Energy Conversion and Management 51 (2010) 1891–1897

Contents lists available at ScienceDirect

Energy Conversion and Management

journal homepage: www.elsevier.com/locate/enconman

Coal flotation optimization using modified flotation parameters

and combustible recovery in a Jameson cell
Hüseyin Vapur a, Oktay Bayat a, Metin Uçurum b,*
a
Çukurova University, Mining Engineering Department, Balcali, 01330 Adana, Turkey
b
Niğde University, Mining Engineering Department, 51100 Niğde, Turkey

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

Article history: This study discusses a new coal flotation optimization approach. It is conducted using modified flotation
Received 13 November 2008 parameters and combustible recovery. The experimental work was evaluated in two stages. In the first
Received in revised form 7 July 2009 stage, recoveries (1, 2, 3, 5 and 8 min of flotation times) of Jameson flotation operating parameters were
Accepted 18 February 2010
fitted to first-order kinetic model, R = R1 [1 ! exp (!kt)] where R was recovery at t time, R1 was ultimate
Available online 20 March 2010
recovery and k was the first-order rate constant to draw the time recovery curves in the experimental
study. Two parameters, the ultimate recovery (R1) and first-order rate constant (k), were then obtained
Keywords:
from the model to fit an experimental time recovery curve. A modified flotation rate constant (Km)
Coal flotation
Flotation kinetics
defined as product of R1 and k, i.e., Km = R1 " k, and selectivity index (SI) defined as the ratio of the mod-
Combustible recovery ified rate constant of coal to the modified rate constant of ash (SI)=Km of Coal/Km of Ash), which could be
Optimization collectively called ‘‘modified flotation parameters”. It was used to determine of the sub and upper values
of operation variables. In the second one, combustible recovery (%) and ash content (%) were used to opti-
mization of the Jameson flotation variables and it was found that d80 = 0.250 mm particle size, 1/1 veg-
etable oil acids/kerosene ratio, 20% solids pulp density, 0.600 L/min wash water rate and 40 cm
downcomer immersion dept could be used to separate efficiently coal from ash. Final concentrate was
obtained with 94.83% combustible recovery and 17.86% ash content at optimum conditions after 8 min
flotation time.
! 2010 Elsevier Ltd. All rights reserved.

1. Introduction tation technologies, including a more compact design and lower

Flotation is one of the most complex mineral processing opera-
tions as it is affected by a very large number of variables. Many of
these are beyond the control of the minerals engineer, and some
cannot be even measured quantitatively with the available instru-
ments. The relations between measured and controlled variables
are intricately related. Sometimes, simultaneously changing vari-
ous component settings will reinforce a particular attribute. In
addition, various component settings can cancel or counteract each
other if changes are not chosen wisely. For coal fines (<0.5 mm),
froth flotation is the most effective method of separating ash form-
ing mineral matter from the carbonaceous material [1].
The Jameson cell is a co-current flotation device, in which a feed
slurry-frother mixture is pumped through a fixed diameter orifice.
Feed injection through the orifice under pressure draws in air due
to a venture effect. Ultra fine bubbles are produced in the down-
comer because of the high shearing provided by the jet action of
the feed slurry [2]. It has major advantages over conventional flo-
* Corresponding author. Tel.: +90 388 225 23 50; fax: +90 388 225 01 12.
E-mail address: cevher@nigde.edu.tr (M. Uçurum).
capital cost, a minimum amount of maintenance and the ability
to operate at elevated temperatures [3].
Flotation kinetics brings together the influence of all the factors
that take place during a flotation process [4], some of the most
important factors required to achieve a good separation are ore
characteristics, machines and equipment involved, hydrodynam-
ics, particle size, pulp density and chemical reagents. Kinetic mod-
els can be used to analyze batch flotation results. Two parameters
which are R1 (ultimate recovery) and k (first-order rate constant)
are obtained from the model fit to an experimental recovery-time
curve. They can be effectively used to evaluate variables affecting
flotation process [5]. Among the many flotation models, the classi-
cal first-order flotation model (Eq. (1)) is widely used and can be
utilized to optimize the design of flotation circuits [5–7].
R ¼ R1 ½1 ! expð!ktÞ' ð1Þ
where R is the recovery of component at time t, R1 is the ultimate
recovery of component and k is the first-order rate constant for the
component.
Optimization of flotation parameters using rate models is not a
new concept. The kinetic model based on time recovery data,

