(6f) New Energy-Efficient and Low-Cost Multicomponent Distillation Configurations | AIChE

(6f) New Energy-Efficient and Low-Cost Multicomponent Distillation Configurations


Giridhar, A. V. - Presenter, Purdue University
Agrawal, R. - Presenter, Purdue University

The synthesis of efficient multicomponent separation networks has posed computational challenges in process design, with distillation networks having the greatest economic impact. One reason for the computational difficulty is that there are a combinatorially large number of possible configurations for a given multicomponent feed. Most configurations designed and built in the last 40 years have occupied a very small fraction of this large search space. Recent work in the literature has shown the energy efficiency of some new configurations, which were developed through a process of discovery, which have not been included in traditional sets of configurations. It is desirable to build a systematic framework that embeds a ?complete? set of configurations, so that advances in energy efficiency are not reliant on discovery processes. In this work, we introduce a novel framework to embed all possible distillation configurations for a given feed subject to the constraint that they use the minimum number of columns (i.e., n-1 columns for an n-component feed). We believe that this framework and the associated method are powerful enough to provide energy-efficient and cost-effective designs of distillation networks when used on a computer, and simple enough to enable humans to verify the quality of the design by hand in reasonable time. We accomplish this by introducing the concept of a ?supernetwork?, which considers only configurations with the minimum number of separation units, both with and without thermal linking. Several of these configurations have not been previously described in the literature. Our detailed comparison of the heat demand of such configurations with traditional sharp-split configurations for dozens of feed configurations has shown savings potential of up to 45%. We also evaluate different optimization methods to search the novel configuration space and optimize the heat demand with respect to composition and flow rates.