodynamics (Kwon et al., 2006) although DAF has been utilized in some flotation
ce

column circuits with good success.

Ac

Despite the many differences between mineral flotation and flotation deinking (Heindel, 1999;

Schulze, 1991, 1994a, b; Pan et al., 1993), all flotation cells operate on similar principles and

three separate processes take place in sequence after hydrophobicising of the desired mineral:

1. aeration, where air bubbles are introduced into the system

2
 2. mixing, where bubbles and particles are intimately mixed to maximize bubble–particle

interaction and

3. separation, where bubbles and bubble–particle aggregates are allowed to separate from the

bulk mixture and skimmed away.

t
ip
The dosage of reagents and their interaction with particles of different size are two major factors

cr
in the flotation process, because the resultant product quantity as well as quality is a measure of

us
their effective interaction. The process of flotation, thus, is an intense interactive “Engineering
Mineral Processing and Extractive Metallurgy Review

System” depicted in Figure 1. From Figure 1, it is clear that for a given ore, the flotation system

an
is interrelated to the chemical, equipment and operational factors and a change in any one of

these factors will result in a change in the over-all flotation response of the material ( Klimpel,
M
1995). With this general macroscopic view of the flotation process as a whole, several distinct

micro processes can be identified. Each of these plays a role in determining how individual solid
ed

particles will respond while they are in the flotation cell, under a given condition of operation.
pt

The entire mineral industry flotation practice is growing with the knowledge of the process and

sub-processes of flotation. With this increased knowledge, more efficient flotation models are
ce

being developed regardless of machine type and material floated. Models developed are being

used for process analysis and simulation and for the improved understanding of the flotation
Ac

process prior to operation of an automatic control system. Models are used for both online and

offline simulation and optimization of circuits. Numerous modelling methods are available, both

commercially and in literature, which can be used to predict the flotation performance. The

application and usefulness of these models in the plant suffer from the uncertainties due to the

3
 complexities of industrial operation (Ross, 1990). In this paper the modelling of both batch and

continuous processes are considered.

2. Modelling Approaches

t
ip
In the initial flotation model of the 1930’s, as reviewed by Dowling et al. (1985) flotation

cr
recovery was expressed as an exponential function of time. Subsequently many flotation models

have been proposed and published in the literature.

us
Mineral Processing and Extractive Metallurgy Review

A flotation model is similar to a chemical kinetic model having the form as follows:

dN 1
dt
= f (k , N i ) = −k1 N 1 − k 2 N 2
m n
an (1)
M
The above model describes the change in particle concentration, N 1 , with respect to time t, as a
ed

function of certain concentration(s), N i , and rate constant(s), k i . The negative sign indicates

decreasing concentration due to flotation of particles out of the cell. The exponents m and n are
pt

the order of the process.

ce

Most researchers agree that flotation is a first order process and a function of only particle

concentration and a rate constant (Kelsall, 1961; Arbiter and Harris, 1962; Mao and Yoon,
Ac

1997).

dN 1
= −kN 1 (2)
dt

4
 The assumption of first order kinetics may not be always valid as there are other factors which

affect the kinetics of flotation such as reagent types and concentrations, airflow rate, impeller

speed, particle size etc. (Jian-Guo-Zhang, 1989). Many researchers have found flotation kinetics

to be second order (Tomlinson and Fleming, 1965; Mori et al., 1983; Hernainz and Calero, 2001;

t
ip
Somasundaran and Lin, 1971; Yalcin and Kelebek, 2011) and non-integral order (Hernainz and

Calero, 2001).

cr
Assuming perfect mixing, if the value of rate constant and residence time within the cell is

us
Mineral Processing and Extractive Metallurgy Review

known, recoveries of each individual species can be calculated from following equation:

R=
kτ
1 + kτ
an (3)
M
Since recoveries are the desired output from flotation models, rate constant is a useful parameter.

Throughout flotation modelling history, the attempts made are to produce a general flotation rate
ed

constant equation.
pt

A comprehensive description on fundamentals of the flotation modelling of mechanical cells has

been presented by King (2001) and for column flotation by Finch and Dobby (1990), Rubinstein
ce

(1995) and Finch et al. (1995).

Ac

Mathematical models describing flotation as a rate process are classified into six categories

(Mehrotra and Padmanabhan, 1990):

a) kinetic models based on chemical analogy

b) probabilistic and stochastic models

5
 c) multiphase models

d) mechanistic models

e) kinetic models with distributed rate constant, and

t
ip
f) continuous flotation models

cr
In 1981, Lynch et al. conveniently defined the flotation models into three categories: empirical,

us
probability and kinetic models.
Mineral Processing and Extractive Metallurgy Review

On the basis of the overall process and sub-processes occurring in flotation, models can be
an
broadly classified under two categories: macro-scale and micro-scale models which have been

used to describe the kinetics of flotation (Mika and Fuerstenau, 1969; Schulze, 1977 and
M
Chander and Polat, 1995). Figure 2 represents general classification of flotation models.
ed

In the subsequent part, only the models of flotation on macro-scale are presented.

2.1. Macro-scale Models

pt
ce

In macro-scale modelling, the overall response of the flotation process is related to various

operating parameters through a set of mathematical equations. The macro-scale models can be
Ac

further subdivided into two categories, namely; theoretical and deterministic.

2.1.1. Theoretical Models

6
 Theoretical models involve theoretical assumptions and complex equations that include a

number of unknown variables which are difficult to determine but yet useful in providing an

insight into the possible events in the flotation process. Theoretical models can be further divided

into two categories: phenomenological and empirical (Woodburn, 1970; Lynch et al., 1981;

t
ip
Herbst and Bascar, 1984).

cr
2.1.1.1 Phenomenological Models

us
Phenomenological models are derived from conservation of mass, momentum, and energy
Mineral Processing and Extractive Metallurgy Review

equations. These models provide relationships between the cause and effect of the process

an
through the use of equations that are derived by understanding the physics and chemistry of the

important sub-processes. These models are usually valid over a wide range of operating
M
conditions, complex and often take much time to develop.
ed

Bergh and Yianatos (2011) and McKee (1991) highlighted several issues that still exist in this

model and concluded that “Multivariate predictive control is ideally the solution for high quality
pt

control”. However, to be applicable without losing its benefits, good measurements, acceptable

regulatory control of local objectives (i.e. base level controls), reliable dynamic models,
ce

explicitly stated process constraints and new methods to promote robustness are needed.

Flotation process models have weaknesses in most of those aspects, one drawback being the lack
Ac

of the complete knowledge of necessary physical parameters for such models.

Phenomenological models can be broadly explained under 3 groups: population balance,

probabilistic and kinetic models (Polat and Chander, 2000).

7
 A. Population balance model

A population balance model (Bascar, 1982; Bascar and Herbst, 1982; and Bascar, 2000) is a

three phase model which represents each mineralogical species and particle size, with each

particle species being classed according to their state in the slurry i.e. free in pulp, attached to

t
ip
bubble in pulp, free in the froth, or attached to bubble in froth. Kinetic equations are used to

cr
relate the transfer of particles between slurry states. Hydrodynamic considerations such as power

dissipation into pulp, gas holdup in pulp, etc. are incorporated into this model. This makes it

us
Mineral Processing and Extractive Metallurgy Review

possible to simulate the effects of manipulated variables, such as air flow-rate, pulp level,

agitation, etc. on a flotation process.

an
Gas dispersion in mechanical flotation cells has traditionally been modelled through empirical
M
correlations using bulk dimensionless numbers. This approach gives little scope for the

prediction of bubble size as it involves a macroscopic overview of the process and does not take
ed

into account the basic sub-processes of gas dispersion occurring in the flotation cell. Hence,

population balance approach is a more effective procedure for the prediction of the bubble size in
pt

such systems (Lee et al., 1987; Prince and Blanch, 1990). The flotation cell is modelled as two
ce

separate zones, the impeller zone and the bulk zone. Bubble breakage occurs primarily in the

impeller zone, while bubble coalescence dominates in the bulk zone.

Ac

8
 B. Probabilistic models

The recovery rate of particles is related to the probability of success, P x , of a sequence of events

which must occur in a flotation cell before a particle can reach the concentrate launder

(Schuhmann, 1942) and for particles of average size x, it is given by:

t
ip
Px = Pc . Pa . F . [x ]. V (4)

cr
Where, P c , P a are the probabilities of particle-bubble collision and adhesion, F is the froth

us
Mineral Processing and Extractive Metallurgy Review

stability factor and V is the cell volume.

an
The Schumann method described only particle-bubble collision and adhesion mechanisms.

Tomlinson and Fleming (1965) added the term P e (Probability of levitation to the base of the
M
froth column without disengagement of particle and bubble) and P f (Probability of drainage from

the froth) and stated that:

ed

F=P e .P f (5)
pt

Probabilistic models are based on relative occurrence of various sub-processes such as collision,
ce

adhesion and detachment. These models can serve as a bridge to integrate micro and macro-scale

models. It assumes a rectangular distribution of rate constant and is described by:

Ac

  1  
Rt = R 1 −  {
 1 − e − K R t } (6)
  − K Rt  

Where, K R is rectangular rate of coefficient; R is (fraction) recovery after long flotation time t.

This model is commonly known as “Klimpel model”.

9
 C. Kinetic models

Kinetic Analysis:

The rate of change of an output parameter for a change in the input parameter, as a function of

t
time is called kinetics. Although the flotation process is complex, in terms of its various internal

ip
mechanisms, it is intrinsically a rate process, since in this process particles are deposited into the

cr
concentrate launder at a rate which is proportional to the concentration of those particles in the

us
pulp. Arbiter and Harris (1962) found that the flotation kinetics is the study of the variation in
Mineral Processing and Extractive Metallurgy Review

amount of froth overflow product with flotation time, and the quantitative identification of all

rate controlling variables. an

In analogy to kinetics, the kinetics of bubble particle collision and coalescence in the pulp may
M
be defined by the following ordinary differential equation:
ed

dC
= − K n C n Cb
m
(7)
dt
pt

Where, C and C b are the concentration of the particles and bubbles respectively; n and m are the
ce

respective orders; t is the flotation time; K is the flotation rate constant. Negative sign indicates

the decreasing concentration of mineral particles from pulp with increasing time.
Ac

If the air supply is constant, any tendency for the change in the bubble concentration is small.

