Gas velocity, m/s

FIG. 14-124 Effect of gas and liquid rates on onset of mesh reentrainment and safe operating regions. To convert meters per second to feet per second, multiply by 3.281. (Calvert, Yung, and Leung, NTIS Publ. PB-248050, 1975.)

FIG. 14-124 Effect of gas and liquid rates on onset of mesh reentrainment and safe operating regions. To convert meters per second to feet per second, multiply by 3.281. (Calvert, Yung, and Leung, NTIS Publ. PB-248050, 1975.)

Superficial gas velocity, m/s (b)

FIG. 14-125 Experimental pressure measured by Calvert as a function of gas velocity and liquid loading for (a) horizontal gas flow through vertical mesh and (b) gas upflow through horizontal mesh. Mesh thickness was 10 cm with 2.8-mm wire and void fraction of 98.2 percent, crimped in alternating directions. To convert meters per second to feet per second, multiply by 3.281; to convert centimeters to inches, multiply by 0.394. (Calvert, Yung, and Leung, NTIS Publ. PB-248050, 1975.)

Superficial gas velocity, m/s (b)

FIG. 14-125 Experimental pressure measured by Calvert as a function of gas velocity and liquid loading for (a) horizontal gas flow through vertical mesh and (b) gas upflow through horizontal mesh. Mesh thickness was 10 cm with 2.8-mm wire and void fraction of 98.2 percent, crimped in alternating directions. To convert meters per second to feet per second, multiply by 3.281; to convert centimeters to inches, multiply by 0.394. (Calvert, Yung, and Leung, NTIS Publ. PB-248050, 1975.)

Table 20-41 [Chemical Engineers' Handbook, 5th ed.)], showing the minimum size of particles collectible in different types of scrubbers at reasonably high efficiencies, is a good selection guide. Cyclonic spray towers can effectively remove liquid particles down to around 2 to 3 |m. Figures 20-112 and 20-113 (Chemical Engineers' Handbook, 5th ed.), giving target efficiency between spray drop size and particle size as calculated by Stairmand or Johnstone and Roberts, should be considered in selecting spray atomization for the most efficient tower operation. Figure 14-126 gives calculated particle cut size as a function of tower height (or length) for vertical countercurrent spray towers and for horizontal-gas-flow, vertical-liquid-flow cross-current spray towers with parameters for liquid drop size. These curves are based on physical properties of standard air and water and should be used under conditions in which these are reasonable approximations. Lack of uniform liquid distribution or liquid flowing down the walls can affect the performance, requiring empirical correction factors. Calvert (R-10) suggests that a correction factor of 0.2 be used in small-diameter scrubbers to account for the liquid on the walls, i.e., let QL/Qg = 0.2 (QL /Qg)aotual. Many more complicated wet scrubbers employ a combination of sprays or liquid atomization, cyclonic action, baffles, and targets. These combinations are not likely to be more efficient than similar devices previously discussed that operate at equivalent pressure drop. The vast majority of wet scrubbers operate at moderate pressure drop [8 to 15 cm (3 to 6 in) of water or 18 to 30 cm (7 to 12 in) of water] and cannot be expected to have high efficiency on particles smaller than 10 |m or 3 to 5 |m respectively. Fine and submicrometer particles can be captured efficiently only in wet scrubbers having high energy input such as venturi scrubbers, two-phase eductor scrubbers, and flux-force-condensation scrubbers.

Venturi Scrubbers One type of venturi scrubber is illustrated in Fig. 17-48. Venturi scrubbers have been used extensively for collecting fine and submicrometer solid particulate, condensing tars and mists, and mixtures of liquids and solids. To a lesser extent, they have also been used for simultaneous gas absorption, although Lundy [Ind. Eng. Chem., 50, 293 (1958)] indicates that they are generally limited to three transfer units. They have been used to collect submicrometer chemical incinerator fume and mist as well as sulfuric and phosphoric acid mists. The collection efficiency of a venturi scrubber is highly dependent on the throat velocity or pressure drop, the liquid-to-gas ratio, and the chemical nature of wettability of the particulate. Throat velocities may range from 60 to 150 m/s (200 to 500 ft/s). Liquid injection rates are typically 0.67 to 1.4 m3/1000 m3 of gas. A liquid rate of 1.0 m3 per 1000 m3 of gas is usually close to optimum, but liquid rates as high as 2.7 m3 (95 ft3) have been used. Efficiency improves with increased liquid rate but only at the expense of higher pressure drop and energy consumption. Pressure-drop predictions for a given efficiency are hazardous without determining the nature of the particu-late and the liquid-to-gas ratio. In general, particles coarser than 1 |m can be collected efficiently with pressure drops of 25 to 50 cm of water. For appreciable collection of submicrometer particles, pressure drops of 75 to 100 cm (30 to 40 in) of water are usually required. When particles are appreciably finer than 0.5 |m, pressure drops of 175 to 250 cm (70 to 100 in) of water have been used.

