SIMULATION OF INDUSTRIAL RECOVERY RECOVERY BY HEAP HEAP LEACHING GOLD J. Tremolada, Tremolada, * R. Dzioba **and J. Menendez Menendez *** (*) Instituto Iberoamericano de Metalurgia Extractiva, Calle Marcahuasi Mz. E´1Lt.2, 3 Era Etapa, Urb.Portada Del Sol,La Molina,Lima 12, Peru Telephone : 511-3657308 E-mail:
[email protected] julio.tremolada@iberometex .com
(**) Departamento de Minería, Facultad de Ciencias Físicas y Matemáticas y Naturales. Universidad Nacional de San Luis, San Luis-Argentina E-mail :
[email protected] (***) Departamento de explotación y prospección de minas,Universidad de Oviedo, España, E-mail :
[email protected] [email protected] ABSTRACT This work contains contains the results of industrial industrial scale simulation/p simulation/predic rediction tion of Heap Leaching process for auriferous ores using data obtained by cyanidation in bottle roll tests. The simulation was made based on unreacted core models, in which which the diffusi diffusion on in the produc productt layer layer contro controls ls the reacti reaction on veloci velocity ty.. This This applie applies s for a diffus diffusion ion contro controlle lled d reacti reaction on where where the kinetic kinetic consta constant nt (k) is sensitive to the particle particle size. The model was developed developed for the determination of metallurgical recoveries of gold in the leach pad, obtained from bottle roll tests. The model allows the industrial prediction of gold recovery, from variables such as: recovery gold from bottle roll tests, leaching kinetic constant from bottle roll tests, particulate size from laboratory tests, industrial particulate size K50 from the leach pad, leaching time and the industrial kinetic constant from the leach pad. Keywor Keywords: ds: Cyani Cyanidat dation ion,, leach leaching ing,, gold, gold, kineti kinetic c consta constant, nt, partic particula ulate te size, size, simulation.
1. INTRODUCTION
important advantage of the heap leaching process is its flexibility: it
Precious metals extraction
have can be applied to crushed Run Of
been carried out for many years Mine (ROM), enclosed tailings or mostly by the traditional cyanidation relatively fine materials. In order to method,
which
presents
great evaluate the leaching behavior of a
selectivity and stability, and is also a material,
experiments
in
batch,
well-suited process for low grade small column or small heap are ores.
This
process
has
low often performed and the results are
operational cost and allows for the extrapolated by using rule of thumb extraction of a cyanide soluble metal to represent the industrial heaps or from the ore by dissolving its metal dumps. content
These
tests
provide
Malhotra, Klimpel and Mular
information about the leaching agent (1991) .
consumption,
the
maximum
In the Heap Leaching process, gold
recovery of the chemical species,
and silver extraction is carried out
and the rate of recovery; however,
by
the scale-up to industrial heaps or
percolating
solution
through
dilute an
cyanide ore
heap
large
columns
are
not
directly
(generally low grade ores) stacked
possible due to the difficulty of
on
reproducing
the
characteristics;
moreover,
impermeable
specifically solutions
to loaded
metals
( pregnant
cyanide
solution
pads
designed
collect
cyanide
with
precious
experiments
physical these
are time and cost
solution).
As
demanding, which stimulated the
percolates
the
development
heap, gold and silver dissolve. An
of
mathematical
models Andrade Lima, (2004) .
. 1.1 UNREACTED CORE MODEL
λ = stoichiometric factor
Case: Solid Spherical Particle
V = molar volume, V = P.M./ρ
Consider a mineral spherical particle
C
that reacts forming a porous solid
concentration in the solution)
=
concentration
(reagent
product, as indicated in Figure 1: D = diffusion coefficient in the porous layer
Solution Reaction Product
r 0
α = reacted mineral fraction
r Unreacted core
t = leaching time Diffusion of reactant inward
Soluble products
The
term
in
the
equation
(1):
2VDC/λr 02 = k, is well-known as the Figure 1. Unreacted core model . kinetic constant of the leaching 1 – 2/3α – (1 – α)2/3 = 2VDC/λr 02[t] Box,J.C. 1984,1986),Chae(1979),Dickson and
processes. The diffusion-controlled shrinking-
Hendrix
core kinetics model is shown:
(1993),levenspiel(1981), ,Prosser(1989),Roman
and
1 – 2/3α – (1 – α)2/3 = K[t] (2)
Benner(1974),Sanchez-chacon(1997 )
(1) Kinetic model, in which:
2. EXPERIMENTAL PROCEDURE In the execution of the present work Bottle
r 0 = initial radius of the mineral particle
Roll
granulometry
Tests
with
under
coarse extreme
leaching conditions were carried out
for
determining
both
maximum
the
kinetic of
model,
allow
the
behavior
the
possible extractions and recoveries
simulation
of
that can be obtained by the heap
precious metals extraction kinetics
leach process.