0196-8904/$ - see front matter ! 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.enconman.2010.02.019
1892 H. Vapur et al. / Energy Conversion and Management 51 (2010) 1891–1897
which uses the extra dimension of rate, has been in vogue since
time immemorial for scaling up of laboratory data. Often, interpre-
tation on the performance of a flotation circuit, based only on R1
(the ultimate recovery) and k (the first-order rate constant of the
component) may lead to wrong conclusions. In such cases, a mod-
ified flotation rate constant Km defined as the product of R1 and k,
i.e., Km = R1 " k and selectivity index (SI), defined as the ratio of the
modified rate constant of valuables to the modified rate constant of
gangue, can be used [8].
Sripriya et al. [8] has been made an attempt to optimize the
batch laboratory froth flotation parameters of fine coal using the
above two concepts, i.e., Km and SI and statistical techniques. A flo-
tation bank containing four Outukumpu cells was optimized using
the results obtained from the laboratory study. The airflow number
and the froth number were used as a basis for scale up. To gauge
the performance of the froth flotation circuit, an efficiency param-
eter called the coefficient of separation (CS) was used. The yield
from the flotation circuit improved, the froth ash reduced and
the rejects ash went up. Xu [5] and Uçurum and Bayat [9] remarks
modified flotation rate constant and selectivity index are useful
parameters in the flotation studies.
In the optimization of flotation, the statistical design of experi-
ments [10,11] has several advantages over the classical method of
treating one variable at a time. The full factorial experiment is
method of design of experiments in which a statistical analysis is
performed to evaluate the significance of the main and interaction
effects as evaluated from the experimental results. In particular,
they are used when several factors have to be studied in order to
determine their main effects and interaction. The experiments
can be conducted in an organized manner and can be analyzed sys-
tematically to obtain much needed information. The information
can be utilized for optimization purpose. A valid optimization
strategy would permit the adjustment of those manipulable vari-
ables, which influence the objective. The statistical techniques
have been used to study the flotation of minerals [12–14] and coal
[15,16].
A review of coal processing literature indicates a lack of statis-
tically based studies on application and/or interaction of reagents
with coals [17]. Naik et al. [18] carried out a 25 factorial design
to study the effect of diesel oil, pH, MIBC, particle size and sodium
silicate on flotation of a low-rank non-coking coal. They found that
the influence of particle size was relatively more significant than
those of various reagents among which the collector had the high-
est influence. Under optimum conditions, the combustible recov-
ery was 88% at 25.4% ash. They also carried out 23 factorial
design experiments [19] on the influence of sodium meta silicate,
kerosene and MIBC with a higher rank coal. They obtained a prod-
uct with 91.11% combustible grade with 95.58% combustible
recovery.
In this study, optimization of coal flotation was made in two
stages. In the first stage, classical first-order flotation model was
used in order to investigate the modified flotation parameters
(Km, SI) of five important Jameson flotation variables, in the second
stage, combustible recovery (%) and ash content (%) were utilized.
This experimental design is a new methodology in the field and
this work can potentially produce a significant contribution to
the literature of the subject.
2. Materials and method
2.1. Coal washing plant intermediate sample
The sample of coal washing plant intermediate product used in
the experiments was collected from Ömerler coal washing plant in
Kütahya, Turkey. The chemical analysis of the sample was given
Table 1. The XRD study of the sample showed that it contained
the following minerals: quartz, serpentine–kaolin group minerals,
illite, smectite group clay minerals and amorphous matter. The cal-
orific value of the sample was 3935 kcal/kg. The volatile matter of
sample was determined as 29.30%. Besides, the ash content and the
moisture content of sample were determined as 33.41% and 8.70%,
respectively.
2.2. Slurry preparation
For a standard flotation test, a sub-sample (0.5 kg) was ground
in a stainless steel mill under dry medium. The ball mill had the
dimensions, 200 ( 200 mm, and was charged with 10 kg of stain-
less steel balls. The grinding times were 10, 15 and 20 min, giving
particle size distributions of d80 = 0.355, 0.250 and 0.106 mm,
respectively. The size analyses of the feed and three individual
fractions are shown in Fig. 1. Kerosene oil (0.8 gr/cm3) and vegeta-
ble oil acids (0.9 gr/cm3) supplied by MarSA-Adana, Turkey were
used as reagents. At different weight ratios of vegetable oil acids
and kerosene were mixed with tap water to obtain stable emul-
sions in condition tank of the Jameson flotation cell for 5 min. Veg-
etable oil acids + kerosene dosage was 10% of the dry solids by
weight. For the preparation of slurry feed for the Jameson cell,
ground fine coal was added in solution and conditioned for 5
min. All flotation experiments were conducted at normal pH
(7.0–7.5) where no flocculation of gangue on the sample was
thought to be observed due to high surface of both at this pH.
The pulp was then floated for 1, 2, 3, 5 and 8 min adding wash
water.
2.3. Jameson flotation cell set up and its operating characteristics
Laboratory batch flotation tests were made in a Jameson flota-
tion cell which was constructed in stainless steel. The Jameson flo-
tation cell can be divided into two main zones. These are
downcomer (2.0 cm in diameter and 100 cm in length) and separa-
tion tank (10 cm in diameter and 75 cm in length). Pulp is condi-
tioned in the 30 L tank and pumped into the downcomer, which
is the primary contacting zone of particles swarmed with bubbles.
There is a nozzle at a relatively high level of the downcomer to pro-
vide a high pressure jetting action. When the slurry passes the noz-
zle, atmospheric air is sucked into the downcomer due to the
venturi effect. The jetting impact of slurry pool in the downcomer
and the air sucked plunge into the separation tank (plunging jet).
In the mixing zone, i.e., top part of the slurry pool in the downcom-
er, the hydrophobic particles have opportunity to collide with and
adhere to fine air bubbles (e.g., 400–700 lm) generated within the
water jet shearing action [20]. The Jameson cell is an effective de-
vice for gas–liquid contacting. Small bubbles are formed in a high-
shear region surrounding the plunging jet, leading to high interfa-
cial area per unit volume of gas [21]. Flotation recovery of Jameson
cells is an important issue and can be variable depending on oper-
Table 1
Chemical compositions and physical properties of the sample (as received) used in the
study.
Property
C (%) 42.56
H (%) 3.13
N (%) 1.65
O (%) 8.13
S (%) 2.42
Moisture (%) 8.70
Ash (%) 33.41
Volatile matter (%) 29.30
Calorific values (Kcal/kg) 3935
 H. Vapur et al. / Energy Conversion and Management 51 (2010) 1891–1897 1893

