Ferronickel Smelting

The Twelfth International Ferroalloys Congress

Sustainable Future

671

June 6 9, 2010
Helsinki, Finland

Ferronickel Smelting

CONSIDERATION OF RELEVANT PROCESS PARAMETERS

In order to evaluate the technical and economical conditions in nickel pig iron production, several
parameters influencing the process should be taken into consideration, such as availability of raw
material, reducing agent, energy source and other economic considerations like capital and
operational costs and the market size and value etc.
2.1

Process Highlights

Ferronickel production both in blast furnace and electric furnace .have been analyzed taking into
consideration the above aspects. For the electric furnace route for ferronickel production, shielded arc
smelting instead of submerged arc operation has been considered.( Figure 1)
Minrio
Ore

Coke
Coque
Zona zone
Lumpy
Granular

Zona
deCohesive
Amolecimento
Zone
e Fuso

CHARGE
Carga
SLAG
Escria

ALLOY
Liga

Zona
de Coque
Estagnado

Zona
de Coque Ativa
Camada
em Amolecimento
e Fuso
Zona
de Combusto
Cadinho
Crucible

Active coke
Zone
Softening
and melting
layer
Combustion
Zone

Zona de
Dropping
Gotejamento
Zone

Figure 1:

Electric furnace Shielded arc

Figure 2:

Blast furnace scheme

In shielded arc practice, as the electrodes are above the slag level, and hence the carbon from
electrodes does not take part in the reduction reactions, which consume only the reducing agent
(coke, charcoal). In such process some selectivity of the reduction of nickel relative to iron can be
achieved [4, 5]
In the blast furnace practice, carbon from coke/charcoal charged is not only reducing agent but also
energy source. In this case, the amount of carbon charged in the furnace is much greater than that
required only for nickel reduction. Due to this the iron oxide reduction reaction will be extended to its
completeness, resulting in iron rich alloy. Figure 2 shows blast furnace scheme
3
Considering the Chinese approach [3], a charcoal blast furnace with useful volume of 200m with
glendons for heating up the blast is chosen for Ferronickel production .This type of plant is very
common in Brazil for the production of Pig iron. It is a low cost and flexible facility. Nearly 7 million
tons of pig iron are produced in such facilities in the Minas Gerais State of Brazil [6].

2.2

Raw Materials

Nickel containing lateritic ores have extensive occurrence all over the world. Figure 3 gives an
idealized section through a lateritic nickel deposit, and the range of application of the available
technologies. In the figure, the term Pyrometallurgy refers to conventional ferronickel smelting.
Nickel pig iron smelting range is going to be analyzed.
The composition ranges and the respective extraction processes are given in the figure. As shown in
the scheme, two main layers are identified the saprolitic, deeper layer and the limonitic shallower
layer. These layers differ from each other both from mineralogical and from chemical aspects and
hence would necessitate different recovery processes. In between the above layers there are
transition layers.
The technological routes as shown in the figure suggest that richer, magnesia bearing ores are more
suitable for to ferronickel electric smelting process whereas the limonitic ones are more suitable for
hydrometallurgical routes [7] [8].
Smelting of leaner ores for the production of Fe-Ni is possible technically but the energy consumption
should increase considerably. This can be counteracted to some extent by an increase in the iron
reduction (up to the limits of the ore iron content), but the nickel content will fall below that required for
ferronickel normal grades. However in this case, one can resort to nickel pig iron smelting. The figure
covers the whole range

The Twelfth International Ferroalloys Congress

Sustainable Future

672

June 6 9, 2010
Helsinki, Finland

Ferronickel Smelting

ore body profile and the suitability of a particular ore quality in respect of a recovery process.

Figure 3:

Section through a lateritic deposit [2][7]

As such, it can be said that nickel pig iron production is more free in terms of ore quality in
comparison with that of conventional ferronickel. If the ore is suitable (qualitatively and quantitatively),
ferronickel smelting in electric furnace or hydrometallurgical processing would be preferred routes. On
the other hand, if the ore quality is not adequate, nickel pig iron production in blast furnace could be
an alternative. Naturally this would depend on demand for low nickel alloys which can be produced
with nickel pig iron as starting material.
2.3

End Uses Stainless Steel Products

Table 1 shows the chemical compositions of selected grades of stainless steels.

