Evaluation of zinc adsorption through batch and continuous scale applying Bayesian technique for estimate parameters and select model.

This work aims to study the efficiency of zinc adsorption onto granular-activated carbon, predicting the mathematical models that best describe the adsorption behavior in a fixed bed column. First, batch scale experiments were performed to evaluate the influence of pH (3 to 6), contact time (5 to 60 min), and absorbent concentration (5 to 25 g L) using synthetic effluent. Fixed bed column experiments were performed by varying the adsorbent concentration (10, 13, 20, and 40 g L) and the effluent flow rate (15 and 20 mL min). Markov Chain Monte Carlo and Bayesian criteria information were applied to describe the phenomena using Langmuir, Freundlich, Temkin, Redlich-Peterson, Sips, Toth, Khan, Radke-Prausnitz, for isotherm models, and Thomas; Yoon-Nelson; Yan; Clark models for breakthrough curve. Adsorption operating best conditions were pH 5, 20 g L of solid, and 50 min of contact time. These parameters allowed 80% of Zn removal, being better described by the Tempki model. In tests on a pilot plant, the Yan model was able to predict the second-order kinetic model, with an increase in the effluent flow and a 50% increase in the bed saturation time with a greater amount of adsorbent solid.