Sei sulla pagina 1di 35

MASS TRANSFER IN GAS ABSORPTION & DIFFUSION We now move to analysis of gas absorption.

Gas absorption essentially involved the transfer of materials from the gas phase to the liquid phase. it is "defined" as the operation in which a gas mixture is contacted with a liquid for the purpose of preferentially dissolving one or more components of the gas mixture and to provide a solution of them in the liquid. The gaseous component is said to be absorbed by the liquid. The transferred component is known as the solute. In the simplest case, gas absorption involves as least 3 components. As an illustration, consider an ammoniaair-water system. The gas contains ammonia-air mixture. Ammonia is the solute and it is very soluble in water while air is not. Hence, by means of contacting the gas mixture with water, ammonia will dissolves preferentially in water, and a solution of ammonia in water (ammonium hydroxide) is obtained. Gas absorption processes are widely used in the industry. It can be used for removing contaminants or impurities from a gas stream. One of the most common examples of gas absorption and stripping is the amine absorption and regeneration unit whereby toxic H2S gas from a fuel gas mixture is removed by

liquid amine (DEA, MEA or glycol) as will be discussed later. Other example include the absorption of CO2 using hot potassium carbonate, or using the solvent Selexol . Yet in other applications, it can be used for gas dehydration when an insoluble gas is dried by contact with a dehydrating liquid. An example is in the drying of chlorine using 98 wt% sulfuric acid. In air pollution control, the various oxides of nitrogen can be removed by absorption with water, sulfuric acid, and organic solutions. Gaseous ammonia can be removed by absorption with water. Equilibrium Distribution (Solubility) Curve The gas absorption process involves the redistribution of solute between the gas phase and the liquid phase when the 2 phases come into close contact and achieves equilibrium condition. The relationship between solute concentration in the gas phase and in the liquid phase at constant temperature and pressure is known as the equilibrium distribution curve, as shown in the Figure below. The presence of the solute in the liquid represent the its gas solubility at the prevailing temperature and pressure.

NOTE : The equilibrium solubility curve for gas absorption as seen here is not the same as the equilibrium curve for distillation. Some of the important differences to observe: in distillation, we have seen that the equilibrium curve simply shows the equilibrium relationship between 2 components: the more volatile and the less volatile. In gas absorption, as noted at the beginning, in the simplest case, there will be 3 components (e.g. NH3 , air and water). the equilibrium solubility curve is plotted for a particular constant temperature. Any point on the same curve represent gas solubility at the same temperature. For the equilibrium curve in distillation, the points represent vapour-liquid equilibrium at different temperatures. there is no 45o diagonal for solubility curves, i.e. the solubility curve can lie anywhere in the x-y plot (or p-y plot, etc). On the other hand, the equilibrium curve for distillation normally lies above the 45o diagonal line (except, of course, for azeotropic mixtures).

the equilibrium solubility curve is not usually plotted over the entire concentration range from 0.0 to 1.0 mole fractions (as in customary done for design of continuous distillation column using McCabe-Thiele Method). This is because generally most gases are soluble in a liquid only over a narrow concentration range.

Before we looked into the design of gas absorption equipment (i.e. tray or packed towers where the gas and liquid streams come into contact), we need to understand how this mass transfer actually take place between the phases. To do so, we turn to analysis using film concept.

Mass Transfer Equation & Film Mass Transfer Coefficients In commercial absorption equipment, both the liquid and the gas are usually in turbulent flow and the film thickness is not easy to determine. Therefore instead of analysis of mass transfer using Fick's Law, it is more convenient to write the molar flux of A using mass transfer equation of the form below: Molar Flux, (NA = Mass Transfer Coefficient x Driving Force)

At a point A (xAL, yAG), we can write the mass transfer equations for each of the phases:

where : molar flux of component A, mole/(area.time) mass transfer coefficients ky in the gas phase concentration driving force ( yAG in the gas phase (mole yAi ) fraction) mass transfer coefficients kx in the liquid phase concentration driving force ( xAi in the liquid phase (mole xAL ) fraction) NA The k-values above are also known as film mass transfer coefficients, and they are usually determined experimentally, or by correlations. Because there are many analogies between heat transfer and mass transfer, many correlations originally derived from heat transfer are used for the prediction of mass transfer coefficients. [In fact, the mass transfer equation is obtained based on the analogy with the heat transfer equation q = Q/A = h (DT); where DT is the temperature difference driving force for heat flow. There are 2 mass transfer equations for 2 different mass transfer coefficients, one in the gas phase and another in the liquid phase; just like the case of a heat exchanger (e.g. double-pipe, or shelland-tube) whereby there is a tube-side heat transfer coefficient and a shell-side heat transfer coefficient. Similar definition can be made using overall mass transfer coefficients.

