Absorbers

Introduction ............................................................... 110 H y d r o c a r b o n A b s o r b e r Design ................................ 110 H y d r o c a r b o n Absorbers, Optimization .................. 112 Inorganic Type ........................................................... 113

109

110

Rules of Thumb for Chemical Engineers

Introduction
A general study of absorption can be confusing since the calculation methods for the two major types are quite different. First, there is hydrocarbon absorption using a lean oil having hydrocarbon components much heavier than the component absorbed from the gas stream. These absorbers may or may not be reboiled. For designing these, one uses equilibrium vaporization constants (K values) similarly to distillation. Another similarity to distillation is the frequent use of fractionating trays instead of packing. Canned computer distillation programs usually include hydrocarbon absorber options. The other major type is gas absorption of inorganic components in aqueous solutions. For this type design one uses mass transfer coefficients. Packed towers are used so often for this type that its discussion is often included under sections on packed towers. However, in this book it is included here.

Source
Branan, C. R., The Process Engineer's Pocket Handbook, Vol. 1, Gulf Publishing Co., 1976.

Hydrocarbon Absorber Design

Because of its similarity to distillation, many parts of this subject have already been covered, such as 1. Tray Efficiency 2. Tower Diameter Calculations 3. K Values As for distillation, shortcut hand calculation methods exist, for hydrocarbon absorption. In distillation, relative volatility (~) values are generated from the K values. For hydrocarbon absorption the K values are used to generate absorption and stripping factors. The 1947 Edmister 1 method using effective overall absorption and stripping factors and the well-known Edmister graphs are very popular for hand calculations. An excellent write-up on this and the Kremser-Brown-Sherwood methods are on pages 48-61 of Ludwig. 2 where n= m Ea = EsAe = Se = Number of theoretical stages in absorber Number of theoretical stages in stripper Fraction absorbed Fraction stripped Effective absorption factor Effective stripping factor

Edmister Method (1947). Briefly, the Edmister absorption method (1947) with a known rich gas going to a fixed tower is as follows:
1. Assume theoretical stages and operating temperature and pressure. 2. Knowing required key component recovery Ea, read Ae from Figure 1 at known theoretical trays n.

3. Assume a. Total mols absorbed b. Temperature rise of lean oil (normally 20-40~ c. Lean oil rate, mols/hr 4. Use Horton and Franklin's 3 relationship for tower balance in mols/hr. This is shown in Table 1. 5. Calculate L1/V1 and Ln/Vn. 6. Use Horton/Franklin method to estimate tower temperatures. This is shown in Table 2. 7. Obtain top and bottom K values.
Table 1 Tower Balance Quantity Rich gas entering at b o t t o m Gas Absorbed Lean gas leaving absorber Gas leaving bottom tray Gas leaving tray 2 from top Lean oil Liquid leaving top tray Liquid leaving bottom tray Symbol
Vn. 1

Equation Known Assumed

V1 = Vn + 1 - AV
Vn =Vn + I(V1/Vn + 1)1/N V 2 - V l / ( V l / V n + 1)1/N

Ae
Ea "-

-me n+l Ae - 1
M+I M+I

n+l

AV
V1 Vn
V2

(1)

Se
ES ~

-Se -1 (2)

L0 L1
Ln

Known L1 = L0 + V2 - V1
Ln = Lo + AV

Se

Absorbers

111

1,0

84 .

0.95, 0.90

o esl
<:

0.60 ' " 0.75 ~

mmmmmm

q~

070 ~
0.65 I

mmm
" -

(se

w=,Mcaim mm wmmm

V,"qr~

0.40!
O~le~ "I"1 = - ~ O"O" l--l* j

a)saeNipnimmme=ammeaammm

,"L.$-.'~ I I ]_j~o,~ o.~5 1 ~ i~r~_..~ ~.~.Ip_|illlln'

l~~i:

1111

I I I I L J - J - ~ --r,

o.=o . ~ . , ~ ~ o.os, ~"FIl

~

0.2 0.3 0.4 0.5 0.6 o.7 o.e 0.9 =.o i2

Values of Ae or Se

i L.J--F.-T-'t'-I-N 1--i""i~-' I i 11 11l I 1I 111 1 1 I1 1 1 i 1 I 1 l II 1I 1 l I

L4

1.6

1.8 2.0

3.0

4.0

5.0

6.0

Figure 1. This graph shows the abosrption and stripping factors, Ea and Es, versus effective values, Ae and S e (efficiency functions). (By permission, W. C. Edmister, Petroleum Engineer, September, 1947 Series to January, 1948.)

