EPA-600/4-77-006
January 1977
DISPERSION OF ROOF-TOP EMISSIONS FROM ISOLATED BUILDINGS
A Wind Tunnel Study
Roger S. Thompson
Meteorology and Assessment Division
Environmental Sciences Research Laboratory
Research Triangle Park, NC 27711
and
David J. Lombard!
Northrop Services Inc.
Research Triangle Park, NC 27711
SCIENCES RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
KESEARCH TRInNbLE PARK, NORTH CAROLINA 27711
-------
DISCLAIMER
This report has been reviewed by the Environmental Sciences Research
Laboratory, U.S. Environmental Protection Agency, and approved for publi-
cation. Mention of trade names or commercial products does not constitute
endorsement or recommendation for use.
11
-------
ABSTRACT
A fluid modeling study of the dispersion of roof-top emissions from
rectangular buildings was performed in the meteorological wind tunnel of
the EPA Fluid Modeling Facility. The basic building shape was a 0.18 meter
cube. Variations included a building twice as wide and buildings twice
and three times as high. Each building was placed in a 1.8 meter, simulated,
neutral atmospheric boundary layer. Low momentum, non-buoyant emissions
were released through an opening at the roof center. Photographs of flow
visualization experiments are presented for a qualitative evaluation of
the building wakes. Concentration measurements at ground level and aloft
are presented for each case at downwind distances within 20 building heights.
Concentrations are presented in a non-dimensional form for ease in application
to full scale situations. The orientation of the wind to the building was
found to significantly affect ground-level concentrations near the building.
A 45° orientation of the cubical building increased the maximum ground-level
concentration by a factor of six over that observed for a perpendicular
orientation.
iii
-------
CONTENTS
Abstract i i i
Figures vi
iyinuo 1 s vi i i
1. Introduction 1
2. conclusions , 2
3. hxperirnental Techniques 3
4. Experimental Results 5
Keferences 9
-------
FIGURES
Number Page
1 Top and side views of building model 11
2 Cubical model in wind tunnel test section during flow visua-
lization experiment 12
3 Mean velocity profiles at location of cubical building with-
out building in position (x=C) and downwind of building
(x=2H,10H) 13
4 Local turbulence intensity at location of building without
building in position (x=0) and downwind of tiie building
(x=2H) 14
5 Flow visualization of Case 1 15
6 Vertical concentration profiles for Cases 1 and 8 (y-0) ... 16
7 Flow visualization of Case 1 (top view) 17
8 Lateral ground-level concentration profiles for Case 1 .... 18
9 Ground-level concentrations for Cases 1, 2, 6, and 9
Cy=0;z=0.025ni) 19
10 Flow visualization of Case 2 20
11 Flow visualization of Case 2 (top view) 21
12 Vertical concentration profiles for Case 2 (y=0) 22
13 Lateral ground-level concentration profiles for Cases 1, 2, 3,
4, 6, and 9 (x=0.9m) 23
14 Vertical concentration profiles for Case 9 (y=0) 24
15 Ground-level concentrations for Cases 1, 3, 4, and 8
(y=0;z=0.025m) 25
16 Flow visualization of Case 3 26
17 Vertical concentration profiles for Case 3 (y=0) 27
VI
-------
FIGURES (Cont'd)
Number Page
18 Flow visualization of Case 4 28
19 Vertical concentration profiles for Case 4 (y-0) 29
20 Flow visualization of Case 5 39
21 Flow visualization of Case 6 31
22 Flow visualization of Case 7 32
23 Vertical concentration profiles for Case 5 (y-0) 33
24 Vertical concentration profiles for Case 6 (y-0) 34
25 Vertical concentration profiles for Case 7 (y=0) 35
vn
-------
SYMBOLS
..Measured tracer concentration, % by vol.
..Maximum tracer concentration for a given profile, % by vol.
