Still the Iron Age

by Vaclav Smil

  High content of free lime prevents the use of some slag in construction, but after separation both materials become usable, with lime best used as fertilizer. Because of its high content of basic compounds (typically about 38% CaO and 12% MgO), ordinary slag is an excellent fertilizer used to control soil pH in field cropping as well as in nurseries and parks and for lawn maintenance and land recultivation; slag also contains several important plant micronutrients, including copper, zinc, boron, and molybdenum.

  Life Cycle Assessments

  Life cycle assessment (LCA) is the most comprehensive approach to the compilation and evaluation of potential environmental impacts of entire product systems throughout their, often complex, history (ISO, 2006). LCA is particularly revealing as it allows us to compare environmental impacts in their entirety rather than, misleadingly, choosing a single (albeit the most important) variable or focusing only on a segment of a complex production process. Consequently, an LCA of steel should start with raw material extraction, include all relevant ironmaking and steelmaking processes, follow the material flow through manufacturing and use, and look at the recycling and disposal of obsolete products (WSA, 2011c).

  Assessed variables range from measures of toxicity to humans and ecotoxicity of water and sediments to nutrient loading (eutrophication), acidification, photochemical ozone creation potential (POCP), and global warming potential. Again, as is the case with energy analyses, data limitations and differences in analytical boundaries and conversion ratios may complicate the comparability of specific results. And, obviously, when comparing specific studies it is necessary to look at identical, or at least very similar, product categories, for example, at heavy-duty structural steel, or at least at a broad category of structural steel.

  There are now many published LCA values for steel. Besides LCAs for specific steel products or applications—for example, for truck wheels (PE International, 2012), tubular wind towers (Gervásio et al., 2014), or bridges (Hammervold, Reenaas, & Brattebø, 2013)—there are also assessments for average environmental impacts of national steel production—such as Thakkar et al. (2008) for India and Burchart-Korol (2013) for Poland—and national LCAs offering specific values for a wide range of finished products. A Canadian LCA (Markus Engineering Services, 2002) provided highly disaggregated impact values for nails; welded wire mesh and ladder wire; screws, nuts, and bolts; heavy trusses; open web joists; rebar rods; HSS, tubing; hot rolled sheet; cold rolled sheet; galvanized sheet; galvanized deck; and galvanized studs. And there are also LCAs for alternative resources in ironmaking (Vadenbo, Boesch, & Hellweg, 2013).

  Not surprisingly, given the commonalities (or outright identities) of major production processes, these published rates, while displaying national and regional differences, are generally in fairly close agreement, but care must be taken to compare the values for the same production routes (not for a product made by the BF/BOF route and another one using EAF) and for similar time periods. National averages and international appraisals of typical impact values suffice for the first-order comparisons with competing materials. LCAs of steel in Western economies show that advancing air and water pollution controls have removed the industry from the list of the most worrisome emitters, and they also indicate generally low or very low impacts in terms of human toxicities and ecotoxicities (on the order of 0.05 mg/t of crude steel), acidification (2–4 kg of SO2 equivalent per tonne of steel), eutrophication, and POCP.

  Carbon emissions: But, not surprisingly, LCAs of steel production also confirm that the sector’s high reliance on coal has made it an important emitter of greenhouse gases. The industry emits mostly CO2, and only small volumes of CH4 are released during coking—typically a mere 0.1 g/t of coke (IPCC, 2006)—sintering, and BF operation. Generation of CO2 is, of course, at the core of iron oxide reduction in BFs, as the oxides of iron react with CO produced by the combustion of coke and coal and produce pig iron and CO2. In addition, the calcination of carbonate fluxes produces CaO, MgO, and CO2. CO2 emissions during steelmaking are comparatively modest because pig iron contains no more than about 4% of carbon to be oxidized, while finished steel retains some of it.

