FIELD INVESTIGATION OF SULFUR DIOXIDE WASHOUT
            FROM THE PLUME OF A LARGE
COAL-FIRED POWER PLANT BY NATURAL PRECIPITATION

              J.M.  Hales, J.M. Thorp and M. A. Wolf

                          TO


          ENVIRONMENTAL PROTECTION AGENCY
            AIR POLLUTION  CONTROL  OFFICE


                CONTRACT NO. CPA 22-69-150
                       MARCH 1971
               ATMOSPHERIC SCIENCES DEPARTMENT
                BATTELLE MEMORIAL INSTITUTE
                PACIFIC NORTHWEST LABORATORY

-------
                           FINAL REPORT

                                ON



           FIELD INVESTIGATION OF SULFUR DIOXIDE WASHOUT
                     FROM THE PLUME OF A LARGE

          COAL-FIRED POWER PLANT BY NATURAL PRECIPITATION

              J. M. Hales, J. M. Thorp and M. A. Wolf


                                TO
                  ENVIRONMENTAL PROTECTION AGENCY
                   AIR POLLUTION CONTROL OFFICE
                    CONTRACT NO.  CPA 22-69-150
                            March 1971
                 ATMOSPHERIC SCIENCES DEPARTMENT

                  BATTELLE MEMORIAL INSTITUTE
                  PACIFIC NORTHWEST LABORATORY
                   RICHLAND, WASHINGTON 99352
Eattelle is not engaged in research for advertising, sales promotion, or
publicity, and this report may not be reproduced in full or in part for

-------
                                FOREWORD

This report was prepared by Battelle-Northwest, Richland, Washington
pursuant to Contract No. CPA 22-69-150 with the Air Pollution Control
Office, Environmental Protection Agency.  The Project Officer was Mr.
Charles R. Hosier, Division of Meteorology.

The research which is reported was conducted by scientific and technical
personnel of the Atmospheric Sciences Department of Battelle-Northwest.
Mr. M. A. Wolf was the Project Director and the principal investigators
were Dr. J. M. Hales and Mr. J. M. Thorp.  Battelle-Northwest personnel
who contributed to the investigation were:

                               W. S. Crews
                               F. 0. Gladfelder
                               D. W. Glover
                               V. T. Henderson
                               0. L. Jackson
                               E. W. Lusty
                               M. C. Miller
                               J. L. Rising
                               R. E. Wheeler
The authors especially want to acknowledge the major contributions of
Mr. Glover and Mr. Gladfelder toward the success of the field experiments.

Air Pollution Control Office personnel of the Air Resources Field Labo-
ratory at Jimmy Stewart Airport, Indiana County, Pennsylvania deserve
special acknowledgment.  Mr. F. A. Shiermeier contributed greatly to
the effectiveness of this study.  Mr. R. R, Seller and Mr. T. J.
Therkelsen provided continued and willing support in the field and lab-
oratory phases of the work.

The cooperation of Mr. Sam McKinney and Mr. Ken Gray of the Pennsylvania
Electric Company, and Mr. Jerry Bucher of the Indiana Tourist Promotion

-------
                           TABLE OF CONTENTS
             FIELD INVESTIGATION OF SULFUR DIOXIDE WASHOUT
                       FROM THE PLUME OF A LARGE
            COAL-FIRED POWER PLANT BY NATURAL PRECIPITATION
                                                                   Page
FOREWORD
LIST OF FIGURES
LIST OF TABLES
NOMENCLATURE
SUMMARY

CHAPTER I
CHAPTER II
CHAPTER III
CHAPTER IV
CHAPTER V
CHAPTER VI

REFERENCES

APPENDIX A
APPENDIX B
INTRODUCTION
DESCRIPTION OF THE FIELD EXPERIMENT
THEORETICAL ASPECTS OF GAS WASHOUT BY RAIN
REVIEW OF FIELD EXPERIMENTS
DISCUSSION OF RESULTS
CONCLUSIONS
APPENDIX C
FIELD DATA SUMMARY
REVERSIBLE WASHOUT -- AN EXAMINATION OF THE
CONSEQUENCES OF LINEARITY, NONLINEARITY, AND
WASHOUT THROUGH DROPLET CAPTURE
SIGN-X ANALYZER CALIBRATION
   i
   v
 vii
  ix
xiil

   1
   7
  23
  61
  83
 107

 111

 117

 207
 213

-------
LIST OF FIGURES
FIGURE
1.1
2.1
2.2
2.3
2.4
3.1
3.2
3.3
3.4
3.5
3.6
3.7
3.8
3.9
3.10
4.1
4.2
4.3
5.1(a)
5. Kb)
5.1(c)
5.2
5.3
5.4
5.5
5.6
TITLE
Plume-Atmosphere Interaction — Sulfur Compounds
Aerosol Washout Coefficients
Distribution of Fluorescein Aerosol Washout
Sulfur Dioxide Washout Coefficients
Sampling Grid
Representation of Simplified Film Theory
Trajectories of Various Sizes of Raindrops
Terminal Fall Velocity of Raindrops as a Function of
Diameter
Sherwood Number as a Function of Drop Size
Schematic of Transport and Washout Mechanisms in a Falling
Drop
Equilibrium Behavior of the Sulfur Dioxide - Water System ,
Experimental Data for Moderately Low Concentration
Absorption-Desorption Behavior for 0.03 cm Diameter Drops
Absorption-Desorption Behavior for 0.1 cm Diameter Drops
Absorption-Desorption Behavior for 0.3 cm Diameter Drops
Estimated Solubility of Sulfur Dioxide in Water
Keystone Area Map
Sulfur Dioxide Concentration in Air and in Precipitation,
Run 20
Sulfur Dioxide Air Concentration and "Dry Deposition,"
Run 21
Peak Sulfur Dioxide Washout Flux from Keystone Plume
Peak Sulfur Dioxide Washout Flux from Other Local Sources
Peak Sulfur Dioxide Washout Flux from Unidentified Sources
Selected Raindrop Spectra, Indiana County, Pennsylvania
Snowflake Character
Comparison of Peak Sulfur Dioxide Washout Concentration and
Collector Exposure Time
Sulfate Washout Fluxes
Rate of Sulfur Dioxide Oxidation
PAGE
3
10
11
13
15
26
31
35
36
38
49
51
52
53
59
66
81
82
89
90
91
93
96
98
104

-------
                            LIST OF TABLES
TABLE                           TITLE                            PAGE
1.1
3.1
3.2

5.1
5.2
5.3

5.4
Atmospheric Factors Which Influence Plume Interactions
Sulfur Dioxide - Air Properties Used in Calculating k
Overall Mass-Transfer Coefficients for Stagnant Falling
Drops
Peak Washout Concentrations and Fluxes
Sulfur Dioxide Washout Flux in Snowfall
Comparisons of Surface Concentrations Under Neutral
Stability Conditions
Fluxes of Washout Sulfate and Sulfur Dioxide
4
37

50
88
94

100
102

-------
                             NOMENCLATURE



a            Raindrop radius, t

c            Concentration  (subscripted) moles/£3

Ca           Space-curve notation

D            Diffusion  coefficient  (subscripted)  £2/t

E            Washout efficiency, dimensionless

f            Probability-density function  for drops in space, l/£

f            Probability-density function  for drops falling through a
             horizontal surface, l/£

g            Function denoting equilibrium relationship, arbitrary units
             depending  upon concentration  units chosen

H,H'         Henry's-law constants, dimensionless and £3/mole respectively

k            "Film" mass-transfer coefficient (subscripted).  k  denotes
             gas-phase  coefficient, kx denotes liquid-phase coefficient;
             moles/£2t

K            Overall mass-transfer  coefficient based on gas-phase driving
 ^           force, moles/£2t

Kl»K2'Ki'K2  Eclu:Llibrium constants  defined by (3.28), (3.29), (3.33),
             and  (3.34).  Units vary depending upon defining equation.

m            Mass-concentration of  particulate, m/£3, or total mass of
             particulate, m

N            Total number of raindrops in  a unit volume of space,
 0           number/^3

N            Total number of raindrops falling through a unit area per
             unit time, number/£2t

r            Radial position coordinate, L

r            Position vector, £

Re           Reynolds number of falling drop, dimensionless

Sc           Schmidt number, dimensionless

-------
 Sh          Sherwood number of falling drop, dimensionless

 t           Time coordinate, t

 u           Wind velocity vector, tit

 v           Raindrop velocity vector, tit
  Si

 v           Terminal fall velocity of drop,  -^ ,   tit

 w           Molar removal rate of pollutant  by washout from a unit
             volume in space, moles/£3t

 x           Downwind position coordinate, t.  Also liquid-phase mass
             fraction (subscripted),  dimensionless

 y           Crosswind position coordinate, t.   Also gas-phase mass
             fraction (subscripted),  dimensionless

 z           Vertical position coordinate, t.

 Z           Amount of pollutant carried  to ground  per unit area-unit
             time,  moles/£ t

 a           Linearized overall mass-transfer coefficient,  moles/£2t

 AY          Spacing between  collectors,  t

 &            Amount  of pollutant delivered to ground by a radius - a
             raindrop,  moles

 v            Kinemetic  viscosity, £2/t
                              SUBSCRIPTS


A           Denotes "species A"

aq          Denotes aqueous condition

b           Denotes bulk mixture

e           Denotes equilibrium condition

E           Denotes "effective"


-------
Denotes interface, or ground-level depending on application




Denotes liquid-phase




Denotes gas-phase

-------
                                SUMMARY

An experimental investigation was conducted in western Pennsylvania to
measure the sulfur dioxide washout by natural precipitation from the
plume of a large, coal-fired power plant.  This information is necessary
in assessing the extent of depletion from the diffusing plume and in eval-
uating the possible consequences arising from deposition at the surface.

The field investigation extended over three one-month periods in the
vicinity of the Keystone Generating Station where 11 to 23 tons of sulfur
dioxide, depending on power output, were released hourly from two 800 ft
stacks.  During rain episodes in October-November, 1969 and April-May,
1970, and during snowfall in February, 1970 there was a total of 22
experimental runs.

Precipitation samplers were placed at intervals of 4° to 12° on sampling
lines encircling the Keystone Station at nominal distances of 1-1/4,
2-1/2, and 4 miles.  The samplers were waste cans with approximately one
square-foot collection area.  A plastic funnel with an attached bottle
was fitted to the waste can for rain collection, while a plastic bag
liner was used for snow collection.  A solution of tetrachloromercurate
(TCM) was contained in the rain collector bottle to prevent sulfur
dioxide loss.  The TCM solution was omitted from the snow collectors on
the basis that low, ambient air temperatures would inhibit loss.

Chemical analyses were performed in a mobile field laboratory.  The sul-
fur dioxide analysis employed the sulfamic acid variation of the West
and Gaeke method.  Use of the Technicon Autoanalyzer permitted rapid
assay of samples containing as little as 0.1 micromole per liter of
sulfur dioxide.  Analyses for pH, sulfate and nitrogen compounds were
performed on selected samples.

Early experimental observations revealed that sulfur dioxide washout was

-------
 much  less than predicted by preexisting theory.  A revised theory was
 developed which  incorporates the physicochemical phenomena of  absorption-
 desorption reversibility and pH limited solubility of sulfur dioxide.
 The basis of  the revised theory is the expression of the sulfur  dioxide
 flux  from a falling drop as
                                           moles
                        (reversible process)
 rather  than  as
                                   i        moles
                            N, = - k  y.,  —5	
                             A      y *Ab  cm sec

                            (irreversible process)

 which had been employed in  the previous theory.  Here y^ is the mole-
 fraction of sulfur dioxide  in the gas surrounding the drop, c.   is  the
 average sulfur dioxide concentration in the drop, g denotes a solubility
 equilibrium relationship, and K  and k  are overall and gas film mass-
 transfer coefficients, respectively.   It should be noted that the  essential
 feature of Equation 1 is the parenthetic term which may be positive  or
 negative, thereby enabling  the treatment of both absorption and desorption;
 obviously this quality is not satisfied by Equation 2.

 Elaboration and extension of Equation 1 demonstrate that for the conditions
 of the Keystone study a great majority of sulfur dioxide absorbed  by the
 precipitation during its passage through the plume should be released
 before it reaches the ground.  Additionally, this theoretical treatment
 indicates more effective washout from low-elevation sources and significant
washout from tall stacks only at greater distances.

-------
A particularly interesting aspect is the influence of pH on sulfur dioxide
washout.  The reversible theory combined with solubility estimates indi-
cates that washout should vary strongly with pH over the range of levels
observed in the field.  Thus, acid-forming air pollutants sorbed by the
precipitation will tend to inhibit sulfur dioxide washout.

A further result of the theoretical analysis is that sulfur dioxide wash-
out through its capture by condensation droplets in the plume is an
unimportant mechanism compared to gas-phase washout.  Based on a supporting
field experiment and laboratory work by others, sulfur dioxide washout
through the mechanism of prior absorption on plume particulate is expected
to be insignificant.

The significant results of the experimental study, which are in agreement
with the revised theory, are:

     1.  Washout of sulfur dioxide was up to two orders of magnitude less
         than predicted by preexisting theory.  No significant difference
         was observed between washout fluxes in rain and snowfall; both
         were clustered about a value of 10 ymoles (n^hr)"1 for precipita-
         tion rate normalized to 1 mm hr"1.

     2.  The observed relationship between washout and precipitation rate
         differed by a factor of two from the relationship derived from
         preexisting theory.  The larger exponent of precipitation rate
         in the observed relationship with washout is interpreted as more
         efficient washout by large drops than is predicted by preexisting
         theory.

     3.  Washout concentrations of sulfur dioxide appeared to increase
         with distance from the stacks.  Preexisting theory predicts con-
         centrations proportional to the vertically-integrated mass of
         sulfur dioxide which should decrease with distance because  of
         lateral diffusion of the plume.


-------
      4.  Washout of background  sulfur dioxide appeared  to  cause an inor-
          dinately strong  interference with plume-washout observations.
          Remote sources such  as  the Homer City Generating  Station and the
          Lucerne coke  ovens,  both about 12 miles southeast of  the Keystone
          Station, contributed heavily to the sulfur dioxide washout.

      5.  Under certain circumstances, when the washout  of  background  sulfur
          dioxide was relatively  high, a depression of sulfur dioxide  con-
          centration was observed in precipitation collected beneath the
          plume.   This  indicated  that sulfur dioxide washout was inhibited
          in some way by the Keystone Station or that it was converted to
          sulfate.

 Comparison of  sub-plume samples with adjacent ones, which were  assumed
 representative of the  background, showed a higher sulfate washout beneath
 the plume.  Analysis of these cases indicates that this is not  a simple
 conversion from sulfur dioxide to sulfate.  Calculation of the  degree of
 oxidation required  to  produce the observed sulfate washout flux showed  that
 previously observed  oxidation rates could account for most of the observed
 values.  The highest values,  which occurred in snowfall, are attributed
 to high oxidation rates and enhanced washout of larger aerosols resulting
 from condensation of plume water vapor on the sulfate particles.   The
 sulfate washout  flux beneath  the plume in snowfall was observed to be
 proportional to  the  square of precipitation rate.

 By virtue  of physicochemical aspects of sulfur dioxide washout,  the maxi-
 mum  concentration in precipitation incident on the ground apparently  is
 reduced through elevation of the source by tall stacks.  However,  sulfate
 washout which may account for greater sulfur removal than sulfur  dioxide
 washout in the vicinity of the source is unaffected by the increased
 height of emission.  At greater distances,  as the plume diffuses  toward
 the ground total washout  of sulfur dioxide is expected to increase and
 to dominate the sulfur  washout processes.   Definition of the resulting
washout distribution remains to be determined.

-------
                               CHAPTER I

                             INTRODUCTION

This report describes an investigation of precipitation washout of sulfur
compounds from the plume of a large coal-fired, electrical generating
plant.  Such considerations are particularly important at the present
time because of the increasing national demand for electrical power,
which has been met in part by the advent of large, mine-mouth generating
stations.

Such operations, by virtue of the enormous quantities of coal they consume,
add considerably to the pollution load of the atmosphere.  Sulfur-compound
emissions, which arise from the combustion of sulfur impurities in the
coal, are of particular concern in this respect.  The most reliable esti-
mates indicate that about 40 percent of sulfur-compound emissions from
human activities arise from coal-fired generating facilities.     In the
three years between 1963 and 1966, the rise in sulfur-compound emissions
from coal-fired power plants amounted to roughly 2.3 million tons per
year (computed as sulfur dioxide) or an increase of 25 percent.

Assessment of the effects of such increases must necessarily consider the
sources, sinks, and dispersion of sulfur compounds as they relate to their
physical and physiological influences on the environment.  To accomplish
this, the Air Pollution Control Office (APCO) initiated the Large Power
Plant Effluent Study (LAPPES).  The precipitation washout study described
in the present report is one facet of the overall LAPPES program.

The LAPPES program is being conducted in western Pennsylvania in the
vicinity of the Keystone (1800 megawatts), the Homer City (1280 megawatts),
and the Conemaugh (1800 megawatts) Generating Stations.  The precipita-
tion washout study reported in this document was confined to the vicinity
of the Keystone plant, although the emission from Homer City was tenta-
tively identified as a background source.


-------
 PRELIMINARY ASSESSMENT OF PROBLEM

 Upon emission  of  the plume to the atmosphere, the sulfur compounds  that
 it contains are subject to a wide variety of physical interactions.
 Intuitively one might expect a given plume component, say a  sulfur  dioxide
 molecule,  to interact with its environment in a number of ways  prior to
 its ultimate deposition on the earth's surface.  Such interactions  may be
 represented as in Figure 1.1. This chart is incomplete in that  it does
 not include some  of the interactions that have been considered  to be
 relatively insignificant (e.g., formation of elemental sulfur by sulfur
 dioxide reduction); the figure is adequate, however, in describing  quali-
 tatively the more important interactions and in lending perspective  to
 the overall removal process.  It is important to note that although  the
 present study  is  concerned primarily with assessing net deposition  rates
 (steps  4-10 and 7-8, on Figure 1.1) these ultimate processes depend  upon
 and reflect all of the previous interactions.

 It is obvious  that the mechanisms depicted in Figure 1.1 will be influ-
 enced markedly by atmospheric conditions.  The factors considered to be
 most  significant in this respect are listed in Table 1.1. Visualizing  a
 superposition  of the atmospheric factors of Table 1.1 upon the  physico-
 chemical interactions of Figure 1.1 gives some indication of the complex-
 ity of  the processes resulting in the ultimate deposition of pollutant
 material.

 STUDY OBJECTIVES

The primary objective of this study was to assess the effectiveness  of
precipitation washout  as a mechanism for removing sulfur compounds from
the plume of  the Keystone plant.   This  assessment is necessary  to eval-
uate the significance  of washout  relative to increasing the delivery of
pollutants to the  earth's surface,  and  to decreasing the burden of sulfur
compounds in the atmosphere.  A  secondary objective of the study was
to evaluate the relative importance of  the various interactions and

-------
                                                           OXIDATION
   SO? SORBED
       ON
   PARTICIPATE
     MATTER
                                                    RAIN SCAVENGING,
                                                   . COAGULATION
                            NOIlVilOdVA3
                             'NOUdHOSJO
Figure 1.1    Plume-Atmosphere Interactions  — Sulfur Compounds

-------
                                TABLE 1.1

          ATMOSPHERIC FACTORS WHICH INFLUENCE PLUME INTERACTIONS

 I.     METEOROLOGICAL PARAMETERS
       A.  Temperature Field
       B.  Humidity Field
       C.  Wind and Turbulence Fields
       D.  Solar Radiation

 II.    PRECIPITATION PARAMETERS
       A.  Precipitation Rate
       B.  Raindrop Size Distribution
       C.  Snow Morphology
       D.  Precipitation Charge Distribution
       E.  Cloud Morphology

 III.   PLUME PERTURBATIONS
       A.  Buoyancy Effects
       B.  Stack Aerodynamics
       C.  Orographic Effects
       D.  Multi-Plume Interactions
atmospheric variables depicted in Figure 1.1 and Table 1.1. These  objec-
tives were to be accomplished by simultaneous field measurements of wash-
out rates and pertinent atmospheric parameters.

Chapter II describes the field experiment and gives the theoretical basis
for its design.  During an early phase of the experimental program, it
became evident that this preliminary theory was inadequate to  explain qual-
itatively the results observed, and a revised theory of sulfur dioxide
washout was developed.  The revised theory is presented in Chapter III.

-------
A review of the field experiments is presented in Chapter IV.  The experi-
mental data are consolidated in Appendix A.

Discussion of the field results in light of the revised theory of Chapter
III is presented in Chapter V and study conclusions are contained in the
final chapter.

-------
                              CHAPTER II

                  DESCRIPTION OF THE FIELD EXPERIMENT

The experimental approach was based primarily upon the method suggested
       (2)
by May.     This technique, which was employed for the measurement of
                                         (3 4)
gas and aerosol washout in earlier works,      involves the collection
of precipitation on a surface sampling line which extends across the
overhead plume.  The mass of material collected by the samplers is com-
pared with the total mass that has passed overhead during the sampling
period.  The basis for this approach and description of the area, sampling
grid, equipment, supporting measurements, and analysis techniques are
presented here.

