ENSC3019
/CHPR8503
:
Week 6 Design of Packed Columns
1 Dr Kevin Li
Recommended reading:
McCabe et al., Unit Operations of Chemical Engineering, Chapter 21
Treybal, R. E. Mass Transfer Operations, 3rd Edn. McGraw-Hill 1955, Chapter 9 Coulson, J. M. and Richardson, J. F. Chemical Engineering, Volume 6: Particle Technology and Separation Processes, 5th Edn. Butterworth-Heinemann 2002
Mass Transfer (MT) across phase interface:
two-resistance model
2
Gas film Liquid film
Bulk gas Bulk liquid
yA,G yA,i xA,i xA,L distance Resistances to diffusion of A:
(i) in the gas phase film (ii) in the liquid phase film
At the interface: assume local equilibrium between yA and xA, no resistance to MT across the interface
(
, ,)
A y A g A iN
=
k
y
−
y
N
A=
k
x(
x
A i,−
x
A L,)
1 ky ∝ 1 kx ∝Mass-transfer coefficients:
an engineering concept that allows us to simplify complex diffusion problems. 3
(
)
A y iN
=
k
y
−
y
Flux (mole/m2/s) Coefficient Driving force (concentration difference)=
×
Since concentration could be defined in different ways, a variety of coefficients can be defined:
Summary of general forms of MT Rates for two-phase films
ky is local MTC for gas phase
yiis mole fraction (of component A) in gas at the gas-liquid interface , y is bulk vapour composition
kx is local MTC for liquid phase
xi is mole fraction (of component A) in liquid at the gas-liquid interface, x is bulk liquid composition
Ky is overall MTC for gas phase
y* is composition of vapour that would be in
equilibrium with the bulk liquid of composition x Kx is overall MTC for liquid phase
x* is composition of vapour that would be in
equilibrium with the bulk vapour of composition y
4
(
)
A y iN
=
k
y
−
y
(
)
A x iN
=
k
x
−
x
(
*)
A xN
=
K
x
−
x
(
*)
A yN
=
K
y
−
y
MTC=mass transfer coefficient. Subscripts A, and G, L dropped here for simplicity.
See McCabe et al. page 547-548. Or if you’re keen for more discussion look at Treybal’s Chapter 5..
m’ is local slope of equilibrium curve
i.e.
1
1
'
y y xm
K
=
k
+
k
(
*)
( ) ' i i m = y − y x − xTutorial 1 Equilibrium for component A between air and water is described by Henry’s law y*=4x. The local mass transfer coefficients are kx=2 mol m-2s-1 and k
y=1 mol m-2s-1. (1) What is the overall mass transfer coefficient for gas
phase?
(2) Evaluate the flux of A between phases at a point in a column where bulk compositions are 0.08 mole fraction in the gas and 0.01 mole fraction in the liquid.
6.1 Packed columns for absorption
Dr Kevin Li
Consultation hours 15:00-17:00Thursdays
2.49A in Civil & Mech Eng building
Equipment for gas-liquid absorption
Need intimate contact between the immiscible
phases to achieve mass transfer (MT) between
phases.
Flux
N
A rate of transfer per unit area of gas-liquid interface
Engineering MT equipment focuses on increasing
Main equipment types
Packed columns
Random
(let to fall randomly into column during installation)Structured
(engineering for lower ΔP, higher cost )Tray columns -
liquid levels on each trayColumn internals
9
GREEN, D. W. & PERRY, R. H. (eds.) (2008).
Perry's chemical engineers' handbook, New York: McGraw-Hill.
Packing material, plus Liquid inlet systems
Liquid & vapour distributors Liquid collecting devices
Packing supports
Good info at manufacturer
Packed columns – random packings
10 Metal pall rings Raschig rings
VSP Inner arc ring
see more images at
Structured packings
www.sulzerchemtech.com
Mellapak
TMwww.sulzerchemtech.com
Tray columns
V-grid www.sulzerchem.com Sieve tray
High performance trays
eg. Shell calming section tray
14
1. Tray columns can be designed to handle a wider range of liquid and gas flow rates. Packed columns are not suitable for very low liquid rates.
