Contemporary Anthropogenic Silver Cycle: A Multilevel Analysis

Interesting findings include the following: (1) several silver-mining countries export ore ... (2) the level of development for a country, as indicate...
0 downloads 0 Views 524KB Size
Download PDF
Environ. Sci. Technol. 2005, 39, 4655-4665

Contemporary Anthropogenic Silver Cycle: A Multilevel Analysis J E R E M I A H J O H N S O N , †,‡ J U L I E J I R I K O W I C , † M A R L E N B E R T R A M , †,§ D . V A N B E E R S , †,∇ R . B . G O R D O N , †,| K A T H R Y N H E N D E R S O N , †,|,⊥ R . J . K L E E , † T E D L A N Z A N O , †,O R . L I F S E T , † L U C I A O E T J E N , †,# A N D T . E . G R A E D E L * ,† Center for Industrial Ecology, Environmental Engineering Program, and Department of Geology and Geophysics, Yale University, New Haven, Connecticut 06511

Anthropogenic cycling of silver in 1997 is presented using three discrete governmental units: 64 countries encompassing what we believe to be over 90% of global silver flows, 9 world regions, and the entire planet. Using material flow analysis (MFA) techniques, the country level cycles are aggregated to produce the regional cycles, which are used to form a “best estimate” global cycle. Interesting findings include the following: (1) several silvermining countries export ore and concentrate but also import silver-containing semiproducts and products; (2) the level of development for a country, as indicated by the gross domestic product, is a fair indicator of silver use, but several significant outliers exist; (3) the countries with the greatest mine production include Mexico, the United States, Peru, and China, whereas the United States, Japan, India, Germany, and Italy lead in the fabrication and manufacture of products; (4) North America and Europe’s use of silver products exceed that of other regions on a per capita basis; (5) global silver discards, including tailings and separation waste, totaled approximately 57% of the silver mined; (6) approximately 57% of the silver entering waste management globally is recycled; and (7) the amount of silver entering landfills globally is comparable to the amount found in tailings. The results of this MFA lay the basis for further analysis, which in turn can offer insight into natural resource policy, the characterization of environmental impact, and better resource management.

Introduction Identifying and quantifying material flows can increase the understanding of resource use, location, discards, and the resulting environmental impacts. It can also provide insight into the flows in and out of stockpiles such as governmental, investor, and producer hoards. Material flow analysis (MFA) * Corresponding author address: c/o Center for Industrial Ecology 205 Prospect St., School of Forestry and Environmental Studies, Yale University, New Haven, CT 06511; phone: +1-203-432-9733; fax: +1-203-432-5556; e-mail: [email protected]. † Center for Industrial Ecology. ‡ Environmental Engineering Program. § Now at European Aluminum Association, OEA/EAA Recycling Division, B-1150 Brussels, Belgium. | Department of Geology and Geophysics. ⊥ Now at bpTT Energy Company, Trinidad and Tobago. O Now at Environmental Protection Agency, Denver, Colorado. # Now at Federal Institute of Technology, Zurich, Switzerland. ∇ Now at Curtin University of Technology, Perth, Australia. 10.1021/es048319x CCC: $30.25 Published on Web 05/07/2005

 2005 American Chemical Society

and substance flow analysis (SFA) can help illuminate the current and future state of material flows and accumulation in the economy and environment. Generally, MFA tracks material intensities and other indicators of economies or industrial sectors on a specific spatial scale (1). SFA tracks the flows of specific substances to identify specific environmental problems and propose remedial/prevention strategies. Using these tools can help one estimate the amount of a resource captured or lost to the environment and thus the balance needed to maintain or approach sustainability (2). Such information can be invaluable when examining the patterns of resource use, identifying potential areas and magnitudes of environmental impact, and developing policies for resource management (3-6). The Center for Industrial Ecology at Yale University conducts MFA/SFA research through the Stocks and Flows (STAF) project, examining the full life cycle of a material from extraction through disposal on a variety of spatial levels: country, regional, and global. Through this multilevel approach, flows of technologically significant materials are quantified, available stocks are determined, and environmental and policy implications are addressed. To date, multilevel cycles have been completed for copper (7-9) and zinc (10, 11). The copper and zinc cycles have been characterized for each significantly contributing country, as well as nine world regions, and the planet as a whole. Several of the regional level cycles for silver have been published as well: Europe (12), CIS (13), and Asia (14). This paper incorporates these regional cycles, adds those from other regions, and presents the silver cycle for the planet as a whole. The use of MFA/SFA analysis is not new. The Danish government, over the last several decades, has conducted over 35 SFA analyses for Denmark (15). In 1991, Baccini and Brunner stated that, with urbanization, the anthroposphere assumes high energy and material fluxes (16). Jasinski quantified the anthropogenic flows of mercury in the United States in 1995, as part of an attempt by the United States Bureau of Mines to assess the environmental impact of mineral dissipation (17). A number of additional countrylevel analyses by various researchers have followed (e.g., refs 18-21). An SFA of silver is of interest for several reasons. Its depletion time, measured as the reserve base divided by the annual consumption rate (22), is estimated to be approximately 30 years using United States Geological Survey reserve base estimates (23). Some would argue that this calculation points to potential long-term supply issues (24). (This school of thought is by no means universal, however, as detailed in Tilton’s “On Borrowed Time?” (25).) The applications of silver include extensive use in the photographic industry, electronics and electrical equipment, jewelry and silverware, solders and brazing alloys, dentistry, decorative foils and threads, and coins. With key industrial applications relying on silver, ensuring an adequate supply while considering viable alternative materials for some applications is essential. If current usage trends continue, above-ground stocks will be of increasing importance in maintaining resource availability. In addition, silver, as a precious metal, has issues related to this time period such as producer hedging and government stockpiling that make its quantification more difficult to determine than is the case for lesser-valued metals. Finally, silver that is lost to ecosystems presents potential toxicity problems (6), so characterizing the magnitude of loss has environmental applications. VOL. 39, NO. 12, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

