Skip to main content

Hogan Steel Archive: Blast Furnaces

The “Hogan Steel Archive,” representing a three-year collaborative effort of the Walsh Library’s Department of Archives and Special Collections and Fordham’s Industrial Economics Research Institute, commemorates and preserves the remarkable steel legacy

Blast Furnaces

The document files relating to blast-furnace technology contain the following information: flow chart and description of the blast-furnace process, operating and design data for U.S. blast furnaces, U.S. sinter-plant operations, lists of blast furnaces, comparative furnace practice, mathematical blast-furnace and hot-blast-stove models, measures to increase iron output and conserve energy, fuel injection and coke rates, fuel rates, pulverized-coal injection (PCI), natural-gas injection, oxygen enrichment, oxy-coal injection, use of oxygen and steam, reformed-gas injection, oil displacement, high-blast temperatures, internal and external combustion chambers, wind rates, blast-furnace top design, hot-metal superheating, charcoal furnaces, mini furnaces, cogeneration with furnace gas, furnace construction and rebuilding, raw materials, tailored furnace burdens, iron-ore selection, dolomite-fluxed pellets, coal washing and furnace productivity, coal savings, slag granulation, raw-materials inventory by airplane, history of ironmaking in the Ohio Valley and Pennsylvania, ironmaking in South America, blast furnace vs. direct reduction and other alternatives, and economic survival of the blast furnace.

Among the Archive’s books and references specifically treating the blast furnace are Productivity in the Blast-Furnace and Open-Hearth Segment’s of the Steel Industry by Father Hogan ; Blast Furnace Technology: Science and Practice by Julian Szekely ; Steel Plant Design, Blast Furnaces by the Carnegie-Illinois Steel Corporation; and Directory of Blast Furnaces of the World, prepared in 1975 by Koppers Company, Incorporated.

Analysis: More than 500 years have elapsed since the primitive forerunners of today’s modern blast furnaces were first used to smelt iron ore in obtaining liquid iron. Today, essentially the same metallurgical process remains the world steel industry’s workhorse, its remarkable longevity insured by constantly evolving process and practice improvements that continue to make the blast furnace an ever more productive and fuel efficient method for satisfying the industry’s growing hot-metal needs.

The blast furnace, essentially a giant vertical shaft, made of steel and lined with refractory brick, is filled or charged from the top with three principal materials, iron ore, coke, and limestone. Representing the furnace burden, the materials are met as they descend by a rising volume of preheated gas from combusting the coke with air, which is blown in under pressure through a series of openings, called tuyeres, toward the base of the shaft. Once lighted, most of today’s blast furnaces operate continuously for a decade or more, before their removal from service to replace their refractory linings.

Coke provides the fuel for smelting the ore, the carbon-rich gas to reduce the ore’s oxide content, and the supporting structure that keeps the furnace burden permeable and free moving. The limestone, a fluxing agent, combines in a liquid state with other unwanted elements in the ore, including silica, alumina, and manganese, as well as with sulfur and ash from the coke, allowing most of these to be removed as a molten mass lighter than iron, called slag. At intervals in the smelting process, the slag is drawn off through the furnace’s cinder notch, an opening just below the tuyeres and a short distance above the tap hole at the bottom of the furnace, where relatively pure molten iron containing some 4% carbon is “tapped” from the furnace.

During the twentieth century, when the world’s annual output of blast-furnace iron increased from some 40 million to 577 million tons, the staying power of the blast furnace stood in marked contrast to the recurring shifts in steel melting and refining technology, which saw steel producers establish and maintain greater flexibility in using increased quantities of ferrous scrap and other ironbearing metallics. In 1900, mostly Bessemer steel was refined almost exclusively from blast-furnace iron, resulting in a hot metal to crude steel ratio of approximately 1100 kg/t on a worldwide basis. During the next 40 years, as the Bessemer gave way primarily to the scrap friendly open-hearth process, the ratio declined to some 700 kg/t, and it then remained at or about that same level through the end of the century.

Notably, during six decades, regardless of fluctuations in world steel activity, and despite additional, major shifts in steelmaking technology from the open-hearth to the basic-oxygen (BOF) and electric-arc furnaces, the world steel industry’s use of blast-furnace hot metal maintained a largely stable relationship to its crude steel output. Within individual countries, depending on the process mix and practices employed to produce steel, the relative use of hot metal frequently varied from the 700 kg/t worldwide norm, with the steel industries of Japan and Brazil among those usually consuming more, and those in the United States , Korea , and the European Union consuming less.

