Environ. Sci. Technol. 2006, 40, 7060-7069
The Contemporary Anthropogenic Chromium Cycle J E R E M I A H J O H N S O N , * ,† LAURA SCHEWEL,† AND T. E. GRAEDEL‡ Program in Environmental Engineering, Yale University, and School of Forestry and Environmental Studies, Yale University
Chromium is an essential engineering metal used in stainless and alloy steels, chemicals, and refractory products. Using material flow analysis, all major anthropogenic chromium flows are characterized for the year 2000, from mining through discard, on three spatial levels: fiftyfour countries, nine world regions, and the planet. Included is the first detailed quantification of chromium in internationally traded finished products and diverse waste streams. Findings include (1) 78% of chromium flow entering final use is added as a net addition to stock on the global level; most countries are close to this figure; (2) the majority of mining occurs in Africa (2400 Gg Cr/yr) and the Commonwealth of Independent States (1090 Gg Cr/ yr), while the major end-users are Asia, Europe, and North America at 1150, 1140, and 751 Gg Cr/yr, respectively; (3) waste flows of chromium are the greatest in Europe (420 Gg Cr/yr), Asia (370 Gg Cr/yr), and North America (290 Gg Cr/yr), but the composition of these waste flows varies greatly among the world regions; (4) releases of chromium by the global system, which total 2630 Gg Cr/yr, are nearly evenly divided among tailings, ferrochromium slag, downgraded scrap, and post-consumer losses; (5) many countries have a heavy foreign dependence on chromium in the all forms, as is demonstrated for the United States. The findings relating to in-use stock changes and finished product trade are relevant to industry, allowing for more accurate planning for future scrap availability. The quantification of releases due to discards and dissipation hold environmental and human health relevance, while the full life cycle international trade assessment addresses local scarcity.
Introduction Chromium is a critical metal used in dozens of products that we rely on every day, but it is seldom used alone. The most common application is in metallurgical end uses in alloys, consuming 90% of virgin chromium. The addition of chromium adds corrosion and oxidation resistance to metals, making steel “stainless”. While other alloying elements, such as nickel and molybdenum, may also be added, chromium is an essential ingredient and no suitable substitute is known. Corrosion resistance extends the life of products, allows industrial activities to occur in harsh environments and with harsh chemicals, and reduces replacement costs. (The Department of Transportation estimated that corrosion cost * Corresponding author phone: +1-203-506-9439; fax +1-203432-5556; e-mail:
[email protected]. † Program in Environmental Engineering. ‡ School of Forestry and Environmental Studies. 7060
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 22, 2006
the United States economy $276 billion in 1998, amounting to 3.1% of the gross domestic product (1).) In addition to metallurgical uses, chromium is also used in refractories and foundry sands for its heat resistance, and in chemicals for leather tanning, pigmentation, and wood preservation. Significant attention has been paid to the adverse health effects of chromium, which are highly dependent on oxidation state. There is sufficient evidence to demonstrate carcinogenicity in humans of hexavalent chromium in the chromate, chromate pigment, and chromium plating industries; limited evidence for carcinogenicity of chromic acid and sodium dichromate; and inadequate evidence for metallic and trivalent compounds (2). Effects of hexavalent chromium exposure include respiratory cancer, kidney damage, and skin irritation. Because the United States does not currently mine chromite ore, but nonetheless relies on chromium for military and aviation applications, it has identified chromium as a critical and strategic metal. A handful of countries, principally South Africa and Kazakhstan, supply most of the virgin chromium to the rest of the world. Material flow analysis (MFA) models for metals at the national and global levels for short time intervals (e.g., 1 year) are a well established approach to material flow studies. The Stocks and Flows (STAF) project at Yale University (3), of which this research project is a part, has created such models for copper (4), zinc (5), and silver (6). Several studies exist that quantify the chromium cycle for a defined region or limited scope. These include a historical study of chromium use in Sweden (7, 8), a Danish study centered on the end uses, disposal of chromium, and trace constituency mobilization (i.e., coal, oil, and cement) (9), and an analysis on the Flemish region of Belgium (10). Gabler (11) and Papp (12) each published studies detailing the recycling of chromium. Papp additionally published “Chromium Life Cycle Study”, which examined the life cycle of both chromium-containing products and minerals, calculated production data for all significantly contributing countries, and made estimates for global releases to the environment through both anthropogenic and natural means (13). Using techniques of material flow analyses (MFA), this article presents anthropogenic chromium cycles for fiftyfour countries or country groups, nine world regions, and the planet as a whole. Tens of thousands of data points were collected from dozens of sources, including industry coalitions, governmental and non-governmental organizations, site visits, and expert opinions. By examining the entire life cycle, from mining through discard, the fate of chromium in products has been determined, releases from the system have been identified, and opportunities for more efficient resource use are presented. This study also lays the foundation for a series of further analyses on such topics as energy use, environmental impact, and future scrap availability.
