MODELLING OF ATMOSPHERIC FLOW AND DISPERSION IN THE … · 2008. 11. 12. · 12th Int. Conf. on Harmonisation within Atmospheric Dispersion Modelling for EREL Regulatory Purposes,

ERELEREL 12th Int. Conf. on Harmonisation within Atmospheric Dispersion Modelling for Regulatory Purposes, 6 – 9/10/2008, Cavtat, Croatia

1

MODELLING OF ATMOSPHERIC FLOW AND DISPERSION IN THE WAKE OF A

CYLINDRICAL OBSTACLE

S. Andronopoulos1, I. Mavroidis2, A. Venetsanos1, J.G. Bartzis3

1Environmental Research Laboratory, Institute of Nuclear Technology and Radiation Protection, National Centre for Scientific Research “Demokritos”, 15310 Aghia Paraskevi Attikis, Greece, e-mail:

[email protected] Open University, Patras, Greece

3Department of Energy Resources Management Engineering, University of Western Macedonia, Greece

mailto:[email protected]


2

Aim of the paper

• Computational study of the interaction of continuous plumes released from point sources with buildings– short-range dispersion of atmospheric pollutants in built-up

areas– Simple, isolated structures: main characteristics of flow and

dispersion• Present case: Cylindrical obstacle (“building”)

– Previously studied case: “Atmospheric dispersion in the presence of a three-dimensional cubical obstacle: Modelling of mean concentration and concentration fluctuations”, by Mavroidis, Andronopoulos, Bartzis, Griffiths, Atmospheric Environment 41 (2007), 2740-2756


3

Tools, data, methods

• CFD code ADREA-HF– Mean concentrations, concentration fluctuations

• Data from field experiments– “Field and wind tunnel investigations of plume dispersion

around single surface obstacles”, by Mavroidis, Griffiths, Hall, Atmospheric Environment 37 (2003) 2903-2918

• Comparisons of calculated with experimental results / model performance

• Study of computed dispersion patterns• Comparisons between cylindrical and cubical obstacles


4

Experimental set up: cylinder

• Dual source:- Ammonia- Propane

• Ammonia source displaced laterally

• Height and diameter of cylinder H=1.15m

• Source and detector height: H/2

• Atmospheric stability: D (neutral)

• Wind speed: 2.6 – 4.5 m/s


5

Experimental set up: cube

• Dual source:- Ammonia- Propane

• Ammonia source displaced laterally

• Height and diameter of cylinder H=1.15m

• Source and detector height: H/2

• Atmospheric stability: D (neutral)

• Wind speed: 4.7 – 6.3 m/s


6

Experimental results: ammonia concentrations

Sensor at ٠.٥H downwind

٠.٠

٠.١

٠.٢

٠.٣

٠.٤

٠.٥

٠.٦

٠.٧

٠.٨

٠xH ٠.٥xH ٠.٧٥xH ١.٠xH ١.٢٥xH ١.٥xH ١.٧٥xH ٢.٠xH

Source displacement

Am

mon

ia c

once

ntra

tion

(non

-dim

.)

CubeCylinder

Sensor at ٢H downwind

٠.٠

٠.١

٠.٢

٠.٣

٠.٤

٠.٥

٠.٦

٠.٧

٠.٨


Source displacement

Am

mon

ia c

once

ntra

tion

(non

-dim

.)

CubeCylinder

Sensor at ٣H downwind

٠.٠

٠.١

٠.٢

٠.٣

٠.٤

٠.٥

٠.٦

٠.٧

٠.٨


Source displacement

Am

mon

ia c

once

ntra

tion

(non

-dim

.)

CubeCylinder

Sensor at ٥H downwind

٠.٠

٠.١

٠.٢

٠.٣

٠.٤

٠.٥

٠.٦

٠.٧

٠.٨


Source displacement

Am

mon

ia c

once

ntra

tion

(non

-dim

.)

