Working paper, presented at the 3rd Biennial Conference of the US Society for Ecological Economics, Tacoma, WA, July 20 -23, 2005 1Efficiency and Production: Historical Trends for Seven Industrial Sectors Jeffrey B. Dahmus and Timothy G. Gutowski Department of Mechanical Engineering Massachusetts Institute of Technology 77 Massachusetts Avenue, Room 35-234 Cambridge, MA 02139 [email protected] and [email protected] Abstract A review of historical data for resource efficiency, production, and total resource usage for seven industrial sectors reveals several clear patterns of activity. For example, over their approximately 50 to 200 year histories the manufacturing, utilities and service sectors reviewed here, all showed significant average increases in all three measures. However, on average, production increases outpaced efficiency increases by substantial margins leading to average increases in total resource usage. For example, while the energy efficiency of pig iron production has improved by an average of over 1% annually for the last 185 years, worldwide production of pig iron over this same period of time has increased by an average of 4% annually. Aluminum production, nitrogen fertilizer production, electricity production from coal, oil, and gas, and air travel all exhibit similar trends. The correlation between efficiency and production, as well as the possible causal relationship between the two, casts serious doubts on the ability of “naturally occurring” efficiency improvements alone to reduce resource consumption and associated environmental impacts. However, efficiency and consumption data from the use phase of automobiles and refrigerators, cases in which efficiency improvements were imposed by the government, show very different trends. For example, in the case of refrigerators, government-mandated efficiency requirements, that were constantly upgraded, reversed previous trends and led to an overall reduction in resource use and impact. A similar reversal was found for automobiles, but policy inaction has allowed the original pattern of increased resource usage to reemerge. The implications of using efficiency improvements to curb total resource usage are discussed.Working paper, presented at the 3rd Biennial Conference of the US Society for Ecological Economics, Tacoma, WA, July 20 -23, 2005 2Introduction In attempting to move towards a more sustainable society, efficiency is often seen as a promising solution. Industrial ecologists, business leaders, engineers, politicians, and others have all embraced efficiency as a means of preserving economic growth while reducing the amount of resources used by the economy. Indeed, efficiency improvements are often seen as “win-wins”, as they allow for economic and societal gains to continue unimpeded while at the same time reducing resource consumption (DeSimone and Popoff 1997, OECD 1998, WBCSD 2000). Technical efficiency improvements are by no means new; “naturally occurring” or “autonomous” efficiency improvements have been taking place for centuriesa. In production activities, process efficiencies are continually being improved, thus allowing more useable output to be produced per unit of input. However, in many cases, while efficiency has improved, production has at the same time also increased, resulting in a situation in which the overall inputs to production have increased, not decreased. In addition to bringing the win-win scenario into debate, these trends have also raised questions regarding the possible causality of these trends. This paper will explore the historical relationship between efficiency and production for seven industrial sectors, ranging from the worldwide production of pig iron to the U.S. consumption of residential refrigeration.b To better frame this relationship in the context of sustainability discussions, a variant of the well-known IPAT equation will be used.c This variant is the IPT form of the equation, where impact (I) is a product of production (P) and technology (T). Expressing technology as the reciprocal of efficiency (e) yields ePI1×= (1) a Note that during the period reviewed here, on average, fuel prices have been falling. Efficiency improvements during these periods generally accompany technology innovations and are called “autonomous” in the economics literature. We use “naturally occurring” as synonymous with “autonomous”. b In the case of supply-side activities such as manufacturing, the relationship between efficiency and production is of interest. In the case of demand-side activities such as product use, the relationship between efficiency and consumption is of interest. c The IPAT equation is a mathematical identity that equates impact (I) to the product of population (P), affluence (A), and technology (T). A recent discussion and historical review of the IPAT equation is provided by Chertow (2000).Working paper, presented at the 3rd Biennial Conference of the US Society for Ecological Economics, Tacoma, WA, July 20 -23, 2005 3 From (1), it is clear that in order to reduce impact, one of four different scenarios must occur: 1) production decreases with no change to efficiency, 2) efficiency increases with no change to production, 3) production decreases while efficiency increases, or 4) efficiency increases faster than production increases. The first three methods of impact reduction all require either a decrease in production or a leveling of production, and are thus not generally considered desirable by many. The fourth method of impact reduction, in which the rate of improvement in efficiency exceeds the rate of growth in production, requires the following condition: PPee ∆≥∆ (2) Production and Efficiency Production and resource efficiency data were obtained for five of the seven industrial sectors studied: three global sectors, namely pig iron production (energy and coke efficiency), aluminum production (energy efficiency), and nitrogen fertilizer production (energy efficiency), and two U.S. sectors, namely electricity generation (coal, oil and gas efficiency), and airplane travel (energy efficiency). Figures 1 through 4 show worldwide efficiency and production data for pig iron, aluminum, and nitrogen fertilizer. Process efficiencies shown in the figures are measured in terms of kilograms of product produced per unit of resource used. Resources are measured in joules of energy used,