Methodology STAF research on technological material cycles for metals is conducted using a framework consisting of four life stages, as shown in Figure 1(a). Production involves the mining, milling, and smelting processes; fabrication and manufacture (F&M) entails the creation of semi-finished and finished products; the use life stage involves the consumer, commercial, and industrial end uses of the finished products; waste management and recycling (WM&R) deals with endof-life. Scrap may be recycled in one of three forms: home (internal) scrap, prompt (industrial, but external) scrap, and old (post-consumer) scrap. Import and export occurs after 10.1021/es060061i CCC: $33.50
2006 American Chemical Society Published on Web 10/17/2006
FIGURE 1. (a): A simplified schematic diagram of a technological resource cycle, with successive life stages plotted from left to right; (b) system boundary diagram for chromium many of the transformations. Underlying these flows is a series of stocks, or reservoirs, that feed or ingest material to or from the system. Such reservoirs include geological resources, production stocks, government stocks, investor stocks, in-use stocks, and post-consumer discards (e.g., landfills). (Some of these reservoirs are minimal or negligible sources for material.) A system boundary diagram was created specifically for chromium (Figure 1b), laying the framework for data collection and cycle generation. Limitations in data availability and tractability prevent us from quantifying all chromium flows with great accuracy. Nonetheless, relatively complete estimates are of substantial utility. To quantify the preponderance of each flow, fifty-four countries or country groups were included in this study, accounting for 94% of the global gross domestic product (GDP), 77% of the world’s population, 98% of the chromite mine production, and over 99% of stainless steel production. The year chosen for this study was 2000, in order to provide a timely understanding of use with adequately available data.
Results Production. The mining of chromite ore and the associated tailings losses occur in several nations with ample geological reserves; the top five mining countries produce nearly 90% of global chromite, as expressed in contained chromium (i.e., excluding non-chromium material). County-specific chromium concentrations (14) were applied to mine production (15, 16) and import/export data (15). Details are provided in the Supporting Information. Ferrochromium (abbreviated FeCr) is produced as an intermediate product for metallurgical uses. The production and import/export of three categories were examined: high carbon charge grade (HCCH), other FeCr, and ferro-silicochrome (15). Country-level Cr concentrations were applied (17) and the FeCr entering F&M was determined by subtracting the net export from domestic production. An estimate of FeCr smelting recovery rates of 80% is assumed for all countries (13), except Finland which was assigned an 85% rate and validated by a site visit. These calculations result in VOL. 40, NO. 22, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
7061
TABLE 1. Internationally Traded Finished Products, as Ranked by Contained Chromium product
HS1996 code
assumed Cr contenta
global trade (Gg Cr)
top exporting countriesb
top importing countriesa
passenger cars stainless steel kitchen items automotive exhaust systems cutlery turbo jets/other gas turbine engines pumps food processing machines tanning chemicals tableware (not knives) taps, cocks, and valves
8703 732393 870892 8211 8411 8413 8438 320290 8215 8481
0.27% 18% 9.8% 11.2% 16% 2.7% 14.4% 17% 12.3% 0.72%
91 88 45 42 40 40 37 37 32 24
Japan, Germany China, Hong Kong US, France China, Hong Kong US, Germany US, Germany Germany, Italy Germany, Italy China, Hong Kong US, Italy
US, Germany US, Mexico Canada, Germany US, Hong Kong US, Germany US, Canada US, Mexico South Korea, China US, Hong Kong US, Canada
a This column represents the Cr concentration of the entire category, i.e., it factors in both chromium-containing and non-chromium containing products. b By mass of chromium.
global production of 2940 Gg Cr in HCCH, 258 Gg Cr in other FeCr, 55 Gg Cr in ferro-silico-chrome, and slag and dust losses of 637 Gg Cr. Details are available in the Supporting Information. (Note: 1 Gg ) 1000 metric tons.) The atmospheric losses were approximately 31 Gg Cr, using an emission factor of 5791 g Cr/t FeCr produced (13). Technology for metals recovery from slag has been developed and is now becoming more widely used. Based on anecdotal evidence (16), this recovery was estimated for South Africa, Turkey, and Zimbabwe. Current use of this recovery has increased since 2000. In 2000, 815 Gg chemical grade ore, 158 Gg refractory ore, and 318 Gg foundry sand ore were used (15). Assuming 44, 40, and 40% chromite content, respectively (18), the contained chromium entering F&M was 245, 43, and 87 Gg Cr. Some governments stockpile natural resources to mitigate the effects of trade problems (16) and net stockpile changes must be quantified in these cases. Fabrication and Manufacture. This study finds that 86% of the chromium entering fabrication is used in stainless steel; thus, changes in the use of stainless steel and its rate of recycling will have a direct and appreciable effect on chromium. It was estimated that 1.4% of the contained chromium used in the production of stainless steel ended up as slag (13) and between 0.12 and 0.30% was lost as atmospheric chromium releases. These are included as a subset of industrial waste (IW). Details are available in the Supporting Information. The most significant form of chromium recycling is in the form of stainless steel scrap. Calculations on the percentage of chromium in stainless steel that is generated from scrap were based on data from Pariser (19). This resulted in a global average of 42% of chromium coming from secondary sources, which compared well to another estimation of 35-40% (12). These findings are not at odds with claims that 60% of stainless steel is recycled: some of the recycled content in stainless steel is from carbon steel, which increases the total amount of recycled iron, but does not add secondary units of chromium. Of the stainless steel scrap that is utilized, an assumption is made that 20% comes from home scrap (not shown on the final cycles), 32% from prompt scrap, and 48% from old scrap. The fabrication of chromium chemicals, refractories, and foundry sands has not been documented on country-level basis. The global capacity for chromium chemical production (20) and ore use, suggests that global chromium chemical production was operating at 74% capacity. It was assumed that Russia was operating well below its capacity (at 50% capacity), the U.S., U.K., Japan, and Turkey were operating above the global average capacity (at 85% capacity), and all countries were realizing 15% losses (13), which are included in the Hazardous Waste (HW) flows due to the prevalence of hexavalent chromium. 7062
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 22, 2006
