Metals, Energy and Sustainability Barry Golding Abstract In his keynote address to delegates attending the 1993 Australasian Institute of Mining and Metallurgy Centenary Conference, Brian Skinner concluded that; “The global resources of minerals mined for purposes other than energy are such that exhaustion will not be a problem a century hence.” His bold statement was in part responsible for this thesis. Sustainability is a difficult concept in mining because most mines are unsustainable. Two concepts are addressed in the chapter devoted to defining sustainability in relation to mining. Are sufficient resources available, and can these resources be developed within the constraints of Ecologically Sustainable Development (ESD)? The economic constraint of increasing human welfare prefaces the requirement that copper be produced at no greater relative cost than it is today. Economic and engineering models are developed for the cost of copper production in an attempt to answer the question: is copper mining sustainable, particularly if energy costs increase? More specifically, will the relative cost of producing copper in the US be higher or lower in 2020 assuming companies adhere to the ESD principles? Mining will be considered sustainable if it enhances human welfare, now or in the future, within the constraint of ESD. The data set for the economic model was constructed mainly from information supplied by the U.S. Geological Survey, the Bureau of Labor Statistics and the U.S. Census Bureau for the period from 1954 to 2002. The Energy Information Administration price forecasts for the US out to 2035 suggest that US copper producers will face, at most, a doubling of the real energy price between 2002 and 2020. This order of magnitude increase is used to estimate the impact of energy price increases on the cost of copper production. A translog cost function is employed to develop an economic model. The independent variables are capital, labour, energy, materials, copper ore-grade and time (K, L, E, M, g, t), with time representing technology change. The cost function is estimated using the Seemingly Unrelated Regression method. The chosen economic model was tested and found to be sufficiently robust and provided the information needed to forecast future movements in the price of copper under the likely scenarios of relative energy price increases and other changes. An engineering model is also developed that compares data for the actual energy required to mine and concentrate a kilogram of copper in the US between 1954 and 2002 with the estimated theoretical amount of energy required. A formula for the theoretical energy required to mine and concentrate copper is developed based on data from many authors; however, the formula is extended to take account of the finer grinding required for lower grade copper ores. The difference between the theoretical energy requirement and the actual energy required to mine and concentrate copper in the US is assumed to be due to technology improvement. The economic model, which is based on the work of previous researchers who have used the KLEM translog cost function to examine historic data, is extended to forecast the future cost of producing copper. Specifically, the relative cost of producing copper in 2020 is estimated. Likewise, the engineering model is extended to estimate the future changes in the energy required to produce a kilogram of copper. Formulae are developed for estimating the likely decrease in the grade of copper ore to be mined and the theoretical energy that will be required to mine and concentrate copper as ore grades decrease. A formula for the estimated reduction in energy required flowing from technology improvements is developed. The energy required to mine and concentrate copper in 2020 is estimated using these formulae. Time, representing the technology trend, is introduced into the economic model as a natural number, implying that the rate of technical change is constant; however, when time is introduced as the natural log of time, implying a decreasing rate of technological improvement, the increase in energy cost is no longer offset by the technology cost reduction, and a twofold increase in the cost of energy will result in a forecast increase in the relative cost of producing copper. The true amount by which technology change can offset energy price increases and mineral grade decreases may well lie somewhere between that predicted from the chosen economic model and that flowing from the model including the technology trend as the natural log of time. However, the findings from the chosen economic model and the engineering model support the proposition that, with a twofold increase in the relative cost of energy, the cost of producing copper will be lower on average in real terms in 2020 than it was in 2002.