4655

FIGURE 1. (a) A simplified schematic diagram of a technological resource cycle, with successive life stages plotted from left to right. (b) The system boundary diagram for the anthropogenic silver cycle. The goals of this study, directed at the contemporary (ca. 1997) cycle, include the following: (i) quantifying the amount of silver extracted, used, lost to the environment, and discarded on a global level; (ii) characterizing the losses of silver from each life stage; (iii) estimating the amount of silver recovered from potential waste streams; (iv) determining the amount entering the stock of specific reservoirs; and 4656

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 12, 2005

(v) comparing and contrasting countries and regions to look for trends in silver flows.

Methodology Silver Cycle Characterization. This paper describes a research study conducted with methodologies similar to those utilized for copper and described in Graedel et al. (9). Figure

1a shows the framework that defines the four stages of the material cycle: production, fabrication and manufacturing, use, and waste management. We term each stage, or box, as a process, reservoir, or stock. The material moves between stocks in flows, the arrows between the processes. Each process can be subdivided further into subprocesses such as mining, alloy fabrication, and recycling. Storage of the material is possible within all of the processes, and can be determined directly from the data or more often, as a calculation based on material inputs and outputs. Each process can have flows into the environment or recycling (scrap) flows. All of the material begins in the Earth’s lithosphere, and eventually ends up in the environment, although often after a long period of time. Using this framework, we produced a life cycle and system boundary diagram for anthropogenic silver (see Figure 1b). Limitations in data availability and simple issues of practicality prevent us from locating and quantifying all of the silver flows with the highest degree of accuracy. Nonetheless, relatively complete estimates are of substantial utility. Thus, the aim of the STAF research is to characterize a minimum of 80% of the silver contained in each flow and each major use. To satisfy this objective, we chose 64 countries or country groups on the basis of gross domestic product (GDP), population, silver production, and silver consumption. We believe that this particular study captures well over 90% of all silver flows. The year chosen for this study was 1997 because we found that the data were essentially complete for this time period and were representative of years in the late 1990s. Each life stage in every cycle should theoretically achieve perfect closure; that is, input flows equal output flows after accounting for any change in stock. Because of data and estimation imperfections, this balance is not always achieved. To reconcile this, we add phantom flows to each of the reservoirs that require them. These flows are represented in our diagrams by values encased in ovals with a dashed arrow pointing into or out of the life stage and, when included in the flow balance, cause perfect closure to be achieved. Phantom flows are needed for most cycles, but often are less than 10% of the total flows throughout the life stage. In several cases, much larger phantom flows exist, leading one to assume that significant data limitations or errors are present and signaling the opportunity for more thorough, accurate data collection and estimation. On the most basic level, material flow analysis is a series of mass balances. Within a bounded system, the change in mass is equal to the inflows minus the outflows, as seen in the equation below.

d (m) ) dt

∑

˘ in) in_flow(m

-

∑

˘ out) out_flow(m

As is the case in an analysis of an element, this equation assumes that no new mass of the given substance is being created or destroyed within the system. The time derivative argument represents the change in mass within the system boundaries, calculated by subtracting the total outflow from the total inflow. From the mass balance perspective, inputs to the silver cycles are from mine production and net import (for nonglobal-level cycles). Outputs include tailings and separation waste, net export (for nonglobal-level cycles), and post-use releases to landfills and the environment at large. Because silver is not being created or destroyed, the change in mass for these cycles occurs entirely through changes in stocks. This is seen through such means as producer hedging, net investment or disinvestment, government stockpile increases or decreases, and additions to or removal from in-use stock.

TABLE 1. Silver Output by Source Metala 2000 primary lead/zinc copper gold other total a

2001

4547 6208 4230 2883 208

25% 34% 23% 16% 1%

4578 6360 4391 2774 249

25% 35% 24% 15% 1%

18 076

100%

18 352

100%

All values in Mg. Source: World Silver Survey (26).

TABLE 2. Silver Fabrication Demand by End Usea 2000 industrial applications 11 728 electrical and electronic 5147 equipment brazing alloys and solder 1189 photographic use 6828 jewelry and silverware 8752 coins and medals 926 total

2001 42% 10 529 18% 4123

39% 15%

4% 24% 31% 3%

4% 24% 33% 3%

28 235 100%

1128 6539 8944 846

26 859 100%

a

All values in Mg. “Electrical and Electronic Equipment” and “Brazing Alloys and Solder” are subsets of “Industrial Applications”. Source: World Silver Survey (26).

