XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3 721
the range of 8 to 9, aligning with the natural pH of the
FPTs slurry. At natural pH levels, pH and mixing rates have
minimal impact on supernatant turbidity (Figure 8-b’’).
Elevated turbidity at extreme pH levels is attributed to the
partial dissolution of carbonate and clay minerals at low pH
values, as observed during flocculation and sedimentation
experiments. Additionally, the increased negative charge of
FPTs at high pH values amplify repulsion forces between
APAM and negatively charged particles (Bahmani-ghaedi
et al., 2022).
Confirmation Tests
An experimental confirmation test was conducted on a
laboratory scale, with three replicates, to confirm and vali-
date the optimal conditions recommended by the experi-
mental design software (numerical optimization) with a
desirability above 0.8 (OD =854). This high overall desir-
ability is an indication identifying the optimal conditions
that maximize or minimize multiple outcomes, facilitat-
ing a comprehensive approach to achieve the best overall
performance. The conditions of the confirmation tests are
as follows: a dose of 80 g/tds, a pH of 8.81 and a mix-
ing rate of 117 rpm. This validation aimed to ensure the
accuracy and reliability of the identified parameters, foster-
ing the most effective flocculation and sedimentation per-
formance concerning ISR, WR, and WT. The repetition
of the experiments in triplicate adds a layer of robustness
to the findings, strengthening the confidence in the sug-
gested optimum conditions and their applicability to real-
world scenarios. The results of three confirmation tests are
depicted in Table 8. The results of the three experiments
were averaged and the standard deviations were reported.
The confirmation test results validated the model,
exhibiting a close match with predictions from the experi-
mental design software. In Table 7, both the experimental
and predicted values of the studied responses are presented.
The experimental outcomes revealed 14.05 cm/min,
85.8%, and 0.97 NTU for ISR, WR, and water turbidity,
respectively, closely aligning with the predicted values of
13.792 cm/min, 85.161%, and 1.86 NTU for ISR, WR,
and water turbidity. This alignment affirms the robust-
ness of the experimental design, highlighting its efficacy in
describing the complex factors influencing the system reaf-
firming its utility in guiding optimized flocculation pro-
cesses for fine tailings dewatering. The minimal variations
observed can be attributed to experimental errors. Notably,
the experimental confirmation outperformed the software
predictions, with the supernatant turbidity being 0.89
NTU less than predicted. While there is a slight 0.639%
difference in WR, and a difference of 0.258 cm/min in the
ISR. The overall agreement underscores the reliability and
accuracy of the developed model. Since the order of impor-
tance for the studied responses is WR ISR WT, the
observed difference in water turbidity is acceptable, and the
model can be considered valid.
CONCLUSION
In this study, flocculation and settling parameters were
optimized for fine phosphate tailings (FPTs) slurry, mainly
comprising carbonates, silicates, fluorapatite, and clay min-
erals. Zeta potential measurements indicated a negative
charge at the slipping plane of FPTs particles. Among the
tested polyacrylamide flocculants, Magnafloc5250 demon-
strated the highest initial settling rate (ISR), water recovery
(WR), and acceptable supernatant turbidity. With using
Magnafloc5250, response surface methodology (RSM)
considered flocculant dose, mixing rate and pH as fac-
tors was employed. Cubic polynomial regression models
for ISR, WR, and turbidity were developed based on 20
experimental runs. Optimal parameters for each response
were determined by RSM, resulting in a flocculant dose of
80 g/tds, pH of 8.81, and mixing rate of 117 rpm, yielding
an ISR of 13.8cm/min, WR of 85.16% and turbidity of
1.86 NTU. Validation experiments affirmed the suitability
and efficacy of central composite design CCD in RSM for
optimizing flocculation and settling parameters for FPTs
slurry. The results reveal clear connections among floccula-
tion parameters, floc properties, and the dewatering effi-
ciency of FPTs.
Table 7. Results of the confirmation experiments
Experiment
ISR
(cm/min) WR (%)
Turbidity
(NTU) Flocs size Water pH
Conductivity
(µS/cm) TDS (ppm)
Conditions Flocculant dose =80 g/tds pH =8.811 MR =117.682 rpm
Average of 3
replicates
14.05 ± 0.09 85.8 ± 0.31 0.97 ± 0.05 Medium 8.72 ± 0.03 560.33 ±
16.26
404.33 ± 18.45
Predicted values 13.792 85.161 1.861 — — — —
the range of 8 to 9, aligning with the natural pH of the
FPTs slurry. At natural pH levels, pH and mixing rates have
minimal impact on supernatant turbidity (Figure 8-b’’).
