Numerical Modeling of Dehydrogenation and Denitrogenation in Industrial Vacuum Tank Degassers

Abstract

Hydrogen and nitrogen are inescapable elements in all commercial steel products and the presence of dissolved hydrogen and nitrogen in liquid steel can cause various problems in most of the final products. In general, the reduction of these elements in liquid steel is required in most steelmaking companies, where vacuum treatment is typically applied to remove these impurities. The main focus of this thesis work has been put on investigating the dehydrogenation and/or denitrogenation behavior in a number of operational vacuum tank degassers (VTD) from different industrial plants. A literature review on various investigations and modeling techniques in the related field was firstly presented in this report. Based on the developed theories and methods that are relatively separate in the open literature, an integrated computational fluid dynamics (CFD) model was built to better understand the degassing process on an industrial scale and more importantly, for accurate predictions that are of considerable importance to industrial process operators. The CFD model consists of two sub-routines for calculating multiphase flows and species transportations, respectively. The commercial CFD package of ANSYS FLUENT was adopted and augmented by various user-defined functions. As for the multiphase sub-model, the standard k-epsilon equations were extended by adding extra source terms to consider the impact of gas injections on turbulence quantities. The sub-model was validated by using literature data for an aqueous system whose similarity represented one of the industrial VTDs studied in this work. With the extended k-epsilon equations, deviations from measured data of axial liquid velocity and turbulent kinetic energy were lower than 13 % and 18 % respectively, whereas the deviations were about 30 % and 85 % with the standard equations. For mass transfer calculations, two fundamental expressions that have been commonly employed to compute mass transfer coefficient in gas-liquid systems were assessed. Comparisons with process data showed that the eddy-cell correlation provides a better prediction under the studied conditions. The versatility of the CFD model was further demonstrated by performing extensive simulations to cover the effect of gas flow rate, initial element (i.e., hydrogen and nitrogen) content and steel compositions on final element content and degassing rate. For hydrogen removal, deviations from measured data in different industrial plants were ranged between 6 % and 14 % and for nitrogen removal, the deviations were generally lower than 13 %.