Production. In the production stage, the silver ore is extracted from the lithosphere, separated from other substances, and refined into concentrate. Because of its geological occurrence, silver is often co-mined with other metals, including lead, zinc, copper, and gold (26). With silver being a byproduct, albeit often a profitable one, the extraction of the primary metal is often the driving force in mine production. Table 1 shows the silver output for source metal, with only 25% of silver being mined in primary silver mines. Once mined, silver is separated from other materials through smelting or various leaching methods (27). Several data sources (28-32) were used to determine a best estimate of losses from tailings. This estimation, set at 20% of the ore removed, was used globally because sufficient mine-specific data were not available. As seen in the cited data sources, primary silver mines often have lower losses due to tailings, ranging from 2 to 12%. We believe that secondary silver mines have higher tailing losses as a percent of silver output. One reference shows a developing country, Laos, retaining only 39% of silver during processing (32). Clearly this is an area in which further information can improve the silver cycle dramatically, although doing so would require country- or even mine-specific data. After this stage, the silver is then further processed into refined silver, at which point it flows to the fabrication and manufacturing process. Scrap from all other processes into production is assumed to go to the refinery. Fabrication and Manufacturing. During fabrication, the refined silver from production or recycled scrap is used to produce silver semimanufactured products and silver alloy semimanufactured products. These products are then used to manufacture finished silver and silver alloy products. Table 2 shows the global end-uses of silver, with the major enduses being industrial, photographic, and jewelry and silverware. Scrap from manufacture is either sent back to fabrication or to the production process for further refining (26). Use. Silver entering use is in the form of finished products or components of finished products. Silver’s unique qualities, such as its strength, malleability, electrical properties, high reflectivity, and aesthetics make it valuable for a variety of uses (26). The use groups selected for this study are electrical VOL. 39, NO. 12, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

4657

and electronic products, photography (including medical X-rays and other images), jewelry and silverware, brazing alloys and solders, and coins and medals. Silver is used on a much smaller scale in several products such as catalysts, dental fillings (less common in areas with higher GDPs), batteries, and superconducting wire. Silver may be used in alloy form, with silver-copper alloys dominating such uses as coins, silverware, and jewelry (27). Waste Management. Silver and silver residues enter the waste management process from either the fabrication and manufacturing process or the use process. The silver enters waste management in one of five flows: waste from electrical and electronic equipment (WEEE); sewage and sewage sludge (S/SS); hazardous waste (HW); industrial waste (IW); and municipal solid waste (MSW). In waste management, the silver is recycled back to the refinery as old scrap, treated and released into the environment, or released into the environment without treatment. The quantities of silver in WEEE were based on an approximation of the amount of silver contained in the circuit boards of discarded electrical and electronic equipment as 0.2% of the circuit boards by weight (33), and the silver waste from the fabrication of circuit boards. The silver discarded in sewage and sewage sludge includes photographic wastes, medical sewage (dentists and hospitals), electroplating, and domestic sewage. Silver in the hazardous waste flow consists of silver oxide batteries and dental amalgams. A Swiss study estimated that 60% of their batteries are recycled (34) and a Chinese study (35) states that 10% of China’s batteries are recycled. Using these percentages, we estimated the recycling rates for developed and developing countries. In dental amalgam wastes, the silver itself has not been quantified; however, the mercury accompanying the silver has been quantified (36). Thus, the silver content was approximated as proportional to the mercury content. In the silver analysis, industrial wastes consist of the wastes from photographic development. The data for all of the regions was estimated using the European silver cycle. In country-level cycles, the waste generation of silver from film and photograph production ranges from 1 to 4 g Ag/(capita‚ year) using an assumed direct relationship between photographic development and GDP per capita in constant prices, up to 10 000 USD per capita. Above that threshold, comparable behavior to the European cycle is assumed and no correlation with GDP is applied. A recycling rate of 55% is employed (37). Municipal solid waste is the waste generated by households, businesses, open areas, and so forth, and contains silver from photographic prints, dental fillings, silverware, coins and other forms of discarded silver, and silvercontaining products. To calculate the silver content in MSW for each country, we obtained the urban metal content and the percent of the population that is urban-dwelling. For countries where these data could not be found, we made estimates using comparable countries. The differences in metal content between urban and rural people were accounted for based on the findings of a previous study (38). Through personal communications with researchers in Zurich, it was found that their MSW had 8% metal content with 0.00075% overall silver content. The silver content number was then scaled down using a linear relationship with overall metal concentration for each of the countries. This provided all of the data needed to calculate the MSW silver flow. Old scrap data were taken from the World Silver Survey (26). Added to these numbers was the amount of silver recycled in association with photographic development. Although this recycling of silver may not be “old scrap” in 4658

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 12, 2005

the traditional sense of the phrase, we believed that this was the most accurate method to represent these flows.

