ENCYCLOPEDIA OF ENVIRONMENTAL SCIENCE AND ENGINEERING - VAPOR AND GASEOUS POLLUTANT FUNDAMENTALS pps

Control devices may broadly be classified according to the physi-cal separation process being used, adsorption: absorption: sig-extraction: distillation: or to the chemical process, nous

V

VAPOR AND GASEOUS POLLUTANT FUNDAMENTALS

The toxic gases produced during combustion and other

chemical processes may be removed by destructive

dis-posal, dispersive dilution or as recoverable side products

The removal path chosen is at the present time motivated

primarily by economics, but public pressure and

aware-ness of environmental problems also influence the choice

This section will concern itself with destructive

posal and/or various recovery processes, the subject of

dis-persion being ably handled in the sections on Air Pollution

Meteorology and Urban Air Pollution Modeling. The main

emphasis will be on principles of gaseous reaction and

removal with the description of equipment for air

pollu-tion abatement covered by pollutant Problems

specifi-cally concerned with the automobile can be found under

Mobile Source Pollution. Although the control principles

to be described below are general, it is usually necessary

to design equipment for each installation because of

varia-tions in physical and chemical properties of effluents;

also, in general, the cost of adding pollution devices to an

existing unit (retrofit) will be higher than if they were placed

in the original design, because of construction difficulty and

downtime

Although the majority of effluent material from

com-bustion occurs in the gaseous state, it is important to

char-acterize the total effluent stream for control purposes For

example, the effluent may be condensible at operating

temperature (a vapour) or noncondensible (a gas), but it

usually is a mixture of the two Particulate matter (solids)

and mists (liquids) are often suspended in the gaseous

stream; if the particles do not separate upon settling they

are called aerosols The considerations in this section deal

with gas or vapor removal only and not with liquid or solid

particle removal

SULFUR DIOXIDE, SO2, AND TRIOXIDE, SO3

Sulfur dioxide is generated during combustion of any

sulfur-containing fuel and is emitted by industrial processes that use

sulfuric acid or consume sulfur-containing raw material The major industrial sources of SO2 are sulfuric acid plants, smelt-ing of metallic ores, paper mills, and refining of oil Fuel com-bustion accounts for roughly 75% of the total SO2 emitted

Associated with utility growth is the continued long term increase in utility coal consumption from some 650 million tons/year in 1975 to between 1400 and 1800 million tons/year

in 1990 Also the utility industry is increasingly converting

to coal Under the current performance standards for power plants, national SO2 emissions are projected to increase approx-imately 15 to 16% between 1975 and 1990 (Anon 1978) The

SO2 emitted from power plants is usually at low concentration (0.5% by volume) However, a 900 MW unit will emit over 13,000 pounds of SO2 per hour for a 1% sulfur coal The SO2emitted from industrial processes is at higher concentrations and lower flow rates The emitted SO2 combines readily with mists and aerosols, thus compounding the removal problem

Information concerning emissions standards is essential

to pollution control engineering design The current US eral SO2 emissions limits for a stack are 1.2 lb/106 BTU for new oil and gas fired plants Also, uncontrolled SO2 emis-sions from new plants firing solid, liquid, and gaseous fuels are required to be reduced by 85% The percent reduction requirement does not apply if SO2 emissions into the atmo-sphere are less than 0.2 lb/106 BTU

fed-Flue gas desulfurization (FGD) methods are rized as nonregenerable and regenerable Nonregenerable processes produce a sludge that consists of fly ash, water, calcium sulfate and calcium sulfite In regenerable pro-cesses, SO2 is recovered and converted into marketable by-products such as elemental sulfur, sulfuric acid or con-centrated SO2 The sorbent is regenerated and recycled

catego-The US Environmental Protection Agency believes the following types of FGD systems are capable of achieving the emissions limit standards: lime, limestone, Wellman-Lord, magnesium oxide and double alkali Due to the pro-cess economics, utility industry prefer the lime/limestone systems Limestone processes constitute about 58% of the current calcium-based capacity in service and under con-struction, and 69% of that planned, which amounts to 63% of

the total (De Vitt et al., 1980) The following reactions take

place in the limestone systems:

CaCO (s) CaCO (aq)

2 CaCO (aq) 2H CaCHCO Ca

The dissolution rate of limestone depends on the pH

values The pH values encountered in practical operations

of limestone systems is in most cases between 5.5 and 6.5

If the limestone systems operate at too low pH, SO2 removal

efficiency will decrease At too high pH, the scale formation

will be promoted Other factors affecting the performance of

limestone systems include solids content, liquid-gas ratio,

and corrosion A discussion can be found elsewhere (De Vitt

2) chemical reagent used

3) end-product produced: saleable product or

dis-posable waste

The reagent, end product, principle of operation, and SO2

removal efficiency of major FGD processes are shown in

to the Stack gas cleaning sections.

Sulfur dioxide reacts slowly with a large excess of oxygen

in the presence of sunlight to form trioxide Gerhard (1956)

showed that the process occurs with O2 at a rate of 0.1–0.2%

per hour and Cadle (1956) with ozone, O3, at 0.1% per day

Niepenburg (1966) illustrates the effects of oxygen in the

waste gas during combustion of oil

The conversion of SO2 to SO3 is believed to be possible

at realistic rates because of the presence, on diverse surfaces,

of Fe2O3 which acts as a catalyst The SO3 has a short lifetime

since it readily combines with water vapor in the atmosphere

to form sulfuric acid

NOx is produced in all combustions which take place using air

as an oxygen supply and in those chemical industries ing nitric acid More than 55% of the total NOx emissions of

employ-20 million tons originate from stationary sources as shown in

from combustion of fossil fuels for utilities Direct related emissions account for only 5% of the stationary source total Approximately 30% of all stationary source NOx is emit-ted by coal-fired utility boilers Uncontrolled NOx emissions from coal-fired sources have been measured in the range 0.53

industry-to 2.04 lb/106 BTU at full load (Ziegler and Meyer, 1979) The

NOx formed in combustion is from fixation of atmospheric nitrogen and/or fuel nitrogen Ermenc (1956) found that at high temperature nitrogen and oxygen combine to form both

NO and NO2 The yield of NO increases from 0.26% at 2800F

to 1.75% at 3800F If the temperature is reduced slowly the reverse reaction will take place, but if the products are quenched

by rapid heat exchange, the reverse reaction rate becomes small and the oxides remain in the exhaust stream The oxide NO can usually be oxidized to form NO2 according to:

2NO O 2→2NO 2

Because this is a tri-molecular gas phase reaction, the centration of NO and NO2 tremendously affects the rate at which the oxidation takes place At low concentration, for example 1–5 ppm in air, the reaction is so slow that it would

con-be negligible except for the photochemical reactions which take place in the presence of sunlight The dioxide also reacts with oxygen to form ozone The existence of nitrogen trioxide

at low concentration in polluted atmospheres is postulated (Hanst, 1971) to form by the reaction with ozone

HNO3—that is, about 200 ppm (Ricci, 1977) The most widely used process for nitric acid plant tailgas cleanup is catalytic decomposition of NOx to nitrogen and oxygen

The current and projected values of the New Source Performance Standards (NSPS) for NOx are discussed later

in this article During recent years N2O formation rates have been the subject of controversy, especially in fuel NOxmechanisms

Oxides of Carbon

The carbon dioxide, CO2, concentration threshold for humans

is 5% (5000 ppm) for an 8-hour exposure This compares with

a normal atmospheric CO2 concentration of 0.03% (300 ppm)

With a perfect stoichiometric combination of pure carbon

in air, a CO2 concentration of about 21% could be attained

Considering the usual dispersion of combustion gases it would

take an unusual isolation to produce a CO2 health hazard

A more detailed description of CO2 consequences may be

found in the Appendix

Incomplete combustion of fuels is the more serious problem, since carbon monoxide, CO, will form This rarely happens in stationary furnaces for which efficiencies of combustion are high and oxygen is available in excess of the theoretical requirements

