Scrubber Design
Scrubber Design
USER MANUAL
Contents
1.0 Introduction
5.0 Nomenclature
6.0 References
Appendix
MNL102A 02/03
Page 2 of 26
SCRUBBER
Fast Chemical Reaction Rating & Design
1.0 Introduction
The software is designed to predict the performance of randomly packed wet gas scrubbing
columns operating counter currently at atmospheric pressure and involving mass transfer
by fast chemical reaction of binary systems.
The software can be used to predict and optimise the performance of an existing column or
for the design of a new column. The package is supported by an extensive physical
property database which is used to predict gas diffusivity, gas viscosity and allowable
scrubber outlet concentrations for the most common gas phase reaction species to be found
in the Fine Chemicals industry. The column design can be based on a specified pressure
drop per metre or as % flood; the calculated column diameter is then adjusted to satisfy
manufacturing constraints and the design is optimised.
The design liquor rate can be determined on column temperature rise from reaction
stoichiometry and heat of reaction or as a multiple design factor of minimum wetting rate.
The stoichiometric volume required is also calculated.
The database for the selected packing is based on Plastic Pall rings in the size range of
15mm to 89mm. Custom selection is also allowed.
The software platform is Microsoft Excel which allows the database for reactant species,
reactions and packings to be readily modified to meet specific customer requirements.
The software is intended as ‘task specific’ and does not allow for special features such as
unit change facilities or adjustment of correlation parameters.
MNL102A 02/03
Page 3 of 26
SCRUBBER
Fast Chemical Reaction Rating & Design
The components to be processed are of two types and are entered as component name in
CAPS and with no spaces e.g. HYDROGENBROMIDE.
Once selection has been made the database automatically sets the design outlet
concentration based on LTEL (Reference 3) or odour threshold limit values. A design
factor allows for further adjustment if required.
Flows
The inert flow and reactant gas flows are entered in kg/hr. Validation has been set at a
minimum flow of 10 kg/hr for the inert flow to prevent calculation error messages.
The gas phase reactant selection sets the liquid phase reactant automatically to allow
prediction of reaction stoichiometry and heat of reaction.
Some gas phase reactants e.g. Amines allow for a selection of scrubbing liquor to be made,
select accordingly.
If gas evolution from the reaction is specified in kg the design stoichiometric volume will
be calculated.
Liquor rate can be determined on column temperature rise or as a multiple of MWR, select
accordingly.
Entering the design maximum temperature rise of liquid across the column sets an initial
liquid rate. Depending upon the application it may be desirable to set the liquor rate as a
multiple factor of the Minimum Wetting Rate subject to satisfying the stoichiometric
requirements.
Packing
Select packing to be used or input customer specified parameters in the units specified.
MNL102A 02/03
Page 4 of 26
SCRUBBER
Fast Chemical Reaction Rating & Design
Operation of Software
It must be appreciated that the calculations interact as the design is optimised. The column
diameter effects the liquid loading and hence the minimum wetting rate which in turn
effects the gas loading and column pressure drop. The height of a transfer unit (HTU) is
also effected by the gas loading.
The designer should ensure on completion of design that all parameters e.g. diameter, %
flood, liquor rate, are consistent i.e. % flood fields equal, liquor rate as required and
diameter fields within range.
Design
Set an initial column diameter. The predicted flood is displayed for this diameter.
If design is based on DESIGN FLOODING set design flood value. Column design results
will display relevant column diameter.
The design liquid rate will be based on design MWR factor or design temperature.
The Initialisation column diameter is now reset to proposed design and the design is
optimised by manual iteration until Design Flood, Predicted Flood and Design Liquid rate
parameters are satisfactory.
If the design is based on Pressure Drop / metre of packing the flooding parameter is
automatically set based on column diameter. The same optimisation procedure is followed
as before.
Rating
Adjust Design Flood to equal Predicted Flood and reset Liquor Rate to maintain desired
value.
