Still the Iron Age

by Vaclav Smil

  Iron Ore

  Iron ore dominates the material input into BFs, and its crustal abundance means that the production of primary iron has not been, and in the foreseeable future will not be, limited by the availability of this key resource. Iron is the third most abundant mineral element in the Earth’s crust, with 5%, following Al at 8.1% and Si at nearly 28%. Its content is between 2% and 3% in sedimentary rocks, 8.5% in basalt and gabbro. Iron is mostly found in oxygenated forms (and O2, with 46.6%, is the most abundant crustal element); contributions by carbonates, sulfides, and silicates are minor. Magnetite (Fe3O4), a dark grey or black oxide that can be found in igneous, metamorphic, and sedimentary rocks, has the highest iron content (72.36%), and its strong magnetism makes it easy to separate it from gangue and to produce a high-quality concentrate. Pure hematite (Fe2O3) contains 69.94% Fe; it, too, can be found in all kinds of rocks, and it is concentrated by gravity and flotation techniques. Hydrous oxide—above all goethite (HFeO2) and lepidocrocite (FeO(OH))—has just over 60% Fe, and siderite (FeCO3) is the most important carbonate ore, with 48.02% Fe.

  Most iron is found in sedimentary ores belonging to ancient (Precambrian) banded iron formation (BIF), metamorphosed sediments including the minerals itabirite, jaspilite, hematite-quartzite, and taconite; these silica-and-magnetite layers can be hundreds of meters thick and can extend for hundreds or even thousands of kilometers. These ores are commonly present in formations that are readily accessible by surface mining, and many large iron ore mines are among the world’s most extensive excavations stripping vegetation from large areas and turning them into landscapes of terraces, pits, and ponds. Ratios of material initially excavated to the final marketed product rise with the quality of shipped ore, and they commonly range from 4:1 to 7:1 for high-grade production. Adriaanse et al. (1997) used the ratio of just 2 tonnes of overburden per tonne of extracted iron ore, but it is difficult to estimate the most likely error associated with this global generalization.

  Average iron content of common ores varies from less than 20% to nearly 70%: weighted mean concentration at producing mines is now just short of 60%, with Brazilian and Australian ores above that mean and the Chinese ores having just 30–40% Fe. America’s leading iron resource, Mesabi Range taconite in Minnesota and other iron ranges in Michigan, has 30–40% Fe and 40–50% SiO2, and its development (and large-scale enrichment) started only in the 1950s in an effort to reduce reliance on imported ore (Cannon, 2011; Kakela, 1981). Lake Superior taconite, upgraded in pellet form, now supplies about 85% of the US iron ore demand. Since 1950, typical run-of-mine production has shifted from coarse hematites and deeper fine hematites during the 1950s and 1960s to coarse and fine hematites and rich itabirites in the 1970s, to fine hematites and rich and poor itabirites during the 1980s and 1990s, to widespread depletion of hematites and poorer itabirites since the year 2000 (Mourão, 2011).

  USGS puts the global resources of iron at more than 230 Gt contained within more than 800 Gt of crude iron ores and reserves at 87 Gt Fe and 190 Gt of crude ore (USGS, 2014). The reserve totals imply global R/P ratios of 61 for crude ore and 27 for its metallic content. Iron ores are found worldwide, but, as with most minerals, massive deposits (as well as high-quality reserves) are unevenly distributed, with Australia (26%), Brazil (18%), and Russia (16%) accounting for 60% of global iron reserves. China, now the world’s largest iron ore producer, has only about 8% of all reserves, while the United States has less than 2.5%. Global extraction totals, available since the beginning of the twentieth century, show more than half a dozen distinct periods. The production of iron ore (gross weight) rose from about 90 Mt in 1900 to 177 Mt in 1913, declined and stagnated during WW I and the first half of the 1920s, reached 201 Mt in 1929, and returned to that level only by 1937. WW II peak output of 235 Mt in 1942 was surpassed in 1950, and then steadily rising demand for steel brought the aggregate to 902 Mt by 1975, the level that was surpassed only in 1987.

  For the remainder of the century the production remained between 925 and 1070 Mt, but then, driven by China’s extraordinary rise in primary iron smelting (causing both rising domestic production and large imports), it grew by nearly 60% in a single decade (in 2010, the output was 2.59 Gt), and then nearly 25% higher, at 3.22 Gt, in 2014 (USGS, 2014). This means that iron ore production is the third most massive extractive enterprise in the world: its aggregate annual output is surpassed only by the total mass of fossil fuels and bulk construction materials.