D ..Diameter of source, m
h . .Building height, rn
L ..Building length, m
3
g ..Source emission rate of tracer, % by vol. m /sec
Re ..Building Reynolds number, dimensionless
U(z) ..Mean wind speed at height z, m/sec
U ..Mean wind speed at building height in absence of building,
m/sec
U ..Mean free stream wind speed, m/sec
oo
p 2
2 ..Mean square of wind speed fluctuation, m /sec
u1
vJ . .Building width, m
w ..Mean effluent speed at source, m/sec
x ..Coordinate in direction downwind from source, m
y ..Coordinate in crosswind direction, m
z ..Coordinate in vertical direction, m
a ..Exponent of wind speed profile, dimensionless
<5 ..Boundary layer thickness, m
6 ..Angle between wind direction and normal to upwind building
face, degrees
2
v ..Kinematic viscosity of air, m /sec
p ..Density of effluent, Kg/rn
3
p ..Density of ambient air, Kg/rn
a
Y ..Non-dimensional concentration coefficient
viii
-------
SECTION 1
INTRODUCTION
because fluid modeling can accurately reflect the complex nature of
turbulent air flow near structures, it is uniquely suited to the study of
the dispersion of matter released from a building's roof-top. In such
studies, two distinct phenomena have been observed downwind of an isolated,
blunt structure. One is the development of an aerodynamic cavity of high
turbulence and reverse flow caused by separation of the air flow over the
building roof. The other is the confinement of roof-top emissions to the
wake or dispersion envelope, which are determined by the shape of the
building and by the approaching wind field.
The dispersion of roof-top emissions was modeled by Halitsky (1968)
in a study of the concentration fields at and near rectangular buildings
placed at various orientations to the wind. He used an approach flow of
uniform mean velocity and low turbulence and an emission source located at
tne center of the roof. Although effluent momentum was varied, all emissions
were non-buoyant. In another study, Yang and Meroney (1970) considered
low momentum releases from all faces of a cubical structure and modeled both
neutral and stratified boundary layers. Robins (1975) and Castro and Robins
(1975) have performed fairly thorough studies of the turbulent wind field
and plume dispersion in the vicinity of a surface mounted cube in a wind
tunnel at Marchwood Engineering Laboratories, England. Conditions for some
of their cases were quite similar to those used for the cubical building in
the present study.
The present study, wnich was performed in the Environmental Protection
Agency's Fluid Modeling Facility, extends previous work by analyzing concen-
tration fields downwind of four different shapes of rectangular buildings.
Block-shaped model buildings were placed in a simulated, neutral atmospheric
boundary layer to determine the indluence of the wind field and building
geometry on the dispersion of roof-top emissions. Only simple building
shapes were used with no other buildings, trees, or hills located nearby.
Air containing a low percentage of methane as a tracer was released from the
center of the roof to model a low momentum, non-buoyant effluent. Ground-
level and vertical concentration profiles were determined downwind by meas-
uring tracer levels in samples taken at various positions.
-------
SECTION 2
CONCLUSIONS
Smoke visualizations and measured concentration profiles are presented
for generalized building shapes and wind directions. The situation was
simplified in tiiat the source was positioned at the roof center and the
effluent was non-buoyant and had negligible momentum. Considering the
cubical building model as the basis for comparision, some genral observa-
tions were made:
1) A change in the direction of the approaching wind significantly
altered the shape and size of the aerodynamic cavity and the dispersion
envelope. These changes resulted in different dispersion patterns and
ground-level concentrations. A wind direction of 45° increased the maximum
ground-level concentration by a factor of six. However, ground-level con-
centrations beyond x = 10H were essentially equal for the 0° and 45° cases.
2) Doubling the width of the building reduced by about 20 percent the
concentrations below the building height (z = H) at a downwind distance of
x = 5H. Tiie ground-level maximum was approximately the same as that for the
cube, but occurred nearer the building. Ground-level concentrations at
distances beyond x = 5H were about 30 percent lower than for the cubical
uuilding.
3) Doubling the length of the building in the direction of the wind
had little effect on the concentrations.
4) Doubling or tripling the height of the building elevated the plume
and lowered the ground-level concentrations. The dispersion envelope over
the building was similar for all three heights tested. Changing the wind
direction from the perpendicular for the cases involving taller buildings
increased the ground-level concentrations near the building.