It has been estimated (Anon., 1970) that slightly more than

100 million tons of CO are emitted annually in the USA, of which the major sources are automobiles (59%), various open burnings (16%), chemical industry (10%) and other transpor-tation means (5%) New York City, with its acute urban traffic problem, has established a first alert at 15 ppm of CO over

TABLE 1 Status of Commercial FGD Processes (Adapted from Princiotta, 1978) FGD Process Reagent End Product Principle of Operation Efficiency(%)SO2 Removal Limestone Scrubbing Limestone (LS) CaSO3/CaSO4 Sludge LS slurry reacts in scrubber

absorbing SO2 and producing insoluble sludge.

80–90

Lime Scrubbing Lime CaSO3/CaSO4 Sludge Lime slurry reacts in

scrubber absorbing SO2and producing insoluble sludge.

at 8-hour period High temperatures favor the equilibrium

dissociation of CO2 to CO, with the latter being very stable

at high temperatures Thus is a CO2–CO mixture is quenched

from its high temperature zone the percentage CO may remain

high, since at lower temperatures longer times are required to

reach equilibrium Rich fuel-air mixtures favor the formation

of CO over CO2 A complete description of CO control

meth-ods may be found in the section Mobile Source Pollution.

Miscellaneous Gases

Compounds of fluorine are known to have negative effects

(Fluorisis) at concentrations as low as 5 × 10−3 ppm They

are generated as waste gases of fertilizer aluminum and

ceramic processes, but are present to a lesser extent in most

flue gases A concentration of 0.1 ppm (vol.) of fluorine has

been set as a maximum permissible value by the American

Conference of Governmental Industrial Hygienists; USSR

standards are roughly one tenth as stringent

Ozone, O3, is one of the strongest gaseous oxidants and is

formed naturally from oxygen during electrical discharges in

the atmosphere and at the high temperatures of combustion

Taken as oxidants, New York City classified an ozone level

above 0.03 ppm as unsatisfactory and above 0.07 ppm as

unhealthy over a 6 hour period Eye irritation commences

at concentrations of about 0.1 ppm Interestingly enough ozone in the lower stratosphere affords part of the protective shield against ultraviolet radiation from the sun, which could destroy land vegetation

O3→O + O.2

Some scientists are concerned that nitric oxide formed by supersonic jets may deplete the ozone supply in the lower stratosphere, eroding the barrier to the destructive rays

High temperature processes involving metal recovery

from ores emit mercury vapor in addition to sulfur dioxide

Mercury is available at concentrations up to a few hundred

ppm (Kangas et al., 1971) during zinc sulfide ore ing for example Hydrochloric and hydrofluoric acids also

process-appear in the roaster gases of such processes No danger levels for mercury vapor have been officially established

in ambient air quality standards

A few limits have been established for less common ants of the process industries in USSR standards given below

pollut-The US ambient air quality standards call for bon concentrations below 160 mg/m3 (0.24 ppm) between 6–9 am

hydrocar-Aldehydes and other oxygenated hydrocarbons are

formed by the action of ozone on unburned hydrocarbons in the presence of sunlight For example,

O31-3 Butadiene→A croleinFormaldehyde

In the above reaction both products have been linked to the severe eye irritation encountered in urban environments

TRANSPORT OF POLLUTANTSThe feed and waste materials of any combustion or chemi-cal process travel through ducts or pipes Control devices may be placed at various stages of the process, depend-ing on the separation technique to be employed It will

be valuable to review the flow and transport behavior for fluids and then the separation methods The important pro-cess variables to be considered are the mass flow rate of the waste gas, its temperature, pressure and composition

