XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3 713
over time to calculate the settling rate of the flocculated
suspension. The initial settling rate (ISR) of particles was
determined from the slope of the linear part of mudline
interface height versus sedimentation time. Following a
20-minute settling period, an aliquot of the supernatant
was extracted for turbidity measurement using a turbi-
dimeter (Hanna Instruments HI-88713-02), reported in
nephelometric turbidity units (NTU). The sedimented part
served to determine the final solid content of the thickened
tailings. The global methodology used for the current study
is highlighted in Figure 1.
Experimental Design and Statistical Analysis
The design of experiments (DOE) and the optimization
of the flocculation-sedimentation process were conducted
using Design Expert ® 13.0.5.0 software (Stat-Ease Inc.,
USA). A central composite design (CCD) was employed
to investigate the effect of flocculant dosage, pH and stir-
ring speed on the flocculation and sedimentation effec-
tiveness (ISR, water recovery and turbidity). The CCD is
well-suited for assessing both the primary effects of each
variable and the interactions between the studied factors, all
within a constrained number of experiments (Ait-khouia et
al., 2022).
A face centred CCD (α =1) comprises 2n factorial runs
(e.g., –1, –1, –1), n axial runs (e.g., α, 0, 0), and nc center
runs (0, 0, 0), resulting in a total of N runs. This study
employed a comprehensive 23 factorial design (8 points),
complemented by 2 axial points per factor at a distance ±
α from the design center (6 points) and replicates of the
design center (6 points). In total, 20 experimental runs
were carried out during the CCD experiments (Eq. 1).
*N n 2 2n 3 2 3 6 20 n
c =++=++=(1)
An empirical model correlating the flocculation effi-
ciency parameters with dewatering process variables was
developed using a cubic function obtained for the study
variables, as presented in Eq 2.
Y x x x x
x x x x x x
x x x x x x
x x x2
i i
i
ii i ij i
i i i
j
i
iii i
i
0
1
3
2
1
3
1
3
1
3
3
123 1 2 3 1 1
2
2
1
3
1 3 1
2
3 2l1 2
2
1 2l3 2
2
3
3 3
2
3
2
1 1
b b b b
b b b
b b b
b b3l2x f
=+++
+++
+++
+++
=====
=
l2
l
l
////
/
(2)
where Y stands for the response variable, while the propor-
tions of the mix ingredients are denoted as xi and xj. The
coefficients for regression are represented as βi, βii, βij and
βiii (constant term, linear terms, interaction terms, qua-
dratic terms and cubic terms) ε is the residual associated
to the experiments. Supplementary information regarding
the measurements and their corresponding responses are
depicted in Table 2. The assessment of the model’s qual-
ity relied on the coefficient of determination (R2) and the
statistical significance of the model was evaluated through
analysis of variance (ANOVA). Coding the studied factors
facilitated the transformation of actual values into dimen-
sionless coordinates. These coded variables were assigned
values of –1, 0, and +1, representing the lowest, central, and
maximum limits for each studied variable. This approach
eliminates the influence of variable magnitudes, enabling
the combination of factors on a dimensionless scale (Sun
et al., 2010).
RESULTS AND DISCUSSION
Characterization Results
The particle size distribution (PSD) curve for the FPTs is
illustrated in Figure 1. It is evident from the graph that
80% of the sample exhibits a particle size smaller than
103 µm, while 50% of the sample is smaller than 53 µm.
Furthermore, 10% of the sample is smaller than 9.4 µm.
These findings suggest that our sample is primarily com-
posed of clay and silt-rich material.
