Estimation of heat transfer coefficient in bubble column reactors using support vector regression

The objective of this study was to develop a unified correlation for heat transfer coefficient (h) in bubble columns without internals for various gas–liquid systems using support vector regression (SVR)-based modeling technique. From the data published in open literature, 366 data points from 10 open sources spanning the years 1976–2007 for h were collected. Generalized SVR-based models have been developed for the relationship between h and prominent design and operating parameters such as column and sparger geometry, gas–liquid thermo-physical properties, operating temperature, pressure, superficial gas velocity, etc. Further, this model for h has been uploaded on the link http://www.esnips.com/web/UICT-NCL. The proposed generalized SVR-based correlations for h has prediction accuracies of 98.56% and average absolute relative error (AARE) of 7.05%. Also, the SVR-based correlation showed much improved predictions when compared with those estimated by empirical correlations in the literature.