EWCE 435

Air Pollution & Control

Lecture 5

Instructor
Maj Kazi Shamima Akter, Ph.D
Department of EWCE, MIST
 Lecture Outline
Introduction to Air Quality References
Models

• Air Quality Modeling • Fundamentals of Air

Pollution – Daniel Vallero,
• Atmospheric diffusion 5th ed.
theories
• Atmospheric Pollution:
History, Science and
Regulations – Mark Z
• Gaussian Plume Jacobson
Models
 Air Quality Modeling
Objectives:

• To predict pollutant concentrations at various locations

around the source

• To assess source impacts and design control strategies

• To predict future pollutant concentrations from sources

after implementation of regulatory programs

• To Identify source contribution to air quality problems

 Air Quality Modeling
DETERMINISTIC APPROACH
• The deterministic mathematical models calculate the
pollutant concentrations from emission inventory and
meteorological variables according to the solution of
various equations that represent the relevant physical
processes.

• Deterministic modeling is the traditional approach for the

prediction of air pollutant concentrations in urban areas.
 Air Quality Modeling
STATISTICAL APPROACH
• Statistical models calculate pollutant concentrations by
statistical methods from meteorological and emission
parameters after an appropriate statistical relationship
has been obtained empirically from measured
concentration.

• Also called regression model (Regression models

describes the relationship between predictors
(meteorological and emission parameters) and pollutant
concentrations)
 Air Quality Modeling
DETERMINISTIC APPROACH
(1) Dispersion/Diffusion Models:
 Uses mathematical formulations to characterize
atmospheric processes that disperse a pollutant emitted
by a source.

 Example: Gaussian Plume Model

(2) Photo-chemical Box Models:

 Long-range air quality models that simulate the
changes of pollutant concentrations in the atmosphere
due to the chemical and physical processes in the
atmosphere.
 Air Quality Modeling
STATISTICAL APPROACH (Empirical Model)
(1) Chemical Mass Balance Method (CMB)

(2) Positive Matrix Factorization Method (PMF)

(3) Land-Use Regression (LUR) Model

(4) Machine Learning

 Atmospheric Transport and Dispersion
• Atmospheric transport is the mechanism that moves the
pollution from a source to a receptor.

• The simplest source - receptor combination is that of an

isolated point source and an isolated receptor.

• A point source may best be visualized as a chimney or

stack emitting a pollutant into the air.

8
 Atmospheric Transport and Dispersion
• The effluent from the stack will flow directly from it to the
receptor when the wind is along the line connecting them
(Figure 3.6).

• The wind is the means by which the pollution is

transported from the source to the receptor.

• If the wind speed is greater than the speed of ejection from

the stack, the wind will stretch out the plume until the
plume speed equals wind speed.

9
 Atmospheric Transport and Dispersion
• Mixing by turbulence, stretching out of the plume and
meandering of the plume from the straight line between
source and receptor (as wind direction fluctuates from its
mean value over the time of transit between the two
points) tend to reduce the plume concentration towards
the receptor.

• The sum of all these processes is called diffusion.

• Substances diffuse according to Fick’s law of diffusion,

wherein the concentration diminishes with distance from
the source. This is known as a concentration gradient.

• Therefore, dispersion is the preferred term when

discussing atmospheric transport at a scale larger than
molecular. 10
 Atmospheric Transport and Dispersion
• If the plume carrying air pollutants is above the height
where ground-based convective and turbulent processes
will bring it down to the ground reasonably close to its
origin, it may travel for hundreds of miles at that height
before being brought to earth by these processes in a
remote community.

• This is known as long-range or long-distance transport.

11
 Dispersion Models and Equations
• Mathematical approaches have been applied to the
problem of dispersion of pollutants into the atmosphere,
although the dispersion process is quite complex and
depends upon several meteorological factors.

• The dispersion is generally three dimensional.

• Several empirical equations have been developed by

several investigations.

• These equations or models estimate the concentrations of

pollutants in the plume at any distance x, y, and z, i.e., in
horizontal down-wind direction, horizontal cross-wind
direction and vertical direction, respectively.

12
 The Gaussian Plume Model
• Most diffusion models use the Gaussian plume idea, which also
is a material balance model.

• In it, one considers a point source such as a factory smokestack

(which is not really a point but a small area that can be
satisfactorily approximated as a point) and attempts to compute
the downwind concentration resulting from this point source.

13
 The Gaussian Plume Model

• The origin of the

coordinate
system is placed
at the base of the
smokestack, with
the x axis aligned
in the downwind
direction.

• The contaminated gas stream (normally called a plume) is shown

rising from the smokestack and then leveling off to travel in the x
direction and spreading in the y and z directions as it travels.
14
 The Gaussian Plume Model
• Such plumes normally rise a considerable distance above the
smokestack because they are emitted at temperatures higher
than atmospheric and with a vertical velocity.