The measurements of concentration as presented should be of use to
numerical modelers in developing algorithms for predicting concentrations
near pollution sources, such as roof-top incinerators, roof exhausts, etc.
For full scale situations that are resonably similar to those of the block
models of this study, the x values of the figures can be applied directly
to predict concentrations.
-------
SECTION 3
EXPERIMENTAL TECHNIQUES
The experiment was performed in the EPA Meteorological Wind Tunnel
(Snyder, 1974; Snyder et al., 1977). This low speed, open circuit wind
tunnel has an 18 meter (m) long, 2.1 m high, 3.7 m wide test section in
which atmospheric boundary layer air flow is simulated. Vortex generating
fins as developed by Counihan (1969) were used to structure the desired
boundary layer. The fins used for this study were 1.8 n high and produced
a boundary layer height (0 of 1.8 m. Two-dimensional roughness slats,
0.019 m high by 0.051 m wide and placed crosswind on 0.475 m centers, were
mounted on the tunnel floor to maintain an equilibrium boundary layer over
the length of tiie test section.
A dimensionless ratio of interest in modeling dispersion near buildings
is the building Reynolds number, Re = HU^/v, where II is the building height,
UH is the mean wind speed at z = H with no building present, and v is the
kinematic viscosity of air. Ideally, the model Reynolds number and the full
scale Reynolds number would be equal. However, in full scale situations,
the Reynolds number is generally orders of magnitude larger than it is possi-
ble to obtain in the wind tunnel. Golden (1961) showed that concentration
fields near building models are invariant with Reynolds number provided the
building Reynolds number is greater than 11,000. In this study, the minimum
Reynolds number for any case was 17,500.
For the model flow to be representative of atmospheric flows, the
mean velocity field and tiie turbulence structure must be simulated. Profiles
of mean velocity measured in a neutral atmosphere can be expressed as a
power law of the form,
U(z) . /Zxa
U V
CO
The value of the exponent, a, is governed by the roughness of the terrain.
For a wind tunnel boundary layer flow, a power law with an exponent charac-
teristic of the modeled terrain is desired. Similarly, components of local
turbulence should be comparable to those in the atmospheric boundary layer.
Mean velocity and local turbulence intensity were measured with a
hot-film anemometer. Vertical profiles were obtained where the building
-------
would be located. The anemometer signal was digitized, linearized, and
processed by a minicomputer. One minute averages at a sampling rate of
6000 samples per second were used.
Eight combinations of building shape and orientation to the wind were
studied (Table 1 and Figure 1). The most extensive measurements were taken
for Case 1, a cubical building (H=0.18 m) with a face perpendicular to the
wind. An additional case, no. 8, was a repeat of case 1 with the mean wind
speed twice that for case 1. It was performed as a check to verify that the
dispersion in the building wake was independent of building Reynolds number.
For all cases the model building was well within the boundary layer.
Flow visualization was used to gain insight into the qualitative fea-
tures of the dispersion. An oil-fog generator was used to produce smoke
which was released from the emission source. Sampling positions for quanti-
tative measurements were based on observed plume trajectories. Figure 2
shows the cubical building model in position in the wind tunnel test section,
the vortex generating fins, the floor roughness elements and scales to the
right of the model. Marks on the scales are 0.05 m apart and the two scales
were 0.9 m apart. Two tungsten lights were used to illuminate the smoke
plume. Still photographs of each case were taken for a permanent record
using a 35 mm camera with a 135 mm lens. The exposure of one second at
f/16 VMS long enough to average most fluctuations of the plume.
For quantitative experiments, a 4.8 percent by volume mixture of methane-
in-air was used as the tracer. The effluent rate, monitored through rota-
meters, was held at 4.0 liters per minute throughout the experiment. With
the source diameter, D, of 0.0254 m, the effluent velocity was 0.13 m/s.
Thus, the ratio of effluent velocity, w, to local velocity, UH, was much less
than 1.0 for all cases (see Table 1) as required for momentum effects to be
neglible. Buoyancy due to the hydrocarbon tracer was also negligible; the
ratio of effluent density to ambient density, ps/Pa» was 0-98 for all cases.