Then, the rate of flotation is the rate at which the concentration of minerals in pulp decreases.

dC
i.e., = −K nC n (8)
dt

10
 M
From the kinetic point of view, concentration C can be defined as: C = . So if the volume is
V

assumed to not vary during the flotation, the problem becomes considerably simpler and the

value of the flotation rate constant may be obtained. In a review on flotation fundamentals,

t
Nguyen and Schulze (2004) have discussed the flotation kinetics and stated that the flotation

ip
kinetics is, in general, between first and second order; first order for the flotation of single

cr
minerals or dilute pulps and second order for low grade ores or more concentrated pulps.

us
Three approaches may be considered to define the flotation rate constant:
Mineral Processing and Extractive Metallurgy Review

(a) First order kinetics (n=1)

(b) Second order kinetics (n=2)

an
M
(c) Non-integral order kinetics
ed

(a) First order kinetics:

According to Polat and Chander (2000), kinetic modelling is based on the assumptions that the
pt

rate of the bubble-particle collision process is first-order, with respect to the number of particles
ce

and that the bubble concentration remains constant. In general, concentration of valuable

minerals through capture of particles by bubbles in flotation is considered to follow the first
Ac

order kinetics. When n=1, equation 2 is analogous to First order kinetic equation and is given by:

dC
= − K1C (9)
dt

Where, K 1 is first order rate constant

11
 Equation 3, when integrated for the first order flotation kinetics (Arbiter and Harris, 1962) leads

to the following expression:

R = R∞ (1 − e − K1t ) (10)

t
ip
Where, R is the cumulative recovery after time t and R ∞ is the maximum recovery after

prolonged flotation time and K 1 is the first order rate constant.

cr
Knowing recovery and time, the overall rate constant, K, could be determined from the following

us
Mineral Processing and Extractive Metallurgy Review

first order kinetics equation:

R = 1 − exp(− Kt ) an (11)

Where, t is the reference flotation time.

M
The extension of above equation for the three-component kinetics model is given by Nguyen and
ed

Schulze (2004), in which the feed is divided into fast, slow and non-floating components:

R = 1 − m f exp(− K f t ) − m s exp(− K s t ) − mtail

pt

(12)
ce

Where, m f , m s and m tail denote respectively the mass fractions of the fast, slow and non-floating

components in the feed.

Ac

Numerous batch tests (Imaizumi and Inoue, 1963; Harris and Chakravarti, 1970; Jameson et al.,

1977; Dowling et al., 1985; Rastogi and Aplan, 1985; Ek, 1992), semi batch tests (Harris and

Quadross-Paz, 1978) and continuous flotation tests (Jowett and Safvi, 1960) have been carried

out which support the use of first order rate equation or its slightly modified versions (Agar and

12
 Barrett, 1983; Xu and Stratton- Crawley, 1996; Yianatos and Hendriquez, 2006; Kelebek and

Nanthakumar, 2007).

Conducting both batch and continuous flotation experiments, Mika and Fuerstenau (1969) and

Bull (1966) observed first order kinetics for ideal mixing, examining a variety of industrial

t
ip
flotation cells of different types and sizes operated at different parameter levels (air rates,

cr
impeller speeds, and froth depths),which was corroborated by Gorain et al. (1999).

us
According to Polat and Chander (2000), the three parameter first order models fit well to
Mineral Processing and Extractive Metallurgy Review

flotation kinetics that incorporate a distribution of flotation rate constants. Hernainz and Calero

an
(2001) found that the removal of solids from the pulp phase is defined by a first order rate

equation. Thus, the basic first order model can be considered quite sufficient for the description
M
of the kinetics (Niemmi, 1995).

(b) Second order kinetics:

ed

dC
When n=2, equation 2 is analogous to second order kinetic equation and given by: = −K 2C 2
pt

dt

(13)
ce

When integrated for the second order flotation kinetics, the equation becomes (Arbiter and
Ac

Harris, 1962):

2
R∞ K 2 t
R= (14)
1 + R∞ K 2 t

Where, K 2 is the second order rate constant.

13
 In literature, there are very few studies that report second order flotation kinetics. Mori et al.

(1983) showed that, at initial stages, the process follows zero order kinetics and at the end, it

follows second order. The stage intermediate to the two extremes have intermediate order.

Recently, Yalcin and Kelebek (2011) found that recovery of both native gold and pyrite from

t
ip
flotation of a pyritic gold ore follows second order flotation kinetics.

cr
(c) Non-integral order kinetics:

us
Studies regarding non-integral order kinetics are rare. Hernainz and Calero (2001), in a
Mineral Processing and Extractive Metallurgy Review

comparative study on flotation kinetics of celestite and calcite, involving three models, i.e. first

an
order, second order and non-integral order, found that while the second order model represented

the data better than first order model, the non-integral order model showed much better fit to the
M
flotation data among the three models.

Kinetic models use empirically fitted parameters based on the observed performance. Kinetic
ed

models that invoke the chemical reactor analogy and consider flotation as a reaction between
pt

bubbles and particles, where the removal of solids from the pulp phase is defined by a first order

rate equation, have received most attention in the literature, since these models can be readily
ce

adopted for development of control strategies in industrial applications and the mathematics is

considerably easier.
Ac

There are currently two major approaches to flotation circuit design and optimisation,

both of which are based on well-established models to predict flotation performance (Mendez et.

al., 2009; Hu et. al., 2013). The first approach describes the flotation kinetics in terms of a first-

14
 order rate process (Mehrotra and Kapur, 1974; Dey et al., 1989; Guria et al., 2005) which is

widely used in industry, to predict the recovery of the floated species as a function of residence

time, and the experimentally determined flotation rate constant.

The second approach considers the recovery of the floated species in terms of both true flotation

t
ip
and entrainment (Franzidis and Manlapig, 1999; Gorain et al., 1998, 1999; Runge et al., 1998;

cr
Savassi, 1998). The recovery by true flotation resembles the first order rate model in the previous

approach, together with empirical froth phase relationships that account for the froth recovery

us
Mineral Processing and Extractive Metallurgy Review

and entrainment effects.

an
There are various kinetic models proposed for flotation process as listed in Table 1. Most of

these models are reduced to classical first order model. Solution of the first order rate equation
M
results in the classical first-order flotation model, as given by equation 5. This model has been

found to give best fit with high confidence levels. Among many flotation models, the classical
ed

first order flotation model is widely used and can be utilized to optimize and design the flotation

circuits (Xu, 1998; Agar et al., 1998; Oliveira et al., 2001; Ucurum, 2009).
pt

Imaizumi and Inoue (1965) working with quartz, chalcopyrite and pyrite systems also used the
ce

classical form of model. Gulsoy (1999) proposed a modified form of first order flotation kinetic

model, incorporating the gangue recovery into the froth (R g ) as:

Ac

R = R∞ (1 − e − Kt ) + R g (1 − R∞ ) (15)

Where, R is flotation recovery; R ∞ is the recovery at infinite time; t is flotation time and K is

flotation rate constant.

15
 Nguyen et al. (1998) developed a first order flotation kinetic model involving the probability of

collision (P c ), the probability of attachment (P at ) and the probability of detachment (P det ) for

single-bubble flotation experiment:

  V 

t
R = 1 − exp − Pc Pat (1 − Pdet )1 +  (16)

ip
  U 

cr
Where, R is recovery; V is particle settling velocity and U is bubble rising velocity.

us
Mineral Processing and Extractive Metallurgy Review

Model parameters cannot be compared among different first order models since they correspond

to a different property of the distribution function for each model. For example, the rate
an
parameter for the classical first order model represents some kind of mean flotation rate while

the rate parameter for the first order model with a rectangular distribution of k’s corresponds to
M
the flotation rate of the fastest floating component.
ed

A good compromise to characterize industrial flotation equipment has been described by Finch

and Dobby (1990), where the collection zone and froth recoveries are considered independently
pt

while estimating the overall flotation performance:

ce

Rc R f
R= (17)
1 − Rc (1 − R f )
Ac

Multi-phase modelling:

Previously, the flotation process was based on a single phase model, considering the flotation

machine as a well-mixed cell where the froth zone was neglected. Subsequently, multi-phase

16
 models, considering first order flotation rates, were introduced in order to account for the

presence of two distinct zones: pulp zone, and froth zone, assuming that the pulp and froth

phases are each ideally mixed and the material transport occurs between two phases in one or

both directions.

t
ip
Woodburn (1970), Lynch et al. (1981), Vera et al. (1999), Alexander et al. (2003) and Seaman et

cr
al. (2004) also considered a two phase model (pulp and froth) to characterize the flotation

process, but still keeping the idea of using a single overall flotation rate constant K to account for

us
Mineral Processing and Extractive Metallurgy Review

the overall cell performance. Emmanouli et al. (2006) presented the basis for the three phase

modelling of an industrial scale DAF tank, using flotation kinetic concepts.

an
Despite the efforts to separate both phases, typically a single overall flotation rate constant is
M
being used (Flint, 1974; Harris, 1978; Harris et al., 1975; Laplante et al., 1983; Deglon et al.,

1999; Deglon, 2003; Savassi, 2005). This is not trivial as K is a complex parameter and
ed

dependent on many factors governing the flotation process such as particle size, degree of

liberation, air flow rate, residence time, reagent concentration, design of flotation cells etc. To
pt

account for this, various continuous distribution functions of K have been introduced (Polat and
ce

Chander, 2000).

According to Imauzimi and Inoue (1963), if the material to be floated consists of particles whose
Ac

rate constants could be expressed by a continuous distribution function F (k), the recovery R at

time t will be equal to:

17
 ∞
R(t ) = R∞ [1 − ∫ F (k ) exp(−kt )dk (18)
0

Where, R ∞ is the ultimate recovery at infinite time

t
ip
According to Polat and Chander (2000), the mineral recovery (R) at time t in a batch flotation

process can be described by the following equation:

cr
∞
R
( )
= ∫ 1 − e − kt F (k )dk

us
(19)
Mineral Processing and Extractive Metallurgy Review

R∞ 0

an
Where, R ∞ is the maximum flotation recovery at infinite time, and k is the kinetic rate constant.

