& EPA
        United States
        Environmental Protection
        Agency
Office of Air Quality
Planning and Standards
Research Triangle Park. NC 27711
EPA-454/R-92-023
(Revises EPA-450/4-88-015)
October 1992
        Air
            WORKBOOK FOR
        PLUME VISUAL IMPACT
      SCREENING AND ANALYSIS
                (REVISED)

-------
                                      EPA-454/R-92-023
o
                 WORKBOOK FOR
             PLUME VISUAL IMPACT
           SCREENING AND ANALYSIS
                    (REVISED)
                 Office Of Air Quality Planning And Standards
                     Office Of Air And Radiation
                  U. S. Environmental Protection Agency
                   Research Triangle Park, NC 27711

                       October 1992

-------
This report has been reviewed by the Office Of Air Quality Planning And Standards, U. S. Environmental
Protection Agency, and has been approved for publication.  Any mention of trade names or commercial
products is not intended to constitute endorsement or recommendation for use.
                                      EPA-454/R-92-023

                                  (Revises EPA-450/4-88-015)
                                             u

-------
                             PREFACE

     This document was first issued in September 1988 as a draft
for public comment.  On February 13, 1991 (56 FR 5900), EPA
issued a Notice of Proposed Rulemaking to augment the Guideline
on Air Quality Models (Revised) with modeling techniques includ-
ing those referred to here.  This document is revised to reflect
these comments and is included in Supplement B to the Guideline.
                                 i 1 i

-------
                                 CONTENTS


Acknowledgments	    i i i
Nomenclature	    vii
1  INTRODUCTION	      1
2  GENERAL CONCEPTS	      5
      What Makes a Plume Visible	      5
      What Causes Plume Contrast	      8
      Plume Effects on Light Transmission	     10
         Plume Contrast Against the Sky	     16
         Plume Contrast Against Terrain	     17
      Plume Perceptibillty	     19
3  LEVEL-1 SCREENING	     21
      Assumptions in Level-1 Screening	     22
      Preparing Level-1 Input	     22
      Exercising the Screening Model VISCREEN	     24
4  LEVEL-2 SCREENING	     39
      Selecting Particle Size Distributions	     39
      Determining Worst-Case Plume Dispersion Conditions	     41
      Accounting for Complex Terrain	     49
      Exercising VISCREEN	     50
      Alternative Use of Plume Visibility Models	     50
5  LEVEL-3 ANALYSIS	     51
      Objectives of Level-3 Analysis	     51
      Suggestions for Level-3 Analysis	     55
         Frequency Distribution of Dispersion Conditions	     55
         Calculating Plume Visual Impacts	     56

-------
      Coupling Magnitude and Frequency	  57
      Interpreting the Cumulative Frequency Curve	  57
      Summarizing Results	  58
      Optional Use of VISCREEN	  58

References  	  59

Appendix A:  Perceptibility Thresholds and Recommended Screening Analysis Criteria for
             Plumes and Haze Layers

Appendix B:  The Plume Visual Impact Screening Model (VISCREEN)

Appendix C:  Examples of Plume Visual Impact Screening and Analysis

Appendix D:  VISCREEN  Listing

Appendix E:  Dispersion Parameter Calculations
                                        vi                            Revised 10/92

-------
                            NOMENCLATURE
     babs  —  Light  absorption coefficient of an air parcel,
             proportional  to  concentrations of nitrogen dioxide
             and  aerosol  (e.g.,  soot)  that absorb visible radiation
             (m-1)
     bext  —  Light  extinction coefficient of an air parcel, the sum of
             absorption and scattering coefficients (nf )
       bR  —  Light  scattering coefficient of particle-free air caused
                                                       1
             by Rayleigh scatter from air molecules (nf )
    bscat  ™  Ll9ht  scattering coefficient resulting from Rayleigh
             scatter (air molecules)  and Mie scatter (particles), the
             sum of bR and bsp (m"*)
 (bext/m)  —  Light  extinction efficiency per unit species mass (m /g)
(b.cat/V)  --  Light  scattering efficiency per unit aerosol volume
                             ?   ^
             concentration (nr/cnr)
      b__  --  Light  scattering coefficient caused by particles only
        ^        1
             (m'1)
        C  —  Contrast at a given wavelength of two colored objects such
             as plume/sky or sky/terrain
     Cm.|n  —  Contrast that is just perceptible, a threshold contrast •
   Cp-|ume  —  Contrast of a plume against a viewing background such as
             the sky or a terrain feature
       Cr  —  Contrast of a terrain feature at distance r against the
             sky
      ACr  —  Change in sky/terrain contrast caused by a plume or extra
             extinction
       CQ  --  Intrinsic contrast of a terrain feature against the sky.
             The sky/terrain contrast at r = 0.  For a black object,
             C0 = -1
                                  vii

-------
         d — Distance between the emission source and the observer (m)
AE(L*a*b*) — Color difference parameter used to characterize the
              perceptibility of the difference between two colors.  In
              the context of this workbook, it is used to characterize
              the perceptibility of a plume on the basis of the color
              difference between the plume and a viewing background such
              as the sky, a cloud, or a terrain feature.  Color
              differences are due to differences in three dimensions:
              brightness (L*) and color hue and saturation (a*, b*)
        F- -- Solar insolation or flux incident on an air parcel within
                                              pi
              a given wavelength band (watt m~* um  )
         I — Light intensity or radiance for a given line of sight and
                                      ?   1   1
              wavelength band (watt m~<:sr~J>unr-L).  Subscripts t and h
              refer to terrain and horizon, respectively.
      Iohl- — Light intensity reflected from an object such as a terrain
                              211
              feature (watt m  sr  ynf )
    p(x,e) — Phase function, a parameter that relates the portion of
              total scattered light of a given wavelength \ that is
              scattered in a given direction specified by the scattering
              angle 9
         Q — Emission rate of a species, such as SC^, or plume flux at
              a given downwind distance, which may be less than the
              emission rate because of surface deposition and chemical
              conversion (g s  ).  Subscripts refer to species
              considered (e.g., SC^, SOT, and particulate)
         r — Distance along the line of sight from the view'ed object  to
              the observer  (m)
        rQ — Object-observer distance (m)
        r_ — Distance from observer to centroid of plume material (m)
        rv — Visual range, a parameter characteristic of the clarity  of
              the atmosphere, inversely proportional to the extinction
              coefficient.   It is the farthest distance at which a black
              object is perceptible against the horizon sky (m)
       rVQ — Background visual range without plume (m)
         t — Time  (s)
                                   vi ii

-------
        u — Wind speed (m s  )
       WD ~ Wind direction
        x — Downwind distance from emission source (m)
xmin»xmax ~" Distance along plume axis from emission source to the
             closest and most distant Class I area boundaries (m)
        x ~ Wavelength of light (m)
        p — Density of a particle (g m~3)
        a ~ Horizontal angle between a line of sight and the plume
             centerline
        8 ~ Vertical angle between a line of sight and the horizontal
        Y — Plume offset angle, horizontal angle between the line
             between the emission source and the observer and the plume
             centerline
        4> ~ Azimuthal line-of-sight angle, horizontal angle between
             the line connecting the emission source and the observer
             and the line of sight
        i|» -- Vertical angular subtense of plume
        x — Concentration of a given species in an air parcel (g m  )
        T — Optical thickness of a plume, the line-of-sight integral
             of the extinction coefficient.  Subscripts refer to the
             component of the total, or plume, optical thickness (e.g.,
             particulate, S0]j, N02)
     [  ] — Denotes the concentration of the species within brackets
        01 — Albedo of the plume or background atmosphere, the ratio of
             the scattering coefficient to the extinction coefficient
        9 -- Scattering angle, the angle between direct solar radiation
             and the line of sight.  If the observer were looking
             directly at the sun, 9 would equal 0°.  If the observer
             were looking away from the sun, 9 would equal 180°.
                                  ix

-------
                             1   INTRODUCTION
This guidance document is designed to assist the user in the evaluation of
plume visual impact as required by the Prevention of Significant Deteri-
oration (PSD) and visibility regulations of the U.S. Environmental Protec-
tion Agency (EPA).  Sources of air pollution can cause visible plumes if
emissions of particulates and nitrogen oxides are sufficiently large.  A
plume will be visible if its constituents scatter or absorb sufficient
light so that the plume is brighter or darker than its viewing background
(e.g., the sky or a terrain feature such as a mountain).  PSD Class I
areas such as national parks and wilderness areas are afforded special
visibility protection designed to prevent such plume visual impacts to
observers within a Class I area.

The objective of this document is to provide guidance on the assessment of
plume visual impacts, including the use of a plume visual impact screening
model (VISCREEN), which can be used to calculate the potential visual
impact of a plume of specified emissions for specific transport and dis-
persion (meteorological) conditions.  VISCREEN can be applied in two suc-
cessive levels of screening (Levels 1 and 2) without the need for exten-
sive input specification.  If screening calculations using VISCREEN
demonstrate that during worst-case meteorological conditions a plume is
either imperceptible or, if perceptible, is not likely to be considered
objectionable (i.e., "adverse" or "significant" in the language of the EPA
PSD and visibility regulations), further analysis of plume visual impact
would not be required as part of the air quality review of a source.  How-
ever, if screening demonstrates'that criteria are exceeded, plume visual
impacts cannot be ruled out, and more detailed plume visual impact analy-
sis to ascertain the magnitude, frequency, location, and timing of plume
visual impacts would be required.  Such detailed plume visual impact
analysis is called Level-3 analysis and is carried out by more sophistica-
ted plume visibility models such as PLUVUE II.  Figure 1 shows a  logic
flow diagram of the three  levels of plume visual impact screening and
analysis.

This guidance document and the screening model VISCREEN are designed to
replace the procedures described in the "Workbook for Estimating Visi-
bility Impairment" (Latimer and Ireson, 1980).  The procedures described
in this document are simplified by use of the screening model VISCREEN,

-------
     Level-1
     Screen
Using  VISCREEN
                     INPUT:
                     • NOx and  paniculate
                       emissions
                     • Background  visual range
                     • Distance to Class  I area
                                   Calculate contrast and   dE
                                      values  on worst-case  ,
                                  assumptions  using  VISCREEN
                                                                   Visual
                                                                impact is  not
                                                                judged to be
                                                                 adverse or
                                                                 significant
     Level-2
      Screen
Using  VISCREEN,
     a  Plume
Visibility   Model
                     INPUT: All of the
                            above  plus:
                     • Worst-case  meteorology
                     • Size  distributions
                                    Calculate  worst-day visual
                                   impacts based on  actual  area
                                   conditions using one or more
                                         of the following:
                                     (1) VISCREEN
                                     (2)  Plume  visibility model
     Level-3
     Analysis
  Using    Plume
    Visibility
      Model
TNPUT: All of the
       above, plus:
• Joint frequency of wind
  speed,  wind direction,
  stability, mixing  depth.
  and background ozone
  concentration  and  visual
  range	
                                                                                          Visual
                                                                                      impact is  not
                                                                                       judged  to be
                                                                                        adverse or
                                                                                       significant
                                                     Calculate magnitude  and frequency
                                                       of occurrence of visual  impact
                                                       using plume  visibility models
                                                                     Visual
                                                                  impact is  not
                                                                  judged to be
                                                                   adverse or
                                                                   significant
      Is
impact judge
to be adverse
or  significant
  by  govern-
                                                                   isual
                                                                impact  is
                                                              judged to be
                                                               adverse  or
                                                                ignificant
                                                        Analyze Alternatives
                                                         • Better  emission controls
                                                         • Alternative  sites
                                                         • Scaled-down  source  size
       FIGURE  1.   Logic flow  diagram for 3-level plume visual  impact analysis.

-------
instead of the hand calculation encouraged by.the earlier document.
VISCREEN is designed to evaluate plume visual effects along the plume's
entire length for two different viewing backgrounds and for two different
sun angles.  One important design feature of VISCREEN that distinguishes
it from the earlier EPA Visibility Workbook is the evaluation of the
potential perceptibility of plumes using recent psychophysical concepts
(see Appendix A).

In addition, to simplify the plume visual impact screening and analysis
process, this guidance is limited to assessing the visibility of a plume
itself, not whether the plume contributes to reductions in general visi-
bility.  Thus, a source's contribution to regional haze is not considered
in this guidance.  Although regional haze is the most extensive and seri-
ous form of visibility impairment throughout the United States and in
Class I areas, it is caused by multiple sources located throughout a
region.  A single emission source may contribute to such a problem but  is
generally not the sole (or even major) contributor.  The protection and
improvement of regional visibility must be achieved through broader regu-
latory action than is possible with the review of a single emission
source.  In addition, regional haze analysis requires a different analysis
tool:  regional dispersion models, rather than plume models.  However,  the
process of assuring that plume visual impacts are not objectionable to
visitors to Class I areas may contribute to the broader visibility
protection issue by limiting industrial source siting near Class I
areas.

These guidelines are designed to be brief and straightforward.  The reader
interested in more detail is advised to refer to the 1980 EPA Visibility
Workbook (Latimer and Ireson, 1980).  In addition, citations for several
references regarding visibility and visibility modeling are provided in
the reference section of this document.  These sources can be consulted if
the reader is interested in the details of visibility modeling, the
derivation of formulas used in VISCREEN, and the broader regulatory, pol-
icy, and technical issues associated with visibility protection.

This guidance document is organized as follows:  Section 2 provides a
brief overview of the concepts used in plume visual impact screening and
analysis including a description of parameters used to characterize the
perceptibility of plumes.  Section 3 provides a step-by-step procedure for
implementation of the simplest, Level-1 screening analysis.  Section 4
provides guidance on Level-2 screening, including the determination of
worst-case meteorological conditions.  Section 5 provides suggestions
regarding the most detailed, Level-3 plume visual impact analysis that  is
required only if a source fails both the Level-1 and -2 screening tests.
A  discussion of plume perceptibility threshold research is presented in
Appendix A.  Technical documentation and a listing of the plume visual
impact screening model VISCREEN are provided in Appendixes B and D,
respectively.  Examples of plume visual impact screening and analysis
calculations are provided in .Appendix C.

-------
                               GENERAL CONCEPTS
In this section we present a brief overview of the concepts required to
understand che technical approach used in plume visual impact screening
and analysis.  More detailed background information can be obtained from
the references cited in the back of this document.

First, we discuss what makes a plume visible.  Then, we present an over-
view of light scattering and absorption in the atmosphere and the emis-
sions that are responsible.  Next we describe the specific geometries
assumed for plume visual impact analysis and present the basic formulas
describing plume visual impact.  Finally, we discuss plume perceptibility
screening criteria.
WHAT MAKES A PLUME VISIBLE

The objective of plume visual impact screening and analysis is to deter-
mine whether or not a plume is visible as an object itself.  To understand
what makes a plume visible, we first ask what makes any object visible.
Any viewed object is visually perceptible to a human observer if the light
emanating from the object and impinging on the retina of the eye is
sufficiently different from light emanating from other objects so that the
difference or contrast between the given object and surrounding objects
(its viewing background) produces a perceptible signal to the optic nerve
and the brain.  Visual perception requires contrast.  Contrast can be
large as in the case of this black type on white paper, or contrast can be
small as in the case of touch-up paint that doesn't quite match.

Since the human eye responds differently to different wavelengths of
light, the eye responds to color as well as brightness.  The range of
wavelengths to which the human eye responds is called the visible spectrum
and ranges from the short-wavelength (0.4 micrometer, ym) blue to the
middle-wavelength (0.55 urn) green to the long-wavelength (0.7 urn) red.
Contrast can be defined at any wavelength as the relative difference in
the intensity (called spectral radiance) between the viewed object and its
background:

                        C = (

-------
where C is the contrast and Iobj and Iback are the light intensities (or spectral radiances) of the
object and its background.

If the viewed object is brighter than its background, it will have a positive contrast.  For
example, a white cloud viewed against a dark blue sky will have a positive contrast.  If the
object is darker than the background, its contrast is negative.  For example, a distant
mountain is usually visible because of a negative contrast against the horizon sky (unless the
mountain is snow-covered, in which case its contrast is generally positive).

Figure 2 illustrates the concept of contrast at different wavelengths with four hypothetical
objects.  Object  1 has spectral radiance distribution defined by It over the visible spectrum.
Because Object  1's spectral radiance is uniform over  all visible wavelengths, it is nominally
white.  Object 2 is darker than Object 1 because spectral radiances at all wavelengths are
lower than those for Object 1.  In addition, Object 2 is a different color because there is
relatively more light at the red end of the visible spectrum than at the blue end. The contrast
of Object 2 against Object 1 is negative at all wavelengths, but blue contrasts are more
negative than both green and red wavelengths.  As a result Object 2 would appear dark red
(brown) compared to Object 1.  Similarly, Object 3 would appear as a dark blue, and Object
4 would appear  as an even darker gray (or black).  If Object  3 were the viewing background
for Object 2, its contrast at the blue end of the visible spectrum  would be negative, while its
contrast at the red end would be positive. Thus, contrasts at  all  wavelengths in the visible
spectrum characterize the brightness and color of a viewed object (such as a visible plume)
relative to its viewing background.

In the plume visual impact screening model VISCREEN, contrasts at three wavelengths (0.45,
0.55, and 0.65 um) are used to characterize blue, green, and red regions of the  visible spec-
trum.  In the plume visibility model PLUVUE II, calculations are performed  for 39 wave-
lengths.  Thus, we can ascertain whether a plume will be brighter or darker or  discolored
compared to its  viewing  background by evaluating its contrasts in the blue, green, and red
portions of the visible  spectrum.  If plume contrast is positive, the plume  is brighter than its
viewing background; if negative, the plume is darker.  If contrasts are different at different
wavelengths, the plume is discolored.  If contrasts are all zero, the plume is indistinguishable
from its background (i.e., imperceptible).
                                                                            Revised 10/92

-------
*
«

t*

•5
*
oc
o
41
a
w
                      Wavelengths used

                        inVISCREEN

                           4
I
0.4
y
N
1
0.45
(blue)

1
0.55
(green)

I
0.65
(red)

I
0.7
\
S
             Varelength of Light   dim)
    FIGURE 2.  Example distributions of light intensity

    of four objects.

-------
WHAT CAUSES PLUME CONTRAST

The contrast of this black text against the white paper is caused by dif-
ferences in the amount of light reflected from the page.  Almost all of
the light impinging on the white paper is reflected, and almost none of
the light impinging on the black ink is reflected; hence, the text has a
large negative contrast (C = -1).   Plume contrast is caused by a somewhat
different set of physical processes:  plume contrast results from an
increase or decrease in light transmitted from the viewing background
through the plume to the observer.

This increase or decrease in light intensity (spectral radiance) is caused
by plume constituents that scatter and/or absorb light.  There are only
two common plume constituents that scatter or absorb light.  Particulates,
depending on their nature, can scatter light or both scatter and absorb
light.  Nitrogen dioxide (N02) absorbs light of all wavelengths in the
visible spectrum but it is a stronger absorber at the blue end of the
spectrum.

We can characterize the atmospheric optical properties of a plume in a
manner analogous to the way plume concentrations are characterized.
Instead of using mass concentration (vg/nr), which is the mass of a given
species per unit volume of ambient air, we use parameters called the light
scattering coefficient (bsca+), the light absorption coefficient (babs),
and their sum, the light extinction coefficient (bext).  These coef-
ficients are essentially the concentrations of the equivalent light scat-
tering, absorption, and extinction cross-sectional area.  They are cross-
sectional area per unit volume of air; hence, their units are m /m  or m   .

These coefficients are similar to concentration in that they are propor-
tional to the mass concentrations of the particulates and N02 that scatter
and/or absorb light; however, since different chemical species have dif-
ferent light extinction efficiencies, there is no simple one-to-one rela-
tionship between mass concentration and light extinction.  For example,
submicron particles between 0.1 and 1 ym are much more effective in scat-
tering light per unit mass than are either smaller or larger particles.
Soot is a stronger light absorber than N02 per unit mass.  Table 1 shows
the light extinction efficiency of several common constituents of plumes
and background atmospheres.  Light extinction coefficient (bext) is the
product of the mass concentration and the light extinction efficiency of
the given species.

Plume visual impact models account for the concentrations of various spe-
cies in a plume  (e.g., N02, submicron particulate, coarse particulate, and
soot) and their  light scattering and absorption properties at various
visible wavelengths  (e.g., blue, green, red).

-------
TABLE 1.  Typical light extinction efficiencies for
constituents of plumes and background atmospheres.

                                   Light Extinction
                                    Efficiency at
                                     x = 0.55 um
Constituent                           (m /g)

Soot                                        13

Hygroscopic fine particles including        4-8
  (SOj) and nitrates (N0~)

Fine particles (0.1 < D < 1 urn)              3

Coarse particles (1 < D < 10 ym)            0.4

Nitrogen dioxide (N02)                     0.17

Giant particles  (D > 10 ym)               < 0.04
Sources:  Latimer et a!., 1978, 1985; Latimer and
          Ireson, 1980

-------
PLUME EFFECTS ON LIGHT TRANSMISSION

Figure 3 shows a schematic of the viewing situation that is mathematically
represented in a plume visual impact model.  A plume of limited dimensions
is embedded in an otherwise uniform background atmosphere.  The observer's
line of sight intersects the center of the plume at distance r_ from the
observer and it intersects a viewing background object (e.g., a mountain)
at distance rQ.  The direct rays from the sun are at angle e with respect
to the line of sight.  The change in the spectral light intensity at any
point along the line of sight (either inside or outside the plume) as a
function of distance r along the line of sight is:
                                                                        (1)
             Ml        W*W            ~TU    **%»%* ^    «*

where

             r = the distance along the line of sight from the object to
                 the observer;

        p"(x, e)= the scattering distribution or phase function for scat-
                 tering angle e (see Figure 3 for definition of 9) modi-
                 fied to account for multiple, as well as single, light
                 scattering;
                                       7
         F5(x) - the solar flux (watt/m /urn) incident on the atmosphere,
     bscat (x) ~ the ^Qht scattering coefficient, which is the sum of the
                 Rayleigh scattering (due to air molecules), bp, and the
                 scattering due to particles, bs_:
                  bscat(x) = bR(x) + bsp(x)    ;                        (2)

      bext (x) = the Ii9nt extinction coefficient, which is the sum of  the
                 scattering, bscat(*)t and absorption, (x) babs,
                 coefficients:

                  "ext ' Wx> + Wx>    •

On the right-hand side of Equation (1), the first term represents  light
absorbed and  scattered out of the line of sight; the second term repre-
sents light scattered into the line of sight.  The values of bscat and
^abs can ^e eva1uated if the aerosol and N02 concentrations and such
characteristics as the refractive index and the size distribution  of the
                                  10

-------
       	-^
'••">iji''!.'		IT	 +
                                                      Viewing
                                                      Background
                                                      Object
FIGURE  3.   Geometry  of plume, observer, viewing  background,
and  sun.
                           11

-------
aerosol are known.   Except  in the cleanest atmospheres,  b$cat Is dominated
by bs_; also,  unless  soot  is  present, b^  is dominated  by the-absorption
coefficient due to  N02.  Scattering and absorption are wavelength-depen-
dent, and effects are greatest at the blue end (x  = 0.4  ym) of thr visible
spectrum (0.4  < x < 0.7  urn).   The Rayleigh scattering coefficient DR is
proportional to x~  ;  the scattering coefficient caused by particles is
generally proportional to  x~n, where 0 < n < 2.  Also, N02 absorption is
greatest at the blue  end.   This wavelength dependence causes the natural
blue sky coloration as well as discoloration of the atmosphere.

For a uniform  atmosphere,  without inhomogeneities  caused by plumes (where
bscat and bext do not vary  with distance r along the line of sight), Equa-
tion (1) can be solved to  find the intensity and coloration of the horizon
sky:
                                  bscat(x)
                             4ir
                                  'ext
                                      (x)
                                           F(x)
                                                                       (4)
The perceived intensity of distant bright and dark objects will approach
this intensity as an asymptote, as illustrated by Figure 4.

Atmospheric coloration is determined by the wavelength-dependent scatter-
ing and absorption in the atmosphere.  The spectral distribution of I(x)
for x over the visible spectrum determines the perceived color and light
intensity of the viewed object.  The relative contributions of scattering
(aerosols plus air)  and absorption (N02) to coloration can be illustrated
by rearranging Equation (1):
          1  dJIxj
        TTIT   dr
                               P(X.O)
                     'scat
(x)
                                      Fs(x)
TOO
Jabs
                      (x)
                                                                       (5)
Note from Equation (4)  that when light absorption is negligible compared
with light scattering  (i.e.,  b
Iho(x), is simply:
                                 .
                                     b   ), the clear horizon intensity,
                                        t
                                p(x,0)
                                       Fs(x)
                                                                       (6)
We now can rewrite Equation (5):
             1
            TTxT  dr
                       = b
                          scat
                              (x)
          TOT
                                             (7)
                                   12

-------
!„(»)
           LIGHT INTENSITY OF HORIZON
                           Objtct-Obstrvcr Distinct r
   FIGURE 4.   Effect of an atmosphere on the  perceived light
   intensity of objects.

-------
Equation (7) is thus an expression relating the effects of light scatter-
ing and light absorption to the change in spectral light intensity with
distance along a sight path.  On the right-hand side of Equation (7), the
first term is the effect of light scattering, and the second terra is the
effect of light absorption (Nt^).  As noted previously, since bscat and
babs (due to N02^ are Stron9 functions of wavelength and are greater at
the blue end (\ = 0.4 um), atmospheric coloration can result.

Equation (7) makes clear that N0£ always tends to cause a decrease in
light intensity since the second term in Equation (7) is always nega-
tive.  However, particles may brighten or darken a plume, depending on
whether the first term in Equation (7) is positive or negative.  If, at a
given point along the sight path, I(x) is greater than the clean horizon
sky intensity I^Q(X), then the quantity in brackets in the first term on
the right-hand side of Equation (7) will be negative, which means that the
net effect of scattering will be to remove light from the line of sight.
This effect would occur if a bright, white cloud or distant snowbank were
observed through an aerosol that did not contain N02«  If, however, I(x)
is less than IhQ(x), then the quantity in brackets in Equation (7) will be
positive, which means that the net effect of scattering will be to add
light to the line of sight.  This effect would occur if a distant, dark
mountain were observed through an aerosol that did not contain NC^; scat-
tering would cause the mountain to appear lighter.  Only light absorption
can cause I(x) to be less than I^Q/*). and whenever I(x) < I^o(x)» scat-
tering will add light to the sight path, thereby masking the coloration
caused by N02 light absorption.

