PRODUCTION OF FERRONICKEL FROM NICKEL LATERITES

IN A DC-ARC FURNACE

H. Lagendijk and R.T. Jones

Pyrometallurgy Division, Mintek, Private Bag X3015, Randburg, 2125, South Africa

E-mail: HermanL@mintek.co.za rtjones@global.co.za

Nickel – Cobalt 97

36th Annual Conference of Metallurgists, Sudbury, Canada, August 1997

ABSTRACT

Laterites and other oxidized nickel ores constitute a very important part of world-wide nickel reserves. In the conventional production of ferronickel from these ores, much fine material is produced which cannot readily be accommodated directly in existing three-electrode or six-in-line AC furnaces. DC-arc furnace technology allows ore particles less than 1 mm in size to be treated directly, thereby improving the overall recovery of nickel without the need for expensive agglomeration techniques. Because of the high moisture content of laterites, the ores should be dried and calcined before smelting. In order to decrease the energy consumption further, the ores could also be pre-reduced. The CO-rich off-gas from the furnace could be used to supplement the energy requirements, and is also a good reducing agent. Because fine ore particles are readily treated in a DC-arc furnace, units such as fluidized beds (which require materials of small particle size) can be used for the pre-treatment stage of the process.

A process has been developed whereby nickel laterites of a wide compositional range can be smelted in a DC-arc furnace, to produce ferronickel. The flexible operation of a DC-arc furnace (especially its lower dependence on electrical properties of the slag, because of open-arc operation, in addition to the ability to run at an optimum slag temperature, due to the open-bath mode of operation) allowed for the successful treatment of ores with a SiO₂/MgO ratio between 1.2 and 3.0, as well as ores containing up to 30 per cent by mass of iron (which tends to cause slag foaming in a conventional immersed-electrode furnace). A frozen lining can be maintained between the molten bath and the refractory lining, in order to minimize refractory wear (especially at high SiO₂ contents). Results of furnace testwork at power levels up to 750 kW are presented. Tests at the 120 kW furnace scale, together with some preheating in a 300 mm diameter bubbling fluidized bed, are also described.

INTRODUCTION

The name laterite (from the Latin, later, ‘a brick’) is used to describe the weathering product of ferruginous rock exposed to strongly oxidizing and leaching conditions, usually in tropical and subtropical regions. This porous, claylike rock comprises hydrated oxides of iron, aluminium, etc.

The ‘oxidized’ ores of nickel constitute by far the world's largest known reserves of this metal. The ores include the true laterites (in which the nickel oxide is intimately associated with limonitic iron oxide) and the silicate ores which often contain the mineral garnierite. These oxidized ores are found in regions of the world where tropical weathering occurs, or where at least sub-tropical conditions have prevailed in past geological times.

Mined laterite ore, being of a porous nature, can hold a large content of free moisture, commonly 25 to 30% H₂O, although it can contain even 40% or more. In addition to this, combined water, which is not completely driven off until a temperature of 700 to 800°C is reached, can amount to up to 15% based on the dry ore weight. Because of the quantity of water present, and because water requires so much energy to evaporate and heat up, it is clear that some pretreatment of the ore (at least drying and calcining) is required before smelting.

Because of the friable nature of laterites, run-of-mine ore is normally screened as the first step of ferronickel production. In the screening operation alone, up to 50% undersize particles may be screened out because of the difficulties in effecting a physical separation of wet laterite at a smaller size than 6 to 8 mm. This material would then require some form of agglomeration (e.g. sintering or pelletizing) before further processing. It is conceivable that the material could be dried before screening to say 1 or 2 mm prior to calcining. However, fines and dust are generated in the drying process, and dust generation in the calcining step is also increased. The large quantities of dust that are generated in the drying and calcining steps are unsuitable for smelting in a conventional electric-arc furnace. Once again, to utilize these dusts, they would need to be agglomerated at considerable expense.

Large quantities of laterite fines and dust have been stockpiled over the years. These can be used profitably as a feedstock for the production of ferronickel.

The chemical composition of laterite usually varies considerably, and this adds to the difficulty in processing this material. Normal electric furnaces, operating with a layer (or partial layer) of feed material on top of the melt, are reportedly difficult to operate with an SiO₂/MgO ratio greater than 2.0, or an iron content of more than 20% by mass, as these conditions can cause operational instability, mainly due to a tendency to slag foaming. These compositional problems can be overcome, to some extent, by blending different ores.