BASIS FOR EXPERIMENTAL CONCEPT

May's technique of analysis is based upon the assumption that washout
occurs as a first-order, irreversible process.  This method can be
visualized most easily by considering a "puff" of plume of mass, m, that
has been emitted from the stack and is drifting downwind across the
sampling line.  By virtue of first-order irreversibility, washout may be
equated to the time rate of plume depletion, dm/dt, provided washout is
the only mechanism acting to deplete the plume.  This process can be
represented by

                               |r=-Atn       ,                  (2.1)

where the constant A is referred to as the washout coefficient.

In terms of the initial mass of the "puff" mQ, the downwind distance x
and the mean wind speed u, Equation 2.1 may be transformed to give

                         dm     /A    ,-»  -Ax/u                     ,0 Ov
                         -r- = -(A m /u) e           .             (2.2)

-------
 This relationship provides  an experimental means  for determining the wash-
 out coefficient for a particular  pollutant,  provided,  of  course, that the
 first-order irreversibility assumption  is valid.   The  measurement of dm/dx
 is accomplished by measuring the  total  mass  of pollutant  contained in pre-
 cipitation, collected in samplers spaced along an arc  around the source
 and extending beyond the edges of the plume.  For sampler spacing AY,
                              dm
                                                                     (2.3)
 where Em.  is the total mass  recovered  from  the samplers whose  individual
 collection area is A and m becomes  the total mass of plume  that  has passed
 overhead during the  sample collection  period.  For short distances  and
 small washout coefficients,  the exponential term of Equation 2.2 can be
 neglected  and
                                  u AY u/(moA)                       (2.4)
 where m  is  the initial mass of the plume emitted during the sampling
 period.

 An  independent calculation of particulate washout can be obtained  if the
 raindrop  size spectra are measured and the collection efficiency of the
 particulate by various raindrop sizes is known.  The efficiency is unity
 if  all  the particulate lying in the path of the raindrop is removed.   This
 relationship, when the particulate is monodisperse, is given by


                         A = it NT I   a2 E(a)f*(a)da                (2.5)
                                  Jo

where E(a) denotes collection efficiency and f*(a) and NT are  the  proba-
bility-density function and total number of drops falling through  a unit
area in unit time.

-------
The close agreement of Equations 2.4 and 2.5 has been shown    for a case
of washout of water soluble particulate by rain.  Owing to the solubility
of the particulate, it was assumed that it behaved similar to water drop-
lets of equivalent a2p, the product of the square of droplet or particle
radius and its density.  Collection efficiencies for the interaction of
water drops and water droplets measured by Kinzer and Cobb^   were used
together with measured rain spectra.  The resulting agreement is shown
in Figure 2.1.
Although the integrated mass from the precipitation collectors is used
for the calculation of the washout coefficient, there is an advantage in
analyzing each collector separately.  This permits better evaluation of
an experiment for it indicates the degree of plume containment, contamin-
ation and dry deposition.  Figure 2.2 shows the mass of the water-soluble
particulate which was recovered from each collector on arcs at 50 ft and
100 ft following the release of 44 grams during a period of 3 minutes in
a rainfall of 8.2 mm-hr"1.     Plume containment was complete.  The near-
ideal distributions indicate that contamination was at a minimum.  An
almost equivalent total mass on both arcs indicated that dry deposition
was also minimal.

The washout coefficient for gases also can be determined with this experi-
mental technique providing the initial assumption of a first-order,
irreversible process is valid.  The calculation of a washout coefficient
for gases from raindrop spectra utilizes the equation
                        4ir N   I   a D.  Sh(a)f (a) da               (2.6)
                            o  /      Ay
where a is the drop radius, D   is the diffusivity of the gas, Sh is the
Sherwood number, and f and N  are the probability-density function and
total number of drops existing in a unit volume of space.

-------
     10
       -2


o
o
*^
o
ra
                  Experimental Results for
               o Fluorescein (azp  -35 u^
                  Experiments on Olympic Peninsula
                  Theoretical Results Using Measured
                  Rain Spectra and Collection
               x Efficiencies of Kinzer and Cobb
                  for Monodisjperse Water Droplets
                             "Best Fit" Line by Engelmann
                             to Theoretical Results for Mono-
                                 disperse Water Droplets
                             (a2p =42 M^gm-cc'^and Rain
                              Spectra from Washington State
                                       and India
                                              10

                              Rainfall Rate (mm -hr"1)
50
            Figure 2.1    Aerosol Washout Coefficients

-------
200
—  ARC A (50 FT  RADIUS)
 —  ARC B (100  FT  RADIUS)
160
      60      40        20         0       20        40       60
       DISTANCE ALONG  ARC  FROM ARBITRARY  CENTERLINE.FEET

-------
 Comparisons  of washout coefficients for gases determined by Equations  2.4
 and 2.6  have been  less satisfactory, in general, than the previously noted
 comparisons  for particulates.  Figure 2.3 is an example of such  a compar-
 ison.  These experiments were conducted to determine the washout coeffi-
 cient  for  sulfur dioxide released from a heated gas cylinder mounted on a
 low tower.(8).  Collectors were located on Arcs A and B at distances of
 50 ft  and  100 ft.  Arc A experimental values were about 20 percent of  those
 calculated from the observed raindrop size spectra.  Closer agreement  was
 noted  on Arc B, but this was attributed to deposition of sulfur  dioxide on
 the collectors from the surface air.

 It appeared, therefore, that the initial assumption might not be completely
 valid  in the case  of sulfur dioxide washout.  Consequently, additional
 measurements beyond those required for particulate washout were  provided
 for the  Keystone experiments.  These measurements are described  in
 subsequent sections of this chapter.

 DESCRIPTION  OF AREA

 The Keystone Generating Station is located in western Pennsylvania about
 35 miles east-northeast of Pittsburgh.   Moderate, westerly winds  prevail
 and extremes of weather are infrequent.   Precipitation is fairly  evenly
 distributed  over the year,  averaging approximately two inches per month.
 The total annual precipitation for this  area is less than elsewhere in the
 Northeast.  During the field experiment  periods, it was observed  that  a
 large number of storms either passed to  the south or north of the area.

 Rolling country extends in  all directions for at least fifty miles from
 the Keystone Station with elevations ranging from 1000 to 1500 ft in the
vicinity of Keystone.   Much of the land  is tree covered with a variety of
hard and soft woods.   Indiana County,  immediately east of the Keystone
 Station is  known as the Christmas  Tree  Capital of the World.  There are
numerous  small  farms,  in addition  to the tree farms, scattered throughout
this area of  low population.   The  largest nearby population center is


-------
              • EXPERIMENTAL, QUILLAYUTE

              O THEORETICAL, COMPUTED FROM
                EXPERIMENT RAIN SPECTRA
u
O)
   10
     -4
UJ
o
            THEORETICAL BEST FIT LINE
              FOR  SELECTED SPECTRA,
            AFTER  ENGELMANN, et al.
                                                     21
                                                                   17A
                                                              16A
   10
     -5
0.1
     Figure  2.3
                                           1.0


                               RAINFALL INTENSITY (mm hr'l)


                          Sulfur Dioxide Washout Coefficients

-------
Indiana, which ie located about 10 miles to the east.  Students at Indiana
University of Pennsylvania constitute about half of its population of
nearly 25,000.

SAMPLING GRIP

The numerous small farms result in a dense network of roads.  This was a
fortunate circumstance for the field operations which required motor vehi-
cles for rapid deployment of collectors over relatively large distances.
Although it is desirable to have the collectors on an arc equidistant from
the Keystone Station, a certain amount of compromise was required by the
existing road pattern.  A collection arc of 90° was anticipated as suffi-
cient coverage for a single experiment, but collector locations were
selected to completely enclose the Keystone Station to provide coverage
for all wind directions.

These encirclements will be referred to throughout this report as arcs.
Arcs A, B, and C were established during the course of the experiments
at nominal distances of 1-1/4, 2-1/4, and 4 miles, respectively.  The
location of these arcs relative to the Keystone Station is shown in Fig-
ure 2.4.  Also shown are the collector locations which were spaced about
4° apart on Arcs A and B and at nearly 12° on Arc C.  Uneven spacing was
necessitated by local conditions which might interfere with successful
operation.

Collectors were placed at consecutive locations which spanned approximately
90° on Arcs A and B in the experiments, or runs, conducted during the first
field period in October-November,  1969.  Apparent failure to contain the
plume during that period resulted in extending the arcs to about 180° by
sampling every other location.   To extend the arc by doubling the number
of collectors would have resulted in increasing the deployment and
retrieval times,  thereby reducing the period for concurrent sampling by
all collectors.   The increased spacing was particularly necessary on Arc
C to provide coverage at that distance.

-------
                                                    30
                                                    •
                             ,." 1 "I''
                                        •
                                        •  •
                                       •   »
                                            \
                                     «


                                 50 .""
                                                         4
                                                         4
                         4
                         4
                         4
                                   40
                                                 «30
                            50 « 4  «
                           •    •
LEGENDS
  3K  RRC fl STRTIBNS
  <  flRC B STRTI0NS
  *  RRC C STRTIONS
  K DENOTES PLflNT  L8CRTIQN
                                        N
SCRLE
  I	
H=  1  MILE
                       Figure 2.4    Sampling Grid

-------
PRECIPITATION COLLECTORS

Different collectors were used for rainfall and snowfall.  In the case  of
rainfall, polyvinyl chloride funnels were fitted to waste cans of approxi-
mately 1000 sq cm collection area.  A polyethylene bottle containing 5  cc
of tetrachloromercurate (TCM) solution was attached to a perforated cap
on the stem of each funnel.  The bottle was removed from the funnel and
capped for transit to the laboratory upon retrieval.

The funnel was replaced by a polyethylene bag for the collection of snow.
The bag fitted into the waste can and snuggly covered its rim.  The TCM
solution was omitted on the basis that the low temperatures would inhibit
sulfur dioxide loss or chemical reaction.  Upon retrieval, the snow was
shaken into a corner of the bag and the open end gathered and tied
securely.

In addition to the primary collectors which were used for sulfur dioxide
and sulfate analysis, identical collectors, except for the TCM solution,
provided samples for pH, nitrate, and nitrite analyses.  These collectors
were located at the background sampling location and at two to four
locations on the arcs during the first field period when only pH was
measured.  The secondary collectors were located at alternate primary
sampling locations during the second and third periods.

ADDITIONAL GRID MEASUREMENTS

Emphasis was placed on the collection of precipitation samples, but addi-
tional measurements were made on the grid during each run to provide
support data.   At the background location and at as many as four arc
locations,  portable generators provided power for the operation of high
volume air samplers, bubblers, and precipitation characterization instru-
mentation.

The high volume samplers were provided to collect airborne sulfate on

-------
filter paper for determination of its concentration in air.  These were
located in standard NASN  (National Air Sampling Network) shelters.  Use
of the high-volume samplers was suspended after early experiments as it
became apparent that the sample sizes obtained were too small for reli-
able analysis.  The bubblers which were operated to determine sulfur
dioxide concentrations in air sampled at a rate of approximately 0.5
liters per minute, controlled by hypodermic needle orifices.  Raindrop
size and electric charge spectrometers were operated on the sampling
arcs and at upwind locations to establish background values.  The impor-
tance of drop size spectra is shown in Equations 2.5 and 2.6.  Charge
may be important if the sulfur is present as particulate.  During the
February period, snowflake character was recorded photographically.

WIND VELOCITY PROFILE

The introduction of heated pollutant into the atmosphere at a height of
800 ft necessitates the definitions of wind velocity to well above that
height.  This was accomplished using single-theodolite pibals released
from the Overlook — a visitor center located about 200 ft above the base
of the Keystone Generating Plant on a hill about one-half mile to the
southeast.  This location provided an unobstructed view of the sounding
balloon.  The 30 gram balloon was inflated to one-half standard free-
lift to reduce the ascension rate, and position readings were taken at
30 sec intervals.  A release was made prior to each run to estimate the
plume centerline for arc deployment and, insofar as practical, hourly
releases were made during the run for continued operational guidance and
documentation.

An average ascension rate was determined with double-theodolite soundings
during fair weather and this rate was assumed for subsequent single-
theodolite soundings.  Except during periods of heavy rainfall, little
deviation from the predetermined ascension rate was noted.  A problem
which did arise, particularly during the February series, was that low
ceilings and condensation from the Keystone Station prevented tracking
to the desired elevation.

-------
AERIAL AND MOBILE GROUND-LEVEL SAMPLING

Aerial sampling was conducted using a Cessna 172 aircraft equipped to mea-
sure and record sulfur dioxide concentration,  temperature, airspeed, humid-
ity, altitude, and time.   These variables  were recorded on magnetic tape
using a DL620A recorder manufactured by Metrodata Systems, Incorporated of
Norman, Oklahoma.  Temperature, airspeed,  humidity,  and altitude measure-
ments were made using a TVH26 instrument package, also manufactured by
Metrodata.  Sulfur dioxide was measured using an analyzer and scrubber,
manufactured by Sign-X, Inc.  of Essex, Connecticut.   Laboratory calibrations
of the sulfur dioxide analyzer are given in Appendix C.

The sampling probe for sulfur dioxide measurement was located on the leading
edge of the right wing, approximately four feet from the fuselage.  The
probe passed into the wing interior and then into the cockpit to the ana-
lyzer.  Humidity and temperature sensors were mounted on the right wing
strut.  The pitot-tube airspeed sensor was mounted on the left wing next to
the aircraft pitot-tube.

Plume sampling was conducted  by flying normal to the plume "centerline"
between two appropriate landmarks at a constant airspeed of 80 mph.  Tra-
verses were flown at vertical intervals of 100 and 200 ft.  During the
Spring phase of the study, it became increasingly evident that sulfur
dioxide concentrations near the ground were more critical in influencing
washout than were those aloft.  For this reason, the sulfur dioxide ana-
lyzer was adapted for mobile  ground-level  sampling and placed in a vehicle.
Sampling was accomplished by  driving along the sampling arcs and recording
sulfur dioxide levels at  each station.

Ground level sulfur dioxide concentrations on Arcs A and B were generally
too low to detect plume location with the  Sign-X analyzer; on Arc C, how-
ever,  the plume could usually be detected, although  peak sulfur dioxide
levels were generally only a  few hundredths of a part per million.

-------
All of the aircraft sampling equipment performed satisfactorily, but air-
craft operation was limited by low ceilings and low visibility during rain-
fall and snowfall.  Integral components of the aircraft instrumentation
were detained enroute to Pennsylvania for the third period investigation
by a Teamster wildcat strike.  This  instrumentation, as well as other
equipment, was not released during April and the aircraft was not employed
during that sampling period.

Airborne particulate matter was sampled on occasion using a Unico cascade
impactor.  Particle morphology and size distributions were similar to
                                                          (Q\
those observed in the comprehensive  study reported by MRI.

LABORATORY ANALYSES

With the exception of some pH measurements all chemical analyses of the
study were performed in a mobile field laboratory, located adjacent to the
APCO field office at Jimmy Stewart Airport  in Indiana County.  All of the
collected samples were analyzed for  sulfur dioxide content.   Selected
samples were analyzed for sulfate (SOT), nitrate (NC)7), nitrite (N0~),
and pH.

Upon their collection, the samples were delivered to the mobile laboratory
for analysis, as quickly as possible, to minimize deterioration.  If it
became necessary to postpone their analysis for more than a few hours the
samples were refrigerated.

SULFUR DIOXIDE ANALYSIS —  Sulfur dioxide was analyzed using the sulfamic
acid variation of the West and Gaeke method.      A 2.5 ml aliquot of
each sample was withdrawn using an individual hypodermic syringe and
transferred into a sampling cup of the Technicon Autoanalyzer for subse-
quent automated analysis.  This technique allowed measurement of samples
as small as 0.1 micromole sulfur dioxide per liter of water.   Tests showed
the results to be reproducible to within about 2 percent.   Overall error
of the method was expected to be less than about 5 percent.

-------
Aliquots  from several samples were withdrawn and analyzed at various  time
intervals  to check for deterioration.  These tests showed that no significant
loss  of sulfur dioxide occurred for several hours after sample collection.

Analysis  of snow  samples was conducted in a somewhat different manner owing
to  the absence of TCM solution during the collection period.  Here the
plastic bags which contained the samples were suspended in the laboratory,
and 2.5 ml of TCM solution was injected into each.  The snow was allowed to
melt  and  the bags were shaken to mix their contents.  Upon complete melting
2.5 ml aliquots were withdrawn from each bag and analyzed for sulfur  dioxide,
using the  techniques described above for rain samples.

SULFATE ANALYSIS — Sulfate content of the samples was analyzed using the
turbidimetric barium chloride method.      This technique was chosen  over
                                     (12)              (13)
the more  elegant barium chloroanilate     and titration     methods because
                                                                      *
it  was found to be the least affected by the presence of TCM solution.   A
Bausch and Lomb Spectronic 20 photometer was employed for the majority of
these analyses.

Sulfate results exhibited a large amount of scatter — behavior typical of
turbidimetric data.   The majority of this scatter, however, was suspected
to  have arisen from artifacts within the samples.   Scatter from purely
analytical factors is expected to be about 15 percent.

NITRATE ANALYSIS —   Nitrate levels of selected samples were measured using
the chromotropic acid method of West and Ramachandran. (14»15) This technique
was time-consuming and tedious,  but was felt to give quite reliable results.
Much of the scatter  that occurred in the results was suspected to have
occurred from artifacts  within the samples.

Nitrate measurements  were made only with samples that did not contain TCM
solution,  and every  effort was made to insure that these analyses were
*                 ~	
 The presence of  TCM solution was felt to be important here because it
tended to  prevent further sulfate production within the sample during
the time period between  collection and analysis.


-------
performed as soon as possible after sample collection.

NITRITE ANALYSIS —  Nitrite was analyzed using a variation of the well-
known Saltzmann method.  Few successful nitrite analyses were completed,
owing to its rapid decomposition.  A certain degree of success was achieved
finally by adding the Saltzmann reagent to the rain sample in the field
immediately upon collection; however, it appears that significant decompo-
sition may have occurred even under these circumstances.

Measurement of the pH of selected samples was accomplished using a pH meter.
Most of these measurements were taken in the field so as to minimize the
time between sampling and analysis.  During the fall 1969 series, pH mea-
surements were taken before and after sparging of the samples with nitrogen.
This gave inconclusive results and was discontinued for the duration of the
study.

-------
                              CHAPTER III

               THEORETICAL ASPECTS OF GAS WASHOUT BY RAIN

The profusion of theoretical publications in the meteorological literature
concerning precipitation washout of atmospheric particulates contrasts
sharply with the relative dearth of such material related to gas washout
phenomena.  Interpretation of the data obtained in this investigation
depends strongly upon the development of an adequate theoretical basis for
the analysis of below-cloud, gas washout processes.  It can be demonstrated
rather conclusively  (cf. Appendix B) that sulfur dioxide washout by rain-
capture of sulfur dioxide-containing fog droplets is insignificant compared
to gas washout through simple gas absorption by rain (cf. steps 2-3-4-10
and 2-4-10 in Figure 1.1).  For this reason Chapter III will deal only with
the simple gas washout process.

The first section deals with the individual rate processes contributing
to the overall gas washout phenomenon.  These are related to the net rate
of washout, and methods for estimating their magnitudes are described.
Subsequently, these methods are employed to define limiting situations in
an attempt to bracket behavior exhibited by real systems.  The section
concludes with a discussion of some of the laboratory methods used pre-
viously to measure gas washout, indicating their various advantages, limi-
tations, and applicability to the present study.

The second section of this chapter deals exclusively with thermodynamic
aspects, i.e., chemical reaction and solubility equilibria as they relate
to washout behavior.

The material presented here pertains specifically to the washout of sulfur
dioxide; however, its applicability is rather general and it can be used
as a preliminary basis for theoretical analysis of all systems in which
gas washout occurs.

-------
RATE PHENOMENA
DEFINITION OF MASS-TRANSFER COEFFICIENTS - It is customary to visualize
the phenomenon of gas absorption by a falling drop to occur as a consecu-
tive, two-step process.  The gas migrates first from the bulk medium to
the drop surface, whereupon it mixes into the interior of the liquid.
Such a visualization will be shown later to be highly superficial; ^it does,
however, suffice to provide a mathematical framework which can be "forced"
to fit physical behavior upon subsequent manipulation of the pertinent
parameters.

For trace-gas washout, this two-step process can be represented mathemat-
ically by the equations
          gas-phase step:
                          N   = - k (v   - v  )                      (3.1)
                          NAo     V Ab   7Ao'
          liquid-phase step:

                          N   = - k (x.  - x., )         .             (3.2)
                           Ao      x  Ao    Ab

 Here N    is  the average molar flux of material A"*" passing through  the
      Ao
 liquid  -vapor interface, and x. and y.  denote mole fractions  of A in the
 liquid  and gas phases, respectively.  The subscripts b  and  o pertain to
 bulk  (average) and interfacial conditions,  k  and k  are known as the
                                             y      x
 gas-phase and liquid-phase mass-transfer coefficients,  respectively.  The
 convention will be used that fluxes passing from the drop are  positive
 entities.
 *                                                                       (16)
  The development here  is based partially upon  that  given by Bird, et al.
  Nomenclature  used  here is  consistent with  that  employed by these authors.
 ^A flux is  defined  here as  the rate  of  the  passage  of material through a
  cross  section of unit area.

