Mathematical Modeling of Site Sensitivity Indices in the Site Selection Criteria for Hazardous Waste Treatment, Storage and Disposal Facility

With the Basel ban, imposing restrictions on Trans-boundary movement of hazardous wastes, coming into force with effect from 1.1.1998, the concept of treatment and safe disposal of hazardous wastes (HWs) assumes paramount significance. The annual HW generation in India is estimated at 0.3-0.5 million tonnes. As per the ideal industrial siting criteria in India, the industry should have enough land available within its premises for the treatment and disposal and or reuse/recycling of the wastes generated from it. However, very few industries in India, mostly in large scale and very few in medium scale, own proper treatment and disposal facilities. A common waste treatment and disposal facility such as treatment, storage and disposal facility (TSDF) for management of hazardous wastes generated from industries, is one of the useful options under such conditions. Few Guidelines issued by Ministry of Environment and Forests under Hazardous Wastes (Management & Handling) Rules, 1989 promulgated under Environment (Protection) Act, 1986 are available in India for selection of best site for TSDF. These guidelines may vary from country to country. A few limitations in the prevailing methodology for site selection of TSDF under Indian conditions with respect to the prediction of Site Sensitivity Indices (SSI) are observed. The present study is aimed at floating few suggestions for the accuracy improvement of the same. A best-fit mathematical model (BFMM) is developed to accurately predict the SSI. Mathematical models are also developed for the assumption (LIAI) given in the Guidelines and another assumption (OLDM) under which the ranking of the sites may take place. The results are encouraging and revealed that, though the methodology given in the Guidelines is closely matching with the present statistical analysis, the present study is superior in terms of simplicity, reliability, accuracy and clarity over the prevailing assumptions. Moreover, the BFMMs proposed in the present study, also predict the asymptotic SSI values of 1 and 0 for all the attributes considered. The results of the second assumption (i.e., OLDM) are highly erratic and we recommend it to be discouraged from practicing it.