WIT Press


Multiple Plunging Jet Aeration System And Parameter Modelling By Neural Network And Support Vector Machines

Price

Free (open access)

Paper DOI

10.2495/WP060581

Volume

95

Pages

10

Published

2006

Size

482 kb

Author(s)

S. Deswal, D. V. S. Verma & M. Pal

Abstract

Plunging jet aeration systems provide a simple and inexpensive method of supplying oxygen for wastewater treatment. Though numerous studies have been reported on aeration with a single plunging jet, very few studies are available in open literature on multiple plunging jet aeration systems. The present work is the result of an extensive laboratory study carried out on single and multiple water jets in vertical and inclined orientations, and different possible configurations of the number and diameter of jets. The effect of single and multiple jets on overall volumetric mass transfer coefficient (KLa) is studied by correlating it with kinetic jet power per unit volume (P/V). It was found that KLa increases with an increase in the number of jets, and also the performance of multiple jets improves with increasing kinetic jet power per unit volume. Neural Network and Support Vector Machine modelling techniques have been applied on the experimental data to determine the significant jet parameters that govern the performance of multiple plunging jets aeration system. Based on these parameters, an empirical relationship for predicting the KLa has been proposed for which the correlation coefficient and root mean square error are 0.96 and 3.27 respectively. Both the modelling techniques have also been used for the prediction of KLa and have been found to work well. The findings of these modelling techniques and empirical relationship are expected to be quite useful in the development of an efficient multiple plunging jets aeration system. Keywords: aeration, multiple plunging jets, overall volumetric mass transfer coefficient, kinetic jet power neural network, support vector machines.

Keywords

aeration, multiple plunging jets, overall volumetric mass transfer coefficient, kinetic jet power neural network, support vector machines.