FIELD OF THE INVENTION

The invention relates to a method for exploiting dusts generated in a ferronickel process. The invention also relates to sintered pellets produced by the method.

BACKGROUND

Nickel-bearing laterite ores, which are used as the main mineral raw material in ferronickel process, are friable and dusty in character. The nickel content of laterites is usually between 0.95 and 3 w-%. Therefore, large amounts of ore material need to be handled in the ferronickel process. The slag to metal ratio in smelting is high, generally even over 10. Consequently, the elec ^¬ tricity consumption of the ferronickel process is high.

Laterite ores contain high amounts of different types of volatile compounds, in general over 10 w-%. One of the volatile compounds is goethite, the structure of which also contains nickel and hydrated silicates of Mg and (Mg,Al) . Nickel can also be present in the magnesium silicates. Further, laterite ores also contain 10-20 w-% moisture depending on the deposit location.

The ferronickel process starts with crushing and screening of the ore, in which connection finely divided ore dust is produced. Today, this dust is mainly stored in heaps, and only a minor part of dust is exploited in the ferronickel process.

Crushed and screened ore is first fed into a drying drum, where part of the ore's moisture content is removed. At this point, a small amount of wet dust is produced. At present the utilization of wet dust is dif ^¬ ficult and non-efficient.

From the drying drum the ore is fed to a rotary kiln where final drying and calcination of the ore and pre-reduction of nickel and iron take place. Due to the high crumbling tendency of the ore, lots of finely divid ^¬ ed material and dust are produced from the ore at this stage .

The amount of finely divided material removed from the rotary kiln is generally over 20 % of the weight of the ore fed to the rotary kiln. The dusts from the ro ^¬ tary kiln are first separated in a dry cyclone and then in a wet scrubber.

The grain size of both the cyclone dust and the scrubber dust is very small, usually less than 50 μιη. The specific surface of the dust is high due to the high po ^¬ rosity of dust particles.

The amount of dust produced in the ferronickel process is substantial and the nickel content in the dust is higher than in the original ore because the dusts are mostly calcined. The dusts also contain carbon which originates from coal that has been used as a reducing agent in the rotary kiln. Substantial amounts of nickel are bound to these dusts. At present the exploitation of nickel which is bound to these dusts is at a very low level due to several difficulties in the dust handling.

PURPOSE

The purpose of the present method is to elimi ^¬ nate, or at least reduce the problems of the prior art.

More precisely, the purpose of the present meth ^¬ od is to provide a new method for exploitation of nickel- bearing dusts formed in a ferronickel process.

SUMMARY

More precisely, the now presented method com ^¬ prises producing a pelletizing feed from dusts of the ferronickel process and from finely divided laterite ore with a particle size less than 3 mm, after which the pelletizing feed is pelletized to produce green pellets. The green pellets are sintered in a steel belt sintering furnace to produce sintered pellets. The steel belt sin ^¬ tering furnace comprises one or more zones for drying, heating, sintering and cooling the pellets. Finally, the sintered pellets are fed either to a rotary kiln for pre ^¬ reduction of nickel and iron or to a ferronickel smelter together with pre-reduced laterite ore received from a rotary kiln.

The steel belt sintering (SBS) process of Ou- totec can be used for the sintering.

According to one embodiment the particle size of the finely divided laterite ore is less than 2 mm.

According to one embodiment, dust from a dry cy ^¬ clone cleaner and/or dust from a wet scrubber can be in- troduced into the pelletizing feed.

According to one embodiment, the amount of fine ^¬ ly divided laterite ore introduced into the pelletizing feed is 0.1-50%, preferably 1-45%, more preferably 5-40% of the weight of the pelletizing feed. It seems based on performed tests (described later herein) that an addition of finely divided laterite ore (crushed ore) clearly im ^¬ proved the strength values of the pellets. It appears easier to produce durable pellets from nickeliferous pro ^¬ cess dusts by adding at least a small amount of finely divided laterite ore.

According to one embodiment, bentonite is added to the pelletizing feed in an amount of 0.1-2%, or alternatively 0.1-1.5%, preferably 0.5-1.2% of the weight of the pelletizing feed. It seems based on performed tests (described later herein) that an addition of bentonite improves the strengths of pellets.

According to one embodiment, the temperature of the drying gas supplied into the drying zone of the steel belt sintering furnace is 100-500°C and the retention time in the drying zone is 10-15 minutes. Alternatively, the temperature of the drying gas supplied into the dry ^¬ ing zone of the steel belt sintering furnace can be 250- 420°C and the retention time in the drying zone can be 10-12 minutes. Also a temperature of 300-420°C could be used .