The term (1 − e − kt ) represents the mineral recovery of a first order process with a variable rate
M
constant k, as function of time. F (k) is the rate constant distribution function for mineral species

with different flotation rates.

ed

In a continuous flotation process, the mineral recovery (R) can be described by the following
pt

general equation (Yianatos and Henriquez, 2006; Yianatos, 2007):

ce

∞ ∞

∫ (1 − e ) F (k ) E (t ) dk dt
R
R∞ ∫0
− Kt
= (20)
0
Ac

Where, F (k) is the rate constant distribution. E (t) is the residence time distribution function for

continuous processes with different mixing characteristics. The mineral recovery depends on the

flotation rate distribution of the floatable minerals, the mixing region in the collection zone and

the actual mean residence time.

18
 Several floatability distribution functions for F(k) such as Gamma, Rectangular, Triangular as

well as sinusoidal distributions have been suggested by Imauzimi and Inoue (1963), Harris and

Chakravarti (1970), Huber-panu et al. (1976), Klimpel (1980), Diao et al. (1992) and Polat and

Chnader (2000). Ferreica and Loveday (2000) considered a double normal distribution to

t
ip
describe the F (k) function. Savassi (2005) developed a new modelling technique in batch

flotation using the Weibull (Rosin-Rammler) distribution to represent the floatability

cr
distribution.

us
Mineral Processing and Extractive Metallurgy Review

The Gamma distribution having only two parameters (a, b) which is capable of representing

different types of distributions; from exponential decay up to normal distribution is given by

Woodburn and Loveday (1965):

an
M
b a +1
F (k ) = . k a . e −b.k (21)
Γ(a + 1)
ed

Rectangular flotation distribution is given as

pt

F (k ) =
1
, 0 ≤ k ≤ k max
k max (22)
F (k ) = 0, k < 0, k > k max
ce

Where, k max corresponds to the maximum flotation rate constant.

Ac

Weibull (Rosin-Rammler) distribution of the rate constant is given by:

n
n −1  k 
n  k  − 
F (k ) =

.  .e  km 
k m  k m
(23)


19
 Where, n is the shape factor and k m is the scale factor of distribution.

Mineral recovery in batch cells given by Yianatos et al. (2010) using Rectangular distribution is:

R  1 − e − k max . t 
= 1 −  (24)

t
R∞  k max . t 

ip
Flotation recovery of a continuous process using Gamma distribution is given by (Yianatos et al.

cr
2010):

us
Mineral Processing and Extractive Metallurgy Review

 ∞  b 
a +1

R = R∞ . 1 − ∫ E (t ).  dt  (25)
 0 b+t  
an
And mineral recovery for batch tests using gamma distribution is given by:
M
a +1
R  b 
= 1−   (26)
b+t 
ed

R∞

According to Yianatos et al (2010), the flotation rate distribution follows a normal distribution in
pt

industrial as well as laboratory batch tests. They showed that a single parameter model, such as
ce

the rectangular model can be as good as the Gamma or Weibull distribution in terms of model fit.

2.1.1.2. Empirical Models

Ac

Understanding the various micro processes involved in the flotation cell, namely collision,

attachment and detachment, is a fundamental step towards the prediction of the rate constant of

flotation kinetics (Schulze, 1984; Dobby and Finch, 1987; Crawford and Ralston, 1988; Hewitt

20
 et al., 1995). Despite recent advances, these micro processes are still not fully understood from a

quantitative viewpoint. As a result, mineral and coal processing engineers have placed emphasis

on the empirical models of flotation kinetics. This empirical method is valuable in many cases,

but lacks in study of essential insights of the bubble particle collection phenomenon. The results

t
ip
of these empirical investigations of flotation kinetics change from experiment to experiment:

first-order, second-order and fractional-order kinetics. It is clear that the change of mineral mass

cr
in a given volume of a flotation cell per unit time is affected by two phenomena; transport

us
processes and true flotation. Transport processes include macro turbulence, convection of liquid
Mineral Processing and Extractive Metallurgy Review

or gas phase and settling of solid phase whereas true flotation includes collision, attachment and

detachment. an
M
Empirical models are specific to a particular ore of universal type. The amount of material

floated is related to input and output variables through suitable mathematical equations.
ed

Statistical methods are used to relate dependent and independent variables and used for

predictive control to estimate the curve fitting parameters. The statistical correlations can be
pt

obtained by normal multiple linear regression methods or spline regression methods (Mular,

1972, Whiten, 1971). The parameters obtained by such an analysis usually do not have any
ce

physical significance. Continuous analysis of plant data and corrective adjustment of the

predictive controller make it adaptive to changing conditions, as it is especially important within

Ac

the context of flotation control, which is prone to non-linear, complex behaviour.

Both online and offline methods are necessary for collection of data for empirical models. In

offline methods, the data is either normal shift or daily operating data, or data from a special

21
 sampling work in which experimental designs are used. In other case, data can be obtained by

laboratory determination of assays and pulp solid contents with some help from online

instrumentation. In the online method, online instrumentation can be used exclusively through

offline analysis of the data. Faulkner (1966) described the initial collection and analysis of data

t
ip
by offline methods.

cr
Empirical models use batch test data and/or plant data to relate flotation rate constant values to

each of the components in the system, i.e., the number of components for each mineral of

us
Mineral Processing and Extractive Metallurgy Review

interests will be system-specific but Runge (2009) indicated that one or two floating and one

non-floating components would generally be sufficient. Assuming that the material has fast and
an
slow floating components, the model is described by:
M
{
Rt = (1 − Fs ) 1 − e
−k f t
}+ F {1 − e }
s
kst
(27)
ed

Where, F s is the fraction of slow floating components; K f , K s are the rate coefficient of fast

floating and slow floating components respectively; R t is the fraction of recovery at time t.
pt

The advantage of these models is that they do not require size by liberation data to classify
ce

particles; whereas the disadvantage is that the grouping of particles is not related to any

measurable physical property and hence cannot be connected to other mineral processing
Ac

operations such as mills or cyclones (Vianna, 2004). Hence their use is limited in certain

circumstances where there is change in ore properties, size distribution or percentage solids of

the feed to a circuit.

22
 Bazin et al. (1994) and Mclvor and Finch (1991) gave examples of empirical models which can

incorporate the change in particle size but both these models are ore-specific.

According to Runge (2009), the following two equations can be used to determine the rate

constants of each component for either a batch test or an industrial cell (where perfect mixing is

t
ip
assumed) if recovery is known.

cr
For batch flotation test data, the recovery is described as:

us
( )
Mineral Processing and Extractive Metallurgy Review

R (t ) = ∑i −1 mi 1 − e ki t
n
(28)

an
Where, R (t) is the overall recovery at time t which is the sum of all the individual component

recoveries; m i is the proportion of a mineral in each component and k i is the rate constant of
M
each component.
ed

Recovery relating plant data is given by:

 kτ 
R(t ) = ∑i −1 mi  i 
n
pt

(29)
 1 + k iτ 
ce

Where, 𝜏 = the residence time for all the cells or bank of cells.
Ac

Recently, Newcombe (2014) established a mathematical relationship to describe the measured

behaviour of the valuable particles within the settling zone of an operating industrial flash

flotation cell linking the first order kinetics rate constant K (for pyrite) for each size class i, to the

residence time within the settling zone, 𝜏 s :

R

23
 Ki = α i e −β τ i s
(30)

Where, 𝛼 𝑎𝑛𝑑 𝛽 are empirical fitted parameters for each size class i.

For empirical models, the froth recovery has been the most important factor to be taken into

t
ip
account as a practical parameter to describe the influence of froth in flotation (Alves dos Santos.

et al., 2014). The model of Gorain et al. (1998) showed that froth recovery was modelled by

cr
exponential decay of the air retention time in froth 𝜏 f :R

us
Mineral Processing and Extractive Metallurgy Review

R f = exp(− β .τ f ) (31)

an
Where, 𝛽 is a parameter which depends on physical and chemical properties of the pulp.
M
Thus, it may be concluded that though theoretical models involve complex equations of

unknown parameters, these models provide insight into the possible events of the flotation
ed

process. Among all the theoretical models, kinetic models are simple, useful, and considerably

easier and found to give the best fit. Kinetic models are found to prevail in almost all flotation
pt

conditions regardless of the ore type and ore characteristics as well as flotation cell

configurations.
ce

2.1.2. Deterministic Models

Ac

The rate of recovery of minerals in flotation relies on the contribution of three factors: (1) the

engineering contributions, i.e. the design of the flotation cell, (2) the hydrodynamic conditions

inside the flotation cell, e.g. bubble velocity, density, turbulence and (3) the process of particle–

bubble interaction. To recover minerals, particles need to first, collide with a bubble (collision

24
 efficiency, E c ), secondly, attach to the bubble (attachment efficiency, E a ) and finally, form a

stable particle-bubble aggregate capable of rising to the top of the flotation cell (stability

efficiency, E s ). These sub-processes, in turn, depend on both the hydrodynamic conditions (E c

and E s ) and the surface chemistry of the mineral particles (E a and E s ) (Duan et al., 2003).

t
ip
So the study of deterministic models can be carried out on the basis of the following parameters:

cr
A. Surface forces and Hydrodynamics:

us
Mineral Processing and Extractive Metallurgy Review

Derjaguin and Duhkin (1961) were probably the first to consider surface forces in developing a

model for bubble–particle interactions. These surface forces included London–van der Waals,
an
ion-electrostatic, hydrophilic and structural forces. The hydrophilic force seems to become zero

when the bubble–particle adhesion is to occur.