CubeCylinder


7

Experimental results: ammonia concentration fluctuations


٠.٠

١.٠

٢.٠

٣.٠

٤.٠

٥.٠

٦.٠

٧.٠


Source displacement

Am

mon

ia c

once

ntra

tion

fluct

uatio

n (c

'/C)

CubeCylinder


٠.٠

١.٠

٢.٠

٣.٠

٤.٠

٥.٠

٦.٠

٧.٠


Source displacement

Am

mon

ia c

once

ntra

tion

fluct

uatio

n (c

'/C)

CubeCylinder

Sesnor at ٣H downwind

٠.٠

١.٠

٢.٠

٣.٠

٤.٠

٥.٠

٦.٠

٧.٠


Source displacement

Am

mon

ia c

once

ntra

tion

fluct

uatio

n (c

'/C)

CubeCylinder


٠.٠

١.٠

٢.٠

٣.٠

٤.٠

٥.٠

٦.٠

٧.٠


Source displacement

Am

mon

ia c

once

ntra

tion

fluct

uatio

n (c

'/C)

CubeCylinder


8

Experimental results: ammonia concentration intermittency


٠.٠

٠.١

٠.٢

٠.٣

٠.٤

٠.٥

٠.٦

٠.٧

٠.٨

٠.٩

١.٠


Source displacement

Am

mon

ia c

once

ntra

tion

inte

rmitt

ency

CubeCylinder


٠.٠

٠.١

٠.٢

٠.٣

٠.٤

٠.٥

٠.٦

٠.٧

٠.٨

٠.٩

١.٠


Source displacement

Am

mon

ia c

once

ntra

tion

inte

rmitt

ency

CubeCylinder

Sesnor at ٣H downwind

٠.٠

٠.١

٠.٢

٠.٣

٠.٤

٠.٥

٠.٦

٠.٧

٠.٨

٠.٩

١.٠


Source displacement

Am

mon

ia c

once

ntra

tion

inte

rmitt

ency

CubeCylinder


٠.٠

٠.١

٠.٢

٠.٣

٠.٤

٠.٥

٠.٦

٠.٧

٠.٨

٠.٩

١.٠


Source displacement

Am

mon

ia c

once

ntra

tion

inte

rmitt

ency

CubeCylinder


9

Computational results: non-dimensional mean concentration

X (H)

Y(H

)

-٥ ٠ ٥ ١٠

-٦

-٤

-٢

٠

٢

٤

٦

C non-dim١٠١٠.٧٠.٥٠.٣٠.١٠.٠١

N٠١

X (H)

Y(H

)

-٥ ٠ ٥ ١٠

-٦

-٤

-٢

٠

٢

٤

٦

C non-dim١٠١٠.٧٠.٥٠.٣٠.١٠.٠١

N٠٢

X (H)

Y(H

)

-٥ ٠ ٥ ١٠

-٦

-٤

-٢

٠

٢

٤

٦

C non-dim١٠١٠.٧٠.٥٠.٣٠.١٠.٠١

N٠٥

X (H)

Y(H

)

-٥ ٠ ٥ ١٠

-٦

-٤

-٢

٠

٢

٤

٦

C non-dim١٠١٠.٧٠.٥٠.٣٠.١٠.٠١

N١١

X (H)

Y(H

)

-٥ ٠ ٥ ١٠

-٦

-٤

-٢

٠

٢

٤

٦

C non-dim١٠١٠.٧٠.٥٠.٣٠.١٠.٠١

N٠٧


10

Computational results: non-dimensional concentration StD

X (H)

Y(H

)

-٥ ٠ ٥ ١٠

-٦

-٤

-٢

٠

٢

٤

٦

StD non-dim١٠١٠.٧٠.٥٠.٣٠.١٠.٠١

N٠١

X (H)

Y(H

)

-٥ ٠ ٥ ١٠

-٦

-٤

-٢

٠

٢

٤

٦

StD non-dim١٠١٠.٧٠.٥٠.٣٠.١٠.٠١

N٠٢

X (H)

Y(H

)

-٥ ٠ ٥ ١٠

-٦

-٤

-٢

٠

٢

٤

٦

StD non-dim١٠١٠.٧٠.٥٠.٣٠.١٠.٠١

N٠٥

X (H)

Y(H

)

-٥ ٠ ٥ ١٠

-٦

-٤

-٢

٠

٢

٤

٦

StD non-dim١٠١٠.٧٠.٥٠.٣٠.١٠.٠١

N١١

X (H)

Y(H

)

-٥ ٠ ٥ ١٠

-٦

-٤

-٢

٠

٢

٤

٦

StD non-dim١٠١٠.٧٠.٥٠.٣٠.١٠.٠١

N٠٧


11

Model evaluation: turbulence closure schemes – k-l (1), k-ε, k-l (2)