DC laterite process

DC-arc furnace technology can be used to provide an economical process for the smelting of nickel-containing laterite ores and dust, to produce a crude ferronickel product that could be refined to saleable ferronickel by conventional techniques.

Mintek has been working on the production of unrefined ferronickel from nickel-containing laterite in DC-arc furnaces since 1993. In this process, lateritic material is fed, together with a carbonaceous reducing agent, to the central region of the molten bath of a cylindrical DC-arc furnace. This feed material is preferably pre-treated, i.e. hot (calcined), and optionally pre-reduced.

DESCRIPTION OF A DC-ARC FURNACE

The development of DC-arc furnace technology at Mintek has been described in detail previously (1). Mintek’s 500 kW pilot plant, being a typical DC-arc furnace with ancillaries, was utilized for some of the larger-scale testwork described in this paper. The DC-arc furnace (Figure 1) comprises a water-cooled refractory-lined cylindrical shell, a conical roof, a graphite electrode, and an anode configuration comprising various pins protruding through the hearth refractory.

The furnace has a single electrode (cathode) situated centrally in the roof and positioned above the molten bath. The electrode may be hollow, and the feed materials may be introduced through the hollow centre of the electrode. The molten bath forms part of the electrical circuit (anode). The return electrode, or anode, consists of multiple steel rods built into the hearth refractories and connected at their lower end to a steel plate which, via radially extending arms, is linked to the furnace shell, and further to the anode cable.

The water-cooled roof is lined with an alumina refractory, and contains the central entry port for the graphite electrode and three equi-spaced feed ports. The outer sidewalls of the furnace are spray-cooled with water to protect the refractories, and to promote (together with a tight control of the power and feed rate of material to the furnace) the formation of a slag freeze-lining within the vessel.

The furnace is fed more or less continuously, and is tapped periodically (or even continuously, if so desired). The furnace is operated at a slightly higher pressure than that of the surrounding atmosphere, in order to substantially exclude the ingress of air into the furnace.

Figure 1 - Schematic diagram of a DC-arc furnace

The solid-feed system for the furnace comprises a batching plant and a final controlled feeding system. The batching plant consists of feed hoppers mounted on load cells, vibratory feeders positioned under the hoppers, an enclosed belt conveyor, a bucket elevator, and a pneumatically activated flap valve to direct the feed to one of two final feed hoppers. The final feeding system is made up of separate centre and side feeding arrangements. The centre feeder comprises a screw feeder discharging into a telescopic pipe attached to the hollow graphite electrode. The side feeders are vibratory, and discharge into feed chutes leading to the feed ports.

The gas-cleaning system consists of a water-cooled off-gas pipe, a refractory-lined combustion chamber, water-cooled ducting, a forced-draft gas cooler, a reverse-pulse bag filter, a fan, and a stack. The condensed fume and dust, which accumulates in the lower conical section of the bag plant, is discharged via a rotary valve into a collecting drum. This dust would, of course, be recycled back to the furnace in an industrial situation.

PRE-TREATMENT

Energy may be recovered from the process by using the thermal calorific value of the off-gases from the furnace to assist in the energy requirements for drying or calcining of the laterite feed. The CO-rich off-gas may also be used for pre-reduction purposes.

It is advantageous to pre-reduce the laterite prior to its introduction into the furnace. In such a case, the laterite ore is dried first, followed by dry milling, and calcining at 700 to 900°C in a fluidized bed. The calcined laterite is then pre-reduced in a fluidized reduction reactor, using a solid carbonaceous or gaseous reductant, at 800 to 850°C, prior to feeding to the furnace bath. As the DC-arc furnace can smelt fine materials, a fluid bed reactor can be linked to the furnace.

The above variant of the process has even greater advantages in cases where the SiO₂/MgO ratio is what would normally be regarded as excessively high and the overall reduction requirement is conducted in the pre-reduction stage. No SiO₂ would be reduced in the pre-reduction step, and accordingly very little silicon appears in the molten metal produced in the furnace. Also, by selective pre-reduction of NiO and Fe₂O₃ in the pre-reduction step, the ratio of Ni to Fe in the metal can be controlled to obtain a high-grade ferronickel. A lower-powered furnace can be used, as the melting requires less energy than is the case for the smelting reactions (especially if some SiO₂ were also to be reduced to some significant extent in the smelting process).