M
The DLVO theory takes into account some of the surface forces occurring in flotation. This
ed

theory combines the van-der Waals force and the electrostatic force into a total surface force

lacking hydrophobic force parameter which is a major contributor to surface forces between
pt

particles and bubbles in a water medium (Yoon and Mao, 1996 and Yoon, 2000). The non-
ce

DLVO theory includes forces such as hydration (Yotsumoto and Yoon, 1993) and hydrophobic

forces (Xu and Yoon, 1990) which play an important role in coagulation. The hydrophobic force
Ac

is the major driving force for bubble–particle adhesion (Yoon and Yordan, 1991; Schimmoller et

al, 1992; Schimmoller, 1993).

Modelling flotation without considering contributions from the hydrophobic force is unrealistic.

In this sense, the surface forces measurement has been used to model pulp flotation rates,

25
 including particle–bubble attachment and detachment, in terms of surface chemistry parameters

(contact angle, zeta potential, Hamaker constant and surface tension), by relating the energy

barriers to the kinetic energies of the bubble–particle interaction (Yoon, 2000).

Yoon and Mao (1996) proposed a first order flotation model accounting for both the

t
ip
hydrodynamic and the surface chemistry parameters, using the application of extended DLVO

cr
theory:

us
 E1    γ lvπR12 (1 − cos θ )2 + E1 
2
1  3 4 Re 0.72  R1 
Mineral Processing and Extractive Metallurgy Review

k = Sb  +   exp −  × 1 − exp − 
15  R2    (32)
4 2  E k     E ' k 

an
Where, S b is the superficial surface area rate of bubbles; R e is the Reynolds number of the

bubble; R 1 is the particle size; R 2 is the bubble size; E 1 is the energy barrier; E k is the kinetic
M
energies of the macroscopic particles; γ lv is the surface free energy at liquid/vapour interface; 𝜃

is the contact angle; E k ’ is the kinetic energy that tears the particle off the bubble surface. Here,
ed

the surface forces include electrostatic, dispersion and hydrophobic forces.

pt

A flotation model derived from first principle and dealing with all dominant surface forces found
ce

in flotation is proposed by Mao and Yoon (1997):

 E   W + E1 
2
1  3 4 Re 0.72  R1 
Ac

k = Sb  +   exp − 1  1 − exp − a 

15  R2
(33)
4 2   E k   E ' k 

These surface forces are the electrostatic, van-der Waals, and hydrophobic forces and are

modelled based upon the extended DLVO theory. But the problem with the above models is that

26
 these are derived in absence of turbulent effects. Turbulence is process-determining in flotation.

While macro-turbulence controls suspension, micro-turbulence controls air dispersion, the rate of

particle-bubble collisions etc. Hence the hydrodynamics optimization of flotation processes (cell

geometry, power input, air flow rate etc.) is necessary. But on the other hand, it provides a

t
ip
valuable relationship between the hydrodynamics of the system and the surface forces.

cr
A flotation model is given by Duan et al. (2003) working with chalcopyrite ore based upon

turbulent characteristics of the flow as well as hydrodynamic forces:

us
Mineral Processing and Extractive Metallurgy Review

 4 7
2

G fr  0.33ε d b 9
9  ∆ρ b  3 1 
  .
k= − 2.39
d b .Vcell 
 υ3
1 an  ρ  υ  c a s
 fl  b

E .E .E (34)
M
This model also included the contributions from the efficiencies of collision, attachment and

stability between particles and bubbles as well as their frequencies of collision. But the model
ed

does not account for the effect of surface forces.

pt

The above equation is similar to the expression for flotation rate constant of a batch-wise

flotation test given by Jameson et al., (1977) and Ralston, (1992) in the absence of turbulent
ce

environment:
Ac

3 G fr . h
K= Ec . E a (35)
2 d bVcell

Where, h is the height of flotation cell and E s is considered as unity.

27
 A semi-empirical flotation model is developed by Sherrel (2004) taking into account probability

of attachment and detachment which incorporates hydrodynamics as well as surface chemistry

parameters in a turbulent environment:

K = β N b PA (1 − PD )RF (36)

t
ip
Where, 𝛽 is the collision kernel (m3.s-1); N b is the number density of bubbles (number per unit

cr
volume, m-3); P A is the probability of attachment; P D is the probability of detachment; R F is the

us
froth recovery factor.
Mineral Processing and Extractive Metallurgy Review

Kostoglou et al. (2006) presented a critical review on the state of the art of flotation fundamental
an
modelling, and also introduced a new generalized particle–bubble aggregation model, developed
M
for a combined gravitational and turbulent flow field.

The particle collision and detachment frequencies in flotation have been related to the turbulent
ed

energy density (Bloom and Heindel, 2002). Also, the number of bubble–particle collisions has

been predicted by CFD simulation of flotation cells (Koh et al., 2000; Koh and Schwarz, 2003;
pt

Koh and Schwarz, 2007).

ce

Most recently, Karimi et al. (2014), working on quartz particles, have given the

expression for flotation rate constant based on CFD model which compares both experimental
Ac

data and the analytical computations reported by Pyke (2004) and Newell (2006):

 4 7
2

7.5 G fr  0.33ε 9 d b 9  ρp − ρf  1
3
K=   . E .E .E (37)
π d bVr  1  ρ  u  c a s
 υ3  f  i


28
 Where G fr is the gas flow rate (cm3/min); d b is the bubble diameter (cm); V r is the reference

volume (cm3); 𝜀 is the turbulent dissipation rate (cm2/s2); 𝑣 is the fluid kinematic viscosity

(cm2/s); ρ p and ρ f are density of particle and fluid respectively; u i is the turbulent fluid velocity

(cm/s).

t
ip
The numerical prediction was validated and then compared with the analytical computations and

cr
experimental measurements of Newell. The analysis showed that the flotation rate constant

increased with particle size of quartz for a range of particle size and this trend is observed in both

us
Mineral Processing and Extractive Metallurgy Review

CFD results and theoretical results. The CFD based model not only captured the trend of

an
experiments for a range of particle sizes but also yielded improvements in the prediction of rate

constant compared to the theoretical calculation of Pyke and Newell.

M
This new CFD-kinetic model can improve the prediction of flotation rate constant and also this

method can be applied as a reliable and cost-effective tool for the flotation modelling. But the
ed

drawback is this model does not account for entrainment phenomenon.B. Entrainment factor:
pt

Entrainment of gangue material into the froth concentrate is an important factor to be considered

in the performance of the flotation cells as it dilutes the desired material and reduces the grade of
ce

the product. The proportionality between the fractional gangue recovery and the fractional water

recovery is often referred to as the entrainment factor, or degree of entrainment, and symbolised
Ac

as Ent (e.g. Subrahmanyam and Forssberg, 1988 and Zheng et al., 2006). The entrainment factor

is defined as the recovery of the entrained species divided by the recovery of water (e.g.

Subrahmanyam and Forssberg, 1988 and Kirjavainen, 1996). According to Alves dos Santos N.

29
 et al. (2014), entrainment can be mathematically defined as the factor that relates free gangue

recovery, R g to water recovery, R w and given by:

R g =ENT.R w (38)

t
Expression for overall recovery due to true flotation is described by Dobby (1984) and

ip
Falutsu and Dobby (1989) which does not include recovery due to entrainment:

cr
Rc R f
R=

us
(39)
Rc R f .(1 − Rc )
Mineral Processing and Extractive Metallurgy Review

an
Savassi (1998) has given expression for recovery incorporating process variables such as water

recovery, R w ; residence time, 𝜏; and entrainment factor in the froth, ENT:

M
PS bτR f (1 − Rw ) + ENT Rw
R=
(1 + PS τR ) + ENT R
b f w
(40)
ed

The compartment model by Savassi (2005) which account for the simultaneous contribution of
pt

true flotation and entrainment to the overall recovery in a conventional flotation cell is given by:

K cz .τ cz . R f . (1 − Rw ) + ENT . Rw
ce

R=
(1 + K cz .τ cz . R f ).(1 − Rw ) + ENT . Rw
(41)
Ac

Where, K cz is the flotation rate constant exclusive for the collection zone; 𝜏 f is the residence time
R

in collection zone; R f is the froth recovery of attached particles; R w is the water recovery from

the feed to the concentrate streams.

30
 A model for froth recovery, R f including the effect of mechanical entrainment is given

by Vera et al. (2002):

R f = (1 − ENT . Rwf )exp(− β .τ f ) + ENT . Rwf (42)

t
Where, R wf is the water recovery in froth; ENT is the mechanical entrainment; 𝛽 is the parameter

ip
to be adjusted; and 𝜏 f is the air retention time in froth.

cr
R

Recently a comprehensive model for froth flotation is developed by Alves dos Santos et

us
Mineral Processing and Extractive Metallurgy Review

al. (2014) linking the compartment model to a set of phenomenological models describing the

froth recovery, the water recovery and the entrainment factor.

C. Residence time distribution:

an
M
Mixing of various phases inside flotation chamber is one of the important parameters as
ed

it affects the particle-bubble collection and detachment process; a high degree of mixing is

necessary to bring solid particles in contact with the bubbles whereas it is deleterious if it
pt

happens in an unsuitable zone. Since the geometrical, operational, and physico-chemical

parameters have a crucial role in the mixing process, their study is desirable for designing and
ce

optimizing the flotation cell. Residence time distributions are measured to establish the gas,

liquid and solid-particle flow patterns in flotation cells. Their results are used to identify regions
Ac

of maldistribution within them.

There are essentially two models which analyze the residence time distribution results:

(a) the axial dispersion model and (b) the mixed zone model (Mavros P., 1992).