MNL102A 02/03
Page 5 of 26
SCRUBBER
Fast Chemical Reaction Rating & Design
The model is based on the specific case of absorption with fast chemical reaction. This
special case assumes that the concentration of reactant gas at the interface is ‘0’ and that
liquid phase resistance to transfer is ‘0’. The model presupposes that there is adequate
reactant available i.e. liquid rate and composition of reacting component in the liquid, to
satisfy the stoichiometric requirement of the reactant gas phase load.
The column loading and mass transfer coefficient are based on conditions at the gas inlet.
The model is therefore based on a constant temperature and gas mass flowrate throughout
the column. In the case of absorption of gas streams with low concentrations of reactant in
the gas phase, as is found in environmental scrubbing of Local Extract Ventilation systems,
the model conditions are satisfied.
However, in the case of absorption of gas streams with a high concentration of reactant in
the gas phase, as is found on reaction vent systems, the model is conservative, i.e.
overpredicts the packed height requirement, which is safe. The justification for this claim is
as follows:-
The overall gas phase mass transfer coefficient Kg varies with temperature
1
Kg 0.56
T
Increasing film temperature decreases Kg. This effect is offset by the dependency of gas
diffusivity on temperature where
K g T 0.315
MNL102A 02/03
Page 6 of 26
SCRUBBER
Fast Chemical Reaction Rating & Design
G
HTU where G v the velocity
K g a P - p lm
The term (P – p)lm approaches unity for dilute gas streams and becomes significant for
concentrated gas streams.
HTU v 0.25
Decreasing gas velocity results in a decrease in the HTU. The HTU could be calculated in
stages up the tower resulting in a lower packed height requirement. This approach has not
been adopted as it is recognised that the correlation for Kg prediction requires a minimum
Reynolds number to be achieved which is probably not achieved towards the top of the
tower, i.e. streamline flow prevails at some point. This maldistribution will result in an
effective increase in the HTU requirement which is difficult to quantify and in any event a
minimum of 600mm distribution allowance is recommended to be included in the design
packed height. The distribution allowance should be reviewed for each application.
v v g vl
The model assumes liquid velocity vl = 0 since at high velocities the correction is negligible
and at low velocities the gas may be in streamline flow as discussed above.
It is recognised that the model is not based on a rigorous analysis of the physical factors
involved. However, the main application is on multipurpose fine chemicals processes which
by their nature have a wide variation in operating conditions which at the design stage may
not even be identified.
The packed height prediction is conservative for concentrated gas streams as no allowance is
made for the increase of (P – p)lm up the tower and this is justified as the increase in cost
involved represents a small percentage of the total cost of the facility, probably less than
5%.
MNL102A 02/03
Page 7 of 26
SCRUBBER
Fast Chemical Reaction Rating & Design
The diffusion coefficient DAB for a binary mixture of gases A and B may be estimated from
the Fuller, Schettler and Giddings Relationship.
where T temperature °K
P pressure atm
MA and MB molecular weights of components A & B
A and B are the atomic diffusion volumes of components A & B
The program uses the atomic diffusion volumes as detailed in Table 3-342 in Reference 4.
This relationship will predict DAB to within 5 to 10% over the normal operating temperature
range for wet gas scrubbing at atmospheric pressure.
The gas viscosity is predicted using the general purpose Chemcad Library equation 102. The
number of coefficients required for this equation are 2 or 3.
A x TB
Viscosity Y
C
1
T
Where A, B and C are coefficients for the gas reactant species from the Chemcad Databank.
The program uses the viscosity of the gas mixture at the inlet conditions where for species i
y i i M i ½
1
y i M i ½
MNL102A 02/03
Page 8 of 26
SCRUBBER
Fast Chemical Reaction Rating & Design
The reactant species A and flowrate WA in kg/hr is entered and molecular weight
MA kg/kmole is read from the physical property database (PPD).
The inert species B and flowrate WB in kg/hr is entered and molecular weight MB kg/kmole
is read from PPD.