Non-dimensional ammonia concentration

٠.٠

٠.١

٠.٢

٠.٣

٠.٤

٠.٥

٠.٦

٠.٧

٠.٨

٠.٩

٠ ٠.١ ٠.٢ ٠.٣ ٠.٤ ٠.٥ ٠.٦ ٠.٧ ٠.٨ ٠.٩

Experimental

Cal

cula

ted N٠١

N٠٢N٠٥N١١N٠٧


٠.٠

٠.١

٠.٢

٠.٣

٠.٤

٠.٥

٠.٦

٠.٧

٠ ٠.١ ٠.٢ ٠.٣ ٠.٤ ٠.٥ ٠.٦ ٠.٧

Experimental

Cal

cula

ted N٠١

N٠٢N٠٥N١١N٠٧


٠.٠٠

٠.٠٥

٠.١٠

٠.١٥

٠.٢٠

٠.٢٥

٠.٣٠

٠.٣٥

٠.٤٠

٠.٠٠ ٠.٠٥ ٠.١٠ ٠.١٥ ٠.٢٠ ٠.٢٥ ٠.٣٠ ٠.٣٥ ٠.٤٠

Experimental

Cal

cula

ted

N٠١N٠٢N٠٥N١١N٠٧


12

Model evaluation : turbulence closure schemes – k-l (1), k-ε, k-l (2)

Non-dimensional ammonia concentration StD

٠.٠

٠.١

٠.٢

٠.٣

٠.٤

٠.٥

٠.٦

٠.٧

٠.٨

٠.٩

٠ ٠.١ ٠.٢ ٠.٣ ٠.٤ ٠.٥ ٠.٦ ٠.٧ ٠.٨ ٠.٩

Experimental

Cal

cula

ted N٠١

N٠٢N٠٥N١١N٠٧


٠.٠

٠.١

٠.٢

٠.٣

٠.٤

٠.٥

٠.٦

٠.٧

٠ ٠.١ ٠.٢ ٠.٣ ٠.٤ ٠.٥ ٠.٦ ٠.٧

Experimental

Cal

cula

ted N٠١

N٠٢N٠٥N١١N٠٧


٠.٠

٠.١

٠.٢

٠.٣

٠.٤

٠.٥

٠.٦

٠.٧

٠.٠ ٠.١ ٠.٢ ٠.٣ ٠.٤ ٠.٥ ٠.٦ ٠.٧

Experimental

Cal

cula

ted

N٠١N٠٢N٠٥N١١N٠٧


13

Model evaluation : turbulence closure schemes – k-l (1), k-l (2), k-ε

Ammonia concentration fluctuation

٠

١

٢

٣

٤

٥

٦

٧

٨

٩

٠ ١ ٢ ٣ ٤ ٥ ٦ ٧ ٨ ٩

Experimental

Cal

cula

ted

N٠١N٠٢N٠٥N١١N٠٧

Ammonia concentration fluctuation

٠

١

٢

٣

٤

٥

٦

٧

٨

٩

٠ ١ ٢ ٣ ٤ ٥ ٦ ٧ ٨ ٩

Experimental

Cal

cula

ted

N٠١N٠٢N٠٥N١١N٠٧

Ammonia concentration fluctuations

٠

١

٢

٣

٤

٥

٦

٠ ١ ٢ ٣ ٤ ٥ ٦

Experimental

Cal

cula

ted

N٠١N٠٢N٠٥N١١N٠٧


14

Model evaluation : turbulence closure schemes – k-l (1), k-l (2), k-ε

٠.٥H ٣.٠H ٠.٥H ٣.٠H ٠.٥H ٣.٠H ٠.٥H ٣.٠HN٠١ ٠.٤٢٣ ٠.٢٣٨ ٠.٧٨٨ ٠.٤٤٥ ٠.٦٢٩ ٠.٤٤٢ ٠.٢٦٢ ٠.١٤٧N٠٢ ٠.٣١٩ ٠.١٧١ ٠.٨٠٦ ٠.٤٦٣ ٠.٦٦٤ ٠.٤٨١ ٠.٢٩٥ ٠.١٦٨N٠٥ ٠.٢٨٨ ٠.١٧٥ ٠.٧٩٠ ٠.٤٥١ ٠.٦٤٢ ٠.٤٥٩ ٠.٢٨٠ ٠.١٥٨N١١ ٠.٣٣٨ ٠.٢٣٨ ٠.٨٠٥ ٠.٤٦١ ٠.٦٦١ ٠.٤٧٨ ٠.٢٨٨ ٠.١٦٤N٠٧ ٠.٢٠٤ ٠.١٢٢ ٠.٧٨٧ ٠.٤٤٩ ٠.٦٣٦ ٠.٤٥٣ ٠.٢٧٦ ٠.١٥٦