Samples were drawn through a 0.002 m diameter sampling probe and passed
through a flame ionization detector to determine tracer concentrations. The
ambient hydrocarbon background was measured before and after each run. Since
the background variation was snail, it was assumed to be linear with time to
compute values to correct the raw data. The output of the flame ionization
detector was digitized and processed on a minicomputer. One minute averages
at a sampling rate of 2000 samples per second were used.
-------
SECTION 4
EXPERIMENTAL RESULTS
Velocity and turbulence measurements were made to evaluate how well
the model flow represented the atmospheric boundary layer. Measurements
were also made at several positions in the field downwind of the building.
The mean freestream speed was equal to 2 m/s (except for Case 8).
Profiles of the mean velocity measured over the release point with the
building removed and at two locations downwind of the building are shown in
Figure 3. The mean velocity profile in the approach to the building (x=0)
is handily approximated by a one-seventh power law. In the far-wake region
of the cubical building (x=10H) the velocity profile has nearly recovered
to tiie approach profile. A one-seventh power law profile is typical of flow
over flat rural topography with low shrubbery (Davenport, 1963). Although
hot-film anemometer measurements are unreliable near the building where tur-
bulence leveTs are high and flow reversals occur, severe distortion of the
mean velocity profile caused by the building is observed in the profile at
x=2H.
Longitudinal turbulence intensities of the approach flow were also com-
parable to field observations made in a neutral atmosphere. The maximum
local turbulence intensity in the approach flow, vV
-------
Castro and Robins (1975) reported on the dependence of the charac-
teristics of the cavity on the thickness of the approach flow boundary
layer. Their surface pressure and velocity measurements indicated that
the flow reattached on a cubical building roof at x - -0.25H, for the
type of boundary layer used in the present study. Observations of the
dispersing smoke in the present study did not substantiate their claim.
The time-averaged photographs are similar to those of Yang and Meroney
(1970) that did not show a reattached flow.
Smoke visualization studies provide insight into gross features of
the dispersion of contaminants but should not be used to make detailed
inferences. When photographed, lighting and exposure differences can
make the same situation appear quite differently. Hence, concentration
measurements of a tracer released from the source must be relied upon for
the quantitative evaluation and comparison of dispersion.
All measured concentrations, C, were made non-dimensional with the
source emission rate, Q, according to the corresponding building height,
H, and mean wind speed at that height, IL;
x = CUH H2/Q.
This should be kept in mind when comparing values for different building
heights. To avoid interference with the surface roughness slats, a height
of z = 0.025 rn was taken as "ground-level" for concentration measurements.
Coordinates (x, y, z) are defined in Figure 1.
Of the building shapes considered, a cubical building is the nost ele-
mentary. Figure 5 demonstrates the behavior of a roof centered release
from a cubical model when the wind is perpendicular to a side (Case 1).
Reverse flow just above the ton of the cube advected some of the effluent
to the upwind edge of the roof. The height of the dispersion envelope deter-
mined from the smoke visualization was approximately 1.5H over the building
and approximately 2.OH at a downwind distance of 5H. Vertical concentration
profiles (Figure 6) indicate similar envelope heights. The maximum concentra-
tion for the vertical profile at a downwind distance of x = 2H occurred at an
elevation of z = H. At distances farther downwind, the peak value indicated
by each profile moved closer to the ground. The lateral spread of the plume
can be observed to be only slightly greater than the building width in the
top view of the visualization (Figure 7). Measured concentrations at x = 2H
and x = 5H show the plume width (plume edge taken as point where C = 0.1C )
to be approximately 2.5H and 3.5H, respectively (Figure 8).
A longitudinal, ground-level profile along the plume centerline (Figure 9)
indicated a maximum ground-level concentration of x = 0.43 at a distance of
x = 5H. Robins (1975) reports values for a case similar to Case 1. His
results, converted to the scaling used here, show a maximum ground-level
-------
concentration x = 0.31 to occur at a distance of x = 3H. The exit momentum
ratio for his model was 6 times that used here and his velocity ratio
was lower than in this study.