The raw material feed rate variables may also be of nificance, as in the desulfurization of fuel oil Control devices may broadly be classified according to the physi-cal separation process being used, adsorption: absorption:

sig-extraction: distillation: or to the chemical process, nous or heterogeneous catalytic reaction In each instance, equations which account for the transport of material and energy must be developed

homoge-In a sense almost any process may be considered as taking place in a pipeline The simplest model of flow is called plug flow and assumes that no mixing takes place along the axis of the pipeline, but that lateral mixing is com-plete Also, this assumes a flat velocity profile exists at each

E ED EDD

Solubility, CC per 100 GMS

acetic acid

longitudinal position, or that the average velocity is the same

at each lateral position

Continuity

If the density, k, of the fluid at a distance along the pipe, Z,

of cross section, S, changes, the velocity must also change as

seen by an elemental mass balance across d Z distance, i.e

setting the mass accumulation rate equal to the sum of net

input and generation rates (see Figure 3)

1 ( )

For steady state results, vS  const  W o and the mass

flow rate becomes the same at all axial positions If the fluid

is incompressible,  const., as for most liquids, vS, the

vol-umetric flow rate, does not vary with position even during

transient conditions

Motion

In a comparable manner an elemental momentum (force)

balance may be made over length dZ, which for

incompress-ible flow reduces to

v Z

FMPFNIMH GLRIE

SU FIGURE 3

For both steady and incompressible flow

dp dZ

F S

g

o z

The equation describes the relation between velocity and

pressure along the pipe The quantities F and F w are the tudes of skin frictional force and force doing work on external surfaces, respectively, both per unit length of pipe

magni-ENERGYThe First Law of Thermodynamics may be written for the

differential element of length, dz, at steady state

dH dz

g g

v g

p T

v

vT z

Component Balance

The equations of continuity, motion and energy often may

be applied to describe the situation in stacks of power plants, in the flow of fuels and effluents, and in the analy-sis of material, momentum and energy requirements of a

pollution producing process To analyze the concentrations

of pollutant it still remains to make component material

balances of n − 1 species within the system (for which n

components exist) For separation processes an additional

phase equation is usually required for transfer of pollutants

between the rich and lean phases

The mass balance on a particular species may be found

for component A by examining the imaginary stationary

dif-ferential element of thickness dz Assuming plug flow we

may derive an expression for c A:

Accum Net Input Generationrate of rate of rate of

R A —production rate of component A, by chemical or

nuclear reaction, moles A formed/(time) (vol.)

v—average fluid velocity

Also, r A is usually some empirical function of c A such as kc A n

for an irreversible decomposition reaction of nth order.

Gas Adsorption

Adsorption is the process by which a solid surface attracts

fluid phase molecules and forms a chemical or physical bond

with them The mechanism of adsorption includes:

1) diffusion of the pollutant from bulk gas to the

external surface of the particles,2) migration of the adsorbate molecules from the

external surface of the absorbent to the surface of the pores within each particle,

3) adsorption of the pollutant to active sites on the

pore

The attraction for a specific gas phase component will depend

on properties such as the concentration of the gas phase

com-ponent, the total surface area of absorbent, the temperature,

polarity of the component and adsorbent, and similar

prop-erties of competing gas molecules Adsorption is used to

concentrate (30–50 fold) or store pollutants until they can

be recovered or destroyed in the most economical manner

Adsorption is an exothermic process The heat of

adsorp-tion for chemical adsorpadsorp-tion is higher than that for physical

adsorption In the former, if the amount of pollutants to be

removed large, it is necessary to remove the heat of

adsorp-tion, since the concentration of adsorbed gas decreases with

increasing temperature at a given equilibrium pressure For

chemical adsorption, properties which affect reaction

kinet-ics will also come into play (see section on Gas Reaction)

Activated carbon, silicon, aluminum oxides, and molecular

sieves make up the majority of commercially significant

adsor-bents Activated carbon is the least affected by humidity and

physically adsorbs nonpolar compounds since it has no great

electrical charge itself The adsorption rate of activated carbon can be increased with chemical impregnation For instance, activated carbon impregnated with oxides of copper and chro-mium are found very useful to remove the hydrogen sulfide in gas streams where oxygen is not present (Lovett and Cunniff, 1974) Alumina and silica materials preferentially adsorb polar compounds Molecular sieves have greater capture efficiencies than activated carbons but they often have a lower retention efficiency and are considerably more expensive

