Study of dynamic performance of heat-integrated distillation columns considering the effect of relative volatility of the mixtures

Abstract

In this paper the dynamic performance of Heat-Integrated Distillation Columns (HIDiC) is presented. The dynamic performance was determined for the optimal HIDiC designs optimized previously using a constrained Boltzmann-based estimation of distribution algorithm. Eight close-boiling mixtures, covering a range of relative volatility (α) from 1.12 to 2.4, were used as case studies. The dynamic behavior was obtained under open and closed-loop process analysis. The results obtained showed that the HIDiC columns undergo worse dynamic performance than their equivalent traditional columns for all case studies. Furthermore, it was notorious that the difference in the dynamic behavior of both configurations kept a relatively uniform trend for most systems. However, a marked difference in the dynamic properties was determined for the mixture close to azeotropic behavior, which experienced a considerably larger control effort.Thus, the novel findings disclosed in this paper show that, although the HIDiC sequences experienced worse dynamics than the conventional columns, the HIDiC configurations reached a stable dynamic behavior for all the range of α of the mixtures under study. Nevertheless, high control difficulties were particularly determined for the HIDiC configuration used to separate the mixture with α close to unity (mixture of xylenes), which in fact is the HIDiC sequence with the best energetic and economic benefits.Hence, the energetic and economic potential of the HIDiC columns is not limited by the dynamic behavior for most case studies analyzed, at least in theoretical terms. However, such potential might be particularly reduced or unharnessed in the separation of the mixture with α near the azeotropic behavior due to the dynamics of the HIDiC column for this separation.So, the findings presented in this paper allow to infer that sequential or simultaneous dynamic studies must be achieved along with the optimization of the HIDiC columns to determine the integral performance (energetic, economic and dynamic) of these configurations.