ERI Home IRR Home Last updated 8/10/04
Iron is tough. It is tough to win from native ore, tough to work and shape, and tough in its application as technology's primary structural metal. And the iron and steel production industry is, appropriately enough, a particularly tough sector for U. S. businesses to survive in.
History had already begun to be recorded well before early metallurgists discovered the conditions necessary to reduce the iron in naturally occurring ores to the metallic state. Copper and bronze had been in widespread use for millennia before iron made an appearance (perhaps around 2,500 BC). The iron smelting process requires intense heat, far higher than can be reached in an open flame. Not until enclosed furnaces had been developed (probably for the original purpose of firing pottery), along with techniques for producing particularly hot flames (through the use of hardwood or charcoal, and of bellows for supplying a stream of air), were the prerequisites in place for iron reduction.
But while iron metal is difficult to make, iron ore is easy to find. Iron is one of the most abundant elements in the earth's crust (about 5% by weight), and usable deposits of iron ore are geographically widespread. Thus, possession of the mine does not confer control of the market. The prize goes to the most efficient producer, the one that can carry out an energy-intensive process, involving the processing of great quantities of raw materials, and can produce a finished material within tightly controlled specifications at the lowest cost. Adding to the already intense pressure on primary producers to minimize production costs are two further factors:
These fundamental facts about the nature of iron and its role in technology and the global economy may help explain the changes the U. S. iron and steel sector has undergone over the past several decades. According to historical statistics from the USGS, the production of pig iron (reduced from ore) has declined from a peak of 91.9 million metric tons in 1973 to 47.9 million metric tons in 2000, and the production of steel (refined iron, with carbon removed and alloying elements added) has declined from a peak of 137 million metric tons in 1973 to 102 million metric tons in 2000. Meanwhile, world consumption had continued to increase (although U. S. consumption has not). Iron and steel production has been steadily moving offshore.
What remains of the industry has advanced in a number of ways. Industry sources claim that "the discharge of air and water pollutants has been reduced by 90%" over the last 25 years. Even allowing for the 48% reduction in iron production and the 26% reduction in steel production during that time, that is an impressive accomplishment. The same sources note that the amount of energy used to produce a ton of steel has decreased by 45%. This progress has come at a price. Yearly capital expenditures on "environmental facilities" (presumably pollution control devices and the like) amount to 15% of total capital outlays, and the cost of operating and maintaining the facilities is $10 - $20 per ton of steel produced. Steel prices are typically in the range of a few hundred dollars per ton, but the amount per ton that companies must spend on environmental operating costs is about the same order of magnitude -- in fact, is slightly bigger than -- their typical profit margin on that ton. Managing environmental costs are survival issues for firms in the iron and steel sector today.
Notwithstanding its sizable investment in environmental infrastructure, the iron and steel sector is still one of the nation's biggest energy users, and still one of the biggest emitters of criteria air pollutants and greenhouse gases, of all U. S. industrial sectors. Its impact remains considerable, and further improvements in environmental performance would be significant. The question is, how much more room does the sector have to give?
Environmental impacts and risks
Quantitative impact data
Effects of existing and future regulations on impacts
Iron and steel making processes proceed through a series of successive refinements. Generally speaking, the earlier in the series, the greater the environmental impact.
The series begins with the refinement not of the iron-bearing material itself, but of the fuel-reactant needed to liberate the iron from its naturally occurring compounds. Once a suitable fuel is available, the iron itself is extracted from the ore. Finally, the crude material is purified and alloyed. These three steps are often termed cokemaking, ironmaking, and steelmaking.
Coking is essentially the cooking of coal. The object is twofold.
