Browse Prior Art Database

GETTING THE MOST OUT OF PACKED DISTILLATION COLUMNS

IP.com Disclosure Number: IPCOM000019406D
Publication Date: 2003-Sep-12
Document File: 6 page(s) / 44K

Publishing Venue

The IP.com Prior Art Database

Abstract

The design of packed distillation columns has to take into account the effects of maldistribution on the overall performance. Various aspects of this phenomenon have been discussed in textbooks, handbooks and review articles [1-4]. The current article summarizes the key findings and provides some guidelines on how to get the most out of packed distillation columns.

This text was extracted from a Microsoft Word document.
This is the abbreviated version, containing approximately 16% of the total text.

GETTING THE MOST OUT OF PACKED DISTILLATION COLUMNS

The design of packed distillation columns has to take into account the effects of maldistribution on the overall performance. Various aspects of this phenomenon have been discussed in textbooks, handbooks and review articles [1-4]. The current article summarizes the key findings and provides some guidelines on how to get the most out of packed distillation columns.

The earliest publication to recognize the importance of liquid channeling [5] shows the effect on a system with a relative volatility of 1.07 containing 10 – 100 stages for two different compositions at the top. It can be seen that larger losses of column efficiency occur when one increases the number of theoretical stages or the fraction of the less volatile component or both at equivalent levels of channeling. Although only random packing was in use at the time of this publication, it is an elementary theoretical calculation on the effect of liquid bypass without mass transfer and it should apply equally well to packed columns with structured packing. Similar calculations can be done for other systems to show that the bottoms of distillation columns are particularly vulnerable to liquid channeling when they contain many theoretical stages.

Major packing vendors like Norton have shown that taller random packed towers suffer proportionately larger losses in efficiency with a reduction in inlet liquid distribution quality in comparison with shorter beds [6]. Moore and Rukovena have suggested a valuable graphical method to evaluate the quality of distribution [7]. These publications have been cited in the review literature; for instance see figure 9.6 [1] and figure 8 [4]. Based on such work it is a common practice in the industry to limit the number of theoretical stages when the separation to be performed is considered to be sensitive to maldistribution. Kister (page 551) [1] specifically teaches that columns containing less than five theoretical stages are relatively insensitive while those with ten or more stages are more sensitive to maldistribution. While much of this is based on random packed towers, the general conclusions are also applicable to columns that use structured packing.

The developments that have taken place in the last four decades have led to an understanding of the implications of maldistribution beyond just liquid channeling and the quality of inlet liquid distribution. As summarized in reference [1], these include composition pinching effects due to local changes in the L/V ratio, a lateral mixing effect which counteracts local pinches, and liquid flow unevenness which develops within the packing which itself is a complex phenomenon that is strongly influenced by wall flow. The work from some key references is described below.

The concept of a “natural distribution” in packing was introduced by Albright in 1984 [8], according to which any initial distribution better than such natural distribution will...