The mathematical expressions used in this document and the plume visual
impact screening model VISCREEN are simply solutions to Equation (1) for
different boundary conditions and for different values of bscat, bext,
p(e) and FS as they are affected by natural and man-made light scatterers
and absorbers.  The plume visibility models use similar formulations, but
most account for multiple scattering effects.*

Now a plume (either ground-based or elevated) may be visible because it
contrasts with a sky viewing background as shown in Figure 5(a) or it con-
trasts with a terrain feature as shown in Figure 5(b).  The plume visual
impact screening model VISCREEN evaluates both of these possible viewing
backgrounds.
 * Multiple  scattering  is light scattered into the line of sight after
   previous  scattering  (i.e., light reflected from terrain features and
   light  scattered from other portions of the atmosphere).
                                   14

-------
          (a)  Plume Visible Against  the  Sky
                //////A////
            (b)   Plume  Visible Against Terrain
FIGURE 5.   Two viewing  situations  in which plumes may be
visible.
                        15

-------
Formulas for contrasts representative of both types of viewing situations can be derived by
solving Equation (1) for appropriate boundary  conditions.

Plume Contrast Against the Sky

Let us consider now the geometry shown in Figure 3, namely, the case of a plume embedded
in an  otherwise uniform background atmosphere.  If we ignore the effects of multiple
scattering, Equation (1) can be solved for the contrast between the plume and the horizon sky
background (see Figure 5a) as observed at distance rp from the plume as follows (Latimer and
Ireson, 1980):
                                                                                   (8)

                                    [1 - ^(-tpiwu)] e*P(-&*

where

          Ih  = spectral radiance of horizon sky (without plume present)

             = spectral radiance of plume  viewed in front of horizon sky
       P     SB average phase function for the plume constituents and the background
                atmosphere

        to    = average albedo of plume and background, where albedo is the ratio of light
                scattering to total  light extinction

             - plume optical thickness along the line of sight (increment above background)
                 plum*
                                           16                             Revised 10/92

-------
      bext = background atmosphere's light extinction coefficient

        r. = distance between plume centerline and observer
Note that,  depending on whether the product of the phase function and the
albedo (poi) for the plume is larger or smaller than that for the back-
ground, the plume will be brighter (C > 0) or darker (C < 0) than the
background  horizon sky.  Also note that the contrast is dependent on the
plume optical thickness (Tplume); as Tp-,ume approaches zero, Cplume
approaches  zero.  Plume contrast also diminishes as the plume-observer
distance rp increases.
Plume Contrast Against Terrain

To characterize the types of visibility impairment represented in Figure
5(b), we need to calculate a change in sky/terrain contrast caused by a
plume:
                                    with plume
             without plume
where
                    with plume
                                  t-plume "  h-plume
                                      I
h-plume
                 without plume
                           Cr = the sky-terrain contrast of a terrain
                                feature at distance r from an observer

                       It, Ih = the spectral radiances of a terrain
                                feature and the horizon sky (unaffected by
                                p1ume)

           *t-plume* ^-plume ~ the sPectral radiances of plumes viewed in
                                front of horizon sky and terrain
For simplicity we assume that the terrain that is viewed behind the plume
has an intrinsic radiance, Igbj, which is a function of the horizon sky
radiance Ih, namely, Iobj = (1 + C0)Ih.  C0 is the intrinsic contrast.
the terrain were black, CQ would equal -1.
                                 If
                                  17

-------
Again solving Equation (1) and ignoring multiple light scattering, we can
derive the following expression for the change in terrain contrast caused
by the plume (Latimer and Ireson, 1980):
-C0   exP<-bextro>   1    '     l
                                                                        (9)
                                    \  +  plume

where rQ = distance between the terrain object and the observer.

Equations (8) and'(9) are the analytical expressions at the heart of  the
plume visual impact screening model VISCREEN.  Careful examination of
these two equations illustrates the following sensitivities:

     1.   Plume contrasts (against both the sky and terrain) increase
          with increasing plume light extinction (i.e., as concentrations
          of particulates and N02 in a plume increase).

     2.   Plume contrasts increase if the line of sight is oriented to
          intersect a larger amount of plume material  (i.e., the  line of
          sight is along the plume center!ine).

     3.   Plume contrasts increase for sun angles and  for particle size
          distributions that tend to maximize the difference (both posi-
          tive and negative) between the phase functions for the  back-
          ground  atmosphere and for the plume.

     4.   Plume contrasts increase if the plume is moved closer to the
          observer.

     5.   Plume contrasts increase with decreasing light extinction of the
          background atmosphere (i.e., with increasing background visual
          range).

     6.   Plume contrasts against terrain are maximum  if the terrain
          object  is relatively close to the observer and the terrain's
          intrinsic contrast is maximum (e.g., if it were black).

 Since  screening calculations are designed to be conservative estimates of
 worst-case  conditions, situations are selected to (1)  maximize the concen-
 trations  and  light scattering efficiencies of optically active plume  con-
 stituents,  the intersection of the line of sight and the plume, the back-
 ground visual range, the intrinsic contrast of terrain objects, and the
 difference  between background and plume phase functions; and (2)  minimize
 the distance  between the observer and the plume.  Once conservative
 estimates of  worst-case conditions are specified, the  plume visual impact
                                   18

-------
screening model VISCREEN uses Equations (8) and (9) to calculate plume contrasts.  If such
contrast values are larger than screening criteria, the possibility that the plume will cause sig-
nificant visual impact cannot be ruled out, and less conservative, more realistic estimates
would be required.
PLUME PERCEPTIBILITY

The perceptibility of a plume depends on the plume contrast at all visible wavelengths.  At a
single wavelength, the contrast between the plume and its surroundings is determined by the
difference in the intensity of the light reaching the observer from each.  Therefore a single
measure, intensity, could be used to quantify contrast if visible light were composed of a
single wavelength.  With a range of wavelengths, a measure of contrast  must recognize both
"overall" intensity, and perceived color, and so perceptibility is really a function of changes in
both brightness and color.  To address  the added dimension of color as well as  brightness, the
color contrast parameter, AE, was chosen for use as the primary  basis for determining the
perceptibility of plume visual impacts in screening analyses. AE provides  a single measure of
the difference between two arbitrary colors as perceived by humans.  This parameter allows
us to make quantitative comparisons of the perceptibility  of two  plumes, even though one
may be a reddish discoloration viewed against a blue sky while the other may be a white
plume  viewed against a dark green forest canopy.

Contrasting  surfaces are detected by human vision using three  types of visual information
(cues).  The trichromatic theory of Helson (1938) and Judd (1940) predicts colors perceived
by human subjects based on the  visual qualities described as brightness (intensity), lightness
(saturation), and color (hue).  Perceived brightness of a colored surface is dependent  upon the
intensity of  the applied illumination.  For example, the brightness of the white of a daisy is
larger for a  daisy in direct sunlight than for a daisy in the shade.  The color or hue of a
surface is dependent on the ratio of the intensity of red to green  light that is reflected.  The
lightness of a color is the strength or density of a color and is  often called the saturation.  An
example of  this cue comes from photography: a properly or slightly underexposed color is
said  to be more saturated than an overexposed color which appears to be washed out by the
addition of white.  Color contrast is therefore made up of differences in  these three visual
qualities (cues).

As implied by its name, the trichromatic theory of color assumes that all shades of color are
composed of three primary colors: red, green, and blue. These primary  colors are not single
wavelengths, but rather an envelope of wavelengths, whose peak intensities occur at
frequencies  we associate with  each of the primary colors.  The purely chromatic character-


                                           19                              Revised 10/92

-------
istics of a perceived color are then described by three numbers (X=red, Y=green, Z=blue) that
represent the intensity of each color in the "mix".  (These are computed as the integration
over the visible spectrum of the product of the intensity of the illumination and the
trichromatic weighting function for each primary color.)

The amounts of red, green, and blue (X,Y,Z) can be used to approximate the three cues used
to quantify the contrast between colored objects. Three empirical mathematical functions of
(X.Y.Z) were defined which quantitatively best capture the qualitative features of the three
cues: brightness, hue,  and saturation.  Each of these three mathematical functions is defined
relative to the one or more components of chromaticity of a reference white card under direct
sunlight (X0,Y0,Z0). For brightness, only a single chromatic  component is needed, and since
the eye is most sensitive  to intensity changes in green, the function for brightness, L*, is
defined in terms of Y. Since hue depends on the red/green reflected intensity ratio, the
function describing hue, a*, is defined in terms of X and Y.  The mathematical function
describing the amount of saturation, b*, is  defined in terms of Y and Z (see equations in
Appendix B).

For each of the three visual cues, the contrast between two  surfaces is simply a difference
between the values of the mathematical functions for each surface.  For example, contrast due
to changes in brightness is defined as the difference in the function for brightness, AL*. The
total color contrast, AE, is taken to be the sum
                                       2 + (Aa ')2

This formulation is based on the following assumptions:

(1) AE depends only on AL, Aa, and Ab;

(2) Differences in contrast cues AL*, Aa*, and Ab* are independent of one another.

Although a AE of 1 and a contrast of 0.02  have been  traditionally assumed to be the threshold
of perceptibility, a survey of the literature (see Appendix A) suggests a broad range of
perceptibility thresholds.  The most sensitive observers are able to detect contrasts or color
changes one-half this magnitude, and the casual observer may require contrast or color
changes more than two times larger than these "traditional" values.  In addition, the literature
suggests that perceptibility thresholds increase for very wide and for very narrow plumes,
with plumes less than 0.02° being essentially imperceptible.  Figure 6 summarizes the range
of perceptibility thresholds supported in the literature.
                                           19a                             Revised 10/92

-------
The plume visual impact screening model VISCREEN is designed to ascertain whether the
plume from a facility has the potential to be perceptible  to untrained observers under
"reasonable worst case" conditions.  If either of two screening criteria is exceeded, more
comprehensive (and realistic) analyses should be carried out. The first criterion is a AE value
of 2.0; the second is a  green (0.55 urn) contrast value of 0.05.  In the  case of sufficiently
narrow or broad plumes, the higher perception thresholds (for diffuse-edged plumes) are used
instead of the above criteria.
                                          19b                              Revised 10/92

-------
  1.0
  0.1
in
2
 0.01
                        Howe11 and Hess (1978) data:
                        Square-wave gratings  (sharp edge)
                        Sine-wave gratings (diffuse edge)
                                         Upper bound (screening criterion)
       Perceptibility
        "Threshold"
                                                                  Data of Malm et  al,
                                                                  (1986) for sharp-
                                                                  edged plumes
 0.001
                               0.1                         1.0

                              Plume Vertical Angular Subtense* (°)
10
    FIGURE   6.   Plume  perceptibility threshold  as a  function  of  plume
    thickness  (Y).   See  definition of 
-------
                               LEVEL-1 SCREENING
This section describes the process of Level-1 plume vi-:'al impact screen-
ing using the screening model VISCREEN and uie brief ii.put required to
initiate the screening process.  Details of the plume visual impact
screening model are provided in Appendix B.
ASSUMPTIONS IN LEVEL-1 SCREENING

Level-1 screening is designed to provide a conservative estimate of plume
visual impacts (i.e., impacts that would be larger than those calculated
with more realistic input and modeling assumptions).  This conservatism is
achieved by the use within the screening model VISCREEN of worst-case
meteorological conditions:  extremely stable  (F) atmospheric conditions,
coupled with a very low wind speed (1 m/s) persisting for 12 hours, with a
wind that would transport the plume directly  adjacent to the observer  (as
shown schematically in Figure 7).
PREPARING LEVEL-1 INPUT

Through the use of default parameters, the input required for Level-1
plume visual impact screening is limited to the following:

     Emission rates of particulates (including soot and primary sulfate)
     and nitrogen oxides (including primary N02)

     Distance between the emission source and (1) the observer, (2) the
     closest Class I area boundary, and (3) the most distant Class  I area
     boundary*
* It should be noted that although VISCREEN is designed primarily for
  assessing plume visual impacts in Class I areas, it can also be applied
  in PSD Class II areas.  In such cases these distances can be specified
  arbitrarily.
                                 21

-------
               Assumed Worst-Case_
               ~  Plume Centerlines
•\
                                                  Boundary of
                                                  Class  ! Area
                                                     Minimum distance
                                                     to Class I area (d)
                                                         , Emission
                                                         Source
FIGURE 7.  Determining distances for Level-1 screening.
                      22

-------
           Background visual range appropriate for the region in which the Class I area is
           located.

Before using VISCREEN, the analyst should summarize the emission rates for

           Primary paniculate matter
           Nitrogen oxides (NOJ
           Primary nitrogen dioxide  (NO2)
           Soot (elemental carbon)
           Primary sulfate  (S£>4°)
SO2 emissions are not required as input to VISCREEN.   Moreover, the issue of secondary
sulfate formation  (5O4") is not treated in VISCREEN because of the limited range of
applicability of a steady state Gaussian dispersion model and because of the uncertainty of
estimating the conversion of SO2 to SO4 in a coherent plume.  More sophisticated plume
visibility models treat both secondary sulfate and nitrate.

These emissions can be provided in any  units convenient to the analyst since VISCREEN will
prompt the analyst for his/her choice of units of mass (e.g., grams, kilograms, metric  tonnes,
pounds, or tons) and time (e.g., seconds, minutes, hours, days, or year).  Thus, emissions can
be specified in g/s or ton/yr or whatever combination is  desired.

Emission rates should be the maximum short-term rates  expected during the course of a year.
The values used for plume visual impact screening generally would be the maximum  emission
rates for which the air quality permit is being applied and would correspond to those  used for
short-term (i.e., 1-, 3-, and 24-hour average) air quality impact analyses.

For almost every emission source, the emission rates of  the last three species (primary NO2,
soot, and sulfate) can be assumed to be zero. However, if NO2 is directly emitted from the
emission source (e.g., from a chemical process  such as a nitric acid plant) as opposed to
being formed in the atmosphere from NOX emissions, this  primary NO2 can be considered.
Even  if primary NO2 emissions are set to zero,  VISCREEN assumes that 10 percent of NOX
emissions is initially converted to NO2 either within the  stack of the source or within  the first
kilometer of plume  transport (Latimer et al.,  1978). If soot is known to be emitted (e.g., if
diesel vehicles are a component of the emissions source), its emission rate should be  provided
separately from that of other particulates. Finally, some sources (such as oil-fired power
plants or smelters) may have a significant component of primary sulfate in a size range that
has maximum light scattering efficiency. If so, primary sulfate (  SOj  ) emissions should be
specified and input  separately from either particulate or  soot. In summary, for most sources

                                          23                             Revised 10/92

-------
the analyst need only input the total participates and NOX emission rates (the first two
categories of emissions required by VISCREEN);  only the small fraction of emission sources
producing nonzero primary NO2, soot, and sulfate requires input of these emissions to
VISCREEN.

Using a topographic map of appropriate scale, the analyst should identify the portion of the
Class I area  that is closest to the emission source  and measure (or compute) the distance
between the  emission source  and this closest boundary.  This distance is the distance between
the emission source and the observer that should be input to VISCREEN (d in Figure 7).
Then the analyst should draw plume centerlines offset by half a 22.5° sector width (i.e.,
11.25°) on either side of this hypothetical, worst-case observer location as shown  in Figure 7.
The  analyst should determine the downwind distance (along these assumed plume centerlines)
to the closest (^J and most distant (x^) Class I area boundaries (even if these two
distances are on opposite sides of the observer).  If either x,^ is greater than d, set x^ equal
to d  for the sake of conservatism.  There may be certain shapes of Class I areas where  the
plume centerlines drawn on opposite sides of the observer cross boundaries more  than once.
In such cases the smallest x^ and the largest x,,,^ should be used to be conservative (see
Figure 8).

The  last input needed to perform a Level-1 screening analysis is the  background visual range
of the region in which the Class I area is located.  Figure 9 provides default background
visual range values for the contiguous United States.  In cases where there  is more applicable
onsite data, source owners should consult with the Federal Land Manager for the  Class I area
in question concerning  appropriate regional background visual range values for input to
VISCREEN or other plume visibility models.

With emissions, distances, and visual range as the only inputs required for Level-1 screening,
the analyst can  exercise the screening model VISCREEN.
EXERCISING THE SCREENING MODEL VISCREEN

The plume visual impact screening model VISCREEN is designed for use on an IBM-
compatible personal computer with minimal memory requirements.  VISCREEN is written in
FORTRAN 77.  VISCREEN can be  run simply by inserting the VISCREEN program diskette
in the A drive and typing A:VISCREEN.  The model first requests  the names of two disk
files (that it will create) to which results  will be written.  These include a summary file,
which will contain a formatted, tabular presentation of results, and a results file, which
includes arrays of results that can be read into spreadsheet programs for further analyses,
plotting, et cetera.

                                          24                             Revised  10/92

-------
       Emission  source
Plume  offsetj
angle t
(= 11.25° for
Level-1  screening)
Plume  Centerline
                      Lines of sight  for
                      every 5° of azimuth f
                             Terrain  viewing background
                             assumed to be at far  edge
                             of plume
                                             n  (downwind distance
                                          to closest Class I area
                                          boundary)
               Boundary of
               Class  I Area'
                                                                                              (downwind distanci
                                                                                      to most distant Class I
                                                                                      area  boundary)
                                                                                           Plume  subtending
                                                                                           22.5 ° sector width
                                                                                              horizontal
                 FIGURE  8,   Geometry of  plume  and observer lines of sight used for plume
                 visual  impact  screening.
                                                    25

-------
                                                                     (/I
                                                                     3
                                                                      o
                                                                       >
                                                                     3   •
                                                                    <—    3
                                                                        "O
                                                                     a»    re
                                                                     c
                                                                     3
                                                                     en c
                                                                     -*  ai
                                                                     u  i
                                                                     c *->
                                                                     O •>-
                                                                         to

                                                                         >
                                                                         
26

-------
The model will request the inputs previously discussed (emissions, dis-
tances between emissions source, observer, and Class I area, and the back-
ground visual range).  It will also ask whether you want default input
parameters.  For Level-1 plume visual impact screening, the analyst should
use the default input offerred in VISCREEN.  Once the analyst has provided
the requested input and confirmed this selection of input, VISCREEN will
begin its calculations.  (Execution may take several minutes if VISCREEN
is run without a math coprocessor.)

After program execution, VISCREEN will display a summary of the Level-1
screening calculations similar to that shown in Figure 10.  All four tests
are based on the screening criteria UE = 2, Cp(x = 0.55 um) =-0.05] and
the perception threshold curve for diffuse-edged plumes shown in Figure 6.
VISCREEN will identify whether the given plume passes or exceeds four
tests.  The first two tests refer to visual impacts caused by plume par-
cels located inside the boundaries of the given Class I area.  The last
two tests are for plume parcels located outside the boundaries of the
Class I area.

The first two tests are used to determine visual impacts when so-called
integral vistas are not protected (or are not of concern in the given
analysis).  An integral vista is a view from a location inside a Class I
area of landscape features located outside the boundaries of the Class I
area.  The Federal Land Manager for a given Class I area should be contac-
ted to determine whether analyses for integral vistas are required.  If
not, the VISCREEN analysis results for plume parcels located outside the
Class I area could be ignored (the last two tests), and results for par-
cels within the Class I area (first two tests) would be used for screen-
ing.  If integral vistas are protected as well as the within-area views,
VISCREEN results for parcels located inside and outside the Class I area
should be used to determine whether the emission source passes the given
level of screening (i.e., all four tests should be used).  For views both
inside and outside the Class I area, calculations are performed for two
assumed plume-viewing backgrounds:  the horizon sky and a dark terrain
object.  VISCREEN assumes that the terrain object is black and located
adjacent to the plume on the side of the center!ine opposite the obser-
ver.  In the example shown in Figure 10, the plume from the power plant
fails all four screening tests.

After the display of the screening test summary, VISCREEN will ask the
analyst whether the calculated results for lines of sight (plume parcels)
with maximum predicted visual impact should be displayed.  If selected,
VISCREEN displays a summary similar to that shown in Figure 11.  This sum-
mary shows calculated plume perceptibility (color difference) A£
parameters for four lines of sight corresponding to plume parcels located
inside/outside of the Class I area and in front of sky/terrain viewing
                                    27

-------
OVERALL RESULTS OF PLUME VISIBILITY SCREENING

SOURCE:  Public Electric Coal #3
CLASS I AREA:  Longview NP

 INSIDE  class I area —
Plume delta E EXCEEDS screening criterion for SKY background
Plume delta E DOES NOT EXCEED screening criterion for TERRAIN background
Plume contrast DOES NOT EXCEED screening criterion for SKY background
Plume contrast DOES NOT EXCEED screening criterion for TERRAIN background

 OUTSIDE class I area —
Plume delta E EXCEEDS screening criterion for SKY background
Plume delta E EXCEEDS screening criterion for TERRAIN background
Plume contrast EXCEEDS screening criterion for SKY background
Plume contrast DOES NOT EXCEED screening criterion for TERRAIN background
 SCREENING CRITERIA:        DELTA E =  2.0
                     GREEN CONTRAST =  .050
FIGURE 10.  Sample VISCREEN screening summary.
                                   28

-------
VIEW    ANGLES (DEGREES)       DIST (KM)  PLUME PERCEPTIBILITY  DELTA E(L*A*B*)
 no      phi  alpha   psi       x     rp         forward        backward
Line of sight with maximum perceptibility for plume viewed
         against a SKY background INSIDE  class I area.
 33     84.4   84.4  ' 1.39   80.0   15.7          4.7 *           2.4 *

Line of sight with maximum perceptibility for plume viewed
         against a TERRAIN background INSIDE  class I area.
 33     84.4   84.4   1.39   80.0   15.7          1.5              .6

Line of sight with maximum perceptibility for plume viewed
         against a SKY background OUTSIDE class I area.
  7     35.0  133.8    .96   63.5   21.6          5.7 *           2.7 *

Line of sight with maximum perceptibility for plume viewed
         against a TERRAIN background OUTSIDE class I area.
  1      5.0  163.8    .29   24.9   55.8          3.4 *           1.2
 * Exceeds screening criteria
             FIGURE 11.  Sample VISCREEN summary for lines of sight
             with maximum plume perceptibility.
                                29

-------
backgrounds.  These four lines of sight were selected by VIS GREEN (from as many as 37
lines of sight for which plume contrast calculations were made) as the plume parcels with
maximum predicted visual impact (i.e., the largest ratio of the calculated plume AE parameter
or contrast to the screening criterion).* The lines of sight (LOS's) are described by a view
number.  The plume is viewed in 5° increments of azimuth  (see Figure 8) starting from the
emission source. Thus, view No. 1 would be the plume parcel 5° to the right (or left) of the
emission source. The last three views or lines of sight are for plume parcels 1 kilometer
downwind from the source and at the nearest and most distant Class I area boundaries.  These
are included to describe the plume appearance for LOS's  nearly across the source, and at the
points of plume entry and exit from the Class I area.  In addition to view number, the lines of
sight are described by three angles (see Figure 12):

            (phi), which is  the azimuthal angle (in degrees) between the line connecting  the
              source and observer and the line of sight;

           a (alpha), the angle (in degrees) between the  line of sight and the plume center-
              line; and

           Y (psi), the vertical angle (in degrees) subtended by the plume (see Figure 3).

In addition, two distances relevant to the  given plume parcel are provided that are critical to
the identification of perceptibility.  The plume parcel's downwind distance (x) and the
distance between the observer and the plume (rp) are provided (in kilometers).  A third
distance is that from the observer to terrain background (r0).

Results are provided for two assumed worst-case sun angles. The "forward scatter" case
refers to a situation in which  the sun is in front of the observer such that the scattering angle
(9) is 10°. Such a sun angle  will tend to maximize the light scattered by plume particulates
and maximize the brightness of the plume.  (In reality, such a sun angle may or may  not
occur during worst-case conditions for the given line of sight).  The "backward scatter" case
refers to a situation in which  the sun is behind the observer such that the scattering angle is
140°.  A plume is likely to appear the darkest with such a sun  angle.  Asterisks denote values
that exceed the screening criteria.
            The largest ratio, rather than the largest AE and contrast values, is used because a
            broad or  narrow plume may have large AE or contrast and yet be imperceptible
            (see Figure 6).


                                           30                              Revised 10/92

-------
Emission
Source
                                           View ing Background
              Observer;
                                                                           22.5
                                                                           Wide Plume
      FIGURE  12 .  Distances and angles that specify a given line  of sight.
                                        31

-------
After displaying the summary of lines of sight with maximum calculated plume visual impact,
VISCREEN asks whether AE's for lines of sight are to be displayed. If this option is
selected, VISCREEN will show all the lines of sight analyzed in the screening procedure.
These results are displayed in order of view number, first for the sky background cases and
second for the terrain background cases.  Several screens of output are necessary to show all
the lines of sight (as many as eight screens, four for each of the two viewing backgrounds).
Figure 13 is a sample of such output.

After viewing the AE summaries and output, the analyst is given the option of viewing plume
contrast values at 0.55 urn.  Plume contrasts at three wavelengths of light are calculated by
VISCREEN, and are written to the results file.  These may be useful in characterizing the
relative brightness and color of the plume compared to its viewing background.  A summary
of lines of sight with maximum negative or positive green contrast is provided (see example
in Figure 14).  Since maximum plume perceptibility may occur for lines of sight different
from those  of maximum plume contrast, the lines of sight summarized here may be different
from those  in the AE summary.  As for the AE summary, asterisks denote contrasts whose
absolute values exceed the screening criterion.  In a fashion similar to that for the AE
summary, VISCREEN gives the analyst the option of viewing the green plume contrast values
for all  lines of sight (Figure  15).  In some cases, because VISCREEN calculates results for
lines of sight every  5 degrees, one or several of the lines of sight may be physically
unrealistic.  The analyst should review each line of sight, paying particular attention to those
for which screening criteria are exceeded, to verify that screening decisions are not based on
unrealistic geometries.  For example, in Figure 13, view number 2 corresponds to a 10° line
of sight (0),  If the view is toward the north then this worst-case impact should be eliminated
because it is associated with an unrealistic geometry. The 10 degree forward scatter scenarios
are only possible for views  to the east (mornings), south (high latitudes and winter periods),
and west (evenings). Screening decisions should be based on the worst case impacts
associated with realistic geometries.

After these VISCREEN outputs are displayed, the analyst is asked whether additional
calculations are to be made with changed emissions, distances, and so on.* Unless the  analyst
is interested in evaluating the effect of alternative emissions or siting distances, additional
VISCREEN analyses will not be needed for Level-1 screening.