100 factors related to particle size have been identified, the net effect of
90 these factors is very difficult to predict. For instance, in particle–
bubble interaction, particle size is known to play a critical role in
80
Cumulative Passing (%)

the probability of particles colliding with bubbles, attachment of

70
particles to bubbles after collision, as well as remaining attached
60 in the pulp phase [24]. This study examined the effects of particle
50 size distribution (i.e., d80 = 0.250 and d80 = 0.106 mm) on the flota-
40 tion behavior of coal in the Jameson cell. Table 3 gives ultimative
recovery (R1), first-order rate constant (k), modified rate constant
30
Feed
(Km), selectivity index (SI) and coefficient (R2) for studied particle
20
10 min. Grinding sizes. d80 = 0.250 mm gives better results for k and Km for coal. Fur-
10 15 min. Grinding
thermore, d80 = 0.250 mm had higher SI (3.63) than d80 = 0.106 mm
20 min. Grinding
0 (2.77). Kinetics analysis of experimental data indicated that the
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 best feed size for the coal should be d80 = 0.250 mm. In the labora-
Size Distribution (mm) tory tests, the size particles were also observed to form smaller and
more stable bubbles, resulting in more efficient flotation. In addi-
Fig. 1. Particle size distribution of sample after three different grinding times.
tion, the first-order kinetic model gave the best fit for particle sizes
experimental data (Fig. 2).
ating parameters and particle size. To assess the flotation recovery Solids concentration is expected to influence solids suspension
in the Jameson cell, the operation within the downcomer and oper- due to its effects on hindered settling, turbulence dampening,
ation within the separation tank should be evaluated separately. and (possibly) viscosity effects [25]. Using different solids concen-
The recovery that occurs within the downcomer is an area that is trations, the effect of solids concentration on the modified flotation
still under active investigation [22]. Experimental details of the flo- parameters was determined for the coal. In order to examine the
tation tests are summarized in Table 2. effect of pulp density on the flotation, 5% and 20% by weight of
pulp densities were used. Table 4 presents ultimate recovery
2.4. Analysis (R1), first-order rate constant (k) and coefficient (R2) obtained from
fitting, including the modified rate constant (Km) and selectivity in-
The hydrophobic particles and tailings were collected, filtered dex (SI). Five percent by weight of pulp density gives higher results
and dried in an oven at 90 ± 5 "C to constant weight and assayed. for k and Km for the coal. Furthermore, 20% solids is strongly lower
Their ash levels were determined at 875 "C. General guidelines out- than 5% solids in terms of rate constant (k = 0.0676) and modified
lined in the ASTM standards were followed. In this study, the frac- rate constant (Km = 0.0730) for ash. The selectivity in flotation be-
tional recoveries after 1, 2, 3, 5 and 8 min of flotation time were tween coal and ash wasn’t observed when the pulp density is en-
fitted to the models. A proprietary curve-fitting program (Curve hanced from 5% to 20% by weight and the selectivity index of the
Expert 1.3, a shareware from http://www.ebicom.net/~dhyams/ pulp densities are almost the same with 3.63 and 3.85. The first-or-
cvxpt.htm [23]) was used to determine k, R1 and R2. der flotation model fits the pulp densities experimental data very
In the second stage the study, the results of the flotation test re- well (Fig. 3) for coal and ash.
sults were evaluated by percentage of combustible recovery. The Water-insoluble hydrocarbons are widely used as collectors to
combustible recovery was calculated using Eq. (2) increase the affinity of coal particles towards the air bubbles. These
! " collectors are basically non-polar oils such as kerosene, crude
M c ð100 ! Ac Þ petroleum, fuel-oil and certain coal-tar distillates [26]. Kerosene,
Combustible recovery ð%Þ ¼ ( 100 ð2Þ
M f ð100 ! Af Þ the zpc of most of the coals are below 5.5 and bears negative charge
in neutral pH [27]. The kerosene droplets also bear negative charge
where Ac ash content of clean coal, Af ash content of feed, Mc mass of
at neutral pH. Therefore, the interaction of coal and kerosene is
clean coal, Mf mass of feed.
possibly due to hydrophobic interaction [28]. The floatability of
coal increases with the adsorption of kerosene and there is increase
3. Results and discussion in recovery. Klimpel and Hansen [29] reported the similar results
but they also observed decrease in flotation rate constant at high
Although the effect of particle size on flotation performance has collector dosage. With the increase in hydrophobicity of coal parti-
been widely studied to date and many important physico-chemical cles, inter-particle attraction will increase and agglomeration is
likely to take place. The decrease in grade is due to flotation of pro-
Table 2
gressively higher mineral matter particles with the increase in
Jameson flotation column design and operating specifications. recovery. The increase in collector dosage also causes high recov-
ery of finer coal particles with very poor selectivity of coal over
Column variable
other mineral matter in the finer particle range. The use of oil im-
Diameter (mm) 100 proves flotation rate of particles of all sizes and specific gravities
Height (mm) 750
Downcomer diameter (mm) 20
but the effect is more for the locked or mineral particles [30]. Veg-
Downcomer height (mm) 1000 etable oil are also light compounds (0.9 g/cm3), less expensive than
Froth height (mm) Variable hydrocarbons and even more important, unlike mineral oils, they
d80 (mm) 0.355, 0.250, 0.106 are renewable and non-polluting energy resources [31]. It is well
pH 7.0-7.5
known that the collector dosage of coal flotation is an important
Vegetable oil acids/kerosene ratios 2/3, 1/1, 3/2
Pulp density (%) 5, 10, 20 variable. The aim of this work is to obtain high calorific value prod-
Wash water rate (L/min) 0, 0.600, 1.0 ucts from coal fines cleaning wastes by flotation with kerosene and
Downcomer immersion depth (cm) 30, 40, 50, 60 vegetable oil acids mixing. The effect of vegetable oil acids/kero-
Conditions time (min) 10 sene ratio on modified flotation parameters was investigated using
Flotation time (min) 1, 2, 3, 5, 8
2/3 and 3/2. Table 5 gives ultimative recovery (R1), first-order rate
constant (k), modified rate constant (Km), selectivity index (SI) and
1894 H. Vapur et al. / Energy Conversion and Management 51 (2010) 1891–1897

