Packed Absorption Tower: Packing Material, Bed Sizing, Internals, Liquids, Operation

Wet devices for soluble contaminants differ mainly in how much fan energy they consume. The US EPA classes them by static loss for particulate collection: under 5 in. w.c. is low energy, 5 to 25 in. w.c. medium, and 25 in. w.c. and above high energy, which is where submicron dust capture begins. On gas duty it still tracks operating cost. Fixed packing sits at the low end and earns its place on vapor duty rather than on dust.

Sizing starts with six numbers, and each has to come from measurement rather than from a nameplate: actual volumetric flow in ACFM at real inlet conditions, concentration of the target compound, the outlet figure written into the permit, stream temperature, moisture content and dust loading. Typical applications sit between 500 and 75,000 scfm, at entering concentrations of 250 to 10,000 ppmv and inlet temperatures of 40 to 100 °F on vapor duty. Above that band solvent evaporation rises and mass transfer falls, so a quench stage ahead of the shell usually pays for itself, though whether it is strictly required depends on the solvent.

Two inputs cause most of the rework. Understate the entering concentration and the bed comes up short at the stack; understate temperature and the shell material is wrong before fabrication starts. Both surface after commissioning, when correction means new hardware rather than a setpoint change.

Solvent choice follows the chemistry. Water alone handles compounds with high solubility, among them ammonia, formaldehyde and light alcohols. For everything else there are two routes. One is a reagent that converts the dissolved species into a salt, which keeps the driving force high up the shell. The other is a solvent the compound is genuinely soluble in: mineral oil recovering hydrocarbon vapor for later stripping is the standard non-aqueous case. The packed tower absorption liquid common to nearly every acid-gas installation in the United States is caustic soda, with sodium carbonate and slaked lime filling the same role in specific plants.

The control window in that table describes the liquor in the sump rather than the reagent as delivered. A 10 to 20 percent caustic solution sits near pH 14 in the drum, and the dosing pump stands between the two.

Reagent use follows stoichiometry, blowdown follows solids. Bleed-off on acid duty starts at 10 to 15 percent solids, past which salt carries over into the outlet stream, and the wasted fraction runs anywhere from 0.1 to 10 percent of total throughput depending on the salt formed. Size the recirculation vessel of a packed gas absorption tower for that bleed rate, the make-up line and the dosing point, and not merely for pump suction.

Read that table as a trade. One-inch ceramic rings deliver 58 ft² of surface per cubic foot at Fp 160; two-inch plastic Tri-Packs deliver 48 ft² at Fp 16. Nearly the same contact area from an element an order of magnitude lower in Fp, pointing to a much lower pressure drop at the same duty. How much lower is a calculation rather than a ratio: resistance depends on gas and liquid loading as well as on the element, so the number to use comes off the supplier's curves at the design point.

The choice of absorption tower packing material is a chemistry and temperature problem before it is an efficiency problem. Carbon steel and standard stainless elements are ruled out by strongly corrosive acid streams, which is why EPA's manual steers acid duty away from metal. High-nickel alloys, titanium and zirconium do the work, at a price that usually settles the argument in favor of plastic or ceramic. Polymer elements are unsuitable above their softening range. Both plastic and metal are limited to an unsupported depth of 20 to 25 feet before the weight of the stack deforms the lower layers, which is why tall beds are split into sections.

Ceramic and glass sit outside that split. They tolerate most hot mineral acids, strong oxidizers and solvents that destroy polymers, and brittleness is the price. The practical reason an absorption tower is packed with glass tubing in bench and pilot work is that glass presents a chemically neutral, readily wetted surface across whatever chemistry a trial throws at it, and it leaves the view through a transparent shell clear while the operator judges wetting and the onset of flooding. Silicate bodies do have exclusions. Hydrofluoric acid and soluble fluorides attack them outright, and borosilicate glass is also taken apart by hot concentrated phosphoric acid and by hot strong caustic; PVDF, PTFE or polypropylene take over in those services.

Element geometry drives packing tower absorption performance more than any other item on the data sheet, and our catalog splits the choice explicitly. TORNADO RP carries random elements — Raschig, Pall, Intalox and Torch-Air rings — at 600 to 30,000 cfm per module. TORNADO SP carries structured blocks over the same range, needs less of them for the same outlet figure, runs at lower resistance and resists plugging better, and it is specified for low entering dust, which structured media needs to keep its channels open. Where the stream carries dust and vapor together, TORNADO FB floats low-density spheres in the irrigation flow, and a mobile arrangement of that kind is markedly less prone to blinding because the elements never stop moving. A packed column in gas absorption tower service can be built any of those three ways, and the dust figure usually decides which.

The design of packed towers for absorption systems then follows a fixed sequence: diameter, solvent rate, depth of packing, resistance, and a check that the unit still works at minimum flow.

Diameter comes from Eckert's modification of the generalized correlation, which returns superficial gas mass velocity at flooding for a given Fp and liquid-to-gas ratio. Practice backs away from that value: the flooding factor runs 0.60 to 0.75, and most installed units operate at 60 to 70 percent of flooding velocity. One sanity check follows immediately: shell diameter should be at least fifteen times the nominal element size, or wall channelling dominates and a smaller element has to be selected.

Irrigation rate is bounded from below by wetting rather than by stoichiometry. The minimum wetting rate is 0.85 ft²/hr for rings above three inches and for structured grids, and 1.3 ft²/hr for everything else; multiply it by solvent density and by the specific area of the media to get the minimum superficial liquid rate in lb/hr·ft². Suppliers quote 2 to 20 gpm per 1,000 cfm as the working band on acid duty, and with highly soluble compounds pump size follows the wetting minimum ahead of the mass balance.