According to one embodiment, the temperature of the heating gas supplied into the heating zone of the steel belt sintering furnace is 1000-1300° C and the re ^¬ tention time in the heating zone is 8-15 minutes. Alter ^¬ natively the temperature of the heating gas supplied into the heating zone of the steel belt sintering furnace can be 1100-1200° C and the retention time in the heating zone can be 9-11 minutes.

According to one embodiment, the temperature of the sintering gas supplied into the sintering zone of the steel belt sintering device is 1150-1350° C and the re- tention time in the sintering zone is 10-16 minutes. Al ^¬ ternatively the temperature of the sintering gas supplied into the sintering zone of the steel belt sintering de ^¬ vice can be 1200-1250° C and the retention time in the sintering zone can be 10-16 minutes.

According to one embodiment, the temperature of the material bed in the sintering zone is in the range of 1300-1500° C. Alternatively the temperature of the mate ^¬ rial bed in the sintering zone can be in the range of 1360-1400° C.

Sintered nickel-containing pellets produced by a method presented herein can have nickel content even higher than 2%, in some cases higher than 4%. The Fe/Ni ratio of the sintered pellets can be lower than 10.0, preferably lower than 5. Alternatively, the Fe/Ni ratio of the sintered pellets can be lower than 5.0, preferably lower than 4.8. The apparent porosity of the pellets can be 20-40%, or 20%-35%, preferably 25-30%.

Producing sintered pellets from ore fines and dust generated in a ferronickel process enables more ef- ficient exploitation of waste materials which have earli ^¬ er been stored in heap. Sintered pellets can be fed into a rotary kiln for pre-reduction before smelting. Alternatively, sintered pellets can be fed to smelting together with pre- reduced laterite ore received from a rotary kiln.

The new method enables cost-effective utiliza ^¬ tion of ferronickel dusts which previously have been stored in heaps.

The method is environmental friendly in that it enables exploitation and recycling of waste materials (the dust produced in the ferronickel process) .

DETAILED DESCRIPTION

Laterite ores are very friable and lots of fines and dust are formed during their handling. The amount of nickel in these fines and dusts is remarkable and valua ^¬ ble. The pelletizing and sintering of the mixtures of fi ^¬ ne ore and dusts have been clarified in the laboratory and pilot plant scales. Pelletizing and sintering were investigated using a pelletizing disc and a batch reactor (certain type pot furnace) which simulated the real pro ^¬ cess. The target was to use a steel belt sintering ma ^¬ chine for the sintering of the pellets. The temperature profiles and retention times in different zones of steel belt sintering machine were specified to reach sintered pellets suitable for smelting.

In the first set of experiments, only dusts from dry cyclone cleaners and wet scrubbers were used in dif ^¬ ferent proportions as a raw material. In the second set of experiments, dry and wet dusts were mixed with finely divided ore in different proportions. Finely divided ore with a particle size less than 2 mm was added to the mix ^¬ ture in an amount up to 50 w-%. Finely divided ore with a particle size less than 3 mm can be used as an alterna ^¬ tive. In some of the test results shown herein finely di- vided ore with a particle size less than 2 mm was added to the mixture in an amount of 30 Chemical analysis of the raw materials used in the tests is shown in Table 1. The composition of dry cy ^¬ clone fines and wet scrubber slime was quite similar. Both the cyclone fines and the scrubber slime contained more nickel and total iron (Fe _to t ) than the laterite ore. The loss of ignition (L.O.I.) and moisture content was lower in the dusts than in the laterite ore.

TABLE 1 Chemical analysis of ore and dusts

The ore was crushed and screened to reach a par ^¬ ticle size less than 2 mm. The grain size of the cyclone dust and scrubber slime was less than 50 ym and less than 44 ym, respectively. The dusts were received from the gas cleaning of a rotary kiln.

Bentonite was added as a binding agent into the pelletizing feed. The amount of bentonite was 0.1-1.0% of the dry weight of the feed. The amount of coke or other carbon-bearing material added to the pelletizing feed was 0-2% depending of the carbon content of the dusts.

The moisture content of the wet pellets was 19- 21% because of the very fine grain size and high porosity of the pellets produced. The cold compressive strength of wet pellets was 1.3-2.0 kg/pellet of 12 mm diameter. The cold compressive strength generally refers to the com- pressive strength when the pellet is cooled, ie. when it is a room temperature (eg. 16°C, 20°C or 24°C) . After drying the cold compressive strength of green pellets was 4-12 kg/pellet of 12 mm diameter. The green pellets were strong enough for sintering.

