Contact
    Start Publications Enhanced mass transfer in a novel external-loop airlift reactor ...
    ←  Publication Finder
    Attension

    Enhanced mass transfer in a novel external-loop airlift reactor with self-agitated impellers

    Year: 2017

    Journal: Biochem. Eng. J., Volume 118, FEB 15, page 53–63

    Authors: Lukic, Natasa Lj.; Sijacki, Ivana M.; Kojic, Predrag S.; Popovic, Svetlana S.; Tekic, Miodrag N.; Petrovic, Dragan Lj.

    Organizations: Ministry of Education, Science and Technological Development of the Republic of Serbia [172025]

    Keywords: Self-agitated; Impellers; External-loop airlift reactor; Volumetric mass transfer coefficient; Non-Newtonian; Artificial neural networks

    In the present study a novel type of internals, self-agitated impellers, were inserted in the riser section of an external-loop airlift reactor to intensify mass transfer rates. The performance of self-agitated impellers was evaluated through comparative analysis of the obtained volumetric mass transfer coefficient values with regard to liquid phase properties and sparger types. Compared to the configuration without impellers, self-agitated impellers considerably improved reactor characteristics by increasing volumetric mass transfer coefficient up to 82% at smaller superficial gas velocities. At higher gas velocities, corresponding to aeration conditions operated in most cultivation, values of volumetric mass transfer coefficient were 20-30% higher with the insertion of impellers. The effective viscosity played a key role in defining the magnitude of the impellers' effect on the improvement of volumetric mass transfer coefficient. Superficial gas velocity and sparger design influenced impellers efficiency as well. The highest improvements of volumetric mass transfer coefficient were achieved in the most viscous carboxylmethyl-cellulose solution by using the least effective type of sparger. Besides proposed empirical correlations, which gave an average relative error of 9.1%, an artificial neural network was also successfully developed to estimate the volumetric mass transfer coefficient. The results showed that neural network model was able to predict volumetric mass transfer coefficient with an average relative error of 4.8%. (C) 2016 Elsevier B.V. All rights reserved.
    View publication