FURNACE OPERATION

The feed rate of materials and the energy input into the furnace are adjusted to achieve and maintain desired bath and tapping temperatures of the slag and metal. Cooling of the furnace walls assists in the formation of a slag freeze lining, which is particularly required in the case where the SiO₂/MgO ratio is greater than 1.5, to protect the refractories from excessive wear.

The carbonaceous reducing agent is added in such quantities that the oxygen in the off-gases is substantially in the form of carbon monoxide, and the nickel content of the slag is below 0.15%. The temperature of the furnace is controlled to between 1500 and 1700°C, depending on slag composition.

EXPERIMENTAL RESULTS

Example 1 - 50 kW tests

At this preliminary small-scale stage of experimentation, smelting tests were carried out on:

a) laterite ore fines and dusts (even 100% less than 100 µm),

b) partially calcined laterite ores, and

c) nickel-containing slags.

Tests were conducted in a 50 kW furnace with an outside diameter of 600 mm, and a refractory lining thickness of 114 mm. The refractory material had a 96% MgO content. The hearth was lined with a chrome-magnesite rammable material to a thickness of 310 mm, and a number of mild steel rods were used to make the DC (anode) electrical connection from the molten bath through the hearth refractory to the anode cable. The molten bath in the furnace was heated to the desired operating temperature, with an initial metal charge.

The feed materials consisted of calcined laterite dust (< 100 µm) from an industrial rotary kiln calciner (for two tests), and laterite ore fines (< 6 mm) dried at 250°C (also for two tests). Charcoal (< 4 mm) was used as the reductant in all these tests. The compositions of the feed materials, and masses of the feed materials and products, are shown in Tables I and II respectively. The feed materials were passed through a feed port in the furnace roof into a reaction zone, and the liquid products were tapped intermittently from the furnace. Some additional tests were carried out using a somewhat smaller furnace shell, in order to increase the number of samples obtainable at this scale of operation. This smaller furnace was equipped with water-cooling of the sidewalls. Results of the smelting tests, showing metal and slag compositions, are presented in Tables III and IV respectively.

Table I - Composition of feed materials for 50 kW tests, mass %

Component	Laterite fines ( dried at 250^oC)	Laterite dust	Charcoal
NiO	1.96	2.45	-
Fe2O3	39.1	36.8	-
MgO	12.2	17.5	0.2
SiO2	30.7	34.4	4.3
Al2O3	5.60	3.83	0.8
CaO	0.50	0.25	0.4
MnO	0.69	0.58	-
Cr2O3	2.51	1.15	-
Fixed carbon	-	0.79	64.0
Moisture	9.17	2.0	5.6
Volatiles	-	-	23.3
Total	102.43	99.75	98.6

Table II - Masses of feed materials and products of the 50 kW tests

		Feed materials, kg					Products, kg
Test Series	Batch	Mild Steel	Laterite fines	Dust	Char-coal	Slag	Ferro-nickel	Slag
A	Start 1 2 3 4 5	1.5		5.9 7.0 7.0 7.0 7.0	0.51 0.52 0.44 0.44 0.37		0.22 0.12 1.86	4.3 9.5 8.3 16.6
B	1 2 3 4 5 6		7.0 7.0 7.0 7.0 10.0 7.0		0.60 0.52 0.44 0.44 0.63 0.37			5.1 4.2 8.6 27.5
C	Start	6.0
	1 2 3			10.0 10.0 10.0	0.63 0.63 0.63		0.08	21.4
Digout							7.00	8.6
D	Start 1 2 3 4 5 6	4.5 + 1.5 Ni	5.0 5.0 5.0		0.11 0.11 0.12 0.70 0.70 0.84	4.9 4.0 5.0	0.20 0.78 0.27 8.00	2.3 2.5 2.1 3.3 2.6 4.9

Table III - Metal compositions from 50 kW tests, mass %

Test series	Batch	Ni	Fe	Si	Cr	P	S	C
A	3 4 5	0.06 0.21 14.70	99.1 99.6 84.3	0.15 0.11 0.05	0.08 0.05 0.06	0.023	0.10	0.05
C	3	7.82	90.5	0.18	0.04			0.10
Furnace contents		8.43	90.9	<0.02	0.05			0.02
D	5 6	23.30 16.60	72.8 83.1	0.75 <0.05	0.20 0.40

