The mine is situated on the Kraaipan greenstone belt and was discovered by Shell in the mid-1980s.
Harmony acquired the operationin 1999 and currently the mine produces some 300 kg of gold a month.
What makes Kalgold challenging is that this Kraaipan orebodyis unique, being unlike the Wits Basin sedimentary deposits and different to the Barberton greenstone ores.
Although the mine is an openpit operation, the average grades of2,0 g/t are not spectacular.
“However, the recently-improved gold price has given us the opportunity to look at lower-grade areas of the orebody,” explains Kalgold general manager Dario Clemente.
His arrival on the mine some 18 months ago was preceded by an unsuccessful sale attempt to Afrikander Leases (Aflease).
However, Aflease managed Kalgold for four months from December 2003 to March 2004.
Clemente, a metallurgist, was brought to Kalgold to bring theplant back to profitability.
What he found was a plant that was less than ten years old but had suffered from different metal-lurgical management styles.
Gold recoveries were sagging at some 65%.
“Twenty months ago Kalgold faced two issues, the price of gold and the plant, which was not delivering either the efficiency or the level of costs required,” explains Clemente.
In the last year, the plant costs have been reduced to about R50/tand the efficiencies are now up to 90%.
By March 2004, the plant was giving about a 70% recovery. However, Clemente began adjust-ing the processing parameters, which, in a space of three months, raised the efficiency to 85%.
Some more changes were made in the plant and, in particular, the control philosophies were improved, and work was put into the aspectsof oxidation, cyanidation and pH control.
What the Kalgold metallurgical team gained was a better insightinto the chemistry of the orebody.
What had not been understood previously, is that the ore contains alkaline minerals, which increase the pH in the process circuit.
“The level of alkaline mineralsvaries from one area of the pit to another. Sometimes there is not much alkaline content in the ore, and we have to add lime,” explains Clemente.
At other times, the amount ofalkaline material pushes the pH in the plant over acceptable limits.
On conventional gold plants that process ore from the Wits Basin, high alkalinity is not a problem,explains Clemente.
However, the greenstone-type ore found in the Kraaipan belt is unique in its chemical make up, and there has been no complete pre-cedent for processing the ore found at Kalgold.
Though Kalgold mines a greenstone-type ore, the mineralogy of the rock differs from the Barberton gold ores.
If the pH climbs too high, thennot only does the plant consume increasing quantities of expensive cyanide, but gold is also lost in the process.
To counter the gold losses, the previous plant management had added more cyanide, to a point where the cost of the cyanide wasalmost more than the revenuederived from the gold recovered.
Whereas the Barberton greenstone ores contain arsenic, which gives its own set of problems, the Kalgold ore contains pyrrhotite and other sulphide components, which tend to increase cyanide consumption, among other effects.
In the pit, as mining moves from one block of ore to the next, thelevels of pyrrhotite can give way to increased iron carbonate (siderite), which influences the level of pH in the plant.
“Once we managed to contain the pH level in a very narrow band, and we understood this performancebetter, we found we could alter processing strategies during a single shift,” elaborates Clemente.
“All of a sudden, we found that the cyanide consumption was going down and the gold recoveries were going up.” At one stage for every ton of ore processed, the plant was consuming three kilograms of cyanide.
Now the plant has dropped this to 0,8 kg/t – a very large costsaving for Kalgold.
Clemente’s first three months on Kalgold were, by his own admission, “a very steep learning curve”.
It was not only gaining an under-standing of the chemistry of the ore and the plant-control systems, but the milling circuit needed to be sorted out as well.
A number of projects needed to be tackled from an upgrade to the crushers, an upgrade to the milling circuit, the leaching phase and the oxygen-injection system that needed an overhaul.
The plant had been designed for 138 000 t/m. However, at the time the plant was struggling to mill 120 000 t/m, and the amount of steel balls being used in milling was high.
Once the parameters had been set up for the mills, the mill throughput went up first to 140 000 t/m and then to 150 000 t/m in July 2005.
The ore at Kalgold is very hard and abrasive. However, Clemente explains that if the milling conditions are optimised, electrical power and steel usage drops.
It took a long time to arrive at the correct steel balls, once that had been done, steel usage, which was three kilograms for every ton of ore, has now been cut by two thirds.
The plant has three ball mills, a large mill that runs at 125 t/h and the two smaller mills at 65 t/h.
“This is a saving of R500 000 a month,” continues Clemente, “which is huge for a small mine such as this one.” In addition, dropping the on-mine costs meant that the cut-off grade dropped and stockpiles of one gram a ton of ore could be process-ed profitably.
Maintenance has been another issue that needed to be solved, Clemente continues.
The plant was constructedon a limited budget, the original thinking being that Kalgold was going to be a short-life mine.
