Proceedings of the 7th Conference on Nuclear and Particle Physics,

11-15 Nov. 2009, Sharm El-Sheikh, Egypt

WORST AIR CONCENTRATION FROM NON-GAUSSIAN

PLUME MODEL
Khaled S.M. Essa and M. Embaby

Mathematics & Theoretical Physics Department, NRC, AEA

P.O. 13759, Cairo-Egypt

The maximum ground level air concentrations due to elevated point source diffusion over
simple terrain are estimated using herimitized advection diffusion equation for linear forms of the
eddy diffusivities which is derived before by Essa et al. (2007). Also, the critical wind speed,
critical plume height, critical downwind distance and critical vertical eddy diffusivity are also
estimated. The results are applied with data taking from first reactor, Inshas, Egypt.

INTRODUCTION

Pasquill and Smith (1983) have obtained the worst case results when the ratio of the
vertical to lateral dispersion coefficients doesn’t depend on downwind distance. Simulation
of photochemical smog in the Melbourne air shed: worse case study is studied by Hess
(1989). A simple prediction formula for maximum ground level concentrations from
coning plumes is investigated by Laless et al. (1977). The effect of the plume rise and wind
speed on extreme value of air pollutant concentration is studied by Essa et al. (2006). The
hermitized concentration from advection diffusion equation where eddy diffusivities are
depending on downwind distance is obtained by Essa et al. (2007). Modeling extreme
concentrations from a source in a turbulent flow over a rough wall is studied by Zheng et
al. (2007).
In this paper, we estimate the worst case hermitized ground level concentration at
centerline is estimated and the result is extended to a plume which is trapped between the
ground and elevated inversion layer. The critical wind speed, critical plume height, critical
downwind distance and critical vertical eddy diffusivity are also estimated. The results are
validated with data of first reactor, Inshas, Egypt.

WORST CONCENTRATION

For hermitized advection diffusion equation, the ground level concentration along the
plume centerline is given by (Essa et al. (2007)):

-343-
 1 
 1
Q  2 H 2   
C ( x, 0, 0)  1  x 2  (1)
U   x 2  

Where Q is the emission rate, U is the wind speed at stack height, H is the effective height
of the emission (hs+ ∆h); h s is the stack height and ∆h is the plume rise, and α, β and γ are
the turbulence parameters depend on the stability.
To get maximum concentration, put C / x  0 . Differentiating equation (1) with
respect to “x”, and equals the result with zero, we get that:
xmax  2 H (2)
After that one gets  2C / x 2 , and substitutes with maximum downwind distance xmax as
follows:
1
 (  3)
Q (1   ) 
 2C / x 2  at xmax  2 H (3)
U   H 4
After substituting the values in Eqn (3), we get a negative value for the second derivative;
the values of the concentration become large.
The vertical maximum eddy diffusivity Kz becomes:
K z   Ux  2 UH (4)

The maximum ground concentration along the plume centerline becomes:

Q 1 
Cmax ( x, 0, 0)  2
1     1(5)
2U   H
The plume rise is ∆h=3woD/U (Briggs, 1969), where wo is the exit velocity of the plume
and D is the inside diameter of the stack. The effective height H is written as:

H = h s+ 3woD/U = h s+ A/U (6)

where A=3woD, A is a number depends on exit velocity and diameter.

In unstable conditions, the empirical relations for turbulence parameters based on
convective velocity w* can be written as (Arya, 1999):

β = 0.31(w*/U)2; γ = 0.16(w*/U)2 (7)

where w* is the scale convective velocity. Also the intensity parameters β and γ for stable
conditions are depending on friction velocity u* (Sharan et al. 1998) as follows:

β = 3.61(u*/U)2; γ = 0.16(u*/U)2 (8)

To get the worst wind speed, taking, C / U  0 , Differentiating equation (1) with
respect to “u”, and equals the result with zero, we get that:

Uworst = 3A/8h s (9)

-344-
Then the worst effective height is given by:

Hworst = h s+A/U = h s+8Ah s/3A = 11hs/3 (10)

The worst vertical eddy diffusivity is given by:

11 2 A
K zworst  (11)
3
Substituting Uworst, Hworst, β, and γ with Eqns (9), (10) and (8) respectively in equation (5),
we get the worst ground concentration along the plume centerline in stable conditions as
follows:

0.130489QA 1 
 1
Cstbleworst ( x,0,0)  1      12)
 u*2 hs3
Finally substituting Uworst, Hworst, β, and γ with Eqns (9), (10) and (7) respectively in
equation (5), we get the worst ground concentration along the plume centerline in unstable
conditions as follows:

0.445298QA 1 
 1
Cunstbleworst ( x,0,0)  1      (13)
 w*2 hs3

WORST TRAPPED CONCENTRATION

If we consider that the plume is trapped under the stack of a point source, the ground
level concentration along the plume centerline is written as (Ragland 1976):

1 
 1 
Q  4 ( H  2 Z i ) 2    
Ci ( x,0,0)  1   (14)
U   x 2  x2 
where Zi is the height of inversion height. The maximum concentration will happen if the
plume rise is just up the inversion layer, hence, we get:
1 
  1 
Q  4 Zi 2    
Ci ( x,0,0)  1   (15)
U   x 2  x2 
Where Zi = hs+A/U, Taking Ci / x =0, to get maximum downwind distance in the form:
xmax  2 Z i (16)
Then the worst ground level concentration along centerline due to inversion layer is given
as:
Q 1 
 1
Ciworst ( x, 0,0)  1      (17)
4U   Z i2

This formula is just half the previous result estimating before (Eqn (5)), where Zi equals the
effective height H if the plume rise is just up the inversion layer. The results of the critical
wind speed and position of the height concentration are equals for non-Gaussian and
trapping formulas while the worst concentration for non-Gaussian concentration is twice

-345-
the trapping concentration. Notice that the worst concentrations depend on the height of
inversion layer and the effective height.

CASE STUDY

It is useful to apply the derived normalized concentration C/Q on the first research
reactor at Inshas. A continuous Ventilation system is provided with the reactor to the areas
where radioactive gases, volatile materials and suspended particles can exist due to either
leakage or airborne radioactivity. The total ventilation rate which could be emitted from the
reactor stack of 43 m height, 1 m internal diameter, and exist velocity 4 m/s is 39965 m3/hr
(Report 53 of Reactor Physics Department).
Table 1 shows the worst ground-level normalized concentration, wind speed,
effective height and distance are tabulated in Table 1 through atmospheric stability
conditions. We find that the wind speed, the effective height and downwind distance are
the same in stable and unstable condirions. The worst normalized concentrations have the
largest value in stable condition, then neutral condition and the minimum worst
concentration happens in unstable case.

Table 1: Worst case conditions for non Gaussian plumes.

Stability C/Q sec/m3 Uworst (m/s) Hworst (m) Xworst(m)

-6 4.45
Neutral 12.60*10 157.7 223.02
-6 3.45
Stable 15.80 *10 157.48 222.71
-6 3.45
Unstable 2.19*10 157.48 222.71

Table 2: Wind speed, friction velocity, effective height, downwind distance and the concentration
at the axis of the plume at the reactor release over emission rate during the year 1999 in neutral
classes.

U (m/s) u* (m/s) Hworst(m) xworst C/Q *10 6

(m) (sec/m3)
5.27 0.33 157.28 222.43 9.07
5.31 0.33 157.26 222.40 9.07
5.34 0.34 157.25 222.39 8.55
6.37 0.4 157.88 223.28 6.17
5.17 0.32 157.32 222.48 9.65
4.45 0.28 157.7 223.02 12.60
5.1 0.32 157.35 222.53 9.65
4.81 0.3 157.49 222.72 10.98
5.3 0.33 157.26 222.40 9.07
4.86 0.31 157.47 222.70 10.28
5.36 0.34 157.24 222.37 8.55
5.19 0.33 157.31 222.47 9.07
5.41 0.34 157.22 222.34 8.55
5.54 0.35 157.17 222.27 8.06
5.2 0.33 157.31 222.47 9.07
5.61 0.35 157.14 222.23 8.06

-346-
 5.79 0.36 157.07 222.13 7.62
6.27 0.39 157.91 223.32 6.49
5.93 0.37 157.02 222.06 7.22
6.01 0.38 157 222.03 6.84
5.41 0.34 157.22 222.34 8.55
5.75 0.36 157.09 222.16 7.62
5.26 0.33 157.28 222.43 9.07