in a closed industrial circuit.
The objective of the present work is
2.1 MINERALOGY
to develop a systematic support The minerals present in the ores : methodology
for
industrial
scale
recovery prediction in heap leach circuits of oxidized auriferous ores. The present study procedures were
Table 1. Mineralogical characterization of ore samples. Minerals
Composition
Quartz
SiO2
the industrial leach pad. The study
CaSi2O8 Al2NaSi3O8 Al Ortopyroxene Silicate of Iron,Mg Zircon ZrSiO4 Sphene CaTiSiO5 Muscovite(Sericite) KAl2(OH, F)2 AlSi3O10 Clays Fosfosilicate of Al Hematite Fe2O3 Rutile TiO2 Goethite FeO(OH) Calcite CaCO3 Barite BaSO4 Anglesite PbSO4 Jarosite KFe3(OH)6(SO)2 Gold Au
contemplates the specification of
Table 2. Percentual mineralogical composition of the samples.
developed
to
predict
the
gold
recovery based on the parameters recovery of gold from bottle roll test, leaching
kinetic
constant
from
laboratory tests, particulate average size from laboratory test, industrial particulate size K50, leaching time and leaching kinetic constant from
experiments from bottle roll tests for the
determination
of
kinetic
parameters and the development of calculation methodologies that, with
Plagioclase
Minerals quartz Goethite Clays Others Pyrite
% 78 10 10 1 Trace 100
The bottle roll tests were carried out by duplicate, and the results, are shown in Table 4. CHEMICAL CHARACTERIZATION The Table 3. Ore samples head grades. Sample
Au (g/t) Antaya 1.020 Socabaya 0.658 Trinidad 0.782
Ag (g/t) 10.450 1.610 2.403
experimental
conditions
of
these tests were:
•
Mass of Ore: 2.00kg dry
•
Dilution (L/S): 1/1
•
Cyanide Force: 200 ppm
IN THE LEACH PAD
•
Slurry’s pH: 10.5
Table 4. Industrial Recovery in 60 days leaching.
•
Leaching Time: 1-2-4-8-12-
RECOVERY FROM INDUSTRIAL
Sample Antaya Socabaya Trinidad
Au (%) 67.85 70.15 61.88
24-48-72-96 hours The
parameters
that
were
determined and calculated are the 2.2
EXPERIMENTAL
METHODOLOGY The samples in their totality were sieved by mesh 2”, the product above 2” was crushed to < 2” in order to obtain uniformity in the sample (100% < 2”).
following ones:
•
Control
of
pH
cyanide
for
the
and
free
Pregnant
solution to 1-2-4-8-12-24-4872-96 hours, readjusting the cyanide as necessary in the extracted aliquot.
volume
of
the
•
Gold and silver analysis in
stays. Thus, before the reaction
pregnant solution for 1-2-4-8-
occurs it is necessary that the
12-24-48-72-96
diffusion of reagents through the
hours
reacted material layer takes place (a
leaching.
first order reaction on the surface of •
Gold and silver concentration the particle’s unreacted in
readjusted
core
is
Pregnant assumed). The model is based on
solution. the supposition that the mineral is •
Gold and silver distribution
isotropic and that both leachants
for
and products are transported by the
1-2-4-8-12-24-48-72-96
hours leaching.
solution that fills completely the particle pores. In addition, it is
•
Cyanide
and
lime assumed
that
the
solution
is
consumptions and gold and sufficient to completely cover and to silver head grade. circulate around the external surface 2.2.1CALCULATION
of the mineral particle.