A model was also developed for refractory and foundry sand production. Estimated production data were available for Canada, Japan, the U.S., and Western Europe (21), and production for other nations was determined utilizing import/export data. A production loss rate of 30% was applied (13). This waste stream, totaling 37 Gg Cr, is also included in HW. Cr metal is a highly refined product that is used in very small quantities in specialized end uses (including superalloys). This study assumes aluminothermic Cr metal production uses ferrochromium as a feed source, whereas electrolytic production uses chromium chemicals. Cr metal production was estimated for 2000 at 21.5 Gg Cr (22) (see the Supporting Information), and the International Chromium Development Association Statistical Bulletin was used to determine import and export flows (15). The percent of chromium used in each end use (23), together with the apparent consumption of stainless steel (24), was used to calculate product manufacture quantities. End use categories and product examples are given below: • Buildings and infrastructure: elevators, railways • Transportation: automobiles, ships • Industrial machinery: heat exchangers, tanks • Household appliances and electronics: washing machines, dish washers • Metal goods and other end uses: cutlery, fasteners Semi-finished and finished product trade play significant roles in the chromium technological cycle. Semi-finished trade included stainless steel, steel alloy, chromium metal, and four chromium chemicals; product trade encompassed 74 chromium-containing products. Using the United Nations Comtrade database, import and export data were collected for each country, and concentrations, based on published literature and expert opinion, were applied. Table 1 shows the top ten products, ranked by chromium content traded. These constituted 52% of the chromium trade in finished products. Over 900 Gg of chromium (nearly 30% of the chromium entering use) undergoes international trade as a finished product (see the Supporting Information). Non-hazardous industrial waste (IW) is defined as the losses during F&M that are not promptly captured in the scrap system. These losses are estimated to be 9.3% of the chromium used in fabrication of stainless steel (including stainless steel slag and dust) and 1.5% of chromium in product manufacturing (11). In addition to these losses, chromium-containing alloys may also be recycled in a lower form (e.g., stainless steel being recycled as carbon steel). This is considered “downgraded scrap,” and the chromium contained therein leaves the chromium cycle when this occurs. High losses of chromium (5% of fabrication and 15% of manufacturing chromium) for downgraded scrap are reported (11). It is assumed that ferritic scrap, which has low or no nickel
content, is downgraded at a much higher rate than austenitic scrap which contains significant concentrations of valuable nickel. Use. The products entering use are determined by the difference of product manufacture within a country and its net export. Globally, metal goods and other uses makes up 30% of the flow, buildings and infrastructure and industrial machinery are 25% each, transportation is 15%, and household appliances and electronics are 5%. As economies develop, accumulation of material as inuse stock generally occurs. The net annual addition to inuse stock is determined by the difference between material entering use and that leaving for WM&R, less any dissipative uses. Little chromium is dissipated during use relative to other flows. Dissipation from 304- and 315-grade stainless steel due to environmental exposure produces annual losses of 0.25-0.3 mg Cr/m2-yr (25). To accurately determine dissipation, the stainless steel in-use stock and the percentage of this steel that regularly comes in contact with rain must be known. These stock calculations are outside the scope of this study, but a rough estimation was conducted, assuming the in-use stock is twenty times greater than the annual use, 10% of this stock resides outdoors, and the average thickness is 1 mm. This order of magnitude estimation leads to annual dissipation of 200 kg Cr, a flow that is negligible from the bulk material perspective, but may be interesting from an environmental perspective. Leather, pigments, and treated wood use chromium dissipatively. Chromium’s ability to migrate in leather is limited (430-980 ppm) (26), so it assumed that chromium enters municipal solid waste at end-of-life. The bulk of pigments used in household goods will likely remain there until their disposal. Yellow road paint (approximately 2.2 wt % Cr) (27) is assumed to completely dissipate, but data are insufficient to estimate these losses. Chromated copper arsenate is used in wood preservation for such end uses as transmission poles, fencing, bridges, and decking, with 60% of chromium chemicals being so employed (18). Hexavalent chromium (as chromic acid) constitutes 47.5 wt % of the CCA wood preservatives, and approximately 2-3% of the chromic acid leaches from treated wood (28). A rough estimate of stock was used (10 years’ flow), with an assumption of 70% wood in contact with water, leading to an annual global dissipation of 7 Gg Cr from treated wood. The remainder of the Cr stays within the wood until its disposal or incineration. Because calculations indicate that rates of dissipation are small, and the uncertainty is great, dissipation flows are not included on our cycle diagrams. Waste Management and Recycling. Flows into WM&R include IW and HW (previously discussed), Waste from electronics and electrical equipment (WEEE), municipal solid waste (MSW), end-of-life vehicles (ELV), construction and demolition debris (C&D), sewage/sewage sludge (S/SS), and other discards. Details are provided in the Supporting Information. Data on bulk per capita flows of WEEE are taken on a regional, and in some cases, country-level basis (29). Weigand et al. calculated the Cr concentration of electroscrap to be 730 mg Cr per kg of electroscrap using composition analysis (30), and this figure was employed to derive the Cr flows. Bulk MSW flows were taken from several sources (31-36) and estimated based on GDP when data were not available. A Cr content of MSW of 210 ppm was applied in developed countries, based on physicochemical analyses of 34 waste streams (30), with an assumed 8% metals content in MSW. For developing countries, the Cr-to-total metal ratio was held constant, but the total metal content of MSW was scaled