Table 3: Wind speed, friction velocity, effective height, downwind distance and the concentration
at the axis of the plume at the reactor release over emission rate during the year 1999 in stable
classes.

u (m/s) u* (m/s) Hworst(m) xworst C /Q *106

(m) (sec/m3)
4.43 0.32 157.71 223.04 9.65
3.81 0.27 157.15 222.24 13.55
4 0.29 157.9 223.30 11.75
4.92 0.35 157.44 222.65 8.06
3.7 0.27 157.24 222.37 13.55
3.57 0.26 157.36 222.54 14.61
3.64 0.26 157.3 222.46 14.61
3.45 0.25 157.48 222.71 15.80
3.6 0.26 157.33 222.50 14.61
3.8 0.27 157.16 222.26 13.55
3.99 0.29 157.01 222.05 11.75
3.89 0.28 157.08 222.14 12.60
3.75 0.27 157.2 222.31 13.55
3.98 0.29 157.02 222.06 11.75
3.47 0.25 157.46 222.68 15.80
4.06 0.29 157.96 223.39 11.75
4.3 0.31 157.79 223.15 10.28
4.31 0.31 157.78 223.13 10.28
4.02 0.29 157.99 223.43 11.75
4.11 0.3 157.92 223.33 10.98
3.94 0.28 157.05 222.10 12.60
3.86 0.28 157.11 222.19 12.60
3.67 0.26 157.27 222.41 14.61

Table 4: Wind speed, friction velocity, effective height, downwind distance and the concentration
at the axis of the plume at the reactor release over emission rate during the year 1999 in unstable
classes.

U (m/s) w * (m/s) Hworst(m) xworst C /Q *10 6

(m) (sec/m3)
4.43 1.33 157.71 223.04 1.91
3.81 1.28 157.15 222.24 2.06
4 1.29 157.9 223.30 2.03
4.92 1.37 157.44 222.65 1.80
3.7 1.27 157.24 222.37 2.09
3.57 1.25 157.36 222.54 2.16
3.64 1.26 157.3 222.46 2.12
3.45 1.24 157.48 222.71 2.19
3.6 1.26 157.33 222.50 2.12

-347-
 3.8 1.27 157.16 222.26 2.09
3.99 1.29 157.01 222.05 2.03
3.89 1.28 157.08 222.14 2.06
3.75 1.27 157.2 222.31 2.09
3.98 1.29 157.02 222.06 2.03
3.47 1.25 157.46 222.68 2.16
4.06 1.3 157.96 223.39 1.99
4.3 1.32 157.79 223.15 1.93
4.31 1.32 157.78 223.13 1.93
4.02 1.29 157.99 223.43 2.03
4.11 1.3 157.92 223.33 1.99
3.94 1.29 157.05 222.10 2.03
3.86 1.28 157.11 222.19 2.06
3.67 1.26 157.27 222.41 2.12

The calculated values of u *, Hworst, xworst and normalized concentrations C/Q for neutral,
stable and unstable conditions are presented in Tables (2), (3) and (4) respectively. The last
columns in the three Tables are given in 48 hours that are the usual continuous operation
time of the reactor. The variations of the normalized concentrations are shown in the last
column of these Tables.
For the worst case concentrations, the trapping type plume is obtained by multiplying
the concentrations of Eqn. (5) by a factor of half. The ratio between trapping and non-
Gaussian concentration is given as follows:

Cimax H2
 (18)
Cmax 2 Zi2
Fig. (1) Shows that a polynomial and straight line decreasing fits well with data in the case
of stable and neutral conditions respectively, but a straight line is little decreasing in
unstable condition. The worst concentration is greater in stable condition than neutral
condition. In unstable condition, the worst concentration is about the same and decreasing.
The values of the worst normalized concentration in stable case is 15.8*10-6 sec/m3 at
x=222.7m,12.60*10-6 sec/m3 at x=223.02m and 2.19*10-6 sec/m3 at x=222.71m in stable,
neutral and unstable conditions respectively.
The fitting equations of the polynomial and linear in stable, neutral and unstable conditions
are as follows:
(C max/Q) stable = 0.65(xmax)3 -0.45(xmax)2 -0.57xmax+12.62 (19)

(C max/Q) neutral = -1.08xmax+249.96 (20)