Table 3
Parameters obtained from model (R = R1 [1 ! exp (!kt)]) fit to data set for coal and ash.

Size distribution, d80 (mm) Coal Ash SI

2 2
R1 k Km R R1 k Km R
0.250 0.9322 1.0484 0.9700 0.9994 0.6040 0.4430 0.2675 0.9992 3.63
0.106 0.9328 0.7280 0.6790 0.9996 0.7909 0.3099 0.2451 0.9981 2.77

100 100
90 90
80
80
70
70
Recovery (%)

60

Recovery (%)
60
50
0.250 mm Exp. Coal 50
40 0.250 mm Model Coal
0.250 mm Exp. Ash 40
30 0.250 mm Model Ash
5% Exp. Coal
0.106 mm Exp. Coal 30
20 0.106 mm Model Coal
5% Model Coal
5% Ash Exp.
10 0.106 mm Exp. Ash 20 5% Model Ash
0.106 mm Model Ash 20% Exp. Coal
20% Model Coal
0 10 20% Exp. Ash
0 2 4 6 8 10 20%Model Ash
0
Flotation Time (min.) 0 2 4 6 8 10

Fig. 2. Fitted to data set for size distribution on coal and ash recovery.
Flotation Time (min.)

Fig. 3. Fitted to data set for pulp densities on coal and ash recovery.