Mass Transfer Dimensionless Groups and Correlations The attached Table below (left) showed examples of similarity of dimensionless groups for heat and mass transfer; and the other Table (right) showed several correlations for mass transfer. Generally, the mass transfer correlations are more complex and difficult to use. In addition, they are very specific in applications and are limited to some simple situations.]

[ See also pp. 66-70, 72-77 in R.E. Treybal, "Mass Transfer Operations", 3rd Ed. for more discussion on analogies between heat and mass transfer, and see pp. 666-674 in McCabe et al, "Unit Operations for Chemical Engineering", 5th Ed. for examples of using various mass transfer correlations ] In the above analysis of mass transfer across an interface, note that since the interface concentrations varies throughout the gas absorption equipment (e.g. a tray column). It is worthwhile highlighting that NA depends on the conditions at the particular point in the column. In other words, NA may vary throughout the entire length of the column.

Mass Transfer Resistance The resistance to mass transfer is defined as the reciprocal of the mass transfer coefficient:

represents the resistance to mass transfer in the liquid phase

represents the resistance to mass transfer in the gas phase [Again, refer to the analogy with heat transfer whereby 1/h is the convective resistance to heat transfer] It is important to know if one of the 2 resistances is controlling the mass transfer. If so, the rate of mass transfer can be increased by promoting turbulence in and/or increasing the dispersion of the controlling phase. Click here for more information.

Analysis of mass transfer resistance at a point P From the 2 mass transfer equations, by rearranging and eliminating NA :

Referring back our analysis of mass transfer from Point P to Point M, we see that the ratio of mass transfer coefficients is actually equal to the slope of line PM. The gradient of the line determines the relative resistances of the 2 phases.

The above equation is useful if one does not know the interface equilibrium concentrations. We can use the above equation to determine the equilibrium concentration at the interface ( xAi, yAi ), i.e. to locate point M, provided that kx and ky are known (or can be calculated using appropriate correlations). We do so by plotting a straight line originating from point P ( xAL, yAG ) with slope - kx / ky. The point of intersection of this line with the equilibrium curve gives point M which yield the values of xAi and yAi . That way we can calculate the flux NA at that particular point. Two-Film Theory and Equilibrium Solubility Curve In the analysis of gas absorption, we are interested in the transfer of materials throughout the entire gas absorption equipment, not just a single location in the equipment. Therefore the two-film theory can be analyzed more effectively by using the equilibrium solubility curve. The concentrations at the interface in the gas ( yAi ) and in the liquid ( xAi ) is represented as a point M on the equilibrium solubility curve. Point M thus has the coordinates ( yAi, xAi ). As we move along the column along the continuous interface, we can trace out an equilibrium curve. [ Remember that one of the assumptions of the two-film theory is that equilibrium condition exists at the interface ] [ A number of other theories have also been used, e.g. penetration theory, surface-renewal theory, surface-stretch

theory, etc. See pp.144-150, J.D. Seader & E.J. Henley, "Separation Process Principles" for more information. ] Very often, the subscript "Ai" is dropped, and the equilibrium curve is simply a relationship between y and x; i.e. y = f(x). The concentrations in the bulk gas phase ( yAG ) and in the bulk liquid phase ( xAL ) is represented as a point P above the equilibrium curve. Point P thus has the coordinates ( xAL, yAG ). Point P is located above the equilibrium curve. [ Remember that it is the departure from equilibrium that provide the driving force for mass transfer ] See the Figure below :

Notation : yAG = composition of A in the bulk gas phase (mole fraction) xAL = composition of A in bulk liquid phase (mole fraction)

( xAi, yAi ) = equilibrium interface compositions (mole fraction)

Analysis of Mass Transfer Process using Two-Film Theory In the gas-phase, the concentration falls from yAG in the bulk gas to yAi at the interface. Thus, there is a concentration driving force for mass transfer from the bulk gas to the gas film to the interface. At the interface, the component A crossed the interface and enters the liquid side. In the liquid-phase, the concentration falls from xAi at the interface to xAL in the bulk liquid. Thus, there is a concentration driving force for mass transfer from the interface to the liquid film to the bulk liquid. The mass transfer process can be represented by the line PM. See the Figure above. Mass transfer can be described by a set of mass transfer equations. Click here for more information.