The ability of the blast furnace to remain the world’s preeminent iron producer for so many years can be attributed to steady gains in output and efficiency, which are exemplified by long-term improvements in the worldwide coke rate. Back in 1900, the world’s blast furnaces consumed, on average, somewhat more than one ton of coke to smelt one ton of iron, high by today’s standards, but down from about 1.5 tons twenty years earlier. By 2000, less than half a ton was needed, lowering the worldwide coke rate to 459 kg/t, with larger, more modern furnaces using above average amounts of supplemental fuel achieving coke rates of 300 kg/t and lower. Worldwide, blast-furnace coke consumption was the same in 2000 as it was 25 years earlier (265 million tons), but blast-furnace iron production was 85 million tons higher. Time and again, the evolution of blast-furnace technology has yielded benefits usually exceeding expectations, and particularly since the 1960’s, has disproved recurring predictions that a substantial share of the world’s iron would soon be made by direct reduction or some alternative smelting process.

This evolution of the blast-furnace process has encompassed: 1) improvements in the furnace, itself, such as substantial increases in furnace size, improved materials handling and burden distribution, higher hot-blast temperatures with humidity control, higher top-gas pressures, higher wind or blowing rates with oxygen enrichment of the air blast, advances in refractories permitting longer furnace campaigns, and computerized operation and process monitoring; 2) improvements in furnace-charge metallics, including the use of beneficiated ores, self-fluxing agglomerates, scrap, and even direct-reduced iron (DRI); 3) the use of metallurgical coke with superior chemical and physical characteristics; and 4) the injection of supplemental fuels, including fuel oil, natural gas, coke-oven gas, tar, pitch, and pulverized coal (PCI).

Despite the success of improvements such as these in reinventing the blast furnace, the world steel industry’s relative use of blast-furnace hot metal faces an uncertain long-term future, due mainly to environmental constraints that will increasingly affect the availability and cost of coke, without which the blast furnace cannot operate. There is still room to push the world coke rate significantly lower, although doing so will become increasingly costly, and because coke, in addition to serving as a fuel and reductant, also supports the blast-furnace burden, there are limits to cutting coke rates, which are already being tested by a growing number of individual blast furnaces.


Blast-Furnace Alternatives: Among the subjects treated in the document files on blast-furnace alternatives are the following: a survey of alternative smelting processes, their investment and manufacturing costs, forecasts of their potential use, COREX process (its use with cogeneration, other off-gas applications, its link to DRI production, its comparison to the blast furnace, C-3000 units, at Jindal Steel, at HANBO, and at POSCO), CPICOR process, cupola smelting, DIOS process, direct ironmaking processes, electric-furnace smelting, ELRED process, FASTMELT process, FINEX process at POSCO, HIsmelt process, Inred process (status report, investment and operating costs), Iron Dynamics process, Jordan process, KR process, its testing at United States Steel, LB-furnace smelting, OxiCup process, Proler reduction processes, REDOX process, REDSMELT process, reduction using magnetic fields, ROMELT process, SKF-Plasmasmelt process, and smelting reduction processes in Sweden.

Analysis: Recognition of the potential economic and environmental advantages to be gained by circumventing the blast furnace’s dependency on coke has given rise to considerable activity aimed at developing coke-free smelting alternatives for producing molten iron of comparable quality. Research and development work of note has been underway since the 1970’s, with the most significant projects to date having been pursued in Germany, Austria, Korea, Australia, Japan, the United States, Russia, and India, all seeking ways to replace the blast furnace.

The COREX process, thus far the only coke-free alternative to have achieved significant commercial status, was developed in Germany by Korf Engineering as the KR process, standing for “kohle reduction” or coal reduction. Korf completed a pilot plant at Kehl-Rhein in 1981, and molten iron was produced from a variety of lump iron ores and pellets using different coals, from high-volatile bituminous to anthracite. The KR process thereafter was acquired by Austria ’s Voest-Alpine, was renamed COREX, and was scaled up to commercial proportions.

In 1985, South Africa ’s Iscor Limited decided to install the first commercial plant to supplement the blast-furnace output at its Pretoria steel plant. Started up in 1988, the COREX unit had a 300-thousand-annual-ton iron capacity. Commercial COREX plants (C-2000 units producing 700 thousand or more tons) have since been started up at the POSCO plant in Pohang, Korea in 1995; at Jindal Vijayanagar Steel Limited in India, where the first integrated steel plant based exclusively on hot metal from two C-2000 units went into operation in 1997-98; and at Saldanha Steel Limited in South Africa in 1998-99.

About 45% of the energy input in the COREX process is used to make iron, with the balance yielding a considerable volume of export gas that must be used apart from the process, itself, either as fuel gas for the steel plant, in a power plant to generate electricity, or to produce DRI. For this reason, most ongoing research and development projects on coal-based smelting seek to develop “direct” technologies with efficient combustion and heat transfer that advances the process, itself. Also departing from COREX, a number of projects also seek to use iron-ore fines instead of more costly lump ore, pellets, or sinter.