  Production of iron ore has shifted with the regional dominance of iron smelting. Europe (including the USSR) was the leader until the 1970s; by the century’s end China was the largest producer, followed by Brazil and Australia; by 2014, China became even more dominant, while Australia produced more than twice as much as Brazil. But comparisons in terms of actual metal content tell a different story: China still leads, but its average of only 30% Fe shrank its 2012 output to 393 Mt Fe, compared to 315 Mt Fe for Australia (whose ores average 62% Fe) and about 258 Mt Fe for Brazil, where the average iron content is 52% (Mourão, 2011; USGS, 2014).

  The magnitude of the largest iron ore deposits and economies of scale in their extraction have led to a steady increase in the trade of this raw material (Polinares, 2012). The shares of traded iron ore rose from less than 40% in the early 1970s to 47% in the year 2000 and to 69% in 2013 (WSA, 2015). Two generations ago, Australia was the largest exporter but it accounted for less than 20% of the total, while Canada was almost as large an exporter of iron ore as Brazil, and Sweden and France were still among the top 10. Some 40 countries now export iron ore, but Australia and Brazil dominate the global market: in 2013 they sold 70% of all traded iron ore, and by 2020 they are expected to control 90% of the world’s seaborne ore trade (Jamasmie, 2014).

  Rio Tinto is Australia’s dominant company, producing from 15 mines, moving increasing shares of extracted ores by giant autonomous trucks, and having 1600 km of rail lines linking them with shipping terminals; the company is now developing the country’s largest integrated mineral extraction project in Pilbara, whose annual capacity is to reach 360 Mt in 2017 (Harding, 2014). Brazil’s production and exports are dominated by good-quality itabirite (50% Fe) in Minas Gerais (nearly 70% of nationwide output), and the rest comes from high-grade hematite ores (60% Fe) in Pará (Fig. 6.1). Vale is by far the largest producer (nearly 85%), followed by CSN, Samarco, and MMX (Brazilian Mining Association, 2015). Vale is also the world’s largest iron ore exporting company, followed by Rio Tinto and BHP Billiton. Intercontinental shipping of iron ore is done mostly in Capesize vessels (capacity of 140,000 deadweight tons), but much larger ships have been used to carry Brazilian ore, with Vale acquiring 35 362-m-long Valemax carriers rated at 400,000 deadweight tons (Vale, 2015).

  Figure 6.1 Aerial view of Carajas iron ore mine in Pará, Brazil. Corbis.

  China has been the largest iron ore importer since 2003, when it surpassed Japan: in 2013, it bought 820 Mt, or 64% of the world’s iron ore exports, and that total accounted for nearly 70% of the ore used in the country (WSA, 2014). Other large importers are Japan (135 Mt in 2013), South Korea, Germany, Poland, and Taiwan. China’s extraordinary growth of iron ore imports was the primary cause for a 15-fold increase in ore price between 2000 (when China imported just 70 Mt) and 2011, when the spot prices peaked at nearly $190/t; this has been followed by a fluctuating decline, with spot prices below $70 at the beginning of 2015, but major producers believe that there will be further increase in demand, particularly in Asia.

  Direct charging of crushed natural iron ore (whose lumps are sized at 6–30 mm for BFs and 6–18 mm for DRI (direct reduction of iron) plants) has become less and less common as the iron content of ores has declined and as various beneficiation (concentration), sintering, and pelletizing processes are used to improve the ore quality and enhance its suitability for efficient reduction (Poveromo, 2006). Efficient operation of BFs requires rich iron burden (preferably in excess of 58% Fe) and minimum shares of iron ore fines (particles of les
s than 5 mm), whose presence would impede gas flow and interfere with the reduction process.

  When narrowly defined, beneficiation entails removal of large quantities of nonferrous minerals, mostly SiO2 (reducing its presence to less than 10% by mass), and it is done by the sequence of milling (crushing, grinding) raw ore, its washing, sorting, sizing, magnetic separation, dewatering, and filtering. Flotation following magnetic separation is often used to upgrade concentrates by reducing their silica content. Broadly defined, beneficiation also includes agglomeration that takes mostly the form of pelletizing or sintering (USEPA, 1994). Sintering is the most common, and also most economical, form of thermal agglomeration.