Average ٠.٣١٤ ٠.١٨٩ ٠.٧٩٥ ٠.٤٥٤ ٠.٦٤٦ ٠.٤٦٢ ٠.٢٨٠ ٠.١٥٨Calc/Exper ٢.٥٣ ٢.٤٠ ٢.٠٦ ٢.٤٥ ٠.٨٩ ٠.٨٤

Experimental Calculated k-εCalculated k-l (١) Calculated k-l (٢)

٠.٥H ٣.٠H ٠.٥H ٣.٠H ٠.٥H ٣.٠H ٠.٥H ٣.٠HN٠١ ٠.٥٣٧ ٠.٢٧٨ ١.٠٤٤ ٠.٨٢٠ ١.٠١٧ ٠.٩٣٥ ٠.٧٦٥ ٠.٤٢٥N٠٢ ٠.٤٧٨ ٠.٢٣٥ ١.٢١٥ ٠.٩٥٩ ٠.٩٧٢ ٠.٩٧٧ ٠.٩٠٨ ٠.٥١٢N٠٥ ٠.٤١٦ ٠.٢٣٦ ١.١٤١ ٠.٨٩٧ ٠.٩٨٧ ٠.٩٦٦ ٠.٨٤٥ ٠.٤٧٥N١١ ٠.٣٧٢ ٠.٢٧١ ١.٢١٠ ٠.٩٥٥ ٠.٩٧٩ ٠.٩٨٣ ٠.٨٩٩ ٠.٥١٠N٠٧ ٠.٣٣٩ ٠.١٩٩ ١.١٢٢ ٠.٨٨١ ٠.٩٩٣ ٠.٩٦١ ٠.٨٢٩ ٠.٤٦٥

Average ٠.٤٢٩ ٠.٢٤٤ ١.١٤٦ ٠.٩٠٢ ٠.٩٨٩ ٠.٩٦٤ ٠.٨٤٩ ٠.٤٧٨Calc/Exper ٢.٦٧ ٣.٧٠ ٢.٣١ ٣.٩٦ ١.٩٨ ١.٩٦

Experimental Calculated k-l (١) Calculated k-ε Calculated k-l (٢)

Propane Concentration

Propane Concentration

StD


15

Model evaluation: comparison between cube and cylinder cases

Experimental Calculated, k-l ( )١ Ratio Calculated, k-ε Ratio٠.٥H ٠.٧٣٩ ٠.٨١٦ ١.١٠٣ ٠.٨٠٩ ١.٠٩٤٣.٠H ٠.٢٤٦ ٠.٤٣٨ ١.٧٧٧ ٠.٤٩١ ١.٩٩٣

Experimental Calculated, k-l ( )١ Ratio Calculated, k-ε Ratio٠.٥H ٠.٣١٤ ٠.٧٩٥ ٢.٥٢٩ ٠.٦٤٦ ٢.٠٥٦٣.٠H ٠.١٨٩ ٠.٤٥٤ ٢.٤٠٢ ٠.٤٦٢ ٢.٤٤٨

Experimental Calculated, k-l ( )١ Ratio Calculated, k-ε Ratio٠.٥H ٠.٥٠٩ ٠.٤٨٩ ٠.٩٦٠ ٠.٣٦٧ ٠.٧٢١٣.٠H ٠.٣٢٢ ٠.٤٠٥ ١.٢٥٦ ٠.٣٥٠ ١.٠٨٥

Experimental Calculated, k-l ( )١ Ratio Calculated, k-ε Ratio٠.٥H ٠.٤٢٩ ١.١٤٦ ٢.٦٧٥ ٠.٩٨٩ ٢.٣٠٨٣.٠H ٠.٢٤٤ ٠.٩٠٢ ٣.٧٠٣ ٠.٩٦٤ ٣.٩٥٧

Average propane concentrations for CUBE

Average propane concentrations for CYLINDER

Average propane concentration StD for CUBE

Average propane concentration StD for CYLINDER


16


Cube, sensor at ٠.٥ H downwind

٠.٠

٠.١

٠.٢

٠.٣

٠.٤

٠.٥

٠.٦

٠.٧

٠.٨

٠.٩

٠xH ٠.٥xH ٠.٧٥xH ١.٠xH ١.٢٥xH ١.٥xH ١.٧٥xH ٢.٠xHLateral Source displacement

Con

cent

ratio

n (n

on-d

im.)