The summary and results files, with filenames as entered by the user when VISCREEN was
invoked, are written to the disk as the program executes.  If multiple runs of VISCREEN are
carried out (e.g., with changed emissions), results for these runs are appended to the end of
the files. The summary
                                           32                             Revised 10/92

-------
                      PLUME DELTA E AGAINST A SKY BACKGROUND




VIEW    ANGLES (DEGREES)       DIST (KM)  PLUME PERCEPTIBILITY  DELTA 'E(L*A*B*)
no
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
phi
5.0
10.0
15.0
20.0
25.0
30.0
35.0
40.0
45.0
50.0
55.0
60.0
65.0
70.0
75.0
80.0
85.0
90.0
alpha
163.8
158.8
153.8
148.8
143.8
138.8
133.8
128.8
123.8
118.8
113.8
108.8
103.8
98.8
93.8
88.8
83.8
78.8
PS1
.29
.42
.55
.66
.77
.87
.96
1.04
1.12
1.19
1.25
1.30
1.34
1.37
1.38
1.39
1.39
1.37
X
24.9
38.3
46.8
52.7
57.2
60.7
63.5
65.9
68.0
69.9
71.6
73.2
74.6
76.1
77.4
78.8
80.2
81.6
rp
55.
43.
35.
30.
26.
23.
21.
20.
18.
17.
17.
16.
16.
15.
15.
15.
15.
15.
8
1
3
1
4
7
6
0
8
8
1
5
1
8
6
6
7
9
forward
2
3
4
5
5
5
5
5
5
5
5
5
5
5
4
4
4
4
.3
.7
.8
.3
.5
.7
.7
.6
.5
.4
.3
.2
.1
.0
.9
.8
.7
.6
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
backward
1
1
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
.4
.6
.0
.3 *
.5 *
.6 *
.7 *
.7 *
.7 *
.7 *
.6 *
.6 *
.5 *
.5 *
.5 *
.4 *
.4 *
.3 *
                        FIGURE 13.  Sample VISCREEN AE output.
                                33

-------
                                             -GREEN PLUME CONTRAST-

VIEW     ANGLES        DISTANCES (KM)        forward       backward screening
 no     phi  alpha      x      rp     ro     contrast      contrast criterion
Line of sight with maximum contrast for plume viewed
         against a SKY background INSIDE  class I area.
 33     84.4   84.4   80.0   15.7   32.0  -  -.004           -.033       .05

Line of sight with maximum contrast for plume viewed
         against a TERRAIN background INSIDE  class I area.
 33     84.4   84.4   80.0   15.7   32.0      .020            .011       .05

Line of sight with maximum contrast for plume viewed
         against a SKY background OUTSIDE class I area.
  2     10.0  158.8   38.3   43.1   57.0     -.008           -.064 *     .05

Line of sight with maximum contrast for plume viewed
         against a TERRAIN background OUTSIDE class I area.
  2     10.0  158.8   38.3   43.1   57.0      .044            .038       .05
 *   Absolute value exceeds screening criteria
              FIGURE 14.  Sample VISCREEN summary for lines of
              sight with maximum plume contrast.
                                34

-------
                     PLUME  CONTRAST  AGAINST  A  SKY  BACKGROUND
VIEW
no
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
ANGLES
phi
95.0
100.0
105.0
110.0
115.0
120.0
125.0
130.0
135.0
140.0
145.0
150.0
155.0
.1
84.4
148.2
alpha
73.8
68.8
63.8
58.8
53.8
48.8
43.8
38.8
33.8
28.8
23.8
18.8
13.8
168.6
84.4
20.6
DISTANCES (KM)
X
83.0
84.5
86.2
87.9
89.9
92.1
94.8
97.9
101.8
106.9
113.9
124.4
142.2
1.0
80.0
120.0
rp
16.3
16.7
17.4
18.3
19.4
20.8
22.6
24.9
28.1
32.4
38.8
48.6
65.7
79.0
15.7
44.4
ro
34.5
36.3
38.6
41.5
45.3
50.3
57.0
66.3
80.0
101.8
141.4
234.5
701.9
79.5
32.0
156.9
forward
contrast
-.004
-.004
-'.004
-.004
-.004
-.004
-.004
-.003
-.003
-.003
-.003
-.002
-.001
.051
-.004
-.002
                                             -GREEN  PLUME  CONTRAST-
                                                           backward screening
                                                           contrast criterion
When you're ready,  please press [ENTER]  for
more lines of sight (Q to quit)
-.032
-.032
-.032
-.031
-.031
-.031
-.030
-.029
-.028
-.026
-.023
-.018
-.011
-.037
-.033
-.020
.05
.05
.05
.05
.05
.05
.05
.05
.05
.05
.05
.05
.05
.11
.05
.05
     FIGURE 15.  Sample VISCREEN summary for all  lines of sight.
                                35

-------
file is designed for inclusion in a report (e.g.,  a PSD permit applica-
tion) describing the results of the analysis.   It  contains all information
needed for a reviewing agency to duplicate the VISCREEN results, including
emissions, particle characteristics, meteorology,  and geometry.
Obviously, reports prepared by the users of VISCREEN should include the
rationale for selecting these inputs, especially if non-default values are
chosen.  The summary report automatically identifies Level-1 analyses by
their use of default values.  Figure 16 shows  an example of a Level-1 sum-
mary report.

The results file is not designed for inclusion in  reports, but rather to
facilitate the user's preparation of additional graphics displays.  Such
displays can be created by commercialTy available  "spreadsheet" programs
and graphics packages, or by user-developed programs.  For example, this
file can be used to plot plume AE as a function of viewing azimuth.  As
described more fully in Appendix B, the results file includes all user
Inputs, as well as VISCREEN-calculated values  for plume-observer geometry
variables (e.g., downwind distance and plume thickness) and all optical
parameters for each line of sight.  Optical parameters include contrast
values at three wavelengths (red, green, and blue), AE, and the applicable
screening criterion for each combination of line-of-sight, scattering
angle, and viewing background.  The file is formatted with spaces separa-
ting variables, allowing it to be read into commercially available
"spreadsheet" programs.  It also includes an entry showing the number of
lines-of-sight for which results are presented.*  This entry can be read
as an  index limit by programs written in FORTRAN or other languages.
     This  is necessary for scenarios in which the user specifies a
     relatively large observer-source-terrain angle, causing VISCREEN to
     calculate results for fewer than the normal 34 Unes-of-sight.
                                    36

-------
               Visual Effects Screening Analysis for
                Source: Public Electric Coal 13
                Class  I Area: Longvlew NP
 Input  Emissions  for
                       Level-1 Screening
Particulates
NOx (as N02)
Primary N02
Soot
Primary S04
10.00 G /S
120.00 G /S
.00 G /S
.00 G /S
.00 G /S
          Default  Particle  Characteristics Assumed
               Transport  Scenario Specifications:
     Background Ozone:
     Background Visual  Range:
     Source-Observer Distance:
     Min. Source-Class  I  Distance:
     Max. Source-Class  I  Distance:
     Plume-Source-Observer Angle:
     Stability:   6
     Wind Speed:   1.00 m/s
   .04 ppm
110.00 km
 80.00 km
 80.00 km
120.00 km
 11.25 degrees
                            RESULTS

 Asterisks (*) Indicate plume  impacts  that exceed screening criteria

          Maximum Visual Impacts  INSIDE  Class  I Area
             Screening Criteria ARE  Exceeded
                                    Delta E       Contrast
 Backgrnd Theta Azi Distance Alpha  Crit   Plume   Crit  Plume
SKY
SKY
TERRAIN
TERRAIN
10.
140.
10.
140.
84.
84.
84.
84.
80
80
30
80
.0
.0
.0
.a
84.
84.
84.
84.
2
2
2
2
.00
.00
.00
.00
4.
2.
1.
.
743*
369*
495
593
.05
.05
.05
.05
-.004
-.033
.020
.011
          Maximum Visual  Impacts  OUTSIDE  Class  I Area
             Screening Criteria ARE  Exceeded
                                     Delta  E       Contrast
 Backgrnd Theta Azi Distance Alpha  Crit   Plume   Crit  Plume
SKY
SKY
TERRAIN
TERRAIN
10.
140.
10.
140.
35.
35.
5.
5.
63
63
24
24
.5
.5
.9
.9
134.
134.
164.
164.
2
2
2
2
.00
.00
.00
.00
5.
2.
3.
1.
657*
662*
406*
197
amma
.05
.05
.05
.05
sacaixv
-.006
-.048
.043
.040
FIGURE  16.   Sample  Level-1  summary  report.
                              37

-------
                               LEVEL-2 SCREENING
As shown in Figure 1, Level-2 plume visual impact screening is done if the
Level-1 results exceed the screening criteria.  The objective of Level-2
screening is identical to that of Level-1—the estimation of worst-day
plume visual impacts—but in Level-2 screening more realistic (less con-
servative) input, representative of the given source and the Class I area,
is provided.  This situation-specific input may include particle size
distributions for plume and background that are different from those used
in the default Level-1 analysis.  Median background visual range based on
on-site measurements rather than the map shown in Figure 9 might be
used.  However, the most important potential difference in input between
Level-1 and Level-2 analysis centers on meteorology and plume transport
and dispersion patterns.  While the Level-1 analysis assumes F stability,
a 1 m/s wind speed, and a wind direction that would carry plume material
very close to the observer, in the Level-2 analysis, meteorological data
and the topography representative of the source area and the Class I area
may suggest that worst-case plume dispersion conditions are different.
SELECTING PARTICLE SIZE DISTRIBUTIONS

If the Level-1 default parameters are selected, VISCREEN assigns best
estimates of particle size and density for the emitted and background
atmosphere participate (see Table 2).  However, some situations may not
be adequately characterized by the default particle size and density
parameters.  In such cases, Level-2 screening should be carried out with
different parameters.

For example, the Level-1 screening default for background fine particles
assumes a mass median diameter of 0.3 urn; however, in certain humid areas,
the background fine particulate mode may be larger (0.5 urn), and in cer-
tain dry desert areas, such as the southwestern United States, the fine
mode may be smaller (0.2 um).  If the analyst has measurements of
background particle size distributions and densities that are different
from default parameters, these site-specific values should be used and
documented.
                                      39

-------
TABLE 2.  Default particle size and density
specifications.  (Source:  Seigneur et al.,
1983)
Particle Type
Background fine
Background coarse
Plume part icu late
Plume soot
Plume primary sulfate
Mass Median
Diameter (ym)
0.3
6
2
0.1
0.5
Density
(g/cm3)
1.5
2.5
2.5
2
1.5
                     40

-------
Also, if information regarding the size distribution of emitted particu-
late is available, this data should be used to specify emitted particulate
sizes and densities.  In many cases, particulate emission rate estimates
for a source will be calculated from emission factors that do not specifi-
cally identify the expected size distribution.  In such cases the default
primary particle size distribution should be used.  If more detailed
information on actual size distributions is available, appropriate non-
default values should be used in Level-2 analyses.  In general, larger
particles (greater than 10 urn in diameter) have relatively small
effects.  Thus, if both PM-10 and TSP emission rates are available, it
will usually be appropriate to use the PM-10 rate for primary particle
emissions.  However, if the TSP emission rate is substantially higher than
that for PM-10, the large particle effects may be appreciable.  In this
case the TSP rate should be used, along with appropriate size distribution
parameters.

Another alternative exists if there are two distinct processes contribu-
ting to primary particle emissions (e.g., fuel combustion emissions from a
boiler and fugitive dust from materials handling), and if there are no
primary sulfate emissions from the source.  In such cases the primary sul-
fate emission input can be used for one of the processes, with appropriate
modification to particle density and size distribution inputs.  If this
approach is used, the data and rationale for each input to the Level-2
analysis should be thoroughly documented by the analyst, and reviewed with
the permitting agency and Federal Land Manager.
DETERMINING WORST-CASE PLUME DISPERSION CONDITIONS

Probably the most important input specification for Level-2 screening
analysis is for meteorological conditions:  the worst-case wind direction
and speed and atmospheric stability.  Therefore, the joint frequency dis-
tribution of these parameters as measured at or near the location of the
emission source or the Class I area is important input for Level-2 plume
visual impact screening.

It is essential to consider the persistence as well as the frequency of
occurrence of these conditions.  For example, plume discoloration will
generally be most intense during light-wind, stable conditions.  However,
the transport time to a Class I area increases as the wind speed
decreases.  As the transport time approaches 24 hours, it is increasingly
probable that the plume will be broken up by convective mixing and by
changes in wind direction and speed; thus it will not be visible as a
plume or a discolored layer.
                                   41

-------
Ideally,  one would prefer to have a meteorological data base with detailed
spatial  and temporal coverage.  However, this is rarely possible because
of cost  considerations.  Several alternative approaches can be used to
fill in  missing data, but they all involve making assumptions.  For
example,  if a complete meteorological data base is available only at the
site of  the proposed emissions source, one might assume that conditions at
the site are representative of conditions at other locations in the
region.   However, in regions of complex terrain, this assumption may not
be appropriate.  Often, data collected at ground level are assumed to
represent conditions at the effective stack height, which is a poor
assumption when the plume is several hundred meters above ground or the
site is  located in complex terrain.

Any assessment of plume visual impacts is limited by the availability,
representativeness, and quality of meteorological data.  The Level-1
screening analysis discussed in the previous section does not require the
user to  input any meteorological data; rather, conservative assumptions
are made regarding worst-case stability, wind speed, and wind direction.
The Level-2 screening analysis assumes that the analyst has at least one
year of  meteorological data from the site of the proposed emissions
source,  a nearby site within the region, or the Class I area(s) poten-
tially affected by emissions.  For a detailed discussion of the meteoro-
logical  data input requirements, refer to the EPA Guidelines on Air
Quality  Models (Revised) (1986) and Supplement A (1987) [EPA 450/2-78-
027RJ.

The meteorological data base discussed previously should be used to
prepare tables of joint frequency of occurrence of wind speed, wind
direction, and stability class similar to those shown in Figure 17.  These
tables should be stratified by time of day.  If meteorological data are
available at hourly  intervals, it is suggested that these tables be
stratified as follows:  0001-0600, 0601-1200, 1201-1800, and 1801-2400.
If data are available twice daily, morning and afternoon data should be
tabulated separately.  With this stratification, diurnal variation in
winds and stability  is more easily discernible.  If meteorological data
are not available, the assumptions regarding meteorology used in the
Level-1 analysis are used to assess impact.

On the basis of maps showing the source, observer location, and topo-
graphy, the analyst  should select the wind direction sector that would
transport emissions  closest to a given class I area observer point so that
the frequency of occurrence of impact can be assessed as discussed
below.  For example, in the schematic diagram shown in Figure 18, west
winds would transport emissions closest to observer A, whereas either
west-southwest or west winds would transport emissions closest to observer
B.  Observer C would be affected by emissions transported by west-
northwest and northwest winds, but primarily by west-northwest winds.
                                   42

-------
      •MORNING  HOURS ONLY  (0001-0600); OTHER SETS
      OF TABLES  FOR OTHER TIMES OF DAY
Stability Class F
Wind Speed (m/s)






C
0
*->
u
0)
•r*
O
•o
S





N
NNE
NE
ENE
E
ESE
SE
SSE
S
ssw
sw
wsw
w
WNW
NW
NNW
Total
0-1 1-2 2-3 3-4 4-5 5-6 6-7 7-8 8-9 9-10 >10 Total













FIGURE 17.  Joint frequency distribution tables required to  estimate  worst-case
meteorological conditions for plume visual  impact.
                                 43

-------
 EMISSIONS
 SOURCE
FIGURE 18.  Schematic diagram  showing emissions source, observer
locations, and wind direction  sectors.
                          44

-------
For situations influenced by complex terrain, determination of this worst-case wind direction
and its frequency of occurrence is much more difficult.  The analyst should use professional
judgment in this determination.  In such situations, determination of the worst-case wind
direction and its frequency of occurrence should be made on the basis of the following
factors:

       Location(s) for which meteorological data were collected relative to terrain features,
       emissions source, and potentially affected class I areas.

       Likely plume trajectories for each wind direction (and possibly wind speed and
       stability) based on either data or professional judgment. For example, potential
       channeling, convergence, and divergence of flows should be assessed (see Figure 19).

The next step is to construct a table (see the example  in Table 3) that shows worst-case
dispersion conditions ranked in  order of decreasing severity and the frequency of occurrence
of these  conditions associated with the  wind direction  that could transport emissions toward
the class I area. Dispersion conditions  are ranked by evaluating the product o*yazu, where ay
and az are the Pasquill-Gifford horizontal and vertical diffusion coefficients for the given
stability  class and downwind distance x along the stable plume trajectory identified earlier,
and u is  the maximum wind speed for the given wind speed category in the joint frequency
table.  Equations that approximately fit the Pasquill-Gifford curves are presented in Appendix
E.  The method presented in Appendix  E should be used to calculate ay and az. The analysis
should be conducted for the following meteorological  conditions:

             Pasquill-Gifford           Wind
             Stability Class             Speed (m/s)

                    F                 1,2,3
                    E                 1,2,3,4,5
                    D                 1,2,3,4,5,6,7,8

The dispersion conditions are then ranked in ascending order of the value <7yazu.  This is
illustrated in Table 3.  The downwind distance in this  hypothetical case  is assumed to be 100
km.  Note that F,l (stability class  F associated with wind speed class 0-1 m/s) is  the worst
dispersion condition, since it has the smallest value of ayazu (1.89xl05 m3/s).  The second
worst diffusion condition in this example is F,2, followed by F,3, E,l, and so on.

The next column in  Table 3 shows the  transport time  along the minimum trajectory distance
from the emissions source to the Class  I area, based on the midpoint value of wind speed for


                                            45                              Revised 10/92

-------
                            BOUNDARIES OF
                            16 CARDINAL WIND
                            DIRECTION
                            SECTORS
                                                          ELEVATED TERRAIN
                                                          SHOWN IN SHADED
            EMISSIONS SOURCE
FIGURE  19.  Example of map showing  emissions  source,  elevated
areas,  and stable plume  trajectories.
                             46

-------
TABLE 3.  Example table showing worst-case meteorological conditions for plume visual
impact calculations
Dispersion
Condition
(stability,
(TyO^u Transport
Time
wind speed) (m3/s) (hours)

F,l
F,2
F,3
E,l
E,2
E,3
D,l
E,4
E,5
D,2
D,3
D,4

1.89xl05
3.78x10*
5. 66x1 0s
5.67x10*
l.lSxlO6
1.70x10*
1.89xl06
2.27x10*
2.84x10*
3.78x10*
5.68x10*
7.57x10*

56*
19*
11
56*
19*
11
56*
8
6
19*
11
8
Frequency (f) and Cumulative
Frequency (cf) of Occurrence*
of Given Dispersion Condition
Associated with Worst-Case
Wind Direction* for Given
Time of Day (oercent)
0-6
f
0.2
0.2
0.2
0.3
0.4
0.3
0.0
0.6
0.2
0.1
0.3
0.2
cf
0.0
0.0
0.2
0.2
0.2
0.5
0.5
1.1
1.2
1.2
1.5
1.7
6-12
f
0.1
0.1
0.2
0.2
0.3
0.1
0.2
0.3
0.4
0.2
0.1
0.1
cf
0.0
0.0
0.2
0.2
0.2
0.3
0.3
0.6
1.0
1.0
1.1
1.2
12-18
f
0.0
0.0
0.0
0.1
0.0
0.0
0.5
0.1
0.5
0.0
0.4
0.3
cf
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.1
0.6
0.6
1.0
1.3
18-24
f
0.2
0.2
0.2
0.2
0.2
0.1
0.1
0.3
0.2
0.3
0.2
0.1
cf
0.0
0.0
0.2
0.2
0.2
0.3
0.3
0.6
0.8
0.8
1.0
1.1
*       Transport times to Class I area during these conditions are longer than 12 hours, so they are not added
        to the cumulative frequency summation.

t       The joint frequency and cumulative frequency of wind direction, wind speed, and stability are
        determined separately for each of the four time periods ( 0-6, 6-12, 12-18, 18-24). For a given time
        period, e.g. 0-6, the sum of all frequencies for all dispersion conditions adds up to 100 percent.

$       For a given Class I area.

Note: Distance downwind, values of cy, CT,, and transport times are based on a distance of 100 km.
                                                47
Revised 10/92

-------
the given wind speed category. For example, for the wind speed category, 0-1 m/s, a wind
speed of 0.5 m/s should be used to evaluate transport time; for 1-2 m/s, 1.5 m/s; and so on.
The times necessary for a plume parcel to be transported 100 km are 56, 19, 11, 8, and 6
hours for wind speeds of 0.5, 1.5, 2.5, 3.5, and 4.5 m/s, respectively.

For the Level-2 screening analysis, we assume it is unlikely that steady-state plume conditions
will persist for more than 12 hours. Thus, if a transit time of more than 12 hours is required
to transport a plume parcel from the emissions source to a Class I area for a given dispersion
condition, we assume that plume material is more dispersed than a standard Gaussian plume
model would predict. This enhanced dilution would result from  daytime convective mixing
and wind direction and speed changes.

To obtain the worst-case meteorological conditions, it is necessary  to determine the dispersion
condition (a given wind speed and stability class associated with the wind direction that
would transport emissions toward the  Class I area) that has a cyyorzu product with a cumulative
probability of 1 percent.  In other words, the dispersion condition is selected such that  the
sum of all  frequencies of occurrence of conditions worse than this  condition totals 1 percent
(i.e., about four days per year). The 1-percentile meteorology is assumed  to be indicative of
worst-day plume visual impacts when the probability of worst-case meteorological conditions
is coupled  with the probability of other factors being ideal for maximizing plume visual
impacts.  Dispersion conditions associated with transport times of more than 12 hours are  not
considered in this cumulative frequency for the reasons stated above.

This process is illustrated by the example shown in Table 3, which indicates that the first  two
dispersion  conditions would cause maximum plume visual impacts because the ayazu products
are lowest  for these  three conditions.  However, the transport time  from the emissions source
to the Class I area associated with each of these dispersion conditions is greater than 12
hours. With the third dispersion condition (F,3), emissions could be transported in less than
12 hours.   The frequency of occurrence  (f) of this condition is added to the cumulative
frequency summation (cf). For this hypothetical example, the meteorological data are
stratified into four time-of-day categories.  The joint frequency distributions of wind direction,
wind speed and stability  are determined separately for each of the four time periods.  Each
time period's frequency distribution is calculated such that the sum of the  frequencies for  all
dispersion  conditions adds up  to 100 percent.  For each time period, the one percentile
meteorology would be determined, solely on the cumulative frequencies for that time period.
Then, the most restrictive of the one-percentile dispersion conditions determined for the 4
time periods would be used as a basis for the Level n  analysis.  The rationale for stratifying
the joint frequencies in this way is to provide conservatism in the calculation and also  to
provide information on the time of day that worst-case plume visual impacts are likely to
occur. By determining  worst-case dispersion in this way, one knows the dispersion conditions
                                           48                              Revised 10/92

-------
for each time period that would be expected to be worse one percent of the hours during that
time-of-day period.

Note that the worst-case, stable, light-wind dispersion conditions occur more frequently during
the nighttime hours.* In our example, the following additional worst-case dispersion
conditions add to the cumulative frequency: F,3; E,3; E,4; E,5; D,3; and D,4.  Dispersion
conditions with wind speeds less than or equal to 2 m/s (F,l; F,2; E,l; E,2; D,l; and D,2)
were not considered to cause an impact because of the long transit times to the Class I area in
this example.  Thus, their frequencies of occurrence were not added to the cumulative
frequency summation.  The result of this example analysis is that dispersion condition E,4 is
associated with a cumulative  frequency greater than or equal to 1 percent and the most
restrictive,  so we would use this dispersion condition to evaluate worst-case visual impacts for
the Level-2 screening analysis for this  example case.

It should also be noted that if the location of the observer in the Class I area is at or near the
boundary of one of the 16 cardinal wind direction sectors, it may be appropriate to interpolate
the joint frequencies of wind speed, wind direction, and stability class from the two wind
direction sectors, on the basis of the  azimuth orientation of the observer relative to the  center
of the wind direction sectors.
ACCOUNTING FOR COMPLEX TERRAIN

If the observer is located on elevated terrain or if elevated terrain is between the emissions
source and the observer, dispersion  patterns may be significantly different from those obtained
from the procedures outlined above. For such situations, adjustments to the worst-case
meteorological conditions determined by these procedures may be necessary.

For example, consider the elevated  terrain feature illustrated by the shaded area in Figure  19.
It is unlikely that a stable plume parcel would remain intact after transport to either Observer
A or B.  Either the stable plume would be transported around the elevated terrain feature,
resulting in a longer plume transport distance, or the plume would be broken up by turbulence
       Although plume visual impact is usually not an issue at night, nighttime dispersion
       conditions need to be considered because maximum plume visual impacts are often
       observed in the morning after a period of nighttime transport.  For these situations, the
       nighttime meteorological conditions are most indicative of plume dispersion when the
       plume is viewed at sunrise.  In cooler seasons, stable stagnant conditions may persist
       during daytime hours also.

                                           49                              Revised 10/92

-------
encountered during the straight-line transport up and over the terrain feature.  Also, stable
plume transport in the direction of Observer C would be blocked by elevated terrain.  On the
other hand, Observer D would be in a position where straight-line stable transport is not only
possible but very likely in the drainage flow off the elevated terrain feature.

Accounting for elevated terrain can be a detailed and time-consuming process, requiring
complex-terrain windfield models and other sophisticated tools.  Although such analytical
options are encouraged, we suggest a simpler  screening approach based on assumed
enhancements to dispersion caused by elevated terrain.

If the observer is located on terrain at least 500 meters above the effective stack height for
stable conditions (Observer C in Figure 19) or such elevated terrain separates the emission
source and the observer (Observers A and B in Figure 19), the  worst-case stability class
should be shifted one category less stable.
EXERCISING VISCREEN

The plume visual impact screening model VISCREEN can be run as described previously for
the Level-1 analysis.  However, for Level-2 analysis, the default parameters are not selected.
The analyst selects panicle size distribution and density parameters  suitable for the source and
region in question (although default particle sizes and densities can  still be used if desired).
Meteorological conditions (stability, wind speed, and plume offset angle) appropriate for the
worst-case analysis are used. If available, visual range and ambient ozone data from locations
near the source area and Class I area can be used instead of Level-1 default values. Median
values of both should be used, if available.
ALTERNATIVE USE OF PLUME VISIBILITY MODELS

As an alternative to the  use of the screening model VISCREEN, the analyst may wish to
apply plume visibility models [refer to EPA Guideline on Air Quality Models (Revised) EPA
450/2-78-027R,  Supplement A, and any future supplements]. Although model input
requirements are more extensive for these more sophisticated models, the models  are expected
to be more realistic (less conservative) than VISCREEN.  Several alternative plume and sun
positions should be tested to assure that realistic worst-case scattering angles are analyzed
(VISCREEN analyses only worst-case scattering angles).
                                           50                             Revised 10/92

-------
                               LEVEL-3 ANALYSIS
In Level-3 analysis, the objective is broadened from conservative analysis
of worst-case conditions to a realistic analysis of all conditions that
would be expected to occur in a typical year in the region that includes
both the emission source and the observer.  Level-3 analysis is no longer
considered screening because it is a comprehensive analysis of the magni-
tude and frequency of occurrence of plume visual impacts as observed at a
sensitive Class I area vista.

