In-Pulp Treatment of Dissolved As from a Gold MinePresented by Bernard AubéCo-author: Lucienne Anctil, Mines Aurizon,

Casa Berardi

2

Casa Berardi

Mines Aurizon, Casa Berardi is a gold mine located on the Ontario-Québec border, 3 hours North of Rouyn-NorandaMines Aurizon re-opened the mine in 2006 at 1,600 tpd,

now operating at 2,000 tpdOre is finely crushed and gold is leached using CN and

carbon in pulp The current plan has the mine operating for another 10

years

3

Casa BerardiTailings and Process Water

RECYCLE PUMPHOUSE

DISCHARGE

In 2008, arsenic concentrations were near discharge limits in Process Water Pond.

A study was initiated to compare options for treatment and control of As.

IRON INJECTION

4

Arsenic Issues at Casa Berardi

A detailed mass balance study was completed to determine the source of As As each overflow from the specific ponds is monitored, it was

possible to evaluate the changes in arsenic loading throughout the system

The mass balance showed that the source of arsenic was the fresh tailings water – As was dissolved in the tailings slurry Analyses within the concentrator showed that the As was

liberated during the cyanidation

5

SAMPLING POINTS

MILL

PROCESS WATER POND

Cell 3

Cell 1

Cell 2

POLISHINGPOND

MINE

FINAL EFFLUENT

Fe Fe

FeArsenic loadings calculated at all sampling points.

The clear source of As was the MILL.

Mass Balance Flowsheet

6

As Treatment Theory

Arsenic can be treated using membrane technologies, adsorption, ion exchange, precipitated as calcium arsenate, and permeable reactive barriersMost common treatment is co-precipitation

with iron (Fe) and solid/liquid separationProven technologyMost cost efficientReliablepH of 4 to 5 known to improve treatment efficiency

7

Arsenic Precipitation

As can be removed as a calcium arsenate [Ca3(AsO4)2] at very high pH but it is not very stable As removed as ferric arsenate,

FeAsO4·xFe(OH)3 is much more stable In clear water at optimum pH, a

molar ratio of Fe:As of at least 3:1 for efficient removal and long-term stability

8

Laboratory Testing for Field Treatment

Some testing was completed at CANMET – MMSL to evaluate the ratio of Fe needed to treat As in clear water at alkaline pH in the field (Polishing Pond) Results showed that at an Fe:As molar ratio of 7.5:1 would

bring the concentrations down to 0.1 mg/L As This is the control that has been applied but this control

would no longer be preferred following the construction of Cell 4

9

Tailings Slurry Treatment

The information gathered suggested that many advantages are gained if As can be controlled at the mill Tests were begun to treat the As in the tailings slurry

immediately following the INCO-SO2 cyanide destruction First trials were attempted at different ratios of Iron

addition. Results were not conclusive due to some success and some failures without knowing the reason for response. Detailed Design of Experiment (DOE) was then

undertaken.

10

DOE Design

The DOE was designed to determine the major effects of many parameters on the arsenic treatment efficiency: Fe:As ratio pH and Two-step pHCN concentrationMixing intensityRetention time Aeration

The target concentration was set at 0.4 mg/L As. Current limit at 0.5 mg/L at effluent, but may become 0.2 mg/LWith 4 times dilution in the pond system, 0.4 mg/L in slurry would

be near 0.1 mg/L in Process Water Pond.

11

DOE Challenges

By far the greatest difficulty during the testing was to obtain the desired ratio of Fe to As because the initial As concentrations varied greatly and rapidly Samples taken the previous day were off by often more than 30% Attempts were made with a colorimeter which was only slightly

betterOn-site AA should soon be available

Another problem was obtaining proper pH control during one long series of tests. This was corrected in following tests.