MeasuredCalc., k-lCalc., k-ε

Cube, sensor at ٢ H downwind

٠.٠

٠.١

٠.٢

٠.٣

٠.٤

٠.٥

٠.٦

٠.٧

٠.٨


Con

cent

ratio

n (n

on-d

im.)


Cylinder, sensor at ٠.٥ H downwind

٠.٠

٠.١

٠.٢

٠.٣

٠.٤

٠.٥

٠.٦

٠.٧

٠.٨

٠.٩


Con

cent

ratio

n (n

on-d

im.) Measured

Calc., k-l(١)Calc., k-εCalc., k-l(٢)

Cylinder, sensor at ٢ H downwind

٠.٠

٠.١

٠.٢

٠.٣

٠.٤

٠.٥

٠.٦

٠.٧

٠.٨


Con

cent

ratio

n (n

on-d

im.) Measured



17


Cube, sensor at ٣ H downwind

٠.٠

٠.١

٠.٢

٠.٣

٠.٤

٠.٥

٠.٦

٠.٧


Con

cent

ratio

n (n

on-d

im.)


Cylinder, sensor at ٣ H downwind

٠.٠

٠.١

٠.٢

٠.٣

٠.٤

٠.٥

٠.٦

٠.٧


Con

cent

ratio

n (n

on-d

im.) Measured


Cube, sensor at ٥ H downwind

٠.٠٠

٠.٠٥

٠.١٠

٠.١٥

٠.٢٠

٠.٢٥

٠.٣٠

٠.٣٥

٠.٤٠

٠.٤٥

٠.٥٠


Con

cent

ratio

n (n

on-d

im.)


Cylinder, sensor at ٥ H downwind

٠.٠٠

٠.٠٥

٠.١٠

٠.١٥

٠.٢٠

٠.٢٥

٠.٣٠

٠.٣٥

٠.٤٠

٠.٤٥

٠.٥٠


Con

cent

ratio

n (n

on-d

im.) Measured



18


Cube, sensor at ٠.٥ H downwind

٠.٠

٢.٠

٤.٠

٦.٠

٨.٠

١٠.٠

١٢.٠


Con

cent

ratio

n flu

ctua

tion

(c'/C

)

MeasuredCalc., k-l(١)Calc., k-ε

Cylinder, sensor at ٠.٥ H downwind

٠.٠

٢.٠

٤.٠

٦.٠

٨.٠

١٠.٠

١٢.٠


Con

cent

ratio

n flu

ctua

tion

(c'/C

)

MeasuredCalc., k-l(١)Calc., k-εCalc., k-l(٢)

Cube, sensor at ٢ H downwind

٠.٠

٢.٠

٤.٠

٦.٠

٨.٠

١٠.٠

١٢.٠


Con

cent

ratio

n flu

ctua

tion

(c'/C

)


Cylinder, sensor at ٢ H downwind

٠.٠

٢.٠

٤.٠

٦.٠

٨.٠

١٠.٠

١٢.٠


Con

cent

ratio

n flu

ctua

tion

(c'/C

)



19


Cube, sensor at ٣ H downwind

٠.٠

١.٠

٢.٠

٣.٠

٤.٠

٥.٠

٦.٠

٧.٠

٨.٠

٩.٠


Con

cent

ratio

n flu

ctua

tion

(c'/C

)


Cylinder, sensor at ٣ H downwind

٠.٠

١.٠

٢.٠

٣.٠

٤.٠

٥.٠

٦.٠

٧.٠

٨.٠

٩.٠


Con

cent

ratio

n flu

ctua

tion

(c'/C

)


Cube, sensor at ٥ H downwind

٠.٠

١.٠

٢.٠

٣.٠

٤.٠

٥.٠

٦.٠


Con

cent

ratio

n flu

ctua

tion

(c'/C

)