It is important to determine the frequency of occurrence of visual impact
because the adversity or significance of impact is dependent on how fre-
quently an impact of a given magnitude occurs.  For example, if a plume is
perceptible from a Class I area a third of the time, the impact would be
considered much more significant than if it were perceptible only one day
per year.  The assessment of frequency of occurrence of impact should be
an integral part of Level-3 visual impact analysis.
OBJECTIVES OF LEVEL-3 ANALYSIS

In this section we discuss how one can determine both the magnitude and
frequency of occurrence of plume visual impact.  This procedure entails
making several runs with a plume visibility model for different values of
the following important input parameters that are likely to vary over the
course of a typical year:

     Emission rates (if variable)
     Wind speed
     Wind direction
     Atmospheric stability
     Mixing depth
     Background ozone concentration
     Background visual range
     Time of day and season
     Orientation of observer, plume, and sun

-------
       Viewing background (whether it is sky, cloud, or snow-covered, sunlit, or shaded
       terrain).

Because of the large number of variables important to a visual impact calculation, several
model calculations are needed to assess the magnitude and frequency of occurrence of visual
impact.  It would be ideal to calculate hourly impacts over the course of a year or more using
hourly  values of the above variables. However, such an extensive data base is rarely
available for use. Even if it were available, the computing costs involved would be
prohibitive.  It is therefore preferable to select a few representative, discrete values for each
of these variables to represent the range (i.e., the magnitude and frequency of occurrence) of
visual impact over a given period of time, such  as a season or year.

It is possible to imagine a worst-case impact condition that would never occur in the real
atmosphere; this condition could be  represented on a cumulative frequency plot, such as that
of Figure 20, as  point A. The impact is great, but it almost never occurs.   If another worse-
case situation less extreme than point A were  selected, the magnitude of impact would be
less, but it might occur with some nonzero frequency, about one day per year, for example
(the reasonable worst-case impacts for Level-1 and Level-2 analyses). It is possible to select
various values of all the important input variables and to assess the frequency with which
those conditions resulting in impacts worse than a given impact would occur.  By this
process, several points necessary to  specify the frequency distribution could be obtained (for
example, points B, C, and D in Figure 20).  With average (50-percentile) conditions, a
negligible impact, as shown at point E in Figure 20, might be found.  In Figure 20, the
ordinate could be any of the parameters used to characterize visibility impairment,  such  as
visual range reduction, plume contrast, blue-red ratio, or AE, and the abscissa could represent
cumulative frequency over a season or a year.

In a visual impact assessment, it is recommended that one select various combinations of
upper-air wind speed, wind direction, and atmospheric stability; background ozone
concentration; and background visual range to specify the frequency distribution of plume
visual-impact as shown in Figure 20.  If one has a large, concurrent data base of all five of
these variables, it would be desirable to calculate a five-way joint-probability distribution
matrix and to  use these joint probabilities to calculate frequency of occurrence of impact.
However, in most situations, such a data base is not available, and one must treat the various
worst-case events as independent probabilities.  With this assumption, the probability of
worst-case impacts can be roughly estimated by multiplying the  independent probabilities.
This can be represented as follows:
                                           52                             Revised 10/92

-------
,„
            25          50           75
      Cumulative  Frequency of  Occurrence  (%)
100
 FIGURE  20.   Example of a frequency distribution of
 plume visual  impact.

-------
                                          V)
                                            (10)
                                 1
where f(y > y1) Is the cumulative frequency of impact y greater than y1,
and f(x^ > x.j') is the cumulative frequency of variable x^ having values
that would cause greater impact than the value x^'.

In such an application, one might obtain an estimate of cumulative fre-
quency by using the joint frequency distribution of upper-air wind speed
and wind direction and the separate frequency distributions of upper-air
stability and other parameters critical to plume visual impact.  For
example, a cumulative frequency distribution of the plume perceptibility
parameter A£ can be estimated as follows:
           f(AE > AE1) = f(u < u1, WD < WD1) •  f(s > s')
                                            (11)
where
                           f (other factors)
            f(AE
    f(u < u1, WD < WD1)
              f(s > s1)
the frequency of occurrence of AE values
greater than AE1.  AE1  is calculated on the
basis of a wind speed u1, wind direction WD1,
stability s1, ozone concentration [O^]', and
visual range ry'.

the frequency of occurrence of wind speeds
less than u1 for wind directions within a
specified value (WD1) of the worst-case wind
direction.

the frequency of occurrence of stabilities
greater than s'.
       f(other factors) = the frequency of occurrence of background
                          ozone concentrations greater than  [03]'  (that
                          would cause higher plume NC^ concentrations),
                          background visual range values greater than  rv',
                          and plume dimensions (oy, CTZ) worse than assumed
                          values  (Pasquill-Gifford).

 Note  that  this equation assumes the statistical independence of winds,
 stability, and other factors.  If enough data are available, joint
 frequency  distributions should be used.  This is especially  important  if
 there are  known  conditions  that contradict the assumption of independence
 (e.g., terrain-induced stable drainage that flows).  Each of the input
                                  54

-------
parameters that are important to the visibility model calculation varies
significantly over the period of a year.
SUGGESTIONS FOR LEVEL-3 ANALYSIS

The most exacting way to obtain plume visual impact cumulative frequency
distributions would be to apply a plume visibility model for every time
period (e.g., every daylight hour or 3-hour period) with the appropriate
emissions., wind speed, wind direction, stability, background ozone, back-
ground visual range, sun angle, and viewing background.  Thus, one would
have a calculation for every daytime period in the course of a year.   If
done every 3 hours, this would be approximately 1460 model applications
(365 days/yr X 12 hr/day of daylight/3 hr « 1460 time periods).  Such  a
method is not practical with current plume visual impact analysis hardware
and software.

Thus, the analyst needs to estimate the plume visual impact cumulative
frequency distribution using a limited set of plume visibility model runs
and appropriate assumptions.  There is no simple procedure that can be
recommended for all Level-3 analyses.  Limited comparisons of Level-3  pre-
dictions with measurements suggest that magnitudes and frequencies of
plume visual impact are reasonably well estimated by the following sug-
gested procedures.  It is recommended, however, that any chosen procedures
for performing a given Level-3 analysis be reviewed by the permitting
authority and the Federal Land Manager of the affected Class I area before
analysis commences.
Frequency Distribution of Dispersion Conditions

A joint frequency distribution of wind speed, wind direction, and  sta-
bility should be prepared separately for the following times of day:  mid-
night to 0600, 0600 to noon, noon to 1800, and 1800 to midnight.   This
breakdown is necessary to identify the time of day of impacts.  These dis-
tributions should be compiled for the entire year (or if possible, two or
more years) and for each of the four seasons.  Seasonal analysis of plume
visual impact may be important for the Federal Land Manager and state to
assess the number of visitors potentially impacted by a given plume.  If
worst-case plume visual impacts occur under stable transport conditions,
they will most likely occur during the early morning hours.  In such
cases, it is recommended that the midnight to 0600 frequency distributions
be given the primary attention in Level-3 analysis.  However, for  com-
pleteness, the 0600 to noon and noon to 1800 distributions should  be used
to characterize the frequency of midday and afternoon plume visual
impacts.
                                  55

-------
Calculating Plume Visual Impacts

Plume visual impacts should be calculated for a representative sample (or possibly each) of
the categories of stability, wind speed, and wind direction in the joint frequency distribution.
Since the objective is to estimate the cumulative frequency curve (similar to that shown in
Figure 20), plume visual impact should be calculated for the most distant plume position
(from the observer) within the given wind direction and the highest wind speed appropriate
for a given category of the distribution.  For example, for the frequency distribution cell
representing F, 0-1 m/s, plume calculations  should be made for 1 m/s, not a lower value, and
for the most distant plume position (11.25°  offset is recommended for the worst-case wind
direction sector).  This  approach is necessary because the abscissa of the cumulative
frequency plot is the frequency of conditions that produce impacts larger than the ordinate
value of plume visual impact magnitude (AE).  Plume visual impact should be calculated for
a number of the cells of the frequency distribution (perhaps 20 or more).  The largest impact
magnitudes are likely to occur for wind directions that would carry the plume closest  to the
observer, light wind speeds, and stable  conditions. To fill in conditions causing lower magni-
tudes (but higher cumulative frequencies), the analyst should identify a sample of wind
directions, wind speeds, and stabilities that represent typical conditions.   For example, all the
72 combinations of 8 plume positions or wind directions (e.g., worst case and three adjacent
22.5° sectors to the left and right, representing plume offset angles of 11.25, 33.75, 56.25, and
78.75°, 3 wind speeds (e.g., 0-2, 2-5, and 5-10 m/s), and 3 stabilities (e.g., F, E, and D) could
be used as the input for 72 plume visibility model runs.  These runs would be made using
median background ozone concentration and visual range values. Sun angles would be speci-
fied  by the date and time of the simulation.  The worst-case sun angles should be determined
by sensitivity analysis for one of the worst-case combinations of meteorological conditions
before the full complement of model runs (72 in our example above) is made.  Since worst-
case meteorological conditions generally occur in the morning, it is suggested that simulation
date/times of an hour after sunrise and  an hour before sunset on 21 March, 21 June, 21
September, and 21 December be analyzed in the sensitivity test, and the  worst-case date/time
be used for all subsequent model runs.  Model runs should be made for the appropriate
viewing backgrounds for each line of sight  and  each plume position.  If terrain is found to be
the plume's viewing background, the appropriate distance between the observer and the
terrain feature should be provided as part of the model input.
                                           56                              Revised 10/92

-------
Coupling Magnitude and Frequency

Each of the (for example, 72) model calculations should be evaluated to
select the two maximum plume AE's for conditions when the plume parcel is
inside and outside the Class I area's boundary, respectively.  (If discus-
sions with the Federal Land Manager of the given Class I area suggest that
only within-area plume parcels are of concern, only the former AE need be
compiled.)  The inside and outside &E's separately should be put in
descending order of magnitude and coupled with the corresponding frequency
of dispersion conditions.  Cumulative frequencies should be added by sum-
ming the individual frequencies (see lable 3).  If a wind direction, sta-
bility, or wind speed class was skipped in the sampling of the cells in
the frequency distribution, the frequencies for all conditions expected to
cause greater plume visual impact should be added and coupled with the
given plume visual impact AE.  Separate magnitude/frequency tables should
be compiled for inside/outside views, each time of day, and each season.
Interpreting the Cumulative Frequency Curve

Cumulative frequency distribution curves of plume visual impacts prepared
using the procedures described in the preceding paragraphs should be
interpreted in light of the assumptions and simplifications underlying the
various steps.  Several factors that can be particularly significant
include the use of median values for visual range and background ozone
concentration; the persistence of stable conditions for long transport
distances; and the use of Pasquill-Gifford coefficients as the sole
determinant of plume dispersion.  For specific cases, the combined effect
of such assumptions can be that estimated frequencies of a specific level
of effects (say, A£ greater than 5) may be higher or lower than would
actually occur.

Cumulative frequency curves based solely on the joint frequency of wind
speed, wind direction, and atmospheric stability ignore the probability of
occurrence of other factors that affect plume visual impacts.  This proba-
bility appears as "f(other factors)" in Equation 11.  In our experience,
wind speed, direction and stability are the principal determinants of
plume visual impacts.  In some cases, however, these "other factors" could
be significant.  Obviously, if data and resources allow, analyses can be
expanded to incorporate joint frequency distributions for all key para-
meters.  However, the number of model simulations required will increase
geometrically with the addition of each new dimension.  For example,
treating three visual ranges (e.g., 50th, 75th, and 90th percent!les)
triples the number of simulations.  Further, the data required to develop
such joint frequency distributions are not available for many areas.
                                  57

-------
No explicit formal guidance can be provided at this time for interpreting cumulative
frequency curves.  The analyst should, however, identify which transport scenarios have both
high visual effects and high frequencies of occurrence.  Similarly, the analyst should verify
that the transport scenarios modeled include those under which visual impacts will be
greatest.  If it is likely that simplifying assumptions may have led to bias in the cumulative
frequency curves, then the factors leading to this  conclusion should be described for
consideration by the permitting agency, the Federal Land Manager, and other reviewers.

Summarizing Results

Cumulative frequency plots  similar to Figure 20 should be made for each season, time of day,
and inside/outside combination.  In addition, the number of mornings and afternoons in each
season that AE's are greater than 2 should be tabulated.
RECOMMENDED MODEL FOR LEVEL-3 ANALYSIS

Plume Visibility Model (PLUVUE ID

The recommended model for a Level-3 analysis is the PLUVUE II model (EPA, 1986). The
PLUVUE II (Seigneur et al., 1984) model uses a Gaussian formulation for transport and
dispersion.  The spectral radiance I(A,) at 39 visible wavelengths (0.36 < X < 0.75 urn) is
calculated for views with and without the plume; the changes in the spectrum are used to
calculate various parameters that predict the perceptibility of the plume and contrast reduction
caused by the plume.  PLUVUE II is designed to perform plume optics calculations in two
modes.  In  the plume-based mode, the visual effects are calculated for a variety of lines of
sight and observer locations relative to the plume parcel; in the observer-based mode, the
observer position is fixed and visual effects are calculated for the specific geometry defined
by the position of the observer, plume, and sun.  For either mode, the model requires the user
to select up to 16 different locations downwind of the emission source.  These distances
determine the locations of the optics calculations along the plume trajectory. For further
information regarding the application of the PLUVUE II model, the updated, abridged version
of the PLUVUE II User's Guide (EPA, 1992) should be reviewed.
Optional Use of VISCREEN

As a low-cost, easy-to-apply, but more conservative estimate of plume visual impact, the
analyst may wish to use VISCREEN as the model for generating plume visual impact
magnitudes in the Level-3 analysis. VISCREEN could be used either in place of, or in
addition to, a plume visibility model.  VISCREEN can also  be used to choose meteorological
scenarios to be further analyzed with a plume visibility model.

                                          58                             Revised 10/92

-------
                                    REFERENCES
Blackwell, H. R.  1946:  Contrast thresholds of the human eye. /. Optical Society of
       America, 36:624643.

Booker, R. L., and C. A. Douglas.  1977:  Visual Range Concepts in Instrumental
       Determination and Aviation  Application. NBS monograph 159, U.S. Department of
       Commerce.

Cornsweet, T.  1970: Visual Perception. Academic Press, New York.

EPA 1986: Guideline on Air Quality Models (Revised) and Supplement A. EPA-450/2-78-
       027R.  Research Triangle Park, NC.

EPA  1992:  Updated User's Guide for the  Plume Visibility Model (PLUVUE II). Research
       Triangle Park, NC.

Faugeras, O. D.  1979:  Digital color image processing within the framework of a human
       visual model. IEEE Trans. Acoust., Speech Sig. Process, Vol. ASSP-27, pp.  380-393.

Gordon, J. I.  1979: Daytime Visibility:  A Conceptual Review.  Scripps Institution  of
       Oceanography, Visibility Laboratory, La Jolla, California.

Hall, C. F., and E. L. Hall.  1977:  A nonlinear model for the spatial characteristics of the
       human visual system. IEEE Trans.  Syst., Man, Cybern., SMC-7, pp. 161-170.

Helson, H.  1938:  Fundamental principles of color vision. I. The principle governing changes
       in hue, saturation, and lightness of non-selective samples in chromatic illumination.  /.
       Exp. PsychoL, 23:439-471.

Henry, R. C.  1979: The Human Observer and Visibility—Modern Psychophysics Applied to
       Visibility Degradation.  View on Visibility—Regulatory and Scientific.  Air Pollution
       Control Association, Pittsburgh, Pennsylvania.

Henry, R. C., and J. F. Collins.  1982: Visibility Indices: A Critical Review and New
       Directions. Prepared for Western Energy Supply and Transmission Associates.
       Environmental Research & Technology, Inc., Westlake Village, California (ERT
       Document P-A771).

                                          59                             Revised 10/92

-------
Howell, E. R., and R. F. Hess.  1978:  The functional area for summation to threshold for
       sinusoidal gratings.  Vision Res., 18:369-374.

Jaeckel, S. M. 1973: Utility of color-difference formulas for match acceptability decisions.
       Appl. Optics, 12:1299-1316.

Johnson, R. W. 1981:  Daytime visibility and nephelometer measurements related to its
       determination. Atmos. Environ., 16:1835-1846.

Judd, D.B.  1940:  Hue saturation and lightness of surface colors with chromatic illumination.
       /. Opt. Soc. Am., 30:2-32.

Judd, D. B., and G. Wyszecki.  1975:  Color in Business, Science, and Industry.  John Wiley
       & Sons, New York.

Koenig, A., and E. Brodhun.  1888, 1889:  Experimentelle  Untersuchungen uber die
       psychophysische Fundamentalformel in Bezug auf den Gesichtssinn.  Sitzungssberichte
       preussischen Akademie der Wissenschaften, 26:917-931; 27:641-644.

Larimer, D. A., and R. G. Ireson.  1980: Workbook for Estimating Visibility Impairment.
       U.S. Environmental Protection Agency, Research Triangle Park, North Carolina (EPA-
       450/4-80-031).

Larimer, D. A., R.  W. Bergstrom, S. R. Hayes, M. K. Liu,  J. H. Seinfeld, G. Z.  Whitten, M.
       A. Wojcik,  and  M. J. Hillyer.  1978:  The Development of Mathematical Models for
       the Prediction of Anthropogenic Visibility Impairment.  U.S. Environmental Protection
       Agency, Research Triangle Park, North Carolina (EPA-4503-78-110a,b,c).

Loomis, R. J., M. J. Kiphart, D. B. Garnand, W. C. Malm, and J. V. Molenar.   1985:  Human
       Perception of Visibility Impairment.  Paper presented at the Annual Meeting of the
       Air Pollution Control Association, Detroit, Michigan, June  1985.

Lowry, E. M.  1931: The photometric sensibility of the  eye  and the precision of photometric
       observations.  /. Optical Society of America, 21:132.

Lowry, E. M.  1951: The luminance discrimination of the human eye. Journal  of the Society
       of Motion Picture and Television Engineers, 57:87.
                                          60                             Revised 10/92

-------
Malm, W., M. Kleine, and K. Kelley.  1980:  Human Perception of Visual Air Quality
      (Layered Haze).  Paper presented at the  1980 Conference on Visibility at the Grand
      Canyon.

Malm, W. C, D. M. Ross, R. Loomis, J. Molenar, and H. Iyer.  1986: An Examination of
      the Ability of Various Physical Indicators to Predict Perception Thresholds of Plumes
      as a Function of Their Size and Shape.  Paper presented  at the Air Pollution Control
      Association International Specialty Conference  on Visibility, Grand Teton National
      Park, September 7-10, 1986.

Mathai,  C. V., and I.  H.  Tombach.  1985: Assessment of the technical basis regarding
      regional haze and visibility impairment.  AeroVironment Inc., Monrovia, California
      (AV-FR-84/520).

Middleton, W.E.K. 1952:  Vision Through the Atmosphere.  University of Toronto Press,
      Toronto, Canada.

Optical Society of America, Committee on Colorimetry.  1963:  The Science of Color.
      Optical Society of America, Washington, D.C.

Seigneur, C. et al.  1984: User's Manual for the Plume  Visibility  Model (PLUVUE II).
      Systems Applications, Inc., San Rafael, California.  NTIS PB84-158302.

Stevens, R. K., T. G.  Dzubay, C. W. Lewis, and R. W. Shaw.  1984:  Source apportionment
      methods applied to the determination of the origin of ambient aerosols that affect
      visibility in forested areas. Atmos. Environ., 18:261-272.

Systems Applications, Inc.  1985: Modeling Regional Haze in the Southwest:   A Preliminary
      Assessment of Source Contribution.  Systems Applications, Inc., San Rafael,
      California (SYSAPP-85/038).

Tombach, I., and D. Allard.  1980:  Intercomparison of visibility measurement methods. /.
      Air Pollut. Control Assoc., 30:134-142.

Tombach, I., and D. Allard.  1983:  Comparison of Visibility Measurement Techniques:
      Eastern United States. Electric Rower Research Institute, Palo Alto, California (EPRI
      EA-3292).
                                          61                             Revised 10/92

-------
                    Appendix  A

PERCEPTIBILITY THRESHOLDS AND RECOMMENDED SCREENING
   ANALYSIS  CRITERIA  FOR  PLUMES  AND  HAZE  LAYERS

-------
                                Appendix A

            PERCEPTIBILITY  THRESHOLDS AND  RECOMMENDED  SCREENING
               ANALYSIS CRITERIA FOR PLUMES AND HAZE LAYERS
INTRODUCTION

The plume from an emissions source is visible to an observer if the con-
stituents of the plume, such as particulates and nitrogen dioxide, scatter
or absorb enough light out of or into the observer's line of sight so that
the plume contrasts with its viewing background.  If this plume contrast
is sufficiently large (either positively, signifying a bright plume, or
negatively, signifying a dark plume), the plume becomes perceptible.
Thus, the objective of plume visibility impact analysis is first to deter-
mine the plume contrast and second to determine whether that contrast will
be perceptible.  Some plumes are not visible because the concentration of
optically active species 1n the plume (i.e., those that scatter or absorb
light) is low.  In addition, other factors, such as the position of the
plume relative to the observer and the nature of the haze through which
the plume is viewed, can affect plume visibility.

The objective of this appendix is to review the literature regarding per-
ceptibility thresholds in order to recommend criteria for use in plume
visibility impact screening and analysis.  A perceptibility threshold
would be a suitable criterion for visibility impact analysis if the policy
objective were to be very strict (i.e., to prevent any visible plumes in a
given location).  A perceptibility threshold also may help to define the
lower bound for less strict criteria (that would prevent significant plume
impacts but allow a few days of marginally perceptible plumes).
PERCEPTIBILITY PARAMETERS

Contrast is the parameter most commonly used in the published literature
to describe the sensitivity of the human eye-brain system.  Contrast is
also the most easily calculated plume visibility parameter, since it can
be based on a single wavelength of light and does not require calculations
at other wavelengths in the visible spectrum as do more sophisticated
parameters.
                                   A-l

-------
Contrast is the relative difference in light intensity (radiance) of two
viewed objects and can be calculated as follows:

                                    !   *  1
where Ij and 1% are the light intensities at a given wavelength for
objects 1 and 2 (e.g., a plume and its viewing background).

Another parameter commonly used in plume visual impact analysis is AE.  AE
is perhaps the best currently available plume perceptibility parameter
because it is based on the human eye/brain system's relative sensitivity
to all wavelengths in the visible spectrum.  It is proportional to the
perceptibility of color differences and is essentially identical to just
noticeable differences (jnd).  A AE value of 1 is commonly taken to be
1 jnd.
OVERVIEW OF THRESHOLD RESEARCH

The issue of defining the conditions under which a plume or haze layer
will be visible is part of the scientific field known as psychophysics.
Psychophysics is the branch of psychology that is concerned with subjec-
tive measurement.  It relates physical stimuli to psychological response.
In the current context, we are interested in the effect that differences
in radiant energy (light) directed toward a human observer (the physical
stimulus) have on the psychological response of the eye/brain system of
the observer.

One of the oldest psychophysical determinations is the minimum physical
stimulus increment that the observer can just barely perceive.  This
increment is called the just noticeable difference, differential thres-
hold, or difference limen.  A just noticeable contrast is also called a
litninal contrast, and contrasts greater or less than this contrast are
called supraliminal and subliminal.  The difference limen is never a
sharply defined value.  Since an observer's sensitivity and attention vary
from moment to moment, it is common to define the limen as a statistical
measure.  For example, tne limen mignt be defined as that stimulus that
could be distinguished 50, 70, or 90 percent of the time.  The difference
limen is often set at a value of 70 percent probability of distinguishing
two stimuli.  One of the oldest laws of psychophysics is Weber's Law,
which states that the difference lirnen is a constant fraction of the
stimulus.  Since contrast is defined as the relative difference in light
intensities of two objects and is itself a ratio, Weber's Law could be
                                     A-2"

-------
stated as follows:   the liminal contrast is a constant (regardless of the
light intensity).   Later research has shown that Weber's Law is only
approximate.

The famous Koschmieder equation that inversely relates the visual range to
the light extinction coefficient was based on the assumption that the
liminal contrast for relatively large, sharp-edged objects observed in
daylight is 0.02,  or 2 percent.  Although there were no scientific data to
support this assumption made by Koschmieder in 1924, this contrast value
is widely used in  visibility work for uniformity in discussion.

One of the largest research studies of liminal contrast was carried out by
Blackwell (1946).   In this study, 19 young female observers made more than
2,000,000 observations, of which 450,000 were suitable for statistical
analysis (Middleton, 1952).  Circular stimuli of various sizes ranging
from 0.6 to 360 minutes of arc were presented to the observers.  These
studies indicated  that for a typical daytime luminance of 100 candle/m ,
the liminal, or threshold, contrast ranges from a low of 0.003 (or 0.3
percent) for stimuli subtending 121 minutes of arc (2°) viewed for unlimi-
ted times to contrasts as high as 0.02 (or 2 percent) for stimuli subtend-
ing 10 minutes of arc viewed for limited periods.  Thus, Blackwell's data
suggest that the human observer is more sensitive than Koschmieder assumed
him to be, at least under laboratory conditions.

The data of Koenig and Brodhun (1888, 1889) suggest that for typical day-
light  luminances,  the liminal contrast is independent of the wavelength of
light  over the range tested (0.43 to 0.67 ym) and is on the order of 0.01,
or 1 percent.  For luminances greater than about 100 candles/m , Lowry
(1931, 1951) reported a liminal contrast of 0.014, or 1.4 percent.

Recent psychophysical research (Cornsweet, 1970; Hall and Hall,  1977;
Faugeras, 1979; Howell and Hess, 1978; Malm et al., 1986) has documented
the fact that the response of the human eye/brain system to brightness
contrast is a strong function of the spatial frequency of the contrast.
Spatial frequency is defined as the reciprocal of the distance between
sine-wave crests (or troughs) measured in degrees of angular subtense of a
sine-wave grating.  Thus, spatial frequency has units of cycles/degree
(cpd).  Any pattern of  light intensities, whether it  is a sine-wave,
square-wave, step-function or  any other pattern, can be resolved by
Fourier analysis into a sum of sine-wave curves of different magnitude and
frequency.  To a first  approximation, the spatial frequencies  (f) corre-
sponding to a Gaussian plume of width  (w) are within the order of magni-
tude  centered on f =  1/w.  The human eye/brain system is most sensitive
to spatial  frequencies of approximately 3 cycles/degree (cpd).   Thus, we
might  expect that plumes of width 0.33° (inverse of 3 cpd) to be the most
easily perceptible.   Figure A-l  summarizes the research of Howell and Hess
                                    A-3

-------
                    1000
                 •$  100
                 s
                 o
                 o
                      10
Sharp edges (square wave)

HoweII and Hess (1978)
                                            Diffuse edges  (sine wave)
                                                              _L
                      0.01
              0.1
     1                   10



Spatial Frequency (cycles/degree)
100
                                                                                                                          0.001
                                                                                               0.01
                                                                                               0.1
                       FIGURE A-l.   Contrast  sensitivity curves as  a function of spatial frequency.
87195

-------
(1978).   The sensitivity of the human eye-brain system drops off signifi-
cantly at high spatial frequency (due to visual acuity) and also to a less
extent at low spatial  frequency (i.e., broad, diffus objects).  The human
visual system is much  more sensitive to images with sharp, distinct edges
(e.g., square waves) than to images with diffuse, indistinct edges (e.g.,
sine waves or Gaussian plumes).  At 3 cpd the human visual system has a
sensitivity* to square waves of 300 (corresponding to a threshold contrast
of 2/300, or 0.0066) and to sine waves of 230 (contrast of 0.0086).  Thus,
at this most sensitive frequency, the eye/brain system is 1.3 times more
sensitive to square waves than to sine waves.  At lower spatial frequen-
cies, the difference in sensitivities increases significantly (see Figure
A-l).  The fall-off in sensitivity at high spatial frequencies  is con-
sistent with the data of Blackwell (1946) and with the known responses of
lens systems such as the human eye (Cornsweet, 1970).

