Flow pattern identification for three‐phase distillation with a nonequilibrium modeling

The fear of the uncertain vapor–liquid–liquid hydrodynamics has guided process engineers away from a three‐phase involvement, of which the most uncertain is the flow pattern of the second liquid phase. Hence, a first‐principle model‐based estimation strategy was proposed to identify the flow pattern in the three‐phase distillation. Representative experiments were performed in a laboratory‐sclae packed column, and the results were sampled as reference data for the estimation. A dynamic nonequilibrium (NEQ) model was built up to consider the inter‐phase mass transfer rigorously. The resulting estimation problem was then discretized using the multiple‐shooting method, to formulate a large‐scale NLP problem with a finite number of decision variables. The estimated results reveal that the flow pattern of the second liquid has a close relationship with the liquid and gas loads, and it could play a critical role for the variation of the separation efficiency inside a three‐phase column. Copyright © 2011 Curtin University of Technology and John Wiley & Sons, Ltd.