Environmental challenges abound throughout the cokemaking process. To promote good volatilization, the bituminous coal going into the process is pulverized, creating a particulate emission source. It is then subjected to intense heat in an oven for a prolonged period (typically 2000o F for anywhere from 14 - 36 hours) in a non-oxidizing atmosphere. The gases coming off the material contain a number of criteria air pollutants (carbon monoxide, nitrogen dioxide, and sulfur dioxide), hazardous air pollutants (such as toluene, naphthalene, phenol, and other aromatics, cyanide compounds, and hydrogen sulfide), and ammonia. Some of this material can be recovered as useful by-products, but losses are inevitable at each processing stage, and a number of RCRA hazardous sludges are associated with various recovery processes. The finished material is then pushed out of the oven (another opportunity for particulate release), and is often quenched with large quantities of water, which must then be disposed of as wastewater.
Cokemaking has historically been the largest source of emissions of all the processes associated with this sector. There has been significant rulemaking activity over the past decade as a result of provisions in the Clean Air Act Amendments of 1990, but the most significant effects of this activity have yet to be felt.
Ironmaking refers to the reduction of metallic iron from the oxide form in which the element is found in nature. The oxygen is transferred to carbon supplied by the coke (and sometimes also from supplementary fuels). Some extra oxygen is also added, in the form of heated air -- this extra oxygen burns additional carbon, the heat liberated by that combustion being used to sustain the intense temperatures needed to drive the iron reduction reaction. The product, called pig iron, contains a high percentage of carbon, along with various other impurities.
The process is carried out in a blast furnace, a large vertical reaction vessel whose basic design has not changed in millennia (see a posted excerpt from an AISE publication for an illustration of an early version, complete with clay blowpipes and goatskin bellows). Although improvements and refinements continue to be introduced, the design is dictated by the requirements of the process, and is unlikely to undergo radical changes in the foreseeable future. The design, in turn, entails environmental impacts that are hard to avoid.
One requirement is established by the extremely high reaction temperatures -- it would be very wasteful to allow the system to cool and then to have to reheat it, except on very rare occasions. The furnace must therefore be operated in continuous mode. As another consequence of the intense heat, the use of any kind of mechanical devices for conveying or mixing materials in the reaction vessel would be problematic, so the blast furnace must rely on gravity and convection to establish the needed flows of reactants and products.
To operate in continuous mode using only natural flows, chunks of ore and coke (along with a limestone flux, to help remove impurities from the molten iron) are added to the top of the vessel, and slowly move down. Hot air is introduced through nozzles around the bottom of the reactor, and the gases flow upward. Through a quirk of carbon chemistry, the air, an oxidizing medium, when present in sufficient quantities to burn carbon partially but not completely, gets transformed as it proceeds up through the furnace into an aggressively reducing medium, carbon monoxide. This molecule is sufficiently hungry for another oxygen that it can strip it from the iron oxide ore, leaving reduced molten iron to flow down toward the bottom of the furnace, where it can be periodically tapped off.
The convection requirement further constrains the physical form of the material that is added to the top of the furnace. As the solid material (unreduced ore and unburned fuel) moves down through the furnace, it must remain sufficiently permeable to allow the gases to move upward through it. The chunks must be strong enough to resist being crushed to powder by the weight of the material above it.
A number of environmental consequences flow from these design constraints. The furnace itself is a copious source of carbon monoxide (a criteria air pollutant), that can leak from the furnace at various openings. There are numerous ways to generate particulate emissions, one of the most important occurring during the tapping of molten iron. As environmental standards have improved, devices such as hoods have been installed over tapholes to collect emissions, but some losses are inevitable.
The need to convert the incoming material into suitably sized and sufficiently strong chunks creates another important set of consequences. We have already discussed impacts associated with the cokemaking process. Often, the ore is also subjected to processes that convert it into a more efficient blast furnace reactant. The most important of these conditioning operations is typically done at the ore mine, rather than at the steel mill: it involves pelletizing the ore by baking ore concentrate with a binder. Another is sintering, typically done at the mill, which involves baking combinations of iron-rich dusts from sources such as pollution control systems and ore fines, plus a various materials like water treatment plant sludges (where iron salts are added to aid coagulation). Both processes are associated with air emissions, from the baking process, as well as from materials handling.
The blast furnace also generates large quantities of slag -- impurities that float on the metal, and are skimmed off. Much of that material can be used in the construction industry (in roadbeds, for example), somewhat mitigating its impact.