To this point, we have discussed perception threshold research  that is
based on the use of contrast as the quantitative parameter.  Before pro-
ceeding to the research of Malm and co-workers that specifically addresses
the perception thresholds of plumes and haze layers, we discuss the limi-
ted work that has been performed using the AE parameter.  The AE parameter
is designed to be proportional to the perceptibility of differences in
brightness and color (Judd and Wyszecki, 1975).  It is generally accepted
that a AE of 1 corresponds to 1 just-noticeable difference (jnd).  Thus, a
A£ of 1 is roughly the liminal, or threshold, color difference.  Applying
the AE formulas to a 2 percent contrast, Latimer et al. (1978)  calculated
a AE of 0.78 and concluded that "AE's less than 1 would be imper-
ceptible."  Jaeckel (1973) presented data on the probability that obser-
vers would accept given color differences as a match.  He found that
approximately 30 percent of observers could distinguish a A£ of about 1,
50 percent could distinguish a AE of 2, and more than 90 percent could
distinguish a AE of 4.

Essentially all of the work specifically addressed to the perception
threshold of plumes and layered haze has been carried out by Malm and co-
workers at Colorado State University.  Their laboratory studies were based
on actual or computer-generated color slides of plumes and layered haze.
* Howell and Hess (1978) define sensitivity as the inverse of modulation
  contrast which is (Ij - I2)/(Ii + 13)•  This definition of contrast  is
  approximately half the contrast defined earlier (1^ - \2)/l2-  Thus, we
  multiply modulation contrast by two to obtain contrasts used for visi-
  bility.
                                    A-5

-------
Malm, Kleine, and Kelley (1980) studied the perception threshold for computer-generated
white and NO2 Gaussian plumes.  The response to white and NO2 plumes resulted in
essentially identical contrasts.  Fifty percent of the observers were able to identify a plume
with a contrast of 0.014 and AE of 2.3, 75 percent a contrast of 0.020 and AE of 3.3, and 90
percent a contrast of 0.025 and AE of 4.1.

The most detailed study to date of plume perceptibility thresholds is the work of Malm et al.
(1986). In this study sharp-edged (square wave) plumes were generated by computer and
overlaid on color slides of a natural scene. Plumes of various contrasts and sizes (ranging
from 0.1  to 3° wide) were shown to observers.  These researchers found  that the detection
thresholds for such computer-generated square-wave plumes were a relatively strong function
of the  vertical plume width.  The highest visual sensitivity was found for 0.36° plumes,  which
is consistent with the previously noted maximum sensitivity at a spatial frequency of 3 cycles
/degree.  Maximum sensitivity was 200 (corresponding to a contrast of 0.005) for the 0.36°
plume, and sensitivities for all size plumes were approximately 100 or greater (contrasts of
0.01 or smaller).  These thresholds were defined at the 70 percent probability of detection
point.  This threshold contrast of 0.005 is consistent with the threshold contrast of 0.007 of
Howell and Hess (1978).

Table  A- 1 summarizes  the research described previously. Under laboratory conditions in
which observers  are attentive and trained, the detection threshold (for 50  percent detection)
for objects of optimum size with distinct edges is in the range 0.003-0.007.  For conditions in
which the stimulus has a  diffuse edge (such as would be the case with a  Gaussian plume) or
is different from the optimum-sensitivity size, threshold contrasts appear  to be higher,
approximately 0.009.  The evidence for AE thresholds is not as clear-cut.  The data of Jaeckel
(1973) and Malm, Kleine, and Kelley (1980) support 70 percent detection thresholds for AE
of 3, while the estimates of Lathner et al. (1978) and the more recent data of Malm et al.
(1986) suggest a AE threshold of less than one.

It is instructive to consider the relationship between contrast (which has been used in most
perception research) and AE. For monochromatic contrasts (those involving brightness change
(AL*), but not color change):
                                    Aa* = A£* = 0

thus

                  A£(L*a*&') = [(AL*)2 * (Aa*)2 + (A*')2]1'2  = AL*
                                          A-6                             Revised 10/92

-------
TABLE A-l.  Summary of contrast and color change threshold data.
Contrast
0.003*
0.014
0.007*
0.009+
0.016§
_ _
—
—
—
0.006
0.009
0.014
0.020
0.025
0.01
0.005**
0.010**
~ —
Delta E
—
—
• w
— _
—
1
2
3
4
1.0
1.5
2.3
3.3
4.2
—
—
1.2
Percent
Detection
50
?
?
?
?
30
50
70
90
10
25
50
75
90
90
70
70
100
Edge Reference
Sharp Blackwell (1946)
Sharp Lowry (1931, 1951)
Sharp Howe 11 and Hess (1978)
Diffuse
Sharp
Sharp Jaeckel (1973)
Sharp
Sharp
Sharp
Diffuse (Malm, Kleine, Kelley (1980)
Diffuse
Diffuse
Diffuse
Diffuse
Sharp Loomis et al. (1985)
Sharp Malm et al. (1986)
Sharp
Sharp
 * The most sensitive contrast reported for largest size of stimulus  and
   largest luminance and longest response time evaluated (probably  the
   minimum possible threshold).

 * The most sensitive contrast reported at a spatial frequency of 3
   cycles/degree.

 § Threshold contrast for sharp objects at low spatial frequencies.

 * Minimum threshold for 0.36° wide plumes.

 * Maximum threshold for all size plumes tested.
                                 A-7

-------
since

                   /Y\1/3
          L* = 116  -H     - 16  ,  and
                       ,,.1/3   /v,l/3]
             = 116   hH
If Y1  = (1  + C)Y  ,
                       *
                               7
                               v \ •
                     AL  = 116(7-]   ((1 + c)    - i]
                                O/
where

  AE(L*a*b*) = color difference parameter

         Al_* = change in perceived brightness

   Y1, Y, YQ = Y tristimulus values for an object (e.g., a plume), a
               viewing background, and a white reference, respectively

           C = contrast between observed object (e.g., a plume) and its
               viewing background

For bright viewing backgrounds (Y = 100), this formula yields the follow-
ing approximate formula;*

                                 AE = 38 C

Thus, the laboratory-derived threshold contrast of 0.009 for diffuse-edged
objects is the equivalent AE of 0.34.  The "traditional" (Koschmieder)
threshold contrast of 0.02 is the equivalent AE of 0.76.  Conversely, the
"traditional" just-noticeable-difference AE of 1 is the same as a contrast
of 0.026.
* This relationship assumes a bright viewing background and object.  A
  smaller A£ would result for darker viewed objects.  This finding  is
  consistent with the psychophysical experimental evidence that suggests
  that higher contrasts are required between two dark objects for them to
  be discerned.
                                     A-a

-------
RECOMMENDATIONS FOR PLUME SCREENING CRITERIA

The concept underlying the use of screening analyses, such as Level-1 and
Level-2, is that for some facilities whose emission rates are sufficiently
low, or that are located far enough away from sensitive areas, it may be
possible to use relatively simple calculations to determine that plume
visual effects will be negligible.  In this way, complex and costly
analytical approaches will only be required for those cases in which they
are needed to determine whether visual impacts are unacceptable.  Per-
ceptibility thresholds establish a lower bound for Level-1 and Level-2
screening criteria.  If, under transport and viewing conditions that con-
servatively describe "reasonable worst case" scenarios, it can be shown
that a plume's visual effects are below the threshold of perceptibility,
there is clearly no need to conduct more sophisticated analyses.

As  noted  in the preceding paragraphs, the perceptibility of an object
(e.g., a plume) may depend on both the observer and the viewing condi-
tions.  Under controlled conditions, trained observers looking for
specific objects having sharp edges may have "ower perception thresholds
than do casual observers  in natural conditions.

The literature suggests that the perceptibility threshold for trained and
attentive observers  in a  laboratory environment is on the order of 0.3-0.7
percent contrast and a a£ of 0.1-0.3.  In the natural environment, the
observer  is likely to be much less sensitive to contrast and color differ-
ences because he or  she  is not specifically "looking for plumes and  haze
layers."  Thus, the  use of laboratory-derived estimates of perceptibility
thresholds as screening criteria would be unnecessarily conservative.

Henry  (Henry,  1979;  Henry and Collins, 1982; Henry, private communication,
1987)  suggests  that  the  field threshold may be  2 to 4 times greater  than
the laboratory  threshold.  Although this speculation  is not based on
empirical  evidence,  it  is consistent with .our experience with  "prevailing
visibility" measurements.  For example, airport visibility observations
appear  to correlate  best  with light extinction  measurements when  a con-
trast  threshold of 5 percent  is  assumed in  the  Koschmieder equation  (Gor-
don,  1979; Tombach and  Allard,  1983).   In  their review of regional haze,
Mathai  and Tombach  (1985) make  the following observation:

      Laboratory and  field experiments of the same  sort gave simi-
      lar  results,  but most  field  data suggest a higher contrast
      threshold  than  do  laboratory  data, most probably because  the
      attention  and  target search  conditions differ.   Field experi-
      ments during  World  War  II  have  suggested a threshold contrast
      of about 0.05  to be more appropriate  for ordinary viewing,
      and  additional  recent  research  (Booker and Douglass,  1977;
                                     A-9

-------
     Johnson,  1981;  Tombach  and  Allard,  1980  and  1983;  Stevens et
     a!.,  1984)  further support(sl  the conclusion that  a value
     near  0.05 is more representative of normal viewing than the
     traditional  0.02 {contrast  threshold used in the Koschmieder
     equation].   Simply explained,  for normal  casual  viewing of
     nonspecific scenic features,  the perception  threshold is
     greater than it is when a specific target is being sought in
     earnest with a  relatively long time or a known distinctive
     form  available  to identify  it.

Laboratory thresholds appear to  underestimate actual  thresholds for the
casual observer (e.g., a visitor to a national park or wilderness area who
is not specifically  searching for  plumes or haze  layers).  The above-
suggested  threshold  contrast of  0.05 is an order-of-magnitude larger than
the laboratory-derived threshold contrast of 0.005.  For bright viewing
objects, a contrast  of 0.05 translates to a AE of approximately 2.

On the other hand, if we use the factor of 4 recommended by Henry (Henry
and Collins, 1982; Henry, personal  communication, 1987) to relate labora-
tory conditions to field conditions, the above-mentioned thresholds
derived for the laboratory convert to the following values for field
observation:  a contrast of - 0.02 and a AE of -  0.8.  These values may be
construed as the approximate "best estimate" thresholds of percepti-
bility.  We emphasize that these values are estimates of the percepti-
bility threshold for the casual  observer in the field; a sensitive obser-
ver may be able to detect plumes having much lower contrasts (0.003-0.007)
and lower AE  (0.1-0.3).

In summary, we suggest that the following values characterize our current
understanding of perceptibility:
                                                 Contrast  TT_A£_

          Lower-bound threshold
             (sensitive observer in laboratory)     0.005    0.2

          Best-estimate threshold
             (sensitive observer in field)          0.02     0.8

          Upper-bound threshold
             (casual observer  in field)             0.05     2

 Figure  A-2  synthesizes the  results of this review of perceptibility  thres-
 holds.  The abscissa  (x-axis)  shows  the  vertical width  (angular  subtense)
 of  a  plume.  The ordinate  (y-axis) is the just-perceptible contrast.   The
                                     A-10

-------
 1.0
 0.1
0.01
                       Howell and Hess (1978) data:
                       Square-wave gratings (sharp edge)
                       Sine-wave gratings (diffuse edge)
                                        Upper bound  (screening criterion)
      Perceptibility
       "Threshold"
                                                                 Data of Malm et  al.
                                                                 (1986) for sharp-
                                                                 edged plumes
0.001
                              0.1                         1.0

                             Plume Vertical Angular Subtense* (°)
10
   FIGURE A-2.   Plume perceptibility threshold  as a  function of plume
   thickness  (f).   See definition  of y  in the Glossary in  the front of  this
   workbook and in  Figure  3.
                                       A-11

-------
two curves show the data of Howe"!! and Hess (1978) for sharp and diffuse
edged objects, and the circles show the data of Malm et al. (1986) for
sharp-edged plumes.  These two separate and independent experiments are
remarkably consistent, indicating that plumes of approximately 0.3-0.5
degree vertical angular width are most easily perceptible.  Diffuse-edged
objects have larger perceptibility thresholds than sharp-edged objects.
The diffuse-edged curve derived from Howell and Hess (1978) data is taken
to be the lower bound of perceptibility for plumes.  The "best estimate"
and "upper bound" are also shown for comparison.  For very thin plumes
(< 0.1° width) and very wide plumes (> 5° width) the Howell and Hess
(1978) data are assumed to define the threshold.

For Levels 1 and 2 plume visual impact screening, we recommend that the
higher set of threshold values (contrast of 0.05; A£ of 2) be used as the
criteria for screening.  For very wide or narrow plumes the Howell and
Hess (1978) diffuse-edge thresholds should be taken as the criteria for
screening (see Figure A-2).
                                     A-12

-------
                    Appendix B




THE PLUME VISUAL IMPACT SCREENING MODEL (VISCREEN)

-------
                                Appendix B

           THE  PLUME  VISUAL  IMPACT  SCREENING  MODEL  (VISCREEN)*
The plume visual impact screening model (VISCREEN) was designed- to provide
the user with a simple, easy-to-use analytical tool for performing Level-1
and -2 screening calculations of potential plume visual impacts,
especially in PSD Class I areas where visibility is a protected value.
VISCREEN is a simple plume visibility model.  The objective of the model
is to calculate the contrast and the color difference of a plume and its
viewing background.  Because VISCREEN is to be used for screening calcula-
tions, it was designed to be conservative (i.e., to overpredict potential
plume-visual impacts).  Therefore, VISCREEN calculates larger plume visual
impacts, for the same input specifications, than do more sophisticated
models such as PLUVUE and PLUVUE II.

VISCREEN is designed to operate on the simplest and most modestly equipped
IBM PC or compatible.  It will operate with 256K memory.  It will utilize
a math coprocessor, if installed, with substantial improvement in execu-
tion speed.  VISCREEN is coded in FORTRAN 77.  A listing of the source
code is presented in Appendix D.  Figure B-l schematically illustrates the
logic flow of VISCREEN.  Each of the major calculation steps in VISCREEN
is described, in succession, in the following sections.
INPUT

Because VISCREEN is designed to be straightforward, the input requirements
have been scaled down to the minimum necessary to describe the variety of
emissions, meteorological and background conditions, and the plume/
observer geometries an analyst is likely to encounter in Level-1 and -2
plume visual impact screening.  Input is requested by screen prompts.
* See Latimer et al. (1978) and Latimer and Ireson (1980) for derivations
  of many of the VISCREEN descriptions and algorithms.
                                    B-l

-------
Emissions
Distances  between
emission  source  and
observer and Class I
area  boundaries
Visual range
Particle size,  density
Meteorology
                         Calculate  background
                             atmosphere's
                          optical   properties
                         Calculate  plume  size,
                         concentration,  NO2
                        formation,  and optical
                              thickness
                                 I
                           Calculate  plume
                            contrasts  and
                          radiances  for  sky
                         and  terrain  viewing
                            backgrounds
                         Calculate plume   AE
                          values  for sky and
                           terrain  viewing
                            backgrounds
                          Determine  whether
                        plume  contrast or   AE
                          values exceed the
                          screening  criteria
                          Display  screening
                          summary and   AE's
                           and  contrasts for
                           worst-cue or all
                            linei  of sight
                           Write  summary
                           and  output  files
                                             Yes
                                     No
                            f    Stop    J
Subroutine
 CHROMA
     FIGURE  B-l.   Logic  now  diagram  of  the  plume  visual  Impact
     screening  model   (VISCREEN).
                                        B-2

-------
VISCREEN first requests names for two disk files that will be created.
The summary file presents an abbreviated summary of inputs and screening
results.  The results file contains all inputs used in the analysis, as
well as geometry, &E, and contrast calculations for each line of sight and
wavelength.  This file is designed to allow its use as input to user-
developed graphics routines or commercially available spreadsheet pro-
grams.

The second set of inputs describes the emissions.  The user is given a
choice of units to input mass emission rates of various species that are
likely to cause visual effects.  The units of these emission rates are
mass per unit time.  Mass can be specified in metric units (grams, kilo-
grams, metric tonnes) or in English units (pounds or tons).  Time can be
specified in seconds, minutes, hours, days, or years.  The emission rate,
whatever the units used, should be the maximum short-term  (i.e., hour)
emission rate likely to occur in the course of operation of the emission
source.  The emissions that almost all analyses will consider are particu-
lates and nitrogen oxides (NOX); however, VISCREEN will also allow the
user to input such.species as (1) primary nitrogen dioxide (NC^) if this
species is directly emitted by the given chemical process, (2) primary
sulfate (SO^) if this species is directly emitted or if a  second particle
size mode needs to be specified, and (3) elemental carbon  (soot) if the
emission source is a diesel engine or other source with incomplete combus-
tion.  However, for the vast majority of commonly encountered emission
sources involving either fugitive emissions or combustion  emissions, the
analyst would only have to input particulate and NOX emissions.

The third set of inputs contains the distances that characterize the view-
ing situation.  The first is the distance between the emission source and
the observer.  The next two are the distances along the plume centerline
from the emission source to the closest and most distant  Class I area
boundaries.  Finally, the background visual range distance is specified.

The next set of  inputs  is requested only  if the user indicates that the
default specifications  built into VISCREEN  (see Table B-l) are not
acceptable for a given  screening analysis.  These include  the particle
size and density for  the emitted particulate and primary  sulfate and for
background fine  and coarse particulate, the background ozone concentration
(used  to calculate NO  to NO^ conversion  in  the plume), wind  speed, atmo-
spheric stability class, and the offset angle between the  plume center!ine
and the line between  the emission  source  and the observer.   For Level-1
analyses,  the default parameters would be used, and none  of  the above
inputs  would need to  be specified.   In most Level-2 analyses the default
particle size and density specifications  would be acceptable; only the
meteorological  input  specifications would have to be changed from the
Level-1 default  values.
                                  B-3

-------
TABLE B-l.  Default specification for VISCREEN.
            Particle Specifications
Mass Median
Type Diameter D (urn)
Background fine
Background coarse
Plume part icu late
Plume sulfate
Plume soot
0.3
6
2
0.5
0.1
Density
o(g/cm3)
1.5
2.5
2.5
1.5
2
Wind speed = 1 m/s
Stability = F
Background [63] = 0.04 ppm
Plume offset angle Y » 11.25'
                    B-4

-------
VISCREEN summarizes the inputs after specification and allows the user to
correct mistakes before proceeding.  If necessary, emission rates are con-
verted to grams per second.
GEOMETRY OF PLUME, OBSERVER, CLASS I AREA,
VIEWING BACKGROUND, AND LINES OF SIGHT

The next section of VISCREEN computes the angles and distances that
describe a variety of lines of sight relevant to the given situation.
Figure 7 (in the main text) illustrates the set of lines of sight  and  geo-
metries for a typical Level-1 and -2 screening calculation.  Figure  B-2
summarizes the angles and distances that describe a single line of
sight.  The plume is always assumed to be 22.5° wide, and the viewing
background is always assumed to be adjacent to the plume on the side of
the plume centerline opposite the observer.  VISCREEN computes plume
visual impacts for lines of sight at 5° increments for the azimuthal
angle $ and for lines of sight corresponding to distance x (along  the
plume centerline) of 1 kilometer, and xm1-n and xmax (the distances along
the plume centerline from the emission source to the closest and most  dis-
tant Class I area boundaries). The angle a and the distances r_ and  r0 are
needed for all subsequent plume visual impact calculations.  These
parameters can be solved for by noting the following relationships that
hold for all triangles:  (1) the sum of the three interior angles  equals
180° and (2) the ratio of the length of a triangle leg to the sine of  the
opposite angle is equal for all three legs of the triangle.

For the lines of sight where  is known, angle a can be solved directly as
follows:
                            a » 180  - y -
Since
        sina    sin   siny   '
 For  lines of  sight where x  is  known,  angle  $ must  be  calculated  as
 follows:
                                  Q-K

-------
Emission
Source
                                           Viewing Background
              Observer,
                                                                           22.5
                                                                           Wide Plume
      FIGURE B-2.  Distances and angles that specify a given line of sight.
                                      B-6

-------
                              tan"1    x sin
                              tan
                                     d - x cos Y


Similar calculations are used to determine the distance rQ.

The plume offset angle y is set to be equal to 11.25° for all Level-1
screening calculations, and most Level-2 screening analyses will be per-
formed for such a plume offset.  VISCREEN can be run for any arbitrary
offset angle between 0 and 180°; however, angles less than 11.25° and
greater than 168.75° are not recommended because these plume positions,
which are extremely rare, result in lines of sight along the plume axis
that are not calculated with precision using the assumptions coded in
VISCREEN.
OPTICAL PROPERTIES OF THE BACKGROUND ATMOSPHERE

The optical properties of the background atmosphere are then calculated
for each of the three wavelengths used in VISCREEN: 0.45, 0.55, and 0.65
ym.  These optical properties include the extinction coefficient and the
phase function for the forward and backward scatter sun angles assumed in
the screening process.  Sun angles are defined by the scattering angle Q,
which is the angle between the line of sight and the direct solar beam.
VISCREEN uses two scattering angles—10 and 140°~to calculate potential
plume visual impacts for cases where plumes are likely to be brightest
(6 = 10°) and darkest (0 = 140°).  Figure B-3 and Table B-2 show typical
phase functions for these two worst-case sun angles for typical size dis-
tributions.  The differences between phase functions for given particle
size distributions and pure air  (Rayleigh scattering) are greatest  in for-
ward scatter (10° is a reasonable estimate of a worst-case bright plume
situation) and in back scatter (140° is a reasonable estimate of a  worst-
case dark plume situation).

The scattering coefficient caused by particles is determined by subtract-
ing the Rayleigh scattering coefficient:
        bsp(x = 0.55 um) = bscat(x = 0.55) - bR(x = 0.55)


where bR(x = 0.55 um) =  11.62  x  10'V1.

On the basis of the data of Whitby and Sverdrup  (1978)  and  calculations  of
Latimer et al. (1978), the fraction of bsp caused by coarse particles  is
assumed to be 0.33.  Thus, we  have
                                  B-7

-------
   10J
in
in
"    D
«  10U
C
O
"Z    5
u
C
3
o.
  10
    -1
  10
    -2
             0 =  10°, forward
             scatter situation
             used in VISCREEN
         — 0.1 urn
                                                               (RAYLEIGH
                                                               SCATTER)
o = 140; back scatter
situation used in
VISCREEN
           NOTE:
             PHASE  FUNCTIONS AT  X = 0.55
             ARE SHOWN  FOR  PARTICLE SIZE  DIS-
             TRIBUTIONS WITH  INDICATED  MASS
             MEDIAN DIAMETER  (DG) AND GEO-
             METRIC STANDARD  DEVIATION  (a  =  2.0)
              20     40.    60      80      100      120
                          Scattering Angle  e  (degrees)
         140
160
180
     FIGURE B-3.   Phase  functions  for  various  particle  size  distributions.
                                  B-8

-------
         TABLE B-2.  Atmospheric optical parameters for various particle size distributions used In VISCREEN.
CD
UD
Particle
Size Mass Median
Index Diameter D(pm)
1
2
3
4
5
6
7
8
9
0.1
0.2
0.3
0.5
1.0
2.0
5.0
6.0
10.0
Phase Functions p(e,x)
n
2.8
2.1
1.6
1.0
0.2
0
0
0
0
bscat/V
(m2/cm3)
1.7
4.5
6.0
6.7
5.0
2.6
0.9
0.8
0.4
Blue (x
e = 10°
5.17
7.76
9.61
11.94
15.09
15.84
10.98
8.39
7.28
= 0.4 pm)
e = 140°
0.330
0.199
0.172
0.169
0.174
0.143
0.082
0.064
0.046
Green (x
e = 10°
4.24
0.49
8.11
10.33
13.64
16.07
13.64
11.67
9.23
= 0.55 pm)
6 = 140°
0.429
0.247
0.193
0.165
0.166
0.156
0.094
0.085
0.055
Red (X
6 = 10°
3.64
5.62
7.14
9.27
12.54
15.47
14.83
12.83
10.55
=0.7 um)
e = 140°
0.517
0.296
0.219
0.175
0.170
0.170
0.136
0.106
0.075

-------
                      bsp-submicron = °-67 bsp   '

                       bsp-coarse = °*33 bsp  •

The phase function for each of the scattering components can now be
determined.  Phase functions for the submicron  and  coarse background aero-
sols are specified in Table B-2.

The Rayleigh scattering phase function (for air)  is a function of the
scattering angle 0, but it is independent of wavelength x and can be
approximated quite well by the following relationship:


                     p(e) = 0.75 [1 + (cos e)2]
The scattering coefficients at different wavelengths (i.e.,  x = 0.45 and
0.65 ym) can be determined from the relationship:
                              • 0.55
where values of n are given in Table B-2 for various particle size distri-
butions and n » 4.1 for Rayleigh scatter.

The average background atmosphere phase functions are calculated for each
wavelength x and scattering angle e as follows:
                 p(x'9) [background
                                          bsp(x) p(x,e)
 PLUME DISPERSION, N02 FORMATION, AND OPTICAL CONDITIONS

 The plume  is treated as a Gaussian distribution in the vertical and a uni-
 form distribution in the horizontal over the width of the 22.5° sector.
 The line of sight is always assumed to be horizontal in VISCREEN; thus,
 optical thickness is calculated as follows:
         __Eiwnv71  P1  + 4N02  IT
Tplume
                                  (2»r" o  U Sin a
                                 B-10

-------
where the summation is over all particles (participate, SO^, and soot),
beX£/V is the light extinction efficiency per unit aerosol volume, and
bext/M is the light extinction efficiency per unit mass of NC^.

