Processing Gold Ore by Flotation

Processing Gold Ore by Flotation


A flotation plant is being erected at the Falcon mine, Rhodesia, to treat ore containing gold and copper. With the exception of the Mt. Morgan, the Etheridge, and the Great Fitzroy mines, Queensland, I have not heard of the flotation process being used successfully to treat ore containing an appreciable amount of gold. The Elmore thick oil process was installed at the Lake View Consols gold mine, Kalgoorlie, several years ago, but was not successful, as the ore was not suitable, and unsuccessful experiments were made by Minerals Separation, Ltd., on ore from the Lancefield mine, Western Australia, which contains mispickel. The Elmore vacuum process was installed at the Cobar gold mines, New South Wales, and at the New Ravens- wood gold mines, Queensland. Both these mines contain copper in the form of sulphide, as well as gold, but the plants only ran a few weeks. I was informed that the plant at the former mine (where the ore contains about $8 gold and 1.5% copper) gave a fair recovery of copper, but left too much gold in the tailing or left enough copper in the tailing to prevent profitable cyanidation of the gold.

To return to the Mt. Morgan mine, the laboratory apparatus had a capacity of one pound of ore at a time, and the results now being obtained in the experimental mill approximate closely those obtained in the laboratory. The object of concentration was, of course, to obtain a concentrate containing as much gold, copper, and iron, and as little silica as possible, commensurate with a good extraction of the gold, because it was found that the less silica the concentrate contained the poorer was the extraction of gold. It costs 13 cents to flux one unit of silica, and it was necessary to steer a middle course. Experiments made with Sonstadt solution on ore from one part of the mine showed that clean quartz (after separation by specific gravity from all mineral) contained not less than $1.50 gold per ton. In practice, of course, it is impossible to float all the mineral and sink all the gangue.

The agitator in the laboratory plant was at first run at 1100 r.p.m., but was afterward reduced to 800. Tests were made with pulps of different proportions, each separate pulp being agitated for the same length of time, that is, 6 minutes, and it was found that there was not much difference, in the extraction of gold and copper, between a pulp containing three parts solution to one of ore, and a pulp containing seven parts solution to one of ore. A pulp of 1 to 1 was too thick and gave poor results. In practice, the thinner the pulp the smaller the capacity of the flotation machine. Tests were also made to ascertain the effect of agitating for different lengths of time. Two tests were made in the laboratory of which I have a note: one for 10 minutes and one for 15 minutes. The ore contained $6.50 gold and 2% copper; 12% of this sample would remain on a 60-mesh screen. The first one gave a concentrate containing $22.70, 9.4% copper, and 18% insoluble, with an extraction of 51% of the gold and 84.5% of copper. The second gave a concentrate containing $20.20 gold, 7.8% copper, and 27% insoluble, with an extraction of 64.5% of gold and 91.8% copper. The gold left in the tailing was probably in the gangue, as the extraction was poorer than usual. As a rule, the longer agitation and separation are continued, the more silicious the concentrate is. In practice, the length of treatment is regulated by the thickness of pulp and the number of boxes in the flotation machine. Tests made to ascertain to what degree fine crushing was necessary showed emphatically that the ore must all pass through a screen of 60 holes to the linear inch if a good extraction is to be obtained, and that the finer it was crushed, at any rate down to 120-mesh, the better the extraction was. Tests showed that when using eucalyptus oil there was no advantage in using an acid solution, but that, on the other hand, slight acidity did no harm. Much of the copper pyrite in the ore readily floats on water without any previous agitation. On treating ore containing $25 gold direct by agitation and flotation, without amalgamating or concentrating on tables, it was proved that fine free gold can be floated by using eucalyptus oil.

A few years ago some experiments were made by crushing in ball-mills and concentrating on Wilfley tables, but they were not successful. Last year it was decided to make a thorough trial of the Minerals Separation process, and a small testing plant was erected in the laboratory. At the same time a full-sized experimental unit, capable of treating 300 to 400 tons per 24 hours, was erected in one of the abandoned chlorination plants. Both sets of experiments were carried out by the metallurgical staff of the Company. After they were finished, a representative of the Australian branch of Minerals Separation, Ltd., paid a visit to the mine and conducted a few tests, which confirmed the results obtained by the mine staff.