A Technology Roadmap from the Department of Energy Office of Industrial Technologies describes a number of potential alternative technologies to the blast furnace for ironmaking, and suggests that 10-15% of the pig iron produced in the U. S. in 2015 will be produced using one of the alternative methods. It is interesting to contrast the implications of these new developments in ironmaking technology for energy efficiency with the implications for environmental impact. Direct reduction, an approach that eliminates the cokemaking step and uses coal, natural gas, or other suitable fuel in relatively unprocessed form, can be advantageous in terms of energy efficiency and productivity. From the environmental point of view, the direct reduction approach has less to offer. The inputs are essentially the same. All of the volatiles in the coal still have to go somewhere. The environmentally acceptable alternatives for them include recovery and reuse, or conversion to less toxic forms. Much of the volatile fraction is hydrocarbon pollutants, which can be oxidized to carbon dioxide (a greenhouse gas, but somewhat less objectionable in terms of toxicity or ground-level ozone forming potential). With a coke oven, the volatiles must be removed from the oven before oxidation, since any possibility of oxidation must be suppressed throughout the oven to keep from destroying the product. As noted above, blast furnaces have oxidizing regions, where combustion of the coke or other fuel occurs, and reducing regions, dominated by carbon monoxide, where metallic iron is reduced from the ore. Conceivably, conditions could be arranged to accomplish much of the oxidation of the volatiles directly in the oxidizing regions of the blast furnace, along with the elemental carbon fraction. This might be marginally preferable to the multistage control approach required with coke ovens, where the volatiles must be removed from the oven and oxidized in a separate pollution control stage (with the attendant inefficiencies and losses). But it is unlikely that direct reduction systems will be optimized for emissions control, at least not until the systems have attained the best possible energy efficiency and productivity improvements, so the direct reduction technology is not likely to become a positive development from an environmental perspective for some time to come.
Iron as it emerges from the blast furnace contains about 4% carbon -- too much for most applications. Steelmaking is the term applied to the last stage of refinement in composition of the end product. In addition to removing the carbon and other impurities, the steelmaking process also involves the addition of various alloying elements to give the finished material the combination of properties desired.
There are two main processes for making steel in current use. The Basic Oxygen Furnace (BOF) is used for processing most of the nation's pig iron production. The Electric Arc Furnace (EAF) is more generally used for recycled material.
In the BOF, atmospheric oxygen is blown through molten metal, either through water-cooled lances inserted from the top of the furnace, or through nozzles located toward the bottom. The oxygen burns off most of the carbon, and the heat of that reaction supplies more than enough energy to maintain the required temperature (in fact, the system must be cooled by the addition of more metal to keep the temperature within the desired range).
The environmental impact of the BOF comes mostly from the carbon monoxide that escapes from the melt, and the particulates that are generated as a result of the gas flows and the transfers of incoming and outgoing materials. These are significant, but are much more amenable to control devices than was the case with the older open-hearth technology. Other impacts include the generation of large quantities of water, both cooling water and water from wet scrubbers used for pollution control. Like the blast furnace, the BOF generates large quantities of slag. Because it is not as dimensionally stable as blast furnace slag, the use of BOF slag is more limited, and large quantities are landfilled.
The Electric Arc Furnace is used by a new breed of steel producer, the "minimill". Unlike the older integrated mills, which combined cokemaking, pig iron production, and steel production on one site, minimills use primarily scrap steel as their raw material, and can therefore skip the cokemaking and ironmaking steps. As recyclers, and as manufacturers that carry out the process with the smallest environmental footprint of the three major sector operations, minimills can presumably claim the environmental high ground. But the EAF is not without impacts of its own.
