Course Notes

Goal and Learning Tools for Class
The goal for this class is for students to develop a better understanding of air quality as it pertains to ambient air quality and emissions to the ambient environment. First principles will be used to solve realistic air quality problems. Also, current events will be discussed as they relate to the content of this class. class Learning tools for this class are lectures, discussions, supplementary materials, assignments, textbook, and exams.
CEE 446 2009 Mark J. Rood, University of Illinois at Urbana-Champaign. All Rights Reserved. 1 CEE 446
Definition of Aerosol
aerosol dispersoid consisting of liquid p g q and/or solid particles that are suspended in a gaseous dispersion medium. Such mixture exhibits stability by separation processes over some relevant time period.
2009 Mark J. Rood, University of Illinois at Urbana-Champaign. All Rights Reserved.
Composition of Dispersion Medium for Dry Ambient Air and Flue Gas An aerosol consists of particles ranging in diameter from 0.001 m to 100 m. The lower limit in particle diameter is due to the particles becoming so small that they behave as gas molecules. The upper limit is due to separation processes that readily remove th larger particles f the l ti l from th i their gaseous dispersion medium.
Flue Gas Generated by Combustion of Coal Concentration Concentration Component Component (by volume) (by volume) Dry Ambient Air N2 O2 Ar CO2 Ne He CH4 SO2 CO NO2 O3 78.08% 20.95% 0.93% 0.035% 18 ppmv 5.2 ppmv 1.2 ppmv 0.1 0 1 ppmv 0.1 ppmv 0.02 ppmv 0.02 ppmv N2 CO2 H2O(g) O2 SO2 NO CO 70% 12% 12% 6% 2000 ppmv 200 ppmv 100 ppmv
[H2O(g)] = 1-3% by volume for ambient air ppmv = parts per million by volume
CEE 446 2009 Mark J. Rood, University of Illinois at Urbana-Champaign. All Rights Reserved. 3 CEE 446 2009 Mark J. Rood, University of Illinois at Urbana-Champaign. All Rights Reserved. 4
Units Used To Describe Concentrations And Ideal Gas Law

Concentrations of gases are typically expressed on a volume per unit volume p p basis or mass per unit volume basis. On a volume per unit volume basis: Va = volume of a gaseous component, a VT = total volume of a gas Volume units f Va and VT are arbitrary V l it for d bit except they must be consistent with each other. Concentration of gases are also described by b mass per unit volume. U i th id l it l Using the ideal gas law and the definition of a mole:
PT Va = naRTT
5 CEE 446 2009 Mark J. Rood, University of Illinois at Urbana-Champaign. All Rights Reserved. 6
ppmv = parts per million by volume V = a 10 6 VT
CEE 446
PT = total pressure Va = volume of pure component a na = moles of component a p R = ideal gas law constant TT = temperatur e of the gas = Ta ma na = MWa ma = mass of component a MWa = molecular weight of component a
Values and Units for Ideal Gas Law Constant
R
0.0821 82.06 62.36 62 36 1.314 0.08314 998.9 0.7302 21.85 555.0 10.73 1545 8.314 18510 1.987 1.987 8314 8.314
Units
(atm)()/(g-mole)(K) (atm)(cm3)/(g-mole)(K) (mm Hg)( )/(g-mole)(K) (atm)(ft3)/(lb-mol)(K) (bar)( )/(g-mole)(K) (mm Hg)(ft3)/(lb-mole)(K) (atm)(ft3)/(lb-mole)(R) (in Hg)(ft3)/(lb-mole)(R) (mm Hg)(ft3)/(lb-mole)(R) (psia)(ft3)/(lb-mole)(R) (p )( ) ( (psfa)(ft3)/(lb-mole)(R) )( ) (Pa)(m3)/(g-mole)(K) (psia)(in3)/(lb-mole)(R) (BTU)/(lb-mole)(R) (cal)/(g-mole)(K) (joule)/(kg-mole)(K) (joule)/(g-mole)(K)
= 0.0821 atm l/g - mole K
PT Va =
ma RTT MWa PTMWa RTT
Concentration of a gaseous component can be described by the number of molecules of the gaseous component per unit volume of the gas sample sample.
Na na = NAV VT VT
ma = Va
ma Va PTMWa ppmv PTMWa = = VT VT RTT 106 RTT
Na = number of molecules of gaseous component a NAV = Avagadro' s number = 6.022 1023 molecules/ mole
CEE 446
Example: How many g of SO2 exist per m3 of gas if the SO2 concentration is 1 ppmv at 298 K and 1 atm?
[SO2 ] =
ma ppmv PTMWa = VT 106 RTT
g (1 atm ) 64.06 g - mole 1 ppmv SO 2 = 6 atm l 10 0 08 98 0.0821 g - mole K (298 K ) = 2.62 10 6 g SO 2 10 6 g 10 3 l l g m3
( g ) MWa = (atomic weight)i

i
[SO2 ] = 2,618
g SO2 m3
MWa = 32.06 + 2 (16.00 ) = 64.06 / mole

Let's look at other applications for the ideal gas law. For a gas containing component a:
Pa VT naRTT = PT VT n TRTT Pa na = PT n T
PT Va naRTT = PT VT n TRTT Va na = VT n T
Then...
Or...
Pa na Va = = PT nT VT
2009 Mark J. Rood, University of Illinois at Urbana-Champaign. All Rights Reserved. 13 CEE 446 2009 Mark J. Rood, University of Illinois at Urbana-Champaign. All Rights Reserved. 14
CEE 446
Mole fraction (ya) and mole ratio (Ya) are also commonly used to describe the concentration of gaseous component a. If nT = na + nb (two component g sample): ( p gas p )
n na Ya ya = a = = = nT nb + na nb + na 1+ Ya nb nb Ya = na na y = = = a nb n T n a n T n a 1 y a nT n T
15 CEE 446
For an ideal gas sample of constant mass and composition that exhibits a change in temperature, pressure and/or volume going from condition 1 t condition 2 f diti to diti 2.
na nb
P1V1 n RT = 1 1 P2 V2 n2RT2
n1 = n2
P T V2 = V1 1 2 P2 T1
2009 Mark J. Rood, University of Illinois at Urbana-Champaign. All Rights Reserved. 16
na nT
CEE 446
For a gas stream flowing at a constant mass feed rate and composition:
& & m1 = m2 & m = mass feed rate = Qg g Qg = volumetric flow rate of gas stream g = mass density of gas stream
m P(MW) = = g V RT Qg1g1 = Qg2 g2 and MW1 = MW2
Q g1
P MW2 P1MW1 = Q g2 2 RT2 RT1
PV = nRT =
CEE 446
m RT MW
17 CEE 446
P T Q g2 = Q g1 1 2 P2 T1
Also, mean molecular weight (MW ) of a gas mixture is described by :
Standard Conditions
Standard conditions are important when working with ambient and industrial gas streams because concentrations of contaminants can vary widely with T, P, concentration of H2O(g), and for combustion flue gases, their concentrations of CO2 or O2. Standard conditions are defined at PS = 1 atm and a specified temperature, TS, that can be 0C, 20C, 32F, 77F . . . . depending on the defined frame of reference.
n MW = (MW )i i n T i Where
ni = moles of component i nT = total moles of gas sample MWi = molecular weight of component i
CEE 446 2009 Mark J. Rood, University of Illinois at Urbana-Champaign. All Rights Reserved.
The concentration of component a in a moist gas stream can be corrected to dry conditions by: A = actual condition D = dry condition a = component a
ma V T,D
ma V T,D
CEE 446
Concentrations of contaminants in flue gases can be corrected to 12% CO2 by volume at dry condition by:
ma = V T,A
ma = V T,A
VT,A V T,D
ma VT,D,12% CO 2
ma 12% CO 2 by volume in dry gas stream = VT,D y% CO2 by volume in dry gas stream
1 1-y H2 O(g)
Example Problem: Application of the Ideal Gas Law Concentrations of contaminants in flue gases can be corrected to 7% O2 by volume by: A gas stream is emitted from a point source at 140C and 1.01 atm. The composition of the gas stream is 200 ppmv nitric oxide (NO), 700 ppmv sulfur dioxide (SO2), 100 ppmv carbon monoxide (CO), 10.9% by volume H2O(g), 5% by volume oxygen (O2), 10% by volume carbon dioxide (CO2), 74% by volume nitrogen (N2), and 3 g m-3 (actual conditions) of particulate matter (PM).
ma VT,D,7 % O 2
O by volume 14% 2 in dry gas stream ma = VT,D O 2 by volume 21% O 2 by volume y i d gas stream - y% i d gas stream t t in dry in dry
CEE 446
At actual conditions: a. Determine the mass concentration of NO, as NO2, at dry, standard temperature, and standard pressure conditions using units conditions, of mg/m3. Assume TS = 25C.
TT, A = 140C PT, A = 1.01 atm
[NO]A
g 30 (1.01 atm ) 200 g - mole = 6 atm l (140 + 273 K ) t 10 0.0821 g - mole K = 1.79 10 - 4 mg m3
26
[NO]A
ppmv a PT, A MWa = 10 6 RTT, A
g l
= 178.7
25 CEE 446
CEE 446
At dry standard conditions:
[NO]S = [NO]A
PT,S TT,A PT,A TS
1 1 yH O 2 ( g)
b. Determine the mass concentration of particulate matter (PM) at dry standard conditions using units of g m-3. At dry standard conditions:
mp VT,D,S 1 1 y H2 O 1 g 1 273 + 140 = 3 3 m 1.01 273 + 25 1 0.109 m = p V T ,A PT,S P T,A TT,A T T,S
= 4.6 g PM at dry standard conditions m3
28
mg 1 atm 273 + 140 K 1 = 178.7 3 m 1.01 atm 273 + 25 K 1 0.109 mg = 275.2 3 m
[NO2 ] = [NO]
CEE 446
MWNO2 MWNO
g 46 mg NO 2 g - mole = 422 = ( 275.2) g m3 30 g - mole

27 CEE 446
mp VT,D,S
Evaluation of Closure between Two Sets of Independent Results

One way to evaluate closure between results from two independent methods that describe the same parameter is the Average Absolute Error (AAE) as defined by: AAE =
1 N xi - xi (1) N i=1 xi
(1) (2)
Chemical Classification of Particulate and Gaseous Contaminants

Composition Sulfur Nitrogen Organic Carbon Particles H2SO4, (NH4)2SO4 NH4NO3 CH3(CH2)6CO2H (Octanoic Acid) C CaCO3 n/a Attached Rn Gases SO2, H2S NH3, HNO3, NO, NO2
CH4, HCHO CFC11(refrigerant), CFC12 (propellant), CFC12 (solvent)
100
Elemental Carbon Inorganic Photochemical Oxidants Radionuclides
n/a CO2, CO, HCl O3

Unattached Rn 222Rn, 218Po, 214Pb, 214Bi, 210Po, 210Pb, 206Pb
30
where N is the total number of matched samples that are used to compare the results from the two methods, and x(1) and x(2) represent the results from method (1) and method (2), respectively.
CEE 446 2009 Mark J. Rood, University of Illinois at Urbana-Champaign. All Rights Reserved. 29
CEE 446
Physical Classification of Particulate Material

Phase: Aerosol particles can exist as a solid and/or as a liquid Pure soot (elemental liquid. carbon) exists as a solid and does not change phase under typical atmospheric conditions. Pure (NH4)2SO4 can exist as a solid or can dissolve in an aqueous solution depending on relative humidity and the chemical and physical history of the air mass containing the aerosol particles.
Size: Atmospheric and industrial aerosol particles can vary in diameter over five orders of magnitude depending on their f origin and the transformations they experience. As previously mentioned, particle diameters can range in diameter from 0 001 m to 100 m 0.001 m.
32
Typical Concentrations of Trace Contaminants in Lower Troposphere

Contaminant Carbon monoxide (CO) Nitrogen dioxide (NO2) Photochemical oxidants (e.g., O3) Sulfur dioxide (SO2) Lead (Pb) Particulate material smaller than 10 m diameter (PM-10) Ammonia (NH3) Nitric acid (HNO3) Formaldehyde (HCHO)
CEE 446
Typical Atmospheric Particle Size Distributions

10 PM PM 2.5 8 Ultrafine (PM ( Nanoparticles (PM 0.01) 6
Droplet Mode Fresh High Temperature Emissions, Organic Carbon, Sulfuric Acid, Metal Vapors
0.1 0 1) 10
Ambient Concentration
Re elative Mass Concentration
0 1 - 50 ppmv 0.1 0.1 ppbv - 0.5 ppmv 0.01 - 0.5 ppmv 1 ppbv - 1 ppmv 10-4 - 10 g/m3 1 1 000 g/m3 1,000 / 1 - 30 ppbv 0.01 - 50 ppbv 0.1 - 50 ppbv
33 CEE 446
Condensation Mode
Condensed Organic Carbon or Sulfuric Acid Vapors, Clean Environment
Sulfate, Nitrate, Ammonium, Organic Carbon, Elemental Carbon, , Heavy Metals, Fine Geological
Geological Material, Pollen, Sea Salt
Nucleation
0
Aitken
Accumulation
Coarse
0.001
0.01
0.1 1 Particle Aerodynamic Diameter (m)
10
100
(John Watson, Desert Research Institute, 2002, JAWMA, 52(6))

Particle Size Distributions

N = total number of particles Let N(dp) = cumulative number distribution
N N ( d p i+ 1 )
Normalized Cumulative Number Distribution

1 N (d p i+ 1 )/N
N (d p )
N (d p) [-] N ( d p i) d p i+ 1 100
N [-] 0 0
N (d p i)/N d p i+ 1 100
d pi
0 0
d pi
d p [ m ]
d p [ m ]
CEE 446
35
CEE 446
36
Let n(dp) = number of particles per unit differential range of particle size ((d(dp)), then
n' (dp ) = dN( dp ) d(dp ) = N( dp i+1 ) N( dp i ) dp i+1 dp i
0.01 n(d p i+1 )/N n (d n(d p ) N [-] 0 0
n(d p i)/N
for a small change between Npi+1 and Npi. Normalizing n(dp) with respect to N results in:
n' ( dp ) N
CEE 446
d pi
d p i+1
d p [m ]
100
dN( dp ) Nd(dp )
37 CEE 446
If n' (dp ) =
dN(dp ) d(dp )
then dN(dp ) = n' (dp )d(dp ) and...

N( dp ) 0
dN(dp ) =
N(dp ) = (
dp 0
dp 0
n '(dp )d(dp )
n '(dp ) ( p ) ( )d(d
Let Y = some characteristic = f(dp), e.g., number length l th interfacial surface area cross-sectional surface area volume mass
where Y = Y(dp )n'(dp )d(dp )
0
as dp
N(d ) N p
There is also interest in evaluating various properties of the aerosol particle size distribution.
10
Characteristic Functions for Particle Size Distributions*

Y Number (Np) Length (Lp) Interfacial surface area (Ap) Cross-sectional surface area (Ap) Volume (Vp) V l Mass (mp) Y(dp) 1 dp d2 p 1/4 d2 p 1/6 d3 p 1/6 d3 p p
Application of Y =
dp = dp =0
Y(dp )n'(dp )d(dp )
Let's look at the total mass of particles in a particular sample.
mP,T = =
41 CEE 446
dp = dp =0 dp = dp =0
mp (dp )n(dp )d(dp ) 1 3 dppn(dp )d(dp ) 6

42
*The above table is valid for spherical particles.

Aerosol Particle Size Averages and Widths of Distributions

It is useful to describe a distribution of particles by some type of "average" value, based on number, area, or mass distribution of particles. d = (based on number) dp, mean = 1 p dpn '(dp )d(dp ) N dp = 0
=
where,
n(dp i ) = num ber of particles within a characteristic size range m(dp i ) = m ass of particles within a characteristic size range m P,T = total m ass of particles dp, m edian = diam eter of particles with 50% of th e particles sm aller and 50% of the particles larger than the specified value dp, m ode = diam eter of particles that occur m ost frequently
d
i =1 i =1
pi
n(dpi )
pi
n(d
pi
d
i =1
pi
n(dpi )
(based on number) (b d b )
dp, mean =
CEE 446
d
i =1
m(dpi ) m p,T
(based on mass)
11
But we need to know more than the "average" size. It would be useful to know how the particle size distribution varies from its "average" size.
0 .0 1 n (d p) N [- ] 0 0 100
variance =
(dp dp, mean ) = 0
n(dp )d(dp )
N 1
= i =1
d p [ m ]
(dpi
d p, mean )2 n( dp i ) N 1
All of these distributions have the same "average" particle diameter but have different widths to each distribution.
standard deviation = = 2
0 .0 1
The standard deviation () represents a characteristic width of a symmetric particle size distribution. 68.2% of the distribution exists between dp, mean - and dp, mean + , with 84.1% of the distribution less than dp, mean + , and 15.9% of the distribution less than dp, mean - .
n (d p ) N [-] 0 0
d p, m ean
d p [ m ]
100
The schematic presented above describes p a symmetric distribution of particles and its correspond ing dp, mean and .
CEE 446
47
CEE 446
48
12
Gaussian Size Distributions

It is useful to apply a continuous function that is normal (Gaussian) to describe particle size distributions.
1 x 2 1 exp 2 2
Gaussian or Normal Distribution

0 .0 1 n (d p ) N [-] 0 0
d p, m ean n (d p ) = f(x ) N
f (x) = n( dp ) N
2 1d d 1 p p, mean exp = 2 2
d p [ m ]
100
CEE 446
49
CEE 446
50
It is also useful to look at the transforme d distributions with the distribution based on ln(dp ) or log(dp ) instead of dp , respective ly. This transforma tion of dp to ln(dp ) or log(dp ) may provide an equation with a better fit to the data than a normal distribution.
There is also interest in the geometric mean diameter (dp, g ) and the geometric standard deviation (g ) when using a log-normal distribution. Such transformation is important because particle size distributions are typically log-normally distributed not normally distributed.
CEE 446
51
CEE 446
52
13
Asymmetric Distribution
0.01 n(dp) N
Symmetric Distribution
0.01 n(ln(dp)) N
n(dp i )ln(dp i ) dp,g = exp i =1 N
and
1/ 2 n( dp i )(ln(dp i ) ln( dp,g ))2 g = exp i =1 N 1
[-]
0 0 dp [m] 100
[-]
0 0.001 ln(dp) [m] 100
CEE 446
The values for dp,g and g, based on mass/area/or number of the particles, can also be determined graphically by using a log-probability plot. Particle diameter is plotted on the log-scale ordinate and cumulative percent mass/area/or number of particles less than the specified diameter is plotted along the abscissa. If the data points lie on a straight line, then the distribution is lognormal. The geometric mass/area/or number mean diameter is the diameter with 50% by mass/area/or number of the particles less than the specified diameter.
Application of Particle Size Distribution Data to Log-probability Paper

dp,g = dp with 50% of
particles < stated value (dp, 50 ) g = = dp,84.1 dp,50 dp,50 dp,15.9
56
14
Terminal Settling Velocity for Spherical Particles

Force Balance
Fd Fb Z Fd = drag force Fb = buoyancy force Fg = gravitational force
Use Newton's 2nd Law of motion v r d(Vp ) r Fz = mp ap = mp d(t) r where ap = acceleration of particle
r v Assume steady state conditions (Vp Vtp )
constant mass of particle, and positive in downward direction.

v v mp d(Vtp ) =0 Fz = d(t) v v v v Fz = Fb Fd + Fg = 0
particle
Fg
CEE 446
57
CEE 446
58
Where,
v g v Fb = mp g p v2 v g A Vtp CD p Fd = 2 v v Fg = mp g
mp = mass of particle g = density of gas p = density of particle v g = gravitatio nal force constant
A = front cross - sectional area of particle p

15
v Vtp = terminal settling velocity of particle CD = drag coefficient
v Solving for Vtp ,

v g mpg1 v p Vtp = 1 p 2 g A CD
1/ 2
v2 g v g A Vtp CD v p p + mp g = 0 mp g 2 p
v Difficult to solve for Vtp because : v CD = f ( Vtp , dp , g , g )

CD = f (Rep )
CEE 446
61
CEE 446
62
in general,
Drag Coefficient Versus Reynolds Number for Spheres, Disks, and Cylinders
( velocity)(length)(density) ( vis cos ity)
Re =
for our case,

Rep = Vtp dp g g
D Drag Coefficient, CD [-]
Spheres Disks Cylinders
where, g = dynamic viscosity of the gas [g/cm-sec] g = g/g = kinematic viscosity

Reynolds Number, ReP [-]

SOURCE: Adapted from Lapple and Shepherd (1940). Ind. Eng. Chem., 32, 605.
16
Look at three regions to determine CD for spheres.
Cunningham Correction Factor

We need to correct the drag force term with the Cunningham Correction Factor (Kc) when the gas no longer acts as a continuous medium medium.
K c = 1 + Kn + exp Kn
Regions Describing Fluid Flow Around a Particle and Corresponding Reynolds Number and Drag C ff Coefficient
Region Stokes (viscous flow) Rep Rep 0.1 0.1 0 1 Rep 2 Transition 2 Rep 500 Newton
CEE 446
CD 24/Rep
24 3 9 2 Re 1 + 16 Rep + 160 (Rep ) ln(2 Rep ) p 24 0.687 Re 1 + 0.15(Rep ) p
where,
500 Rep 2 x 105
0.44
65 CEE 446
Kn= Knudsen Number = f(T,P, gas composition) f(T P = 1.257 = 0.40 = 1.10
Criterion to determine if Kc needs to be determined is based on the Knudsen Number, Kn. 2 g if Kn = > 0.01 dp Then Kc should be calculated because it is > 1.0 where, h g = mean free path of the gas dp = particle diameter
g =
g 0.499 g ug,molecule
= 0.066 m at 20 o C and 1 atm for air
= f(T, P, gas composition) where,

g = dynamic gas vis cos ity g = gas density ug,molecule = mean speed of the gas molecules
68
17
8RT ug,molecule = MW g
1/ 2
The drag force term can now be corrected with Kc as presented below.
where,
R = ideal gas law constant T = temperatur e MW g = mean molecular weight of gas
r r g Ap Vtp 2CD A Fd = 2K c
v Solving for Vtp
r g mpg 1 K c r p Vtp = 1 p 2 gA CD
1/ 2
70
v Let' s simplify the equation for Vtp , assuming p >> g , spherical particles, and Stokes' region.
v g mp g 1 K c v p Vtp = 1 g A CD p 2
1/ 2
mp =
1 3 dp p 6 1 2 A = dp p 4 24 g 24 CD = = Rep Vtp dp g
where,
1
g p
1 3 v dp p g (1) K c v2 6 Vtp = 1 24 g 1 2 g dp v 2 4 Vtp dp g

CEE 446
18
Example Problem: Calculation of Particle
Terminal Settling Velocity Calculate the terminal settling velocity of a 0.2 m diameter particle that results from gravity. The density of the particle is 2 g/cm3. Air temperature and pressure are 400 K and 1.01 atm, respectively. , p y
v 2 v dp pK c g Vtp = 18 g
CEE 446
73
CEE 446
74
where,
Assuming Stokes' Region, spherical particles, and p >> g
p = 2 g cm3
v 2 v dp pK c g Vtp = 18 g
dp = 0.2 x 10 4 cm g = f(T, P, compositio n) assume g = f(T) for air at 1 atm g = 0.0825 kg g g = 2.292 x 10 4 (cm sec) m hr v cm v g = 980 k sec 2
CEE 446
75
CEE 446
19
Calculate Kn
2 2 g dp g 0.499 g ug,molecule
g =
P MW g RT atm l 0.0821 (400K) g moleK

.
Kn = g =
1.01atm (1 01atm ) 29
g g mole
= 0 89 0.89
g l
= 0.89 103
g cm3
78
ug,molecule
8RT = MW g
1/ 2
kg m 2 ( 400K ) (8) 8.314 x 10 3 sec 2 kg mole K = kg ( ) 29 kg mole m = 540.4 sec = 540.4 x 10 2

CEE 446
1/ 2
g 2.292 x 10 4 cm sec g = g cm (0.499 ) 0.89 x 10 3 540 .4 x 10 2 3 sec cm = 9.55 x 10 6 cm 2 g dp (2)(9.55 x 10 6 cm) (0.2 x 10 4 cm)
Kn =
cm sec
79 CEE 446
= 0.95 > 0.01
80
20
Calculate Kc
K c = 1 + Kn [ + exp - (
)] Kn 1.1 )] 0.95
v Now calculate Vtp

cm -4 2 g (2.31) 980 (0.2 x 10 cm) 2 2 cm3 sec V = tp -4 g (18) 2.292 x 10 cm sec - 4 cm = 4.39 x 10 (positive downward) sec
K c = 1 + 0.95 [1.257 + 0.4 exp - ( K c = 2.31
CEE 446
81
CEE 446
82
Check Rep
Rep = Vtp dpg g
Scattering and Absorption of Light by Particles and Gases

Reduction in our ability to see through the atmosphere is the most obvious manifestation of air pollution. Such reduction in visibility is due to the interaction of visible light with particles and gases in the atmosphere. Visibility in the ambient environment can be y described by visual range. Visual range the distance at which a large black object just disappears from view.
g -4 cm -4 -3 4.39 x 10 sec (0.2 x 10 cm) 0.89 x 10 cm3 = g -4 2.29 x 10 cm sec = 3.41 x 10-8 << 0.1
Therefore, Stokes' Law assumption is valid.