Elevated turbidity at extreme pH levels is attributed to the
partial dissolution of carbonate and clay minerals at low pH
values, as observed during flocculation and sedimentation
experiments. Additionally, the increased negative charge of
FPTs at high pH values amplify repulsion forces between
APAM and negatively charged particles (Bahmani-ghaedi
et al., 2022).
Confirmation Tests
An experimental confirmation test was conducted on a
laboratory scale, with three replicates, to confirm and vali-
date the optimal conditions recommended by the experi-
mental design software (numerical optimization) with a
desirability above 0.8 (OD =854). This high overall desir-
ability is an indication identifying the optimal conditions
that maximize or minimize multiple outcomes, facilitat-
ing a comprehensive approach to achieve the best overall
performance. The conditions of the confirmation tests are
as follows: a dose of 80 g/tds, a pH of 8.81 and a mix-
ing rate of 117 rpm. This validation aimed to ensure the
accuracy and reliability of the identified parameters, foster-
ing the most effective flocculation and sedimentation per-
formance concerning ISR, WR, and WT. The repetition
of the experiments in triplicate adds a layer of robustness
to the findings, strengthening the confidence in the sug-
gested optimum conditions and their applicability to real-
world scenarios. The results of three confirmation tests are
depicted in Table 8. The results of the three experiments
were averaged and the standard deviations were reported.
The confirmation test results validated the model,
exhibiting a close match with predictions from the experi-
mental design software. In Table 7, both the experimental
and predicted values of the studied responses are presented.
The experimental outcomes revealed 14.05 cm/min,
85.8%, and 0.97 NTU for ISR, WR, and water turbidity,
respectively, closely aligning with the predicted values of
13.792 cm/min, 85.161%, and 1.86 NTU for ISR, WR,
and water turbidity. This alignment affirms the robust-
ness of the experimental design, highlighting its efficacy in
describing the complex factors influencing the system reaf-
firming its utility in guiding optimized flocculation pro-
cesses for fine tailings dewatering. The minimal variations
observed can be attributed to experimental errors. Notably,
the experimental confirmation outperformed the software
predictions, with the supernatant turbidity being 0.89
NTU less than predicted. While there is a slight 0.639%
difference in WR, and a difference of 0.258 cm/min in the
ISR. The overall agreement underscores the reliability and
accuracy of the developed model. Since the order of impor-
tance for the studied responses is WR ISR WT, the
observed difference in water turbidity is acceptable, and the
model can be considered valid.
CONCLUSION
In this study, flocculation and settling parameters were
optimized for fine phosphate tailings (FPTs) slurry, mainly
comprising carbonates, silicates, fluorapatite, and clay min-
erals. Zeta potential measurements indicated a negative
charge at the slipping plane of FPTs particles. Among the
tested polyacrylamide flocculants, Magnafloc5250 demon-
strated the highest initial settling rate (ISR), water recovery
(WR), and acceptable supernatant turbidity. With using
Magnafloc5250, response surface methodology (RSM)
considered flocculant dose, mixing rate and pH as fac-
tors was employed. Cubic polynomial regression models
for ISR, WR, and turbidity were developed based on 20
experimental runs. Optimal parameters for each response
were determined by RSM, resulting in a flocculant dose of
80 g/tds, pH of 8.81, and mixing rate of 117 rpm, yielding
an ISR of 13.8cm/min, WR of 85.16% and turbidity of
1.86 NTU. Validation experiments affirmed the suitability
and efficacy of central composite design CCD in RSM for
optimizing flocculation and settling parameters for FPTs
slurry. The results reveal clear connections among floccula-
tion parameters, floc properties, and the dewatering effi-
ciency of FPTs.
Table 7. Results of the confirmation experiments
Experiment
ISR
(cm/min) WR (%)
Turbidity
(NTU) Flocs size Water pH
Conductivity
(µS/cm) TDS (ppm)
Conditions Flocculant dose =80 g/tds pH =8.811 MR =117.682 rpm
Average of 3
replicates
14.05 ± 0.09 85.8 ± 0.31 0.97 ± 0.05 Medium 8.72 ± 0.03 560.33 ±
16.26
404.33 ± 18.45
Predicted values 13.792 85.161 1.861 — — — —