A different wind direction produces a significant change in the way the
roof-top emissions disperse. When a corner of the cube was pointed into the
wind (Case 2, Figure 10), flow reversals atop the building did not occur as
in Case 1. Emissions were swept directly over the downwind building corner
with no reverse flow over the building roof (Figure 11) and more smoke reached
ground-level immediately downwind of the building than was observed for Case 1.
With this building orientation, the dispersion envelope did not extend far
above the building, nor did it grow rapidly with increasing downwind distance.
The vertical concentration profiles illustrate the differing characteristics
(See Figure 6 and 12). Maximum concentrations for each profile occurred at
ground-level for distances greater than x = 5H. However, for the profile at
x = 2H, the maximum concentration occurred at an elevation of approximately
z = 0.66H. The smoke envelope was wider for Case 2 than for Case 1 (at x =H,
the observed width was 1.5H for Case 1 and 2H for Case 2). This is in part
due to the increased projected area of the building into the wind. The flow
separated at the corners of the building (Figure 11) to produce standing
vortices which also increased the width of the dispersion envelope. Ground-
level concentrations were much higher for Case 2 than for Case 1 up to about
x = 5H but they were nearly enual by x = 10H (Figure 9).
For cases where cubical buildings are oriented at other angles to the
mean flow, emissions may be expected to produce concentration levels that are
between those of Case 1 and Case 2. Case 9 was such a configuration with
the building rotated to an angle 6 - 22.5°. The plume for this case was
found to be slightly skewed about the line y = 0 (Figure 13). The longitudi-
nal concentration profile (Figure 9) and the vertical profiles (Figure 14)
were measured directly downv.ind of the release point. In Figure 9 one can
see that near the building the concentration levels for case 9 fell between
those of Cases 1 and 2. Farther downwind, concentrations for all three cases
approached the same levels. The vertical profiles of Case 9 (Figure 14)
were similar to those of Case 2 beyond x = (JH.
Case 8 was a repeat of Case 1 at twice trie freestrean velocity. It was
performed as a check on the Reynolds number independence of the dispersion.
A longitudinal, ground-level concentration profile (Figure 15) and a verti-
cal profile at x = 5H (Figure 6) are practically identical to the corresponding
profiles for Case 1. The close agreement of these profiles indicates that the
dispersion field did not change substantially with a doubling of the building
Reynolds number.
A block with its width, W, twice the other dimensions (Case 3), produced
an envelope of somewhat greater vertical extent (Figure 16) than for Case 1.
For Case 3 the vertical concentration profiles (Figure 17) had shapes similar
-------
to those for Case 1. However, concentrations below the building height
(z=H) were lower by approximately 20 percent. This effect can be attributed
to more lateral and vertical spreading of the plume. The ground-level concen-
trations along the plume centerline (Figure 15) were lower by about 30 percent
at locations far from the building (x > 5H). The maximum ground-level value
was about the same as for Case 1, but it occurred closer to the building.
Case 4, a model building with its length in the direction of the wind
twice the other dimensions, had a similar vertical bound on the envelope
(Figure 18). The vertical profiles for Case 4 (Figure 19) were also similar
to those for Case 1. Concentrations near the plume centerline for the x = 2H
profile were only slightly higher for Case 4. The ground-level concentrations
along the plume centerline were only slightly lower than the Case 1 (Figure 15),
The building that was twice as high as its other dimensions, Case 5, had
an envelope directly over the building (Figure 20) which was similar to that
of Case 1; but its height did not increase downwind of the building. Also
illustrated by Figure 20 is that little of the smoke reached the ground
immediately downwind of the building. With this building rotated 45° (Case 6,
Figure 21), however, a considerable amount of the smoke reached the ground.
The flow observed over the building roof for Case 6 was similar to that for
Case 2. The emissions were swept directly over the downwind corner of the
roof into the cavity of reverse flow.
Case 7 (Figure 22), involved the use of a building three times as high
as the building used in Case 1, and had about the same growth in vertical
bound as Case 5. Essentially no smoke reached the ground near the building.