  The process starts with the preparation of a raw mix of iron ore and spillage fines (sinter feed is typically 0.15–6 mm), coke fines or coal, and flux (lime, limestone, dolomite); water is added, and the mix goes into a sintering machine (onto a traveling grate), where it is ignited and the coke in the mixture provides the necessary heat (1300–1480 °C) to melt the surface and agglomerate the mix, forming a porous cake that is cooled and crushed into lumps of 15–25 mm suitable for BF burden (Outotec, 2015a). Typical material input is 2.5 t of raw mixture to produce a tonne of sinter. The process was developed to treat fine waste in the early twentieth century and sintered ores are now the dominant material charged into BFs, used to produce about 70% of all hot metal.

  Adding basic flux to sinter (that is, producing self-fluxing material) improves its physical and metallurgical properties (it acts as a binder and improves fine particle agglomeration), reduces the melting temperature of the ore, and eliminates direct charging of limestone or dolomite. Typical sintered ores contain about 57–58% Fe, 7–8% CaO, up to 2% of MgO, and 4–5% SiO2. Because sinter is prone to crushing during transportation, its production is usually done on ironmaking sites from ores shipped from domestic sources or brought by bulk carriers to seaside iron and steel mills (but care must be taken when shipping iron ore as well because its fines can liquefy in a ship’s hold). Sintered ore may be the only burden, but more often the charge is composed of a mixture of sintered and pelletized ore, sometimes with a small amount of raw ore.

  In contrast to sinter, which is porous and brittle, iron pellets, produced by agglomerating very fine-grained iron ore with binders (bentonite, small amount of limestone), are compact and hard and can be transported over long distances without crushing. Pellets (spheres of 9–16 mm in diameter) are formed from wetted ground ore in balling drums. Then they go into indurating furnaces, moving on a traveling grate on a 30–55-cm-thick bed to be heated (usually by natural gas burners) and then cooled and stored for shipment (Outotec, 2015b). Their thermal treatment (induration) produces a strong and fairly uniform charge that contains more than 60–65% Fe, 4–5% silica, and traces of Al, Mn, and P. Pelletization is done usually right at the mining sites or at the exporting ports, and processing units have capacities of up to 7.25 Mt/year. Global production of beneficiated ore rose from just over 900 Mt in 1990 to 2.33 Gt in 2013 (Schmöle, Lüngen, & Noldin, 2014).

  Metallurgical Coke

  Decline in specific coke requirements has been accompanied by substantial expansion of pig iron smelting and hence by absolute increase of coal extraction destined for ironmaking. Coking coal should have low ash and low sulfur content, and coals of different volatilities are blended to control the quality of the final product and the volume of by-product gases (Díez, Alvarez, & Barriocanal, 2002; Mussatti, 1998). As coal was displaced in railroad and water-borne transportation by liquid fuels, and as the main source of household, commercial, and industrial heating by fuel oil and natural gas, it retained its worldwide importance in only two major markets: in centralized electricity generation and as the source of metallurgical coke.

  Worldwide extraction of coking coal was just above 1 Gt in 2013, with China being by far the world’s largest producer with 527 Mt (Fig. 6.2). Australia, with 158 Mt, was a distant second, and expansion of ironmaking in Asia led to rising exports of coking coal: Australia (154 Mt in 2013), the United States, and Canada have been the largest sellers, while China (77 Mt in 2013), Japan, and India have been the largest buyers (WCA, 2015). Coal, delivered by trains and barges or imported in bulk carriers to coastal mills, is first pulverized (to just 0.15–3 mm), blended, and charged from special cars moving above empty hot ovens. Thermal distillation of coal (carbonization at about 1100 °C in oxygen-deficient atmosphere) proceeds in sealed coking ovens with heat transferred from hot brick walls (Valia, 2014; Fig. 6.3).

  Figure 6.2 Piles of coal ready for coking at an iron and steel mill in Hefei (Anhui province), in China. Corbis.

  Figure 6.3 Coke ovens at JFE’s Fukuyama Works (Hiroshima province). Reproduced by permission from JFE Steel.