2. The efficiency and performance of a tray column can be more accurately predicted.
3. Easier to make provisions for withdrawal side streams in plate columns.
4. Fouling & cleaning: can install manholes on trays. However, may be easier to replace packing when fouled.
Plate columns can be designed with more assurance - some doubt that good liquid distribution can be maintained in a packed column.
It is easier to provided cooling or heating in a plate column – coils directly on plates.
Coulson and Richardson Vol 6. list some of the factors which influence choice of trays or packing in a column:
15
Trays/Plate columns vs. Packed
columns
5. For corrosive liquids a packed column will be cheaper than a plate column (due to materials).
6. The liquid hold-up is lower in a packed column. Important if amount toxic or flammable liquid needs to be keep low for safety.
7. Packed columns are more suitable for foaming systems 8. The pressure drop per equilibrium stage can be lower for
packed columns.
Column internals –
process design
16
Process design or process tech support to operation needs to consider:
Type of contacting device
Number equilibrium stages Height of packing required
Pressure drop
Fouling
MT Rate, r
A, for absorption per unit volume of
packed column
kya is local MTC for gas phase on unit volume basis
yiis mole fraction (of component A) in gas at the gas-liquid interface , y is bulk vapour composition
kxa is local MTC for liquid phase on unit volume basis
xi is mole fraction (of component A) in liquid at the gas-liquid interface, x is bulk liquid composition
Kya is overall MTC for gas phase on unit volume basis
y* is composition of vapour that would be in
equilibrium with the bulk liquid of composition x
Kxa is overall MTC for liquid phase on unit volume basis
x* is composition of vapour that would be in
equilibrium with the bulk vapour of composition y
17
(
)
A y ir
=
k a y
−
y
(
)
A x ir
=
k a x
−
x
(
*)
A xr
=
K a x
−
x
(
*)
A yr
=
K a y
−
y
See McCabe et al. page 579
Which MTC and rate equation?
Can use any of the four basic rate equations to
design an absorption column, but the gas-film
coefficients are often used.
We’ll follow McCabe et al. and use K
ya here.
Calculation of packing height
(dilute gas)
19
a = interfacial area/unit volume of column
A = cross-sectional area column (m2)
ZT = total height of packed section
(
*)
yVdy
K a y
y
Adz
−
=
−
L, x2 V, y1 x1 y2 dz x x+dx y+dy yMass balance on component A across differential volume dz.
Assume:
• dilute gas change in molar flow V is neglected
Rate loss solute from gas = Rate gain solute by liquid
Let’s do a dimension analysis here. How do we get this equation from NA?
Calculation of packing height
20
Rearranging and integration of the mass balance equation:
2 * 1 t y
V
dy
Z
K aA y
y
=
−
∫
We now have an equation to calculate the total height of packing, ZT, based on concentration driving force (y-y*), gas flow rate and the gas phase MTC:
See McCabe et al. page 580-581
ZT = (height transfer unit) x (number units)
21
See McCabe et al. page 580-581
2 * 1 y t
d
V
A
Z
y
y
a
y
K
−
=
∫
change in gas conc.
average driving force
Number of transfer units
NOy
Subscript Oy shows based on overall gas phase driving force.
Height of transfer unit
HOy
Units of length.
The height of packing needed to achieve:
change in gas conc.
driving force
=
Operation line and equilibrium line Graphic integration: 1/(Y – Y*) as a function of Y
Gas film:
Liquid film:
Overall gas:
Overall liquid:
Four sets of HTUs and NTUs
23
See McCabe et al. page 583
i y
y
dy
N
y
−
=
∫
i xx
dx
N
x
=
−
∫
* Oxdx
N
x
x
=
−
∫
* Oydy
N
y
−
y
=
∫
/
y yV
A
H
a
k
=
/
x xL A
H
a
k
=
/
Oy yV
A
H
K
a
=
/
Ox xV
A
H
K
a
=
ENSC3019
6.2 Determination of Column Height
Values of height of transfer
unit
25
See McCabe et al. page 580-581
Values of HOy are system dependent.
Sometimes available for a particular system directly in the literature, or could be measured in pilot-plant studies.
But, often need to estimate height of transfer units from
empirical correlations for individual MTCs or individual heights of a transfer unit.
Evaluating the
integral for N
Oy?