-------
Equations 3.1 and  3.2  find  their  theoretical basis  in the simplified film-
theory visualization as  characterized in Figure 3.1.   Here the two con-
secutive steps occur across the thin liquid and gaseous  films  bordering
the drop interface.  From Pick's  law of  diffusion

                                    dyA           dxA
                    N.   - - c  D,,  -=-= = -  c  D^, -r-^        ,       (3.3)
                     Ao      y  Ey  dr       x  Ex dr          '

which may be written

                   c D_                   c  D_
           «        y  Ey ,          x      x Ex ,         ..            / o / \
           NAo	6     (yAb - ^	g— (XA0 - XAb}    •       (3'4)
                     y                     x
which reduces to Equations 3.1  and  3.2  upon making  the  obvious  substitutions
of the mass-transfer  coefficients.   Here  D_,   and D  denote  effective
                                          liy       Jix
diffusivities in the  gas  and liquid  phases, and 6   and 6   are  the  liquid
and gas "film thicknesses."  c denotes total  molar concentration.  None
of the entities comprising k   and  k  should be expected to vary  appreciably
                            x      y
with the mole fraction  of A, hence this simplified film theory implies that
the mass-transfer  coefficients should  not vary with the concentration of
the gas being washed  out.  Such behavior  may  or may not be observed  in the
complex physical situations that occur in nature;  the  reasons  for  this
will be discussed  in  a  later section.
Usually Equations  3.1 and  3.2  are not  employed  directly  to  determine mass
transport rates because of the difficulty  in determining the interfacial
concentrations x.  and y.  .  This problem  is overcome by defining a new
coefficient K  based upon  the  overall  concentration driving-force between
the bulk gas and the bulk  of the liquid;
                          NAo - - Ky
Here y.  is the mole fraction of component A  that would exist in  the gas
      
-------
                                                                         BULK, TOTAL  CONCENTRATION = C,

                                                                     v  =Y
                                                                  .-'•••-. '«  ' /
ro
ON
I

                                                                          GAS  FILM - THICKNESS 6  ,

                                                                       ^EFFECTIVE DIFFUSIVITY D  £$
iGAS-LIQUlD  INTERFACES*
         DROP  INTERIOR, --—-^-—-""^j"  -
                                                                           r  =  a,  X  = x, Y
         TOTAL  CONCENTRATION = Cy _~- r_:~7


         -          -;=:--'~ - XA = xAb"-

                                                                  LIQUID  FILM - THICKNESS  6  ,|
                                                                                            X  ';
                                                                  EFFECTIVE DIFFUSIVITY  D

-------
known as the overall mass-transfer coefficient based on the gas-phase
driving force,  y   is determined from x   using equilibrium data, which
may be expressed in functional form as

                              yA = g(xA)           .                  (3.6)

In the special case where this relationship is linear, Equation 3.6
reduces to

                             yA = HXA              .                  (3.7)

                                                                (18)
Systems conforming to Equation 3.7 are said to obey Henry's law;
H is known as the Henry's-law constant.  It should be emphasized here that
solutions of sulfur dioxide in water do not obey Henry's law; equilibrium
behavior of such solutions is examined in more detail in the second section
of this chapter.

RELATIONSHIP BETWEEN THE MASS-TRANSFER COEFFICIENTS —  The equilibrium
relationship, Equation  3.6, can be employed to express the overall mass-
transfer coefficient in terms of k  and k  by combining Equation 3.1,
3.2, and 3.5.  The outcome is given by
                               y    x    y
m  denoting the slope of  the equilibrium curve "somewhere" between
and x. ; and
     Ao
                                   Ab    Ao
                                                                     (3'9)
For systems obeying Henry's  law m   is simply  the constant, H.

-------
Various types of washout behavior can be identified on the basis of Equa-
tion 3.5.  If K  is truly a constant independent of yAb and yAe> the
process is said to be first order.  If, further, A is annihilated completely
upon contacting the drop surface (total retention), yAe = 0 and the washout
process  is first orderT irreversible.  Often,  when Ky is not constant, the
process can be "linearized" by choosing some intermediate value of the mass-
transfer coefficient, a, such that

                         NA£) = - a (yAb - yAe)    , approximately.   (3.10)
Equation 3.10 is said to express the phenomenon as a psuedo first-order
process.

RELATIONSHIPS BETWEEN MASS-TRANSFER COEFFICIENTS.  WASHOUT RATES. AND WASH-
OUT COEFFICIENTS —  The ultimate objective of determining values of the
mass-transfer coefficients is their utilization to compute net washout
behavior.  The net washout rate, w, defined as the rate of removal from the
air of material A by precipitation within a unit volume element, v, can be
expressed in terms of the mass-transfer coefficient by summing the mass-
transfer rates for all of the drops in v.  Thus, if the precipitation is
composed of spherical, noninteracting drops
                               J°
                               Jo
            • 00
w = - 4irN   I   a2 f (a) N.  (a) da
         o  I            Ao
            'o
                             f   a2  f(a)Ky(a)(yAb- y^) da   ,      (3.11)
                            Jo
 The assumption of spherical drops  has  been shown elsewhere(9) to be suf-
 ficiently valid for the purposes of this  investigation.  Noninteraction is
 considered to be a reasonably valid assumption for most rain situations
 insofar as gas washout is  concerned owing to the high rates of diffusive
 transport relative to the  occurrence rate for drop interactions (i.e.,
 collisions, competition for material,  wake interactions, etc.).

-------
where a denotes drop radius, N   is  the  total number  of  drops  in v,  and
f(a) is the probability density  function  for raindrops  in v.

As noted in Chapter II, most previous analyses  of  atmospheric washout
processes have employed the assumption  of total retention by  the drop and
have expressed the washout rate  in  terms  of an  overall  washout coefficient,
A.;
                              w =  A.  yAb  cy          ,                (3.12)
                   (first-order,  irreversible  process).
Here the subscript  i has  been  added  to  emphasize  that  the assumption of
irreversibility has been  employed.   A.  can be  expressed  in  terms of the
individual washout  coefficients  for  drops of different sizes X. as follows,
provided that  the drops do  not interact.
                        L  =  N  /   f(a)  A.(a)  da     .                (3.13)
                         i     o  I         i
                               *0
From Equations  3.11,  3.12,  and  3.13,  the  following  expressions can be
obtained relating  the washout coefficient to  the  overall mass-transfer
coefficient, K  .
              y
                                       a2  K  f(a)  da                  (3.14)
                               y   Jo
                    (first-order  irreversible process)
                    (first-order  irreversible process)     .           (3.15)

-------
For the washout of gases,  irreversibility is generally a poor assumption
and the use of A± should be discouraged.   This can be overcome by redefining
the washout coefficient in terms of a reversible process, thus
                           w    r   Ab   ^Ae   y       '

illustrating that reversible washout can be treated equally well in terms
of mass-transfer coefficients or in terms of a properly-defined washout
coefficient.  Because of the danger of confusing reversible and irreversible
washout coefficients, the theoretical development is continued mainly in
terms of the mass-transfer coefficient.   This is in concordance with the
related field of chemical engineering, where most of the  pertinent work
regarding gas washout by falling drops has been performed up to the present
time.

The net amount of material A carried to the ground by washout can be related
to K , provided again that the assumptions of sphericity  and noninteraction
are valid.  For this purpose, it is convenient to visualize precipitation
being collected in a bucket located somewhere beneath the plume as shown
in Figure 3.2.  Drops of different sizes entering the collector will have
passed through the plume in different trajectories depending upon their
terminal velocities and interactions with the wind.  These trajectories may
be defined by the family of space curves C (x,y,z).  Also, the drop velocity
                                          3.
v (x,y,z) is defined as dr /dt where r  denotes a position vector origin-
 3.                        33
ating at the collector and tracing out the space curve C  .
                                                       3

For the present, it is assumed that wind parameters and plume concentrations
are fixed in time .  The total amount of A picked up by a drop of radius a
 Time fluctuations in wind parameters and plume concentrations will give
 rise to time-averaged values  of yA.   Such averaged values can be employed
 in this analysis in lieu of "instantaneous" measurements; however, such a
 treatment is totally valid for first-order systems only.  A discussion of
 this problem is given in Appendix B.

-------
I
OJ
                  SPACE CURVES Ca(x.y.z) DENOTING
                  TRAJECTORIES OF DIFFERENT-

                  SIZED DROPLETS

-------
during its time of passage, t , then is given by


                              fS 
-------
 It is emphasized here that Z is an instantaneous entity, and the total
 amount of A collected over a given sampling period must be expressed in
 terms of an additional integral in time.  If wind and plume parameters
 fluctuate appreciably during the sampling period, one must again contend
 with the problem of averaging as discussed in Appendix B.

 Equation 3.19 is an important result.   It indicates how K  is used to
 calculate washout to the earth's surface, and it shows some of the require-
 ments for determining K  experimentally in the field.   If  K  were known,
                        y                                   y
 one could employ Equation 3.19 to calculate Z on the basis of rain,  wind,
 plume concentration, and solubility data.  Generally the converse is not
 true.  That is,  by measuring Z in the  field,  one cannot determine K  on
                                                                    y
 the basis of Equation 3.19 alone, unless some additional information
 regarding the nature of K  is known.   This is a problem that arises  con-
 sistently whenever the interpretation  of differential  quantities  on the
 basis of integral data is attempted.   Equation 3.19 indicates, however,
 that data pertaining to NT,  f*(a),  yAfe,  VQ,  Z,  ZQ,  and solubility are
 necessary (but probably not  sufficient)  for determining K   from field
 measurements.

ESTIMATION OF INDIVIDUAL COEFFICIENTS ~  Equations 3.11 and 3.19 indicate
how the overall mass-transfer coefficient, K  , may be employed to deter-
mine the washout rates w and Z.  K  , in turn, depends upon the individual
coefficients k  and k  as indicated by Equation 3.8.  From this, it is
              x      y
apparent that the degree of success in analyzing gas washout behavior
should depend largely upon an ability to estimate these coefficients.  The
present section is addressed to this task.

Gas-Phase Mass-Transfer Coefficients —   The  transport  of gas to  the  sur-
face of a falling drop occurs primarily by diffusion and convection.  This
relative simplicity  contrasts sharply with complexities exhibited during
the process of aerosol washout where the  additional effects of inertia,
electrical charging, diffusiophoresis, and thermophoresis may become

-------
important/20'2*0  For the case of purely convective-diffusive transport
the gas-phase mass-transfer coefficients can be calculated from the general-
ized semiempirical expression
                         Sh
=2+0.6 Re1/2 Sc1/3
                                                                    (3.20)
where
                         Sh =
  2k  a
    y
  D.  c
   Ay  y
Sherwood number,
                         Re
                               2a v.
           =  Reynolds number,
and
                         Sc
           =  Schmidt number.
Here c  denotes total gas concentration,  D.   is the diffusion coefficient
in the gas, v  is the terminal velocity of the drop, and v is the kinematic
                                               "                      (22)
viscosity of air.  Equation 3.20 often is referred to as the Frbssling
equation.  It has been shown to describe physical behavior with reasonable
accuracy over all conditions of practical interest to the present investiga-
tion.<23)
Since the terminal fall velocity v  can be expressed as a function of drop
size (cf. Figure 3.3), the Sherwood number and, therefore, k  can be
represented in terms of drop size and the appropriate gas properties.  Such
a representation is provided in Figure 3.4, which is essentially a plot of
Equation 3.20, based upon the physical properties of sulfur dioxide and air
at 20°C, shown in Table 3.1.

-------
 -800
 -600 -
If)
•>•

o
O
o
 J-400 -
  -200 -
                         0.1                0.2

                           DROP  DIAMETER, cm
   1600
                                                                - 1000
                                                                -  500
                                                                         in
                                                                         o
                                                                         _i
                                                                         o
                                                                -  100
0.3
     Figure  3.3    Terminal Fall Velocity of Raindrops as a Function
                   of Diameter -  Data of Gunn and Kinzer'19)

-------
                    0.3
                    0.2
                  E
                  u
I
UJ
Q.

O
                    0.1
 10


SHERWOOD NUMBER
                                                               20
                                                                 30
                                                           2ka

-------
                               TABLE 3.1
          SULFUR DIOXIDE - AIR PROPERTIES USED IN CALCULATING k
             	                                      V
    Diffusion Coefficient   Kinematic Viscosity   Gas Concentration
          cm2/sec                cm2/sec             moles/cm3
           0.195                  0.1505            4.152 x 10~5
Liquid-Phase Mass-Transfer Coefficient - General Comments — The ease with
which gas-phase coefficients can be estimated is not matched by similar
behavior exhibited within drops.  The already complex processes of liquid
mixing in a falling drop and their relation to mass-transfer are compli-
cated even further by film effects at the interface and by the possibility
of chemical reaction within the liquid.  Such effects are shown pictorially
in Figure 3.5.

Surface film effects, which are expected to impede circulation within the
drop and offer an additional resistance to mass transfer, depend in a com-
pex way on trace impurities contained by the drop and upon specific inter-
actions between the dissolving gas and the liquid.  Chemical reaction in
the liquid phase also may complicate the analysis enormously depending on
the relative rates of mass transfer and chemical reaction.  There are
several reasons for this.  The phenomenological expression for the rate
of reaction may be highly complex — a problem that is compounded if
several reactions occur simultaneously.  Furthermore, chemical reaction
usually violates the consecutive behavior exhibited by the mass-transfer
steps of the process.  In addition, it results in the formation of new
species which subsequently mix and react further depending upon their own
physical nature.  Finally, because reaction rates often are related to
concentration in a nonlinear fashion, the transport expressions cannot be
evaluated reliably in terms of average liquid concentration as implied by
Equation 3.2.


-------
STREAMLINE
DROPLET SURFACE
CIRCULATION PATTERNS
                       CHEMICAL
                        REACTION
                                                   +r
LIQUID-PHASE
TRANSPORT
                                                      INTERFACIAL
                                                      TRANSPORT
                                                      GAS-PHASE
                                                      TRANSPORT
      Figure 3.5   Schematic of Transport and Washout Mechanisms
                  in a Falling Drop

-------
Obviously, such behavior is difficult to analyze mathematically.  Because
of the highly specific nature of these interactions, this behavior defies
any generalized empirical approach similar to that employed for gas-phase
transport.  The remaining portions of this section examine the signifi-
cance of each of the processes by considering first the problem of mass
transfer inside a drop where neither chemical reaction nor interfacial
resistance occur; the significance of these further complications then are
considered in view of the more simplified behavior.

Liquid-Phase Mass Transfer in the Absence of Chemical Reaction —  A fall-
ing drop experiences frictional drag on its surface which tends to induce
patterns of internal circulation (cf. Figure 3.5).  Such circulation,
combined with a variety of possible secondary mixing effects  (e.g.,
oscillations and thermal perturbations induced by the drop falling through
a temperature gradient), combine with molecular diffusion processes to
effect a transfer of material within the drop.  Such behavior is difficult
to analyze theoretically; some insight to its significance, however, can
be obtained by considering the two limiting cases of zero and infinitely-
rapid convection within the drop.

For the case of rapid convection, the liquid mass-transfer coefficient
becomes large.  Equation 3.8 then reduces to

                                 K  = k            ,                (3.21)
                                  y    y

which can be evaluated using Equation 3.20.  In the absence of irreversible
chemical reaction, this behavior characterizes the most rapid rate of mass
transfer possible, and establishes, therefore, an "upper limit" of behavior
exhibited under atmospheric conditions.

For the case of zero convection, diffusion is the sole mechanism for trans-
port.  If the diffusion coefficient of A in the liquid, DAX, is known, the
rate of transport in a drop can be determined by solving the appropriate
forms of the continuity equation(    for a binary mixture.  Such solutions


-------
are
    given in the- literature in terms of point concentrations within the
drop.(24).  These can be integrated over the drop volume to give
                                        (XA  (X)  - XA,) dX + x. .      0.22)
and
   N   = _ _^^  ,   x  (t) _ x   - txe'"1- /  e  (x  (X) - x  ) dX  ,
    Ao       a    / *  Aov      Ai        J       Ao       Ai
                   n=l
                                                                    (3.23)
where
                                  D
                              a = _p (m;)2           .              (3.24)
                                  a2

Equations 3.22 and 3.23 pertain to spherical drops initially containing
mole fractions x.. of A, which is distributed uniformly throughout the
drop interiors.  Surface concentration x.   may vary with time in any pre-
scribed manner.  Simplified solutions for constant x.   and/or short expo-
                                              (24 257°
sure times may be obtained from the literature   '   , or they may be
deduced from Equations 3.22 and 3.23.  In the absence of surface resistance,
Equations 3.22 and 3.23 represent the slowest rate of mass transfer possible.
These, therefore, combined with equations characterizing the rapid mixing
case provide upper and lower limits for washout behavior.  In the absence
of irreversible reaction and/or surface resistance, all atmospheric wash-
out behavior should fall somewhere between these two limits.

Nonlimiting  behavior, as mentioned previously, is difficult to analyze.
If circulation within the drops could be described exactly, one could
attempt to derive and solve a binary continuity equation for the system  and


-------
thereby predict N. .  Alternatively, one might employ the simpler but less
exact technique of applying surface renewal and penetration theory.   '
Unfortunately, however, theoretical attempts to describe flow and mixing
patterns within drops have met with somewhat limited success, a problem
resulting from physical complications such as drop oscillation, turbulent
flow, and nonidealities occurring at the gas-liquid interface.

For the case of small drops in the Stokes regime,  (Re «1), (cf. Figure
3.5), Hadamard^27)and Rybczynski'2  , (cf., Levich^29^) have derived an
equation characterizing drop circulation in the absence of vibration and
surface nonidealities.  Subsequently, Kronig and Brink     have formulated
an idealized transport equation based upon the Hadamard-Rybczynski form-
ula, and employed the solution to estimate mass-transfer rates.  Kronig
and Brink's solution, which depends upon the assumption that transport by
circulation is rapid compared to that by diffusion, indicates that the
liquid-phase mass-transfer coefficient is increased by a factor up to
2.5 if ample circulation is present.  Most experimental results, however,
                                                           (31 32 33)
indicate that for large drops this factor may be 10 or more   '  '    —
a considerable deviation which has been attributed to increased mixing
through oscillation of the drops as well as to extrapolation outside of
the Stokes regime.

Surface phenomena also are expected to play a rather important role in
                                     (29)
complicating mixing behavior.  Levich     employs the Hadamard-Rybczynski
equation to demonstrate that when minute quantities of surface-active
agent are present, this material is swept to the lee of the drop where it
congregates and impedes further circulation.  Surface agents also may
cause changes in drop oscillatory behavior; implications of such inter-
actions with regard to washout behavior in a contaminated atmosphere are
obvious.

"Interfacial" resistance over and above that normally exhibited by the
liquid phase may be an additional manifestation of surface-active materials

-------
within the drop.-  Such resistances have been measured experimentally for
various surface-active agents in conjunction with the sulfur dioxide water
system/34'35^  Because of the variety of possible types and quantities of
agents occurring in natural rain, however, such results are almost impos-
sible to utilize insofar as washout problems are concerned.

One should note that in the total absence of circulation, any film resis-
tance decreases the liquid-phase mass-transfer coefficient to a value below
that predicted by Equations 3.22 and 3.23, which was suggested previously
as a lower limit.   For most practical situations, however, net effects are
such that K  is larger than that predicted by Equations 3.22 and 3.23;
hence these equations are employed to give some tentative idea as to the
lower limit magnitude.

        The considerable amount of additional work that has been published
with regard to film effects and drop convection has been reviewed elsewhere
(23 32 33)
   '  '    and is not considered further here.  In view of the complexity
of this problem, it is apparent that the success of any related analysis
must depend rather heavily upon experimental measurements.

The Influence of Chemical Reaction —  Dissolution of gaseous sulfur
dioxide molecules  in water is accompanied by reversible chemical reaction.
The absolute nature of this reaction has been the subject of some con-
                                 /og\
jecture.  Earlier  it was supposed v  ' that the scheme
                            S02g + H20 t H2S03                       (3.25)
                                               HSO-                  (3.26)
                        HSO-  + H20 J  H30  + S0=                    (3.27)

-------
characterized true reaction behavior, K' and K' denoting  equilibrium
contants for the first and second ionization, respectively.
                                [H 0] [HSO~]
                          K
and
                                  [H2S03]
                                [H 0+][SO~]
                          Kl -- - - 1-          .               (3.29)
                                    ~
                                 [HSO~]
Subsequent evidence     has indicated, however,  that the compound H^SO,,
"sulfurous acid," does not exist to any appreciable extent in water
solution; rather, un- ionized molecules of dissolved sulfur dioxide appear
to take the place of H^SO., in Equations 3.25  to  3.29 to give the scheme

                        S02  + H20  t  H20 +  S02a                  (3.30)

                                    Kl    +
                       SO-   + H_0  £  H,0  + HSO-                 (3.31)
                         2aq    2.       o        j

                                    K
                        HSO~ + H20  2 H30+ +  S0°                   (3.32)
where
                                tH,0+] [HSOJ
and
                                [H 0
                          K  = -J - i            .               (3.34)
                                  [HS03]
Such evidence, which has been based largely upon various types of spec-
troscopic data, has indicated also the presence of trace amounts of
the pyrosulfite ion, HS20~, and various hydrates, S02  • nH20.  Other

-------
investigators,. (37) however, cite vapor-pressure data to argue that H2S03,
or some related species, must be present in solution.  At the present
time, this conflict is not completely resolved, although the opponents
of un-ionized sulfurous acid appear to have the bulk of evidence in their
favor.

