Oct-2010

Improvements in random packing performance

Developments in random packing design aim for increased efficiency and capacity. Random packing is used extensively in gas processing and high-pressure distillation applications because of the distinct benefits it offers.

Izak Nieuwoudt, Christine Corio and Jeff DeGarmo
Koch-Glitsch

Article Summary

The high liquid rates in these applications, and the high vapour-to-liquid density ratios in high-pressure distillation, generally exclude the use of structured packing. Plus, pressure drop and cost considerations count against the use of trays in several of these applications.

Random packing has a long history and had an interesting evolution. Glass balls were used as random packing as early as the 1820s and, by the 1850s, these were replaced with pumice stone or pieces of coke. In the 1880s, ceramic balls were used as random packing in azeotropic distillation towers and, by the start of the 1900s, were replaced with hollow ceramic and metal balls. Raschig realised the importance of increasing the void fraction of the packing and having its internal surface area take part in mass transfer. In 1914, he patented the use of thin-walled metal cylinders as random packing.¹

It was soon realised that the Raschig ring could be improved by perforating the cylinder wall to improve liquid and vapour communication between the inside and outside of the packing element. From the 1920s to 1960s, several versions of this packing were introduced.^2,3,4 The 1962 version became known as the Pall ring.⁴ The Cascade Mini-ring was introduced in 19715 as an extension to the Pall ring idea. This packing was a perforated cylinder with a lower height-to-diameter ratio. In 1977, IMTP random packing was developed.⁶ This represented a change in direction from the ring-shaped structure, being a saddle-shaped element. In 1991, Fleximax saddle-shaped random packing⁷ was introduced. The Beta Ring, from 1992, represented a return to the ring-shaped structure,⁸ and the wavy-shaped Raschig Super Ring was introduced in 1995.⁹

Through all these developments, IMTP cemented its place as the premier high-performance random packing. In evaluating the performance data of all of these packing types, it became evident that the refiner had to make a choice when it came to revamping a tower: give up some efficiency for extra capacity or give up some capacity to get extra efficiency. But why can we not have both?

Random packing characteristics
The Pall ring was a significant advance because it opened up the interior surfaces of the ring to vapour and liquid flow compared to the earlier Raschig ring. A modified Pall ring with a reduced height-to-diameter ratio was later introduced as the CMR ring. The CMR ring was believed to orient itself preferentially in the packed bed in a manner that reduces pressure drop. IMTP packing elements combine the increased surface area advantages of the Super Intalox Saddle with the interior projecting fingers of the Pall ring. Fifteen years ago, the Raschig Super Ring was introduced. It is claimed that this packing promotes the spreading of liquid films as opposed to a combination of films and droplets. The primary aim of a random packing element is to generate a surface area where liquid and vapour can be contacted. It is important that the liquid and vapour are brought into contact, as elements that are shielded and only see vapour or liquid do not contribute to the mass transfer process. The packing element must also be shaped in a way that the flow path of the vapour is not too tortuous, since this will increase the pressure drop. Upon careful investigation, it became clear that all of the aforementioned packing elements had some of these shortcomings. Koch-Glitsch set out to develop a random packing that would overcome these shortcomings and give users improved capacity and efficiency.

Development of an improvedrandom packing
The considerations outlined above led to the development of the Intalox Ultra.10 Figure 1 shows an element of the packing.

Intalox Ultra random packing comprises a pair of curved side strips with inner and outer arched ribs extending from and between the side strips. The ribs are not all the same shape and some are discontinuous. The side strips are typically flanged to provide strength. Several other strengthening features are built into the element to give it a high strength-to-weight ratio. The packing element is shaped in a way that discourages nesting of one packing element with another. Nesting reduces mass transfer efficiency and can promote liquid and vapour channelling within the packed bed. The ribs of this packing are oriented in space to create an even distribution of surface area in the volume occupied by the packing element. The random packing has open, easily accessible fluid flow paths that facilitate low pressure drop and good vapour-liquid contact, and presents a relatively uniform surface area distribution when viewed at multiple angles, which yields orientation-independent performance.

Random packing performance
Since IMTP random packing has had a very large installed base, it is important to know how Intalox Ultra performance stacks up. Figures 2–5 show the pressure drop and height equivalent to a theoretical plate (HETP) of Intalox Ultra A packing under total reflux conditions in a system comprising light hydrocarbon isomers. Intalox Ultra A is a random packing of nominally 1.5in size. For comparison, data are also shown for IMTP 40 and IMTP 50. The IMTP packing data were obtained in the same pilot plant under the same conditions as the Intalox Ultra A data. The vapour density was 1.06 lb/ft3 and the liquid density 33.6 lb/ft3. The tower diameter is 15.3in and the bed height was approximately 10ft.

Figure 2 shows the measured efficiency (HETP) as the tower throughput is increased while being operated at total reflux (L/V=1). Figure 2 shows one advantage of the new packing: it has almost the same capacity as IMTP 50 and yet has a significantly lower HETP, almost rivaling that of IMTP 40 packing. Figure 3 is a plot of the measured pressure drop as a function of tower throughput while being operated at total reflux (L/V=1). It shows that the capacity and pressure drop of Intalox Ultra A are close to that of IMTP 50 and significantly better than that of IMTP 40.
The performance of the nominally 2in-sized Intalox Ultra L random packing is shown in Figures 4–5. Figure 4 shows that the HETP of Intalox Ultra L is the same as that of IMTP 50, yet its capacity is significantly higher. From Figure 5, it can be seen that Intalox Ultra L exhibits lower pressure drop and higher capacity than IMTP 50. The pressure drop and capacity of Intalox Ultra L rival those of IMTP over the number 50.

From these performance graphs, it is evident that Intalox Ultra can be used to improve throughput or product purity in the case of a revamp. In the case of a new tower, it can be used to shrink the size of the vessel.