  Coal decomposes at temperatures below 475 °C as it forms plastic layers, first near hot brick walls and then moving toward an oven’s center; higher temperatures bring releases of tar and aromatic hydrocarbons and coke shrinks and stabilizes at temperatures between 600 °C and 1100 °C. The coking process lasts 15–18 h to make BF coke and is longer (25–30 h) for foundry coke. Finished incandescent coke is pushed through open battery doors and rapidly quenched (in wet or dry process), crushed, and screened, and it is then ready for metallurgical use. Average lump size is about 5 cm across; European and US coke have less than 10% of ash (but up to 13.5% in China), less than 3% of moisture, and less than 1% of sulfur. One tonne of coking coal yields 600–800 kg of BF coke, 50–100 kg of coke breeze (undersize screenings smaller than 1 cm), roughly 300–350 m3 of coke oven gas, 30–45 L of tar, 10–13 L of ammonium sulfate, and 50–130 L of ammonia liquor (Sundholm et al., 1999).

  Mining of metallurgical coal accounted for about 12% of global coal extraction in 2013 (WCA, 2014), and, as repeatedly noted, decline in specific coke use has been accompanied by rising volume of pulverized coal directly injected into BFs. Global production of metallurgical coke rose from about 344 Mt in 2000 to 596 Mt a decade later, and it reached 685 Mt by 2013, while China’s rising dominance in ironmaking has raised the country’s share of the overall output to 35% in the year 2000, to 65% a decade later, and to 70% in 2013 (Jones, 2014). During the first decade of the twenty-first century aggregate coke output declined in Europe (by more than 20%) and North America (by a similar margin), increased slightly in Latin America, and, thanks mostly to China, expanded 2.3 times in Asia, with the global production rising from nearly 345 Mt in the year 2000 to more than 590 Mt in 2010 and to 685 Mt in 2013.

  Fluxing (slag-making) materials are the easiest ones to secure: they come from two abundant and widely distributed crustal minerals, from limestone (CaCO3) and dolomite, CaMg(CO3)2. In 2012, the United States had nearly 2000 limestone and 150 dolomite quarries (compared to only about 400 granite quarries). USGS provides detailed data for limestone use in construction but withholds data on flux stone in order not to disclose any proprietary information (USGS, 2014). In any case, the two minerals are in abundant supply, and their calcining releases CO2 and yields CaO and MgO, the two dominant metallurgical fluxes that used to be charged directly with lump ore but now are overwhelmingly incorporated into self-fluxing sintered or pelletized ores. Fluxes combine with SiO2 present in the ore and absorb sulfur and phosphorus, two undesirable ingredients of pig iron.

  Hot pig iron is, of course, the dominant input into BOFs: it can make up to 90% of the total charge, with much smaller additions of steel scrap (the largest heat sink during BOF operation) and fluxing materials as well as small amounts of ferroalloys. Fluxing materials should contain about six times more CaO than SiO2, and there should be enough MgO to saturate the slag and hence reduce the chemical erosion of the furnace’s MgO lining. The lining’s longevity is enhanced (to more than 20,000 heats per campaign) by slag splashing after tapping, that is, by blowing the residual liquid slag with pressurized nitrogen introduced through the oxygen lance. Ferroalloys, including manganese, silicon, aluminum boron, and titanium, are added from overhead bins to the ladle.

/>   Material Balances of Integrated Steelmaking

  The latest estimates by the World Steel Association for typical material inputs required to make a tonne of crude steel by the integrated steelmaking route (BF and BOF) are as follows: 1400 kg of iron ore (mostly as sinter or pellets or their combinations), 800 kg of coal (most of it converted to coke), 300 kg of fluxing materials (mostly for BF), and 120 kg of recycled steel. For a large BF producing 10,000 t/day of iron and supplying an adjacent large BOF, annual requirements add up to 5.11 Mt of ore, 2.92 Mt of coal, 1.09 Mt of flux materials, and nearly 0.5 Mt of steel scrap. A large integrated steel mill thus receives every year more than 9.5 Mt of materials that must be eventually introduced in a virtually punctiform fashion through the sealed top of a BF and into the open top of a BOF.

  In global terms in 2012 (with roughly 1170 Mt of pig iron and 1100 Mt of crude steel produced in integrated mills by BF–BOF sequence), this added up to about 1.64 Gt of iron ore, nearly 900 Mt of coal, about 330 Mt of fluxing materials, and 130 Mt of steel scrap. Extracting, producing, transporting, and handling this mass of materials requires a vast global system of mining, shipping, processing, and storage that has to operate without interruption. Inevitably, these massive inputs are transformed not only into the desired valuable metal (crude steel ready for further processing into specific alloys and semifinished and finished products) but also into massive solid or voluminous gaseous waste streams that must be either captured and reused or collected and recycled.