12

DOE Testing

Overall, with nearly one year of tests, 141 trials were completed. Initial As concentration varied considerably:

2421181512963

30

25

20

15

10

5

0

Concentration initiale d'As (mg/L)

Fréq

uenc

e

Mean 12.27StDev 4.146N 141

Freq

uenc

y

Initial As concentration (mg/L)

13

302520151050

60

50

40

30

20

10

0

Consigne du Ratio

Rat

io r

éel

DOE Testing - Ratios

The Setpoint ratios were difficult to attain due to variability of feed concentration:

Rea

l rat

io

Setpoint Ratio

14

10987654

13

12

11

10

9

8

7

6

5

4

Consigne de pH

pH d

e ré

acti

on r

éel

Ca(OH)2CaO

ChauxType de

DOE Testing - pH

The Setpoint pH was normally controlled except for one series of tests where quicklime was used:

Mea

sure

d re

actio

n pH

Setpoint pH

15

DOE Testing – CN

Cyanide was most often below 2 mg/L (79 of 141 tests) It occasionally exceeded 10 mg/L due to operation of

INCO-SO2 system It was once above 80 mg/L as the CN destruction system

was down

16

5035302520151050

8

7

6

5

4

3

2

1

0

Ratio Fe : As réel (plage)

Moy

enne

des

con

cent

rati

ons

d'A

s (m

g/L)

DOE Results - Ratio

Clearly, the higher the ratio, the better the treatment. High ratios are expensive.

Aver

age

Ars

enic

Con

cent

ratio

ns (m

g/L)

Actual Molar Fe:As Ratio (range)

17

6050403020100

9

8

7

6

5

4

3

2

1

0

Ratio réel

Conc

entr

atio

n A

s fin

ale

(mg/

L)

DOE Results - Ratio

These actual results show some successful treatment (As<0.4 mg/L) at ratios of less than 20:1

Ars

enic

Con

cent

ratio

ns (m

g/L)

Actual Molar Fe:As Ratio

18

13121110987654

9

8

7

6

5

4

3

2

1

0

pH de réaction

Conc

entr

atio

n A

s fin

ale

(mg/

L)

DOE Results – pH

Lower pH values are better Note that low As values at pH>11 are calcium arsenate

Ars

enic

Con

cent

ratio

ns (m

g/L)

Treatment pH

19

90.035.025.015.07.53.51.0

5

4

3

2

1

0

CN plage

Moy

enne

As

(mg/

L)

101520253035

plageRatio

DOE Results – CN

CN Concentrations affect treatment efficiency at low ratios but less so at high ratios – CN complexing consumes Fe.

Aver

age

Ars

enic

Con

cent

ratio

n (m

g/L)

CN Concentration range (mg/L)

Ratio Range

20

DOE Results – Other Parameters

Aeration did not show any effects, but all tests were at high pHMixing intensity was not a significant factor Negative effects were measured at higher retention times

when there was a high pH. At expected pH values, retention time was not a significant factor.

21

40353025201510

0.4

0.3

0.2

0.1

0.0

Ratio molaire Fe : As réel

Conc

entr

atio

n fin

ale

d'A

s (m

g/L)

6789

pH plage

6.07

6.02

6.08

8.038.05

8.128.03

9.18

7.45

6.96

7.70

DOE Results – Successful Tests at Ratios < 40:1

Results shown when CN is well treated (<2 mg/L)A

rsen

ic C

once

ntra

tion

(mg/

L)

Fe:As Treatment Ratio

pH Range

22

9.59.08.58.07.57.06.56.0

50

40

30

20

10

pH de réaction

Rat

io m

olai

re F

e :

As

DOE Results – Successful Tests

All of these tests had As<0.4 mg/L. The regression shows the required Fe:As ratio at given pH values. More tests are needed to confirm or better define this relationship.

Mol

ar F

e:A

s R

atio

Treatment pH

23

Full-Scale Treatment

On-going with testing, ferric sulphate addition of 3.5 L/min in the tailings pump box has shown significant improvement in dissolved As fed to the tailings ponds In the past 4 months, the average dissolved As

concentration in the tailings discharge has been 0.43 mg/L Feed As concentrations in this time averaged 11.3 mg/L This essentially works, but it is an overdose of Fe at a

significant cost

24

Next Step

A mix tank is to be built in between the cyanide destruction and the tailings pump box – for treating the As with pH control The system will be automated and the As analysis will

need to be performed regularly to optimise treatment Although performance is improved at lower pH values,

consideration must be given to issues concerning the tailings placement pH and additional pH modification required in the field

25

Conclusions

Although the iron addition rates may be higher, it is possible to treat As in the tailings slurry at reasonable addition rates. A compromise must be made between pH for tailings

disposal and As treatment efficiency. This will allow arsenic to be disposed of with the tailings

and eliminates the need for a treatment plant and sludge management.