down using country-specific data when available and estimates based on GDP when such data were not available. ELV flows are calculated using country-specific data on passenger cars and commercial vehicles in use in 1992 (37), chosen because the profile of cars in use at this time would be entering end-of-life in 2000. The bulk ELV flows were calculated with a weighted median product lifetime of 13.5 years and an average car weight of 1530 kg (38). An independent check of these results for European countries differed by only 15% (31). A concentration of 0.27% Cr in automobiles was determined based on 19 kg of stainless steel (with an estimated 17% Cr concentration) per automobile plus 0.91 kg of Cr not used in stainless steel applications (38). Data on bulk flow for C&D were available for 28 countries from four sources (39-42). The mean for these countries with less than $3000 GDP was 16 kg C&D discard/capitayear; for $3000-20 000 the mean was 260 kg, and for countries about $20 000, the mean was 420 kg. For the 26 countries where data were not available, these means were applied. A waste stream analysis found the average chromium concentration to be 150 mg Cr/kg discard (43). Country-level data for Europe and the United States were obtained for S/SS (44, 45). For other countries, the average per capita generation rate of 19.5 kg/capita was used, scaled down according to the percent of people connected to sanitation services (46). Chromium concentrations ranged from 25 to 157 mg Cr/kg discard (44), with an approximate average of 60 mg Cr/kg discard used for the countries where specific data were not available. Other discards include industrial machinery and large transportation units (e.g., planes, trains, and ships). Industrial machinery discards were estimated assuming a 5% longterm growth rate in stainless steel, a 25-year lifetime of equipment (USGS, personal communication, 2006), and a constant market share of industrial machinery within stainless steel. This led to a discard flow equaling 30% of industrial machinery entering use in 2000, containing 250 Gg of chromium. The discard of planes, trains, and ships, using a 30-year lifetime (USGS, personal communication, 2006) led to a discard flow containing 30 Gg of chromium. The export of old scrap was calculated using the United Nations Comtrade database, resulting in approximately 300 Gg Cr traded. The authors are aware that much informal scrap trade does exist, and more detailed information on these flows would improve the rigor of this study. Data Quality and Data Utility. The data that were used to characterize the chromium cycles were large in number and variable in quality. Production and F&M tend to have the highest quality data, whereas that for trade, scrap, and WM&R are scarce and less certain. That being said, a cycle in which the flows are accurate to ( 25% is still of substantial utility and can be used to inform policy makers and engineers. Country-Level Cycles. Country-level cycles were generated for every country and country group in this study (see the Supporting Information). After accounting for net changes in stocks, each life stage should achieve perfect closure: input flows equal output flows. Due to data imperfections, such balance is not seamlessly achieved. To reconcile this discrepancy, phantom flows are added to each of the reservoirs to achieve closure and are represented in the diagrams by values encased in ovals with dashed arrows pointing into or out of the life stage. These flows are needed for most cycles, but are often less than 10% of the total flows throughout the life stage. In cases where larger phantom flows exist, significant data improvements are needed. Country-level cycles vary greatly and much can be inferred by their examination. The cycle for South Africa shows a country utilizing its vast mineralogical resources, and exporting ore, concentrate, and FeCr. Use in finished products is quite small. China imports raw materials, adds VOL. 40, NO. 22, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
7063
7064
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 22, 2006
FIGURE 2. Regional-level chromium cycles, ca. 2000, for (a) Africa, (b), Asia, (c) Commonwealth of Independent States, (d) Europe, (e) Latin America and the Caribbean, (f) the Middle East, (g) North America, (h) Oceania, and (i) Antarctica. The values encased in the ovals represent phantom flows required to achieve closure (see text). FeCr ) Ferrochromium; IW ) Industrial Waste; HW ) Hazardous Waste. All values are given in Gg Cr/yr. value, and exports the resulting products. Since 2000, Chinese stainless steel melting capacity has multiplied; its share of stainless steel production has increased from 2.3% of global production in 2000 to an estimated 19% in 2005, with further expansions planned. The United States has no active mining operations and significant net imports of ferrochromium, semi-finished products, and finished products. The net import of products (106 Gg Cr/yr) tells only part of the story: the U.S. has the largest magnitudes of both import and export of chromium contained in products, at 216 and 110 Gg Cr/yr, respectively. The United States is also the only country in this study with a documented change in government stock. In 2000, the U.S. government sold 90 Gg Cr in the form of ore (55 Gg Cr), ferrochromium (35 Gg Cr), and chromium metal (0.17 Gg Cr). Chromium losses in different parts of the life cycle vary by nation. Those in South Africa, a developing nation with ample natural resources, are overwhelmingly from mining and ferrochromium smelting in the production life stage. The U.S. sees the bulk of the chromium losses occurring in post-consumer discards, while China is evenly split between the two. Regional-Level Cycles. Nine regional chromium cycles are shown in Figure 2 Africa (Figure 2a) is dominated by South Africa’s flows, with some additional production activity occurring in Zimbabwe. Little post-production activity is apparent, with very small flows entering use. The Asian chromium cycle (Figure 2b) shows a region active in all life stages, with net import of raw materials, value adding processes, and net export of semi-finished and finished products. The large scope of Japanese stainless steel production coupled with the burgeoning growth of the Chinese stainless steel industry in the past several years will likely make this region the dominant force in stainless steel fabrication. Figure 2c shows that the Commonwealth of Independent States (CIS) has a cycle that is similar to Africa’s. Substantial chromite ore is mined, predominately in Kazakhstan, and low magnitude flows exist post-production. Significant chromium chemical manufacture occurs in this region, explaining why the HW flow for CIS is among the highest of the regions. The European cycle (Figure 2d) shows little mining activity, all of which occurs in Finland. Large import flows of ore, concentrate, and ferrochromium feed the industrial sector, which meets the needs of the continent and allows for a net export of semi-finished and finished products. Latin America and the Caribbean (Figure 2e), the Middle East (Figure 2f), and Oceania (Figure 2h) all have relatively