Most of the available experimental evidence suggests that the velocities
of reactions 3.31 and 3.32 are rapid, although there are some indications
                                     (38)
to the contrary.   Wang and Himmelblauv  '  measured ionization kinetics
of reaction 3.31 using a radioisotope technique and concluded that this
reaction should proceed slowly,  approaching completion on the order of
one minute.  Practically all of the evidence in the remaining literature
is less quantitative in nature, but suggests that the results of Wang
                                                      (39)
and Himmelblau are in error.  Lynn and his co-workers,     for instance,
measured the conductivity of a rapidly mixed solution of sulfur dioxide
and water at various times following the initial mixing of the components,
and found the ionization to be at least 90 percent complete in about
0.1 second.  Toor and Chiang     obtained additional experimental data
which indicated that these reactions should be essentially complete in
0.003 second, agreeing well with Saal's     earlier upper-limit value
                                                              (37)
of 0.004 second for ionization reactions of this type.  Thomas
argues that these measurements may have included some diffusion and
secondary reaction effects which rendered these reaction-time estimates
conservatively high.

Perhaps the most  dramatic demonstration of the rapidity of the hydrolysis
reactions is simply the rapid response of conductivity analyzers such as
the Davis and the Sign-X, which depend upon these reactions for their
operation.   This  is especially apparent with the Sign-X sulfur dioxide
scrubber, which removes sulfur dioxide from gas streams by exposure to
ion-exchange resin in water for extremely short residence times.  If
sulfur dioxide did not hydrolyze very rapidly, such scrubbers would be
inoperable.

-------
                   (42)
Whitney and Vivian    have  suggested  that hydrolysis  times for sulfur
dioxide in water may be somewhat greater  than  the preceding arguments
indicate; but their analysis has been  criticized subsequently by Lynn,
       (39)
et al.     who claim that the "reaction"  effect apparent in their
absorption data was probably caused by an inappropriate choice of a
liquid diffusion coefficient.
The disagreement between the measurements of Wang and Himmelblau and
those of the other workers is difficult to explain.  From the description
of the experiment given in their paper it appears that an undetected
problem with sulfur dioxide partitioning to the gas phase as samples
were withdrawn from their reservoir may have been a factor.  This is
purely conjecture and difficult to verify.  Precipitation washout of
sulfur dioxide, based on the validity of Wang's and Himmelblau1s data,
                                      (43)
has been examined by Miller and DePena

It appears that there is a need for further experimental work to resolve
these questions, although it seems reasonable to analyze sulfur dioxide
washout assuming rapid hydrolysis.  Under such circumstances it has been
         (44)
suggested     that absorption can be treated as psuedo-physical phenomenon.
The problem remains, however, of accounting for the multiple species
formed by the reactions, each of which will diffuse at its own rate and
recombine and/or dissociate depending upon conditions in its immediate
environment.
For the limiting case of rapid mixing, such difficulties present no
particular problem, since differences in molecular diffusion rates become
unimportant and concentrations will be uniform throughout the drop.  For
the opposite limiting case, however, Equations 3.22 and 3.23, which
pertain to a binary mixture, are not strictly applicable.  A more elabo-
rate set of equations, pertaining to conservation of species within the
multicomponent system could be set forth.  Because of their complexity
and because of uncertainties of various physical properties, however, the

-------
value of such an effort is doubtful.   It seems more expedient to choose a
mean diffusivity and treat the sulfur dioxide water system as a pseudo-
binary mixture to obtain estimates of the lower limes of washout behavior.

The complications of ionization reactions insofar as nonlimiting mixing
behavior is concerned are obvious.  To the authors' knowledge there have
been no serious attempts to examine this behavior in detail theoretically.

 In  addition to the reversible, rapid ionization reactions that occur when
 a sulfur dioxide molecule comes in contact with water, slower, irreversible
 reactions may take place.  Sulfate formation is undoubtedly the most
 important of the irreversible reactions, although many aspects of this
 formation process are not well understood at the present time.  Junge and
     (45)
 Ryan     have shown experimentally that trace amounts of catalyst in the
 drop influence reaction behavior strongly.  Ammonium hydroxide, in addi-
 tion to chlorine salts of manganese, iron, copper, and cobalt are active
 as  catalysts for the oxidation.  Sodium chloride, in contrast, appears to
 have little catalytic activity.  For pure water solutions, Junge and Ryan
 found that sulfur dioxide oxidation takes place very slowly — behavior
 that has been observed also in this study.

 Junge and Ryan noted that for a given ambient sulfur dioxide concentration,
 the rate of sulfate formation decreased with decreasing pH.  Since sulfate
 formation results generally in a lowering of pH, this reaction tends to
 decrease in rate as it progresses, coming to a virtual halt at a pH of
 about 2.  It has been suggested (cf., Scott and Hobbs^  ') that the
 observed retardation occurs simply because of the decreased solubility
 of sulfur dioxide under these circumstances; Junge and Ryan indicate,
however, that a more complex type of autoinhibition may be involved.

Junge and Ryan point out that if a (chemically) basic material is present
so that a suitably high pH is maintained, the sulfate-forming reaction
will persist.  These authors have demonstrated this experimentally using


-------
ammonium ion, a common constitutent of atmospheric precipitation.

Expressions for mass-transport rates to drops, wherein irreversible
reaction occurs, are available in the literature for the previously
described limiting cases    *  '   .  Such expressions, however, are often
contingent upon the validity of Henry's law.  In addition, these solutions
have not dealt  with the complications posed by variable catalytic activ-
ity and autoinhibition.

                                  (45)
Previous experimental measurements     have indicated that rates of sul-
fur dioxide oxidation  (half-lives of the order of hours) should be slow
compared to rates of physical absorption as the drop falls through the
rapidly changing concentration field of a plume.  In view of such behav-
ior, it appears permissible to neglect altogether the oxidation of sulfur
dioxide under such circumstances, at least for situations wherein plume
dimensions are not abnormally large, and the drops are sampled while still
carrying a significant fraction of the sulfur dioxide collected enroute.

For circumstances wherein the time rate of change of sulfur dioxide con-
centration in the neighborhood of a drop is slow, the opposite may be
true.  That is, chemical reaction may be the most significant factor con-
trolling the amount of sulfur dioxide being washed out.  Such situations
may exist for in-cloud washout or for cases of below-cloud washout
involving long falls through regions where sulfur dioxide concentrations
do not change rapidly.  Estimations for washout of sulfur dioxide under
such conditions have been published by previous authors

ILLUSTRATIONS OF LIMITING WASHOUT BEHAVIOR ~  The two "limiting" cases
of washout behavior noted previously are characterized by Equations 3.8,
3.20, 3.21, 3.22, and 3.23.  These cases are examined by calculating and
comparing the washout rates predicted by both.  One method of accomplish-
ing this is simply to consider the rate of takeup (or loss) by a single
drop falling in a region of known concentration.  For simplicity, an

-------
initially clean* drop is considered which falls through a region wherein
the concentration is maintained at 1 part sulfur dioxide per million  parts
of air  (PPM)  (yAT = 10~6).  Subsequently, this same drop, now saturated
               AD
with sulfur dioxide, is considered as it falls in clean air  (yAb = 0).

Total transport to the drop may be found by integrating the flux expres-
sions, Equations 3.1 and 3.2 with respect to time and multiplying by the
drop surface area.  This must be performed in conjunction with equilibrium
data of the form given by Equation 3.6 using a trial-and-error procedure.
The equilibrium relationship used here is based upon the data of Terraglio
and Manganelli,     and is shown in Figure 3.6.

Mass-transfer coefficients for the various cases considered here are  shown
in Table 3.2.  Here the gas-phase coefficients have been calculated from
the Frossling Equation 3.20, and the overall coefficients were determined
from Equation 3.5.  Liquid-phase calculations are based on an assumed
diffusion coefficient of 1.7 x 10~5 cm2/sec
Absorption-desorption curves based on these calculations are shown in
Figures 3.7, 3.8, and 3.9.  These curves are based upon rather idealized
conditions of plume geometry, yet they illustrate vividly several points
which characterize gas-washout phenomena.  Perhaps, the most significant
aspect illustrated by these curves is the ability of the drop to desorb
gas under appropriate conditions.  Thus, if a drop falls through a plume
and emerges into cleaner air before reaching the ground, it may release
a majority of the gas that is absorbed in more concentrated regions.
Such behavior will tend to lower the altitude of the sulfur dioxide plume.
The significance of this effect has not been investigated thoroughly;
preliminary calculations indicate that, under extreme circumstances, it
may be appreciable.

Figures 3.7, 3.8, and 3.9 show that the rates of absorption decrease
toward zero as equilibrium is approached, a direct consequence of the

-------
E
a.
a.
o
t—I
I-

-------
                               TABLE 3.2
      OVERALL MASS-TRANSFER COEFFICIENTS FOR STAGNANT FALLING DROPS
   Fall  Time
      sec
 Liquid  Concentration
	ymoles/L	
                                                         K
                  Moles/cm sec
                 Absorption    Desorption
                           Absorption    Desorption
                      	 0.3 mm Diameter Drop 	
       0.0
       0.3
       0.6
       0.9
       1.2
       1.5
      0.0
      2.0
      4.0
      6.0
      8.0
     10.0
      0.0
     20.0
     40.0
     60.0
     80.0
    100.0
   0.0
  39.2
  61.0
  75.9
  86.9
  95.3
139.7
112.9
 96.7
 84.1
 76.6
 72.7
              0.001258
              0.000475
              0.000389
              0.000322
              0.000280
              0.000246
                      	 1.0 mm Diameter Drop 	
  0.0
 38.5
 56.3
 68.6
 78.0
 85.7
    ,7
    .4
139.
110.
 95.4
 84.0
 74.8
 67.0
0.000891
0.000212
0.000152
0.000126
0.000110
0.000099
                      	 3.0 mm Diameter Drop 	
  0.0
 53.1
 71.1
 83.3
 92.6
100.0
139.7
 93.8
 76.7
 64.4
 54.8
 47.0
              0.0000650
              0.0000715
              0.0000518
              0.0000437
              0.0000392
              0.0000364
              0.001258
              0.000469
              0.000453
              0.000459
              0.000551
              0.000601
0.000891
0.000235
0.000214
0.000209
0.000209
0.000211
              0.000650
              0.000109
              0.000101
              0.000102
              0.000105
              0.000110
reversibility of the washout process.   Such behavior points out the
inapplicability for gas washout of the first-order irreversible model
employed in most previous theoretical  washout analyses.  Contrary to
assumptions employed in previous work,(52'53) the first-order irreversible
model is not generally applicable for  gas washout analysis even for
situations involving trace-gases.

-------
   150
   TOO
  ..,      i.ST-npnF

~f      — — -~3^=»
                 ,  IRREVERSIBLE
                                REVERSIBLE  -  WELL MIXED DROPLET
                                REVERSIBLE  -  STAGNANT DROPLET
   150
o
o

 CM
o
in
LU

_)
CL

O
o:
   100
    50
           1ST-ORDER,  IRREVERSIBLE MODEL
REVERSIBLE   STAGNANT  DROPLET
                     REVERSIBLE - WELL MIXED DROPLET
                           I    	I
                           1                   2

                          FALL DISTANCE, METERS
Figure 3.1   Absorption-Desorption Behavior for 0.03 cm Diameter Drops

-------
      150
      100
      150
  UJ
  O
  o
  o
   CM

  °   100
Q_
O
cn
a

    50
                             REVERSIBLE  -  WELL MIXED DROPLET


                          1ST-ORDER, IRREVERSIBLE
                                REVERSIBLE - STAGNANT DROPLET
                           1ST-ORDER, IRREVERSIBLE MODEL
                                REVERSIBLE - STAGNANT DROPLET
                                REVERSIBLE - WELL MIXED
                                                  DROPLET
                             J_
                                               1
                             20                  40

                            FALL DISTANCE, METERS
                                                                   60
Figure  3.8   Absorption-Desorption Behavior for 0.1 cm Diameter Drops

-------
    150
    100
     50
  o
  s:

                                        REVERSIBLE  - WELL  MIXED
                                                           DROPLET
                               1ST-ORDER, IRREVERSIBLE
    150
   CM

  5?
    100
  Q-
  O
  at
  o
     50
                             REVERSIBLE - STAGNANT DROPLET
                     1ST-ORDER, IRREVERSIBLE MODEL
                          REVERSIBLE - STAGNANT DROPLET
                          REVERSIBLE - WELL MIXED DROPLET

                             i        	I	
                            100                200

                           FALL DISTANCE, METERS
300
Figure  3.9   Absorption-Desorption Behavior for 0.3 cm Diameter Drops

-------
Curves corresponding to the first-order,  irreversible model are shown in
Figures 3.7, 3.8, and 3.9 for comparison.   From these, it is seen that
application of the irreversible model for gases is valid only for condi-
tions where the curves coincide.  Such conditions might be approached for
the situation of a small plume bordering  the earth's surface; however,
one would not expect these conditions to  be valid in the present study.
Inapplicability of the first-order, irreversible model is unfortunate
in the sense that it invalidates, for most cases of gas washout, the more
convenient method of analyzing washout behavior that was described in
Chapter II.  As a consequence, more physical data are required before a
complete examination of the process can be performed, a fact that was
indicated previously by Equation 3.19.

The upper and lower "limit" curves of Figures 3.7, 3.8, and 3.9, as
discussed previously, bracket the behavior to be expected from real
systems.  Except for very large drop sizes, the limits of physical behav-
ior are confined to fairly narrow regions, and, in a sense, these can be
used as approximations in lieu of absolute knowledge of true behavior.
"Patched" solutions of this type have been applied previously to other
                                   (54)
types of physical science problems.      In this context, the variation
of the mass-transfer coefficient, as exemplified by Table 3.1, should be
mentioned.  From this table, one can determine limits for the overall
coefficients and, if desired, estimate a  mean value to be used for
linearization in Equation 3.10.  For cases wherein K  varies widely, this
mean should be chosen carefully in view of the physical conditions of
interest.

An additional point concerns the sharp dependence of absorption behavior
on drop size.  In view of such behavior one would expect a dramatic differ-
ence in sulfur dioxide level between raindrops of different sizes, depend-
ing upon the history of the plume-raindrop interaction.  Experimental
field measurements of sulfur dioxide content as a function of drop size,
not available at present, would be of high interest in this regard.

-------
LABORATORY MEASUREMENTS OF SULFUR DIOXIDE WASHOUT — The  complexity
exhibited by Equation 3.19 suggests that  controlled  laboratory  experiments,
rather than field studies, might be more effective  in  analyzing washout
behavior theoretically.  This equation is rewritten here  for convenience.
     Z = 4TrNT  /  a2 f*(a)
               o            c       va
If, for example, experiments  could be conducted with uniform drops falling
vertically in air of uniform  concentration,  the integrals of Equation 3.19
could be simplified to
              Z - -                                  >             (3.35)
which reduces considerably  the ambiguity involved in determining K  .

Laboratory experiments in this area can be divided into four different
categories, listed as follows:

     1.  Studies of drops suspended on solid supports.
     2.  Studies of drops suspended by moving gas.
     3.  Studies of isolated drops falling through gas.
     4.  Spray chamber studies.

Each of these types of experiments has its individual disadvantages.
Studies of drops mounted on solid supports (e.g., hypodermic needles,
filaments) have the strong disadvantage that natural circulation patterns
are invariably perturbed.  Such perturbations can be avoided with the
second and third types of studies if care is taken to ensure that the
newly formed drop is completely stablized in its environment prior  to


-------
contact with gas.  For the case of isolated, falling drops, this usually
necessitates a long stabilization column, where the oscillations induced
during drop formation are allowed to dampen out and terminal velocity is
approached prior to contacting the gas.  Some insight with regard to the
effects of oscillations of newly formed drops can be obtained from the
                          (31)
results of Garner and Lane

Experiments involving single drops suspended in air have the serious
disadvantage that, because of the inherently small sample size, they are
not generally amenable to studies where low concentrations are involved.
Isolated, falling drop studies overcome this disadvantage by combining
several  drops into one sample.  Spray studies share this advantage; how-
ever, such experiments are plagued by several marked disadvantages,
including unavoidable drop perturbation, nonconstant fall velocity, and
the presence of a spectrum of drop sizes.

Experimental studies of the washout of trace sulfur dioxide by sprays have
been conducted by Georgii and Beilke.^  '  '  The above mentioned diffi-
culties, however, have rendered these authors' results of limited value
to the present study.  Isolated falling drop experiments involving sulfur
dioxide washout have been performed by Johnstone and Williams.      These
studies, however, were confined to rather high sulfur dioxide concentra-
tions and involved additive-containing drops.  In this context, the sus-
                                      (31)
pended drop studies of Garner and Lane     should be mentioned.  These
studies also involved washout from gases of relatively high concentration.

The main use of the results of the studies described above is their common
qualitative indication that both gas-phase and liquid-phase mass-transfer
resistance are important in atmospheric sulfur dioxide washout.  Further
implications from these experiments of a quantitative nature are dubious
insofar as trace-gas washout is concerned.  Further data from an isolated,
falling drop experiment would be of extremely high interest in this
regard.

-------
THERMODYNAMIC ASPECTS - CHEMICAL AND SOLUBILITY EQUILIBRIA

Solubility relationships, in addition to their importance in influencing
transport behavior, are useful in estimating maximum and minimum washout
rates in the absence of actual washout data.  If the maximum concentration
in a plume is known, for example, then one can state immediately that the
concentration of washout gas in the precipitation is no greater than the
equilibrium concentration corresponding to that maximum.  Similarly, a
lower limit can be established on the basis of the minimum concentration
experienced by the drop, provided, of course, that there is no loss by
reaction within the drop.

Dispite its importance to the understanding of sulfur dioxide washout,
there is relatively little data pertaining to the solubility of trace
concentrations of this gas in water.  The bulk of sulfur dioxide solubil-
ity data that does exist pertains to higher concentrations than those of
interest here.  With the exception of some fragmentary information per-
taining to air concentrations down to about 200 ppm, the only low con-
centration data that have been published are those of Terraglio and
Manganelli, which were illustrated earlier in Figure 3.6.

Normal atmospheric sulfur dioxide concentrations are of the order of
0.01 ppm or less, and it is of interest to possess solubility data down
through this region.  Such information can be estimated roughly by
extrapolating the existing data to the concentration levels of interest,
provided that certain assumptions with regard to equilibrium behavior
                                        (58)
are made.  On the basis of previous work^    it appears justifiable to
assume Henry's law to be valid for the equilibrium between gaseous and
aqueous molecular sulfur dioxide  (Equation 3.30).  Thus
                            [S02]  = H[S02]                          (3.36)
On the basis of available data, H = 1.0 at  20°C, provided  [S02]  and
                                                               g

-------
[S02]   have units of ppm and  ymoles/1, respectively.

Previous work indicates also that the second dissociation constant is so
small that it can be neglected for all practical purposes.  From this
work, one may estimate the first dissociation constant, K^, to be about
1.7 x 104 iamoles/1 at 20°C.

Two types of hydrogen ion may exist in the solution, that arising from
the  ionization of sulfur dioxide and that originating elsewhere (e.g.,
from the presence of dissolved sulf ate) .  It is useful to differentiate
between these; the latter will be referred to as "excess" hydrogen ion.
Thus,
                               [H3°+]S0  + [H3°+]ex      '          (3.37)
Combination of Equations 3.33, 3.36, and 3.37 provides an expression for
ccn , the total concentration of dissolved sulfur dioxide:
 SO-

         CSO- = [SO-1   + [HSO~]
            Z      i aq       j
                [S02]g   -[H30]ex ± yCH]^ + 4Kl[S02]g/H
                       T                         -               .    (j, JO}
                  H
Solutions to Equation 3.38 are plotted in Figure 3.10.  Here the parameter
denoting "excess pH" is simply the pH arising from the presence of excess
hydrogen ion, i.e.,

                    excess pH = -Iog10 l
-------
1000
0.01
                                                                A
                                                                •
                                                                o
                                             EXCESS  pH
                                             EXCESS  pH
                                             EXCESS  pH
                                             EXCESS  pH
                                             EXCESS  pH
6
5
4
3
2
    0.001
0.01                  0.1                   1
      AIRBORNE S02  CONCENTRATION ppm
     10

-------
this would lead one to expect a general lowering of measured washout rates
with decreasing pH.  Secondly, this dependence would be expected generally
to decrease the effectiveness of washout in the plume, since the plume
itself contains acid-forming material (e.g., sulfates, nitrates) which
will be washed out and serve to lower pH.  Such an effect explains the
otherwise paradoxical situation where sulfur dioxide concentrations in
rain that has passed through the plume are actually lower than those in
rain that has not.  This phenomenon is suspected to have contributed
significantly to behavior exhibited by much of the data of this study.
Since pH lowering by washout in the plume should be more significant the
higher the pH of the incoming rain, this effect should be more pronounced
under circumstances wherein background pH is high.

Finally, it should be noted that in-plume washout of acid-forming materials
complicates the calculation of sulfur dioxide washout enormously.  Refer-
ring to Equation 3.5, it is seen that the equilibrium relationship is no
longer fixed, but varies with time as the acid-forming substances are
accumulated.  This suggests that in order to calculate sulfur dioxide
washout, one must have substantial information pertaining to washout of
other atmospheric species,  making any thorough analysis under such con-
ditions a truly formidable problem.