Cylinder, sensor at ٥ H downwind

٠.٠

١.٠

٢.٠

٣.٠

٤.٠

٥.٠

٦.٠


Con

cent

ratio

n flu

ctua

tion

(c'/C

)



20


E٠٣, YS = ٠ H

٠.٠

٠.١

٠.٢

٠.٣

٠.٤

٠.٥

٠.٦

٠.٧

٠.٨

٠.٩

١.٠

٠ ٢ ٤ ٦ ٨ ١٠ ١٢Downwind distance from obstacle (H)

Con

cent

ratio

n (n

on-d

im)

ExperimentalCalc., k-lCalc., k-ε

N٠١, YS = ٠ H

٠.٠

٠.١

٠.٢

٠.٣

٠.٤

٠.٥

٠.٦

٠.٧

٠.٨

٠.٩

١.٠


Con

cent

ratio

n (n

on-d

im)

ExperimentalCalc., k-l(١)Calc., k-εCalc., k-l(٢)

E٠٤, YS = ٠.٥ H

٠.٠

٠.١

٠.٢

٠.٣

٠.٤

٠.٥

٠.٦

٠.٧

٠.٨

٠.٩

١.٠

٠ ٢ ٤ ٦ ٨ ١٠ ١٢Downwind dstance from obstacle (H)

Con

cent

ratio

n (n

on-d

im)


N٠٢, Ys=٠.٥ H

٠.٠

٠.١

٠.٢

٠.٣

٠.٤

٠.٥

٠.٦

٠.٧

٠.٨

٠.٩

١.٠


Con

cent

ratio

n (n

on-d

im)



21


E١٠, YS = ١ H

٠.٠٠

٠.٠٥

٠.١٠

٠.١٥

٠.٢٠

٠.٢٥

٠.٣٠

٠ ٢ ٤ ٦ ٨ ١٠ ١٢Downwind distance from obstacle (Η)

Con

cent

ratio

n (n

on-d

im)


N٠٥, Ys = ١ H

٠.٠٠

٠.٠٥

٠.١٠

٠.١٥

٠.٢٠

٠.٢٥

٠.٣٠


Con

cent

ratio

n (n

on-d

im) Experimental


E٠٦, YS = ١.٥ H

٠.٠٠

٠.٠٥

٠.١٠

٠.١٥

٠.٢٠

٠.٢٥


Con

cent

ratio

n (n

on-d

im)


N١١, Ys = ١.٥ H

٠.٠٠

٠.٠٥

٠.١٠

٠.١٥

٠.٢٠

٠.٢٥


Con

cent

ratio

n (n

on-d

im)



22


E٠٧, YS = ٢ H

٠.٠٠

٠.٠٢

٠.٠٤

٠.٠٦

٠.٠٨

٠.١٠

٠.١٢

٠.١٤

٠.١٦


Con

cent

ratio

n (n

on-d

im)


N٠٧, YS = ٢ H

٠.٠٠

٠.٠٢

٠.٠٤

٠.٠٦

٠.٠٨

٠.١٠

٠.١٢

٠.١٤

٠.١٦


Con

cent

ratio

n (n

on-d

im)



23


E٠٣, YS = ٠ H

٠

٠.٢

٠.٤

٠.٦

٠.٨

١

١.٢

١.٤

١.٦


Con

cent

ratio

n flu

ctua

tion

(c'/C

)


N٠١, YS = ٠ H

٠.٠

٠.٢

٠.٤

٠.٦

٠.٨

١.٠

١.٢

١.٤

١.٦


Con

cent

ratio

n flu

ctua

tion

(c'/C

)


E٠٤, YS = ٠.٥ H

٠.٠

٠.٢

٠.٤

٠.٦

٠.٨

١.٠

١.٢

١.٤

١.٦

١.٨

٢.٠


Con

cent

ratio

n Fl

uctu

atio

ns (c

'/C)


N٠٢, Ys = ٠.٥ H

٠.٠

٠.٢

٠.٤

٠.٦

٠.٨

١.٠

١.٢

١.٤

١.٦

١.٨

٢.٠


Con

cent

ratio

n flu

ctua

tion

(c'/C

) ExperimentalCalc., k-l(١)Calc., k-εCalc., k-l(٢)


24


E١٠, YS = ١ H

٠

١

٢

٣

٤

٥

٦


Con

cent

ratio

n Fl

uctu

atio

n (c

'/C)