21
Visibility Degradation
Glacier National Park, Montana
Another way to describe visibility is with Meteorological Range. L = where, L = meteorological range [L] 3 .9 ext
7.6
12.0
21.7 Glacier National Park vista*

65.3
Unit:g/m3
ext = light extinction coefficient [L-1]
Source: *William Malm (1999), Introduction to Visibility, National Park Service, pp. 79 **http://www.search.com/reference/Air_pollution
Visibility degradation is caused by the scattering and absorption of light with particles and gases in the atmosphere.
light scattering: radiant energy is re-radiated at the
light absorption: radiant energy is absorbed
same wavelength (light) but with different intensities and different directions. Light scattering by particles depends on wavelength of light (light), particle diameter (dp), cross-sectional area of the particles (dp2 /4), and refractive index of the particles (Rp).
and then transformed into other forms of energy such as heat, chemical reaction, and/or emission of light at a different light. Transmission of light through the atmosphere is described by the following figure.
88
22
Change In Light Radiance During Transmission In Atmosphere

sun Particles & gases dispersed in air
Observer I + d(I) d(I) d(x) L

Object I
The observer is looking at some object at a distance L. The light's radiance (or normalized intensity), that is detected by the observer is due to radiance from the object and additional radiance from sunlight that is scattered into the observer's sight path. The sunlight that is scattered into the observer's sight path is called air light, A. A is defined as radiance per unit g , p distance of transmission. The change in light radiance in the atmosphere can be described by the following equation.
d(I) = ( ext I)d( x ) + Ad( x )
where,
ext = light extinction coefficient [L-1]
I = radiance of light [W/(L2 steradian light)] A = air light [W/(L2 steradian light L)] d(x) = displacement [L]
The value of A is difficult to evaluate and is ignored with this treatment of visibility degradation. degradation Letting A = 0 provides an 0, equation called the Beer-Lambert Law as described schematically in the following figure.
CEE 446
91
CEE 446
92
23
Schematic Describing Beer-Lambert Law
For a small change in distance, defined by d(x), and A = 0:
I (x = x) = I ( ) d(I) Observer x=x d(x) L
I (x = x0) = I0 (
d(l) = -(ext l) d( ) + A d( ) ( d(x) d(x)
Object x = x0
d(l) = lo l
l
x xo
ext d(x) = - ext L
l ln = - ext lo
( x - xo )
l = lo exp (- ext L)
The product of ext and L (extL) is defined as optical depth.

ext i related t th amount of light scattered is l t d to the t f li ht tt d and absorbed by particles and gases in the atmosphere. ext is described by the following equation. ext = sp + sg + ap + ag
where,
sp = light scattering coefficient for particles sg = light scattering coefficient for gases ap = light absorption coefficient for particles ag = light absorption coefficient for gases
CEE 446
95
CEE 446
96
24
Attenuation of light in urban atmospheres is dominated by scattering and absorption of light by particulate matter. The extinction coefficient for particulate matter (p) can be described with Mie-Lorenz theory.
p =
where, p = light extinction coefficient for particles that can be applied to light absorption (ap) or light scattering (sp) [L-1] K = light absorption (Kap) or light scattering (Ksp) efficiency factor for particulate material [-] d(Nv(dp)) = differential number concentration of particles with diameter, dp
i =1
(K i )
particle cross - sectional area unit volume of gas i
2 dp (d(N (d ))) = 0 (K) v p 4
CEE 446
It is now useful to sum all of the particles' crosssectional areas as shown in the following figure.
Schematic Describing Beer-Lambert Law when Considering Particulate Matter and Gases
I (x = x) = I d(I) Observer x=x d(x) L
where,
d(Nv (dp )) d(dp ) = nv (dp )
d(Nv (dp )) = nv ( dp )d( dp )
I (x = x0) = I0
For scattering of light by particles:

Object
sp = 0
x = x0
2 dp n (d ) d(d ) K sp p 4 v p
100
25
where, Ksp = light scattering efficiency factor for particulate material nv = n per unit volume of gas [L-4] (per unit particle size range)
(Ksp)(dp2/4) is a good example of a weighted distribution, Y(dp). Ksp depends on the optical size parameter () and refractive index of the particles (Rp).
= dp light g
Rp = depends on particle composition [-]

= optical particle size [-]
Values for K can be determined graphically or analytically. An example of a graph describing Ksp is presented below.
K sp
This same equation is described in a different form as equation (1-7) in Air Pollution: Its Origin and Control as presented below.
sp = Ni K i ri2
i =1 n
K sp p
x=
CEE 446
dp ligh t
(van de Hulst, 1957)

26
where,
sp= light scattering extinction coefficient [L-1]
The light scattering coefficient for a gas and its dependence on wavelength of light () is described by
sg, = C g R 2 - 1 p 4
Ni = number concentration of particles with a given radius. (NOT the cumulative number concentration of particles up to size ri) [L-3] Ki = light scattering efficiency factor [-] ri = particle radius [L]
)2 (6 + 3)
(6 7 )
where, Cg = 3.077 x 10-24 m3 at STP (0 C, 1 atm) Rp = index of refraction for air = 1 000294 [-] 1.000294 [ ] at STP = anisotropy constant for air = 0.035 [-] = wavelength of scattered light [m]
Dependence of sg and the other extinction coefficients on temperature and pressure is described by:
T P sg (T2 , P2 ) = sg (T1, P1 ) 1 2 T P 2 1
Absorption of visible light by particles in the atmosphere is typically dominated by soot particles. Therefore:
ap-soot = (ap-soot )(Csoot ) ap-soot = light absorption coefficient for soot [m-1 ]
ap-soot = 10 m2/g = light absorption efficiency factor for soot
when using absolute units (e.g. Kelvin, total pressure) )
g Csoot = mass concentration of soot 3 m

27
Absorption of visible light by gases in the atmosphere is dominated by NO2, although the overall contribution to visibility degradation by NO2 in the atmosphere is typically very small.
The light absorption coefficient for NO2 (ag-NO ) is described by:

2
ag-NO = (ag-NO ) (CNO )

2 2 2
where,
ag-NO
2
= light absorption efficiency for NO2 = 2.92 x 10-4 m-1 ppmv-1 2 92

2
CNO = concentration of NO2 [ppmv]
CEE 446
109
CEE 446
110
The same physics and chemistry are applicable to exhaust gases and particles in ducts as they apply to ambient aerosol. Continuous emission monitoring systems (CEMS) are used i d t t it i t d in ducts to measure the opacity of the exhaust stream.
I Opacity (%) = 1 - (100%) I0 = (1 T )(100%)
Numerous field studies have occurred to determine extinction budgets for visible light at different locations around the world. Results from studies in Denver, Colorado and Houston, Texas are presented below.
where T = fractional transmittance of light across the duct.

28
Percent of Attenuated Light Associated with Specific Aerosol Particulate Matter & Gases
Denver, winter 1978
NO2 6% soot 31%
c
Methods to Determine Health Effects Caused by Air Contaminants

Adverse effects of air contaminants are studied by three primary methods: Epidemiological Studies of Human Populations
Study effects to humans and other animals that have occurred (evidence-based medicine) Typically study effects of contaminants at low concentrations during extended periods of time Complicating factors due to co-existence of multiple contaminants
Houston, summer 1980

other 7%
NO2 5%
other 23%
(NH4)2SO4 & H2O 20%
SO4, assoc. ions & H2O 48%
carbon 11% soot 13%
carbon 19%
NH4NO3 17%
Sloane, C.S. (1989) Visibility in Urban Smog, General Motors Research Report, GMR7003, 16 pp.
Laboratory Studies of Humans that are Exposed to Known Air Contaminants
Laboratory Studies of Animals and Plants That are Exposed to Known Air Contaminants
Study humans responses to exposure of individual components (exercise concentrations (exercise, of specific contaminants, degree of sensitivity) Laboratory measurements are more specific and sensitive than field measurements Short duration of intensive experiments Study humans before, during, and after exposure
Study animals and plants before, during, and after exposure Greater flexibility with respect to contaminant type, concentration and duration of exposure Difficult to extrapolate results from animals and plants t humans l t to h Difficult to extrapolate results from short-term high-concentration exposures to long-term lowconcentration effects
CEE 446
29
Air Quality Laws and Regulations

Typically air quality laws are mandated by the federal government in the form of the Clean Air Act (1963), and its amendments in 1970, 1977, and 1990. These laws are then interpreted to develop regulations such as th N ti h the National A bi t Ai Q lit l Ambient Air Quality Standards (NAAQS), New Source Performance Standards (NSPS),
and National Emission Standards of Hazardous Air Pollutants (NESHAP). The regulations are based on considerations such as type of contaminant, concentration of contaminant, mass emission rate of contaminant, averaging time, cost and/or availability of air quality control technology.
118
Clean Air Act Amendments of 1970

Goal: Clean air by July 1975 Established United States Environmental Protection Agency (USEPA) Provided funds to support research and programs at the state and regional level Establish National Ambient Air Quality Standards (NAAQS)
Development of State Implementation Plans (SIPs) Require Reasonable Available Control Technology (RACT) for select existing sources Establish New Source Performance Standards ( (NSPS) for select new, modified, or ) , , reconstructed sources
120
30
Enforcement of National Emission Standards ( ) for Hazardous Air Pollutants (NESHAPs) from 1973 to 1989, the number of hazardous air pollutants (HAPs) increased from 3 to 8. NESHAPS limit emissions of materials that may cause an increase in fatalities for in serious, irreversible, or incapacitating illness.
Requirement of Best Available Control Technology (BACT) that is determined on a case-by-case basis considering economic, environmental, and energy constraints ( , gy (RACT < NSPS < BACT< MACT (Maximum Achievable Control Technology) Establish air quality monitoring requirements Establish fines and criminal penalties Establish standards for aircraft Establish more strict automobile emission standards
CEE 446
121
CEE 446
122

States must submit revised SIPs Classification of areas with respect to NAAQS Establish policy pertaining to non nonattainment areas
Expansion of the concept of Prevention of Significant Deterioration (PSD) New major facilities must apply BACT standards Establish Good Engineering Practice g g (GEP) for stack height
CEE 446
123
CEE 446
124
31

Urban Air Quality: Ambient air quality regulations
pertaining to ozone, carbon monoxide, and particulate matter with diameters < 10 m (PM10) ti l t tt ith di t were strengthened. There were 96 cities that failed existing ozone standards and were ranked from marginal to extreme. New control strategies included upgraded inspection and maintenance programs for automobiles, vapor recovery systems, transportation controls to reduce the amount of vehicle miles driven in an urban area,
and more stringent controls at stationary sources. Carbon monoxide standards were not met in 41 cities and those cities were ranked from moderate to serious. Possible control strategies included the initiation or upgrading inspection and maintenance programs for vehicles and transportation controls. The 72 cities that did not meet the PM10 standard needed to implement RACT and other types of control strategies.
Permits: USEPA's ability to enforce
standards was enhanced by requiring that a facility obtain a five year operating permit for the entire facility. Also facilities were required to pay permit fees that reimburse state agencies for costs incurred to operate the th permit program. it
128
32
May 1991 July 1992 November 1993 December 1993 November 1994
November 1995
November 1996 November 1997
Time Schedule to Develop Title V Permitting Requirements EPA proposes operating permit rule EPA issues final regulations setting minimum requirements for state permit programs State agencies must submit a plan to EPA that describes the states program and regulations to implement Title V requirements EPA advises states as to whether their plans are complete as submitted EPA must issue final approval or disapproval of state agency programs and regulations Sources must have submitted a Title V permit application to the respective state agency within 12 months following EPA approval of the state program. State must have acted on one third of permit applications received. State agencies may require submittals before this deadline Twenty four Twenty-four months following EPA approval of the state program, states must act on second third of permit applications received Thirty-six months following EPA approval of the state program, states must act on final third of permit applications received
Motor Vehicles: Allowable emissions of
hydrocarbons, carbon monoxide, and nitrogen oxides were reduced for 1994 vehicles. Also, emission standards needed to be met for longer time periods. Emphasis was also placed on the use of reformulated gasoline to reduce the aromatic content of gasoline used in heavily polluted urban environments Oxygenated fuels environments. (e.g., alcohol blended fuels) were also to be sold during the winter in cities with elevated carbon monoxide concentrations.
CEE 446
Hazardous Air Pollutants (HAPs): Emissions
Acid Rain: A market-based system was
of 189 hazardous air pollutants (e.g., carcinogens, mutagens carcinogens mutagens, and reproductive toxins) were to be reduced within 10 years. The list of HAPs is revised with time. The list of toxic materials were to be published by USEPA within one year of the p y passage of the g 1990 Clean Air Act Amendments. MACT requirements are to be developed for specific source categories.
implemented to reduce the amount of sulfur dioxide emitted into the atmosphere from utility power plants. The amount of nitrogen oxides emitted from utility power plants was reduced by implementing p performance-based standards.
132
33
Ozone Depletion: Restrictions pertaining to
chloro-fluoro-carbons (CFCs) will be more strict than previously defined by the Montreal Protocol. Production of carbon tetrachloride and methyl y chloride will be phased out by 2000, methyl chloroform will be phased out by 2002, and CFC production will be phased out by 2030. Use of CFCs in "nonessential" applications will be prohibited after November 1992. Warning labels are also required on all containers and products that contain CFCs and other ozone-depleting materials.
Enforcement: USEPA can assess penalties
up to $200,000 and require that violations be corrected without filing a court case. "Field citations" can also be issued up to $5,000/day by USEPA inspectors. Recipients of the penalties and citations are entitled to an administrative hearing if they disagree with USEPA's findings Violations findings. of environmental emergency orders can be as high as $25,000/day.
Titles of the Clean Air Act Amendments of 1990
Also, premeditative release of hazardous air pollutants that place a person in imminent danger can result in 15 years in jail and a fine up to $1x106. Negligent releases are punishable by up to one year in jail and a fine. $10,000 can be awarded to citizens who provide information leading to criminal p g convictions or civil penalties associated with the Clean Air Act of 1990.
Provision for Attainment and Maintenance of Title I National Ambient Air Quality Standards Provisions Relating to Title II Mobile Sources Title III Hazardous Air Pollutants Title IV Acid Deposition Control Title V Permits Title VI Stratospheric Ozone Protection
Title VII
Provisions Relating to Enforcement Miscellaneous Provisions Clean Air Research Disadvantaged Business Concerns Job Displacement Provisions
Title VIII Title IX Title X Title XI
CEE 446
135
CEE 446
136
34
National Ambient Air Quality Standards (NAAQS)a

Carbon Monoxide (CO) Primary Standards* Standards
Effects Associated With Exposure To Carbon Monoxide

Concentration Exposure Effects 80-97 80 97 mg/m3 / 1h hr Sensitive subjects h S iti bj t have reduced exercise capacity 17-21 mg/m3 8 hr Dizziness and 3 115 mg/m 8 hr headaches Impairment of maximum 97 241 97-241 mg/m3 1 hr sustained exercise in healthy human subjects
mg/m { 10 mg/m 40 / None
3 3
for 8 hour for h f 1 hour
Secondary Standard
___________________________
*Aggravation of cardiovascular disease
Limits oxyhemoglobin concentration in blood. No secondary standard for CO

___________________________
given for 25C and 760 mm Hg http://www.epa.gov/air/criteria.html

aValues
Lead (Pb)
Primary Standard* and Secondary Standard

_________________________
Nitrogen Dioxide (NO2 )

___________________________
0.15 g/m3 for 3 months 1.5 15 for quarterly average g/m3
100 g/m3 for 1 year
*Increased concentration of Pb in blood Increased exposure to Pb causes impairment of porphyrin metabolism (synthesis of hemoglobin), learning disabilities, and seizures.
*Risk of acute and chronic respiratory disease NO2 also contributes to the formation of photochemical oxidants.
CEE 446
139
CEE 446
140
35
Effects Associated With Exposure To Nitrogen Dioxide

Concentration C t ti 150-188 g/m3 207 g/m3 Exposure E Over several years 8 hr Effects Eff t Increased incidence of acute respiratory disease in children and their parents Reduction in ventilatory function for asthmatics
Ozone
Primary Standard* and y Secondary Standard
___________________________
0.12 ppmv for 1 hour pp 0.075 ppmv for 8 hour
*Eye irritation and respiratory function impairment
CEE 446
141
CEE 446
142
Effects Associated With Exposure To Photochemical Oxidants

Concentration 98 g/m3 255 g/m3 Exposure Effects 4 hr Vegetation damage Daily Subjects with respiratory maximum disease are impaired
Particulate Matter with Diameter < 10 m (PM10 )

___________________________
150 g/m3 for 24 hour
*Reduction in ventilatory function
CEE 446
143
CEE 446
144
36
Particulate Matter with Diameter < 2.5 m (PM2.5 )

Primary Standard and Secondary Standard
___________________________
Effects Associated With Exposure To PM10

Concentration 60-180 / 60 180 g/m3 TSP with SO2 ith and relative humidity (RH) effects 150 g/m3 TSP (RH < 70%) Effects Corrosion of steel and zinc plates
15.0 g/m for 1 year 35 g/m3 for 24 hour
1997-2002: 1997 2002: review of PM2.5 health effects 2012-2017: attainment of PM2.5 standards should be obtained
Visibility reduced to 5 miles 3 100-130 g/m TSP with SO2 > Increased incidence of respiratory disease 0.05 ppmv 3 750 g/m TSP with [SO2] > Excessive deaths 0.27 ppmv
CEE 446
146
Sulfur Dioxide (SO2 )
Primary Standard and Standard* Secondary Standard

___________________________
year {80 g/m for 124 h 365 g/m f / for hour

3 3
Effects Associated With Exposure To SO2

Concentration 105-236 g/m3 288-497 g/m3 (24 hr mean) 785 g/m3 Effects Annual mean exposure t 185 g/m3 A l to / smoke is associated with increased occurence of respiratory symptoms and lung disease Associated with increased rate of hospitalization of sensitive population groups ( ld l ) l ti (elderly) Tree damage
{1,300 g/m
for 3 hour
*Aggravation of chronic lung disease and increased risk of acute and chronic respiratory illness
CEE 446
147
CEE 446
148
37
Summary Of New Source Performance Standards (NSPS)a for New, Modified, or Reconstructed Municipal Solid Waste Combustors with Combustion Capacities Over 225 Mg/dayb (large MWCs)
see Table 2-4b, p.79 in Air Pollution: Its Origin and Control
___________________________
Good Combustion Practice
Maximum load level demonstrated during g dioxin/furan performance testing Maximum particulate matter control device inlet temperature demonstrated during dioxin/furan testing
aFederal bBased
Register, 40 CFR, part 60 on Best Demonstrated Technology

CEE 446
CO concentration and averaging times depends on type of combustion chamber [50-150 ppmva (4-24 hr)], e.g., modular starved and excess air [50 ppmva (4 hr)] American Society of Mechanical Engineers or State certification for supervisors, operator training, and training manuals
_____________________________
Organic Emission Limit
Dioxins/Furans [13 ng/dscma] Based on good combustion practice: spray dryer, fabric filter, carbon injection
Metal Emission Limits Particulate matter [24 mg/dscma]
Opacity [10% (6 min)] Cd / Pb / Hg [0.02 / 0.20 / 0.080 mg/dscma] or = [0 02 0 20 0 080 85%

____________________________
7% by volume O2, dry basis [ ] = Standard

aAt
At 7% by volume O2, dry basis, and at STP = 25C, 1 atm [ ] = Standard

a
38
Acid Gas Emission Limit SO2 [80% removal efficiency or 30 ppmva (24 hr)] HCl [95% removal efficiency or 25 ppmva] NOx Emission Limit NOx [150 ppmva (24 hr)] (180 ppmva for
Fugitive Emissions
< 5% opacity at ash handling system

Monitoring Requirements
SO2 with continuous emission monitoring system (CEMS) [24 hr geometric mean] NOx with CEMS [24 hr arithmetic mean] Opacity with CEMS [6 min average]
____________________________
y ) first year) Selective non-catalytic reduction

____________________________
At 7% by volume O2 dry basis [ ] = Standard

2009 Mark J. Rood, University of Illinois at Urbana-Champaign. All Rights Reserved. 153 CEE 446
[ ] = Standard
CEE 446
General Methods to Control the Generation and Emissions of Air Contaminants Into the Atmosphere
CO load and temperature with CEMS [4 or 24 hr average] Particulate matter, dioxins/furans, Cd, Pb, Hg, and HCl [Annual stack test and fugitive emissions]
____________________________
Modify process Modify feed stream Demand side management Modify social norms Ancillary air pollution control devices Shutdown source
[ ] = Standard
CEE 446
155
CEE 446
156
39
Capture, Transport, and Removal of Contaminants from Gas Streams

Once contaminants are generated by a source, y p y yp they need to be captured by some type of hood and transported in a duct system to an air quality control device. The hood and duct system should capture and transport the contaminants without excessive dilution. Increasing the volumetric flow rate of the flue gas stream results in increased operating and capital costs for the air quality control device.
The "control velocity" is the desired velocity of the gas at the contaminant's null point to cause the contaminant to be captured with a reasonable margin of safety. These values can be obtained in industrial hygiene handbooks. The necessary gas flow rate in the hood to achieve the appropriate "control velocity" at the contaminant's null point can be determined with equations developed by Dalle Valle (1953, 1963) and Hemeon (1963).
Fluid Streamlines and Lines of Constant Velocity for a Gas Stream Captured by a Hood
Ah hood e n tr a n c e u g,x n u ll p o in t x ( h o o d c e n te rlin e ) flu i d s tr e a m li n e
An empirical expression relating the velocity of the gas outside of a circular hood entrance to the velocity of the gas at the hood entrance is presented below.
ug, x ug,h - ug, x g = 0.1 A h x2
Q g ,h u g ,h
li n e o f c o n s ta n t v e lo c it y
CEE 446
159
CEE 446
160
40
where, ug,x = gas velocity at displacement x from hood inlet [ft/min]

ug,h = gas velocity at hood inlet [ft/min]
Ah = cross-sectional area of hood inlet [ft2] x = displacement between null point and inlet to hood [ft]
For hood openings that are square, round, or rectangular with no restriction to the gas flow, the following expression can be applied to relate volumetric fl l t l t i flow rate of the gas in the t f th i th hood to the velocity of the gas at a particular location outside of the duct. The following equation is valid for an aspect ratio of the hood opening up to 3 to 1. 1 Qg,h = ug,x (10 x2 + Ah)
162
where, Qg,h = volumetric flow rate of gas at the inlet of hood [ft3/min] ug,x = gas velocity at displacement x from hood inlet [ft/min] x = displacement between null point and inlet of hood [ft] Ah = cross-sectional area of hood inlet [ft2] If a rectangular hood opening is bounded on one side as presented below, then the equation describing the volumetric flow rate in the hood is as follows.
10 x 2 + 2A h Q g,h = ug,x 2
CEE 446
163
CEE 446
164
41
Schematic of Hood Opening with Gas Flow Blocked on One Side

Ah hood entrance Q g,h x (hood centerline) line of constant velocity boundary (e.g., wall, floor) u g,x null point fluid stream lines
Once the contaminant is captured within a hood then it needs to be transported through the duct system to the air quality control device and fan assembly. The resulting gas stream is then exhausted to the outside environment or possibly returned to the indoor environment if the gas stream is adequately cleaned.
CEE 446
165
CEE 446
166
Duct System with Air Quality Control Device and Induced Draft Fan
A B C D
The fans that are used to provide the pressure differential to draw the gas stream through the duct system can be of a centrifugal or axial type.
F E
air quality control device fan
CEE 446
42
Centrifugal Fan
fan enclosure
gas flow
Axial Fan
fan blade
outlet
external motor
fan motor protective screen
CEE 446
169
CEE 446
170
Parameters to consider when choosing a fan is the required volumetric flow rate, pressure drop, rate of rotation of the fan blades, power requirements, explosion proof capabilities, and resistance to physical attrition and chemical reaction.
A duct system needs to be balanced to achieve the appropriate gas flow rate at each hood entrance. The static regain balance method or the blast gate method can be used to balance the duct system's gas flows.
CEE 446
171
CEE 446
172
43
Static regain balance method is a trial and
error method that is not very flexible, more difficult to use, but does not require dampers in the exhaust ducts.
Blast gate method relies on the use of
Removal Efficiency for an Air Quality Control Device