  These iron- and steelmaking CO2 emissions cannot be eliminated as long as we rely on BF and BOF, and the only way to control them would be their capture and permanent storage. In contrast, CO2 emissions associated with ore mining, agglomeration, coking, and electricity consumption can be reduced by improving efficiencies of relevant conversions. And, of course, higher rates of scrap-based steelmaking are another source of reducing CO2 emissions. Specific emissions, all cited in t CO2/t of liquid steel, are: from 1.4 to 2.2 t for integrated steel mills in the West (typically about 1.8–2.0), but as much as 3.5 in India; 1.4 to 2.0 t for natural gas-based direct reduction processes, but as much as 3.3 in India for DRI using low-quality coal; and just 0.4 to 1.1 t for scrap-based smelting in EAFs (Gale & Freund, 2001; IEA, 2012; OECD, 2001; USEPA, 2008).

  Chen, Yin, and Ma (2014) put the 2012 mean at 2.3 t CO2/t of metal for the BF/BOF route and 1.7 t CO2/t of metal for EAFs (this high rate is due to China’s overwhelmingly coal-based electricity generation). But Thakkar et al. (2008) put direct emissions at only 2.01 t CO2/t for some of India’s large integrated steel mills, while according to Burchart-Korol (2013) the average Polish BF/BOF route emissions are as high as 2.46 t and EAF emissions are at 913 kg of CO2/t. Typical direct European emissions listed by Pardo et al. (2012) are (all in t CO2/t of crude steel) 2.27 for BFs, about 0.2 for BOFs, 0.24 for EAFs, between 0.08 and 0.09 for different hot mills, and just 0.008 (8 kg) for cold mills. CO2 emissions in Germany in 2013 averaged 1.466 t/t of product when measured in terms of finished steel products (22% reduction since 1990) and 1.328 t/t in terms of crude steel (Stahlinstitut VDEh, 2014). Average specific CO2 emissions of Canada’s iron and steel industry show a decline from 2.13 t/t of output in 1990 to 1.72 t in 2011 (Nyboer & Bennett, 2013).

  About 70% of all emissions from the BF/BOF sequence originate in preparing the charges into BFs and in their now prolonged operation. All of the following rates are expressed in kg CO2 per tonne of steel, and the shares of CO3 in the overall volumes of generated gases are in parentheses (IEA, 2012). Preparation of self-fluxing sinters emits mostly between 200 and 350 kg CO2 (290 kg might be a good average, with CO2 just 5–10% of the gas volume); lime kilns preparing CaO flux release 57 kg (30%); modern coking keeps the emissions below 300 kg (average 285 kg, 25%); generating hot blast in stoves adds about 330 kg (25%); and BF gas carrying away the products of iron ore reduction amounts to 1255 kg of CO2 equivalent, and their combustion in an adjacent electricity-generating plant releases about 700 kg/t (CO2 being about 20% of the flue gas). Finally, releases attributable to hot rolling and to BOFs add, respectively, about 85 and 65 kg. The total CO2 emissions thus come to at least 1.8 t per tonne of rolled coil (to be used in making cars or appliances).

  Calculating the global total of the steel industry’s CO2 emissions and expressing it as a share of global anthropogenic releases of the gas are exercises in unavoidable approximations. For example, IPCC (2007) put the industry’s share at 6–7% of anthropogenic CO2 emissions, and IEA (2008) put it at 4–5%. Assuming global averages of 2.1 t CO2 for integrated steelmaking (dominated by Chinese production) and 1 t CO2 for EAFs would yield 2012 emissions (with roughly 1.1 Gt of integrated and 0.45 Gt of EAF steel output) of 2.75 Gt. This would have been nearly 8% of total anthropogenic CO2 emissions in that year (about 35.6 Gt), more than 8% of all emissions attributable to the combustion of fossil fuels (about 33 Gt), and about 25% of all emissions from industries (11.5 Gt).

  My simple calculations are confirmed by the Steel CO2 Model by McKinsey (2014): it attributes 8% of the world’s 2011 CO2 emissions to steel (direct contribution of 5.6%, electricity generation 0.7%, and mining of ores, coal, and limestone 1.7%). That works out to about 31% of all industrial emissions estimated by McKinsey. Similarly, Hidalgo et al. (2003) put the share of the sector’s CO2 emissions at about 28% of the EU’s to
tal industrial releases. The iron and steel industry thus contributes twice as much as the emissions from chemical syntheses, and about 60% more than the production of cement and 45% more than the world’s oil and gas industry (electricity generation, with nearly 25%, is the largest contribution resulting from the combustion of fossil fuels).