Originally, processing was by means of a heap-leachoperation. However, as theoxidised ore gave way to sulphide ore at greater depth, the mine bought two secondhand mills to launch a carbon-in-pulp process.
“We have to live with designconstraints such as the size of the sumps and the fact that there are no tar roads,” explains Clemente.
When Clemente arrived on the mine, equipment failures werefrequent, and thus a significant amount of money was spent on bringing the plant up to standard.
However, the payback started almost immediately and, within three months, the investment in sound maintenance had more than paid for itself.
For example, presently, the mills are achieving better than 90% availability.
With the success, a number of other things rectified themselves.
The engineering section welcomed the attention dedicated to maintenance issues and, with increasing success, became morepositive towards maximising the plant performance.
The high costs, poor recoveries and excessive downtime had sapped the morale of the plant operators and, combined with pressure toperform, this manifested itself in discipline problems and disputes.
With the plant now operating at a greatly improved level, a sense of pride has been restored.
“We have not had a dispute withthe unions in six months,” he adds.
Further improvements in re-coveries were brought about by some technical changes.
Some of the ores contained clay, which stops the carbon used in the process from absorbing the gold, a term referred to as ‘preg-robbing’.
When high levels of clay were encountered, it was found that the residue gold in the tailings stream would also increase.
A simple change to the process stream solved the clay problems and pushed the recoveries from 85% to 90%.
In the years that the plant was struggling with efficiency issues,the mine was trucking ore to a blending pad near the primary crusher and stacking it in separate ore piles, depending on the grade.
These strategic (strat) ore stockpiles would be blended together to try and present the Kalgold plant with a constant grade.
There was a fear from the plant operators that certain grades of strat ore were ‘poison’ and that, if fedinto the plant, the recoveries would drop through the floor.
During a period when mining was forced to stop, the strat ores were fed to the plant and good recoverieswere achieved.
About four months ago, a decisionwas made to discontinue thesecondary handling of the ore as the cost implications were proving to be significant.
With the improvements made to the plant, the process stream is now far more tolerant of differences in the ore feed, and thus the need for creating large strat ore stockpiles has fallen away, elucidates Clemente.
Only the low-grade grade ore below one gram a ton is stockpiled.
The operational efficiency in producing a single product has saved the mine significant money.
“With the plant performing well, it opens up the scope for us to mine some lower-grade ore, for example the A Zone, which is now inside our cut-off,” says Tseliso Ntabe, a Kalgold mining engineer.
The A Zone has a lot of internal waste and it is not always possible to selectively mine this,” he adds.
Some of the shift foremen who have been at Kalgold for longer, were sensitive to the changes in the process but their knowledge had not made its way up the management chain, continues Clemente.
They were hesitant to implement changes, as they were understand-ably concerned about gold losses.
However, now that they have a better understanding of the process, they run the plant at a high level of efficiency with a bare minimum of supervision.
While the recoveries have gone up, the residues have come down from 0,7 g/t to 0,2 g/t.
A year ago, the processing cost per ton dropped from R80/ton to R50/ton. Since then, the processing cost has stayed constant in spite of increases in the costs of all plant consumables, such as steel, for example, which has gone up by 20%.
Gold production has increased year on year by 24% while working costs have dropped from R89 600/kg a year ago to R67 800/kg this year.
These factors have seen Kalgold move from being a loss-makingoperation to making a healthyprofit in the closing months of 2005. (See graph on page 27)
Making the most ofremaining reserves
Though the Goldridge mining lease area has four deposits, the largest and most southerly is the so-called ‘D Zone’, which has been the traditional mining target.
The pit has been mined since 1996 in a series of waste-stripping exercises to expose fresh ore. The waste-stripping exercises are colloquially known as ‘pushbacks’, where the eastern wall of the pit is mined and new access rampsestablished.
Though the original pit design catered for a final pit depth of250 m, Kalgold has recently consulted with some Australian mine design consultants who have shown them some techniques, which will allow the pit floor to be minedsome 50 m deeper.
This extra pit depth is not to be sneezed at as it will account foranother 300 000 tons of ore.
Currently, the production from the pit varies between 300 000 t/m and 500 000 t/m. What affects production is the availability of the mining machinery and rain.
“When it rains, we take all the equipment out of the pit, as bad weather can cause sidewall failures,”explains Kalgold’s mining managerChristo Pothas, a man with immense experience in hard-rock gold-mining.
The present pit has a life of about 18 months.
The mine recently completedits ‘pushback three’ and is now moving to complete the finalpushback four.
A basic tenet of the Harmony ‘way’ is the slogan ‘the orebody dictates’, and at Kalgold it dictates very directly the ratio of waste stripping to ore that will keep Kalgold profitable.
However, with modern techniques such as Whittle optimisation, much of the chance element has been taken out of pit design.
The D Zone orebody is on average 55 m wide and dips at 70