It is important to be able to describe the removal efficiency for an air quality control device or the removal efficiency for two air quality control devices that are located in series.
dampers in the duct system to balance the gas flows within the duct system This system. method is more flexible to retrofit, easier to design, but requires dampers in the exhaust system.
174
Steady State Mass Flow Rates of a Contaminant for an Individual Air Quality Control Device
& min ( dp ) = mass per unit time of a contaminant with a particular diameter entering air quality control device & mcol ( dp ) = mass per unit time of a contaminant with a particular diameter that is removed from the gas stream & mout ( dp ) = mass per unit time of a contaminant with a particular diameter that penetrates through the air quality control device
min(dp)
air quality control device
mout(dp)
mcol(dp)
CEE 446
44
( dp ) =
The graded removal efficiency ((dp)), based on mass of the contaminant, for an air f th t i t f i quality control device is defined by:
& mcol ( dp ) & min ( dp ) & & min (dp ) mout (dp ) & min ( dp ) & mout (dp ) & min ( dp )
= 1-
= 1 - P(dp )
where, P(dp) = graded penetration of a contaminant p with a particular diameter
where,
(dp) = graded removal efficiency ((dp)) per p g [L unit particle size range [ -1]
The overall removal efficiency (T) of a contaminant for an air quality control device is:
& min,T = total mass feed rate of
contaminant into air quality control device [m/t]
T =
& min (dp ) (dp ) d(dp ) & min,T

CEE 446
45
For a discrete particle size distribution, T is described by:

T =
Steady State Mass Flow Rates of a Contaminant for Two Air Quality Control Devices in Series
& min ( dpi ) (dpi ) & i=1 min,T
min,1(dp)
There are many cases where two air quality control devices are located in series to remove a contaminant from a gas stream.
air quality control device (1)
air quality control device (2)
mout,2(dp)
mcol,1(dp) l1 mout,1(dp)
mcol,2(dp) l2 min,2(dp)
CEE 446
181
CEE 446
182
where,
& min,1(dp ) = mass per unit time of a contaminant with a particular diameter entering the first air quality control device & mout,1(dp ) = mass per unit time of a contaminant with a particular diameter penetrating the first air quality control device
& mcol,1(dp ) = mass per unit time of a contaminant with a particular diameter collected by the first air quality control device (dp )1 = fractional removal efficiency of a co ta contaminant with a pa t cu a a t t particular diameter, based on mass of the contaminant, for the first air quality control device
46
& min,2 (dp ) = mass per unit time of a contaminant with a particular diameter entering the second air quality control device & mout,2 (dp ) = mass per unit time of a contaminant with a particular diameter penetrating the second air quality control device
& mcol,2 (dp ) = mass per unit time of a contaminant with a particular diameter collected by the second air quality control device (dp )2 = fractional removal efficiency of a co ta contaminant with a pa t cu a a t t particular diameter, based on mass of the contaminant, for the second air quality control device
The graded removal efficiency, based on mass of the contaminant, for two air quality control devices ((dp)1,2),
(dp)1,2 = 1 - P(dp)1,2
(dp )1,2 = 1 - P(dp )1,2

& mout,2 (dp ) & min,1(dp )
where, P(dp)1 2 = graded p ( 1,2 g penetration of a contaminant with a particular diameter, based on mass of the contaminant, for both air quality control devices
= 1-
188
47
& & mout,2 (dp ) = (min,2 (dp ))(1 2 (dp )) & = (mout,1(dp ))(1 2 (dp )) & = (mout,1(dp ))(P2 (dp )) & & mout,1(dp ) = (min,1 (dp ))(1 1(dp )) & = (min,1(dp ))(P1(dp )) & & mout,2 (dp ) = (min,1(dp ))(P1(dp )) (P2 (dp ))
(dp )1,2 = 1 -
& mout,2 (dp ) & miin,1( dp ) & min,1 (dp )(P1(dp )(P2 (dp )) & min,1(dp )
(dp )1,2 = 1 -
= 1 - P1(dp ) P2 (dp )
CEE 446
189
CEE 446
190
Therefore, overall removal efficiency for both devices (T,1,2) is described by:
Mechanisms to Remove Particulate Contaminants from Gas Streams

The primary mechanisms for removal of particulate matter from gas streams is Brownian motion, interception, and impaction.
T,1,2 =
& min,1(dp ) (dp )1,2 d(dp ) & min,T
T,1,2 =
& min,1(dpi ) (dpi )1,2 & min,T i=1
CEE 446
191
CEE 446
192
48
Removal of Particles from a Gas Stream with a Collector Body via Brownian Motion, Interception, and Impaction
Brownian Motion impaction
Enhancement of these mechanisms can occur b using external f by i t l forces such as h electrostatic and/or centrifugal forces.
interception collector body fluid streamline

Separation and Removal of Particulate Contaminants from Gas Streams

Particulate contaminants are typically removed from industrial gas streams with the use of: Settling Chambers (gravitational force) Cyclones (centrifugal force) Wet Collectors (Brownian motion, interception, and impaction) Electrostatic Precipitators (electrostatic force) Fabric Filters (Brownian motion, interception, and impaction)
Settling Chamber Schematic of a Gravitational Settling Chamber

L W
_ ug Vtp _ ug = Qg/A CSA
Qg
shelves
ACSA = HW
n = number of channels
49
Settling chambers allow sufficient residence time for the particles to settle vertically in the gas stream and be captured by horizontal shelves. The vertical motion of the particles with respect to the gas is caused by the force of gravity. Flow within the settling chamber is characterized as laminar or turbulent with the Reynolds number for the settling chamber (Rechamber) Rechamber is ). < 2,300 for laminar flow and > 4,000 for turbulent flow.
Rechamber =
where, h
(u)(L)(g ) g
u = characteristic velocity = ug ug = superficial gas velocity in the settling chamber Qg A CSA

Qg = actual volumetric flow rate of the aerosol passing through the settling chamber ACSA = cross-sectional area of the settling chamber L = characteristic length = Dh Dh = hydraulic diameter of a duct = 4 (Acsa/perimeter)
ug =
Qg A CSA
Qg WH
Qg W Hn
Qg / n W H
W = width of the settling chamber idth f th ttli h b H = height of the settling chamber
H = height of one channel in the settling chamber
n = number of channels in the settling chamber

CEE 446
50
Dh =
4W H 2W H = 2(W + H) W + H
with one channel ( (n = 1) ) with n channels
Rechamber
Qg 2W H g W Hn W + H = g 2Qgg n(W + H) g
( )
2Qgg
Once Rechamber is known, then one of two known models can be used to describe particle removal by the settling chamber.
(nW + H) g
CEE 446
201
CEE 446
202
Laminar Flow Condition (Rechamber < 2,300)
Schematic of One Channel within a Settling Chamber

upper shelf particle _ ug
The graded collection efficiency, , for the settling chamber is described by

(dp ) = Z(dp ) H
where,
H
Vtp L
Z = vertical displacement of particle due to gravity during the gas residence time in the settling chamber H = height of one channel
lower shelf
51
Z(dp ) = Vtp (dp ) t

Z can be determined by equating the time that is allowed for the particle to settle vertically to the gas residence time in the chamber. Z(dp) = vertical displacement of particle due to gravity during the gas residence time in the settling chamber Vtp(dp) = terminal settling velocity of the particle t = time allowed for particle to settle in the chamber
CEE 446
205
CEE 446
206
L t= ug
Z(dp ) = Vtp (dp )
L ug
L = length of the shelves in the settling chamber t = time allowed for particle to settle in the chamber
ug =
Qg A CSA
Qg Wn H
Z(dp ) =
LVtp (dp ) W H Wn Qg
CEE 446
207
CEE 446
208
52
This is the same equation as Eq. 5-22 in Air Pollution: Its Origin and Control with n = 1.
(dp ) =
Z(dp ) H
LVtp (dp )W H )Wn Qg H

Acollection = L W n
Vtp (dp ) A collection ( Qg
(dp ) = (
Vtp (dp )LWn Qg

(dp ) =
CEE 446
209
CEE 446
210
Turbulent Flow Condition (Rechamber > 4,000)
Material Balance Used to Describe Particle Removal with Turbulent Flow

m p,v x W m p,vo H V tp L m p,v
m p vx + p,v x
ug
m p,vL
The turbulent flow model assumes well , mixed conditions in the vertical direction, reintrainment of particles is negligible, and the particles are moving in the horizontal direction with the gas stream.
212
53
Applying a material balance in the horizontal direction, for one channel:
(ug )(A CSA )(mp,v x + x mp,v x ) = (mp,v )(Vtp (dp ))(A collection ) (ug )( HW )( mp,v ) = (mp,v )(Vtp (dp ))(W x)
particulate matter removal rate between two vertical planes
particulate matter deposition rate to the bottom of the channel
let as
mp,v d(mp,v) x d(x)
ug HWd(mp,v ) = mp,v Vtp (dp )Wd( x )
CEE 446
213
CEE 446
214
Apply separation of variables:
P(dp ) =
mp,vL mp,v o
d(mp,v ) mp,v
mp,v
Vtp (dp ) ug H
x =L
d( x )
Vtp (dp )L = exp ug H
L mp,v o
d(mp,v ) mp,v
= x =o
Vtp (dp ) d( x ) ug H
-Vtp (dp )L (dp ) = 1 - P(dp) = 1- exp ug H
but
ug =
Qg n HW
mp,vL ln mp,v o
CEE 446
Vtp (dp )L = ug H
215 CEE 446
n = number of channels
216
54
- V (dp )Ln HW (dp ) = 1 - exp tp Qg H

Acollection = WLn for multiple shelf settling chamber
- Vtp(dp )A collection (dp ) = 1 - exp Qg
This is the same equation as Eq. 5-55b, p.224 in Air Pollution: Its Origin and Control with n = 1.
CEE 446
Characteristics of Settling Chambers

Advantages No N moving parts i t Low capital cost Low operating cost Does not require use of additional materials Disadvantages Large volume requirement Low collection efficiency for dp < 10 m
LVt W d = 1 exp Q g
Example (Gravitational Settling Chamber)
Define variables given in problem. dp = 20 m p = 1 g/cm3 Qg = 5 x 104 m3/hr gas = dry air T = 298 K P = 1 atm n = 20 L = 10 m W = 10 m H = 0.2 m (dp = 20 m) = ?
Calculate the collection efficiency of a gravitational settling chamber for 20 m diameter particles with density of 1 g/cm3. Total gas flow rate for the chamber is 5 x 104 m3/hr. The carrier gas for the particles is dry air at 298 K and 1 atm. The chamber y has 20 shelves that are 10 m long and 10 m wide with vertical shelf spacing of 0.2 m.
CEE 446
55
Calculate the Reynolds Number for the Chamber

Re chamber = 2 Qg g n( W + H) g )
Re chamber =
(2)(5 x 10 4 m3 / hr )(1.185 kg / m3 ) (20 )(10 m + 0.2 m)(0.0663 kg / m hr ) turbulent flow
= 8,761 > 4,000
g = 0.0663 kg/m hr g (1 atm) 29 P MW g mole g = = RT atm l 0.0821 (298 K ) g mole K = 1.185 g/l = 1.185 kg/m 3
Determine collection efficiency for turbulent flow condition.

- V A (dp ) = 1 - exp tp collection Qg
where,
Vtp = 1.2 cm/sec = 43.2 m/hr p (p. 211, Fig. 5 - 8 in textbook) A collection = n L W = (20)(10 m)(10 M) = 2,000 m 2
(43.2 m/hr)(2,00 0 m2 ) (dp ) = 1 - exp (5 x 10 4 m3 / hr ) = 0.8224
= 82% for dp = 20 m
CEE 446
223
CEE 446
224
56
Cyclones (Centrifugal Force)
Gravitational force is useful to remove large particles (dp > 10 m) from gas streams but is t i not very effective for smaller particles. ff ti f ll ti l The centrifugal force can be used to achieve larger removal efficiencies for smaller particles. The gas stream is forced to change its direction with the particles following a different direction. The centrifugal force causes the particle to
be transported in a different direction than the gas stream allowing for separation and collection of particles from gas streams. streams
Typical cyclones are vane-axial or involute as described in the following figures figures.
CEE 446
225
CEE 446
226
Individual Vane-axial Cyclone

outlet for clean gas stream inlet for particle laden gas stream inlet for particle laden gas stream
Involute Cyclone and its Standardized Proportions

H Sc
L1 = 2Do
De
L1
L2 = 2Do H = Do/2
vortex finder
vane
Do
L2
Sc = Do/9 De = Do/2 Dd = Do/4
outlet for particles

Dd
W = Do/4 (width of cyclone inlet)

228
57
Top View of Involute Cyclone

R
There is interest in determining the particle removal efficiency of an involute cyclone with standardized proportions as described below.
R R* Ri
Ro
W
where, Ro = outer radius of cyclone Ri = inner radius of cyclone W = Ro - Ri = width of the cyclone's inlet R* = minimum or critical radius for which a particle of diameter, dp, will just reach the outer wall of the cyclone and be removed from the gas stream The graded collection efficiency of an involute cyclone, (dp), can then be described by:
(dp ) = Ro R * Ro R * = R o Ri W
CEE 446
231
CEE 446
232
58
Particle Trajectory in a Cyclone of Standardized Proportion

A force balance can be used to describe the radial velocity of a particle that is located in a gas stream that is turning with radius R.

Vp, tang R
Vp, rad
particle
using a force balance on a particle

r r mp ap = Fi
mp v d(Vp ) d(t) v v = Fac Fd
Assume the particle is spherical
v2 v Vp, tang Fac = centrifugal force = mp R v Vp, tang = tangential velocity vector of particle R = radius of curvature
where, r ap = acceleration of particle
v Fd = drag force assuming Stoke's Law force, v = 3 g dp Vp, rad v Vp, rad = radial velocity vector of particle
236
59
It is assumed that particle mass remains constant and steady-state condition is achieved very rapidly such that:
v d( Vp, rad )
d( t )
mp
v v Fac = Fd v2 Vp, tang R v = 3 g dp Vp, rad
mp
=0
and
v2 v mp Vp, tang Vp, rad = 3 g dpR
CEE 446
237
CEE 446
238
Assuming spherical particles with density, p,

mp = 1 3 dp p 6
It is also assumed that

v Qg Vp, tang = ug = WH
1 3 v2 dp p Vp, tang v Vp, rad = 6 3 g dpR
=
CEE 446
v2 2 dpp Vp, tang

18gR
239 CEE 446
with radius, R, as some value between the inner radius, Ri, and outer radius Ro, of the radius radius, cyclone.
240
60
v Vp, rad is then described by:
Therefore,
Vp,rad =
Ro R * t
Vp,rad
2 2 Ro R* dp p ug = = t 18gR
where t is the time period for the particle to be transported from R* to Ro.
Ro R * =
2 2 dp p ug t
18gR
CEE 446
241
CEE 446
242
It is now necessary to determine t. where,

displaceme nt of the gas stream in the tangential direction t = (superficial gas velocity) circumfere nce of the vortex with radius R (number of turns) = (superficial gas velocity) =
CEE 446
Ne = number of equivalent turns of the g q gas stream in the cyclone

= L 1 L1 + 2 H 2
(2R)(Ne ) ug
2009 Mark J. Rood, University of Illinois at Urbana-Champaign. All Rights Reserved. 243 CEE 446 2009 Mark J. Rood, University of Illinois at Urbana-Champaign. All Rights Reserved. 244
61
where, H = height of the inlet to involute cyclone g y L1 = height of the cylindrical portion of cyclone L2 = height of the conical p g portion of cyclone y
Therefore,
2 2 dp p ug (2R )(Ne )
Ro R =
18 gR ug
2 dp p ug Ne
9 9 g
CEE 446
245
CEE 446
246
and,
2 R R* dpp ugNe (dp ) = o = W 9gW
then,
(dp ) =
2 dppQgNe
but,
ug = Qg WH
9gW 2H
CEE 446
247
CEE 446
248
62
The expression for (dp) is difficult to apply because (dp) can be calculated to be > 100% and there were numerous assumptions in its derivation. Therefore, define dp,50 as the particle diameter that is collected at (dp) = 0.5.
(dp ) = 0.5 =
2 dp,50pQgNe
9gW 2H
dp,50
9 g W 2H = 2p Q g Ne
1/ 2
CEE 446
249
CEE 446
250
This expression for dp,50 is in the same form as equation (5-63) on page 229 in Air Pollution: Its Origin and Control. Now the problem is how to apply the equations to calculate collection efficiency for the cyclone A calibration curve can be cyclone. used for cyclones of standardized proportion.
Calibration Curve for Cyclones of Standardized Proportions

C
(dp)
B A
dp / dp,50
63
dp,50 can be readily calculated given the geometry of the cyclone, density of the particles, particles viscosity of the gas and volume flow rate of the gas stream. dp/dp,50 values can then be determined for the particle size distribution of interest. Values of (dp) can then be read from the calibration curve for calculated values of dp/dp,50. The overall collection efficiency, T, can then be calculated.
m (d ) & in pi (dp ) T = i & i=1 m T,in
Pressure drop (P) across air quality control devices is also important because operating cost of the device can depend heavily on pressure drop. y p p
254
An empirical expression describing pressure drop (P) for cyclones is presented below.
where, P = pressure drop [m H2O] g = inlet superficial gas velocity [m/sec] g = gas density [Kg/m3] K = empirical constant = 16 for standard tangential inlet or = 7.5 for vane axial entry H = height of cyclones inlet [m] W = width of cyclones inlet [m] cyclone s g = gravitational force constant [9.8 m/sec2] L = density of liquid water [1,000 Kg/m3] De = diameter of cyclones outlet for gas stream [m]
P =
u g
2 g
2gL
HW = (PT 2 PT1 ) D2 e
CEE 446
255
CEE 446
256
64
Characteristics of Involute Cyclones

Advantages No moving parts Low maintenance requirements Low operating and capital cost Removes liquid or solid particles that are not sticky Disadvantages
Typical Values for Involute Cyclones

(dp ) 50% for dp = 1 10 m
Low collection for dp < 1m
Do = 0.5 to 3 m ug = 10 20 m/sec
CEE 446
257
CEE 446
258
Example (Cyclone)
An involute cyclone of standard proportions with a 2 m diameter is operated at a gas flow rate of 10 m3/sec. The gas stream consists of air at 500 K and 1 atm. The calibration curve for the cyclone is presented in Figure 5-17 (p.230) of Air Pollution: Its Origin and Control. What is the collection efficiency for a 10 m diameter particle?
Cyclone of Standard Proportions Do = 2 m 3 Qg = 10 m /sec gas = air T = 500 K P = 1 atm calibration curve Figure 5-17 5 17 (dp = 10 m) = ?
260
65
calculate dp,50
where,
g = 0.0969 kg g = 0.0269 m hr m sec
dp,50
9 g W 2H = 2p Q g Ne
1/ 2
W=
Do 2 = = 0 .5 m 4 4 Do 2 = =1 m 2 2
H=
CEE 446
261
CEE 446
262
p = ? assume p = 1 g m3
g cm3 dp,50
= 1 x 10
g (9) 0.0269 (0.5 m)2 (1 m) m sec = m3 g ( )(6) (2)1 x 10 6 10 m3 sec = 1.257 x 10 - 5 m = 12.67 m
1/ 2
Ne =
2 Do L 2 1 2 Do + =6 L1 + 2 = 2 Do H 2
CEE 446
263
CEE 446
264
66
calculate dp/dp,50
dp / dp,50 = 10 m = 0.79 12.67 m
Wet Scrubbers (Brownian motion, Interception, and Impaction)
Read (dp = 10 m) Figure 5-17

(dp = 10 m) 38%
Wet scrubbers use Brownian motion, interception, interception and impaction to remove particles from gas streams. They also cause hygroscopic particles to grow in size such that impaction of particles onto the scrubber droplets becomes more effective. Wet p scrubber models, in particular for venturi scrubbers, assume the dominant mechanism for particle removal is impaction.
CEE 446
Optimal scrubber droplet diameter ranges from 500 m to 1,000 m when considering impaction. Impaction of particles onto a collector body is characterized by the dimensionless impaction number (Ni) also known as Stokes number (Stk). This number is a ratio of the particle's stop distance (xs) to particle s a characteristic length of the collector body (e.g., diameter of the scrubber droplet (dc)).
Stk = Ni =
x s Vpo = dc dc
where,
= relaxation time
Vpo = initial velocity of the particle dc = diameter of the collector body (e.g., diameter of a sphere or cylinder)
268
67
Particle Decelerating in a Quiescent Gas

The relaxation time () can be determined by applying a f force balance on a particle that is decelerating in the horizontal direction as described in the next slide.
fluid st ea streamline e quiescent gas ug = Vpo

ug decelerating particle
CEE 446
269
CEE 446
270
The stop distance (xs) needs to be determined to calculate Stk. Newtons 2nd law can be used to describe a decelerating particle of constant mass moving in the horizontal direction and assuming Stokes law (Rep < 0.1).
mp = d( Vp ) d( t ) =
1 3 dp p 6 = d( Vp ) d( x ) Vp
d( Vp ) d( x ) d( x ) d( t )
mp d(Vp ) d(t)
3 g dp Vp Kc
3 gdp Vp d( Vp ) 1 3 Vp = dp p 6 d( x ) Kc
CEE 446
271
CEE 446
272
68
Use separation of variables and solve for d(x).

x=xs v p =0
p
Therefore,
=
2 dp pK c
x = 0
d( x ) = v
2 dp pK c
= v po
18 g
d( Vp )
18 g
2 dp pK c V = V xs = po 18 g po
and
Stk =
2 dp pK c Vpo
18 gdc
CEE 446
273
CEE 446
274
Particle Collection Efficiency Versus Stokes Number

1
There are many types of wet scrubbers. The three primary types of wet scrubbers are counter-current spray towers, cyclonic wet p y , y scrubbers, and venturi scrubbers. Counter-current spray towers typically operate with liquid collector bodies (scrubber d l t ) t ( bb droplets) traveling d li downward and d d the gas stream containing the particulate contaminant material traveling upward.
0 Stk
69
Counter-current Spray Tower

outlet for clean gas stream demister d i t spray nozzles particle inlet for water stream scrubber droplet inlet for gas stream containing particulate matter outlet for liquid stream containing particulate matter
Cyclonic scrubbers, or wet cyclones, atomize droplets of water with a spray bar that is located along the centerline of the cyclone. These droplets then collect particulate matter as they are transported to the outer edge of the cyclone. The liquid also allows cleansing of th walls i th cyclone. l i f the ll in the l
CEE 446
Cyclonic Wet Scrubber

clean gas stream scrubber droplet
spray bar
particle particle laden gas stream filtered liquid

CEE 446
Venturi scrubbers work by accelerating the gas stream to velocities around 50 to 150 m/sec. m/sec The gas stream accelerates because the duct that contains the gas stream is constricted similar to that of a venturi shown below.
excess liquid with particles

70
Venturi Scrubber and Cyclone Separator

inlet for gas stream containing particulate contaminants outlet for clean gas stream
particles cyclone separator inlet for liquid throat scrubber droplets outlet for liquid containing particulate contaminants
As the gas stream is flowing through the venturi, liquid is injected into the beginning of the venturi or at the throat entrance of the venturi. The liquid is atomized due to the hi h t i Th li id i t i d d t th high velocity gas stream flowing past the inlets for the fluid. At the instant that the liquid is atomized, the resulting droplets are at the same velocity as the wall of the venturi. Such venturi condition allows the largest possible relative velocity between the particles in the gas stream and the large droplets that have been
atomized in the venturi. This large relative velocity between the scrubber droplets and particulate contaminants allow for impaction of the th particulate contaminants onto th scrubber ti l t t i t t the bb droplets. The particulate material then becomes part of the scrubber droplets allowing much easier removal of the contaminants from the gas stream because of the large size of the scrubber droplets. The scrubber droplets are typically removed from the gas stream with a centrifugal separator.
Equations are developed in Air Pollution: Its Origin and Control that describe the particle removal efficiency that can be achieved by a venturi scrubber and the pressure drop that results from atomizing the scrubber droplets and accelerating the droplets to approach the same velocity as the g pp y gas streams. Assumptions used in the development of the venturi scrubber model include:
71
Gas velocity is constant throughout the length of the venturi's throat Flow is one dimensional, incompressible, and adiabatic Particles are collected by the atomized droplets by impaction Droplets are uniformly distributed along the cross section of the venturi's throat
Diameter of the scrubber droplets remain constant Volume ratio of the scrubber droplets to the gas stream is small Pressure forces around the scrubber droplets are symmetrical and therefore ignored
CEE 446
285
CEE 446
286
By considering a force balance on the atomized droplets, an equation can be developed that describes how the scrubber droplets accelerate along the length of the venturi's throat and the pressure drop caused by accelerating the droplets. Such force balance results in the following g equation for pressure drop.
2 Q P = l ug l Qg
CEE 446
287
CEE 446
288
72
where,
P = -change in pressure accross the length of the venturi's throat (-(P2 - P1))
= pressure drop
= correction factor for droplets lost to walls of venturi = 0.85
ug = velocity of gas in the venturi s throat venturi's Ql = volume ratio of liquid to gas flow rates Q g
Q 2 P = (1.02 x 10 3 )(ug ) l Qg P = cm H2 O ( l ) = [ ] ug = [cm / sec ] Ql = [ ] Qg
290
Particle collection efficiency achievable by a venturi scrubber has been described by considering the removal of particles due to impaction, pressure drop, contaminant particle diameter, viscosity of the gas stream, and densities of the scrubber liquid droplets d d l t and contaminant particles as t i t ti l described in the next slide.
2 2 Q (6.3 x 104 )l pK c dp ug l f 2 Qg (dp) = 1 exp 2 g where,
( dp ) = graded particle collection efficiency [-]
l p
= density of scrubber droplets = density of particulate contaminant
Kc = Cunningham correction factor for particulate contaminant [-]

73
where, dp = diameter of particulate contaminant f = experimental coefficient = 0.1 to 0.4 [-]

g = gas viscosity
Characteristics of Wet Scrubbers

Advantages Possible to achieve high particle collection efficiency Humidifies and cools gas stream Possible to simultaneously remove particles and gases (even sticky particles) ti l ) Variable throat area of venturi allows selection of particle collection efficiency
293 CEE 446
Disadvantages Requires the use of a liquid medium that needs to be treated High pressure drop for high particle collection efficiencies Corrosion and precipitation can occur
Use any set of dimensionally consistent units.
Liquid can freeze
CEE 446
294
Typical Values for Venturi Scrubbers

m m < ug < 100 sec sec l m < Ql l < 2.0 3 Qg m
Example (Venturi Scrubber)
50
0.2
5 10%(dp ~ 0.1 m) < (dp ) < 100%(dp ~ 50 m) 01 10 cm H2 O( l ) < P < 70 cm H2 O( l )
Emissions from an industrial plant contain 0.4 m diameter particles that are dispersed in air The density of the air. particles is 1.5 g/cm3. The gas stream is at 77C and 1 atm. These emissions are treated by a venturi scrubber that is operating at 40 cm H2O(l) pressure drop p g p p with a gas velocity in the throat of the venturi of 100 m/sec. The venturi is using water as its scrubbing fluid. Estimate the
CEE 446
74
liquid to gas flow ratio and the range of particle collection efficiencies that are expected for this venturi scrubber. dp = 0.4 m gas = air
p = 1 5 g/cm3 1.5
P = 1 atm P = 40 cm H2 O ug = 100 m / sec l = 1g / cm3 Ql =? Qg (dp = 0.4 m) = ?