Table 2 Tower Temperatures

Temperature
Rich gas Lean oil

Symbol
Tn.l To AT Tn

Equation
Known Known

10. Read Eai values from Figure 1. 11. Calculate mols of each component absorbed. 12. Compare to assumed total mols absorbed and reassume lean oil rate if necessary.
Edmister Method (1957). Edmister has developed an improved procedure 4 that features equations combining absorption and stripping functions as follows:
V1 - ~)aVn+l + (1 -- ~)s)ko
L1 - l~sLm+, + (1 - ~)a )Vo

inlet

Temperature rise
B o t t o m Tray

Assumed
T n = Tn+ 1 +

AT (Vn+ 1 - V 2 )

Top Tray

T1

T1= Tn -

AT (Vn+ 1 - V l )

(Absorption Section) (Stripping Section)

(9) (10)

where 8. Calculate absorption factors for each component i at the top and bottom
ATi -- L 1 / K l i Vl (3)

ABi - L n / K n i Vn For stripping factors

STi -- V, K , i / L , SBi -- Vn K n i / L n

(4)

L1 = Liquid from bottom stripping tray Lm+l = Liquid to top stripping tray * a -" 1 - Ea, fraction not absorbed 0s = 1 - Es, fraction not stripped Vo = Vapor to bottom stripping tray Other symbols are defined in Tables 1 and 2. Figure 1 and Equations 3-8 are used as before. V~ and L1 are found from Equations 9 and 10. The improved procedure is better for rigorous solution of complicated absorber designs.
Lean 0il. The selection of lean oil for an absorber is an economic study. A light lean oil sustains relatively high lean oil loss, but has the advantage of high mols/gal compared to a heavier lean oil. The availability of a suitable

(5)
(6)

9. Obtain Aei from

Ae -

lAB (AT + 1) + 0.2511/2 - 0 . 5

(7)

Similarly
S e - [ST(S B "1"1)+ 0.251 '/2 - 0.5 (8)

112

Rules of Thumb for Chemical Engineers

material has a large influence on the choice. A lean oil 3 carbon numbers heavier than the lightest component absorbed is close to optimum for some applications. In natural gas plant operations, however, the author generally sees a lean oil heavier by about 10-14 carbon numbers.
Presaturators. A presaturator to provide lean oil/gas contact prior to feeding the lean oil into the tower can be a good way of getting more out of an older tower. Absorber tray efficiencies run notoriously low. A presaturator that achieves equilibrium can provide the equivalent of a theoretical tray. This can easily equal 3-4 actual trays. Some modem canned computer distillation/absorption programs provide a presaturator option.

Sources

1. Edmister, W. C., Petroleum Engineer, September, 1947 Series to January, 1948. 2. Ludwig, Applied Process Design For Chemical and Petrochemical Plants, 2nd Ed., Vol. 2, Gulf Publishing Company, 1979.

3. Horton, G. W. and Franklin, B., "Calculation of Absorber Performance and Design," Ind. Eng. Chem. 32, 1384, 1940. 4. Edmister, W. C., "Absorption and Stripping-factor Functions for Distillation Calculation by Manual- and Digital-computer Methods," A.L Ch.E. Journal, June 1957. 5. Fair, James R., "Sorption Processes for Gas Separation," Chemical Engineering, July 14, 1969. 6. Zenz, E A., "Designing Gas-Absorption Towers," Chemical Engineering, November 13, 1972. 7. NGPSA Engineering Data Book, Natural Gas Processors Suppliers Association, 9th ed., 1972. 8. Norton, Chemical Process Products, Norton Company, Chemical Process Products Division. 9. Treybal, R. E., Mass Transfer Operations, McGrawHill Book Co., Inc., New York, 1955. 10. Rousseau, R. W. and Staton, S. J., "Analyzing Chemical Absorbers and Strippers," Chemical Engineering, July 18, 1988. 11. Diab, S. and Maddox, R. N., "Absorption," Chemical Engineering, December 27, 1982. 12. Branan, C. R., The Process Engineer's Pocket Handbook, Vol. 1, Gulf Publishing Co., 1976.