  There are several effective ways to achieve considerable reductions of specific CO2 emissions, mainly thanks to the combination of the just reviewed decline in energy intensity of pig iron production and capture and reuse of CO2-rich BF gases, and in many countries, and notably in the United States, also thanks to the rising share of inherently less carbon-intensive scrap-based steelmaking. DRI aside, EAF steelmaking (increasingly in mini-mills) is the only large-scale commercial option to eliminate the use of coke, but its extent is obviously limited by scrap availability and price. When compared to a typical integrated mill, the energy requirement of a scrap-based mini-mill is just 50% (11 GJ/t vs. 22 GJ/t), carbon emissions are as little as one-quarter (0.5 t CO2/t vs. 2.0 t CO2/t), and the total material flux is less than one-tenth as large (0.25 t/t vs. 2.8–3.0 t/t).

  Expansion of EAF steelmaking has been, despite the significant overall growth of the metal’s global output, a major contributor to a relatively modest growth of industrial CO2 emissions. Further gains of coke-free steelmaking are likely: post-2010 availability of cheap natural gas (produced by hydraulic fracturing of shales) in the United States and Canada led some experts to expect that half of North America’s BF/BOF capacity will be replaced by DRI/EAF in 15 years (Laplace Conseil, 2013). Significant gains could still be achieved by near-universal adoption of the best existing practices. Given already high-energy conversion efficiencies, many specific reductions are modest, but their combination would yield improvements on the order of 10–15%, with the largest gains resulting from the installation of the best steam turbines in mill power plants, maximum use of pulverized coal injection, use of coke dry quenching, and BOF heat and gas recovery (Pardo et al., 2012).

  Injection of pulverized coal has been the most successful, and now widely used, option to reduce typical coke charges. Coke dry quenching began in a few plants during the 1970s, with the pioneering installations at the NSC Yawata works able to handle 56 t/h; Japanese data show that it reached about 60% of all operations by 1990 and that it became the standard practice by 2013 (Tezuka, 2014). Red-hot (1200 °C) coke is charged into a cooling tower where its heat content is exchanged with the bottom-blown circulating inert gas, and the gas is used to generate steam in an adjacent water boiler. Most of Japan’s dry-quenching plans (installed largely during the 1980s) have processing capacities of 140–200 t of coke per hour, while the largest plants in China can produce 260 t/h (NSSE, 2013).

  Coke dry quenching recovers waste heat equal to about 0.55 GJ/t of coke, and, moreover, using higher quality coke made by dry quenching reduces a typical BF coke charge by 0.28 GJ/t and cuts down on dust emissions (Worrell et al., 2010). Relatively smaller energy gains would come from universal scrap preheating, sinter plant waste heat recovery, optimized sinter/pellet ratios, oxy-fuel burners in EAFs, and pulverized coal injection (Lee & Sohn, 2014). And about 10–30% of all input energy leaves EAF as hot exhaust gas, but its capture and reuse are challenging due to its high dust content. Estimated costs of CO2 reductions range widely, depending on the targeted process, national peculiarities, and extent of controls, but they are no less than $50/t of CO2 and could be well above $100/t.

  Additional emission cuts will require new approaches, and the EU now supports a number of ultra-low CO2 steelmaking (ULCOS) projects whose eventual aim is to cut the emissions by half (JRC, 2011). The leading techniques include top gas recycling BF and HIsarna and ULCORED processes. Top gas recycling returns the generated gas into the furnace as a reducing agent instead of preheated air, and the first demonstration plant should be ready around 2020. The HIsarna process relies on preheated coal and partial pyrolysis for melting in a cyclone and on a smelter vessel for final ore reduction, but its commercial introduction is not foreseen before 2030. ULCORED would produce directly reduced solid iron and use pure oxygen instead of air, reducing gas produced from either methane or coal syngas, and remove CO2 by pressure swing absorption or amine washers (Knop, Hallin, & Burström, 2008). Again, the process is not expected to operate until 2030.