The use of electricity as the major heat source is a tradeoff. It is cleaner at the point of use, to be sure. But one pays a significant price in efficiency to convert heat energy into electricity, only to turn it back into heat. An overall efficiency of 35% (electric power out divided by heat energy in) is quite good for electric power generators. Turning it back into heat can be done much more efficiently. But for every calorie consumed by the electrodes in the EAF, about three calories of heat had to be expended at the power generator. If the heat came from fossil fuels, the greenhouse gas impact from the EAF is correspondingly three times greater than the impact from furnaces burning carbon directly. An Energy and Environment Profile of the U. S. Iron and Steel Industry (DoE-OIT, August, 2000) describes a growing trend for newer EAF operations to depend increasingly on burning carbon present in the melt to help generate the heat, reducing electricity consumption. This trend improves the greenhouse gas consumption picture, but increases the emissions due to the carbon combustion and the gas flow (since its mode of operation is approaching that of the BOF). Furthermore, the use of electric power generates nitrogen oxides in the arcing that occurs between the electrodes. In fact, comparison of the emission intensity of many of the criteria air pollutants, including nitrogen and sulfur oxides, carbon monoxide, particulates, and VOCs, shows, surprisingly enough, an order of magnitude higher emission per ton of steel from the EAF than from the BOF. (These comparisons are based on figures from the American Iron and Steel Institute, as recorded in the DoE-OIT Profile. It should be noted that these estimates include the emissions from the generation of the electricity consumed by the EAF, not just the emissions generated at the EAF itself.)
Although cokemaking and ironmaking account for much of the environmental impact from the iron and steel sector, the downstream processing of the steel as it emerges from the BOF or the EAF also involves significant emissions at a number of points. Materials transfer of molten steel, such as in casting processes, is generally accompanied by the generation of particulates. Alloying can also generate emissions, particularly when desulfurization is involved. Various surface finishing properties, including scale removal, and various chemical treatments such as pickling with hydrochloric acid, each have characteristic emission profiles.
Greenhouse gas emissions from a number of energy-intensive sectors have been calculated by NCMS on the basis of 1998 fuel consumption data from the Energy Information Administration (EIA) of the U. S. Department of Energy, and from greenhouse gas inventory data from the EPA (Inventory of U.S. Greenhouse Gas Emissions and Sinks). The results are posted in a summary document, Greenhouse Gas Estimates for Selected Industry Sectors. According to that calculation, the iron and steel sector (NAICS code 331111) was responsible for 224.8 Tg CO2 equivalent (1 Tg, or teragram = 1 million metric tons), of which 157.1 Tg was due to fuel consumption, and 67.7 Tg to non-fuel uses (such as the release of carbon dioxide from the use of limestone, calcium carbonate, as a flux material). This makes the sector one of the most significant contributors to greenhouse gas emissions of all industry sectors, second only to petroleum refining of all industry sectors at the six digit NAICS level. Iron and steel would even outweigh most other NAICS industry sectors aggregated at the three digit level -- apart from its own sector (primary metals, NAICS 331), only the chemicals sector, NAICS 325, would exceed it in annual greenhouse gas emissions.
Air emissions data for certain key criteria pollutants (ozone precursors) are available from the National Emission Trends (NET) database (1999), and hazardous air pollutant emissions data are available from the National Toxics Inventory (NTI) database (1996 is the most recent year for which final data are available). These numbers are direct emissions -- they do not include emissions associated with electric power consumed by a facility, but generated off site. For SIC code 3312 (blast furnaces and steel mills), the total emissions are:
(The criteria pollutants selected for inclusion in this analysis are those whose contributions to smog are the most injurious to health.)
Among all sectors (not just industry sectors) at the four digit SIC aggregation level, SIC 3312 is second only to electric power as a generator of fine particulates (PM2.5), and is in the top ten as a source of nitrogen oxides. It ranks eleventh as a source of VOCs.
About 100 gallons of wastewater are generated for every ton of coke produced. This is a combination of water that has come into direct contact with the hot coke as it emerges from the oven, plus water that is used at various stages in subsequent treatment of the quench water to remove various contaminants. This represents the most heavily impacted water, but it is only a small fraction of the total water requirements of the overall process. According to AISI figures cited in the DoE-OIT Profile, about 75,000 galloons of water are needed to produce a ton of steel. Since producing one ton of pig iron requires about a half ton of coke, the most heavily impacted wastewater fraction represents only a small percentage, less than 0.1%, of the total water requirement.
