PROCESSING OF ORES

Technical Field

T ^h is ⁱ nvention relates to the processing o _f nic _k e _l ^d epos ⁱ ts, particularly those deposits where the nickel occurs pre ^d om ⁱ nantly in the sulphide mineral pent _l an _d ite

Background Art

The processin ^g of sulphide nickel deposits currently f ^f ooilllowedH ^bg y ^rlnd fl ⁱⁿ o ⁹ tati ^t o ⁰ n ^l i ⁱ n ^be "wh ^t i ^e ch *g ^e an _g u ⁿ e ^lCke m ^l iner ^« a-lWs ^d ar.eI re ^j ecte ^d . _It is particularly important to reject nearly all of the minerals that contain magnesium _b ecause the r«ore M ^g O there is in a nickel concentrate the hi _g her the temperature for su ^b sequent smelting, owing to the effect whrch magnesia has on slag viscosity. Any increase in smelt-n ^g temperature increases energy costs an _d re _d uces ^{tne llfe of f} urnace refractories.

The approach usually used in flotation is to arrange the chem-cal con ^d itions to be as favourable as possible for the flotation of nickel sulphides and as unfavoura _b le as poss ^ib le for recovery of magnesium _b earin _g minerals part ⁱ cularly those minerals such as talc which are naturally strongly floatable. _For example, copper suiphate is often used to increase the rate of flotation

SUBSTITUTE SHEET

of nickel minerals while polysaccharides such as starches or guar gums are added to suppress the floatability of talc. This strategy produces a highly efficient separation of nickel from non-sulphide minerals with a minimum concomitant loss of nickel. However, there is almost no selective rejection of those sulphides that contain little or no nickel.

Until recently there was little incentive to reject gangue sulphides such as pyrrhotite. Indeed, the presence of such sulphides in concentrates was often considered desirable because of the small amount of nickel they sometimes contain and because their reaction with oxygen produces a great deal of heat and less fuel is therefore needed for smelting. The only real penalty for recovering the gangue sulphides was the cost of transporting them to the smelter. That the smelting of such concentrates produced large amounts of sulphur dioxide was rarely considered a real problem.

Over the last few years this situation has changed dramatically. High sulphur dioxide emissions are now no longer environmentally acceptable and more and more stringent restrictions are being imposed on the amount of sulphur that smelters can emit. As a result, much research has been directed towards developing a flotation method that allows the selective rejection of gangue sulphides and of magnesium bearing minerals. By rejecting as much of the ^" gangue sulphides as possible before smelting, nickel production for a given emission of sulphur can be maximised.

Disclosure of the Invention

An object of the present invention is to provide a flotation process for the selective rejection of

v»

enhanced compared to those in a mild steel mill and we presume that some such interaction is the reason pyrrhotite floats less strongly.

Where it is not practical to use stainless steels in industrial grinding mills, steels that are not highly reactive may be used. An oxygen containing gas such as air may be supplied during grinding to ensure that conditions are not strongly reducing. A mild steel mill with mild steel grinding media and having ports permitting access of air during grinding may be used. A rubber lined mill charged with corrosion resistant steel grinding media could be used.

The pH of most nickel flotation pulps ranges from slightly acidic to slightly alkaline and at such pH values sulphides usually float rapidly with collector alone in the absence of any interference from minerals such as talc. The rate of flotation of the sulphides can be increased still further by adding copper sulphate before the collector but once copper sulphate is added it is not possible to select between pentlandite and pyrrhotite.

If instead of adding copper sulphate, alkali is added to raise the pH to a value of not less than about 9, pyrrhotite floats much less strongly and there is some decrease in the recovery of pentlandite. By analysing products from tests done under such conditions we have discovered that nearly all ^" of the nickel that is lost is of a size greater than 75 μm while the decrease in recovery of pyrrhotite results mostly from a decrease in the rate of flotation of particles of intermediate size (10 to 75 μim) . Recognising this, we have applied a technique known as split conditioning, which was developed earlier in our laboratory (Anthony, Kelsall and Trahar,

SUBSTITUTE SHEET

Preferably the relatively coarse fraction has a size greater than about 75 μm.

The pulped ore may contain about 20% to 65% solids.

While many types of alkalis and collectors may be used, soda ash is particularly preferred as the pH modifier and the preferred collector is n-amyl xanthate. Sodium carbonate or lime may also be used as the pH modifier however lime may lead to some loss of selectivity between the sulphides. A mixture of alkalis and of collectors may be used. The water used in this step should have little dissolved copper.