The ease of adsorbent regeneration depends on the nitude of the force holding the pollutants on the surface of adsorbent The usual methods for the adsorbent regeneration include stripping (steam or hot inert gas), thermal desorp-tion, vacuum desorption, thermal swing cycle, pressure swing cycle, purge gas stripping, and in situ oxidation

mag-In many respects the equilibrium adsorption tics of a gas or vapor upon a solid resemble the equilibrium solubility of a gas in a liquid For simple systems, a single curve can be drawn of the solute concentration in the solid phase as a function of solute concentration or partial pressure

characteris-in the fluid phase Each such curve usually holds at only one specific temperature, and hence is known as an isotherm Five types of commonly recognized isotherms are shown by the curves in Figure 4 There are three commonly used mathemat-ical expressions to describe vapor or gas adsorption equilib-rium: the Langmuir, the Brunauer-Emmett-Teller (BET), and the Freundlich isotherm The Langmuir isotherm applies to adsorption on completely homogeneous surfaces, with neg-ligible interaction between adsorbed molecules It might be surmised that these limitations correspond to a constant heat

of adsorption The Freundlich isotherm can be derived by assuming a logarithmic decrease in heat of adsorption with fraction of coverage Gas adsorption is an unsteady state pro-cess The curve of effluent concentration as a function of time

is commonly referred to as the break-through curve It usually has an S shape The break-through curve may be steep or rela-tively flat, depending on the rate of adsorption, the adsorption isotherm, the fluid velocity, the inlet concentration, and the column length The time at which the break-through curve first begins to rise appreciably is called breakpoint

The design of an adsorption column requires prediction

of the breakthrough curve, and thus the length of the tion cycle between elutions of the beds, given a bed of certain length and equilibrium data Because of the different forms

adsorp-of equilibrium relationship encountered, and the unsteady nature of the process, prediction of the solute break-through curve can be quite difficult At present, detailed design of adsorption columns is still highly dependent on pilot scale evaluations of simulated or real systems

Before discussing the method of predicting the through curves, one should consider the isotherm For Langmuir

break-isotherm (Langmuir, 1917), if it is assumed that A 1 reacts with

an unoccupied site X0 to form adsorbed component X1,

FIGURE 4 Types of adsorption isotherms.

the equilibrium adsorption concentration, C, is obtained in

terms of the gas phase concentration C1 and the total

adsorp-tion site concentraadsorp-tion, C0

C C

Here, K1 is the adsorption equilibrium constant which varies

only if temperature varies

Diatomic molecules such as chlorine might be expected

to simultaneously adsorb and dissociate on adjacent sites

Such an adsorption might be described symbolically by

The latter may be referred to as competitive adsorption

con-centration for the adsorption types described thus far

Another isotherm finding wide use, particularly in layer adsorption, is that of Brunauer, Emmett and Teller (1938), the BET equation

multi-C C

If neither the Langmuir or BET equations are satisfactory,

a plynomial fit to adsorption data may be required

For adsorption the particles are usually dumped into a column as a packed bed (Figure 6) In commercial adsorp-tion columns, equilibrium concentrations are not attained uniformly, but for convenience the rate of adsorption is

assumed to be proportional to C See below under Reaction.

The time t3 in which breakthrough (Cig at the exit equals the permissible set amount) occurs may be established by analyzing the differential component mass balance assuming that the transport of material again is governed by diffusion through a film

N ik C g( igC ig*) (12)

in which C ig* is the gas phase concentration in equilibrium

with the absorbed concentration Cis at the same elevation

M

G F E D

M C

FIGURE 6 (a) Pictorial representation (b) Schematic model showing a differential element over which a mass balance is made (c) Pollutant concentration as a function

of time at various heights.

