XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3 1887
parameters constant. It is evident that the leaching temper-
ature and catalyst dosage interplayed significantly for nickel
recovery, and that leaching temperature and catalyst dosage
clearly dominated other parameters. It can be observed that
the concentration of sulfuric acid didn’t significantly influ-
ence the recovery as expected.
Process Optimization and Model Validation
For every response, the Minitab RSM-based optimization
program was used to determine the parameter ranges that
would yield the best recovery of the important minerals.
Based on the examined experimental data, optimization
studies were conducted. The objective was to increase the
leaching recoveries by changing the process factors at a
set desirability, as given in Table 12. As a result, the soft-
ware produced five highly desirable ideal options for every
response (see Table 13). For every response, two leaching
tests were conducted, considering the solution with the best
leaching recoveries and the solution with the best process
conditions. Table 14 summarizes the experimental recov-
eries. The validation tests strongly agreed with the antici-
pated value, demonstrating the attractiveness, correctness,
and dependability of the RSM model. Overall, the valida-
tion investigations showed that the RSM models could be
Table 12. The numerical process optimization for the desired response within variable range
Process Response Goal Lower limit Upper limit
Grind time − In range 0 10
Catalyst dosage − In range 0 30
H
2 SO
4 concentration − In range 0 1
Leaching time − In range 4 12
Leaching temp. − In range 20 80
Ni and Cu recoveries (%)Maximize − 98
Table 13. The solutions obtained for the numerical optimization of Ni and Cu
Run
Order
Grind Time,
mins
Catalyst
Dosage, wt.%
H
2 SO
4 Concentration,
M
Leaching
Time, h
Leaching
Temp., °C
Response,
%Desirability
Ni
1 7.9798 30 0.59596 12 80 97.0675 0.88344
2 9.08342 30 0.593405 12 80 96.8807 0.86009
3 8.68526 29.9725 0.626672 11.9681 79.9954 96.8807 0.86008
4 9.68763 30 0.581449 12 80 96.6311 0.82888
5 9.9567 30 0.56715 12 80 96.4853 0.81066
Cu
1 0.9246 14.9536 0.75711 12 80 98 1
2 7.7945 14.1361 1 4 80 98 1
3 1.3586 30 0.23704 8.0068 80 98 1
4 9.7279 30 0.15405 7.8476 80 98 1
5 10 29.7441 0 11.9954 77.8803 98 1
Table 14. The optimum process variables, model prediction and validation recoveries
Element Selected Run
Predicted Recovery,
%
Experimental
Recovery, %Error, %
Ni 1 97.07 98.92 1.87
3 96.88 93.68 3.42
Cu 1 98 99.14 1.15
2 98 98.27 0.27
parameters constant. It is evident that the leaching temper-
ature and catalyst dosage interplayed significantly for nickel
recovery, and that leaching temperature and catalyst dosage
clearly dominated other parameters. It can be observed that
the concentration of sulfuric acid didn’t significantly influ-
ence the recovery as expected.
Process Optimization and Model Validation
For every response, the Minitab RSM-based optimization
program was used to determine the parameter ranges that
would yield the best recovery of the important minerals.
Based on the examined experimental data, optimization
studies were conducted. The objective was to increase the
leaching recoveries by changing the process factors at a
set desirability, as given in Table 12. As a result, the soft-
ware produced five highly desirable ideal options for every
response (see Table 13). For every response, two leaching
tests were conducted, considering the solution with the best
leaching recoveries and the solution with the best process
conditions. Table 14 summarizes the experimental recov-
eries. The validation tests strongly agreed with the antici-
pated value, demonstrating the attractiveness, correctness,
and dependability of the RSM model. Overall, the valida-
tion investigations showed that the RSM models could be
Table 12. The numerical process optimization for the desired response within variable range
Process Response Goal Lower limit Upper limit
Grind time − In range 0 10
Catalyst dosage − In range 0 30
H
2 SO
4 concentration − In range 0 1
Leaching time − In range 4 12
Leaching temp. − In range 20 80
Ni and Cu recoveries (%)Maximize − 98
Table 13. The solutions obtained for the numerical optimization of Ni and Cu
Run
Order
Grind Time,
mins
Catalyst
Dosage, wt.%
H
2 SO
4 Concentration,
M
Leaching
Time, h
Leaching
Temp., °C
Response,
%Desirability
Ni
1 7.9798 30 0.59596 12 80 97.0675 0.88344
2 9.08342 30 0.593405 12 80 96.8807 0.86009
3 8.68526 29.9725 0.626672 11.9681 79.9954 96.8807 0.86008
4 9.68763 30 0.581449 12 80 96.6311 0.82888
5 9.9567 30 0.56715 12 80 96.4853 0.81066
Cu
1 0.9246 14.9536 0.75711 12 80 98 1
2 7.7945 14.1361 1 4 80 98 1
3 1.3586 30 0.23704 8.0068 80 98 1
4 9.7279 30 0.15405 7.8476 80 98 1
5 10 29.7441 0 11.9954 77.8803 98 1
Table 14. The optimum process variables, model prediction and validation recoveries
Element Selected Run
Predicted Recovery,
%
Experimental
Recovery, %Error, %
Ni 1 97.07 98.92 1.87
3 96.88 93.68 3.42
Cu 1 98 99.14 1.15
2 98 98.27 0.27