Ferramentas quimiométricas aplicadas na avaliação de estratégias que visam a remoção de íons fluoreto em efluente de drenagem ácida de mina

Abstract

The goal of this study is to provide improvements in the treatment of effluents from acid rock drainage (ARD) of the Mineral Treatment Unit (Caldas / MG) of the Nuclear Industries of Brazil (INB). Due to the presence of fluoride ions in the effluent, the costs of inputs and waste management are in the order of millions of reais per year. Thus, the search for alternative solutions or an optimized condition is of paramount importance, making the current treatment system more efficient from a financial, environmental and social point of view. Of the nine materials initially tested for fluoride removal, five were selected because of the best performance: gypsum, manganese dioxide, alumina, aluminum sulfate and aluminum polychloride. Following, a selection of variables that influence the removal of fluoride was done considering: Dosage of Hydrated Lime; Stage of Addition; Time and Material Dosage, by means of fractional factorial design for each material (5 x 23-1 ). Then, with the variables and materials selected, the work was directed to obtain a response surface that allowed identifying the optimum working region, considering the responses: final fluoride concentration, costs and mass generation of solids. For this, Doehlert type design was used with three variables (13 experiments + 3 replicates) and the evaluated responses were transformed, according to criteria, through desirability function. After obtaining the response surfaces and after the models validation for the best performance materials: gypsum and hydrated lime, the efficiency of the use of gypsum - solid effluent from the treatment of ARD - was tested on continuous scale by a pilot plant operation. It was concluded that the addition of gypsum in two different points in the treatment reached the objective of fluoride concentration (less than 10.0 mg / L) with lower cost of inputs. The proposal allows cost reduction in the order of hundreds of thousands of reais per year.