TBNBFIGURE 5

The equilibrium concentrations have been determined for

most commercially available adsorbents and typical pollutants

and are presented as either Langmuir or BET isotherms In

general such equations take the form

The generation term is excluded, as it is assumed that

chemi-cal reaction is not taking place in the system To develop an

expression for the net rate of accumulation of the mass of A i

in the column, that is, the rate expression for the adsorption

process, one also needs the function of the concentration of

A i in the fluid phase

Considering a differential column segment and writing the

continuity relationship for pollutant A i in each phase for this

differential section, one gets for the solid phase in this section

11

M a : is the molecular weight of A1

P s: is the averaged density of the solids

: is the fractional voidage in the bed

For the gas phase in this differential segment,

V z: is the superficial velocity of the fluid

These partial differential equations (13)–(15) may be

solved simultaneously by numerical analysis using difference

formulas to approximate the partial derivatives In such a way the breakthrough curves of hazardous organic vapors may be predicted for a given adsorbent

Smoothed computerized results were plotted on Figure 7 for five different compounds having Langmuir type behavior

on activated carbon under the same hypothetical operating conditions If one wishes to attain a 90% removal of certain organic vapor, one could easily see from Figure 7 that diethyl ether requires the shortest re-cycle time and methyl isobutyl ketone the longest among the five materials on the graph

pre-Here (RT/V s )ln f s /f g is plotted versus N s (in which:—gas

constant, T—temperature, K, V s—molar volume of

adsor-bate, cc/mole, f s and f g—fugacites of adsorbate and gas

and N—amount of gas adsorbed, g—moles/gm adsorbent

Hydrocarbons and SO3 adsorb readily on activated carbon

SO2 has a maximum retention of 10 wt.% on carbon at 20C, 760 torr Ozone decomposes to oxygen on carbon (Ray and Box, 1950)

adsorp-tion on silica gel Activated carbon has significantly better equilibrium properties than does silica gel (vis Figure 9 vs

Figure 13)

Other results for activated carbon and zeolites may be found

in the book by Strauss (1968) Basic facts about adsorption properties of activated charcoal, system types and components and applications are discussed by Lee (1970) He tabulated data on the air purification applications for inexpensive, non-regenerative, thin bed adsorbers and for regenerative systems, and discusses the design of a solvent vapor recovery system

I Diethyl ether

II Acetone III Carbon disulfide

IV MEK

V Methyl isobutyl ketone

V IV

III III II I

50 0

0.5 1.0

“Activated carbon filters were used to concentrate

atmospheric mixtures of acrolein, methyl sulfide, and

n-propyl mercaptan Removal efficiency and carbon

capac-ity for each of the odor compounds were investigated

using two different carbones, Cliffchar (4–10 mesh) and

Barnebey–Cheney (C-4) A closed system was devised to

establish a known atmospheric odor concentration for each

filter run Solvent extraction techniques were employed to desorb and recover the odor compounds from the carbon filters All quantitative analyses were conducted with gas liquid chromatography utilizing the hydrogen flame ion-ization detector The removal studies conducted indicate that the efficiency of removal of a carbon filter is essen-tially 100% up to the point of filter breakthrough This breakthrough point is governed by the filter’s capacity for

a particular compound This study indicated that the filter capacity is dependent both on the type of carbon employed and the particular odor compound adsorbed Solvent recov-ery of the odor compounds from the carbons varied from 0

to 4.5% for the mercaptan up to 96 to 98% for acrolein Per cent recovery was found to vary for a given odor compound with different carbons and for a given carbon with different odor pollutants.” (Brooman and Edgerley, 1966.)