26 2 * 1 t yd
V
Z
A
y
y
a
y
K
−
=
∫
Simplest case - Straight operating & equilibrium lines.
Can evaluate NOy by: change in gas conc.
log mean driving force
Oy N =
(
) (
(
) (
)
)
(
)
* * * 1 2 * 1 * 2 ln lm y y y y y y y y y y − − − − = − −(
) (
)
2 2 1 * * 1 y Oy y lmdy
y
y
N
y
y
y
y
−
=
=
−
−
∫
Log mean driving force
For details on the integration above, see Coulson & Richardson Vol2.
Example H2S scurbber problem & solution provided at end of these set of slides.
Challenges and discussions
27 What if gas is not dilute?
Where do I get values of Mass Transfer Coefficients?
Affects of temperature and pressure?
What if there’s a chemical reaction as well as absorption? E.g. amine absorption for acid gas removal?
Example 13.1 H
2S scrubber
28
Gas from a petroleum distillation column has its H2S concentration reduced from 0.03 kmol H2S /kmol inert hydrocarbon to 1 % of this value by scrubbing with triethanolamine-water solvent in a
countercurrent tower, operating at atmospheric pressure and 300 K. The equilibrium relation for the solution is described by Ye=2X.
Solvent enters the tower free of H2S and leaves containing 0.013
kmol H2S /kmol solvent. If the flow of inert gas is 0.015 kmol/s.m2 of
tower cross-section, calculate: (a)Height of absorber required
(b)Number of transfer units NOG (or Noy)required
The overall coefficient for absorption KYa is 0.04 kmol/s.m3 (unit
Example 13.1 solution (1)
29 Data:
1) Equilibrium expression Ye=2X
2) Top of column conditions Y2 = 0.03 x 0.01 = 0.0003
L
s, X
2Vs, Y
1X
1Y
2 absorber3) Bottom of column conditions Y1 = 0.03
X1 = 0.013 Y1e=0.026
Driving force = Y1-Y1e = 0.004
X2 = 0 Y2e = 0
Example 13.1 solution (2)
Logarithmic mean driving force:
(
)
0.004 0.0003 0.004 ln 0.0003 0.0037 0.00143 2.59 e lm Y −Y = − = =Mass balance on H2S in gas film:
(rate moles lost from gas) = (rate mass transfer)
(
1 2)
(
)
s G e lm
V Y
−
Y S
=
K aP Y
−
Y
SZ
Where S is the cross section area
Example 13.1 solution (3)
31
And we can rewrite in terms of lumped overall coefficient:
Then:
Solve for Z:
Z = 7.8 m
K
Ga P = K
Ya = 0.04 kmol/s m
3V
s(Y
1- Y
2) = K
Ya (Y - Y
e)
lmZ
Example 13.1 solution (4)
32
Now calculate height of transfer unit:
Number of transfer units:
N
OG= 21
20.7
OG OGZ
N
H
=
=
Which is another expression of NOy Which is another
expression of HOy
Example 13.1 alternatives solutions
33
If your love calculus you could solve analytically:
N
OG= 21.1
2 1 Y OG e YdY
N
Y
Y
=
−
∫
Calculate H
OGas before. Then calculate Z.
Example 13.1 alternatives solutions
34
Could do a graphical-numerical solution (eg.
trapezoidal rule or Simpson rule to find the N
OG)
Looking forward:
Plate columns vs. Packed columns
• Coulson and Richardson Vol 6. suggest the following advantages/disadvantages for Plate vs Packed:
• Plate columns can be designed to handle wider range of liquid and gas flow rates
• Packed columns not suitable for very low liquid rates • The efficiency and performance of a plate column can
be more accurately predicted
Looking forward:
Plate columns vs. Packed columns
• Plate columns can be designed with more assurance - some doubt that good liquid distribution can be
maintained in a packed column.
• It is easier to provided cooling or heating in a plate column – coils directly on plates.
• Easier to make provisions for withdrawal side streams in plate columns.
• Fouling by solids – can easily install manholes on plates – small columns however – may be easier to replace
packing when fouled.
Looking forward
:
Plate columns vs. Packed columns
• For corrosive liquids a packed column will be cheaper than a plate column (due to materials).