.
T = 77C = 350 K
CEE 446
297
CEE 446
298
Ca cu ate the qu d Calculate t e liquid to gas ratio. at o

2 Q P = l ug l Q g 2 Q = (1.02 10 3 )ug l Q g
If P is in units of cm H2O(l), ugis in units of cm/sec, and Ql has the same units as Qg.
Ql P = 2 Q g (1.02 x 10 3 )ug
CEE 446
299
CEE 446
300
75
where,
P = 40 cm H2 O( l )
. .
( 40 cm H2O( l ) ) Ql = Q g (0.85 )(1.02 x 10 3 )(1 x 10 4 )2
= 0.85 ug = 1 x 104 cm sec
Ql = 4.61 x 104 Qg
CEE 446
301
CEE 446
302
where, Calculate particulate material collection efficiency.

2 2 Q (6.3 x 104 )l pK c dp ug l f 2 Qg (dp ) = 1 exp 2 g
l = 1 g/cm3 p = 1.5 g/cm3 Kc = ? dp = 0.4 x 10-4cm f = ? (ranges 0.1 to 0 4) 0 1 0.4)
CEE 446
303
CEE 446
304
76
P = 40 cm H2O g = 0.0748 kg/m hr = 2.078 10-5 kg/m sec = 2.078 10-4 g/cm sec
1.1 K c = 1 + Kn 1.257 + 0.4 exp Kn Kn = 2 g dp g 0.499 g ug,molecule
g =
CEE 446
305
CEE 446
306
8RT ug,molecule = MW g
1/ 2
10 3 g J 8.314 ( 8 ) 8 314 ( 350 K ) Kg g-mole K = g ( ) 29 g-mole = 505.5 m cm = 5.055 10 4 sec sec
1/ 2
g =
P MWg RT
g (1 atm) 29 g mole = atm l 0.0821 (350 K ) g mole K
= 1.009
g g = 1.009 x 10 3 l cm3
(note : J = Kg m
CEE 446
/ sec 2
)
77
g 2.078 x 10 4 cm sec g = g cm (0.499 )1.009 x 10 3 5.055 x 10 4 3 sec cm = 8.165 x 10 6 cm = 0.082 m
1 .1 K c = 1 + (0.408 )1.257 + 0.4 exp 0.408 = 1.52
Kn =
(2)(0.082 m) = 0.408 > 0.01 (0.4 m)
g g 4 2 4 cm 2 4 2 (6.3 x10 4 ) 1 1.5 (1.52)(0.4x10 cm) (1x10 sec ) (4.61x10 )(f ) cm3 cm3 (dp ) = 1 exp 2 g 4 2.078 2 078 x10 cm sec
CEE 446
309
CEE 446
310
Electrostatic Precipitation (Electrostatically Enhanced Impaction)
(dp) 1 - exp[ -2.4536 f2 ]
(f = 0.1) = 2% ( (f = 0.4) = 28% )
for dp = 0.4 m
Electrostatic precipitation was initially used as an industrial air quality control device in USA and Europe during the early 1900s. Sulfuric acid particles were removed from gas streams by electrically charging the particles and then subjecting the particles to a large electric field, causing the particles to migrate toward a collection surface that allows for impaction and removal of the particles from the gas stream.
78
The particles pass through a volume that has a large electrical potential (~ 50 kV) applied across a spacing of about 10 cm. This large field strength (electrical potential/distance between electrodes) causes the release of free electrons from the discharge electrode. These electrons have a large amount of energy as a result of the very high but localized electrical high, field strength. As the electrons migrate away from the discharge electrode, they impact against gas molecules causing ionization of
the gas molecules and release of more free electrons. Such generation of the large number concentration of free electrons is called electron avalanche. Eventually the free electrons are transported beyond the localized high field strength near the discharge electrode causing a decrease of their kinetic energy. As the electrons decelerate, they attach themselves to particles instead of generating more free electrons.
The mechanisms that are responsible for deposition of electrons onto aerosol particles are diffusion and field charging. Diffusion charging dominates for particles with diameters smaller than about 0.3 m while field charging dominated for particles with diameters larger than about 0.3 m. Diffusion charging will be ignored in this course course. Therefore, the equations will underpredict particulate collection efficiency for particles smaller than about 0.3 m.
There are typically two configurations for electrostatic precipitators ( S ) tubular (ESPs), and flat plate electrostatic precipitators as described in the following schematics.
316
79
Tubular Electrostatic Precipitator

clean gas stream discharge electrode p
collection electrode particles ti l particle laden gas stream

Tubular electrostatic precipitators are used to remove liquid droplets from gas streams (e.g., droplets of sulfuric acid). Tubular ESPs are not as difficult to model as flatplate ESPs, but flat-plate ESPs are much more commonly used (e.g., coal fired power plants). Flat plate ESPs can be plants) Flat-plate
318
Two-Stage Flat-Plate Electrostatic Precipitator

configured as a two-stage or a single-stage design. Two-stage ESPs charge the particles first and then remove the particles in a different volume of the ESP. Singlestage ESPs charge the particles and remove the particles from the gas stream in the same volume.
collection electrodes _ ug p
particles
negative discharge electrodes
y x
CEE 446
319
CEE 446
320
80
One-Stage Flat-Plate Electrostatic Precipitator

particles H discharge electrodes D collection electrodes L n = 2 n = number of channels
_ ug
Axonometric View
collection electrodes _ ug p p
discharge electrodes
Particulate collection efficiency for a flat-plate single-stage ESP operating under laminar flow conditions can be described by the fl diti b d ib d b th following equation that was derived with the same type of approach that was taken for laminar flow settling chambers. Here, particle migration is in the horizontal direction caused by the electric field strength in the ESP.
particles
CEE 446
Top View
(dp ) =
p A collection Qg
n = number of channels H = height of collection electrodes L = length of collection electrode Qg = volume flow rate of gas stream
where, where p = electric migration velocity of the particle toward the collection electrode
Acollection = total collection area for the particles = 2nHL

ESPs are typically operating under turbulent flow conditions. Particulate collection efficiency for a flat-plate single-stage ESP ffi i f fl t l t i l t can be described by applying the following material balance.
81
Material Balance Applied to a TurbulentFlow Single-Stage Flat-Plate Electrostatic Precipitator

mp,vx D mp,vo H _ ug mp,vL x mp,vx+x
L mp,v and p collection electrode

CEE 446
This model assumes well-mixed conditions in the y and z directions with a concentration gradient of particulate mass concentration in the x direction for the bulk gas stream. This concentration gradient is due to the loss of particles to the collection electrodes of the ESP.
z x
discharge electrode
Therefore, a material balance results in the following relationship.
p p particulate (A ) particulate ( A ) mass flux x + x CSA x + x mass flux x CSA x x direction = + particulate ( A CSA ) y mass flux y y direction
mass per unit time of particles lost from the gas stream between x and x + x
mass per unit time of particles deposited onto the collection electrodes between x and x + x
CEE 446
327
CEE 446
328
82
(mp,v x + x ug ) (HD) (mp,v x ug ) (HD)
(ug )(HD)(mp,v x + x mp,v x ) = mp,v p 2Hx
= [(mp,v p )(2H x )]
let as
mp,vx+x mp,vx d(mp,v) x d(x)
where, p = migration velocity of the particle toward the collection electrode t d th ll ti l t d
(ug )(HD)d(mp,v ) = mp,v p 2Hd(x)
CEE 446
329
CEE 446
330
Assuming p f(x) Use separation of variables and integrate the resulting equation.
mp,v =mp,v
mp,vL 2L = p ln uD mp,v g o mp,vL mp,v o 2L = exp p uD g
mp,v =mp,v oL
d(mp,v ) mp,v
= x =0
x =L
p 2d( x ) ugD
P( dp ) =
2L (dp ) = 1 P( dp ) = 1 exp p uD g
83
but ug =
Qg nHD
p 2nLH (dp ) = 1 exp Qg
p A collection ( dp ) = 1 exp Qg
but b t Acollection = 2 LH 2nLH
CEE 446
333
CEE 446
334
This equation is referred to as the DeutschAnderson equation. Assumptions include constant migration velocity for a particle of a given diameter, uniform mass concentration of particulate material in the y and z direction, no reintrainment of particles from the collection electrodes, no transport of particles above or below the collection electrodes ("sneakage") ("sneakage"), and a constant gas stream velocity in the x _ direction (ug = constant).
Corrections to the Deutsch-Anderson equation for non-ideal conditions have been developed as presented below.
p A collection (dp ) = 1 1+ Qg
n'
where, 3 n' 5.
336
84
p A collection (dp ) = 1 exp p Qg
where, x 0.5.
If the overall collection efficiency (T), Qg, and Acollection are determined experimentally, then p y, p can be calculated with the DeutschAnderson equation with p = p when considering overall particle collection efficiency. p is referred to as the "effective or drift migration velocity.
CEE 446
337
CEE 446
338
If there is interest in knowing (dp), then particle size distribution measurements need to be measured upstream and downstream of the ESP or p(dp) needs to be calculated. p(dp) can be calculated using Newton's second law, assuming constant particle mass and steady-state conditions.
Assuming Newton's second law, perpendicular to th t the collection electrode. ll ti l t d

d(p ) d( t )
mp
= Felectrical Fdrag
CEE 446
339
CEE 446
340
85
where, Felectrical = force exerted on particle due to g electric field strength Fdrag = drag force exerted on particle mp d(p ) d( t )
Determine Felectrical Felectrical = qpEce
where, =0 qp = electric charge of particle [C] Ece = electric field strength at collection electrode
Felectric = Fdrag
Assume saturation charge of the particles due to field charging to determine qp.
2 qp = Pp' oEpc dp
Dp = dielectric constant of the particulate material [-] (e.g., Dp = 6.9 for neoprene, neoprene and Dp = 3.8 for quartz) 38 o = permittivity of free space = 8.854 x 10-12 C/V m (Coulombs Law) Epc = electric field strength where p g particle is charged [V/m] dp = diameter of particle [m]
where,
Pp' =
3Dp
(Dp + 2)
CEE 446
343
CEE 446
344
86
Determine Fdrag: Assume Stoke's Law and correct Fdrag with the Cunningham correction factor factor.
Fdrag = 3 gdp p Kc
p can now be determined.

Felectrical = Fdrag g
2 Pp o Epc dp Ece =
3gdpp Kc
p =
Pp oEpc dpEceK c 3g
CEE 446
345
CEE 446
346
The electric field strength (E) depends on location but that dependence on location is p "second order." Therefore, E is assumed _ constant and equal to E = Epc = Ece. E= mean electical potential V = displacement between D/2 discharge and collection electrodes
347 CEE 446
Therefore,
p =
Ppo E 2dpK c 3g
CEE 446
348
87
Characteristics of Electrostatic Precipitators

Advantages Collects liquid or solid particles Disadvantages Particulate collection efficiency depends on composition of particles
Typical Values For Single-Stage FlatPlate Electrostatic Precipitators

30 kV V 50 kV 0.1 m D 0.3 m 1 m H 10 m 0.01 m m p 0.2 sec sec
Low operating cost due to High capital cost compared low pressure drop and low to fabric filtration electrical consumption No liquid effluent for dry q y ESPs High particulate mass collection efficiency Complicated to design and p g operate Requires large electric field strength
CEE 446
349
CEE 446
350
m m 0 .5 ug 3 sec sec 300 m2 1,000 am 3 min A collection 3,000 Qg m2 1,000 am 3 min 2
1. 0 W am min
3
L 1 .5 H
Pc W 30 Qg am3 min
where, A collection = specific collection area for ESP Qg

(am3 = actual m3) where,

Pc = specific corona power for ESP Qg
88
Example (Electrostatic Precipitator)
2 Dp 8 P 5 cm H2O(l) 2 cm/sec p 20 cm/sec overall > 85 to 99+ %
An electrostatic precipitator (ESP) is used to separate and remove particles from a gas stream that is generated by a combustion chamber. Twenty percent by mass of the particles have a diameter of 0.3 m while the balance of the particles have a diameter of 1 0 m. The temperature, 1.0 m temperature pressure, and mean molecular weight of the gas stream is 140C, 1 atm, and 29 g/gmole. The dielectric constant of the
CEE 446
20% by mass of particles are 0.3 m in diameter particles, mean electric field strength in the ESP, collection electrode spacing, total gas flow rate number of channels and total rate, channels, collection electrode area are 3, 4 kV cm-1, 20 cm, 100 m3 sec-1, 50, and 8,000 m2, respectively. The overall length to height ratio of the collection electrodes is 1.1. Determine the overall particulate mass collection efficiency for the ESP. 80% by mass of particles are 1.0 m in diameter
T = 140o C P = 1 atm MWg = 29 g/g-mole D'p = 3

CEE 446
89
E = 4 kV/cm = 400,000 V/m D = 20 cm Q g = 100 m / sec A collection = 8,000 m 2 n = 50 L/H = 1.1

Determine Rechamber, ESP

ugDh g g Qg DnH 4DH 2(D + H)
358
Rechamber, ESP = h b
ug =
Dh =
Rechamber , ESP
Qg DH 2 (g ) DnH D + H = Rec = g 2Qgg n(D + H)g
g =
P(MW g ) RT
Rec =
g (1 atm) 29 103 l g mole = m3 atml 0.0821 (413K) g mole K = 8.55 x 102 g m3
where,
Q g = 100
CEE 446
m3 m3 = 3.60 x 10 5 sec hr
359 CEE 446
n = 50
90
D = 0.2 m H=?= H= A collection 2nL
g (T = 413K ) = 0.0845
kg m hr
A collection 2nH(1.1)
1/ 2
Therefore,
m3 kg 0.855 ( 2) 3.6 10 5 hr m3 Re c = kg (50 ) (0.2 m) + (8.528 m ) 0.0845 m hr = 8.528 m
361 CEE 446
A H = collection 2n(1.1) 2n(1 1)
8,000 m 2 H= (2)(50)(1.1)
CEE 446
1/ 2
= 1.7 10 4 >> 4 10 3
The gas stream is turbulent and the DeutschAnderson equation should be used to q determine the graded collection efficiency. A (dp ) = 1 exp p collection Qg
Acollection and Qg are given. Therefore, need to calculate p(dp) to determine (dp). Ppo E 2dpK c 3g
p =
CEE 446
363
CEE 446
364
91
where,
Pp =
3D (3)(3) p = = 1. 8 (D + 2) (3 + 2) p C Vm
dp,2 = 1.0 x 10 6 m 1.10 K c,i = 1 + Kni 1.257 + 0.4 exp Kni Kni = 2 g dp,i
o = 8.854 x 10 12 V m
E = 400,000
dp,1 = 0.3 x 10 6 m
g =
g 0.499 g ug,molecule kg m hr
ug,molecule
8RT = MW g
1/2
g (T = 413K ) = 0.0845 = 2.335 x 10 5
kg m sec
3 g J (8) 8.314 (413K) 10 g mole K kg = g ( ) 29 g mole = 549.1 m sec
1/2
CEE 446
367
CEE 446
368
92
5 kg 3 g 6 m 10 10 2.335 10 m sec kg m g = g m (0.499 ) 8.55 x 102 3 549.1 sec m = 0.10 m
Kn dp = 0.3m =
2 g dp,i
(2)(0.10) = 0.667 > 0.01 ( 0 .3 )
Kn dp =1m =
(2)(0.10 ) = 0.200 > 0.01 (1)
Therefore, Therefore
Therefore, calculate Kc,i for dp = 0.3 m and 1.0 m.
CEE 446
369
CEE 446
370
1.10 K c,dp =0.3m = 1 + Kn1.257 + 0.4 exp = 1.89 Kn K c,dp =1.0m = 1.25
p,dp =0.3m =
Pp oE2 dpK c 3 g
2
C V (1.8) 8.854 10 12 400,000 (0.3 10 6 m)(1.89) Vm m = kg (3) 2.335 10 5 m sec
p,i can now be calculated.

p,dp =0.3m = 0.021m / sec p,dp =1.0m = 0.046 m / sec
CEE 446
371
CEE 446
372
93
m 2 0.021 (8,000 m ) sec ( dp = 0.3m ) = 1 exp 3 m 100 sec = 0 814 0.814
m 2 0.046 (8,000m ) sec ( dp = 1.0m ) = 1 exp 3 m 100 sec = 0 975 0.975
CEE 446
373
CEE 446
374
Filtration (Brownian Motion, Interception, and Impaction)

m overall = in,i (i ) i mT
overall = (0.2)(0.814) + (0.8)(0.975) = 0.943
overall = 94.3%
One method that is very effective at removing particles from gas streams is filtration. The gas stream, that contains particulate material that needs to be removed from the gas stream, is passed through some type of media that allows penetration of the gas stream but captures the particulate material. The media can consist of a single sheet of woven material
CEE 446
94
(e.g., fiberglass, cotton, and nylon), membranes (teflon matrix), fritted stainless steel, steel and cartridges (e g high efficiency (e.g., particulate air (HEPA) filters). The choice of fabric is dependent on the composition of the gas stream and particulate material, temperature of the g stream, and the p gas , acceptable levels of particulate collection efficiency and pressure drop.
Initially the filter itself performs most of the filtration of the particles from the gas streams. streams As more of the particles are filtered from the gas stream, the particles form a mat of particulate material referred to as a "filter cake." The filter cake allows for more extensive filtration of the particles from the gas stream, due to Brownian motion, interception, and impaction, as depicted in the following schematic.
CEE 446
Formation of Filter Cake on Fabric Filter

particles ug
_ _
filter cake
ug particles
filter
filter filter with filter cake

379 CEE 446
clean fabric
CEE 446
Such enhanced filtration of the particulate material is useful for achieving high p particulate collection efficiencies, but , pressure drop across filter cake can be prohibitive. Such increase in pressure drop is reduced by providing cleaning of the filter cake from the filter. Mechanisms used to clean the filter cake from the filter includes shaker, reverse air, pulse jet, and replacement of cartridges.
95
There are numerous models that describe particulate collection efficiency achieved by filtration. These models have a difficult time realistically describing the contribution of the filter cake to particulate collection efficiency. Seeing that collection efficiencies for filters are typically very high (> 99%) we will not discuss the theory of particle removal from gas streams in this course.
Although accurate particulate collection efficiency is difficult to predict, the pressure drop across the fabric and filter cake is possible. Darcy's Law is used to relate pressure drop to mass concentration of particulate material in the incoming gas stream, stream superficial gas velocity viscosity of velocity, the gas stream, and filtration coefficients.
382
where, Pressure drop (P) across a filtration unit is described by summing the individual pressure drop components that exist in series as the gas stream flows through the filtration unit.
P 'T =
PT = pressure drop for the filtration unit = -(P2-P1) Penc = P due to the enclosure of the filtration unit Pfilter cake = P due to the filter cake Pfilter = P due to the filter
P '
i
P 'T = P 'enc + P 'filter cake + P 'filter
CEE 446
383
CEE 446
384
96
Typically, Penc << Pfilter cake + Pfilter Therefore, PT Pfilter cake + Pfilter P due to the filter cake and filter can be described individually using Darcy's Law.
Evaluation of Pfilter cake

xp Acsa _ ug mp,v filter
particles filter cake
xf
CEE 446
385
CEE 446
386
P 'filter cake =
where,
g xp ug Kp
Qg = volumetric flow rate of gas stream ACSA = net cross-sectional area of the filter Kp = filter cake permeability factor Such expression is helpful but it is necessary to determine xp and Kp. xp can be determined by using a material balance for y g the particulate material that is captured with particulate collection efficiency = 100%.
g = dynamic viscosity of gas xp = thickness of filter cake _ ug = superficial gas velocity =

Qg A CSA
CEE 446
387
CEE 446
388
97
mass of particulate material removed from gas stream
mass of particulate material collected by filter
where, mp,v = mass concentration of particulate material in the inlet gas stream t = amount of time that filtration has occurred c = mass density of filter cake
(mp,v ug )(A CSA )(t)( ) = (c )(A CSA )(xp ) =1 0 1.0
CEE 446
389
CEE 446
390
Therefore,
xp = mp, v ug t c
let
Rp = filter cake resistance factor [sec -1] =
g ug mp,v ug t p Kp c
g K p c
and
P'filter cake = P
then
2 Pfilter cake = R pmp, v ug t
CEE 446
391
CEE 446
392
98
Evaluation of Pfilter
where, g = dynamic gas viscosity xf = thickness of filter _ ug = superficial gas velocity Kf = filter permeability factor
P 'filter =
g x f ug Kf
CEE 446
393
CEE 446
394
let
kg R f = filter resistance factor 2 m sec gx f = Kf
Such expression can be further simplified by taking advantage of the definition of filter drag, S,
Si =
P'T ug
then
P ' filter = R f ug
and the definition of areal dust loading, Wp:

Wp = mass of particles in filter cake A CSA
and
2 P'T = Rp mp,v ug t + R f ug
= mp, v ug t
CEE 446
99
Therefore,
S=
PT Rp Wp ug R f ug = + ug ug ug
Filter Drag versus Areal Loading of Particulate Material
Se
S = Rp Wp + R f
This resulting equation is in the linear form of y = mx + b with y = S, m = Rp, x = Wp, and b = Rf.
S Wp
non-linear due to initial formation of filter cake
Sr
Wp
By letting Rf = Se, the effective residual filter drag, then: S = RpWp + Se
Sometimes, Rp is referred to as K2 in the S literature and it describes the ratio of Wp in the linear regime of S = f(Wp).
Values for parameters related to fabric filtration are available in the literature. These values provide an order of magnitude estimate of what to expect for a particular application. If values for the necessary parameters are not available in the literature, then a pilot-plant study can provide the values of the necessary _ parameters (Rp = K2, mp,v, ug, Se) required for characterizing a larger-scale filtration unit.
CEE 446
100
Characteristics of Filtration
Advantages Disadvantages High particulate collection High pressure drop efficiency over a wide range compared to ESP of particle diameter Wide choices of filters depending on application Simultaneous removal of particles and gases Particulate removal efficiency does not depend on d td d composition of dry particles Lower capital cost than an ESP Higher annual cost than an ESP Static charge can result in an explosion Space requirements can be large Will not work well with particles that adhere to fabric (fabric blinding)
401 CEE 446
Typical Values for Fabric Filtration Units

0 .1 Qg am3 min am3 min 10 A CSA m2 m2
2 cm H2O( l ) P 25 cm H2O( l )
0o C T 800o C
CEE 446
10 cm
diameter of filter tube
Example (Fabric Filter)
30 cm
length of fabric 10 m th f f b i 0.5 m l tube 1 number of 20 compartments
0.1
Nmin Nmin < Rp = K 2 < 100 gm gm
A fabric filter will be used downstream of a spray dryer to remove particulate material from a gas stream. Actual concentration of particulate matter is 10 g m-3 upstream of the fabric filter. A pilot plant was built and operated downstream of the spray dryer to characterize the behavior of the filter cake and fabric under actual conditions. Pilot plant data are presented below.
CEE 446
101
Filter Drag (S) [cm H2O(l) cm-1 sec] 8.0 13.2 15.3 16.8 18.3 19.9 21.4 22.9 24.5
Areal Dust Loading (Wp) [g m-2] 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
a. Determine the value of the effective residual filter drag (Se) with units of cm H2O(l) cm-1 sec sec. See figure on next page.
Se = 12.2 cm H2O(l) cm-1 sec
CEE 446
405
CEE 446
406
Filter Drag versus Areal Loading of Particulate Material