Gas Absorption

Adsorption is a diffusional process that involves the transfer

of molecules from the gas phase to the liquid phase because

of the contaminant concentration gradient between the two phases Adsorption of any species occurs either at the sur-face of the liquid film surrounding the packing or at the bubble surface when the gas is the dispersed phase When

a gas containing soluble components is brought into contact with a liquid phase an exchange of the soluble components will occur until equilibrium in a batch system or steady state

in a flow system is attained Adsorption may involve only a simple physical solubility step or may be followed by chem-ical reaction for more effective performance The latter is usually used for flue gas desulfurization and denitrification

Rates of adsorption depend on the solubility of the gas At equilibrium, for gaseous species of low or moderate solu-bility, the partial pressure of the component is related to its liquid mole fraction according to Henry’s law,

p i  HX i where H is Henry’s constant Both partial pressures and mole

fraction may be related to concentration

p i  C ig RT and X i  C il /C 1

At constant temperature, C ig  H ′C il , where H ′  H/C t RT. If

the gas is highly soluble in the liquid, H will be small The

solubility of the gas is affected by the concentration of ions

in the solution at the interface Van Krevelen and Hoftijzer (1948) proposed an empirical equation to correct the effect

of concentration of ions on Henry’s constant

The rate of mass transfer is proportional to both the interfacial area and the concentration driving force The proportional constant is known as the mass transfer coeffi-cient Because material does not accumulate at the interface (Figure 14) the flux in each phase must be the same Thus the rate of transfer per unit area is

N i  j g (C ig  C igI )  h L (C iLI  C iL ).

HC DCC DHC ECC EHC FCC

PS MxRTYUONXNY

EC EC GH GH IL IL

FIGURE 8 The system nitrous oxide-carbon

McBain and Britton: J Am Chem Soc 52, 2217 (1930)

( Fig 14 ).

A BA

10 20 30 40 50 60 70 80

CA DA EA FA GA HA

MO O TYX TX PP M-N-L- xR

FIGURE 9 Absorption isotherms for the system

CO2-carbon (note that t c  31C).

C CB CBB CBBB

Y × CBE A

A-B DB FB HB LB CBB

EIE-D U DMH-M U DIE-D U CMF-I U CLB U CGB U CBB U II-F U

S U XSYS

ID LD MD ND OD EDD

In most industrial adsorption processes, the gas is reacted

with some substance to form a semistable compound in

the liquid phase This technique permits a great deal more

gas to be adsorbed per gallon of liquid circulated, and, in

most instances, will increase the mass transfer coefficient

where E  enhancement factor, k L  mass transfer

coef-ficient in absorption E  1 for physical adsorption The

enhancement factor can be found elsewhere (Astarita, 1967;

Danckwert, 1970; Sherwood et al., 1975).

Consider a system whose equilibrium line is straight over the range of compositions which need be considered

The mass transfer rate may be described in forms of pseudo concentration values, thus,

N iK C g( igC ig*)K C L( iL* C iL),

where K g and K L are overall mass transfer coefficient in gas and liquid phase, respectively If concentrations at equilib-rium can be represented by Henry’s law

C ig*  H C iL,then

H K

    

If the gas is highly soluble in the liquid, H′ will be small and

K g  k g Hence the mass transfer rate is

N iK C g( igC ig*)

and absorption is said to be gas phase controlled

In cleaning an effluent stream of low pollutant species concentration physical absorption alone is often insufficient

to produce the required removal and a reactant may be added

to the absorbing solution to enhance the rate For example potassium permanganate has been found to be an excellent absorbent for NO (see excellent review by C Strombald (1988))

FIGURE 13 Adsorption isotherms for the system CO2-silica gel (note that t c  31C)

Patrick, Preston and Owens: J Phys Chem

OMYNXU PTX TPR

x

CB FB

HB IB CBB

FIGURE 12

The design of absorbers involves the estimation of

column diameter, height, and pressure drop The column

diameter is fixed by the contaminated gas flow rate The

determination of the height of the two-phase contacting zone

involves an estimation of the mass transfer coefficients, the

alternating use of equilibrium concentration relationship,

and the law of mass conservation For nonisothermal or

adi-abatic operating conditions, the law of energy conservation

needs to be considered There are many types of equipment

and configurations for absorbers or scrubbers For example,

McCarthy (1980) discussed the scrubber types and

selec-tion criteria

For packed towers the interfacial area, a, differs from

the packing surface area, a T, because the packing is not

always completely wetted A fraction of the surface may

not be active in mass transfer Also, stagnant pockets will

be less effective than flowing streams The correlation of

Onda et al (1968) may be used to estimate the value of

interfacial area

a a

L a

L a g

Where s c represents the critical surface tension above

which the packing can’t be wetted The values of sc for

various packing materials are shown in Table 4 (Onda

that are not accounted for in the values of a Values of a

Pall rings are underestimated about 50% according to Charpentier (1976)