The amount of collector and the proportion of the collector added to the coarse fraction need to be determined for each ore, but generally speaking more than half of the collector should preferably be added to the coarse fraction. In our laboratory tests three quarters of the collector was added to the coarse fraction and the other quarter to the fine fraction. This proportion may be varied but it is important that most of the collector is added to the coarse size.

The split size for conditioning and the conditioning time need to determined for each particular ore. In our tests, pulps were wet screened over a 75 μm sieve and each size fraction was conditioned with collector for 5 minutes. In practice a cyclone or a cylinder-cyclone could be used to classify the pulp. Two stages of classification might also be used to ensure a precise size split. A few minutes of conditioning would probably be sufficient in most cases.

A gangue depressant can also be added either during or after split conditioning but such a reagent must be used

SUBSTITUTE SHEET

judiciously because some nickel can be lost if too much is added.

A talc depressant such as guar gum may be used, again judiciously. It has been found that the addition of small amounts of Guartec diminishes the extent to which talc slows down the rate of pentlandite flotation. Preferably most of the talc depressant, if used, is added to the fine fraction

Any suitable frother may be used in the flotation step, for example Propylene Glycol. The type and amount of ^f rother if any, need to be determine _d for each ore. For some ores, a frother might not be needed at all, or the Plant water available at the site may produce strong frothing obviating the need to use a frothing agent.

The rate of pentlandite flotation after split conditioning is not quite as rapid as that in the presence of copper sulphate and collector and a longer time is required for flotation. in laboratory tests, flotation times have been up to 16 minutes.

A further rejection of non-sulphides in a second stage may be necessary before the concentrate produced by the process of the second aspect is acceptable for smelting. This can be accomplished readily using a method which is in principle very similar to that used in conventional processing. By adding an activator such as copper sulphate and a collector to the concentrate produced by the process in accordance with the second aspect, the sulphides can be floated very rapidly while the floatability of the non-sulphides can be suppressed using a suitable depressant. By including such a stage, concentrates assaying only about 2 percent MgO can be produced without much additional loss of nickel.

SUBSTITUTE SHEET I

Smelters usually require that concentrates assay less than about 4 percent MgO.

Thus in yet a third aspect the invention provides a process in accordance with the second aspect comprising the further step of treating the concentrate produced by process of the second aspect with a collector for sulphides, an activator and optionally a depressant for non-sulphides and subjecting the treated product to froth flotation to produce a sulphide-containing froth.

The activator may be copper sulphate. We emphasise he_-e again that while copper sulphate increases the rate of flotation of pentlandite, once it has been added no further selection between pentlandite and pyrrhotite is possible. This is why copper sulphate is not used until after much of the pyrrhotite has been rejected. The depressent may be a talc depressant such as guargum.

The amounts of copper sulphate, collector and talc depressant needed are much the same as for conventional processing. In laboratory tests, we have achieved excellent results using 250 g/t of copper sulphate, 150 g/t of amyl xanthate and 550 g/t of the talc depressant Guartec. Other collectors and gangue depressants may also be used. Lowering the pH for the second stage might be advantageous in some circumstances. Sulphuric acid may be a suitable acid to use.

Particularly good results can be achieved by operating the second stage in a rougher-cleaner configuration. It is important to minimise the recovery of water in cleaning because rejection of the non-sulphides is ultimately limited by entrainment.

SUBSTITUTE 1

Pneumatic cells such as column cells are reported to give lower water recoveries and to be particularly efficient for cleaning and might therefore be better suited than conventional cells for the cleaning of concentrates.

We have also found that the presence of large amounts of talc ⁱ n nickel ores significantly reduces the rate of flotation of the sulphides, particularly when copper sulphate is not added. This problem can be overcome by allowing the talc to float without collector before sulphide flotation. Talc is naturally strongly floatable and the only reagent needed for such a pre-float is a frother. Frothers that give brittle froths, such as methylisobutylcarbinol (MIBC), are preferred. It may even be possible to pre-float without a frother.

Thus in yet a fourth aspect the present invention provides a process for the recovery of pentlandite from a sulphide ore containing pentlandite and pyrrhotite wherein the ore ⁱ s comminuted and subjected to froth flotation in the absence of a collector to produce a talc-containing froth and a sulphide-containing tailing and th _e reafter sub ^j ecting the tailing to one or more sulphide flotation step.

The sulphide-containing tailing produced by the process of the fourth aspect may be used as the feed for the processes of the invention in its first, second and third aspects.