Table IV - Slag compositions from 50 kW tests, mass %

Test	Batch	NiO	FeO	SiO2	Cr2O3	MgO	CaO	MnO	Al2O3
A	2 3 4 5	0.11 0.06 0.04 0.19	43.1 42.8 41.9 40.2	20.3 21.4 23.2 17.7	5.45 4.98 4.88 6.23	27.6 27.0 27.0 30.9	0.21 <0.02 <0.02 <0.02	0.52 0.53 0.53 0.47	4.67 3.99 4.05 4.72
B	3 4 5 6	1.37 1.35 1.04 1.18	36.7 37.9 38.5 37.3	25.4 24.9 26.8 28.0	4.74 4.69 3.70 4.12	26.5 24.8 24.3 25.3	0.40 0.46 0.49 0.40	0.57 0.60 0.63 0.59	5.89 5.80 6.14 5.55
C	3	0.15	29.9	29.3	4.09	28.3	0.39	0.65	4.73
D	1 2 3 4 5 6	0.24 0.88 0.01 0.03 0.71 0.19	32.3 33.1 33.4 23.0 26.0 16.5	25.0 24.8 26.5 34.1 34.3 36.7	3.73 4.21 3.13 3.83 2.85 3.55	29.3 29.8 28.3 28.4 27.4 32.9	0.55 0.55 0.64 0.70 0.76 0.70	0.96 0.72 0.74 1.12 1.00 0.95	7.10 6.43 6.93 8.51 8.16 8.83

From the above results, it can be seen that even extremely fine (< 100 µm) nickel-containing laterite can be utilized effectively in a DC-arc furnace. Nickel levels in the slag could be lowered below 0.1%, when laterite fines and dusts were smelted, as well as when nickel-containing slags were processed.

Relatively high power fluxes at this small scale, and high slag temperatures around 1700°C, caused substantial erosion of the MgO-based sidewall refractory material, as is evident from the differences between the SiO₂/MgO ratios of the feed materials and that of the tapped slags, as summarized in Table V.

Table V - Summary of 50 kW test: Power, power flux per unit hearth area, tapping temperatures, as well as the SiO₂/MgO ratios of the feed materials and that of the slags

Test No.	Furnace ID, m	Average Power, kW	Water- cooling (Y/N)	Power flux, kW/m^2	SiO2/MgO ratio in the feed material	SiO2/MgO Ratio in the Slag	Temperature, °C
A	0.372	47.5	N	437	1.96	0.90	1710
B	0.372	46.4	N	427	2.52	1.08	1693
C	0.372	47.6	N	438	1.96	1.04	1676
D (Heats 1-3)	0.200	29.1	Y	926	1.02	0.87	-
D (Heats 4-6)	0.200	27.9	Y	888	2.52	1.18	-

Example 2 - 120 kW test

A 120 kW DC-arc furnace was used for the processing of laterite calcine blends, with the objective being the study of the smelting behaviour of materials with SiO₂/MgO ratios ranging from 1.2 to 3.0. The smelting of laterite types having a range of Fe contents from 15 to 20% was tested at the same time. A part of the test was devoted to the smelting of lateritic material that was preheated in a bubbling fluidized-bed reactor linked to the DC-arc furnace.

High-quality magnesite refractories were used for the sidewalls and hearth. The internal diameter of the furnace was 760 mm. The outer shell was spray water-cooled to protect the sidewall refractories. Two tapholes were used; the lower taphole for the metal product, and the upper for slag drainage. The two-taphole system ensured efficient slag-metal separation.

The smelting campaign was conducted on a continuous basis over a period of seven days, and about 7.2 tons of material was smelted. Prior to the smelting testwork, the laterite ores were calcined to reduce the loss on ignition (LOI) value from 11.5% (dry basis) to a residual LOI of 6.5%.

As a separate condition, about 600 kg of the feed materials were preheated (prior to smelting) to about 600°C in a fluidized bed, using liquefied petroleum gas (LPG). A schematic layout of the fluidized bed is shown in Figure 2. An enlargement of diameter in the upper portion of the reactor allowed solids to drop out of the gas, in order to decrease the carryover of fines in the off-gas.

