See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/274082166

Cyanide Recovery and Detoxification Study on Seabridge Gold's KSM Project

Conference Paper · January 2015

CITATION READS

1 1,021

5 authors, including:

Mariam Melashvili C.A. Fleming

University of British Columbia - Vancouver SGS
16 PUBLICATIONS 76 CITATIONS 38 PUBLICATIONS 996 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

POX treatment for refractory gold ores View project

Center for Advanced Separation Technologies View project

All content following this page was uploaded by Mariam Melashvili on 27 March 2015.

The user has requested enhancement of the downloaded file.

 Cyanide Recovery and Detoxification Study on Seabridge Gold’s KSM Project

*Mariam Melashvili1, Chris Fleming1, Inna Dymov1, Jim Smolik2, Dean Lindsay3

1
SGS Canada Inc.
185 Concession Street
Lakefield, Ontario K0L 2H0
(*Corresponding author: [email protected])
2
Seabridge Gold Inc.
106 Front Street, Suite 400
Toronto, Ontario M5A1E1
3
Dean G. Lindsay Consulting Ltd.
378 – 55 Street
Delta, BC. V4M3J5

ABSTRACT
This paper presents results from the proposed cyanide detoxification circuit for Seabridge
Gold’s KSM project. The proposed flowsheet uses a two-stage CCD (counter-current
washing/decantation) circuit to transfer the flotation products CIL(carbon-in-leach) barren solution to a
SART process (Sulphidization, Acidification, Recycling of precipitate, and Thickening of precipitate)
to recover copper and free cyanide from the copper cyanide complex. The SART barren solution is
then treated by an AVR process (Acidification, Volatilization of HCN gas, and Reneutralization) to
concentrate the cyanide that was liberated in SART for recycling to leach. The washed CIL barren pulp
(the CCD underflow) is treated with the SO 2 /Air process to oxidize residual cyanide.

The goal of this complex flowsheet is to recover and recycle as much cyanide as possible and
produce a final effluent containing less than 0.5 mg/L copper and CN WAD at the lowest cost. The
SO 2 /air process is capable of achieving low residual CN WAD (<1 mg/L) in the final product, but the
persistence of aqueous copper above the target level has been detected in the treated solution in some
instances. The expected high concentrations of copper and cyanide in the KSM CIL barren ( >200
mg/L each species) and the difficulties this might create for the SO 2 /air process was a key factor in the
decision to consider including the washing, SART and AVR steps in the project flowsheet.

The target concentrations of CN WAD and Cu were readily achieved after aging the SO 2 /air
treated pulp, as would occur in a tailing or polishing pond. This suggests no additional treatment will
be required to sustain low levels of these species in the final plant effluent.

KEYWORDS

cyanide destruction, cyanide recovery methods, stability of aqueous copper in CIL pulp
 INTRODUCTION

The KSM Project is a proposed gold/copper mine located 65 kilometers northwest of the town
of Stewart, B.C at the site of one of the largest undeveloped gold resources in the world. The KSM
Project is made up of four large gold and copper deposits. The proposed mine has a 50-55 years life.
Part of development work was conducted at SGS Minerals, Lakefield site. This work included
a preliminary assessment of cyanide destruction and recovery methods designed to meet strict limits
regarding cyanide and copper levels in the plant’s final effluent. The assessment was conducted in two
stages. In the first stage, testing was conducted on a solution with a similar composition to the
anticipated wash solution from two-stage counter-current washing of the KSM cyanide leach residue.
A variety of cyanide destruction methods were evaluated, including SO 2 /air, peroxide, Caro’s acid and
activated carbon adsorption to reduce the concentrations of aqueous cyanide and copper to the target
levels. In addition, the SART process followed by AVR was evaluated to recover the cyanide for
recycling purposes, and to lower the concentrations of copper and cyanide entering the cyanide
destruction stage of the overall process.

The second stage included preparation of actual feed material for testing, by conducting
carbon-in-leach (CIL) experiments on flotation products – a cleaner scavenger tailing sample and a
rougher concentrate sample. This test program consisted of feed sample preparation, gold recovery in
batch cyanidation/CIL experiments and simulation of CIL residue washing by appropriate dilution and
decantation. Cyanide destruction testing on the CIL washed residue was then conducted using the
SO 2 /air process, followed by polishing treatment using carbon adsorption and peroxide oxidation.
Recovery of cyanide and copper from the wash solution was also evaluated using the SART and AVR
methods.