Results Country-Level Silver Cycles. The characteristics of the silver cycles for individual countries or country groups varied widely, even among countries with comparable economies and cultures. Figure 2 depicts a sampling of six country-level diagrams, each possessing distinct characteristics and telling a unique story. Because of space limitations, the cycles for each of the 64 countries cannot be shown, but they have been made available in the Supporting Information. Because we rely on many data sources to construct these cycles, it is important to gauge the confidence of the results. Following the approach advocated by Moss and Schneider (39), we assigned confidence levels to each silver flow on the basis of the collective judgment of the authors regarding the reliability and quality of the data. Moss and Schneider advocate a five-point scale for confidence levels, but we contracted this to three designations: low, medium, and high uncertainty. For the regional- and country-level diagrams, we indicate the degree of confidence by varying the form of the flow arrows, as shown in Figure 2. Figure 2a is the contemporary silver cycle for Peru, where all results are given in Mg Ag/yr (note: 1 Mg ) 1 metric tonne). This cycle shows a country extracting ample mineralogical silver resources and exporting silver ore and concentrate as well as manufactured silver goods. Figure 2b, the silver cycle for Mexico, shows another country with significant mining operations. The closure balance of Peru and Mexico around fabrication and manufacturing is proportionally larger than the other countries and, we believe, reflects data insufficiencies for import and export. Similar to Peru, Mexico exports silver ore and concentrate, but this country shows a substantial net import of silver goods. India’s silver cycle, shown in Figure 2c, shows only low levels of mine production but a very high import of refined silver. This silver is used in manufacturing operations within India, meeting high demand for jewelry, silver threads, and silver foils. Significant flows into in-use stock are seen for this country. Figure 2d, the silver cycle for the United States, shows a country in which moderate mine production and releases from production stock are used to meet the high demand for silver in the manufacturing life stage, resulting in a net export of silver in products. Figure 2e is the contemporary silver cycle for France. France’s silver mine production is negligible, so it imports refined silver. This silver is used within the country to meet the internal demand for silver-containing products, resulting in a near net balance of silver product import and export. Germany’s silver cycle (Figure 2f) shows heavy import of silver ore, concentrate, and refined silver and an active manufacturing sector that produces an excess of silver products to be exported. Some interesting overall observations can be made concerning the country-level cycles. Several silver-mining countries are net exporters of ore and concentrate but are importers of silver-containing semiproducts and products. This indicates that their natural resources are adequate but that they do not significantly contribute to the value-adding processes associated with silver fabrication and manufacture. Overall, the dominant forces in mining are Mexico, the United States, Peru, and China, whereas the dominant forces in fabrication and manufacture are the United States, Japan, India, Germany, and Italy. The level of economic development for a country, as determined by the GDP per capita, is usually a fair indicator of silver use, but several significant outliers exist. For example, India, with a low GDP per capita when compared to the developed world, is a major consumer of silver products. It is also significant that in almost every case,

FIGURE 2. Country-level silver cycles, ca. 1997, for (a) Peru, (b) Mexico, (c) India, (d) the United States, (e) France, and (f) Germany. Import and export figures for concentrate may include ore. The values encased in ovals represent “phantom flows” required to achieve closure (see text). All values are given in Mg Ag/yr. including countries not shown here, silver is being added to in-use stock in rather large proportions.

Independent States (CIS), (g) Oceania, (h) Asia, and (i) Antarctica.

Regional-Level Silver Cycles. The flows for each of the country-level cycles were aggregated to produce the nine regional cycles. The regional-level contemporary silver diagrams are shown in Figure 3. The nations of the world were divided into the following nine regions, consistent with previous STAF cycles for copper (9) and zinc (40): (a) North America, (b) Latin America and the Caribbean, (c) Europe, (d) Africa, (e) the Middle East, (f) the Commonwealth of

Viewing these nine regional-level cycles in concert provides information on which areas of the world have, and exploit, mineralogical silver wealth, use silver in manufacturing processes, employ silver in use, and dispose of silver. North America’s cycle, shown in Figure 3a, illustrates the highest regional level of mine production as well as silver releases through producer hedging and net disinvestment. Exports of refined silver in this region exceed imports by VOL. 39, NO. 12, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

4659

FIGURE 3. Regional-level silver cycles, ca. 1997, for (a) North America, (b) Latin America and the Caribbean, (c) Europe, (d) Africa, (e) the Middle East, (f) the Commonwealth of Independent States (CIS), (g) Oceania, (h) Asia, and (i) Antarctica. Import and export figures for concentrate may include ore. The values encased in ovals represent phantom flows required to achieve closure (see text). All values are given in Mg Ag/yr. 4660

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 12, 2005

approximately 4000 Mg. A substantial amount of silver product manufacture occurs within this region, but it is nearly met by the high demand of the use life stage. Import of silver contained in semimanufactured and finished products is comparable to that of the exports, resulting in a near-perfect silver trade balance of these goods. Of the silver and silver alloy products that enter use, 35% enters in-use stock, 12% is recycled photographic waste, and 53% enters waste management. Of the amount entering waste management, including imported old scrap, 36% is recycled and 64% is sent to the environment, either directly or in landfills. Figure 3b shows another region, Latin America and the Caribbean, that has a wealth of mineralogical silver. In this region, however, the vast majority of silver is either immediately exported in the form of ore or processed into concentrate or refined silver before leaving the region. Nearly one-half of the semimanufactured and finished products that enter use in this region are imported. Approximately 775 Mg of silver enter use, an amount dwarfed by Asia (Figure 3h), North America (Figure 3a), and Europe (Figure 3c). Europe mines nearly 2000 Mg of silver, predominately in Poland and, to a lesser extent, Sweden. These domestic resources are not nearly enough to meet the demand of the manufacturing sector, so substantial imports of ore, concentrate, and refined silver are needed. Silver entering use for this region is nearly identical to that of North America, at 6000 Mg Ag/yr. Significantly different, however, is the fate of this silver. Much more (56% versus 35%) enters in-use stock, whereas only 31% (as compared to 53%) enters waste management. Africa, shown in Figure 3d, paints quite a different story. Although mine production is relatively low, a majority of mined silver is quickly exported. Later in the material cycle, one sees a net import of silver-containing products, showing that the value-added processes occurred external to Africa. Although this tells an interesting story for Africa, the magnitude of these numbers is small on the global scale. Similarly, the Middle East’s silver cycle, Figure 3e, shows small flows from a global context. Demand, however, requires the net import of both refined silver and silver contained in semimanufactured and finished products. Larger quantities of silver are mined in the CIS, as shown in Figure 3f. Net exports of silver are seen in every form: ore and concentrate, refined silver, semimanufactured and finished products, and old scrap. A similar cycle is shown for Oceania in Figure 3g. Again, strong silver use in the manufacturing sector leads to a net export of refined silver and silver products. In this case, a net import exists for ores and concentrates as well as old scrap. Asia, hosting over one-third of the world’s population, shows the largest flow of silver entering use (refer to Figure 3h). Mine production and releases from production stock fall far short of meeting this demand, so significant imports of refined silver are required. A high proportion of the silver entering use (72%) stays in the reservoir as in-use stock. The rest is recycled as new scrap (8%) or sent to waste management (20%). This “hoarding” of silver may be due in part to India, with 83% of its silver entering use remaining as stock. The penchant of Indians for silver jewelry, an end-use product that often stays in stock, may be the cause. The use of material in Antarctica follows a unique path. Mineral extraction and product manufacture do not occur on the frozen tundra; all of the products must be imported. As shown in Figure 3i, imported products either remain in Antarctica or are shipped “off ice” for disposal elsewhere. Although the magnitude of these flows are insignificant from a global perspective, this region provides an isolated case study with thorough import/export documentation. Global-Level Silver Cycle. When summing the regional cycles, net import and export should equal zero. This indeed