As mentioned in the Company’s annual report, these flotation experiments were successful, the extraction being higher and the costs lower than expected. The company is now building the first unit of a plant to treat 1000 tons per 24 hours. The ore will be crushed by rock-breakers, Symons disc crushers, rolls, and tube-mills. It will then be concentrated on Wilfley tables, after which it will go through a second set of tube-mills, thence to the flotation machines. It is presumed that no royalty will be payable on the Wilfley concentrate. This concentrate will either be briquetted or sintered in a Dwight-Lloyd machine, and smelted in blast-furnaces along with the ‘copper ore’ and ironstone and limestone fluxes. The Company has no reverberating furnaces.


Many oils were tested, and, generally speaking, it was found that only essential oils gave a coherent froth and good extraction, other oils like petroleum, oleic acid, and lubricating oils tending to form granules which sank. The. best results were obtained from eucalyptus, closely followed by ‘ Essential C’ and Pinus lam us vulgaris. Oleic acid, which was used for years at Broken Hill on zinc ore with hot solution, and gave good results when tried on this ore with neutral and acid solutions, gave an enormous froth and floated most of the silica. A mixture containing 95% of eucalyptus and only 5% of oleic acid gave a concentrate containing 47% silica, showing the power of the oleic to float silica. Experiments were afterward made with a mixture of oils, and one combination (known as Mt. Morgan mixture) was found to give a better extraction of both gold and copper than any of the individual oils, and at less expense. When the sample was all crushed to pass 80 mesh, an extraction of 80% of the gold and 90% of the copper could be obtained every time, with a concentrate containing about 25% insoluble, which can be reduced to 10% by re-treatment. Hot solutions and a solution containing 1% of common salt were found to be detrimental to good recoveries.


A test on a sample, crushed to pass a screen of 120 holes per linear inch, containing $37 gold and 4.8% copper, gave a recovery by flotation alone of 90% of the gold and 98.5% of the copper, but left $8 gold in the tailing. The concentrate carried 44% insoluble matter, which could be reduced by re-treatment. A different oil (eucalyptus) would have given a poorer recovery and a cleaner concentrate.

Tests made on ore containing $9 gold, 3.5% copper, and 45% insoluble, showed that after crushing to pass 60 mesh and treating by direct flotation, an extraction of 82% of the gold and 96% of the copper could be obtained, with a concentrate containing only 21% insoluble. No doubt with finer crushing even better recoveries would be had. These results leave tables and vanners far behind. It was found decidedly advantageous to re-use the solutions.

A Wilfley table was erected in the mill, some tests made, and the tailing treated by flotation in the laboratory. Sometimes these tailing samples were dried before flotation, and sometimes they were not. It was invariably found that a better extraction was obtained from those which had not been dried, as no matter how carefully the operation was conducted, some of the iron pyrite got sufficiently oxidized to resist flotation, and it carried some of the gold.

In some of the tests the crushed ore was concentrated by panning in the laboratory, and afterward subjected to flotation. In this case the water in the laboratory was used, which did not come from the same source as the water used in the mill. It was noticed that the longer the sample was allowed to remain in the water after panning, the worse the subsequent flotation was. For example, where flotation took place immediately after vanning, the residue assayed $2.60 gold and 0.30% copper, but where tailing from panning was allowed to remain under water for 6 hours before flotation, the residue assayed $3.10 gold and 0.67% copper. An analysis of this water was made, and this incident shows what might happen in a mill where the ore is in contact with bad water for some hours before reaching the flotation machine, such as the time it is going through rolls, Chilean mills, tube-mills, and classifiers, over tables and through thickening devices, and perhaps through secondary tube-mills. The water in question was neutral, both before and after coming in contact with the ore.

Some tests were made both in mill and laboratory in which air was drawn into the agitation boxes through pipes fixed vertically in the corner with the top open to the air and the bottom ending in a bent pipe terminating under the impeller of the agitator. No improvement was, however, noticeable.

Grading tests were conducted on crude ore and flotation products. They showed that as regards crude ore, after crushing either in mill or laboratory, the finest grade of concentrate or ore was the richest and the coarsest grade of tailing was richest, both in gold and copper. The fact that the finest grade of tailing was the poorest shows that this process will float the finest sulphides successfully.