For comparison purposes, it was included a typical composition of one 300 Series steel.
Table 1:
Grades
(AISI)
201
203
204
205
304

Chemical characteristics of selected stainless steels.[10]

C (max)
0.15
0.15
0.15
0.12-0.25
0.08

Mn
5.50-6.50
7.5-10.0
6.5-9.0
14 15.5
2.0 max

Cr
16-18
17-19
15.5-17.5
16.5-18.0
18.0 -20.0

Ni
3.5-5.5
4.0-6.0
1.5-3.5
1.0-1.75
8.0-10.5

N
0.25 max
0.25 max
0.05-0.25
0.32-0.40
-

Cu
2.0-4.0
-

For the production of 300 series stainless steels, which contain 8% Nickel and above would require
high grade Ferronickel normally produced in the electric furnace.
The 200 Series is being produced since the 1930s and a lot of development work was done during
1940s and 1950s in USA, due to shortage of nickel at the time, This grade has became popular
because of its lower price and suitability to some specific applications with even better results
compared to 300 series steels.
The construction sector is the major consumer of 200 Series. 200 Series is already preferred in
various applications [9] where it can be used. It constitutes over 60% of all stainless steels for this
sector. In the last few years, 200 series market share is about 10% of the stainless steel market.
Nickel content in the 200 Series steels is between 0 - 6%. As can be seen, these nickel contents are
compatible with the nickel content range of nickel pig iron which could be the starting material for that
purpose.
2.4

Reference prices

In order to check the feasibility of the blast furnace process, it is advisable to consider as a reference
basis a stable scenario. Otherwise the evaluation could lead to false conclusions based on a
particular condition. Hence it is assumed, that the conventional Ferronickel production would be in

The Twelfth International Ferroalloys Congress

Sustainable Future

673

June 6 9, 2010
Helsinki, Finland

Ferronickel Smelting

equilibrium with the demand, on a worldwide basis. Under these conditions, the price of nickel
(contained in the alloy) is stable or within the normal market fluctuations. Figure 4 presents LME
(London Metals Exchange) nickel prices variation in the last twelve months. From the graph, one can
easily recognize the crisis period and the future recovery trend. As can be seen, there is lot of
fluctuation and neither the lower region nor the peak of figure 4 are good reference points but there is
a trend to stability in the near future . It is not easy to forecast precisely the future, but as a first
approach, a hypothetical scenario has been adopted in this work, considering the mean value of the
graph.
US$/ton

Figure 4:

Nickel prices (cash buyer) 12 months Source LME - September, 24th, 2009 [11]

PROCESS SIMULATION

In order to identify the conditions in which nickel pig iron produced in the blast furnace can compete
with conventional ferronickel production in electric furnace , both processes were simulated,
employing programs based on thermochemical models where mass and energy balances were
made considering the typical operating conditions of each one.
The reference nickel ores are considered as coming from an ore body whose profile and
characteristics are shown in figure 3. Naturally this ore body is not real, but its structure is typical of
many of them, around the world. From this ore body, the ores for the two smelting simulation routes
are chosen.
The first one to be simulated is the conventional ferronickel production in electric furnaces.
3.1

Conventional Ferronickel Production (Electric Furnace)

The process is based on RKEF technology (Rotary Kiln Electric Furnace). In this simulation, ore
composition is chosen for the required range of Nickel content. From figure 3 it is seen that the most
suitable ore comes from the garnierite/saprolite/serpentine layer. The mean composition of this layer
is given in table 2. It is also feasible to employ lower grade ores from the intermediate layer, and that
the lower assumed limit of the composition is given in table 2.
Table 2:

Selected layers ore composition deeper layers

Main layer

Ni

Co

Fe

Cu

SiO2

MgO

Cr2O3

2.500

0.028

13.710

0.010

29.770

20.600

0.70

Limit layer

1.800

0.028

20.000

0.010

35.000

18.000

0.70

These compositions are referred to the ore after calcining.