NOTE : The bulk concentrations yAG, xAL are not equilibrium values, otherwise there would be no diffusion of A.

REMINDER : The two-film theory and equilibrium curve can be expressed in other ways, e.g. in terms of partial

pressure (for the gas phase) and concentration (for the liquid phase); but the analysis for them is the same as outlined before for mole fractions ( x and y). Refer to the Figure below which showed the two-film theory and equilibrium curve expressed in partial pressures and concentrations.

Overall Mass Transfer Coefficients The previous definitions for molar flux NA require the knowledge of the interface concentrations. Since experimental sampling of the concentrations at the interface is very difficult or virtually impossible; it is more useful to define the mass transfer equation using overall mass transfer coefficients KX and KY :

xA* is the concentration (mole fraction) in liquid phase that is in equilibrium with yAG. yA* is the concentration (mole fraction) in vapor phase that is in equilibrium with xAL. Driving force for mass transfer: ( yAG - yA* ) in the gas phase (as indicated by line PC) and ( xA* - xAL ) in the liquid phase (line PD). See the Figure below.

NOTE : for most situations the asterisk '*' is not shown. [ Again, there is similarity to heat transfer - the overall heat transfer coefficient which can be defined based on outer surface area of the tube (UO) or on the inner surface area of the tube (Ui). Here we have the overall mass transfer coefficient based on the gas side or based on the liquid side ]

Relationship between film and overall mass transfer coefficients: It can be shown that kx, ky, KX, and KY are related through the following equations:

where m" is the slope of line segment DM, and m' is the slope of line segment MC as shown. If the equilibrium line is straight, then m' = m". kx , ky , KX , and KY all change with positions in the tower. The above equations can be used in gas solubility analysis.

NOTE: units for kx , ky , KX , and KY varies with the way the mass transfer equation is written (vapor phase or liquid phase) and the driving forces used, e.g. mole fractions ( y or x ), mole ratios ( X or Y ), weight fraction, partial pressures (p), or concentrations (c) etc. Analysis of Gas Solubility Recall the relationship between overall and film mass transfer coefficients, and that the 1/k represents the mass transfer resistance. If m' is small (i.e. the equilibrium curve is very flat), the term m'/kx is not significant, therefore: and the major resistance to diffusional mass transfer lies in the gas phase and the mass transfer is said to be gas-phase controlled. In this case, solute A can be interpreted as being very soluble in the liquid: at equilibrium, a small concentration of A in the gas will bring about a very large concentration in the liquid.

If m" is large (i.e. the equilibrium curve is very steep), then the term 1/m"ky is insignificant, therefore,

and the majority of resistance to mass transfer resides in within the liquid. The mass transfer is said to be liquid-phase controlled. Solute A is relatively insoluble in the liquid: a very large concentration of A in the gas phase is required to provide even a small change of concentration in the liquid. [See pp.110, R.E. Treybal, "Mass Transfer Operations", 3rd Ed.]

For some examples of gas solubility analysis, click here. Determination of the packed height of a column (see later) most commonly involves the overall gas-phase coefficients because the liquid usually have a strong affinity for the solute so that resistance to mass transfer is mostly in the gas. [ See pp.319 - 321, J.D. Seader & E.J. Henley, "Separation Process Principles" ]

Examples of Gas Solubility Analysis The Figure below shows the solubilities of some gas-air mixtures in water (partial pressure of solute vs. liquid mole fraction of solute).