The objective of the testwork was to identify and prove a conceptual flowsheet that would
consistently produce a final effluent solution containing < 0.5 mg/L weak acid dissociable cyanide
(CN WAD ) and <0.5 mg/L copper.

Materials and methods

The recovery of cyanide was accomplished with the SART process, and the cyanide was then
concentrated for recycling to leach with the AVR process. SART testing was conducted in a 3L glass
reactor. The mixing was provided by an overhead agitator. The pH was maintained at the target value
with a 20% solution of sulphuric acid. The precipitant reagent was prepared by dissolving NaSH (69%
purity) in DI water. A flocculant (Magnafloc 156) was added to assist settling of the fine copper
sulphide precipitate. The feed for AVR was prepared by settling the precipitate, decanting the
supernatant and filtering the thickened precipitate. Air was then bubbled through the clear supernatant
solution to volatilize the hydrogen cyanide gas dissolved in solution and transfer it to a caustic
scrubber solution. The concentration of accumulated cyanide in the scrubber was monitored by
conventional titration against silver nitrate in the presence of rhodanine indicator.
 The cyanide destruction test using the SO 2 /air process was conducted in a 1L reactor with the
overhead agitator. Hydrated lime slurry was used to maintain the target pH. The airflow was
controlled to maintain a concentration of > 4 mg/L DO (dissolver oxygen) in solution. Sodium
metabisulphite (as the source of SO 2 ) was added by pumping solution at the flow rate and
concentration calculated to achieve the target ratio of SO 2 to CN WAD . The residual CN WAD was
monitored by picric acid method. At first, a batch test was conducted to produce a reactor full of
treated solution with low residual cyanide, which was then used as the starting solution for a
continuous test. The treated product in the reactor from one test was used as the starting material for
the next test.

Four different methods of polishing dilute solutions, after most of the copper and cyanide had
been removed, were tested. These methods involved treatment with Caro’s acid, hydrogen peroxide,
ferrous sulphate and activated carbon. In each of these experiments except for the activated carbon test,
a glass reactor with a magnetic stirrer was used. The Caro’s acid (25% peroxo-mono-sulphuric acid)
reagent was prepared in advance using a procedure provided by mixing concentrated grades of H 2 SO 4
(minimum 95%) and H 2 O 2 (50-70%). In the peroxide tests, a ~3% H 2 O 2 solution was prepared and
added in the amount required to achieve the target dosage. In the experiments with ferrous sulphate, a
solution of analytical grade FeSO 4 *7H 2 O was prepared and added at the target ratio of iron to aqueous
copper. Carbon adsorption polishing tests were conducted using fresh pre-attritioned activated carbon
(Calgon GRC-22). The carbon was added to the solution or pulp and either placed on rolls or placed
in a glass column through which the cyanide solution was pumped upflow. Samples were collected at
the top of the column and analyzed to determine the kinetics of copper and cyanide adsorption.

Stage 1- Treatment of Plant Wash Water

The objective of the test program was to investigate various treatment options to produce a
final treated product containing less than 0.5 mg/L of residual Cu and CN WAD . Testwork was
performed on the simulated CCD plant O/F (overflow) wash water sample and on several diluted CCD
plant wash water solutions. These were prepared by adding gypsum saturated water to generate target
copper concentrations of 90 mg/L (Solution A), 5 mg/L (Solution B), and 2 mg/L (Solution C), to
simulate polishing treatment of the tailings pond effluent. The overall test program and the results
achieved are summarized in Figure 1.
 Feed: CCD O/F Wash Water
Caro's (1)
Wash Water SART (1) AVR (1) Cu 0.1 mg/L
Cu 145 mg/L Cu 0.4 mg/L Cu 0.5 mg/L CN 1.2 mg/L
CN 176 mg/L CN 156 mg/L CN 14 mg/L
H2 O2 (1)
Cu 0.1 mg/L
CN 0.4 mg/L

Solution C Solution B Solution A SO2 /air (1)

