Manufacture of Silico-manganese

Just as in the case of chromium, silicon reduces the solubility of carbon in manganese alloys; Table 6 gives the carbon content in equilibrium with various silicon contents of silicon-manganese alloys. These figures will vary somewhat with the amount of time the alloy is allowed to stand in the ladle, because some carbides (probably SiC) will float out of the melt and form a slag or scum on the surface.

It will be evident, therefore, that if silicon-manganese is reacted with a manganese ore or a high manganese oxide slag, the silicon will be removed and a “refined” or medium to low carbon ferro-manganese will be produced. As the desiliconisation is done with manganese oxide, the manganes so reduced is carbon-free, so that the carbon content of the refined ferro-manganese is considerably lower than that of the silico-manganese from which it is made.

Silico-manganese is, therefore, produced as an intermediate product for further processing, but quite a large production of the output is sold directly to the steel industry as a deoxidant.

Raw material stock yard at FAP, Joda

There are a number of possible ways of making these alloys, from just mixing high-carbon ferro-manganese with, say, a 90 per cent Si alloy, to smelting the high-carbon alloy with quartz. Classically, high-silica ores of manganese, which are generally available at a discount per unit of Mn, are smelted with coke plus adjustments of quartz and/or manganese ore to get the correct metallurgical balance. With the tendency, however, towards using the fluxless process of making the high-carbon alloy, the integration of this latter process with silico-manganese production has been increasingly prevalent in recent times. This means that the high-manganese slags from ferro-manganese manufacture, which run from 26 to 42% Mn, form an important part of the burden for making silico-manganese. These slags, being very low in iron(<1%Fe), can be blended with really low-grade manganese ores with, say, 32% Mn, 25% SiO₂ and 15 Fe.

Ferro-manganese being collected in ladle

In South African practice, rich slag from the primary smelting operation contains 50.2% MnO, 0.59% FeO and 29.4% SiO₂. To make a 15 to 20 per cent silicon alloy, it is mixed with a siliceous manganese ore of the analysis (Mn 38.3%, Fe 8.42%, SiO₂ 21.8%); and the burden per tonne of alloy consisted of: Manganese ore: 1.107 tonnes, Enriched slag: 1.094 tonnes, Coke: 0.86 tonnes, Quartz: 0.14 tonnes, Energy consumption: 5770 kWh/tonne.

The above charge made an alloy containing Mn: 69.6%, Si: 16.8%, C: 1.68%.

For making a higher-silicon alloy, namely the 20 to 25% Si grade, the ore is changed to a more siliceous variety containing 33.5% Mn, 26.4% SiO₂, and 9.4% Fe. The charge contains the following proportions per tonne of the final alloy: Manganese ore: 1.196 tonnes, Enriched slag: 1.103 tonnes, Coke: 0.97 tonne, Quartz: 0.088 tonne, Electrical energy: 6800 kWh/tonne.

The alloy producted analysed Mn: 67.05%, Si: 21.6%, Co: 0.65%

The proportions of ore, slag and quartz can be varied quite widely and will depend upon how much slag is available for the operation. For instance, the Russians use the following charge proportion to produce a 20 per cent Si alloy:

Rich manganese slag 40% Mn, 30% SiO₂: 300 parts
Manganese ore (48% Mn) : 100 parts
Quartz : 80 parts
Coke breeze : 90 parts
Steel scrap : 20 parts

It can readily be calculated that the carbon efficiency with such a charge is very high, due to the fact that only one atom of oxygen has to be removed from Mn++ in the slag. The Russians point out that a high slag burden is readily smelted under cool top conditions, and closed-roof furnaces can be used with gas recovery.