POSCO started to address in-process heat transfer and fine-ore compatibility in 1990, when it formed the “New Project 1 R&D Team,” spearheaded by its affiliate the Research Institute of Industrial Science and Technology (RIST). In 1992, the team and Voest-Alpine started to investigate a modification of COREX known as the FINEX process. The development of FINEX was supported by the Korean steel industry through its Korea New Steel Technology Research Association, with POSCO having a lead role and using its COREX plant as a commercial research laboratory to test the feasibility of using fine ores.

In the FINEX process, fine ore is reduced to 80-90% metallic iron in a series of fluidized beds, and the remaining iron oxides are reduced, carburized, and liquified with the formation of a slag in a melter-gasifier, which is analogous to that employed in the COREX process. As coal is gasified with oxygen in the melter-gasifier, the resulting reduction gases are supplied to the fluidized-bed system, with the remaining gas being similar in quality to that from COREX.

In February 1998, POSCO and RIST initiated work at Pohang on a FINEX pilot plant with a daily capacity of 150 tons. The plant went into operation in July 1999, and based on its performance, POSCO in December 2000 started construction of a FINEX demonstration plant to produce 600 thousand tons of molten iron annually. Completed and placed in operation in June 2003, the plant successfully demonstrated process and facility reliability over the following year, leading to a decision by POSCO to build a large-scale commercial plant.

In August 2004, ground was broken at Pohang for a commercial, US$1.1-billion FINEX plant with an expected annual capacity of 1.5 million tons. Compared to a new blast-furnace complex of the same scale, complete with coke ovens and sintering facilities, POSCO anticipates that its FINEX plant will cost 8% less to build, will deliver molten iron at a 17% savings in production costs, and will drastically reduce the usual emissions that result from the blast-furnace process.

Another direct-smelting alternative, the HIsmelt process, after undergoing more than 20 years of development and testing at multiple pilot plants, is approaching commercialization at Kwinana , Western Australia . HIsmelt employs a converter-like vessel with water-cooled sidewalls and a series of inlets around its lower perimeter through which downward-angled lances inject iron-ore fines, coal fines, and powdered lime. The materials are injected at high velocity directly into a molten bath (1450 deg C), which releases smelting gases that are burned above the bath by hot, oxygen-enriched air (1200 deg C), delivered by a hot-blast stove system and blown through a top-mounted lance. A mixture of metal and slag droplets and splashes erupts from the bath into the top space within the vessel and carries heat back to the bath to sustain the process. Upon injection, the ore, coal, and flux achieve significant penetration into the bath, leading to the dissolution of carbon into the metal and smelting reduction of the iron ore.

HIsmelt originated in Germany in 1982, when Klockner and Australia’s CRA Ltd. tested smelting reduction in a 60-ton, oxygen-steel converter and then built a small-scale pilot plant at Maxhutte Steelworks in southern Germany, where trials were conducted from 1984 to 1990. In November 1993, CRA and U.S.-based Midrex Corporation commissioned a research and development facility at Kwinana to scale up the process and use ore fines from Australia ’s Pilbara region. A 2.7-meter vertical smelter was built and commissioned in 1996-97, the results achieved in two years of operation eventually leading to a joint-venture agreement to install a commercial plant.

Construction of the plant commenced at Kwinana in January 2003, the project being pursued as a joint venture with four international partners: Australia ’s Rio Tinto, Ltd., which owns HIsmelt Corporation Pty. Ltd. and has a 60% share in the project; Nucor Corporation of the United States (25%); Japan ’s Mitsubishi Corporation (10%); and the Shougang Group of China (5%). The commercial joint-venture plant will employ a 6-meter diameter vessel and have an annual capacity to produce 820 thousand tons of hot metal, which will be cast into pigs of 4 to 6 kg. The plant is scheduled for completion by the end of 2004.

Among other blast-furnace alternatives, the ROMELT process currently produces molten iron, albeit on an experimental basis, at a plant in India . A single-stage process without pre-reduction, ROMELT uses non-coking coal to smelt a variety of ironbearing materials, including ore fines and waste oxides, in a horizontal furnace with both bottom and upper tuyeres, the former used to blow oxygen-enriched air to agitate and heat a swirling bath of molten slag in which reduction occurs, and the latter to blow pure oxygen to produce post-combustion heat.

ROMELT was developed in Russia by the Moscow Institute of Steel and Alloys and has been installed at an experimental facility in the state of Chattisgarh , India . Built by the National Mineral Development Corporation Ltd., the plant has a capacity of 300 thousand annual tons, and the company is working to scale up the process to the 1.5-million-ton level.

Two additional alternative-smelting projects of significance have been pursued in Japan and the United States , namely the DIOS Project and the AISI Direct Steelmaking Program. In both cases, considerable research and development was carried out over a number of years starting in 1988 and 1989, and although important progress was made in establishing process fundamentals, the projects essentially were concluded once pilot-plant trials had been completed.