Figure 2 - Schematic diagram of the 300 mm ID bubbling fluid bed reactor, which was linked to the 120 kW DC-arc furnace

For all tests in which laterite was fed directly to the furnace, the laterite was screened to a particle size of less than 8 mm. In the case of heating the feed materials to 600°C in the fluidized bed, screening was carried out to a size range of less than 2 mm.

Table VI summarizes, on a test condition basis, the masses and critical chemical analyses of the feed materials and the products.

Table VI - Feed material and product masses, with chemical compositions for the 120 kW test

Test stage No.	Laterite mass, kg	Coal mass, kg	Ni in laterite, %	Fe in laterite, %	Laterite SiO2/MgO ratio	Slag mass, kg	Ni in slag, %	Fe in slag, %	Slag SiO2/MgO ratio	Metal mass, kg	Ni in metal, %
1	1685	142	1.57	17.3	1.53	1456	0.17	14.5	1.45	114	6.3
2	1750	128	1.54	14.9	1.23	1240	0.04	10.3	1.19	126	16.1
2 FBR	919	58	1.36	14.9	1.23	727	0.30	15.3	1.29	50	16.8
3	1300	80	1.57	19.4	1.52	1067	0.15	18.8	1.41	36	16.8
4	900	54	1.49	15.0	1.73	841	0.12	14.9	1.60	67	22.1
5	600	36	1.22	20.3	3.03	340	0.12	14.1	2.18	34	20.9
Dig- out	-	-	-	-	-	161	0.34	12.0	1.67	122	20.3

Slag temperatures varied between 1600 and 1700^oC, depending on the chemical composition. During this test, conducted at power fluxes between 280 and 300 kW/m² of hearth area, a substantial improvement in the match between slag and feed SiO₂/MgO ratios was obtained, as opposed to Example 1, where the power fluxes had been substantially higher. However, there was some erosion of the refractories, especially at high SiO₂/MgO ratios.

Example 3 - 500 kW test : laterite calcine smelting

The aim of this test was to gather further information, in addition to that obtained from the test described in Example 2, on the smelting of laterite calcine in a DC-arc furnace. Further data was required with respect to reductant addition, energy consumption, pre-baked electrode consumption, refractory performance, dust carry-over, and arc length–voltage characteristics.

To meet these objectives, a five-day smelting campaign was conducted on Mintek’s 500 kW DC-arc furnace, during which the following conditions were investigated:

1. Coarse calcine (2 to 10 mm) with a SiO₂/MgO ratio of 1.29, and 16% iron content

2a. Fine calcine (< 2 mm) with a SiO₂/MgO ratio of 1.78, and 23% iron content

2b. Fine calcine (< 2 mm) with a SiO₂/MgO ratio of 1.88, and 26% iron content.

Additional heats were completed, to determine arc length–voltage characteristics, the effect of a higher power level (750 kW) on furnace performance, and the feasibility of feeding through the bore of the graphite electrode.

The screening and blending of the laterite calcine (< 0.05% LOI) was done at Mintek. In order to obtain the desired chemical compositions for the various test conditions, additions of hematite, silica sand, and calcined magnesite were made.

In total 30.8 tons of material was processed, with the overall mass balance given in Table VII.

Table VII - Overall mass balance for 500kW test on laterite calcine smelting

Warm-up metal	Feed material, t		Products, t
	Calcine	Coal	Slag	Metal	Dust
0.50	28.89	1.44	26.81	1.03	0.55

The dust losses to the furnace off-gas system amounted to about 2%, for both the fine and coarse materials.

Averaged operating details for the campaign are summarized in Table VIII.

Table VIII - Averaged operating details for 500kW tests on laterite calcine smelting

	Condition
Process parameter	1	2a	2b	750 kW	Centre feed
Mass fed, t	10.66	7.84	3.93	2.42	1.05
Feed rate, kg/h	327	405	403	792	353
Carbon addition, %	2.0	2.6	2.4	2.0	2.1
Slag temperature, °C	1682	1603	1565	1538	1486
Metal temperature, °C	1465	1520	1477	-	-
Power input, kW	500	470	405	720	370
Measured rate of energy loss, kW	170	154	122	180	170
Voltage, V	170	178	178	250	200
Current, kA	2.92	2.06	2.28	2.89	1.85
Resistance, mW	58	67	78	87	108
Power flux, kW/m^2 hearth area	377	362	305	543	-
Energy consumption, kWh/kg feed	1.4	1.1	1.0	-	-
Thermal efficiency, %	54	61	64	-	-

- Not measured/calculated

The overall electrode consumption was 3.1 kg/MWh or 5 kg/t calcine, and the average voltage drop in the arc, when side feeding was employed, was determined as 7 V/cm.