A small amount of nickel is recovered to the talc concentrate owing to the non-selective contribution whi _r -h entrainment makes to the recovery of all minerals However, this loss can be kept to a minimum by cleaning the talc concentrate and by minimising the amount of water recovered during flotation. Control of water recovery may

SUBSTITUTE SHEET

be achieved by keeping pulp level and aeration rate low in the cleaning step. Additionally the use of froth modifiers such as tri-butyl phosphate may be considered although as such modifiers can sometimes adversely affect flotation, they should be tested beforehand in a laboratory preferably using talc from the plant.

Column cells might again be suited more than conventional cells for cleaning.

A second option for handling the talc concentrate is to recycle it to a final stage which is in accordance with the fourth aspect of the present invention where sulphides are floated from talc- Recycling the concentrate in such a manner would provide an opportunity for any floatable nickel to be removed whilst ensuring that the feed to the pyrrhotite rejection stage contains little fast floating talc.

A third option is to reject part of the talc concentrate and recycle the rest in the manner just described. In this instance treatment is by reverse flotation of cleaned talc concentrate so as to produce a talc product-unfloated fraction-that can be rejected and a sulphide concentrate that can be recycled to the final stage; reagents, and minerals contacted with reagents must be removed from the pre-float circuit. The activator, collector and depressant for non-sulphides used in the final stage may be added in the reverse cleaner instead of being added to the final stage.

For ores that contain small amounts of talc or that contain only poorly floatable non-sulphides a pre-float might not be necessary; the floatability of such minerals might be suppressed readily by adding a suitable depressant during sulphide flotation.

SUBSTITUTE SHEET j

In a preferred aspect the ore subjected to the process in accordance with the invention is ground and/or classified so as to minimise particles of a size less than about lOμm.

An important finding from our research is that poor floatability of fine (-10 μm) nickel places a very real limitation on the nickel recovery ultimately attainable. This limitation exists in our new process (more than half of the nickel we fail to recover in our laboratory tests is less than 10 μm in size) just as it does in standard laboratory tests using convention methods and in all of the plants for which we have recovery-size data.

We recognise the limitation imposed by the difficulty of selectively recovering fine nickel and seek to handle the problem by excellence in both grinding and classification. By efficient and precise grinding and classification and by careful regrinding of selected size fractions of selected streams the amount, of nickel ground to very fine sizes can be minimised.

To apply the present invention most efficiently the particle size dependence of flotation should be known for each of the ores to be treated. This size dependence can be established by careful sizing of flotation products. Data from such sizings allow the optimum size range for flotation to be determined (see, for example Figure 5).

Brief Description of the Drawings

Figure 1 is a general flowsheet showing one arrangement of steps in a process in accordance with the invention

SUBSTITUTE SHEET I

Figure 2 is a general flowsheet showing a second arrangement of steps in a process in accordance with the invention.

Figure 3 is a general flowsheet showing a third arrangement of steps in a process in accordance with the invention.

Figure 4 is a general flowsheet showing a third arrangement of steps in a process in accordance with the invention.

Figure 5 is a graph showing recovery-size dependance for one ore using a flotation method in accordance with the invention. N/P/S is a measure of pyrrhotite behaviour and NSG is a measure of the behaviour of all the non-sulphide gangue minerals including talc.

Figure 6 is a graph showing Nickel- N/P/S selectivity curves (ratio of average constants) for flotation of an ore ground in an open mild steel mill (test 1) and in a closed stainless stell mill (test 2 ) .

Figure 7 is a graph showing Nickel-MgO selectivity curves (ratio of average rate constants) for flotation of an ore ground in an open mild steel mill (test 1) and in a closed steel mill (test 2).

Modes for Using the Invention

A scheme including the talc pre-float and the first sulphide flotation stage of the second aspect of the invention is shown in Figure 1.

Referring to Figure 1, the feed ore, which has preferably been ground under substantially non-reducing conditions,

SUBSTITUTE SHEET 1

is supplied to a talc pre-float rougher which may be, for example, a series of Agitair cells. In this particular arrangement, the talc recovered from the pre-float is discarded whilst the sulphide-containing tailings are subjected to split conditioning, that is, the sulphide-containing tailing is separated into a relatively coarse fraction and relatively fine fraction and both fractions are conditioned with potassium n-amyl xanthate, the major proportion of the xanthate being used to treat the coarse fraction, and then the two conditioned fractions are recombined.