Cu 2.1 mg/L Cu 4.9 mg/L Cu 88 mg/L Cu 2.0 mg/L
CN 2.0 mg/L CN 5.4 mg/L CN 104 mg/L CN 1.0 mg/L

Carbon (2) H2 O2 (2) FeSO4 (2) Carbon (1)

Cu 0.2 mg/L Cu <0.1 mg/L Cu 0.07 mg/L Cu <0.1 mg/L
CN 0.5 mg/L CN <0.1 mg/L CN <0.1 mg/L CN <0.1 mg/L

Figure 1- 1st Stage Treatment of CCD O/F Wash Water

Treatment of the undiluted CCD wash water by SART at pH 3 with 120% of the stoichiometric
requirement of NaSH precipitated copper and reduced its concentration in solution to <0.5 mg/L. In the
following AVR step, >90% of the cyanide in solution was stripped to the caustic scrubber, but residual
cyanide remained at an unacceptably high level of 14 mg/L. Polishing treatment of the AVR barren
solution was therefore evaluated using Caro’s acid and hydrogen peroxide. Treating the AVR barren
solution with 5 times the stoichiometric requirement of Caro’s acid was unsuccessful in reducing the
CN WAD level to the target <0.5 mg/L, but the AVR barren solution responded well to polishing
treatment using 10 times the stoichiometric requirement of H 2 O 2 . The CN WAD and Cu concentrations
were reduced to <0.5 mg/L.

Solution A (104 mg/L CN WAD , 88 mg/L Cu) responded reasonably well to treatment using the SO 2 /air
process. Conducting the continuous test at pH 8.5, with 1 hour retention time, while adding 4.8 g SO 2
per gram CN WAD , produced a treated solution containing 1 mg/L CN WAD and 2 mg/L Cu. Contacting
this solution with 15 g/L activated carbon reduced the residual CN WAD and Cu to well below their
target <0.5 mg/L concentrations. Treating the SO 2 /air solution product with ferrous sulphate, at
approximately 11 moles iron per mole copper, was also successful in attaining the target <0.5 mg/L
residual CN WAD and Cu.

Cyanide destruction with hydrogen peroxide was tested on the diluted Solution B (5.4 mg/L CN WAD ,
4.9 mg/L Cu) with approximately 10 times the stoichiometric required of H 2 O 2 . This method was
successful in reducing the concentrations of CN WAD and Cu in solution to < 0.5 mg/L.

Activated carbon was evaluated as a treatment option for the very dilute Solution C (2 mg/L CN WAD ,
2.1 mg/L Cu). Treating Solution C in a fluidized bed column, at a solution flowrate of 10 carbon bed
volume per minute and a solution to carbon mass ratio of ~2000, reduced the Cu concentration from 2
mg/L to <0.5 mg/L. However, the effluent still contained 0.79 mg/L CN WAD . Passing the solution
through the same carbon bed a second time reduced the CN WAD to <0.5 mg/L.

Stage 2- Treatment of Flotation Products

This part of the program included (i) feed sample preparation, (ii) cyanidation/CIL treatment
of the feed, (iii) simulation of residue washing by appropriate dilution and decantation, (iv) cyanide
destruction on the CIL washed residue using the SO 2 /air method and (v) polishing treatment using
carbon adsorption and peroxide oxidation. In addition, cyanide recovery using the SART and AVR
methods was evaluated on the wash solution. The proposed flowsheet and the results achieved are
summarized in Figure 2.

Feed: 50:50 Blend of Flotation Concentrate and Scavenger Tailing

CN Scrubber
CN 103 mg/L
O/F
CIL pulp Wash 1 st stage Wash 2 nd stage SART (2) AVR (2)
Cu 175 mg/L Cu 74 mg/L Cu 26 mg/L Cu 2 mg/L Cu 0.5 mg/L
CN 240 mg/L CN 113 mg/L CN 25 mg/L CN 78 mg/L CN 1.1 mg/L
U/F

SO2 /air (2) SO2 /air (3) SO2 /air (4) H2 O2 (4)
Cu 5.1 mg/L Cu 2.5 mg/L Cu 0.4 mg/L Cu 0.2 mg/L
CN <0.1 mg/L CN <0.1 mg/L CN 0.1 mg/L CN 0.2 mg/L