S [c H2O(l) / (cm cm m/sec)]
Se Wp
b. Determine the value of the filter cake resistance factor (Rp = K2) with units of (( ((cm H2O(l))/(cm/sec))/(g/m2). )( )) (g )
Rp = 24.5 15.3 S = Wp 4 . 0 1 .0 3.07 cm H2O( l ) = (cm / sec) ( g / m2 )
Sr
Wp [g / m2]
102
Summary for Particle Removal Devices

Gravitational settling chambers, centrifugal cyclones, wet scrubbers, electrostatic y , , precipitators, and fabric filters are devices that we discussed in order of increasing overall particle collection efficiency. Typically, as particle collection efficiency increases, so does the cost and complexity of the air quality control devices. Each device has its appropriate application. Care must be taken when considering composition of the gas stream and particles, temperature, pressure, volumetric gas flow rate, variability of the gas stream and particles, particle size distribution, capital cost, and operating cost.
CEE 446
409
CEE 446
410
Separation/Transformation and Removal of Gaseous Contaminants from Gas Streams with Ancillary Air Quality Control Devices
The separation/transformation and removal of gases from gas streams is similar to the separation and removal of particles from gas streams. A "driving" force is required for driving gaseous contaminants just like particulate contaminants. Now the driving force is in the
form of a concentration gradient. A "sink" is provided in the air quality control device to deplete the concentration of the gaseous contaminant at a particular location, resulting in a concentration or chemical potential gradient of the gaseous contaminant. The concentration gradient causes mass transfer t ti di t t f of the contaminant from the "bulk" gas medium towards the "sink."
CEE 446
103
Ancillary air quality control devices that are used to separate/transform and remove gaseous contaminants from gas streams include: Combustion -- gaseous contaminant is oxidized to other less noxious compounds (e.g., thermal oxidation of CO to CO2). Adsorption -- gaseous contaminant reacts physically/chemically with a solid medium (e.g., activated carbon and benzene in air).
Absorption -- gaseous contaminant reacts physically/chemically with a liquid medium (e.g., wet scrubber to remove SO2 from flue g gas with Ca(OH)2 slurry). ( ) y) Condensation -- gaseous contaminant exhibits a phase change from a gas to a liquid by decreasing temperature resulting in the reduction of the saturation partial pressure of the contaminant below the contaminant's actual partial pressure (e.g., mechanical or cryogenic refrigeration of benzene in air).
Catalysis -- conversion of gaseous contaminant to a less noxious compound with a catalyst (e.g., nickel oxide catalyst on alumina to chemically reduce NO to N2 and H2O). Selective Non-Catalytic Reduction -- mix gaseous contaminant with a reactant to chemically transform gaseous contaminant t h i ll t f t i t to a less noxious material (e.g., mix NH3 with NO in flue gas stream at ~1000C to chemically reduce NO to N2 and H2O).
Biological Transformation -- expose gaseous contaminant to biological film resulting in conversion of gaseous f contaminant to a less noxious material (e.g., C6H6 in humidified air is passed through a bed containing biofilm resulting in the formation of CO2, H2O and biomass). biomass)
416
104
Molecular Sieve -- solid material, such as aluminosilicate, can be prepared with well defined pore sizes ranging from a few nm p g g to 100's of nm that allow selective penetration of molecules up to a maximum molecular size into the sieve. Once the sieve is saturated, it can typically be regenerated allowing for capture and reuse of the materials that were trapped in the sieve.
Instead of discussing all of the methods g described above, there will be more in depth discussions of combustion, adsorption and absorption.
418
Combustion
Terminology
Fuel: Solid, liquid, or gas containing primarily C--C and C--H bonds that are broken during combustion, converting chemical to thermal energy. Oxidant: Oxidizing agent for combustion reaction (O2). Diluent: Inert carrier gases (N2 and excess O2) that do not oxidize in the combustion reaction.
Combustion Stoichiometry of Fuels Containing only C & H

(combustion in air)
CxHy + -1(x + y/4)O2 + 3.76 -1(x + y/4)N2 xCO2 + (y/2)H2O + 3.76 -1(x + y/4)N2 + [(-1 - 1)(x + y/4)]O2
valid for 1.0 where = equivalence ratio (mfuel / mair )actual = (mfuel / mair )stoichiometric
1 Excess air = 1 (100% )
Fuel + Oxidant + Diluent Combustion Products + Diluent

420
105
Example Problem: Combustion of Ethane

What are the concentrations (% by volume) of CO2, H2O, N2, and O2 in the flue gas stream of a burner fueled by pure ethane (C2H6) at an equivalence ratio of = 0.8? Assume N2 is conserved (i.e., no chemical reaction with O2)
Solution:
1) Determine the stoichiometry of reaction where: x = 2 (carbon) y = 6 (hydrogen) = 0.8 (eq. ratio)
C2H6 + 4.375 O2 + 16.45 N2 2 CO2 + 3 H2O + 16.45 N2 + 0.875 O2
CEE 446
421
CEE 446
422
2) Determine the volume fraction of each gaseous end product: tot. mol. formed = 2 + 3 + 16.45 + 0.875 = 22.325 moles Thus, Thus mole fraction of each component as component, defined in the stoichiometric equation is:
C(CO2) = 2/22.325 C(H2O) = 3/22 325 3/22.325
= 0.090 = 9.0% = 0 134 = 13 4% 0.134 13.4%
C(N2) = 16.45/22.325 = 0.737 = 73.7% C(O2) = 0.875/22.325 = 0.039 = 3.9% Note: S N t Sum of the above concentrations is 100% f th b t ti i
CEE 446
423
CEE 446
424
106
What if = 0.5? (i.e., if MORE air is added) Total moles of product = 34.82 C(CO2) = 5.7% [9.0% at = 0.8] 5 7% [9 0% 0 8] C(H2O) = 8.6% [13.4%] C(N2) = 75.6% [73.7%] C(O2) = 10.1% [3.9%] That is why air quality regulations are standardized to dry 7% O2 or dry 12% CO2, to account for dilution air.
Combustion Stoichiometry of Fuels Containing C, H, S, and H2O

1 1 CxHySz + wH2O + -1 O2 + 3 76 -1 N2 H 3.76
xCO2 + (y/2 + w)H2O + zSO2 + 3.76 -1 N2 + [(-1 - 1) ]O2 where = [x + (y/4) + z]
426
Combustion Stoichiometry of Fuels Containing C, H, S, O, N, and H2O

The same assumption apply except that O reduces th oxygen d d the demand f th combustion d for the b ti air and the N forms N2. CxHySzOrNp + wH2O + -1 O2 + 3.76 -1 N2 xCO2 + (y/2 + w)H2O + zSO2 + (3.76 -1 + P/2)N2 + [(-1 - 1)]O2 where = [x + (y/4) + z (r/2)]
Example: Finding Equivalent Molecular Formula of Fuel, Given Ultimate Analysis

Reworked WW&D Ex. 6-7: A bituminous coal is being burned at 23% excess air ( = 0.813) in a coal-fired boiler. The ultimate analysis of the coal showed the following percentages by mass: C = 63.5%, H = 4.07%, S = 1.53%, O = 7.46%, N = 1.28%, H2O = 15.00%, ash = 7.16%. Determine the concentrations of CO2, H2O, N2, O2, and SO2 in the gases exhausted from the combustor.
107
element C H S O N H2O ash
% by mass 63.5 4.07 1.53 7.46 1.28 1 28 15.00 7.16 sum: 100%
MW [g/g-mole] 12 1 32 16 14 18
g-mol element per 100 g fuel 5.292 4.070 0.048 0.466 0.091 0 091 0.833
g-mol element per g-mol C 1.000 0.769 0.009 0.088 0.017 0 017 0.157
The concentrations of CO2, H2O, N2, O2, and SO2 can be determined by the overall stoichiometric combustion reaction, where: x = 1 0 (carbon) 1.0 y = 0.769 (hydrogen) = 1.20125
Result:
z = 0 009 (sulfur) 0.009 w = 0.157 (moisture) = 0.813
Thus, our equivalent molecular formula for this fuel is: C1.000H0.769S0.009O0.088N0.017 + 0.157 H2O + 7.16 g ash per 100 g fuel
Total moles product = 7.38 mole / g-mol C C(CO2) = 13.9% C(H2O) = 7.55% C(N2) = 74.7% C(O2) = 3.71% C(SO2) = 0.13% Sum = 100% (check!)
Adsorption
Adsorption consists of a system containing select adsorbents and adsorbates. The adsorbent is the solid adsorbing medium that provides sites for selective removal of gaseous contaminant from the gas phase (e.g., silica gel, activated carbon, and zeolites (crystalline metallic compounds)). ( y p )) The adsorbate is the gaseous material that can be selectively removed from the gas stream by the adsorbent.
The two mechanisms for adsorption are physical and chemical adsorption. Physical Adsorption -- gaseous molecules attach themselves to the surface of the adsorbent with an intermolecular attractive force (van der Waals force). Such reaction is reversible and exothermic with 2 to 20 kJ of energy released per mole of adsorbed adsorbate (e.g., H2O(g) onto silica gel and C3H8 onto activated carbon).
108
Chemical Adsorption -- gaseous molecules attach themselves to the surface of the adsorbent with valance forces. Such forces reaction is typically irreversible and exothermic with 20 to 40 kJ of energy released per mole adsorbed adsorbate ( g, (e.g., conversion of SO2 to SO3 on activated alumina).
Quantitative Description of the Adsorption Zone for a Fixed Bed

It is useful to describe how quickly an adsorption zone moves through the length of an adsorption vessel and the thickness of the adsorption zone. Such characterization allows prediction of how long an adsorbent can b used b f l d b t be d before it needs to be regenerated.
CEE 446
433
CEE 446
434
Schematic of an Adsorption Zone Moving through Fixed Bed of Adsorbent

Xsat
m a,o/Vg,T adsorption zone adsorption bed
where,
ma = mass of adsorbate in bulk gas Vg,T phase per unit volume of gas
X
0
ma,o / Vg,T
m a,o/Vg,T 0 m a,o/Vg,T 0
Xsat X Xsat
clean adsorbent
X = mass ratio of adsorbed adsorbate to adsorbent Xsat = saturated mass ratio of adsorbed adsorbate to adsorbent at ma ma,o = Vg,T Vg,T
X
x saturated adsorbent
CEE 446
109
It is now useful to determine the velocity of the adsorption zone (uad). Use a material balance around a control volume with the adsorbent flowing through a stationary adsorption zone and 100% of the adsorbate removed from the gas stream in the adsorption zone.
mass feed rate of adsorbate in carrier gas
mass deposition rate of adsorbate onto adsorbent
m & mg,T a,o Vg,T =m X & ad sat g,T
CEE 446
437
CEE 446
438
where,
& m g,T = total mass feed rate of gas stream t
ma,o = mass of adsorbate (component a) Vg,T in incoming gas stream per unit volume of gas g,T = density of total gas
& mad = mass feed rate of adsorbent passing through the stationary adsorption zone
Xsat = saturated mass ratio of adsorbed adsorbate to adsorbent at ma ma,o = Vg,T Vg,T
CEE 446
439
CEE 446
440
110
but
& mad = ad A CSAuad
ACSA = cross-sectional area of the adsorption bed uad = velocity of the adsorbent = velocity of the adsorption zone Therefore,
m & m g,T a,o V g,T = A ad CSA uad X sat g,T
where, h ad = apparent density of the adsorbent = true(1-) true = density of adsorbent with no void spaces = void fraction of adsorbent
The Freundlich equation can be used to ma relate to X at equilibrium conditions. Vg,T

ma V = X g, gT
*
Where and are empirical constants determined experimentally at equilibrium conditions. For the saturated portion of the adsorption bed:
ma,o = Xsat Vg,T
CEE 446
443
CEE 446
444
111
Rearranging the Freundlich equation at
then using the equation from the material balance and the Freundlich equation:
m a ,o ma = V g,T V g,T m = a ,o V g,T /

1
m & mg,T a,o Vg,T g,T
X sat
= A u ma,o ad CSA ad Vg,T
CEE 446
445
CEE 446
446
Solving for uad,

m & m g,T a,o V g,T = g,T ad A CSA ma,o Vg,T
1
If the adsorption zone thickness, , is << the length of the adsorption zone, then the amount of time that will pass before breakthrough is described by: g y
tB = L uad
uad
uad =
& m g,T g,T ad A CSA
( )1
ma,o Vg,T
( 1)
But if is not << L, then, , ,

tB =
447 CEE 446
L uad
448
CEE 446
112
It is now necessary to determine if is significant when compared to L.
Mass Balance for Adsorbate in Adsorption Zone

adsorption zone saturated adsorbent clean adsorbent
ACSA
A material balance around the adsorption zone can be used to determine . Such derivation is valid for a dilute gas stream with concentration of the adsorbate in the inlet gas stream < a few percent by volume.
ma,o / Vg,T ma / V g,T
ug
uad
Xsat
ma,1 x1
x
ma,2 X
0 0
x2
L x
450
x0
& & ma,2 ma,1 =
& m g,T ma,2 ma,1 g,T Vg,T Vg,T
as then
& & & ma,2 ma,1 d(ma ) m ma,2 ma,1 d a V Vg,T Vg,T g,T & d(ma ) = & m g,T ma d g,T Vg,T
Assuming gas phase mass transfer resistance controls the rate at which the gas phase adsorbate is removed from the gas stream: mass flux of gas & & ma,2 ma,1 = A CSA phase adsorbate where, ACSA = cross-sectional area of adsorption bed
CEE 446
451
CEE 446
452
113
* K ma ma ( x x ) & & ma,2 ma,1 = A CSA 1 V 2 V g,T g,T
x2 - x1 = horizontal displacement across the adsorption zone

as x 2 x1 d( x )
where,
ma = equilibrium concentration of gas Vg,T phase adsorbate corresponding to X
*
then
& & & ma,2 ma,1 d(ma )
K = mass t transfer coefficient f ffi i t
[min-1] [ i 1
* m m & d(ma ) = A CSA K a a d( x ) V V g,T g,T
CEE 446
453
CEE 446
454
Therefore,
& & m g,T ma m g,T ma = & d(ma ) = d dV g,T g,T Vg,T g,T & d(ma ) = & m g,T g,T m d a V g,T
Substituting the Freundlich equation,

*
ma V = X g,T
* ma ma d( x ) = - A CSA K Vg,T Vg,T
CEE 446
455
CEE 446
456
114
into the above equation results in the following expression.
It is still necesary to relate X to
ma to solve Vg,T the differential equation. From the previous
& d(ma ) =
& m g,T g,T
m d a V g,T
material balance for some value of X,
m = - A CSA K a X d( x ) Vg,T
& m g,T ma = ad A CSA uad X g,T Vg,T
CEE 446
457
CEE 446
458
then
Therefore,
m & m g,T a V g,T X= g,T ad A CSA uad X=
ma V g,T m ( )1 a,o Vg,T
(1)
but
& m g,T
g,T ad A CSA
1 m a,o ( 1)
uad =
( ) Vg,T
m 1 m = a ( ) a,o V V g, gT g,T m X = a Vg,T

(1 )
m a,o V g,T
(1 )
CEE 446
459
CEE 446
460
115
Substituting the expression for X into the expression for d(ma),
Separating variables,
x1= x0
& d(ma ) =
& m g,T
g,T
m d a V g,T =-
g,T A CSA K d ( x ) & m g,T

ma = Vg,T 1 o
ma Vg,T
1 m m m = - A CSA K a a a,o d( x ) V V V g,T g,T g,T
ma Vg,T
1 m a ma ma,o Vg,T Vg,T Vg,T
m d a V g,T
CEE 446
461
CEE 446
462
multiplying through by
1 ma,o V g,T , and letting P = 1 ma,o V g,T
ma V g,T , ma,o V g,T
reversing the limits of the integral on the right side of the equation, and integrating the left side of the equation results in, q ,
x xo =
& m g,T
g,T A CSA K
Po
d(P) P P
CEE 446
463
CEE 446
464
116
Integration of the right side of the above equation results in

& Po 1 1 P(1) mg,T ln + x xo = ln A K P ( - 1) 1 Po 1 g,T CSA
with P = P0 = 0.99 at x0 and P = P at x = x
& 0.99 1 1 P (1) m g,T ln x xo = ln + (1) A g,T CSA K P - 1 1 0.99
This equation is in the same form as Eq. 6-11 on page 311 of Air Pollution: Its Origin and Control
Letting = x - x0 with P = 0.01 and P0 = 0.99

& m g,T 0.99 1 1 0.01(1) ln = x xo = ln + ( 1) A g,T CSA K 0.01 ( 1) 1 0.99
Regeneration of the Adsorption Bed

Eventually the adsorbent will need to be regenerated or replaced with a new bed without regeneration. Regeneration of the adsorbent is preferable instead of replacement and disposal, if reasonably possible. Regeneration can occur by using elevated t l t d temperature, steam, and/or t t d/ vacuum.
and tB can be determined tB = L uad
CEE 446
467
CEE 446
468
117
As thermal regeneration occurs, the adsorption zone is replaced with a desorption zone with the desorption zone moving in the opposite direction taken by the adsorption zone. Such approach inhibits the slow migration of high molecular weight adsorbates through the bed that are difficult to desorb.
It is useful to be able to describe the amount of time required to regenerate the bed of adsorbent, adsorbent the same approach is taken as the case for adsorption except we look at the velocity of the regeneration zone, uR , and the Freundlich equation describing the equilibrium constraints during regeneration.
CEE 446
469
CEE 446
470
where, Now, Now uR =

& m g,R
g,R ad A CSA
1 R R
ma Vg,T
(R 1)
R
y p g uR = velocity of the desorption zone during regeneration mg,R = mass feed rate of gas during regeneration g,R = density of gas during regeneration
CEE 446
471
CEE 446
472
118
ad = density of adsorbent ACSA = cross-sectional area of the vessel R, R = empirical constants used with the Freundlich equation that are applicable during regeneration of the adsorbent
ma = concentration of adsorbate in gas V g,T R phase during regeneration of
then
t BR = L uR
with tBR = time required to regenerate the adsorption vessel L = length of adsorption vessel f uR = velocity of desorption zone
adsorbent
Characteristics of an Adsorption Unit

Advantages High collection efficiency of gaseous contaminant f t i t for properly designed system Allows for capture and reuse of gaseous materials Disadvantages Pressure d P drop can b become significant
Typical Values for Adsorption Units

10 m dgranules 0.5 cm m2 specific m2 2,500 g surface area g 1 pore size 1,000 1 m m ug 10 min min
Preferential adsorption of other gas phase constituents can occur System can become Lifetime of adsorbent can complicated if designed for be on the order of years ti ti continuous operation Useful to remove dilute to Pilot-plant or laboratory data concentrated adsorbates is needed to adequately from gas streams design system
10
CEE 446
475
CEE 446
476
119
Example (Adsorption)
A gas stream containing acetone and air needs to be treated before it is emitted into the atmosphere. The partial pressure of acetone in the gas stream is 5 mm Hg. The gas stream is at 100C and a total pressure of 1 atm. The total mass feed rate of the 1 gas stream is 1 5 kg sec-1. Acetone will be 1.5 removed from the gas
stream via adsorption with granular activated carbon. The values for ad, , , 3 17.3 3 K, L, K L and ACSA are 380 kg m-3, 17 3 kg m-3, 2.23, 40 sec-1, 3.5 m, and 7 m2, respectively. Determine the amount of time, in unit of hour, that the bed can remove acetone from the gas stream before the occurrence of g breakthrough.
CEE 446
477
CEE 446
478
[C H 3 C O C H 3 ] =
5 m m Hg = 6.58 x 10 3 m ole fraction 760 m m H g
= 17.3 kg / m3 = 2.23 K = 40 sec 1 L = 3. 5 m
T = 100 o C = 373 K P = 1 atm & m g,T = 1.5 kg / sec ad = 380 kg / m 3
MW
= (3 )(12) + (6 )(1) + (1)(16 ) = 58 g/g-m ole
A CSA = 7 m 2 tB =
479 CEE 446
(L ) uad
480
CEE 446
120
ma,o ppmv a P(MW )a = Vg,T RT 10 6 5 6 (1 atm ) 58 g 10 760 g mole = atm l 10 6 0.0821 (373 K ) g mole K = 0.0125 g kg = 0.0125 3 l m
481 CEE 446
uad
m & m g,T 1 a,o V g,T = g,T ad A CSA
( )
n g MW = i (MWi ) = 6.58 x10 3 58 g mole i nT g + 1 - 6.58 x 10 -3 29 g mole
= 29.2
g g mole
482
CEE 446
g,T
g (1 atm) 29.2 g mole P(MW ) = = RT atm l (373 K ) 0.0821 g mole K
uad
kg kg kg 1.5 sec 17.3 3 0.0125 3 m m = kg kg 2 0.954 3 380 3 (7 m ) m m
(1 2.23)
((2.23 1) / 2.23)
= 1 89 x 10 4 1.89
= 0.954
CEE 446
g kg = 0.954 3 l m
483 CEE 446
m sec
484
121
& mg,T P 1 1-P (-1) = ln 0 + ln K A P -1 1-P (-1) CSA 0 g, T kg 1.5 0.99 1 1-0.011.23 sec ln = + ln 1.23 0.954 kg 40 sec1 7m2 0.01 1.23 1-0.99 m3
tB =
L ( 3.5 0.046 ) m = = 1.83 x10 4 sec uad 4 m 1.89 x 10 sec
)(
tB = 5.08 hr
= 4.59 x 10- 2 m = 0.046m
CEE 446
485
CEE 446
486
Absorption
Mixing of a gas with a liquid phase can allow mass transfer of soluble gaseous g components from the bulk gas phase into the liquid phase as a result of absorption. Physical absorption occurs when the soluble gaseous specie, a, dissolves in the liquid phase (e.g., dissolving SO2 in H2O(l)).
SO2 and air H2O(l) +

12H2SO3 + HSO31 + SO32
CEE 446
487
CEE 446
488
122
Chemical reactions in the liquid phase can occur to remove the dissolved form of specie a (e.g., Ca2+ + SO32- + 2H2O(l) CaSO32H2O(s)).
SO2 and air Ca2+ + SO32- + 2H2O(l) CaSO32H2O(s)
CEE 446
489
CEE 446
490
Dominant Physical / Chemical Reactions SO2(g) SO2(aq) SO2(aq) + H2O(l) H2SO3(l) H2SO3(l) H+ + HSO32HSO3- H+ + SO32
Ca(OH)2(s) Ca2+ + 2 OHCa2+ + SO32- + 2H2O(l) CaSO32H2O(s)
Overall Chemical Reaction Ca(OH)2(s) + SO2(g) + H2O(l) CaSO32H2O(s)

C a S O
3
2 H
2
O + 2
1 O
C a S O
4
+ 2 H
2
CEE 446
491
CEE 446
492
123
Intimate mixing of the gas and liquid phases can occur by spray towers, plate/tray y p y ,p y towers, venturis, bubblers, spray dryers, and packed towers.
Spray towers atomize droplets, that typically consist of liquid water, at the top of the tower. The droplets are large enough that p g g their terminal settling velocity is larger than the gas stream's superficial velocity that is typically in the upward direction. Such configuration, with the gas flowing in the opposite direction as the droplets, is referred to as a counter-current contactor.
CEE 446
493
CEE 446
494
Counter-Current Spray Tower Absorber

outlet for clean gas stream demister spray nozzles inlet for liquid stream sorbent droplet inlet for gas stream containing gaseous contaminant
Plate/tray tower absorbers allow intimate mixing as the gas flows upward through the liquid as the liquid flows horizontally over each plate/ tray. Such configuration is referred to as a cross-current contactor.
outlet for liquid stream with dissolved gaseous contaminant
CEE 446
495
CEE 446
496
124
Plate/Tray Tower Absorber

outlet for clean gas stream demister inlet for liquid stream
inlet for gas stream containing gaseous contaminant