2
coefficient (R ). Modified flotation rate (Km) of 3/2 vegetable oil
acids/kerosene ratio was higher (0.6892) than lower vegetable oil
acids/kerosene ratio (0.4452) for coal. In the mean time, Km values
of ash have the same characteristic trend with coal. As for selectiv-
ity index (SI), 2/3 and 3/2 vegetable oil acids/kerosene ratio gave
almost the same results with 1.98 and 2.07, respectively. The mod-
el could present the time recovery curves of flotation tests for both
collector dosages quite well for coal and ash (Fig. 4).
The froth phase of the flotation system, to a great extent, deter-
mines the separation performance, as the grade (ratio of desired to
total solids recovered) of the product (concentrate) depends pri-
marily on its structure and stability. The froth also contributes to
the recovery (fraction of valuable solids recovered from the pulp
to the concentrate) achieved, since the amount of desired material
drop-back from the froth, together with the kinetics of the pulp
phase, determines the recovery. Wash water is traditionally added
to column flotation cells, but it has found some, though more lim-
ited, use in the operation of more conventional flotation cells. The
froth phase is extremely important in the operation of a flotation
cell, seeing that, it is critical in determining the amount of un-
wanted gangue collected to the concentrate and thus the purity
of product. The role of the wash water is to prevent pulp entrain-
ment into the concentrate. In this sense, the minimum wash water
should just supply the water the concentrate. However, an excess
of wash water increases froth mixing and short circuiting, thus
decreasing froth cleaning, particularly for shallow froth depths. It
is also important to verify the circuit water balance, because an ex-
cess of water will cause problems downstream [32]. Fig. 5 shows
the recovery of coal and ash as a function of flotation time and
model fits which shows the model fits the experimental data quite
well for coal and ash at the two different wash water rates. Table 6
presents ultimative recovery (R1), first-order rate constant (k),
modified rate constant (Km), selectivity index (SI) and coefficient
(R2) from the fittings for coal and ash. In the laboratory Jameson
flotation tests with the studied wash water rates, 0 L/min gave bet-
ter result for k and Km as 0.9510 and 0.8090, respectively. When in-
creased wash water rate from 0 L/min to 1.0 L/min, k and Km values
of ash decreased from 0.6310 to 0.1185 and from 0.3800 to 0.1134,
respectively. By way of addition, increasing wash water rate from
0 L/min to 1.0 L/min did not increased coal/ash selectivity (2.13
for 0 L/min, 2.45 for 1.0 L/min). Compared to mechanical cells,
the functions of producing bubbles and particle–bubble collision/
attachment in a Jameson cell are done separately inside the down-
comer. The following steps occur within the downcomer: (1) the
jet created by the slurry passing through the orifice and promotes
the inducement of air into the downcomer; (2) the shearing action
of the jet generates fine bubbles and transports them through the
mixing zone; (3) the particles and the bubbles collide and attach to
each other and subsequently travel down the downcomer; through
the pipe flow zone; (4) bubbles are removed by hydrostatic pres-
sure from the downcomer creating a vacuum for further air
entrainment [33]. The immersion depth of the downcomer is also
an important parameter, which determines froth depth and froth
residence time [34]. In this study, 30 cm downcomer immersion
depth was compared with 50 cm downcomer immersion depth

Table 4
Parameters obtained from model (R = R1 [1 ! exp (!kt)]) fit to data set for coal and ash.

Pulp Density (wt.%) Coal Ash SI

R1 k Km R2 R1 k Km R2
5 0.9322 1.0484 0.9700 0.9989 0.6040 0.4430 0.2675 0.9959 3.63
20 1.1005 0.2554 0.2811 0.9888 1.0840 0.0676 0.0730 0.9896 3.85
 H. Vapur et al. / Energy Conversion and Management 51 (2010) 1891–1897 1895

Table 5
Parameters obtained from model (R = R1 [1 ! exp (!kt)]) fit to data set for coal and ash.

Vegetable oil acids/kerosene ratios Coal Ash SI

2 2
R1 k Km R R1 k Km R
2/3 0.8777 0.5072 0.4452 0.9994 0.6300 0.3563 0.2245 0.9994 1.98
3/2 0.8900 0.7743 0.6892 0.9998 0.6165 0.5384 0.3322 0.9978 2.07

100 rate constant (Km), selectivity index (SI) and coefficient (R2). 30 cm
90 appear to have a higher modified rate constant (0.3799) while
50 cm has lowered a value (0.2811) for coal. In the mean time,
80
Km of 30 cm was higher (0.1620) than 50 cm (0.0730) for ash. In
70 terms of coal/ash selectivity index, 50 cm immersion depth gave
Recovery (%)

60 a much better result than 30 cm with a value of 3.85. Results in

50
2/3 Exp. Coal Fig. 6 confirm that the model fits the experimental data very well
2/3 Model Coal
for both collectors.
40 2/3 Exp. Ash
2/3 Model Ash
30 3/2 Exp. Coal 3.1. Optimization
20 3/2 Model Coal
3/2 Exp. Ash
10 3/2 Model Ash
According to kinetics results, !0.250 mm was determined as
the best particle size fraction. In the second study of optimization,
0
0 1 2 3 4 5 6 7 8 9 !0.355 mm, upper-size of !0.250 mm, and !0.106 mm, sub-size
Flotation Time (min.) of !0.250 mm, were used. The experimental conditions were kept
constant between tests and the best results were obtained with
Fig. 4. Fitted to data set for vegetable oil acids acids/kerosene ratio on coal and ash !0.250 mm particle size in terms of combustible recovery and
recovery.
ash content for 8 min flotation time. As it can be seen from the re-
sults, illustrated Fig. 7, which combustible recovery (%) and ash
content (%) were 94.97%, 24.55%, respectively.
100
For determination of optimum solids ratio of pulp, three ratios,
90
5 wt.%, 10 wt.% and 20 wt.%, used in the tests. The selectivity index
80 (SI) values of 5 wt.% and 20 wt.% were very close (3.65 and 3.85).
70 Thus, 10 wt.% was tested as a middle value of solid ratios. The cal-
Recovery (%)