Sintering of green pellets was carried out in simulated steel belt sintering equipment. Process parame ^¬ ters were adjusted based on the composition of the sin ^¬ tering feed and the properties of the resulting product.

The compression strength of sintered pellets was in the range of 80-210 kg/pellet in different parts of the product bed. The diameter of the pellets was a little smaller than 12 mm because the pellets were a little bit compressed due to the high porosity of green pellets.

The results of chemical analysis of sintered pellets are shown in Table 2. The nickel content of sin ^¬ tered pellets was over 4 w-%. Iron (Fe _to t) was mainly pre ^¬ sent as Fe ³⁺ . The Fe/Ni ratio was about 4.6 at its best.

In the steel belt sintering equipment the gas temperature in the drying stage was 250-420° C. The re ^¬ tention time in the drying stage was 10-12 minutes be ^¬ cause of the high moisture content of the pellets.

Table 2 Chemical analysis of the sintered pellets.

The temperature of heating gas supplied into the heating stage was 1100-1200° C. The retention time in the heating stage was 9-11 minutes. Product bed temperatures between 1270 and 1370° C could be reached in the heating stage, depending on the composition of the feed and the amount of combustible coal present.

The temperature of the sintering gas supplied into the sintering stage was 1200-1250° C. The retention time in the sintering stage was 10-16 minutes. The tern- perature of the bed was between 1360 and 1400° C during tests where pellets of good quality were produced.

Sintered pellets were highly porous. The appar ^¬ ent porosity of the pellets was in the range of 25-30%, which is why the reducibility of the pellets was good. In pre-reduction tests carried out at 800° C with CO gas and 5% coke in the mixture, the degree of nickel metalliza ^¬ tion was over 60% and the degree of iron metallization was over 40%.

The abrasion resistance of the pellets was high enough to withstand a further treatment in a rotary kiln followed by smelting, or just smelting together with pre- reduced laterite ore received from a rotary kiln.

To get a better understanding of the effective- ness and functioning of the process pelletizing and sintering tests in the pilot scale were performed. A de ^¬ scription of the test is explained in the following. These test results verify functioning of the process and give the real basis for the designing of the commercial scale process.

Batch pelletizing tests were performed. The pur ^¬ pose of batch pelletizing was to study the pelletizing properties of a mixture of three different materials and the mixtures of them. The chemical analyses of the test materials were as shown herein in Table 1 already.

Pelletizing was carried out on a disc with a di ^¬ ameter of 2.0 m and a depth of 0.30 m. The maximum rota ^¬ tion speed of the disc was 20 rpm. The ground laterite ores and dusts were mixed with a binding agent and fine coke in some tests. The mixing time was 10 minutes. A little water was added into the mixer in some tests.

The batch was fed manually onto the disc. The pelletizing feed was moistened with water sprayers ac ^¬ cording to the formation of the pellets. The pelletizing time was about 50-60 minutes. The desired size of the pellet (ie. diameter of the pellet) was 12 mm. The rota- tion speed of the disc was adjusted according to the for ^¬ mation of the pellets.

The quality of the green pellets was determined by measuring the cold compressive strengths of the green and dry pellets. The moisture content of the green pel ^¬ lets was also measured.

Batch sintering tests were performed using a batch sintering system illustrated in Fig. 1 that included a gas (eg. butane) burner 21, a burning or combustion chamber 22, a sintering reactor 28 and gas lines 23. The gas lines 23 were equipped with the water-cooled valves 24 for leading the combustion gases to the reactor 28 and the off-gases (or out-gases) via gas lines 27 to the gas cleaning system (not shown in the figure) . 25 is an inlet for extra air for controlling the gas temperature.

The sintering process was controlled continuous ^¬ ly by an automatic process control system. The reactor 28 includes a metal grate 26 and is lined with refractory and insulating castable mixes.

Fig. 1 shows the positions of several thermocou ^¬ ples installed inside the pellet bed to measure the tem ^¬ peratures during the sintering process. 3 and 11 indicate thermocouples of the feed gas, 4,5,7,12 and 13 indicate thermocouples of the pellet bed, 8 indicates a thermocou- pie of the protection layer, 9 indicates a thermocouple of the grate and 10 indicates a thermocouple of the off- gas. In the batch sintering tests, the temperature pro ^¬ files inside the pellet bed during sintering were meas ^¬ ured. These values give information on the sintering tem- peratures. Also there was a pressure drop during the test from the upper part towards the lower part of the sinter ^¬ ing reactor.