Cu CIP- 24 h Aging -24 h Caro's (2)

Cu <0.1 mg/L Cu <0.1 mg/L Cu 0.3 mg/L
CN <0.1 mg/L CN - mg/L CN 0.6 mg/L

Figure 2 - 2nd Stage Flowsheet for Flotation products

The CIL procedure was performed using 1000 mg/L NaCN in the presence of 28 g/L activated carbon.
The final CIL leach solution contained 175 mg/L Cu and 240 mg/L CN WAD . Two-stage counter-current
washing of the CIL residues was simulated by diluting the CIL pulp with an appropriate volume of
gypsum-saturated water, followed by decantation and filtration. Hence, gypsum saturated water was
used to simulate the recycle of AVR barren solution to the CIL wash circuit.

The solution recovered in the simulated two stage CCD wash circuit contained 74 mg/L copper and
113 mg/L cyanide, and this solution was treated by SART at pH 3 with 120% of the stoichiometric
requirement of NaHS. The copper concentration in solution was reduced to 2 mg/L. Treatment of the
SART product solution by AVR (at the same pH of 3) to recover the liberated cyanide, lowered the
cyanide concentration from 78 mg/L to ~ 1mg/L and further reduced copper to 0.5 mg/L. The results
indicated that barren solution containing low levels of residual cyanide and copper (~1 mg/L CN wad
and 0.5 mg/L Cu) could be produced in this SART/AVR flowsheet, although further polishing
treatment of the AVR barren solution with hydrogen peroxide may be needed to meet the CN limit for
discharge to the environment of <0.5 mg/L.

The solution portion of the slurry in the thickener underflow from the simulated two stage CCD wash
circuit contained 26 mg/L copper and 24 mg/L cyanide, It was noticed during testing that the aqueous
copper was unstable in the simulated CIL wash thickener underflow, and its concentration had to be
readjusted back up to the target concentration of ~25 mg/L before cyanide destruction testing
commenced. In order to do that, sufficient CIL decant solution was added to the washed pulp.

Three continuous SO 2 /air cyanide destruction tests were conducted on the simulated thickener
underflow slurry, targeting different residual copper concentrations in each test (ranging from 0.4 to
5.1 mg/L). The final CN WAD concentration was 0.1mg/L or lower in each test. It was observed during
this testing that the products of these continuous SO 2 /air tests were also quite unstable and lost most of
their soluble copper content by the time of polishing testing. Therefore, in one polishing test,
sufficient CIL decant solution was added to the SO2/air treated slurry to increase the copper and
CN WAD concentration back up to ~ 4 mg/L for carbon adsorption tests. One bottle-on-rolls test was
conducted using 30 g/L activated carbon, while another was carried out without activated carbon - to
assess the effect of pulp aging. The concentrations of Cu and CN WAD decreased to < 0.1 mg/L after 24
hours in the test conducted with activated carbon. However, even in the test in which no carbon was
added, the final solution contained < 0.1 mg/L Cu after standing for 24 hours. This suggests that
simple aging of cyanide destruction products may be sufficient to meet the copper and CN WAD criteria
for discharge to the environment.

Stability of CIL pulp

A series of tests was devised to evaluate the stability of aqueous copper in the CIL pulp. The
test program consisted of CIL cyanidation of 50:50 blend of the cleaner scavenger tailings and
flotation concentrate, followed by two stage washing of the CIL pulp using gypsum saturated water. A
portion of the washed pulp was allowed to age for 12 days in a covered pail. The remainder of the CIL
washed pulp was treated with SO 2 /Air to destroy the cyanide and the SO 2 /air treated pulp was placed
in three bottles. The pulp in the first bottle was allowed to age for 12 days without mixing. The pulp
in the second bottle was contacted with 30 g/L activated carbon on rolls for 12 days. The third bottle
was rolled without the addition of activated carbon for 12 days. The flowchart for the testwork and the
results achieved are summarized in Figure 3.
 Feed: 50:50 Blend of Flotation Concentrate and Scavenger Tailing

CIL pulp 2 stage Wash SO2 /air Carbon Static

Cu 193 mg/L Cu 26 mg/L Cu 1.3 mg/L Cu <0.1 mg/L
CN 346 mg/L CN 23 mg/L CN <0.1 mg/L CN <0.1 mg/L