is nearly the case, but minor adjustments and assumptions were needed. The initial global cycle resulted in a net export of silver concentrate and a net import of refined silver and semimanufactured and finished products. The remaining net export and the remaining net import were nearly identical, within 3% of each other. The export flow of this concentrate was instead sent to the fabrication and manufacture life stage, eliminating the import that existed there. The second, and final, adjustment that needed to be made to eliminate the remaining global import dealt with the trade of old silver scrap. Because all of the old scrap import/export data were taken from one source (the United Nations Comtrade Database), one would expect the net import to equal zero on a global level. This, however, was not the case. Importers of old scrap were significantly more conscientious at reporting than exporters, resulting of a net imbalance of 2000 Mg of silver scrap being imported. This may be due, in part, to informal scrap reclamation in developing countries for which recordkeeping is limited. These issues, as well as phantom flows, were dealt with through a system of equations that created mass balance around each life stage. In the system of equations below, please refer to Figure 1a for variable definitions. A mass balance was conducted around the production life stage

OR - TS + SP - Px - Po + MP + WP ) 0 the fabrication and manufacture life stage

MI + WM - Mx - Mo - MP - MW ( Mphantomflow ) 0 the use life stage

UI - S - Ux - Uo ) 0 and the waste management life stage

WI - Wx - WS - La + MW ( Wphantomflow ) 0 Because there is no import or export at the global level, these flows equal zero:

Px ) Mx ) Ux ) Wx ) 0 Phantom flows are also sent to zero:

Mphantomflow ) Wphantomflow ) 0 Qualitative assessments were made on the accuracy of each flow, and then changes were made to achieve closure. For the production life stage, flow Po was deemed the least accurate and was altered to achieve closure. Subsequently, flows Mo, Uo, and La were adjusted, resulting in the final best estimate global-level silver cycle shown in Figure 4. When examining the global silver cycle, several noteworthy results emerge. Global silver discards, including tailings and separation waste, totaled approximately 57% of the silver mined. Approximately half of the silver entering waste management globally is recycled. The other half, that is lost to useful commerce, has two fates. Of the 6100 Mg Ag/yr released from waste management, approximately 2100 Mg Ag is dissipated directly into the environment, and approximately 4000 Mg Ag/yr enters landfills. This amount is found to be nearly identical to the tailing waste of the production stage. As has been demonstrated, for example, in San Francisco Bay (41, 42), determining the sources and magnitude of silver releases is vital in attempts to improve ecosystem health. The characterization of the global cycle for silver in Figure 4 allows us to compute several characteristic performance VOL. 39, NO. 12, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

4661

FIGURE 4. Global-level best-estimate silver cycle, ca. 1997. All values are given in Mg Ag/yr.

TABLE 3. Performance Ratios for Global Metal Cycles ratio

definition

utilization efficiency prompt scrap ratio accumulation ratio recycling ratio reprocessing ratio

Ψ ) 1 - (Mw/Mo) Π ) Mp/Po R ) S/Ui F ) Ws/(Mw + Uo) Ω ) Wp/Ws

a

This work.

b

silvera copperb zincc 0.99 0.06 0.49 0.57 1.00

1.00 0.05 0.67 0.53 0.67

0.91 0.07 0.67 0.42 0.30

Reference 8. c Reference 36.

ratios. These are given in Table 3 along with data for the global cycles of copper and zinc. The information for the other metals permits us to examine the influence of cost on the ratios, given that silver is more costly than copper and copper is more costly than zinc. Consider first the utilization efficiency, a measure of the fraction of material entering manufacturing that emerges in products other than discards. This ratio is near unity for both silver and copper, but closer to 0.9 for zinc. This is due to the higher losses of zinc in industrial processes, and may be caused, in part, to the lesser effort to capture this lower-valued material. The prompt scrap ratios, reflecting input to manufacturing that must be returned for reprocessing, is similar for the three metals at 0.05-0.07. These values may represent typical manufacturing losses that are recaptured in prompt scrap for industrial processes that utilize metals. As technological material cycles for other metals are completed, we will better understand if this figure is independent of the metal being examined. The accumulation ratio, which measures the degree to which flow into use is added to the stock of final users on a net basis, is significantly lower for silver than for the other metals. This is a reflection of the large fraction of silver used in photography, much of which is removed during processing. If this calculation were to do be done on the specific end uses of silver, we would expect a much higher accumulation ratio for silverware and jewelry and a much lower ratio for photographic silver. The recycling ratio describes the amount of the metal entering waste management that is reclaimed as old scrap, whereas the reprocessing ratio calculates the percent of old scrap that goes back to production (instead of being “reused” in the fabrication or manufacturing of products). Recycling and reprocessing ratios clearly show the influence of economics, being highest for silver and lowest for zinc. It is obvious that enhanced value in discards results in more efficient recovery and reuse. Examining specific flows and comparing countries and regions provides added insight. Intuitively, mine production 4662

