&EPA
           United States
           Environmental Protection
           Agency
          Office of Air Quality
          Planning and Standards
          Research Triangle Park NC 27711
EPA-450/4-80-023
July 1981
           Air
Guideline for Determination
of Good Engineering Practice
Stack Height (Technical
Support Document for the
Stack Height Regulations)

-------
                                          EPA 450/4-80-023
              GUIDELINE FOR DETERMINATION OF
          GOOD ENGINEERING PRACTICE STACK HEIGHT
(TECHNICAL SUPPORT DOCUMENT FOR STACK HEIGHT REGULATIONS)
                         July 1981
           U.  S.  Environmental  Protection Agency
            Office of Air, Noise, and Radiation
       Office of Air Quality Planning and Standards
       Research Triangle Park,  North Carolina  27711

-------
                                ACKNOWLEDGEMENT
     This document was primarily prepared  by Alan  Huber,  formerly of the
Monitoring and Data Analysis Division,  Office of Air  Quality Planning and
Standards, Environmental  Protection  Agency,  now assigned  to the Agency's
Meteorological and Assessment Division,  Environmental  Sciences Research
Laboratory.   Comments and suggestions  by Joseph Tikvart and James Dicke of
OAQPS were significant.   Changes to  the July 31, 1978, draft were based
mainly on comments and suggestions received  during the period of public
comment.

-------
                         TABLE OF CONTENTS
                                                            Page
ACKNOWLEDGMENT	,  .  ii
1.0  OVERVIEW AND  RECOMMENDATIONS	   1
2.0  TECHNICAL  BASIS  FOR GEP  STACK HEIGHT  .'	   5
     2.1   Description  of Aerodynamic Effects	   5
     2.2   Building Effects	•  .  .	   7
-  -   2.3   Quantitative Rationale for GEP
           Equation and Excessive Concentration	15
     2.4   Terrain  Influences	24
     2.5   Minimum  Stack Height	26
3.0  DETERMINATION OF  GEP  STACK HEIGHT	29
     3.1   Initial  Assumptions 	  29
     3.2   Simple Structures	  30
           3.2.1 Low  Structures	31
           3.2.2 Tall  Structures	31
     3.3   Complex  Structures	35
           3.3.1 Tiered Structures	35
           3.3.2 Tapered Structures  	  37
           3.3.3 Groups of Structures	41
     3.4   Terrain  Obstacles	.43
     3.5   Framework for Demonstrating  GEP  Stack  Height.  .  .  43
4.0  AIR  QUALITY ESTIMATES	47
REFERENCES	53
Appendix  A.   Annotated Bibliography  	 A-l
                                 tit

-------
1.0  OVERVIEW AND  RECOMMENDATIONS
     As required by section 123 of the Clean Air Act Ammendments of
1977, the Administrator has promulgated regulations (40 CFR Part 51) to
assure that  the control of any air pollutant under an applicable imple-
mentation plan shall not be affected by (1) stack heights that exceed
good engineering practice or  (2) any other dispersion technique.  Good
engineering  practice (GEP)'is defined with respect to stack heights
in section 123 of  the Clean Air Act Amendments of 1977, 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  atmospheric downwash, eddies and wakes
which may be created by the source itself, nearby structures or nearby
terrain obstacles."
     The GEP definition is based on the observed phenomenon of disturbed
.atmospheric  flow near structures or terrain obstacles.  The GEP definition
identifies the minimum stack  height at which significant adverse aerodynamic
effects are  avoided.  Some finite stack height is therefore necessary to
insure that  emissions do not  result in excessive ground-level concentrations
due  to adverse aerodynamic effects.
     This guideline provides  technical support for the definition and
specification of GEP stack height as found in the stack height regulations
for  sources  near building structures and for reasonable minimum stack
heights.  The remainder of this section contains applicable definitions
of GEP stack height.  Section 2 summarizes available technical information
and  provides a  technical basis for the GEP definitions.  Section 3
provides  examples  of factors  and methods to be applied in determining
                                1

-------
GEP stack height.   Section 4 identifies  the procedures  and the assumptions
that should be used in determining  emission limitations based on GEP
stack height.   An  annotated bibliography is included  that provides a
representative selection of statements found in  the scientific literature
concerning the stack height for which adverse aerodynamic effects may be
a problem.
     The scientific literature in general  indicates that a case specific
review is integral to assuring the  prevention of adverse aerodynamic
effects near a given source.  However, the literature also identifies
generalized formulations which are  designed to establish the minimum
stack height to prevent this .phenomenon.   The following guidance  is based
on these generalized findings:
          (1)   GEP stack height to  minimize the  adverse impact of building
structures is  defined by the formulation
                    Hg=H+1.5L                                 (1)
Where:  H  = Good  engineering practice stack height
         H = Height of the structure or  nearby structure
         L = Lesser dimension (height or projected width) of the structure
             or nearby structure.
Both the height and width of the structure are determined from the
frontal area of the structure, projected onto a  plane perpendicular to
the direction  of the wind.  If the  structure is  asymmetrical, the GEP
stack height should be based on the plane projection  lying upwind from
the source (stack) which results in the  greatest justifiable height.
The plane projection may have a multitude of heights  or widths,
 As new studies and research are reported,  additional  guidance will  be
 provided.

-------
particularly for a multilayered structure.  Each combination of the
height, H, and lesser dimension (height or width), L, must be evaluated
for each segment of the structure to determine which one results in the
greatest GEP stack height as defined by Equation (1).  Adjacent and
nearby structures whose plane projections lying upwind from the source
are overlaying should be considered as one structure.  Likewise, structures
aside of each other should be considered as one structure if their
distance of separation is less than their smallest dimension (height or
width).
     The downwind area in which a nearby structure is presumed to have a
significant influence on a source should be limited to five times the height
or width of the structure, whichever is less.  Thus, application of Equation (1)
should be limited to emission points within 5L of the building structure.
The area of influence becomes diminishingly small as the height to width
ratio of a structure increases.  Thus structures such as stacks and radio
or TV transmission towers should not be considered in GEP stack height deter-
minations.  Assumptions associated with the determination of GEP stack
height and appropriate examples are presented in Section 3.  Complex
structures with a multitude of heights and widths are covered in Section 3.3.
Where concern exists for possible significant effects on sources from  a
distance greater than 5L, a wind tunnel or field study should be conducted
unless an analogy to a similar study is available.
          (2)  The GEP stack height required to minimize the adverse
effects of elevated terrain should be determined on a case-by-case
basis.  Field studies designed to evaluate the specific situations under

-------
the variety of adverse meteorological  conditions  are  the best source of
information.  Where field studies  are  not possible, comparable fluid
model studies are acceptable.  A framework for demonstrating GEP stack
height is presented in Section 3.5.
          (3)  There are circumstances where  pollutants  from a major
source may be released near ground level  or from  a  stack with no supporting
structure on which to base GEP calculations.   To  avoid  natural atmospheric
effects which may cause excessive  concentrations  around  very low level
sources, a stack height of 65 meters is defined as  good  engineering practice,
without demonstration of necessity for any source (see  Section 2.5).
          (4)  There may be cases  where a plume from  a  source is expected
to impact on downwind terrain whose elevation exceeds the GEP stack height
elevation calculated from Equation (1) or from a  fluid  modeling/field study.
In such situations the allowable stack height is  that at which the concen-
tration on the terrain feature is  the  same as the maximum ground-level con-
centration for emission limitations purposes, calculated for terrain no
greater than the appropriate GEP stack height determined under (1), (2) or
(3) above.
     Thus the allowable stack height to avoid plume impaction will in no case
result in a less restrictive emission  limit than  for  a  source in a flat
terrain.   If the source owner operator elects not  to fully utilize this
stack height allowance, the emission limitation must  insure that ambient air
quality standards and PSD increments will not be  violated.  Guidance for mak-
ing air quality estimates is contained in Section 4.0.
 Flat terrain is defined as a setting without significant meteorological
 complexities, e.g., topographic features do not exceed the physical stack
 height of the source being modeled.

-------
2.0  TECHNICAL BASIS FOR GEP STACK HEIGHT
     2.1  Description of Aerodynamic Effects
          Atmospheric flow is disrupted by aerodynamic forces in the
immediate vicinity of structures or terrain obstacles.  The aerodynamic
forces evolve from interacting frictional forces and pressure gradients
induced by the local obstruction.  The surface friction and pressure
gradients combine to retard the atmospheric surface layer flow enough to
produce regions where the flow is locally distorted, causing an area of
stagnation (cavity} to develop.  The flow within the stagnant region is
highly turbulent and conceptually perceived as circulating eddies.  The
outer boundary of the eddy or cavity region extends from the point of
separation to reattachment downwind, as shown in Figure 1.  The wake is
defined as the entire region of the flow field that is disturbed by the
obstacle.  The upper boundary of the wake is called the "envelope", as
shown in Figure 1.  The reattachment point is taken as the ground-level
position where the flow is no longer drawn back towards the backside of
the building.  Downwind, beyond the reattachment, the flow readjusts
itself to a boundary layer appropriate to local surface roughness.  For
sharp-edged obstacles the flow distinctly separates at the leading
edges.  For rounded obstacles the point of separation can vary greatly.
The disrupted flow near either building structures or terrain obstacles
can both enhance the vertical dispersion of emissions from the source
and reduce the effective height of emissions from the source.  For
elevated sources, these aerodynamic effects tend to cause an increase in
the maximum ground-level concentrations.

-------
   UNDISTURBED REGION
                                                                <-t ~  ^  c
                                                                ^ WAKE REGION
Figure 1.  Diagrammatic outline of the envelope and cavity regions in the wake of a building (vertical section)

-------
          Additional discussions of the aerodynamically induced disruption
around obstacles can be found, for example, in Hunt, ejt al_. (1978);
Cermak (1976); Halitsky (.1968); Scorer (1968); and Batchelor (1967).
The complex pattern of flow around a rectangular block is depicted by
Figure 2.' A review of the literature clearly indicates that the aerodynamic
influences and the extent of the wake are highly dependent on the particular
shape and design of the obstruction.  The extent of the wake also depends
on the'characteristics of the approaching atmospheric flow.  Presently,
theoretical and quantitative understanding of the extent of obstacle
influences are limited.  Further examinations of the extent of influence
for a wide range of structures and terrain obstacles are needed.
     2.2  Building Effects
          The scientific literature in .general indicates that a case
specific review is integral to assuring the prevention of adverse aerodynamic
effects in the immediate vicinity of a given source.  However, the
literature also identifies generalized formulations which are designed
to establish minimum stack heights to prevent this phenomenon.  One such
formulation is the "2.5 times rule," which specifies that stacks designed
to discharge their effluent at least 2.5 times the height of the highest
nearby structures would escape building influences.  This rule arose
during the early part of this century as a practical formula.  Hawkins
and Nonhebel (.1955). report that the rule had been successfully used by
the British electricity generating industry during the previous 20
years.  A British government report (Beaver, 1954), which summarizes the
informed opinion at that time, presents the 2.5 times rule as successfully
used in practice.  According to Sutton (I960) the rule was probably
                                 7

-------
CO
                              Figure 2.  Depiction of flow pattern around a rectangular block as presented
                              by Woo et al., 1977 and Hunt et al., 1978.

-------
originally deduced by Sir David Brunt from W. R. Morgan's study of the
height of disturbances over a ridge in connection with an investigation
into the disaster of an airship.  Marks' Standard Handbook for Mechanical
Engineers (Baumeister, e_t aj_.  1978) states that "a ratio of stack
height to building height of 2-1/2 to 1 or more is commonly used to
avoid entrapment of the plume in the vortex of adjacent buildings and
the associated high ratios of ground-level concentration."
          No matter what its origins, the rule can be generally supported
by scientific literature.  In some instances where application of the
2.5 times rule was considered impractical individual evaluations of the
specific case have been made.  Host of these studies were conducted as
scale model studies in a wind tunnel where the design-parameters could
be easily adjusted to determine the necessary stack placement and height.
Unfortunately, field studies have been limited to a few case-specific
problems.  The following are among the most significant findings from
studies of building wake effects.
          Evans (19571 estimated the smoke visualized shape and size of
the cavity region for nearly two hundred variations of basic building
shapes in a wind-tunnel study.  He found that regardless of the height
of the building the pattern of the air going over the top of the buildings
appeared the same.  Examination of the published sketches shows the
cavity to extend from the ground vertically to about 1.5 times the
height of the building.  In the case of pitched roofs the height scale
should be taken as the height of its apex.  When the width of the building
was increased from 1 to 8 times its height, the downwind extent of the
cavity increased from 2 to 5 times the building height.  As the width of

-------
the building was further increased to 28 times  its  height, the downwind
extent was found to increase at a somewhat smaller  rate to 9 times the
building,height.
          Wind-tunnel  tests defining the influence  of block-type structures
on smoke emissions from roof-mounted chimneys were  conducted by Lord,
e_t aj_. (1964).  An examination of their results shows that the height
of the cavity is nearly equal to the building height plus one-half times
the building height or width, whichever is less.   However, the maximum
vertical extent of the disturbed flow above the cavity was found to  be
equal to the building height plus up to 3 times the building height  or
width, whichever is less.
          Halitsky (1958) reviewed several wind-tunnel studies of flow
near structures.  One of the studies (Halitsky, et_ aj_._  1963) demonstrated
that the wake in the lee of a rounded building  is not as great as that
found in a study of sharp-edged buildings (.Halitsky, 1963).  Meroney and
Yang (1971) found that for a stack less than 1.5 times the building
height the plume was downwashed into the lee side of the building.   When
the stack height was increased to 2.0 times the height of the building
the influences were greatly diminished.
          A formulation that prescribes the stack height sufficient  to
avoid significant building influences has been  presented by Lucas (.1972)
and Briggs  0973)..  They state that a stack should equal the height  of
the building plus 1.5 times the height or width, whichever is less.
Snyder and  Lawson (19761 in a series of wind-tunnel tests showed that
this formulation is adequate for a stack close  to a building whose
height is three times its width, and for a building whose width  is twice
its height.
                                 10

