Distillation has a high potential for energy savings accounting for around 40-70% of capital and operating costs in petrochemical and commodity manufacturing. Recently, complex heat-integrated column configurations have been found to have better energy efficiency than simple separation trains. Therefore, we present a computer-aided method to systematically design complex column networks to separate multi-component mixtures with lowest energy demand. The proposed computational method exploits a thermodynamically motivated problem transformation entitled temperature collocation. This method drastically reduced the search space for separation synthesis and produces rigorous column profiles equivalent to solutions of the mass, equilibrium, summation and energy (MESH) equations. The method also traverses safely pinched regions in which many existing techniques are known to fail. The methodology applies to any type of vapor-liquid equilibrium model including ideal, non-ideal and azeotropic mixtures without limitations in the number of components.
We have addressed the problem of synthesizing complex column configurations. Feasible designs can be identified with our design algorithms and optimal configurations are identified in terms of energy efficiency or a combination of capital or energy cost. We will demonstrate case studies for the separation of up to ten components as well as the systematic design of thermally coupled side stripper and rectifier columns. We will discuss an exhaustive search for complex separation networks for a quaternary separation problem. These robust synthesis algorithms will be shown to automatically synthesize a distillation network for given product purity requirements. Entire separation flowsheets are generated with rigorous thermodynamic models without the need to introduce limiting simplifications as is the case with existing shortcut techniques. The computer procedure robustly converges to the desired purity targets, unless the desired target is thermodynamically impossible to realize. The computational approach guarantees to identify realizable columns with a finite number of trays and operating conditions. All results shown in the presentations are validated rigorously with commercial flowsheet simulator software packages AspenPlus