The chemical composition of the FPTs sample, deter-
mined through the XRF method, is presented in Table 3. As
indicated, the sample is notably abundant in CaO, SiO2,
Table 2. Experimental process variables, units, levels, corresponding natural and coded values
Variable Units
Levels
Performance Parameters −1 0 +1
X1: Dosage g/tds 40 70 100 Initial settling rate (cm/min)
Water recovery (%)
Water turbidity (NTU)
X
2 :pH — 6 8 10
X
3 :Mixing rate rpm 60 100 140
over time to calculate the settling rate of the flocculated
suspension. The initial settling rate (ISR) of particles was
determined from the slope of the linear part of mudline
interface height versus sedimentation time. Following a
20-minute settling period, an aliquot of the supernatant
was extracted for turbidity measurement using a turbi-
dimeter (Hanna Instruments HI-88713-02), reported in
nephelometric turbidity units (NTU). The sedimented part
served to determine the final solid content of the thickened
tailings. The global methodology used for the current study
is highlighted in Figure 1.
Experimental Design and Statistical Analysis
The design of experiments (DOE) and the optimization
of the flocculation-sedimentation process were conducted
using Design Expert ® 13.0.5.0 software (Stat-Ease Inc.,
USA). A central composite design (CCD) was employed
to investigate the effect of flocculant dosage, pH and stir-
ring speed on the flocculation and sedimentation effec-
tiveness (ISR, water recovery and turbidity). The CCD is
well-suited for assessing both the primary effects of each
variable and the interactions between the studied factors, all
within a constrained number of experiments (Ait-khouia et
al., 2022).
A face centred CCD (α =1) comprises 2n factorial runs
(e.g., –1, –1, –1), n axial runs (e.g., α, 0, 0), and nc center
runs (0, 0, 0), resulting in a total of N runs. This study
employed a comprehensive 23 factorial design (8 points),
complemented by 2 axial points per factor at a distance ±
α from the design center (6 points) and replicates of the
design center (6 points). In total, 20 experimental runs
were carried out during the CCD experiments (Eq. 1).
*N n 2 2n 3 2 3 6 20 n
c =++=++=(1)
An empirical model correlating the flocculation effi-
ciency parameters with dewatering process variables was
developed using a cubic function obtained for the study
variables, as presented in Eq 2.
Y x x x x
x x x x x x
x x x x x x
x x x2
i i
i
ii i ij i
i i i
j
i
iii i
i
0
1
3
2
1
3
1
3
1
3
3
123 1 2 3 1 1
2
2
1
3
1 3 1
2
3 2l1 2
2
1 2l3 2
2
3
3 3
2
3
2
1 1
b b b b
b b b
b b b
b b3l2x f
=+++
+++
+++
+++
=====
=
l2
l
l
////
/
(2)
where Y stands for the response variable, while the propor-
tions of the mix ingredients are denoted as xi and xj. The
coefficients for regression are represented as βi, βii, βij and
βiii (constant term, linear terms, interaction terms, qua-
dratic terms and cubic terms) ε is the residual associated
to the experiments. Supplementary information regarding
the measurements and their corresponding responses are
depicted in Table 2. The assessment of the model’s qual-
ity relied on the coefficient of determination (R2) and the
statistical significance of the model was evaluated through
analysis of variance (ANOVA). Coding the studied factors
facilitated the transformation of actual values into dimen-
sionless coordinates. These coded variables were assigned
values of –1, 0, and +1, representing the lowest, central, and
maximum limits for each studied variable. This approach
eliminates the influence of variable magnitudes, enabling
the combination of factors on a dimensionless scale (Sun
et al., 2010).
RESULTS AND DISCUSSION
Characterization Results
The particle size distribution (PSD) curve for the FPTs is
illustrated in Figure 1. It is evident from the graph that
80% of the sample exhibits a particle size smaller than
103 µm, while 50% of the sample is smaller than 53 µm.
Furthermore, 10% of the sample is smaller than 9.4 µm.
These findings suggest that our sample is primarily com-
posed of clay and silt-rich material.
The chemical composition of the FPTs sample, deter-
mined through the XRF method, is presented in Table 3. As
indicated, the sample is notably abundant in CaO, SiO2,
Table 2. Experimental process variables, units, levels, corresponding natural and coded values
Variable Units
Levels
Performance Parameters −1 0 +1
X1: Dosage g/tds 40 70 100 Initial settling rate (cm/min)
Water recovery (%)
Water turbidity (NTU)
X
2 :pH — 6 8 10
X
3 :Mixing rate rpm 60 100 140