A bench scale UASB reactor (160 l) was used in stimulus-response assays in order to evaluate its behaviour when submitted to cyclical sinusoidal variations of the influent flow rate. Assays were carried out with eosin Y as the tracer in different operational conditions: constant mean influent flow rate of 16 l.h-1 and hydraulic residence time of 10 h (assays 1 and 2), and influent flow rate submitted to sinusoidal variation of 40% (assays 3 and 4) and 60% (assay 5). The variation of tracer concentration in the effluent was adjusted by the theoretical models of dispersion (low and high) and N-continuous stirred tank reactors in series. Complete mixture behaviour was noted to the application of the constant influent flow rate and a plug flow one to the application of the cyclical sinusoidal variations of the influent flow rate. The low dispersion model demonstrated the best mathematical adjustment in the RTD.
Diffusion; bioreactor; hydrodynamics; tracer; tailing phenomena