Butane gas was burned with air or with oxygen enrichment, depending on the process stage. The batch sintering process comprised the following stages: i. dry ^¬ ing with combustion gases, ii. heating with combustion gases using oxygen enrichment, iii. sintering with com- bustion gases using oxygen enrichment, and iv. cooling with air.

For each zone, the amount of gas and retention time was pre-selected . The temperature of the combustion gases was controlled by the amount of butane, air ratio and oxygen enrichment. The combustion gases entered the reactor from above. Cooling air 5 was ducted into the re ^¬ actor from above. Moist pellets (the area indicated by reference 29, Fig. 1) were charged into the reactor on a bottom layer (the protective layer, indicated by refer ^¬ ence 30) of the chromite sintered pellets. The weight of the moist pellets was about 15 - 18 kg.

The sintered pellets were discharged from the reactor after the program was finished and the system was cooled (to room temperature) . The sintered pellet bed was divided into three different sections (upper, middle and lower layers) for laboratory tests. The bottom layer was changed after every test.

Chemical analysis of the raw materials used in the tests is shown in Table 1. The nickel contents of the cyclone dust and scrubber slurry are clearly higher than that of the ore as shown in Table 1. The dusts were al ^¬ most totally calcined and thus their nickel content was high .

In some of the tests bentonite was used as bind ^¬ ing material. Table 3 shows the chemical analysis of the bentonite used in the tests.

Table 3 Chemical analysis of bentonite.

Chemical analysis of the fine coke used in pilot plant batch sintering tests is shown in Table 4. Table 4 Chemical analysis of fine coke. Vola- Cfix A1 ₂ 0 ₃ MgO CaO S Na ₂ 0 K ₂ 0 Fe _tot S10 ₂ tiles [%] [%] [%] [%] [%] [%] [%] [%] [%]

1.5 78.3 3.3 0.1 0.2 0.8 0.2 0.2 0.9 5.8

Table 5 shows the screen size of the Ni laterite ore .

The grain size of the original ore samples were below 25 mm and the average grain sizes were about 5 mm. The grain size of crushed ore was less than 2.38 mm and it contained 51.5 w-% less than 74ym.

Both dusts were very fine and the measured spe ^¬ cific surface areas were 16.14 m ² /g for cyclone dust and 35.05 m ² /g for scrubber slurry. The grain size of cyclone dust was less than 50ym and the average size 15ym and those of scrubber slurry were correspondingly 44ym and 4.6ym.

Table 5 Screen analysis of ore.

The screen analysis of the fine coke for the pelletising and sintering was as follows:

41 % -74 ym

25 % -37 ym

The grain size of bentonite was below 40ym.

Main phases of test materials are presented in Table

Table 6 Main phases of the test materials.

Phase Laterite Cyclone Scrubber ore fines slurry

Mg ₃ Si ₂ 0 ₅ (OH) ₄ , X X clinochrysotile

(Mg,Al) ₃ (Si,Al) ₂ 0 ₅ (OH) ₄ X

lizardite

The grain structure of the ore was loose and plenty of breaks appeared. The grains also contained dif ^¬ ferent phases. Therefore it is not durable against abra- sion and substantial dust formation in the rotary kiln can be predicted.

The microstructure of the ore mainly consisted of different silicates and iron oxides. The dusts con ^¬ tained different calcined magnesium silicates, silica, magnetite and hematite. Nickel appeared as trevorite in the dusts.

Table 7 shows the tested batches and the mois ^¬ ture content of the wet pellets. As can be seen in some of the batches about 30% finely divided laterite or was added to the pelletizing feed.

Table 7 Batch composition of the tests.

At first, pelletising tests were performed in the laboratory, but the strengths of the green pellets were too weak. Some pelletizing tests were carried out with a larger disc situated in the pilot plant. With the larger disc, it was easier to make compact and satisfac ^¬ tory strength of the pellets. This kind of effect is fre ^¬ quent when very light and fine materials are pelletized. The feed mixture and average strengths (wet and dry strengths) of the pelletizing pellets (a larger disc) are presented in Table 8. The amount of bentonite in these tests was 1 % and pre-wetting degree in a mixer was 5 %. Table 8 Batch composition, moistere content and wet and dry strength of the pellets.

where

1) crushed ore to - 2 mm

2) when the diameter of the pellet is 12 mm then F _j

where

D = the diameter of the measured pellet

F _b = the compressive strength of the measured pellet

The wet and dry strengths of the pellets in the selected tests were high enough for the sintering pro ^¬ cess. It seemed that an addition of bentonite improved the strengths of pellets. An addition of crushed ore (of about 30%) clearly improved the strength values of the pellets. The moisture content in pelletizing of dusts was about 20

The cold compressive strengths of the sintered pellets in the selected tests are presented in Table 9 (proportioned to 12 mm in diameter) .