WA / M A
Inlet mole fraction of A yA
WA / M A WB /M B
WA WB
Mean molecular weight M AB
WA / M A WB /M B
M AB 273
Inlet gas density 1
22.4 273 T
M 273 mg
Inlet concentration Component A y A 10 6 A
22.4 273 x T m3
If the LTEL value is given in EH40 (Reference 3) this is read directly. If only STEL is
available this value is taken and 0.10 (STEL) is read. For malodorous components such as
amines the low odour threshold value is taken.
MNL102A 02/03
Page 9 of 26
SCRUBBER
Fast Chemical Reaction Rating & Design
The liquor composition is set by the program based on reactant species A. This selection is
set as follows:-
Based on the reaction selected the heat of reaction is set from the PPD.
MNL102A 02/03
Page 10 of 26
SCRUBBER
Fast Chemical Reaction Rating & Design
If the heat of reaction is significant the liquor rate is set using the liquor temperature rise
across the column as the basis:-
H R WA / M A
Liquor rate Wl
(t1 t 2 ) C p l
Where
H R heat of reaction kcal/kmole
WA / M A component A flowrate kmoles/hr
t1 - t 2 liquor temperature rise K
C p l liquor specific heat kcal/kg K
WA (n l M l ) kg
Stoichiometric rate Wl
M A C P l 0.01 hr
The wetting rate in the column is a function of the column diameter and the packing
characteristics.
4 Wl m3
d 2 l a hm
Minimum wetting rate = 0.0792 m3 / h m for Plastic Pall rings which is taken as a constant
The design MWR factor applied will depend on the specific application but should be
greater than 1.3.
MNL102A 02/03
Page 11 of 26
SCRUBBER
Fast Chemical Reaction Rating & Design
The liquor volume to satisfy stoichiometric requirements is calculated from the mass of
reactant gas predicted release.
w A nl Ml
Stoichiometric volume m3
M A C l 0.01 l
The actual circulation tank volume selected will depend on the number of reactions to be
carried out without a liquor change, heat balance considerations and the safety factor applied
for moles of liquid reactant left remaining.
MNL102A 02/03
Page 12 of 26
SCRUBBER
Fast Chemical Reaction Rating & Design
In Figure 1 we consider the rate of absorption across an element dz in a tower of area A and
with a packing specific volume a.
Na dz K g aP (y y*) dz GA dy LA dx
G dy dy
Packed height Z
K g ap (y y*)
where (y y*)
is defined as the number of overall gas transfer units (NTU) and represents an integrated
value of the number of times the driving force must be divided into the change in gas
concentration. It therefore describes the difficulty of the separation.
G
The group is defined as the height of a transfer unit (HTU).
K g aP
For the special case of absorption followed by fast chemical reaction (Reference 2, page 94)
where y* = 0.
ln(1 y1 )
Nog = ln
ln(1 y 2 )
This expression is accurate for gas compositions varying over a wide range and for low
concentrations reduces to
y
Nog = ln 1
y2
MNL102A 02/03
Page 13 of 26
SCRUBBER
Fast Chemical Reaction Rating & Design
ln 1 y1
NTU N og ln
ln 1 y 2
Thus calculation of the inlet mole faction y1 of the reactant gas and the selection of the
design outlet mole faction y2 sets the NTU.
To predict the outlet concentration of an existing tower the effective NTU is calculated from
the HTU and the existing packed height minus an allowance for distribution.
ln 1 y1
y 2 exp
exp NTU
MNL102A 02/03
Page 14 of 26
SCRUBBER
Fast Chemical Reaction Rating & Design
A correlation has been developed for predicting the gas film coefficient in a falling film
tower of 1 in diameter.
v g
K g 0.04 Re -0.25 Sc -0.5 (Drift Factor)
P
Where,
g = density of soluble gas at temperature and total pressure of the gas mixture g/cm2
vd
Reynolds Number Re Schmidt Number Sc
D
v = gas velocity relative to liquid surface cm/s
The correlation is valid for conditions of turbulent flow (Re > 5000).
In the literature the Re exponent varies between –0.17 and –0.25, thus the correlation value
is conservative. The Sc exponent varies between –0.5 to –0.56, however, Morris and
Jackson (Reference 2) consider –0.5 to be more appropriate for absorption.