Near building, ground-level concentrations for Cases 5 and 7 were not high
enough above background to obtain reliable measurements. This was in agree-
ment with the smoke visualization photographs.
Vertical profiles for Cases 5 and 6 are shown in Figures 23 and 24. With
this taller building rotated 45°, concentrations aloft at x = H were roughly
three times those that occurred with the building perpendicular to the wind.
The vertical profile flattens out so that the maximum occurs at the ground
level near x = 2H for Case 6 but not until after x = 4H for Case 5. The
vertical profiles for Case 7 (Figure 25) were similar to those for Case 5.
-------
REFERENCES
Castro, I. P. and A. G. Robins (1975): The Effect of a Thick Incident Boundary
Layer on the Flow Around a Small Surface Mounted Cube. Marchwood Engineering
Laboratories Report R/M/N795, Central Electricity Generating Board, England.
Counilian, J. (1969): An Improved Method of Simulating an Atmospheric Boundary
Layer in a Uind Tunnel. Atrn. Env. 3^ p. 197-214.
Davenport, A. G. (1963): The Relationship of Wind Structure to Wind Loading.
Paper 2, Proc. of Conf. on "Uind Effects on Buildings and Structures", (Nat.
Phys. Lab., Teddington, Middlesex, June 1963) HMSO, London, 1965, p. 54-102.
Golden, J. (1961): Scale Model Techniques. M. S. Thesis, College of Engr.,
NYU, May.
Halitsky, J. (1968): Gas Diffusion Near Buildings, in Meteorology and Atomic
Energy, 1968, edited by David H. Slade, p. 221-255.
Harris, R. I. (1968): Measurement of Wind Structure at Heights up to 598 ft.
above Ground-Level, Symp. Wind Effects on Buildings and Structures, Loughbo-
rough Univ. Tech. (Dept. of Transport Technology).
Huber, A. H. and W. H. Snyder (1976): Building Wake Effects on Short Stack
Effluents. Preprint Volume of the American Meteorological Society 3rd Symposium
on Atmospheric Turbulence, Diffusion and Air Quality. Raleigh, NC, October 19-
20, 1976.
Robins, A. G. (1975): Plume Dispersion in the Vicinity of a Surface Mounted
Cube. Marchwood Engineering Laboratories Report R/M/R220, Central Electricity
Generating Board, England.
Snyder, W. H. (1974): Fluid Modeling Program of the Meteorology Laboratory,
U.S. Environmental Protection Agency. Chapter 31 of Proceedings of the Fifth
Meeting of the Expert Panel on Air Pollution Modeling. NATO-CCMS, N. 35.
Snyder, W. H., R. S. Thompson, and R. E. Lawson, Jr. (1977): The EPA Meteoro-
logical Wind Tunnel: Design, Construction, and Operating Details. Environmental
Protection Agency, Research Triangle Park, NC (In preparation).
Yang, B. T. and R. N. Meroney (1970): Gaseous Dispersion Into Stratified Building
Wakes. FDDL Technical Report, CSU, CER70-71BTY RNM-8.
-------
Table 1. GEOMETRY AND DEFINING PARAMETERS FOR EACH CASE
Case
1
2
3
4
5
6
7
8
9
Shape, m
L
0.18
0.18
0.18
0.36
0.18
0.18
0.18
0.18
0.18
W
0.18
0.18
0.36
0.18
0.18
0.18
0.18
0.18
0.18
H
0.18
0.18
0.18
0.18
0.36
0.36
0.54
0.18
0.18
Parameter
e,degrees
0
45
0
0
0
45
0
0
22.5
H/6
0.1
0.1
0.1
0.1
0.2
0.2
0.3
0.1
0.1
Uu,m/sec
n
1.46
1.46
1.46
1.46
1.62
1.62
1.70
2.92
1.46
w/UH
0.090
0.090
0.090
0.090
0.081
0.081
0.078
0.045
0.090
UHH
Pr- H
Kc
V
17,500
17,500
17,500
17,500
38,900
38,900
61 ,000
35,000
17,500
a L = length, W = width, H = height.
b 0 = building"orientation to wind, H/6 = ratio of building
height to boundary layer height, IL = mean wind speed at
building height, w/U,, - ratio of effluent velocity to local
velocity, Re - Reynolds number.