FIG. 14-126 Predicted spray-tower cut diameter as a function of sprayed length and spray droplet size for (a) vertical-coun-tercurrent towers and (b) horizontal-cross-flow towers per Calvert [J. Air Pollut. Control Assoc., 24, 929 (1974)]. Curve 1 is for 200-|m spray droplets, curve 2 for 500-|m spray, and curve 3 for 1000-|m spray. QL/QC is the volumetric liquid-to-gas ratio, L liquid/m3 gas, and uG is the superficial gas velocity in the tower. To convert liters per cubic meter to cubic feet per cubic foot, multiply by 10-3.

Sprayed horizontal length, Z, m (b)

FIG. 14-126 Predicted spray-tower cut diameter as a function of sprayed length and spray droplet size for (a) vertical-coun-tercurrent towers and (b) horizontal-cross-flow towers per Calvert [J. Air Pollut. Control Assoc., 24, 929 (1974)]. Curve 1 is for 200-|m spray droplets, curve 2 for 500-|m spray, and curve 3 for 1000-|m spray. QL/QC is the volumetric liquid-to-gas ratio, L liquid/m3 gas, and uG is the superficial gas velocity in the tower. To convert liters per cubic meter to cubic feet per cubic foot, multiply by 10-3.

FIG. 14-127 Prediction of venturi-scrubber cut diameter for hydrophobic particles as functions of operating parameters as measured by Calvert [Calvert, Goldshmid, Leith, and Mehta, NTIS Publ. PB-213016, 213017, 1972; and Calvert, J. Air Pollut. Control Assoc., 24, 929 (1974).] uG is the superficial throat velocity, and AP is the pressure drop from converging to diverging section. To convert meters per second to feet per second, multiply by 3.281; to convert liters per cubic meter to cubic feet per cubic foot, multiply by 10-3; and to convert centimeters to inches, multiply by 0.394.

FIG. 14-127 Prediction of venturi-scrubber cut diameter for hydrophobic particles as functions of operating parameters as measured by Calvert [Calvert, Goldshmid, Leith, and Mehta, NTIS Publ. PB-213016, 213017, 1972; and Calvert, J. Air Pollut. Control Assoc., 24, 929 (1974).] uG is the superficial throat velocity, and AP is the pressure drop from converging to diverging section. To convert meters per second to feet per second, multiply by 3.281; to convert liters per cubic meter to cubic feet per cubic foot, multiply by 10-3; and to convert centimeters to inches, multiply by 0.394.

One of the problems in predicting efficiency and required pressure drop of a venturi is the chemical nature or wettability of the particulate, which on 0.5-|m-size particles can make up to a threefold difference in required pressure drop for its efficient collection. Calvert (R-9, R-10) has represented this effect by an empirical factorf which is based on the hydrophobic (f = 0.25) or hydrophilic (f = 0.50) nature of the particles. Figure 14-127 gives the cut diameter of a venturi scrubber as a function of its operating parameters (throat velocity, pressure drop, and liquid-to-gas ratio) for hydrophobic particles. Figure 14-129 compares cut diameter as a function of pressure drop for an otherwise identically operating venturi on hydrophobic and hydrophilic particles. Calvert (R-9) gives equations which can be used for constructing cut-size curves similar to those of Fig. 14-127 for other values of the empirical factor f. Most real particles are neither completely hydrophobic nor completely hydrophilic but have f values lying between the two extremes. Phosphoric acid mist, on the basis of data of Brink and Contant [Ind. Eng. Chem., 50, 1157 (1958)] appears to have a value off = 0.46. Unfortunately, no chemical-test methods have yet been devised for determining appropriate f values for a par-ticulate in the laboratory.