The conditioned sulphide concentrate is treated in two stages, stage 1 being the first stage of sulphide flotation and stage 2 the second. The conditioned sulphide concentrate is first treated in a stage 1 rougher at pH 9 in the absence of copper sulphate to selectively reject some of the pyrrhotite and recover a pentlandite-containing concentrate. The pentlandite-containing concentrate is then subjected to conventional sulphide flotation in a second stage rougher-cleaner arrangement in the presence of copper sulphate, potassium n-amyl xanthate and a talc depressant such as guartec to produce a final nickel concentrate and tailing streams which are rejected.

In practice, the tailings from both sulphide flotation stages (streams a, b and c in Figure 1) would be re-cycled or subject to further treatment so as to maximise nickel recovery. The most appropriate place to recycle such streams and the best methods of further treatment need to be determined on a case-by-case basis. For example, for ores that are difficult to grind or v/hich are finely disseminated much of the pentlandite in the tailings might be locked with other minerals. Regrinding of composite

SUBSTITUTE SHEET

particles would therefore be necessary before a further stage of flotation to recover the nickel.

Figure 2 illustrates an alternative arrangement of stages in which the talc-containing concentrate floated in the pre-float rougher is subjected to cleaning in a pre-float cleaner and where the tailings are returned to the pre-float rougher and the talc rejected.

Figure 3 illustrates another possible arrangement wherein the talc-containing concentrate from the talc pre-float rougher is combined with the sulphide-containing concentrate floated from the stage 1 rougher.

Figure 4 illustrates a further possible arrangement in accordance with the invention wherein part of the talc concentrate is rejected in a "reverse cleaner" and the rest is recyced to the final stage of the process where talc is rejected from the sulphides.

In order that the invention might be understood more readily the following non-limiting examples are provided.

Example 1 Effect on Sulphide Flotation Stage 1 Results of Grinding in Mills Constructed of Different Steels

The sample used was from a nickel deposit containing 2.38% Ni, 14.5% MgO, 6.79% S and 13.5% Fe. Mineralogical analyses indicated that the nickel occurred almost exclusively as pentlandite and that pyrrhotite was the predominant sulphide gangue mineral. The principal magnesium bearing minerals were talc, magnesite and silicates of the chlorite group. Microprobe analyses

SUBSTITUTE SHEET

showed that none of the gangue minerals contained more than about 0.5% nickel and that most contained much less.

Laboratory ball mills constructed of either stainless steel or of mild steel with media of similar type were used to grind the ore. Changing from one mill to the other produced little change in the particle size distribution after grinding. Products from grinding were 80 percent by weight passing about 75 μm. The ore was ground in 500 gram lots at 67 percent solids using distilled water.

A modified Denver laboratory cell was used for flotation Ow ⁱ ng to the presence of talc, tests included a 2 minute pre-float in which the only reagent added was a frother. The tailing from this pre-float was then wet-screened over a 75 _μm sieve and the two size fractions separately conditioned with collector before flotation.

Either potassium n-amyl xanthate (KnaX) or potassium ethyl xanthate (KeX) was used as collector. Three quarters of the collector was added to the sieve over-size and the remaining quarter was added to the under-size.

The frother was polypropylene glycol (Cyanamid Aerofroth 65) made up as a 0.25 percent solution. An initial dose of 2ml of frother solution was added before the talc pre-float and thereafter 2ml per minute was added from an automatic pump to maintain an active froth column.

The flotation gas was bottled air (a synthetic mixture of 0 ₂ and N ₂ ) and was supplied at a flow rate of 8 litres/minute. The pH was adjusted to 9 and controlled automatically at this value using a dilute sodium hydroxide solution. All water was distilled.

SUBSTITUTE SHEET j

- 16 -

Concentrates and tailings were sampled in a standard manner and assayed for Ni, S and MgO.

Results from the tests are given in Table 1. In this table the letters A and R refer to the cumulative concentrate assay and to the cumulative recovery of each species named. These values include losses to the talc pre-float. The behaviour of the non-pentlandite sulphur (N/P/S) minerals was calculated assuming that pentlandite contains 33. ₀ percent nickel and 33.0 percent sulphur. As has already been indicated, pyrrhotite was the predominant sulphide gangue mineral and it is therefore assumed that non-pentlandite sulphur is mostly pyrrhotitic sulphur although it is known that the ore contains some pyrite.

Comparing tests 1 and 2 in Table 1 shows that the pyrrhotite floated less strongly from ore ground in the stainless steel mill than from ore ground in the mild steel mill after the addition of n-amyl xanthate. The change in floatability was substantial with the N/P/S recovery being lowered by more than 13 percent. When ethyl xanthate rather than n-amyl xanthate was added, the pyrrhotite again floated less strongly from ore ground in the stainless steel mill (compare tests 3 and 4 in Table

1).