Dephosphorisation of Liquid High Carbon Ferro-manganese

High carbon ferro-manganese produced at FAP, Joda contains more than 0.25% P and for the production of special quality steels (less than 0.015% ‘P’); it is essential to restrict the ‘P’ content of the Fe-Mn used to less than 0.15%. Accordingly, detailed investigations were conducted at Tata Steel to dephosphorise the ferro-manganese using a calcium silicide based reagent. It was possible to achieve more than 50% degree of dephosphorisation with less than 0.15% ‘P’ in the alloy after treatment at 1400^oC (which was the optimum treatment temperature), Figure 1. Injection of cored calcium silicide wire proved to be the most effective reagent particularly for high carbon ferro-manganese, Figure 2. Detailed laboratory investigations were completed for optimising the parameters such as composition of the reagent, amount to be added, effect of variation in silicon contents, treatment temperature and treatment time. It is to be mentioned that for 0.10% decrease in [P], the premium obtained is around Rs. 500 per tonne of ferro-manganese.

Conclusions

Operational features of Fe-Mn production with regard to theoretical considerations, furnace design,operating practices, raw materials characteristics, pre-treatment of raw materials, mass balance, slag composition, composition of the alloy produced, power consumption, electrode consumption, etc., have been described in detail for assessing the level at which FAP, Joda is presently operating or should be operating.

A brief account has been given for the production of silico-manganese.

Based on the operating experiences and intensive literature search, strategies to be adopted for reducing the production cost at FAP, Joda are summarised below:

• Under the existing operating conditions, it should be possible to achieve a specific power consumption of 2600 kWh during consistent/stable operation. Use of sinters would be advantageous from the point of view of power consumption. No significant decrease in power consumption is envisaged by using manganese ore briquettes. However, briquetting would certainly help in utilising the finer fractions of the ore.

• For further decreasing the specific power consumption, adoption of fluxless (self fluxing) operation is recommended.

• Though it is possible to run the electric furnace with an MnO content of the slag as low as 15 to 16 per cent, it would mean high slag basicity, high power consumption and high fume losses.

• MnO rich slag obtained from fluxless operation can be used in one of the existing furnaces at FAP, Joda for producing silico-manganese. Alternatively, silico-manganese can be refined to produce low and medium carbon ferro-manganese.

• Use of pre-heated (up to 950^oC) charge in a specially designed reactors (such as rotary kiln/ vertical shaft furnace) can bring down the specific power cosumption to less than 3000 kWh. These reactors can also be converted to reduction reactors for supplying pre-reduced manganese ore into electric smelting unit.

• Use of hot pre-reduced ore from the above reduction reactors (rotary kiln/vertical shaft furnace) integrated to smelting unit can bring about a drastic reduction in specific power consumption (around 2000 kWh).

• For producing low phos-phorus (less than 0.20%) ferro-manganese, CaO/BaO based flux is recommended for the dephosphorisation of liquid high carbon Fe-Mn. But the melt will have to be desiliconised to less than 0.20% Si before dephosphorisation. Injection of Ca - Si cored wire in the melt can reduce the ‘P’ in liquid ferro - manganese to less than 0.15%. Laboratory investigations have been completed for optimising the parameters such as composition of the reagent, amount to be added, effect of variation in silicon contents, treatment temperature and treatment time. It is to be mentioned that for 0.10% decrease in [P] the premium obtained is around Rs. 500 per tonne of ferro- managanese.

• The use of partially burnt coal (Jhama) as a partial substitute to coke needs to be looked into and tried in actual practice as it is expected to increase the charge resistance.

Asish Kumar Roy was born in 1965 in Kolkata. He had his schooling in Kokata in New Alipore Multipurpose School and completed his Graduation in Metallurgical Engineering from Jadavpur University in 1986. He joined Ispat Alloys Limited (Ferro alloys unit at Balasore, Orissa) in 1986. He has experience of 18 years in Ferro alloys production and technology of various products like silicon metal, ferro silicon, silco manganese, ferro manganese, calcium silicide, ferro chrome, etc. He had also got the opportunity to work with Birla Jute and Carbides (Kolkata), ICCL (Chowdar), IMPEX Group (West Bengal) in different positions. He was with Tata Steel from December- 2002 as Head of Ferro Alloys Plant at Joda, but has since left the company.

HOME