(C max/Q)unstable = -0.07xmax+17.69 (21)

-348-
 20.00
6
Neutral *10 [c/q=-1.08x+249.96]
3 2
Stable* 106 [c/q=0.65x -0.45x -0.57x+12.62]
Maximum concentration /emission rate (sec/m3 )

6
16.00 Unstable *10 [c/q=-0.07x+17.69]

12.00

8.00

4.00

0.00

222.00 222.40 222.80 223.20 223.60

Maximum downwind distance (m)

Figure 1: Variation of worst concentration over emission rate (Cmax/Q) and the maximum
downwind distance (xmax)in neutral, stable and unstable conditions.

Fig. 2. Shows the ratio of worst ground-level trapping to non-Gaussian worst concentration
equals half non-dimensional distance H/Zi. The dimensionless of ground-level trapping to
non-Gaussian concentration in stable condition is larger than in neutral and unstable
conditions. The fitting of the ratio increases in stable condition from 0.4 to 0.6 but in
neutral and unstable conditions the ratio increases from 0.0 to 0.01.

-349-
 0.6

Maximum Trapped concentration /maximum concentration

Neutral
Stable
Unstable

0.4

0.2

0.0

0.88 0.92 0.96 1.00 1.04 1.08

Effective height/ mixing height
Figure 2: Ratio of ground-level trapping to non-Gaussian concentration via non-dimensional
distance.

CONCLUSIONS

The worst normalized concentration value is high in stable conditions, then neutral
condition and the minimum worst concentration happens in unstable case. The normalized
concentration depends on the effective height and inversion height.
The fitting worst concentration is greater in stable conditions than neutral conditions.
In unstable condition, the fitting worst concentration is nearly the same and decreasing. The
values of the worst normalized concentration in stable case is 15.8*10-6 sec/m3 at
x=222.7m, 12.60 sec/m3 at x=223.02m and 2.19*10-6 sec/m3 at x=222.71m in stable,
neutral and unstable conditions respectively.
The worst concentration for trapping plume is obtained by multiplying the worst
ground level concentration along the plume centerline by a factor half.
The dimensionless of ground-level trapping to non-Gaussian concentration in stable
condition is larger than in neutral and unstable conditions. The fitting of the ratio increases
in stable condition from 0.4 to 0.6 but in neutral and unstable conditions the ratio increases
from 0.0 to 0.2.

-350-
 REFERENCES

[1] Arya S. P., Air pollution meteorology and dispersion, Oxford University Press, New
York, (1999).
[2] Hess G. D., “Simulation of photochemical smog in the Melbourne air shed: worse case
Study”, (1967), Atmospheric Environment, Vol 23, issue 3, (1989), 661-669.
[3] Khaled S. M. Essa, Fawzia Mubarak, and Sawsan, E. M. Elsaid, (2006), “Effect of the
plume rise and wind speed on Extreme Value of air pollutant concentration”, Meteorology
and Atmospheric Physics, 93, 247-253, (2006).
[4] Khaled S. M. Essa, and Maha S. El-Otaify, “Mathematical model for hermitized
atmospheric dispersion in low winds with eddy diffusivities linear functions down wind
distance”, Meteorology and Atmospheric Physics, 96, 265-275, (2007).
[5] Laless D. P., Karras G. S., “Simple prediction formula for maximum ground level
concentrations from coning plumes”, Atmospheric Environment, Vol 11, issue 11,
1977,1117-1119, (1967).
[6] Pasquill F. and Smith F.B., Atmospheric diffusion. Ellis Horwood Ltd, Halstead press,
Chichester, (1983).
[7] Ragland K. W., “Worst –case ambient air concentrations from point sources using
theGaussian plume model”, Atmospheric Environment, 10, 371-374, (1976).
[8] Sharan M., Yadav A. K., “Simulation of diffusion experiments under light wind, stable
conditions by a variable K-theory model”, Atmospheric Environment, 32, 3481-3492,
(1998).
[8] “Reactor Physics Department”, Report 53, (1965).
[9] Zheng Tong xie, Paul Hayden, Alan G. Robins, peter R. Voke, “Modeling extreme
concentrations from a source in a turbulent flow over a rough wall”, Atmospheric Environ-
ment, Vol 41, issue 16, 3395-3406, (2007).

-351-