• For Gaussian plume calculations the plume is assumed to be

emitted from a point with coordinates (0, 0, H), where H is
called the effective stack height, which is the sum of the
physical stack height, h in Fig. 6.3) and the plume rise, ∆h (in Fig.
6.3).

• For the moment we will assume that we are dealing with a

point source located at (0,0, H) that steadily emits a non-
buoyant pollutant at emission rate, Q (normally in g/s).

15
 The Gaussian Plume Model
• Let us assume the wind blows in the x direction with velocity u
and that this velocity is independent of time, location, or
elevation.

• The problem is to compute the concentration due to this source

at any point (x, y, z) for x > 0.

16
 Gaussian Puff : Two-dimensional Spreading

…………. (1)

• Eq. (1) is the basic Gaussian plume equation.

• This name comes from the fact that the exponential terms have
the same form as the Gauss normal distribution function, which
is widely used in statistics.

17
 Gaussian Puff : Two-dimensional Spreading

…………. (1)

• Eq. (1) is the product of three terms.

• If we set y = (z - H) = 0, then the two right-most terms will be exp

0= 1, which shows that the fist term is the concentration on the
centerline of the plume.

• The values of σy and σz increase with downwind distance, so that

this centerline concentration decreases with downwind distance.

• The second term shows how the concentration decreases as we

move in the horizontal, sidewise, ±y, direction from the plume
centerline.
18
 Gaussian Puff : Two-dimensional Spreading

…………. (13)

• As the second term involves y2, it is the same for moving in the
+y or -y direction. It is always ≤ 1.0.

• The third term is similar to the second one, but it shows how the
concentration decreases as we move vertically away from the
elevation of the plume centerline (z = H). It also is symmetrical
and always ≤1.0.

19
 Gaussian Puff : Two-dimensional Spreading
Example 1:
A factory emits 20 g/s of SO2 at height H. The wind speed is 3
m/s. At a distance of 1 km downwind, the values of σy and σz
are 30 m and 20 m, respectively. What are the SO2
concentrations at the centerline of the plume, and at a point
60 meters to the side of and 20 meters below the centerline?

…………. (13)

Solution:
The centerline values are those for which y = 0 and z = H, so
both of the terms in the exponential are zero. Since exp 0 =1,
the exponential term is unity.

At the centerline
20
 Gaussian Puff : Two-dimensional Spreading
At the point away from the centerline, we must multiply the
preceeding expression by

therefore,

21
 Gaussian Puff : Two-dimensional Spreading
• To use the Gaussian plume equation one must know the
appropriate values of σy and σz.
• From Eqs. (10) and (11), we would expect them to have the form

…………. (10)
…………. (14)
…………. (11)

• It seems reasonable to assume that Ks would depend on wind

speed and on the degree of atmospheric turbulence, which is a
function of wind speed and degree of solar heating (insolation)
and perhaps some other factors.

• It is also reasonable to assume that for any given degree of

insolation the value of K will be linearly proportional to the wind
speed; i.e., Ky/u and Kz/u are constants. 22
 Gaussian Puff : Two-dimensional Spreading
…………. (14)

• Thus from Eq. (14), for any given meteorological condition, each
of the σ should be proportional to the square root of the
downwind distance.

• However, experimental evidence does not agree well with this

prediction.

• The available data have been correlated by Turner and by others

and presented in the form of plots of log σy and log σz vs. log x.

23
 Gaussian Puff : Two-dimensional Spreading
• For each atmospheric condition,
the plots would be straight lines
with slope 1/2 .

• The best correlations of the

experimental results illustrate
that on such plots the horizontal
dispersion coefficient σy forms a
family of straight lines (for
various atmospheric
conditions)

• But these lines have the slopes of

0.894 instead of 0.50 (from
predicted derivation) (Fig. 6.7).

Figure 6.7: Horizontal dispersion coefficient, σy as a function

of downwind distance from the source, x for various stability
categories. 24
 Gaussian Puff : Two-dimensional Spreading

• The vertical dispersion coefficient

σz forms a fan-shaped pattern for
various atmospheric conditions.

Figure 6.8: Vertical dispersion coefficient, σz as a function of

downwind distance from the source, x for various stability
categories. 25
 Gaussian Puff : Two-dimensional Spreading

• A - F correspond to different levels of atmospheric stability.

• On a clear, hot summer morning with low wind speed, the sun heats the ground, which in
turn heats the air near it, causing that air to rise and thus to mix pollutants well.

• The atmosphere is unstable, and the values of σy and σz will be large.

• On a cloudless winter night, the ground cools by radiation to outer space and thus cools
the air near it.

