No abstract

The performance of a draft-tube spouted bed bioreactor (DT-SBB) packed with hydrogel particles for biomass immobilization has been used to treat a phenolic wastewater in continuous mode. The biomass support particles were made of a cyclodextrin-based polymer and then seeded with an acclimated mixed culture. Due to the low density of such particles and also the bioreactor design, the pumping energy required to maintain a moving bed of the resulting bioparticles was very low and a recirculation flow was not necessary. The inlet phenol concentration and the phenol inlet loading rate were the operating variables studied. Phenol removal efficiency was used to monitor the process. The DT-SBB showed a high removal capacity (up to 2.8 kg-phenol/m 3 d) with a high efficiency (>99%). The removal capacity of the DT-SBB was found to be limited by the availability of dissolved oxygen in periods with high phenol elimination rate. The removal efficiency decreased significantly for higher loading rates (>3.1 kg/m 3 d) because of phenol inhibition. The evolution of the distribution of microbial populations was also investigated. A predominance of gram-negative bacteria (especially the genera Pseudomonas and Acinetobacter) was observed during the periods of maximum degradation.

A draft-tube spouted-bed (DTSB) reactor was equipped with an auxiliary aeration device to provide air into the annular region and thereby improve the oxygen transfer efficiency. The effects of the total air flow rate and its distribution between spout and annulus, the liquid phase viscosity (water and carboxymethyl cellulose solutions), and the solid holdup (glass and cyclodextrin polymer beads) on the oxygen transfer efficiency were discussed. The oxygen transfer coefficient increased with the air flow rate and the ratio of air flowing through the annulus, whereas it decreased with increasing viscosity and solid holdup. A correlation was proposed to predict the transfer coefficient in DTSB reactors with primary and auxiliary aeration. A good fitting was achieved between the experimental data and those estimated with the model.