The ferronickel smelting procedure is well known [2][4][12] and will not be described here.
The main features considered are as follows:
- Shielded arc procedure

The Twelfth International Ferroalloys Congress

Sustainable Future

674

June 6 9, 2010
Helsinki, Finland

Ferronickel Smelting

The amount of carbon charged corresponds to what is the stechiometrically necessary to

generate CO for NiO reduction plus an additional amount to guarantee the aimed nickel
recovery value.
- The SiO2/MgO ratio is maintained within the 1.4 and 1.9 limits and the FeO content depends
on the alloy requirement and energy consumption (This depends on the f charged carbon).
Many other factors have to be considered but the ones mentioned above are sufficient to characterize
the process. The aimed nickel content in the alloy is 30%, with a lower limit of 20%. Reducing agent is
a conventional metallurgical coke (C 85%, Ash 8%).
Process simulation was carried out employing a program based on a thermochemical model based on
mass and energy balances, considering the typical conditions of a conventional ferronickel smelting
operation and facility. In the simulation dry calcined ore is charged for the electric furnace [12].
Based on the alloy and slag aimed characteristics, the program calculates the mass inputs and
outputs. In this program, both the operational requirements (calcination temperature, SiO2/MgO ratio,
alloy composition etc) and the thermodynamic ones (slag/alloy partitions, equilibrium conditions,
interaction coefficients, etc), are processed and adjustments are made till the aimed results are
achieved. These results are presented in table 3
Table 3:

Main results - Ferronickel Smelting Main ore layer

Main Parameters

Alloy

Kg/t ore

Kg/t alloy

Inputs
Calcined ore

Slag

Co

0.26%

CoO

0.02%

Si

0.18%

SiO2

47.68%

1000

11686

Fe

69.64%

Cr2O3

1.09%

Coke

34

398

Cu

0.14%

NiO

0.51%

Electrode

70

Cr

0.17%

MgO

33.02%

Ni

29.54%

FeO

17.34%

Outputs
Alloy

86

1000

0.06%

Cu2O

0.00%

Slag

701

8194

Selectivity

1.83%

SiO2/MgO

1.44%

This slag has a high melting point, what is a normal feature in ferronickel smelting. This can be seen
in figure 5. In this figure results of this work and those of main ferronickel facilities worldwide are
plotted together. As can be seen, there is a very good fitting with actual results in similar plants, which
confirms the validity of the procedure.
The energy balance resulted in an electrical energy consumption of 11294 kWh/t alloy. Using the
same program and considering that the ore charge is from the limit layer, the results are presented in
table 4.
Table 4:

Main results - Ferronickel Smelting Limit ore layer

Main Parameters

Alloy

Kg/t ore

Kg/t alloy

Inputs
Calcined ore

Slag

Co

0.24%

CoO

0.01%

Si

0.20%

SiO2

48.86%

1000

10982

Fe

79.21%

Cr2O3

0.95%

Coke

43

476

Cu

0.14%

NiO

0.32%

Electrode

66

Cr

0.16%

MgO

25.15%

Outputs

22

Ni

19.99%

FeO

24.41%

Alloy

91

1000

0.06%

Cu2O

0.00%

Slag

804

8834

Selectivity

SiO2/MgO

1.94%

The Twelfth International Ferroalloys Congress

Sustainable Future

675

220.58%

June 6 9, 2010
Helsinki, Finland

Ferronickel Smelting

This slag melting point is lower than in previous case because of the higher FeO content. This can be
seen in figure 5.
The energy balance resulted in an electrical energy consumption of 13888 kWh/t alloy.
From the diagram, the melting point of the calculated slag (main layer) is around 1670C (which
means an operational temperature of 1700C) and this is 1470C for limit layer (which means an
operational temperature of 1500C).

Figure 5:
3.2

Main layer

Limit layer

Phase diagram ferronickel slags [2]

Nickel Pig Iron Production In Blast Furnace

Charcoal blast furnaces of medium and small and medium sizes have been operating in Brazil since
many decades (up to 200m3 internal volume). In these the burden consists of iron ore or
agglomerate, coke or charcoal and flux. The blast is heated in glendons up to 700C [7].
In this case, the energy necessary to the process is supplied by carbon, which is also consumed in
reduction reactions leading to almost complete reduction of iron and as such there is very little or no
FeO in the slag. Two main restrictions can be recognized from this. The first one is the impossibility
of improving the Ni/Fe ratio in alloy, in comparison to the ore, what explains the typically low nickel
content in the alloy produced in blast furnace. The second is the low FeO content in slag , having
mostly silica and magnesia, causing high slag melting temperatures (figure 5), that are not
achievable in a small blast furnace with low blast temperatures, without oxygen enrichment of the
blast . However fluxes like limestone can be added to lower the slag temperatures As such the slag
practice adopted in the electric furnace smelting, does not apply to the blast furnace practice
Considering the discussions on ore body from figure 3, it is clear that ores from any level could be
employed in blast furnace smelting. However as has been pointed out richer ores are normally
processed in electric furnace or by hydrometallurgical route and the ores poor in nickel content can be
processed in blast furnace.
There are many aspects to be considered in the evaluation of ore burden, mainly hematitic in the
conventional blast furnace process of pig iron production. From that point of view the present ore
material containing nickel really cannot be considered as ore It is rather a waste that is removed
to allow the exploitation of the higher grade layers. The chemical analysis of this layer is given below
(table 5). Adopting the same criteria considered earlier in the ferronickel simulation, the limit for the
ore to be used for nickel pig iron production would be somewhere in the intermediate layer, reported
in figure 3. The limit layer considered has ore characteristics given in table 5.