FIGURE 3. Global-level best-estimate chromium cycle, ca. 2000. FeCr ) Ferrochromium; IW ) Industrial Waste; HW ) Hazardous Waste. All values are given in Gg Cr/yr. small flows throughout their entire life cycle. The three regions have modest mining operations, export most of their ore and concentrate, and have small net imports of finished products. Of these three regions, the only stainless steel producing country is Brazil; the rest of the nations rely on import of semi-finished and finished products. Figure 2g shows the North American chromium cycle. This region has no mining operations and relies heavily on the import of ore, concentrate, and semi-finished and finished products. Because no mining and little ferrochromium production occurs here, this region does not have large flows of production wastes. Antarctica’s chromium cycle (Figure 2i) was based on a unique case study by Klee that examined all shipments into and out of the continent (47). The flows are tiny compared to the other regions, but their assessment renders this study a true global perspective on chromium use. Global Cycle. A best-estimate chromium cycle (Figure 3) was created by aggregating the regional cycles and eliminating the residual “global import” and phantom flows. Details are provided in the Supporting Information. The cycle shows that of the chromium removed from the lithosphere, approximately 14% is lost as tailings, 12% is lost as ferrochromium slag, and 74% is sent to fabrication. Of the latter amount, 90% is for metallurgical uses, while 7% and 3% are used for chemicals and refractories, respectively. Prompt and old scrap feeds the stainless steel industry about 1100 Gg of chromium per year. It is highly significant that the rate of chromium loss in downgraded scrap equals the amount in old scrap that is recycled for its chromium content. The net addition to in-use stock equals 78% of the chromium entering use. Of all flows in all forms of chromium entering WM&R, over one-half is recycled as old scrap. The four flows VOL. 40, NO. 22, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
7065
tions of chromium emissions were conducted. Air emissions from coal combustion were 1.7-11 Gg Cr/yr in electricity production and 1.1-7.6 Gg Cr/yr in industrial uses. Oil combustion led to minor releases for electricity production and up to 3.0 Gg Cr/yr in industrial applications. Total fossil fuel atmospheric losses are between 3.4 and 22 Gg Cr/yr. Losses due to bottom and fly ash are much more significant: these range from 160 to 470 Gg Cr/yr, and could equal up to 20% of the previously characterized losses. Consistent with previous approaches for zinc (5) and silver (6), and utilizing the variables defined in Figure 1a, the following indicators were calculated for chromium. Utilization efficiency (Ψ): Manufacturing output less waste as a fraction of manufacturing output. FIGURE 4. Chromium entering use versus chromium entering inuse stock on a country and regional basis. All values are given in Gg Cr/yr. of chromium losses are all nearly equal in magnitude: 28% tailings, 24% ferrochromium slag, 25% industrial downgraded scrap, and 22% discard management releases to landfills and the environment. While we have captured the major flows of anthropogenically mobilized chromium originating in chromite ore, we have not included the mobilization due to trace contaminants in fossil fuels. Using the methodology and emission factors set out by Nriagu and Pacyna (48), coupled with energy statistics from the International Energy Agency (49), estima-
Ψ ) (Mo - MD - Qs,M)/Mo ) 0.64 Accumulation ratio (R): Addition of in-use stock as a fraction of chromium entering use.
R ) S/Ui ) 0.78 Recycling ratio (F): Old scrap as a fraction of discard management input.
F ) Ds/(MD + Uo) ) 0.54 Analysis. Figure 4 plots the net addition to stock as a function of the chromium entering use. The global average rate of 78% is mirrored on the country level. It is important
FIGURE 5. Regional flows of chromium into (a) use; (b) discard management. The relative areas of the circle represent the magnitude of the total flows. CIS ) Commonwealth of Independent States; BI ) Buildings & Infrastructure; TR ) Transportation; IM ) Industrial Machinery; HAE ) Household Appliances & Electronics; MG ) Metal Goods and Other Uses; C&D ) construction and demolition debris; ELV ) end-of-life vehicles; MSW ) municipal solid waste; S/SS ) sewage/sewage sludge; WEEE ) waste from electronics and electrical equipment; IW ) industrial waste; HW ) hazardous waste; “Other” includes industrial capital equipment and non-automobile transportation discards. 7066
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 22, 2006
FIGURE 6. International trade of chromium: (a) Net trade of chromium in ore, concentrate, ferrochromium, semi-finished products, finished products, and scrap; United States’s trade of chromium in (b) ore and concentrate; (c) ferrochromium; (d) stainless steel (semi-finished); (e) finished products; (f) the total of these four groups. All values are given in Gg Cr/yr. to note that these variables are not completely independent: one of the eight waste streams (“other”) is a function of the flow into use. (If the dependent waste stream is removed, the correlation between chromium entering use and net addition to stock is just as strong, with an R2 value of 0.9926.) This correlation helps illustrate the cross-cultural phenomenon that chromium products have not reached saturation with consumers, and new products are being added everywhere to in-use stock. Although Figure 4 shows that the net addition to stock is highly dependent on the flow of finished products into use, a country’s GDP is not an ideal indicator of chromium use (see details in the Supporting Information). Not only is chromium used in a wide variety of products, but regions utilize these products at widely disparate rates and patterns. Figure 5a shows the chromium entering use regionally, subdivided by end use categories. (The area of the circle in the figure is proportional to the magnitude of the flow entering use in this region.) Europe, Asia, and North America dominate, with relatively comparable flow magnitudes, but quite different proportions of end use categories: Europe and Asia use higher proportions of chromium for buildings and infrastructure and for industrial machinery than does North America, which has a higher share in transportation. Figure 5b shows the results of the discard stream analysis, utilizing a similar display approach. (Because old scrap is
drawn from these flows, the total of discards leaving use, by itself, is not an appropriate benchmark for environmental performance.) Europe, Asia, and North America dominate the global discard flows, and again we see significant differences in which subflows dominate. Globally, 36% of the total flow comes from IW, 23% from “other discards”, 14% from ELV, 12% from MSW, 7% from C&D, 6% from HW, and less than 1% from S/SS. Africa and Asia have a much larger share of IW than the global average, likely due to stainless steel manufacturing. Europe and Oceania have larger shares of “other discards”, most notably retired stainless steel industrial equipment. The share of ELV discards in North America is twice that of the global average, likely due in part to the high car ownership rate in this region. With large production volumes of chromium chemicals, CIS had a large share of its discards in the form of HW. The regional disparities in the chromium cycle extend to extraction as well. Chromite ore deposits are located heterogeneously around the world and several nations have emerged as major suppliers. Nearly 90% of the chromite ore that is mined (measured as contained chromium) comes from five countries: South Africa, Kazakhstan, India, Zimbabwe, and Turkey. The net importing countries are a broader group. The top seven make up two-thirds of the total net import flow of chromium in all forms and are, in descending order, the United States, Japan, China, France, Germany, VOL. 40, NO. 22, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
7067
United Kingdom, and Italy. Figure 6a shows the net trade of chromium for countries in the STAF country selection. Although historic trade relations between the exporting and importing countries have proven to be stable, political unrest or other episodic incidents in South Africa or other major producers could quickly lead to a global price spikes or shortages. By fully understanding how each nation utilizes metals, and in what forms they import and export them, the impacts of severe trade disruptions can be anticipated. An example is shown in Figure 6b-f for the United States. Each subfigure shows the flows of chromium between the U.S. and its major trade partners in one of its forms: ore and concentrate, ferrochromium, stainless steel (semi-finished), and finished products. These diagrams show the predominant dependence of the U.S. on chromium supply in all forms: ore, concentrate, and ferrochromium from Africa, ferrochromium from CIS, stainless steel from Europe, and finished products from Asia. Environmental Releases. The major anthropogenic losses of chromium have been quantified as part of this work. Inferences can be made about their fate based on the source of their release. There are several competing environmental factors that dictate the conversion between trivalent and hexavalent chromium, including concentrations of chromium species and oxidizing or reducing agents, the electrochemical potentials of the redox reactions, complexing agents, ambient temperature, light, sorbents, acid-base reactions, and precipitation reactions (50). Atmospheric releases, predominately in aerosol form, come from processing [FeCr production (31 Gg Cr/yr), stainless steel production (24-59 Gg Cr/yr), and fossil fuel combustion (3.4-22 Gg Cr/yr)], as well as leather tanning, electroplating, refuse incineration, cement manufacture, and wood preservation, (all of which are small but not quantified). The total atmospheric releases for the three quantified sources ranges between 58 and 112 Gg Cr/yr, which is consistent with an independent estimate of 74 Gg Cr/yr (but made approximately 20 years ago) (51). Roughly one-third of atmospheric releases of chromium are thought to be in its hexavalent form (52). Essentially all atmospheric chromium is removed via wet deposition, with the metal present in roughly equal concentrations of the hexavalent and trivalent forms (52). Chromium losses to ground and surface water are almost always linked to industrial sources, such as mining and milling operations, metal plating, wood treatment, and leather tanning operations (53). These emissions can have environmental consequences, but the overall flow magnitudes are thought not to be large relative to the global chromium system and are not dealt with in our analysis. The chromium present in the form of solid discards may be recycled as old scrap, sent to landfills, or released directly to the environment. Downgraded scrap (670 Gg Cr/yr) still resides in useful products, but only as a trace constituent. It is possible that some modest fraction may be returned to the chromium cycle, as bounded by this study, through the use of carbon steel scrap in new stainless steel production. FeCr slag (640 Gg Cr less 31 released to the atmosphere) may be used as a byproduct or landfilled. IW (450 Gg Cr) includes stainless steel slag and non-recycled metal waste from F&M. This material would typically be landfilled, unless a suitable use (such as slag in asphalt) can be found. ELV (170 Gg Cr), C&D (81 Gg Cr), and other discards (280 Gg Cr) have high recycling rates and predominately well-defined uses for chromium, and thus large portions of these flows do not enter the environment. WEEE (36 Gg Cr) and MSW (150 Gg Cr) are assumed to have lower recycling rates, with a higher percentage of the chromium sent to landfills or informally disposed. Fly and bottom ash (160-470 Gg Cr/yr) may be used as a byproduct (e.g., in concrete) or landfilled. 7068
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 22, 2006
Discussion When considering how humans use metals, much of the focus has traditionally centered on mining and refining operations. Once the metal enters a finished product, tracking of the metal as such is virtually nonexistent; there are many products, they are widely dispersed, and the metal content varies widely. This study has quantified these flows for chromium and amassed them into cohesive cycles. The benefits of doing so are numerous, including: identifying leaks in the system where material is being dissipated indicating opportunities for mitigating the effects of global or local scarcity determining human health and ecosystem concerns linking material flows to energy and water use understanding demand and the services provided Figure 5b shows the varying importance of the chromiumcontaining waste streams in each region. Armed with this information, one can now target waste reduction on a regionspecific basis. The format of this study also allows for the mitigation of waste tailored to the spatial scale at which it must occur. For example, dissipation during use is a spatially scattered loss of chromium, while tailings losses are localized at mine sites. By examining not only the trade flows of natural materials, but also of semi-finished and finished products, the true net import reliance is better understood. Chromium has been deemed a critical and strategic metal by many governments because of its uses in military equipment and aviation. MFA provides quantitative and qualitative information relevant to this vulnerability, and would serve as an excellent resource in advising stockpile sales or buildup. Trade flow quantification identifies the final user and producer; countries and regions; in many cases they are not the same. By importing raw materials from developing nations, the developed world has effectively outsourced a considerable amount of the environmental impact associated with these production processes. The data sets generated by this study would thus prove quite valuable in calculating the true costs of material use from cradle to grave. Along these lines, through the quantification of releases to the environment, the health and ecosystem effects of chromium releases can be better understood. A full epidemiological study would need to be conducted, but understanding the bulk releases to the environment can greatly inform such work. The mobilization and manipulation of metals consume enormous amounts of energy and water. Effective energy and water policies should include a thorough understanding of the entire life cycles of consumer goods, including the manipulation of metal in all its forms. With this perspective, the true energy and water costs of a product can be better understood.
Acknowledgments This material is based upon work supported by the National Science Foundation under grant nos. BES-9818788 and BES0329470. We thank the advisory board for the Stocks and Flows Project, Barbara Reck for her work on characterizing waste streams, Tao Wang and Barbara Reck for their work on stainless steel, and the United States Geological Survey Mineral Commodity Specialists, especially John Papp.
Supporting Information Available Additional details on methodology, results, country-level cycles, and analysis are available. This material is available free of charge via the Internet at http://pubs.acs.org.
Literature Cited (1) Corrosion Costs and Preventive Strategies in the United States, FHWA-RD-01-157; Federal Highway Administration: Washington, DC, 2002.
(2) International Agency for Research on Cancer, Monographs on the Evaluation of Carcinogenic Risks to Humans, Chromium and Chromium Compounds; IARC: Paris, 1990. (3) Harper, E. M.; Johnson, J.; Graedel, T. E. Making metals count: Applications of material flow analysis. Environ. Eng. Sci. 2006, 23(3), 2006, 493-506. (4) Graedel, T. E.; Van Beers, D.; Bertram, M.; Fuse, K.; Gordon, R. B.; Gritsinin, A.; Kapur, A.; Klee, R. J.; Lifset, R. J.; Memon, L.; Rechberger, H.; Spatari, S.; Vexler, D. Multilevel cycle of anthropogenic copper. Environ. Sci. Technol. 2004, 38, 12421252. (5) Graedel, T. E.; van Beers, D.; Bertram, M.; Fuse, K.; Gordon, R. B.; Gritsinin, A.; Harper, E. M.; Kapur, A.; Klee, R. J.; Lifset, R.; Memon, L.; Spatari, S. The multilevel cycle of anthropogenic zinc. J. Ind. Ecol. 2005, 9 (3), 67-90. (6) Johnson, J.; Jirikowic, J.; Bertram, M.; Beers, D. v.; Gordon, R. B.; Henderson, K.; Klee, R. J.; Lanzano, T.; Lifset, R.; Oetjen, L.; Graedel, T. E. Contemporary anthropogenic silver cycle: A multilevel analysis. Environ. Sci. Technol. 2005, 39, 4655-4665. (7) Bergba¨ck, B.; Anderberg, S.; Lohm, U. A reconstruction of emission, flow, and accumulation of chromium in Sweden 1920-1980. Water, Air, Soil Pollut. 1989, 48, 391-407. (8) Anderberg, S.; Bergba¨ck, B.; Lohm, U. Flow and distribution of chromium in the Swedish environment: A new approach to studying environmental pollution Ambio 1989, 18, 216-220. (9) Helweg, C.; Rasmussen, J. O., Mass Flow Analysis of Chromium and Chromium Compounds; Danish Environmental Protection Agency: Strandgade, Denmark, 2004. (10) Timmermans, V.; Holderbeke, M. V. Practical experiences on applying substance flow analysis in Flanders: bookkeeping and static modelling of chromium. J. Cleaner Prod. 2004, 12, 935945. (11) Gabler, R. C. A Chromium Consumption and Recycling Flow Model; United States Geological Survey: Denver, CO, 1994. (12) Papp, J. Chromium Recycling in the United States in 1998; circular 1196-C; U.S. Geological Survey: Denver, CO, 2001. (13) Papp, J. Chromium Life Cycle Study; information circular 9411; U.S. Bureau of Mines; U.S. Government Printing Office: Washington, DC, 1994. (14) Riekkola-Vanhanen, M., Finnish Expert Report on Best Available Techniques in Ferrochromium Production; The Finnish Environment Institute: Helsinki, Finland, 1999. (15) ICDA Statistical Bulletin; International Chromium Development Association: Paris, 2003. (16) Papp, J. U.S. Geological Survey Minerals Yearbook: Chromium; USGS: Denver, CO, 2002. (17) ICDA, International Chromium Development Association; http://www.icdachromium.com/, last accessed January, 2006. (18) The Economics of Chromium; Roskill Information Services Limited: London, England, 1993. (19) Pariser, H. End Use of Nickel 1980-2001; Alloy Metals and Steel Market Research: Xanten, Germany, 2002. (20) Papp, J.; Lipin, B. Chromium; open file report 01-381; U.S. Department of the Interior, U.S. Geological Survey: Denver, CO, 2001. (21) The Economics of Chromium; Roskill Information Services Limited: London, England, 2000. (22) Lofthouse, H. The World of Chromium Metal; London & Scnadinavian Metallurgical Co Limited, Merallurg Performance Materials Group: Rotherham, UK, 2001. (23) Pariser, H. Chrome End Use, Alloy Metals & Steel; ICDA: Paris, 2002. (24) World Stainless Steel Statistics; INCO: Toronto, 2004. (25) Odnevall Wallinder, I.; Lu, J.; Bertling, S.; Leygraf, C. Release rates of chrmoium and nickel from 304 and 316 stainless steel during urban atmospheric exposure - a combined field and laboratory study. Corros. Sci. 2002, 44, 2303-2319. (26) Hansen, M.; Rydin, S.; Menne, T.; Duus Johansen, J. Quantitative aspects of contact allergy to chromium and exposure to chrometanned leather. Contact Dermatitis 2002, 47, 127-134. (27) Adach, K.; Tainosho, Y. Characterization of heavy metal particles embedded in tire dust. Environ. Int. 2004, 30, 1009-1017. (28) Stook, K.; Tolaymat, T.; Ward, M.; Dubey, B.; Townsend, T.; Solo-Gabriele, H.; Bitton, G. Relative leaching and aquatic toxicity of pressure-treated wood products using batch leaching tests. Environ. Sci. Technol. 