  Chapter 8

  Ubiquitous Uses of Steel

  Sectoral Consumption and the Quest for Quality

  Abstract

  Steel is a truly ubiquitous material: there is no industrial enterprise that would not, directly or indirectly, rely on it. There is no modern construction activity that can proceed without it: even if a building were to be constructed without any steel members or without any nails, by using doweled wood components, the wood would have to be cut by steel saws, and the building’s basement would have to be dug by steel machinery, or at least by steel shovels. There is no commercial or household activity that would not, or ultimately, owe its existence to it. There are no means of transportation that could function without it: even airplanes made solely from aluminum and composite fibers have to take off from and land on steel-reinforced runways, to say nothing about the metal used to build smelters and machines producing aluminum and carbon composites.

  Keywords

  Steel uses; construction; infrastructure and buildings; fuels and electricity; transportation; industrial equipment and consumer products; stainless steel

  Steel is a truly ubiquitous material: there is no industrial enterprise that would not, directly or indirectly, rely on it. There is no modern construction activity that can proceed without it: even if a building were to be constructed without any steel members or without any nails, by using doweled wood components, the wood would have to be cut by steel saws, and the building’s basement would have to be dug by steel machinery, or at least by steel shovels. There is no commercial or household activity that would not, ultimately, owe its existence to it. There are no means of transportation that could function without it: even airplanes that would be made solely from aluminum and composite fibers, they would have to take off from and land on steel-reinforced runways, to say nothing about the metal used to build smelters and machines producing aluminum and carbon composites.

  And this dependence only grows as societies proceed along the secular trajectory of economic development, but its forms obviously change as new uses arise and new steels are introduced to meet specific demands. Consequently, along with its enormous expansion, steel used during the past 150 years has seen major shifts in final destinations. During the 1870s and 1880s rails were the dominant product made of Bessemer steel. By 1900 the most expansive stage of worldwide railway construction was nearing its end; rails were still the leading finished product, but the twentieth century saw several major shifts in final steel uses.

  At its beginning buildings with steel skeletons were still limited to a relatively small number of skyscrapers in a handful of the US cities, and the use of reinforcing steel in concrete roads, dams, and buildings was in its infancy. Car manufacturing, 8 years before launching the Ford Model T, was still a small-scale, artisanal industry akin to assembling bicycles that was producing a few thousand vehicles a year that only rich people could afford. As a result, there was still no large-scale construction of concrete roads reinforced with steel. There were no washing machines and only a small number of electric stoves; food preservation using steel cans was relatively common, but that was in part because there were still no affordable electrical refrigerators. A century later the global steel output was dominated (about 55% of the total output in 2013) by hot-rolled flat pieces required to make industrial and transportation machinery and equipment and household goods, and reinforcing bars for construction accounted for nearly 15% of all shipments (WSA, 2015).

  No comprehensive statistics according to consistently defined categories are kept for the metal’s final consumption because steel industry’s reporting ends with its output of intermediate products. Wo
rld Steel Association offered the following division for 2011: construction 51.2%; mechanical machinery 14.5%; metal products 12.5%; automotive 12%; other transport 4.8%; electrical equipment 3.0%; domestic appliances 2% (WSA, 2012). Cullen et al. (2012) attempted a comprehensive and accurate mapping of the global steel supply chain for the year 2008: it traces the progress in five steps, from reduction to steelmaking, casting, rolling/forming, and fabrication.

  Not surprisingly, there are many data and classification problems inherent in such complex exercises. For example, the two published values for reinforcing bar production in 2008 were 147 Mt and 210 Mt, while Cullen et al. (2012) ended up with a solved value (required to balance the total for hot-rolled long products) of 174 Mt. The final product of their analyses is a Sankey diagram, a revealing visual presentation of the scale and the complexity of steel flows. They aggregated end-use products into four major categories: vehicles, industrial equipment, construction, and metal goods. Construction was by far the largest category with 54.8% of all final steel uses, divided between buildings (32.9%) and infrastructure (21.9%). Metal goods and industrial equipment were about equal (16.3% and 16.2%), with the former split into appliances (2.8%), food packaging (just 0.7%), and other goods, and the latter divided into mechanical equipment (13.1%) and electrical equipment (3.1%). The fourth major category, labeled vehicles, claimed 12.7% of the total use, with 8% for cars, 2.8% for ships, and 2% for trucks.