31
 (1) The axial dispersion model:

If there is very little mixing, the axial dispersion model is considered. This model states that

mixing is due to departures from ideal plug flow, caused by disturbances in the flow front. The

model can be used for column flotation and not for the mechanically agitated cells. An approach

t
ip
to suppress axial mixing is the introduction of horizontal baffles, made from perforated plates

cr
(Kawatra and Eisele, 2001). Axial dispersion models use parameters such as Peclet number and

vessel dispersion number to define residence time distribution profiles in flotation columns

us
Mineral Processing and Extractive Metallurgy Review

(Mavros, 1993a).

an
An equation representing axial dispersion model for the concentration C of some species

is given by:
M
∂C ∂ 2C ∂C
= D 2 −u (43)
∂t ∂x ∂x
ed

Where, D is axial dispersion coefficient; x is the distance along which the concentration varies.
pt

The above equation can be rewritten in dimensionless form as follow:

ce

∂C ∂ 2 C ∂C
= Nd − (44)
∂θ ∂z 2 ∂Z
Ac

Where, N d is the dimensionless dispersion number and N d = 𝐷�𝑢𝐿 (Levenspiel, 1972), where, L

is the vessel length; z is the dimensionless linear dimension; 𝜃 = 𝑡�𝜏 , where 𝜏 is the mean

residence time in the vessel and 𝜏 = V�Q , where V is cell volume and Q is volumetric flow rate

32
 entering the cell. This model, with dispersion number as its sole parameter is more appropriate to

use but the disadvantage is the lack of a precise method for the determination of the dispersion

number. A perusal of literatures available regarding the use of axial dispersion models indicates

that the axial dispersion models despite the limitations have been applied successfully to explain

t
ip
the phenomenon in column flotation. (Finch and Dobby, 1990; Mills and O’Connor, 1992;

Alford, 1992; Mankosa et al., 1992; Mavros, 1993a, 1993b; Tuteja et al., 1994; Tokoro and

cr
Okano, 2001; Jena et al., 2009; Quinn and Finch, 2012;). (2) The mixed zone model:

us
Mineral Processing and Extractive Metallurgy Review

This model is considered when the cell is either a single perfectly mixed region or when

it behaves like a series of perfectly mixed zones connected in series, like the bank of
an
mechanically agitated cell. The theoretical residence time distribution profile for any number of
M
cells (N) in series is given by the following equation (Levenspiel, 1972):

N ( Nθ ) N −1
E (θ ) = exp(− Nθ ) (45)
ed

( N − 1)!

Once the mean residence time 𝜏 and variance 𝜎 2 of an experimental residence time distribution
pt

are determined, the value of N may be calculated by following:

ce

τ2
N= (46)
σ2
Ac

These models are relatively simple as the number of parameters is usually small, yet these

are able to represent a variety of residence time distribution corresponding to widely different

mixing conditions. This model is capable of dealing with all kinds of flow patterns through the

33
 construction of models incorporating plug flow as well as mixed regions. Another advantage of

this model is that in case of complex ore systems, involving fast- and slow-floating minerals as

well as gangue, the determination of the type and number of zones may be used for examination

of possible operation improvement. Theoretical residence time distribution profiles for various

t
ip
mixed zone models are available in literature (Joshi and Sharma, 1979; Mehrotra and Saxena,

1983; Frew, 1984; Dobby and Finch, 1985; Frew and Davey, 1985; Wilson and Frew, 1986;

cr
Mavros et al. 1989;Yianatos et al., 1993).

us
Mineral Processing and Extractive Metallurgy Review

D. Floatability:

an
A different approach to model the flotation process considers the specific floatability of particles

as a function of particle size, degree of liberation and chemical adsorption of reagents (Niemi,
M
1995). Current flotation models classify particles on the basis of floatability. Floatability is a

property of the feed stream and a function of both particle and stream parameters (Welsby et al.,
ed

2010; Runge, 2009; Vianna, 2004). The floatability response of a flotation system cannot be

measured directly and must be calculated from the observed response of an ore. According to
pt

Welsby et al. (2010), physical properties such as size and liberation are directly related to the
ce

flotation response. Runge (2009) stated that floatability is not dependant on operating conditions

within the cell, while Vianna (2004) related floatability to cell hydrodynamic conditions such as
Ac

gas dispersion, agitation and turbulence. The means of classifying particles on the basis of

floatability relies on the fact that all particles within a particular class/component float at a

similar rate. The majority of models are based on the assumption that once a particle is assigned

34
 a floatability value, this value does not change throughout the process within a flotation circuit

(Runge, 2009).

Based on extensive research, relationships between ore floatability and other parameters of the

flotation process have been developed; of these the most significant relationships in predicting

t
ip
performance is given by Gorain et al. (1997) and Savassi (1998). Gorain et al. (1997) found a

cr
linear relationship between the flotation rate (K), the bubble surface area flux (S b ), the recovery

in the froth (R f ) and the ore floatability (P):

us
Mineral Processing and Extractive Metallurgy Review

K = Sb R f P (47)

an
Taking particle size into account, Nguyen et al. (1998) developed a model for flotation
M
recovery, R, of particle diameter D p collected by a bubble of diameter D b :

4n
R=
Hπ (D p + Db ) C p
(48)
ed

Where, n is the number of particles of diameter D p collected per bubble; C p is the concentration
pt

of particles of diameter D p in flotation pulp; H is the height of the pulp volume.

ce

A property based model for recovery by bubble-particle collection has been developed by
Ac

Ralston et al. (2007) which allows the relative effects of particle size, particle contact angle and

cell hydrodynamics:

R pb (t ) =
Kt
(49)
1 + K coll t

35
 Where, R pb (t) is the recovery of the valuable mineral phase by bubble capture, as a function of

pulp residence time (t) in the collection zone and K is the collection zone rate constant.

This model can account for changes in feed properties such as particle size and mineral

hydrophobicity. Here, recovery of each mineral phase across the cell (R cell ) can be calculated

t
ip
using the following equation:

cr
Rcell = Rattached
(100 − Rent ) + R (50)
ent
100

us
Mineral Processing and Extractive Metallurgy Review

Where, R attached is the recovery of mineral phase actually reporting to the concentrate launders;

an
R f is the recovery of mineral phases across the froth; and R ent is the recovery of mineral phases

by entrainment.
M
3. Summary
ed

The modelling and simulation of flotation process has received particular attention from the
pt

mineral processing community for almost three decades. An overview of the literature

pertaining to this field shows that much has been achieved, but also that some technologies still
ce

remain to be transferred to industry. Despite the great advances observed in terms of process
Ac

knowledge, the mechanisms and principles as well as the design and scale-up of industrial

flotation cells, are still not fully understood and remains relatively inconclusive and not

comprehensive. New approaches, extension of current models and revived works will probably

guide future investigative efforts.

36
 Since the study of all the sub-processes and the corresponding cause and effect relationships is

not possible in case of micro-scale models, emphasis is given on the study of macro-scale models

as it deals with the overall response of the material to the flotation process.

Though theoretical models are useful in providing insight into the possible events in the flotation

t
ip
process, they involve complex equations of unknown parameters which are difficult to

cr
determine.

us
There are several experimental models. Some experimental models include capillary and
Mineral Processing and Extractive Metallurgy Review

hydrostatic forces, the forces of gravitation, buoyancy and machine acceleration and capillary

an
pressure ignoring surface forces while some models include all dominant surface forces based

upon extended DLVO theory and ignore surface chemistry parameters. Some models consider
M
turbulent characteristics of the flow as well as hydrodynamic forces while some consider

quiescent environment, whereas other models account for thermodynamics. Some models
ed

incorporate both hydrodynamics as well as surface chemistry parameters in a turbulent

environment but lack in entrainment factor and residence time distributions whereas some
pt

models include only residence time distributions, lacking other important parameters such as
ce

surface forces, entrainment phenomenon etc.

Empirical models are ore-specific. They use batch test data and/or industrial data to relate
Ac

flotation rate constant values to each of the components in the system. Their use is limited to

certain circumstances where there is change in ore properties, size distribution or percentage

solids of the feed to a circuit. The results of the empirical investigations of flotation kinetics

37
 change from one experiment to another. This empirical method is valuable in many cases, but

lacks in study of essential insights of the bubble particle collection phenomenon.

4. Conclusion:

t
ip
v From literature, it is clear that first order kinetic models hold good for both batch and

cr
continuous flotation processes whereas second order and non-integral order kinetic models are

rare.

us
Mineral Processing and Extractive Metallurgy Review

v Kinetic models are based on carrying out experiments taking certain known inputs into

an
consideration from which recovery can be predicted.

v If possible, the controller, either human or computer, may be able to improve the
M
recovery by modifying those inputs. These models can also find the maximum recovery within
ed

certain input ranges.

v Kinetic models have the ability to correlate data as well as give insight into the process.
pt

These models are more useful and can be applied to operating plants, with satisfactory outcomes.
ce

v Kinetic models allow more effective benchmarking and optimisation of flotation circuits

at both laboratory and plant scale.

Ac

v From all the kinetic models listed in Table 1, the classical first order kinetic model is the

best model and can be utilized to optimize the flotation process as it suits to both batch and

continuous flotation processes with high confidence levels.

38
 References

Alexander, D.J., Franzidis, J.P. and Manlapig, E.V., 2003. Froth recovery measurement in plant

scale flotation cells, Miner Eng, 16: 1197–1203.

t
Alford, R.A., 1992. Modelling of single flotation column stages and column circuits. Int. J. of

ip
Miner. Process, 36 (1992), pp. 155-174

cr
Alves dos Santos, N., Savassi, O. N., Antonio, E., Clark P. and Afonso, H. M., 2014. Modelling

us
flotation with a flexible approach-Integrating different model to the compartment model. Miner.
Mineral Processing and Extractive Metallurgy Review

Eng. (2014) Vol. 66–68, p. 68–76

an
Agar, G.E., and Barrett, J.J., 1983. The use of flotation rate data to evaluate reagents. CIM Bull.

76 (851), 157–166.
M
Agar, G.E., Chia, J. and Requis, L., 1998. Flotation rate measurements to optimize an operating

circuits, Minerals Engineering 11 (4) (1998) 347–360.

ed

Arbiter, N., and Harris, C.C.. 1962. In: D.W. Fuerstenau (Ed.), Froth Flotation, AIME, Ann
pt

Arbor, MI, 1962, p. 215 Fiftieth anniversary volume.

ce

Bascar, O.A., 1982. Modelling and computer control of a flotation cell. PhD thesis, University of

Utah, Salt Lake City, Utah.

