Connecting countries to climate technology solutions
English Arabic Chinese (Simplified) French Russian Spanish Yoruba

Smelt reduction for iron and steel sector

Technology group:
Smelting reduction processes are the latest development in pig iron production. These processes combine the gasification of coal with the melt reduction of iron ore. Energy intensity of smelting reduction is lower than that of blast furnace, as the for coke production is not needed and the need for ore preparation is reduced.

The iron and steel sector is the second-largest industrial user of energy, consuming 616 Mtoe in 2007 and is also the largest industrial source of CO2 emissions. The five most important producers – China, Japan, the United States, the European Union and Russia – account for over 70% of total world steel production. Smelting reduction is a term assigned to a group of upcoming ironmaking processes which aim at overcoming certain fundamental problems of the existing blast furnace route. These problems include dependence on large scale operation, reliance on coking coal and prepared raw materials as well as environmental pollution.


The iron and steel sector is the second-largest industrial user of energy, consuming 616 Mtoe in 2007 and is also the largest industrial source of CO2 emissions. The five most important producers – China, Japan, the United States, the European Union and Russia – account for over 70% of total world steel production. A main technique used in the iron and steel industry is smelting, where the use of smelting is to produce a metal from its ore. This includes iron extraction (for the production of steel) from iron ore, and copper extraction and other base metals from their ores. Smelting reduction is a term assigned to a group of upcoming ironmaking processes which aim at overcoming certain fundamental problems of the existing blast furnace route. These problems include dependence on large scale operation, reliance on coking coal and p#mce_temp_url#repared raw materials as well as environmental pollution (Basu et al. 1995). They can be considered as the latest development in pig iron production and omit coke production by combining the gasification of coal with the melt reduction of iron ore (Worrell and Nellis 2007).

Feasibility of technology and operational necessities

Smelting reduction usually produces hot metal from ore in two steps. Ores are partly reduced in the first step and then final reduction and melting takes place in the second stage. The entire process is carried out in two separate process reactors (the reduction shaft and the melter gasifier). Raw materials (a mixture of lump ore, pellets and/or sinter) are charged into the reduction shaft and reduced to approximately 93 % of metallized DRI by a counterflow of process gas. Discharge screws convey the direct reduced iron into a melter gasifier where, apart from final reduction and melting, all metallurgical metal and slag reactions take place. The frontrunner of this process is COREX from the VAI group (Gojic and Kozuh 2006).

Similar Processes under development include [A detailed explanation of these processes can be found in COREX, CCF, DIOS, AISI, and HISmelt (Price et al. 2001).

  • DIOS (Direct Iron Smelting Reduction) was developed by the Japan Iron and Steel Federation (JISF) , the Centre of Coal utilisation and a consortium of eight Japanese steelmakers. The DIOS system has three fluidised furnaces .Iron ore is preheated in the first of two fluidized bed reactors in series and pre-reduced to 15-25% in the second reactor using cleaned offgas from the smelter. Dust removed from the smelter, off-gas and fines removed from the gases leaving the fluidized bed reactors are injected back into the smelter. In addition, a small amount of coal fines, of the order of 50 kg per tonne of hot metal production, is injected into the smelter offgas to cool the offgas and provide additional CO and H2 for pre-reduction.
  • In AUSMELT smelt reduction process, lump ore or ore fines are fed continuously into a converter along with lump coal and flux. Fine coal , oxygen and air are injected through a top lnce to allow submerged combustion. The degree of oxidation and reduction is controlled by adjusting fuel to air and coal ratios as well as the proportion of fine coal injected down the lance. All reactions are completed in a single reactor.
  • In HISMELT smelt reduction process Coal is injected through bottom tuyers into a molten bath. Carbon is rapidly dissolved and reacts with oxygen from incoming iron ore to from carbon monoxide and iron. This reaction is endothermic and therefore to keep the process going additional heat has to be supplied .This is achieved by reacting carbon monoxide released from the bath with oxygen from top injection of air. Reacted hot gasses exit the vessel and are used in a fluidised bed to pre- heat and pre-reduce incoming ore.

The Russian ROMELT process does not involve any pre-reduction step. The smelter has a water-cooled roof and sidewalls in contact with slag and conventional refractories in contact with the metal. A mixture of air and oxygen is injected through two rows of tuyeres. Coal and ore are fed by gravity. The system, simple and robust. ROMELT consumes more energy than other smelting processes due to the lack of pre-reduction and extensive water cooling.