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 12, 2005

occurs where geological deposits are present. It is in the later life stages where more interesting patterns emerge. Figure 5a shows the silver entering use at the country level for the countries and country groups of STAF-Europe (labeled using ISO country codes), whereas Figure 5b shows the regions of the world. Within Europe, Italy (IT) leads the region in silver entering use, closely followed by the United Kingdom and Ireland (UK/IE) and Germany (DE). Regionally, Asia towers over all of the other regions, using 9900 Mg of silver annually. Europe and North America show comparable amounts, with all of the other regions using much less. Averaged throughout STAF-Earth, we find that approximately 49% of silver entering use is added to in-use stock. Figure 6 details this specific flow from several perspectives. Figure 6a shows the silver entering in-use stock for the regions of the world. Again, we see Asia towering over the other regions, but the margin between Europe and North America has increased significantly. These absolute quantities signal the greatest opportunity for the eventual collection of silver in in-use stock collection. Figure 6b shows the ratio of silver entering use to silver entering in-use stock for the countries and regions in our data set. Among the countries that retain a significantly higher percentage of their silver entering use are Mexico (86%), India (84%), and Japan (72%); several that retain a much lower fraction include the United States (21%), France (28%), and Germany (34%). Figure 7 details the flows into waste management. Figure 7a shows these flows regionally, as well as the respective per capita flows. The wealthier regions are clearly those with the highest per capita discharges of silver. Figure 7b shows the breakdown of these flows by waste management substream. It is important to note that the sum of the substream flows is less than that of Figure 7a, because recycled photographic waste is not included. It is obvious that more silver is lost in municipal solid waste, where potential recovery is likely to be problematic. The same is true of silver in sewage sludge. More tractable are the quantities of silver discarded in industrial or electronic waste streams. As discussed in the methodology section, silver is neither created nor destroyed, and changes in mass occur through changes in stock. Considering only these inflows, outflows, and changes in stock (i.e., a “black box” approach), a simplified picture of silver use can be painted. In Figure 8, these flows are shown for each region, normalized on a per capita basis. STAF-Earth is shown, which includes the 64 countries or country groups examined in this study. Those nations account for 4.91 billion of the 5.82 billion people

FIGURE 5. Silver entering use for (a) STAF-Europe and (b) STAF-World. All values are given in Mg Ag/yr.

FIGURE 6. (a) Silver entering in-use stock for STAF-World. (b) Silver entering in-use stock versus silver entering use. All values are given in Mg Ag/yr.

FIGURE 7. (a) Silver entering waste management, including scrap from use that is immediately recycled. (i) Total flow of silver in Mg/yr. (ii) Per capita flow of silver in g/(yr‚capita). (b) Silver flow into waste management by waste stream. All values are given in Mg Ag/yr.

that inhabited the Earth in 1997 (43). The Earth portion of Figure 8 assumes that all silver has been accounted for and is thus distributed over the entire 1997 population of 5.82 billion people. This is done, in part, to show the effect on the per capita global flows of many of the impoverished nations that were not examined. Import or export, shown on a net basis, incorporates all of the markets and forms of silver (e.g., silver concentrate, silver products, etc.). We see from this figure that the average European has the highest flow into in-use stock (over three

times the global average), followed by the average inhabitant of Oceania and North America. To be expected, mining and separation waste closely followed mine activity and areas of natural abundance. The highest per capita post-use discard rate is found in North America, more than double that of all of the other regions. O, symbolizing other sources and sinks for silver stocks, is comprised of government acquisitions or releases, producer hedging, and investment or disinvestment. It is interesting to note that all regions that show a net flow for these sources indicate an outflow of silver from these stockpiles. VOL. 39, NO. 12, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

4663

photographic purposes is decreasing, both as a consequence of the consumer shift to digital photography (44) and of efforts to develop silver-free photosensitive media (45). Conversely, the use of silver in electronic solder is increasing in some regions due to regulations aiming to eliminate lead-based solders (46), and jewelry-related applications appear robust. More than most industrial metals, therefore, the cycle of silver needs periodic updating, across a variety of governmental and spatial levels. The level at which data are available also affects the methodology and results of a multilevel study such as this. Producer hedging and net investment figures were made available only on a global level. Thus, once the global cycle was complete, these flows were systematically allocated to the countries where the flows seemed most probable. This was primarily done through qualitative comments from the literature and indications from closure balances. It is difficult to evaluate the level of uncertainty in this approach. Despite these data limitations, the value of material cycles is not significantly compromised. This study provides detailed accounts of silver entering stocks, including discards that can potentially be recaptured. Should supply become restricted, reducing such losses or harvesting dormant stocks would become essential. In addition, silver is primarily mined as a coproduct, or “hitchhiker” resource, with other metals, and it is the demand for those metals that drive mine production (23). This relative inflexibility of supply, the lack of good substitutes for certain end uses, and silver’s high value in an unpredictable financial market makes a detailed accounting useful.