• The air forms an inversion layer, making the atmosphere stable and inhibiting the
dispersion of pollutants, so the values of σy and σz will be small. 26
 Gaussian Puff : Two-dimensional Spreading
• According to Martin:

27
 Gaussian Puff : Two-dimensional Spreading
• According to Martin:

28
 Gaussian Puff : Two-dimensional Spreading
Example 2:
Estimate the values of σy and σz at a point 0.5 km downwind
from a pollutant source on a bright summer day with a wind
speed greater than 6 m/s.

Solution:
From Table 6.1, for a bright summer day the incoming solar
radiation is "strong," so we use stability category C.

Then, using Figs. 6.7 and 6.8, we read (for x = 0.5 km) σy =
56 m and σz = 32 m.

29
 Modifications in Basic Gaussian Plume Model
Effect of Ground:

…………. (13)

• At present Eq. (13) is the best simple prediction method for the
concentration in plumes considerable distances above the
ground.

• However, we are generally most interested in concentrations at

ground level because that is where most people and property
are exposed.

• The blind application of Eq. (13) at or near ground level gives

misleadingly low results. It indicates that pollutants continue to
disperse at any value of z, even at z less than zero.
30
 Modifications in Basic Gaussian Plume Model
• The ground damps out vertical dispersion. The upward and
downward random atmospheric eddies that spread the plume in
the vertical direction cannot penetrate the ground. Thus, vertical
spreading terminates at ground level.

• To account for this phenomenon in calculations, it is commonly

assumed that any pollutants that would have carried below z =
0, if the ground were not there are "reflected" upward as if the
ground were a mirror.

• Thus, the concentration at any point is due to the plume itself

plus what is reflected upward from the ground.

31
 Modifications in Basic Gaussian Plume Model

• This method is equivalent to assuming that a mirror-image

plume below the ground transmits as much up through the
ground surface as the aboveground plume would transmit down
through the ground surface if the ground were not there.

• In this case Eq. (13) takes the form of Eq. (15) as follows:

…………. (15)
32
 Modifications in Basic Gaussian Plume Model
Example 3:
A factory emits 20 g/s of SO2 at height H. The wind speed is 3
m/s. At a distance of 1 km downwind, the values of σy and σz
are 30 m and 20 m, respectively. What are the SO2
concentrations at the centerline of the plume, and at a point
60 meters to the side of and 20 meters below the centerline
of the plume, for the cases of H= 20 m and H=30 m?

…………. (15)

33
 Modifications in Basic Gaussian Plume Model

…………. (15)
Solution:
For H = 20 m, we are computing the concentration at the ground
level itself.

From Eq. (15) we see that at z = 0, the two terms in the brackets
at the right are identical, and each is equal to the value that that
term had in Example 1.

Thus our answer is exactly twice that in the second part of

Example 1; i.e., 145 x 2 = 290 µg/m3.

This is a general result. for z = 0, Eq. (15) always gives exactly

twice the value given by Eq. (13). 34
 Modifications in Basic Gaussian Plume Model

…………. (15)
For H = 30 m

35
 Modifications in Basic Gaussian Plume Model
Ground Level Modification:

…………. (13)

…………. (15)
• Equation (15) is correct for ground level or any elevation
above it.

• For large values of z, the contribution of the (z + H)2 term

becomes negligible and the result is practically identical
with that from Eq. (13).
36
 Modifications in Basic Gaussian Plume Model
Ground Level Modification:

…………. (15)

• Most often we are interested in ground-level concentrations.

• If we substitute z = 0 into Eq. (15) and simplify, we find

…… (16)

37
 Modifications in Basic Gaussian Plume Model
Ground Level Modification:

…………. (13)

…… (16)

• Equation (16) We may consider Equation (16) as the

"ground level modification” of Eq. (13), taking reflection at
the ground surface into account.

• Although Eq. (13) is the basic Gaussian plume equation,

Eq. (16) is the single most widely used estimating equation
because it applies directly to the problem of greatest
practical interest. 38
 Modifications in Basic Gaussian Plume Model
Ground reflection and temperature inversion:

…… (17)

39
 Modifications in Basic Gaussian Plume Model
Simplification of Gaussian Plume Equations
(i) Concentration at ground level (z = 0) with no hm (with
ground reflection)
…… (16)

(ii) Concentration at ground level (z = 0), in the downwind

horizontal direction along the centerline of the plume (y =
0) with no hm (with ground reflection )

…… (18)

40
 Modifications in Basic Gaussian Plume Model
Simplification of Gaussian Plume Equations
(iii) z = 0, y = 0, no hm and emission at ground level (h = 0)
(with ground reflection)
…… (19)

41
 Modifications in Basic Gaussian Plume Model
Estimation of Peak Downwind Concentration
• The simplest way would be using a spreadsheet program
to calculate C (x,0,0) as a function of x, using the following
equation –

• And finding peak downwind concentration.