Ac

Bascar, O.A., 2000. An interactive dynamic flotation model framework. Proceedings of XIV

International Mineral Processing Congress, Rome, Italy, p. 21–31.

Bascar, O.A. and Herbst, J.A., 1982. Dynamic modelling of a flotation cell with a view toward

automatic control. XIV IMPC, CIM, Toronto, Canada.

39
 Bazin, C., Grant R., Cooper, M. and Tessier, R., 1994. Prediction of metallurgical performance

as a function of fineness of grind. In: Proceedings 26th Canadian Mineral Processors, Ottawa.

Bergh, L.G. and Yianatos, J.B., 2011. The long way toward multivariate predictive control of

flotation processes. Journal of Process Control 21, 226–234.

t
ip
Bloom, F. and Heindel, T.J., 2002. On the structure of collision and detachment frequencies in

cr
flotation models, Chem Eng Sci, 57: 2467–2473.

Bull, W.R., 1966. The rates of flotation of mineral particles in sulphide ores, Proc. Austr. Inst.

us
Mineral Processing and Extractive Metallurgy Review

Min. Metall., 254, 1966, pp. 69–78.

an
Chander, S. and Polat, M., 1995. Coal flotation kinetics: Interactions between physical and

chemical variables. In: Plenary Lecture, Proc. Int. Conf. on Mineral Proc.-Recent Advances and
M
Future Trends, Kanpur-India, pp. 615–631.

Crawford, R.and Ralston J., 1988. The influence of particle size and contact angle in mineral
ed

flotation .Int. J. Miner. Process., 23 (1988), pp. 1–24

pt

Deglon, D.A., 2003. A novel attachment-detachment kinetic model, in Proc of the Strategic

Conference, Flotation and Flocculation: From Fundamentals to Applications, Ralston, J., Miller,
ce

J. and Rubio, J. (eds). 109–116, (Snap Printing, Australia).

Ac

Deglon, D.A., Sawyerr, F. and O’Connor, C., 1999. A model to relate the flotation rate constant

and the bubble surface area flux in mechanical flotation cells, Miner. Eng, 12(6):599–608.

Diao, J., Fuerstenau, D.W. and Hanson, J.S., 1992. Kinetics of Coal Flotation. SME-AIME

Annual Meeting, Phoenix, AZ.

40
 Derjaguin, B. V., and Duhkin, 1961. S. S., Theory of flotation of small and medium size

particles. Trans. Inst. Min. Metall. 70, 221–246.

Dey, A., Kapur, P.C. and Mehrotra, S.P., 1989. A search strategy for optimization of flotation

circuits. International Journal of Mineral Processing 26, 73–93.

t
ip
Dobby G.S. and Finch J.A.; 1985. Mixing Characteristics of industrial flotation columns.

cr
Chemical Engineering Science. 40(1985), pp.1061-1068

Dobby G.S. and Finch J.A, 1987. Particle size dependence in flotation derived from a

us
Mineral Processing and Extractive Metallurgy Review

fundamental model of the capture process, Int. J. Miner. Process., 27 (1987), pp. 241–263.

an
Dowling, E.C., Klimpel, R.R. Aplan, F.F., 1985. Model discrimination in the flotation of a

porphyry copper ore. Miner. Metall. May, pp. 87-1-01.

M
Duan J., Fornasiero D. and Ralston J., 2003. Calculation of the flotation rate constant of

chalcopyrite particles in an ore, Int. J. Miner. Process. 72 (2003) 227– 237

ed

Ek, C., 1992. Flotation kinetics, Mavros, P. and Matis, K.A. (eds). In Innovations in Flotation
pt

Technology, Vol. 208, 183–209 (Nato ASI Series, Kluwer Academic Publishers, Dordrecht, The

Netherlands).
ce

Emmanouil, V., Skaperdas, E., Karapantsios, T.D. and Matis, K.A., 2006. Towards three-phase
Ac

modeling of a dissolved air flotation tank, in Proceedings of the XXIII International Mineral

Processing Congress, Istanbul, Turkey, 3–8 September.

Falutsu, M. and Dobby, G.S., 1989. Direct measurement of froth drop back and collection zone

recovery in a laboratory flotation column. Minerals Engineering 2 (3), 377–386.

41
 Faulkner, B. P., 1966. Computer control improves metallurgy at Tennessee Copper’s flotation

plant, Mining engineering, 18, pp. 53-57.

Ferreira, J.P. and Loveday, B.K., 2000. An improved model for simulation of flotation circuits.

Minerals Engineering 13 (14/15), 1441–1453.

t
ip
Finch, J.A. and Dobby, G.S., 1990. Column Flotation, 1st edition, (Pergamon Press, London,

cr
UK).

Finch, J.A., Uribe-Salas, A. and Xu, M., 1995. Column flotation, in Matis, K.A. and Marcel, D.

us
Mineral Processing and Extractive Metallurgy Review

(eds). Flotation Sciences and Engineering, 291–330 (New York, USA).

an
Flint, L.R., 1974. A mechanistic approach to flotation kinetics, Trans Inst Min Metall, 83,90–95.

Franzidis, J.and Manlapig, E., 1999. A new, comprehensive, and useful model for flotation. In:
M
Proceedings of a Symposium held at the Annual SME Meeting, pp. 413–423.Frew J.A., 1984.

Back mixing in industrial flotation banks. 1. Effect upon the determination of flotation rate
ed

coefficients. International Journal of Mineral Processing.13(1984), pp. 239-250.

pt

Frew J.A. and Davey K.J., 1985. Backmixing in industrial flotation banks. 2. Effect upon

metallurgical performance. International Journal of Mineral Processing. 15(1985), pp. 239–

ce

250Gorain, B.K., Franzidis, J.P.and Manlapig, E.V., 1997. Studies on impeller type, impeller

speed and air flow rate in an industrial scale flotation cell. Part IV – Effect of bubble surface area
Ac

flux on flotation performance. Miner. Eng. 10, 639–654.

Gorain, B.K., Franzidis, J.P. and Manlapig, E.V., 1999. The empirical prediction of bubble

surface area flux in mechanical flotation cells from cell design and operating data. Minerals

Engineering 12, 309– 322.

42
 Gorain, B.K., Harris, M.C., Franzidis, J. P.and Manlapig, E.V., 1998. The effect of froth

residence time on the kinetics of flotation. Minerals Engineering 11 (7), 627–638.

Gulsoy, O.Y., 1999 An evaluation of the entrainment modelling approaches in flotation, Mining,

Bulletin of chamber of mining engineers of Turkey, 1999, 34, 13-30 (in Turkish).

t
ip
Guria, C., Verma, M., Mehrotra, S.P.and Gupta, S.K., 2005. Multi-objective optimal synthesis

cr
and design of froth flotation circuits for mineral processing, using the jumping gene adaptation

of genetic algorithm. Industrial & Engineering Chemistry Research 44, 2621–2633.

us
Mineral Processing and Extractive Metallurgy Review

Harris, C.C., 1978. Multiphase model of flotation machine behaviour. Int. J. Min. Proc. 5, 107–

129.
an
Harris, C.C. and Chakravarti, A., 1970. Semi-batch flotation kinetics: species distribution
M
analysis. Trans. AIME 247, 162–172.

Harris, C.C., Chakravarti, A. and Degaleesan, S.N., 1975. A recycle flow flotation machine
ed

model, Int J Miner Process, 2(1): 39–58.

pt

Harris, C.C. and Cuadross-Paz, A., 1978. Species interaction in flotation: a laboratory-scale

semi-batch study. Int. J. Min. Proc. 5, 267–283

ce

Heindel, T. J., 1999. Fundamentals of flotation deinking. TAPPI Journal, 82(3), 115–124.
Ac

Herbst, J.A.and Bascar, O.A., 1984. Mineral processing control in the 1980s. In: J.A. Herbst,

D.B. George, K.V.S. Sastry (Eds.), Control ’84. SME-AIME, New York, NY, pp. 197–215.

Hernainz and Calero, 2001. Froth flotation: kinetic models based on chemical analogy. Chemical

Engineering and Processing 40 (2001) 269–275.

43
 Hewitt, D., Fornasiero and D., Ralston, J., 1995. Bubble–particle attachment, J. Chem. Soc.

Faraday Trans., 91 (13) (1995), pp. 1997–2001.

Hu, W., Hadler, K., Neethling, S.J. and Cilliers, J.J., 2013. Determining flotation circuit layout

using genetic algorithms with pulp and froth models. Chemical Engineering Science 102 (2013)

t
ip
32–41.

cr
Huber-Panu, I., Ene-Danalache, E. and Cojocariu, D.G., 1976. Mathematical models of batch

and continuous flotation. In: Fuerstenau, M.C. (Ed.), Flotation, A.M. Gaudin Memorial Volume,

us
Mineral Processing and Extractive Metallurgy Review

vol. 2, AIME, New York, pp. 675–724.

an
Imauzimi, T. and Inoue, T., 1965 Proc. Int. Min. Proc. Congress, 6th, Cannes, 1963

Imauzimi, T. and Inoue, T., 1963. Kinetic consideration of froth flotation. Proc. VI Int. Min.
M
Proc. Congress, Cannes.

Jameson, G.J. and Nam, S., Young, M.M., 1977. Physical factors affecting recovery rates in
ed

flotation. Miner. Sci. Eng. 9 (3), 103– 118.

pt

Jena, M. S., Biswal, S. K., Mandre, N. R. and Venugopal, R., 2009. Study on axial dispersion

coefficient of water phase in column flotation. X Intl. Conf. [MPT-2009], Bhubaneswar.

ce

Jian-Guo Zhang, 1989. Factors affecting the kinetics of froth flotation. PhD thesis, Department
Ac

of Mining and Mineral Engineering, University of Leeds.

Joshi, J. B. and Sharma, M. M. 1979. A circulation cell model for bubble columns.Chemical

Engineering Research and Design, 57(1979) . pp. 244-251

44
 Jowett, A. and Safvi, S.M., 1960. Refinements in methods of determining flotation rates. Trans.

AIME 217, 351–357.

