Artificial neural networks for the treatment of tanneries effluents

Abstract

One of the main problems that concerns companies dedicated to leather tanning and dressing lies in the high contaminant concentrations of the effluent that they release into the sewage network, which exceed the maximum admissible values ​​(VMA) established in DS 010-2019 -VIVIENDA.In line with this reality, an alternative is presented to reduce the contaminant load of Biochemical Oxygen Demand (BOD5), Chemical Oxygen Demand (COD) and Total Suspended Solids (TSS) present in tannery waters through the use of a effluent treatment system that is controlled by an artificial neural network that automatically determines the dosing of inputs for the treatment of effluents in tanneries and depending on the combination of effluents from the tannery processes, the chemical inputs are dosed.The methodology used consists of removal of sedimentable organic load through a grease trap and sedimentation pond, oxidation of sulfur organic matter, pH regulation, coagulation and flocculation according to the dosage indicated in the neural network and an aeration system. With a sample of 5 m3 / day, it was possible to reduce the parameters by 70%, obtaining a concentration of BOD5, COD and TSS around 211.3 O₂ mg / L, 790 mg / L and 109.7 mg / L respectively; meanwhile, 0.007 mg / L of hexavalent chromium, 0.005 mg / L of total cyanide, 1.9 mg / L of settleable solids, 6.5 mg / L of oils and fats and 740.9 mg / L of sulfates were obtained; all under the validated treatment system unlike conventional methodologies that only treat the effluents for each process separately.