Dynamic modeling methods for characterizing substance flows in the economy permit a multi-year analysis because they treat time-series data. This paper analyses the complete life cycle flows of copper extracted and used during the twentieth century in North America using time series production (primary and secondary) and fabrication data, and a dynamic product residence time model to determine discard flows. The major motivation for looking at historical inventories of substance stocks and flows is to understand how and in what proportion materials are used, how they may dissipate into the environment, and how they partition into certain reservoirs (e.g., landfills), which can ultimately lead to assessing the feasibility of mining landfills for recovery of valuable materials. Results for the last century (1900-1999) show that about 160 Tg of copper was extracted from the lithosphere and two anthropogenic stocks have formed, i.e., in use (70 Tg) and in landfills, tailings and slag reservoirs (85 Tg). When comparing the amount of copper utilized in North America this century to that worldwide throughout human history, we found that North America accounted for about 30% of all extracted copper. At present, natural reserves of copper amount to 113 Tg in North America, however, only about 50% can be feasibly extracted. The copper cycle shows that over the 100 year period, 40 Tg of copper were collected and recycled from post-consumer waste, and 56 Tg accumulated in landfills or were lost through dissipation, and 29 Tg of copper in waste was produced in the form of tailings and slag and stored in waste reservoirs. Unless the copper stored in waste repositories can be economically extracted, that resource is lost to society. The residence time model shows a significant rise in the rate at which copper is placed in landfills from post-consumer waste in the period 1940 (270 Gg Cu/year) to 1999 (2790 Gg Cu/year). This trend may intensify in the future because of the increasing rate of electronic equipment use and correspondingly shorter residence times, and the absence of an efficient collection and processing infrastructure for retired electronics. Uncertainties in product end-uses and product residence times are high, and additional data are needed to confirm copper waste generation. (c) 2005 Published by Elsevier B.V.
