Aplicação da fluidodinâmica computacional no estudo hidrodinâmico de um reator anaeróbio híbrido (UAHB) abiótico com validação experimental

Abstract

The main objective of this work was to evaluate the hydrodynamic behavior of a hybrid anaerobic reactor by means of computational fluid dynamics and hydrodynamic stimulus-response tests. For a study, a 17,7 L hybrid anaerobic filled with mats of interlaced vinyl fiber was used. In order to evaluate the hydrodynamic behavior and its hydraulic characteristics and anomalies, three hydrodynamic tests were performed for each of the hydraulic detention methods tested (HRT of 4 h, 8h and 12 h) using a yellowish eosin tracer, and the total duration of the tests was three times the theoretical HRT. For the computational fluid dynamics simulations, the ANSYS-CFX 19.0 software, academic version, was used. From the hydrodynamic tests, it was possible to recognize the variation curves of the eosin concentration over time, mean advances of the concentration peaks of 46%, 66% and 53% for the HRT of 4 h, 8 to 12 h, respectively, being indicative of the presence of non-interior paths of the reactor. The curves also allowed an observation of the slow decay of tracer marking over time, characterizing the long tail effect, besides the presence of some peaks along the curves, being an indicative of strong internal recirculation inside the reactor. Based on the adjustment of the experimental data to the uniparametric models, it was observed that the predominant flow regime in the reactor is that of complete mixing. By calculating the hydraulic characteristics and anomalies, it was verified that in most cases the positive volumes of dead zones, which may have been influenced by the slow decay of the tracer concentration curve. The Reynolds number indicated a laminar flow regime, there were no hydraulic short-circuits in the three HRT tested and it was observed that the majority of the tests had unsatisfactory hydraulic efficiencies.The CFD simulations allowed to obtain the volume of dead zones in the reactor, which was approximately 13 % in the three simulated HRT, besides allowing to verify the presence of paths preference in the flow of the fluid, however there was no validation of the simulations, due to the relative error being above 50 % between the volume of real and simulated dead zones.