The iron and steel sector generates tens of millions of tons of solid wastes and residues annually (39 million tons in 1997, according to the DoE-OIT Profile). Blast furnace slag, at about 13 million tons, was the largest single contributor to this stream.
The most significant direct risks posed by the iron and steel sector are in the air quality area, primarily as a source of fine particles.
The sector is also a major contributor to nitrogen oxide and volatile organic compound (VOC) emissions. Among the criteria air pollutants, these categories are the most direct contributors to ozone generation. However, there are other sources of VOCs and NOx, particularly in urban areas (from vehicle exhausts, for example) that are comparable to or larger than blast furnaces and steel mills in terms of volume of pollutant generated in a typical geographical location. Thus iron and steel operations are one source among many, and are liable to feel significant, but not extraordinary, pressure from the surrounding communities to ameliorate their VOC and NOx emissions.
If there is an area in which the iron and steel sector seems particularly vulnerable to potential environmental liability, it would be as a generator of fine particles. The special concern with particles less than 2.5 microns in diameter, which are small enough to get trapped in the lungs and cannot be easily cleared through normal processes once lodged, is relatively new, but will probably be growing in public consciousness in the coming years. Iron and steel operations are the pre-eminent industrial source of fine particles. Electric power generation is the only sector generating more fine particle emissions, and it is a somewhat different case. Iron and steel process are associated with fugitive emissions less easily captured, or dispersed through tall stacks, than the contained combustion characteristic of the electric power sector. The problem is visible as it happens. The shower of glowing sparks that accompanies the transfer of hot metal has been in the past an impressive display characteristic of iron and steelmaking, and a source of pride. In a possible future spotlight of media concern, with a public conditioned to the the adverse effects of fine particle emissions, that paradigmatic picture could become a public relations problem for the industry.
Another noteworthy aspect of risk associated with the iron and steel sector is the unique status of coke oven emissions. When one scans down the list of substances defined as Hazardous Air Pollutants (HAPs) in Section 112 of the Clean Air Act, one sees:
An EPA fact sheet helpfully informs us that "coke oven emissions include coal tar, creosote, and coal tar pitch". In any case, the material admixture included under the "coke oven emissions" rubric, considered as a definable, if composite, substance in its own right, has well-studied toxicity properties, and has been the subject of extensive toxicological research. It is classified by EPA as a known human carcinogen.
Among all the environmental standards that apply to the iron and steel industry, air quality regulations have had the most profound effect on the sector since the age of environmental regulation began in earnest (ca. 1970). As indicated above it is likely that much of the future regulatory activity affecting this sector will also concern air quality. Accordingly, most of the following discussion will focus on air emissions regulations.
Cokemaking was the subject of a NESHAP finalized in 1993. Additional Maximum Achievable Control Technology (MACT) standards for certain aspects of coke production (pushing, quenching and battery stacks) were proposed in 2001. However, new cokemaking facilities are not being constructed, so virtually all of the anticipated improvement from these regulations will have to come from retrofits applied to existing installations. Figures available from industry sources stress the amount of money now being spent on environmental compliance, but there seems to be less hard data available on precisely how much of an impact reduction, specifically for the cokemaking stage, that this expenditure has bought.
Since cokemaking is generally lumped into SIC 3312 along with the other typical operations of an integrated iron and steel facility, and since the available EPA air quality databases are broken down by SIC code, it is difficult to measure the effect that the new regulations have had so far specifically on cokemaking. One can, however, compare data from the National Emissions Trends (NET) database on the EPA AirData website for 1999 with the same data for 1996, to see what kind of progress has been made for the 3312 SIC code in general. The results are tabulated below for a number of pollutants.