60 culated results of combustible recovery and ash content (%) for the
50 three solid ratios were given in Fig. 8. The combustible recovery
0 L/min. Exp. Coal and ash content were 94.97% and 24.55% respectively for 5 wt.%.
40 0 L/min.Model Coal
0 L/min. Exp. Ash
For 10 wt.% as a middle value, the combustible recovery was
30 0 L/min. Model Ash 85.00% and ash content was 20.94%. The best results was obtained
1 L/min. Exp. Coal
20 1 L/min. Model Cola at 20% pulp density with 15.30% ash content and 92.01% combus-
1 L/min. Exp. Ash
10 1 L/min. Model Ash tible recovery. Hence, 20 wt.% was selected as an optimum solid
0 ratio.
0 1 2 3 4 5 6 7 8 9 In order to establish the individual effect of vegetable oil acids/
Flotation Time (min.) kerosene ratio, 2/3, 1/1 and 3/2 vegetable oil acids/kerosene ratios
were used. 2/3 and 3/2 vegetable oil acids/kerosene ratios had
Fig. 5. Fitted to data set for wash water rate on coal and ash recovery.
close selectivity index (SI) values (1.98 and 2.04). Because of this,
1/1 was tested as a middle value of vegetable oil acids/kerosene ra-
tio. In the second phase of experimental studies flotation tests
on the separation of coal from ash using modified flotation param- were carried out to obtain optimum vegetable oil acids/kerosene
eters. Table 7 presents (R1), first-order rate constant (k), modified ratio and the results were given in Fig. 9. As seen in Fig. 9, the best

Table 6
Parameters obtained from model (R = R1 [1 ! exp (!kt)]) fit to data set for coal and ash.

Wash water rate (L/min) Coal Ash SI

R1 k Km R2 R1 k Km R2
0 0.8504 0.9510 0.8090 0.9967 0.6030 0.6310 0.3800 0.9972 2.13
1.0 0.8752 0.3176 0.2780 0.9918 0.9565 0.1185 0.1134 0.9893 2.45

Table 7
Parameters obtained from model (R = R1 [1 ! exp (!kt)]) fit to data set for coal and ash.

Downcomer immersion depth (cm) Coal Ash SI

R1 k Km R2 R1 k Km R2
30 0.9764 0.3890 0.3799 0.9989 0.8617 0.1890 0.1620 0.9959 2.34
50 1.1005 0.2554 0.2811 0.9888 1.0840 0.0676 0.0730 0.9896 3.85
1896 H. Vapur et al. / Energy Conversion and Management 51 (2010) 1891–1897

100 min had almost the same selectivity index (SI) values (2.12 and
90 2.45) and, selected as the limit values. 0.6 L/min was used as a mid-
80
dle wash water rate. The test results of combustible recovery and
ash content for the three water rates of froth washing were given
70
in Fig. 10. The best results obtained at 0.600 L/min wash water rate.
Recovery (%)

60 As it can be seen from the results, illustrated in the figure, which

50 ash content was decreasing from 33.41% to 13.68% giving a
40 90.17% combustible recovery.
30 30 cm Exp. Coal The last study for optimization of coal flotation was conducted
30 cm Model Coal
30 cm Exp. Ash on determination of optimum downcomer immersion depth. The
20 30 cm Model Ash
50 cm Exp. Coal former results of kinetic studies showed that the best value of
10 50 cm Model Coal
50 cm Exp. Ash depth was 50 cm. The middle value of depth, therefore, was se-
50 cm Model Ash
0 lected 50 cm. The depth values of 40 cm and 60 cm were deter-
0 1 2 3 4 5 6 7 8 9
mined to be used as the limit values of the downcomer depth.
Flotation Time (min.) The test results of combustible recovery and ash content for the
Fig. 6. Fitted to data set for downcomer immersion depth on coal and ash recovery.
four downcomer depths (30, 40, 50 and 60 cm) of the Jameson col-
umn were given in Fig. 11. It was showed that 40 cm as optimum
downcomer depth with 17.86% ash content and 94.83% combusti-
100 ble recovery was obtained at 8 min flotation time.
Combustible Recovery/Ash Content, %

90
80

70 100

60 Combustible Recovery, % Combustible Recovery/Ash Content, % 90

Ash Content, % 80
50
40 70

30 60
Combustible Recovery, %
20 50 Ash Content, %

10 40

0 30
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35
20
Particle size (mm)
10
Fig. 7. The effect of particle size on recovery, ash rejection and efficiency index 0
(froth height: variable; pH 7–7.5; pulp density: 5%; vegetable oil acids/kerosene 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6
ratio: 1/1; wash water rate: 0.600 L/min; downcomer immersion depth: 50 cm;
flotation time: 8 min).
Vegetable oil acids/kerosene ratio

Fig. 9. The effect of vegetable oil acids/kerosene ratio on recovery, ash rejection and
efficiency index (froth height: variable; pH 7–7.5; particle size: !0.250 mm; pulp
100 density: 20%; wash water rate: 0.600 L/min; downcomer immersion depth: 50 cm;
Combustible Recovery/Ash Content, %

flotation time: 8 min).

