Predicting spatial distribution of heavy metals in an abandoned phosphogypsum pond combining geochemistry, electrical resistivity tomography and statistical methods.

Affiliation

Departamento de Ingeniería Minera, Geológica y Cartográfica, Universidad Politécnica de Cartagena, Paseo Alfonso XIII, 52, 30203, Cartagena, Murcia, Spain. Electronic address: [Email]

Abstract

One of the wastes generated in fertiliser production from phosphoric rock is phosphogypsum, whose mismanagement lead to environmental and health risks. Therefore, a detailed evaluation of the chemical composition of phosphogypsum is necessary to determine effective means of its management. Due to the high amount of generated waste, the cost and time consumed for this characterisation by chemical analysis is limiting. Hence, efficient tools should be developed to predict the chemical composition of this waste. Thus, this study aims to: 1) determine the physic-chemical characterisation of phosphogypsum pond using geochemical and geophysical techniques and 2) predict the heavy metals spatial distribution through statistical models. Results show that the most concentrate metal is chromium with a maximum of ≈900 mg.kg-1 and cadmium is the least concentrated (maximum ≈23 mg.kg-1). The Electrical Resistivity Tomography revealed the superposition of two layers. The top one (waste) presents low resistivity (≈17Ω.m) while the bottom layer shows higher resistivity (>124Ω.m). Metal concentrations and resistivities were combined by applying non-linear regression models. Cr showed the strongest correlation (R2 = 0.68), yielding an accurate model that was used for revealing the spatial distribution of the highest Cr concentrations in the pond, with the consequent reduction of expensive traditional methods.

Keywords

Correlation,Electrical resistivity tomography,Fertilizers,Phosphogypsum,Regression models,

OUR Recent Articles