FIGURE 8. Regional and global per capita mass balance diagrams. “STAF-Earth” displays the final results distributed over the population of the 64 countries examined. “Earth” displays these same results distributed over the entire global population. Population data taken from the World Bank (43). All values in g Ag/capita.

Discussion The findings present what we believe to be the first comprehensive multilevel anthropogenic silver cycle. As with any material flow analysis conducted at these large levels, our findings can never be perfect or complete; data limitations exist, and estimations must be made. Evaluating the uncertainty is an appropriate next step, but is complicated by the wide array of data sources used and, frequently, the absence of independent data to verify results. One observation that can be made about the reliability of the findings is that production and fabrication and manufacturing results are more reliable than those of use and waste management. As a basic guideline, we estimate the level of uncertainty to be low for most of the flows in the production and fabrication and manufacture life stages, medium for the scrap flows, and high uncertainty when dealing with waste management. Also, the magnitude of the closure balances can serve as an indicator of the accuracy of the results. It is important to note that a small closure balance is necessary but not sufficient to indicate accurate data. However, those countries or regions showing large closure balances immediately signify data insufficiencies. This work represents a snapshot in time of the silver cycle, which is itself changing rapidly. The use of silver for 4664

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 12, 2005

Perhaps most importantly, studies such as this lay the groundwork for further analysis of natural resource use. Economic analysis could provide models to indicate what prices would trigger increased recycling or if externalities (such as losses to the environment or ample in-use stocks) should be considered. Applying energy analysis to these cycles would allow manufacturers to make informed decisions on material choice in relation to energy consumption (e.g., what material to use for a product, or whether to use virgin or recycled material). With soaring energy costs, the knowledge of how energy is linked to a material is inextricably tied to its final price. For materials with detrimental environmental or health and safety effects, these cycles could serve as a foundation for fate and transport studies. Armed with the knowledge of the magnitude of flows into each reservoir, as well as the form of the substance, scientists could focus their efforts where the most significant impact can be made. Thus, substance flow analyses, in part and as a whole, lend themselves to policy considerations at every stage of the decision-making process.

Acknowledgments We thank the staff of the United States Geological Survey and the World Silver Survey for their assistance in data gathering and evaluation, as well as the International Precious Metals Institute for partial funding. We would also like to acknowledge M. Buttazzoni, M. DePalo, J. Errecart, and K. Zyla for their contribution to the European silver cycle.

Supporting Information Available The material cycles for all 64 countries, 9 regions, and the planet as a whole. This material is available free of charge via the Internet at http://pubs.acs.org.

Literature Cited (1) Bringezu, S.; Moriguchi, Y. In A Handbook of Industrial Ecology; Ayres, R., Ayres, L., Eds.; Edward Elgar Publishing Limited: Cheltenham, UK, 2002; pp 79-90. (2) Graedel, T. E.; Klee, R. J. Getting serious about sustainability. Environ. Sci. Technol. 2002, 36, 523-529. (3) Lifset, R. J.; Gordon, R. B.; Graedel, T. E.; Spatari, S.; Bertram, M. Where has all the copper gone? The stocks and flows project, part 1. JOM 2002, 54, 21-26. (4) Spatari, S.; Bertram, M.; Fuse, K.; Graedel, T. E.; Rechberger, H. The contemporary European copper cycle: 1 year stocks and flows. Ecol. Econ. 2002, 42, 27-42. (5) Adriaanse, A.; Bringezu, S.; Hammond, A.; Moriguchi, Y.; Rodenburg, E.; Rogich, D.; Schu ¨ tz, H. Resource Flows: The Material Basis of Industrial Economies. World Resources Institute, 1997. (6) Bianchini, A.; Grosell, M.; Gregory, S. M.; Wood, C. M. Acute silver toxicity in aquatic animals is a function of sodium uptake rate. Environ. Sci. Technol. 2002, 36, 1763-1766. (7) Bertram, M.; Graedel, T. E.; Rechberger, H.; Spatari, S. The contemporary European copper cycle: Waste management subsystem. Ecol. Econ. 2002, 42, 43-57. (8) Graedel, T. E.; Bertram, M.; Kapur, A.; Reck, B.; Spatari, S. Exploratory data analysis of the multilevel anthropogenic copper cycle. Environ. Sci. Technol. 2004, 38, 1253-1261. (9) 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. (10) Gordon, R. B.; Graedel, T. E.; Bertram, M.; Fuse, K.; Lifset, R.; Rechberger, H.; Spatari, S. The characterization of technological zinc cycles. Resour., Conserv. Recycl. 2003, 39, 107-135. (11) Spatari, S.; Bertram, M.; Fuse, K.; Graedel, T. E.; Shelov, E. The contemporary European zinc cycle: 1-year stocks and flows. Resour., Conserv. Recycl. 2003, 39, 137-160. (12) Lanzano, T.; Bertram, M.; DePalo, M.; Wagner, C.; Zyla, K.; Graedel, T. The contemporary European silver cycle: One year stocks and flows. Submitted for publication, 2004. (13) Henderson, K. The contemporary silver cycle for CIS countries: Using industrial ecology to evaluate silver flows. J. Young Invest. 2003, 9, available online at http://www.jyi.org/volumes/ volume9/issue1/articles/henderson.html. (14) Johnson, J.; Bertram, M.; Henderson, K.; Jirikowic, J.; Graedel, T. E. The contemporary Asian silver cycle: 1-year stocks and flows. J. Mater. Cycles Waste Manage. 2005, 7, in press. (15) Hansen, E.; Lassen, C. Experience with the use of substance flow analysis in Denmark. J. Ind. Ecol. 2003, 6, 201-219. (16) Baccini, P.; Brunner, P. H. Metabolism of the Anthroposphere; Springer: New York, 1991. (17) Jasinski, S. The material flow of mercury in the United States. Resour. Conserv. Recycl. 1995, 15, 145-179. (18) Moriguchi, Y. Recycling and waste management from the viewpoint of material flow accounting. J. Mater. Cycles Waste Manage. 1999, 1, 2-9. (19) Eurostat Economy-Wide Material Flow Accounts and Dervied Indicators: A Methodological Guide. Office for the Official Publications of European Communities, 2001. (20) Bringezu, S. In Perspectives on Industrial Ecology; Bourg, D., Erkman, S., Eds.; Greenleaf Publishing: Vienna, 2003. (21) Palm, V.; Jonsson, K. Materials flow accounting in Sweden: Material use for national consumption and for export. J. Ind. Ecol. 2003, 7, 81-92. (22) Graedel, T. E.; Allenby, B. R. Industrial Ecology, 2nd ed.; Prentice Hall: New Jersey, 2003. (23) Hilliard, H. In USGS Mineral Commodities Summaries; United States Geological Survey: Reston, VA, 2004.