In the experimental mill the ore is crushed in rock-breakers and Krupp dry-crushing ball-mills without drying. This plant was formerly used to crush oxidized ore for chlorination and, being on the spot, it was naturally utilized in preference to buying new machinery. The crushed ore drops into a bin at the bottom of which are two Challenge feeders. These deliver the ore into a launder where it is met by a stream of water which carries it direct to a six-compartment Minerals Separation machine. Each spindle is driven by a half-crossed belt, thus eliminating the noise and grease incidental to the old Broken Hill method of gearing. The machine is of the Hoover single-level type, by which one man can attend to all the flotation boxes. The concentrate was collected at first in circular wooden vats with filter-bottoms of cocoa matting, and later in shallow rectangular concrete tanks which formed part of the old chlorination works. The whole plant is extremely simple and requires very few men to run it. It has not been found practicable to use a screen finer than 35 mesh on the ball-mills. It is found that the gold, copper, and iron contents are greater in the concentrate overflowing from No. 1 box and that they gradually decrease until No. 6 is reached, while the silica content increases from 10% in the concentrate from No. 1 box to about 50% in that from No. 6. About 56 hp. is required to drive the agitators at 350 revolutions per minute.

As it is intended to use “Wilfley tables in the new mill to assist in recovering the iron pyrite in the ore for fluxing and other purposes, two of these machines were placed in the experimental mill and some tests made to find out what results may be expected of them. Taking an average of several tests on ore from different parts of the mine, the grading of the ‘table feed’ was as follows: 10% remained on 60 mesh, and 19% passed through 60 but remained on 120 mesh. It contained $4.50 gold, 1.8% copper, 9% iron, and 76% insoluble. The concentrate assayed $17 gold, 2.9% copper, 34% iron, and 18% insoluble; the recoveries were 33% of the gold, 13% of the copper, and 38% of the iron. No doubt, had the pulp been classified and the fine material passed over slime tables or vanners, better results would have been obtained, but the Company does not intend to use mechanical concentrators for the slime, preferring to rely on the flotation process, so it was not worth while experimenting with them.

During the flotation experiments with eucalyptus oils some tailing was produced which contained a fair amount of gold, and attempts were made to recover some of this by amalgamating and cyaniding. It was found that no extraction by amalgamation was possible, nor was any extraction by cyaniding possible without either roasting or finer grinding. On unroasted tailing assaying $3 gold and 0.44% copper, after crushing to pass 120 mesh, separating the slime, and leaching the sand for 9 days, an extraction of only 60c. per ton was obtained with a consumption of 3.6 lb. of cyanide per ton. On a different tailing crushed to pass 80 mesh, which after slime was separated assayed $2.90 gold and 0.30% copper, an extraction of $1 was obtained in 5 days with a consumption of 2 lb. of cyanide.

Samples of slime were treated by agitation and washed by decantation, and gave slightly better extractions, but the consumption of cyanide went up to 6 or 7 lb. The strength of solution used in these tests was 0.10% KCN. It should perhaps be noted that all samples of flotation tailing had been dried before being tested by cyanidation.

Effect of Roasting

Two samples of sand from tailing were roasted and treated by percolation. The value was $3. The roasting reduced the sulphur to 0.5%. Although the copper and iron were oxidized by roasting, the consumption of KCN was less than in treating the unroasted tailing, which was contrary to expectation. “With three days treatment, the residue was reduced to $1 per ton, and about one-third pound of copper was dissolved from each ton of tailing by the cyanide. The consumption of cyanide was 1.4 lb. per ton, so that the extraction was higher and the loss of cyanide less than in treating unroasted tailing. Speaking from memory, I think that attempts to regenerate the cyanide in solution by means of sulphuric acid and lime were not very successful. The solution contained 0.05 gram copper per litre.

These cyaniding tests were merely done for information, as it is not expected that the tailing from the new mill will be profitable for cyaniding. The subject of extracting gold from flotation tailing arose a few years ago at the Cobar gold mines, as already mentioned, but in that case the difficulty was overcome by selling the mine, which contained highly silicious ore, to a company which owned a smelter, and had, or thought it had, plenty of basic ore for flux. Unfortunately, the amount had been overestimated and the problem is still unsolved—but that is another story.