The amount of N02 in the plume is calculated by assuming that 10 percent
of initial NO is converted to N02 via the reaction with Q£ and the rest is
titrated with ambient 03.  For conservatism, the solar photodissociation
of N02 and the further reaction of N02 to form HN03 (realistic assumptions
for stable plume conditions near sunrise) are ignored.  In this conversion
the plume concentration of NOX is calculated as follows:
                   (N0v
                           (2w)l/2
N02 concentrations in the plume are calculated as follows:


                        + [N02]p    ,    if [NOX] > h


                      [NOX] +  [N02lp    ,    if  [NOX] < h



where
              [N02] =
        - plume center line ^ concentration,
      h = 0.1  [NOX] + [031,
  [N02P] - primary  (directly emitted) N02
        = background ozone concentration
The scattering efficiency for each particle size mode is taken from the
bS(.at/V shown in Table B-2.  Scattering at different wavelengths is scaled
using the parameter n (also shown in Table B-2) as follows:


                          ^crat                ' '  v "n
                        =* -^p (x * 0.55 um)
                                  B-ll

-------
The Light absorption efficiency for NC^ was taken by averaging efficien-
cies centered on the three wavelengths (x = 0.45, 0.55, 0.65 ym) to obtain
the following light absorption efficiencies:  0.691, 0.144, and 0.015
m /g.  The light absorption efficiency of soot was assumed to be 10 m/g
for all wavelengths.

To avoid gross overestimates of plume optical thickness (and potential
division by zero) for small a's, the minimum a is assumed to be 5°.  The
effect of limited persistence of worst-case stable meteorological condi-
tions is treated in VISCREEN by assuming that input stable dispersion con-
ditions (with stability categories of E and F) persist for a maximum of 12
hours.  For plume parcels located in positions that would require longer
transport times, additional dispersion is assumed by increasing the wind
speed for the given plume parcel so that the transport time exactly equals
12 hours.  This is a crude way of accounting for stable plume breakup
after long transport times.  Plumes with stability classes of A, B, C, or
0 are assumed to persist for all transport times including those greater
than 12 hours.
PLUME CONTRAST

The contrast of the plume against sky and black terrain viewing  back-
grounds is calculated conservatively by considering  single  scattering  and
ignoring multiple scattering  (Latimer and Ireson,  1980) as  follows:
 Plume Contrast
    'plume
                    plume
                             -  1
                 'background
 Reduction  in  Sky/Terrain  Contrast  Caused  by  Plume
            exp(-bextro)
 where
1 -
    Pplume'  pbackground =
  average phase functions for plume and back-
  ground atmosphere, respectively.  p~ is a
  function of x and 0,
                                 B-12

-------
                  ID ,     = ratio of light scattering to light extinction
                   plume         ,
                             in plume

             "background = 1 (assumin9 no absorption)


These contrast values are calculated for each wavelength (x = 0.45, 0.55,
and 0.65 ym) and each scattering angle (o = 10 and 140°).

Light intensities for later use in calculating the color difference
parameter A£ are calculated from the contrast values as follows:

The sky background light intensity for each scattering angle (e) and wave-
length of light (x) is calculated as follows:

                                  FS(X) p(x,  e)
                            !sky =      4^


where FS(X) is the radiant flux from the sun  (see Glossary in front of
Workbook).

Similarly, a white reference is
                              Ao -  27-    •


The light intensity of the plume against the sky is  (Latimer et al,,
1978):


                     plume-sky ~ *•     plume'   sky


The light intensity of the terrain background  (assumed to be black) viewed
at distance rQ is


                  Wain * U - exp(-bext ro>  Jsky   •


The light intensity of the plume viewed against the  dark terrain viewing
background is then


                 1 plume terrain = terrain * ACr *sky
                                   B-13

-------
PLUME AE VALUES

The color difference parameter AE is calculated from the three light intensities using the
following equation:
where
                              L' = 116 (Y/YJ113 - 16  ,
                             a* =500 MM   -JL
                             fc'-MOM-    -  —
                                           /a,)
 In these equations, the tristimulus values X0, Y0, Z0 define the color of the nominally  white
 object-color stimulus from a perfectly diffuse reflector normal to the direct solar beam (I0
 defined above).  Calculations are normalized such that Y0 equals a typical midday
 illumination of 100 candle/m2 and X0 = Z0 = 100 candle/m2.  The AL*, Aa*, and Ab* refer to
 the difference in these three functions between the plume and its viewing background (either
 sky or terrain).
 The three chromaticity tristimulus weighting functions  x y z were determined for each of
 the wavelengths by averaging the values shown in Figure B-4 over the wavelengths centered
                                         B-14
Revised 10/92

-------
on X = 0.45, 0.55, and 0.65 urn. These average weighting factors and other parameters used
in VJSCREEN are summarized in Table B-3.
COMPARISON OF CALCULATIONS WITH SCREENING CRITERIA

The calculated contrast and AE values are compared to the default screening criteria described
in Appendix A (i.e., AE = 2, contrast = 0.05, and the Howell and Hess curve for diffuse-edge
objects) or to user-specified criteria.  The vertical plume dimension for each line of sight is
calculated using the following formula:
                                       B-14a                          Revised 10/92

-------

-------
       400
 500        600
Wavelength, \ (nm)
  Source:   Judd and Wyszecki  (1975).
FIGURE B-4.    Weighting values x(x), y(x), Z(A)
                  B-15

-------
TABLE B-3.  Average chromaticity tristimulus weighting
functions, NC^ light absorption efficiency, and solar flux
used in VISCREEN.

             	Wavelength \	
             Blue, 0.45 urn    Green, 0.55 ym    Red, 0.65 ym
Parameter    (0.36-0.50 ym)   (0.51-0.60 ym)   (0.61-0.74 ym)
X
y
~z
bahe - N07/M
0.1196
0.0935
0.7012
0.691
0.6317
0.8229
0.0159
0.144
0.1838
0.0753
0.0000
0.015
   Fs            1712             1730             1414
(watt nf2 sr'1)
                                B-16

-------
                                    -
                               =  tan
                                           P

This plume width is used with the diffuse-edge curve in Figure A-2 to
determine the minimum perceptible contrast.  The green (x = 0.55 ym) plume
contrast for each scattering angle and viewing background is then compared
to this minimum perceptible contrast and the screening contrast threshold
of 0.05.  A secondary test is made by comparing the plume AE with the
minimum perceptible AE and the screening A£ of 2.  If the plume contrast
is greater than both contrast values, or the plume AE is greater than both
A£ thresholds, the given line of sight fails the screening test.  To find
the minimum perceptible AE, the equivalent AE is calculated for an object
having the minimum perceptible (Howell and Hess) contrast from the sky and
the terrain viewing backgrounds (see Appendix A).  This approach is
believed to provide a conservative underestimate of threshold AE,
especially for dark terrain viewing backgrounds for which higher contrasts
may be needed to distinguish a plume.
OUTPUT

VISCREEN generates a summary file and a results file in addition to the
screen display during a user's sessions.  The files store the inputs and
results to provide the user with a record of a run.  The user is queried
by VISCREEN for the names of the files.

The summary file (see Figure 16) is designed to provide a concise, single-
page summary of the inputs and the results of a run.  Only the user-
supplied inputs are provided; defaults are not printed to save space.  The
results included in the summary file display maximum visual  impacts inside
and outside of the Class I area, with any screening criteria exceedances
indicated.  Format of the summary file is self-explanatory.

The results file contains all inputs and results that may be used for
other analyses.  The format of the results file is designed  to allow the
file to be imported into commercially available spreadsheet  programs.  In
this way, users can design their own tabular and graphical displays of
VISCREEN results.  The content and output formats for the results file are
provided in Table B-4.  There are three sections to the file.  The first
section (records 1 through 11) contains the inputs for the run.  Section
two (records 12 and 13+) contains the number of lines of sight (record 12,
included so that a user-developed program can know how many  records of
information to read), followed by one record for each line of sight (LOS).
These LOS records include LOS geometry information, screening threshold,
and AE values.  The final section is similar to the second,  in that it
                                 B-17

-------
Table B-4.   Output format for the VISCREEN results file
Record
No. Contents
1 Source name
2 Class I area name
3 Mass unit
= 1 grams
= 2 kilograms
= 3 metric tonnes
= 4 pounds
= 5 tons
Time flag
= 1 seconds
= 2 minutes
= 3 hours
= 4 days
= 5 years
4 Particulate emission rate
NOx emission rate
N02 emission rate
Soot emission rate
S04 emission rate
5 Source-Observer distance
M1n. Source-Class I distance
Max. Source-Class I distance
Background visual range
6 Default flag
= 1 used default value
= 0 user input value
Background Fine Particulate Density (g/cm3)
Background Fine Particulate Size index (urn)
= 1 0.1
= 2 0.2
= 3 0.3
= 4 0.5
= 5 1.0
= 6 2.0
= 7 5.0
= 8 6.0
= 9 10.0
Format
A
A
15





15





F10.3
F10.3
F10. 3
F10. 3
F10.3
F10.3
F10.3
F10.3
F10.3
15


F10.3
15









         Default flag                                    15
         Background Coarse  Particulate  density           F10.3
         Background Coarse  Particulate  size  index        15
                                 B-18

-------
Table B-4 (concluded).   Output format for the VISCREEN results file
Record
  No.    Contents                                      Format

    8Default flag15
         Plume Participate density                      F10.3
         Plume Particulate size index                   15
    9    Default flag                                   15
         Plume Soot density                             F10.3
         Plume Soot size index                          15
   10    Default flag                                   15
         Plume Primary S04 density                      F10.3
         Plume Primary S04 size index                   15
   11    Default flag                                   15
         Background Ozone (ppm)                         F10.3
         Wind speed (m/s)                               F10.3
         Stability index                                15
   11A   (Record not included in Level-1 runs)
         Default flag, plume offset angle               I5.F10.3
   12    Number of lines of sight                       15
   134-   Line of sight (LOS)                            IX, 12
         LOS classification                             12
           = 0  outside Class I area
           = 1  inside Class I area
         Azimuthal angle                                F8.1
         Angle between LOS and plume                    F7.1
         Source-observed plume distance                 F7.1
         Observer-observed plume distance               F7.1
         Observer-terrain distance behinde plume        F7.1
         PSI                                            F5.2
         Green Contrast threshold                       F7.3
         Screening threshold and delta E for:
           sky, forward scatter                         2F7.2
           sky, backward scatter                     "  2F7.2
           terrain, forward scatter                     2F7.2
           terrain, backward scatter                    2F7.2
   14    Number of lines of sight                       15
   15+   Line of sight number                           IX,I2
         LOS classification                             12
         Azimuthal angle                                F8.3
         Green Contrast threshold                       F7.3
         Green contrast for:
         forward scatter, sky and terrain background    2F7.3
         backward scatter, sky and terrain background   2F7.3
         Blue contrast (as for green)                   4F7.3
         Red contrast (as for green)                    4F7.3

Note: Records 13+ and 15+ are repeated for each line of sight:

-------
begins with the number of LOS's,  and is followed by one record for each
LOS.  These contain LOS identifiers, the green contrast screening
criterion, and green,  blue and red contrast values .for each line of sight
and viewing background.  Figure B-5 shows an example of a complete results
file.
CONSERVATISM OF VISCREEN

VISCREEN is designed for use in Level-1 and -2 plume visual impact screen-
ing calculations.  The objective of the screening exercise is to identify
emission sources that have the potential to cause adverse visibility
impairment.  Because these sources can be analyzed further (with more
sophisticated models) in a more detailed manner (e.g., using Level-3
analysis), the screening model should yield output that is consistently
conservative.  That is, it should calculate plume visual impacts that are
likely to be greater than those that would actually be encountered and
those that would be calculated in Level-3 analysis.  This conservatism is
necessary to avoid approving an emission source that passes a screening
test, but could have problems that would be revealed by a more detailed
analysis.  It also eliminates the need for facilities with negligible
effects to carry out more complicated and costly assessments.

VISCREEN was designed to be conservative by making the following model
assumptions:

     1.   It is assumed that the line of sight is horizontal so that it
          Intersects the most plume material.  Nonhorizontal lines of
          sight intersect less plume material because horizontal disper-
          sion of plumes exceeds vertical dispersion, especially under
          stable conditions.

     2.   N02 conversion is conservatively treated by assuming the plume
          is uniformly mixed in the 22.5° sector.  This enhanced disper-
          sion mixes the plume with more ambient Oj, resulting in greater
          conversion.  However, the assumed enhanced dispersion does not
          decrease the line-of-sight integral of plume material for the
          assumed horizontal viewing conditions.  Only the vertical
          dimensions of the plume determine the magnitude of the plume
          material that intersects the horizontal line of sight.

     3.   Worst-case sun (scattering) angles are assumed.  The forward
          scatter case (9 = 10°) yields very bright plumes because the
          sun  is placed nearly directly in front of the observer.  This
          geometry would rarely occur 1n reality.  The backward scatter
          case  (e =  140°) yields the darkest possible plumes.  Thus, the
                                 B-20

-------
00
i
          'Public Electric Coal #3
          'Longvlew NP
              1    1
10.000 120.000
80.000 80.000
1 1.500 3
1 2.500 8
1 2.500 6
1 2.000 1
1 1.500 4
1 .040 1
1 11.250
34
1 0
2 0
3 0
4 0
5 0
6 0
7 0
8 0
9 0
10 0
11 0
12 0
13 0
14 0
15 0
16 0
17 1
18 1
19 1
20 1

5.0
10.0
15.0
20.0
25.0
30.0
35.0
40.0
45.0
50.0
55.0
60.0
65.0
70.0
75.0
80.0
85.0
90.0
95.0
100.0

163.8
158.8
153.8
148.8
143.8
138.8
133.8
128.8
123.8
118.8
113.8
108.8
103.8
98.8
93.8
88.8
83.8
78.8
73.8
68.8
.000 .000 .000
120.000 110.000
.000 6

24.9
38.3
46.8
52.7
57.2
60.7
63.5
65.9
68.0
69.9
71.6
73.2
74.6
76.1
77.4
78.8
80.2
81.6
83.0
84.5

55.8
43.1
35.3
30.1
26.4
23.7
21.6
20.0
18.8
17.8
17.1
16.5
16.1
15.8
15.6
15.6
15.7
15.9
16.3
16.7

66.3 .29
57.0 .42
50.3 .55
45.3 .66
41.5 .77
38.6 .87
36.3 .96
34.5 1.04
33.1 1.12
32.1 1.19
31.4 1.25
30.9 1.30
30.6 1.34
30.6 1.37
30.9 1.38
31.4 1.39
32.1 1.39
33.1 1.37
34.5 1.35
36.3 1.32

.050
.050
.050
.050
.050
.050
.050
.050
.050
.050
.050
.050
.050
.050
.050
.050
.050
.050
.050
.050

2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00

2.29
3.70
4.78
5.29
5.55
5.65
5.66
5.61
5.53
5.43
5.32
5.22
5.11
5.01
4.91
4.82
4.73
4.64
4.55
4.45

2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00

1.42
1.62
1.97
2.26
2.47
2.60
2.66
2.68
2.68
2.66
2.62
2.58
2.54
2.50
2.45
2.41
2.36
2.32
2.26
2.21

2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00

3.41
3.18
2.86
2.55
2.38
2.27
2.18
2.11
2.04
1.98
1.91
1.84
1.78
1.71
1.64
1.56
1.48
1.40
1.32
1.22

2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00

1.20
1.03
.92
.87
.87
.88
.87
.86
.84
.82
.79
.76
.73
.70
.66
.63
.59
.55
.50
.46
                                              FIGURE B-5.  Example  results  file.

-------
oa
i
ro
21 1
22 1
23 1
24 1
25 1
26 1
27 1
28 1
29 1
30 0
31 0
32 0
33 1
34 1
34
1 0
2 0
3 0
4 0
5 0
6 0
7 0
8 0
9 0
10 0
11 0
12 0
13 0
14 0
15 0
16 0
17 1
18 1
19 1
20 1
105.0
110.0
115.0
120.0
125.0
130.0
135.0
140.0
145.0
150.0
155.0
.1
84.4
148.2

5.000
10.000
15.000
20.000
25.000
30.000
35.000
40.000
45.000
50.000
55.000
60.000
65.000
70.000
75.000
80.000
85.000
90.000
95.000
100.000
63.8
58.8
53.8
48.8
43.8
38.8
33.8
28.8
23.8
18.8
13.8
168.6
84.4
20.6

.050
.050
.050
.050
.050
.050
.050
.050
.050
.050
.050
.050
.050
.050
.050
.050
.050
.050
.050
.050
86.2
87.9
89.9
92.1
94.8
97.9
101.8
106.9
113.9
124.4
142.2
1.0
80.0
120.0

-.006
-.008
-.008
-.007
-.006
-.006
-.006
-.005
-.005
-.005
-.005
-.004
-.004
-.004
-.004
-.004
-.004
-.004
-.004
-.004
17.4
18.3
19.4
20.8
22.6
24.9
28.1
32.4
38.8
48.6
65.7
79.0
15.7
44.4

.043
.044
.042
.038
.035
.033
.031
.029
.028
.026
.025
.024
.023
.022
.021
.021
.020
.019
.018
.017
38.6
41.5
45.3
50.3
57.0
66.3
80.0
101.8
141.4
234.5
701.9
79.5
32.0
156.9

-.055
-.064
-.064
-.059
-.054
-.051
-.048
-.045
-.043
-.041
-.039
-.037
-.036
-.035
-.034
-.034
-.033
-.033
-.032
-.032
1.27
1.22
1.15
1.08
1.00
.91
.81
.71
.60
.49
.37
.04
1.39
.53

.0.
.o:
.o:
.0;
.0;
.0;
.0;
.0:
.0:
.0:
.0:
.0:
.0:
.01
.0]
.0]
.01
.Q]
.01
.01
                                                            .050   2.00   4.35
                                                            .050   2.00   4.23
                                                            .050   2.00   4.08
                                                            .050   2.00   3.90
                                                            .050   2.00   3.67
                                                            .050   2.00   3.38
                                                            .050   2.00   3.00
                                                            .050   2.00   2.51
                                                            .050   2.00   1.91
                                                            .050   2.00   1.22
                                                            .050   2.00    .55
                                                            .109   6.13   5.37
                                                            .050   2.00   4.74
                                                            .050   2.00   1.48
040  -.026    .021
038  -.045    .030
034  -.059    .035
029  -.065    .037
025  -.068    .038
022  -.069    .038
020  -.069    .038
018  -.068    .037
017  -.067    .036
015  -.066    .034
014  -.065    .033
013  -.064    .032
013  -.062    .031
012  -.061    .029
012  -.060    .028
Oil  -.059    .027
Oil  -.058    .025
010  -.057    .024
010  -.056    .022
010  -.054    .020
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.60
2.00
2.00
2.14
2.07
1.98
1.87
1.73
1.56
1.35
1.09
.80
.50
.28
1.33
2.37
.60
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
5.99
2.00
2.00
1.13
1.02
.91
.79
.67
.53
.38
.22
.07
.00
.00
5.19
1.49
.05
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.56
2.00
2.00
.41
.35
.29
.23
.18
.13
.09
.06
.02
.00
.00
1.92
.59
.01
-.036
-.064
-.084
-.092
-.097
-.098
-.098
-.097
-.096
-.094
-.092
-.090
-.088
-.087
-.085
-.083
-.082
-.080
-.079
-.077
.020
.029
.035
.036
.036
.036
.035
.033
.032
.030
.029
.028
.026
.025
.024
.023
.022
.020
.019
.018
.040
.040
.036
.031
.027
.024
.022
.021
.019
.018
.017
.016
.016
.015
.015
.015
.014
.014
.014
.014
.047  -.035   .037
.045  -.035   .031
.041  -.031   .025
.036  -.027   .020
.033  -.024   .017
.031  -.021   .015
.029  -.019   .013
.027  -.018   .012
.026  -.017   .011
.024  -.016   .010
.023  -.015   .010
.022  -.014   .009
.022  -.014   .009
.021  -.013   .008
.020  -.013   .008
.019  -.013   .008
.019  -.013   .007
.018  -.012   .007
.017  -.012   .007
.017  -.012   .007
                                                  FIGURE B-5  Continued

-------
21 1 105.000
22 I 110.000
23 1 115.000
24 1 120.000
25 1 125.000
26 1 130.000
27 1 135.000
28 1 140.000
29 1 145.000
30 0 150.000
31 0 155.000
32 0    .141
33 1  84.375
34 1 148.157
.050
.050
.050
.050
.050
.050
.050
.050
.050
.050
.050
.109
.050
.050
-.004
-.004
-.004
-.004
-.004
-.003
-.003
-.003
-.003 '
-.002
-.001
.051
-.004
-.002
.016
.015
.013
.012
.010
.007
.005
.002
.001
.000
.000
.050
.020
.000
-.032
-.031
-.031
-.031
-.030
-.029
-.028
-.026
-.023
-.018
-.011
-.037
-.033
-.020
.009
.009
.008
.007
.006
.005
.004
.002
.001
.000
.000
.049
.011
.000
-.053
-.052
-.050
-.048
-.045
-.042
-.037
-.031
-.024
-.015
-.006
.006
-.058
-.018
.018
.016
.014
.011
.008
.005
.003
.001
.000
.000
.000
.011
.026
.000
-.075
-.073
-.071
-.068
-.064
-.059
-.053
-.044
-.034
-.021
-.009
-.010
-.082
-.026
.016
.014
.012
.010
.007
.005
.002
.001
.000
.000
.000
.011
.022
.000
.014
.014
.014
.014
.014
.014
.014
.014
.013
.011
.008
.101
.014
.012
.016
.015
.014
.012
.011
.009
.007
.004
.001
.000
.000
.102
.019
.001
-.012
-.012
-.012
-.012
-.012
-.012
-.012
-.012
-.011
-.010
-.007
-.059
-.013
-.010
.007
.006
.006
.006
.005
.004
.003
.002
.001
.000
.000
.098
.007
.001
                                        FIGURE B-5  Concluded

-------
     screening calculations are likely to yield the brightest and
     darkest possible plumes.   It is left to more detailed PLUVUE
     modeling to identify realistic worst-case sun angles that would
     occur at specific times of interest.

4.   Multiple scattering is ignored in VISCREEN.  Light scattered
     into the line of sight from directions other than directly from
     the sun tend to slightly decrease the plume contrast for the
     worst-case sun angles assumed.

5.   For terrain viewing backgrounds, the terrain is assumed to be
     black (the darkest possible) and located as close to the obser-
     ver and the plume as possible.  This assumption yields the
     darkest possible background against which particulate plumes are
     likely to be most visible.  In reality, terrain viewing back-
     grounds (if indeed terrain is behind the plume) would be less
     dark and would be located farther from the observer.

6.   Meteorological conditions are assumed to persist for at least 12
     hours.  After 12 hours, some additional dispersion is assumed in
     VISCREEN (by increased wind speeds), but the plume is still con-
     sidered to remain intact.  More realistic treatment of the per-
     sistence of worst-case dispersion conditions would most likely
     yield lower plume visual impacts.

7.   Default meteorological conditions assumed for the most conserva-
     tive Level-1 screening (F, 1 m/s, Y - 11.25°) are extreme and
     are expected to be more conservative than worst-case conditions
     identified in the more realistic Level-2 and -3 analyses.

8.   The screening threshold (A£ = 2; contrast of 0.05) was selected
     at the upper bound of the perceptibility threshold, representing
     a reasonable estimate for casual observers in the field.

-------
                      Appendix C




EXAMPLES OF PLUME VISUAL-IMPACT SCREENING AND ANALYSIS

-------
                                     Appendix C

      EXAMPLES OF PLUME VISUAL-IMPACT SCREENING AND ANALYSIS
The objective of this appendix is to assist the reader in understanding how specific screening
and analyses might be carried out in different situations and at different levels of analysis.
The detailed instructions provided in the text of this document are not repeated here.  Rather,
the examples are accompanied by limited commentary  so that the reader obtains an overview
of different plume visibility screening alternatives.  Any application of plume visual-impact
screening and analysis technology will differ depending on the circumstances of the given
scenario.

This appendix provides examples of visibility screening and more detailed analyses for five
different scenarios:
LEVEL-1  AND LEVEL-2 SCREENING

       1.    The first example that was presented in Latimer and Ireson (1980), a coal-fired
            power plant, for which Level-1 and -2 screening calculations were performed.

       2.    The second example that was presented in Latimer and Ireson (1980), a cement
            plant, for which Level-1 and -2 screening calculations were performed.

       3.    A paper mill located very close to a Class I area for which Level-1 and -2
            screening calculations were  performed.
LEVEL-3 ANALYSIS

       4.    A large coal-fired power plant located 90 km from a western national park, for
            which Level-1, -2, and -3 screening and analyses were carried out.

                                         C-l                            Revised 10/92

-------
       5.    A very small emission source located extremely close to a western Class I area,
            for which all three levels of screening and analysis were performed.
EXAMPLE  1:  COAL-FIRED POWER PLANT (1980 WORKBOOK EXAMPLE 1)

This example is based on a hypothetical coal-fired power plant proposed for a site
approximately 70 km from a Class I PSD area in  Nevada. The emission rates for this
hypothetical power plant are projected to be 25 g/s of particulates, 380  g/s of nitrogen oxides
(as N02), and 120 g/s of sulfur dioxide.  Figure C-!  shows the relative locations of the
proposed site and the Class I  area.  The Federal Land Manager has identified the view toward
the mountains to the west.as integral to the visitors' experience of the Class I area.

For conservatism, the observer is  placed on the boundary of the Class I area closest to the
power plant, which in this case is at the southwestern corner of the Class I  area.  (Although
more visitors would  be located at the visitors' center, the Federal Land  Manager has stated
that all locations in  the Class  I area are of interest because of widespread visitor use.)  From
measurements made  off of a topographical map (see Figure C-l), the distance from  the
proposed plant site to this closest corner is 70 km. Since the lines drawn at an  11.25° angle
on both sides of the  line between the plant site and the nearest corner of the Class I area are
outside the Class I area, the closest Class I  area boundary is also selected to be 70 km, for
conservatism.

Exhibit C-l  shows the results of the VISCREEN analysis for this example.  The source fails
the Level-1  test with a maximum AE of 17.8, nearly  nine times the screening threshold.   Its
maximum contrast of -0.140 (for  the backward-scattering scenario) is nearly identical to the
1980 Workbook Level-1 screening calculation  of -0.146.  The plume is also predicted to  be
visible against terrain with a contrast of +0.107 (for the forward-scattering scenario), a
slightly higher value than  the 0.0814 calculated in the 1980 Workbook.

To characterize worst-case meteorological conditions for Level-2 screening, we obtained
meteorological data from an airport 100 km west  of the proposed power plant.  Although the
intervening  terrain is not flat, we  judged that the 850-mb wind and stability data are the best
available data source. For the trajectory passing to the northwest of the Class I area,
                                           C-2                             Revised 10/92

-------
 o
 I
 CO
                                   Scale In kilometers
                                      \//\  CLASS I AREA
n
O-
S
K)
Figure C-l.   Relative locations of Example 1 proposed power plant and Class I area for
             example I, screening example (where y - 11.25° and i|> = aziinuihal angle of
             observer line of sight).