From the solubility plot, note that: at the same temperature (10 oC) HCl is more soluble than NH3, which in turn is more soluble than SO2. solubility of any gas is dependent on temperature. In most cases (but not always), the solubility of a gas decreases with increasing temperature. NH3 is less soluble at 30 oC than at 10 oC. different gases have different solubilities e.g. HCl is more soluble than SO2 at the same temperature (10 oC). the relatively insoluble gas is high in concentration in the gas phase, i.e. high partial pressure at equilibrium. Conversely, very soluble gas has low partial pressure. In many practical cases, only one component in the gas mixture is relatively soluble in the liquid, e.g. in the NH3-Air-H2O system above, since NH3

is relatively more soluble than air in water, thus, pAir >> pNH3 . [ Equilibrium solubility curves for other liquids are also available ]

PACKED COLUMN FOR GAS ABSORPTION (DILUTE SYSTEMS) Gas absorption can be carried out in a packed column. We will look at both column diameter and packed height in this Section. The Figure below showed a typical gas-liquid flow in a packed column. Click here for more information of packings.

Determination of column diameter involves the analysis of pressure drop across the packed bed.

As for packed height, the design used in the early days was based on the HETP method. This was largely replaced by the Method of Transfer Units.

Comparison between number of theoretical trays, HETP and Method of Transfer Units The Number of Transfer Units (NTU) and Height of Transfer Units (HTU) such as NOG, HOG should not be confused with the number of theoretical trays (N), and the height equivalent to theoretical plate (HETP) respectively. When the operating line and equilibrium line are straight and parallel: NTU = N ; and HTU = HETP Otherwise, the NTU can be greater than or less than N as shown in the Figure below:

When the operating line is straight but not parallel, we have the following relationships:

where

is the Absorption Factor.

Packed Column & Packings Besides tray column, distillation (as well as other unit operations such as gas absorption, liquid-liquid extraction, etc.) can also be carried out using packed column filled with packings. Various types of packings made of different types of materials of construction are available, and both random and structured packings are commonly used. Examples of random packings as shown in the Figures below - left and right - are Raschig rings, Pall rings, Berl saddles, etc.

Random vs. Stacked Random packings, as the name implied, are dumped into a column during installation and allowed to fall in random. Small packings poured randomly into a vessel is certainly the more popular and commonly employed form of packed-tower design. However, in certain instances where exceptionally low pressure drop

and very high flowrates are involved, stacked or oriented packings have also been used. See the Figure below. However, only those packings of cylindrical shape and with a diameter larger than 3-inch would be practical to install in a stacked form. Two types of arrangement are possible: triangular (diamond) pitch or square pitch.

The different packings has several basic characteristics that make them suitable for gas-liquid contacts. Dry vs. Wet Random Packing In dry packing application, the packings are allowed to drop into the column via the (a) chute-and-sock method, or (b) rope-and-bucket method. Dry packing avoids high hydrostatic liquid head, and prevents the introduction of water into a dry process. It is also quicker and less expensive than wet packing, and it minimises rusting of metal packings. In any case, it is not suitable for plastic packings, as plastic typically floats on water. Wet packing applications are preferred when the packings are constructed of breakage-prone materials, such as ceramic or carbon. The column is first filled with water and the packings are gently poured down the column. The water cushions the fall and promotes randomness of settling. This tends to increase column

capacity and improve the column pressure drop characteristics. Wet packing also minimises compression and mechanical damage to packing materials. The main disadvantage is the need to remove the water after loading and dry the packings. [ Back on Top ]

Structured Packing Structured packings are considerably more expensive per unit volume than random packings. They come with different sizes and are neatly stacked in the column. Structure packings usually offer less pressure drop and have higher efficiency and capacity than random packings. 2 examples are shown in the Figures below - left and right.

Besides packings, there are other column internals. [ Back on Top ]

Packing Height : The Method of Transfer Units

A newer concept in the analysis of packed column centred on the method of transfer units. This method is more appropriate because the changes in compositions of the liquid and vapour phases occur differentially in a packed column rather than in stepwise fashion as in trayed column. In this method, height of packings required can be evaluated either based on the gas-phase or the liquidphase. The packed height (z) is calculated using the following formula: z=NxH where N = number of transfer units (NTU) - dimensionless H = height of transfer units (HTU) - dimension of length The number of transfer units (NTU) required is a measure of the difficulty of the separation. A single transfer unit gives the change of composition of one of the phases equal to the average driving force producing the change. The NTU is similar to the number of theoretical trays required for trayed column. Hence, a larger number of transfer units will be required for a very high purity product. The height of a transfer unit (HTU) is a measure of the separation effectiveness of the particular packings for a particular separation process. As such, it incorporates the mass transfer coefficient that we have seen earlier. The more efficient the mass transfer (i.e. larger mass transfer coefficient), the smaller the value of HTU. The values of HTU can be estimated from empirical correlations or pilot plant tests, but the applications are rather restricted.