The Twelfth International Ferroalloys Congress

Sustainable Future

676

June 6 9, 2010
Helsinki, Finland

Ferronickel Smelting

Table 5:

Selected layers ore composition shallower layers

Overburden

Ni

Co

Fe

Cu

SiO2

MgO

Cr2O3

0.8

0.064

35.0

0.010

44.0

3.0

0.70

Limit layer

1.6

0.064

35.0

0.010

38.0

10.0

0.70

The simulation was carried out employing a program developed for pig iron production in small blast
furnaces with a thermochemical model involving heat and mass balances.
As the small blast furnaces for pig iron production mostly operate with charcoal as fuel in Brazil the
present simulation is also done with the same reducing agent, whose characteristics are the same as
employed in Brazilian small blast furnaces (C 70%; Ash 3%; VM 21%).
As already mentioned, almost all iron and nickel are reduced to the metallic phase. In order to achieve
a low melting point and free running slag, limestone should be introduced in the burden so that the
CaO+MgO/SiO2 ratio is about 0.8. Inputs and outputs calculated by the program, as well as the
resulting slag and alloy, are presented in table 6
Table 6:

Main results - Nickel Pig Iron Overburden layer

Main Parameters
Kg/t alloy
Inputs
Ore
2760
Charcoal
1016
Limestone
2086
Outputs
Alloy
1000
Slag
2450

Co
Si
Fe
Cu
Cr
Ni
P
C

Alloy
0.13%
0.28%
93.71%
0.00%
0.19%
2.19%
0.10%
3.50%

Slag
CoO
SiO2
Cr2O3
NiO
MgO
FeO
Cu2O
CaO/SiO2

0.00%
49.31%
1.03%
0.01%
8.68%
1.30%
0.00%
0.8%

The slag, phase is of CaO-MgO-SiO2 system, in the region of pseudowolastonite, whose

characteristics are compatible with the operation of small blast furnaces with relatively low blast
temperatures. Figure 6 shows the slag composition of the system.
The simulation with the limit layer ore was carried out employing the same program and the operating
conditions, reducing agent etc, as in the previous case. The main results, from this simulation are
presented in table 8
Table 7:

Main results - Nickel Pig Iron Limit layer

Main Parameters
Kg/t alloy
Inputs
Ore
2760
Charcoal
1016
Limestone
2086
Outputs
Alloy
1000
Slag
2450

Co
Si
Fe
Cu
Cr
Ni
P
C

Alloy
0.18%
0.38%
90.17%
0.00%
0.20%
6.01%
0.15%
3.00%

CaO
SiO2
Cr2O3
NiO
MgO
FeO
Cu2O
CaO/SiO2

Slag
33.21%
47.44%
1.15%
0.03%
16.99%
0.93%
0.00%
0.8%

The slag referred to the ore in the limit layer, also of the CaO-MgO-SiO2 system, is in the region of
akermanite, whose characteristics are compatible with the operation of small blast furnaces with
relatively low blast temperatures. Figure 6 shows the position of the slags in that system.

The Twelfth International Ferroalloys Congress

Sustainable Future

677

June 6 9, 2010
Helsinki, Finland

Ferronickel Smelting

Figure 6:

Overburden layer

Limit layer

Phase diagram nickel pig iron slags [15]