The improvement in the selectivity of separation of nickel from pyrrhotite when the stainless steel mill is substituted for the mild ^' steel mill can be seen clearly when the rate data for test 1 (mild steel mill grind) and for test 2 (stainless steel mill grind) are plotted in the form of selectivity curves (Figure 6). Assuming first order kinetics, the slope of the lines is the ratio of the average rate constants for the flotation of the nickel mineral and for the N/P/S minerals. A line of 45° indicates no selectivity; a line down the ordinate (the

SUBSTITUTE SHEET I

vertical axis) indicates perfect selectivity between Ni and N/P/S.

Plotting nickel-MgO selectivity curves for the same tests (Figure 7) shows that there is little or no change in the selectivity of separation of pentlandite from magnesium bearing minerals.

Table 1 - Stage 1 Sulphide Flotation Results after Grinding in a Steel Mill and after Grinding in a Stainless Steel Mill.

SUBSTITUTE SHEET

Example 2 Split Conditioning

The floatability of coarse nickel at pH 9 can be raised by using split conditioning. In this example the collector was divided between two parallel conditioning stages and flotation carried out for 16 minutes. The split conditioning approach was compared with the conventional process using copper silphate and ethyl xanthate and the results are shown in table 2.

Table 2 A comparison between the Efficiency of Split Conditioning and of adding CuSO-, for the separation of Nickel from N/P/S (16 min results)

Ni N/P/S MgO N/S/G H--

It will be seen from Table 2 that for equivalent Nickel recovery the split conditioning examples (35 and 27) recover less pyrrhotite as compared to the the conventional method (37) for the same level of pentlandite recovery.

Example 3 Results from a Laboratory Test using the Invention.

j SUBSTITUTE SHEET |

The first test of Example 1 was repeated but this time a second stage of sulphide flotation was added to reject non-sulphides from the concentrate. This second stage included both roughing and cleaning. The flotation time for the rougher was 8 minutes and for cleaning the time of flotation was 6 minutes. Pulps were conditioned before both the rougher and the cleaner and the order of addition of reagents was as follows:

Rougher: pH of the pulp adjusted to 7 using sulphuric acid

100 mg of CuS0 _/| 5H ₂ 0 conditioned for 2 minutes

20 mg of KnaX conditioned for 2 minutes

150 mg of Guartec conditioned for 2 minutes

Cleaner:

10 mg KnaX conditioned for 2 minutes 15 mg Guartec conditioned for 2 minutes

Additions of 20 mg of KnaX and 50 mg of Guartec were also made after 2 and 4 minutes of rougher flotation while additions of 10 mg of KnaX and 5 mg of Guartec were made after 2 and 4 minutes of cleaner flotation. Frother was added as required during both roughing and cleaning.

A particular effort was made to minimise entrainment in cleaning. To reduce the water recovery the pulp level was lowered, a minimum of frother was added and the froth scraping rate was decreased from every 5 seconds to every 10 seconds.

The results for each stage of the test are as follows:

Stage 2 100 mg CuS0 ₄ .5H_0 Rougher 50 mg KnaX

250 mg Guartec 8 minutes flotation pH 7

Of the nickel originally in the sample, 81% was recovered to the final (stage 2 cleaner) concentrate at a grade of 16.5%. This is a particularly good result for this ore and by way of comparison, standard laboratory lotation methods produce concentrates assaying only about 9% nickel at 80% recovery.

It should also be recognised that:

(a) none of the tailings were recycled, that is, the nickel which reported to the rougher tailings and the cleaner tailings were treated as a loss, and

SUBSTITUTE SHEET

- 21

(b) we used only a single stage of grinding without classification. As a consequence about 20% of the nickel was ground to less than 10 μm in size and of this we recovered less than half.

With further treatment and with recycling of products it is not unreasonable to e _P ect that the result can be improved still further, particularly when the limitations imposed by the poor floatability of the fine nickel are addressed as outlined in step 5 of our Disclosure of the Invention.

References

Anthony, R.M. , Kelsall, D.F., and Trahar, W.J., 1975. The effect of particle size on the activation and flotation of sphalerite. Proc. Australas, Inst.Min.Metall. , Vol.254, pp.47-58.

Trahar, W.J., 1981. A rational interpretation of the role of particle size in flotation. Int. J. Miner. Process., Vol.8, pp.289-327.

SUBSTITUTE SHEET