• When a computer is not readily available, peak downwind

concentration can be estimated using the following chart
and the equation –

42
 Modifications in Basic Gaussian Plume Model
Estimation of Peak Downwind Concentration

• If the stability class and effective stack height (H) are

known, then one can estimate –
 distance of peak and
 (Cu/Q) max from the chart.

• Then using the above equation, Cmax can be estimated.

43
 Modifications in Basic Gaussian Plume Model
Estimation of Peak Downwind Concentration

44
 Modifications in Basic Gaussian Plume Model
Example 4: Ground Level Concentration
A stack emitting 80 g/s of NO2 has an effective stack height of
100m. The wind speed is 4m/s at 10 m, and it is a clear summer
day with the sun nearly overhead. Estimate the ground level
NO2 concentration –
(a) directly downwind at a distance of 2 km
(b) at a point downwind where NO2 is maximum
(c) at a point located 2 km downwind and 0.1 km of cross-
downwind axis

Solution:
(a) Here, Q = 80 g/sec
H = 100m
V0 = 4.0 m/s and clear summer day, ∴ Stability class B
p = 0.15
45
Modifications in Basic Gaussian Plume Model

46
Modifications in Basic Gaussian Plume Model

47
Modifications in Basic Gaussian Plume Model

48
 Effective Stack Height
• The effective height of the stack (chimney) consists of its
actual height (h) plus the height (∆h) to which the plume
rises above the stack before levelling out as shown in the
Fig.

49
 Effective Stack Height
• Plume rise is caused primarily by buoyancy and
momentum of exhaust gas and stability of atmosphere.

• Buoyancy results when exhaust gases are warmer than the

ambient and/or when the molecular weight of the exhaust
is lower than that of air

• Momentum is caused by the mass and velocity of the

gases as they leave the stack.

50
 Effective Stack Height
Plume rise/ height estimation:
•There exists several equations for calculating the plume
height (∆h).

•The USEPA recommends the following model.

51
 Effective Stack Height
Plume rise/ height estimation:
• For neutral or unstable conditions (stability class A – D)

52
 Effective Stack Height
Plume rise/ height estimation:
• For stable conditions (stability class E and F)

• The quantity of S is a stability parameter with units of s-2

and is given by –

53
 Modifications in Basic Gaussian Plume Model
Example 5: Stack Height Estimation
A power plant has a 100 m stack with inside radius of 1m.
The exhaust gases leave the stack with an exhaust velocity of
10m/s at a temperature of 220°C. Ambient temperature is 6°C.
Wind speed at effective stack height is estimated to be 5m/s,
surface wind speed is 3m/s and it is a cloudy summer day.
Estimate the effective height of this stack.

Solution:
Here, Ts = 220 + 273 = 493 K
Surface wind speed = 3m/s and cloudy summer day
∴ Stability class = C
Ta = 6 + 273 = 279 K

54
 Modifications in Basic Gaussian Plume Model
Now, F = gr2vs (1 – Ta/Ts)

55
 Modifications in Basic Gaussian Plume Model
Example 6: Stack Height Estimation
A 750 MW coal fired power plant has a 250 m stack with
inside radius of 4 m. The exit velocity of the stack gases is
estimated at 15m/s, at a temperature of 140°C (413K). Ambient
temperature is 25°C (298K) and wind at stack height is
estimated to be 5 m/s. Estimate the effective height of the
stack if –
(a) the atmosphere is stable with temperature increasing at the
rate of 2°C/km
(b) the atmosphere is slightly unstable, class C.
Solution:
Here, Ts = 140 + 273 = 413 K
Surface wind speed = 5 m/s
Ta = 25 + 273 = 298 K
56
 Modifications in Basic Gaussian Plume Model
Solution:

57
 Modifications in Basic Gaussian Plume Model
Solution:

58
 Modified Gaussian Plume Model
For Line Sources (e.g. Road):

For simplicity, consider –

(i) infinite length source at ground level
(ii) Wind blowing perpendicular to the line.

Examples of Line Sources

(i) Motor vehicles travelling along a straight section of a
highway
(ii) Agriculture burning along the edge of a field
(iii) A line of industrial sources on the bank of a river
59
 Modified Gaussian Plume Model
For Line Sources (e.g. Road):

(a) No ground reflection

(b) With ground reflection

where,
QL = source emission rate per unit length of road (g/sec-m)

60
 Modifications in Basic Gaussian Plume Model
Example 7: Ground Level Concentration of line source
Cars travelling at 55 mph speed at 75 m apart are emitting
5g/mile of CO. The wind speed is 3.5 m/s and perpendicular to
the road. Estimate ground level concentration of CO at a
distance 300m downwind. Consider atmosphere to be
adiabatic.

Solution:
As the atmosphere is adiabatic, that means the atmosphere is
neutral, hence stability class is D.
For x = 300 m, σz = 15 m

61
Modifications in Basic Gaussian Plume Model

62