METHODOLOGY 2.2.2 UNREACTED CORE MODEL A simple way to understand the Model
Applied
in
Experimental
heterogeneous kinetic mechanisms Tests For Industrial Simulation: is using the unreacted core model, described by Levenspiel (1981). It
1 – 2/3α – (1 – α)2/3 = K’[t]
assumes that when advancing the
Model
Kinetic
reaction front through a spherical α = Recovery Gold from bottle roll particle, it keeps its form and size, tests when the reacted mineral layer
K lab = Leaching Kinetic constant
from bottle roll tests
Take the results
obtained
from the recovery of gold from bottle roll test.
X lab = Particulate average size from bottle roll tests
Through
a
granulometric
analysis of the sample the X 50 ind = Industrial particulate Size average
size
of
mineral
X50 in the leach pad particles is determined in the t = Leaching Time K ind( dia-1 )= Industrial
laboratory. Kinetic
constant in the leach pad
For each lixiviation process there is a partial recovery of gold at a laboratory level,
where
therefore Kind( dia-1 )= K lab x( X lab / X 50 ind )2
in
the
equation
1 – 2/3α – (1 – α)2/3 = K’[t] the value of α is replaced by the
X = particle granulometric size
partial recovery of gold. The use of Excel Software solver for prediction of the industrial recovery
Later the partial values for
gold of the leach pad have relevant
the model of each lixiviation
importance.
process are graphed and by the slope the kinetic constant
Methodology of the used unreacted
of the laboratory lixiviation
core model
can be determined (K lab).
Later through granulometric analysis of the mineral fed to
a cell or industrial module of
from
the
the leach pad the X 50
1 – 2/3α – (1 – α)2/3 .
model
industrial of the leach pad
Finally, using the correlations
can be determined. of KT and of the model the
Next the kinetic constant of
gold recovery rate of the
the industrial lixiviation on the
leach can be predicted using
leach pad is determined by
Excel Solver software.
the relation Kind ( dia-1 )= K lab x( X lab / X 50 ind )2 .
For each projected lixiviation process on the leach pad the corresponding
projected
value of K ind x T can be determined.
The
The independent gold recoveries for each bench in cyanidation tests are shown in Table 5.
value of each K T is
correlated
3. RESULTS
with
the
corresponding value obtained
AVERAGE RECOVERY GOLD FROM BOTTLE ANTAYA: 89.10% Gold
AVERAGE RECOVERY GOLD FROM BOTTLE SOCABAYA: 85.01% Gold
AVERAGE RECOVERY GOLD FROM BOTTLE TRINIDAD: 78.53% Gold
Simulated Recovery Au % 80
r e v o c e R %
70 60 50 40 30 20 10 0 0
20 40 Leaching Time ( Days ) Trinidad
Socabaya
60 Antaya
Figure 2. Results of the Industrial Simulation based on bottle roll cyanidation tests.
4. OBSERVATIONS AND CONCLUSIONS Table 6. Summary gold recovery bottle roll, industrial leach pad, and simulation with unreacted core model
ORE BENCH ANTAYA SOCABAYA TRINIDAD
Recovery Au (%) Bottle Roll Test 89.10 85.01 78.53
Recovery Au (%) Industrial Leach pad 67.85 70.15 61.88
Recovery Au (%) Simulation leach pad gold recovery 67.34 70.06 61.58
Percentage of difference of the gold recovered between the industrial leach pad and the simulated leach pad with unreacted core model 0.75 0.13 0.48
A model methodology, that integrates information from bottle roll cyanidation tests, such as recovery gold from bottle roll tests , leaching kinetic constant from laboratory leaching test, particulate average size from laboratory test, industrial particulate size k50 in the leach pad, leaching time and the industrial kinetic constant in the leach pad, ore characteristics , has been developed . It allows us to simulate and predict precious metals industrial recoveries in the operation of a closed circuit of leaching in heap. The process simulation given by the model demonstrates its technical feasibility to be applied in industrial heap leaching circuits. The forecasted industrial recoveries for Antaya (67,34 %), Socabaya (70,06 %) and Trinidad (61,58 %) were very similar to the recoveries industrial with 0.75%,0.13%,and 0.48% of
margen of
difference respectively. In relation to the bottle roll cyanidation tests with coarse granulometry, which
determine maximum possible extraction under extreme
leaching conditions, average gold recoveries as high as 89,10%, 85,01%, 78,53% for Antaya, Socabaya, and Trinidad respectively were obtained. The effect of the sample composition of the ore bench has relevant high importance.
the simulations of an industrial gold heap leaching from bottle roll test showed that the model can successfully predict the general features of the process time evolution and can be used to general design studies.
ACKNOWLEDGEMENTS
The authors wish to thank Institute Iberoamerican of Extractive Metallurgy , and University of San Luis, Argentina, and the department of exploitation and exploration of mines,University of Oviedo, Spain for the support on this project and the permission to publish this paper. REFERENCES
[1] Andrade Lima L.R.P.(2004), A mathematical model for isothermal
heap and
column leaching .Brazilian journal of chemical engineering, vol 21,no 3, Sao Paulo, July/Sept . [2] Auck,Y. And Wadsworth, .(1974), Physical and Chemical factors in Copper
Dump leaching. Solutions Mining Symposium, AIME . [3]Bartlett R.W.(1992), Solution Mining: Leach and Fluid Recovery of Materials.
Gordon and Breach. Science Publishers, Reading, U.K. [4]Box J.C. and Prosser A.P.(1986), "A general model forv the reaction of several
minerals and several reagents in heap and dump leaching". Hydrometallurgy 16: 77-92. [5] Box,J.C. and Yusuf,R. (1984).. Simulation of heap and dump leaching process,
Proceedings of the Symposium on Extractive metallurgy,Melbourne,Australia, p.117-124 [6]Brimhall D.B. Wadsworth M.E.(1973), Oxygen Consumption in Dump Leaching.
Transactions, Vol. 254, pp. 68-75, Society of Mining Engineers, AIME .
[7]Chae, D. And Wadsworth, M.(1979), Modeling of the Leaching of Oxide Copper
Ores. US Bureau of Mines.. [8]Dickson D.G. and. Hendrix J.L(1993). A general model for leaching of one or
more solid reactants from porous ore particles. Metall.trans.B 24B: 157-169. [9] Dickson D.G. and. Hendrix J.L(1993). A mathematical model for heap leaching
of one or more solid reactants from porous ore pellets. Metall.Trans.B 24B: 10871102. [10]Kelley,R.J. Zárate,G.(1995), Heap leaching of copper oxide ores - scale-up
between laboratory and pilot-plant results. Cobre 95 vol 3, 743-755 .
[11]Levenspiel O.(1981), "Ingeniería de las Reacciones Químicas". 2ª ed., Reverte
S.A., Barcelona, España, pp. 385-415. [12]Malhotra, D.,Klimpel, R.R., Mular, A.L(1991). “ Evaluation and optimization of
Metallurgical Performance “. Society of Mining,Metallurgy and Exploration Inc Littleton ,Colorado. [13]Petersen,J and Dixon D.G. (2002). Systematic modelling of heap leaching
processes for optimisation and design, Proceedings of the EPD Congress 2002 and Fundamentals of Advanced Materials for Energy Conversion, Edited by Taylor, P.R., Chandra, D., and Bautista, R.G., TMS, p.757-771. [14]Prosser A.P. (1989), "Simulation of gold heap leaching as an aid to process
development. Precious Metals 89 ,TMS,Warrendale,P.A. 121-135 . [15]Roman R.J., B.R. Benner and G.W. Becker(1974), "Diffusion Model for Heap
Leaching and its Applications to Scale-Up". Trans. Soc. Min. Engrs. AIME, Vol. 256, p. 247-252. [16]Sanchez-Chacon and Lapidus G.T.(1997),"Model for Heap Leaching
of gold
ores by cyanidation". Hydrometallurgy 44: 1-20 . [17]Shafer J.L., M.L. White and C.L. Caenepeel(1979), "Application of the Shrinking
Core Model for Copper Oxide Leaching". Min. Eng. Vol 31(2), p. 165-171.
TABLES AND FIGURE LEYENDS
Table 1. Mineralogical characterization of ore samples. Table 2. Percentual mineralogical composition of the samples. [2] Table 3. Ore samples head grades. Table 4. Industrial Recovery in 60 days leaching. Figure 1. Unreacted core model. Figure 2. Results of the industrial simulation based on bottle roll cyanidation tests