It is emphasized that the arguments of this section are based upon extrap-
olated data, the accuracy of which is rather questionable.  Experimental
measurements of solubility as a function of excess pH in this concentra-
tion region would be of high value in examining these phenomena in further
detail,  and in validating the implications of the present discussion.

-------
                               CHAPTER  IV

                      REVIEW OF FIELD EXPERIMENTS

Chapter IV contains an overall review of the  field experiments. The pri-
mary data — concentrations of sulfur dioxide  recovered in the precipitation
collectors — are tabulated in Appendix  A.  Graphical representations of
sulfur dioxide distributions on the grid are  included there also.  On
these, concentrations are represented by radial bars of proportionate
length extending from the collector locations.  Vertical profiles of wind
velocity, determined principally from pibal soundings at the Overlook,
appear in Appendix A to account, in part, for the observed distributions.

Run data in Appendix A are grouped according  to the experimental period in
which the run was conducted.  A table preceding each group summarizes the
runs relative to the operations, wind velocity, precipitation rate, sulfur
dioxide emission from Keystone, and sulfur dioxide recoveries on the grid.

Frequent reference is made to the data  of Appendix A throughout this
chapter.

REVIEW OF FIELD EXPERIMENTS

Experimental runs were conducted during three separate periods of approxi-
mately one month duration:  October-November, 1969; February, 1970; and
April-May, 1970.  Autumn and Spring runs were conducted in rain and the
Winter runs all were conducted in snow.  The  Spring runs were added to
the program because of a particularly dry Autumn.

FIELD PERIOD 1 —  The first three weeks of the Autumn period were char-
acterized by warm, clear weather, which provided ample time for prepara-
tion for field experiments, but delayed the initial experiment to test the
adequacy of those preparations.  The first run was conducted at night on
October 20 and the final four runs occurred on the mornings of November
1 and 2.

-------
Autumn rains were accompanied by low clouds, which resulted in insuffi-
cient ceilings for aircraft operation during four of the five runs.  The
one occasion for which aircraft use was possible provided an opportunity
to develop aircraft sampling capability under experimental run conditions.
All equipment operated satisfactorily, resulting in the measurement of
temperature, humidity, and sulfur dioxide concentration from near stack
height to cloud base in traverses about one mile downwind of the Keystone
Station.

It was planned that definition of the stack plume geometry would be
provided photographically.  However, the plume is not normally visible,
except for condensation of its vapor, because of the highly efficient
removal of fly ash by the electrostatic precipitators.  When conditions
are proper for persistence of condensate in the stack plume, it normally
is obscured by the cooling-tower condensate plumes.

Pibal soundings showed that the wind direction varied widely with height;
in one case, exceeding 100° within 2000 ft.  This questioned the adequacy
of sampling on 90° of arc, since a wider distribution of sulfur dioxide
in precipitation must be assumed with such large directional shear.
Consequently, the sampling arcs were extended during the later experi-
mental periods.

Subsequent analysis of the first period data showed sulfur dioxide con-
centrations in the collectors to be distributed erratically.  Neither a
quasi-normal distribution along the arcs, nor a decrease in concentration
with distance was evident.  Wide variability between adjacent collectors
and alignment of troughs on one arc with peaks on the other were common.
Large differences between peak and background generally were not observed.
The absence of reduced values at the extremes of the arcs suggested that
the plume had not been contained.

Expansion of the grid with sampling to a greater distance and increased
lateral coverage was deemed necessary to contain the plume and document

-------
the possible increase of concentration with distance.  Both actions were
taken in Field Periods 2 and 3.  The reduced likelihood that meaningful
correlation of washout and precipitation characteristics could be achieved,
due to the wide variability in washout, resulted in conducting precipi-
tation characterization at only one location during the subsequent periods.
Precipitation charge spectra were taken  infrequently and no attempt at
their correlation with washout has been made.

Run 1 —  Deployment of the precipitation collectors began in moderate
rain following the passage of a thunderstorm which occurred about two
hours ahead of its forecast.  Less than half were in position when the
surface wind measurement at the Overlook showed a shift from 280° to 250°.
The wind shift was accompanied by a sharp decrease in rainfall intensity,
and a number of collectors deployed on the southern portion of the arcs
after this time received insufficient precipitation for analysis.

The fact that the collectors, which sampled the heavier rainfall during
the period when the surface wind direction placed the plume south of them,
did not give evidence of dilution leads to the conclusion that only the
surface wind exhibited a northerly component.  Both the 1900 EST Pitts-
burgh sounding, which showed a wind direction of 230° near the stack
height elevation, and the 240°-250° winds measured at the Overlook follow-
ing the weather disturbance suggest that the upper winds were west-south-
westerly throughout the run.

Figure A.2 and Tables A.2 and A.3 show significantly lower concentrations
beyond Stations A-17 and B-16, which were installed following the surface
wind shift at 2140 EST.  Exceptions to this are Stations A-26 and B-17,
whose very high values are attributed to the exhaust of gasoline-powered
generators at those locations.  High readings which appear at A-21, A-22,
A-25, B-20, B-21, and B-24 are unexplained, but may be due to unobserved
surface winds which retained their northerly component after the wind
shift.

-------
An alternative.explanation of the observed distribution is that the wash-
out sulfur dioxide was derived from an upwind source which blanketed the
entire grid.  This would result in a fairly homogeneous background dis-
tribution and, supposedly, the Keystone plume washout would be superposed
upon it.  It might be argued from the distributions shown that no addition
to the background is obvious and that a deficiency is more likely.  The
discussion of subsequent runs will also note this apparent negative influ-
ence of the Keystone plume.

Run 1 pointed up several deficiencies in the experimental design.  Whereas
this method of sampling had been used successfully in small-scale experi-
ments, a more complex situation exists on this larger scale.  Wind shear,
for example, becomes an important problem when a buoyant plume is emitted
from an 800 ft stack.  The necessary grid size to investigate the release
of sulfur dioxide from a tall stack introduces the problem of sufficiently
rapid collector deployment and retrieval to attain concurrent sampling at
all stations and assure plume containment throughout a run.

Runs 2, 3 —  Runs 2 and 3 were conducted on a morning of warm front
precipitation and a cloud base sufficiently high for aircraft operation
during Run 2.  The aircraft traverses from 2000 ft to 3000 ft MSL showed
sulfur dioxide at all levels in peak concentrations of 1-4 ppm.  The
                                                     o
temperature lapse rate was only slightly stable — 2.4 C/1000 ft — and was
conducive to the observed mixing at this time.  Orientation of the inter-
cepts of sulfur dioxide from the Keystone stack was along the wind direc-
tion shown by the pibal sounding at 0725.  At all levels, the wind was
between 165° and 170° at that time.

The surface distribution of sulfur dioxide in the collectors, Shown in
Figure A.4, indicates a minimum concentration on Arc A at a bearing of
330° to 345°from the Keystone stack.  Larger values lie to the north,
beneath the aircraft intercepts of the plume, and also to the south.
Washout concentrations on Arc B do not show a similar dip, but increase
generally from south to north.  The dip in concentration on Arc A may be


-------
the result of plume splitting at  the stack which  is discussed in a later
section.

It is unlikely, under the observed wind profile,  that the sulfur dioxide
distribution from north to north-northeast of  the Keystone  Station origin-
ated at Keystone.  It is suggested that the source of this  distribution
was the Homer City Generating Station, which is appropriately oriented
some 12 miles further upwind, as  shown in Figure  4.1.  One measurement
of sulfur dioxide by the aircraft in a traverse at 2500 ft which extended
north of the other traverses was  apparently due to an intercept of the
Homer City plume.

The absence of washout sulfur dioxide north of Keystone in  the Run 3
distribution, shown by Figure A.6, is due to the  termination of the rain
before northern portions of the arcs were deployed.  The higher concen-
trations to the west of Keystone  are a result of  continued backing of
the winds with time at increased  heights as shown by the 0930 pibal in
Figure A. 3.

Runs 4, 5 —  Rainfall was heavier and the overcast thicker and lower
than on the preceding day when Runs 2 and 3 were  conducted.  Aircraft
operation was not feasible.  Run  4 was conducted  during a period of heavy
precipitation with the rate exceeding 9 mm-hr"1 for a brief period when
all collectors were positioned.   The storm had nearly passed before the
change-out of collectors had been completed for Run 5.  Throughout the
morning, the raindrops were quite uniform in size at about one-half
millimeter diameter.  The air was near saturation.

Approximately 90° of shear was shown by the pibal sounding during Run 4
with directions ranging from east-northeast at the Overlook to south-
southeast about 1000 ft above the stack exit.  Wind speeds were lighter
than on the preceding day and the buoyant cooling tower plume was observed
to interact with the stack plume.

-------
                                              ARC B
I
r
                                                     ARC C
                                                     KEYSTONE GENERATING STATION
                                                                             INDIANA
                            1    2

                            MILES
LUCERNE COKE
    OVENS
                                                  HOMER CITY
                                              GENERATING STATION

-------
Based upon the observed winds, containment should have been nearly com-
plete, but high values at the arc ends make uncertain the degree of
containment.  During the period of the runs, observed  winds had insuffi-
cient directional range to account for the distribution at the arc
extremities.  Clearly, the winds were suitable to transport the Homer
City plume over the arcs.  This apparently is the source of much of the
observed washout sulfur dioxide, since the Homer City furnaces were fired
throughout this day.

Lower volumes of collected rainfall on Arc B in Run 5 are the result of
decreased rainfall rate before complete deployment of that arc.  Total
sulfur dioxide recoveries on Arcs A and B were essentially equal for
Run 4.  Since concentrations also were comparable on the two arcs, this
points up the general contamination of the entire sampled grid.

One feature which was quite prominent in both runs is the decrease in
concentration on Arcs A and B beneath the expected plume path.  In fact,
a line through the troughs exhibits a curvature suggestive of the veering
wind.  Unlike Run 2, this low concentration region cannot be accounted
for by splitting of the plume since there is little evidence that adjacent,
high concentrations can be attributed to the Keystone plume.   It appears,
therefore, that the Keystone Station actually had a negative effect on
washout in this instance.  It was also noted that minimum sulfur dioxide
values were associated with maximum sulfate values and this is discussed
in Chapter V.

FIELD PERIOD 2 —  Suitable weather for experimentation in snow occurred
only during the first half of February.  Seven runs were made in a variety
of snow situations.  Again, the distributions of sulfur dioxide concen-
tration in the precipitation were erratic.  The apparent lack of plume
containment by the samplers, which was indicated in the first period, was
again noted, but general obscuration of the plume washout by the washout
of background sulfur dioxide now was considered a more likely explanation,
since sampling arcs had been extended.


-------
Along with the extension of the sampling arcs, field procedures were
altered to provide shorter duration for deployment and retrieval of pre-
cipitation collectors.  Ideally, the sampling should be concurrent at all
collectors during a run to minimize the effect from variations in precip-
itation with time.  An extended period for change-out between consecutive
runs further complicates the analysis since sampling on the first run is
continuing while sampling on the second run is beginning.  Two or three
crews of two men  each were used for deployment and each arc was deployed
in turn.  This was in contrast to the first period operation when two men
concurrently deployed both arcs.  Despite extension of the arcs by 100
percent and the addition of 50 percent more collectors, the time was cut
significantly.  The difference between the average total sampling duration
for each collector and the concurrent sampling duration for all collectors
was reduced from 52 minutes during the first period to 26 minutes during
the second period.  Additional manpower was made available by the reduced
emphasis on precipitation characterization and as a result of the inabil-
ity to use the aircraft during low ceiling and low visibility conditions
in snowfall.

Photographic documentation of snowflake characteristics was provided for
each run of this period through the use of special equipment.  Snow con-
ditions differed on the four run days and these differences are clearly
shown in the photographs which are discussed in Chapter V.

Run 6 —  The first run in the snow washout series was conducted during
a moderate snowfall.   The 0700 EST Pittsburgh sounding showed a weak
inversion extending to about 2000 ft above the surface, indicating a very
stable air mass.  Surface temperatures were 15°-20°F during the run and
the snow consisted of small, dry grains and crystals.  Surface winds were
west-northwesterly and the Pittsburgh rawinsonde showed north-northwesterly
winds aloft.  Pibal soundings and other support operations were not pro-
vided as a consequence of the occurrence of suitable experimental weather
less than a day after crew arrival in Pennsylvania.

-------
Approximately 189° coverage was provided on Arcs A and B.  Primary col-
lectors, those for sulfur dioxide and sulfate, were spaced at double the
interval used during the first field period.  Secondary collectors were
located at alternate primary collector stations to provide samples for
pH, nitrate, and nitrite analyses.

The distribution shown by Figure A.12 exhibits a multimodal character and
is seen to be quite erratic.  Distributions at the northeast end of Arcs
A and B and at the southwest end of Arc A are believed to be unrelated
to the Keystone emission considering the existing wind regime.  All of
these collectors were located in narrow road cuts or along steep hill-
sides which under the existing conditions of flow and stability might
tend to concentrate vehicle exhaust and sulfur dioxide from other sources
in the vicinity of the collectors.

The plume appears to have been contained, particularly if the outliers
can be ignored.  A further indication of containment was a visual obser-
vation from Arc B during the run which placed the plume near the center
of the arc.  This position was aligned along the near-surface wind
direction.  It is suggested that the second mode, south-southeast of
Keystone, was associated with the upper level winds.  As in Run 2, there
is a suggestion of plume splitting at the stack, which permits two
distinct areas of washout.  Additional support for this phenomena is given
in a later section.

Runs 7, 8 —  Precipitation on the day of Runs 7 and 8 was quite heavy
and relatively steady.  However, Tables A.15 through A.18 show large
variations between collectors in the volumes of precipitation collected.
This is a problem common to all precipitation samplers because of local
wind speed variations.  It is more pronounced for snow because of its
low terminal velocity and the degree of reentrainment.  Consequently,
the effective area of the collector is reduced by high wind and increased
by drifting snow.  It is probable that the large volumes collected were
the result of drifting snow.

-------
Wet amorphous clusters of snow were mixed with light rain and surface
temperatures remained above freezing during Runs 7 and 8.  The air was
near saturation and moderately stable.  A pibal sounding in the morning
showed a wind direction of 138° near the surface, veering to 178° at
approximately double the stack height.

The arcs which were centered north-northwest of the Keystone Station pro-
vided about 180° coverage.  Collector spacing was the same as on the
preceding run and secondary collectors were used in a like manner.  The
extent of coverage in relation to the observed winds should have ensured
plume containment.  In addition, three separate sightings of the plume
from the arcs placed it northwest of Keystone, yet the distributions,
shown in Figures A.14 and A.16, extend over all sampling positions on
both arcs.   If any evidence of the plume exists to the northwest of
Keystone, it is an apparent decrease in concentration as noted previously
with south-east flow.

Both units  at Homer City were operating on this day, and it is apparent,
therefore,  that washout from the Homer City plume obscured the Keystone
washout. All collectors were downwind of Homer City  based on the
observed winds.  The sulfur dioxide contained in collectors on the north-
east leg of Arc A was due, probably, to washout from the Homer City plume.
Maximum values of about 10 ymole-liter"1 are attributed to this source at
a distance  of about 13 miles.

Runs 9. 10, 11 — Temperatures were slightly above freezing at the surface
and the 0700 EST Pittsburgh sounding showed below freezing temperatues
aloft, so that all precipitation fell as snow.  The air was saturated and
the ceiling was 500 ft, or less.  Following Run 11, a pibal was tracked
to stack height, but previous pibals were tracked only 30 seconds due to
the low ceiling.  All observed winds were between 270° and 290°.

The indicated winds were westerly at the surface, with speeds increasing
through the day.  Higher elevation winds were poorly defined even on a


-------
regional scale since Pittsburgh's wind soundings were not complete and
the morning pattern throughout the eastern seaboard was poorly defined.
Northerly flow in western Pennsylvania was the most probable at lowest
elevations, though northeasterly could not be ruled out.  Within the
first 2000 ft above the surface, easterly and southeasterly winds also
were indicated prior to noon.  The sulfur dioxide distributions of
Figures A.18, A.20 and A.22 fail to provide indication of a preferred
wind direction, though Arc A was completely sampled in Run 9.

As indicated by collector volumes, the precipitation was quite uniformly
distributed on Run 9, but became somewhat erratic later in the day with
increased wind speed.  Washout patterns were quite similar for these
three consecutive runs, but concentrations were noticeably lower on Runs
10 and 11.  This was due perhaps to the increased wind speed, which
decreased the airborne concentration by stretching the plume.

The uniformity of sulfur dioxide distribution again points to a source
other than the Keystone Station, but the poorly defined wind field makes
the source unidentifiable.  Low sulfate measurments shown in Tables A.20,
A.23, and A.24 suggest that the background air was clean.  Based on the
observed near-surface winds, the air previously passed over open farm-
land.  It should be noted also that sulfate concentration peaks were
found at Stations B-22 and B-26 in Run 9 and A-18 and A-20 in Run 11.
These stations are downwind of Keystone based on the observed winds.
Sulfur dioxide concentrations in collectors B-22, B-24, B-26 of Run 9
and A-18 of Run 11 had the lowest values.  This relationship between
sulfur dioxide and sulfate concentrations which was observed in Runs
4 and 5, also, is discussed further in Chapter V.

Run 12 —  Arc B alone was employed in this run, which sought to contrast
the Keystone plume with the background for an unambiguous wind flow.
Pibal soundings showed the winds veering from east-southeast near the
surface to south-southeast beneath the inversion base indicated at 2500

-------
ft MSL on the'1900 EST Pittsburgh sounding.   The Pittsburgh sounding also
showed neutral stability in the layer below the inversion.  Surface tem-
peratures were below 20°F and temperatures above freezing showed nowhere
in the sounding.  Light precipitation fell as crystalline snowflakes.
        j
The sulfur dioxide concentrations in the collectors were low, as they
had been in Run 6.  Unlike that run, the distribution consisted largely
of washout from sources other than Keystone.  The broad distribution
along the southern portion of Arc B could not have originated at Keystone
with the observed winds.  The Keystone plume was observed over Arc A at
a bearing of 300° from Keystone and the peak value for the entire arc,
as shown in Figure A.24, corresponded with this observation and with the
stack height wind direction.

Two factors appear to rule out Homer City as the source of washout south
of Keystone.  On the basis of washout during prior runs on the arcs,
Homer City contributes significantly higher  concentrations than Keystone
and this would certainly be the case during  Run 12 when both units at
Homer City were operational.  The concentrations south of Keystone, how-
ever, are about equal to the peak concentration northwest which is
attributed to Keystone.  Also, for Homer City washout to be evident on
the southern portion of the arc, it would also be evident on the western
portion with which it is aligned.

The probable source for the washout south of Keystone in Run 12 is the
coke ovens which are located at Lucerne, south of Indiana.  This source
is shown in Figure 4.1.  The highest concentrations on the arc south of
Keystone lie at an azimuth of 290° from Lucerne.  It is interesting that
this source, which is located at a distance of over 8 miles, contributes
as much sulfur dioxide washout on Arc B as the Keystone Station at a
distance of 2 miles.  Assuming that the release rate of sulfur dioxide
from the coke ovens, is well below that of Keystone, the increased surface
air concentration from this surface release must be a significant factor.

-------
FIELD PERIOD 3 —  Precipitation during April was more evenly distributed
than during the previous two periods, although appreciable rain fell on
only three days.   On one of these days, three runs were conducted; the
first during passage of an isolated cell, and the last during passage of
a vigorous cold front.  Large differences in the sulfur dioxide washout
were observed between these runs.

An innovation during the April period was the use of the sulfur dioxide
analyzer, which was intended for use in the aircraft, for sampling of the
near-surface air.  It was mounted on a truck and driven on the arcs to
locate high concentrations from the Keystone Station and other sources.
These tests are described in a later section.

Arc C, which had been installed during the February period, was utilized
in four of the April-May runs, Arc B in three others.  A dry run to
evaluate the collector response to sulfur dioxide air concentration was
also made.  In addition, one intended run, Run 17, failed to receive
precipitation, and another, Run 13, received such small amounts as to be
unsuitable for further consideration.

Support equipment was sparse as a result of a wildcat Teamster strike
when the equipment was in transit.

Runs 14, 15 —  Heaviest sulfur dioxide recoveries were realized on Runs
14 and 15 which were conducted under an easterly wind flow.  Winds prior
to Run 14 were east-southeast at all levels.  Throughout the morning, the
winds backed so that by the end of Run 15 they were essentially from the
east, the final pibal showing lower elevation winds with a northerly
component and upper winds with a southerly component.  Wind speed
increased significantly above stack height though this may have been only
apparent — due to the heavier, wet pibal.  The air was essentially sat-
urated throughout a deep layer from the surface and the thermal stability
was about neutral.

-------
Moderate precipitation was quite steady over the entire grid.  The synoptic
map, in fact, showed precipitation was widespread northwestward from an
intensifying storm system near Norfolk, Virginia.

