Lumped kinetic models for single ozonation of phenolic effluents

In this work, the mineralization of a phenolic wastewater was investigated by several lumped kinetic models to predict the total organic carbon depletion during single zonation of a mixture of six phenolic compounds present in olive oil mill wastewater. Firstly, the kinetic equations were developed in terms of lumped compounds representing the parent pollutants and reaction intermediates according different models: First-Order Kinetic (FOKM, with one oxidation step), Two Step First Order Kinetic (TSFOKM, including also and intermediate path), Modified First-Order Kinetic (MFOKM, adding initial non-oxidizable pollutants to the FOKM), Lumped Kinetic (LKM, accounting for non-oxidizable species production). Generalized Kinetic Model (GKM, where, besides the direct mineralization step, formation and degradation of intermediates provide a three step alternative model) and Modified Generalized Kinetic Models (MGKM, a five step proposal with the addition of two paths accounting for non-oxidizable species possible to be produced from the oxidation of raw or intermediate compounds). Secondly, the behavior of these various models was studied at different operational conditions. The FOKM failed to describe the process while TSFOKM incorporates the drawback of requiring the choice of a transition point between the two types of oxidation behaviors. MFOKM, assuming the presence of refractory compounds in the initial wastewater, was rejected since all the parent phenolic compounds were totally removed from the effluent along the treatment. LKM gave good agreement with the experimental results at low ozone inlet concentrations, due to the high recalcitrant character of the organic intermediates that were formed. Nevertheless, for higher ozone inlet concentration some by-products degradation occurs, which is not predicted by this model. Finally, the GKM fitted satisfactorily the laboratorial data with errors in the range of +/- 10% and, even with the shortcoming of involving a large number of parameters, MGKM was the most reliable model to describe this complex process. (C) 2010 Elsevier B.V. All rights reserved.