Environmental S<.ieriqj?s
Laboratory \
Research Triangle Parfc
/
-
tlT
Determination of
Good-Engineering-
Practice Stack
Height
A Fluid Model
Demonstration
Study for a Power
Plant
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are
1 Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4 Environmental Monitoring
5 Socioeconomic Environmental Studies
6 Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8 "Special" Reports
9 • Miscellaneous Reports
This report has been assigned to the ECOLOGICAL RESEARCH series. This series
describes research on the effects of pollution on humans, plant and animal spe-
cies, and materials Problems are assessed for their long- and short-term influ-
ences Investigations include formation, transport, and pathway studies to deter-
mine the fate of pollutants and their effects. This work provides the technical basis
for setting standards to minimize undesirable changes in living organisms in the
aquatic, terrestrial, and atmospheric environments
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161
-------
EPA-600/3-83-024
April 1983
DETERMINATION OF GOOD-ENGINEERING-PRACTICE STACK HEIGHT
A Fluid Model Demonstration Study for a Power Plant
by
Robert E. Lawson, Jr.
and
William H. Snyder
Meteorology and Assessment Division
Environmental Sciences Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
ENVIRONMENTAL SCIENCES RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NC 27711
-------
DISCLAIMER
This report has been reviewed by the Environmental Sciences
Research Laboratory, U.S. Environmental Protection Agency, and
approved for publication. Approval does not signify that the contents
necessarily reflect the views and policies of the U.S. Environmental
Protection Agency, nor does mention of trade names or commercial
products constitute endorsement or recommendation for use.
The authors, Robert E. Lawson, Jr. and William H. Snyder, are physical
scientists in the Meteorology and Assessment Division, Environmental Sciences
Research Laboratory, U.S. Environmental Protection Agency, Research Triangle
Park, NC. They are on assignment from the National Oceanic and Atmospheric
Administration, U.S. Department of Commerce.
ii
-------
PREFACE
This report was prepared for the purpose of demonstrating the application
of a fluid modeling approach to the determination of good-engineering-practice
stack height. The approach follows the recommendations set forth in the Guide-
line for Use of_ Fluid Modeling tc> Determine Good Engineering Practice Stack
Height (EPA, 1981).
-------
ABSTRACT
A study using fluid modeling to determine good-engineering-practice (GEP)
stack height for a power plant installation is discussed. Measurements are
presented to describe the simulated boundary layer structure, plume-dispersion
characteristics in the absence of the model plant building, and the maximum
ground-level concentration of effluent downstream of the source, both with and
without the model plant building. Analysis of the maximum ground-level concen-
trations shows that, in this case, a stack height of 64.1 m meets the current
GEP criteria for 100% plant-load conditions.
IV
-------
CONTENTS
Abst ract i v
Figures vii
Symbol s i x
Acknowledgements xi
1. Introduction 1
2. Technical Approach 3
3. Examination Of Topography, Meteorological Parameters,
And Selection Of The Area To Be Modeled 4
3.1 Topography 4
3.2 Meteorological parameters 4
3.3 Selection of modeled area 5
4. Evaluation And Justification Of Modeling Criteria 7
4.1 Similarity criteria 7
4.2 The model 9
5. Evaluation Of Simulated Boundary Layer 11
5.1 Boundary layer simulation 11
5.2 Dispersion comparability tests 13
6. Determination Of GEP Stack Height 16
6.1 Dispersion in the absence of the building 17
6.2 Dispersion in the presence of the building 21
6.3 Determination of GEP stack height 23
6.4 Plume rise 26
6.5 Discussion of results 27
7. Summary 31
References 32
Table 1 34
Appendices
A. Description Of Facilities And Instrumentation 55
A.I The Fluid Modeling Facility Wind Tunnel 55
A.2 Instrumentation 56
A.2.1 Velocity measurements 56
A.2.2 Concentraction measurements 56
A.2.3 Data acquisition system 57
A.2.4 Volume flow measurements 58
-------
CONTENTS (continued)
Appendices
B. Concentration Measurements For Stack Heights
Of 54.2 m, 68.8 m, 72.3 m, And 90.3 m „ 61
C. GEP Stack Height For 50% Plant-Load Conditions 67
D. Raw Data Listings 70
-------
FIGURES
NUMBER PAGE
1 Topography, meteorological tower locations, and
area modeled 35
2 Wind frequency distributions 36
3 Cumulative frequency distribution of wind speeds
for northwest winds under neutral stability
(valley tower location) 37
4 Vertical temperature profile in wind tunnel test
section 38
5 Top and side views of the building 39
6 Schematic of the boundary-layer simulation system 40
7 Velocity profiles for the simulated atmospheric boundary
layer 41
8 Turbulence intensity and Reynolds stress profiles 42
9 Lateral uniformity of mean velocity (a) and longitudinal
turbulence intensity (b) 43
10 Surface concentration profiles (A) compared with
Pasquill-Gifford C and D stability 44
11 Vertical concentration profiles compared with Pasquill-
Gifford C stability 45
12 Lateral concentration profiles compared with Pasquill-
Gifford C stability 46
13 Plume widths compared with Pasquill-Gifford curves .... 47
14 Flow visualization with and without the primary facility
model (paraffin-oil smoke source).
Hs = 64.1 m, 100% plant load 48
15 Vertical profiles of mean velocity and longitudinal
turbulence intensity downstream of the model
building 49
vn
-------
FIGURES (continued)
NUMBER PAGE
16 Vertical profiles of vertical turbulence intensity (a)
and Reynolds stress (b) downstream of the model building ... 50
17 Surface concentration profiles with (A) and without
(n) the building. Stack height 64.1 m 51
18 Vertical concentration profiles with (A) and without
(n) the building. Stack height 64.1 m 52
19 Vertical concentration profiles with (A) and without
(n) the building. Stack height 64.1 m, downstream
distances of 1.5 km and 1.7 km respectively , 53
20 Lateral concentration profiles with (A) and without (n) the
building. Stack height 64.1 m, downstream distances of 1.5 km
and 1.7 km respectively 54
-------
SYMBOLS
d displacement height [L]
D stack diameter [L]
H structure or obstacle height [L]
HB building height [L]
Hg GEP stack height [L]
Hs stack height [L]
L lesser dimension (height or width) of structure [L]
Q tracer volumetric flow rate [L^/T]
Reg building Reynolds number
Rej effluent Reynolds number
u1 streamwise fluctuating velocity [L]
u* friction velocity [L]
U wind speed [L/T]
UB wind speed at top of building [L/T]
Us wind speed at stack height [L/T]
U „ free-stream wind speed [L/T]
W stack effluent exit velocity [L/T]
W vertical fluctuating velocity [L/T]
x streamwise coordinate [L]
y cross-stream coordinate [L]
IX
-------
SYMBOLS (continued)
z vertical coordinate [L]
z0 roughness length [L]
6 boundary layer depth [L]
e gravel size [L]
v kinematic viscosity [L2/T]
Ps effluent density [M/L3]
Pa density of ambient air [M/L3]
ay horizontal dispersion parameter [L]
oz vertical dispersion parameter [L]
x concentration [M/L3]
X1 concentration at ground level [M/L3]
-------
ACKNOWLEDGMENTS
The assistance and cooperation of the following individuals are
gratefully acknowledged: Mr. R.D. Jones for his painstaking efforts in
collecting the data; Mr. Alan Huber for his many helpful discussions and
comments; Ms. Carolyn Coleman and Ms. Eileen Ward for their patience in
typing and assembling this report. Special thanks are due the Tennessee
Valley Authority for providing the meteorological and plant operations
data on which this demonstration study was based.
-------
SECTION 1
INTRODUCTION
Section 123 of the Clean Air Act Amendments of 1977 defines Good-Engi-
neering-Practice (GEP) stack height as "the height necessary to insure that
emissions from the stack do not result in excessive concentrations of any
air pollutant in the immediate vicinity of the source as a result of atmos-
o
pheric downwash, eddies and wakes which may be created by the source itself,
nearby structures or nearby terrain obstacles". The purpose of this study
was to determine the GEP stack height for a power plant installation using
fluid modeling techniques. The model was based on an existing facility, the
TVA Widows Creek Plant, for which plant operating conditions, meteorological
parameters, and detailed topographical maps were available. Almost every
installation will have features that are unique; topographical and meteoro-
logical parameters are the most common of these features. Nevertheless, the
fluid modeling approach is practical, and, if applied properly, should be
useful in power plant design.
The general working rule for GEP stack height is:
Hg = H + 1.5L
where Hg is the GEP stack height, H is the height of the structure or nearby
obstacle, and L is the lesser dimension (height or width) of the structure or
nearby obstacle. Regulations to implement Section 123 of the 1977 Clean Air
Act Amendments allow stack heights near structures as determined by the above
equation to be used in establishing an emissions limitation plan.
Fluid modeling techniques may also be used to determine GEP stack heights
1
-------
needed to prevent excessive pollutant concentration in the vicinity of the
source. The maximum ground-level concentration measured in a model that
includes nearby structures or terrain obstacles is termed "excessive" when
it is 40% or more in excess of the maximum ground-level concentration
measured in a model that does not include downwash, wake, or eddy effects
produced by the nearby structures or terrain. The basic document that stip-
ulates requirements for fluid modeling GEP studies is the Guideline for Use
of Fluid Modeling to Determine Good Engineering Practice Stack Height (here-
after referred to as the "Guideline") (EPA, 1981). A more detailed reference,
Guideline for Fluid Modeling of Atmospheric Diffusion (Snyder, 1981), provides
technical standards for evaluation of various aspects of this study.