HydrocarbonAbsorbers, Optimization
This section is a companion to the section titled Fractionators-Optimization Techniques. In that section the Smith-Brinkley 1 method is recommended for optimization calculations and its use is detailed. This section gives similar equations for simple and reboiled absorbers. For a simple absorber the Smith-Brinkley equation is for component i: where

hi- correlating factor; if the feed is mostly liquid, use

Equation 1 and if mostly vapor, Equation 2. hi - (K~//Ki)(L/L')[(1 - S n)/(1 - S m)]i hi - (L/L')[(1- Sn ) / ( 1 - Sm )]i
Nomenclature

(1) (2)

f _ (1- sN)+qs(SN -- S)
1
-

S N+I

where

f i - (BXB/FXF)i Smi- K'iV'/L' Sni- KiV/L

For a reboiled absorber:

f_

(1-- anN-M) Wqs( SnN-M - an )

(1 -- SnN-M) -+-hSnN-M (1 -- SmM+I)

B - Bottoms total molar rate, or subscript for bottoms F - Feed total molar rate, or subscript for feed f i - Ratio of the molar rate of component i in the bottoms to that in the feed Ki- Equilibrium constant of component i in top section = y/x K'i-Equilibrium constant of component i in bottom section- y/x L - Effective total molar liquid rate in top section L ' - E f f e c t i v e total molar liquid rate in bottom section

Absorbers

113

M = Total equilibrium stages below feed stage including reboiler N = Total equilibrium stages including reboiler and partial condenser qs = Fraction of the component in the lean oil of a simple or reboiled absorber S = Overall stripping factor for component i Sm = Stripping factor for component i in bottom section Sn = Stripping factor for component i in top section V = Effective total molar vapor rate in top section

V' - Effective total molar vapor rate in bottom section X - Mol fraction in the liquid Y - Mol fraction in the vapor

Sources
1. Smith, B. E., Design of Equilibrium Stage Processes, McGraw-Hill, 1963. 2. Branan, C. R., The Process Engineer's Pocket Handbook, Vol. 1, Gulf Publishing Co., 1976.

Inorganic Type
Design of inorganic absorbers quite often involves a system whose major parameters are well defined such as system film control, mass transfer coefficient equations, etc. Ludwig 1 gives design data for certain well-known systems such as NH3-Air-H20, C12-H20, CO2 in alkaline solutions, etc. Likewise, data for commercially available packings is well documented such as packing factors, HETR HTU, etc. If KGAvalues are available for a known system, those of an unknown system can be approximated by KCA ( u n k n o w n ) - KGA (known)(Dvunkn~ Dvknown where KCA -- Gas film overall mass transfer lb mols/hr (ft 3) (ATM) Dv -Diffusivity of solute in gas, ftZ]hr coefficient, ~

Film Control. The designer needs to know whether his system is gas or liquid film controlling. For commercial processes this is known. In general, an absorption is gas film controlling if essentially all resistance to mass transfer is in the gas film. This happens when the gas is quite soluble in, or reactive with, the liquid. Ludwig 1 gives a listing of film control for a number of commercial systems. If a system is essentially all gas or liquid film controlling, it is common practice to calculate only the controlling mass transfer coefficient. Norton 2 states that for gas absorption, the gas mass transfer coefficient is usually used, and for stripping the liquid mass transfer coefficient is usually used.

Diffusivities. The simplest gas diffusivity relationship

is the Gilliland relationship. Dv - 0.0069

T3/2(1/MA + 1/MB ),/2

P(V~/3 + V~/3)2

where T = Absolute temperature, ~

Mass Transfer Coefficients. General equations for mass transfer coefficients are given in various references if specific system values are not available. These must, however, be used in conjunction with such things as packing effective interfacial areas and void fractions under operating conditions for the particular packing selected. It is usually easier to find KGAfor the packing used with a specific system than effective interfacial area and operation void fraction. Packing manufacturers' data or references, such as Ludwig ~ can provide specific system KGA or KLA data. Tables 1-6 show typical KGAdata for various systems and tower packings.