497 CEE 446
Venturi absorbers allow intimate contact between the gas stream and the atomized droplets as the droplets accelerate along the length of the venturi's throat. Venturis are used primarily for removal of particles from gas streams, but they can also be used to remove very soluble gases (e.g. HF) from gas streams.
CEE 446
498
Venturi Absorber and Cyclone Separator

inlet for gas stream containing gaseous contaminant outlet for clean gas stream
throat of venturi scrubber inlet for liquid sorbent b t droplets
cyclone separator
outlet for liquid containing dissolved gaseous contaminant
Bubble absorbers intimately mix the gas stream and liquid by bubbling the gas stream through a vat of liquid. University of Illinois i Illi i is operating a b bbl absorber at ti bubble b b t Abbott Power Plant to remove SO2 from the gas stream generated by the combustion of high sulfur content coal ([S] 3% by mass). Chemical absorption occurs as the SO32- in the liquid phase (typically H2O(l)) reacts with dissolved CaCO3 (Ca2+) and O2 to form CaSO4, a solid byproduct.
CEE 446
125
Bubble Absorber
outlet for clean gas stream
inlet for gas stream containing gaseous contaminant
demister
bubble outlet for liquid t ith stream with dissolved gaseous contaminant
inlet for liquid stream
Spray dryer absorbers, often referred to as wet / dry absorbers, atomize slurry droplets into the gas stream. As H2O(l) evaporates from the droplets, l bl d l t soluble gases absorb i t th b b into the droplets and react with dissolved reactants such as Ca2+ formed from Ca(OH)2 in the droplets. Eventually all of the free H2O(l) evaporates leaving dry particles consisting of CaSO3 and unreacted Ca(OH)2. These dry particles are then typically removed from the gas stream with a fabric filter or ESP.
CEE 446
Spray Dryer Absorber and Particle Removal Device

atomizer inlet for gas stream containing gaseous contaminant inlet for liquid stream sorbent droplet
outlet for clean gas stream fabric filter or electrostatic precipitator solid reaction end-products
spray y dryer hopper outlets for solid reaction end-products
Packed tower absorbers typically have an inert packing material with liquid flowing over the packing. The gas stream is then passed through the wet packing material allowing a large amount of interfacial surface area between the gas and liquid.
CEE 446
503
CEE 446
504
126
Packed Tower Absorber

outlet for clean gas stream demister spray nozzles inlet for liquid stream sorbent droplets packing material inlet for gas stream containing gaseous contaminant
Mass transfer across a gas-liquid interface is typically described with two-film theory two film (Lewis, 1916).
CEE 446
505
CEE 446
506
Two-Film Theory Applied to the GasLiquid Interface

gas-liquid interface liquid film gas film bulk liquid phase xA
where, yAG = mole fraction of component A in the bulk gas phase yAi = mole fraction of component A in the gas phase at the gas-liquid interface xAi = mole fraction of component A in the liquid phase at the gas-liquid interface
yAG yA bulk gas phase yAi G
xAi
xAL L
displacement
127
xAL = mole fraction of component A in the bulk liquid phase G = gas phase film thickness L = liquid phase film thickness
Mass transfer, using two-film theory, applies the following relationship:

mole = mass transfer driving flux A coefficient force A
CEE 446
509
CEE 446
510
For the liquid phase: NA = kL(CAi - CAL) NA = kx(xAi - xAL) CA = [mole/liter] xA = [mole fraction]
For the gas phase: NA = kG(PAG - PAi) NA = ky(yAG - yAi) PA = [ [partial pressure] i l ] yA = [mole fraction]
CEE 446
511
CEE 446
512
128
where, NA = molar flux of component A [mole / L2 t] k = interfacial mass transfer coefficient corresponding to the appropriate driving force and phase CA = liquid phase concentration of component A [mole / liter] xA = liquid phase concentration of component A [mole fraction] PA = partial pressure of component A in the gas phase [partial pressure] yA = gas phase concentration of component A [mole fraction]
CEE 446
513
CEE 446
514
These expressions describing mass transfer can be difficult to apply because of the small pp y scale of the film thickness ( ~ 0.1 mm). Therefore, overall mass transfer coefficients have also been used instead of interfacial mass transfer coefficients as described below.
For the liquid phase: NA = KL(CA* - CAL) CA = [mole/liter] NA = Kx(xA* - xAL) xA = [mole fraction]
CEE 446
515
CEE 446
516
129
For the gas phase: NA = KG(PAG - PA*) PA = [ [partial pressure] i l ] NA = Ky(yAG - yA*) yA = [mole fraction]
where, NA = molar flux of component A [mole / L2 t] K = overall mass transfer coefficient corresponding to the appropriate driving force and phase CA* = equilibrium liquid phase concentration of component A corresponding to yAG
CEE 446
517
CEE 446
518
xA* = liquid phase mole fraction of component A that is in equilibrium with yAG PA* = gas phase partial pressure of component A that is in equilibrium with xAL yA* = gas phase mole fraction of component A that is in equilibrium with xAL
The interfacial and overall mass transfer coefficients are interrelated by the following expression as described in Air Pollution: Its Origin and Control (p. 321, eq. 6.23).
1 1 1 = + H Ky ky kx
where, H = Henrys law constant

130
Instead of discussing how these general equations can be applied to all of the previously listed absorbers, we will narrow our description of absorbers t counter-current d i ti f b b to t t packed absorbers. When characterizing a packed absorption tower, tower it is of interest to describe the liquid to gas ratio (L/G), pressure drop (P), height of the packing (Z), and cross sectional area of the packing (ACSA).
Description of Inlet and Outlet Flows for a Counter-Current Absorber

Lm,2 x2 Lm s X2 m,s (2) Gm,2 y2 Gm,c Y2 control volume (1)
Lm,1 x1 Lm,s X1
Gm,1 y1 Gm,c Y1
522
where, Lm = total liquid molar flow rate [mole/time] Gm = total gaseous molar flow rate [mole/time] Lm,s = total liquid solvent molar flow rate [mole/time] Gm,c = total carrier gas molar flow rate [mole/time]
x = mole fraction of contaminant in the liqiud phase [nA/nT](l) y = mole fraction of contaminant in the gas phase [nA/nT](g) X = mole ratio of contaminant in the liqiud phase [nA/ns](l) Y = mole ratio of contaminant in the liqiud gas [nA/nc](g)
131
Applying a material balance for the control volume around the absorber.
moles of moles of
= component A component A into contol volume out of contol volume
Such expression is useful, except Gm1, Lm2, Gm2, and Lm1 can change, making it difficult to apply the relationship. Therefore, use pp y p , inert carrier gas and solvent flow rates instead of total flow rates of the liquid and gas streams. Gm,cY1 + Lm,sX2 = Gm,cY2 + Lm,sX1
Gm,1y1 + Lm,2x2 = Gm,2y2 + Lm,1x1
CEE 446
525
CEE 446
526
Rearranging and solving for Y2,

Gm,c (Y2 Y1 ) = Lm,s ( X2 X1) Lm,s (Y2 Y1 ) = Gm,c (X2 X1 )
Lm,s Y2 = X X1 + Y1 Gm,c 2
This expression is called the operating line for f a counter-current absorber. U of th t t b b Use f the operating and equilibrium lines for the system allows determination of the liquid to gas ratio.
or Y=
CEE 446
Lm,s (X X1) + Y1 Gm,c

132
Operating and Equilibrium Lines for a Counter-Current Absorber

operating line Y1 Y Y Y2 X
For a typical application Y1, Y2, and X2 are specified with X1 depending on the liquid to gas ratio. The minimum liquid to gas ratio is determined by the slope of the operating line that passes through (X2, Y2) and the point where Y1 intersects the equilibrium line. li
equilibrium line X2 X
X1
Operating Lines and Equilibrium Line for a Counter-Current Absorber

operating line corresponding to minimum liquid to gas ratio Y1 Y
Y2 equilibrium line X2
CEE 446
The absorption tower would have to operate at equilibrium condition at the inlet of the gas stream and outlet of the liquid stream to satisfy a minimum liquid to gas ratio. Such case is optimal to minimize the consumption of liquid but is extremely difficult to achieve, requiring a very tall , q g y absorption tower.
X1
133
Therefore, L m,s L = m,s 1 G G m,c design m,c min
L m,s = liquid to gas ratio that should be G m,c design achieved according to the design L m,s G = minimum liquid to gas ratio m,c min determined from the slope of the operating line that passes through (X2,Y2) and the point where Y1 intersects the equilibrium line
1 = safety factor typically ranging between 1.3 and 1.7

Pressure Drop versus Gas Flow Rate

Pressure drop can be determined given that L m,s has been quantified. The G m,c design following plot describes how pressure drop per unit height (P/Z) varies with gas flow rate (G) and liquid flow rate (L).
L3 L2 L1
flood point
log (P/ Z)
load point
dry d packing ki
log (G)
134
log
Extensive experimental research has resulted in empirical correlations describing pressure drop dependence on the liquid and gas flow rates in a counter-current packed absorber. One set of generally accepted set of results is presented in Air Pollution: Its Origin and Control as Figure g g 6-16 and Tables 6-1 and 6-2.
Correlation Describing Pressure Drop per Unit Height of Packing for a CounterCurrent Packed Absorber
flooding
(G)2F(L)0.1w gcGLL
increasing P/ Z
L G log G L
where, G = mass flux of gas stream [lbm/(ft2 sec)]
gc = gravitational force constant (32.2 lbm ft/(lbf sec2)) g = density of gas [lbm/ft3] L = density of the liquid [lbm/ft3] L = mass flux of the liquid stream [lbm(ft2 sec)]
F = packing factor (Table 6.1, p. 327, in Air Pollution: Its Origin and Control [-]) L = viscosity of the liquid [centipoise] W = density of water [lbm/ft3]
CEE 446
539
CEE 446
540
135
L G G L can be determined noting that cross sectional area of the tower drops out of L . G The value of the terms describing the
The absorber should not operate at flooding condition but some Gdesign value < Gflooding. Where, Where Gdesign = 2 Gflood where, where 2 = flooding safety factor ranging between 0.4 and 0.7
ordinate (vertical axis) can then be determined at flooding conditions. All values for the variables are know except for G. Solve for G, the gas mass flux that would occur at flooding condition (Gflood).
Now the cross-sectional area of the tower (ACSA) can be determined.
A CSA =
G G design
Such approach is valid for dilute concentrations of soluble contaminant in the gas phase (< ~1% by volume). If the absorber is treating a y ) g more concentrated gas stream, the procedure to determine ACSA should be performed at the bottom and top of the absorber and the largest value for ACSA should be used for further calculations.
CEE 446
543
CEE 446
544
136
The pressure drop per unit height of the tower can now be determined by using Gdesign and Ldesign to calculate the terms along the abscissa (horizontal axis) and ordinate. The pressure drop per unit height for the column can now be determined by reading its value from the line of constant pressure drop per unit height of packing that intersects the lines originating from the values describing the abscissa and ordinate.
Another technique to estimate pressure drop per unit height of packing is presented in the following empirical equation equation.
2 P (nL / L ) (G) 8 = 10 m 10 G Z
546
where, P = pressure drop per unit height of Z packing material [(lbf/f 2)/f ( /ft )/ft] m,n = constants available in Table 6-2, p. 328 in Air Pollution: Its Origin and Control L = mass flux of liquid [lbm/(ft2 hr)] G = mass flux of gas [lbm/(ft2 hr)] L = density of the liquid stream [lbm/ft3] G = density of the gas stream [lbm/ft3]
CEE 446
547
CEE 446
548
137
It is now necessary to determine the height of the absorber to describe the size of the absorber and the total pressure drop caused by the wet packing material. Mass transfer depends on the following expression: NA = (K)(C)
where, NA = molar flux of component A K = mass transfer coefficient C = concentration gradient of contaminant
CEE 446
549
CEE 446
550
Unfortunately, K and C vary along the length of the absorption tower. Therefore, it is necessary to consider changes in the absorber as a function of height in the absorber by using a material balance.
moles of component moles of component A transferred into = A transferred out liquid phase of gas phase per unit time per unit time
cross moles A transferred interfacial into liquid phase area sectional volume area of interfacial (time) area packed tower
change in moles A height of = in gas phase packing per unit time
CEE 446
551
CEE 446
552
138
NA a ACSA d(Z) = -d(ACSA Gm yAG) where, NA = molar flux of component A

a= interfacial surface area unit volume of packing in the absorber
d(Z) = differential packing height Gm = molar flux of gas stream yAG = mole fraction of component A in the bulk gas phase
ACSA = cross-sectional area of the packed tower

but, NA = Ky (yAG - yAG*) Solving for d(Z), d(Z)

Gm d(y AG ) K y a(1 y AG )(y AG y * ) AG
therefore,
K y a y AG y * d( Z ) = d(G y AG ) AG m = G d(y AG ) m (1 y AG )
555 CEE 446
d(Z) =
CEE 446
556
139
then, It would be useful to change the form of the expression for d(Z) by including the lnmean value for (1 - yAG).
(1 y AG ) d ( y AG ) G m Lm L d(Z) = K y a(1 y AG ) (1 y )(y y * ) AG AG AG Lm
Z Z = Z 2 d(Z)
1
(1 - y AG )Lm
(1 - y * ) - (1 - y AG ) AG = (1 - y *AG ) G In (1 - y AG )
y (1 y AG )Lm d(y AG ) G m = AG,1 K y a(1 y AG )Lm y AG,2 (1 y )(y y * ) AG AG AG
CEE 446
557
CEE 446
558
where, Z = HtOGNtOG and HtOG = overall height of a mass transfer unit based on gas phase mass transfer resistance NtOG = overall number of mass transfer units based on gas phase mass transfer resistance
HtOG is determined using correlations with the dimensionless Schmidt number (Sc). It can be shown that,
1 1 H = + Ky ky kx
140
can be transformed to
1 1 G G m m K a(1 y ) = k a(1 y ) AG Lm y AG Lm y H G L m m + L k x a(1 x AL )Lm (1 y AG )Lm m (1 x AL )Lm
and is equal to
HtOG = HtG +
( ) H Gm (HtL ) 1- x AL Lm Lm (1 - y AG )Lm
for dilute systems,

1 x AL 1 y 1 AG
CEE 446
561
CEE 446
562
therefore,
HtOG HG m ( = HtG + (HtL ) L m
where, , , = constants for a given type of packing and mass fluxes of liquid and gas (Table 6-3, p. 337) , = constants for a given type of packing and mass flux of liquid (Table 6-5, p 338) 6-5 p.338) = dynamic viscosity of liquid [lbm/(ft hr)]
and
H tG =
(G ) (L )
S c (g)
L H tL = L
CEE 446
S c (l)
563 CEE 446
141
Sc = Schmidt number =
( D)
Values for Schmidt numbers and diffusion coefficients for select species in air and water are provided in Table 6 4 p 338 of 6-4, p. Air Pollution: Its Origin and Control. To be able to calculate Z, NtOG must be de e determined. Numerous numerical a d ed u e ous u e ca and graphical techniques can be used to determine NtOG. A graphical technique will be used in this course.
where, = dynamic viscosity = densit density D = diffusion coefficient for component A
CEE 446
NtOG = y
y AG,1
AG,2
(1 y AG )Lm d ( y AG ) (1 y AG )( y AG y * ) AG
therefore,
NtOG
y AG,2 ( y
y AG,1
y AG,1
d( y AG )
AG
y* ) AG = total change in Y mean driving force
for dilute systems, 1 - yAG (1 - yAG)Lm
y AG,2 ( Y
d( YAG )
* YAG )
AG
NtOG can be determined graphically by the following schematic.
CEE 446
567
CEE 446
568
142
Determination of the Overall Number of Mass Transfer Units Based on Gas Phase Mass Transfer Resistance
operating line Y1 C F D C=D Y*AG H YAG
G Y Y2 A B X2
CEE 446
E=F E
A=B X
equilibrium line X1
569 CEE 446
For a gas-phase mass-transfer resistance controlled absorber, the number of mass transfer units can be estimated by drawing a vertically intermediate line that is equidistant between the equilibrium and operating lines (A = B and C = D). Units are then counted off from the lowest point p of the operating line (G) with horizontal distances such that the horizontal distance between the
operating and intermediate line is equal to the length of the line surpassing the intermediate line (E = F) A vertical line is F). then drawn to the operating line and another step is drawn until point H is surpassed horizontally. The number of units for this example is 1.3. Once NtOG is p determined then,
Z = HtOG NtOG
and
P P = ( Z ) Z
Therefore, Therefore (L/G)design, ACSA, Z and P are Z, specified for a particular counter-current packed bed absorber.
CEE 446
143
The overall approach to determine (L/G)design, ACSA, Z, and P is to: 1) Plot Y vs. X 2) Graphically determine (L/G)min with equilibrium and operating lines 3) (L/G)design = 1(L/G)min
4) Use (L/G)design, F and Figure 6-16 to determine Gflood 5) Gdesign = 2(Gflood) 6) ACSA = G/Gdesign for worst case condition 7) Determine P/Z from Figure 6-16 or P /Z 6 16 with the empirical equation and corresponding data in Table 6-2
CEE 446
Example (Absorption)
8) Determine HtOG with Sc(l), Sc(g), G, L, L, and appropriate constants 9) Determine NtOG graphically with the equilibrium and operating lines 10) Z = HtOG NtOG 11) P = (P/Z)(Z)
A gas stream passes counter-currently through a random packed absorption tower at a constant L/G value of 5 kg liquid per kg of gas. Ceramic Raschig rings with diameter of 2.54 cm are used as the packing material. Assume the density of the liquid, d th li id density of the gas, viscosity of it f th i it f the liquid, and 2, are constant and equal to 1 g cm-3, 1 10-3 g cm-3, 1.0 centipoise, and 0.5, respectively. Determine the
CEE 446
144
pressure drop per unit height of packing in units of cm H2O(l) per meter of packing in the absorption unit using Fig. 6-16 in Air Pollution: Its Origin and Control.
l = 1
g lb = 62.4 m 3 cm ft 3 g lb = 6.87 x 10 2 m 3 cm ft 3 P using Fig. 6 - 16 Z

1/ 2
g = 1.1 x 10 3
L 5 kg liquid = constant G kg gas
2 = 0.5, want
Ceramic Raschig Rings at d = 2.54 cm F = 155

6.87 x 10-2 L g = 5 G l 62.4
= 0.166
578
Using Figure 6-16 (Gflood )2 F (L )0.1 w = 0.10 gc G L L with
(0.10)gc G L Gflood = F(L )0.1
1/ 2
lbmft lb lb 6.87 x 10 2 m 62.4 m (0.10) 32.2 2 3 lbm sec ft ft 3 = (155 )(1 centipoise )0.1 = 2.98 x 10 1 lbm ft sec
2
1/ 2
L = w
580
145
G design = (0.5 )G flood lb = (0.5) 2.98 x 10 1 2 m ft sec = 1.49 x 10 1 lbm ft sec

2
2 0.1 (G design ) F(L ) w L gc GL
(1.49 x 10 1)2 (155 )(1)0.1 62.4 (32.2)(6.87 x 10 2 )(6.24) 62.4
= 0.025
CEE 446
581
CEE 446
582
From Figure 6-16

P = Z (0.27 in) 2.54 ( ft ) 12 in ft 2.54 cm in
cm in 1m 100 cm
ordinate 0.025 P in H2O 0.27 abscissa 0.166 Z ft packing
= 2.3
cm H2O m packing
CEE 446
583
CEE 446
584
146
Meteorology
Once particulate and gaseous contaminants are emitted into the atmosphere, they are y transported until they are transformed into other materials or removed from the atmosphere by wet or dry deposition. The mechanisms that cause the transformation and removal of contaminants in the atmosphere are very dependent on
meteorology. Therefore, we need to discuss meteorology before discussing how contaminants are transported through the atmosphere. t h The atmosphere is described by layers with the troposphere existing within the lowest 10 km of the atmosphere The stratosphere atmosphere. stratosphere, mesosphere, and thermosphere exist above
CEE 446
585
CEE 446
586
the troposphere. The troposphere (turning or changing sphere) contains approximately 80% of the atmosphere's mass and virtually all of its H2O. The stratosphere (layered sphere) is very stable and has been extensively studied because of its ability to absorb UV radiation. The layers of the atmosphere and their temperature profiles are presented below:
Layers and Vertical Temperature Profiles for U.S. Standard Atmosphere
CEE 446
587
CEE 446
SOURCE: Wallace, J. M. and Hobbs, P. V. (1977). Atmospheric Science: An Introductory Survey, Academic Press, NY, NY, p. 23.
588
147
It is important to consider how atmospheric pressure depends on height, especially for the first 10 km of the atmosphere The atmosphere. dependence of pressure on height can be described with the hydrostatic equation and ideal gas law as described below:
hydrostatic equation
589 CEE 446
where: P = total pressure g = density of gas (air) g = gravitational force constant Z = height
d(P) = g g d( Z )
CEE 446
590
but,
g = f (P, T, MW )
where,
MW = mean molecular weight of gas (air) g 29 g - mole R = 8.314 J g mole K
therefore,
P MW g = (ideal gas law) RT
T = temperatur e [K ]
CEE 446
591
CEE 446
592
148
P MW g d(P) = d(Z) RT d(P) MW g P = RT d(Z) MW g RT constant at T = T
d(P) MW g = o P RT
z
o
d(Z)
P ln Po
MW g ( Z Zo ) = RT
594
letting
(Z Zo ) P = Po exp RT / MW g
CEE 446
593
CEE 446
letting
H=
RT = scale height MW g
for
T = 10o C J 8.314 (273 + 10 K) g mole K H= g m 3 kg 29 9.8 10 g g mole sec 2 = 8,279 m = 8.3 km
then,
(Z Zo ) P = Po exp H
CEE 446
595
CEE 446
596
149
At Po = 1 atm, at Z o = 0 km, and T = 10 o C :
z P = (1 atm) exp 8.3 km
Such equation indicates that pressure decreases exponentially with height as confirmed in the following schematic for the lowest 100 km of the atmosphere.
CEE 446
597
CEE 446
598
Dependence of Air Density and Pressure on Geometric Height

It is a so important to co s de the depe de ce s also po a o consider e dependence of atmospheric temperature on height. Such consideration is important with respect to stability of the atmosphere.
CEE 446
SOURCE: Weast, R. C., et al., ed. (1985-1986). CRC Handbook of Chemistry and Physics, CRC Press, Boca Raton, FL, F-153. 2009 Mark J. Rood, University of Illinois at Urbana-Champaign. All Rights Reserved.
599
CEE 446
600
150
where, The dependence of temperature on height of the atmosphere can be described with the first law of thermodynamics and the hydrostatic equation. For a dry volume of air, the first law of thermodynamics indicates, d(q) = d(u) + d(w)
d(q) = heat transferred to the gas sample unit mass of dry gas sample change in internal energy of gas sample unit mass of dry gas sample work completed by the gas sample unit mass of dry gas sample
d(u) =
d( w ) =
CEE 446
601
CEE 446
602
1 d( w ) = P d g
where, h= therefore, therefore

d( q) = d(u) + d( w ) 1 d(P) 1 = d(h) P d + P d g g g
where, P = total pressure g = density of dry gas sample and

u =h P g
1 g
enthalpy of the gas sample unit mass of dry gas sample
d(u) = d(h) Pd( )

CEE 446
d(P) g
603 CEE 446
= d(h)
d(P) g
604
151
but, Cp = specific heat of the gas sample at constant pressure [J/ C kg]
d(h) (h) = (T ) P d( T )
therefore,
d( q) Cp d(T ) d(P) g
For an adiabatic process occurring to this volume of gas, d(q) = 0, and,

d(P) g
for the lowest 10 km of the atmosphere and d(h) Cp d(T)

Cp d( T )
606
Now substituting d(P) = -g g d(Z) from the hydrostatic equation into the above equation,
Cp d(T) gd(Z) d(Z) m sec 2 J 1004 o C kg 9.8
The dry adiabatic lapse rate (ad) is defined as

o d(T) C 5.4o F = 9.76 = d(Z) km 1,000ft
d(T) g = d(Z) Cp
ad =
9.76
CEE 446
C km
152
From experimental meteorological data,

T 217 K 288 K 7 .1 K = = Z 10 km 0 km km 7.1o C = km
Therefore, the atmospheric lapse rate is

env d(T) 7.1o C d(Z) km
But, ad > env !!!! Let's look at the wet adiabatic lapse rate, s, for a moist volume of air that is lifted adiabatically in the atmosphere.
CEE 446
609
CEE 446
610
Dew Point Temperature and Dry Adiabatic Lapse Rate Line Dependence on Height
lifting condensation level (LCL) Pressure (log scale) dry bulb temperature at constant d dew point p temperature
It is now necessary to consider the effect of condensing water vapor using the first law of thermodynamics and the hydrostatic equation.
d( q) = Cp d(T ) d(P) g
Temperature (linear scale)
CEE 446
611
CEE 446
612
153
but,
d(P) = g d ( Z ) g d( q) = Cp d( T ) + g d ( Z )
where,
L = latent heat of condensati on for water vapor 2.2 x 10 6 J kg
with
d( q) = L d ( W )
d( W ) =
change in mass of water vapor unit mass of dry gas sample
CEE 446
613
CEE 446
614
but,
L d ( W ) = Cp d( T ) + g d ( Z ) L d( W ) d( T ) g = + Cpd( Z ) d( Z ) Cp
g Cp d (T ) = L d ( W ) d (Z) 1+ Cp d ( T )
but,
g = ad Cp
d(T ) L d ( W ) g = d( Z ) Cp d ( Z ) Cp L d ( W ) d ( T ) g = Cp d (T ) d ( Z ) Cp
and
d( T ) = s d( Z )
616
154
where, s = saturated adiabatic lapse rate

d( T ) ad = d( Z) 1 + Ld( W ) Cpd(T ) C , for dry air s ad km
617 CEE 446
Another important physical meteorological consideration is potential temperature (). the temperature of an air parcel if it expands or compresses adiabatically from its initial temperature and pressure to a p essu e, pressure, P0, of 1,000 mb. o ,000 b
s =
For humid air s 4

CEE 446
618
An equation for can be derived using the first law of thermodynamics, the hydrostatic equation, and the ideal gas law.
d(q) = Cp d( T ) d(P) g d(P) g
Cp d( T ) =
d(P) g
but,
P MW RT RT d (P) P MW
g = Cp d ( T ) =
d( q) = 0 = Cp d(T )
CEE 446
619
CEE 446
620
155
Using separation of variables,