-------
          Peterka and Cermak (.1975) present an evaluation of mean velocity
and turbulence characteristics in the wake of buildings based on wind-
tunnel studies.  They found for wider buildings, that the mean velocity
defect and turbulence excess did not begin until 3 to 5 building, heights
downstream while the decay began almost immediately downstream of tall,
narrow buildings.  Differences were found in the flow behind a rectangular
shaped building (height to width ratio of 2.44) when oriented perpendicular
to the approach flow compared to that when oriented at 47 degrees to the
approach flow.  The mean velocity defect decayed fairly rapidly over the
first 20 building heights in both cases.  However, for the 47 degree
case, an excess of 3 to 4 percent of the freestrearn velocity remained
constant to 80 building heights downwind.  No evidence of a turbulence
excess or defect was found at such a great distance.  The existence of a
mean velocity defect to 80 building heights is believed evidence of a
vortex pair with axes parallel to the flow direction which are a remnant
of the corner vortices formed at the leading roof corner.  The vertical
profiles of mean velocity defect and turbulence intensity excess which
are reported for the perpendicular case, show values less than 5 percent
at all heights greater than 2.5 times the building.
          Hansen and Cermak (.19.75)_ and Woo, Peterka, and Cermak (.1976)
present additional wind-tunnel measurements of mean velocity and turbulence
characteristics in the wakes of structures.  The results are similar to
those discussed above.  The downstream extent of the recirculation
.Ci.e., cavity! region determined from mean velocity and turbulence
measurements is identified in Woo, Peterka, and Cermak 0976) for a
                                  11

-------
range of model  sizes and test conditions.   In  most cases,  th.e downstream
extent was found equal  to 3 to 5  building  heights  except for tall,
narrow structures whose downstream extent  was  much less.
          Robins and Castro (.19771 examined the  wind-tunnel  flow field
in the vicinity of a model  cube.   The  flow around  the cube was found to
be highly dependent on  orientation.  Strong vortices  generated by the
top leading edges were  found for  an approach flow  at  45 degrees to the
building edge.   They found the cavity  region to  extend to  1.5 times the
building height for an  approach flow perpendicular to the  building edge
and to 2 times  the building height downwind for  an approach flow at 45
degrees to the  building edge.  The downwind zone,  where the flow was
significantly affected, extended, however, to  5  building heights for
both cases.  The effluent from a  stack 2.5 times the  building height
having a stack  exit velocity 3 times the wind  speed,  was found to be
insignificantly affected by the building for an  approach flow perpendicular
to the building edge and to result in  a 20 percent increase in maximum
ground-level concentrations for an approach flow at 45 degrees to the
building edge.
          Huber and Snyder Q976). evaluated a  series  of wind-tunnel
studies designed to examine building wake  effects  near a building whose
width was twice its height.  The  size  of the cavity was found to be
approximately 1.5 building heights above ground  level in the vertical
and 2.5 building heights downwind.  In evaluating  the building influence
on dispersion,  aerodynamically generated turbulent flow was found to
rapidly decay in the region 3 to  10 building heights  downwind.  The most
                                 12

-------
significant disturbed flow occurred within 5 building heights downwind.



A significant building influence on ground-level concentrations was



found for cases with the stack less than 2 times the building height.



The building influences were found to be significantly reduced for a



stack 2.5 times the height of the building.



          In the vicinity of building structures where mechanically



generated turbulence dominates the undisturbed atmospheric flow, wind-



tunnel 'modeling has been found to be very reliable.  However, near the



outer boundaries of the wake, differences can be signficant.  In the



above early studies of Evans 0957) and Lord, e_t a_l_. (1964) no attempt



was made to simulate an atmospheric boundary layer.  Thus, preference



should be given to the results in the most recent studies.



          A review and evaluation of the current literature as reflected



above and in the annotated bibliography reveals a consensus that the



height of the cavity downwind of structures extends to the height of the



structure plus 0.5 times the height or width, whichever dimension is



less.  However, significant influences on plume behavior are found to



extend farther.  The well established 2.5 times rule is found to be the



consensus as the stack height necessary to avoid significant effects for



buildings whose projected width is greater than its height, although



individual studies show some deviation.  For tall buildings, where the



width is less than the height, the stack height need only be equal to



the height of the building plus 1.5 times its width.  Thus, the good



engineering practice stack height has been determined to be equal to the



height of the structure plus 1.5 times the height or width, whichever is
                                  13

-------
less.  This determination is  most  applicable  to  sharp-edged structures.



The extent of significant effects  for  rounded structures  are likely not



as great as those for sharp-edged  structures, although  there is  very



little information available.



          The downwind extent of the highly  turbulent region where there



are significant effects-is,  unfortunately, not as  well  defined.   Based



on the current literature, it is recommended  that,  for  the purposes of



determining GEP stack height,  the  downwind extent  of the  highly  turbulent



region be taken downwind of the  lee  side  as  5 times the height or width



of the structure, whichever is less, i.e., 5L from Equation (1).  This



choice is most applicable to  a structure  whose width is less than 10



times its height.  In situtations  where  the  structure is  wider than 10



times its height, there may be significant adverse effects extending



farther downwind.  The distance, 5L, generally corresponds to the cavity



length.  Most sources that are so  close  to a  structure  will likely be



greatly affected if their height is  less  than GEP  stack height as determined



above.  Sources at increasingly  greater  distances  would need a decreasingly



lower stack height in order not  to be  significantly affected. This



would be especially true for  highly  buoyant  sources whose emissions



would rapidly rise to heights well  above  the  disturbed  flow.  General



rules for defining a GEP stack height  for sources  at distances greater



than 5L are not presently feasible.  Where concern exists for possible



significant effects on sources greater than  a distance  5L, a wind tunnel



or field study should be conducted unless some reference  to a similar



study is available.






                               14

-------
          Evaluation of the wind-tunnel results of Evans (1957)  indicates
that for extremely wide buildings the maximum extent of adverse  effects
likely do not extend beyond 10 times their height.  In the wind-tunnel
studies and literature review reported by Kuber, et a!.  (1976) on flow
over two-dimensional obstacles a maximum extent of 10 times the  obstacle
height was found, for the cavity region, except in the case of very thin
obstacles where the extent was found to be much greater.  Hosker (1979)
has made an extensive review of the literature and has developed an
empirical estimation procedure for cavity length behind two and  three
dimensional sharp-edged rectangular buildings.  His presentation could
be used as a guide to indicate where a demonstration study for credit
beyond 5L is likely justified.  However, additional factors which are
not presently understood may effect the downwind extent of significant
influence.  Again, very little information was found for rounded structures-
which are unlikely to have as great a downwind influence as do sharp-
edged structures.  However, hemisphere shaped obstacles and sharp-edged
obstacles placed with a 45 degree orientation to the approach wind have
been both shown by Peterka and Cermak (1975). to have weak vortex patterns
which may persist far downstream.  It is not known, however, what effect
the weak vortex pattern could have on emissions from stacks.
     2.3  Quantitative Rationale for GEP Equation and Excessive Concentration
          Little of the literature on building effects presented above
and in the annotated bibliography contains specific data that can be
used in evaluating building influences.  Design stack height near buildings
has been based mostly on theory and experience with minimal supporting
                                 15

-------
 data.  Also, some of the data  available cannot be used because no measurements
of concentrations in the absence  of buildings  were taken for comparison.
Specific data available from the  literature concerning cavity and wake
height are summarized in Figures  3 and 4.
          In Figure 3, the cavity height (h )  is found to be well represented
by
                         hc = H + 0.5L.                (2).
The scatter of data appears evenly distributed about the line with slope
equal to 1.0.  The three sets of  data included in Figure 3 were taken
from wind-tunnel studies where  smoke was used  to visualize the region
where flow was circulating.
          Figure 4 presents data  from studies  defining the necessary
stack height in the absence of  any plume rise  to avoid some wake effect.
These data are only qualitatively useful since no measure of the significance
of the effect on air quality problems can be inferred.  The wake height (h  )
estimate has been used above to define GEP stack height as formulated by
                         h  = H + 1.5L.                (3).
The data from Meroney and Yang  (1971). and Lord, e_t al_. (1964) came from
observations of the plume center!ine visualized through smoke.  The wake
height estimate was defined as  the minimum plume centerline height found
to be unaffected by the building.  The other data are from an examination
of vertical concentration profiles.  For these data, the wake heights
were defined as the plume center!ine height where profiles both with and
without the building were judged  to be essentially the same.  One must
6e very careful in interpreting the data in Figure 4.  The visualized

                                 16

-------
   2.6
   2.4
   2.2
   2.0
   1.8
<
(—
o

o
cc
u.
a
LU
   1'6
in
LU
   1.4
   1.2
   1.0
   0.8
          OLORD,ET AL (1364)
       — DHUBER&SNYDER (1976)         	

        (~A-H EVANS (1957)

          CAVITY HEIGHT ESTIMATED FOR
       — 90 BLDG. TYPES. AVG = 1.5         	
                     STD DEV = 0.36
          hc = H+0.5L (EQ2)              Q
          WHERE: hc = CAVITY HEIGHT
       	        H = SLDG.HEIGHT        	
                 L  = LESSER DIMENSION
                      (BLDG. HEIGHT OR WIDTH)
     0.8        1.0       1.2        1.4
                hc/H (EQUATION 2)
                                       1.5   1.6
   Figure 3. Cavity height estimate.
                     17

-------
<
Q
    4.5

    4.4

    4.2

    4.0

    3.8

    3.6

    3.4

    3.2

    3.0
DC  2.8
LL.
2  2-6
    2.4
co  22
uu   •
 §
.c
3:  2.0

    1.8

    1.6

    1.4

    1.2

    1.0
O UKEGUCHI, ETAL. (1967)
A HUBER&SNYDERU976)
O SNYDER & LAWSON (1976)
O MERONEY& YANG (1971)
O JENSEN & FRANK (1965)
A SHERLOCK & STALKER (1940)
O LORD, ETAL. (1964)
                               hw=H-H.5L (EQ3)
                            WHERE:  hw = WAKE HEIGHT
                                    H =BLDG. HEIGHT
                                    L = LESSER DIMENSION
                                       (ELDG.HEIGHT OR WIDTH)
                                                                         s
                                                                         o
                                                                         a
                                                                         o-
      1.1       1.3        1.5       1.7        1.9       2.1
                                   hw/H (EQUATION 3}
                                                               2.3
                                                          2.5
 Figure 4. Wake height estimate.
                                          18

-------
studies can be strongly biased by the observer's eye and are extremely



sensitive to the density of the smoke.  The information from concentration



profiles is influenced strongly by where the traverse through the plume



is made and the judgment in determining what constitutes a significant



concentration difference.  In all these studies a higher stack would



have been required if the objective were to determine the height at



which there was no building wake effect on the emissions. 'Most of the data



presented in Figure 4 came from studies which did not fully consider



proper simulation of atmospheric flow.  Influences due to building



effects would be diminished in highly unstable and/or turbulent atmospheric



conditions.



     The data presented in Figure 4 show Equation 3 to approximate the



lower bound of these measurements.  Although the consensus opinion in



the scientific literature strongly supports using Equation 3 to determine



GEP stack height, actual studies could show the need for a much taller



or lower stack depending on one's interpretation of what is a significant



influence and on the effect of possible plume rise.  To more precisely



define that height for a specific stack, ground-level measurement both



in the wake of and in the absence of the building are needed to assess



the increase in maximum concentrations.  The ground-level measurements



must be sufficient to determine the location of the maximum concentration



which may occur at a different position in the wake of the building than



found in absence of the building.  The increase in maximum concentration



is simply the difference between the maximum concentration found in the



wake of the building and that found in absence of the building.  This





                                 19.

-------
concentration increase can be  assessed  to  determine  whether the increase
constitutes an excessive concentration.  • In  practice,  successive runs
varying the physical  stack height would  be conducted until  the concentration
increase due to building influence is  less than  the  excessive concentration
criteria.
     Only three data  sets having  ground-level  measurements  that included
increased maximum concentrations  were  found  in the  literature which can
be used to determine  excessive concentration.  Snyder (1979) reported on
recent, additional EPA data at the May  30, 1979  public hearing on the
stack height regulation.  All  data are  presented in  Figures 5 and 6.  A
theoretical estimate  (Britter, e_t_ aj_.  1976)  of increased maximum is also
presented.  The theoretical estimate assumes the building is very much
wider than it is high, and should be considered  as  providing an upper
estimate.   For all data, the plume rise  was  very small and  thus plume
centerline height is  nearly equal to stack height.   In all  cases, the
simulated atmospheric flow is  likely typical of  that which  occurs for
high wind, neutrally  stable situations.   Thus  differences among the data
are due to change in  stack height, building  size, or building orientation.
     The maximum ground-level  concentrations downwind of the building
and in absence of the building were used to  form the concentration
ratios in Figures 5 and 6.  The maximum ground-level concentrations
occur naturally at different positions.   The data in Figures 5 and 6, as
presented by Snyder (1979)., show that  higher concentrations downwind of
buildings depend quite strongly on building  width.   Ground-level maximum
concentrations associated with a stack 2.5 times the cubical and the
                                 20

-------
ro
             Q
             ca
             O
Q
_l
m
                3.5  -
                 3  -
                2.5  -
1 i 1 i 1 1 1 I 1
	 BRITTER.ETAL. (197G)
1 THEORY: W»H
- O i
1 ' D HUBER AND SNYDER (1976): W/H =
1 \
\ *VV/H=°° ° UKEGUCHTETAL. (1967): W/H = 2
\ ^ A ROBINS AND CASTRO (1977): W/H =
— \ t
\SNYDER (1979):
^ A. W/H = 0.33
\ G W/H = 1
3 * 0 W/H = 3
\



2


1





                1.5
                           1.25
                          WHERE: Hs = STACK HEIGHT
                                  H = BLDG. HEIGHT
                                  W=BLDG. HEIGHT
                       1.5
1.75
2.25

HS/H
2.5
2.75
3.25
                      Figure 5. Comparison of increased maximum ground-level concentrations for 0° building
                      angle cases.
3.5

-------
ro
ro
H


3.5




d 3
Q
_J
OQ
O
x 2.5

d
Q
_J
CO
x 2

1.5


1
I I I I I I I
	 BRITTER
. THEORY
i
, I I
, ETAL. (1976)
W»H
^^
A A ROBINS AND CASTRO: W/H = 1
» \
\ » SNYDER

(1979):
\ \ A W/H =0.33
~ \ ^ • W/H = 1
\ v • W/H = 3
\ \ W/H = - WHERE;
\ \
\ \
\ \
N, \
N. ^-H
A "^-^
\ ^^"^^A
X. ••— 	
A 	

I I Is* I I I I
1 1.25 1.5 1.75 2 2.25 2.5 2.75
	
Hs = STACK HEIGHT
H = BLDG. HEIGHT _
W = BLDG. WIDTH
i

~"~
I
* -— Jg, _
I """"
,
— — o
I I
3 3.25 3.
                                                              HS/H


                             Figure 6. Comparison of increased maximum ground-level concentrations
                             for 45° building angle cases.

-------
wide buildings oriented perpendicular to the approach wind (.i.e., zero



degrees building angle) are found to be increased by roughly 20 to 40



percent by the building wake.  The theoretical estimate suggests an 80



percent increase as an upper limit.  The data for the same buildings



oriented 45 degrees to the approach flow are found to have maximum



ground-level concentrations increased by roughly 40 to 80 percent.  The



differences are due to the presence of longitudinal vortices in the wake



of buildings having a 45 degree orientation to the approach wind as



discussed in Section 2.2.  The 80 percent increase found for the building



with W/H = 3 and having a 45 degree building angle very likely represents



the maximum effect of changing building orientation since, for wider



buildings, the longitudinal vortices generated at the sides of the



building would be less likely to interact.  Also the data are for a



source centered on the building and having no plume rise.  These conditions



should result in the greatest potential effect.