The liquid phase mass transfer coefficient can be mated using Mohunta’s equation (1969)

r m

m r

where P : total pressure, atm

M : gas molecular weight

D G : solute gas diffusively

C  2.3 for d ,  1.5 cm

C  5.23 for d  1.5 cm

Tray type towers have also been used successfully

Bubble cap plates correlations have been proposed by Andrew (1961)

TABLE 4 Critical surface tension of packing materials Material sc dynes/cm

in which:

S  effective liquid area on plate, cm2

u  superficial gas velocity, cm/sec

a  interfacial area/unit area of plate

A conservative estimate for the liquid phase coefficient in a

sieve plate may be obtained from the Equation of Claderbank

and Moo–Young (1961)

k l0 31gv1 1 3 d v l

1 2 3

( ) (/ / )/ cm/sec

where vl is the kinematic liquid viscosity For CO2 and water

at ordinary temperature, k l≈ 0.01 cm/sec Typical sieve plate

N i values are about 10−4 g mole/cm2-sec-atm

“Each tray of Figure 15 has a series of drawn orifices fitted

with a cage and cap The orifice has a flared entrance This

reduces the dry pressure drop allowing a greater percentage of

the work expended to be utilized for scrubbing The floating

cap maintains the scrubbing efficiency even with variations in

gas flow as wide as 40–110% of capacity

Gas enters the vessel flowing upward through the valve

trays Liquid is introduced on the top tray and flows across

each tray over a weir and then to a sealed downcomer to the

next lower tray A level is maintained over each tray by the

weir The upward flowing gas is given a horizontal component

by the cap causing atomization of the liquid on the tray The

froth formed consisting of a myriad of small droplets traps the

particles and absorbs or reacts with acid or alkaline vapors

The liquid agitation on the tray surface prevents buildup This

has been demonstrated by the many successful applications in

the Petrochemical Industry involving tarry solids and liquids

Each tray has a 1½ W.G pressure drop A typical

instal-lation with four (4) trays would require less than 8″ W.G

pressure loss.”

An excellent review of gas phase absorption may be

found in the work of Danckwerts (1970) The molecular

dif-fusivities in the vapor phase D r , and in the liquid D1, may

be found from existing correlations, for example see Bird

et al. (1960) Unlike the solid in adsorption the liquid

sol-vent in absorption usually leaves the system where it can be

regenerated Hence a steady state plug flow analysis in either

phase in terms of overall coefficients is possible

The required tower height is then given by either of the

fol-lowing integral relationships:

in which: a is the surface area of contact per unit volume

of bed; L and G are per superficial mass velocities Further

discussion on the subject may be found in the work of

Cooper and Alley (1994)

PROPERTIES OF ABSORBENTSHenry’s law constants for CO, CO2, NO and H2S, are pre-sented in Table 5 Lower values of H such as those for H2S correspond to higher solubility values Table 6 contains spe-cific wt fraction absorbed at equilibrium vs gas partial pres-sure for both ammonia and SO2 in water; Table 7 has similar material for HCl

Highly soluble materials have absorption rates which are controlled by diffusion through the gas phase (see Table 8)

REACTIONProcesses exist for catalytically removing gaseous pollutants

by either forming harmless products or products more nable to recovery The behavior of most catalytic reactions requires a more substantial analysis than homogeneous sys-tems because of the presence of at least two phases One of the FIGURE 15 Flexitrary Scrubber (Courtesy of Koch Engineering).