New methodology for hazardous waste classification using fuzzy set theory - Part I. Knowledge acquisition

In the literature on hazardous waste classification, the criteria used are mostly based on physical properties, such as quantity (weight), form (solids, liquid, aqueous or gaseous), the type of processes generating them, or a set of predefined lists. Such classification criteria are inherently inadequate to account for the influence of toxic and hazard characteristics of the constituent chemicals in the wastes, as well as their exposure potency in multimedia environments, terrestrial mammals and other biota. Second, none of these algorithms in the literature has explicitly presented waste classification by examining the contribution of individual constituent components of the composite wastes. In this two-part paper, we propose a new automated algorithm for waste classification that takes into account physicochemical and toxicity effects of the constituent chemicals to humans and ecosystems, in addition, to the exposure potency and waste quantity. In part 1, available data on the physicochemical and toxicity properties of individual chemicals in humans and ecosystems, their exposure potency in environmental systems and the effect of waste quantity are described, because they fundamentally contribute to the final waste ranking. Knowledge acquisition in this study was accomplished through the extensive review of published and specialized literature to establish facts necessary for the development of fuzzy rule-bases. Owing to the uncertainty and imprecision of various forms of data (both quantitative and qualitative) essential for waste classification, and the complexity resulting from knowledge incompleteness, the use of fuzzy set theory for the aggregation and computation of waste classification ranking index is proposed. A computer-aided intelligent decision tool is described in part II of this paper and the functionality of the fuzzy waste classification algorithm is illustrated through nine worked examples. (c) 2007 Elsevier B.V. All rights reserved.