90

80

70

60
100
Combustible Recovery/Ash Content, %

50 Combustible Recovery, %
Ash Content, % 90
40
80
30
70
20
60
10 Combustible Recovery, %
50
Ash Content, %
0
0 5 10 15 20 25 40
30
Pulp density (%)
20
Fig. 8. The effect of pulp density on recovery, ash rejection and efficiency index
10
(froth height: variable; pH 7–7.5; particle size: !0.250 mm; vegetable oil acids/
kerosene ratio: 1/1; wash water rate: 0.600 L/min; downcomer immersion depth: 0
50 cm; flotation time: 8 min). 0 0.2 0.4 0.6 0.8 1 1.2
Wash water rate (L/min.)
vegetable oil acids/kerosene ratio was 1/1 with 93.60% as combus-
Fig. 10. The effect of wash water ratio on recovery, ash rejection and efficiency
tible recovery and 14.06% ash content for 8 min flotation time. index (froth height: variable; pH 7–7.5; particle size: !0.250 mm; pulp density:
For determination of optimum wash water rate firstly, kinetic 20%; vegetable oil acids/kerosene ratio: 1/1; downcomer immersion depth: 50 cm;
studies were carried out. The water rates of 0 L/min and 1.0 L/ flotation time: 8 min).
 H. Vapur et al. / Energy Conversion and Management 51 (2010) 1891–1897
100
Combustible Recovery/Ash Content, %
90
80
70
60
Combustible Recovery, %
50
Ash Content, %
40
30
20
10
0 [13] Cilek EC, Yilmazer BZ. Effect of hydrodynamic parameters on entrainment and
20 30 40 50
Downcomer immersion depth (cm)
Fig. 11. The effect of downcomer immersion depth on recovery, ash rejection and
efficiency index (froth height: variable; pH 7–7.5; particle size: !0.250 mm; pulp
density: 20%; vegetable oil acids/kerosene ratio: 1/1; wash water rate: 0.600 L/min;
flotation time: 8 min).
4. Conclusion
The results of this study are as follows:
) The results of the kinetic studies are in good agreement with the
first-order rate equation of the form R = R1 [1 ! exp (!kt)] for
coal and ash. The use of kinetic analysis for the evaluation of
the effect of variables on flotation separation has been demon-
strated. Firstly, modified rate constant and selectivity index
were all used to study coal flotation optimization in the Jameson
cell. Then the combustible recovery (%) and ash content (%) were
successfully utilized in developing new approach coal flotation
optimization.
) A product with 17.86% ash content and 94.83% combustible
recovery could be obtained after 8 min concentration time from
this type of coal flotation study at d80 = 0.250 mm particle size,
1/1 vegetable oil acids/kerosene ratio, 20% solids pulp density,
0.6 L/min wash water rate and 40 cm downcomer immersion
depts.
) It is expected that the coal flotation optimization work will pro-
vide a significant contribution to the field, i.e., that the work
either will provide something new to the field or will improve
some existing knowledge or methodology in the field.
References
[1] Aplan FF. The historical development of coal flotation in the United States. In:
Parekh BK, Miller JD, editors. Advances in flotation technology. SME; 1999. p.
269–87.
[2] Jameson GJ. A new concept in flotation machine design. Miner Metall Process
1988:44–7.
[3] Yan Y, Jameson GJ. Application of the Jameson cell technology for algae and
phosphorus removal from maturation ponds. Int J Miner Process
2004;73:23–8.
[4] Lazic P, Calic N, Boltzman S. Model of flotation kinetics. In: Proceedings of the
XXI international mineral processing congress 2000; B8a. p. 87–93.
1897
[5] Xu M. Modified flotation rate constant and selectivity index. Miner Eng
1998;11(3):271–8.
[6] Agar GE, Chia J, Requisc L. Flotation rate measurements to optimize an
operating circuits. Miner Eng 1998;11(4):347–60.
[7] Oliveira JF, Saraiva SM, Pimenta JS, Oliveira APA. Kinetics of pyrochlore
flotation from Araxa mineral deposits. Miner Eng 2001;14(1):99–105.
[8] Sripriya R, Rao PVT, Choudhury RB. Optimization of operating variables of fine
coal flotation using a combination of modified flotation parameters and
statistical techniques. Int J Miner Process 2003;68:109–27.
[9] Uçurum M, Bayat O. Effects of operating variables on modified flotation