• The liquid hold-up is appreciably lower in a packed
column – important if amount toxic or flammable liquid needs to be keep low for safety
• Packed columns are more suitable for foaming systems • The pressure drop per equilibrium stage can be lower
for packed columns – impt. vacuum distillation • Packing cheaper for small columns, d < 0.6 m
Review:
Plate columns vs. Packed columns
• Coulson and Richardson Vol 6. suggest the following advantages/disadvantages for Plate vs Packed:
• Plate columns can be designed to handle wider range of liquid and gas flow rates
• Packed columns not suitable for very low liquid rates • The efficiency and performance of a plate column can
be more accurately predicted
Plate column easy to think of in # of stages,
what about packed?
40 Vn+1 y n+1 Vn yn Vn-1 yn-1 L n+1 x n+1 Ln xn Ln-1 xn-1 n + 1 n n - 1 ideal actual N N = η
?
“Ideal stage” stage-by-stage determinationH
eight
E
quivalent to a
T
heoretical
P
late
(
HETP
)
Column height is determined from # of theoretical plates and the height equivalent to a theoretical plate (HETP)
41 0 1 0 1 xB xD α = 4 7 stages
Example: 7 Theoretical Stages
If the HETP is 0.5 m then...
3.5 m
packed ideal
How to determine an HETP
• Typically determined through empirical data • General values for random packing
– 0.3 to 0.6 m
• Smaller packing can have lower values but also less capacity
• Structured packing can have much improved HETP
– 0.1 to 0.2 m
• Typically no fundamental prediction for HETP
Random and Structured Packing
43
Plastic Tripak (Jaeger Products Co.)
Metal Tripak (Jaeger Products Co.)
Section of expanded metal packing
Sections of expanded metal packings placed altenatively at right angles (Denholme Co.)
Structured packing elements for small colums with wall
wipers at the periphery
Random - larger HETP
Structured - smaller HETP (better separation with smaller column height)
Example: HETP for iso-octane/toluene with
Intalox packing
• HETP given in terms
of a flow capacity
factor
• #25, 40 50 refer to
packing sizes of 1,
1.5, 2 inches
44 superficial velocity Recommended design velocity: 20% less than when HETP rises rapidlyWetted area key to good separation
• The better the wetted area the lower the HETP
– Thus structured packing typically better than random
• Areas of high liquid flow tend to have low vapour flow and vice versa
• Liquid will also tend toward the outside
• Also means redistribution can be important
– Recommended design practice of redistribution every 3 to 4 m
Liquid distributors/redistributors
• Liquid distributor – distribute liquid evenly over column (feed and reflux from condenser)
• Redistributors - collect liquid that has migrated to the walls and redistribute it evenly over packing or even out any other maldistribution
46
Recommended examples from textbooks
read, understand, try to do the problems yourself
47 McCABE et al:
Examples: 21.1, 21.2, 21.3, 21.4, 21.5, 21.6, 21.7 BENÍTEZ, J. (2009):
Examples 6.1, 6.4, 6.5, 6.6, 6.7, 6.8
SEADER, J. D. & HENLEY, E. J. (2006). Example 7.1, 7.2, 7.3, 7.4, 7.6
Treybal, R. E. (1981); illustration 9.10
6.4 DETERMINATION OF
COLUMN DIAMETER
-- APPLICABLE TO BOTH DISTILLATION AND ABSORPTION COLUMNS
REFERENCE FOR
ASSIGNMENT 1
48 Not presented
Determination of Column Diameter
• Column diameter D is a function of the volumetric flow rate V and velocity u of the gas entering the column
• 𝐷𝐷 = 𝜋𝜋𝑢𝑢4𝑉𝑉
• For a given task, gas flow rate V is known, and then unknown parameter is velocity u.