     Thus, it is anticipated for most situations, that maximum ground-



level concentrations downwind of building structures should not be



increased by more than 40 to 80 percent if the stack is equal to 2.5



times the building height.  Data for the tall thin building (W/H = 0.33)



shows that a stack much less than 2.5 times the building height is



needed to avoid increases.  GEP stack height as given by Equation 1 is



equal to 1.5 times th.e building height for the 0 degree building angle



case and equal to 1.7 times the building height for the 45 degree building



angle case.  The increase in maximum ground-level concentration from



such stack heights was found by Snyder (J9.79) to be increased by less



than 40 percent.





                                23

-------
     2.4  Terrain Influences
          Elevated terrain can  be much  larger than  most building structures,
Atmospheric phenomena on these  scales  can have a  great influence on the
development of aerodynamic forces,  beyond those found in the wake of
low-lying structures.  Very few definitive evaluations of the extent 'of
significant adverse effects in  the wake of terrain  obstacles are found
in current literature.
          The review of published field studies presented by Huber, e_t
al.  (1976) strongly supports the assertion that,  on the leeward side of
a mountain ridge, a circulating eddy with strong  downwash and dispersion
characteristics can exist.  Many of these studies are contained in the
annotated bibliography.  However, information that could define the
point where the flow separates  and the  size and extent of the cavity was
not found.  The point of separation appears to be a function of mean
flow speed and direction, atmospheric  stability,  downs lope and upslope
angle of the ridge sides, and the location of the ridge with respect to
surrounding terrain.
          For a particular situation,  the greatest cavity occurs when
flow separation occurs at the ridge apex.  Both field studies and fluid
modeling results confirm a natural expectation that the more obtrusive
the ridge, the larger the cavity region.  Obstructions with salient
features should exhibit definite separation at their edges under all
atmospheric conditions.  The size of the cavity region is greatest for
isolated ridges with steep sloping sides.  Stable atmospheric conditions
act to restrict the size and extent of the cavity region.  Under highly
                                 24

-------
stable flows other phenomena, such as lee waves and rotors, may be
found.  Terrain features that most adversely affect flow are two-
dimensional in nature.  Lateral air motion around a hill under neutral
stability results in a smaller eddy size than would be observed for a
two-dimensional ridge.
          Sporn and Frankenberg 0966) and Frankenberg (1968) recognized
the potential for adverse terrain influences in the late 1960's when
their pioneering experience with tall stacks began.  A wind-tunnel study
was conducted for the Clifty Creek plant since preliminary evaluations
indicated that there would be unusual difficulties from an aerodynamic
standpoint.  An abrupt rise of the terrain to a plateau approximately
100 m above plant level was found in the prevailing downwind direction.
The authors indicate that the results of the wind tunnel study showed
that stacks with a gas exit velocity of 36 m/s and a height twice the
plateau height (200 m) would be adequate to insure that the plume would
not intercept the boundary layer flow along the hillside and be immediately
brought to the ground.   The Kyger Creek plant presented no special
terrain problems so the stack height was determined from diffusion
calculations only.  The results of the analyses at Clifty Creek and
Kyger Creek were used as a guide in determining the necessary stack
design for newer facilities.  For example, the stacks at the Cardinal
Power Plant were constructed 251.8 m high; this makes them about  1.5
times the height of the surrounding terrain, Frankenberg, e_t a]_.  0970).
          Williams and Dowd  09.69.).. report that wind-tunnel studies of
gaseous diffusion have been used in many cases to help determine  stack
                                 25

-------
heights.   It has been observed,  however,  that  for  scaling ratios larger
than 600:1,  consistent and repeatable  results  become  difficult to obtain.
          A  recent study," "Plume Dispersion  in Complex Terrain," by
Johnson and  Mage 0978) was found to provide some  specific cases applicable
to assessment of potential terrain effects  for two American Electric
Power generating plants.  The  stack of the  Mitchell  Power Plant is more
than 2.5 times higher than the maximum terrain features in the vicinity
of the plant, while the stacks at the  Kammer Power Plant are nearly
equal to the elevations of the surrounding  terrain.   The horizontal
spread of the plume from the stacks of the  Kammer  Power Plant were found
on the average to be twice as  large as the  spread  found for the Mitchell
Power Plant.
          Because of the complex air flow over terrain and the general
uniqueness of each situation,  no simple definition of GEP stack height
is possible as has been recommended for building and other structures.
Until further studies better define the extent of  the region where
significant terrain influences can affect nearby sources, determination
of GEP stack height in the vicinity of terrain obstacles should be made
on a case-by-case basis.
     2.5  Minimum Stack Height
          In the case of very low structures or where there is essentially
no structure to which a stack is attached,  application of the 2.5 times
rule may yield answers which have little or no meaning.  Isolated release
points may require some physical height for security, safety or other
public health reasons.  Excessive ground-level concentrations may result
                                 26

-------
from low level releases, due to adverse meteorological phenomena in the
lower few tens of meters above the surface.  The specific height of the
lower few tens of meters above the surface, often called 'the surface
boundary layer,' varies not only with certain meteorological factors but
also among the definitions used by micro-meteorologists such as Sutton
Cca. 50 m}(.1953), Busch (30 m or so) (1973), and others.  In this layer,
the vertical atmospheric structure is largely a function of thermal and
mechanical turbulence generated at the surface, i.e., surface heating by
the sun or cooling by terrestrial radiation, and the surface roughness
caused by obstacles to air flow.
          To minimize the influences of these natural atmospheric
effects, one alternative is to consider that good engineering practice
should not preclude the construction of stacks up to a reasonable height
of 65 meters.  This will certainly minimize the deleterious effects of
stable and/or stagnant conditions, and allow reasonable dilution to take
place in the short travel time to nearby locations by permitting a
wider spectrum of atmospheric eddy sizes to act in the dispersion process
significantly without contributing to problems which arise from long
range transport and transformation of pollutants.  It should be noted,
however, that reasonable stack heights will not eliminate instantaneously
high concentration peaks associated with looping plumes.  Eigsti (1979)
shows that emissions from stacks whose heights are less than 65 m are
not likely to contribute significantly to  the overall loading of sulfate
in  the atmosphere.  However, for taller stacks, the  increase in height
can contribute  significantly to additional sulfate formation and transport.
                                   27

-------
This should also apply to other potential  chemical  transformation mechanisms
in the atmosphere.
          Thus, it is recommended  that the Equation (.1),  H  = H  + 1.5L,
be applied unless the resulting height is  less  than 65 meters.  If this
is the case, a stack height credit up  to  65 meters  height could be
allowed.
                                 28

-------
3.0  DETERMINATION OF GEP STACK HEIGHT
     3.1  Initial Assumptions
          GEP stack height is designed to insure that emissions from a
stack do not result in excessive concentrations as a result of aerodynamic
effects from nearby structures or terrain features.  It is assumed that
the wind speed and direction that may result in wakes, eddies or downwash
are possible for stacks less than GEP height.  Excessive ground-level
concentrations will not result when:  (1) the emission point is well
above the disturbed flow, C2) the effluent rise is sufficiently great
to keep a significant part of the effluent plume above the disturbed
flow or (3) the wind direction places the stack outside the area of
disturbed flow.
          GEP stack height as determined by Equation 0) does not consider
plume rise.  However, plume rise should not be significant in the
determination of GEP stack height if, under high wind speeds, plume rise
near the source is negligible.  For most sources, even those with a
relatively high exit velocity, a wind speed of 15-20 m/s will result in
significantly reduced plume rise and thus increase the potential for
adverse effects and the need for some stack height.  Therefore, the
critical conditions for determining GEP stack height for most sources
are considered likely to be high winds associated with neutral atmospheric
stability with little plume rise near the sources.
          Sources situated within 5 times the lesser dimension of the
building structure (.1-6-» 5L downwind from the trailing edge of the
structure} are presumed near enough to the building to base GEP stack
height on Equation 1.  Where concern exists for possible significant
                                29

-------
effects on sources at greater  distance,  a  wind  tunnel  or field study
should be conducted unless some  reference  to  a  study  for a  similar
situation is available.  -Likewise  a  demonstration  could be  used to
demonstrate insignificant effects  for  stacks  of height less than that
given by Equation 1.   A  demonstration  is necessary for determining GEP
stack height near elevated terrain.  A framework for  demonstrating GEP
stack height is presented in Section 3.5.   The  "Guideline for Use of
Fluid Modeling to Determine Good Engineering  Practice Stack Height"
(.EPA, 1980) provides  specific  guidance to  be  followed.  Field studies
should be designed and evaluated on  a  case-by-case basis since the
complexities of field studies  do not make  it  feasible to propose specific
criteria.
     3.2  Simple Structures
          GEP stack height has been  defined to  be  equal to  the height of
adjacent or nearby structures  plus 1.5 times  the structure  height or
width, whichever is less.  Both  the  height and  width  of the structure
are determined from the  frontal  area of the structure, projected on a
plane perpendicular to the direction of the wind.   If the structure is
asymmetrical, the GEP stack height should  be  based on the plane projection
lying upwind from the source  (.stack) which results in the greatest
justifiable height.
          In some situations  the projected area may be very irregular,
thus resulting in a multiplicity of scales.  Note  that structural protuberances
are seldom a significant factor in determining  GEP stack height.  For
the purpose of determining GEP stack height,  nearby is limited to 5
structure heights or widths, whichever is  less, downwind from the trailing
edge, of the structure.
                              30

-------
          Figure 7 illustrates applications to three types of buildings.
A GEP stack should have a height equal to the upper edge of the shaded
regions of the vertical cross-section if the stack lies within the
associated shaded region of the horizontal cross-section.  Note for both
the tall, thin structure and the short, long structure, the expected
sphere of influence is less than that found for the moderately tall
cubical structure.
          3.2.1  Low Structures
                 The nearby region of adverse influence downwind, R, for
a uniform low structure (one whose width all around is greater than its
height) is easy to determine.  It is 5 times the structure height,
downwind in all directions from the trailing edge of the building.  The
vertical extent of disturbed flow is 2.5 times the structure height
throughout the entire vicinity of the structure.  Thus GEP stack height
is defined as 2.5 times the structure height.  This determination for a
low structure is presented in Figure 8 where the sphere of influence is
outlined.  Figure 8 also depicts the maximum projected structural width, W,
 affecting each of the four given sources.  Mote that these projected
widths are only valid for a wind which is perpendicular to the actual or
the cross sectional surfaces.  Since the projected width for all directions
is greater than the height, the width scale is not a factor in determining
GEP stack height.
          3.2.2  Tall Structures
                 The width scale becomes the significant factor in
determining  GEP stack height whenever the structure is taller than it
is wide.  In Figure 9, the structure is tall and thin Cone whose lateral
dimensions are  less than its height)..  The determination of the structural
                                 31

-------
                        TOP VIEW

                                                   H2H
                                                   H1-
                                                   HS-
GEP2=1.15 H2
GEP^Z.5 H!
GEP=2.5 H
                                     •W2
                       SIDE VIEW
Figure 7. Determination of the GEP stack height near three types of structures.
                                 32

-------
                                                    TOP VIEW
                      i . i : -i rrn .nit iTTTTTrrrrn-JTrrrrn n 11 rrrrrn -
2.5H-GEP
STACK HEIGHT
                                                    SIDE VIEW
Figure 8. Determination of the maximum projected structure width and
associated region of adverse influence for four stacks  placed near a low
structure.
                              33

-------
                 Rl = 5W}
                                                  1.75H-GEP STACK HEIGHT
                                                 ' BASED ON WIND DIRECTION 3
Figure 9. Determination of the projected structure width and associated region of
adverse influence for four possible wind directions near a tall, thin structure.
                                     34

-------
width and resulting presumed aerodynamically effected nearby region for



four wind directions is presented in Figure 9.  The nearby region, R, is



5 times the projected width, downwind from the trailing edge of the



structure.  Note that the extent is highly dependent on the wind direction.



The GEP stack height for a tall structure is determined to be equal to



the structure height plus 1.5 times the projected structure width.



Thus, GEP based only on the side view in Figure 9 wou.ld be equal to 1.75



times the hetght of the.structure.  .Since the projected width of the



structure is dependent on the wind direction, all directions projecting



downwind towards the source need to be assessed.  The maximum allowable



GEP for sources near a tall structure is then equal to the structure



height plus 1.5 timas the maximum projected structure width.



     3.3  Complex Structures



          3.3.1  Tiered Structures



                 Figure 1Q presents a more complex, tiered structure.



For this situation, tier 1 by itself has a nearby region, R, extending



downwind for five heights.  Th.e addition of tier 2, which is equal in



height to tier  1, causes both the vertical and downwind extent of  the



region of significant influence to double since the height scale is the



overall height  which still is less than the width.  The projected  area



downwind of tier 3 which is placed above tier 2 has a height 4 times



greater than its width, as can be seen from examination of Figure  10.



However, the downwind region of influence of tier 3 extends downwind



less than the  influence of tier 2.  Should a source be located directly



downwind of tier 3, although out of its influence, GEP is then based on



the influence  of tier 2.  Note that the vertical and downwind extent of



influence of tier 2 totally engulfs the influence of tier 1.  However,



the across flow extent  is, of course, greater.

-------
 Wi=5HI
W3=0.25K3
                         TOPViEW
                     WIND DIRECTION
GEP3=1.4H3.
r; cp«-7 cu«
GEPi=2.5Hi-



IT77 TTTf, <*///////// rTTT7\
|3 /,/,„/,„„/,„„„.
[Jn _--.-.-,-.. J >
3 ^1
1 ^
A\\\\ in rTTrii iiriiimt
^ tl^S 1 ;
J

/
/
/
i
////// y /////////'//..'
\
\
\

                            SIDE VIEW
Figure 10.  Variation in the determination of the region of adverse
influence for additions to a tiered structure.
                                36

-------
                 For the situation presented in Figure 10, GEP stack

height is equal to GEP-, (1.41-L) for all sources downwind of tier 3 and

placed within R^.  GEP for sources farther downwind but not beyond R? is
      /'
equal to GE?2 (.Z.SHg).  For sources outside of the projected width of

tier 2,  however within the projected width of tier 1 and downwind distance


R-l,  the GEP stack height is equal to GEP] (.2.5^).  Other orientations

of the building to the wind can result in different determinations of

GEP stack height where the projected width is less than its height.  For

the building design in Figure 10., only the influences of tier 3 change

th.e GEP determination since only its projected width is less than its

height.

                 The influence of tiers has been assumed to be complementary.

Very little information relative to such situations is found in the

present scientific literature.  The influence of tiers may not be exactly

complementary since additional tall tiers, similar to tier 3 in the

above example, may result in some streamlining of the flow around the

lower tiers and thus some reduction in their effects.  Since such effects

are likely minimal, it is recommended  that, until further evaluations

are reported, the effects of tiers should be considered totally additive


as presented here.  A demonstration should be provided for special

situations where support for the above assumption is desirable or in

situations where there is concern for  additional complications.

          3.3.2  Tapered Structures

                 A tapered structure is presented in Figure 11.  This


situation differs from the tiered structure only in that there is now a

continuous array of widths and heights to consider.  The four selected

heights and widths are representative.  Above the Hp level of the structure


                                 37

-------
CO
Co
                    W2 = 2H2

                    (A'3 = 0.5H3

                    W4 = 0.14H4
                                                                    	  	 _   .«< _     3
                                                                   H^                   i.iniui!
                                                                                                  SIDE VIEW
                     Figure 11.  Determination of the region of adverse influence for a tapered structure.

-------
the width is less than its height.  Since the width decreases further



v/ith increasing height both the determination of GEP stack height and



the region of adverse influence which are now functions of the width



have little effect on the determination of GEP stack height.  Figure 11



shows that the overall determination of the region of adverse influence



is set by the H~ level and the overall GEP stack height set by the hU



level for sources within R-, are only slightly greater than.that set by



the \\2 level.  The overall GEP stack height is the outermost boundary of



the outlined regions presented in Figure 11.  The region near the top is



not a significant factor in determining GEP stack height in this situation



since the width is negligible.



              Frgure 12 presents a cooling tower structure.  In this



application, use of the width at the top of the cooling tower was found



to result in the outermost boundary of influence and thus the overall



determination of GEP stack, height.  The downwind region of adverse



influence is, therefore, given in Figure 12 as 5 times the width at the



top of the tower structure and the GEP stack height is given as the



height plus 1.5 times the width at the top.



              As is the case for tiered structures, very little information



relative to such situations is found in the present scientific literature.



In addition, as pointed out in the literature review presented in Section



2.2, the influences of rounded obstacles may not be as great as that



found for sharp-edged buildings as a result of variation of the flow



separation line.  The above determination is recommended until further



evaluations are reported  in the scientific literature.  A demonstration



should be provided for special situations where support for the above



assumption is desirable or in situations where there is concern for



additional complications.        3^

-------
o
                                                                        R = 311 •
                                                                                                     771
                                                                                                      n
                                                                                                     //I
                                               WIND DIRECTIO
t>
                                                                                              TOP VIEW
                          1.9H-GEP STACK HEIGHT
                                            \niiuniiiiiii in iiuiin 1'im ill I tu in /////////////////////,
                                          I
                                                                                              SIDE VIEW
                     Figure 12. Determination of the region of adverse influence near a cooling tower structure.

-------
              3.3.3  Groups of structures
                     Figure 13 presents an evaluation of the region of
adverse influence downwind of a group of structures.  The top view shows
the projected downwind extent for three wind directions.  The effects of
adding building 3 is shown as the added region of influence beyond that
shown for building 2.  The downwind extent of the region of adverse
influence is equal to five times'the height or projected width of the
building, whichever is less.  The influence of nearby buildings is
assumed to be exactly complementary, similar to that shown for tiered
structures.  Where the projected widths of adjacent buildings do not
overlay but whose lesser projected dimension (height or projected width)
of either building is greater than the projected distance of separation,
treat the gap as if it were filled by a structure equal in height to the
lesser projected height.  This is demonstrated in the front view.  The
distance of separation between building 1 and building 2 is too large
while building 2 and building 3 are assumed to be sufficiently close to
be treated as a single building for purposes of determining GEP stack
height.  The side views shew all three buildings to be simply complementary.
GEP stack height based only on the front view and the side views is
presented in Figure 13.  As for single structures, the maximum allowable
GEP is equal to that resulting from an evaluation of all wind directions.
              The influence of groups of buildings has been assumed to
be complementary.  Very little information relative to such situations
is found in the scientific literature.  The above general procedure is
recommended until further evaluations are reported from which more
specific guidance may evolve.  A demonstration should be provided for
special situations where support for the above assumption is desirable
or in situations where there is concern for additional complications.
                                 41

-------
ro
                                                                                                       REGIONS OF ADVERSE INFLUENCE:
                                                                                                                WIND DIRECTION
                                                                                                                            DO"
                               Figure 13. Determination of region of adverse influence for a group of buildings.

-------
     3.4  Terrain Obstacles



          GEP stack height for new major sources and proposed stack



height increases for existing sources to minimize the adverse impact of



terrain obstacles should be determined on a case-by-case basis.   Field



studies designed to evaluate the specific situation under the variety of



adverse meteorological conditions are most preferred.  Where field



studies are not practicable, comparable fluid model (wind tunnel, water



channel, etc.) studies or mathematical analyses that are developed are



acceptable.



     3.5  Framework for Demonstrating GEP Stack Height.



          As presented above, case-by-case demonstrations should be used



for GEP determinations for terrain obstacles.  A demonstration should be



provided for special situations where support for the preceding  assumptions



is desirable or in.situations where there is concern for additional



complications near buildings.  A demonstration may be used to justify a



stack taller than the GEP definition CEquation 1) when excessive concentrations



are due to the influence of nearby structures.  Where field studies are



not practicable, comparable fluid model Cwind tunnel, water channel,



etc.I studies or mathematical analyses that are developed are acceptable.



In field studies and fluid model or mathematical simulation, some quantitative



evaluation of the obstacle's influence is a necessary part of demonstrating



GEP stack height.  Quantitative analyses are necessary since GEP stack



height is to be based on "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."



          Comparable fluid model studies require certain similarity



criteria to be considered.  Discussion of similarity criteria can be



                                 43

-------
found, for example, in Snyder (1981);  Snyder (1972);  Sundaram, et a!.
(1971); Cermak (1970); and Halitsky (1968).   Comparable mathematical
analyses that are developed must also  satisfy physical  laws and be well
supported by field or fluid model data.   Nondimensional parameters that
characterize the flow in the atmosphere  must be matched in the model
medium.  This consideration is necessary to  insure that the flow in the
medium accurately simulates that in the  atmosphere.
          Modeling simulations rely on the continuing development and
refinement of state-of-the-art techniques.  The specific criteria and
procedures for an adequate modeling GEP  determination are presented in
separate guidance documents.  Specifications for such fluid model demon-
strations are found in the "Guideline  for Use of Fluid Modeling to
Determine Good Engineering Practice Stack Height" (EPA, 1980).  This
guideline is based on a separate guideline entitled,  "Guideline for
Fluid Modeling of Atmospheric Diffusion" (Snyder, 1981), which reviews
the fundamental principles and practical applications of fluid modeling.
         >.
The aim of that guideline is to establish the capabilities and limitations
of fluid modeling, and to establish EPA standards for the conduct of
fluid modeling studies.
          As the state-of-the-art improves,  future guidance may require
additional data and/or specific critical assessments.  For this reason,
reviewing agencies should establish a  requirement that a study plan be
submitted prior to the conduct of the  study so that the latest EPA
quality assurance procedures and guidance will be considered.
          In some situations field studies may be desired in conjunction
with or as support for a fluid model study.   Proposed field studies
should be designed and evaluated on a  case-by-case basis since the

                                     44.

-------
complexities  of field studies do not make it feasible to propose specific
criteria.  The following discussion presents, generally, the essential
components for demonstrating a GEP stack height by a field study.
          The causeCs). and magnitude of the disturbed flow used to
justify the GEP stack height should be clearly identified.  In the case
of an isolated building, this can be easily accomplished by documenting
the release of visible smoke at ground level  and on top of the building
to demonstrate the general region of influence.  Influences caused by
atmospheric phenomena such as oscillations in the flow and inversion
breakup are not creditable toward determining a GEP stack height.
          A field determination of GEP stack height requires experiments
to determine the concentration patterns from two release points—one
with the structureCs). and/or terrain; the other in the absence of structure(s)
and/or terrain.  This means there must be a location near the site of
the source where the atmospheric flow is similar except for differences
caused by structures and/or terrain near the source.  A monitoring array
must be arranged to clearly identify the maximum concentrations downwind
of similar releases at both sites.  Meteorological instrumentation must
be placed upwind of both sites to show that the approaching atmospheric
flow is similar.  In areas where the upwind fetch at both sites is
similarly homogenous with no nearby obstructions such as buildings or
elevated terrain, one may expect similar approach flows.  A light wind,
stable atmospheric flow is very sensitive to external influences, often
resulting in great differences between even close sites.  Generally,
moderate to high wind speeds with near neutral stability conditions can
be expected to result in the more severe wakes, eddies and downwash.
                                 45

-------
46

-------
4.0  AIR QUALITY ESTIMATES
     When any dispersion model is used for determining an emission
limitation, it is the intent of the stack height regulation that the
stack height specified for use be no greater than the GEP height.  The
GEP stack height based on the physical configuration of the source and
any nearby structure or upwind terrain feature should be determined by
the procedures in the preceding section.
     Higher estimates of air quality impact may result for sources with
GEP stack, height lower than the existing stack height or those which do
not make full use of the terrain impaction allowance.  Therefore, for
certain sources, more stringent emission limitations than currently
required by SIPs may be required.  However, the guidance presented here
is consistent with the stack heights regulation and with the intent of
Section 123 of the 1977 Clean Air Act Amendments.
     In some cases, a greater GEP height may be justified, based on
nearby terrain effects.  Specifically, a GEP stack height based on upwind
        -.
terrain features near the source may be used for determining an emission
limitation if adequate justification is demonstrated through a field
study or fluid model study.  When the above procedures are inapplicable
or yield a GEP height less than 65 meters, the GEP height should be
specified as 65 meters for model input.
     In the event that the actual stack height is less than the GEP
height, the stack height specified for use with the dispersion model
should be limited to the actual height.  In that case, it is possible
that excessive pollutant concentrations may occur in the immediate
vicinity of the source due to atmospheric downwash, eddies and wakes
created by the source itself  or nearby structures or terrain features.
                                      47

-------
Such adverse effects should  be  accounted  for when  estimating  the air
quality impact of a source.   Guidance  concerning such  effects has been
provided in several reports  (Huber  and Snyder,  1976; Huber,  1977; Budney,
1977; and Bowers, e_t al_.,  1979).
     For GEP stack height, on the other hand, such adverse atmospheric
effects likely are avoided and  specific modeling techniques  can be
recommended for estimating the  air  quality  impact  of a source.   A simple
screening analysis may first be conducted to eliminate from  further
consideration those new sources that clearly will  not  cause  an air
quality problem.  Screening  procedures (Budney, 1977)  provide a conservative
estimate of maximum concentrations; i.e., a margin of  safety is incorporated
to insure that maximum concentrations  will  not  be  underestimated.  If a
more refined analysis is necessary  for a  new source or if a  SIP revision
is being considered for an existing source, the analysis  should be
consistent with techniques recommended in the "Guideline  on  Air Quality
Models"-(EPA, 19781.  The Guideline makes specific recommendations concerning
air quality models, data bases  and  general  requirements for  concentration
estimates.
                                    48

-------
Case Examples

     There are three  basic  situations  in which modeling analyses will be

applied to emissions  from GEP  stacks.

      Case 1:  Terrain less  than GEP height.

           Recommendation:   Apply 2 "flat-terrain"'dispersion modeling

           technique,  as  recommended in Section 4.3 of the  Guideline  on

           Ai.r Quality Models.   Use the GEP stack height as  the  specified

           physical  stack height input to the model.
  GEP
    l
     Flat  terrain is defined  as  a  setting without significant meteorological
     complexities, e.g..topographic features do not  exceed the physical  stack
     height of the source being  modeled.

                                     49

-------
Case 2:   Terrain Greater than GEP  height.
     Recommendation:   As discussed earlier,  a  GEP  stack is  theoretically
     high enough to avoid downwash,  eddies and wakes  caused by nearby
     structures and elevated terrain upwind  of the source.   However, even
     though the stack is tall enough, or  the source is  located so as to avoid
     adverse aerodynamic effects,  there  is still the  possibility of
     plume interaction with elevated terrain features further downwind.
     High concentrations may occur on the downwind elevated terrain due
     to the effluent plume coming  close  to or  impacting it.
          The steps for determining  the  stack  height  allowance to avoid
     excessive concentrations due  to plume impaction  are:
          1.  Calculate the maximum ground-level concentration using the
     GEP stack height in a "flat terrain" dispersion  modeling technique.
          2.  Calculate the maximum concentrations on each  terrain feature
     that is above GEP stack height.   Some techniques for estimating
     ambient concentrations on elevated  terrain have  been identified
     (Burt, 1977 and Egan, 1975) and should  be considered as discussed in
     the "Guideline on Air Quality Models."
          3.  If the maximum concentration on  elevated  terrain exceeds the
     maximum "flat terrain" concentration, calculate  the stack height for which
     the elevated terrain concentration  equals the "flat terrain" concentration.
     This is the maximum allowable stack  height for modeling and should be
     used in subsequent analyses if such  a stack is actually constructed.
          4.  If the source owner  or operator  does not  propose a maximum
     allowable stack height, further modeling  with the  proposed stack height
     may be required to adjust the "flat  terrain"emission limitation so that
     the ambient air quality standards and/or  PSD  increments are protected.
                                      50

-------
  CEP HEIGHT,
f
                      WIND
                 MAXIMUM CONCENTRATION fOfc'FLAT TERRAIN'1
  MAXIMUM

 ALLOWABLE
  HE\GHT
 SSF HEIGHT
*;•'/'; >  i '  i J  >-)
                           "MAX,
       WHEK 7L.V s li,   >THE CORReSP«»WPlN6- STACK HE.16-HT /S THE
                 j  /xnASt, >

                 AU-OWABLt Fop,
        IF PROPOSED OR ACTUAL STACK HE\G-HT JS NOT AT THE


        ALLOWABLE HEIGHT, XiAX MUST NOT EXCEED
                EXAMPLES FOR  CASE ^ .
                                    51

-------
Case 3:  Multiple Source Impacts
     Recommendation:  Many situations  are anticipated in which there
     will be significant contributions to ambient concentrations due to
     sources other than the one  in  question.   In  such cases, first
     estimate the air quality impact of the source in question, as
     discussed for Cases 1 and 2.   Then superimpose the air quality
     impact of the other sources  to estimate the  total air quality
     impact.  For the calculation of contributions from other sources,
     GEP-based emission rates should be used in conjunction with GEP
     or allowable stack heights  as  input to the modeling assessment.
     The emission limitation for the source in question generally should
     be determined such that the  National Ambient Air Quality Standard
     and allowable concentration  increments will  be met even after
     natural background and the  additive impact of other sources
     are considered.  Guidance is available for estimating contributions
     from other sources.  (Budney,  1.977 and EPA,  1978).
                                      52

-------
                            REFERENCES

Batchelor, G. K., 1967:  An Introduction to Fluid Dynamics.   Cambridge
     University Press, Cambridge,  (Great Britain}, 325-331.

Baumeister, T., E. A. Avollone, and'T. Baumeister, III,  (editors),  1978:
     Marks' Standard Handbook For  Mechanical  Engineers.   McGraw-Hill,
     New York, Chapter 18, page 16.

Beaver, S. H. (Chairman), 1954:  Report of Government Committee on  Air
     Pollution.  Her Majesty's Stationary Office, CHD.  9322,  November.

Bowers, J. F., J. R. Bjorklund, and C. S. Cheney, 1979:   Industrial
     Source Complex (ISC). Dispersion Model User's Guide.   (EPA-450/4-79-031),
     Environmental Protection Agency, Research Triangle  Park, NC  December.

Briggs, G. A., 1973:  Diffusion Estimation for Small  Emissions.  Atmospheric
     Turbulence and Diffusion Laboratory, MOAA, Oak Ridge,  TN, (Draft)
  '   ATDL No. 75/15.

Britter, R. E., J. C. R. Hunt and  J. S. Putteek, 1976:   Predicting
     Pollution Concentrations Near Buildings  and Hills.   Symposium  on
     Systems and Models in Air and Water Pollution, September 22-24,
     London.

Budney, L. J., 1977:  Procedures for Evaluating Air Quality Impact  of
     New Stationary Sources.  Guidelines for  Air Quality Maintenance
     Planning and Analysis:  Volume 10 (EPA-450/2-77-001, OAQPS Guide-
     line Number 1.2-029R), Environmental Protection Agency,  Research
     Triangle Park, NC  October.

Burt, E. W., 1977:  Valley Model User's Guide.  (EPA-450/2-77-018),
     Environmental Protection Agency, Research Triangle  Park, NC, September.

Busch, N. E., 1973:  On the Mechanics of Atmospheric Turbulence.  In
     I'orkshop on Micrometeorology, by American Meteorological Society,
     Science Press, Ephrata, PA. Chapter 1.

Cermak, J. E., 1970:  Laboratory Simulation of the Atmospheric Boundary
     Layer.  American Institute of Aeronautics and Astronautics, 3rd
     Fluid and Plasma Dynamics Conference, Los Agneles,  CA, June 29 -
     July 1, NO. 70-751.

Cermak, J. E., 1976:  Aerodynamics of Buildings.  In Annual Review of
     Fluid Mechanics, Vol. 8, Van Dyke, M. And W. G. Vincenti (Co.  Ed.),
     Annual Reviews,  Inc., Palo Alto, CA, 75-106.

Cramer, H. E., and J. F. Bowers, Jr., 1976:  West Virginia Power Plant
     Evaluation.  Prepared for U.  S. EPA Region III, Philadelphia,  PA,
     Hay.


                                 53

-------
Counihan, J., 1969:  An Improved  Method  of  Similating  an  Atmospheric
     Boundary Layer in a Wind  Tunnel.  Atmospheric  Environment,  3_,
     197-214.

Egan, 8. A. 1975:  Turbulent Diffusion in Complex Terrain.   Lectures on
     Air Pollution and Environmental  Impact Analysis,  American Meteorological
     Society, Boston, MA.

Eigsti, S.  L., 1979:  An Assessment  of the  Potential  Effects of Stack
     Height on Sulfate Formation  and Sulfur Deposition.   U.  S. Environmental
     Protection Agency, Office of Air Quality  Planning and  Standards,
     Research Triangle Park, NC  December,  18  pp.

Environmental Protection Agency,  1978:   Guideline on Air  Quality Models.
     (EPA-450/2-78-027) Office of Air Quality  Planning and  Standards,
     Research Triangle Park, NC  April.

Environmental Protection Agency,  1980:   Guideline for  the Use of Fluid
     Modeling to Determine Good Engineering Practice  Stack  Height.
     CEPA-450/4-80-015) Office of Air Quality  Planning and  Standards,
     Research Triangle Park, NC.

Evans, B. H., 1957:  Natural Air  Flow Around Buildings,  Research Report
     No. 59, Texas Engineering Experiment Station,  Texas  A&M College
     System.

Frankenberg, T. T., 1968:   High Stacks for  the Diffusion  of Sulfur Dioxide
     and other Gases Emitted by Electric Power Plants, Am.  Ind.  Hyd. Assoc.
     J_., 29, 181-185.

Frankenberg, T.'T., I. Singer, and M. E. Smith,  1970:   Sulfur Dioxide in
     the Vicinity of the Cardinal  Plant  of  the American Electric Power
     System.  Proc. 2nd Int. Clean Air Cong. Washington,  DC.

Halitsky, J., 1963:  Gas Diffusion Near  Buidlings.  ASHRAE  (Trans.),
     69_, '464-484.

Halitsky, J., 1968:  Gas Diffusion Near  Buildings.  Meteorology and
     Atomic Energy - 1968, D.  H.  Slade  (Ed.},  Chapter  5-5.

Hansen, A.  C. and J. E. Cermak, 1975:  Vortex-Continuing  Wakes of
     Surface Obstacles.  Project  Themis  Technical Report, No. 29,
     December 1975, CER75-76 ACH-JEC16.

Hawkins, J. E. and G. Nonhebel, 1955:  Chimneys  and the Dispersal of
     Smoke.  J. of the Institute  of  Fuel, 28_,  530-545.

Hosker, R.  P., Jr., 1979:   Empirical  Estimation  of  Wake Cavity Size
     Behind Block-Type Structures.  American Meteorological  Society 4th
     Symposium on Turbulence,  Diffusion, and Air Pollution,  Reno, Nevada,
     January 15-18, pp. 603-609.


                                  54

-------
Kuber, A. H. and H. H. Snyder, 1976:   Building Wake Effects  on  Short
     Stack Effluents.   Third Symposium on Atmospheric Turbulence  Diffusion
     and Air Quality,  Raleigh, MC, Oct.  19-22, 235-241.

Huber, A. H., W. H. Snyder, R. S.  Thompson,  and R.  E. Lawson, Jr.,  1976:
     Stack Placement in the Lee of a  Mountain Ridge.   U.  S.  Environmental
     Protection Agency, EPA-600/4-76-047, Research  Triangle  Park,  NC,  Sept.

Huber, A. H., 1977:  Incorporating Building/Terrain Wake  Effects  on Stack
     Effluents.  AMS-APCA Joint Conference on Applications on Air Pollution
     Meteorology, Salt Lake City,  Utah,  Nov.  29 - Dec.  2, 353-356.

Hunt, J. C. R., C. J.  Abell, J. A. Peterka,  and H.  Woo,  1978:   Kinematical
     Studies of the Flows Around Free or Surface Mounted  Obstacles;
     Applying Topology to Flow Visualization.  J. Fluid  Mech.,  3_,
     Part 1, 179-200.

Johnson1, F. G., and D. T. Mage, 1978:  Plume Dispersion  in Complex
     Terrain.  Presented at the Annual Meeting of the APCA,  Houston,
     Texas, paper No.  78-73.10.

Lord, G. R., W. D. Baines, and K.  J.  Leutheusser, 1954:   On  the Minimum
     Height of Roof-Mounted Chimneys, Results of an Exploratory Wind-Tunnel
     Study.  Report TP-6049, Technical Publication  Series, Dept.  of
     Mechanical Engineering, University of Toronto.

Lucas, D. H., 1972:  Choosing Chimney Heights in the Presence of  Buildings.
     Porceedings of the International Clean  Air Conference,  Melbourne,
     Australia5 May 15-18, 47-52.

Meroney, R. N. arid B.  T. Yang, 1971:   Wind-Tunnel Study  on Gaseous  Mixing
     Due to Various Stack Heights  and Injection Rates Above  an  Isolated
     Structure.  USAEC Report No.  COO-2053-6.

Peterka, T. A. and J.  E. Cermak, 1975:  Turbulence  in Building  Wakes.
     Fourth International Conference  on Wind Effects  on  Buildings  and
     Structures, London, United Kingdom.  Colorado  State  Univ.
     Report No. CPE 74-755, AP-JEC 34.

Robins, A. G. and  I. P. Castro, 1977:  A Wind Tunnel  Investigation  of
     Plume Dispersion in the Vicinity of a Surface  Mounted Cube-I'.  The
     Flow Field, II.  The Concentration Field.  Atmospheric  Environment,
     17., 291-311.

Scorer, R. S., 1968:  Air Pollution.   Pergamon Press, Oxford, England,
     107-123.

Snyder, W. H., 1972:  Similarity Criteria for the Application of  Fluid
     Models to the Study of Air Pollution Meteorology.   Boundary-Layer
     Meteorology,  3_, 113-134.

                                  55

-------
Snyder, W.  H.  and R.  E.  Lawson,  Jr.,  1976:   Determination  of a  Necessary
     Height for a Stack  Close to a  Building—a Wind Tunnel  Study.
     Atmospheric Environment, 10, 683-691.

Snyder, W.  H., 1979:   Testimony  on  Behalf of the U. S.  Environmental
     Protection Agency at the Public  Hearing on Proposed Stack  Heights
     Regulations, May 31  1979.

Snyder, W.  H., 1981:   Guideline  for Fluid Modeling of Atmospheric  Diffusion.
     (EPA-600/8-81-009).   U.  S.  Environmental Protection Agency,  Environmental
     Sciences Research Laboratory,  Research  Triangle Park,  NC  27711.

Sporn, P. and T. T.  Frankenberg, 1966:   Pioneering Experience with High
     Stacks on the OVEC  and American  Electric Power Systems, Presented
     at the International Clean  Air Congress, London, Paper No.  IV/9.

Sundaram, T. R., G.  R. Ludwig, and  G.  T.  Skinner, 1979:  Modeling  of the
     Turbulence Structure of the Atmospheric Surface Layer.  American
     Institute of Aeronautics and Astronautics, 9th Aerospace Sciences
     Meeting, New York,  NY, No.  71-136.

Sutton, 0.  G., 1953:   Micrometeoroloqy McGraw-Hill, NY.

Sutton, 0.  G., 1960:   Discussion before the  Institute,  in  London,  23.d,
     May 1960.  J. Institute of  Fuel,  33_, 495.

Ukeguchi, N., H. Sakato, H. Okamoto,  and Y.  Ide,  1967:   Study on  Stack
     Gas Diffusion.   Mitsubishi  Technical Bulletin No.  52,  August.

Williams, D. H., Jr., and J.  T.  Dowd,  1969:  Design and Construction
     Features of the 1600 MW Mitchell  Plant. Combustion.  August,  19-23.

Woo, H. B.  C., J. A.  Peterka, and J.  E.  Cermak, 1976:   Wind-tunnel
     Measurements in the Wakes of Structures, Technical  Report  for NASA
     Marshall Space Flight Center,  NASA CR-2806,  Colorado  State Univ.
     Report No. CER75-76HGCW-JAP-JEC40.
                                       56

-------
                            APPENDIX A

                      ANNOTATED BIBLIOGRAPHY
Sherlock, R. H. and E. A. Stalker, 1940:  The Control of Gases in the
Wake of Smokestacks.  ASME Journal ,  62_, 455-458.

     A wind-tunnel investigation was used to determine whether an addi-
tion of the height of the existing stacks would prevent downflow of
stack gases into the area surrounding the Crawford Station of the
Commonwealth Edison Company, Chicago.  An additional study of the
nature and cause of the behavior of the gas in the wake of smokestacks
is reported.  The turbulent region immediatly adjacent to the downstream
surface of the stack was found to cause plume downwash.  If the gases
thus brought down come within the influence of the turbulence flow over
the roof of the building, they were then quickly brought to the ground
behind the building.

     Zero downwash into the wake of the smokestack was observed when the
stack gas exit velocity was greater than twice the wind velocity.
Downwash was approximately one stack diameter below the top of the stack
when the stack gas exit velocity was only twice the wind velocity.  The
model study of Crawford Station demonstrated the need for a stack increase
of 50 feet to prevent downwash from any direction, provided that the gas
velocity is high enough to prevent the first step of downwash.  This
additional increase results in the stack being approximatley 2.5 times
the highest part of the building structure.


Davidson, W. F., 1959:  Studies of Stack Discharge Under Varying Con-
ditions.  Combustion, 23(4), 49-51.

     The problem encountered in designing stacks for the new Astoria
Station in New York City is reviewed.  Design of the stack to have a
height greater than 2.5 times the height of the power station is stated
as a long time recognized "rule of thumb".  However, the author believes
that, despite the importance of this factor, except for stacks of limited
height and the number of investigations made, it is still impossible to
give any rules or criteria that  can be used with reasonable assurance
to predict the stack performance of a new station.  Thus, carefully
planned wind-tunnel tests seem to be required.  In the case of Astoria
Station, increase in stack height was originally limited by nearby
airport runways.  A wind-tunnel model was tested to determine the necessary
exit gas velocity to provide a sufficient plume height to minimize
adverse building effects.  A special stack nozzle was designed to keep
velocity of the exit gas equal to the full load parameter regardless of
the actual load.
                                A-l

-------
Strom, G. H., 1952:  Wind-Tunnel  Techniques  Used  to  Study Influence of
Building Configuration on Stack  Gas  Dispersal.   Industrial  Hygiene
Quarterly, 13, 76-80.

     Wind-tunnel  experimentation  is  presented as  a research tool  that
has yielded answers-difficult,  if not impossible,  to obtain by other
means.  Stack gas dispersal  in  the presence  of  buildings  and other
nearby structures is given as  the most frequently investigated problem
in the wind tunnel.  Wind-tunnel  modeling  is suggested when use of
empirical rules for stack height such as requiring a stack  to be 2.5
times the building may lead to  unnecessarily high and costly structures.
Discussion of wind-tunnel modeling methods and  criteria then follow.


Beaver, S. H. (Chairman), 1954:   Report of Government Committee on Air
Pollution.  Cdm.  9322.  Her Majesty's Stationery  Office.

     A committee  was appointed  in July, 1953, with the following terms
of reference:

          "To examine the nature, courses  and effects of air
          pollution and the efficacy of present preventive
          measures; to consider what further preventive measures
          are practicable; and  to make recommendations."

     Discussion of desirable stack height  is taken from Appendix VI.

                            APPENDIX VI

        The Influence of Chimney Design and  Height on the
        Drsperston of Flue Gases  prom Industrial  Chimneys

            Memorandum by the  Industrial Sub-Committee

INTRODUCTION

     The original function of  high chimneys  was to create draught for
the furnaces.  With the introduction of mechanically created draughts
early in the century, many factories were  equipped with only short
chimneys and as a consequence  smoke dispersal was not good.  More
recently, however, there has been a trend  towards use of high chimneys
in order to improve dispersion  by discharge  into  the higher levels of
the air.

     We have found that the information on a chimney design and height
and the effect of chimney height on probable conditions on  the ground to
the lee of the chimney is widely scattered and  in general inaccessible
to industrial engineers.  We have therefore  felt it  necessary to go into
the subject in some detail in  this appendix. The following is a summary
of the best informed opinion at present, but further investigation may
cause these opinions to be revised.

                                 A-2

-------
1.    Down-draught

     When a wind blows across a building or a hill a down-draught is
created on the lee side.  (.1) It is important that chimneys should
discharge their smoke high enough for it to escape these down-draughts
if possible.

     A rule used successfully for about 20 years by the Electricity
Industry is that the height of a chimney shall be at least 2-1/2 times
the height of the highest adjacent building.  When the chimney is sited
in hilly country or among buildings which make it impracticable to apply
the "2-1/2 times" rule, wind tunnel tests on models may be necessary to
determine where to site the chimney and how high to make i.t to avoid
down-draughts.  Pending further research on the subject, a good working
rule for low buildings is to make the chimney not less'than 120 feet
high -- though discretion  must of course be exercised for small install-
ations.

2.    Down-wash

     Down-wash is the drawing downward of chimney smoke by the system of
stationary vortices or eddies that form in the lee of a chimney when a
wind is blowing.  If the velocity of emission of the smoke is not great
enough to overcome down-wash some of the smoke will be drawn by these
eddies down into the down-draughts of the buildings beneath.

     The down-draught will then carry the smoke to the ground.  Ex-
periments have shown that down-wash will not occur if the velocity of
emission is sufficiently high.  It is clear to us that further research
on the design of chimney mouths is required.

     Reference (2)_ gives a graph showing for a given wind speed the
minimum velocity of emission for avoiding down-wash.

3.    Chimney height and dispersal of smoke and gases

     At whatever height smoke is discharged, gravity will eventually
bring the larger particles of dust and soot to the ground.  Moreover,
because of the natural turbulence and mixing of the atmosphere, a propor-
tion of the finer particles and gases in the smoke will reach the ground,
although their motion is unaffected by gravity.  The higher the point of
discharge the greater will be the dilution of the gases and dust by the
time they reach the ground.
                               A-3

-------
Corby, G. A-> 1954:   Airflow Over  Mountains:  A  Review of the State of
Current Literature.  Quart.  J.  Roy.  Met.  Soc., 80_,  49.1.

     The work of J.  Forchtgott,  who gathered about .35  different sets of
observations involving five different  mountain ridges  located in Bohemia
is reviewed.  Mountain airflow is  classified into  four main types:
01 undisturbed streaming,  (.2). standing  eddy streaming, (3)_ wave streaming,
and ('4 ]_ rotor streaming.   The case of  standing eddy  streaming corresponded
to the situation of boundary layer separation at the ridge apex with
cavity formation in the lee.  This type  of  flow  is reported to have been
observed frequently.  Forchtgott implied that this situation was predominant
under moderate wind speed and wind shear conditions.  Even for the  cases
with smooth waves above,  some form of  turbulent  wake was found in the lee
of the ridge.  No discussion of the extent  of the  region of modified
airflow is presented.
                              A-4

-------
Hawkins, J. E. and G. Nonhebel, 1955:  Chimneys and the Dispersal of
Smoke.  J. of  the Institute of Fuel, 28_, 530-545.

     To avoid  parts of a smoke plume being blown rapidly to the ground
by local disturbances of the wind, the authors report that it is necessary
to choose minimum heights of chimney and exit velocities of flue gases
which 'are related to the height of surrounding buildings, diameter at
chimney and local ground contour.  Disturbances of the atmospere set up
by the wind flowing past the chimney and over buildings can, under
certain circumstances, draw the smoke rapidly to the ground so that the
efficiency of  the chimney as a smoke disperser is much impaired.  The
.region of so-called "down-draughts" is stated to stretch from the top of
the windward face of the building, rise to about twice the building
height and stretch for about six times the height downwind of the building.
These dimensions are stated to be approximate and to i'ncrease with
cross-wind width of the building.  Also similar effects occur in the lee
of. hills.

     It is reported that a committee appointed by the Electricity Commissioners
f_Great Britain] proposed the rule that, to discharge flue gas clear of
down-draughts, chimneys should be 2.5 times the heights of the highest'
adjacent building.  "This rule has been used successfully by the electricity
generating industry during the last 20 years, although there is some
evidence that  at high wind speeds cool gas plumes can be brought down by
down-draught even though the chimney height satisfies the 2.5 times
rule."  The usefulness of the wind-tunnel tests as an indication of how
high the chimney height should be to avoid down-draught, in difficult
cases is stated.  For large plants in complicated locations, advice is
given to obtain confirmatory data by observation of the spread of smoke
from smoke generations and observations of the trajectories of "zero-
buoyancy" balloons.  It is noted that when a chimney is discharging into
a region of down-draughts and  turbulence behind a building, changes in
the velocity of emission or temperature of the flue gas as it emerges
from the chimney will make little or no difference to conditons on the
ground.  The work of Sherlock and Stalker (.1950) is referenced in determining
the necessary  exit velocity to avoid the drawing-down of the smoke plume
by the chimney wake.  Also stated is the likelihood that a more intense
wake-region will occur for a square-shaped chimney in comparison to the
circular chimney.
                              A-5

-------
Scorer, R. 5., 1955:  Theory of Airflow Over Mountains:   IV-Separation
of Flow from the Mountain Surfaces.   Quart.  J.  Roy.  Met.  Soc., 81,  340-
350.

     According to the author, the flow separation  point  is stationary
when there is a salient edge at the  top of a hill  or ridge.   Numerous,
but limited, field studies relating  to the zone of recirculation and
instances of intense mixing and general down-draughting  in the leeward
regions of ridges are cited.  Details are insufficient to draw firm
conclusions relating to formation of separated  flows.   No specification
of the size of the region modified is given. Three types of flow separation
in mountainous areas are discussed.   These are  (1) air-mass  (i.e.,
valley flow independent of the flow  aloft),  (2) two-dimensional aerodynamic
type (.I.e., flow over a ridge), and  (3) three-dimensional aerodynamic
type (.i.e., flow around an isolated  peak).  In  general,  the  influences
of a three-dimensional hill are reported to be  less than that of a  two-
dimensional ridge.  Also, katabatic  winds tend  to  reduce the likelihood
and size of the region of separated  flow, whereas  anabatic winds should
enhance the size of any region.


Evans, B.. H., 1957:  Natural Air Flow Around Buildings.   Research Report
No. 59, Texas Engineering Experiment Station, Texas A&M College System.

     The shape and size of the downwind eddy caused by the model building
was determined in a wind tunnel study for nearly two-hundred variations
of the basic building shape.  The downwind eddy was defined  as the area
between the building and the point downwind of  the building  where some
particles of the air close to the ground are found to flow upwind toward
the building.  Smoke patterns were used to determine the observed dimensions
of the eddy.  The shape of the building, the roof type,  the  position of
openings, and the orientation with repsect to the wind,  were all found
to have an effect on the air flow over the building.  Several significant
findings are reported.  It was found that, regardless of the height of
the building, the pattern of the air going over the top of a tall building
appeared the same.  For pitched roofs, the depth of the downwind eddy
increased due to the increase in the height of  the building.  When the
building was extended in the downwind direction, the depth of this
downwind eddy decreased.  When the width of the building (perpendicular
to the wind direction), was increased from one times its height to eight
times  its height, the downwind depth of the eddy increased from 2 to 5.25
times  its height.  As the width of the building was further increased to
28 times its height, the downwind depth of the  eddy increased at a
somewhat smaller rate to 8.75 times  its height.
                              A-6

-------
Scorer, R. S., 1959:  The Behavior of Chimney Plumes.  Int. J. of Air
Pollution, 1_, 198-220.

     The 2.5 times rule concerning chimney heights is presented as being
a well-known commendable rule because it is comprehensible as a practical
working rule:  it has no precise theoretical  justification, and if
experience proved it to be inadequate it could be changed by Act of
Parliment!  It is also argued that architects should accept the chimney
heights necessary for the proper dispersal of pollution as a requirement
and desvgn buildings with the chimney as an integral part instead of as
an undesirable appendage.  Also in the lee of a cliff there may be
eddies into which, if a chimney is sited in the downdraught of the eddy,
the plume may be carried down to the ground bodily.  This is more serious
than being diffused down by ambient turbulence.  A case at Hope Cement
Works near Sheffield is di;scussed.  A problem of downdraught was solved
by installing a 150. meter chimney which reaches above the eddies downwind
of the nearby hill.
                               A-7

-------
Nonhebel,  G., 1960:   Recommendations  on  Heights  for  New Industrial
Chimneys.   J. Institute of Fuel,  33.,  479-495.

     A review of the present state  of knowledge  and  experience,  and
recommendations are  put forward as  the basis  of  discussion between
industrialists and those responsible  for the  administration of the  Clean
Air Act of 1956.  This technical  review  was felt necessary since no
detailed technical advice had so  far  been issued by  any governmental
department to assist those frequently faced with difficulty in deciding
the height of chimney required under  the provisions  of this Act.

     Appendix VI of  the Beaver Report (1954)  is  referenced as providing
guidance on technical considerations  governing the  height of chimneys.
Where a chimney rises from or is  adjacent to  a high, large building, the
recommended height is stated to be  at least 2.5  times the height of the
building.   For small plants (.reference to very low  buildings appears to
be intended) the Beaver Report (.1954) makes the  recommendation that
chimney heights be not less than  120  feet high.   The author goes on to
point out that, where there is a  choice  in the orientation of a long •
building to which is attached a chimney, the  longitudinal axis should be
at right angles to the prevailing wind.   It is suggested that when  a
chimney of a large plant is to be built  among a  group of high buildings
which makes it costly to apply the  "2.5  times rule," the only satisfactory
solution is to make  tests with models in a wind  tunnel to determine its
minimum height and its position with  respect  to  the  buildings.  For
small installations  where the chimney plume is not  expected to be
seriously affected by downdraughts  exerted by a  neighboring building, a
sliding scale of minimum stack height from 50 feet  to 120 feet for
plants with steam output up to 33,000 Ib/hr is given.  This minimum
height is suggested  to insure adequate dispersal of  flue gases and  is
based on specific estimates of maximum desirable ground-level concentrations
A stack gas discharge velocity of 1.5 times the  wind velocity is referenced
as sufficient to keep the center!ine  of  the plume from being drawn  below
the chimney top.  The impracticalility of achieving  the necessary discharge
velocity in relatation to very high wind velocities  is noted as not too
important  since dispersion of the  gases is  increased under such conditions.
It is suggested that consideration  be given  to  increasing the velocity
of the exit gas by addition of aerodynamically designed nozzles to  the
chimney top.
                                   A-8

-------
Sutton, 0. G., 1960:  Discussion before the Institute, in London, 23.d,
May 1960.  J. Institute of Fuel, 3_3, 495 (comment).

     It is pointed out that the 2.5 times rule be strictly applied only
to a building which is very long across wind, and only near the central
point.  Sutton believes the origin of the rule was deduced by Sir David
Brunt from W. R. Morgan's study of the height of disturbances over a
long ridge, in an investigation into the disaster of the airship R. 101;
if a wind were blowing perpendicular to the longside of a building, the
disturbances should extend upwards to about 2.5 times the height of the
roof.  Anotier significant point raised by Sutton was that, since it is
impossible to take every factor into account in the mathematics of
atmospheric turbulence, the only thing to do is look at a situation with
the aid of scaled-down models.


Scorer, R. S. and C. F. Barrett, 1962:  Gaseous Pollution from Chimneys.
Int. J. of Air and Water Pollution, 6_, 49-63.

     the wake region of the building is given by a vertical circular
cylinder centered on the building of height 2.5 times the height of the
building and of horizontal radius equal to 3.5 times the width of the
building.  For a building whose width is less than its heights, the wake
region is of height 2.5 times the maximum width.
Skinner, A. L., 1962:  Model Tests on Flow from a Building Ventilation
Stack.  Atomic Energy Establishment, Winfrith, Report AEEW-W 227.

     Wind tunnel tests were conducted on a model of a building to assess
the minimum requirements for a stack which would effectively disperse
the ventilation air clear of the building wind eddies and also avoid
recirculation into the inlet grille,  A stack 2.25 times the average
roof height was found to be just sufficient.


Davidson, B., 1963:  Some Turbulence and Wind Variability Observations
in the Lee of Mountain .Ridges.  J. Appl. Meteor., 2C4), 463-472.

     The results of a number of balloon releases made in two valleys in
Vermont are reported.  Balloon releases were made at several positions
along the sides of ridges that had approximately 20 degree slopes.
Balloon paths were determined using theodolites.  The limited results
could not be used to confirm a point of separation of the extent of a
leeward cavity region.  The extreme turbulence generated in the lee of
the ridges, however, appeared to be dissipated at most elevations at a
distance of 4 to 6 ridge heights downind.
                                 A-9

-------
Thomas, F.  W.,  S. B. Carpenter,  and  F.  E.  Gartrell,  1963:   Stacks—How
High?  JAPCA, 13(5), 198-204.

     TVA experience has demonstrated that  when stacks are  less than
twice the height of the main powerhouse structure,  the plume may, during
high velocity wind, be caught in the turbulent vortex sheath and brought
to the ground level in relatively high  concentrations very near the
plant and sometimes re-enter the building  air supply.  Also, extensive
wind-tunnel tests are stated to  have demonstrated that downwash does not
pose_ a problem where the stack height is at least 2.5 times the height
of the powerhouse or other nearby structures and  appropriate efflux
velocities are provided.


Buettner, K. J.  K., 1964:  Orographic Deformation of Wind  Flow.  Uni-
versi-ty of Washington, Seattle,-Washington.-  Prepared for  U.S. Army
Electronics Research and Development Laboratory,  Fort Monmouth, New
Jersey, under Project No. 1AO-11001-B-021-01, Contract No. DA 36-039-SC-
89118, 70 p.

     The general features of flow over  a ridge are treated theoretically
and experimentally.  A ridge station was constructed on the lee side of
the Ipsut Pass area of Mount Rainier National Park in Washington as part
of a study of the effect of terrain  obstacles on  the fallout of particulate
matter through the atmosphere.  Tracer  particles  of zinc sulfide were
released and collected.  Data were collected for  5 days during which the
airflow approach was perpendicular to the  ridge.   During the period of
experimental set-up, only light  to moderate winds were observed.  The
most common wind field occurrence is reported as  a "Vortex sheet flow"
with the airstream separating from the  ridge top  and forming a wake zone
in the lee of the ridge.  For this flow, the.wind field was constant
above and zero below a plane representing  the wake zone.  Only a small
amount of particulate penetrated down through the horizontal vortex
sheet.  A contaminant released in the calm zone is reported to meander
in an unpredictable manner.  Previously a  lee eddy with the main airstream
moving first horizontally away from the ridge, then down,  and then up
again close to the valley bottom was visually observed.  At this site,
such a flow pattern was believed to  exist  only for strong  winds.  Laminar
flow complicated by thermal winds is reported to  occur when stable
settled conditions prevail and the gradient wind  at ridge  level was less
than 6 knots C3.1 meters per second).
                                A-10

-------
Eimern, J., R. Karschon, L. A. Razumova, and G. W. Robertson, 1964:
Windbreaks and Shelter-breaks.   Word Meteorological Organization Technical
Note No. 59..

     Part of this report summarizes the literature on the influence of
shelterbelts on air flow.  The region leeward of shelterbelts is reported
to have reduced winds, and a degree of turbulence and eddying of the
flow in the lee.  According to one reference, the air flow is affected
up to even three or four times the height of the belt.

     Most of the literature is concerned only with defining the downwind
extent of the region of reduced winds.  The literature offers a wind
range of distances.  It is reported that, according to West European,
North American, and Russian experiences, the rule of thumb applies that
the shelter zone extends to 30 times the obstruction height.  However,
for. a_wind_ reduction of 20 percent and more, jthe effect is noted to only
20 times the height.  Moreover, the extent is very dependent on the
permeability, shape and width of the belt, roughness of the ground
surface, thermal stratification of the air.  No discussion defining the
vertical extent that shelterbelts can effect stack effluents is given.


Lord, G. R., W. D. Baines, and H. J. Leutheusser, 1964:  On the Minimum
Height of Roof-Mounted Chimneys, Results of an Exploratory Wind-Tunnel
Study.  Report TP-6409, Technical Publication Series, Dept. of Mechanical
Engineering, University of Toronto.

     Wind-tunnel tests of smoke emission from roof-mounted chimneys on
both block-type and pyramidal  structures are described.  The tests were
performed in a constant velocity low turbulence wind field.  The wind
velocity v/as equal to the stack emission speed.  Four conditions defining
a minimum stack height are given, each corresponding to a different
degree of plume distortion by the structures.  For a given stack location,
building configuration, and wind direction, the height of the stack
necessary to meet each of the four conditions is reported.

     A discussion of building wake effects is included.  The point is
made that, even if the source is above the wake, the effluent may later
enter the region of influence.  At several building heights downstream,
the turbulent region is stated to be about twice the building cross-
section.  For the tests, the stack was placed over the center of the
building.  The vertical extent of building influences was found to scale
with the building width for tests where the building height is greater
than its width.  The height above the building of the stack at which
smoke began to be entrained to the stagnant wake of the building was 0.5
times the building width.  For the tests when the building width was
greater than the building height, the vertical extent of the building
influences were similar to above definitions, however, with the height
scale replacing the width scale.
                                 A-ll

-------
Moses, H., G. H. Strom,  and J.  E.  Carson,  1964:   Effects  of Meteorological
Engineering Factors on Stack Plume Rise.   Nuclear Safety,  6_, 1-19.

     This paper contains a review  and  disucssion  of  several  reports
concerning desirable stack height  near buildings  and terrain.   Movies of
smoke flow patterns over buildings with small  stacks at Argonne National
Laboratory were said to illustrate "cart wheels"  forming  on the lee side
with a diameter several  times the  height of the building  and thus providing
high concentrations of contaminant.  The wind-tunnel  studies of air flow
around buildings by Evans (.1957)   HaTrtsky C1952). and Strom (1962), are
discussed.  The likely origins  of  the  2.5  times rule'of thumb, which has
been used by the British Electricity  Industry  since  the 1930's is presented
in light of comments of Sutton  0960}.   It is  reported that the Dutch
require that a stack must only   be 1.5 times  the  height of the highest
building in the neighborhood.  It  is concluded that  no elementary rule,
such'as a" 1/5 or 2.5' times ruler can""be "applied to all  situations.   The
air flow in mountainous areas is stated to be  quite  complicated with
terrain irregularities located   many stack heights upwind and downwind
influencing plume motions.  It  is  suggestsed  that, whenever a potential
pollution problem results from  an  effluent emitted by a stack located in
all but perfectly uniform terrain, wind-tunnel studies should be considered
Gloyne, R. W., 1965:  Some Characteristics  of the  Natural  Wind and Their
Modification by Natural and Artificial  Obstructions.   Scientific Horticul-
ture, XVII, 7-19.

     Some characteristics of wind  field modification  by natural  obstructions
are reported.  An eddy flow 2 barrier heights in  vertical  extent and 10
to 15 barrier heights in horizontal  extent  to the  leeward  side of a
"near solid" barrier was diagrammed.   At ground  level,  the region of
disturbed flow extended to about 30 barrier  heights.   Downwind of a
steeply sloped, wooded hill with a wind blowing  at right angles  to its
length, the disturbed flow is reported to also extend downwind to about
30 times its height.  Additional discussions  relevant to wind modifications
were also presented, and the point is made  that  each  case  must be assessed
separately.  Slope angle and thermal  stability and wind speed were
influential factors in determining the extent of terrain induced dis-
turbances.


Jensen, M., and N. Frank 1965:   Model-Scale Tests  in  Turbulent Wind.
Danish Technical Press, Copenhagen.

     A large number of systematic  wind-tunnel studies of concentration
downwind from an isolated chimney  and a chimney  on a  house are reported.
An evaluation of the data indicates some building  influence even for a
stack height three times the house height.
                                    A-12

-------
Halitsky, J., G. A. Magony, and P. Hal pern,
Topographical Effects.  New York University
Laboratory Report No.  "iR-66-5, 75 p.
1966;   Turbulence Due to
 New York, Geophysical
     Comparisons between the author's wind tunnel model results and
Davidson's (1963) field observations in the lee of Green Peak, Vermont
are reported.  Best Agreement resulted for the higher model wind speeds
suggesting that tests of this type be run with a minimum ridge height
Reynolds number of 1  x 105.  1^« field observations of a cavity and wake
flow generally fitted the model te^i results.   The boundary layer and
upstream turbulence conditions were not -imulated in the wind tunnel
tests.
Ukeguchi, N., H. Sakata, H. Okamoto, and Y. Ide, 1967:
Gas Diffusion.  Mitsubishi Technical .Bulletin~No. 52.
             '^idy on Stack
     i he authors reported that downdraughts occur where the structures
and/or buildings stand near the-stack, but these can be prevented on the
whole with the increase of stack height to 2.5 times greater than the
structures and/or buildings surrounding the stack.  They stressed that
downdraughts produce very high ground level concentrations, depend on
the layout of structures and/or buildings, and must be avoided.
tunnel  study examined the influence of a nearby building complex
plume diffusion and found only a small effect when the stack v/as
times the building height and a negligible effect when the stack
over 3 times the building height.  No general rules are given as
applicable to the effects of topography; thus wind-tunnel models
used to assess air quality impact.
                     A wind-
                     on
                     2.5
                     was
                     being
                     are
World Meteorological Organization, 1967:  The Airflow over Mountains.
WHO, Geneva, Switzerland.  Report No. 98, 43 p.

   •  The World Meteorological Organization technical note concludes
that, over rugged terrain, whether the flow aloft is smooth or other-
wise, it usually rests on a turbulent wake.  Although little descriptive
detail of such regions is presented in the report, many photographs
showed the wave structures above the wakes, as revealed by cloud formations
                                  A-13

-------
Berlyand, M.  E.,  1968:   Meteorological  Factors  in  the  Dispersion of Air
Pollutants in Town Conditions.   Symposium of  Urban Climates and Building
Climatology,  Brussels.

     The author mentions that the  character of  air motion changes considerably
near hilly relief and can substantially influence  pollutant dispersion.
The increase  of concentration was  reported to sometimes  occur even if
the pollutant sources are located  on elevated places,  but near leeward
slopes where  wind velocity decreases sharply  and downward currents
arise.  -He -states- that at present  numerical solution  of  the equations of
motions and wind tunnel  experiments  are carried out for  each case.
Experiments on models of separate  plants and  buildings have permitted
determination of zones i.n which downward currents  and  pollutant stagnations
are possible.  The "2.5 times rule"  is  referenced  as  the recommended
stack height  in order to avoid considerable increases  of concentration.
Halitsky, T., 1968:  Gas
Atomic Energy - 1968, D.
          Diffusion  Near  Buildinas.   Meteorology and
          H.  Slade  (Ed.),  Chapter  5-5.
     A detailed discussion of flow separation and  wake formation near
buildings is presented.   The introduction of a building into a backgound
flow- is stated to cause  changes  in the velocity and pressure fields.
The new fields are called aerodynamically distorted,  with the amount of
                                              the  distorted and
                                              literature review
                                                 the
                                                 of -
'low1
distortion measured by the difference  between
background properties. .The author  presents  a
near characteristic structures.   It appears  that  the  flow downwind of
sharp-edged buildings  is  disrupted  to  a  greated extent  than .for rounded
buildings.  No definition of the  vertical  or horizontal  extent of the
building wake which could be used to determine the  height of  a stack
sufficient to avoid adverse influence  is  presented.
Scorer, R. S.,
pp 107-108.
1968:   Air Pollution.   Pergamon  Press,  Oxford,  England,
     The author discusses the consequences of a  separated flow in the
wake of obstacles.  Several examples of adverse  influences on chimney
effluents in the wake of buildings and steep hills  are presented.  The
examples are quite descriptive of the problem; however, no specific
definitions to the size and extent of'wake effects  are given.  It is
suggested that chimney tops be cleanly shaped without elaborate decoration
or increase in exterior size and that the efflux velocity should be
enough only to prevent downwash into the v/ake and cold inflow.  It is
noted that devices have been employed to prevent chimney downwash.  The
author also states that, if chimneys need to be  short, there are many
devices which can be employed to prevent separation at salient edges.
One such device is shown.
                                A-14

-------
Strom, G. H., 1968:  Atmospheric Dispersion of Stack Effluents.  In:
Air Pollution. Vol. I, Stern, A. C. (Ed.), Academic Press, New York,
uiapte.r «t

     A brief discussion of the effects on plume dispersion induced by
terrain and b'ui.ldi.ngs is presented.  The results of several wind tunnel
experiments, are presented.  Th.e need for experimental procedures is
stated since there are no accurate analytical procedures.  The adverse
effects were seen to be greater when the wind was normal to the long
dimension of the Building.  The desirability of designing stacks high
enough to have the plume remain clear of the highly turbulent regions is
stated.  No specific definitions of the extent of the highly turbulent
regions is presented.  Evans (.1957) is referenced as providing guidance
when experimental data is not available for specific cases.


Forsdyke, A- G., 1970:  Meteorological Factors in Air Pollution.
Technical Note No. 114, World Meteorological Organization, Geneva,
Switzerland.

     The following sentence is the only mention of stack height in
relation to the effect of building eddies which, if the chimney is not
h.igh. enough, will bring high concentrations of the pollutant down to
ground level in puffs.  "To overcome this effect it is required in some
countries th.at the chimney height shall be at least two and one half the
freight of the building from which it rises."


Pooler, F., Jr., and L. E. Niemeyer, 1970:  Dispersion from Tall Stacks:
An Evaluation.  Presented at the Second International Clean Air Congress,
^asAington, D. C.  December 6-11, 1970, Paper No. ME-14D.  31 p.

     The authors present, as part of a study evaluating dispersion from
tall stacks, several situations in which unexpectedly high ground level
concentrations could be associated with mountain lee effects.  On days
with, neutral flow, the plume from a stack located 13 ridge heights
downwind from a 450. m ridge was carried down to ground level within a
very short distance.  This phenomenon could well be a result of the
strong downwash that occurs near the leeward edge of a standing eddy.
                               A-15

-------
World Meteorological  Organization,  1970:   Urban  Climates  and Building
Climatology.  Proceeding of the  Symposium on  Urban  Climates and Building
Climatology, Jointly organized by  the  World Health  Organization and WHO,
Brussels, October 1968,  WMO Technical  Note No.  108,  109.

     Concern for potential  adverse  building effects  upon  plume dispersion
was mentioned in several of the  symposium presentations.   Only one of
the authors alluded to the  "2.5  times  rule" as  referenced by Hawkins and
Nonhebel (1955).  One of the general conclusions as  reported by T. J.
Chandler was that -"there 'is an-urgent  need to define much more vigorously
the physics of the urban surface—particularly its  thermal and aerodynamic
properties."  He also concluded  that wind measurements within the cubic
of the city are clearly dependent  upon very  local  conditions which
"makes it very difficult to use  such field observations. to construct any
general theory although simple models  of  airflow around single structures
may s'tiu "prove "of "practical use"."   Wind tunnel "and  similar laboratory
techniques have a very real contribution  to make in  these enquiries."


Meroney, R. N. and B. T. Yang,  1971:   Wind-Tunnel  Study on Gaseous
Mixing Due to Various Stack Heights and Injection Rates Above an Isolated
Structure.   USAEC REport No. COO-2053-6.

     This wind-tunnel study examines the  influence  of a simple cubical
structure on the dispersion of a tracer gas  released from short stacks
at varying heights ,and exhaust velocities.  Both smoke visualization and
quantitative concentration  measurements were  made.   The conclusions of
this study include;

     C.1 ).  For a stack less  than  1.5 times the building height, high
exhaust velocities cannot prevent  some immediate downwash.
          As the stack height increases,  the effect of building en-
trainment decreases.  Exhaust velocities, for stack heights greater than
twice the building height, apparently need only be high enough to avoid
downwash behind the stack itself.

     C.3)  Building orientation apparently aggravates entrainment even
for a simple cubical structure, however,  the effect is not a major
consideration here.  (For more complicated building complexes, the
influences may be more significant.)
                               A-16

-------
Orgill, M. M., J. E. Cermak, and L. 0. Grant, 1971:  Laboratory Simu-
lation and Field Estimates of Atmospheric Transport - Dispersion Over,
Mountainous Terrain.  Colorado State University, Fort Collins, Colorado.
Technical Report No. CER70-71MM-JEC-LOG40.

     An extensive literature review relating to both field and fluid
modeling studies and a discussion as to how mountainous .terrain can
alter atmospheric airflow is presented.  The authors report that,'for
neutral airflow over a mountain, a large semipermanent eddy occurs on
the lee side.  An area in the central Rocky Mountains of Colorado was
chosen for a field and laboratory study of transport and dispersion over
irregular terrain.  Two different atmospheric conditions were simulated:
the thermal stability used in the wind tunnel model was near-neutral in
the lower levels and stable in the upper levels for one'case and totally
neutral. throughout.for the other.case.  Field, data yielded information
on the mean velocity and dispersion characteristics over the local
terrain.  Totally neutral atmospheric stability conditions were observed
on only one day.  No sepcific information as to where and when boundary
layer separation occurs or the size or shape of the cavity region in the
lee of ridges is reported in either the field or laboratory study
results.  The purpose of the report is to generalize on flow patterns in
complex terrain on a much larger scale.