-------
                      Level-1 Screening
 Input Emissions  for
    Participates
    NOx (as  N02)
    Primary  N02
    Soot
    Primary  S04
    25.00  G   /S
   380.00  G   /S
      .00  G   /S
      .00  G   /S
      .00  G   /S
     **** Default  Particle Characteristics Assumed
               Transport Scenario Specifications:
     Background  Ozone:
     Background  Visual Range:
     Source-Observer  Distance:
     Min.  Source-Class I Distance:
     Max.  Source-Class I Distance:
     Plume-Source-Observer Angle:
     Stability:    6
     Wind  Speed:   1.00 m/s
                        .04 ppm
                     170.00 km
                     70.00 km
                     70.00 km
                     90.00 km
                     11.25 degrees
                           RESULTS

 Asterisks (*)  indicate plume impacts that exceed screening  criteria

          Maximum  Visual  Impacts INSIDE  Class I  Area
             Screening Criteria ARE Exceeded
                                    Delta E       Contrast
 Backgrnd Theta  Azi  Distance Alpha Crit  Plume   Crit  Plume
SKY
SKY
TERRAIN
TERRAIN
10.
140.
10.
140.
84.
84.
84.
84.
70
70
70
70
.0
.0
.0
.0
84.
84.
84.
84.
2
2
2
2
.00
.00
.00
.00
17
10
8
4
.807*
.828*
.852*
.004*
.05
.05
.05
.05
-.005
-.140*
.107*
.041
          Maximum  Visual  Impacts OUTSIDE Class I  Area
             Screening Criteria ARE Exceeded
                                    Delta E       Contrast
 Backgrnd Theta  Azi  Distance Alpha Crit  Plume   Crit  Plume
  SKY      10.
  SKY     140.
  TERRAIN  10.
  TERRAIN 140.
35.
35.
15.
15.
55.6
55.6
41.0
41.0
134.
134.
154.
154.
2.00 20.370*
2.00 11.101*
2.00 15.827*
2.00  4.791*
.05   -.007
.05   -.207*
.05    .205*
.05    .143*
EXHIBIT  C-l.   Level  1  screeninq  analysis  for  Example  1.
                     C-4

-------
we tabulated winds from the south'west and west-southwest for both morning and afternoon
soundings.  From these tabulations, a frequency of occurrence (Table C-l) was developed.
The cumulative frequency entries show that on three to four days per year conditions with
c>ya2u values of 7.5xl05 m3/s (E  stability, 2 m/s) can be expected. Note that the bulk of the
contribution to the cumulative  frequency  (0.9 percent out of 1.0 percent) represents the  1200
GMT E,2 dispersion conditions.  This corresponds to approximately 5:00 a.m. LST. Note
also that the afternoon sounding  frequency of E,2 dispersion conditions was relatively  high
(0.6 percent, or about two days per year).

Exhibit C-2 summarizes the VISCREEN analysis  using the meteorological conditions of E
and 2 m/s (less extreme than the Level-1  F and 1 m/s).  The maximum plume perceptibility
for plume parcels located within the Class I area occurs  when the sun is in front of the
observer (forward-scatter conditions) and the plume is observed against the sky. For these
conditions, the plume AE is 8.9,  about 4.5 times  larger than the screening threshold. Given
the geometry shown in Figure  C-l, the possibility could not be ruled out that such  a forward-
scatter situation would occur.  Even if such a sun angle  were not possible, the second  test for
a backward scatter sun angle indicates that the plume would be quite visible, exceeding both
the AE and the green contrast screening thresholds. The even larger impacts calculated for
plume parcels outside the Class I area are relevant in this example since they could occur
within an identified integral vista.  The maximum green contrasts for the plume parcels
located outside the Class I  area were 0.231  in forward scatter and -0.129 in backward  scatter.
These values require careful interpretation, however, as they are  for the line of sight through a
plume parcel only 1  km from the source.

Although not shown here, a Level-3 analysis would be required for  this plant because  of the
failure of both the Level-1  and -2 tests for lines of sight within the Class I area.
EXAMPLE 2:  CEMENT PLANT AND RELATED OPERATIONS (1980 WORKBOOK
           EXAMPLE 2)

A cement plant has been proposed, along with related quarrying, materials handling, and
transportation facilities,  for a location 20 km from a Class I area. Terrain in the vicinity is
relatively flat, and no external vistas from the Class I  area (a  national park) are considered
integral to park visitor experience. Visibility at some  locations within the park boundaries is
of concern, however.

The point in the Class I area closest  to the proposed site is shown in Figure C-2 as Point A.
This point is 20 km away from the proposed


                                          C-5                             Revised 10/92

-------
TABLE C-l.   Frequency of Occurrence  of SW and WSW Winds by Dispersion Condition
and Time of Day
Dispersion
Condition ayo\u Transport
(stability, Time
wind speed) (m3/s) (hours)

F,l
F,2
E,l
F,3
E,2
E,3
D,l
E,4
E,5
D,2
D,3
D,4

1.29xl03
2.57xl05
3.75xl03
3.86xl05
7.50xl05
1.12xl06
1.16X106
1.50xl06
1.87xl06
2.32x1 06
3.49xl06
4.65xl06

33
11
33
7
11
7
33
5
4
11
7
5
Time of Day (percent)1
OOZ 12Z
f2
0.1
0.1
0.2
0.0
0.6
0.6
0.4
0.4
0.2
1.6
3.4
2.4
cf3
0.0
0.1
0.1
0.1
0.7
1.3
1.3
1.7
1.9
3.5
6.9
9.3
f
0.2
0.0
0.3
0.1
0.9
1.4
0.3
1.2
1.8
0.8
1.2
1.5
cf
0.0
0.0
0.0
0.1
1.0
2.4
2.4
3.6
5.4
6.2
7.4
8.9
 1. OOZ refers for midnight Greenwich Mean Time (GMT) and 12Z refers to noon GMT.

 2. Frequency

 3. Cumulative Frequency

 4. Persistence of stable meteorological conditions for over 12 hours is not considered likely. Therefore,
 conditions requiring greater than 12-hour transport time are not included in the cf contribution.


 Note:  Distance downwind, values of o , az, and transport times are based on a distance of 70 km.
                                               C-6
Revised 10/92

-------
              *** User-selected Screening Scenario  Results *"
Input  Emissions  for
   Particulates
   NOx  (as  N02)
   Primary  N02
   Soot
   Primary  S04
 25.00  G  /S
380.00  G  /S
   .00  G  /S
   .00  G  /S
   .00  G  /S
              PARTICLE CHARACTERISTICS
              Density       Diameter
Primary  Part.
Soot
Sulfate
2.5
2.0
1.5
              Transport Scenario Specifications.
    Background  Ozone:
    Background  Visual Range:
    Source-Observer Distance:
    Hin.  Source-Class I Distance:
    Max.  Source-Class I Distance:
    Plume-Source-Observer Angle:
    Stability:   5
    Wind  Speed:   2.00 m/s
                     .04 ppm
                  170.00 km
                   70.00 km
                   70.00 km
                   90.00 km
                   11.25 degrees
                          RESULTS
Asterisks  (*)  indicate plume impacts that exceed  screening criteria

        Maximum Visual Impacts INSIDE  Class  I Area
            Screening Criteria ARE Exceeded
                                   Delta E      Contrast
Backgrnd  Theta Azi Distance Alpha Crit  Plume    Crit  Plume
SKY
SKY
TERRAIN
TERRAIN
10.
140.
10.
140.
120.
120.
84.
84.
80.
80.
70.
70..
6
6
0
0
49.
49.
84.
84.
2
2
2
2
.00
.00
.00
.00
a.
5.
4.
1.
,925*
.312*
.050*
.763
.05
.05
.05
.05
-.002
-.070*
047
.017
         Maximum Visual Impacts OUTSIDE Class  I  Area
            Screening Criteria ARE Exceeded
                                   Delta £      Contrast
Sackgrnd  Theta Azi Distance Alpha Crit  Plume   Crit  Plume
SKY
SKY
TERRAIN
TERRAIN
10.
140.
10.
140.
0.
0.
0.
0.
1.
1.
1.
1.
.0
.0
.0
.0
169.
169.
169.
169.
2
2
2
2
.00
.00
.00
.00
18.
4
15.
948*
.808*
.292*
6.160*
.05
.05
.05
.05
.231*
-.129*
.166*
.151*
EXHIBIT  C-2.    Level  2 screening  analysis  for Example  1
                          C-7

-------
               Proposed
               Cement  Plant
                                                            lass I
                                                           Area Boundary
                                20
                               Scale
                                              I
40 km
Figure C-2.   Relative locations of Example 2 proposed cement plant and Class I area.

-------
plant.   Lines drawn 11.25° on either side of the line between the site and
Point A intersect the Class I area boundary at distances (for conservatism
in Level-1 screening) of 23 and 25 km.  Since these distances are greater
than the minimum distance, the minimum distance to the Class I area boun-
dary (xmi-n) is set equal to 20 km, as suggested by the Workbook.  The most
distant Class I area boundary (xmax) for analyses on Point A is 80 km away
from the cement plant site.

On the basis of discussions with the Federal Land Manager, the closest
point that is likely to be visited within the Class I area is 58 km away
from the site (Point B).  The two dashed lines shown in Figure C-2, which
are drawn at 11.25° on opposite sides of the line connecting the plant
site and Point B, intersect the closest boundary at 40 and 44 km and the
most distant boundary at 117 and 90 km.  For conservatism, xmin is set at
40 km and xmax is set at 117 km.  Also for conservatism, Level-1 analysis
was performed using Point A, while Point B was used for Level-2 analysis.

The proposed project would cause elevated emissions from numerous process
points and ground-level emissions of fugitive dust.  (Estimated emissions
rates and particle-size distributions are shown in Table C-2.)  In the
Level-1 and -2 screening, for conservatism, all the elevated and ground-
based emissions were lumped together as 1f they originated from a single
source.  Thus, the particulate emissions were specified as the sum of the
process and fugitive emissions.   In the Level-1 analysis, Level-1 default
particle specifications were used rather than the known particle size
distributions.  Exhibit C-3 summarizes the VISCREEN analysis results.
Since integral vistas are not protected at this Class I area, only the
within-park impacts were relevant.  Even so, every case considered—
forward and backward scatter as well as sky and terrain viewing
backgrounds—showed an  impact exceeding the Level-1 screening criteria.
Thus, further screening and analysis were warranted.

The Level-2 analysis separately specified the process and fugitive emis-
sions with their known  particle-size distributions (while still assuming
the two plumes overlapped).  This was carried out by letting the primary
particulate signify the fugitive  emissions and the primary sulfate signify
the process emissions.  Particle  sizes were specified to agree with Table
C-2.  The  less severe worst-case  meteorology was found to be 0 and 1
m/s.  Exhibit C-4  shows that VISCREEN calculated impacts were not in
excess of  the screening criteria.  The marked difference in Level-1 and
Level-2 results arises  in part from the less conservative meteorology and
geometry of the Level-2 scenario.  A major factor also, however, is the
significant change in particle size characteristics used for the fugitive
emissions.
                                 C-9

-------
                    ***
    Input Emissions  for

       Participates      4
       NOx (as  N02)      2
       Primary  N02
       Soot
       Primary  S04
 Level-1  Screening
                                              ***
.93
.72
.00
.00
.00
MT /DAY
MT /DAY
MT /DAY
MT /DAY
MT /DAY
        **** Default  Particle  Characteristics Assumed

                  Transport  Scenario  Specifications:
        Background Ozone:
        Background Visual  Range:
        Source-Observer Distance:
        Min.  Source-Class  I  Distance:
        Max.  Source-Class  I  Distance;
        Plume-Source-Observer Angle:
        Stability:   6
        Wind  Speed:   1.00 m/s
                  .04  ppm
                60.00  km
                20.00  km
                20.00  km
                80.00  km
                11.25  degrees
                               RESULTS

    Asterisks (*)  indicate plume  impacts  that  exceed  screening  criteria

             Maximum Visual  Impacts  INSIDE  Class  I Area
                Screening Criteria ARE Exceeded
                                       Delta  E        Contrast
    Backgrnd Theta Azi  Distance  Alpha  Crit   Plume    Crit   Plume
SKY
SKY
TERRAIN
TERRAIN
10.
140.
10.
140.
145.
145.
84.
84.
28
28
20
20
.5
.5
.0
.0
24.
24.
84.
84.
2
2
2
2
.00
.00
.00
.00
18
4
27
4
.245*
.677*
.724*
.859*
.05
.05
.05
.05
.287*
-.186*
.279*
.134*
             Maximum Visual  Impacts  OUTSIDE  Class  I Area
                Screening Criteria ARE Exceeded
                                        Delta  E        Contrast
    Backgrnd Theta Azi  Distance  Alpha  Crit   Plume    Crit   Plume
     SKY      10.   10.     9.6   159.   2.00  22.273*
     SKY     140.   10.     9.6   159.   2.00   5.425*
     TERRAIN  10.   35.    15.9   134.   2.00  30.404*
     TERRAIN 140.   35.    15.9   134.   2.00   6.276*
                            .05    .346*
                            .05   -.224*
                            .05    .326*
                            .05    .190*
EXHIBIT C-3.  Level  1 screening analysis for  Example 2.
                          C-10

-------
TABLE C-2.  Estimated project emissions.
Emissions
                                      Emissions Rates
Parti oil ate Matter

   Process Sources                     0.395 MT/day
     (effective stack height = 50 m)
     DG = 1 um

     p  = 2 g cnf3

   Fugitive Emissions
     DG = 10 um

     p  = 2 g cm~^

Sulfur Oxides
     (effective stack height = 50 m)

Nitrogen Oxides
     (effective stack height = 50 m)
                                       4.54 MT/day
                                       7.26 MT/day


                                       2.72 MT/day
                         C-ll

-------
              ***  User-selected  Screening  Scenario  Results  ***
Input Emissions  for
   Particulates
   NOx (as  N02)
   Primary  N02
   Soot
   Primary  S04
   .54
   .72
   .00
   .00
   .40
MT /DAY
MT /DAY
MT /DAY
MT /DAY
MT /DAY
              PARTICLE  CHARACTERISTICS
              Density        Diameter
Primary Part.
Soot
Sulfate
2.0
2.0
2.0
       9
       1
       5
              Transport  Scenario  Specifications:
    Background  Ozone:
    Background  Visual  Range:
    Source-Observer Distance:
    Min.  Source-Class  I  Distance:
    Max.  Source-Class  I  Distance:
    Plume-Source-Observer Angle:
    Stability:    4
    Wind  Speed:   1.00 m/s
                     .04 ppm
                   60.00 km
                   58.00 km
                   40.00 km
                  117.00 km
                   11.25 degrees
                           RESULTS

Asterisks (*)  indicate plume  impacts  that  exceed  screening  criteria

         Maximum Visual  Impacts  INSIDE  Class  I Area
          Screening Criteria  ARE NOT  Exceeded
                                   Delta  E       Contrast
Backgrnd
SKY
SKY
TERRAIN
TERRAIN
Theta
10.
140.
10.
140.
Azi Distance Alpha
35.
35.
35.
35.
46.
46.
46.
46.
1
1
1
1
134.
134.
134.
134.
Crit
2
2
2
2
.00
.00
.00
.00
Plume
.657
.307
.724
.155
Crit
.05
.05
.05
.05
Plume
.003
-.012
.009
.006
         Maximum Visual  Impacts  OUTSIDE  Class  I  Area
          Screening Criteria ARE NOT  Exceeded
                                   Delta  E        Contrast
Backgrnd Theta Azi  Distance Alpha  Crit   Plume    Crit   Plume
                                                 .05    .008
                                                 .05   -.013
                                                 .05    .018
                                                 .05    .018
SKY
SKY
TERRAIN
TERRAIN
10.
140.
10.
140.
0.
0.
0.
0.
1.0
1.0
1.0
1.0
169.
169.
169.
169.
2.00
2.00
2.00
2.00
.802
.421
1.988
.636
   EXHIBIT  C-4.  Level  2 screening analysis  for Example 2,
                     C-12

-------
EXAMPLE 3:   PAPER MILL

A paper mill is proposed near a Class I area (see Figure C-3).  Anticipa-
ted paper mill emissions are shown in Table C-3.

The closest point in the Class I area is Point A, which is 7.8 km from the
mill.  However, Point B is the location in the Class I area that is
closest to the mill, relatively frequently visited, and unobstructed by
tree cover.  Point A was used for Level-1 screening and Point B for Level-
2 screening.

Although a plume-rise analysis shows that the plume from the  largest emis-
sion source (the power boiler) would not be at the same elevation as
plumes from other sources, and, thus, that plumes would not overlap, for
conservatism all emissions are lumped together as a single plume.  Exhibit
C-5 shows the result of Level-1 VISCREEN calculations for this plume and
the closest Class I area boundary.  With plume A£ values ranging from 10.2
to 25.7 for views against the sky (views of distant terrain were not pos-
sible at this Class I area), the screening clearly shows the  significant
potential for adverse plume visual impacts.  The plume contrast values
indicate that the plume would be bright (positive contrast) in forward
scatter (sun in front of observer) and dark (negative contrast) in back-
ward scatter (sun behind observer).

An analysis of on-site data indicated that the worst-case meteorology
could be characterized by F and 3 m/s, rather than the F and  1 m/s assumed
in Level-1 screening.  Exhibit C-6 summarizes VISCREEN results using this
meteorology and Point B geometry (see Figure C-3).  Although  impacts are
substantially lower (ranging from AE's of 4.0 to 8.6), they are still
considerably above the Level-2 screening criteria for both scattering
angles assumed.  Since the plume-rise analysis indicated that the plume
from the largest emitter at the mill would not overlap plumes from other
sources, a final analysis was performed with emissions from this single
largest emission source—the power boiler.  Exhibit C-7 summarizes the
VISCREEN results.  AE's range from 2.2 to 4.7, down considerably from the
more conservative Level-1 and -2 analyses, but still considerably in
excess of the screening threshold.  Thus, a Level-3 analysis would be war-
ranted in this case, and the possibility of adverse plume visual impact
could not be ruled out without additional analysis.
 EXAMPLE 4:  POWER PLANT IN THE WESTERN UNITED STATES

 A power plant  located  in the western United States north of  a Class  I  area
 was scheduled  to be expanded from two to four units of 400 MWe each.
 Table C-4 summarizes the emissions for the base and expanded scenarios for
                                     C-13

-------
    Paper   Mill
t
                    7.8 km
  N
   FIGURE C-3.   Relative locations of paper mill  and  Class  I area used in
   example 3.
                                         C-14

-------
TABLE C-3.  Paper mill stack emissions data.
                Stack    Stack     Exit       Exit
                Height   Diam.   Velocity     Temp.
                 (Ft)    (In)    (Ft/Sec)     (°F)
             Emissions
          (Metric Tons/Day)
        PM
SO,
NO,
Power Boiler    200      144     25.36
Recovery Boiler 275      114     94.06
Smelt Tank      250       72     23.00
Lime Kiln       260       50     26.02
  Total:
155     1.022
380      .491
155      .130
160      .087
1.756   2.027
4.069   1.560
 .064
 .091    .454
        1.72    5.97
        4.03
                                    C-15

-------
                       Level-1 Screening
 Input Emissions  for

    Particulates
    NOx (as N02)
    Primary N02
    Soot
    Primary S04
1.72  MT /DAY
4.03  MT /DAY
 .00  MT /DAY
 .00  MT /DAY
 .00  MT /DAY
     **** Default  Particle Characteristics Assumed

               Transport Scenario Specifications:

     Background Ozone:                  .04 ppm
     Background Visual  Range:         60.00 km
     Source-Observer  Distance:         7.80 km
     Min. Source-Class  I Distance:     7.80 km
     Max. Source-Class  I Distance:    13..00 km
     Plume-Source-Observer Angle:     11.25 degrees
     Stability:   6
     Wind Speed:   1.00 m/s

                           RESULTS

 Asterisks (*)  indicate plume impacts that exceed  screening criteria

          Maximum  Visual Impacts INSIDE  Class I Area
             Screening  Criteria ARE Exceeded
                                    Delta E       Contrast
 Backgrnd Theta Azi  Distance Alpha Crit  Plume   Crit   Plume
SKY
SKY
TERRAIN
TERRAIN
10.
140.
10.
140.
153.
153.
84.
84.
13.
13.
7.
7.
0
0
8
8
16.
16.
84.
84.
2,
2
2
2.
.00
.00
.00
.00
25.
10.
34.
5.
677*
235*
701*
013*
.05
.05
.05
.05
.201*
-.245*
.247*
.086*
          Maximum Visual  Impacts OUTSIDE Class I  Area
             Screening  Criteria ARE Exceeded
                                    Delta E       Contrast
 Backgrnd Theta Azi  Distance Alpha Crit  Plume   Crit   Plume
SKY
SKY
TERRAIN
TERRAIN
10.
140.
10.
140.
2.
2.
2.
2.
1.0
1.0
1.0
1.0
167.
157.
167.
167.
2
2,
2,
2,
.00
.00
.00
.00
31
3
52
16
.191
.757
.827
.779
*
*
*
*
.05
.05
.05
.05
.577*
- 337-
.597'
.564-
EXHIBIT  C-5.    Level  1 screening  analysis  for Examole  3,
                        C-16

-------
                  *** User-selected Screening Scenario Results ***
     Input  Emissions  for
        Particulates
        NOx  (as  N02)
        Primary  N02
        Soot
        Primary  S04
   .72
   .03
   .00
   .00
   .00
MT /DAY
MT /DAY
MT /DAY
MT /DAY
MT /DAY
                   PARTICLE CHARACTERISTICS
                   Density       Diameter
     Primary  Part.
     Soot
     Sulfate
2.5
2.0
1.5
       6
       1
       4
                  Transport Scenario Specifications:
         Background Ozone:
         Background Visual Range:
         Source-Observer Distance:
         Min.  Source-Class I Distance:
         Max.  Source-Class I Distance:
         Plume-Source-Observer Angle:
         Stability:   6
         Wind  Speed:   3.00 m/s
                     .04 ppm
                   60.00 km
                    9.30 km
                    8.00 km
                   13.00 km
                   11.25 degrees
                               RESULTS

     Asterisks  (*)  indicate plume  impacts that exceed screening criteria

             Maximum  Visual  Impacts  INSIDE  Class I Area
                 Screening Criteria ARE Exceeded
                                        Delta E       Contrast
     Backgrnd  Theta Azi  Distance Alpha Crit  Plume   Crit  Plume
SKY
SKY
TERRAIN
TERRAIN
10.
140.
10.
140.
144.
144.
47.
47.
13
13
8
8
.0
.0
.0
.0
25.
25.
122.
122.
2
2
2
2
.00
.00
.00
.00
8
3
15
1
.558*
.984*
.596*
.948
.05
.05
.05
.05
.062*
-.076*
.105*
.034
              Maximum  Visual  Impacts OUTSIDE Class I Area
                 Screening Criteria ARE Exceeded
                                        Delta E       Contrast
     Backgrnd  Theta  Azi  Distance Alpha Crit  Plume   Crit  Plume
SKY
SKY
TERRAIN
TERRAIN
10.
140.
10.
140.
1.
1.
1.
1.
1
1
1
1
.0
.0
.0
.0
167.
167.
167.
167.
2
2
2
2
.00
.00
.00
.00
19.
5.
36.
9.
745*
156*
760*
265*
.05
.05
.05
.05
.335*
-.204*
.403*
.294*
EXHIBIT  C-6.  Level  2 screening analysis for Example 3 (all  emissions)
                         C-17

-------
   Input  Emissions for

      Participates     1.02  MT /DAY
      NOx (as  N02)     2.03  MT /DAY
      Primary  N02       .00  MT /DAY
      Soot             .00  MT /DAY
      Primary  S04       .00  MT /DAY
                PARTICLE CHARACTERISTICS
                Density       Diameter

   Primary Part.     2.5            6
   Soot             2.0            1
   Sulfate          1.5            4

                Transport Scenario  Specifications:

       Background Ozone:                  .03 ppm
       Background Visual Range:         60.00 km
       Source-Observer Distance:         9.30 ton
       Min.  Source-Class I Distance:     8.00 ton
       Max.  Source-Class I Distance:    13.00 ton
       Plume-Source-Observer Angle:     11.25 degrees
       Stability:   6
       Wind Speed:   3.00 m/s

                             RESULTS

   Asterisks (*) indicate plume impacts that exceed screening  criteria

            Maximum Visual Impacts  INSIDE  Class I Area
              Screening Criteria ARE  Exceeded
                                     Delta E       Contrast
   Backgrnd Theta Azi Distance Alpha  Crit  Plume   Crit  Plume
SKY
SKY
TERRAIN
TERRAIN
10.
140.
10.
140.
144.
144.
47.
47.
13.
13.
8.
8.
0
0
0
0
25.
25.
122.
122.
2.
2.
2.
2.
00
00
00
00
4
2
10
1
.724*
.184*
.096*
.150
.05
.05
.05
.05
.041
-.044
.064'
.020
            Maximum Visual Impacts OUTSIDE Class I Area
               Screening Criteria ARE  Exceeded
                                     Delta E       Contrast
   Backgrnd Theta Azi Distance Alpha  Crit  Plume   Crit  Plume
SKY
SKY
TERRAIN
TERRAIN
10. 1.
140. 1.
10. 1.
140. 1.
1
1
1
1
.0
.0
.0
.0
167.
167.
167.
167.
2.
2.
2.
2.
00
00
00
00
14
3
29
6
.179
.630
.335
.406
*
*
*
*
.05
.05
.05
.05
.236*
-.144*
.306*
.192*
EXHIBIT  C-7.   Level  2  screening analysis  for
Example  3 (power boiler  emissions).
                       C-18

-------
TABLE C-4.   Emissions parameters for Example 4 Power Plant.
            Parameter
    Emissions per Unit
Unit 1 or 2   Unit 3 or 4
Stack height (ft)
             (m)

Flue gas flow rate (acfm)
                 (nr/sec)

Flue gas temperature (°F)
Particulate emissions
   Density (g/cnr)
   Mass median diameter (ym)
   Geometric standard deviation
   Flue gas concentration
      (tig/m3)
   Flue gas opacity (%)
   Mass emissions rate (g/sec)

   Nominal control efficiency (%)

Sulfur dioxide ($02) emissions
   Flue gas concentration (ppm)
   Mass emissions rate { g/sec)

   Nominal control efficiency (%)

Nitrogen oxide emissions
   Flue gas concentration (ppm)
   Mass emissions rate (as N02)  (g/sec)
      600
      183

   1,555,980
      734

      138
      332
     99.5
       93
      132

       80
      366
      372
   600
   183

1,555,980
   734

   138
   332
      2.0          2.0
      1.7          1.7
      1.5          1.5

    25,100       10,100
      20            9
      18.4          7.4
  99.8
   47
   66

   90
   314
   319
                                   C-19

-------
each boiler unit.  Figure C-4 summarizes the geometry of the plant, the Class I area, and
typical stable plume trajectories.  The Federal Land Manager was concerned about the view
from the observer location shown in this figure, because from this vantage point an observer
has an unobstructed view north, where a plume from the power plant would probably be
transported.  Since the vista of concern and the Class I area itself are both elevated relative
to the position of stable plumes, it was felt that stable plume transport  into  the Class I area
was unlikely, but that a view of a stable  plume, as shown in Figure C-4,  would be of concern.