Table IX provides information on the most important chemical analyses of the composite calcine and smelter products, as well as the nickel recoveries, which have been calculated on the assumption that dust carry-over could be recycled and does not constitute a loss of nickel.

Table IX - Chemical analysis and nickel recoveries from 500 kW test on laterite calcine smelting

Condition	Composite calcine analyses			Slag analyses			Alloy nickel content	Nickel recovery, %
	Ni, mass%	Fe, mass%	SiO2/MgO ratio	Ni, mass%	Fe, mass%	SiO2/MgO ratio	Ni, mass%
1	1.48	16.1	1.29	0.20	16.0	1.19	43.0	86
2a	1.23	23.2	1.78	0.10	20.8	1.55	26.9	95
2b	1.14	25.7	1.88	0.16	23.3	1.76	24.8	89

The nickel content of the composite calcine was diluted under condition 2, owing to the raw-material additions made to attain the desired chemical compositions. The selective reduction ratio was high, 7.4 and 6.1 for conditions 1 and 2 respectively, where this ratio is defined as the nickel recovery to metal divided by the iron recovery to metal.

Example 4 - 500kW test: Rotary kiln calciner dust smelting

The structure of this test campaign was very similar to that described in Example 3, but rotary kiln calciner dust was used as feed material. The dust was dry, with 6.5% LOI. The following conditions were investigated:

1. Dust only, with a SiO₂/MgO ratio of 1.71 and 26% iron content. Under this condition, additional heats were completed to determine arc length-voltage characteristics, and the feasibility of feeding through the centre of the hollow electrode.

2. Dust together with hematite and silica sand additions to increase the SiO₂/MgO ratio to 2.20 and the iron content to 32%.

The overall mass balance for the test is given in Table X.

Table X - Overall mass balance for 500 kW test on rotary kiln calciner dust smelting

Feed material, t			Products, t
Warm-up metal	Dust	Charcoal	Slag	Metal	Dust
0.47	25.80	0.89	21.38	1.30	1.22

About 4.5 per cent of the ultra-fine dust feed material (with a d₅₀ of approximately 35 µm) carried over to the off-gas system.

Table XI presents averaged operating parameters for the test campaign.

Table XI - Averaged operating details for the 500 kW test on rotary kiln calciner dust smelting

	Condition
Process parameter	1a	1b	Centre feed	2a	2b
Mass fed, t	4.30	11.32	1.44	5.11	1.61
Feed rate, kg/h	265	289	307	337	196
Carbon addition, %	2.9	3.4	3.4	3.8	3.8
Slag temperature, °C	1554	1576	1533	1526	1522
Metal temperature, °C	-	1540	-	1475	1500
Power input, kW	373	374	364	372	267
Measured rate of energy loss, kW	130	110	110	105	115
Voltage, V	170	170	180	170	170
Current, kA	2.19	2.20	2.02	2.19	1.57
Resistance, mW	78	77	89	78	108
Power flux, kW/m^2 hearth area	281	281	274	280	200
Energy consumption, kWh/kg feed	1.40	1.29	1.21	1.10	1.37
Thermal efficiency, %	61	66	66	67	55

- Not measured

The results were very similar to those obtained in Example 3. The electrode consumption was 3.4 kg/MWh (corresponding to 5 kg/t dust), and the mean voltage drop in the arc, during around-the-electrode feeding, was 6.0 V/cm. The electrical resistivity of the slag was calculated to be approximately 0.78 Wcm.

Important chemical analyses of the composite dust, and slag and metal products, are presented in Table XII. Also given is the recovery of nickel, once again based on the assumption that the only loss of nickel is to the slag phase.

Table XII - Chemical analyses of dust feed and products, together with the nickel recoveries obtained during the 500 kW test on rotary kiln calciner dust smelting.