PLASMAMELT involves the reactions in a coke-filled shaft furnace with tuyeres spaced symmetrically around the lower part of the furnace. The shaft is completely filled with coke. Plasma generators and equipment for injection of metal oxides mixed with slag forming material and possibly reductants are attached to the tuyeres. In front of each tuyere a cavity is formed inside the coke column where reduction and smelting take place. At regular intervals the produced slag and metal are tapped from the bottom of the shaft furnace. In the case of iron ore smelting the off-gas from the furnace, consisting mainly of carbon monoxide and hydrogen, can be used for pre reduction of the ore.In other applications of the process, such as reclaiming of alloying metals from baghouse dust, the produced gas is utilised as a fuel gas.If the raw material contains metals with high vapour pressures, for example zinc and lead, these metals leave the furnace with the off-gas which is then passed through a condensor where the metals are recovered from the gas.

The current version of smelt reduction technology is most suitable for medium-scale integrated plants, which are mainly found in developing countries (IEA 2008). Nevertheless, the development of such processes demands time and money and these countries lack capital and support infrastructure, hence they are often discouraged by the perceived risks involved in new technologies The period of development, from the experimental stage to that of a reliable industrial capacity, can be ten years or more. The cost of developing a direct reduction process to the point where it can be considered a proven technology could range from 30 to 100 million US dollars (Gojic and Kozuh 2006). The higher process risk associated with the new smelt reduction processes will result in a much slower start-up curve, and higher commissioning costs.

Status of the technology and its future market potential

Direct reduction processes were introduced into industrial service in the late 1950s. A number of groups around the world, particularly in Europe, Japan, Australia, South Africa and the USA, are engaged in research and development on several process concepts. Many such processes have been successfully tested in pilot/demonstration plant scale and a few of them have been commercialized. Currently, only the COREX process (Voest-Alpine, Austria) is commercial and operating in South Africa, South Korea and India, and under construction at Baosteel in China. The specific COREX process uses agglomerated ore, which is pre-reduced by gases coming from a hot bath. The pre-reduced iron is then melted in the bath. The process produces excess gas, which is used for power generation, DRI-production, or as fuel gas. The FINEX technology allows the use of ore fines, but the first commercial plant is now under construction and hence not included in the best practice. Likewise the first commercial plant using the HISmelt process is under construction in Australia (Worrell and Neelis 2007). Still, Smelt-reduction technologies are, for the most part, niche applications and they are intended to provide sources of hot metal in small capacities (less than 1 Million Tons/year) (MIDREX).

A future solution for the smelt reduction technologies can be the hydrogen smelting reduction, which is researched in Austria (Hiebler and Plaul 2004), but the deployment is due to start after 2025 (IEA 2010). Smelt reduction technologies also allow the integration of Carbon Capture and Storage (CCS) into the production of iron, especially when the process is nitrogen free (IPCC 2007).

Based on the IEA (2010), the deployment status can contribute to an increase in shares of smelt reduction technologies from 3% in 2015 to 18% in 2030 and 31% in 2050. More in detail, the current R&D status involves heat exchange improvement in FINEX, a new configuration of HISMELT to lower coal consumption, integration of HISMELT and ISARNA, while current demonstration projects take place for FINEX and HISMELT, a project for producing reduced pellets will be active in 2015 and a demonstration plant with smelter will function by 2020.

How the technology could contribute to socio-economic development and environmental protection

Smelt reduction processes can assist industrial development of the iron and steel sectors in medium sized installations in developing countries. Smelt reduction, which integrates ore agglomeration, coke making and iron production in a single process, offering an energy-efficient alternative at small to medium scales (IPCC 2007). As an example, Price et al. (2001) demonstrate that South Africa developped the COREX process, as it possesses large reserves of suitable iron ores, but only small reserves of coking or metallurgical coals. However, coals suitable for smelt reduction are available in quantity, and the coal is available “on the doorstep” of the iron ore reserves. The smelting reduction process can use iron ores high in alkali content, as found in South Africa (Wintrell, 1992). Hence, the COREX process allowed the economic use of the local iron ore and coal reserves (Price et al. 2001). In general, the hot metal produced by the smelting reduction process is cheaper than the one obtained by the blast furnace route.

Contribution of the technology to economic development (including energy market support)

In general, the energy consumption from smelt reducing technologies is reduced because production of coke is abolished and iron ore preparation is reduced. As an example, due to its specific properties the COREX export gas can replace natural gas for the majority of applications including power generation, DRI production, heating, and generation of synthesis gases for the chemical industry.The best practice values for the COREX plant are based on the commercially operating plant at POSCO’s Pohang site in Korea. The plant coal consumption is around 29.4 GJ/thm (100 kgce/t), 75 kWh/t (9.2 kgce/t) hot metal electricity and 526 Nm3/t hot metal of oxygen. It exports offgases with an energy value of 13.4 GJ/t (457 kgce/t) hot metal (Sorrell and Neelis 2007).