N٠٥, Ys = ١ H

٠

١

٢

٣

٤

٥

٦


Con

cent

ratio

n flu

ctua

tion

(c'/C

)


E٠٦, YS = ١.٥ H

٠

١

٢

٣

٤

٥

٦

٧

٨


Con

cent

ratio

n Fl

uctu

atio

n (c

'/C)


N١١, Ys = ١.٥ H

٠

١

٢

٣

٤

٥

٦

٧

٨


Con

cent

ratio

n flu

ctua

tion

(c'/C

)ExperimentalCalc., k-l(١)Calc., k-εCalc., k-l(٢)


25


E٠٧, YS = ٢ H

٠

٢

٤

٦

٨

١٠

١٢

١٤


Con

cent

ratio

n Fl

uctu

atio

n (c

'/C)


N٠٧, YS = ٢ H

٠

٢

٤

٦

٨

١٠

١٢

١٤


Con

cent

ratio

n flu

ctua

tion

(c'/C

)



26

Summary and conclusions (1)

• Computational simulations of atmospheric dispersion experiments around isolated obstacles in the field

• CFD code ADREA-HF: dispersion of positively or negatively buoyant gases in complicated geometries

• Single cylindrical obstacle normal to the mean wind direction and two upwind sources of ammonia and propane, with the ammonia source located at different lateral positions

• Concentrations and concentration fluctuations for both gases were calculated by the model and compared with the experimental results

• Comparisons of experimental and model results with the case of dispersion around an isolated cubical obstacle are also presented and discussed


27


• Analysis of experimental results:– Variation of ammonia concentrations, concentration

fluctuations and intermittency as the source is displaced laterally

• Source at 0. H and 0.5 H: higher concentrations for cube• Sharper decrease of concentration with source displacement• Concentration peaks for source at 0.5 H (2, 3 and 5 H downwind)• Cylinder: fluctuations increase, cube: fluctuations peak for

source at 1.5 H off• Intermittency increases with source displacement, more sharply

for cube• Computational results: k-l (2 versions for cylinder)

and k-ε turbulence closure schemes tested


28


• Model performance evaluation:– “Scatter” plots for cylinder case

• Ammonia concentrations– k-l(1) and k-ε similar performance, around factor-of-2– k-l(1) more points inside the factor-of-2 range, large

overestimation for cases with large source displacement• Ammonia concentration StD

– Most points inside the factor-of-2 range for all models• Ammonia concentration fluctuations

– Most points inside the factor-of-2 range for all models– k-l(2) model results show little variation


29


• Model performance evaluation:– Propane results statistics:

• “Ensemble” averages (same source position, stability conditions, similar wind)

• The model performance for the cube case was better than for the cylinder case

• The ratio Calc. / Exper. for concentrations is 2 times higher and for concentration StD is 3 times higher in the cylinder case

– Ammonia concentration variation as the source is displaced laterally

• k-l(1) and k-ε model results are similar and in general overestimate. Some peaks are predicted, only for the cube

• k-l(2) model results are better for small source displacements but show little variation


30


• Model performance evaluation:– Ammonia concentration fluctuation variation as the source is

displaced laterally• Models do not predict the experimental peak for the cube for

source displaced 1.5 H off the centreline• k-l(2) model results show very little variation

– Ammonia concentration profiles downwind• Experimental values are higher for the cube than for the

cylinder close to the obstacle and decrease more sharply• k-l(1) and k-ε results are similar and overestimate the

experimental values for source placed at 0, 0.5, 1 and 1.5 H off the centreline

• k-l(2) agree better for source placed at 0, 0.5, 1 H but are worse for source placed at 1.5 and 2 H off the centreline


31


• Model performance evaluation:– Ammonia concentration fluctuations profiles downwind

• Model results peaks are “weaker” than the experimental and occur closer to the obstacle

• For the cylinder case no peak is observed in the experimental data for source displacement above 1 H off centreline

• k-l(2) model results are better for smaller source displacements, k-l(1) and k-ε are better for larger source displacements.

• Future work:– More detailed analysis of the differences in results and

model performance between cubical and cylindrical obstacle cases

Documents

MODELLING OF ATMOSPHERIC FLOW AND DISPERSION IN THE … · 2008. 11. 12. · 12th Int. Conf. on Harmonisation within Atmospheric Dispersion Modelling for EREL Regulatory Purposes,