EVALUATION OF SIMULATION RESULTS

The parameters values from the simulations enable the evaluation of feasibility of Nickel Pig Iron
production small blast furnaces and the product competitivety in relation to the conventional
ferronickel production
On facilities subject, a conventional ferronickel smelter includes, besides electric furnace, ore drier
and a calcining rotary kiln. For alloy refining a ladle furnace facility (de-S) and an oxygen converter
type vessel would be required (de-P and de-C).
Charcoal based small blast furnace facility is very simple and does not require drying and calcining
facilities) Charcoal has little sulphur and hence ladle furnace is not necessary.
In the particular case of Brazil, there are many such small blast furnaces for the production of pig iron
and these can be adapted to nickel pig iron smelting.
The main items of operating costs are given in table 9 and 10 corresponding, respectively to the
electric furnace route (conventional ferronickel production) and blast furnace route (nickel pig iron)
Average price from figure 4 has been taken as a reasonable and acceptable reference point. The
value arrived at for the price of nickel on this basis is around US$ 16000.00/ton. Another price that
must be considered for market stability is that of the ore. Even though this may be smelters property,
its value must be taken into consideration in the calculations. Ore price is normally fixed relative to the
nickel content, considering a non speculative scenario, parameters are the LME price, the ore grade
and the payable premium, and this price can be calculated according to following relation [12]:
Ore Price = LME Nickel Price x Nickel Grade x (Nickel Grade x Payable Premium)
Payable premium depends on market conditions, nickel grade, mine characteristics etc.
Table 8:

Main cost items Rotary Kiln / Electric Furnace Route

Ore (t)
Energy(kWh)
Electrode(t)
Coke(t)
Charcoal(t)
TOTAL

Main layer: Niore-2.5%; Nialloy-30.0%

u/t alloy
US$/u
US$/t alloy
11.686
100.00
1168.60
11294
0.083
937.40
0.070
500
35.00
398
250.00
99.50
1260
266.70
336.00
2576.50

The Twelfth International Ferroalloys Congress

Sustainable Future

678

Limit layer: Niore-1.8%; Nialloy-20.0%

u/t alloy
US$/u
US$/t alloy
10.982
51.84
569.30
13888
0.083
1152.70
0.066
500.00
33.00
476
250.00
119.00
1260
266.70
336.00
2210.01

June 6 9, 2010
Helsinki, Finland

Ferronickel Smelting

Table 9:

Main cost items Blast Furnace Route

Ore (t)
Charcoal (t)
Limestone(t)
TOTAL

Ovb layer: Niore-0,8%; Nialloy-2,2%

u/t alloy
US$/u
US$/t alloy
2.760
10.24
28.26
1.016
266.70
270.93
2.086
150.00
312.10
612.10

Limit layer: Niore-1,6%; Nialloy-6,0%

u/t alloy
US$/u
US$/t alloy
3.790
40.96
155.24
1.641
266.70
437.60
2.052
150.00
307.80
900.64

Based on these figures and considering the Ni contents both in ores and alloys, the cost of the Ni
units can be calculated and put together, for comparison purposes in figure 7.

US$/ Ni unity (cost&price)

Cost and price - Ni unity

350
300
250

FeNi cost

200

FeNi price

150

Ni Pig cost

100

Ni Pig price

50
0
0

0,5

1,5

2,5

Ore grade - %Ni

Figure 7:

Variation of the cost of Ni in alloy with Ni ore content

The cost lines were calculated dividing the total cost values in each condition by the Ni content in
alloy (FeNi and Ni Pig)
The commercial values were calculated considering both the Ni and the Fe content, according to the
following formulas:
FeNi
(Alloy price/t ) = (Ni price) x %Ni + (Pig Iron price)x (100 - %Ni)/%Ni [14]
Where: %Ni is the alloy Ni content.
As can be seen, this price considers the iron values, that werent valuated in the past.
Ni Pig Iron
(Alloy price/t) = Ni price x %Ni + (Pig Iron price) x %Fe
Where: %Ni and %Fe are the Ni and Fe contents in alloy. In this case, the iron value calculations
consider that this product is in fact a pig iron.
The LME price considered for nickel is US$16000.00/t and that for iron is the pig iron price, FOB
Brazilian port, in August/2009 that was considered as US$ 350.00/t.
The Nickel values in the alloys were considered as 90% LME, for FeNi and 85%LME for Ni Pig Iron.
According to the graph, the range of each product seems clear and suggests a complementary
condition between the two kinds of product.
As can be seen, the differences between cost and value are almost constant.

The Twelfth International Ferroalloys Congress

Sustainable Future

679

June 6 9, 2010
Helsinki, Finland

Ferronickel Smelting

The Twelfth International Ferroalloys Congress

Sustainable Future

680

June 6 9, 2010
Helsinki, Finland