Avoid Problems with Random Packing

Careful internals' selection is crucial for absorbers and scrubbers.

By Dirk Willard, contributing editor

Share Print Related RSS

A trough seemed like a good idea at the time. Using it in the acid absorber would allow a 5:1 turndown, compared to the 3:1 of spray nozzles. The trough was trouble from the beginning. We couldn't keep it flat, so water spilled out unevenly. Debris lodged in the slats at the bottom caused uneven liquid distribution. Eventually, the plant manager got rid of my "bright" idea.

A good column design leads to a smooth startup.


Designing random-packed towers still is an art. So, let me share a few insights, some learned the hard way.

The key components in any packed tower are: 1) the vapor dis-engagement space; 2) the feed distribution system; 3) the hold-down; 4) the supports; 5) the liquid distributors; 6) the vapor inlet; and 7) the sump and liquid re-distributor.

Let's start with the vapor dis-engagement space, i.e., the area above the spray nozzle(s). Install a demister. Chevron demisters seldom are suitable; instead use a 12-in. mesh pad of layered coarse/medium/fine strips and introduce fresh feed at 0.5 gal/min/ft2, spraying upward at the pad bottom. Mesh pads can foul in a few months if the feed is dirty. Make sure to anchor pads with plastic rope to prevent bypass. Leave about one-quarter of the vessel diameter for disengagement between the centerline of the horizontal exit and the top of the mesh pad.

As far as feed distribution, one spray nozzle always is best — it obviates maldistribution caused by pressure differences among multiple nozzles. If one nozzle isn't viable (e.g., because of the space needed for vapor disengagement), then lay out the piping to control pressure at each individual nozzle. A full-cone 90° nozzle works best. Ask the vendor for the droplet distribution expected for about 5-psig drop, the low end of the operating curve, and at what pressure 15% of the droplets will be less than 15-microns in size. Leave enough space for vapor disengagement above the nozzle: the same vertical space needed for the spray to hit the vessel wall.

The hold-down, or bed-limiter, need only be webbing. A scrubber should have fairly low gas superficial velocities: ~1,500 lb/hr/ft2. Absorbers require tighter control — one I designed used 1,300 lb/hr/ft2. The relationship between the liquid and gas superficial velocities is reversed in an absorber with a higher gas rate; in a scrubber the liquid rate always is higher. I usually design for a 7-gal/min/ft2 irrigation rate and try to specify the spray nozzle and pump for at least 15 gal/min/ft2 or the tower flooding capacity. You'll need the extra flow to account for poor column performance or changes in the packing. Absorber flow requires careful regulation — too low and acid absorption creates enough heat to burn and melt packing.

The internals — packing, gas distributor and, most importantly, liquid redistributor — are key to good tower performance. Specify as small a packing as possible. I prefer a gas-injection-type gas distributor with a layer of large packing over the top. The liquid redistributors should carry no more than 14 ft of packing, preferably less, if possible. I recommend a full redistributor because it serves both as packing support and redistributor. For fouling resistance, avoid weir-trough and orifice-type units.

Sump capacity should suffice to allow at least 10 minutes for a drop from the high alarm, typically 90%, to a low alarm. I set the low-low alarm at the point when the recirculation pump cavitates. High-high is at the bottom of the overflow at 98%.

Now, let's turn to the random packing itself. Modern packing maximizes the contact between liquid and vapor but often sacrifices mechanical integrity by using a tendril drip lattice design. In a study of seven plants, I showed how a 3-in. packing failed to perform because of collapse of these delicate lattices; I recommended smaller and sturdier packing. Plastic packing, the preferred choice in corrosive environments, is limited to below 400°F for fluoropolymers, 180°F for polypropylene (PP), and 230°F for 15%-glass-filled PP.

Finally, let's briefly consider the height of a transfer unit (HTU). Because random packing can't compete against trays in efficiency, it's best to go with a simple definition (p. 598 in the 4 ed. of Coulson and Richardson's "Chemical Engineering Design," vol. 6): for 1-in. packing, use an HTU of 2 ft; for 1½-in., use 2½ ft; and for 2-in., use 3 ft. For larger than 2-in. packing, use 3 ft, and for packing less than 1-in., use 1 ft.

Packed towers offer an inexpensive alternative to trayed columns if cleverly designed.

 



DIRK WILLARD is a Chemical Processing Contributing Editor. You can e-mail him at dwillard@putman.net.

Share Print Reprints Permissions

What are your comments?

You cannot post comments until you have logged in. Login Here.

Comments

No one has commented on this page yet.

RSS feed for comments on this page | RSS feed for all comments