["Principles of Unit Operations" 2nd Ed., Foust et al, p.391] The calculation of packing height follows the same nomenclature as before and this is shown in the Figure below. In this Section, we will focus on the applications of the equations rather than any derivation of them. Determination of the packed height can be based on either the gas-phase or the liquid-phase.

For the gas-phase, we have: z = NOG x HOG

and KY is the overall gas-phase mass transfer coefficient. "a" is the packing parameter that we had seen earlier (recall the topic on column pressure drop, e.g. Table 6.3) that characterize the wetting characteristics of the packing material (area/volume). Normally, packing manufacturers report their data with both KY and "a" combined as a single parameter. Since KY has a unit of mole/(area.time.driving force), and "a" has a unit of (area/volume), the combined parameter KY a will have the unit of mole/(volume.time.driving force), such as kg-mole/(m3.s.mole fraction). As seen earlier, other than mole fraction, driving force can be expressed in partial pressure (kPa, psi, mm-Hg), wt%, etc. y1* is the mole fraction of solute in vapour that is in equilibrium with the liquid of mole fraction x1 and y2* is mole fraction of solute in vapour that is in equilibrium with the liquid of mole fraction x2 . The values of y1* and y2* can be obtained from the equilibrium line as previously covered (see section on Two-Film Theory). See the Figure below.

(y1 - y1*) is the concentration difference driving force for mass transfer in the gas phase at point 1 (bottom of column) and (y2 - y2*) is the concentration difference driving force for mass transfer in the gas phase at point 2 (top of column). [ Point P (x, y) as shown is any point in the column. The concentration difference driving force for mass transfer in the gas phase at point P is (y - y*) as shown previously, this time no subscripts are shown. ] NOTE: Both equilibrium line and operating line are straight lines under dilute conditions.

Alternatively, equilibrium values y1* and y2* can also be calculated using Henry's Law ( y = m x, where m is the gradient) which is used to represents the equilibrium relationship at dilute conditions. Thus, we have: y1* = m x1 ; y2* = m x2 [ Back on Top ] Similarly for the liquid-phase we have: z = NOL x HOL

and KX is the overall liquid-phase mass transfer coefficient, and "a" is the packing parameter seen earlier. Again, normally both KX and "a" combined as a single parameter. Likewise, x1* is the mole fraction of solute in liquid that is in equilibrium with the vapour of mole fraction y1 and x2* is mole fraction of solute in liquid that is in equilibrium with the vapour of mole fraction y2 . Refer to Figure 134 for finding values of x1* and x2* from the equilibrium line.

Alternatively, x1* = y1 /m and x2* = y2 /m. (x1* - x1) is the concentration difference driving force for mass transfer in the liquid phase at point 1 (bottom of column) and (x2* - x2) is the concentration difference driving force for mass transfer in the liquid phase at point 2 (top of column). Using either gas-phase or liquid-phase formula should yield the same required packing height :

[ For more info on packed column design, see Chp. 4, "Process Plant Design", J.R. Backhurst & J.H. Harker, or for applications of various packed columns, refer to "Random Packings and Racked Towers", R.F. Strigle Jr. ] HETP (Height Equivalent to a Theoretical Plate) As we have noted, instead of a tray (plate) column, a packed column can be used for various unit operations such as continuous or batch distillation, or gas absorption. With a tray column, the vapours leaving an ideal plate will be richer in the more volatile component than the vapour entering the plate by one equilibrium "step". When packings are used instead of trays, the same enrichment of the vapour will occur over a certain height of packings, and this height is termed the height