(24) Kesler, S. E. Mineral Resources, Economics, and the Environment; Macmillan: New York, 1994. (25) Tilton, J. E. On Borrowed Time? Assessing the Threat of Mineral Depletion; Resources for the Future: Washington, DC, 2003. (26) The Silver Institute, World Silver Survey, 2002. (27) Renner, H. Ullman’s Encyclopedia of Industrial Chemistry; WileyVCH Verlag GmbH & Co., 2002. (28) Atmaca, I. T. Data from the Federal Institute of Geoscience and Natural Resources, Germany, 2002. (29) Sievers, H. Ressourcenorientierte Gesamtbetrachtung von Stoffstroemen matallischer Rohstoffe, Dissertation work at the Technical University of Aachen, Germany, 2001. (30) Yahoo Press Release, http://biz.yahoo.com/cnn/041115/ 898de19e68faedd4c824d13c14bae047_1.html (accessed January 12, 2005). (31) Mining Technology, http://www.mining-technology.com/ projects/kghm/ (accessed January 12, 2005). (32) Wu, J. The Mineral Industries of Cambodia and Laos; United States Geological Survey, 2003. (33) Siemers, W.; Vest, H. Environmental Handbook: Environmentally Sound Electroscrap Disposal and Recycling; Deutsche Gesellschaft fur Technische Zusammenarbeit: Eschborn, Germany, 1999. (34) Vest, H.; Jantsch, F. Umwelt-Handbuch: Umweltvertraegliche Batterieentsorgung und Verwertung; Deutsche Gesellschaft fur Technische Zusammenarbeit (in German): Eschborn, Germany, 1999. (35) China Education and Resource Network, http://www.edu.cn/ 20011120/3010913.shtml (accessed July 01, 2003). (36) Mercury Headworks Analysis for 2000, Palo Alto RWQCP, http:// www.city.palo-alto.ca.us/cleanbay/pdf/mercmassbal.pdf (accessed August 31, 2004). (37) Abwasserrelevante Silberstrome in Wien, Im Auftrag der Magistratsabteilung 22 (in German), Ressourcen Management Agentur, 2000. (38) What a Waste: Solid Waste Management in Asia, World Bank, Urban Development Sector Unit, 1999. (39) Moss, R. H.; Schneider, S. H. In Guidance Paper on the CrossCutting Issues of the Third Assessment Report of the IPCC; Pachuri, R., Taniguchi, T., Tanaka, K., Eds.; World Meteorological Organization: Geneva, 2000; pp 33-51. (40) Graedel, T. E.; van Beers, D.; Bertram, M.; Fuse, K.; Gordon, R. B.; Gritsinin, A.; Harper, E.; Kapur, A.; Klee, R. J.; Lifset, R.; Memon, L.; Spatari, S. The multilevel cycle of anthropogenic zinc. J. Ind. Ecol., in press, 2004. (41) Squire, S.; Scelfo, G. M.; Revenaugh, J.; Flegal, A. R. Decadal trends of silver and lead contamination in San Francisco Bay surface waters. Environ. Sci. Technol. 2002, 36, 2379-2386. (42) Kimbrough, D. E.; Wong, P. W.; Biscoe, J.; Kim, J. A critical review of photographic and radiographic silver recycling. J. Solid Waste Technol. Manage. 1996, 23, 197-207. (43) World Bank Database, http://devdata.worldbank.org/dataonline/ (accessed May 15, 2004). (44) Metals - Gold falls on profit-taking; base metals drift on fundliquidation. AFX News Limited, September 26, 2003. (45) Marshall, J. L.; Telfer, S. J.; Young, M. A.; Lindholm, E. P.; Minns, R. A.; Takiff, L. A silver-free, single-sheet imaging medium based on acid amplification. Science 2002, 297, 1516-1521. (46) Roos, G. Passives makers actively eliminate lead. Electronic Engineering Times, Feb 3, 2003.

Received for review October 28, 2004. Revised manuscript received April 3, 2005. Accepted April 7, 2005. ES048319X

VOL. 39, NO. 12, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

4665