Condition	Composite calcine analyses			Slag analyses			Alloy nickel content	Nickel recovery, %
	Ni, mass%	Fe, mass%	SiO2/MgO ratio	Ni, mass%	Fe, mass%	SiO2/MgO	Ni, mass%
1a	1.63	25.7	1.71	0.22	27.8	1.61	32.2	89
1b	1.63	25.7	1.71	0.09	24.8	1.61	25.1	95
2a	1.22	32.0	2.19	0.11	32.9	1.88	24.8	92
2b	1.22	32.0	2.19	0.11	32.9	2.20	24.8	92

The selective reduction ratio (recovery of nickel to metal divided by recovery of iron to metal) was 7.7 when calculated over the whole test campaign, which illustrates the selective reducing capability of the DC-arc furnace while maintaining a high recovery of nickel. The crude ferronickel also contained on average 0.045% Si, 0.01% C, 0.3% S, and 0.055% P, therefore necessitating further removal of sulphur and phosphorus to meet the ISO specifications of 0.03% S maximum and 0.03% P maximum.

DC-ARC FURNACE PROCESS ADVANTAGES

The superior performance of a DC-arc furnace, compared with an AC three-electrode furnace or a six-in-line furnace operated with immersed electrodes, for this type of process, is due to a number of factors. These have been presented previously (1-3). The most important advantage of the process is the fact that a wide range of laterite ores (in terms of particle size and chemical composition) can be processed.

Feeding of fine material

Problems have previously been noted with regard to the feeding of laterite ore fines to a conventional shielded-arc AC furnace, leading to decreased power and production levels (4). On the other hand, the DC-arc furnace has the industrially-proven ability to process fine ores very successfully (1). Even dusts can be utilized directly in the production of ferronickel, due to the highly stable nature of the DC arc.

Wide compositional range for feed

In the DC-arc furnace, the bulk of the electrical resistance is located in the arc, and the smelting process is conducted with an open bath. The furnace operates under open-arc conditions, with the electrode positioned above the bath, so the resistivity of the slag has little influence on the supply of energy to the furnace bath. Therefore, the bath temperature can be controlled to minimize the tendency to slag foaming. Effective energy supply depends less on slag composition, allowing the slag chemistry to be optimized for the best recovery and minimum flux addition (instead of for the required electrical characteristics). Furthermore, the relatively high iron oxide content of some laterites, would (if they were not blended with lower iron-containing laterites) result in high iron-containing slags with high electrical conductivity, which do not permit effective energy generation in the melt when using a slag-resistance furnace.

Effective control

An additional advantage of the open-arc, open-bath smelting mode of operation is the effective control of the reductant addition, as there is no direct contact between the graphite electrode and the melt.

It is possible (and desirable) to maintain a layer of frozen material in contact with the sidewalls, in order to protect the refractory lining of the furnace.

CONCLUSIONS

· DC-arc furnace technology has been successfully applied to the production of ferronickel from laterite ores and dusts.

· Pilot-plant testwork at Mintek at power levels of up to 750 kW has demonstrated that laterites of wide compositional range (with respect to iron content and SiO₂/MgO ratio) can be smelted effectively to produce crude ferronickel at high nickel recoveries.

· The smelting of dusts (100% < 35 mm) recovered from rotary kilns used to calcine nickel laterite ores has also been demonstrated successfully at this scale.

ACKNOWLEDGMENTS

This paper is published by permission of Mintek.

REFERENCES

1. R.T. Jones, N.A. Barcza, and T.R. Curr, “Plasma Developments in Africa”, Second International Plasma Symposium: World progress in plasma applications, EPRI (Electric Power Research Institute) CMP (Center for Materials Production), Palo Alto, California, 9-11 February 1993.

http://www.mintek.co.za/Pyromet/Plasma/Plasma.htm

2. R.T. Jones, D.A. Hayman, and G.M. Denton, “Recovery of cobalt, nickel, and copper from slags, using DC-arc furnace technology”, International Symposium on Challenges of Process Intensification, 35th Annual Conference of Metallurgists, CIM, Montreal, Canada, 24-29 August 1996, 451-466.

http://www.mintek.co.za/Pyromet/Cobalt/Cobalt.htm

3. H. Lagendijk, A.F.S. Schoukens, P. Smith, and P.W.E. Blom, “The production of ferronickel from nickel containing laterite”, South African Patent , ZA 94/6071, 12 August 1994.

4. T. Ma, J. Sarvinis, N. Voermann, B. Wasmund, J. Sanchez, and O. Trifilio, “Recent developments in DC furnace design”, International Symposium on Challenges of Process Intensification, 35th Annual Conference of Metallurgists, CIM, Montreal, Canada, 24-29 August 1996, 169-182.