-------
SECTION 2
TECHNICAL APPROACH
Bearing in mind that the height of the stack is creditable as 6EP if the
maximum ground-level concentration in the presence of the nearby building or
obstacle is 40% greater than that measured in its absence, the ultimate
objective of this study is to simply examine maximum ground-level concentra-
tions as a function of stack height, in the presence and absence of a nearby
structure or terrain obstacle. Other criteria specified in the Guideline
must be met in order to validate the fluid-modeling approach. Certain steps
specified in the Guideline must be followed when conducting any GEP fluid
modeling study:
1. Examination of the topography and meteorological parameters,
and selection of the area to be modeled.
2. Evaluation and justification of modeling criteria.
3. Evaluation of the test facility in the absence of buildings,
other surface structures, or large roughness and/or elevated
terrain.
4. Determination of the GEP stack height.
5. Documentation of the facility operation, instrumentation
used in the study, and associated parameters.
These steps were followed and are reported in the following sections.
-------
SECTION 3
EXAMINATION OF TOPOGRAPHY, METEOROLOGICAL PARAMETERS,
AND SELECTION OF THE AREA TO BE MODELED
3.1 TOPOGRAPHY '
The plant is in a river valley which extends southwest to northeast
(figure 1); the river is southeast of the plant. A prominent ridge is
located across the river, approximately 1.6 km southeast of the plant, and
it parallels the river. This ridge rises rather abruptly (15° slope) to a
plateau 250 m higher than the plant. To the northwest of the plant lies
an area of gently rolling hills that extends approximately 7 km to an
irregular plateau of about 300 m. The surface in this area is characterized
by stands of pine trees and agricultural fields. The primary plant struc-
ture is a semi-rectangular building 36.3 m high, 159.5 m long, and 75 m wide.
The longest dimension of the structure is parallel to the river (i.e., on a
southwest-to-northeast line). Surrounding the building is generally flat
terrain interrupted by several small auxiliary control buildings and an
electrical distribution area. A second plant building is approximately 400 m
northeast of the primary structure. The stack in question is between the
primary structure and the river, 84.5 m southeast of the primary structure.
3.2 METEOROLOGICAL PARAMETERS
Meteorological data for one year were used to determine the air flow and
stability characteristics in the plant vicinity. The locations of the metero-
logical monitoring towers are shown in Figure 1. The valley-site tower is
about 1.2 km southwest of the plant; the mountain-site tower is approximately
-------
4 km southeast of the plant, and is situated on the plateau. Wind speed,
wind direction, and temperature data were recorded on the towers at heights
of 10 m and 61 m. The data available from these towers consisted of joint
frequency distributions of wind speed and direction, by stability class.
The wind frequency distributions for all stability categories and for neutral
stability for both valley and mountain locations are presented in Figure 2.
The mountain site data reflected a reasonably uniform distribution for all
stability classes; for neutral conditions, the predominant wind direction
was from the northwest. The valley site data demonstrated the strong influ-
ence of the valley in channeling the air flow; the predominant wind for all
stability classes was along the valley. In neutral stability, there was
again a strong component along the valley, with a secondary maximum for
northwest winds. Analysis of the climatological data by Hanna (1980) showed
the roughness length characteristic of the upstream fetch to be approximately
1.0-1.5 m for the mountain site; the valley site exhibited values of 0.7-1.6 m
for flow along the valley and 0.2-0.6 m for flow normal to the valley. Hanna
also pointed out that these values are probably representative of the fetch
out to a distance of about 600 m from the plant building, and are reasonable
considering the type of surface features surrounding the tower sites.
3.3 SELECTION OF MODELED AREA
The area within 100 m of the primary plant building was modeled in detail.
The terrain outside this 100-m radius was modeled with surface roughness
elements to a distance of 3.5 km upstream and downstream, and 0.8 km either
side of the primary plant building. This area is outlined on Figure 1. The
study was conducted with northwest winds under conditions of neutral atmospheric
stability.
-------
The building effects and ridge effects represent two different areas
of study. In the present study, only the building effects during high-
wind-speed, neutral conditions were examined. This limits the demonstra-
tion to situations without the complication of downwind terrain. In addi-
tion, proper scaling of the ridge southeast of the plant would have lead
to a very small building model. This would have introduced the complica-
tion of small building Reynolds number. To determine the effects of plume
dispersion and possible impingement on downwind terrain, further research
will be necessary.
The northwest wind direction was chosen because it is normal to the
largest dimension of the building. Under this condition, the dimensions
of the building wake are greatest, and, hence, the downwash effect due
to the building is maximized (Snyder and Lawson, 1976). The free-stream
wind speed was selected by plotting the cumulative frequency distribution
of wind speed for northwest winds under neutral stability (Figure 3).
According to the Guideline, the design wind speed should be less than
the speed that is exceeded less than 2% of the time for the given wind
direction. This 98th-percentile wind speed is 8.0 m/s for the 61-m valley
tower site. This corresponds to a free-stream wind speed of 11.6 m/s
(Section 4.2). The fetch which characterizes flow from the northwest is
representative of gently rolling hills covered with stands of pine and
agricultural fields. The surface roughness length is on the order of 0.2
to 0.6 m (Hanna, 1980). This appears consistent with the values of surface
roughness length provided in the Guideline for surface types between pal-
metto and pine forest. Again referring to the Guideline, this range of
7 7
surface roughness lengths corresponds to a u* /Uoo ratio from 0.0023 to
0.0026 and a power law index from 0.18 to 0.22.
-------
SECTION 4
EVALUATION AND JUSTIFICATION OF MODELING CRITERIA
4.1 SIMILARITY CRITERIA
As specified in the Guideline there are five parameters in addition
to geometric similarity that are relevant to modeling atmospheric flow.
These are the Rossby number, Peclet number, Reynolds-Schmidt product,
Froude number, and Reynolds number.
The Rossby number can be ignored here, because it represents the
effects of the Coriolis force and is significant only when modeling
effects greater than 5 km downstream from the source. The maximum down-
stream distance from the source that was modeled in the present study
ranged from 3 to 4 km.
The Peclet number and Reynolds-Schmidt product are indicators of the
importance of turbulent diffusivity, compared with molecular diffusivity
(thermal and mass diffusivities, respectively). According to the Guide-
line, thermal and mass diffusivities are assumed to be negligible if the
Reynolds number is high enough that advection and large scale turbulent
motions are the primary mechanisms for dispersion.
The Froude number indicates the relative importance of inertial and
buoyant forces. There are two Froude numbers that must be considered:
the Froude number of the flow in the wind tunnel and stack Froude number.
To model a neutrally stable (adiabatic) atmospheric flow in the wind tun-
nel, the Froude number of the flow in the test section must be infinite.
This is equivalent to requiring isothermal flow in the tunnel. Figure 4
-------
shows the vertical profile of temperature in the wind tunnel test section
used in the study. The slight temperature gradient is of the order of
-0.3°C in the lowest half-meter. This corresponds to a Froude number of
approximately 57 which, for practical purposes, is neutral flow. The
stack Froude number is an indicator of the buoyancy of the effluent in
the ambient air.. For precise scaling of buoyant releases, the Froude
number of the model stack must match the Froude number of the prototype.
The Guideline specifies that the ratios of stack diameter to building
height, effluent density to ambient air density, and efflux speed to
crosswind speed be matched between the model and prototype. The Guide-
line does not require that ratios of effluent buoyancy be matched. Hence,
the stack Froude number was ignored.
The Reynolds number is the ratio of inertial to viscous forces acting
on an air parcel. For a given fluid, strict matching of the model and
prototype Reynolds numbers requires that the reference velocity be in-
creased in direct proportion to the decrease in scale. As this is imprac-
tical for large reductions in scale, the principle of Reynolds-number
independence is invoked to enable modeling under such conditions. Basi-
cally, the principle of Reynolds number independence states that the
pattern of turbulent flow is similar at all sufficiently high Reynolds
numbers. Two Reynolds numbers were considered in this study, the build-
ing Reynolds number (Reg) and the effluent Reynolds number (Res).
ReB = UBHB/v
where Ug is the wind speed at the top of the building, HB is the building
height, and v is the kinematic viscosity of air. For sharp-edged
buildings, the critical building Reynolds number, according to the Guide-
line, is 11,000. For the present model, which was sharp-edged, the building
8
-------
Reynolds number was 13,400; hence, demonstration of Reynolds-number
independence was not required.
The effluent Reynolds number is defined as:
Res = WsD/v
where Ws is the efflux velocity, D is the stack diameter, and v is the
kinematic viscosity of the effluent. A sufficiently high effluent Reynolds
number ensures that the effluent is turbulent; effluent from full-scale
stacks is almost always turbulent. In the present study, the effluent
Reynolds number of the model stack was approximately 8000 for full-load
and 4000 for half-load plant operating conditions. The Guideline speci-
fies that the effluent Reynolds number should preferably exceed 15,000,
and that if it is below 2000, the flow should be tripped to induce turbu-
lence. Since the effluent Reynolds number fell between these two values,
the flow was tripped. A thin, internally serrated washer was inserted 10
stack-diameters upstream from the stack exit to ensure fully turbulent
effluent flow.
4.2 THE MODEL
Values of parameters in the prototype and in the model are listed in
Table 1. Figure 5 shows top and side views of the model building used in
the study. The scale ratio selected, 1/430, was based on several consid-
erations. Compromises were necessary to meet the opposing requirements
of high Reynolds number and the limited length of the wind tunnel test
section.
The other similarity criteria that were considered were the ratios
of roughness length to boundary-layer depth, and building height to bound-
ary-layer depth (Table 1). The model boundary layer had a depth of 0.9 m
-------
and a roughness length of 0.74 mm and fit a power law profile with an
index of 0.2 (Section 5.1). At a scale of 1/430, this provided a simu-
lated roughness length of 0.32 m and a simulated boundary-layer depth of
387 m. The boundary-layer depth was somewhat lower than the Guideline
recommendation, but it was consistent with examples shown by Davenport
(1963). Also, because the ratio of the boundary-layer depth to the
building height was large, the relatively thin boundary layer was not
expected to significantly influence the results (Snyder, 1981). As no
data were available on the full-scale boundary-layer depth, the rough-
ness length was used as the primary means of comparing prototype and
model. Measurements of turbulence spectra were not attempted; hence, no
data were available for determining integral scales for either the proto-
type or model.
The model building and stack diameters were scaled to be geometri-
cally similar to the prototype. Concentration measurements were initially
made without roughening the model surface, because separated flow was
expected to predominate near the sharp-edged model. In later tests, the
model was covered with small gravel of size e = 20 v /u*(~ 1.5 mm). No
differences in measured concentrations were observed due to roughening
the model. The effluent conditions to be modeled were based on full-load
plant-operating conditions, and both the effluent density ratio and the
effluent to wind speed ratio of the model were matched to the prototype.
Table 1 also lists the stack effluent parameters. The values of the
ratios of effluent to wind speed were in excess of 1.5, which means that
they were high enough to preclude stack downwash.
10
-------
SECTION 5
EVALUATION OF SIMULATED BOUNDARY LAYER
The arrangement evaluated in this section is the model flow in the
absence of buildings, other surface structures, or large roughness and/or
elevated terrain. The same arrangement, but with the plant building pre-
sent, was used in the GEP stack height determination test.
5.1 BOUNDARY-LAYER SIMULATION
The method used for generating a simulated atmospheric boundary
layer followed that of Counihan (1969) and is shown schematically in
Figure 6. The arrangement consisted of a castellated barrier at the
entrance of the test section, to trip the flow, followed by elliptical
wedge vortex generators (which aid in shaping the velocity and turbulence
intensity profiles), and surface roughness elements (which prescribe the
surface layer characteristics and aid in maintaining the boundary layer
once generated). Detailed geometry of the barrier, vortex generator, and
roughness scheme used to generate the boundary layer for this study can
be found in Castro and Snyder (1980). The surface roughness for the
boundary layer consisted of discrete blocks, 27 mm X 27 mm X 18 mm high
covering approximately 25% of the surface area. The advantage of the
barrier, generator, and roughness scheme is that it produces a thick
simulated boundary layer that develops very slowly after an initially
rapid development.
As indicated in Section 4.1, the flow in the test section was essen-
tially isothermal. The profile of Figure 4 is representative of condi-
11
-------
tions during all tests. Since the entire test facility was enclosed in a
temperature-controlled room, no significant temperature fluctuations
occurred during the tests.
Figures 7-9 show the measured boundary-layer characteristics. Figure
7a shows the development of the velocity profile between 6 m and 15 m
from the leading edge of the roughness. The development of the mean
velocity profile is obviously quite rapid initially and quite slow there-
after, as there is little variation beyond 6 m. Using a displacement
height.of 18 mm, the mean velocity profiles all fit very nicely to a 0.2
f i
power law. A semi logarithmic plot of the mean velocity profiles was used
to determine the roughness length and surface shear stress (Figure 7b).
All data are shown; however, only those points representative of approx-
imately the lowest 250 mm (100 m in the prototype) of the simulated
boundary layer were used to determine the best-fit log law. The roughness
length was found to be 0.74 mm, and the square of the friction velocity,
Q 9
u* , was 0.045 (m/s) . In Figure 8a, the longitudinal and vertical com-
ponents of turbulence intensity are plotted as functions of height. The
general shapes of the longitudinal turbulence intensity profiles are con-
sistent with examples in the Guideline, but the absolute values of inten-
sity are about 10% greater. The ratio of the vertical to the longitudinal
component in the surface layer was approximately 0.5, and this is consis-
tent with values typically found in the atmosphere. Figure 8b shows the
shear stress normalized by the surface shear stress determined from the
mean velocity profiles. Although there is considerable scatter, the data
tend to collapse near the surface around a value of 1. The difference
between this boundary layer and the naturally grown boundary layer
12
-------
presented in the Guideline is that this one had a thicker constant stress
region and greater shear stress in the early stages of development.
The model building was located 7 m downstream from the beginning of
the roughness. This assured a reasonably well-developed boundary layer and
a test section long enough to allow downstream measurements in the area of
the maximum ground-level concentrations. Lateral profiles of mean velocity
and longitudinal turbulence intensity (Figures 9a and 9b) were measured at
several heights in the area between the model and the end of the study area
(15 m). The greatest peak-to-peak variations occurred near the end of the
study area and are attributed to effects of the tunnel exit section.
5.2 DISPERSION COMPARABILITY TEST
To establish dispersion characteristics of the simulated boundary
layer, concentration profiles were measured in the downstream, lateral,
and vertical directions through a neutrally buoyant plume. The source
was a porous, sintered-bronze ball of 14 mm diameter. The ball was
supported on the upstream side by a small-diameter tube, to minimize the
influence of the support on the uniform tracer flow issuing from the ball.
Ethylene gas was used as the tracer. The ball is effectively a point
source. By using the ball instead of the standard stack, the problem of
determining the effective stack height (i.e. physical stack height plus
plume rise) was eliminated. Measurements were made with a model stack
height of 233 mm, corresponding to a full-scale stack height of 100 m.
The resulting measurements were converted to equivalent full scale concen-
o
trations in the form xUs/Q (m~ )> for comparison with dispersion estimates
using Pasquill-Gifford stability categories C and D (Turner, 1970). The
13
-------
vertical and lateral concentration profiles were used to calculate the
mass balance of tracer at downwind positions.
Figures 10 - 13 present the data for the (100 m) stack. Figure 10
shows ground-level concentrations measured downstream with an estimate
of the same using Pasquill-Gifford C and D stability classes. Pasquill-
Gifford stability category C provides the best fit to the data for posi-
tions near and upstream of the maximum ground-level concentration. The
comparison demonstrated that the boundary layer was slightly more turbu-
lent than desired. Figures lla and b show vertical profiles of concen-
tration at four downstream locations (0.5 km, 1.0 km, 2.0 km, and 3.5 km)
and predicted profiles, based on Pasquill-Gifford stability category C
and a 100-m effective stack height. The use of 100 m as the effective
stack height is supported by the vertical profile nearest the source.
Here again, data from near the source fit the Pasquill-Gifford curves
reasonably well. But, as the downstream distance increased, the values
departed significantly from the curves. The rapid mixing of the plume
into the surface layer is quite apparent. Lateral concentration distri-
butions were measured at heights corresponding to the peak concentration
found in the vertical profiles at 0.5 km, 1.0 km and 3.5 km from the
source (Figure 12). Again the data have been compared with predicted
concentrations using Pasquill-Gifford stability class C and 100-m effec-
tive stack height. The results are similar to those obtained for the
vertical profiles; the Pasquill-Gifford predictions fit reasonably well
near the source, but depart substantially farther downstream.
Using these vertical and lateral profiles together with the vertical
14
-------
profiles of velocity, the quantity of tracer passing through each down-
wind cross section per unit time was calculated from
// [C(y,z) U(z)/Q]dydz.
Ideally, this quantity should be near unity. Calculated values ranged
from 0.99 nearest the source to 0.88 at the greatest downstream distance.
Figure 13 shows the variation in vertical and lateral spread of the plume
(oz and ay, respectively) with distance. The values of the lateral and
vertical plume widths were derived from the concentration profiles by
assuming Gaussian and reflected-Gaussian distributions, respectively.
The solid lines represent Pasquill-Gifford stability categories C and D.
The values of both oz and ay closely approximated those obtained with C
stability, approaching D stability farther downstream.
In summary, the boundary layer dispersive characteristics were most
closely approximated by Pasquill-Gifford stability class C (i.e., slightly
unstable). The rate of decay beyond 1 km was slightly lower than that
estimated from Pasquill-Gifford, resulting in some broadening of the con-
centration peak, but having little effect on the peak value. These
results may be attributed, in part, to the larger roughness length and
elevated source position of the model as compared to the smaller roughness
lengths and ground-level sources most appropriate to estimates using the
Pasquill-Gifford scheme. The growth rates of Oz and Oy with downstream
distance were more consistent with those found by McElroy (1969) and Vogt
(1977) in studies of dispersion over urban areas with roughness lengths
on the order of 1 m.
15
-------
SECTION 6
DETERMINATION OF GEP STACK HEIGHT
The determination of GEP stack height was based on the effect of the
primary plant structure immediately upstream from the stack. The effect
of the building was initially examined by flow visualization using a par-
affin-oil smoke source. There was little observable difference in the
plume characteristics except for very low stack heights where the plume
was quite obviously directly entrained into the wake of the model building.
Figure 14 shows two representative photos of the flow visualization tests
with a stack height of 64.1 m and 100% plant load operating conditions.
Plume rise estimated from these photos is approximately 45 m at 3.5 build-
ing heights downstream, and 65 m at 12 building heights downstream.
The model flow in the absence of the building was the same as that
documented in Section 5; thus, no further measurements were needed to
characterize the background flow. Flow characteristics in the presence
of the model building are shown in Figures 15 and 16. Vertical profiles of
mean velocity, longitudinal turbulence intensity, vertical turbulence
intensity, and Reynolds stress were measured at three locations downstream
of the model. Due to the response characteristics of the hot-wire ane-
mometer, measurements taken in highly turbulent flows will reflect signif-
icant errors, and measurements taken in areas where flow reversal is
frequent will reflect gross errors. Such measurements can be used only
to compare areas of substantial flow distortion with and without the
model building in place.
16
-------
To find the GEP stack height, data were collected for four stack
heights, 54.2 m, 68.8 m, 72.3 m, and 90.3 m. Excess concentrations of
simulated pollutant were then analyzed and interpolated to find the stack
height corresponding to the GEP criterion of 40% excess concentration.
The GEP stack height was found to be 64.1 m. Documentation for this GEP
stack height is included in sections 6.1 - 6.4; additional data used to
determine^ the GEP stack height are included as Appendix B. Appropriate
justification has been provided where deviations from the Guideline were
deemed prudent.
6.1 DISPERSION IN THE ABSENCE OF THE BUILDING
The model stack was placed in the wind tunnel, and dispersion
characteristics were measured in the absence of the model building. A
model stack height of 149 mm was used, corresponding to a full-scale
stack height of 64.1 m or 1.8 building heights. The stack effluent
density ratio, ps/Pa» and the effluent-to-wind-speed ratio, WS/US, were
matched between prototype and model. A mixture of helium, air, and
ethylene was used to model the effluent; the ethylene served as the
tracer, and the helium reduced the density of the model effluent mixture
to 0.694. A two-minute sampling time with the free-stream wind speed of
4 m/s yielded reasonably stable average concentration values. Since the
ethylene initially tended to cool the effluent significantly, a heat
exchanger, consisting of a coil of copper tubing immersed in a container
of water, was used to maintain the ethylene flow at nearly constant tem-
perature. No heat exchanger was required for the air or helium flow.
All concentration measurements were again converted to form xUs/Q (m~2).
17
-------
Longitudinal, vertical, and lateral concentration profiles were
measured both with and without the model building, Figures 17 - 20.
These plots were combined in order to facilitate direct comparison of the
building's effects on downstream concentrations. Vertical concentration
profiles were measured at three downstream distances: the location of
maximum ground-level concentration, half-way between the source and the
ground-level maximum, and approximately 3.5 building heights downstream
of the source. Lateral concentration profiles were measured at the loca-
tion of the ground-level maximum, both at the surface and at the elevation
of the maximum concentration found from the corresponding vertical con-
centration profile.
Fewer vertical and lateral profiles were measured than specified by the
Guideline. This is justifiable because the terrain is uniform; hence the
plume is transported downstream with no significant lateral or vertical
departure from the source location other than that due to plume rise or
building effect. A model that included complex terrain features would
certainly require a greater number of measurements to isolate adequately
the location and value of the maximum concentration. The maximum ground-
level concentration value in this study has been determined beyond a
reasonable doubt.
The longitudinal ground-level concentration measurements without the
building (Figure 17) showed that the peak concentration occurred approxi-
mately 1.7 km downstream of the source. Several measurements were taken
near this location, not only to quantify the value of the peak concentra-
tion but also to determine the extreme (peak-to-peak scatter) concentra-
tion values. The peak value was 1.09 (±0.08) x 10"^ m~^; thus, repeated
measurements were within ±7% of the peak concentration. As shown in
18
-------
Section 6.4, plume rise near the location of the ground-level maximum
concentration was approximately 66 m, thus giving an effective stack
height of approximately 130 m. Using this effective stack height,
Pasquill-Gifford curves were constructed for stability classes C and D
(Figure 17). When these curves are compared with the data, their results
are very similar to those found in the dispersion comparability tests. C
stability most closely approximates the experimental data, but again under-
estimates the peak concentration value. D stability grossly underestimates
the peak concentration. This comparison again reflects the fundamental
problem of comparing standardized dispersion parameters, which were based
on ground-level sources and small roughness lengths, to situations charac-
terized by elevated sources and relatively large roughness lengths. Since
the comparison is very poor for D stability, only C-stability curves are
shown in Figures 18 - 20 for comparison with the vertical and with the
lateral concentration profiles.
Figure 19 shows the vertical concentration profile taken along the
plume centerline at the location of maximum ground-level concentration.
The data from nearest the surface are similar to the peak concentration
value found from the longitudinal profile, while the concentration gradi-
ent in the vertical profile is seen to be very weak. This weak concen-
tration gradient implies that errors in evaluating the peak ground-level
concentration due to inaccuracies in sensor location were likely to have
been quite small. Plume rise was on the order of 66 m. A more precise
determination of plume rise was difficult due to the rather weak concen-
tration gradient. Figure 18 shows the vertical concentration profile
taken at downstream distances of (0.13 km) and (0.86 km). Plume rise
19
-------
near the source was quite well defined, and was approximately 46 m at
0.13 km downstream and 61 m at 0.86 km downstream. Pasquill -Gifford
curves for C stability are shown for comparison with the vertical profiles
in the absence of the building. Once again these standardized curves offer
a rather poor comparison to the experimental data.
Figure 20a shows the lateral concentration profiles taken at ground
level through the location of peak concentration determined by the longi-
tudinal profile. The value of the peak concentration was again consistent
with the value taken from the longitudinal profile. The lateral gradient
was somewhat weaker than that found in the vertical profiles. The peak
concentration fell directly on the plume centerline; hence, beyond a
reasonable doubt, the ground-level concentration peak has been located.
Figure 20b shows similar profiles at the same downstream location but at
an elevation corresponding to the maximum concentration found from the
vertical concentration profile of Figure 19. Again, Pasqui 11-Gifford
curves for C stability have been provided for comparison.
20
-------
6.2 DISPERSION IN THE PRESENCE OF THE BUILDING
The model building was placed in the wind tunnel and concentration
measurements were made for direct comparison with those made in the
absence of the model building. Comparison of the two sets of data pro-
vided a direct assessment of the influence of the building's wake on
downstream concentration. The stack height was again set at 149 mm (64.1
m full-scale); effluent conditions were identical to those used in the
absence of the building (described in section 6.1). A two-minute averag-
ing time was again found to yield reasonably stable average concentration
values at the free-stream wind speed of 4 m/s.
A longitudinal profile of concentration downstream of the source was
used to locate the maximum ground-level concentration. Subsequently, ver-
tical and lateral concentration profiles were measured in order to esta-
blish beyond a reasonable doubt both the location and value of the ground-
level maximum concentration. Vertical profiles were measured at three
downstream locations: the location of the ground-level maximum, half-way
between the source and the ground-level maximum location and approximately
3.5 building heights downstream of the source. Lateral concentration pro-
files were measured both at the surface and at the height corresponding
to the elevated maximum concentration found from the corresponding vertical
profile.
The longitudinal ground-level concentration measurements in the pre-
sence of the building (Figure 17) showed that the location of the peak
concentration moved slightly closer to the source while the value of the
peak concentration increased by approximately 40%. Several measurements
21
-------
were again taken near the location of the maximum concentration, in order
to determine the limits within which the concentration values fell. The
maximum ground-level concentration was found to be 1.53 (±0.11) x 10~5
m-2, with the peak-to-peak variation near the maximum on the order of ±7%
Figure 19 shows the vertical concentration profile taken along the
plume centerline at the location of maximum ground-level concentration.
The data nearest the surface were in agreement with the peak concentration
value found from the longitudinal ground-level profile. The concentration
gradient near the surface was quite weak due to the increased turbulent
mixing in the wake of the building. Plume rise, while difficult to esti-
mate, was roughly 46 m, or 20 m less than that found in the absence of the
building. Figure 18 shows the vertical concentration profiles taken at
0.13 km and 0.86 km. Plume rise near the source was again reasonably well
defined and was approximately at 40 m at 0.13 km downstream and 56 m at
0.86 km downstream.
A lateral profile of concentration was taken through the ground-
level maximum (Figure 20a). The peak concentration was again similar to
that found from the longitudinal ground-level measurements, both in loca-
tion (i.e. along the plume centerline) and in value. Several measurements
were taken near the peak value in order to demonstrate that the random
scatter in the measured values was of the same order of magnitude as
that found from similar measurements near the peak of the longitudinal
profile. Figure 20b shows the lateral profile taken through the plume
centerline at the elevation corresponding to the peak of the vertical
concentration profile. The measured concentrations were again consistent
with those found from the vertical profile. In all cases, measurements
22
-------
at the extreme lateral positions as well as the extreme vertical posi-
tions showed significant increases in random scatter due to the inter-
mittent nature of the plumes near the edges. Peak values in all cases
were at least two orders of magnitude greater than background; hence,
the profiles were relatively smooth near the peaks.
6.3 DETERMINATION OF GEP STACK HEIGHT
The measurements described in sections 6.1 and 6.2 showed that, for
a stack height of 149 mm (64.1 m full-scale), the maximum ground-level
concentration in the absence of the building was located approximately
1.7 km downstream of the source and had a value of 1.09 (±0.08) x 10~5
m~2. With the building in position, the maximum ground-level concentra-
concentration was approximately 1.5 km downstream and had a value of 1.53
(±0.11) x 10~5 nr2. The effect of the building was thus two-fold; the
location of the maximum ground-level concentration moved approximately 13%
closer to the source, and the value of the maximum ground-level concentra-
tion increased by 40%. A stack height of 64.1 m, or approximately 1.8 build-
ing heights, is then GEP in terms of excess concentration.
The plume rise measurements show that the effect of the building was
primarily to lower the mean height of the plume. The additional turbu-
lence created by the building wake tends to promote mixing of the lower
portion of the plume, thus effectively transporting more tracer to the
ground surface. The result of this mixing action was reflected in the
weaker vertical concentration gradients and higher ground-level concen-
trations in the presence of the building. The lateral concentration
23
-------
profiles of Figure 20 show that there was only a slight increase in the
width of the concentration distribution at the surface.
When the general working rule for GEP stack heights
Hs = H + 1.5L
is applied to this case, the expected value for GEP stack height is 2.5
building heights or 90.3 m. With the fluid modeling approach, the value
is 64.1 m (1.8 building heights), considerably lower than 90.3 m. Two
factors contribute to the difference. The major factor is that the
effluent-to-wind-speed ratio for the conditions modeled (100% plant load)
was considerably greater than that on which the general working rule is
based (WS/US = 3.5 for the present case versus 1.5 for the general working
rule). This additional momentum results in greater plume rise, hence
decreasing the effectiveness with which the turbulent building wake can
entrain the plume and thereby bring more tracer to the surface. Robins
(1975), for example, showed that for a stack of 1.8 building heights
above the center of a cubical building, the excess concentration was about
35% when the Ws/Uoo = 0.55, but less than 20% when Ws/Ua> = 3.0. In order
to ascertain the effect of lowering the effluent-to-wind-speed ratio,
measurements were made with 50% plant-load conditions (Appendix C). The
fact that these results were in closer agreement with the general working
rule substantiates the importance of the large effluent-to-wind-speed
ratio.
The second important factor is the location of the source relative
to the building. In the present case, the source was located approxi-
mately 2.5 building heights downstream of the building; the general work-
ing rule for GEP stack height is based on sources located immediately
adjacent to or directly on top of the building. This separation between
24
-------
the source and the building would certainly reduce the effect of the
building wake on a plume emitted from the source. Huber et al. (1980)
found that, for a model building with a height equal to one-half its
length, the recirculating cavity region extended approximately three
building heights downstream of the building. Using this figure as a
guide, a source 2.5 building heights downstream may be outside the
immediate influence of the most highly turbulent region of the building's
wake. Barrett et al. (1978) found that the maximum ground-level concen-
tration was reduced by approximately 15% when a stack of 1.8 building
heights was moved from the center of a building to 2.5 building heights
downwind. Their results, however, apply to cubical buildings oriented at
45° to the wind. Quantification of the effect of the separation of
source and building is not possible in the present case without further
experimentation. The implication is that the wake effect on the plume
will be reduced when source and building are separated. The scatter in
the maximum concentration values, while on the order of ±7% both with
and without the building, resulted in a relatively large range of scatter
in the excess concentrations. In fact, using the extremes of +7% error
with the building and -7% without the building, a worst-case excess con-
centration of 60% (as opposed to 40%) was indicated. This corresponds to
a possible worst-case GEP stack-height error of about ±4 m. A more
realistic and certainly more reasonable way to estimate the error in
excess concentration is to assume that the errors in maximum concentration
are normally distributed. The standard deviation is then on the order of
one-sixth of the peak-to-peak value or about 2.3%. Using this standard
error in the maximum concentration, the excess concentration is 47% ( as
25
-------
opposed to 40%), which corresponds to a GEP stack height error of about
±1.5 m. This estimate of the error in measurement of excess concentra-
tion is based on a relatively small sample size and may be statistically
questionable; however, it does provide some idea of the accuracy with
which the excess concentration and, hence, GEP stack height can be deter-
mined.
6.4 PLUME RISE
The vertical concentration profiles without the building (Figures 18
and 19) were used to compare measured plume rise with calculated plume
rise. Nearest the source (x = 0.13 km), the measured plume rise was 46
m. At downstream distances of 0.86 km and 1.7 km, the measured plume
rise was 61 m and 66 m, respectively. Applying Briggs1 (1975) plume rise
formulation as specified in the Guideline, the predicted rise at these
downstream distances was 49 m, 92 m, and 115 m. Agreement between pre-
dicted and observed plume rise is very good near the source (46 m v. 49).
The discrepancies at the greater downstream distances are attributable to
two factors: first, only the effluent density ratio and not the stack
Froude number was modeled, thereby reducing the effects of buoyancy; and
second, the large turbulence intensities measured in the simulated bound-
ary layer lead to more rapid dispersion between the surface and the top
of the stack. This had the effect of decreasing the distance over which
turbulent entrainment occurs and as a result, the final rise ^was reached
earlier than would be expected from a less turbulent boundary layer.
With the building model in place, plume rise near the source was
reduced to approximately 40 m. Farther downstream, the differences in
26
-------
plume rise with and without the building were difficult to quantify
because of the weak vertical concentration gradients; however, the mean
plume height did appear to be reduced. The data show, that the mean plume
height at 0.86 km downstream was approximately 120 m; therefore, the
plume rise was about 56 m. At the location of the ground-level maximum,
turbulence from the wake of the building had mixed the plume almost uni-
formly to the surface. Remnants of the elevated maximum can be seen at
about 110 m, thus indicating plume rise of about 45 m. In the presence
of the building, then, the plume initially rose much as it did in the
absence of the building, however, shortly downwind it was more quickly
mixed into the surface layer so that the mean plume height was lower.
6.5 DISCUSSION OF RESULTS
The effect of the building's presence was to move the location of
maximum ground-level concentration closer to the source, thereby increas-
ing the value of the maximum ground-level concentration and lowering the
mean plume height. Since the effluent-to-wind-speed ratio was sufficiently
large to preclude stack downwash, these effects must be attributable to
the wake of the building. More precisely, it was the far wake which was
primarily affecting the plume. Huber et al. (1980) pointed out that there
are three distinct flow regimes found near a building in neutral flow.
The undisturbed region, or free-stream, is the area outside the strongest
influence of the building. Streamlines in this region may be slightly
distorted as they pass over the building, but the turbulence levels
remain essentially unchanged. On the lee side of the building, the flow
is separated and a highly turbulent cavity region may exist. In this
27
-------
cavity region, the flow near the surface is opposite the direction of
mean flow. A plume which becomes trapped in this cavity is quickly mixed
throughout the entire volume of the cavity by the highly chaotic flow.
The extent of this cavity may vary from 2.5 to 10 building heights down-
stream, and depends upon the precise building shape and orientation.
Downstream of the cavity region, a highly turbulent wake is found which
decays with downstream distance. This far wake region is primarily
responsible for the increase in maximum ground-level concentration in the
present case. Even though the low-pressure cavity region may tend to
limit initial plume rise somewhat, the effective stack height is suffi-
ciently great to prevent the plume from being directly entrained into the
cavity. This conclusion is supported by the longitudinal profiles of
Figure 17. The additional turbulence downstream of the cavity, however,
leads to more rapid diffusion of the plume, thus transporting it toward
the surface. In the vertical concentration profiles at 0.86 km downstream
and at the location of the maximum ground-level concentration, the verti-
cal gradient of concentration was weakened considerably by the mixing
action of the turbulence.
While the differences observed with and without the building have
been shown to be related to the building wake effect, there are other
factors that may result in adverse concentrations downstream of the
source.
For the conditions examined in this report, the wind direction was
perpendicular to the longest face of the building. This provided the
greatest opportunity for wake effects by producing the largest wake.
Though the frequency of occurrence is low, wind directions near 45° from
28
-------
perpendicular may lead to trailing vortices, created by the building.
Castro and Robins (1977) and Robins and Castro (1977) have shown, by
indirect means, that a swirling wake consisting of two longitudinal,
rotating vortices can produce downwash on the centerline. A plume with-
in this wake could be drawn toward the surface, increasing ground-level
concentrations substantially (Thompson and Lombardi, 1977). There is
little available information with regard to the effect of building
asymmetry or meandering of wind direction on these vortices; their
strength and duration are difficult to estimate. Our cursory flow visual-
ization exercises, however, did not reveal any effects of building
asymmetry or meandering wind direction on plume behavior.
A second consideration with regard to wind direction is that winds
from the east or southeast may downwash in the lee of the ridge that is
1.6 km southeast of the plant. Such downwash could again force the plume
closer to the surface and create adversely high ground-level concentra-
tions. Indeed, Lott (1982) reported that the highest surface concentra-
tions measured during a field study at this site were caused by terrain-
induced downwash resulting from neutral, southeasterly winds. As pointed
out in section 3.3, a separate series of tests using a different model
scale would be required to further evaluate the effects of this ridge.
The effects of stability are not so important with respect to the
area immediately adjacent to the building. However, during very stable
periods a plume from the stack may be transported with little vertical or
horizontal dispersion, and may impact on the surrounding terrain. The
most likely area for impact is again the ridge southeast of the plant;
the conditions most likely to result in the highest surface concentrations
29
-------
are where the plume elevation is just above a "blocked" layer (Baines,
1979), such that the plume grazes the surface of the slope as it is trans-
ported to the top. An evaluation of these effects, however, would require
additional laboratory work and considerably more extensive meteorological
data.
The effects of plant load were discussed in Section 6.3, and it
should be emphasized that plant load is a very important factor in the
determination of the GEP stack height. At very low plant loads, plume
rise may be minimal and stack downwash may occur. When the plant load is
very high, and winds are light the additional plume rise may mean that
the plume is transported well downstream before encountering the envelope
of the building wake. In this case the building influence will be mini-
mized. For most installations, the average plant load will fall some-
where between these two extremes and may show considerable diurnal and
seasonal variation. Where data are available, a correlation of plant
operating conditions with wind direction and stability may be useful in
justifying more appropriate plant load conditions. In the present demon-
stration study, only 100% plant-load operating conditions were considered
in detail. Appendix C shows that a GEP stack height of 90.3 m (2.5 build-
ing heights) could be justified as GEP on the basis of 50% plant-load.
-------
SECTION 7
SUMMARY
A fluid model study was conducted in a wind tunnel to determine the
Good Engineering Practice (GEP) stack height for a power plant installa-
tion operating under 100% plant-load conditions. A stack height of 64.1
m was shown to meet the current GEP critera.
The meteorological conditions simulated were northwest winds (the
direction perpendicular to the face of the building) and neutral stability.
Surface characteristics were modeled using surface roughness elements to
simulate the terrain approximately 3.5 km upwind and downwind of the model
building. The ratios of effluent density to ambient density and effluent
speed to wind speed were matched between model and prototype, and the
building Reynolds number was sufficiently high to ensure that the flow
around the building was Reynolds-number independent. The background dis-
persion characteristics in the absence of the model were shown to conform
most closely to Pasquill-Gifford stability category C, slightly unstable.
Plume rise near the source was adequately described by Briggs1 formulation.
The effect of the building wake has been shown to decrease plume rise,
decrease the downstream distance to the point of maximum ground-level, and
increase the magnitude of the maximum ground-level concentration by 40%.
Vertical and lateral concentration profiles both with and without the model
building have been provided in order to show that the maximum ground-level
concentration in each case has been determined beyond a reasonable doubt.
The error in the measurement of excess concentration was on the order of ±7%.
The observed differences in maximum ground-level concentration with and with-
out the building were shown to have resulted from the influence of the build-
ing wake.
31
-------
REFERENCES
Baines, P.G., 1979. Observations of stratified flow past three-dimensional
barriers. J. Geophys. Res. 84 (C12): p. 7834-7838.
Barrett, C.F., Hall, D.J. and Simmonds, A.C., 1978. Dispersion from Chim-
neys downwind of cubical buildings - A wind-tunnel study, Warren Spring
Lab. presented at the NATO/CCMS 9th International Meeting on APMA, Toronto,
Aug. 28-31.
Bearman, P.M., 1971. Corrections for the effects of ambient temperature
drift on hot-wire measurements in incompressible flow, DISA Information,
no. 11: 25-30.
Briggs, G.A., 1969. Plume Rise, Critical Review Series, U.S. Atomic Energy
Commission. TID-25075. National Technical Information Service, Springfield,
VA, 81pp.
Briggs, G.A., 1975. Plume Rise Predictions. ATDL No. 75/15, Atmospheric
Turbulence and Diffusion Laboratory, NOAA Environmental Research Laboratory,
Oak Ridge, TN, 53pp.
Castro, I.P. and Robins, A.G., 1977. The flow around a surface-mounted
cube in uniform and turbulent streams. J. Fluid Mech. 79 (pt. 2): 307-335.
Castro, I.P. and Snyder, W.H., 1980. Three Naturally Grown and Simulated
Boundary Layers. Fluid Modeling Facility Internal Report, U.S. Environ-
mental Protection Agency, Research Triangle Park, NC, July, 200pp.
Counihan, J., 1969. An improved method of simulating an atmospheric bound-
ary layer in a wind tunnel. Atmos. Envir. 3: 197-214.
Davenport, A.G., 1965. The relationship of wind structure to wind loading.
In: Proceedings of the Conference on Wind Effects on Buildings and struc-
tures, National Physics Laboratory, HMSO, London, pp. 54-102.
Environmental Protection Agency, 1981. Guideline for Use of Fluid Modeling
to Determine Good Engineering Practice Stack Height. EPA-450/4-81-003.
U.S. Environmental Protection Agency, Research Triangle Park, NC. 47pp.
Hanna, S.R., 1980. Measured sigma-y and sigma-theta in complex terrain
near the TVA Widows Creek, Alabama, Steam Plant. Atmos. Envir. 14 (4):
401-407.
Huber, A.M., Snyder, W.H. Thompson, R.S. and Lawson, R.E. Jr., 1980. The
Effects of a Squat Building on Short Stack Effluents. EPA-600/4-80-055.
U.S. Environmental Protection Agency, Research Triangle Park, NC, 118pp.
32
-------
Khurshudyan, L.H., Snyder, W.H. and Nekrasov, I.V., 1981. Flow and Dis-
persion of Pollutants over Two-Dimensional Hills: Summary Report on Joint
Soviet-American Study. EPA-600/4-81-067. U.S. Environmental Protection
Agency, Research Triangle Park, NC., 143pp.
Lott, R.A., 1982: Terrain-induced downwash effects on ground level S02
concentrations. Atmos. Envir. 16 (4): 635-642.
McElroy, J.L., 1969. A comparative study of urban and rural dispersion.
J. Appl. Meteorol. 8 (1): 19-31.
Robins, A.G., 1975. Plume Dispersion in the Vicinity of a Surface Mounted
Cube, Central Electricity Generating Board, Research Department Report.
R/M/R 220. Marchwood Engineering Laboratories, April.
Robins, A.G. and Castro, I.P., 1977. A wind tunnel investigation of plume
dispersion in the vicinity of a surface mounted cube, I. The flow field.
Atmos. Envir. 11 (4): 291-297.
Snyder, W.H., 1979a. Testimony on Behalf of the U.S. Environmental Protec-
tion Agency, Presented at the Public Hearing on Proposed Regulatory Revi-
sions to the 1977 Clean Air Act Stack Height Regulations, Wash., DC, May
31, 12p.
Snyder, W.H., 1979b. The EPA Meteorological Wind Tunnel: Its Design,
Construction, and Operating Characteristics. EPA-600/4-79-051. U.S.
Environmental Protection Agency, Research Triangle Park, NC, 78pp.
Snyder, W.H., 1981. Guideline for Fluid Modeling of Atmospheric Diffusion.
EPA-600/8-81-009. U.S. Environmental Protection Agency, Research Triangle
Park, NC, 200pp.
Snyder, W.H. and Lawson, R.E. Jr., 1976. Determination of a necessary
height for a stack close to a building - a wind tunnel study. Atmos.
Envir. 10 (9): p. 683-691.
Thompson, R.S. and Lombardi, D.J., 1977. Dispersion of Roof-Top Emissions
from Isolated Building: A Wind Tunnel Study. EPA-600/4-77-006. U.S.
Environmental Protection Agency, Research Triangle Park, NC, 44pp.
Turner, D.B., 1970. Workbook of Atmospheric Dispersion Estimates. Office
of Air Programs, Publication Number AP-26, U.S. Environmental Protection
Agency, Research Triangle Park, NC.
Vogt, K.J., 1977. Empirical investigations of the diffusion of waste air
plumes in the atmosphere. Nuclear Techno!. 34: 43-57.
33
-------
TABLE 1. PROTOTYPE AND MODEL PARAMETERS
Parameter
Scale
Free-Stream Wind Speed, U00(m/s)
Boundary Layer Depth, 6(m)
Roughness Length, z0 (m)
u*2/^
Power Law Index
z0/S
Prototype
Model
Stack Diameter, D (m)
Plant Load (%)
Effluent Velocity, Ws (m/s)
Effluent Temperature (°K)
Ambient Temperature (°K)
Density Ratio (Ps/pa)
Effluent-to-Wind-Speed Ratio
at Stack Exit (Hs = 64.1 m)
1
) 11.6
400(calculated)
0.2-0.6
0.0023-0.0026
0.18-0.22
0.00050-0.00150
0.0056-0.0167
8.2
100
28.1
422
293
0.694
1/430
4.0
0.9
0.00074
0.0028
0.2
0.00082
0.0088
0.01905
100
9.5
293
293
0.694
3.47
3.47
34
-------
OJ
•O
-------
S-
•4->
to
•r~
~a
o
c:
cr
cu
s-
-o
c
36
-------
>-
o
.9 4
.8 4
WIND SPEED (m/s)
1
Figure 3. Cumulative frequency distribution of wind speeds for northwest
winds under neutral stability (valley tower location).
37
-------
1.35 i i i i I i i i i I ii i i i i i i . 1 i , i i i , i i i i i ii .
1.2 -
1.05 -
.9 -
z - d
6
.75 -
.6 -
.115 -
.15 -
0 -f
20 21
A
A
A
f
A
A
I
A
A
I
I
A
A
A
t
A
A
A
c'b
TEMPERATURE (°C)
29 SO
Figure 4. Vertical temperature profile in wind tunnel test section,
38
-------
t 4
i -
^ 2 I
~° c^
r
0) &
T3 O
•H -H
CO ts.
c:
o
in
c
o>
ra
o
00
IJD
(oig'600^992
M
c
to
e
•H
CTi
CX Q)
O -H
H >
60
C
3
«
C
O
o
0)
en
0)
O to
to co
5 c
-------
I
-p
in
>>
to
s-
>
(O
-o
{=
3
O
OJ
o
o
-p
to
O
OO
UD
3
CD
40
-------
S-
£
ro
CO
T3
C
3
o
.a
u
O)
.c
D-
to
o
a
a
o
o
a
o
a
a
2 -a
N
-p
td
I I I I I I I I I I I I I I I I I I I I I t I I I I I I I I I I I I I I I I I I I I I I I I I IJ-
(1)
.c
-p
S-
0)
g
a.
o
o
ai
a>
en
LO
O
41
-------
0
..a..
Nl
CVJ*
CO
CO
O)
to
to
>
to
O)
o
O)
00
(U
en
N
42
-------
(It)
-
.
-
-
-
-
-
-
'
I
„
-
-
-
•
-
-
— i 1 1 1 1 r
s
0
o
o
^>
o
0
o
0,
X
o
o
o
o
o
s
0
o
0
0°
0
1 1 1 i 1 L
1 ' ' ' * ' ' ' '
o n <:
o n <"
o a
C
O!
O
c:
o;
T3
3
C
O
o
o
o
o
o
o
o
°0
0°
0°
00
O <
o o
o %
o £
o %
O
N
80 o o
o in oo
•— in i—
un LO oo
c
-------
10-s -•
(rrf2)
JO'6 -
10
-7
.1
Distance (km)
Figure 10. Surface concentration profiles (A) compared with
Pasquill-Gifford C and D stability.
44
-------
LO o o LO
« O i— CVJ CO
<3DOO
o
a
a
o
o
o
co
o
o
f\J
o
a
LT>
O
00
o
oo
o
C\J
t
T3
O
o- ^1
re
o_
T3
O)
(B
CL
o
o
N
in
i- to
a
o
LO
o
o
o
o
no
o
a
t\j
o
o
N
45
-------
•o
o
CU
I
3
CT
in
Q.
4J
'%
•o
to
o
o
-------
VI
O \ '•
I
CD
O
E
tO
S-
3
O
"O
s_
O
M-
H-
O
(O
Q.
CD
O
c
(O
T3
o;
(O
Q.
O
O
4J
73
O)
CO
s_
3
en
N
47
-------
Figure 14. Flow visualization using paraffin-oil smoke source. H =64.1m,
100% plant load.
48
-------
LO
00 LO
CvJ **• O1
II II II
OQ CQ CQ
X
O
LT>
on
Ln
o
•o
I
N
LU
LU
_J
ZD
m
GJ
o
01
3
_Q
S-
10
"O
3
CD •
C CD
O E
C -r-
O T3
O O
-------
LO
OT
LO
O
D
01
0
O ° &
u n
o
<
a
"
n n o n
c
o
o
LD
•a
1
N
c
C •!-
CTi
II
03 ,-.
or *•€
yf ^J
0 <^
4-
EtO
-o
O.I—
O
"« >>
U (U
O)
3
50
-------
io-
(nf2)
10-6 -.
ID"7
Distance (km)
Figure 17. Surface concentration profiles with (A) and without (D)
the building. Stack height 64.1m.
51
-------
oo
1
o
X
CM
r
X
o
o
I/)
o
o
o
o
o ^:
•r- O"-
•P T-
(O OJ
s- ^:
O) O
O (O
E •!->
O 1/5
O
r— O)
(O E
O T-
••- T3
+J r—
O) 5
00
O)
N
52
-------
500
100
300 --
z (m)
200 -
100 -
0 .0025 .005 .0075 .01 .0125 .015 .0175 .02 .0225 .025
(m"2 X TO"3)
Figure 19. Vertical concentration profiles with (A) and without (D)
the building. Stack height 64.1m, downstream distances of
1.5km and 1.7km respectively.
53
-------
•o
c
CD fO
.c:
•*-> £
O 10
-C QJ
4-> O
•a w
to -5
re
to
0) «<
O «3-
S- 10
D.
O 01
•I— •!— •
•!-> OJ >>
03 SZ i—
S- ^ >
C O -i-
O) ra 4->
O +J O
c: co a>
o a.
o • v>
C7! O)
•— c i-
(T3 -i-
i- T3 E
0) i — ^^
-»-> -r- r^
(O 3
_i .n i—
o
CM
CU
-------
APPENDIX A
DESCRIPTION OF FACILITIES AND INSTRUMENTATION
A.I The Fluid Modeling Facility Wind Tunnel
This study was conducted in the Environmental Protection Agency's
Meteorological Wind Tunnel. It is an ultra-low speed, open-return wind
tunnel with a test section 2.1 m high, 3.7 m wide and 18.3 m long. Air
enters the test section through a flow-straightening honeycomb and four
turbulence-reducing screens. A plenum chamber just prior to the 2.8-to-l
contraction allows turbulence in the wake of the screens to decay. An
adjustable ceiling allows compensation for blockage effects of models and
achievement of a zero-pressure gradient in the test section. Transparent
windows form the sides of the test section to facilitate flow visualiza-
tion. An instrument carriage provides the capability for positioning a
probe anywhere in the test section with an accuracy of ± 1 mm. Controls
and readout for the carriage are conveniently located at an operator's
console. After the test section, the air passes through an acoustic si-
lencer, a rectangular-to-round transition section, the fan, a diffuser,
and another acoustic silencer before being exhausted back into the room.
The tunnel is driven by a 75 kilowatt AC motor with eddy-current coupler
for speed control of the 1.8 m diameter fan. This apparatus provides
steady speeds in the test section of 0.3 to 8 m/s. The motor and fan
assembly is enclosed in an acoustic silencer to provide a low noise level
in the laboratory. Further details of the wind tunnel and its operating
characteristics are described by Snyder (1979b).
55
-------
A. 2 INSTRUMENTATION
A.2.1 Velocity measurements
Mean velocity, turbulence intensity, and shear-stress profile data
were obtained with 751, Inc. model 1054A constant-temperature anemometers
in conjunction with model 1241-T1.5 x-array hot-wire probes (end-flow
style). Calibrations were performed in the free-stream with the sensor
mounted on the instrument carriage. The reference velocities for calibra-
tion were obtained with a Dwyer model 160-24 pitot-static tube1; the differ-
ential pressure was monitored with an MKS Baratron capacitance manometer
(model 310BH sensor head with model 170M electronics unit). Yaw calibra-
tions of the sensors were performed in separate series of tests in an
instrument calibration tunnel.
Temperature near the sensor location was monitored both during cali-
bration and routine operation by a Hewlett-Packard model 2801A quartz ther-
mometer. Analog output from the anemometers was converted to digital form
by a 12 bit analog-to-digital converter. The resulting data were processed
on a Digital Equipment Corp. PDP-11/40 minicomputer. Two-minute averages
at a sampling rate of 1000 samples per second yielded reasonably stable
mean values (± 1% on mean velocity). Further details of the hot-wire and
data-processing systems are given by Snyder (1979b).
A.2.2 Concentration measurements
A hydrocarbon tracer technique was used to measure concentrations
downwind of the source. The technique employed CP grade ethylene (minimum
purity of 99.5 mole percent) as the tracer source. Concentrations were
measured with Beckman model 400 flame ionization detectors (FIDs), operated
in the continuous sampling mode.
56
-------
The FIDs were calibrated using 1% certified (Scott Environmental
Technology, Inc.) "span" gas; zeroing was accomplished with "zero" air
(< 1 ppm hydrocarbons). The FIDs were shown in a separate series of tests
to be linear over four decades of concentration (Khurshudyan et al, 1981).
The samples to be analyzed were drawn from a "rake" of tubes which was
mounted on the instrument carriage to allow convenient positioning. The
sample flow rate was 200 cm^/min. Five analyzers were used simultaneously
to speed the process of acquiring data. One of the five constantly mon-
itored background concentration upstream of the source. Analog output
from the FIDs was also digitized to 12-bit precision for processing by the
minicomputer.
A.2.3 Data acquisition system
All laboratory data were collected using a Digital Equipment Corp.
PDP-11/40 minicomputer. Anemometer calibrations were performed over the
velocity range of interest (typically 6 to 9 points over the range 1 to 5
m/s). During calibration, the computer was used to fit a King's Law form
of equation to the calibration data. This best fit relation was then used
to generate a "look-up" table for conversion of voltage to velocity during
routine operation. A typical calibration curve is shown Figure A-l. The
hot-wire anemometer was typically sampled at 1000 samples per second, and
data reduction took place between samples; hence, real-time outputs of
velocity, intensity, and shear stress were available. Temperature compen-
sation was accomplished using the method of Bearman (1971), which required
occasional modification of the look-up table as temperature in the test
section (room temperature) changed.