Karimi, M., Akdogan, G. and Bradshaw, S.M., 2014. A computational fluid dynamics model for

the flotation rate constant. Minerals engineering. Vol. 69, pp 214–222

t
ip
Kawatra, S. K., and Eisele, T. C., 2001. Coal Desulfurization: High-Efficiency Preparation

cr
Methods, Taylor and Francis, New York, pp. 360

Kelebek, S. and Nanthakumar, B., 2007. Characterization of stockpile oxidation of pentlandite

us
Mineral Processing and Extractive Metallurgy Review

and pyrrhotite through kinetic analysis of their flotation. Int. J. Miner. Process. 84 (1–4), 69–80.

an
Kelsall, D.F., 1961. Application of probability in the assessment of flotation systems. Bull. Instn.

Min. Metall. 650, 191–204.

M
King, R.P., 2001, Modeling and Simulation of Mineral Processing Systems, 1st edition,

(Butterworth-Heinemann, Oxford, UK).

ed

Kirjavainen, V.M., 1996. Review and analysis of factors controlling the mechanical flotation of
pt

gangue minerals. Int. J. Mineral Processing 46 (1–2), 21–34.

Klimpel, R.R., 1980. Selection of chemical reagents for flotation. In:Mullar, A.L., Bhappu, R.B.
ce

(Eds.), Mineral Processing Plant Design, 2nd edn. AIME, New York, pp. 907–934.
Ac

Klimpel, R. R. (1995). The Influence of Frother Structure on Industrial Coal Flotation,

HighEfficiency Coal Preparation (Kawatra, ed.), Society for Mining, Metallurgy, and

Exploration, Littleton, CO, pp. 141-151

45
 Koh, P., Manickam, M. and Schwarz, M.P., 2000. CFD simulation of bubble–particle collisions

in mineral flotation cells. Minerals Engineering 13, 1455–1463.

Koh, P. and Schwarz, M., 2003. CFD modelling of bubble–particle collision rates and

efficiencies in a flotation cell. Miner. Eng. 16 (11), 1055–1059.

t
ip
Koh, P. and Schwarz, M., 2007. CFD model of a self-aerating flotation cell, Int. J. Miner.

cr
Process. 85 (2007) 16–24

Kostoglou, M., Karapantsios, T.D. and Matis, K.A., 2006. Modeling local flotation frequency in

us
Mineral Processing and Extractive Metallurgy Review

a turbulent flow field, Advances in Colloid and Interface Science, 122: 79–91.

an
Kwon, S.B., Park, N.S., Lee, S.J., Ahn, H.W. and Wang, C.K., 2006. Examining the effect of

length/width ratio on the hydro-dynamic behaviour in a DAF system using CFD and ADV
M
techniques, Water Science and Technology, 53(7): 141–149.

Laplante, A.R., Toguri, J.M. and Smith, H.W., 1983. The effect of air flowrate on the kinetics of
ed

flotation. (II) The transfer of material from the froth over the cell lip, Int J Miner Process, 11:

221–234.
pt

Lee, C. H., Erickson, L.E. and Glasglow, L.A., 1987. Bubble breakup and coalescence in
ce

turbulent gas-liquid dispersions. Chem. Eng. Communications, 1987, 59, 65-84

Levenspiel, O., 1972. Chemical reaction engineering, 2nd ed., Wiley, Newyork, p. 273.
Ac

Lynch, A.J., Johnson, N.W., Manlapig, E.V., Thorne and C.G., 1981. Mineral and coal flotation

circuits, their simulation and control. In: D.W. Fuerstenau (Ed.), Developments in Mineral

Processing. Elsevier, Amsterdam, 291 pp.

46
 Mankosa, M., Luttrell, G. and Adel,G., Yoon, R., 1992. A study of axial mixing in column

flotation, Int. J. Miner. Process., 35 (1992), pp. 51–64

Mao L. and Yoon R.H., 1997. Predicting flotation rates using a rate equation derived from first

principles; Int. J. Miner. Process, 51 (1997) 171-181.

t
ip
Mavros, P., 1993a. Mixing in flotation columns. Part I: axial dispersion modeling. Miner. Eng., 6

cr
(1993), pp. 465–478Mavros, P., 1993b. Mixing in flotation columns. Part III. Internal

circulation, Miner. Eng., 6 (1993), pp. 1167–1182

us
Mineral Processing and Extractive Metallurgy Review

Mavros, P., Lazaridis, N.K. and Matis, K.A. 1989. A study and modelling of liquid-phase mixing

an
in a flotation column. International Journal of Mineral Processing. 26(1989), pp. 1–16McIvor,

R.E. and Finch, J.A., 1991. A guide to interfacing of plant grinding and flotation operations.
M
Miner. Eng. 4, 9–23.

McKee, D.J., 1991. Automatic flotation control - a review of 20 years of effort. Minerals
ed

Engineering 4 (7–11), 653–666.

Mehrotra S.P.and Kapur P.C., 1974. The effects of aeration rate, particle size and pulp density on
pt

the flotation rate distribution. Powder Technology, 9, 213-219

ce

Mehrotra S.P. and Padmanabhan, N.P.H., 1990. Flotation kinetics — a review, Trans. Indian

Inst. Met. 43 (1990) 9–21.

Ac

Mehrotra S.P. and Saxena A.K., 1983. Effects of process variables on the residence time

distribution of a solid in a continuously operated flotation cell. International Journal of Mineral

Processing. 10(1983), pp. 255–277

47
 Mendez, D.A., Glvez, E.D. and Cisternas, L.A., 2009. State of the art in the conceptual design of

flotation circuits. International Journal of Mineral Processing 90, 1–15.

Mika, T.S. and Fuerstenau, D.W., 1969. A microscopic model of the flotation process. VIII. Int.

Mineral Proc. Congress, Leningrad, Paper No. S-4, pp. 246–269.

t
ip
Mills, P. and O'Connor C.T., 1990. The modelling of liquid and solids mixing in a flotation

cr
column. Miner. Eng. 5 (1992), pp. 567–576

Mori, S. et al., 1983. Kinetic studies of fluoride flotation. XV Int. Miner. Process. Congr.

us
Mineral Processing and Extractive Metallurgy Review

Cannes, France, 1983.

an
Mular, A. L., 1972. Empirical modelling and optimization of mineral processes. Mineral science

and engineering, 4, No. 3, pp. 30-42.

M
Newell, R., 2006. Scale-up criteria in flotation cells. PhD Thesis, Ian Wark Research Institute,

University of South Australia.

ed

Newcombe, B. 2014. A phenomenological model for an industrial flash flotation cell; Minerals
pt

Engineering 64 (2014) 51–62.

Nguyen, A.V. and Schulze, H.J., 2004. Colloid Science of Flotation, Chapter 30, in Surfactant
ce

Series Vol. 118. Marcel Dekker, Inc, New York.

Ac

Nguyen A. V., Ralston J. and Schulze H. J., 1998. On modelling of bubble particle attachment

probability in flotation, Int. J. Miner. Process. 53 (1998) 225–249.

Niemmi A. J., 1995. Role of kinetics in modelling and control of flotation plants, Powder

technology 82 (1995) 69-77.

48
 Oliveira J.F., Saraiva S.M., Pimenta J.S. and Oliveira A.P.A, 2001. Kinetics of Pyrochlore

flotation from Araxa Mineral Deposits. Minerals Engineering, Vol. 14, No.1, pp. 99-105.

Pan, R., Paulsen, F. G., Johnson, D. A., Bousfield, D. W. and Thompson, E. V., 1993. A global

model for predicting flotation efficiencies: Model results and experimental studies. TAPPI

t
ip
Pulping Conference (p. 1155). Atlanta: TAPPI Press.

cr
Mavros, P., 1992. Validity and limitations of the closed vessel analytical solution to the axial

dispersion model. Miner Eng., 5(9):1053 (1992)

us
Mineral Processing and Extractive Metallurgy Review

Polat, M. and Chander, S., 2000. First-order flotation kinetics models and methods for estimation

an
of true distribution of flotation rate constants. Int. J. Miner. Process. 58, 145–166.

Prince, M.J. and Blanch, H.W., 1990. Bubble coalescence and break-up in air-sparged bubble
M
columns, A.I.C.H.E. Journal, 36, (1990), 1485-1499.

Pyke, B., 2004. Bubble–particle capture in turbulent flotation systems. Ian Wark Research
ed

Institute, University of South Australia.

pt

Quinn J. and Finch, J., 2012. On the origin of bi-modal bubble size distributions in the absence

of frother. Miner. Eng., 36 (2012), pp. 237–241

ce

Ralston, J., 1992. In: Laskowski, J.S. and Ralston, J. (Eds.), Chapter 6 in Colloid Chemistry in
Ac

Minerals Processing. Elsevier, Amsterdam, pp. 203– 223.

Ralston J., Fornasiero D., Grano S., Duan J. and Akroyd T., 2007. Reducing uncertainty in

mineral flotation—flotation rate constant prediction for particles in an operating plant ore, Int. J.

Miner. Process. 84 (2007) 89–98

49
 Rastogi, R.C. and Aplan, F.F., 1985. Coal flotation as a rate process. Min. Metall. Proc. Aug.,

137–145.

Ross, V.E., 1990. A study of the froth phase in large-scale pyrite flotation cells. Int. J. Miner.

Process. 30, 143–157.

t
ip
Rubinstein, J.B., 1995. Column Flotation: Processes, Designs and Practices, 1st edition (Gordon

cr
and Breach Science Publishers, Basilea, Switzerland).

Runge, K., 2009. Modelling of ore floatability in industrial flotation circuits. PhD Thesis,

us
Mineral Processing and Extractive Metallurgy Review

University of Queensland.

an
Runge, K.C., Alexander, D.J., Franzidis, J.P., Morrison, R.D. and Manlapig, E.V., 1998.

JKsimfloat – a tool for flotation modeling. In: 1998 AusIMM Annual Conference – The Mining
M
Cycle, pp. 361–370.