Cp MW R T
d(T) P =1,000 mb d(P) = P T P P ln = ln T P
P C MW = T P P
Cp MW R
for dry air
R = 0.288 CP MW
0.288 0 288
R P = T exp ln Cp MW P
P = T P
622
Height [k km]
Height [k km]
0.5
0.5
Height [k km] 298 303
The ability of the atmosphere to be unstable, neutral, and/or stable can have an enormous effect on air quality. Such effect can be discussed by comparing
d(T ) for actual environmental env , to ad conditions d( Z )
Environmental Temperature Profiles for Stable, Neutral, and Unstable Conditions

1 1 1
0.5
0 293 298 303 Temperature [K]
0 293 Temperature [K]
Stable
Neutral
Unstable
env < ad
env = ad
env > ad
temperature profile for an adiabatic process temperature profile for actual environment
CEE 446
623
CEE 446
624
156
The change in with respect to height can be used to describe the stability of the atmosphere as presented below and discussed in Air Pollution: Its Origin and Control.
T T = Z Z env Z ad
if
<0 Z =0 >0
unstable atmosphere neutral atmosphere stable atmosphere
CEE 446
625
CEE 446
626
Thermal Inversions Caused by Subsidence, Topography, or Frontal Passage

An extremely stable atmospheric condition d(T ) is positive. A thermal positive d( Z) d(T ) inversion occurs when is positive (env d( Z) is negative). Inversions can occur by four can occur when , p g p y, mechanisms: subsidence, topography, frontal passage, and radiation inversion.
1 Height [km] ] Height [km] ] Height [km] ]
warm air
2
warm air
1
warm air cool air
0.5
cool air
1.5
cool air
0.5
mountain
293
298
303
Temperature [K]
Subsidence
Topography
Frontal Passage
actual environmental temperature profile direction of air transport
CEE 446
627
CEE 446
628
157
Thermal Inversion Caused by Radiative Heating and Cooling

1 Height [km m]
early afternoon
Height [km m]
early morning
Height [km m]
midmorning
0.5
increasing time
0.5
increasing time
0.5
actual environmental temperature profile modification to temperature profile as air cools (night) and warms (morning) near the earths surface
The wind velocity profile is another important characteristic of the atmosphere. The wind s wind's velocity depends on height above ground, stability of the atmosphere, and roughness of the earth's surface. An empirical power law is typically used to describe the dependence of wind speed on p p height above ground and atmospheric stability.
630
Z ug,2 = ug,1 2 Z1 where, ug,2 = wind speed at height Z2
ug,1 = wind speed at height Z1, which is typically at Z1 = 10 m P = function of atmospheric stability and surface roughness, as described in Air Pollution: Its
Origin and Control, Table 3-3, p. 129
Another aspect of atmospheric stability is the maximum mixing depth (MMD). MMD describes the maximum height at which the convective mixed layer is expected to be limited. Such information provides how high the convective mixed layer may raise for a given day. Maximum MMDs are g y expected during the summer and minimum MMDs are expected during the winter.
632
158
Maximum Mixing Depths

1 Height [k km]
Example (Meteorology)
MMD
Height [k km]
1 Height [k km]
MMD
0.5
MMD
0.5
0.5
To
1 293 298
T o
303
To
1 293 298
T o
303
1 293 Temperature [K]
To
298
T o
303
Temperature [K]
Temperature [K]
MMD = maximum mixing depth = constant dry adiabatic lapse rate line = actual environmental temperature profile To = actual air temperature at ground level To = normal maximum air temperature at ground level of that month
The surface-level air temperature is 22C while the normal maximum surface temperature for that month is 30C. At an elevation of 600 m, the air temperature is 17C and varies linearly with height as confirmed with a radiosonde measurement. Determine the maximum mixing depth (MMD) in units of meters.
634
17 o C 22o C T = 0.6 km Z env 9.76 o C T = km Z ad
MMD is the height where Tenv = Tad such that g T passes through 22C (To) at Z = 0 m Z env and T passes through 30C (To) at Z = Z ad 0 m.
CEE 446
635
CEE 446
636
159
Tenv = m1 Z + b1 17 o C 22 o C o km = 0.6 km ( Z [k ]) + 22 C Tad = m 2 Z + b 2 9.76 o C o = km ( Z [km]) + 30 C

Equating Tenv = Tad and solving for Z,

Z [km] = 30 - 22 17 - 22 + 9.76 0.6
Z = 5.6 km = MMD
638
Schematic of a Wind Rose

A wind rose is another technique used to describe the atmosphere at a particular location. A wind rose provides the wind speed, wind direction, and the frequency of the wind's speed and direction.
8% 4% calm 10 %
calm
wind speed [km/hr] at Z = 10 m 15 5 15 15 30

640
160
Atmospheric Dispersion of Gaseous and Particulate Contaminants

Once a gas stream is exhausted into the atmosphere, atmosphere it is transported downwind of the source. The contaminants can be removed from the atmosphere during transport by dry deposition, wet deposition, and by transformation to another material. y Dry deposition occurs when the dispersed material is removed from the atmosphere
by some object on the earth. Wet deposition occurs when the contaminant is captured in clouds ("wash out ) or by ( wash out") precipitation that is falling through the atmosphere ("rain out") and is then deposited to the earth.
CEE 446
641
CEE 446
642
Dispersion of contaminants in the atmosphere can be treated by numerous techniques. Box, Eulerian, Lagrangian, and semi-empirical semi-theoretical statistical techniques will be discussed in this course. Emphasis will be placed on the statistical technique, seeing that it is commonly used to model the dispersion of contaminants in the atmosphere.
The box model is the simplest approach to predict the resulting concentration of a contaminant in the atmosphere that results from the emission of that contaminant into the atmosphere. This model is useful to discuss conceptually, but has limited applications due to the necessary assumptions used in the development of the model. The model considers a contaminant that is emitted into the
CEE 446
161
Box Model for Dispersion of a Contaminant in the Atmosphere

atmosphere with some source strength (Q1) and flux (F1) into a box that has a certain length (L) width (W), and h i ht (H) Th l th (L), idth (W) d height (H). The _ wind has a constant speed (ug) as it passes through the box as presented below. The resulting downwind concentration of contaminant i as predicted by the box model is defined as C2.
H ug
_
F2 C2
F1 Q1 W L
CEE 446
645
CEE 446
646
where, L W H = length of box [L] = width of box [L] = height of box [L] = LWH m = mass of contaminant [m]
ug = wind speed
L t
m Q = source strength of contaminant i T

m F = flux of contaminant i 2 L t = Q cross - sectional area
Vbox = volume of box [L3]
m C = concentration of contaminant i 3 L
162
Using a material balance on contaminant i,

d(C) 1 d(m) = d( t ) Vbox d( t ) = = = =
CEE 446
Assuming steady-state conditions,

1 Q1 d(C) =0= LW C 2 ug W H d( t ) LWH LW
1 m1 m 2 t Vbox t 1 Vbox (F1 A 1 F2 A 2 )
Rearranging and solving for C2

Q1 ug WH
1 (F1 LW F2 WH ) LWH 1 Q1 LW C 2 ug WH LWH LW

649 CEE 446
C2 =
650
Example (Box Model)
A city is located at the base of a steep canyon. The canyon is 1 km deep. The city has been accused of violating the eight-hour NAAQS for CO at a receptor site located downwind of the city. These violations occur when an inversion layer exists 500 m above the city. The ground city level receptor site is located within the canyon and is 80 km downwind of the city.
The canyon and city are 20 km wide. The city is 10 km long. A typical wind speed for 1 the canyon is 5 m sec-1. The concentration of CO upwind of the city is 8,000 g m-3 when the NAAQS for CO is violated. The city emits 375 g m-2 sec-1 of CO into the atmosphere. Estimate the concentration of p CO at the receptor site when there is an inversion layer 500 m above the city.
652
163
Box Model to Predict CO Concentration Downwind of Urban Site
Applying a material balance

d(C) 1 d(m) = d( t ) Vbox d( t )
F1 C1 F2 L2 L
ug
F3 C3
= =
1 (F1 H W + F2 L 2 W F3 HW ) LHW 1 (C1 ug HW + F2 L 2 W C3 ug H W ) LHW
CEE 446
653
CEE 446
654
Assume steady-state conditions and solve for C3.

C3 = C1 ug HW + F2 L 2 W ug HW F2 L 2 ug H
g 375 (10,000 m) 2 m sec g C3 = 8,000 3 + m m (500 m) 5 sec = 9,500 g m3
= C1 +
CEE 446
655
CEE 446
656
164
A box model is useful to get an understanding of how source strength, wind speed, and vertical cross sectional speed cross-sectional area of the box play roles in the resulting concentration of contaminant in the atmosphere. Unfortunately, the model does not consider change in wind speed g p with height, concentration gradients within the source's plume, chemical reactions,
stability of the atmosphere, deposition of the contaminant, and topography. Therefore, we need more d t il d models Th f d detailed d l to adequately describe dispersion of contaminants in the atmosphere.
CEE 446
657
CEE 446
658
Applying conservation equations with a Lagrangian frame of reference considers an air sample containing the contaminant p g relative to the moving fluid. The frame of reference moves with the sample as the sample is transported in the atmosphere. The model could consider the physical properties of the sample and the sources, sinks, and chemical reactions that influence the sample during transport.
Applying conservation equations with an Eulerian frame of reference considers an air sample containing the contaminant relative to a fi d coordinate system. Th f fixed di t t The frame of f reference would be stationary while viewing the air mass as it passes through each grid within the spatial domain. Once again, the model could consider the physical properties of the air mass and the sources, sinks, and chemical reactions that would influence the sample during transport.
165
Models using conservation equations with the Eulerian or Lagrangian frames of reference are useful to predict how contaminants are transported, transformed, and then removed from the atmosphere. However, these models require large amounts of information and computing p g capabilities to adequately execute the models.
The semi-empirical semi-theoretical statistical technique, also known as the Gaussian dispersion model is a much simpler and more readily available model when compared to the more general models using conservation equations. The Gaussian model has also been approved pp by State and Federal environmental protection agencies for environmental engineers to predict local air quality that
CEE 446
would be effected by the operation of a source of air contaminant. The Gaussian model is used to predict air quality conditions as needed for facilities to obtain certain permits to construct and operate a source or set of sources at a facility or region. y g
The Gaussian dispersion model was developed by Sutton as part of the Chemical Defense Establishment (1918). Further theoretical considerations about atmospheric turbulence was then considered by Taylor (1921). Hay and Pasquill (1957) developed Taylor's results to refine the Gaussian dispersion model for transport of contaminants in the atmosphere. Numerous field experiments
CEE 446
166
were then performed to see how contaminants were actually transported in the atmosphere and how well the p experimental results agreed with the modeled results. Further improvements in the Gaussian dispersion model then occurred with respect to transport near the source of the contaminant. Considerations of plume rise associated with the plume's momentum and buoyancy at the outlet of
the exhaust stack were considered by Briggs and others (1950, 1969). More recently Gaussian dispersion models have been modified to consider topography and chemical reactions of the contaminants that would occur during atmospheric transport.
666
Dispersion of a Contaminant from an Elevated Point Source

ug (z, stability class) z plume centerline
C( x,y,z )
1 y 2 1 z H 2 Q = exp 2ug y z 2 z 2 y
(x, y, z) x h (x, y = 0, z = 0)
where, h C(x,y,z) = concentration of contaminant i at location x, y, and z relative to source Q ug

667 CEE 446
H h ug (z, stability class) = wind velocity [m/sec]
(x, y, z = 0)
= source strength of contaminant i = wind speed at physical height of stack (h)

H = effective height of stack [m] h = physical height of stack [m] y

CEE 446
h = plume rise [m]
167
y = standard deviation of the plumes probability distribution function along the y axis, also referred to as the plumes dispersion coefficient along the y axis z = standard deviation of the plumes probability distribution function along the z axis, also referred to as the plumes dispersion coefficient along the z axis H = effective height of plumes centerline
The primary factors used to determine the downwind concentration of a contaminant is the location of the receptor relative to the source, the source strength for the contaminant, wind speed at the height of the physical stack (h), stability of the atmosphere, the effective plume height, p , p g , and topography.
670
The source strength of the contaminant can be determined by numerous techniques. For example, the emission rate of the example contaminant from the source can be determined experimentally with stack tests, using USEPA AP-42 emission factors and relevant parameters describing the p p g process and its ancillary equipment, or with the use of a material balance on the process.
Wind speed along the physical height of the stack (h) can be determined using the power law expression described in the meteorological section of this course. Such relationship requires information about atmospheric stability, surface roughness, and the height where the predicted wind speed is needed Such height should needed. correspond to h. Wind speeds are typically measured from meteorological towers at a height of 10 m.
CEE 446
168
The wind speed measured at Z = 10 m can then be used with information about atmospheric stability (Table 3 3 p 129 in 3-3, p. Air Pollution: Its Origin and Control), surface roughness, and the height of the physical stack to determine the appropriate wind speed that would be used in the p Gaussian dispersion model.
Dispersion coefficients, y and z, depend on the distance the receptor is downwind from the source type of topography and the source, topography, stability of the atmosphere. Values for y and z can be determined using PasquillGifford curves.
CEE 446
673
CEE 446
674
Dependence of Dispersion Coefficient Along y Axis (y) on Stability, Surface Type, and Distance Downwind from the Source
Dependence of Dispersion Coefficient Along z Axis (z) on Stability, Surface Type, and Distance Downwind from the Source
Surface Roughness - - - - urban ______ rural
Surface Roughness - - - - urban ______ rural
SOURCE: Wark, Warner, & Davis (1998) Air Pollution: It's Origin and Control, Addison Wesley, 3rd ed. p. 153.
169
The appropriate stability class that should be used with the figures describing the dispersion coefficients can also be determined by the following table.
Conditions for Determining the Stability Class of the Atmosphere DAY NIGHT Wind speed, u Solar Radiation Cloudinesse a (m/sec) Strongb Moderatec Slightd > 4/8 Cloud < 3/8 Cloud <2 A A - Bf B 2-3 A - Bf B C E F 3-5 B B - Cf C D E 5-6 C C - Df D D D >6 C D D D D
a b c
Surface wind speed is measured at 10 m above the ground Corresponds to clear summer day with sun higher than 60 above the horizon Corresponds to a summer day with a few broken clouds, or a clear day with sun 35-60 above the horizon d Corresponds to a fall afternoon, a cloudy summer day, or a clear summer day with the sun 15-35 above the horizon e Cloudiness is defined as the fraction of sky covered by clouds f For A-B, B-C, or C-D conditions, average the values obtained for each Note: A, very unstable; B, moderately unstable; C, slightly unstable; D, neutral; E, slightly stable; F, stable. Regardless of wind speed, class D should be assumed for overcast condition, day or night
CEE 446
677
CEE 446
678
The dispersion coefficients can also be determined analytically using coefficients that fit the equations to the experimental results. The form of the equations are lt Th f f th ti presented below with the values of the coefficients presented in the following table. The value of b is 0.894.
Empirical Constants Used to Determine Values for y and z

Stability A B C D E F a 213 156 104 68 50.5 34 c 440.8 106.6 61.0 33.2 22.8 14.35 x 1 km d 1.941 1.149 0.911 0.725 0.678 0.740 f 9.27 3.3 0 -1.7 -1.3 -0.35 c 459.7 108.2 61.0 44.5 55.4 62.6 x 1 km d 2.094 1.098 0.911 0.516 0.305 0.180 f -9.6 2.0 0 -13.0 -34.0 -48.6
SOURCE: D.O. Martin, J. Air Pollu. Control Assoc. 26, no. 2 (1976):145.
y = axb z = cx d + f
SOURCE: Wark, K. and Warner, C. F., Harper Collins Publishers, 2nd ed. (1981). Air Pollution: It's Origin and Control, p. 115.
y and z are in units of m. More detailed constants are provided in text by Wark, Warner, & Davis (1998) Air Pollution: It's Origin and Control, Addison Wesley, 3rd ed. p. 153-157 for rural and urban surfaces.
Note: X is in units of km,
170
The current form of the Gaussian dispersion model does not consider the influence of the earth s surface on the dispersion of the earth's plume. In reality the surface of the ground can influence plume dispersion by absorbing and or reflecting the plume. A p plume reflecting from the earth's surface is g described below.
Reflection of a Plume from an Elevated Point Source Considering an Imaginary Source and Linear Superposition
z actual source H h region of reflection x actual plume centerline
-h -H
(x, y = 0, z = 0)
imaginary plume centerline imaginary source
CEE 446
681
CEE 446
682
The Gaussian plume equations from the real and the imaginary sources are then summed to determine the resulting concentration of contaminant downstream of the point source.
Contribution from the actual source (Cact(x,y,z)).

1 y 2 1 z H 2 Q = exp 2 u g y z 2 z 2 y
Cact ( x,y,z )
CEE 446
683
CEE 446
684
171
Contribution from both sources Contribution from the imaginary source (Cimag(x,y,z)).
C( x,y,z ) = Cact ( x,y,z ) + Cimag ( x,y,z ) 1 y 2 1 z H 2 Q = exp 2 ug y z 2 y 2 z 1 y 2 1 z + H 2 + exp - 2 y 2 z
Cimag( x,y,z )
1 y 2 1 z + H 2 Q exp = 2 ug y z 2 y 2 z
CEE 446
685
CEE 446
686
This approach assumes that no removal of contaminant occurs at the earth's surface. Such equation should be used at z > H/2.15. q This constraint is based on the assumption that the earth's surface starts to influence the dispersion of the plume when the plumes concentration at the earth's surface is 10% of the value that exists at the centerline of the plume.
There is also interest in determining localmaximum ground-level concentrations of contaminants that are emitted from elevated point sources t compare those l t d i t to th results with NAAQS. If the predicted maximum concentration is less than the corresponding NAAQS, then calculations do not need to be done over the entire spatial domain. Graphical and analytical methods are available to
CEE 446
687
CEE 446
688
172
determine local-maximum ground-level concentrations of contaminants that are emitted from elevated point sources. The receptor exists at ground level along the centerline of the plume (y = z = 0). As presented in the following figure, the graphical technique requires information about atmospheric stability class, effective stack height (H), wind speed (ug), and source strength (Q). Given that information, the distance downwind (xmax)
where the local-maximum concentration occurs and the parameter (Cug/Q)max can be determined from the vertical and horizontal axes, respectively. The localmaximum ground-level concentration of the contaminant can then be determined knowing ug and Q. This local-maximum concentration can also be determined analytically using the following equation and table of constants for that equation.
Local-Maximum Ground-Level Concentration and its Location
C ug = exp a + b (In (H)) + c (In (H))2 + d (In (H))3 Q max
173
Values of Constants Used to Calculate

where, g H = effective stack height [m] a,b,c,d = constants available in the following table C ug -2 Q = [m ] max
Stability Class A B C D E F a -1.0563 -1.8060 -1.9748 -2.5302 -1.4496 -1.0488
C ug Q max
Coefficients b c -2.7153 0.1261 -2.1912 0.0389 -1.9980 0 -1.5610 -0.0934 -2.5910 0.2181 -3.2252 0.4977
d 0 0 0 0 -0.0343 -0.0765
SOURCE: R. Ranchoux, J. Air Pollu. Control Assoc. 26, no. 11 (1976): 1089.
CEE 446
693
CEE 446
694
Another more limited method to determine local-maximum ground-level concentrations of contaminants is to use the Gaussian dispersion model when considering reflection of the plume from the ground. As before, the receptor would exist at ground level along the centerline of the p g plume (y = z = 0).
c ( x,y,z) )
1 y Q 2 y = e 2 ug y z
2 1 z H 2 1 z +H 2 z 2 z +e e
at y = z = 0
CEE 446
695
CEE 446
696
174
For moderately unstable to near neutral conditions

C( x,y =0, z =0 ) = 1 H 2 Q exp (0)( 2) exp 2 z 2 ug y z 1 H 2 Q exp 2 z ug y z y z constant =
C( x,y =0, z =0 ) =
C( x,y =0, z =0 )
1 H 2 exp = 2 z ug 2 z Q
CEE 446
697
CEE 446
698
For constant ug, Q, H, and stability class. C(x,y=0,z=0) = f(z(x))
Therefore, evaluate
d(C( x,y =0, z =0 ) ) d( z )
=0
d(C x,y =0, z =0 ) ) d( z )
2 1 H 2 2 1 H 2 H Q Q 2 z 3 e z 3 = 2 z e + 2 ug ug z z
=0
Rearranging and solving for z,
z =
H 2
CEE 446
699
CEE 446
700
175
z can now be calculated for a given effective stack height (H). xmax is then determined using the resulting z value, the specified atmospheric stability class, and the Pasquill-Gifford curve describing z versus x. Once the location of the site where the local-maximum concentration occurs has been determined, then the determined maximum concentration can be calculated with the following equation.
C( x,y =0, z =0 )
1 H 2 Q = exp ug y z 2 z
702
Not only is there interest in evaluating how the earth can influence dispersion of a plume, plume but there is also interest in how an inversion layer that exists above the plume will influence plume dispersion. The inversion layer can act as an impermeable ceiling to the p g plume causing an inflection of g the plume toward the earth as presented below.
Influence of an Inversion Layer Aloft and the Earth on The Dispersion of a Plume from an Elevated Point Source
z xL temperature profile bottom of inversion layer z
ug H
reflections
elevated point source

704
176
The concentration of the contaminant downwind of the elevated point source can be determined with li b d t i d ith linear superposition and iti d imaginary sources as presented in the following equations.
C(x,y,z) =
1 y Q exp 2 ug y z 2 y
1 z H + 2NL 2 j exp N=1 z 2
2 1 z + H 2NL 2 + exp 1 z H + 2NL + exp z z 2 2
1 z + H + 2NL 2 + exp - z 2
Turner, D.B. (1994) Workbook of Atmospheric Dispersion Estimates: An Introduction to Dispersion Modeling, 2nd Edition, Lewis Publishers, p. 2-18.
where,
N = number of reflections [-] L = height of the inversion layer base [L]
For x 2xL it is assumed that uniform vertical mixing exists. xL is the distance downwind from the elevated point source where the plume is influenced by the inversion layer. When uniform vertical mixing occurs between the earth and base of the inversion layer, the previous equation y , p q converges to the following equation.
CEE 446
707
CEE 446
708
177
C ( x 2 xL , y , 0 z L ) =
1 y 2 Q exp 2 y 2 y Lug
Determination of the Downwind Distance from Point Source Where the Base of the Inversion Layer Influences Plume Dispersion
C(x = xL, y = 0, z = L) = ( ) (0.1)C(x = xL, y = 0 z = H) 0, base of inversion layer
Such an equation (eq. 4-37 in textbook) is useful, but xL needs to be described as presented in the following figure.
centerline of plume 2.15z z L H C(x = xL, y = 0, z = H) vertical concentration profile of contaminant in plume
xL x
CEE 446
709
CEE 446
710
Considering the previous figure, L = H + 2.15 z 2 15 at C(x = xL, y = 0, z = L) = (0.1)C(x = xL, y = 0, z = H) Therefore, Therefore z =
CEE 446
L H = (0.47)(L H) 2.15
711 CEE 446
z can now be determined given the height of the inversion layer's base and the effective height of the plume. Knowing the stability class of the atmosphere below the inversion layer and z, the Pasquill-Gifford curve describing the dependence of z on distance downwind from the point source and stability class can be used to determine xL. Once xL is determined, then
712
178
the values for x that are valid (x 2xL) can be used to calculate the concentration of the contaminant downwind of the elevated point source with the equation considering concentration of the contaminant after two plume inflections from the base of the inversion layer. y
The previous equations pertaining to dispersion of contaminants in the atmosphere assume that contaminants act the same as the dispersion medium. Such case is generally true for unreactive gaseous contaminants with temperature, pressure, molecular weights, and initial momentum similar t th t of th surrounding i il to that f the di environment. As these parameters deviate between the contaminants and surrounding
CEE 446
environment, there needs to be careful consideration of their effects. We will not discuss how these fairly unusual cases could be considered, but will acknowledge that caution should be taken for those types of conditions. Another important consideration is the effect of particle size on transport of p p particulate contaminants in the atmosphere. As the particles increase in size, they behave less and less like the dispersion medium.
Transport of large particles in the atmosphere is treated by considering plume dispersion from an elevated source without reflection from the earth.
1 y 2 1 z H 2 Q exp = p 2 ug y z 2 z 2 y
C( x,y,z )
716
179
The gravitational force exerted on larger particles is considered with the particles' terminal settling velocities.
Vtp = ug = Zp t x t
where, Vtp = terminal settling velocity of particle Zp = vertical displacement of particle ertical due to Vt,p t = change in time y p ug = horizontal velocity of atmosphere at the physical stack height x = horizontal displacement of atmosphere due to ug
CEE 446
717
CEE 446
718
Equating change in times (t) and solving for Zp ,

