Abstract
Process design alternatives can be compared in terms of their effluents and energy consumption. Approaches can be used to account for differences in regional exposure and toxicological effect parameters. A significant number of methodologies have been published in the literature to achieve this objective. Selection of a methodology is typically subjective and influenced by resource availability. Often criteria are not used to guide designers in their selection or to steer associated research activities. In this paper, the merits and applicability of five types of methodology are considered using a hierarchical framework. The hierarchical framework presented is applicable for approaches that are based on toxicological impacts on a regional scale.

Introduction
A significant need has been emphasised in management standards (ISO 14001, BS 7750, EMAS), in legislation (IPC in Europe), by non-governmental organisations and in corporate guidelines for the development of straightforward methodologies to enable the comparison of process design alternatives in terms of environmental impacts. These approaches typically do not indicate that an emission will result in an impact but provide a relative scale for comparison in terms of "contributions" to potential impacts.

Impacts associated with emissions may be classified into resource depletion, human health, ecological degradation, economic effects and social effects. Comparison of process alternatives in terms of all impact categories reflects the objective of sustainable development and can involve the use of holistic frameworks like Life Cycle Assessment (LCA)1 to address associated emissions off-site. In this paper we focus on the options available for comparison in terms of contributions to potential toxicological impacts on human health and the ecosystem from process emissions.

A reasonable consensus exists for comparison in terms of global-scale issues like ozone depletion and global warming potentials. 2 However a significant number of methodologies have been published to facilitate comparisons in terms of contributions to potential toxicological impacts on human health and ecosystem. These range in complexity from simple summations of effluent quantities to detailed, site-specific assessments. The selection of a methodology is often subjective and influenced by resource availability (time constraints, in-house knowledge, etc.). No unifying criteria are available to guide process designers in their selection or to steer research activities. In this paper a hierarchical framework is used to consider the merits and applicability of the different types of methodologies.

The hierarchical framework presented is applicable for the comparison of emissions in the context of potential toxicological impacts to human health and ecosystems on a regional scale. It is assumed that the magnitude and concentration of emissions are normally restricted by legislation to limit actual impacts resulting from individual discharges on a local scale. Secondary impacts associated with photochemical ozone creation, breakdown products, acidification, and eutrophication are similarly not addressed. The strengths and weaknesses of the different types of approaches in the hierarchy are illustrated using a case study.

Conclusion
A significant number of approaches have been presented in the literature to facilitate the comparison of process design alternatives in terms of potential contributions to toxicological impacts on a regional scale. These approaches typically do not indicate that an emission will result in an impact but provide a relative scale for the comparison of effluents. A hierarchical framework and guidelines are presented in this paper to help process designers consider the merits and applicability of the different types of methodologies. The hierarchy is based on the degree of representation of associated environmental mechanisms. However, the suitability of an approach will depend on the relative environmental behaviour of the chemicals considered, the quality of available data, the comprehensiveness of the model and the ability of more resource-intensive techniques to provide an improvement in discrimination.

The direct summation of effluent quantities requires relatively minimal resources, however its applicability is limited for chemicals in effluent streams that exhibit significantly different exposure and effect behaviour. For these cases, a number of techniques are available ranging from weighting by toxicological data to the use of sophisticated models. The concept of using multimedia fate and exposure models in conjunction with effect data to facilitate comparison on a regional scale in terms of contributions to potential toxicological impacts appears very promising, particularly for gaining insights into important mechanisms and providing comparison across species groups. Nevertheless, intrinsic uncertainty is usually undefined and the ability of available data to provide a consistent basis for use in relative comparison applications is questionable. The use of multimedia models and the interpretation of the results therefore remain strongly dependent on scientific judgement. Considerations include model applicability, comprehensiveness, intrinsic uncertainty, parameter selection, data quality, degrees of conservatism and the extent of relevant validation.