Evaluating the environmental impact of products and production processes: a comparison of six methods

The desire of environmentally-conscious consumers and manufacturers to choose more environmentally benign products and processes has led to the development of life cycle assessment (LCA) and design for environment (DfE). In both of these areas, attention has focused initially on the development of inventories of emissions and raw materials consumption for particular products and processes. A number of methods for the comparison and evaluation of an inventory's dissimilar pollution loads and resource demands have been proposed, but no satisfactory solution has yet been identified. This paper compares the structure and properties of six different methods. The health hazard scoring (HHS) system uses the analytical hierarchy process (AHP) to weight workplace toxic effects and accident risks. The material input per service-unit (MIPS) aggregates the mass of all the material input required to produce a product or service. The Swiss eco-point (SEP) method scores pollutant loadings based on a source's contribution to an acceptable total pollution load and an environmental scarcity factor. The sustainable process index (SPI) determines the area that would be required to operate a process sustainably, based on renewable resource generation and toxic degradation; an extension of the dilution volume approach. The Society of Environmental Toxicology and Chemistry's life-cycle impact assessment (SETAC LCA) impact assessment method aggregates pollutants with similar impacts to equivalency potentials (measured in kg CO2 equivalent, kg benzene equivalent etc.) and uses decision analysis to assign weights to different adverse impacts. The environmental priority system (EPS) characterizes the environmental damage caused by equivalency potentials and expresses it in monetary terms, derived from environmental economics. Despite their use for the same purposes, the six methods differ in what they try to achieve, in the effects they consider, in the depth of analysis, in the way values influence the final score, and in use of ordinal or cardinal measures of impact. Two problem areas are identified: (1) to varying degrees, each of the methods has the potential to recommend an alternative that actually has a higher impact than other alternatives; (2) for some of the methods the data requirement is so extensive and the tolerance of imperfect data is so low that the application of the method for reasonably sophisticated products or processes would be too complicated.