The Canadian Malartic gold mine in Quebec has reduced its cyanide use by 20 per cent without compromising recovery by removing cyanide injection points from its grinding circuit, following a study done together with the research consortium COREM.

“We didn’t have any problems with the circuit or recovery,” said Renée Dupéré, metallurgical coordinator at Canadian Malartic, which is jointly owned by Agnico Eagle Mines Limited and Yamana Gold Corporation. “We just wanted to lower our costs.” The operation, which mines a low-grade porphyry deposit, has a throughput capacity of 55,000 tonnes per day.

In the original flowsheet for Canadian Malartic’s processing plant, the cyanide used to extract gold was injected into the SAG mill at the start of the grinding circuit – standard practice in gold processing plants. The reasoning behind this was that early cyanidation could limit the residence time required in leach tanks further down the line, keeping the circuit size small. But COREM’s work found that much of the cyanide injected at the grinding stage was being consumed in reactions with steel grinding media and other minerals in the ore. Switching cyanide injection to the pre-leach thickener and the leach circuit resulted in less wasted cyanide.

A joint research project

The project emerged from discussions among COREM members, which include Canadian Malartic and other domestic gold producers. Finding ways to reduce cyanide consumption is a longstanding shared interest; about 99 per cent of cyanide introduced is consumed by some mineral other than gold.

“What we did here was find one cause of that cyanide being consumed by some unwanted substance,” said Driss Mrabet, a researcher of extractive metallurgy at COREM. In this case, the main culprit was iron, which bonds with cyanide to form ferrocyanide.

In 2014, COREM ran both lab tests at its facility and on-site plant surveys seeking to pinpoint the source of cyanide consumption and concluded that it was the milling stage. Dupéré said this technical assistance was critical.

“One of the reasons we didn’t do this step when we started the mill was that we were not able to analyze where the ferrocyanides were coming from,” she said, since it does not take a long time for ferrocyanide to precipitate. And although many labs are able to analyse ferrocyanide, they do not take into account the maturation of the sample. “COREM did a great job.”

Mrabet agreed: “It’s not easy to find somebody that can do the analysis onsite. I don’t think any mine has the ability to do that. They have to use outside services. If you collect samples and send it to a lab, the time it will take will create a discrepancy between the situation in the plant and the result from the lab. What COREM did is we developed a methodology that we can bring to the site and do a proper characterization so that we really know what happened in the solution.”

Armed with the ability to analyze the chemical composition of plant samples, COREM set about creating experimental conditions. For one set of tests, a mill reactor was built to simulate Canadian Malartic’s SAG mill. The researchers added sodium cyanide to the mill reactor and then let the milled ore sample leach in a glass tank for 48 hours. For the sake of comparison, they added cyanide directly to ore in a glass tank for the same period of time.

At the start of the 48-hour leach period, the level of free cyanide was already lower in the milling test, at a concentration of 121 milligrams per litre (mg/L) compared to 175 mg/L. Over the course of the leach period, the rate of gold recovery was slower in the milling test. A separate analysis showed that much more ferrocyanide was being produced. Taken together, these tests suggested iron in the mill was eating up cyanide.

That left another question: Where was the iron coming from? The top two suspects were iron sulfide minerals in the ore and the steel grinding media. To isolate the source, COREM compared ferrocyanide levels after running cyanide through the mill in four different scenarios: with steel balls plus ore, with steel balls alone, with ceramic balls plus ore and with steel balls plus silica. Comparison between steel and ceramic balls suggested that steel grinding media were probably the major culprit, with the sulphide ore playing a small part as well. The steel plus silica test generated far more ferrocyanide than steel alone, leading researchers to conclude that mechanical attrition of the steel balls was promoting iron-cyanide reactions.

Moving the cyanide addition point

The results of the COREM labwork prompted Canadian Malartic to change how it added cyanide. The plant flowsheet includes a SAG mill, two pebble crushers and three additional ball mills, followed by a pre-leach thickener and then cyanidation tanks.

Instead of adding cyanide at one single location in the SAG mill, Canadian Malartic now divides a smaller quantity of cyanide between the pre-leach thickener and the cyanidation tanks. “And we had enough time in the leaching circuit to leach all the gold that is in the ore,” said Mrabet.

“We still have cyanide in the grinding circuit,” added Dupéré. “But it’s due to recycling process water.” That does not pose such a problem, because COREM’s lab tests had also shown that concentration mattered. The high concentration of cyanide in the SAG mill – more than 250 mg/L – created conditions for cyanide consumption, but the new circulating concentration varied between 60 and 95 mg/L.

The result: the rate of gold dissolution in the grinding circuit stayed the same despite lower concentrations of cyanide. Overall gold recovery, and efficiency of recovery, stayed the same.

The project at Canadian Malartic wrapped up in 2015. The company declined to reveal how much the operation has saved following the flowsheet change.

Wider applicability

As a consortium project, the research COREM and Canadian Malartic have done together should benefit other mines. “This work aimed to attract attention to study this aspect of cyanide addition in the grinding circuit,” said Mrabet, “because the reflex is to directly add cyanide to the head of the circuit, to minimize retention residence time.”

In assessing the helpfulness of this approach, COREM concluded that it was better to “just keep a low level of cyanide that allows a continuous leaching in the grinding circuit, and assure a certain recovery without enhancing side reactions that come from cyanide.”

Dupéré said their results probably apply to any gold project that does, or can, circulate cyanide in its grinding circuit. She added the caveat that Canadian Malartic’s specific ore had played a part in the outcome as well. “Yes, we have a consumption of cyanide due to the grinding media, but we also have a part that was from the ore,” she said, “and we were able to reduce both of the ferrocyanides that were generated with the ore and with the grinding media.”