10
-------
TOP
SIDE
Figure 1. Top and side views of building model.
-------
Figure 2. Cubical model in wind tunnel test section during flow visualization
experiment.
12
-------
0.6
0.5
0.4
0.3
0.2
0.1
o x-0 (NO BUILDING)
a x = 2H
A x=10H
U/U =(z/5)1/7
CUBICAL
-BUILDING
HEIGHT
I I
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Figure 3. Mean velocity profiles at location of cubical building without building
in position (x=0) and downwind of building (x=2H, 10 H).
13
-------
0.6
0.5
0.4
0.3
0.2
0.1
CUBICAL
-BUILDING
HEIGHT
1 I
ox = 0(l\IO BUILDING)
Dx = 2H
A FIELD DATA (HARRIS, 1968)
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35
U
Figure 4. Local turbulence intensity at location of cubical building without
building in position (x=0) and downwind of the building (x=2H).
14
-------
Figure 5. Flow visualization of Case 1.
15
-------
3.5
2.5t>-
1 T
o CASE1,x = 2H
D CASE1,x = 5H
A CASE1,x = 8H
CASE 8, x = 5H
- \\\.
0.75 1 1.25 1.5 1.75
Figure 6. Vertical concentration profiles for Cases 1 and 8. (y = 0.)
16
-------
Figure 7. Flow visualization of Case 1 (top view).
17
-------
x 0.2
-2.5 -2 -1.5 -1 -0.5 0 0.5 1 1.5 2
Figure 8. Lateral ground-level concentration profiles for Case 1.
-------
2.5
x 1.5
0.5
oCASEI
a CASE 2
a CASE 6
CASE 9
10
x/H
12
14
16
18
20
Figure 9. Ground-level concentrations for Cases 1, 2, 6, and 9. (y = 0; z = 0.025 m.)
19
-------
Figure 10. Flow visualization of Case 2.
20
-------
Figure 11. Flow visualization of Case 2 (top view).
-------
2.25
1.75
1.25
0.75
0.25
Figure 12. Vertical concentration profiles for Case 2. (y = 0.)
22
-------
1.6
1.4
1.2
0.8
0.6
0.4
0.2
3
2
0
y/H
Figure 13. Lateral ground-level concentration profiles for Cases 1, 2, 3, 4, 6,
and 9. (x = 0.9 m.)
23
-------
2.25
1.75
1.25
0.75
0.25
Figure 14. Vertical concentration profiles for Case 9. (y = 0.)
24
-------
0.45
0.4
0.35
0.3
0.25
0.2
0.15
0.1
0.05
0
oCASE1
aCASE 3
A CASE 4
CASE 8
I I
0
10 12
x/H
14 16 18 20 22
Figure 15. Ground-level concentrations for Cases 1, 3, 4, and 8. (y = 0; z = 0.025 m.)
25
-------
Figure 16. Flow visualization of Case 3.
26
-------
2.5
o x = 2H
a x = 5H
x = 8H
0 0.15 0.3 0.45 0.6 0.75 0.9 1.05 1.2 1.35
Figure 17. Vertical concentration profiles for Case 3. (y = 0.)
27
-------
Figure 18. Flow visualization of Case 4.
28
-------
2.25
1.75 -'
1.25
0.75
0.25
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 2.2
Figure 19. Vertical concentration profiles for Case 4. (y = 0.)
29
-------
Figure 20. Flow visualization of Case 5.
30
-------
Figure 21. Flow visualization of Case 6.
31
-------
Figure 22. Flow visualization of Case 7.
32
-------
1.75
1.25
o x=H
D x = 2.5H
= 4H
0.75
0.25
Figure 23. Vertical concentration profiles for Case 5. (y = 0.)
33
-------
1.75
1.5
ox =
o x = 2.5H
1.25
0.75
0.5
0.25
2.5
7.5
10 12.5 15 17.5
Figure 24. Vertical concentration profiles for Case 6. (y = 0.)