Depth is the product of the number of transfer units and the height of a transfer unit. A worked case from the EPA Air Pollution Control Cost Manual makes the arithmetic concrete: 22,288 acfm at 100 °F carrying 1,871 ppmv HCl, 99 percent removal, two-inch ceramic Raschig rings at Fp 65.

The last line covers the bed alone, being 0.83 in. w.c. multiplied by 10.3 ft of depth. Inlet and outlet losses, the mist eliminator and the connecting duct sit on top of it, and the fan is sized on the sum. A deep bed cut for 99 percent removal also runs well past the 2 to 8 in. w.c. band in the first table.

Overall packed absorption tower height in that case reaches 26.1 ft, from the correlation H = 1.40 × depth + 1.02 × diameter + 2.81, valid between 2 and 12 ft of diameter and 4 to 12 ft of media. The extra fifteen feet above the rings is inlet plenum, distributor space, mist eliminator and sump, and it decides whether the unit clears the building steel.
Turndown is where packed tower absorption quietly fails. At half design flow the irrigation rate often drops under the wetting minimum, dry channels open, and outlet concentration rises while the differential gauge reads lower than ever. Two-speed or variable-frequency pump control, sized against the wetting figure rather than flow, holds the bed wet across the range.

Standard procedure assumes a dilute single solute, isothermal behavior and a straight equilibrium line. The absorption of concentrated mixtures in packed towers breaks all three: heat of solution and heat of reaction raise solvent temperature along the height, equilibrium shifts the wrong way, and outlet concentration drifts up. Aqueous systems partly self-correct, since adiabatic saturation cools the stream as water evaporates, which is why cooling coils are rare on water duty and routine on amine and strong-acid duty. Multicomponent absorption in a packed tower is sized on the controlling species, normally the least soluble one, and the remaining components are verified against that geometry rather than designed for separately.

Most performance complaints trace to distribution rather than to the media. A bed works only over the area that stays wetted, and an element that never sees liquid is a bypass channel with no resistance.

Distributors split into gravity types, meaning orifice trays and weirs, and pressure types, meaning nozzles and perforated pipes. Nozzles are the most common and the most troublesome: they generate fine droplets that entrain into the outlet, they plug on suspended solids, and they need generous feed rates to compensate for uneven coverage. Orifice trays give better contact and tolerate fouling, at the cost of some resistance to upward flow and a strict leveling tolerance.

Redistribution is not optional on deep beds: published practice calls for a redistributor every 8 to 20 feet of random packing, because liquid migrates to the wall and leaves the core dry. The support plate under each section needs free area greater than the void fraction of the media above it; an undersized support is a frequent cause of flooding that begins below the packing and gets misdiagnosed as an overload.

Above the elements, high velocity can fluidize the top layer and carry pieces into the distributor. A restrainer bolted to the wall handles plastic elements; fragile ceramic usually gets a floating weighted plate that settles as the stack compacts. The mist eliminator sits above all of it, and its face velocity has to be checked against actual outlet flow rather than the design point, or re-entrainment cancels part of the capture just achieved.

The sump completes the loop. A packed tower absorption column with automated dosing needs the pH probe in the recirculation line rather than in still liquid, the make-up point away from the probe, and enough residence volume that a dosing pulse mixes before returning to the nozzles. Packed tower gas absorption is stable only when the solution chemistry is stable, and probe placement decides that.

Material is chosen against the wet chemistry, the peak temperature and the site environment, in that order.

Polymer shells cover most vent and tank duty, while steel extends the envelope to 570 °F where the chemistry allows it. Lined steel with acid-resistant brick remains the answer for hot mineral acid service. Relative to fiberglass, published cost factors run 0.50 to 0.90 for PVC, 0.80 to 1.10 for polypropylene and 1.10 to 1.75 for 304 stainless, which usually settles the plastic-versus-alloy argument early in a project.

An industrial packed absorption tower quotation should list the shell, gas inlet and outlet connections, liquid inlet and drain, distributor and redistributor, support plates, restrainer, media type and volume, mist eliminator, internal piping, sump volume, pump and fan duty, instrumentation, and platforms and ladders. Anything missing there becomes a change order. For a compact packed absorption tower on tank vent service the same list still applies, only smaller: our tank ventilation unit runs to 3,000 cfm at 2.5 in. w.c. with a 1/8 HP immersible pump and a plastic wetted path throughout.

Record clean differential at commissioning, with the media wetted and design flow established; that number is the reference for the rest of the equipment life. The design band is 0.5 to 1.0 in. w.c. per foot of media; a slow climb past it means scale or solids in the media, and a sudden drop means channelling, a collapsed support or a hole in the distributor.

Media life equals shell life under clean conditions and falls into the one-to-five-year range where corrosion, fouling and breakage are active. Scale control is a liquid-side job: conductivity-triggered blowdown, solids held under the 10 to 15 percent bleed point, and a wash cycle on the schedule instead of on the alarm. Sampling the packed tower absorption liquid for chlorides and suspended solids each month catches a materials problem long before the shell shows it.

Energy is a diagnostic too. In the worked case above, fan power computes to 32 kW and the recirculation pump to 4.4 kW at the design point; a fan creeping upward at constant flow is measuring fouling before the stack analyzer does. Annual inspection of a packed tower absorber covers distributor holes, media level and condition, mist eliminator, support plate, probe calibration and nozzle spray pattern.