Pressure drop in a venturi scrubber is controlled by throat velocity. While some venturis have fixed throats, many are designed with variable louvers to change throat dimensions and control performance for changes in gas flow. Pressure-drop equations have been developed by Calvert (R-13, R-14, R-15), Boll [Ind. Eng. Chem. Fundam., 12, 40 (1973)], and Hesketh [J. Air Pollut. Control Assoc., 24, 939 (1974)]. Hollands and Goel [Ind. Eng. Chem. Fundam., 14, 16 (1975)] have developed a generalized pressure-drop equation.

The Hesketh equation is empirical and is based upon a regression analysis of data from a number of industrial venturi scrubbers:

where AP is the pressure drop, in of water; Ugt is the gas velocity in the throat, ft/s; pg is the gas density, lb/ft3; At is the throat area, ft2; and L is the liquid-to-gas ratio, gal/1000 acf.

Calvert (R-15) critiqued the many pressure-drop equations and suggested the following simplified equation as accurate to ±10 percent:

2pu2

981gc

AP is the pressure drop, cm of water; pt and pg are the density of the scrubbing liquid and gas respectively, g/cm3; Ug is the velocity of the gas at the throat inlet, cm/s; Qt/Qg is the volumetric ratio of liquid to gas at the throat inlet, dimensionless; lt is the length of the throat, cm; C'a is the drag coefficient, dimensionless, for the mean liquid diameter, evaluated at the throat inlet; and di is the Sauter mean diameter, cm, for the atomized liquid. The atomized-liquid mean diameter must be evaluated by the Nukiyama and Tanasawa [Trans. Soc. Mech Eng. (Japan), 4, 5, 6 (1937-1940)] equation:

where at is the liquid surface tension, dyn/cm; and |t is the liquid viscosity; P. The drag coefficient Ca should be evaluated by the Dickinson and Marshall [Am. Inst. Chem. Eng. J., 14,541 (1968)] correlation Ca = 0.22 + (24/NRei)(1 + 0.15 NR6) The Reynolds number, NRei, is evaluated at the throat inlet considerations as dtGg/| g.

All venturi scrubbers must be followed by an entrainment collector for the liquid spray. These collectors are usually centrifugal and will have an additional pressure drop of several centimeters of water, which must be added to that of the venturi itself.

Other Scrubbers A liquid-ejector venturi (Fig. 17-49), in which high-pressure water from a jet induces the flow of gas, has been used to collect mist particles in the 1- to 2-| m range, but sub-micrometer particles will generally pass through an eductor. Power costs for liquid pumping are high if appreciable motive force must be imparted to the gas because jet-pump efficiency is usually less than 10 percent. Harris [Chem. Eng. Prog., 42(4), 55 (1966)] has described their application. Two-phase eductors have been considerably more successful on capture of submicrometer mist particles and could be attractive in situations in which large quantities of waste thermal energy are available. However, the equivalent energy consumption is equal to that required for high-energy venturi scrubbers, and such devices are likely to be no more attractive than ven-turi scrubbers when the thermal energy is priced at its proper value. Sparks [J. Air Pollut. Control Assoc., 24, 958 (1974)] has discussed steam ejectors giving 99 percent collection of particles 0.3 to 10 | m. Energy requirements were 311,000 J/m3(8.25 Btu/scf). Gardenier [J. Air Pollut. Control Assoc., 24, 954 (1974)] operated a liquid eductor with high-pressure (6900- to 27,600-kPa) (1000- to 4000-lbf/in2) hot water heated to 200°C (392°F) which flashed into two phases as it issued from the jet. He obtained 95 to 99 percent collection of submicrometer particulate. Figure 14-128 shows the water-to-gas ratio required as a function of particle size to achieve 99 percent collection.

Effect of Gas Saturation in Scrubbing If hot unsaturated gas is introduced into a wet scrubber, spray particles will evaporate to cool and saturate the gas. The evaporating liquid molecules moving away from the target droplets will repel particles which might collide with them. This results in the forces of diffusiophoresis opposing particle collection. Semrau and Witham (Air Pollut. Control Assoc. Prepr. 7530.1) investigated temperature parameters in wet scrubbing and found a definite decrease in the efficiency of evaporative scrubbers and an enhancement of efficiency when a hot saturated gas is scrubbed with cold water rather than recirculated hot water. Little improvement was experienced in cooling a hot saturated gas below a 50°C dew point.

Energy Requirements for Inertial-Impaction Efficiency

Semrau [J. Air Pollut. Control Assoc., 13, 587 (1963)] proposed a "contacting-power" principle which states that the collecting efficiency of a given size of particle is proportional to the power expended and that the smaller the particle, the greater the power required.

Was this article helpful?

## Post a comment