34
-------
1.75
1.5
1.25
0.75
0.5
0.25
ox = 0.66H
nx = 1.6GH
A x = 2.66H
1.5
4.5
7.5
10.5 12
Figure 25. Vertical concentration profiles for Case 7. (y = 0.)
35
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1 REPORT NO
EPA-600/4-77-006
3 RECIPIENT'S ACCESSI ON-NO.
4 TITLE AND SUBTITLE
BUILDINGS
uF KUUF-TuP EMISSIONS FKOM ISOLHTLJ
A Wind Tunnel Study
5 REPORT DATE
January 1977
6. PERFORMING ORGANIZATION CODE
7. A'JTHOR(S)
i\oaer 5,
David J.
Tnoiiipson > ESRL, EPA
Loiiiuarui, Nortnrop Services,
Inc.
8. PERFORMING ORGANIZATION REPORT NO.
Fluid Modeling Report
No. 3
9 PERFORMING ORGANIZATION NAME AND ADDRESS
environmental Sciences Researcn Laboratory
office of Research and Development
U. S. environmental Protection Agency
Research Triangle Park, North Carolina 27711
10 PROGRAM ELEMENT NO.
1AA603
11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
environmental Sciences Research Laboratory
Office of Research and Development
U. S. Environmental Protection Agency
Research Triangle Park. North Carolina 27711
13. TYPE OF REPORT AND PERIOD COVERED
In-house
14 SPONSORING AGENCY CODE
EPA-ORD
15. SUPPLEMENTARY NOTES
16. ABSTRACT
A fluid modeling study of the dispersion of roof-top emissions from
rectangular uuilciings was performed in the meteorological wind tunnel of
the LPA Fluid Modeling Facility. The basic building shape was a 0.18 meter
cuoe. Variations included a building twice as wide and buildings twice
anu tnree times as nigh. Each building was placed in a 1.8 meter, simulated,
neutral atmospneric boundary layer. Low momentum, non-buoyant emissions
were released tnrougn an opening at the roof center. Photographs of flow
visualization experiments are presented for a qualitative evaluation of
tne oinluing wakes. Concentration measurements at ground level and aloft
are presented for eacn case at downwind distances within 20 building heights.
concentrations are presented in a non-dimensional form for ease in application
to full scale situations. Lie orientation of the wind to the building was
found to significantly affect ground-level concentrations near the building.
a 4o° orientation of the cubical building increased the maximum ground-level
concentration uy a factor of six over that observed for a perpendicular
orientation.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
*Air pollution
^Atmospheric diffusion
Air flow
*boundary layer
*buildings
*£nvironmental simulation
wind tunnel
b.IDENTIFIERS/OPEN ENDED TERMS
c. COSATl 1 ield/Group
13B
04A
20D
13M
14B
13 DISTRIBUTION STATEMENT
RELEASE TO PUbi_IC
19 SECURITY CLASS (This Report)
UNCLASSIFIED
21 NO. OF PAGES
44
20 SECURITY CLASS (This page)
22. PRICE
EPA Form 2220-1 (9-73)
36
-------
Q) to -. -»
&«-»> «
o 3 x Q)
S 5' o ^
">
s SB'
ris °
I1~ st
»i
0) O
t>
t1$
O -n -n
"2
§^ r
5 5 0)
c < c
z > en
3 m Z
< ,- m
^ 0) ^
s Sy
TJ -
r &
0 o
< o
m o
CO
T5
0)
O
-n
OB o
o c
o i
3T 3"
l
n
(A
(fi
p>
r>
(D
CD
O 2.
c:
00
O rn
3; Z
o <
CD 33
o O
5 33 -
o' « rn
QJ W
Q)
3
CD
0)
-^
O
Z
H
>
O ° ^ -D
? 3 « :o
°' s. Q. g
oi 3 ° rn
w 3 « O
05 H? CD ^
"ifo
-h "D '-
1 >
5 o
x rn
z
o
u
u
c
01
m
Z
o o
I 3
m
m H
D
r
T)
5 -Tl
O rn
H rn
rn
Q
rn
Z
O
------- |