Table. Emissions data, in tons per year, selected pollutants, for SIC Code 3312
Data from EPA AirData Website
|Pollutant||1996 data||1999 data|
The 1999 results are the same as those quoted in the "Quantitative impact data" section above, with a few additional items. PM10 is a measure of somewhat less fine particles (diameter below 10 micron) than PM2.5 (diameter below 2.5 micron). The latter is considered the more serious health risk, as mentioned above. Ammonia information is available on the website, and may be a useful indicator for this discussion, since it is presumably a direct function of cokemaking operations, that being the most significant source of ammonia of all of the operations covered under SIC 3312.
If there was any substantial improvement in performance, as opposed to expenditure, as a result of regulations introduced in the early 1990s, it is hard to spot in these data. Increased production does not account for the increased emission; indeed, according to figures from the USGS, U. S. production of pig iron -- the greatest source of iron and steel related emissions -- actually declined from 49.4 million metric tons in 1996 to 46.3 million metric tons in 1999, although production of steel increased from 95.5 to 97.4 million metric tons during that year.
There is no question that iron and steel facilities have come a long way in environmental performance since the pre-EPA days. The industry's willingness and ability to approach environmental regulation as a collaborative venture has also improved substantially since the EPA's early forays into steel regulation. But this may be a good moment in history to see whether the dose-response relationship between new regulations and subsequent environmental improvement has been consistent in recent years with what it has been in the past.
The situation may change if a NESHAP covering integrated iron and steel production, proposed in 2001, becomes final. Completion was anticipated in 2001, but this has apparently not occurred (as of 10/31/02). [Note: the final rule was published on 5/20/03]. Rules affecting the industry are a bit of a political hot potato at the moment, due in part to the controversy over protective tariffs versus free trade now affecting the industry, and new environmental rules are likely to be caught up in that issue.
Although water quality and solid waste regulation is not expected to have as significant an effect on the industry in the foreseeable future as air emissions regulation, it should be noted that there has been considerable activity in the past in both water and solid waste rulemaking. In the water quality area, new effluent limit guidelines have been finalized, which will replace those which have been in effect since 1982 (and amended in 1984). The rule was signed by the Administrator on April 30, 2002. (Additional information is available from an EPA index page to background information on the rule.) The final version was published in the Federal Register in October. In conjunction with this analysis, it is interesting to note that a study focusing on wastewater aspects of iron and steel production prepared in 1995 by the EPA Office of Water states (p. xiv) "most process wastewaters from basic steelmaking operations are generated as a result of air emission control and gas cleaning" -- i. e. that air emissions are the root cause of much of the sector's water quality impact, consistent with the conclusions above.
In the area of solid and hazardous wastes, the major contributor on a tonnage basis to the considerable volume of solid waste generated by the sector is slag. Other sources include scale from rolling operations. Contrary to the situation with wastewater, dusts collected from air pollution control equipment constitute a somewhat smaller portion of the total solid waste stream, in terms of raw quantity. Much of this material is nonhazardous. However, there are several RCRA wastes associated with cokemaking (many generated in conjunction with recovering products from quench water), and one with the Electric Arc Furnace (K061 covers specifically "emission control dust/ sludge from the primary production of steel in electric furnaces"). There are no RCRA listed categorical wastes associated specifically with blast furnace or with basic oxygen furnace operations. As far as the near-term future is concerned. there do not appear to be any fundamental changes on the horizon for the sector regarding solid and hazardous waste.
For a clearly written and well-illustrated source on the cokemaking, ironmaking, and steelmaking processes, see the on-line presentation provided by AISI, at http://www.steel.org/learning/howmade/howmade.htm
The EPA Office of Enforcement and Compliance Assurance (OECA) has prepared a Sector Notebook on the Iron and Steel Industry, available at http://www.epa.gov/compliance/resources/publications/assistance/sectors/notebooks/iron.html
"Steel" is one of the Department of Energy Office of Industrial Technologies (DoE-OIT) "Industries of the Future". The DoE-OIT website at http://www.oit.doe.gov/steel/ provides:
A search engine specifically covering the steel industry, created by AISE, is located at http://www.steellinks.com/
The EPA program offices have prepared a number of documents in conjunction with rulemaking in their respective areas. The following documents are posted on the web; they may be found useful in evaluating the technical constraints associated with industry processes: