Environ. Sci. Technol. 2008, 42, 3835–3842
Illuminating Tungsten’s Life Cycle in the United States: 1975-2000 E . M . H A R P E R * ,†,‡ A N D T . E . G R A E D E L † Center for Industrial Ecology, School of Forestry and Environmental Studies, 205 Prospect Street, Yale University, New Haven, Connecticut 06511, and Department of Chemical Engineering, Mason Laboratory, Yale University, P.O. Box 208267, New Haven, Connecticut 06520-8267
Received November 13, 2007. Revised manuscript received December 26, 2007. Accepted January 14, 2008.
Tungsten’s unique properties, such as its high density and melting point, are manifest in a variety of technological applications, and attention has recently centered upon the health and environmental effects of tungsten. Utilizing material flow analysis, the amounts of tungsten produced, fabricated and manufactured, entering use, and entering waste management in the United States from 1975-2000 were determined, with the inclusion of net trade, and with attention paid to the way that tungsten partitions into in-use reservoirs and the quantities of tungsten that enter the environment as a result of anthropogenic processes. The results from two approaches are presented: the first, and typical, approach for historical material flow analysis studies, whereby tungsten use in products over time was approximated using end-use sectors used by the United States Bureau of Mines and the United States Geological Survey (the “end-use sector model”) and a second, more detailed approach whereby the historical pattern of end-use products was determined (the “finished product model”). These models represent the first life cycle of tungsten that takes into account both finished product trade and varying residence times for products entering use. For the cumulative time period of 1975-2000, approximately 60% of the supply to fabrication and manufacturing processes in the United States consisted of imports. The tungsten embodied in products used in applications other than transportation is discarded (i.e., enters the waste stream, which encompasses both material that is eventually landfilled and that is eventually recycled) in the same year it enters service. Of its items entering end-of-life for the cumulative time period, 50-70% of tungsten is estimated to be landfilled. In both models, the largest flows of tungsten to the environment were estimated to originate from post-consumer discards. Approximately 8-17 times as much tungsten entered landfills than was estimated to be generated as waste from fabrication and manufacturing processes; for tailings, these ratios are even greater. Tungsten is an example of a critical mineral resource where the majority of the United States’ supply comes from imports, it has vital uses in manufacturing, and it has a high rate of loss at end-of-life.
* Corresponding author; fax: + 1-203-432-5556;
[email protected]. † Center for Industrial Ecology. ‡ Department of Chemical Engineering. 10.1021/es070646s CCC: $40.75
Published on Web 04/18/2008
2008 American Chemical Society
e-mail:
Introduction Light bulbs, drill bits, and aircraft counterweights play important roles in many people’s lives and, perhaps less obviously, they all utilize an interesting metal: tungsten. Tungsten’s unique properties, including the highest melting point of all metals (3422 °C) and a high density (19.3 g/cm3 at 20 °C), carve its niche in technology, and its diverse uses include armaments, radiation shields in medical applications, catalysts in petroleum refining, heating elements, and fireproofing agents. Environmental questions about tungsten came to light in 2002 when tungsten was nominated for study with high priority to the National Toxicology Program for toxicology and carcinogenesis studies by the Centers for Disease Control and Prevention’s National Center for Environmental Health (1), and tungsten has been the subject of environmental research studies (e.g., 2). This nomination was based upon data showing elevated tungsten body burdens in residents of Fallon, Nevada, and the fact that there are limited data available to assess the potential longterm health effects of tungsten exposure (3, 4). Tungsten has been considered for use in recreational and military “green” ammunition, making it even more relevant to study its use and related losses to the environment. Tungsten has a moderate depletion time (defined as reserves divided by current annual use), estimated at approximately 100 years (5). The reservoir for materials used by technological societies has historically been primary or virgin sources (i.e., ore deposits). Alternate reservoirs or secondary sources exist, namely in materials or products in use, stored, discarded, or dissipated over the years by governments, corporations, and individuals. Characterizing these reservoirs is important, both for increased use as a secondary material and in terms of the mounting evidence that resource loss by dissipation or by disposal in landfills can lead to degradation of environmental quality. Quantifying sources of secondary tungsten may also lead to increased recycling, which has the potential to reduce water and energy demand, as well as the current dependence upon virgin stocks. Material flow analysis (MFA) is the methodology most commonly used to study material use within a set of boundaries (i.e., spatial, process, and temporal). Through utilization of MFA, regional technological materials cycles may be established. These cycles typically trace the material through several stages within a defined region, and quantify how a material is used and any resulting environmental emissions. Within this framework, a stock may be defined as an accumulation of the material being analyzed, and a flow indicates the rate of transfer of the material being analyzed. The Stocks and Flows (STAF) project at Yale University is a large-scale project focusing on technological materials cycles for a variety of spatial and temporal scales, and the general framework considers four life stages: production, fabrication and manufacturing (F&M), use, and waste management. The typical approach to historical modeling using MFA is determining use of a material by defining end-use sectors (e.g., transportation); for the purposes of this study, the sectors used are those used by the United States Bureau of Mines (USBM) and the United States Geological Survey (USGS). It is undeniably more desirable to know the exact products that enter use, rather than these broad sectors. If this were achieved, the analysis then more clearly highlights a variety of potentials for environmental improvement, from controlling environmental emissions from product production, to managing products at their end-of-life phase, to demonVOL. 42, NO. 10, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3835
strating potentials for recycling. All this being said, defining and quantifying end-uses is a challenge that has always plagued MFA research. While it is common to find statistics for end-use sectors for various metals (e.g., the percentages of total use used in transportation) or for materials (e.g., the percentage of total use of tungsten attributed to tungsten carbide), it is an enormous challenge to definesparticularly quantitativelysthe percentages of total use of a metal that may be attributed to individual products. In this study, two approaches were taken: the first, and typical approach, whereby tungsten use in products over time was approached using end-use sectors (the “end-use sector model”, or EUSM) and a second, more detailed, approach whereby the historical pattern of end-use products was determined (the “finished product model”, or FPM). Together, the two models produce a historical technological tungsten cycle for the United States that, in turn, serves as a foundation from which to further study related resource availability and environmental issues.
Materials and Methods Introduction. The tungsten cycle is a complex one and, in developing it, scores of tungsten flows were estimated. Many of these (e.g., tailings) were completed using traditional MFA methods. It is these more traditional aspects that are the focus of this section. The development and outcomes of the less traditional aspects of both methodologies are thoroughly treated elsewhere (6). Conceptual details for both models may be found in Figure S1 in the Supporting Information section. Production. The analysis for production was identical for both models (see Figure 1 a) for details). The subprocesses in production are mining tungsten-bearing ore (i.e., scheelite and wolframite) and its subsequent concentration (through milling and beneficiation processes). Tungsten ore and concentrate are traded materials, and may also enter or leave government or industrial stockpiles. Tailings at this stage enter the environment. Recovery rates of 70-90% were cited by Smith (7), and a recovery rate of 90% was used in this study due to the fact that this process is widely known to be very efficient. The USGS Historical Statistics for Tungsten in the United States (8) was used to identify data sources. Mine shipment data were from the USBM for data from 1975 to 1994 and the USGS for data from 1995 to 2000, with 1988 to 1994 data withheld for proprietary reasons (9). Fabrication and Manufacturing (F&M). The subprocesses subsequent to producing concentrate in production include, but are not limited to, a variety of chemical processing steps (detailed in Figure 1 b), as adapted from Smith (7)). Calculating the amount of tungsten used by F&M (commonly termed “apparent consumption”) represents tungsten in intermediate forms (e.g., ammonium paratungstate (APT) and tungsten carbide powder) that, with further processing, becomes finished products (e.g., broaches and light bulbs) which, in turn, either enter use or are exported. Calculating the amount of tungsten used by F&M was identical for both models, and was according to the formula traditionally used by the USBM and USGS (8) as follows: Amount of tungsten used by F&M ) primary (mine) shipments + secondary production + total imports - total exports + total stock change + total government shipments. Secondary production, included in this calculation, is assumed by the USBM and USGS to be equivalent to scrap consumption. Using this formula, the amount of tungsten used by F&M for each year in the period of 1975-2000 was calculated, and the aggregate value for the entire time period was compared to the same quantity for this period calculated from values published by the USBM and USGS, taking into account both trade and losses to the environment. The difference between these two values is the “phantom flow”, which accounts for discrepancies such as withheld data. Data 3836
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 10, 2008
sources used in this calculation are detailed in Table S1 in the Supporting Information. Data were recorded in as disaggregated a form as possible, but were then aggregated to be sure that they were equivalent to those values published in USGS Historical Statistics for Tungsten in the United States (8). Once confirmed, the aggregated data then constituted the desired categories for this study. For trade, the categories were ore and concentrate, ammonium tungstate (mainly comprised of APT), tungsten carbide powder, and other (e.g., tungsten and tungsten alloy powders, unwrought tungsten, and wrought tungsten other than wire). The stocks held by producers, and industrial consumers and/or dealers, were aggregated into concentrate, APT, tungsten carbide powder, and other (e.g., tungsten powders and chemicals). The most common starting process in F&M is the conversion of APT from concentrate, whereby a minimum of 96% recovery is achieved (7). Losses incurred in converting APT to tungsten oxides and subsequently to metal powder are less than 1%, and conversions of metal powder to tungsten carbide powder may result in a loss of approximately 1%. The processing of tungsten beyond milling and beneficiation is highly efficient due to the fact that most tungsten fabrication is done using well-controlled powder metallurgy processing procedures. For other subprocesses, the intrinsic value of tungsten materials dictates that manufacturing processes be conducted efficiently so that most of the scrap that is generated is recovered. Higher losses (i.e., up to 15%) (7) have been reported in the production of tool steels, superalloys, and other steels and nonferrous alloys when the source of tungsten is low-grade ferrotungsten. Because the amount of tungsten used in alloys is much less that that used in other products, this high loss rate was not considered. Proportionate flows in the different F&M processes and the level of data aggregation in the analysis led to a recovery rate of 96% being assigned to the amount of tungsten used by F&M for the entire period. The subprocesses that take place in F&M after fabrication are concerned primarily with production of finished products. Two models, the EUSM and the FPM, were used to characterize these processes. A detailed description of both models may be found in elsewhere (6), and details appropriate to this paper are found in Table 1. The primary difference between the EUSM and the FPM is the way in which domestic manufacturing of finished products is estimated. In the EUSM, what enters use is divided into end-use sectors (e.g., transportation), whereas in the FPM what enters use is a variety of individual finished products (e.g., reamers). Import and export of finished products was identical for both models (for details, see Harper (6)), and products that were considered are detailed in Table 1. Once import and export product quantities were obtained on a mass basis, tungsten flows were estimated based upon data from a variety of sources and informed estimates (6). Conceptual details for translating monetary bulk flow statistics to tungsten flows may be found in Figure S2 in the Supporting Information section. Estimating domestic finished product manufacture for the EUSM began with using USBM and USGS data for the distribution of tungsten among end-use sectors (10). To determine what enters use, imported products were added and exported products were subtracted. Estimating domestic finished product manufacture for the FPM was based upon data from the United States Census Bureau (USCB). Most of these data are available every five years. As such, 1977 data was used to estimate product manufacture for 1975-1979, 1982 for 1980-1984, 1987 for 1985-1989, 1992 for 1990-1994, and 1997 for 1995-2000. In the cases of light bulbs, motor vehicles, aircraft, and aircraft engines, annual data were available. The data are presented in a highly disaggregated
FIGURE 1. Detailed subprocesses within the life cycle stages of tungsten in the United States for (a) production; (b) F&M (adapted from Smith (7)); and (c) use and waste management. and comprehensive way, and are generally given in terms of monetary value. Through using the corresponding trade statistics, these monetary values were converted into physical flows (6). The products included in the analysis are listed in Table 1. The bulk product flows were converted into tungsten flows using a variety of informed estimates (6). Use and Waste Management. The length of time that the material stayed in use (Figure 1 c) for each sector (in the EUSM), and for each product (in the FPM) are summarized in Table 1. Products that have reached their end-of-life phase typically enter discard streams (e.g., municipal solid waste stream), with a portion of the discards being recycled. The
USGS developed a list of product lifetimes (11) with the input of mineral commodity specialists and manufacturers. These lifetime data were supplemented by data from LightBulbs Direct (12) the National Lighting Product Information Program (13), and Lassner and Schubert (14) for light bulbs. Recycling rates were assigned to specific sectors and products (15). All of the recycling rates provided by Sandvik (15) were given as ranges for cemented carbide tools; these were extended to tools made from all tungsten-bearing materials (e.g, various types of steel). Tungsten waste and scrap import and export were included for Waste Management, with data from the USBM and USGS (9). For exports VOL. 42, NO. 10, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3837
TABLE 1. Details Regarding the Products Considered for Each Analysis, And Their Associated Lifetimes and Recycling Ratesa
analysis type end-use sector model finished product model and finished product trade
end-use sector model
end-use sector metalworking machinery and equipment metalworking machinery and equipment
mining and construction machinery and equipment
end-use sector model, finished product model, and finished product trade
lamps and lighting
end-use sector model, finished product model, and finished product trade
transportation
finished product model and finished product trade end-use sector model
estimated lifetime in both time-residence models (years)
productb
recycling rates (% of discards from use recycled)
assumed to be cutting tools and dies
less than 1c
50(15)
broaches
less than 1c
50(15)
reamers drawing and extruding dies, and pressing, stamping, and punching tools (for metal cutting; e.g., stamping dies) cutting tool inserts end milling cutters other milling cutters threading taps, dies and chasers band saw blades circular saw blades hacksaw blades drills
Less than 1c
50(15)
less than 1c
75(15)
less than 1c less than 1c less than 1c
50(15) 50(15) 50(15)
less than 1c
50(15)
1c 1c 1c 1c
50(15) 50(15) 50(15) 50(15)
less than 1c
15(15)
assumed to be blanks, tips, sticks, and plates used as inserts in tools photographic, large, and miniature incandescent bulbs, and electrical discharge bulbs
less less less less
than than than than
• To be conservative, the lowest values in the ranges of rated life for incandescent and fluorescent lighting were used to calculate lamp lifetimes in years. It was assumed that each lamp is on for 6 hours per day(16). • The metal halide lamp was used as a proxy for electrical discharge lamps. It was assumed that each lamp is on for 11 hours per day(13). • The lifetimes used in the analysis were 0.5 years for incandescent lamps, 3.7 years for fluorescent lamps, 1.4 years for compact fluorescent lamps (this category is only present in the end-use sector model), and 2.5 for other discharge lamps. Because the other lifetimes were so short, only the fluorescent lamps and other discharge lamps were modeled with the time-residence model. The others were assumed to enter and leave use in a one-year interval.
negligible(7)
motor vehicles
10(11)
negligible
aircraft
5(11) (given for engine parts; engine portion in the finished product model) 25(11) (given as the average aircraft lifetime, and so used for the counterweight portion in the finished product model; for tungsten in the entire aircraft in the end-use sector model)
90 (for both engines and counterweight portions)
aircraft engines
5(11) (given for engine parts)
90
10(11)
negligible
electrical and electronic machinery and equipmentd
a The values are global averages and are given for the following categories: cutting tools, mining tools, and large tools. For the finished product model and finished product trade analyses, most categories are aggregates of many products. For example, in both of the aforementioned analyses, approximately 33 types of light bulbs and three types of cutting tool inserts were included. c This estimate was used after surveying the list of lifetimes provided by the USGS (11), in which the product lifetimes of drill tool bits and metal-cutting inserts/tool bits were all estimated at or under one year. Because these product lifetimes were so short, the time-residence models were not used to estimate the length of time they stayed in use. Instead, they were assumed to enter and leave use in a one-year interval. This approach may slightly underestimate flows into use and to recycling. d For tractability, these products were not included in the finished product model or in finished product trade. b
3838
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 10, 2008
FIGURE 2. The 1975-2000 cumulative United States tungsten cycle, where EUSM is end-use sector model, FPM is finished product model, and APT is ammonium paratungstate. Data sources for this analysis are presented throughout the text. Mine shipments were withheld by the USBM and USGS for 1988-1994. The accumulation of stock for this entire time period was estimated at 22 Gg W for the EUSM and at 83 Gg W for the FPM. for the entire time period, unwrought tungsten was reported in an aggregate category with scrap. The 2003 data, in which these two items were reported separately, were used to disaggregate export data for the whole time period. From 1975 to 1982, for this category, gross weights were provided, rather than contained tungsten, so a factor of 80% was used to estimate contained tungsten. The amounts of obsolete scrap generated each year were determined using recycling rates for the various products or sectors included in the models.
Results
FIGURE 3. 1975 and 2000 United States tungsten cycles, where EUSM is end-use sector model, FPM is finished product model, and APT is ammonium paratungstate. Data sources for this analysis are presented throughout the text.
The cumulative tungsten cycle for 1975-2000 for the United States is shown in Figure 2, and cycles for 1975 and 2000 appear in Figure 3. Cycles for each year were generated, and all 26 cycles are presented in Figure S3 in the Supporting Information. Data for mine shipments in the United States were available from 1975 to 1987 and from 1995 to 2000. In 1975, the United States’ mine production was 2.5 gigagram (Gg) tungsten (W) (9) a value that satisfied approximately 40% of the amount of tungsten used by F&M. By 1987, these values had dropped to 0.03 Gg W (9), satisfying 0.3% of this value. For the years 1995-2000, no mine production was reported in the United States (9). With a recovery rate of 90%, and mining tapering-off in the United States, tungsten emissions from tailings for the entire time period were estimated at 3 Gg W. The EUSM treats historical domestic manufacture using a sector approach, with data from the USBM and USGS (10). Traded individual finished products were grouped into sectors, and net import was added to or net export was subtracted from domestic manufacture, in order to estimate the quantity that enters use. Accordingly, the profiles for domestic manufacture, use, and discards to waste management were on a sector basis. The main focus of the FPM was to estimate the domestic manufacture of individual finished products so that more accurate, detailed profiles could be obtained for tungsten entering use and waste management. In order to see how the FPM analysis compared to the EUSM
data, domestic manufacture of individual products was grouped into sectors, and may be seen in Figure 4 (note that, as detailed in the Materials and Methods, these data are available every 5 years and were extended in 5 year intervals to the study period). Much more detailed profiles (for domestic manufacture, and tungsten entering use and waste management) were generated using this model. The detailed manufacturing profile for the FPM is shown in Figure 4. The analysis employed for every type of trade, including finished product trade, was identical for both models, and is described in detail elsewhere (6). The analysis relied heavily upon USBM and USGS statistics for semifinished products (see Table S1 in the Supporting Information for detailed sources)anduponUSCBstatisticsforfinishedproducts(17–19). Techniques were developed to treat the varying levels of aggregation and a lack of mass values (i.e., economic values only were available for many data) for finished product trade (6). Figure 5 highlights various flows over time. Figure 5 a) shows the amount of tungsten extracted (i.e., mine shipments plus estimated tailings) in the United States over time. For the 1975-2000 cumulative time period, the United States was a net importer of tungsten in every category examined (i.e., ore and concentrate, APT, other, finished products, and scrap), except for tungsten carbide powder. By far, the greatest majority of tungsten imported was in the form of ore and VOL. 42, NO. 10, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3839
FIGURE 4. Estimates for domestic manufacture for the finished product model, with sources for this analysis presented throughout the text. The differences between 100% and the total sector uses in the figure represent amounts in the electrical and electronic machinery and equipment sector and in other sectors that could not be captured using the available statistics. concentrate. Trade of finished products over time is shown in Figure 5 b). The United States was a net exporter of tungsten in finished products in early years and, in later years, imported progressively larger quantities of tungsten in finished products. The United States was a net importer of tungsten every year of the study, and imports constituted a significant portion of its tungsten supply. The following calculation was completed for each study year, and results (in terms of percentages) are in Figure 5 c): tungsten supply from net trade ) net trade ÷ (mine shipments + net trade + stock entering F&M (including government shipments) + scrap entering F&M). Because the largest amounts of tungsten are used in products with short lifetimes (the only products with relatively long lifetimes are aircraft and motor vehicles), the accumulation of stock in use was small. The profile for the FPM is shown with the individual products aggregated into sectors in order to allow comparison between the models. The largest emissions to the environment are in the form of consumer discards, with 156 Gg W landfilled in the EUSM and 78 Gg W landfilled in the FPM, for the cumulative time period. These values are significantly larger than the 9 Gg W of environmental emissions estimated to result from F&M processes. The total amount of mine shipments in the United States from 1975 to 2000 (excluding withheld data) is 26 Gg W (data from the USBM and USGS (9)), which is 3 times less than the estimate landfilled for the FPM and 6 times less than that estimated for the EUSM. Amounts landfilled, as a function of total discards, are displayed in Figure 5 d). Comparisons between the amounts estimated to be landfilled and net imports for both models may be made (see Table S2 in the Supporting Information), which is especially interesting in years in which more tungsten is discarded than imported.
Discussion In this paper, we have presented a characterization of the way that tungsten has been used in the United States over 3840
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 10, 2008
a 25-year period (1975-2000). The two models used, the EUSM and the FPM, rely upon a very large quantity of data from a variety of sources. The authors consider the detailed treatment of finished product trade and domestic production an important feature of this research. Quantifying trade of finished products is a notoriously challenging task but, only by doing so, can the amount of tungsten entering use be accurately assessed. For example, if just domestic production of finished products is considered, and a large amount of a certain product containing tungsten is exported, one would overestimate the amount of tungsten entering use. The detailed historical account of trade is a unique aspect of this study. One of the questions that drove this work is this: Is the traditional method of using a sector approach for historical MFA studies “good enough”? The uncertainties posed by the sector approach are multifold. They include assigning lifetimes to an entire sector (to model how long the products stay in use), as well as an unclear understanding of exactly what products constitute a sector. In trying to answer this question, trade and domestic manufacturing statistics were examined at a very detailed level. A number of assumptions were made throughout the study, as detailed in the text, particularly when approaching varying aggregation throughout the years, translating monetary values to physical mass quantities, and deriving tungsten flows from bulk product flows. In the end, when the results for domestic production, or manufacture, in the FPM were aggregated into sectors, they compared reasonably well with the data from the USBM and USGS used in the EUSM as follows: • Metalworking machinery and equipment: This sector ranged from 23 to 62% of the amount of tungsten used by F&M in the FPM and from 39 to 58% in the EUSM. • Construction and mining machinery and equipment: This sector ranged from 9 to 12% of the amount of tungsten used by F&M in the FPM and from 18 to 37% in the EUSM.
FIGURE 5. Patterns of tungsten extraction, net finished product trade, and discards in the United States over time. (a) Extraction (identical for both models, with data from the United States Bureau of Mines and the United States Geological Survey, as detailed in the text), (b) Net finished product trade (identical for both models, with analysis details in the text; negative numbers designate net export, and positive numbers designate net import), (c) Percentage of supply attributed to net imports, and (d) Landfill discard profiles for the EUSM and the FPM; in each graph, the right y-axis refers to the shaded gray area that represents the cumulative tungsten landfilled in all products, whereas the left y-axis refers to the lines that represent tungsten discards on a yearly basis. The key designating the colored lines is intended to show summed values, in most cases (for example, yellow indicates the sum of tungsten discarded from the transportation, metalworking machinery and equipment, and construction and mining machinery and equipment sectors). • Transportation: This sector ranged from 12% to 50% of the amount of tungsten used by F&M in the FPM and from 4 to 11% in the EUSM. • Lighting: In the FPM, this sector was consistently 1% of the amount of tungsten used by F&M, whereas it ranged from 6 to 12% in the EUSM. The major benefit of closely detailing the end-uses is that the products entering Use and, eventually, the discard stream into Waste Management are characterized. Should future need dictate further utilizing discards as secondary resources, having such a model would inform resource management choices.
These models represent the first life cycle of tungsten that takes into account both finished product trade and varying residence times for products entering use. For the cumulative time period of 1975-2000, approximately 60% of the supply to fabrication and manufacturing processes consisted of imports. For each use, except for in the transportation sector, the tungsten embodied in products enters waste streams (a portion of which is recycled) in the same year it enters service. Long-term in-use stocks, thus, is relevant only for products held in the transportation sector. By far the greatest losses in the cycle occur at the post-consumer stage, rather than during production or VOL. 42, NO. 10, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3841
F&M. This work demonstrates that tungsten is a critical mineral resource in which the majority of its supply comes from imports, it has vital uses in manufacturing, and it has a high rate of loss at end-of-life. We anticipate that this study and, in particular, the methodology for the FPM, will provide a foundation for other materials to be studied in a more comprehensive way, particularly materials for which detailed, quantitative statistics regarding their end-uses (either by sectors or by products comprising these sectors) are not readily available. Such studies would likely focus on materials of potential environmental concern, materials whose environmental emissions are expected to increase, and materials whose processing lends itself to high levels of waste, energy, or water usage. As with any historical study, this one describes where the United States has been. It should most appropriately be viewed as an important step in illuminating the potential of where the country could go, particularly in terms of environmental emissions and the sustainable use of tungsten.
Acknowledgments We acknowledge support from the National Science Foundation under grants BES-9818788 and BES-0329470 and additional support from the National Science Foundation, P.E.O. International, and the National Italian American Foundation. We thank members of the Stocks and Flows Project at Yale University for interesting discussions and perspectives, Ernst Worrell for valuable discussions about methodological details, and the United States Geological Survey Mineral Commodity Specialists, particularly Kim Shedd, for invaluable data and input.
Supporting Information Available Annual cycles for the historical period analyzed (i.e., 1975-2000), a table listing data references, two figures intended to illuminate methodological reasoning and details, and a table comparing the amounts of tungsten estimated to be landfilled and those estimated to be imported. This material is available free of charge via the Internet at http:// pubs.acs.org.
3842
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 10, 2008
Literature Cited (1) NTP. Nominations to the testing program: nominations reviewed 2003. National Toxicology Program. http://ntp.niehs.nih.gov/ ntpweb/index.cfm?objectid)25BEBA08-BDB7-CEBAFC56EAD78615ADCF#13 (accessed January, 2006). (2) Koutsospyrosa, A.; Braidab, W.; Christodoulatosb, C.; Dermatasb, D.; Strigulc, N. A review of tungsten: From environmental obscurity to scrutiny. J. Hazard. Mater. 2006, 136, 1–19. (3) CDC. Cancer clusters: Churchill County (Fallon), Nevada exposure assessment, Centers for Disease Control and Prevention. http://www.cdc.gov/nceh/clusters/Fallon/ntp.htm (accessed January, 2006). (4) NIH. Tungsten and selected tungsten compounds: review of toxicological literature, 2003 report prepared for National Institute of Environmental Health Sciences, National Institutes of Health. http://ntp.niehs.nih.gov/ntp/htdocs/Chem_ Background/ExSumPdf/tungsten.pdf (accessed January, 2006). (5) ITIA. Tungsten; International Tungsten Industry Association: Brussels, Belgium, 1997. (6) Harper, E. A product-level approach to historical material flow analysis studies: Tungsten as a case study. J Ind. Ecol. 2008, . in press. (7) Smith, G. R. Materials Flow of Tungsten in the United States, Information Circular 9388; United States Department of the Interior, Bureau of Mines: Washington, DC, 1994. (8) DiFrancesco, C. A.; Shedd, K. B. Tungsten, open-file report 01– 006; United States Geological Survey: Reston, VA, 2004. (9) USBM/USGS. Minerals Yearbook, 1975–2002; United States Bureau of Mines and United States Geological Survey: Reston, VA, 1975-2002. (10) USBM/USGS. Mineral Commodity Summary, 1975-2000; United States Bureau of Mines and United States Geological Survey: Reston, VA, 1975–2000. (11) USGS. Estimates of Lifetimes of Various Products. Personal communication, United States Geological Survey, 2005. (12) LBD. Glossary, LightBulbs Direct. http://www.lightbulbsdirect.com/article_view.asp?ArticleID)7 (accessed May, 2005). (13) NLPIP. Metal Halide Lamps Questions and Answers, National Lighting Product Information Program. http://www.lrc.rpi.edu/ programs/nlpip/lightingAnswers/mwmhl/characteristics4.asp# (accessed May, 2005. (14) Lassner, E.; Schubert, W. D. Tungsten: Properties, Chemistry, Technology of the Element, Alloys, and Chemical Compounds; Kluwer Academic/Plenum Publishers: New York, 1999. (15) Sandvik. Recycling rate estimates (global averages). Personal communication, 2007. (16) DOI. Greening the Department of the Interior, Green Energy and Water Programs, Department of the Interior. http:// www.doi.gov/greening/energy/bulbs.html (accessed May, 2005). (17) USCB. Foreign Trade Statistics (Exports: Document 410); United States Census Bureau: Washington, DC, 1975–1977. (18) USCB. Foreign Trade Statistics (Exports: Document 446); United States Census Bureau: Washington, DC, 1978–2005. (19) USCB. Foreign Trade Statistics (Imports: Document 246); United States Census Bureau: Washington, DC, 1975–2005.
ES070646S