Table 9 Cold compressive strength of the sintered pellets .

The strength of the sintered pellets was good in many tests when the process parameters were in the de ^¬ sired values. The strength was good both with fine dusts and with dusts with coarser crushed laterite ore. Good quality pellets were formed in the central area of the batch throughout the pellet bed (vertical) . Close to the walls the pellets did not react enough due to about 100°C lower temperature than in the central area. This was due to the irregular radial temperature profile in the batch reactor. However, in a continuous process (such as the steel belt sintering (SBS) process of Outotec) the tem ^¬ perature at the edge areas are very even to central area temperatures, and thus should not have the drawbacks of the test reactor.

Fine coke was also added to the feed mixture in some tests and it resulted in metallic nickel formation. Then the reduction gases erupted from the pellets and caused fragmentation. Molten phases were also noticed on the surface of the pellets after the tests. After notic ^¬ ing these phenomena, no coke was further added to the feed mixture.

The chemical analyses of the two selected batch ^¬ es represent well the quality of the sintered pellets. The values are presented in Table 10 as w ~6.

It should be noticed that the volatile compo ^¬ nents, ie. L.O.I (here hydroxides) have decomposed and removed from the pellets. Iron was well oxidized in these tests, with about 90 % of the iron was in ferric form. Table 10 Chemical analysis of the sintered pellets,

The porous structure of the pellets increases the reducibility of nickel and iron compared to the pri ^¬ mary laterite ore. Nickel appeared to be in ferrite grains and in forsterite. Some ferrite particles were rich in nickel. The ferrites contained in points plenty of chromium. Abrasion tests of sintered dust pellets resulted in considerably high amount of dust with a size of less than 0.59 mm. This result was not surprising, because of the earlier mentioned irregular temperature profile in the bed, and accordingly this should not be an issue in a con ^¬ tinuous process. The pellets from the central area of the pellet bed have good abrasion resistance, which corre ^¬ sponds to the situation in a commercial scale.

Some physical values (porosity and density) of the sintered pellets from the tests are presented in Ta ^¬ ble 11.

Table 11 Apparent porosity and density of the sintered

pellets .

The apparent porosity values were high and it advances the high reduction.

Pre-reduction tests of the sintered pellets were performed in a vertical tube furnace where coke and CO- gas were used as a reductant. This tube furnace simulated a vertical pre-reduction vessel. The furnace was equipped with a grate and gas flew upwards through the pellet bed. Three thermo elements were installed inside the bed at different heights from the top of the furnace.

The furnace was heated for two hours to 800°C and kept for two hours at this temperature. Then inert gas was led into the furnace and the heating current was switched off.

Sintered pellets were highly porous. The appar ^¬ ent porosity of the pellets was in the range of 25-30%, which is why the reducibility of the pellets was good.

In the pre-reduction tests carried out at 800°C with CO gas and 5% coke in the mixture the degree of nickel metallization was over 60% and the degree of iron metallization was over 40%.

The reduction rate can be increased by optimiz ^¬ ing the conditions. The structure of the pre-reduced pel- let was dense enough for subsequent handling and the dis ^¬ integration was small.

In the test with the equipment shown in Fig. 1 the retention time 12 min in the drying zone was long enough for these wet pellets without any disintegration. The retention times in preheating was 9 - 10.5 min, and especially in sintering was 12 - 15 min, and these also varied according to the test results and the visual ob ^¬ servations .

A feed gas temperature of max 420 - 440°C was found suitable for drying the pellets. The boundary layer between the fresh and bottom pellets was also dried.

The maximal feed gas temperature in the preheat ^¬ ing zone was 1130 - 1190°C, depending on the retention time. The maximal temperature of the upper bed was then in the range of 1300°C - 1370°C. The maximal temperature of the middle part of the bed was 1270 - 1370°C. The tem ^¬ perature of the boundary layer was 1270°C - 1330°C.

The maximal feed gas temperature in the sinter ^¬ ing zone was 1200 - 1230°C. Then the maximal temperature of the upper bed was 1370 - 1400°C. The maximal tempera ^¬ ture of the middle part of the bed was 1360 - 1380°C. The temperature of the boundary layer was 1360 - 1370°C.

In a real continuous process also other values can be used (other than those used in the test) that have been presented herein.

It is obvious to a person skilled in the art that with the advancement of technology, the basic idea of the method presented herein may be implemented in var ^¬ ious ways. The method and its embodiments are thus not limited to the examples described above; instead the method may vary within the scope of the claims.