Extending Mass Transfer Correlations For Packed Three-Phase Distillation Simulation: A Hierarchical Parameter Estimation Methodology

Liang Chen,Shuqing Wang,Jens-Uwe Repke,Günter Wozny
DOI: https://doi.org/10.1016/S1570-7946(09)70294-9
2009-01-01
Abstract:Simulation of three-phase (vapor-liquid-liquid) distillation is still lacking of reliable mass transfer correlations to describe the rate-based (or nonequilibrium, NEQ) mass transfer process. The objective of this paper is to extend existing mass transfer correlations for three-phase distillation simulation in packed column. The transfer parameters in these correlations are classified into two categories: packing-related parameters and flow-related parameters, which are estimated using a (double-layer) hierarchical methodology based on our comprehensive experimental researches. The experimental database consists of 162 different data sets (both two-phase and three-hase distillations) of n-Butanol/Water/n-Propanol system for three different packings (Montzpack B1-350, Raschig Super-Ring 0.3, Rombopak 9M). The estimation problem with multi-data-set and multi-packing turns out to be a large-scale optimization subjected to a large number of nonlinear model equations. Three widely used correlations in packed column (i.e. Billet & Schultes correlation, Shi & Mersmann correlation, Rocha correlation) are adjusted for rate-based mass transfer calculation of three-phase distillation. Substantial improvements on the simulation predictive ability are observed by using the NEQ model with the adjusted correlations.
What problem does this paper attempt to address?