As the time constant of the FIDs is on the order of 0.5 second, these
57
-------
units were sampled at a rate of one sample per second. Two-minute averages
again provided stable mean values. The FIDs have linear response, so that
the generation of mean values was straightforward. Zero and span values
were recorded at the beginning and end of each test to assure that analyzer
drift was not a problem. Background values were substracted from each
sample to account for background drift.
All data files were stored on disk for later processing and preserved
on magnetic tape.
A.2.4 Volume flow measurements
Ethylene, helium and air were mixed prior to injection into the model
stack in order to obtain correct density and velocity ratios. The flow
rates of these gases were measured and continuously monitored using Meriam
laminar-flow elements (LFEs); the differential pressure was observed on
Meriam micromanometers. Figure A-2 shows the typical apparatus for in-
jecting gases into the tunnel. Calibration of the LFEs was accomplished
using a volumetric flow calibrator (Brooks model 1050A 101), which had a
rated accuracy of 1/2%. Where significant back-pressures were anticipated,
as for example, with the porous ball stack, the back-pressure was monitored
as a check on system integrity.
58
-------
6.5
I"I T~T T 1 f~ T
E = anemometer output voltage
r
6.0
5.5
5.0
4.5
4.0
3.5
3.0
2.5
U = tunnel speed (m/s)
PROBE H359-2
SLOPE = 1.468
INT. = 2.784
ALPHA = 0.450
CALIB.TEMP. = 23.9°C
WIRE TEMP. = 200°C
o CALIBRATION POINTS
• ZERO FLOW VOLTAGE
I i I I I i I i I I I I I I i i i I I
-j
-i
j
0.5
1.0
Ua
1.5
2.0
Figure Al. Typical calibration curve for hot-wire anemometer.
59
-------
MANOMETER!" "1
A i
r~T
^"MANOMETER
B
C2H4
LAMINAR
FLOW
ELEMENT
, POROUS
_ BRONZE
SPHERE
DIA.=14mm
Figure A2. Diagram of source and flow measurement apparatus
used for dispersion comparability test.
60
-------
APPENDIX B
CONCENTRATION MEASUREMENTS FOR STACK HEIGHTS OF
54.2m, 68.8m, 72.3m, AND 90.3m
Figures Bl through B4 present the concentration measurements made
over a range of stack heights for the purpose of initially determining
the appropriate GEP stack height. The measurements were carried out in
the same manner as those for the GEP stack height described in sections
6.1 and 6.2. Note that the effluent-to-wind-speed ratio differs slightly
for each stack height presented. Figure B5 shows the results of these
measurements in terms of percent excess concentration versus stack height,
where excess concentration is defined as the ratio of the maximum ground-
level concentration with the building to the maximum ground-level
concentration in the absence of the building minus one.
61
-------
H- in
I
i
I I .
q.
C! J „ .
I-J --J
o-
CM
-ft
i- -
. ,-,*
ii- ^
i-
if d
i
en
c:
•r-
r—
"r~
JD
-C
cu ^
o o
C "3
O -M
<_5 IT)
CO
CD
en
CM
I
62
-------
I
F
J
n
a :\
Q
=:]D
n ,1
IT)
O
en
c
>> -P
•o
3
_Q
(U
o in
— :o
X
CM
i
cr E
<1 .
CM
=sro
O II
t '
£
A U
dl— 1
___
o
O
O
CJi—
1- •
a. co
c •
o •»->
-M cn
(O -r-
S-
-------
LO
I
d 4,
=1 D
C
•5
o-
--- 00
CO
4-> •
3 CO
0 II
ill
U
IM
o
C vi
es
o X
'
-o «
ETJ
« (O
ft)
O
tOO
CT
fO •!—
S- 0)
o o
C ro
04->
c_> oo
cu
O
CQ
CL)
O1
to
s
— h
- L i™l,,xl™-i«« L™,.-J
X
tO
i
r-
64
-------
ID-4
ID'5 T
i
gi
(nf2)
io
-8
10-
.1
r
Distance (km)
10
"XUS H
(m'2)
10-6
-0.5
(km)
Figure B4. Concentration profiles with (D) and without (A)
the building. Stack height 90.3m, 100% plant load,
W/US=3,22.
65
-------
60
E
o
-------
APPENDIX C
6EP STACK HEIGHT FOR 50% PLANT LOAD CONDITIONS
As described in section 6.3, a direct application of the working rule
for GEP stack height indicates an expected value of 90.3 m or 2.5 building
heights as compared with the experimentally obtained value of 64.1 m or 1.8
building heights. Two reasons were suggested for this difference: the large
effluent-to-wind-speed ratio, and the location of the source relative to the
building. While the location of the source relative to the building is
fixed, the effluent-to-wind-speed ratio varies with plant load. This means
that a determination of GEP stack height at reduced plant load would provide
an indication of the sensitivity of GEP stack height to the effluent-to-
wind-speed ratio. To pursue this idea, experiments were undertaken to
determine the excess concentration appropriate to a stack height of 90.3 m
and 50% plant-load conditions. The experimental arrangement was identical
to that used for 100% plant-load conditions, except that the effluent-to-
wind-speed ratio was reduced by one-half. This provided an effluent-to-
wind-speed ratio for the 90.3-m stack of 1.61. The resulting measurements
are shown as figures Cl and C2. The excess concentration in the presence
of the building is approximatley 35%. This value is consistent with other
investigations, and therefore lends support to the rather low GEP stack
height determined for 100% plant-load conditions.
67
-------
o>
c
•r—
"O
o
o
u")
o
r—1
•="7- I
o "
<*- E
O CO
s_ •
0.0
E
O -4->
IS|
<0 -r-
i- OJ
C
O .^
o o
C (O
o +•>
O V7
rtfTl- ' u J
e
u Q
<]
n
D
. j. ,1- UU- L -J-..-i™ .!
re
^ =>
C71
0-
68
-------
10
-1
10
-S
xUs
(nf2)
io-B
10
-7
lO'8
-0.5
z=80m
.5
y (km)
Figure C2. Concentration profiles with (ID).,and without (A)
the building. Stack height 90.3m, 50% plant load.
69
-------
APPENDIX D
RAW DATA LISTINGS
In order to reduce printing costs, the raw data listings have not
been included with this report. Listings of the raw data are available
from the authors on request.
70
-------
TECHNICAL REPORT DATA
(Please read JnHrufiium on the rcvcnc before completing)
REPORT NO. 2.
.TITLE AND SUBTITLE
DETERMINATION OF GOOD-ENGINEERING-PRACTICE STACK HEIGHT
A Fluid Model Demonstration Study for a Power Plant
. AUTMOR(S)
Robert E. Lawson, Jr. and William H. Snyder
PERFORMING ORGANIZATION NAME AND ADDRESS
Environmental Sciences Research Laboratory
Offi'ce of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
2. SPONSORING AGENCY NAME AND ADDRESS
Environmental Sciences Research Laboratory — RTP.NC
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
3. RECIPIENT'S ACCESSION-NO.
5. REPORT DATE
6. PERFORMING ORGANIZATION CODE
8. PERFORMING ORGANIZATION REPORT NO.
10. PROGRAM ELEMENT NO.
CDTA1D/02-1313 (FY-83)
11. CONTRACT/GRANT NO.
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
EPA 600/09
.SUPPLEMENTARY NOTES
1. On assignment to the Environmental Protection Agency from the National Oceanic and
Atmospheric Administration, US Department of Commerce
ABSTRACT
A study using fluid modeling to determine good-engineering-practice (GEP)
stack height for a power plant installation is discussed. Measurements are presented
:o describe the simulated boundary layer structure, plume dispersion characteristics
ja the absence of the model plant building, and the maximum ground-level concentration
>f effluent downstream of the source, both with and without the model plant building.
Analysis of the maximum ground-level concentration shows that, in this case, a stack
:ieight of 64.1m meets the current GEP criteria for 100% plant load conditions.
KEY WORDS ANO DOCUMENT ANALYSIS
DESCRIPTORS
-
DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
b. IDENTIFIERS/OPEN ENDED TERMS
19. SECURITY CLASS (This Rtport) ~
UNCLASSIFIED
20. SECURITY CLASS (This page)
UNCLASSIFIED
c. COSATI Field/Group
-.
21. NO. OF PAGES
22. PRICE
Form 2220-1 (S-73)
-------
^
Q C
a> -' CL S
g. ,. ct
s~5s §
Z ^ £ Q-
a
-; w
~» 3
o o
o
> ^
ct ft
z
o
f—
>
o
D
TJ
0
3
-1
c
z
H
-<^
m
13
r
O
m
D
m
Z
P
H
<
T>
O
TJ
_
H
rn
C
tn
m
0*
CJ
Q
( i
O
Tl
Tl
n
r
03
C
Z
m
c/i
co
^
^'
D
n
5
^
CD
O
3
O
£>.
cn
r-j
(D
CO
O
(D
3
— *
0
"*
rn
3
irontr
!6
3
s
33
(D
fD
^
O
rr
5"
~+,
o
3
CJ
5
3
q
IT1
^
0
~*~
Z)
'T1
•t;
"T
in
13
a
, — ,
tt
c.
'"H
T""1
^
~1
»„
'j"}
Tt
*^*
33
C)
z
"T1
>
33
-T«
, ^
*~ \
~*
£*
C)
~n
*"'.
•1
Z
0
J)
H
.J
01
------- |