Design of Dividing‐Wall Columns by Mixed‐Integer Nonlinear Programming Optimization. Issue 5 (5th April 2017)

• Record Type: Journal Article
• Authors: Franke, Meik B.
• Other Names: Fricke Armin guestEditor.; Schöneberger Jan guestEditor.
• Is Part Of: Chemie Ingenieur Technik. Volume 89:Issue 5(2017:May)
• Page Start: 582
• Page End: 597
• Publication Date: 2017-04-05
• Subjects: Dividing‐wall column -- Ethanol dehydration -- Extractive distillation -- Heterogeneous azeotropic distillation; Chemical engineering -- Patents -- Periodicals; Chemical engineering -- Periodicals; Chemical industry -- Periodicals; Chemistry, Technical -- Periodicals; 660.05
• DOI: 10.1002/cite.201700005
• Languages: English
• ISSNs: 0009-286X

Abstract

Abstract: Dividing‐wall columns (DWCs) offer an interesting alternative to conventional distillation towers. Savings of up to 30 % in both energy and capital cost are reported. While these high savings are realistic in some applications, DWCs offer less or no advantage at all in other applications. In any case, it is important to make a fair comparison between the conventional system and the DWC based on an optimized process design. Mixed‐integer nonlinear programming (MINLP), especially the presented modified Generalized Benders Decomposition approach, could be the tool to make this comparison, as shown here. Abstract : The design of dividing‐wall columns is a difficult task, because a large number of degrees of freedom has to be fixed. Mixed‐integer nonlinear programming (MINLP), especially the presented modified Generalized Benders Decomposition approach, is able to solve these demanding optimization problems, as shown by means of three industrial case studies.