• Gas velocity is often determined by the viable pressure drop in the column (which is related to operation cost). • Larger velocity higher pressure drop higher
operation cost
• Smaller velocity lower pressure drop larger column diameter and higher capital cost
Centre for Energy - “energy for today and tomorrow”
P
∆
Gy
log
dry Loading point Flooding pointDesign considerations: Pressure drop and flooding
G – mass flow per unit area (Gy-gas, Gx-liquid)
For packed column
Gx G’x
Centre for Energy - “energy for today and tomorrow” Liquid inlet
Liquid outlet Gas inlet Gas outlet
Some flooding description
•A visual build-up of liquid on the upper surface of the packed bed
• A rapid increase in liquid hold-up with increasing gas rate
• Formation of a continuous liquid phase above the packing support plate
• A considerable entrainment of liquid in the outlet vapour
• Filling of the voids in the packed bed with liquid
Design considerations: Pressure drop and flooding
www.see.ed.ac.uk
Centre for Energy - “energy for today and tomorrow”
(McCabe, Smith, Harriott)
Design considerations: Pressure drop and flooding
Gy Gx
L
Centre for Energy - “energy for today and tomorrow”
McCabe, Smith, Harriott
Pressure drop analysis: Eckert graph
Design considerations: Diameter of packed towers
Flooding line Gy : Mass flow of gas per unit area Gy = u ρv Pressure drop in inH2O/ft of packing (brackets: mm H2O/ m of packing)
Normally
* Moderate to high pressure distillation = 0.4 to 0.75 in water / ft packing = 32 to 63 mm water / m packing * Vacuum Distillation = 0.1 to 0.2 in water / ft packing = 8 to 16 mm water / m packing
* Absorbers and Strippers = 0.2 to 0.6 in water / ft packing = 16 to 48 mm water / m
Centre for Energy - “energy for today and tomorrow” 𝑢𝑢2𝜙𝜙𝜙𝜙 𝑔𝑔 𝜌𝜌𝑉𝑉 𝜌𝜌𝐿𝐿 𝜇𝜇𝐿𝐿0.2 𝑤𝑤𝐿𝐿 𝑤𝑤𝑉𝑉 𝜌𝜌𝑉𝑉 𝜌𝜌𝐿𝐿 0.5 In a flooding line, u becomes umax
u, dry column velocity (m/s); umax, flooding point velocity (m/s); g, acceleration of gravity (m/s2); φ, packing
factor (1/m); ψ, liquid density correction coefficient, i.e.
density of water versus density of the liquid ψ = ρH2O/ρL; μL, viscosity of liquid (mPa s), wL and wV, liquid and vapor mass flowrate (kg/s).
Centre for Energy - “energy for today and tomorrow”
Design considerations: Diameter of packed towers
Sinnott
Other different graphs
Given L, V (mass flow rates)
Select pressure drop
determine u
select packing
Double check pressure drop
ENSC3019/
CHPR8503
Topic 3
Solid-fluid separations
56
Dr Kevin Li
Recommended reading:
H. Pierson & B. Perlmutter, Settle Down (Part 1). The Chemical Engineer (TCE), 2010, June pp48-50. H. Pierson & B. Perlmutter, The solution is clear (Part 2). TCE, 2010, July/August pp53-55.
Chapters 28 & 29 of McCabe et al., Unit Operations of Chemical Engineering, 7th Edn. McGraw Hill 2005
We will look at:
Sedimentation & Settling processes
57
Important solid handling processes we won’t
study here:
• Filtration & screening processes
• Size reduction
• Solids mixing
Examples of solid-fluid separations
Oil and gas industry hydrocyclones
58
Separate sand and other solids from water or other liquids
Examples of solid-fluid separations
Coal-fired power station (filter-bags)
– Particulates from flue gases
Gravity classifiers
Separate particles of the same density but different
particle sizes.
60
Feed
Liquid + fine particles overflow
Coarse particles sink, picked up by
Examples of solid-fluid separations
Food and beverage industry (filter)
– Separate curd (solids)
– from whey (liquid)
Properties and handling of particulate solids
Size
Shape
Density
Size and shape of particles
For regular shaped particles we can easily define size and shape.