Yasuo, I., 1971:  Atmospheric Diffusion Theory of Factory Exhaust Smoke
and Its Applications.  Water Engineering Series, published by the Japan
Society of Civil Engineers, Hydraulics Committee.

     The author presents equations for providing air quality estimates
that are intended for flat land.  When the stack height is less than
2.5 times the height of buildings (or the mountains near the stack), it
is suggested that the exhaust gas will be swept down into the turbulence
area caused by the buildings.  When this phenomena occurs, simulation
methods using wind tunnels and other special techniques are used.
                                  A-17

-------
Lucas, D. H., 1972:  Choosing Chimney Heights  in  the Presence of Buildings.
Proceedings of the Interantional  Clean Air Conference,  Melbourne Australia,
May 15-18, 1972,  27-52.

     A chimney 2.5 times the height of any adjacent building is reported
to follow the widely accepted rule of thumb to avoid effects by building
turbulence.  The fact that the building width  must also be relevant in
deciding the effect of the building is discussed.   The  essential dif-
ference for a tall thin  building  is that flow  around the building reduces
the effect of flow over  the buiiding.  It Is generalized for all buildings
that a building wake has a height above the building of 1.5 times the
height of width of the building,  whichever is  less.  The extent of the
turbulent wake is reported to be  pronounced for a distance downwind of
approximately five building heights or half-widths, whichever is less.
While there is no abrupt cut-off  in fact, it is considered convenient to
take" the "effect as declining progressively to  zero from 5 to 10 building
heights or half-widths,  whichever is less.


Schultz, J. G., 1972:  Self Pollution of Buildings.  The ASME Proceedings
of the 1972 National Incinerator  Conference, New York,  NY.  June 4-7,
1972, 201-210.

     It is suggested that good design for a chimney orexhaust system is
to locate them above the eddy area.  Otherwise, there will be recycling
of exhaust products in to the air intake to contaminate the entire
building.  The vertical  extent of the eddy over a cubical building is
given according to Evans (1957) as 1.5 times the building width.


Shingi, K., 1972:  Wind  Tunnel Experiment on Ascent Height of Exhaust
Gas.  Central Research Institute  of Electric Power Industry Report,
71053 (Translated from Japanese), 26 p.

     The results of wind tunnel exepriments on the ascent height of
exhaust gas from thermal and nuclear pov/er plants are reported, and
studies are made of the  ascent height with relation to down-washing,
down-draught, and the stack type.  The laws of wind tunnel similarity
are also discussed.  It was found that stack down-washing does not occur
if the ratio between the exhaust gas speed and wind speed is more than
two.  For the power plants studied, down-draught in the wake of the
building did not occur even when the stacks were much lower than 2.5
times the building hieght, if the exhuast gas  rate was  large enough.
The author comments that the 2.5 times law does not have a theoretical
basis making it applicable to all cases.
                                 A-18

-------
American Society of Mechanical Engineers, 1973:  Recommended Guide for
the Prediction of Airborne Effluents.  Smith, M. (Ed.), New York, AMSE,
85 p.

     One section of the book discusses the influence of buildings and
irregular terrain.  It is reported that few quantitative diffusion
experiments have been made in irregular terrain; however, visual observations
of plume behavior in a variety of situations have been made.  The plume
from a stack placed in the cavity leeward of a valley ridge is said to
become thoroughly diffused before passing downwind to the wake region
where the flow was in the direction of the upper wind.  The air flow
disturbed locally by buildings is shown to influence that portion of the
plume which penetrates the disturbed flow region.  Changes in building
shape and orientation to the wind are reported to affect the cavity
dimensions and flow to a marked degree, but the gross dimensions of the
displacement"zone'and wake for'sharp-edged"buildings appear to be a
function primarily of the frontal area of the building presented to the
wind.  Also for rounded buildings, both the displacement zone and wake
are smaller than for sharp-edged buildings since separation usually
occurs downwind of the center of the buiilding where the direction of
the surface flow just prior to separation is horizontally downwind  •
rather than normal to the wind.  No quantitative definitions of th
vertical or downwind extent of the region of adverse influences near
buildings or terrain are given.
                               A-19.

-------
Briggs, G. A., 1973:  Diffusion Estimation for Small  Emissions.  Atmos-
pheric Turbulence and Diffusion Laboratory, NOAA, OAK Ridge, TN, (Draft)
ATDL No. 75/15.

     A method for estimating air quality concentrations for emissions
influenced L-y buildings is presented.   The plume is considered to be
within the region ---f building influence only when the'estimated source
height is less, than the build:^ height plus 1.5 times the building
height or width, whichever is less.   The "<~avity" region where there is
circulation of the flow within the _wak,e. of the. fci-
-------
Smith, D. G., 1975:  Influence of Meteorological Factors Upon Effluent
Concentrations On and Near Buildings with Short Stacks.  Presented at
the. 6.8tH Annual Meeting of the Air Pollution Control Association,
Boston, fjass., June 15-2Q, 1975, Paper No. 75-26.2.

     Field data of concentrations from stack emissions  near a scaled-
down model of an industrial building is presented.  The tests were
conducted for selected conditions of atmospheric stabilility, aerodynamic
roughness of upwind fetch", and wind orientation angle of the building.
Tfre exit velocity was greater than twice  the wind  speed for all  tests to
eliminate stack downwas'h as available.  The study  was designed to measure
the. amount of effluent reaching the building and ground surfaces in the
downwind wake cavity of the building under a variety of stack heights.
Concentrations along the lee wall of the  building  were measurable, even
when th.e stack was 2 to 2.5 times the building height.  However, much
nigh.er concentrations were found when that "stack was less  than 1.5 times
the building height.


Efrttter, R. E., J. C. R. Hunt, J. S. 'Puttock, 1976:  Predicting  Pollution
Concentrations Near Buildings and Hills.  Presented at  the Conference on
Systems,  and Models in Air and Water Pollution, at  the Institution of
Measurement, London, Sept. 22-24, 1976.

     Several simple mathematical representations of different parts of
th.e flow field near buildings and hills are presented.  These models are
Based on theoretical arguments applicable to two-dimensional flow.
Reliable calculation methods for' the mean turbulent flow around  obstacles
(.three-dimensional i;s implied), are stated to' not exist.  The effects of
the distorted flow-, in the wake behind two-dimensional  bluff surface
obstacles, in a turbulent boundary layer,  upon emissions of various
freight and downwind locations is evaluated.  A source elevated to only
1.5 times the obstacle height is found to be greatly influenced  unless
it is placed farther than 10 obstacle heights downwind.  The influence
upon a source elevated to 2.5 times the obstacle height is found to be
-much less, however, the effect extends to sources  as far downwind as 20
obstacle heights.  No significant effect  is found  for source heights
that are greater than 3 times the obstacle height.
                             A- 21

-------
Huber, A..H., W. H. Snyder, R..S.  Thompson, and R. E.. Lawson, Jr., 1976:
Stack Placement in the Lee of a Mountain Ridge.  U. S. Environmental
Protection .Agency EPA-6.QO./4-J6-.047, R,e.s,ea,rcl TrVa.ngle Park:, NC.

     A wind tunnel study was conducted to examine the effects the
highly turbulent region in the lee of a two-dimensional mountain ridge.
Smoke visualization and hot film anemometry measurements showed that  the
cavity size and shape were minimally affected by the'thickness and
turbulence intensity of the approach, boundary layer flow.  The size  of
the region-.of-.strong circulation in. the lee of the model ridge was found
to be strongly- dependent upon the upwind terrain and the gross topographic
features .(angles!, of the downs lope.  The largest cavity was found to
e:xtend to two ridge'Heights in the vertical and to ten ridge heights
downwind.  A stack 2.5 times the height of the ridge is stated to avoid
the h.ighly turbulent region of the cavity proper.  It is implied that a
taller stack; may- be necessary to avoid all'wake effects since part of
the plume can, in only a short distance, spread downward into the wakes.
Tfte need for studies of the behavior of plumes from sources placed
downwind of the cavity region is stated since the turbulence intensity
downwind of the cavity was found to be still significantly greater than
in the undisturbed flow.


Huber, A. H. and W. H. Snyder, 1976:  Building Wake Effects on Short
Stack Effluents.  American Meteorological Society, Third Symposium on
Atmospheric Turbulence Diffusion and Air Quality, Raleigh, NC  Oct. 19-
22, 1976, 235-241.

     A wind tunnel study was conducted to examine building wake effects
on effluents from stacks near a building whose width is twice its height.
Some discussion of th.e building influences on the plume dispersion is
presented,  for those sources having an effective stack height less than
2.0. building  heights, very significant effects upon measured ground
level concentrations were found.  Visual observations of smoke were also
-made in  order to assess the Building influence upon stack emissions.
There was significant reduction in building effect for the most elevated
stack wh.ich was 2.5 times the building.
                             A-2 2

-------
Snyder, W. H. and R. E. Lawson, Jr., 1976:  Determination of a Necessary
Height for a Stack Close to a Building—a Wind Tunnel Study.  Atmospheric
Environment, 10, 683-691.

     Wind tunnel tests shov/s a stack 2.5 times'the building height is
adequate for a building whose width perpendicular to the wind direction
is greater than its.height, but unnecessary for a tall, thin building.
Smoke was used for flow visualization and quantitative concentration
measurements of a tracer gas emitted with the stack effluent were made
downwind of the building.   For a tall, thin building, application of an
alternative to the 2.5 times rule (Briggs, 1973) was shown to be ade-
quate.  Thus, it is concluded that a sufficient stack height in order to
not have the plume entrained into the wake of the building is equal  to
the building height plus 1.5 times the building height or width, whichever
is less.
Frost, W. and A. M. Shahabi, 1977:  A Field Study of Wind Over a Simulated
Block Building.  NASA CR-2804 prepared by the Univ.  of Tenn.  Space
Inst., Tullahoma, Tenn., March.

     A field study of the wind over a building 2.4 m (.deep)  x 3.2 m
(.high} x 26.8 m  (long) is reported.  The study was designed  to provide a
fundamental understanding of mean wind and turbulence structure of the
wind field.  Eight instrumented towers were placed in the region both
upwind and downwind of the building.  Horizontal  and vertical wind
sensors were placed at the 3, 6, 12, and 20 meter levels.  Approximately
100 experimental runs have been conducted.  Hand held smoke  candles and
anemomenters were used to define the extent of the region of recirculating
flow downwind from the building with its long side oriented  perpendicular
to the flow.  The downwind extent was about 12 ± 2 building  heights.
This is compared to values of 13-16 building heights reported for similar
two-dimensional  laboratory tests.  The smoke patterns indicate that the
wake extends to  a height of approximately 1.5-2 building heights.  The
values of the velocity components at the 3 m level were strongly influenced
by the building, but at the 12 m (^3 building heights) level the influence
was not apparent.
                              A-23

-------
International Atomic Energy Agency, 1977:  Guideline for Atmospheric
Dispersion Estimates.  Vienna, Austria.

     It is reported that the motion of effluents near bluff bodies, such
as buildings, is affected by'distortion of the windfield.  Stacks at
least twice the height of the tallest adjacent building are usually
necessary- except when the discharges are insignificant.  Because of the
great -variety- of possible terrain conditions, a generalized treatment of
the. effects of features such as hills or valleys is stated as not feasible,
since the exact flows will be extremely site-dependent.  The use of
"flui.d "flow-modeling 'is suggested as providing some help in estimating
the plume trajectory near hilly terrain.


Robins, A. G. and  I. P  Castro, 1977:  A Wind Tunnel Investigation of
Plume Dispersion in the Vicinity of a Surface Mounted Cube-I.  The Flow
field,  H-.  Th.e Concentration Field.  Atmospheric Environment,
17, 29.1-311.

     Experiments investigating both, the flow field and plume behavior
downstream of an isolated surface mounted cube in the Marchwood Engi-
neering Laboratory wind tunnel are reported.  The wake air flow was
found to be strongly affected by upstream turbulence.  For both a 0° and
45P orientation of the building into the wind, the effective wake zone
in a turbulent boundary layer extended upwind to about five times the
height  of the. cube.  The region of reversed flow extended downwind to
1.5 freights for wind angle, 9, of 0°, and 2 heights for e of 45°.  The
mean velocity- deficit was reported to extend to twice the building
height  for both the Q° and 45° orientation.  A tracer gas was emitted
from a  stack, over  the roof center.  The stack extended from building
height  to 2.5 times the building height.  The influence of the building
was found to be detectable for 6=0 degrees and a low stack emission
rate; however, for a ratio of emission velocity to wind speed of 3:1,
the influence was  negligible for a stack height 2.5 times the building
freight.  For 8 = 45° the influence of the cube was detectable for all
the stack heights  and emission velocity ratios.  It is concluded that
much, work remains  to be done on the influence of nearby buildings on the
behavior of chimney plumes.  Also, it is especially important to model
correctly- the approach flow when undertaking wind tunnel  investigations
of diffusion in the vicinity of isolated buildings.
                               A-24

-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-450/4-80-023
2.
4. TiTLE AND SUBTITLE
Guideline for Determination of Good Engineering
Practice Stack Height (Technical Support Document
for the Stack Height Regulations)
7. AUTHOR(S)

9. PERFORMING ORGANIZATION NAME.AND ADDRESS
U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
12. SPONSORING AGENCY NAME AND ADDRESS


3. RECIPIENT'S ACCESSION NO.
5. REPORT DATE
July 1981
6. PERFORMING ORGANIZATION CODE
8. PERFORMING ORGANIZATION REPORT NO.
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
This report provides background information used to develop a means of
computing good engineering practice (GEP) stack height per the requirements
of Section 123 of the Clean Air Act, as amended. The report also summarizes
the application of the structure-based formula to determine GEP stack
height under different general building formations.
17.
a. DESCRIPTORS
Air Pollution
1 Good Engineering Practice
i
IS. D!STR'SUT'O\ STATEMENT
j Unlimited
i
t
i
i
KEY WORDS AND DOCUMENT ANALYSIS
b. IDENTIFIERS/OPEN ENDED TERMS
Air Pollution Control
Stack Height
19. SECURITY CLASS (This Report j
Unclassified
20. SECURITY CLASS /This page)
Unclassified
c. COSATl F-ield/Group
13-B
21. NO. OF PAGES
22. PRICE
SPA Form 222C-; (Rev. 4-77)    PREVIOUS  EDITION is OBSOLETE

-------