The distributions of sulfur dioxide shown in Figures A.26 and A.28 fail to
define the plume clearly at a direction of 280° from Keystone as visually
located by an observer on Arc B.  The sulfur dioxide distribution, which
obviously was derived from a source other than Keystone, could not have
been associated with Homer City under the existing wind regime.  Although
the city of Indiana lies immediately upwind of Keystone, distribution over
the entire grid at a near-constant concentration suggests a more distant
source or more diffuse sources; the latter explanation is preferred.  A
single source would necessarily be very distant considering the apparent
directional constancy of the wind.  Travel over a long distance likely
would diminish the sulfur dioxide content of the air mass by conversion
to sulfate.  However, the sulfur dioxide concentration in the precipita-
tion was high and the sulfate content was low.  A number of minor domestic
sources undoubtedly were available throughout the area since temperatures
on this day were about 45°F.

Given the information that the Keystone plume was visually located 280°
from the source, evidence of washout was sought in the sampling data.
Maximum sulfur dioxide concentrations were found in Run 14 at Stations
B-62, B-64, and B-66, which lie between 260° and 275°.  Again in Run 15,
the maximum on that portion of the arc is found at Station B-64.  If these
are indeed evidence of the Keystone plume, the washout from it is about
3.5 to 5.0 ymole-liter"1 and constitutes 20-25 percent of the total.

Run 16 --  Run 16 was the first to utilize Arc C.  Although the day looked
promising for precipitation, the 0700 EST Pittsburgh soundings showed only
a thin saturated layer at 10,000 ft.  It also showed moderately stable air
below stack height and an inversion above it.  Most of the rain fell in a
half-hour period during collector deployment so that not all collectors

-------
received sufficient amounts for analysis.  Winds during the major rain
period were easterly near the surface and veered to southeast about 2000
ft above the surface.  At still greater heights, probably above the
inversion, winds veered through south to west-northwest.

Figure A.30 shows the washout sulfur dioxide distribution.  The cluster-
ing of concentration measurements reflects the short period of rainfall,
since the collectors shown are those which each team deployed before the
rain terminated.  Maximum concentrations south of Keystone can be attri-
buted either to Homer City or to the Lucerne coke ovens depending on
whether the stack height or surface winds are considered.  Perhaps these
concentrations are a combination of washout from both sources.  North and
east of Keystone, the distributions are probably due to diffuse sources
near Indiana.  The temperatures on this day were around 45°F and domestic
heating units would be expected to emit considerable amounts of sulfur
dioxide into the quite stable surface layer.  West-northwest of Keystone,
the distribution is peaked at about 290° and is thought to be washout both
from the Keystone plume and diffuse upwind sources.  The Keystone plume
was sighted between Stations C-21 and C-22, aligned with the stack height
wind direction, when the collectors were being deployed.  Unfortunately
Station C-21 was deployed too late to receive any precipitation.

Runs 18, 19, 20 —  Although the total sampling duration spanned eight
hours for this run series, the bulk of the precipitation was associated
with the passage of a strong isolated thunderstorm in the morning and the
frontal passage in the afternoon.  Both rainfall rate and raindrop size
spectra were measured at the Overlook.

The 0700 EST Pittsburgh sounding showed a deep layer of moist air which
was slightly stable.  Winds at the Keystone Station were initially south-
west, veering to west-southwest by early afternoon and switching to west-
northwest and ultimately to northwest with the frontal passage at 1555
EST.

-------
Within ten minjates in the morning during Run 18, rainfall rates increased
from about 1 mm-hr"1 to over 15 mm-hr"1  and then decreased to about 0.2
mm-hr-1.  The spatial variability is shown by the collector volumes in
Table A. 32.  Although all collectors were deployed at the time of the
heavy rain, those to the southeast received no rain while those to the
north received their full capacity during Run 18.  Run 19 collectors were
exposed for about five hours between Runs 18 and 20.  They sampled, to
differing degrees, a final shower associated with the morning thunderstorm
and a portion of the rain which preceded the front.  The collectors
retrieved first from Run 19 probably did not sample the prefrontal rain.

All collectors were in position for Run  20 during the major periods of
precipitation associated with the frontal passage.  Since they were all
in position when the rain stopped, the collectors sampled the entire post-
frontal rain.  Variability of rain volume in the collectors was least for
this run.  At the Overlook, the postfrontal rainfall rate attained a value
of 30 mm-hr"1 for a brief period.

Extremely low washout concentration of sulfur dioxide was observed in
Run 18 — the lowest of the three runs.  Values above the low background
appeared to be randomly distributed and  were probably due to local con-
tamination.  There was no evidence of washout from the Keystone plume.

Run 19 showed a marked increase in washout concentration, but it again
showed no evidence of washout from the Keystone plume.  Whatever vari-
ability did exist was probably due to the differences of individual
collector sampling times in relation to  the temporal and spatial vari-
ability of rainfall during the five hours of exposure in Run 19.

Run 20 exhibited uniformly high concentrations of washout sulfur dioxide.
Though higher in concentration, the distribution was similar to Runs 9,
10, and 11, for which northwesterly winds also prevailed; though lower in
concentration, the distribution also was similar to Runs 14 and 15 for
for which the distributions were attributed to diffuse sources.  The


-------
marked maximum  shown  at  Station C-8 in Figure A.36,  which also  appears  in
Figure A.34  of  Run  19, is  assumed to be related to some local condition.
Although  Station C-8  was located downwind of keystone relative  to  the
postfrental  wind direction,  such was not the case during Run 19, when the
wind was  southwesterly.

Prefrental wind  flow was  across  the  Pittsburgh metropolitan area, whereas
postfrontal wind  flow was down  the  corridor  to  the west-northwest between
major cities.  Flow down this same  corridor  in  Runs  9,  10, and 11 resulted
in substantial sulfur dioxide washout  concentrations, but only moderate
sulfate concentrations.  Sulfate washout  concentrations  in Run 20 were
about half the prefrental concentrations.  Since Run 20  includes both
prefrontal and postfrontal precipitation,  it  is expected that the com-
parison of each would show even more striking differences.

The more complete sulfur dioxide concentration  data  were subject to an
analysis to  separate  the prefrontal and postfrontal  contributions to Run
20.  This was possible due to a clear  demarcation of  the frontal possage
time and complete postfrontal sampling.   The  distribution of rainfall
between prefrontal and postfrontal  as  determined at  the Overlook, was
used to proportion the individual collector volumes.  These volumes
together with the total  sulfur  dioxide mass recovered in each collector
were sufficient,  using the method of least squares,  to determine pre-
frontal and postfrontal  sulfur  dioxide concentrations.  The values deter-
mined were 1.2 and 10.3  ymoles-liter for  the  prefrontal and postfrontal
rains,  respectively,  and it appears that  the  cleaner postfrontal precipi-
tation is much more effective in the washout  of sulfur dioxide.

Run 21 — Run 21 was  a dry run  to measure, under fair weather conditions,
the effect on the precipitation  collectors of the surface air concentra-
tion of sulfur dioxide from the  Keystone  stack emission.  The 0700 EST
Pittsburgh sounding showed a subsaturated air mass which with the moist
adiabatic lapse rate would result in a thermally stable atmosphere.
Rawinsonde winds were light westerly at stack height and surface winds

-------
at the Overlook veered through the period of the run from 190° to 240°.
During the run, the plume was observed to cross Arc A between Stations
A-7 and A-9, which would indicate a wind direction of about 250°.

Distributions on Arc A were confused by a road resurfacing operation,
which apparently contaminated a number of collectors.  Discounting all
those beyond Station A-10, which were noted to contain dirt, peaks remain
at 50° and 60° from Keystone.  Peaks are noted also on Arc B at these
azimuths.  On both arcs extremely low concentrations appeared between
these peaks, suggestive of the plume splitting which was inferred by wash-
out sulfur dioxide distributions in Runs 2 and 6.

Run 22 —  The final run was conducted on a day with intermittent, moder-
ate rainfall.  The 1900 EST Pittsburgh sounding showed a saturated surface
layer to about 4500 ft with neutral stability at stack height.  The
plumes were observed initially to extend from the stacks on a bearing of
120°, and a short arc was set up centered on that azimuth.  Subsequently,
the winds backed and the plumes were not contained.

The sampling arc did extend across approximately the southern half of the
mean position of the plumes.  The resulting washout concentrations were
very low with a peak value of only 0.6 ymole-liter"1 despite the operation
of both units at Keystone.

SUPPORTING STUDIES

In addition to the major field experiments just described, supporting
studies were conducted to determine the degree of sulfur dioxide sorption
on plume particulate and the level of sulfur dioxide in near-surface air.

The tests to measure sorption on particulate were performed using two
identical electrostatic precipitators, manufactured by the Mine Safety
Appliances Company, operated simultaneously under identical conditions
except for the high voltage, which was turned off on one unit.

-------
Each precipitator contained an insert of filter paper which acted as a col-
lecting surface inside the collection tube.  The inserts were impregnated
with a coating solution consisting of 4 grams sodium hydroxide and 10 ml
glycerine diluted to 100 ml with water.  This solution provided an irrever-
sible "sink" for any sulfur dioxide molecule incident upon the collector
surface.

Since the precipitator operating at high voltage collected essentially all
of the particles and the other collected essentially none, the difference
between collected sulfur dioxide levels gave a measure of particulate-
bound sulfur dioxide.  Simultaneous operations of TCM bubblers provided
a measure of ambient air concentration of sulfur dioxide.

Sulfur dioxide content of the exposed inserts was determined by extracting
their contents in TCM solution using a laboratory blender, neutralizing
with hydrochloric acid, and processing by means of the West and Gaeke
technique.

Since the equipment was not amenable to aircraft use, operation was limited
to times when the plume approached ground level.  Of five attempts at
measurement during the fall period, only one was successful in sampling a
reasonably concentrated region of the plume.  This experiment, conducted
on the morning of October 29, is summarized as follows:

        Average sulfur dioxide concentration in air        0.034 ppm
        Amount of sulfur dioxide passed into each
        precipitator during sampling period               19.9   ymoles
        Amount of sulfur dioxide retained by pre-
        cipitator - High voltage "ON"                      0.748 pinole
        Amount of sulfur dioxide retained by pre-
         ipitator - High voltage "Off"                     0.764 ymole

These results indicate that very little sulfur dioxide existed in sorbed
form in the atmosphere.  The fact that the "voltage off" value exceeds
the "voltage on" value arises from scatter in the experimental results.

If the reliability of the measurements is assumed to be about ± 10 percent,


-------
less than 0.4 percent of the sulfur dioxide existed in the sorbed state.
This result is in accordance with previous laboratory measurements, which
have indicated that the proportion of sulfur dioxide sorbed on particulate
should be quite small under most circumstances.   '

The sampling of near surface air to determine sulfur dioxide concentrations
was performed with the Sign-X Analyzer mounted on a truck.  Traverses of
the arcs were made during the light rain early in Run 20 and during the no-
rain condition of Run 21.

Two peaks were measured on Arc C during the rain condition traverse just
prior to frontal passage.  Passage of the front was recorded at 1555 by a
wind shift at the Overlook from 230° to 290°.  Figure 4.2 shows the air
concentrations, measured in that traverse, offset for the 60° wind shift
and superposed on the precipitation concentrations of sulfur dioxide which
were determined.  There appears to be considerable similarity between these
curves.  The bimodal distribution suggested here may be evidence of the
plume splitting which was inferred by the washout distributions of Runs 2
and 6.

The concentrations of sulfur dioxide in air and in the collectors for Run
21, the dry run, are shown in Figure 4.3.  Bimodality is shown here also.
Observing the higher concentration on Arc B than on Arc A in the afternoon
(p.m.) traverse, it is somewhat surprising  that the dry deposition isn't
significantly higher on Arc. B.  However, the dry run extended over six
hours from mid-morning when another traverse of Arc A (a.m.) showed much
higher values than in the afternoon.  When the plume is mixed to the ground
in a short distance, as in the morning traverse, lower air concentration
values would be expected at greater distances.  Thus, the nearly equal values
of dry deposition on the two arcs result from differences in mixing through
the sampling period which tended to equalize exposures on the two arcs.

The mobile sampling of near-surface air concentration proved quite success-
ful in relating this 'parameter to washout and dry deposition of sulfur
dioxide and in documenting the plume bimodality.

-------
   30
   25
o
s

§  20
z
o
   15
<
CtL
UJ
O
O
CJ
x
o
£  10
         —WASHOUT  CONCENTRATION

         	 AIR CONCENTRATION
                                                                0.04
                           0.03
      /     \~'
           0.02
                                                                0.01
                     I
I
I
                   60              120             180

                    BEARING FROM KEYSTONE,  DEGREES



             Figure 4.2   Sulfur Dioxide Concentrations  in
                         Air and  in Precipitation, Run  20

-------
                                                                  JO.06
      2  H
 a
 ae
                                  AIR  CONCENTRATION
      3  H
                      40           50            60
                     BEARING  FROM  KEYSTONE,  DEGREES
70
Figure 4.3   Sulfur  Dioxide Air Concentration and "Dry Deposition", Run 21


-------
                                CHAPTER V

                          DISCUSSION OF RESULTS

Chapter III presented the development of a physicochemical theory for
sulfur dioxide washout which included mechanisms that were not considered
in the preexisting theories of gas washout by natural precipitation.
Chapter IV pointed up the extremely complex washout behavior which was
observed in the field experiments.  Chapter V discusses the significant
field results and relates them, where feasible, to the revised theory.
The theory alone provides insight into the behavior which can be expected.
It might be well, therefore, to present first those considerations which
have an effect on the overall results of the field experiments.

APPLICATION OF THE THEORY

Two mechanisms which are of particular importance to the washout of sul-
fur dioxide are the reversibility of the mass-transfer process and the
solubility control exerted by the hydrogen ion concentration, or pH, of
the rain.  Consequences of each of these mechanisms are discussed relative
to the washout of sulfur dioxide from an elevated plume.

ABSORPTION-DESORPTION OF SULFUR DIOXIDE —  The reversible nature of the
mass-transfer process implies the existence of some degree of liquid-phase
resistance not accounted for in the preliminary theory.  Moreover, the
desorption of sulfur dioxide in less concentrated regions beneath the
plume should be expected to lower the net washout rates.  On the basis of
Figures 3.7, 3.8 and 3.9, one would expect desorption to be particularly
prominent for plumes at the height of the Keystone stacks (800 ft), but
much less so for plumes at lower heights.  This helps to explain the much
larger, irreversible, washout coefficients observed in the small scale
experiments discussed in Chapter II.  If fall distance beneath the con-
centrated region of the Keystone plume is comparable to the stack height,

-------
desorption frqjn all but the largest raindrops should result In precipita-
tion concentrations of sulfur dioxide in equilibrium with ground-level air
concentrations.  Thus, one can expect that increased plume height will
result in decreased washout, which is contrary to the prediction of the
preexisting theory that washout is independent of plume height.  In addi-
tion, increased vertical dispersion should enhance washout for a given
elevation of the plume centerline.  Conditions can be postulated having
the opposite effect, but they are unlikely under circumstances exemplified
by the Keystone plume.  The enhancement of washout by increased vertical
mixing at fixed plume elevations may be viewed simply as a consequence of
lessening the distance of raindrop fall "beneath" the plume, thereby pro-
viding less opportunity for desorption.

It is of interest also to examine the effect of lateral dispersion of the
plume.  For a fixed plume height and a constant intensity of vertical
mixing, increased lateral dispersion will both dilute and broaden the
plume so that it will be encountered by a greater number of drops.  The
significance to the total washout depends on which effect is more pro-
nounced.  Preexisting theory predicts that they are compensating.  In
exploring this question, it is convenient to visualize first the simpli-
fied situation of a deep plume, bordering the ground, of uniform concen-
tration c   .  Rain falling through such an idealized plume will contain
sulfur dioxide at a saturation value c.   ,  corresponding to c.  .  If
                                      Ax]/        r     °     Ay i
lateral dispersion is increased so that  the plume is widened to twice its
original width, the concentration will be lowered to

                             c.   = CA  12
                              Ay     Ay..

and twice as many drops will encounter the  plume.

Referring to Figure 3.6, it is observed  that, because of the curvature of
the equilibrium line,

-------
                              ,   > c.  /2
                              Ax»    Ax-
hence more sulfur dioxide will be removed from the wider, more dilute
plume.  Such reasoning can be extended to situations wherein equilibrium
is not attained by the drops.  Here, again because of the curvature of
the equilibrium line, mass-transfer driving forces will tend to decrease
with concentration in a manner such that washout is enhanced by increased
lateral dispersion.

It is well to note here the differences between the effects of vertical
and lateral dispersion.  Increased vertical dispersion enhances washout
by reducing the fall distance beneath the plume and thereby reducing the
sulfur dioxide desorption.  Lateral dispersion results in a lowering of
drop  concentration, but a net increase in washout still occurs owing to
exposure to a greater number of drops.

Although plume dispersion is caused primarily by turbulent mixing, an
additional mechanism which arises from desorption and fractionation of
the precipitation, after plume encounter, may be important.  The wide
range of terminal velocities of the individual precipitation elements
results in a wide range of trajectories following plume passage.  This is
amplified under conditions of vertical shear in the wind field.  Lighter
elements, in particular, are carried large distances after exposure in
the plume.  Consequently, the plume is broadened and deepened thereby
enhancing the washout.  Furthermore, washout at the ground will be related
to the complex set of trajectories and exhibit a generally erratic appear-
ance.

The above conclusions can be summarized as follows:

    1.  For a given plume-centerline elevation, increased mixing in any
        direction will enhance sulfur dioxide washout.
    2.  For constant conditions of mixing, increased elevation of the

-------
         plume centerline will tend to retard sulfur dioxide washout.

For situations encountered during this study, the effects of variations of
plume centerline elevation with distance would be expected to be minimal
compared to those arising from increases in dispersion.  Hence, a general
increase in washout with downwind distance should be expected — an effect
that has been observed quite regularly in the field results.  It is prob-
able that this effect is much more dramatic at still greater distances
than were sampled in this study.  Moreover, even though washout rates
observed in the first four miles downwind were rather meager, this mech-
anism of sulfur dioxide removal may become highly significant at greater
distances from the source.

SOLUBILITY OF SULFUR DIOXIDE ~  Figure 3.10 shows how the pH of water
establishes the degree of sulfur dioxide solubility.  Consequently, sulfur
dioxide levels in rain should decrease with decreased pH of the rain.  pH
is determined by the dissolved, nonvolatile compounds which are taken up
by the rain during its formation and fall.  The pH is decreased, generally,
in its passage through the plume by the washout of nonvolatile acid-form-
ing compounds, mainly sulfates and nitrates.

The influence of the pH of the incident rain should be noted.  Nonvolatile
material should be washed-out irreversibly and the amount of pickup should
be roughly constant for similar plume and rain morphologies.  However, pH
is the negative logarithm of hydrogen ion concentration, so that a given
amount of acid-forming material will be more significant in reducing the
pH if the incoming rain has a high pH.

The significance of the rain pH is that low values will limit the solubil-
ity of sulfur dioxide thereby reducing its absorption or enhancing its
desorption.   The greatest washout should occur from a pure sulfur dioxide
plume in a clean atmosphere and far less from a typical power plant plume
in an already polluted atmosphere.

-------
SULFUR DIOXIDE WASHOUT

There were few, if any, runs in which the sulfur dioxide distributions
were sufficiently well-defined for the calculation of washout coefficients.
Given a single source of known output and containment of its plume, this
is a straight-forward calculation under the assumption of an irreversible,
first order process.  However, none of these requirements were met in this
study.  Background values were often high and variable presumably because
of a number of large and small sulfur dioxide sources in the area.  Con-
sequently, separation of the component of washout from the Keystone Sta-
tion plume was generally not possible.

In those cases, where sulfur dioxide washout could be related to other
sources on the basis of the observed wind field, containment was incom-
plete and the sampler configuration was not suitable for the calculations
of the washout coefficient from those sources, either.  However, where
components of the distribution were identifiable with sources, either
Keystone or others, the observed peak values of concentration were taken
for analysis.

These values together with the precipitation volume collected and the
duration of exposure of the collector were sufficient to define the wash-
out flux of sulfur dioxide and the concurrent precipitation rate.  A
summary of these data is presented in Table 5.1 together with identifica-
tion of the probable sources of the concentration maximums.  Specific
sources are the Keystone and Homer City generating stations and the
Lucerne coke ovens.  A general source identified as "Area" is noted for
those run distributions which were sufficiently homogeneous to indicate
the entire grid was blanketed from diffuse sources at a moderate distance.

Figure 5.1(a-c) show the washout flux as a function of precipitation rate
for the various sources.   The two identified sources, Homer City and
Lucerne, are plotted together.  The unidentified area sources and Keystone
are each plotted separately.  The relationship between the flux and the


-------
                                TABLE 5.1
                 PEAK WASHOUT CONCENTRATIONS AND FLUXES
                                                   SO,
Run  Station
No.

 1
 1
 2
 2
 3
 3
 4
 4
 5
 5
 6
 6
 7
 7
 8
 8
 9
 9
10
10
11
11
12
12
14
14
15
15
16
16
18
19
20
22
No.

A13
B12
A64
B89
A51
B69
A60
B63
A57
B68
A22
B42
A2
B90
A2
B58
A50
B47
A68
B30
A38
B47
B60
B46
B64
B64
B64
B64
C22
C14
C29
Cll
C18
B24
 Apparent
  Source

Area
Area
Homer City
Homer City
Homer City
Homer City
Homer City
Homer City
Homer City
Homer City
Keystone
Keystone
Homer City
Homer City
Homer City
Homer City
Area
Area
Area
Area
Area
Area
Keystone
Lucerne
Keystone
Area
Keystone
Area
Keystone
Lucerne
Area
Area
Area
Keystone
 Cone.
.ymole..
'•liter'1
  6.5
  9.8
  8.8
 16.7
  9.0
  6.5
 12.3
 10.4
 10.8
  8.8
  1.7
  4.4
 12.8
  7.2
  9.5
 12.1
 13.2
 11.8
  8.8
  9.8
  7.8
 11.1
  3.5
  3.2
  5.0
 16.0
  2.0
 14.0
 13.1
 27.8
  1.6
  8.7
 11.0
  0.6
       Exposure
Volume Duration
 (ml)    (min)
  74
  66
  52
  32
  16
  20
 122
  89
  91
  77
  83
  43
 134
 149
  72
  27
  72
  90
  92
 106
  65
  84
  24
  25
 245
 245
 225
 225
  28
  20
 262
  50
 212
 118
 56
 55
122
126
117
114
 68
 66
107
 98
139
136
 86
 98
 73
 51
 81
 93
 94
 96
115
129
139
164
107
107
123
123
 25
 23
 76
 47
 72
215
Flux
m -hr
5.7
7.9
2.5
2.8
0.8
0-8
14.7
9.4
6.1
4.7
0-7
0-9
13.3
7.3
6.2
5.0
7.8
7.6
5.8
7.2
3.0
4.8
0.4
0.3
7.6
24.4
2.4
17.1
9.9
16.2
3.7
6.2
21.6
2.2
r i ecj.pj.LciLj.uii
Rate
,mm.
0.88
0.80
0.28
0.17
0.09
0.12
1.20
0.90
0.57
0.52
0.40
0.21
1.04
1.01
0.66
0.35
0.39
0.65
0.65
0.74
0.38
0.43
0.12
0.10
1.53
1.53
1.22
1.22
0.75
0.58
2.30
0.71
1.96
0.36

-------
    20
CM
 I
 S-
 s:
 CO
 LU
 _l
 o
 o
 :E
 CO
   0.3
  x   ARC A
  •   ARC B
  D   ARC C
CIRCLED VALUES ARE
WASHOUT BY  SNOW
      0.05     0.1                          1
              PRECIPITATION  RATE, mm-hr"1
Figure 5.1(a)  Peak Sulfur Dioxide Washout Flux From Keystone Plume

-------
     30
        - x
HOMER CITY  GENERATION  STATION

LUCERNE  COKE OVENS
         CIRCLED  VALUES ARE
         WASHOUT  BY SNOW
                                i   I  i  i i  i I
               0.1
                 PRECIPITATION RATE,  mm-hr
                             1 .0
                               -1
3.0
Figure 5.l(b)  Peak Sulfur Dioxide Washout Flux From Other Local Sources

-------
     50
  CM
   I
   E


   s-
   o
   3:
   CO
    0.5
            CIRCLED VALUES ARE  WASHOUT BY SNOW
    '\
I  i  I I I
        0.01    0.1
J  I  I I I
                                   1
                  PRECIPITATION  RATE,  mm-hr
                                              -1
Figure 5.1(c)  Peak Sulfur Dioxide Washout Flux From Unidentified Sources

-------
precipitation rate is shown best by the Homer City washout contributions,
Figure 5.1(b), which are distributed over a wide range of precipitation
rates.  The slope of the line, which was drawn as a visual best fit, is
measured as 1.23.  This slope is seen to be compatible with the data from
the other sources as well.

The earlier works by Chamberlain     and Engelmann     show slopes less
than unity for the theoretical relationship of washout coefficient and
precipitation rate.  Various raindrop size spectra were used in the cal-
culation of their curves which were based on a first order, irreversible
model.  The selected raindrop size spectra shown in Figure 5.2 which were
taken during washout runs at Keystone, are not markedly different from
spectra that they employed.  The significant difference between the
observed slope of Figure 5.1 and the slopes defined by preexisting theory
testifies to the inadequacy of the first order, irreversible approximation
for the washout of sulfur dioxide.  As shown in Chapter III, many factors
which are not accounted for in the simplified preexisting theory affect
the gain and loss of sulfur dioxide by the precipitation.

The curve through the washout flux data for sources at Homer City and
Lucerne is shown on the other figures to provide ready comparison of the
three sets of data.  Although the curve fits the data from the diffuse
sources of Figure 5.1(c), it is seen to lie above all but one of the data
points in Figure 5.1(a) which were attributed to the Keystone Station
contribution.   This one point which lies within the grouping of data for
the other sources is from a collector position on Arc C, approximately 4
miles from Keystone.  The lower values for Arc A and Arc B samplers are
attributed to the low near-surface concentration of sulfur dioxide at
short distance from an elevated source.  All data points in Figure 5.1(b)
represent fluxes 10-15 miles from their sources where surface concentra-
tions can be assumed to have been increased by the downward diffusing plume.

No apparent difference between washout of sulfur dioxide by rain and by

-------
VO
V
     1   —
eg
LU
I—
LU

•a:
I—I
Q

D_
o
   0.2
                                                       0.5mm-hr
                                                                  -1
                                   i   I  Illilll
I	I   I    I	I     I
        0.1         1
                                              10
        50
90

-------
snow is seen in Figures 5.1(a-c).  However, when the values for snow are
normalized to unit precipitation rate using the experimentally determined
slope, Runs 6 and 12 demonstrate appreciably lower values as is shown in
Table 5.2.
                                TABLE 5.2
               SULFUR DIOXIDE WASHOUT FLUX IN SNOWFALL
                                                      (a)

Date

2-03-70

2-09-70



2-10-70





2-14-70

Run
No.

6
6
7
7
8
8
9
9
10
10
11
11
12
12
Station
No.

A-22
B-42
A-2
B-90
A-2
B-58
A-50
B-47
A-68
B-30
A- 38
B-47
B-60
B-46
Apparent
Source Temperature
(°F)
Keystone 15-20
Keystone
Homer City 35
Homer City
Homer City
Homer City
Area 33
Area
Area
Area
Area
Area
Keystone 20
Lucerne
Sulfur Dioxide
Flux
(pinole- m~2-hr 1 )
2.1
6.5
12.5
7.2
10.2
17.5
14.9
13.0
9.9
10.3
9.6
13.1
5.0
5.5
(a)
   Flux values normalized to unit precipitation rate.
Several explanations are possible for this observed difference.  Certainly
a factor on Arc A of Run 6 is the aforementioned necessity for the plume
growth to be sufficient to raise the near-surface sulfur dioxide concen-
tration.  The large 'difference between Arc A and Arc B concentrations
indicates that the plume had not reached the ground in sufficient

-------
concentrations on Arc A.  An inversion was present to well above the
stack in this run and such behavior would be expected.  During Run 12,
however, neutral stability was indicated by the Pittsburgh rawinsonde and
the mixing should have been adequate.  The washout flux from the Lucerne
coke ovens, a surface source, is comparable to that from the Keystone
Station in Run 12.

It should be noted that Runs 6 and 12 both occurred on days with the
temperature well below freezing, whereas, the other snow runs were con-
ducted with surface temperatures above freezing.  Snow characteristics
consequently were different, and this is shown by Figure 5.4(a-d),
photographs of the snow on each of the snow run days.  Figures 5.4(a) and
(d), which were taken during Runs 6 and 12, respectively, show the dry,
crystalline character of the snow during these runs.  Figures 5.4(b) and
(c) show, as clearly, the wet amorphous nature of the snow during Runs
7 and 9.  The light shading across Figure 5.4(b), in fact, is melt water
or rain which also fell.

Whereas sulfur dioxide washout by wet snow, falling in near-freezing con-
ditons, is comparable, apparently, to washout by rain, it is suggested
that dry snow is less effective in the removal.  Whether the effect is
solely one of temperature cannot be determined from the data which are
grouped closely both above and below freezing.  It is possible that higher
washout is associated with a water film on the snow.  Further investiga-
tion of this effect is necessary.

Some comparison of the observed fluxes with those predicted by the pre-
existing model are possible recongnizing that Equation 2.4 can be written
as,

                         Em /NAt  = AR/(NAYu)                       (5.1)
                           i

where R is the rate of emission of sulfur dioxide, t is the sampling time

-------
                              (d)  RON 12

-------
and N is the number of samplers which receive washout from the emitter.
The other symbols were defined earlier.  The term on the left is the
average flux.  Using the washout coefficient of IQ-4 sec"1 given by
           (52)
Chamberlain     as appropriate for a precipitation rate of 1 mrn-hr'1 and
assuming an emission rate of 3 x 103 gm-sec'1, a wind speed of 10 meters
sec'1 and a plume width of 1500 meters on Arc B, a flux of approximately
1000 ymoles (n^hr)"1 is calculated.  Observed values were seen to be
less by factors of 50 to 100.

As noted earlier, Run 21 was conducted to ascertain the direct influence
of sulfur dioxide concentrations on the collectors — a measure of the
dry deposition.  If it is assumed that the peaks observed in Figure 4.3
were due to the Keystone plume alone, except for the observed background
of 0.1 pinole-liter"1, the peak value represents a mass of 0.016 ymole
delivered to the collector in four hours.  At the deposition rate of
0.004 pmole-hour"1, few runs would be influenced by the effect.  Exceptions
are collectors which were exposed for long periods in light, intermittent
precipitation.  It should be noted, however, that the mobile sulfur
dioxide analyzer, which was operated at the station receiving the peak
deposition, showed air concentrations between 0.05 and 0.1 ppm, values
which were much higher than detected during washout runs.

Figure 5.4 presents the sulfur dioxide concentration data of Table 5.1
plotted against the total time of collector exposure — not the precipi-
tation periods only as tabulated there.  The peak value of  27.8 ymole-
liter"1 is the one most likely to have been increased by dry deposition
of sulfur dioxide.  This collector was exposed for 207 minutes and
collected 20 ml of precipitation.  Using the deposition rate determined
from Run 21, the concentration resulting from dry deposition is determined
to be 0.7 pmole-liter-1, which is only 2.5 percent of the observed con-
centration.  Consequently, dry deposition is not considered a serious
problem.

-------
  50[- WASHOUT FROM KEYSTONE  PLUME  ON:

                  A  ARC A

                  0  ARC B

                  V  ARC C
on
LU
  10
o
z:
3.
o
z
o
  0.5
      •

  •    <

i*   ••
                             11
                                A
                                              i
              50        100       150       200        250       300

                            EXPOSURE TIME, min
         Figure 5.4    Comparison of Peak Sulfur Dioxide Washout
                      Concentration and Collector Exposure Time

-------
The grouping of values near 10 ymole-liter-1  in Figure 5.4 emphasizes
the dependence of washout sulfur dioxide  concentration on surface air
concentration and precipitation acidity as discussed previously.  This
value is in equilibrium with air concentrations of 0.01  to 0.03 ppm when
the precipitation has a pH value between  4.0  and 4.5.  These ranges of
air concentration and pH were observed.

A further observation emphasized in Figure 5.4 is the necessity for
sampling the Keystone plume at greater distances where its proximity to
the ground results in greater washout.  Only  the Arc C measurement of the
Keystone plume shows a washout concentration  comparable  to those from
more distant  sources of presumably lower emission rates.

It is useful to compare surface concentrations resulting from various
sources in light  of the importance of the sulfur dioxide concentration
near the ground in washout.  The method of Smith and Singer     was used
for this comparison.  A neutral atmosphere bounded by stability Classes B..
and C and a wind speed of 5 meters-sec"1  at all levels were assumed.  For
the stack release, a plume rise of 600 ft was determined using the method
                              ( 62^
suggested by Carson and Moses.      The heat  emission rate was taken as
2.5 x 104 kilocalories/sec, an approximate value appropriate for the
Keystone or Homer City Stations.  The results are given in Table 5.3 in
terms of parts per million of sulfur dioxide  per ton released per hour.
Surface concentrations can be calculated  if the source term is defined.
This can be done readily for the generating stations, but the output from
the coke ovens is unknown.  The similarities  in observed washout values
and in calculated values for air concentration for Homer City and Lucerne
suggests that their emission rates are comparable.

The extremely low value at Arc A resulting from the Keystone plume under
the more stable Class C is due to the low diffusion rate which prevents
mixing of the plume to the ground at that distance.  Other values in the
table yield concentrations of 0.01 to 0.2 ppm for typical emission rates

-------
                               TABLE 5.3


                                                     (a)
                COMPARISONS  OF SURFACE  CONCENTRATIONSv/
                   UNDER NEUTRAL STABILITY CONDITIONS


                                    	Stability  Class -
         Condition(b)                     Bl
             1                        3.47 x  10~3      2.19  x 10-1

             2                        1.82 x  10-2      9.91  x 10-1*

             3                        9.81 x  10~3      1.48  x ID-3

             4                        1.31 x  10~3      8.60  x 10-3

             5                        1.34 x  10-3      1.13  x 10~2
        Values are in units of ppm-hr-ton"1.  The product of  these
        values and the source strength in tons-hr~' is the surface
        concentration in ppm.


        Condition 1.  Concentration at Arc A from Keystone source
                      (4,000 ft)
        Condition 2.  Concentration at Arc B from Keystone source
                      (12,000 ft)
        Condition 3.  Concentration at Arc C from Keystone source
                      (22,000 ft)
        Condition 4.  Concentration at Arc B from Homer City  source
                      (80,000 ft)
        Condition 5.  Concentration at Arc B from Lucerne source
                      (80,000 ft, ground release)
of 10-20 tons per hour from the generating stations.  These concentrations

are appropriate to retain sulfur dioxide concentrations of 1 to 50 ymoles-

liter"1 in the precipitation under equilibrium conditions.


A prominent feature of the distributions of washout sulfur dioxide con-

centration is the absence of positive correlation with the Keytone plume.

This, apparently, is a consequence of the shallowness of the plume at

short distances from the source which permits the least time for sulfur >

-------
dioxide sorption and  the greatest  time  for  its  desorption.  An  additional
feature, less easy to reconcile, is  the negative  correlation of washout
beneath the plume.  This was  observed in Runs 4 through  11.  Except for
Run 6, which is discussed  later, these  "troughs"  occurred in a  homoge-
neous distribution of sulfur  dioxide washout from other  sources.  It
appears, therefore, that the  washout of sulfur  dioxide from these sources
was inhibited in some way  by  the Keystone Station, or that the  sulfur
dioxide was converted to sulfate.

SULFATE WASHOUT

Sulfate washout values were examined for the runs which displayed
decreases in sulfur dioxide washout beneath the expected, plume position.
Peaks were noted at or adjacent to all  stations which had minimum sulfur
dioxide washout values.  Sub-plume and  background washout flux values  of
sulfate and sulfur dioxide are shown in Table 5.4.  The value from a
single station was tabulated  if the maximum sulfate and minimum sulfur
dioxide fluxes occurred at the same  collector station.  If they occurred
at adjacent stations, the  average value for the two stations was used.
In all cases the background value was determined by the average of values
at the three stations on each side of the central station or pair of
stations.

The sulfate washout flux attributed  to  the plume is the difference between
the sub-plume and the background fluxes.  It is uncertain that the plume
sulfate washout flux  is derived totally from the stack emission.  It is
possible that background sulfate washout is enhanced by exposure to the
plume.  The additional humidity within  the plume may, for example, con-
dense on the sulfate.  More effective washout would result for the droplet
than for the submicron sulfate particle.  However, the sub-plume washout
flux diminished by the background washout flux is referred to here as  the
plume washout flux.

-------
                                                 TABLE 5.4
o
ho
Run
No.


 4
 4
 5
 5
          7
          7
          9
         11
         11
FLUXES OF WASHOUT SULFATE AND
Station
No.
Precipitation
Rate
, , -1,
(mm-nr ;
SULFUR DIOXIDE
Sulfur Dioxide Flux
Background
/ -• -2
(ymole m
Sub-Plume
hr'1)
Sulfate
Background
(mg m 1
Flux
Sub-Plume
ir )
RAIN
A-55
B-79, B-80
A-53, A-54
B-78

B-52
B-72
B-22
A-18
B-24
1.16
0.79
0.54
0.42

1.38
1.18
0.77
0.34
0.34
7.10
5.51
4.12
1.12
SNOW
8.02
2.35
5.64
' 1.60
1.79
2.36
1.99
1.03
0.10

5.93
1.26
2.66
1.01
1.24
2.79
2.49
1.60
1.20

6.30
5.52
2.94
1.73
1.08
4.24
3.16
2.30
1.74

25.10
12.50
6.18
- 2.64

-------
Washout fluxes for both the background sulfate and the plume sulfate are
shown in Figure 5.5.  The background values in both rain and snow and the
plume values in rain are observed to be  linearly related to the precipi-
tation rate.  This relationship has been observed previously for aerosol
washout.  '     The washout of sulfate from the plume by snow, however,
appears to be dependent on the square of precipitation rate.  Unquestion-
ably, this relationship stems from an interaction between the snow and
some special property of the plume.  This property may be the humidity
which provides high supersaturation relative to the snow.

The question of sulfate origin cannot be answered conclusively, but it
is possible to estimate the extent of sulfur dioxide oxidation required
to explain the observed sulfate flux.  Equation 5.1 can be rewritten as

                             JR1 = FNAYu/A                          (5.2)

where, R1 is the sulfur, as sulfate, emission rate and F is the mean sul-
fur, as sulfate, washout flux over the plume width, NAY.  Calculations
were made using run data including flux  values from Table 5.4 which were
normalized to unit precipitation rate with appropriate slopes for rain
and snow.  The washout coefficient value of 5 x 10~6sec~1, which is
appropriate for submicron particle washout at unit precipitation rate,
was assumed.     These results were then divided by the Keystone Station
sulfur emission rate during the runs to  obtain the percent oxidation
required.

Figure 5.6 compares the calculated oxidation as a function of travel time
to the collector with oxidation data presented by Gartrell, et al.
who measured oxidation at various distances with an instrumented aircraft.
Their finding, that catalytic oxidation  either occurs at a high rate,
1-2 percent per minute, or does not occur, was borne out by Baldwin,
et al/65^ and Arin, et al. (66) who found essentially no change in sulfur
dioxide beyond one kilometer.  The overall agreement of calculated

-------
    20
    10
s-
.c
en
E
o
n:
to
   0.5
                                                SLOPE = 2
                                             SLOPE = 1
SULFATE WASHOUT FROM PLUME
        • IN RAIN

        x IN SNOW
                          .    I  ....  I
                                                SLOPE  =  1
                                            BACKGROUND SULFATE WASHOUT
                                                    • IN RAIN

                                                    x IN SNOW
       0.1
                                      1 .0
                           10
                          PRECIPITATION RATE, mm-hr"1
              Figure 5.5   Sulfate Washout Fluxes

-------
   50
   40
to
o
Q
UJ
M
X
o
x
o
a
a:
   30
   20
CO

It-
CD

h-
Z
UJ

S  10
UJ
Q.
-  X  RESULTS OF GARTRELL.et.al.[63]
   •  CALCULATED FROM  SULFATE WASHOUT
      FLUX MEASUREMENTS  USING WASHOUT
      COEFFICIENT  OF 5xlO"6  SEC.
   4-  DENOTES SNOW WASHOUT
                  X  X1
                               _L
                              10           15

                             TIME,  MINUTES
                                                 20
25
             Figure 5.6   Rate of  Sulfur Dioxide Oxidation

-------
oxidation with the results of Gartrell is striking.  The generally higher
             »
calculated values which are derived from the snow washout flux data may
be due either to enhanced oxidation or to more efficient washout.  The
use of the submicron washout coefficient would be inappropriate if con-
densation on the sulfate nuclei resulted-in substantial growth.

Assuming that only one to two percent of the sulfur is emitted as sulfate,
*  ' it appears probable that the observed sub-plume sulfate resulted
from catalytic oxidation of sulfur dioxide and that Its washout was
enhanced by condensation of plume water vapor on the sulfate nuclei when
temperatures were near freezing.  The fact that maximum decreases in sul-
fur dioxide occurred during rain, and maximum increases in sulfate
occurred during snowfall indicates that the two are not well related.
However, the decreased sulfur dioxide in the sub-plume collectors is
likely a consequence of the low pH of the precipitation caused by the
sulfate acidity.  More sulfur dioxide might be retained in a snowflake
than in a raindrop because the sulfate particle sorbed on the former
would exhibit a local effect, whereas in a raindrop, it would mix and
totally inhibit the sulfur dioxide retention.  Higher background s'ulfate
washout flux in snowfall probably is due to the higher sulfate concentra-
tion in the air during winter as reported by Junge.

It appears from the foregoing that, whereas the washout of sulfur dioxide
from the Keystone Station may be light near the source, sulfate washout
may be appreciable.  For example, Table 5.2 shows that the washout flux
of sulfur dioxide is on the order of 10 ymoleCn^hr)"1, while Figure 5.5
shows sulfate washout flux of 5 mgdi^hr)"1 for a comparable precipitation
rate.  Comparing the sulfur removal, the sulfate washout accounts for
about five times as much as the sulfur dioxide washout.  As noted earlier,
the sulfur dioxide washout should increase with distance, whereas the data
of Table 5.4 suggest that sulfate washout will decrease with distance.

