Resin for Gold Extraction.pdf
(
335 KB
)
Pobierz
Cognis
Corporation
Mining Chemicals Technology
Ion-Transfer Chemicals Technology
ISO 9002 Certified
Cognis AuRIX
100 Resin for Gold Extraction
Engineering Cost Study and Pilot Plant Investigations
by
G.T. Fisher, R.G. Lewis, and M.J. Virnig
Cognis Corporation
J. M. W. Mackenzie
Cognis Australia Pty Ltd
M. R. Davis
Lycopodium Pty Ltd.
Table of Contents
1.
Abstract
1
2.
Introduction
1
3.
Metallurgical Investigations
2
4.
Pilot Plant Operation
7
5.
Cost Estimation
10
6.
Summary
12
7.
Acknowledgements
13
8.
References
14
ABSTRACT
Preliminary capital and operating cost estimates (±25%) comparing conventional carbon-in-
solution (CIS) with resin-in-solution (RIS) using AuRIX® 100 for the recovery of gold and
silver from typical heap leach solutions have been presented previously (Davis
et. al.
,
1999). The objective of this paper is to update this cost comparison based on additional
laboratory testwork and pilot scale test work at a heap leach operation in Sonora, Mexico.
Preliminary bench-scale testwork was carried out with AuRIX® 100 using real and
synthetic leach solutions to determine the technical viability of the resin for gold/silver
recovery and to provide process design parameters for the preliminary RIS circuit. The
scope of the resin testwork was sufficient to support the level of accuracy applied in the
capital and operating cost estimates. No parallel testwork was carried out with activated
carbon. Industry standard design data and operating data from existing CIS operations
were used as a basis for the comparison.
Both the capital and operating cost comparisons indicate a substantial savings for a RIS
circuit using AuRIX® 100 (~ 16 to 17% in terms of capital and 27 to 40% in terms of
operating costs) over a conventional CIS circuit depending upon solution flows and grade.
Dore analysis indicated that selectivity of the AuRIX® resin for gold and silver over copper
(the major base metal present) was favorable compared to the production plant using CIS
at the mine in Mexico. Dore produced by the AuRIX® 100 resin in the pilot plant assayed
as 90.01 % Au, 9.37 % Ag, and 0.62% Cu compared to 68.7% Au, 25.9% Ag, and 5.5% Cu
produced in the CIS production plant.
1. INTRODUCTION
AuRIX® 100 is a weak base ion exchange resin developed by Cognis for the recovery of
gold from typical cyanide leach liquors (Kordosky
et al
, 1993, Virnig
et al
, 1996). It is a
typical styrene- divinylbenzene resin bead functionalized with a guanidine functional group.
Guanidines are very strong organic bases having an intermediate basicity between that of
simple amines and quaternary amines. Simple amines are not sufficiently basic to be
protonated at the typical pH values of cyanide-leach liquors (pH 9 to 11) and are not
effective gold extractants. Quaternary amines are extremely strong organic bases since
they carry a permanent positive charge and are very effective gold extractants. Due to this
permanent positive charge, the stripping of aurocyanide from quaternary amine based
resins requires substitution by competitive ion exchange with another anion such as
Zn(CN)
4
2-
or conversion of the aurocyanide to a cationic thiourea or thiosulfate complex
that no longer associates with the positively charged quaternary amine functionality. Both
of these approaches require treatment of the resin with acid, necessitating the handling of
hydrogen cyanide. This also results in wide swings in pH and ionic strength between the
extraction and stripping stages, contributing to osmotic shock and increased resin
degradation.
The guanidine functionality is sufficiently basic to deprotonate water to form a guanidinium
cation. This cation can form an ion-pair with aurocyanide resulting in gold extraction from
solution. By increasing the basicity of the aqueous phase the guanidinium cation is
converted to the neutral guanidine functionality. The neutral guanidine functionality no
longer forms an
ion-pair with aurocyanide resulting in gold stripping from the resin. Since
both the extraction and stripping stages are carried out under alkaline conditions, AuRIX®
resin does not undergo significant volume changes (2-3%) between extraction and
stripping minimizing the potential for osmotic shock. In a variety of testwork, AuRIX® 100
resin beads have been shown to be very tough and resistant to abrasion and attrition
(Virnig
et al,
1996). In addition, the large diffuse nature of the guanidinium cation results in
greater selectivity for aurocyanide over other base metal cyanides than typical quaternary
amine based extractants.