2005, 39, 155-163. (29) Siemers, W.; Vest, H. Environmental Handbook: Environmentally sound electroscrap disposal and recycling; Deutsche Gesellschaft fur Technische Zusammenarbeit: Eschborn, Germany, 1999. (30) Weigand, H.; Fripan, J.; Przybilla, I.; Marb, C. Composition and contaminant loads of household waste in Bavaria, Germany:
Investigating effects of settlement structure and waste management practice. In Proceedings of the Ninth International Waste Management and Landfill Symposium, Sardinia, Italy, October 6-10, 2003 (31) European Topic Centre on Resource and Waste Management, Waste Base, http://waste.eionet.eu.int/wastebase, last accessed November, 2005. (32) What a Waste: Solid Waste Management in Asia; World Bank, Urban Development Sector Unit: Washington, DC, 1999. (33) Henderson, K. The contemporary silver cycle for CIS countries: Using industrial ecology to evaluate silver flows J Young Invest 2003, 9; http://www.jyi.org/volumes/volume9/issue1/articles/ henderson.html. (34) Wagner, T. Urban Cleaning, Solid Waste Management in Brazil; Stat-USA: Washington, DC, 2002. (35) Stat Canada, Waste Management industry: government and business sectors; http://www.statcan.ca/Daily/English/020425/ d020425b.htm, last accessed November, 2005. (36) United States Environmental Protection Agency. Municipal Solid Waste in the United States: 2000 Facts and Figures; USEPA: Washington, DC, 2002. (37) United Nations Statistical Yearbook; UN: New York, 2004. (38) Staudinger, J.; Keoleian, G. A. Management of End-of Life Vehicles (ELVs) in the US; report No. CSS01-01; University of Michigan: Ann Arbor, MI, 2001. (39) European Commission. Construction and demolition waste management practices, and their economic impacts; Report to DGXI; EU: Brussels, 1999. (40) Characterization of Building-Related Construction and Demolition Debris in the United States; Franklin Associates: Prairie Village, KS, 1998. (41) van Beers, D.; Bertram, M.; Fuse, K.; Spatari, S.; Graedel, T. E. The contemporary Oceania zinc cycle: one-year stocks and flows. J. Material Cycles Waste Manage. 2004, 6, 125-141. (42) van Beers, D.; Bertram, M.; Fuse, K.; Spatari, S.; Graedel, T. E. The contemporary African zinc cycle: One-year stocks and flows. Afr. J. Environ. Assess. Manage. 2003, 7, 21-40. (43) Brunner, P. H.; Stampfli, D. The material balances of a construction waste sorting plant. Waste Manage. Res. 1993, 11, 27-48. (44) Commission of the European Communities. Report from the Commission to the Council and the European Parliament of the implementation of community waste legislation; Directive 86/ 278/EEC on sewage sludge; 2000. (45) Environmental Protection Agency. Biosolids Generation, Use, and Disposal in The United States; EPA: Washington, DC, 1999. (46) World Health Organization. Global Water Supply and Sanitation Assessment, 2000 report; WHO: Geneva, 2000. (47) Klee, R.; Material Flow Analysis of Industrial Systems in Antarctica, Doctoral Disseration, Yale University, 2005. (48) Nriagu, J. O.; Pacyna, J. M. Quantitative assessment of worldwide contamination of air, water and soils by trace metals. Nature 1988, 333, 134-139. (49) International Energy Agency. Energy Balances of Non-OECD Countries: 2001-2002; IEA: Paris, 2004. (50) Kimbrough, D.; Cohen, Y.; Winer, A.; Creelman, L.; Mabuni, C. A. Critical Assessment of Chromium in the Environment. Crit. Rev. Environ. Sci. Technol. 1999, 29, 1-46. (51) Pacyna, J. M.; Nriagu, J. O. Atmospheric emissions of chromium from natural and anthropogenic sources, In Chromium in the Natural and Human Environments; Nriagu, J. O., Nieboer, E., Eds.; Wiley and Sons: New York, 1988; pp 105-123. (52) Kieber, R.; Eilley, J.; Zvalaren, S. Chromium speciation in rainwater: temporal variability and atmspheric deposition. Environ. Sci. Technol. 2002, 36, 5321-5327. (53) Calder, L., Chromium contamination of groundwater, In Chromium in the Natural and Human Environments; Nriagu, J. O., Nieboer, E., Eds.; Wiley and Sons: New York, 1988.
Received for review January 10, 2006. Revised manuscript received July 26, 2006. Accepted July 28, 2006. ES060061I VOL. 40, NO. 22, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
7069