Aging -12 days 30 g/L Carb.Rolls 0g/L Carb.Rolls

Cu 0.1 mg/L Cu <0.1 mg/L Cu <0.1 mg/L
CN 0.2 mg/L CN <0.1 mg/L CN <0.1 mg/L

Figure 3 - Copper Stability Program

The results indicated that the CIL washed pulp was chemically unstable. The concentrations of copper
and CN WAD in solution after 12 days of aging decreased from ~25 mg/L to 0.1 mg/L and 0.2 mg/L,
respectively. Treating the CIL washed pulp with SO 2 /air produced a pulp containing 1.29 mg/L Cu and
< 0.1 mg/L CN WAD . This SO 2 /Air treated pulp was also unstable. The soluble copper concentration in
the SO 2 /air treated pulp decreased to < 0.5 mg/L after only 1 day in all three tests
.

DISCUSSION

In the SART process, the addition of NaHS under acidic condition precipitates copper as
copper sulphide and liberates cyanide as HCN gas in solution. The reaction is presented below:

2Cu(CN) 3 2- + S2- + 6H+ = Cu 2 S + 6HCN [1]

Based on the data presented in Table 1, the recovery of copper was highest (99.7%) when its
initial concentration was highest.
Table 1 - The SART Summary

Test Solution Analysis, mg/L Recovery, %

CNT CNWAD CNF CNS CNO N as Cu CN Cu
NH3 +NH4
Wash Water Treatment
Wash W. 187 176 50 270 25 - 145
SART (1) 162 156 156 280 12 - 0.4 88.6 99.7
Flotation Products Treatment
Wash 1st 122 98 34 150 22 0.9 74
SART (2) 140 78 73 150 4.6 8.0 2.0 74.5 97.3
st
Wash 1 (1) 125 113 37 140 19 0.5 74
SART (3) 120 92 86 150 4.1 6.4 6.0 76.1 91.9

The AVR tests results are presented in Table 2. The rate of HCN stripping from the SART
product solution was slow under ambient conditions. In test AVR (3), only 89% of the cyanide had
been stripped after 5 hours. Extending the air stripping time to 8 hours in AVR (2), improved cyanide
stripping efficiency to almost 99%.
 Table 2 - The AVR on SART treated Solution

Test Residence Solution Analysis, mg/L Recovery

time CNT CNWAD CNF CNS CNO Cu N as % CN of Feed
min NH3 +NH4 CNWAD
Flotation Products Treatment
SART (2) 140 78 73 150 5 2.0 8.0
AVR (2) 480 31 1.1 3 150 <1 0.53 8.3 99.0
SART (3) 120 92 86 150 4.1 6.0 6.4
AVR (3) 300 42 19 14 140 <1 2.51 8.0 88.9

In the SO 2 /air method, the weak acid dissociable cyanide is oxidized to cyanate and copper
precipitates as copper hydroxide.
2CN- + Na 2 S 2 O 5 + 2O 2 + 2OH- → 2CNO- + Na 2 SO 4 + SO 4 2- + H 2 O
[2]

Cu2+ + 2OH- → Cu(OH) 2 [3]

Table 3 - The SO 2 /air Summary

Test Retention Solution Analysis, mg/L Reagent

Time pH CNT CNWAD CNO CNS N as Cu SO2
min NH3 +NH4 g/g CNWAD
Wash Water Treatment
Solution A 108 104 17 160 87.8
SO2 /air (1) 57 8.7 4.8 1.0 120 120 12.7 2.04 4.6
Flotation Products Treatment
Wash 2nd 25.0 25.9
SO2 /air (2) 33 8.5 - <0.1 - - - 5.06 5.2
SO2 /air (3) 33 8.5 - <0.1 - - - 2.45 5.5
SO2 /air (4) 64 8.5 - 0.12 - - - 0.37 5.5

The results indicated that the process is sensitive to the reaction time. By comparing tests run on the
same feed material (Wash 2nd), it can be seen that the target copper concentration of <0.5 mg/L could
only be attained by doubling the residence time from ~0.5h to ~1h.

The relative solubility of aqueous copper (2-5 mg/L) when almost all the cyanide is gone (<0.1 mg/L)
could be related to the formation of another copper-complexing ligand as a cyanide decomposition
product. For example, the hydrolysis of cyanate ions yields carbonate and ammonium ions in solution,
and ammonium ions are known to form a strong complex with copper, which is stable in alkaline
solution. The relevant reactions are shown below:

CNO- + 2H 2 O = NH 4 + + CO 3 2- [4]
 NH 3 + H 2 O = NH 4 + + OH-
[5]
2+ 2+
Cu + 4NH 3 = Cu(NH 3 ) 4 [6]
An E H -pH diagram was constructed using the HSE program for the same copper and ammonia
concentrations as in Test SO 2 /air (1), and this is shown in Figure 4.

Figure 4 - Stability diagram using HSE thermochemistry program

The recommended pH range for the SO 2 /air procedure (pH 8.5–9) falls precisely within the narrow
pH stability domain for the Cu(NH3) 2 2+ and Cu(NH4) 2 2+ complexes. Hence, thermodynamics predict
the likelihood that the copper ammonium ion complexes will be stabilized under the specific pH and
E H conditions required for the SO 2 /air process.

Caro’s acid is a mixture of hydrogen peroxide and sulphuric acid and delivers a higher
oxidation potential and faster reaction rate than peroxide on its own, and therefore does not require the
addition of a catalyst. But the reaction requires the addition of hydroxide (lime) to maintain a weakly
alkaline pH. Caro’s acid oxidises cyanide and thiocyanate species according to these reactions below:

CN- + H 2 SO 5 = CNO- + H 2 SO 4 [7]

SCN- + 4H 2 SO 5 + 10OH- = 5SO 4 2- + CNO- + 9H 2 O [8]

The results are presented in Table 4 and indicate that, under the conditions tested, Caro’s acid was
unable to attain the 0.5 mg/L target level required for cyanide. Thiocyanate was effectively oxidized,
which would be a disadvantage of this method in those jurisdictions in which the removal of CNS is
not required.

Table 4 - The Caro’s acid polishing treatment of AVR barren solution

 Test Solution Composition, mg/L
pH EH CNT CNWAD Cu CNO CNS
Wash Water Treatment
AVR (1) 2.4 379 42 14.3 0.50 <1 261
Caro's (1) 8.4 438 8.8 1.2 <0.05 94 <2
Flotation Products Treatment
AVR (2) 3.3 - 31 1.10 0.53 <1 150
Caro's (2) 9.2 311 25 0.58 0.31 36 62

The hydrogen peroxide process offers relatively fast reaction kinetics, simple control
instrumentation, adaptation to wide swings in feed solution chemistry and a dependable product
quality. In this method, the oxidation of cyanide is optimum under neutral to alkaline conditions and
copper is added as a catalyst. The relevant equations are shown below:

CN- + H 2 O 2 = CNO- + H 2 O [9]

CN- + 2Cu2+ + 2OH- = CNO- +2Cu+ + H 2 O 2 [10]
+ 2+ -
Cu + H 2 O 2 = 2Cu + 2OH [11]
The oxidation reaction is generally conducted at the natural pH of the effluent (around pH 10) and
usually does not require pH control, because H 2 O 2 exhibits only a mild acidic character.

The final treated solutions contained < 0.5 mg/L Cu and CN WAD in all the tests, regardless of the initial
cyanide and copper concentrations. A small amount of thiocyanate was also oxidized, but this clearly
would not be the process to select if thiocyanate removal was required.

Table 5 - Peroxide Treatment Summary

Test H2 O2 Solution Composition, mg/L

% Stoich pH EH, mV CNT CNWAD Cu CNO CNS
Wash Water Treatment
Solution B 8.1 381 6.35 5.4 4.89 0.9 8.7
H2 O2 (3) 1095 8.0 419 <0.1 <0.1 0.14 8.5 <2
Sol. B 8.1 381 6.35 5.43 4.89 0.9 8.7
H2 O2 (2) 2190 8.0 405 <0.1 <0.1 <0.05 7.9 <2
AVR (1) 2.5 413 42.1 14.3 0.50 <1 261
H2 O2 (1) 1000 8.6 368 12.6 0.4 0.10 9.3 240
Flotation Products Treatment
AVR (2) 3.4 354 31.0 1.10 0.53 <1 150
H2 O2 (4) 4200 8.0 364 23.2 0.20 0.24 <1 150

Only two tests were run using ferrous sulphate to lower cyanide and copper concentrations,
according to the reaction mechanisms shown in the equations below. The data showed that this would
be an effective polishing treatment when applied to low grade copper and cyanide solutions. It was
possible to reach very low levels of 0.07 mg/L Cu and < 0.1 mg/L CN WAD when using ferrous sulphate
at a high Fe:Cu molar ratio of 11.

Fe2+ + 6CN- = Fe(CN) 6 4- [11]

Fe(CN) 6 4- + 2Cu2+ = Cu 2 Fe(CN) 6 [12]

Table 6 - Ferrous Sulphate Treatment Summary

TestFeSO4 •7H2 O Solution Analysis, mg/L Recovery, %

Molar Fe/Cu CNT CNWAD CNS CNO Cu Fe Cu
Wash Water Treatment
SO2 /air (1) 4.8 1.0 120 120 2.04 <0.05
FeSO4 (1) 5.6 <0.1 <0.1 99 130 1.91 <0.05 6.4
FeSO4 (2) 11.2 <0.1 <0.1 88 130 0.07 <0.05 96.6

Using activated carbon as a polishing treatment was proposed since the plant will be based on the
CIL process. It is believed that activated carbon lowers copper and cyanide concentrations in solution
both by catalyzing the oxidation of cyanide to cyanate and by adsorption of copper cyanide onto the
carbon. The results are summarized in Table 7, and show that this method is capable of removing free
and complexed cyanide to very low levels providing the starting concentrations of copper and cyanide
are quite low.

Table 7 - Activated Carbon Adsorption

Sample Time Solution Analysis, mg/L

h Cu CNT CNWAD CNO CNS NH3 /NH4
Flotation Products Treatment
SO2 /air (2) 0 3.82 11.7 4.84 53 120 1.6
Cu CIP 2 0.45 - 0.55 - - -
(30 g/L Carbon) 4 0.11 - 0.14 - - -
8 <0.05 - <0.1 - - -
24 <0.05 0.46 <0.05 31 45 3.3

The investigation to determine the stability of aqueous copper in CIL washed pulp showed
that both copper and cyanide precipitated naturally from the CIL tails over time. Their concentrations
decreased from ~25 mg/L to < 0.5 mg/L after 12 days of aging (Table 8). The concentration of
unreacted sulphides in CIL pulp was high (~ 22%) and it is unclear whether copper precipitated from
the CIL tails as a CuCN species or as a Cu 2 S, or even possibly as a copper thiocyanate species
(CuSCN).
 Table 8 - Stability of Leached Pulp

Test Solution Analysis, mg/L

Cu CNWAD
Flotation Products Treatment
Wash 193 346
stage 1 56.2 98.5
stage 2 25.9 23.0
Aging 25.9 23.0
1 days 12.6 19.2
4 days 18.2 17.5
12 days 0.14 0.24

CONCLUSIONS

Various cyanide destruction and cyanide recovery methods have been tested on the CIL tailings
that will be produced when Seabridge Gold’s KSM project is commercialized. The primary objective
of the study was to establish a flowsheet that would produce treated waste water meeting the copper
and cyanide concentration limits (<0.5 mg/L) for direct discharge to the environment. A secondary
objective was to develop a flowsheet that achieved this objective at the lowest overall cost. Whilst the
scope of the investigation did not include a detailed cost analysis of the various flowsheet options, the
test results achieved allowed some general conclusions on the various options to be drawn.

 Cyanide recovery by SART and AVR was shown to be technically feasible. Copper in
the CIL tailings solution (after solid/liquid separation) is recovered at an efficiency of
>95% as a precipitate of Cu 2 S, which can be sold to a smelter to achieve additional
revenue. Cyanide is recovered at an efficiency of >90%, initially as HCN in the SART
discharge, and then as NaCN after AVR, for recycling to leach. It is unlikely that the final
treated solution after SART and AVR will be suitable for direct discharge to the
environment and a final polishing step will likely be needed to lower cyanide (in
particular) to the <0.5 mg/L target level.

 The SART/AVR flowsheet will have the highest capital cost of all the options, and the
value of the cyanide and copper recovered in the process will have to carry the significant
capital cost of solid/liquid separation (at least two stages of CCD), SART and AVR. This
circuit is unlikely to be economically feasible for low-copper concentrations in the pyrite
rougher CIL circuit and can be bypassed on those occasions. However, if a higher-
cyanide-soluble copper sulphide product is treated, such as a first cleaner tailings, much
higher concentration of copper and cyanide will result, and the KSM operation will likely
benefit from incorporation of the SART and AVR circuit.
  Cyanide destruction by the SO 2 /air method applied directly to the CIL tailings slurry is
likely to provide the best overall economics when copper in solution is relatively low
(<100 mg/L) because of the relatively low capital cost involved (one or two agitated
tanks) and the relatively low cost of reagents (SO 2 , lime, air). The results indicated that at
least one hour residence time in the cyanide destruction tank(s) will be needed, and that
the consumption of sulphur dioxide will be about 5g SO 2 per gram of CN WAD. The
results also suggest that, even if the target cyanide and copper concentrations are not
reached in the agitated cyanide destruction tanks, their concentrations will drop further in
the tailings dam, due to slow precipitation of CuCN, CuSCN or Cu 2 S. As a result of this
aging process, it may be possible to attain the target concentrations in the tailings dam
without the need for final polishing. If final polishing of the discharge stream is needed to
lower either or both cyanide and copper levels to <0.5 mg/L, both activated carbon and
ferrous sulphate will be very effective.

 Of the alternative oxidants tested, hydrogen peroxide of CIL tailings slurry, is likely
(along with the SO 2 /air process) to produce the best economics (reasonably low capital
and operating costs). It will also achieve the technical objectives of oxidizing cyanide and
precipitating copper to something close to the required discharge limits. If final polishing
of the discharge stream from the tailings dam is needed, either activated carbon or ferrous
sulphate would be very effective in this case too.

 Both SO 2 /air and peroxide are quite selective oxidants for cyanide, and oxidize very little
thiocyanate. If thiocyanate removal is not required by local environmental regulations,
this will be a significant cost advantage for these two oxidants because the oxidation of
thiocyanate consumes vast quantities of oxidant.

 Caro’s acid was the least effective of the oxidants tested as far as cyanide destruction was
concerned, but was a very efficient oxidant for thiocyanate. Oxidation of thiocyanate is
likely to lead to very high oxidant (peroxide) consumption with this process, and the
requirement for near neutral conditions for the reaction will also lead to high lime
consumption in neutralizing the sulphuric acid component of Caro’s Acid. The
economics of this process are therefore likely to be the least favourable of those tested.

ACKNOWLEDGEMENTS

The authors wish to thank Dave Matthews and Kathleen Ellis for their assistance in laboratory testing
and data processing. Thanks extend to Cuong Trang for invaluable support throughout the project
development, data discussions and final reporting.
 REFERENCES
Byerley, J.J. Enns, K.. Trang, C.V. and Le, V.T. (1988). A Treatment Strategy for Mixed Cyanide
Effluents-Precipitation of Copper and Nickel Cyanides. Randol Gold Forum. Pp331- 339
Johnson, A.R. McQuenn, T.M. and Rodolfa, K.T. (2005). Species Distribution Diagram in the
Copper-Ammonia System: Am Updated and Expanded Demonstration Illustrating Complex Equlibria.
Journal of Chemical Education. Vol. 82 No.3 pp 408-414
Mudder, T.I. Botz, M.M. and Smith, A. (2001) Chemistry and Treatment of Cyanidation Wastes.
Mining Journal Books LTD. Chapter 6. pp 327- 333.
Ford, K.J.R. Fleming, C.A. Henderson, R.D. (2008). Application of the SART Process to Heap
Leaching of Gold-Copper Ores at Maricunga, Chile. Proceedings of the 40th Annual Canadian Mineral
Processors Conference. pp 591-613.
Fleming, C.A. (2014) SGS Gold Workshop-Environmental Aspects. Chapter 7. pp1-30.

View publication stats