The objective of this paper is to evaluate the potential for replacing activated carbon with
AuRIX® 100 for the recovery of gold/silver from pregnant leach solutions (PLS) typical of
heap leach operations. To provide a quantitative comparison between the two
technologies, two operating points representing the extremes in solution flow and grade
generally experienced in heap leach operations (See Table 1) were selected as the basis
for the CIS and the RIS circuit designs.
Table 1. Basis of Comparison
Case
Circuit
Type
PLS
Grade
(mg
Au/L)
PLS
Flow
(m
3
/hr)
Annual Au
Production
(oz/yr)
1
CIS
0.5
400
50,000
2
CIS
2.0
100
50,000
3
RIS
0.5
400
50,000
4
RIS
2.0
100
50,000
The parameters used in the RIS circuit designs have been based on the results of
preliminary resin testwork and knowledge of conventional ion exchange circuits. The
parameters used in the CIS circuit designs have been based on industry standard values
and CIS plant operating data.
2. METALLURGICAL INVESTIGATIONS
2.1 Hydraulics
The low bulk density of the resin and the regular spherical shape results in a low specific
fluidization velocity limiting the upflow flowrate that can be passed through a resin
contactor. Initial bed fluidization occurred at an approximate specific upflow velocity of 3.5
m
3
/m
2
hr. Specific upflow flowrates up to 20 m
3
/m
2
/hr can be used if sufficient provision is
made in the contactor design to allow for the resulting bed expansion to prevent resin
carryover.
In a downflow configuration, the pressure drop across the resin bed was very low (< 1.5
kPa/m at 100 BV/hr) allowing the circuit to be designed for a high specific downflow
flowrate with a low pressure drop and minimal bed compaction.
Based on these results, the preliminary RIS circuit design incorporates downflow for
extraction and upflow for elution. The downflow configuration for extraction allows the RIS
circuit to operate over a wide range of flowrates without the concern over bed expansion
limitations. The upflow configuration for elution will permit the flushing of any trapped
particulate material from the resin column and partially expand the bed to prevent
compaction and potential short-circuiting.
2.2 Freundlich Isotherm
Measured volumes of high grade (200 ml) and low grade (1,000 ml) pregnant leach
solutions (PLS) were contacted with five different masses of pre-conditioned resin for a
period of 48 hours to ensure equilibrium was attained. The resin had been pre-conditioned
with dilute caustic solution and pre-loaded to ~600 g Au/t dry resin to simulate an
anticipated plant barren resin. The residual solutions were then assayed for metal values
and free cyanide. For the Mexico mine leach solution, 4 resin masses were contacted each
with 19.0 L of solution pumped through a small column for 40 hours. Residual solutions
were assayed for metal values. The resultant resin and solution data was then used to
determine the equilibrium isotherm, which can be described by the following Freundlich
equation:
[Au]
Resin
= a [Au]
Soln
b
The Freundlich constants and loading capacity of the resin for the low grade, high grade
and Mexico mine heap leach solution are summarized in Table 2.
Table 2. Summary of Freundlich Constants
Data
Freundlich Constants
Loading Capacity
(gAu/t)
1
a
b
Mexico Mine
solution
(0.42 mg Au/L)
1,218
0.456
2,364
Low Grade
(1.56 mg Au/L)
2,858
0.458
10,097
High Grade
(7.82 mg Au/L)
2,323
0.304
16,550
1
Dry weight of resin basis
The AuRIX® resin has very good loading characteristics at low gold solution grades (< 1.0
mg/L Au) resulting in high loaded resin grades (up to 7,000 g Au/dry tonne of resin).