MA, MB = Molecular weights of the two gases, A and B

P = Total pressure, ATM VA, VB = Molecular volumes of gases, cc/gm mol

Height of Overall Transfer Unit. Transfer unit heights are

found as follows"
Gm H o G --

KGAPAvR
Lm

H o L --

KLApL

114

Rules of Thumb for Chemical Engineers

Table 1

KGA For Various Systems 4

KGA
Solute Gas Br2 Cl 2 CI2 Absorbent Liquid 5% NaOH Water 8% NaOH 4% NaOH Water Water Water Water Water 4% NaOH Water Water 11% Na2CO3 Lb mols/(hr 9 ft 3 9 atm)

Table 4 Overall Mass Transfer Coefficient CO2/NaOH System s Plastic Tower Packings

KGA
Packing (Ib-mol/h 9 ft 3 9 atm) 2.80 1.92 1.23 2.37 2.64 2.25 2,09 1.23 #1 Super Intalox| Packing #2 Super Intalox| Packing #3 Super Intalox | Packing Intalox| Snowflake| Packing 1 in, Pall Rings 189in. Pall Rings 2 in. Pall Rings 389in. Pall Rings

CO2
HBr HCHO HCI HCN HF H2S NH3 SO2 SO2

5.0 4.6 14.0 2.0 5.9 5.9 19.0 5.9 8.0 5.9 17.0 3.0 12.0

Table 5 Overall Mass Transfer Coefficient CO2/NaOH System s Ceramic Tower Packings

K~ Table 2
Relative KGA For Various Packings 4 Type of Packing Super Intalox| Intalox| Saddles Hy-Pak| Pall Rings Pall Rings Maspac Tellerettes Raschig Rings Material Plastic Ceramic Metal Metal Plastic Plastic Plastic Ceramic Relative KGA 1.00 0.94 1.11 1.06 0.97 1.00 1.19 0.78 Packing 1 in. Intalox| Saddles 189in. Intalox| Saddles 2 in. Intalox| Saddles 3 in. Intalox| Saddles 1 in. Raschig Rings 189in, Raschig Rings 2 in. Raschig Rings 3 in. Raschig Rings (Ib-mol/h 9 ft 3 9 atm) 2.82 2.27 1.88 1.11 2.31 1.92 1.63 1.02

Table 6 Overall Mass Transfer Coefficient CO2/NaOH System s Structured Tower Packings

Table 3 Overall Mass Transfer Coefficient CO2/NaOH System 5 Metal Tower Packings

K~
Packing Intalox| Structured Packing 1T Intalox| Structured Packing 2T Intalox| Structured Packing 3T (Ib-mol/h * ft s * atm) 4.52 3.80 2.76

KGA
Packing #25 IMTP| Packing #40 IMTP| Packing #50 IMTP| Packing #70 IMTP| Packing 1 in. Pall Rings 189in. Pall Rings 2 in. Pall Rings 389in. Pall Rings #1 Hy-Pak| Packing #189 Hy-Pak| Packing #2 Hy-Pak| Packing #3 Hy-Pak| Packing (Ib-mol/h 9 ft 3 9 atm) 3.42 2.86 2.44 1.74 3.10 2,58 2.18 1.28 2.89 2.42 2.06 1.45

where

HoG,HOL--Height of transfer unit based on overall gas

or liquid film coefficients, ft Gm, Lm - G a s or liquid mass velocity, lb mols/(hr) (ft 2) KGA,KLA-Gas or liquid mass transfer coefficients, consistent units P A r R -" Average total pressure in tower, ATM DE --Liquid density, lb/ft 3

Absorbers

115

Number of Transfer Units. For dilute solutions the number of transfer units NoG is obtained by

X - M o l fraction in the liquid at the same corresponding point in the system as Y 1, 2 - Inlet and outlet of the system, respectively
Sources

Y1-Y2
( Y - Y * ) I - ( Y - Y*):
N o G --

In (Y - Y *)1 (Y-Y*)2

where

( Y - Y*) = Driving force, expressed as mol fractions Y = Mol fraction of one component (solute) at any point in the gas phase Y* = Mol fraction gas phase composition in equilibrium with a liquid composition, X

1. Ludwig, E. E., Process Design for Chemical and Petrochemical Plants, Vol. 2, Gulf Publishing Co., 1965. 2. Norton Chemical Process Products, Norton Company, Chemical Process Products Division. 3. Branan, C. R., The Process Engineer's Pocket Handbook, Vol. 1, Gulf Publishing Co., 1976. 4. Branan, C. R., The Process Engineer's Pocket Handbook, Vol. 2, Gulf Publishing Co., Houston, Texas, 1983. 5. Strigle, R. E, Packed Tower Design and Applications, 2nd Ed., Gulf Publishing Co., Houston, Texas, 1994.