equivalent to a theoretical plate (HETP). As all sections of the packings are physically the same, it is assumed that one equilibrium (theoretical) plate is represented by a given height of packings. Thus the required height of packings for any desired separation is given by (HETP x No. of ideal trays required). HETP values are complex functions of temperature, pressure, composition, density, viscosity, diffusivity, pressure drop, vapour and/or liquid flowrates, packing characteristics, etc. Empirical correlations, though available to calculate the values of HETP, are restricted to limited applications. The main difficulty lies in the failure to account for the fundamentally different action of tray and packed columns. ["Chemical Engineering Vol.2", 4th Ed., Coulson & Richardson, p.508-509] In industrial practice, the HETP concept is used to convert empirically the number of theoretical trays to packing height. Most data have been derived from small-scale operations and they do not provide a good guide to the values which will be obtained on full-scale plant. [ For more information on HETP prediction, see pp.1-355 of "Handbook of Separation Techniques for Chemical Engineers", 3rd Ed., P.A. Schweitzer , or pp.335, "Separation Process Principles", J.D. Seader & E.J. Henley ]

This method had been largely replaced by the Method of Transfer Units.

Column Diameter and Pressure Drop In determining the column diameter, we need to know what is the limiting (maximum) gas velocity that can be used. This is because the higher the gas velocity, the greater the resistance that will be encountered by the down-flowing liquid and the higher the pressure drop across the packings. Too high a gas velocity will lead to a condition known as flooding whereby the liquid filled the entire column and the operation became difficult to carry out. High pressure will crush and damage the packings in the column. We will begin our analysis by examining the relationship between the gas pressure drop and gas velocity. Refer to the Figure below that shows a typical gas pressure drop in a packed column.

The horizontal axis is the logarithmic value of the gas velocity G, and the vertical axis is the logarithmic value of pressure drop per height of packing [ pressure drop in a packed bed is the result of fluid friction that is created by the flow of gas and liquid around the individual solid packing materials ]. Note: Each packing has its own characteristics pressure drop chart as reported by the manufacturer - for example, see the Figure above (right). [ Back on Top ]

Analysis of Gas Pressure Drop in Packing With a dry packing (i.e. no liquid flow, L = 0), pressure drop increases as gas velocity increases according to the linear relationship as shown by line a-a. This is a straight line on a loglog plot. With liquid flowing in the column, the packings now become wetted (irrigated). Part of void volume in the packings now filled with liquid, thereby reducing the cross-sectional area available for gas flow. At the same gas velocity, the pressure drop is higher for wetted packings compared to dry packings. For example, compare the case for L = 0 vs. L = 5. The line for DP/L under wetted condition lies to the left of line a-a. For a constant liquid flow (say L = 5), at low to moderate gas velocity G; the pressure drop characteristics is similar to that of dry packings, i.e. section b-c of the plot is still straight on log-log plot. Up to this point, there is an orderly trickling of the liquid down the packings. There is no observable liquid being trapped among the packings (no liquid hold-up). As the gas velocity is increased further, the pressure drop increased. Some liquid started to be retained in the packings. When point c is reached, the quantity of liquid retained in the packed bed increases significantly. There is a change in slope

of the line at point c as pressure drop increases more rapidly with G. Point c is known as the loading point, as liquid starts to accumulate (load) in the packings. From point c to d to e, there is a sharp increase in pressure drop at higher G: there is a greater amount of liquid hold-up, a gradual filling of the packing voids with liquid (starting at the bottom of the column), and the column is slowly "drowned" in the liquid. At point e, there is another sharp change in the slope. At this point the entire column is filled liquid and the gas now has to bubble through the liquid in the packing voids. The gas pressure drop is now very high. Point e is known as the flooding point. The gas velocity at this point is known as the flooding velocity (limiting velocity). Points to note : - at constant liquid rate, gas pressure drop increases with gas velocity. - at constant gas velocity, the gas pressure drop is higher at larger liquid rate. - each liquid rate has its own loading and flooding points. - at higher liquid rate, the loading and flooding points occur at lower gas pressure drop. Operation of a gas absorption column is not practical above the loading point. For optimum design, the recommended gas velocity is 1/2 of the flooding velocity. Alternatively, some design can be based on a specified pressure drop condition, usually well below the pressure drop at which flooding would occur.

Potrebbero piacerti anche