2.3 Loading Kinetics
The high grade and low grade PLS were contacted with a measured mass of pre-
conditioned resin in bottlerolls. The solutions were sampled at 0.5, 1, 2, 4, 8 and 12 hours
and assayed for Au, Ag, Cu and Zn. Loading kinetics were also done on feed solution from
the heap leach operation in Mexico. The data was then plotted as ln [Au]
Resin
versus ln t
resulting in a straight line from which the Nicol-Fleming rate constant, k, and the equilibrium
loading factor, n, could be derived (Nicol
et al
, 1984). The Nicol-Fleming rate equation is
commonly used to model carbon circuits. It is:
[Au]
Resin
= k[Au]
Soln
t
n
where [Au]
Resin
is the concentration of gold on the resin in g/t and [Au]
Soln
is the time
weighted average gold concentration in solution in mg/L.
The Nicol-Fleming rate constants for the resin at the two solution grades, actual measured
rate constants for carbon derived from plant data and the typical values used in the design
of CIS circuits are shown in Table 3.
Table 3. Nicol-Fleming Rate Constants
“k” (h
-1
)
Case
“n”
AuRIX® @ 1.2 g resin /L (high grade)
AuRIX® @ 0.24g resin/L (low grade)
AuRIX® @ 0.23g resin/L (Mexico Mine)
Fairview CIP Data
Activated C Design
201
449
645
260
150
0.33
0.37
0.51
0.52
0.6
The low “n” values for the two resin bench scale solution tests indicate that the adsorption
rate of gold onto the resin was limited by the equilibrium loading capacity of the resin (i.e.
insufficient resin was used in the kinetic tests). Without consideration of any effect that the
equilibrium loading capacity may have had on extraction kinetics of the resin, the results
indicate that AuRIX® has significantly faster extraction kinetics than activated carbon as
shown in Figure 1.
4000
3500
3000
Mexico Mine
2500
Carbon Data
2000
Carbon Design
1500
Low grade (1.56 mg Au/L)
1000
500
0
0
5
10
t hr
Figure 1. Resin versus Carbon Rate Curves
2.4 Continuous RIS Trial- Bench Scale
2.4.1 Extraction
Nine perspex columns were filled with pre-conditioned resin to give a resin bed volume of
29 mls (24.55 mm x 61 mm) in each column. Four of the columns were connected in
series in a downflow configuration. The remaining columns were held separate to be used
as the tails column at the end of each loading cycle. A synthetic PLS (~1.9 mg Au/L) was
pumped through the four columns in a downflow direction at 60 bed volumes per hour (29
ml/min). The circuit was pre-loaded for 48 hours prior to the
start of the 5-day trial. The
bench scale pilot plant was operated in a carousel mode with the lead column
being removed from the circuit and a fresh column being added to the tails position.
The second column became the lead column.
start of the 5-day trial. During the trial,
the circuit was operated on a 24 hour cycle time: the lead column was removed and a fresh
tails column added to the circuit every 24 hours. The final cycle was extended from 24 to
66 hours to evaluate the maximum operating time of the 4 column circuit before
breakthrough was observed on the tails column. The circuit solution profiles for Day 5 and
for the breakthrough test are summarized in Figures 2 and 3 respectively.
The gold tenor of the solution exiting the tails column was consistently less than 0.05 mg
Au/L even given the high gold grade of the pre-conditioned (barren) resin (575 g/dry
tonne). Four stages in series with a solution retention time of 1 minute per stage (60 BV/hr
flow rate) are sufficient to ensure an acceptably low tails solution tenor for a heap leach
operation at a PLS tenor up to 2 mg Au/L.
Plik z chomika:
ola54
Inne pliki z tego folderu:
Extraction of Gold From Igun Gold.pdf
(5691 KB)
Gold formation processes.pdf
(283 KB)
Gold Alloy.pdf
(4837 KB)
EXTRACTION AND BENEFICIATION.pdf
(631 KB)
Gold Refining Plant.pdf
(987 KB)
Inne foldery tego chomika:
zegarki
Zgłoś jeśli
naruszono regulamin