Z p = Vtp x ug
z (H Z p ) z H z z
The vertical coordinate can now be modified in the same manner as when elevated stack height was considered.
z H Vtp x ug z
The Gaussian dispersion model for elevated point sources emitting large particles is now described by,
CEE 446
180
c ( x,y,z )
2 z H Vtp x 2 ug 1 y Q 1 p = exp 2ug y z 2 y 2 z
C( x,y,z )
2 x H Vtp ug Q 1 = exp 2 2ug y z z
If there is interest in knowing the concentration of the particles along the centerline of the plume and at the earth's surface, then y = z = 0.
Such expressions are great, but what is meant by a large particle? It is recommended in Air Pollution: Its Origin and Control that particles with cm dp > 20 m Vtp = 1.2 sec
should be treated as large particles. We will be a bit more conservative and consider large particles as those with cm dp > 10 m Vtp = 0.3 sec
There could also be interest in determining the downward flux (F(x,y,z)) of particles from the atmosphere toward the earth Flux of earth. particles can be considered by evaluating the product of the particle's mass concentration (C(x,y,z)) with a certain particle diameter and its terminal settling velocity (Vtp).
CEE 446
723
CEE 446
724
181
F( x,y,z ) = C( x,y,z ) Vtp

2 z H Vtp x 2 ug Q Vtp 1 y 1 = exp 2 ug y z 2 y 2 z
The discussions so far about plume height has referred to an effective plume height which is the sum of the physical stack height plus plume rise above the stack The physical stack. stack height relative to the earth can be readily determined; it is plume rise above the stack that is difficult to quantify on a fundamental yet realistic basis. Numerous relationships describing plume rise are presented in Air Pollution: Its Origin and Control. A summary of those relationships are presented below.
CEE 446
Plume rise when considering initial vertical momentum of the plume and its buoyancy was studied by Carson and Moses (1969) and resulted in the following relationship for plume rise above the exhaust stack.
h = 0.029 ug,s ds u g,a + 2.62 (Q h ) ug,a
1/ 2
where,
h = plume rise [m] m ug,s = vertical stack gas velocity sec ds = inside diameter of stack [m] ug,a = h i horizontal velocity of atmosphere at t l l it f t h t m the exhaust stack' s outlet sec
CEE 446
182
kJ Qh = heat emission rate the exhaust stack sec & = m gCp (Ts Ta )
Ts = temperature of exhaust gas stream at the outlet of the stack [K] Ta = temperature of the atmosphere at the outlet of th stack [K] tl t f the t k Although this expression exists, it should be noted that it does not consider atmospheric stability and the authors indicated that more research needs to be completed to adequately describe plume rise above an exhaust stack.
& m = mass flow rate of stack gas from kg exhaust stack sec
Cp = specific heat at constant pressure of kJ exhaust gas stream kg K
Research pertaining to large stacks for electrical generating facilities resulted in the following "Holland equation." The equation slightly underestimates actual plume rise. ug,s ds Ts Ta 3 1.5 + 2.68 x 10 P ds T ug,a s
h=
Where all variables and units are the same as described above, except that P represents atmospheric p p p pressure in unit of millibars. Although not specifically stated, assume P is the total atmospheric pressure at the earth's surface. When considering atmospheric stability, Carson and Moses were able to provide the following relationships.
CEE 446
731
CEE 446
732
183
h =
u g,s ds u g,a
( Q h )0 . 5 u g,a
3.47 0.35 0 35 -1.04 5.15 2.64 2 64 2.24
These equations use the same definition of variables and units as presented above. Further research pertaining to plumes emitted from coal-burning electricalgenerating facilities operated by the Tennessee Valley Authority resulted in the following relationship (Carpenter, et al., 1970 and Briggs, 1969).
Atmospheric Stability
Unstable < 0 Z Neutral 0 Z

Stable > 0 Z
CEE 446
733
CEE 446
734
114C(F)1/ 3 h = ug,a
K = potential temperatur e gradient , z m K K - 0 001 0.001 0.013 m m z F= g ug,s d2 (Ts Ta ) m 4 s 3 4Ta sec
where,
h = plume rise above exhaust stack' s outlet [m] C = 1.58 - 41.8 z
g = gravitatio nal force constant m = 9.8 sec 2

CEE 446
184
The equations presented above do not consider plume rise as a function of horizontal displacement downwind of the p exhaust stack. The following relationships have been developed by Tennessee Valley Authority to consider plume rise dependence on horizontal displacement downwind from the exhaust stack (Montgomery, et al., 1972).
h=
xF1/ 3 ug,a
2.5 0.56 xmax 3,000 m
Atmospheric Stability Neutral K K < < 0.16 0.17 100 m Z 100 m Moderately Stable K K < 0 .7 < 0.16 100 m Z 100 m Very Stable K K < 1.87 < 0 .7 100 m Z 100 m
3.75
0.49
2,800 m
13.8
0.26
--
CEE 446
737
CEE 446
738
Time Averaging Concentrations

where, x = horizontal displacement downwind of exhaust stack [m] m4 F = same function as described above 3 sec The previously described results were used to determine concentrations of contaminants averaged over 10 minute time periods. If averaging times up to 2 hr is desired then the following power law relationship is used to predict concentrations of contaminants over the longer time period.
CEE 446
739
CEE 446
740
185
t C 2 = C1 1 t2
t1 = 10 minute averaging time t2 = averaging ti i time b t between 10 minutes i t and 2 hours q = constant, ranging between 0.17 and 0.2
where, C2 = concentration of contaminant averaged over a time period up to 2 hours C1 = concentration of contaminant averaged over a 10 minute time period
CEE 446
741
CEE 446
742
Software developed by USEPA (i.e. Screen2) describes the dependence of C2 on averaging time by assuming that C1 is based on a 1 hr average, not 10 min. The 1 hr averaged concentration is then multiplied by 0.9, 0.7, 0.4 or 0.03 to 0.08 for averaging times of 3 hr, 8 hr, 24 hr, and an annual average, respectively. respectively Using this approach typically provides conservative estimates of time weighted concentrations.
If mean concentrations averaged over time periods greater than two hours are desired, then more complicated Gaussian dispersion models should be used that include temporal variability of meteorology and parameters related to source strengths. t th
CEE 446
743
CEE 446
744
186
Line Sources
Examples of a line source are a river and a road
line source y
where,
m q = source strength of line source t L number of vehicles mass of pollutant emitted / vehicle = time length of line source number of vehicles / time mass of pollutant emitted / vehicle e.g. = time distance vehicle traveled / time
C( x,z =0 ) =
ug
x
1 H 2 2q exp z ug 2 2 z
y is not considered due to sources linear emission that is treated as a constant
H = effective height of source [L]
CEE 446
745
CEE 446
746
line source y
45 90
Instantaneous or Puff Dispersion

ex: explosion of railroad tank car, spill of toxic organic compound from truck time of release is << time of transport
ug
x
C(x z = 0) = (x,z
2q z ug
1 H exp 2 z 2 sin
C( x,y,z ) =
Qp
(2)3 / 2 x y z
1 y y 0 2 y e
1 x x 0 2 x e
2 1 z z 0 2 1 z + z 0 2 z + e 2 z e
the last term includes reflection from the ground

187
where, C(x,y,z) = concentration at (x,y,z) Qp = mass of contaminant emitted (x0,y0,z0) = location of center of puff at time of interest to determine C(x,y,z) x, y, z = dispersion coefficients, assume coefficients x = y (table 4.7 of textbook, p 175)
for a ground-level source such as a puff traveling along x-axis, (i.e., for wind traveling i x di ti only) t li in direction l )
( x 0 , y 0 , z 0 ) = ( ug t, 0, 0)
for ground-level receptor, immediately down wind of source f y=0 z=0
CEE 446
750
ug
CEE 446 Air Quality Engineering

x t = t1 t = t2
t=0
Lecture: Detailed Summary of Class Lectures

Mark J. Rood
Department of Civil and Environmental Engineering University of Illinois at Urbana-Champaign
2009 University of Illinois Board of Trustees. All Rights Reserved.
751
C( x,y =0,z =0 ) =
1 x ug t 2 exp (2)3 / 2 x y z 2 x 2Qp
for multiple puff sources, use linear superposition

188
Detailed Review Topics for CEE 446

Definition of an Aerosol Typical Composition of Dry Air and Flue Gas Concentrations of Contaminants and the Ideal Gas Law
Volume based concentrations (e.g., ppmv) Definition of a mole Mass based concentrations (e.g., g/m3) Mean molecular weight of a gas Example problem (concentration unit conversions using the ideal gas law) Pa/PT = na/nT = Va/VT Mole fraction Mole ratio Changes in volume and gas flow rates with changes in total pressure and temperature
Standard Conditions
Temperature Pressure (760 mm Hg) [H2O] [CO2] or [O2] Example problem (actual to standard conditions)
Evaluation of Closure between Two Sets of Independent Results

Average Absolute error
Chemical Classification of Particulate Matter and Gaseous Contaminants

Sulfur, nitrogen, organic material, inorganic material, elemental carbon, photochemical oxidants, radionuclides
CEE 446
Physical Classification of Particulate Material

Phase Size
Gaussian/Normal distribution Log-normal distribution Log-probability graph
Typical Concentrations of Trace Contaminants in Lower Troposphere Tri-Modal Particle Size Distribution (plus ultra-fine particle mode) Particle Size Distributions
Number distribution Cumulative number distribution Normalized cumulative number distribution Characteristic functions (Moments about a distribution) Aerosol particle size averages and widths of distributions Mean, median, and mode Variance and standard deviation
Terminal Settling Velocity of Particles

Force balance and Newton's second law Newton s Drag coefficient Reynolds number (L, V, , ) Stokes, Transition, and Newton regions Cunningham correction factor Knudsen number Mean free path of the gas molecules Mean speed of g molecules p gas
CEE 446
756
189
Scattering and Absorption of Light by Particles and Gases

(Visibility Degradation, Opacity, and Climate Change)
Visual range Meteorological range Light scattering by particles (sp) and gases (sg) Light absorption by particles (ap) and gases (ag) Air light Beer-Lambert law Light scattering and absorption by particles and gases (Mie-Lorenz Theory) Determine sp with particle size distribution, particle composition p (refractive index), and wavelength of light Determine sg with composition of gas, total pressure, and temperature Determine ap with light absorption efficiency factor and concentration of light absorbing particulate materials (black carbon)
Determine ag with light absorption efficiency factor and concentration of light absorbing gaseous materials (e.g., NO2) Ideal gas law can be used to correct for temperature and pressure of the light scattering and adsorption coefficients Optical depth Field t di d Fi ld studies describing li ht extinction b d t ibi light ti ti budgets Opacity for stack gases
Methods Used to Determine Health Effects Caused by Air Contaminants

Epidemiological studies of human populations Clinical studies involving laboratory exposures of humans Laboratory animal exposures
Air Quality Laws and Regulations

Air Pollution Control Act of 1955, Clean Air Act of 1963, and Air Quality Act of 1967 Clean Air Act Amendments of 1970
CEE 446
USEPA, research funds at state and regional levels, National Ambient Air Quality Standards (NAAQS), New Source Performance Standards (NSPS), National Emission Standards for Hazardous Air Pollutants (NESHAPs), State Implementation Plans (SIPs), Reasonable Available Control Technology (RACT), and Best gy (BACT) ) Available Control Technology ( Clean Air Act Amendments of 1977 Revised SIPs, NAAQS area classifications, non-attainment policy, air quality monitoring requirements, fines and penalties, aircraft emissions, stricter automobile emission standards, expansion of PSD, application of BACT to new sources, and GEP Clean Air Act Amendments of 1990 Urban air quality, permits, motor vehicles, air toxics, acid rain, ozone depletion, depletion and enforcement
New Source Performance Standards (NSPS)
Municipal Solid Waste Incinerators (example)
General Methods to Control the Generation and Emissions of Air Contaminants into the Atmosphere
Modify Modif process Modify feed stream Demand side management Modify social norms Ancillary air pollution control equipment Shutdown source
Capture, Transport, and Removal of Contaminants from Gas Streams

Control velocity (ug,x) Null point Relationship between Qgh, ug,x, Ah, and x
National Ambient Air Quality Standards (NAAQS)

CO, Pb, NO2, O3, PM10/PM2.5 and SO2
CEE 446
759
CEE 446
760
190
Centrifugal and axial fans Induced versus forced draft Static regain and blast gate methods to balance duct system
Separation and Removal of Particulate Contaminants from Gas Streams

Gravitational settling chamber (gravitational force) Reynolds number for a chamber (Rechamber) Hydraulic diameter (Dh) (dp) for laminar and t rb lent flo = (VTP, Acollection, Qg) turbulent flow T based on mass of particulate matter Example problem: gravitational settling chamber Cyclone (centrifugal force) Vane axial and involute type cyclones Involute cyclone with standard proportions (dp) = (dp, dp,50(g, W, H, p, Qg, Ne) and calibration curve) T based on mass of particulate matter P = (ug, g, g, K, l, H, W, De2) Characteristics of involute cyclones Typical values for involute cyclones Example problem: standardized involute cyclone Wet Scrubber (Brownian motion, interception, and impaction) Stoke's number (Stk) = Impaction number (Ni)
Removal Efficiency for an Air Quality Control Device

Material balance to determine grade efficiency ((dp)) and overall efficiency (T) for single air quality control device Material balance to determine (dp) and T for two devices in series
Mechanisms to Remove Particulate Contaminants from Gas Streams

Brownian motion Interception Impaction Enhancement of removal mechanisms using electrostatic and/or centrifugal forces
CEE 446
Stop distance (xs) Relaxation time () Relative velocity Counter-current and cyclonic wet scrubbers Venturi scrubber and cyclone separator Pressure drop P = f(, l, ug2, Ql, Qg) drop, Liquid to gas ratio (Ql/Qg) Collection efficiency, (dp) = f(l, p, KC, dp2, f 2, P, g2) Characteristics of wet scrubbers Typical values for venturi scrubbers (superficial gas velocity, liquid to gas flow rate, particle collection efficiency, pressure drop) Example problem: venturi scrubber Electrostatic Precipitator (ESP) (electrostatically enhanced impaction) Mechanisms controlling electrostatic precipitation Tubular and flat-plate ESP configurations Two-stage and one-stage flat-plate ESPs
Discharge and collection electrodes Collection efficiency for laminar flow ESP (dp) = f(wp, Acollection, Qg) Collection efficiency for turbulent flow ESP (dp) = f(wp, Acollection, Qg) Non-ideal Non ideal conditions Re-entrainment of particles Sneakage around collection electrodes Non-uniform gas velocity Empirical corrections to Deutsche-Anderson Equation Migration velocity compared to effective or drift migration velocity Newton s Newtons second law using Fdrag and Felec Saturation charge of particle due to field charging Dielectric constant of particulate material Migration velocity of particle Characteristics of electrostatic precipitators Typical values for ESPs (voltage, electrode spacing and height, migration velocity, superficial gas velocity,
CEE 446
191
specific collection area, specific corona power, dielectric constant, pressure drop, effective migration velocity, collection efficiency) Example: ESP Filtration (Brownian motion, interception, and impaction) Types of filters yp Filter cake Darcys Law Pressure drop caused by filter cake (Pfilter cake) and filter (Pfilter) Filter cake permeability factor (Kp) Mass balance on filter cake Filter cake resistance factor (Rp = K2) Filter permeability factor (Kf) Filter resistance factor (Rf) Filter drag (S) Residual filter drag (Sr) Effective residual filter drag (Se = K1) Areal loading of particulate material (wp)
Characteristics of filtration Typical values for fabric filtration Example: fabric filtration
Summary for Particle Removal Devices

Gravitational settling chamber centrifugal separator wet chamber, separator, scrubber, ESP, fabric filtration
Separation/Transformation and Removal of Gaseous Contaminants from Gas Streams

Introduction Sink Bulk gas phase Typical Methods Used to Separation/Transformation and Removal of Gaseous Contaminants from Gas Streams Adsorption Absorption Combustion
766
Condensation Catalysis Selective non-catalytic reduction Biological transformation Molecular sieve Combustion Fuel, oxidant Fuel oxidant, carrier gas (diluent) Equivalent ratio Stoichiometry for C, H, S, O, N, and H2O containing fuels and air Equivalent molecular forms of complex fuels Ultimate analysis Example problem Adsorption Physical adsorption Chemical adsorption Adsorption zone Material balance on adsorption zone Freundlich equation Velocity of adsorption zone Breakthrough time
Thickness of adsorption zone Regeneration of adsorption bed Characteristics of an adsorption unit Typical values for adsorption (diameter of granules, specific surface area pore size, superficial gas velocity) area, size Example: adsorption Absorption Physical absorption Chemical absorption Contactors Spray tower Plate/tray tower Venturi scrubber Bubbler Spray dryer Packed towers Two-Film theory Molar flux
192
Interfacial mass transfer coefficients Overall mass transfer coefficients Counter-current packed tower Material balance Operating line Equilibrium line Minimum liquid to gas ratio Pressure drop per unit height of packing material Cross-sectional area of packing material Height of packing material Example problem: absorption
Meteorology
Dependence of temperature on height Dependence of pressure on height Scale height Dependence of air density on height Environmental lapse rate Dry adiabatic lapse rate Lifting condensation level Saturated adiabatic lapse rate Potential temperature Stability of the atmosphere Formation of thermal inversion layers Subsidence Topography Frontal passage Radiative heating and cooling Wind velocity profile Maximum mixing depth
Summary for Separation/Transformation and Removal S f S ti /T f ti dR l of Gaseous Contaminants from Gas Streams
Combustion Adsorption Absorption
Wind rose Example problem: maximum mixing depth
Atmospheric Dispersion of Gaseous and Particulate Contaminants

E i i Emission, t transport, t t transformation, and removal of f ti d l f contaminants Dispersion models Box Model Example problem: box model Models based on conservation equations Eulerian frame of reference Lagrangian frame of reference Gaussian Models Dispersion from an elevated point source Dispersion coefficients Pasquill-Gifford plots Empirical constants Reflection of plume from ground
Local-maximum ground-level concentration Graphical technique Empirical constants d(C((x,y = 0, z = 0)))/d(z) = 0 Reflection of plume from inversion layer and ground Required effective stack height Dispersion of large particles in a plume Flux of particulate matter Plume rise dependence on momentum, buoyancy, atmospheric stability, and horizontal displacement g g Time averaging concentrations Example problem: plume dispersion from an elevated source Dispersion from a line source Dispersion from a puff source
193
Summary of Course
Background information, particle size distributions, terminal settling velocity of particles, standard conditions, classifications of contaminants, scattering and absorption of light, regulations, air quality engineering for particulate contaminants, combustion, air quality engineering for gaseous contaminants, meteorology, atmospheric dispersion
CEE 446 Air Quality Engineering Lecture: Overview Summary of Class Lectures
Mark J. Rood
Department of Civil and Environmental Engineering University of Illinois at Urbana-Champaign
2009 University of Illinois Board of Trustees. All Rights Reserved.
CEE 446
773
Overall Review Topics for CEE 446

Definition of an Aerosol Typical Composition of Dry Air and Flue Gas Concentrations of Contaminants and the Ideal Gas Law Standard Conditions Evaluation of Closure between Two Sets of Independent Results Chemical Classification of Particulate Matter and Gaseous Contaminants
Physical Classification of Particulate Matter Typical Concentrations of Trace Contaminants in Lower Troposphere Tri-Modal Particle Size Distribution (plus lt fi T i M d l P ti l Si Di t ib ti ( l ultra-fine particle mode) ti l d ) Particle Size Distributions Terminal Settling Velocity of Particles Scattering and Absorption of Light by Particles and Gases Methods Used to Determine Health Effects Caused by Air Contaminants Air Quality Laws and Regulations
CEE 446
775
CEE 446
776
194
National Ambient Air Quality Standards (NAAQS) New Source Performance Standards (NSPS) General Methods to Control the Generation and Emissions of Air Contaminants i t th At Ai C t i t into the Atmosphere h Capture, Transport, and Removal of Contaminants from Gas Streams Removal Efficiency for an Air Quality Control Device Mechanisms t R M h i to Remove P ti l t C t i Particulate Contaminants f t from G Gas Streams
Separation and Removal of Particulate Contaminants from Gas Streams (gravitational settling, cyclone, venturi scrubber, fabric filter, electrostatic precipitator) Separation/Transformation and Removal of Gaseous Contaminants from Gas Streams (Combustion, (Combustion Adsorption, Absorption) Meteorology Atmospheric Dispersion of Gaseous and Particulate Contaminants Summary of Course
CEE 446
777
CEE 446
778
195

Course Notes

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Course Notes

Uploaded by

Copyright:

Available Formats

Goal and Learning Tools for Class

2009 Mark J. Rood, University of Illinois at Urbana-Champaign. All Rights Reserved.

Units Used To Describe Concentrations And Ideal Gas Law

ppmv = parts per million by volume V = a 10 6 VT

2009 Mark J. Rood, University of Illinois at Urbana-Champaign. All Rights Reserved.

Values and Units for Ideal Gas Law Constant

= 0.0821 atm l/g - mole K

2009 Mark J. Rood, University of Illinois at Urbana-Champaign. All Rights Reserved.

ma RTT MWa PTMWa RTT

ma Va PTMWa ppmv PTMWa = = VT VT RTT 106 RTT

2009 Mark J. Rood, University of Illinois at Urbana-Champaign. All Rights Reserved.

ma ppmv PTMWa = VT 106 RTT

g (1 atm ) 64.06 g - mole 1 ppmv SO 2 = 6 atm l 10 0 08 98 0.0821 g - mole K (298 K ) = 2.62 10 6 g SO 2 10 6 g 10 3 l l g m3

( g ) MWa = (atomic weight)i

MWa = 32.06 + 2 (16.00 ) = 64.06 / mole

2009 Mark J. Rood, University of Illinois at Urbana-Champaign. All Rights Reserved.

m P(MW) = = g V RT Qg1g1 = Qg2 g2 and MW1 = MW2

P MW2 P1MW1 = Q g2 2 RT2 RT1

2009 Mark J. Rood, University of Illinois at Urbana-Champaign. All Rights Reserved.

Also, mean molecular weight (MW ) of a gas mixture is described by :

2009 Mark J. Rood, University of Illinois at Urbana-Champaign. All Rights Reserved.

2009 Mark J. Rood, University of Illinois at Urbana-Champaign. All Rights Reserved.

ppmv a PT, A MWa = 10 6 RTT, A

2009 Mark J. Rood, University of Illinois at Urbana-Champaign. All Rights Reserved.

2009 Mark J. Rood, University of Illinois at Urbana-Champaign. All Rights Reserved.

At dry standard conditions:

PT,S TT,A PT,A TS

mg 1 atm 273 + 140 K 1 = 178.7 3 m 1.01 atm 273 + 25 K 1 0.109 mg = 275.2 3 m

g 46 mg NO 2 g - mole = 422 = ( 275.2) g m3 30 g - mole

2009 Mark J. Rood, University of Illinois at Urbana-Champaign. All Rights Reserved.

2009 Mark J. Rood, University of Illinois at Urbana-Champaign. All Rights Reserved.

Evaluation of Closure between Two Sets of Independent Results

Chemical Classification of Particulate and Gaseous Contaminants

Elemental Carbon Inorganic Photochemical Oxidants Radionuclides

n/a CO2, CO, HCl O3

2009 Mark J. Rood, University of Illinois at Urbana-Champaign. All Rights Reserved.

Physical Classification of Particulate Material

2009 Mark J. Rood, University of Illinois at Urbana-Champaign. All Rights Reserved.

Typical Concentrations of Trace Contaminants in Lower Troposphere

Typical Atmospheric Particle Size Distributions

Condensed Organic Carbon or Sulfuric Acid Vapors, Clean Environment

Geological Material, Pollen, Sea Salt

0.1 1 Particle Aerodynamic Diameter (m)

(John Watson, Desert Research Institute, 2002, JAWMA, 52(6))

2009 Mark J. Rood, University of Illinois at Urbana-Champaign. All Rights Reserved.

Particle Size Distributions

Normalized Cumulative Number Distribution

2009 Mark J. Rood, University of Illinois at Urbana-Champaign. All Rights Reserved.

2009 Mark J. Rood, University of Illinois at Urbana-Champaign. All Rights Reserved.

0.01 n(d p i+1 )/N n (d n(d p ) N [-] 0 0

then dN(dp ) = n' (dp )d(dp ) and...

2009 Mark J. Rood, University of Illinois at Urbana-Champaign. All Rights Reserved.

Characteristic Functions for Particle Size Distributions*

Y(dp )n'(dp )d(dp )

Let's look at the total mass of particles in a particular sample.

mp (dp )n(dp )d(dp ) 1 3 dppn(dp )d(dp ) 6

*The above table is valid for spherical particles.

2009 Mark J. Rood, University of Illinois at Urbana-Champaign. All Rights Reserved.

Aerosol Particle Size Averages and Widths of Distributions

2009 Mark J. Rood, University of Illinois at Urbana-Champaign. All Rights Reserved.

(dp dp, mean ) = 0

2009 Mark J. Rood, University of Illinois at Urbana-Champaign. All Rights Reserved.

2009 Mark J. Rood, University of Illinois at Urbana-Champaign. All Rights Reserved.