parameters in the mineral separation. Sep Purif Technol 2007;55:173–81.
[10] Box GEP, Hunter WG, Hunter JS. Statistics for experiments. New York: Wiley;
1978.
[11] Akhanazarova S, Kafarov V. Experiment optimization in chemistry and
chemical engineering. Moscow: Mir Publishers; 1982.
[12] Rao GV, Mohanty S. Optimization of flotation parameters for enhancement of
grade and recovery of phosphate from lowgrade dolomitic rock phosphate ore
from Jhamarkota, India. Miner Metall Process 2002;19(3):154–60.
60 70 flotation performance. Miner Eng 2003;6:745–56.
[14] Martinez LA, Uribe SA, Carrillo PFR, Coreno AJ, Ortiz JC. Study of celestite
flotation efficiency using sodium dodecyl sulfonate collector: factorial
experiment and statistical analysis of data. Int J Miner Process 2003;70(1–
4):83–97.
[15] Rao TC, Pillai KJ, Vanangamudi M. Statistical analysis of coal flotation—a
prelude to process optimization. IX International coal preparation congress,
New Delhi; 1982. p. c2-1–12.
[16] Mohanty MK, Honakar HQ. Performance optimization of Jameson flotation
technology for fine coal recovery. Miner Eng 1999;12(4):367–81.
[17] Kelebek S, Demir U, Sahbaz O, Acar A, Cinar M, et al. The effects of
dodecylamine, kerosene, and pH on batch flotation of Turkey’s Tuncbilek
coal. Int J Miner Process 2008;88:65–71.
[18] Naik PK, Reddy PSR, Misra VN. Optimization of coal flotation using statistical
technique. Fuel Process Technol 2004;85:1473–85.
[19] Naik PK, Reddy PSR, Misra VN. Interpretation of interaction effects and
optimization of reagent dosages for fine coal flotation. Int J Miner Process
2005;75:83–90.
[20] Evans GM, Atkinson BW, Jameson GJ. The Jameson cell. In: Matis G, editor.
Flotation science and engineering. Marcel Dekker; 1995.
[21] Evans GM, Bin AK, Machniewski PM. Performance of confined plunging liquid
jet bubble column as a gas–liquid reactor. Chem Eng Sci 2001;56:1151–7.
[22] Harbort GJ, Cowburn JA, Manlapig EV. Recovery interactions between the froth
zone, pulp zone and downcomer within a Jameson cell. In: Membrey W, editor.
Proceedings of the 10th Australian coal preparation conference, Pokolbini;
2004. p. 91–101.
[23] http://www.ebicom.net/~dhyams/cvxpt.htm.
[24] Spedden HR, Hannan WS. Attachment of mineral particles to air bubbles in
flotation. Miner Technol 1984;12:TP 2354.
[25] Westhuizen AP, Deglon DA. Solids suspension in a pilot-scale mechanical
flotation cell: a critical impeller speed correlation. Miner Eng 2008:21621–9.
[26] Wojcik W, Janczuk B, Bialopiotrowicz T. The influence of an apolar collector on
the contact angle, detachment force and work of adhesion to the coal surface
in agglomeration flotation of low rank coal. Fuel 1990;69:207–10.
[27] Aplan FF. Coal flotation. In: Fuerstenau MC, editor. Flotation—AM gaudin
memorial volume, vol. 2. SME; 1976. p. 1235–64 [chapter 45].
[28] Mishra SK. Improved recovery of fine coal by flotation process. In: Mishra SK,
Klimpel RR, editors. Fine coal processing. USA: Noyes Publications; 1987. p.
110–35.
[29] Klimpel RR, Hansen RD. Chemistry of fine coal flotation. In: Mishra SK, Klimpel
RR, editors. Fine coal processing. Park Ridge (New Jersey, USA): Noyes
Publications; 1987. p. 78–109.
[30] Olson TJ, Aplan FF. The flotability of locked particles in a coal flotation system.
In: Proc 2nd int congr appl mineralogy in min industry; 1984. p. 367.
[31] Alonso MI, Valdes AF, Tarazona RMM, Garcia AB. Coal recovery from coal fines
cleaning wastes by agglomeration with vegetable oil acids effects of oil type
and concentration. Fuel 1999;78:753–9.
[32] Neethling SJ, Cilliers JJ. Simulation of the effect of froth washing on flotation
performance. Chem Eng Sci 2001;56:6303–11.
[33] Cowburn J, Harbort G, Manlapig E, Pokrajcic Z. Improving the recovery of
coarse coal particles in a Jameson cell. Miner Eng 2006;19:609–18.
[34] Tas!demir A, Tas!demir T, Öteyaka B. The effect of particle size and some
operating parameters in the separation tank and the downcomer on the
Jameson cell recovery. Miner Eng 2007;20:1331–6.