63
Cube
l 3 26
Volume
l
Area
l
=
=
Sphere
3 24
3
2
4
2
d
Volume
d
Area
π
π
=
=
Relative sizes of particulate matter
Shape of irregular particles
Sphericity 656
p s p pd
S V
φ
=
Eq. 28.1 McCabe et al. p 967
d
p= nominal diameter of one particle
V
p= volume of one particle
S
p= surface area of one particle
1 for a sphere
Sphericity of some materials
66Material
ΦMaterial
Φ Spheres, cubes, short cylinders (L=dp) 1.0 Ottawa sand 0.95 Raschig rings (L=dp) 0.33-0.58 Coal dust 0.73 Berl saddle (L=dp) 0.3 Crushed glass 0.65 Mica flakes 0.28Description of populations of
particles
• In practice, size distribution is a histogram
• Distribution curve by mass, number and surface can differ dramatically
• Which distribution we would use is dependent on the end use of the information
Differential VS cumulative
distribution
2 basic principles of
separation
To separate liquid from solids, or solids from
liquids there are only 2 mechanisms available:
(1) Use a screen or porous medium that retains
one component and allows others to pass
(2) Use differences in sedimentation rates as
particles (or drops) move through a gas or
liquid
70 Settling / sedimentation Screen / filter Gravity Centrifugal force Heavy media Flotation Magnetic force Screens Filters
Crossflow eg. membranes
Separation by
Gravity Pressure Vacuum Expression
Gravity sedimentation processes
Three broad functional operations
(1) Classifier
Separate solids into two fractions
(2) Clarification
Remove a relatively small quantity of suspended
particles to produce a clear effluent
(3) Thickening
To increase concentration of solids in a feed
stream
Selecting a separation method
1. Define the problem– Is liquid or solid the valuable product? – How clear does liquid need to be?
2. Establish process conditions
– Particle size, concentrations, flowrates – How long do particles take to settle?
3. Make a short list of appropriate equipment types
Clarifiers and thickeners
Convert dilute slurry of fine particles into a clarified
liquid and a concentrated suspension.
Often performed in large open tanks.
73
Cessnock Wastewater Treatment Works
Batch sedimentation process
74 (1)B
Time
(2)B
A
C
D
(5)A
D
(3)B
A
D
C
(4)A
D
C
Rate of separation
75 Clear liquid interface heightFlocculation
particles < few microns dp settle slowly
Agglomerate particles faster separation
76 Flocculation for waste water treatment
How flocculation works? Videos
https://www.youtube.com/watc h?v=5uuQ77vAV_U
Equipment - thickeners
77 http://www.filtration-and-separation.com/
Motion of a particle in air
The forces acting on a particle in a fluid
Eq(1) (2) (3) (4)
ρ: density of particulate or fluid, kg·m-3
Fd: drag force, kg·m·s-2
ma: sum of the forces acting on the particle a: downward acceleration of the particle, m·s-2
The drag force increases as the velocity of the particle increases, until it reaches the
terminal settling velocity, the sum of force
Stokes’ Law
The relationship between velocity and drag
force:
where, μ is the fluid viscosity, Pa·s or kg·m-1·s-1
Substitute eq (6) into eq (5):
Which is commonly referred to as Stocks’ Law.
George Gabriel Stokes
(6)
In class tutorial 1
Compute the terminal settling velocity in air of a spherical particle with
diameter of 1 μ and 10 μ, respectively. Density of the particle is 2000 kg·m-3, air density 1.2 kg·m-3, viscosity 1.8 x 10-5 kg·m-1·s-1.
V = 9.81 · (10-6) 2 · (2000-1.2)/ (18 · 1.8 x 10-5 ) = 6.05 x 10-5 m/s V = 9.81 · (10-5) 2 · (2000-1.2)/ (18 · 1.8 x 10-5 ) = 6.05 x 10-3 m/s
Estimate how long will it take for the particle to settle down to the
ground level, if it falls from a 3000 m altitude. Assume no convection, no rainfall.
Terminal settling velocity for spherical particles with specific gravity =
Gravity Settler
A gravity settler is simply
a long chamber through which the contaminated gas passes slowly,
allowing time for the particles to settle by gravity to the bottom.
Very effective for very
dirty gases with heavy particles (metallurgical).
The average velocity equals volumetric flow rate divided by cross sectional area:
Physical Model
Easy mathematical analysis and typical model for devices
using similar devices, i.e. cyclones and electrostatic precipitators. H Vavg Vt L captured captured escaped Chamber floor
Deriving the model
Assumptions for plug flow
a) The horizontal velocity of the gas in the chamber is
equal to V
avgeverywhere and constant
b) The vertical component of the velocity of the
particle is equal to the terminal settling velocity due to
gravity, V
tc) If a particle settles to the floor, it stays there and is
not-entrained
d) Particle size distribution is uniform, no interaction
with each other
Traverse time of particle in the flow direction is
t = L / Vavg
Vertical settling distance =
t·Vt = Vt·L / Vavg
So all the particles with vertically settling distance smaller than H will
settle on the floor.