Savassi, O.N., 1998. Direct estimation of the degree of entrainment and the froth recovery of
ed

attached particles in industrial flotation cells. PhD Thesis, University of Queensland.

pt

Savassi, O.N., 2005. A compartment model for the mass transfer inside a conventional flotation

cell. Int. J. Miner. Process., 65–79.

ce

Schimmoller, B. K., Luttrell, G. H., and Yoon, R. H., 1992. ‘‘Proceedings, Engineering
Ac

Foundation Conference on Dispersion and Aggregation, March, 1992.’’

Schimmoller, B. K., Luttrell, G. H., and Yoon, R. H., 1993. ‘‘Proceedings, VXII Int. Mineral

Processing Congr., Sydney, Australia, May 23–28, 1993,’’ Vol. 3, p. 751.

50
 Schuhmann, R., 1942. Methods for steady state study of flotation problems, J. Phys. Chem., 46,

pp. 891-902

Schulze, H.J., 1977. New theoretical and experimental investigations on stability of bubble-

particle aggregates in flotation: a theory on the upper particle size in floatability. Int. J. Min.

t
ip
Proc. 4, 241–259.

cr
Schulze, H.J., 1984. Physico-Chemical Elementary Processes in Flotation – An Analysis from

the Point of View of Colloid Science Including Processes Engineering Considerations. Dev. in

us
Mineral Processing and Extractive Metallurgy Review

Mineral Processing, Vol. 4 (Fuerstenau, D.W., Advisory Ed.), Elsevier, Amsterdam, 348 pp.

an
Schulze, H. J., 1991. The fundamentals of flotation deinking in comparison to mineral flotation.

1st Research Forum on Recycling (p. 161). Toronto: CPPA Press.

M
Schulze, H. J., 1994a. Comparison of the elementary steps of particle/bubble interaction in

mineral and deinking flotation. Eighth International Conference on Colloid and Surface Science.
ed

Schulze, H. J., 1994b. Zur hydrodynamik der flotations-elementarvorgange. Wochenblatt Fur

Papierfabrikation, 122(5), 160, 162, 164–168.

pt

Seaman, D.R., Franzidis, J.P. and Manlapig, E.V., 2004. Bubble load measurement in the pulp
ce

zone of industrial flotation machines-a new device for determining the froth recovery of attached

particles, Int J Miner Process, 74: 1–13.

Ac

Sherrell, I. M., 2004. Development of a Flotation Rate Equation from First Principles under

Turbulent Flow Conditions. PhD thesis, Virginia Polytechnic Institute and State University.

51
 Somasundaran, P. and Lin, I. J.,1971. Method for Evaluating Flotation Kinetics Parameters.

Transactions of SME, Vol 254(1973)

Subrahmanyam, T.V. and Forssberg, E., 1988. Froth stability, particle entrainment and drainage

in flotation – a review. Int. J. Mineral Processing 23, 33–53.Tokoro, C., and Okano, Y., 2001.

t
ip
Boundary conditions for column flotation—a study by transfer function representation of an

axial diffusion model. Minerals Engineering, 14 (1) (2001), pp. 49–64

cr
Tomlinson, H.S., and Fleming, M.G., 1965. Flotation rate studies. International Mineral

us
Mineral Processing and Extractive Metallurgy Review

processing Congress, 6th Cannes, Pergamon Press, Oxford – New York, pp. 563-579.

an
Tuteja R.K., Spottiswood D.J. and Misra V.N., 1994. Mathematical models of the column

flotation process: a review. Miner. Eng, 7(1994), pp.1459–1472.

M
Uçurum, M., 2009. Influences of Jameson flotation operation variables on the kinetics and

recovery of unburned carbon; Powder Technology 191 (2009) 240–246

ed

Vera, M.A., Franzidis, J. P. and Manlapig, E.V., 1999. Simultaneous determination of collection

zone rate constant and froth zone recovery in a mechanical flotation environment. Minerals
pt

Engineering 12 (10), 1163–1176.

ce

Vera, M. A., Mathe, Z.T., Franzidis J. P., Harris M.C., Manlapig E.V. and O'Connor C.T.,

2002. The modelling of froth zone recovery in batch and continuously operated laboratory
Ac

flotation cells. International Journal of Mineral Processing, 64, 135–151

Vianna, S., 2004. The effect of particle size, collector coverage and liberation on the floatability

of galena particles in an ore. PhD Thesis, University of Queensland.

52
 Welsby, S., Vianna, S. and Franzidis, J. P., 2010. Assigning physical significance to floatability

components. Int. J. Miner. Process. 97, 59–67.

Whiten, W. J., 1971. Model building techniques applied to mineral treatment processes. Aus.

I.M.M. Symposium on automatic Control Systems in Mineral Processing Plants, University of

t
ip
Queensland, Brisbane.

cr
Wilson S.G. and Frew J.A., 1986. Effect of operating variables upon liquid backmixing in a

laboratory flotation bank. International Journal of Mineral Processing, 16 (1986), pp. 281–

us
Mineral Processing and Extractive Metallurgy Review

298Woodburn, E.T., 1970. Mathematical modeling of flotation process. Minerals Sci. Eng. 2 (2),

3–17.
an
Woodburn, E.T. and Loveday, B.K., 1965. The effect of variable residence time on the
M
performance of a flotation system. Journal of the South African Institute of Mining and

Metallurgy 65, 612–628.

ed

Xu, M., 1998. Modified flotation rate constant and selectivity index, Miner. Engg. 11 (3) (1998)

271–278.
pt

Xu, M. and Stratton-Crawley, R., 1996. Rate Constant and Selectivity Measurements in a
ce

Flotation Column. In:Gomez, C.O., Finch, J.A. (Eds.), In Column '96.Montreal, Canada,p.81–93.

Xu, Z., and Yoon, R.H., 1990. A study of hydrophobic coagulation. J. Colloid Interface Sci. 134,
Ac

427-434 (1990).

Yalcin, E. and Kelebek, S., 2011. Flotation kinetics of a pyritic gold ore. International Journal of

Mineral Processing 98 (2011) 48–54.

53
 Yianatos, J.B., 2007. Fluid flow and kinetic modelling in flotation related processes: columns

and mechanically agitated cells. Chemical Engineering Research and Design 85 (A12), 1–13.

Yianatos, J.B., Bergh, L.G., Duran, O.U., Diaz, F.J. and Heresi, N.M., 1993. Measurement of

residence time distribution of the gas phase in flotation columns. Proceedings Miner.Eng., Cape

t
ip
Town, S. Afr. pp.25-27.

cr
Yianatos, J.B., Bergh, L.G., Vinnett, L., Contreras, F. and Díaz, F.J, 2010. Flotation rate

distribution in the collection zone of industrial cells. Minerals Engineering 23 (2010) 1030–1035.

us
Mineral Processing and Extractive Metallurgy Review

Yianatos, J.B. and Henriquez, F.H., 2006. Short-cut method for flotation rates modelling of

an
industrial flotation banks. Minerals Engineering 19, 1336–1340.

Yoon, R. H., and Yordan, J. L., 1991. The critical rupture thickness of thin water films on
M
hydrophobic surfaces. J. Colloid Interface Sci. 146 (1991), 565-572.

Yoon, R. H., 2000. The role of hydrodynamic and surface forces in bubble-particle interaction,
ed

Int. J. Miner. Process, 58(1–4): 129–143.

pt

Yoon, R.H. and Mao, L., 1996. Application of Extended DLVO Theory, IV Derivation of

Flotation Rate Equation from First Principles. Journal of Colloid and Interface Science 181, pp.
ce

613–626 (1996).
Ac

Yotsumoto, H. and Yoon, R. H., 1993. Application of extended DLVO theory: I. Stability of

rutile suspensions. J. Colloid Interface Sci. 157(1993), 426.

Zheng X., Johnson N.W. and Franzidis, J. P., 2006. Modelling of entrainment in industrial

flotation cells: Water recovery and degree of entrainment. Miner. Eng, 19 (2006), pp. 1191–1203

54
 Table 1: Various flotation models based on kinetics

1st
Model Classical order
R = R∞ (1 − e − kt )
1 kinetic model

t
ip
Model 1st
 1 
order flotation
R = R∞ 
k t 1 − e − kt
( )


cr
2 model with rectangular
 2 
distribution of

us
Mineral Processing and Extractive Metallurgy Review

floatability

Model 1st order 2 stage kinetic

an  k3   k 3*  
R = R∞  * 
(1 − e ) − 
−k3*t
(1 − e )
−k3t
3 model
 *
M
 k3 − k 3   k3 − k3  
ed

[ ]
Model 1st order reversible
k 4 R∞
R= 1 − e −( k 4 + k − 4 )t
(k 4 + k − 4 )
pt

4 kinetic model
ce
Ac

55
 Model Fully mixed reactor
 
5 model  
R = R∞  1 −
1 
 t 
 1+ 

t
 k5 

ip
cr
Model Gas/solid kinetic
k6t
R =

us
Mineral Processing and Extractive Metallurgy Review

6 adsorption model
(1 + k 6 t )m
an
 k7t 
Model Improved gas/solid
M
7 adsorption model R = R∞  
1 + k 7 
t
ed

Model 2nd order kinetic model

 R∞ 2 k ∞ t 
pt

8 R =  
1 + R ∞ kt 
ce
Ac

Model 2nd order kinetic model

 
9 with rectangular
R = R∞ 1 −
1
{ln (1 − k9t )}
 k9t 
distribution of

56
 floatability

Model Three parameter fast

(
R = (1 − φ ) 1 − e
− k10 f t
) + φ (1 − e − k10 s t
)

t
ip
10 and slow floating

component

cr
( )+ φ (1 − e )

us
R = ( R∞ − φ ) 1 − e
− k11 f t
Mineral Processing and Extractive Metallurgy Review

Model Four parameter fast and − k11 s t

11 slow floating

component an
M
 
Model Three parameters
e − k12 s t − e − k12 u t
R = R∞ 1 − 
(k12u − k12 s t )
12 kinetic model with
 
ed

particle floatability

proportional to sized
pt

distribution
ce

  λ  
Model Three parameter

R = R∞ 
1 −  λ + t  p 
Ac

13 gamma distribution

   

57