MNL102A 02/03
Page 15 of 26
SCRUBBER
Fast Chemical Reaction Rating & Design
P
Drift Factor =
P - p lm
Where p = partial pressure of reactant gas in atmospheres.
`
pg pi
P - plm
P pi
ln
P - pg
Where pg and pi are the actual and interfacial partial pressures of the soluble gas
respectively.
For the special case of fast chemical reaction pi = 0 and for a system pressure of P = 1 we
have
y1 Note p1 = y1
P - plm
1
ln
1 - y1
MNL102A 02/03
Page 16 of 26
SCRUBBER
Fast Chemical Reaction Rating & Design
The program uses the Morris and Jackson gas film coefficient correlation in the following
modified form where the units for Kg are lb mole/hr ft2 atm and the Reynolds number
dimension parameter d = 0.083 ft representing the falling film column diameter.
0.25 0.5
v vd P
Kg = 0.04
RT D p BM
lb mole
Kg = gas film coefficient
hr ft 2 atm
ft 3 atm
R = gas constant = 1.314
lb mole K
T = absolute temperature °K
ft
v = gas velocity
hr
lb
= viscosity
ft hr
lb
= density of gas mixture
ft 3
ft 2
D = diffusivity
hr
y1
where p BM for p i 0
1
ln
1 - y1
MNL102A 02/03
Page 17 of 26
SCRUBBER
Fast Chemical Reaction Rating & Design
The gas velocity v is assumed to be equal to the velocity in the empty tower divided by
fractional voidage of the packing ().
Where
v v g vl v l 0 assumed
vg ft
v
3600 sec
Wt 3.281 ft
v
t A sec
Wt 0.025
Re
A
To scale up from the falling film column a gas film packing factor Rg, to compensate for the
increase in turbulence in a packed tower, is applied to give a corrected gas film coefficient.
Kcg = Rg x Kg
For random rings Rg varies from 2.5 to 3.3 (Reference 1, Table 6). The program uses a
value of 3.0 which is conservative.
MNL102A 02/03
Page 18 of 26
SCRUBBER
Fast Chemical Reaction Rating & Design
1
g l g 9.81 2 kg
G '
Y F
4 0.12
sec m 2
Wt
Area A m2
3600 G '
0.5
4
d 1000 A mm
MNL102A 02/03
Page 19 of 26
SCRUBBER
Fast Chemical Reaction Rating & Design
The column diameter and loading is determined from the generalised pressure drop
L
correlation due to Sherwood Eckert in the form (Reference 7, Figure 4).
G
The graph axes are defined as follows:-
1
L g 2
X dimensionless
G l
Y
g l g g 4
dimensionless
G ' 2 F 0.12
cps
Curves are presented for different pressure drops in mm wc/m packed height namely
4, 8, 21, 42 mm/m, flood. Each curve is broken down into two straight line equations to
enable model predictions to be made.
As before G’ at flood is calculated and hence flooding velocity for the diameter selected.
MNL102A 02/03
Page 20 of 26
SCRUBBER
Fast Chemical Reaction Rating & Design
5.0 Nomenclature
MNL102A 02/03
Page 21 of 26
SCRUBBER
Fast Chemical Reaction Rating & Design
6.0 References
MNL102A 02/03
Page 22 of 26
SCRUBBER
Fast Chemical Reaction Rating & Design
Appendix I
MNL102A 02/03
Page 23 of 26
SCRUBBER
Fast Chemical Reaction Rating & Design
G’2 L’1
y2 x1
y - dy x
dz
y x - dx
G’1 L’2
y1 x2
Column Area A
Packing specific volume a
Na dz K g aP (y y*) dz GA dy LA dx
MNL102A 02/03
Page 24 of 26
SCRUBBER
Fast Chemical Reaction Rating & Design
Appendix II
MNL102A 02/03
Page 25 of 26
SCRUBBER
Fast Chemical Reaction Rating & Design
MNL102A 02/03
Page 26 of 26