Level-1 and -2 analyses were carried out using ViSCREEN.  These analyses indicated that
adverse visibility impairment could not be ruled out.  As a result, a Level-3 analysis was
performed. PLUVUE II was run for several plume transport scenarios to characterize the
cumulative frequency distribution of plume visual impact for mornings in the four seasons.
Since the calculated plume visual impact magnitudes were to be coupled  with  the cumulative
frequency of conditions worse than the indicated impact, plume positions for each wind
direction sector modeled were selected so that the plume impact was the  minimum for the
given sector (see Figure C-5).  Plume visual impacts  were  calculated as a function of azimuth
of view. The maximum plume AE (over all  the possible azimuths) was determined for each
plume transport scenario corresponding to given meteorological conditions. The individual
scenarios were ordered in descending value of AE. The cumulative frequencies for each
season were plotted and these results are summarized in Table C-5.  For  every season except
one (Fall, AE threshold = 5), the number of mornings which exceed the AE threshold are
greatest for Units 1 through 4.  On average,  the largest  number of mornings which exceed the
threshold AE occur in the winter, followed by fall, summer, and spring.
EXAMPLE 5:  CONSTRUCTION SITE NEAR A CLASS I AREA

A facility was proposed to be located only 1.9 km from the eastern boundary of a Class I area
(see Figure C-6).  Three phases of construction or operation were identified.  Each of these
phases (PI, P2, and P3) has its own set of emissions (see Table C-6).  Because diesel engines
were used during construction, emissions of NOX and soot were relatively high.  In addition,
fugitive dust emissions from the construction vehicles' disruption of the native soil were high.
However, these emissions would have relatively high particle sizes.

Level-1 and -2 screening was performed, using VISCREEN, for each of the three phases of
construction/operation. For every emissions, sun  angle,  and viewing background scenario,
impacts were calculated to be considerably in excess of the screening thresholds. Thus, a
Level-3 analysis was performed. Figure C-7 shows the  plume  trajectories that were modeled
for each of three observer locations.  Using the PLUVUE II model, a sensitivity analysis was

                                         C-20                            Revised 10/92

-------
                                               Elevated
                                               Terrain
                               Directions of
                               Stable Plume
FIGURE  C-4.  Location of Example  4 power plant relative to
Class I area.
                              C-21

-------
FIGURE C-5.  Plume trajectories corresponding to various wind
directions used in the visibility impact assessment.
                               C-22

-------
TABLE C-5.  Summary of the frequency of occurrence  of power plant plume visual
impact predicted for Example 4.
Number of Mornings with AE(L*a*b*) Greater
2.5
Season
Winter
Spring
Summer
Fall
Annual Total
Units
1 and 2
4
1
2
3
10
Units
1 through 4
6
2
3
5
16
Units
1 and 2
2
< 1
1
4
4
5
Units
1 through 4
3
1
1
2
7
than Indicated Value

Units
1 and 2
< 1
0
0
< 1
1
10
Units
1 through 4
1
0
0
< 1
< 2
                                         C-23
Revised 10/92

-------
Figure C-6.   Source and observer locations for Example 5.
                            C-24
Revised 10/92

-------
TABLE C-6.  Emissions used as PLUVUE-II input for three phases of construction and
operation (tons per day).
Phase
Phase 1 Construction (pi)
Phase 2 Construction (P2)
Phase 3 Operation (P3)
NOx
0.86
2.75
0.58
Diesel
Exhaust
0.06
0.28
0.01
Fugitive
Oust
0.15
0.E1
0.24
                                 C-25                           Revised 10/92

-------
Figure C-7a.  Plume orientations for which plume visual impacts were calculated from the
perspectives of individual observer—observer No. 1.
                            C-26
                                                             Revised 10/92

-------
Figure C-7b.   Observer No. 2.
                       C-27
Revised 10/92

-------
Figure C-7c.  Observer No. 3.
                           C-28
Revised 10/92

-------
carried out to determine the emitted species most responsible for plume visual impacts.  As
shown in Table C-7, all three species  (diesel exhaust or soot, NOX; and fugitive dust) were
important contributors;  however, soot and NOX appeared to be the largest contributors because
both species absorb light, which results in dark plumes. Because of the large number of wind
speed/wind direction/stability scenarios for which the plume would be visible, over 200
PLUVUE  II runs were  made. Table C-8 summarizes the output from one of these runs.  For
the west southwest wind direction, the plume perceptibility threshold (AE) is exceeded up to a
distance of 5 km, for west winds the AE threshold is exceeded up to  7 km, and for east
northeast winds the AE threshold is never exceeded.  The green contrast value never exceeds
the .05 threshold.

For each run the maximum AE was selected from all the lines of sight that were modeled.
Tables C-9 and C-10 summarize these maximum AE's.  AE's were ordered by descending
value (see Table C-11)  and coupled with frequencies of meteorological conditions  (see Table
C-12). Plumes were  predicted to be visible almost every day from observer location #1.
Plumes were also predicted to be visible from observer locations #2 and #3, but at lower
frequencies.
                                         C-29                             Revised 10/92

-------
TABLE C-7. Sensitivity of plume visual impact to emitted species.
Visual Range Blue-Red
Reduction (%) Ratio
Base Case
Diesel Exhaust Only
NOX Only
Fugitive Dust Only
15.2
9.8
5.7
1.7
0.987
0.988
0.998
0.996
Plume
Contrast
-0.016
-0.015
-0.011
-0.005
AE(L*a*b*)
0.641
0.586
0.497
0.175
Run Description:
  Spring  0800 AM
  Wind  direction  - E
  Wind  speed » 2  m/s
  Stability - D
  Observer II
  Emissions:  Phase Construction (PI)
  Downwind distance:   3 km
                                      C-30
Revised 10/92

-------
TABLE C-8.  Examples of PLUVUE-II output.
 EM1SS DBS   DATE   TIME STAB  WS    WD
                            (M/S)










WO
w










wo
ENE










1 12/21 0800
DISTANCE (KM)
SKY BACKGROUND
REDUCTION OF VISUAL
RANGE (%)
BLUE-RED RATIO

PLUME CONTRAST AT
0.55um
PLUME PERCEPTIBILITY
DELTA E(L*A*B*)


DISTANCE (KM)
SKY BACKGROUND
REDUCTION OF VISUAL
RANGE (%)
BLUE-RED RATIO

PLUME CONTRAST AT
0.55 urn
PLUME PERCEPTIBILITY
DELTA E(L*A*B*)


DISTANCE (KM)
SKY BACKGROUND
REDUCTION OF VISUAL
RANGE (%)
BLUE-RED RATIO

PLUME CONTRAST AT
0.55 ym
PLUME PERCEPTIBILITY
DELTA E(L*A*8*)
D
1.


.497
.919

-.032

3.110


1. .


.550

.918

-.030

3.203


1.


.215

.965

-.011

1.315
2 WSW
3.


.454
.937

-.028

2.492


3. "


.541

.935

-.024

2.629


3.


.063

.983

-.007

.585

5.


.500
.948

-.026

2.212


5.


.595

.945

-.023

2.352


5.


.133

.971

-.014

1.102

7.


.548
.960

-.024

1.873


7.


.642

.957

-.021

2.015


7.


.201

.970

-.017

1.243

10.


.646
.975

-.022

1.436


10.


.736

.972

-.018

1.574


10.


.312

.976

-.019

1.182

15.


.862
.992

-.018

.894


15.


.938

.989

-.015

1.007


15.


.533

.991

-.019

.859
                           C-31                          Revised 10/92

-------
      TABLE C-9.  Summary of maximum AE's calculated for each of the PLUVUE II runs for


      Observer #1.
n
f


p.
>—*
^3

to
                                   Phase 1                      '                Phase  2                    'Phase 3

Stab.
0
0
0
0
0
0
D
0
E
E
E
E
F
F
F
0
0
0
E
E
E
E
F
F
F
0
0
0
0
0
D
0
0
0
0
0
0
0
0
D

US.n/a
2
2
2
2
2
1
3
5
1
2
3
5
2
3
5
1
3
5
I
2
3
5
2
3
5





2
2
2
2
3
3
3
5
5
5

UO
E
usu
u
ENE
UNU
U
U
U
U
u
u
u
u
u
u
ENE
ENE
ENE
ENE
ENE
ENE
ENE
ENE
ENE
ENE
USU
U
ENE
E
UNU
USU
U
ENE
E
USU
E
UNU
USU
E
UNU
Winter
8 an noon
1.2 0.9
3.1 2.3
3.2 2.3
1.3 0.9
I.I 0.8
S.0
2.4
1.6
4.9
3.0
2.3
1.6
4.0
3.0
2.1
2.1
1.0
0.7
1.9
1.2
0.9
0.7
1.4
1 . 1
0.6















Spring Sunner Uinter
4 pn 8 an noon 4 pn 8 an noon 4 pn 8 am noon 4 pn 8 an
1.2 I.I 0.9 1.0 0.9 0.9 1.8 2.5 2.1
3.0 2.7 2.2 2.5 2.1 2.3 4.2 5.7 4.8
3.0 2.2 2.5 2.1 2.3 4.3 5.6
1.2 0.9 1.0 0.9 1.9 2.4
1.0 0.8 0.9 0.7 0.8 2.E 1.9 2.4

4.6
3.3








2.0
1.4







8.2
8.4
3.8
3.6
3.9
5.7
5.9
2.5
2.4 2.1
4.5
1.9
2.0
3.3
1.3
1.4
Spring Sunner Uinter
noon 4 pn 8 an noon 4 pn 8 an
1.8 1.9
4.1 4.8 4.4
4.1 4.8 4.0 4.4
1.8 2.0 1.9
1.8 2.1 1.8 1.9 0.6

1.0
0.7








0.4
0.3







2.0
2.0
0.8
0.B
0.9
1.3
1.3
0.5
0.5
1.0
0.4
0.4
0.7
0.3
0.3

-------
TABLE C-10. -Summary of maximum AE's calculated for each of the PLUVUE-II runs for
Observers #2 and #3 for each phase.

Stab.
0
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D

US.w/s
2
1
L-
"1
4_
2
i
t~
•^
i.
2
~\
t.
2
•n
t.
1
1
\
i
1
t
3
3
3
3
3
3
5
5
5
5
5
5
1

UID
NNU
N
NNE
NE
ENE
E
ESE
SE
SSE
S
NNU
NNE
SE
NNE
E
ESE
NNW
ENE
BE
NNE
E
ESE
NNU
ENE
SE
NNE
E
ESE
ENE

PI
0.5
0.4
0.4
0.5
1 .0

0.4
0.2
0.2
0.2
0.7
5.4
3.2



0.3
0.7
0.2



0.2
0.5
0. 1




Obs.*2
P2
1 .0



2. t


0.9


1 .6

17. 1



1 . 1
1 .7
7.9



0.B
2.0
1 .2



24.8

P3




0.5


0.2


0.4

17.3



0.7
0.5
0.4



0.4
1 .S
1 .0



25.2

Pi
0.5
0.4
0.4
0.4
(2.4
0.7
4.4
0.5
0.3
0.2



0.6
3.3
10.0



0.3
0.5
3.3



0.2
0.4
2.3

Obs.*3
P2






6.6




1 .2


18.0
27.3



0.6
1 .0
5.3



0.4
1 .2
3.9


RO


0.2


0.4
2.7






0.3
18.2
28.5



0.2
0.4
2.1



0.1
1 .0
3.0

 NOTE:  All  runs  performed with a winter morning (080Q)
        sun  angle.
                                      C-33
Revised 10/92

-------
TABLE C-ll.  Transport scenarios ordered by maximum plume AE for each observer location
and phase of construction and operation.
TABLE
Stab. US,
D
E
F
D
D
F
E
D
E
F
D
E
E
D
F
D
D
E
F
D
D
E
F
D
E
3-5
m/s
1
1
2
2
n
J
n
L
3
3
5
1
1
5
5
2
2
2
2
3
n
f.
3
3
S
5
5
3
WD
U
W
U
U
WSU
W
U
U
U
U
ENE
ENE
U
U
ENE
ENE
E
ENE
ENE
UNU
ENE
ENE
ENE
ENE
ENE
TABLE 3-5 d
Stab. US,
0
0
D
0
D
D
D
D
D
D
D
D
D
0
0
D
D
D
m/s
1
1
2
1
3
«1
l»
5
2
-1
4
2
2
3
•7
C
3
5
2
2
5
UD
NNE
SE
ENE
NNU
ENE
NE
ENE
NNU
ESE
N
NNE
NNU
S
SE
NNU
SSE
SE
SE
Obs.tl
Pi
5.0
4.9
4.0
3.2
3. I
3.0
3.0
2.4
2.3
2.1
2.1
1 .E
1.S
1,6
1 .4
1 .3
1 .2
1.2
1 . 1
1 .1
1 .0
0.9
0.8
0.7
0.7
Obs.12
Pi
5.4
3.2
1 .0
0.7
0.7
0.5
0.5
0.5
0.4
0.4
0.4
0.3
0.2
0.2
0.2
0.2
0.2
0.1
TABLE3-5b ob..t.
Stab. US.m/s
D
D
D
D
D
D
0
D
D
D
D
D
D
D
D
D
D
D
D
D
1
1
2
">
f-
3
3
1
1
1
5
5
2
2
2
3
Z
3
5
5
5
UD
U
USU
U
WSU
U
usw
UNU
ENE
E
U
USU
UNU
ENE
E
ENE
UNU
E
ENE
UNU
E
P2
8.4
8.2
5.9
5.7
4.6
4.5
3.9
3.8
3.S
3.3
3.3
2.6
2.5
2.4
2.0
2.0
1 .9
1 .4
1 .4
1.3
                                      TABLE  3-5f   Oba.»2
Stab.
0
D
D
D
D
D
D
D
D
D
D
US ,n/s
1
1
5
5
3
2
3
3
1
5
i
UD
ENE
SE
ENE
SE
NNU
ENE
ENE
SE
NNU
NNU
SE
P3
25.2
17.3
1 .6
1.0
0.7
0.5
0.5
0.4
0.4
0.4
0.2
1
Stab.
D
D
D
D
D
D
D
D
D
D
D
D
D
0
D
D
D
D
D
D
ABLE 3-
WS.ii/s
1
1
n
4
2
3
3
1
1
1
5
5
2
2
2
3
3
3
5
5
5
5c
WD
USW
W
WSW
W
W
usw
UNW
E
ENE
USU
W
UNU
E
ENE
UNW
E
ENE
ENE
E
UNU)
Dbs.t!
P3
2.0
2.0
1 .3
1 .3
1 .0
1 .0
0.9
0.8
0.8
0.7
0.7
0.6
0.5
0.5
0.4
0.4
0.4
0.3
0.3
0.3
TABLE 3-5e   Obs.*2
                                                                Stab.  US,m/s   WD   p2
                                                                     D
                                                                     D
                                                                     D
                                                                     D
                                                                     D
                                                                     D
                                                                     D
                                                                     D
                                                                     D
                                                                     0
                                                                     D
                                                                     D
1
1
3
-i
5
3
1
5
3
2
L.
5
         ENE
          SE
          SE
         ENE
         ENE
         ENE
         NNW
          SE
         NNW
         NNW
          SE
         NNW
24.8
17. 1
 7.9
 -1 i

 2.0
 1 .7
 1 .6
 1 .2
 t . !
 1 .0
 0.9
 0.6
                                           C-34
   Revised 10/92

-------
TABLE C-11.  Concluded
      TABLE 3-5g  Obs.ts
Stab.
D
D
D
D
D
D
D
D
D
D
D
D
D
D
0
D
D
D
D
WS.m/s
1
2
3
1
5
2
1
n
4.
2
3
5
-)
*•
2
2
2
3
-l
i.
5
•?
i.
WD
ESE
ESE
ESE
E
ESE
E
NNE
SE
NNW
E
E
NE
NNE
ENE
N
NNE
SSE
NNE
S
Pi
10.0
4.4
3.3
3.3
2.3
0.7
0.6
0.5
0.5
0.5
0.4
0.t
0.4
0.4
0.4
0.3
0.3
0.2
0.2
      TABLE 3-5h  Obs.*3




Stab. WS.n/s   WD  P2
D
D
D
D
D
D
D
D
D
D
1
1
2
3
5
5
1
3
3
5
ESE
E
ESE
ESE
ESE
£
NNE
E
NNE
NNE
27.3
18.0
6.6
5.3
3.9
1 .2
1 .2
1 .0
0.6
0.4
     TABLE  3-5i    Qbs.«3



Stab. WS,m/s   WD   P3
• D
D
D
D
D
D
D
D
D
D
D
D
I
1
5
2
3
5
3
L.
]
3
2
5
ESE
p
ESE
ESE
ESE
E
E
E
NNE
NNE
NNE
NNE
25.5
18.2
3.0
2.7
2. 1
1 .0
0.4
0.4
0.3
0.:
0.2
0. 1
                                        C-35
                                        Revised 10/92

-------
TABLE C-12. Frequency of worst-case morning plume AE's for observers #1, #2, and #3 in
Class I area.
                            Delta  E
Wind
Speed
( n/s )
Wind
Direction

Max
PI
•

Avg.

P2
Max .
P3
Avg.
Max.
Avg.
OBSERVER $1
I
1
2
3
5
WSW.W,
NE...
'NE...
NE...
NE.. .
WNW
SE
SE
SE
SE
5.
4.
3.
1 .
0.
0
9
0
0
7
4.8
4.7
2.8
0.9
0.6
8
3
2
1
1
.2
.6
.4
.9
.3
8.0
3.6
2.4
1.9
1.3
2
0
0
0
0
.0
.8
.5
.4
.3
1 .9
0.8
0.5
0.4
0.3
OBSERVER #2
1
1
->
4.
3
5
ENE ,E .
NE...
NNE...
NNE.. .
NNE ...
ESE
SE
SSE
SSE
SSE
5.
3.
1 .
0.
0.
4
2
0
7
5
1 .5
0.9
0.2
0.2
0.1
24
17
7
2
1
.8
. 1
.9
->
.2
5.4
4.2
1 .9
1 .2
0.5
25
17
1
1
0
.2
.3
.6
.0
.4
4.5
3.6
0.4
0.3
0.2
OBSERVER #3
1
1
2
3
5
SE.ESE
NE...
NNE. .
NNE. .
NNE. .
,SSE
SSE
.3
.S
.S
10.
4.
3.
0.
0.
0
4
3
5
2
3.9
1.5
1 .0
0.2
0.2
27
18
6
0
0
.3
.0
.6
.6
.4
8.1
4.0
3.0
0.5
0.4
28
18
2
0
0
.5
.2
.7
1
• *i
. 1
6.1
3.3
1.0
0.2
0. 1
                                                     Frequency of Occurrence(%)
                                                     « « — JT» «V •« «• » •*• ^ _V*^HMWflB _» « _ ^ M

                                                     Ann. Wint.Spr. Sun.  Fall
                                                      9.8 17.B  3.4  3.7  13.B
                                                     31 .4 49.B 12.9  15.4  45.9
                                                     65.0 80.3 42.5  59.8  77.4
                                                     77.8 84.0 60.3  80.8  87.0
                                                     86.1  86.9 74.3  93.0  91.4
                                                      1.0  0.8  0.1
1.4   1.9
2.3
                                                      2.2  2.4  0.4  t.o   £.
                                                     14.2 11.1  9.8 23.8   5.8
                                                     17.4 15.1 14.3 29.9  16.3
                                                     19.0 15.5 16.3 34.1  16.3


1
2
2
3
8
2
.3
. 2
.3
.2
2
2
12
14
.5
.6
.8
.0
1
1
12
17
.1
. 1
.4
.3
2
4
29
36
.3
.7
.0
.9
3
4
20
22
.3
. 1
.3
"1
                                                     24.0 15.2 19.2 41
      1.5
                                         C-36
                                                                       Revised 10/92

-------
                                   Appendix D

                              VISCREEN LISTING
      The source code is now made available through the OAQPS Technology Transfer
Network SCRAM Bulletin Board (919-541-5742).
                                      D-l                           Revised 10/92

-------
                Appendix I-





DISPERSION PARAMETER CALCULATIONS
                                              Revised 10/92

-------
                                      Appendix E

                    DISPERSION PARAMETER CALCULATIONS
       Equations that approximately fit the Pasquill-Gifford curves (Turner, 1970) are used to
calculate ay and O7 (in meters) for the rural mode.  The equations used to calculate ay are as
follows:

            oy = 465.11628 (x) tan(TH)(E-l)

where:
            TH = 0.017453293 [c - d ln(x)](E-2)

In Equations (E-l) and (E-2) the downwind distance x  is  in kilometers and ay is in meters.
The coefficients c and d are listed in Table E-l.  The equation to calculate Cz is as follows:
            a, = axb(E-3)

where the downwind distance x is in kilometers and az is in meters. The coefficients a and b
lire given in Table E-2.
                                          E-l                             Revised 10/92

-------
                             TABLE
        PARAMETERS USED TO CALCULATE PASQUILL-G1FFORD av


Pasquill
Stability
Category
A
B
C
D
E
F
C7y = 465.11628
TH = 0.017453293 [c
C
24.1670
18.3330
12.5000
8.3330
6.2500
4.1667
(x)tan(TH)
- d ln(x)]
d
2.5334
1.8096
1.0857
0.72382
0.54287
0.36191
where a  is in meters  and x is in kilometers
                                E-2
Revised 10/92

-------
                      TABLE E-2
  PARAMETERS USED TO CALCULATE PASQUILL-GIFFORD CTZ
Pasquill
Stability
Category x (km)
A* <.10
0.10 - 0.15
0.16 - 0.20
0.21 - 0.25
0.26 - 0.30
0.31 - 0.40
0.41 - 0.50
0.51 - 3.11
>3.11
B* <.20
0.21 - 0.40
>0.40
C* All
D <.30
0.31 - 1.00
1.01 - 3.00
3.01 - 10.00
10.01 - 30.00
>30.00
a z( meters) = ax*1
a
122.800
158.080
170.220
179.520
217.410
258.890
346.750
453.850
**
90.673
98.483
109.300
61.141
34.459
32.093
32.093
33.504
36.650
44.053
(x in km)
b
0.94470
1.05420
1.09320
1.12620
1.26440
1.40940
1.72830
2.11660
**
0.93198
0.98332
1.09710
0.91465
0.36974
0.81066
0.64403
0.60486
0.56589
0.51179
If the calculated value of a  exceed 5000 m, a  is set to
5000 m.

crz is  equal to 5000 m.
                          E-3
Revised 10/92

-------
                 FABLE E-2 (Continued)
PARAMETERS USED TO CALCULATE PASQUILL-GIFFORD a.
Pasquill
Stability
Category x
E <
0.10
0.31
1.01
2.01
4.01
10.01
20.01
(km)
.10
- 0.30
- 1.00
- 2.00
-4.00
- 10.00
- 20.00
- 40.00
>40.00
F <
0.21
0.71
1.01
2.01
3.01
7.01
15.01
30.01
.20
- 0.70
- 1.00
- 2.00
-3.00
- 7.00
- 15.00
- 30.00
- 60.00
>60.00
az( meters) = ax3
a
24.260
23.331
21.628
21.628
22.534
24.703
26.970
35.420
47.618
15.209
14.457
13.953
13.953
14.823
16.187
17.336
22.651
27.074
34.219
(x in km)
b
0.83660
0.81956
0.75660
0.63077
0.57154
0.50527
0.46713
0.37615
0.29592
0.81558
0.78407
0.68465
0.63227
0.54503
0.46490
0.41507
0.32681
0.27436
0.21716
                         E-4
Revised 10/92

-------
                                   TECHNICAL REPORT DATA
                                 ad Ir.Mrjciioia on me reverse Pc/ore LOmniclingl
I  H M1 f •) H T N O
  HPA-454/R-92-023
                                                           J RECIPIENT'S ACCESSION NO.
4 I I TLk AND SUBTITLE
 Workbook for Plums Visual Irapact Screening  and Analysis
              5 REPORT DATE

                October  1992
                                                           6. PERFORMING ORGANIZATION CODE
7  AUTHOR(S)
                                                           8. PERFORMING ORGANIZATION REPORT NO
9. PERFORMING ORGANIZATION NAME AND ADDRESS
  Sigma Research Corporation
  196  Baker Avenue
  Concord, MA  01742
                                                           10. PROGRAM ELEMENT NO.
              11. CONTRACT/GRANT NO.

                Contract 68D90067
                Work Assignment WA 3-3
12. SPONSORING AGENCY NAME AND ADDRESS
                                                           13. TYPE OF REPORT AND PERIOD COVERED
  Office of Air Quality Planning and Standards
  Technical Support Division
  U.S.  Environmental Protection Agency
  Research Triangle Park, NC  27711	
              14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
  This document is a revision of the Workbook for Plume Visual Impact Screening and
  Analysis,  EPA-450/4-88-015.   Work Assignment Manager:  Jawad S.  Touma
16. ABSTRACT
  The Prevention of Significant Deterioration and visibility regulations  of the U.S.
  Environmental Protection Agency  (EPA) require the evaluation of  a  type  of visibility
  impairment which can be traced to a single  source or small group of  sources known as
  "plume blight."  This workbook presents current EPA guidance on  the  use of screening
  procedures to estimate visibility impairment due to plume blight and is an update
  and a revision to the earlier book.  It includes the screening model (VISCREEN)  that
  can be run on a personal computer.  The VISCREEN model is used for both Level-1 and
  Level-2 screening analyses, and  is designed to evaluate plume visual effects along
  multiple lines of sight across the plume's  length for two different  viewing back-
  grounds and for two different scattering angles.  It also provides for  the
  evaluation of the potential perceptibility  of plumes using recent  psychophysical
  concepts.  The workbook provides the technical basis for the model and  contains
  several example applications to  illustrate  the use of these methods.  This document
  was issued as a draft for public comment and is now being revised  to reflect these
  comnents.
17.
                               KEY WORDS AND DOCUMENT ANALYSIS
                 DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS [c!COSATI Held/Group
  Air Pollution
  :ieteorology
  Air Quality Dispersion fiDdel
  Visibility
  Aerosols
  Nitrogen Dioxide
 New Source Review
 Air Pollution Control
    13B
     4A
     4B
18 DISTRIBUTION STATEMENT

  Release Unlimited
19. SECURITY CLASS (This Report!
 Unclassified
j 21 NO. OF PAGES
     150
                                             20. SECURITY CLASS (Tin's page]

                                               Unclassified
                                                                        22 PRICE
SPA Form 2220-1 (Rev. 4-77)
                      PREVIOUS EDITION
                                     OBSOLETE

-------