Figure 3: World best practice energy intensity for the smelt reduction (Source: Worrell and Neelis 2007)

Furthermore, on a global scale, a study in 2000 estimated the 2010 technical potential for energy efficiency improvement with existing technologies at 24% (De Beer et al., 2000) and that an additional 5% could be achieved by 2020 using advanced technologies such as smelt reduction and near net shape casting (IPCC 2007).

Contribution of the technology to protection of the environment

The main benefits of the smelt reduction processes concern its use of a wide range of iron ores, elimination of environmentally unfriendly coking plants, very good hot metal quality for all high-quality steel applications, possible utilization of export gasses, outstanding environmental compatibility. The hot metal and slag tapping are comparable to those in conventional blast furnace practice. The COREX slag for instance has a composition similar to the blast furnace slag and can be used in a likely manner (e.g. in the cement industry).


According to the IEA (2008), the smelting reduction techniques can contribute to a Cost of CO2 avoided at a range of 50-100 USD/ tonne by 2020. Today’s smelt reduction generates substantial amounts of surplus off-gas, typically about 9 GJ/t of product. Re-using the off-gases of the smelt-reduction plant could lead to significant additional CO2 reductions.

The current potential of CO2 reduction from smelt reduction technologies is estimated at around 30-100 MtC by the end of 2010 (IPCC 2007). The total potential energy saving from all best available technologies in the iron and steel industry is 133 Mtoe, equivalent to 421 Mt CO2 on the basis of current production levels. These potentials are technical and the economic potentials are significantly below these levels as achieving these savings will require re-build or major refurbishments. In some regions with small-scale production and low-quality indigenous coal and iron ore, the reduction potential will be particularly difficult to achieve. China accounts for 55% of the potential energy saving, although a number of other countries have higher potential in terms of energy reductions per unit of steel produced. The average global potential is 4.1 GJ per tonne of crude steel, equivalent to 0.3 tCO2/ tonne of steel produced (IEA, 2010).

Financial requirements and costs

In general, the smelt reduction technologies can be cost effective, as they provide a high operational flexibility with respect to output, production stops and raw material changes, and also lower hot metal production expenditures of up to 20 % in comparison with a blast furnace of similar capacity. Actual costs of smelt reduction techniques vary substantially among the different techniques employed. It can be said that smelt-reduction processes may result in a theoretical lower cost to produce liquid steel using coal and fine iron ore. However, operating cost is never the only factor by which investors determine their selection of process route (Klauwon, no date). Still, as a reference, most smelt-reduction processes use a capital cost of US$ 300 per annual MT hot metal production, whereas a MIDREX HOTLINK plant producing 720,000 tons/year of 700 deg. C DRI would only cost approximately US$ 170 per annual ton of hot DRI. Larger HOTLINK plants have a capital cost as low as US$ 138 per annual ton of hot DRI. The liquid iron cost per ton is higher than the theoretical hot metal operating cost of the new smelt-reduction processes.


  • Basu, P., Syamaprasad, U., Jouhari, A.K., Say, H.S. 1995. Smelting reduction technologies for direct ironmaking. Mineral processing and extractive metallurgy review 12 (2-4), 223-255.
  • De Beer, J.G., Harnisch, J., Kerssemeeckers, M. 2000. Greenhouse gas emissions from iron and steel production. IEA Greenhouse Gas R&D Programme, Cheltenham, UK.
  • Gojic, M., Kozuh, S. 2006. Development of direct reduction processes and smelting reduction processes for the steel production. Kem. Ind. 55(1), 1-10.
  • Goodman, N. 2007. Operations at the HIsmelt Kwinana Plant. HIsmelt Corporation.
  • Hibler, H., Plaul, J.F. 2004. Hydrogen plasma smelting reduction – an option for steelmaking in the future. METABK 43(3), 155-162.
  • IEA, 2008. Energy Technology Perspectives - Scenarios and Strategies to 2050. International Energy Agency, Paris, France.
  • IEA, 2010. Energy Technology Perspectives - Scenarios and Strategies to 2050. International Energy Agency, Paris, France.
  • IPCC, 2007. Intergovernmental Panel on Climate Change Fourth Assessment Report: Working Group III report ‘Mitigation of Climate Change’. IPCC, Switzerland.
  • Klauwon, R. (no date). Competitive liquid steel from HOTLINKTM and scrap. MIDREX.
  • Price, L., Phylipsen, D., Worrell, E. 2001. Energy use and carbon dioxide emissions in the steel sector in key development countries.Environmental Energy Technologies Division, Lawrence Berkeley National Laboratory. LBLN-46987.
  • Wintrell, R., 1992. “Why Corex ?” Proc. 1992 Ironmaking Conference, New York: American Institute of Mining, Metallurgical, and Petroleum Engineers.
  • Worrell, E., Neelis, M. 2007. World best practice energy intensity values for selected industrial sectors. ECOFYS/ Ernest Orlando Lawrence Berkley National Laboratory, LBNL-62806.