The fraction of particles that will be captured, is
Fractional collection efficiency =
η = Vt·L / (Vavg·H) (8)
To compute the efficiency-particle diameter relationship, we replace
the terminal settling velocity in eq (8) with the gravity–settling relations described by Stock’s law, finding
for Mixed flow (practical)
Assumption
Gas flow is totally mixed in the z direction but not in
the x direction, as most real gas flows are turbulent.
Collection efficiency
(10) or, (11)
In class tutorial 2
Compute the efficiency-diameter relation for a gravity settler that has
H =2m, L = 10m, and Vavg = 1 m/s for both the plug and mixed flow models, assume Stocks Density of the particle is 2000 kg·m-3, air
density 1.2 kg·m-3, viscosity 1.8 x 10-5 kg·m-1·s-1.’ law.
A: We can get the result using only one computation and then using
ratios. For a 1 micron particle in plug flow:
Particle diameter, μ η plug flow η mixed flow 1 0.000303 0.000303 10 0.03 0.03 30 0.27 0.24 50 0.76 0.53 57.45 1.00 0.63 80 1.94 0.86 100 3.03 0.95 120 4.36 0.99 57.45 μ
Calculation results
Plug flow settling VS mixed flow
Dust gas in Clean gas out
Dust gas in Clean gas out
Plug flow gravity settler
Limitation of gravity settler
Only effective for particles with diameter >100 micron
(fine sand, mineral particle) but not for particles of air
pollution (PM
10)
To increase the collection efficiency substantially and practically, by substituting some other force for the gravity in driving the particles from the gas stream to the collecting surface
Centrifugal force
If a body moves in a circular
path with radius r and
velocity Vc along the path, then it has angular velocity
ω =Vc/ r
Centrifuge force =
acceleration, substitution of g
Example
A particle is travelling in a gas stream with velocity of
18 m/s and radius of 0.3 m. What is the ratio of
centrifugal force to the gravity force acting on it?
Centrifugal Separator (Cyclone)
Substituting the centrifugal acceleration of the
gravitational one into Stocks’ law, eq (7), and drop the
buoyancy term, we find:
This is the settling velocity under centrifuge
(13)Structure of cyclones
Similar to gravity settlers, in
the form of two concentric helices.
Only the outer helix
contributes to collection
Particles get into the inner
helix escape uncollected
Dimensions are typically
based on the diameter D0 of outer helix. Taken as ratios to D0. Gas inlet width, Wi = 0.25
Model details
During the outer spiral of the gas, the particles are driven
to the wall by centrifugal force, where they collect, attach to each other, and form larger agglomerates and slide down the wall by gravity and collect in the dust hopper in the
bottom.
The inlet stream has a height Wi in the radial direction,
equivalent to the height H of pure gravity settler
The length of the flow path is NπD0, where N is the
number of turns that gas traverse the outer helix (normally set as N = 5), analogues to the length of gravity settler L.
Collection efficiency of cyclones
Substitute H =Wi and L = NπD0 into gravity settler
equation (9) & eq (11), finding:
Further substituting the centrifugal Stokes’ law eq (13) into
above equations, finding:
plug flow (14) mixed flow (15)
In class tutorial 3
Compute the efficiency-diameter relation for a cyclone separator that
has Wi = 0.15 m, Vc = 18 m/s, and N =5, for both block and mixed flow assumptions, assuming Stocks’ law.
Particle diameter, μ η plug flow η mixed flow 1 10 30 50 57.45 80 100 120
For very small particles < 5 micron
An industrial multiclone dust collector
diytrade.com
B&W's Multiclone dust collector made of a number of parallel small cyclone
Cut diameter
Measure of the size of the particles caught and the size
passed for a particular particle collector.
Cut diameter is the diameter of a particel for which the
efficiency curve has the value of 0.5, i.e. 50%
Substitute η = 0.5 into Stocks’ law plug flow model,
finding:
Other dust collectors