-------
                              CHAPTER VI

                              CONCLUSIONS

This investigation of sulfur dioxide washout from plumes emitted by tall
stacks has provided significant new information on the washout of gaseous
effluent.  The theoretical development of Chapter III has provided insight
into the physicochemical influences which are important in sulfur dioxide
washout.  It has shown that liquid phase phenomena can substantially alter
the washout effectiveness of rain.  Pates of sulfur dioxide sorption and
desorption are dependent on the size and chemical composition of the rain-
drop as well as on the instantaneous concentration of sulfur dioxide in
the air.  Highly acid  (low pH) rainwater has a markedly reduced affinity
for sulfur dioxide, which is highly soluble in a neutral solution.  Large
drops rapidly reach their equilibrium concentration only if vigorous
mixing is present within the drop, whereas small drops attain equilibrium
rapidly by molecular diffusion alone.  In the absence of an irreversible
chemical reaction to retain the sorbed sulfur dioxide, it will desorb in
a region of lower sulfur dioxide concentration.  Thus, drops falling
below the plume will continually desorb their sulfur dioxide content.
The smallest drops collected will be essentially in equilibrium with the
surface air, while larger drops will retain some "memory" of the plume.

It follows, therefore, that washout from a plume is highly dependent on
the geometry of the plume and the characteristics of the rain.  Clean
rainwater falling through a low elevation, concentrated plume will exhibit
concentrated washout, while acid rainwater falling through either a low
or a high elevation, concentrated plume will fail to record its presence.
However, if the plume from a high elevation source is spread in the ver-
tical so as to approach the ground, clean rainwater will remove substan-
tial sulfur dioxide from it as well, since a large change in air concen-
tration will result in a smaller change in equilibrium concentration.
This fact also results in greater depletion of the airborne sulfur dioxide

-------
in a plume which is widely dispersed laterally since the areal coverage
over-compensates for the reduced concentration.

It should be emphasized that appreciable sulfur dioxide cannot be removed
unless surface air concentrations are substantial and the precipitation
acidity is low.  Consequently, the maximum washout of sulfur dioxide
emitted from a tall stack is likely to occur at the distance of maximum
surface air concentration when background air pollutants do not lower the
precipitation pH.  Furthermore, since other pollutants from the stack
emission tend to decrease the precipitation pH, they also inhibit sulfur
dioxide washout.  It has been concluded from the theoretical analyses
that direct gas absorption is the most important washout mechanism.
Washout of aerosols containing sorbed sulfur dioxide is of negligible
consequence in this respect.  Precipitation as snow has not been consid-
ered in the theoretical development, but from the experimental results,
similarities in the washout of sulfur dioxide are evident.

Analysis of the experimentally determined fluxes of sulfur dioxide has
shown a relationship to precipitation rate which is inconsistent with
the first order, irreversible assumption used in the preexisting theory
of gas washout.  Whereas, the theory showed washout increasing at a lower
rate with high precipitation rates, the experimental data show dispropor-
tionately more effective washout at higher precipitation rates.  This
inference of higher washout efficiencies for larger drops is in agreement
with the theoretical development of Chapter III.  There was no apparent
stratification of the data according to the precipitation type — both
rain and snow washout was equivalent for a given precipitation rate, as
noted above.  Washout flux was estimated to be proportional to the 1.23
power of precipitation rate.  Closer scrutiny of the snowfall data did
indicate lower washout flux values during dry snow runs, but too many
variables are present to draw conclusions from those sparse data.

Washout flux values were consistently less than predicted from preexisting

-------
theory and lowest values were associated with emissions from the Keystone
Station, except in one instance.  On arc C, flux value's comparable with
other sources were observed.  Although this arc is 4 miles from the
Keystone Station and suitably located to receive substantial air concen-
trations of sulfur dioxide under some conditions, it is suspected that
a portion of this flux value can be attributed to an upwind source.
Fluxes from upwind sources — some identified and some not —  were
clustered about a value of 10 ymolesfti^-hr)"1 normalized to a precipita-
tion rate of 1 mm hr"1.  For a pH value between 4.0 and 4.5 this value
is in equilibrium with air concentrations of 0.01 to 0.03 ppm.  These
ranges of air concentration and pH were observed generally, though higher
concentrations were observed and also computed using atmospheric diffusion
techniques.  Maximum observed fluxes were below 30 ymoles^-hr)"1.
Fluxes up to two orders-of-magnitude greater are predicted by the pre-
existing theory.

An interesting feature of several runs was increased sulfate and decreased
sulfur dioxide washout fluxes, compared with background values, in a
region beneath the expected trajectory of the Keystone plume.  The
decrease in sulfur dioxide flux is not sufficient to account for the
increase in sulfate flux, nor are they quantitatively related in any
simple way.  Maximum decrease in sulfur dioxide flux occurred in rain,
while maximum increase in sulfate flux beneath the plume occurred in
snowfall.  Decreased pH from sulfate sorption apparently caused the sul-
fur dioxide deficiency.  It is uncertain that all of the sulfate was
derived from the Keystone Station, although calculations of the required
sulfur dioxide oxidation in the plume were in general agreement with
oxidation rates observed by others.  Highest values of sulfate which
occurred beneath the plume during snowfall are attributed to rapid
oxidation of the sulfur dioxide and to enhanced washout of droplets which
condensed about the sulfate particles as nuclei.  A portion of these
sulfate nuclei may have been contributed by upwind sources.

-------
The different? nature of the sub-plume sulfate flux during snowfall is
demonstrated by the relationships of sulfate flux to precipitation rate.
Both background and sub-plume sulfate washout fluxes in rain and the
background flux in snow were linearly related to precipitation rate.  How-
ever, the sub-plume, sulfate washout flux in snow increased as the square
of the precipitation rate.  Consequently, there appears to be a unique
interaction between the plume and snow or between the plume and precipi-
tation during near-freezing conditions.  As a result, if it is assumed
that sub-plume sulfate flux minus background sulfate flux can be attri-
buted to the Keystone Station, the washout of sulfur as sulfate, is
greater than the washout of sulfur as sulfur dioxide close to the
elevated source during snowfall.

It has been shown, that sulfur dioxide washout from distant, major sources
and even from a number of minor, disperse sources readily obscures the
washout from a major, elevated sulfur dioxide source nearby.  This is
largelv a consequence of the low retention of sulfur dioxide by precip-
tation after its passage through an elevated plume.  Therefore, tall
stacks exercise a positive action in reducing the maximum concentration
of sulfur dioxide in precipitation.  No similar benefit is likely for
the washout of sulfate which may be appreciable in the vicinity of the
sulfur dioxide source.  Higher total sulfur dioxide washout is likely to
occur at greater distances where diffusion of the plume results in greater
concentration in the near-surface air.  It has not been shown where and
to what degreee the washout of sulfur dioxide from the Keystone Generating
Station is maximized.  However, considerable clarification of the complex
process of sulfur dioxide washout has been provided, and this new infor-
mation can be applied toward resolution of the problem.

-------
                               REFERENCES
 1.  Air Quality Criteria for Sulfur Oxides. National Air Pollution Con-
     trol Administration, Publication AP-50.  1969.

 2.  F. G. May.  "The Washout of Lycopodium Spores by Rain," Quart. J.
     Roy. Meteor. Soc.. 84, 451-458.  1958.

 3.  R. J. Engelmann.  "Precipitation Scavenging Studies,"  Hanford
     Radiological Sciences Research and Development Annual Report for
     1964. BNWL-36, pp. 1.54-1.59, Battelle-Northwest, Richland, Wash.
     January, 1965.

 4.  R. J. Engelmann and R. W. Perkins.  "Snow and Rain Washout Coeffi-
     cients for Process Plant Radioiodine Vapor," Nature. 211, 62.
     July 2, 1966.

 5.  M. A. Wolf and M. T. Dana.  "Experimental Studies in Precipitation
     Scavenging," Pacific Northwest Laboratory Annual Report for 1968
     to the USAEC Division of Biology and Medicine. BNWL-1051, Part 1,
     pp. 18-25, Battelle-Northwest, Richland, Washington.  November 1969.

 6.  G. D. Kinzer and W. E. Cobb.  "Laboratory Measurements of the Growth
     and Collection Efficiency of Raindrops," J. Meteor. 13, 295-301.
     1956.

 7.  M. T. Dana and M. A. Wolf.  "Experimental Studies in Precipitation
     Scavenging," Pacific Northwest Laboratory Annual Report for 1967
     to the USAEC Division of Biology and Medicine, BNWL-715 3, pp. 128-
     140, Battelle-Northwest, Richland, Washington.  October 1968.

 8.  M. T. Dana, M. A. Wolf and L. A. du Plessis.  "Field Experiments in
     Precipitation Scavenging,"  Pacific Northwest Laboratory Annual
     Report for 1969 to the USAEC Division of Biology and Medicine,
     BNWL-1307, Part 1, pp. 77-81, Battelle-Northwest, Richland, Wash.
     June 1970.

 9.  B. L. Niemann, M. C. Day and P. B. MacCready.  Particulate Emissions,
     Plume Rise and Diffusion from a Tall Stack. Final Report to National
     Air Pollution Control Administration, Contract CPA-22-69-20, Meteo-
     rology Research Inc., Altadena, California.  1970.

10.  F. P. Scaringelli, B. E. Saltzmann and S. A. Frey.  "Spectrophoto-
     metric Determination of Atmospheric Sulfur Dioxide," Anal. Chem.,
     .39, 1709.  1967.

11.  Selected Methods for the Measurement of Air Pollutants, USPHS
     Environmental Health Series 999-AP-ll.  1965.

-------
12.  R. J. Be-rtalacini and J.  E.  Barney.   "Colorimetric Determination of
     Sulfate with Barium Chloroanilate,"  Anal. Chem., 29^, 281.  1957.

13.  F. P. Scaringelli and K.  A.  Rehme.   "Determination of Atmospheric
     Concentrations of Sulfuric Acid Aerosol by Spectrophotometry,
     Coulometry and Flame Photometry," Anal. Chem.,  41, 707.  1969.

14.  P. W. West and T. P. Ramachandran.  "Spectrophotometric Determination
     of Nitrate Using  Chromotropic Acid," Anal.  Chim. Acta., 35, 317.
     1966.

15.  P. W. West.  "Chemical Analysis of  Inorganic Particulate Pollutants,"
     Air Pollution. A. C. Stern,  Editor,  Academic Press, New York and
     London.  1968.

16.  R. B. Bird, W. E. Stewart and E. N.  Lightfoot.   Transport Phenomena.
     Wiley, New York.  I960.

17.  A. Fick.  "Uber Diffusion,"  Pogg. Ann.. 94,  59.  1855.

18.  C. F. Prutton and S. H. Maron.  Fundamental Principles of Physical
     Chemistry, McMillian, New York.  1951.

19.  R. Gunn and G. D. Kinzer.  "The Terminal Velocity of Fall for Water
     Droplets in Stagnant Air,"  J. Meteor., 6, 246.  1949.

20.  W.G.N. Slinn and J. M. Hales.  "A Reevaluation of the Role of Thermo-
     phoresis as a Mechanism of In- and  Below-Cloud Scavenging," Submitted
     to J. Atmos. Sci.  1970.

21.  P. Goldsmith and F. G. May,   "Diffusiophoresis and Thermophoresis in
     Water Vapor Systems," Aerosol Science, C. N. Davies, Editor, Academic
     Press, New York.  1966.

22.  N. Frossling.  "The Evaporation of  Falling Drops," Gerlands Beitr.
     Geophys.. 52. 170.  1938.

23.  C. L. Pritchard and S. K. Biswas.  "Mass Transfer from Drops in
     Forced Convection," Brit. Chem. Engr., 12, 879.  1967.

24.  H. S. Carslaw and J. C. Jaeger.  Conduction of Heat in Solids,
     Clarendon Press, Oxford.   1959.

25.  J. Crank.  The Mathematics of Diffusion. Oxford Univ. Press, Oxford.
     1956.

26.  R. Higbie.  "Rate of Absorption of Gas into a  Still Liquid," Trans.
     Am. Ins t. Chem. Engrs., 31, 365.  1935.

-------
   27.   Hadamard.   Title unknown.  Comp. Rend.. 154. 1735.  1911.

   28.   Rybczynskl.  Title unknown.  Bull. Intern. Acad. Sci. Craeovie A,
        40.   1911.

   29.   V.  G. Levich.  Physicochemical Hydrodynamics, Prentice-Hall, N.J.
        1962.

   30.   R.  Kronig and J. C. Brink.  "On the Theory of Extraction from Falling
        Droplets," Appl. Sci. Res.. A2, 142.  1951.

     .   F.  H. Garner and J. J. Lane.  "Mass Transfer to Drops of Liquid
VJ       Suspended in a Gas Stream, Part II:  Experimental Work and Results,"
        Trans. Inst. Chem. Engrs., 37, 162.  1959.

   32.   P.  Harriot.  "A Review of Mass Transfer to Interfaces,"  Can. J.
        Chem. Engr.. 40, 60.  1962.

   33.   R.  J. Brunson and R. M. Wellek.  "Mass Transfer Within Oscillating
        Liquid Droplets," Can. J. Chem. Engr.. 48, 267.  1970

   34.   R.  E. Plevan and J. A. Quinn.  "The Effect of Monomolecular Films on
        the Rate of Gas Absorption into a Quiescent Liquid," A.I.Ch.E.J^, 12,
        741.  1966.

   35.   S.  Whitaker and R. L. Pigford.  "Response of a Gas-Liquid Interface
        to Concentration Pulses," A.I.Ch.E.J., 12, 741.  1966.

   36.   M.  Falk and P. A. Giguere.  "On the Nature of Sulphurous Acid,"
        Can. J. Chem.. 36, 1121.  1958.

   37.   W.  J. Thomas.  "The Absorption of Sulphur Dioxide in Water:  Part I -
        Theory," Trans. Inst. Chem. Engrs., 42, T3.  1964.

   38.   J.  C. Wang and D. M. Himmelblau.  "A Kinetic Study of Sulfur Dioxide
        in Aqueous Solution with Radioactive Tracers," A.I.Ch.E.J., 10, 574.
        1964.

   39.   S.  Lynn, J. R. Straatemeier and H. Kramers.  "Absorption Studies in
        the Light of Penetration Theory," Chem. Engr. Sci.. 4_, 49.  1955.

   40.   H.  L. Toor and S. H. Chiang.  "Diffusion Controlled Chemical Reaction,"
        A.I.Ch.E.J.. 5., 339-  1959'                           •**

   41.   R.N.J. Saal.  "The Velocity of  Ionic Reaction," Chem. Abs., 2:2, 1263.
        1968.

   42.   R.  P- Whitney and J. E. Vivian.   "Absorption of S02 in Water,"  Chem.
        Engr. Prog.. 45, 323.  1949.

-------
^43.   J. M. Miller  and R.  G. De Pena.   "The Rate  of  Sulfate Ion Formation
       in Water Droplets  in Atmospheres  with Different  Partial Pressures
       of S02," Paper  presented at  Second  International Clean Air Confer-
       ence, Washington,  D. C.  1970.

  44.   R.A.O.  Nijsing, R. H. Hendrksz and  H. Kramers.   "Absorption of Gases
       in Liquids,"  Chem. Engr. Sci.. 10,  88.   1959.

  45.   C. E. Junge and T. G. Ryan.   "Study of  the  S02 Oxidation in Solution
       and  its Role  in Atmospheric  Chemistry," Quart. J.  Roy. Meteor. Soc..
       84,  46. 1958.

  46.   W. D. Scott and P. V. Hobbs.  "Formation of Sulfate  in Water Droplets,"
       J. Atmos.  Sci.. 24,  54.  1967.

  47.   A. K. Postma.   "The  Effect of Solubilities  of Gases  on Their Scaven-
       ging by Raindrops,"  Precipitation Scavenging. 1970,  U.S.  Atomic
       Commission, Division of Technical Information.   Available from Clear-
       inghouse.   1970.

  48.   P. V. Danckwerts.  "Absorption by  Simultaneous Diffusion and Chemical
       Reaction into Particles of Various  Shapes and into Falling Drops,"
       Trans.  Faraday  Soc., 47. 1014.  1951.

  49.   G. Astarita.  Mass Transfer  with  Chemical Reaction,  Elsevier,  New
       York.   1967.

  50.   P. W. Summers.  "Scavenging  of S02  by Convective Storms in Alberta,"
       Precipitation Scavenging, 1970, U.S. Atomic Energy Commission, Divi-
       sion of Technical  Information.  Available from Clearinghouse.   1970.

  51.   F. P. Terraglio and  R. M. Manganelli.   "The Absorption of Atmospheric
       Sulfur  Dioxide  by  Water Solutions," J.  Air  Poll. Control Assoc.,  17.
       403.  1967.

  52.   A. C. Chamberlain.   Aspects  of Travel and Deposition of Aerosol and
       Vapor Clouds. AERE-HP/R-1261, Atomic Energy Research Establishment,
       Harwell, Great  Britain.  1953.

  53.   R. J. Engelmann.   "The Calculation  of Precipitation  Scavenging,"
       Meteorology and Atomic Energy 1968. D.  N. Slade, Editor,  U. S. Atomic
       Energy  Commission, Oak Ridge, Tenn. 1968.

  54.   S. W. Churchill.   "Bounded and Patched  Solutions for Boundary-Value
       Problems," A.I.Ch.E.J.. 11,  431.  1965.

  55.   H. Georgii, and S. Beilke.   Atmospheric Aerosol  and  Trace-Gas Wash-
       out,  AD 634-907.  1966.

-------
56.  S. Beilke.  "Untersuchungen Uber das Auswaschen Atmospharischer
     Spurenstaffe Durch Niedenschlage," Berichte Inst. Met. Geophys., 19.
     1970.                              	tLJL~   —

>57.  H. F. Johnstons and G. C. Williams.  "Absorption of Gases by Liquid
     Droplets," Ind. Eng. Chem.. 31, 993.  1939.

58.  W. B. Campbell and 0. Mass.  "Equilibrium in S02 Solution,"  Can. J.
     Res., 2^ 42.  1930.

59.  J. M. Hales.   Gas-Aerosol Interactions and the Related Problem of
     Synergism. USPHS Report, Field Studies Branch, Division of Air Pollu-
     tion.  1965.

60.  B. M. Smith, J. Wagman and B. R. Fish.  "Interaction of Airborne
     Particles with Gases,"  Environmental Science and Technology. 3, 358.
     1969.

61.  M. E. Smith and I. A. Singer.   "An Improved Method of Estimating
     Concentrations and Related Phenomena from a Point Source Emission,"
     J. Appl. Meteor.. .5. 631-649.  1966.

62.  J. E. Carson and H. Moses.  "The Validity of Currently Popular Plume
     Rise Formulas,"  Proceedings of the USAEC Meteorological Information
     Meet ing, C. A. Mawson, Editor, AECL-2787, pp. 1-15, Atomic Energy of
     Canada, Limited, Chalk River, Canada.  September 1967.

63.  M. T. Dana.  "Rain Scavenging of Soluble Dye Particles,"  Precipita-
     tion Scavenging, 1970, U. S. Atomic Energy Commission, Division of
     Technical Information.  Available from Clearinghouse.  1970.

64.  F. E. Gartrell, F. W. Thomas and S. B. Carpenter.  "Atmospheric
     Oxidation of Sulfur Dioxide in Coal-Burning Power Plant Plumes,"
     Am. Ind. Hyg. Assn. J., ,24, 113-120.  1963.

65.  R. D. Baldwin, L. Cohen, J. Forrest, B. Manowitz, L. Newman,
     M. E. Smith, M. Sternberg, and W. D. Tucker.  The Atmospheric
     Diagnostics Program at Brookhaven National Laboratory.  Second
     Status Report, BNL-50206.  November 1969.

66.  M. L Arin, C. E. Billings, R. Dennis, J. Driscoll,  D. Lull,
     F. A. Record, P. Warneck, and J. E. Wilder.  Study of Reactions of
     Sulfur in Stack Plumes, GCA-TR-69-12-6, GCA Corporation, Bedford,
     Mass.  December 1969.

67.  C. E.  Junge.  Air Chemistry and Radioactivity,  Academic Press,
     New York and London.  1963.

68.  A. G. Gibbs, J. M. Hales and W.G.N. Slinn.  "Nonlinear Effects in
     Gas Scavenging," to be submitted to J. Atmos. Sci.   1971.


-------
                          APPENDIX A




                      FIELD DATA SUMMARY







LIST OF FIGURES                                            119




LIST OF TABLES                                             121

-------
                            LIST OF FIGURES
                                                                Page
A.I     Wind Profiles, Run 1                                    124
A.2     Sulfur Dioxide Distribution, Run 1                      125
A.3     Wind Profiles, Run 2                                    128
A. 4     Sulfur Dioxide Distribution, Run 2                      129
A.5     Wind Profiles, Run 3                                    132
A.6     Sulfur Dioxide Distribution, Run 3                      133
A.7     Wind Profiles, Run 4                                    136
A.8     Sulfur Dioxide Distribution, Run 4                      137
A.9     Wind Profiles, Run 5                                    140
A. 10    Sulfur Dioxide Distribution, Run 5                      141
A.11    Wind Profiles, Run 6                                    146
A. 12    Sulfur Dioxide Distribution, Run 6                      147
A. 13    Wind Profiles, Run 7                                    150
A.14    Sulfur Dioxide Distribution, Run 7                      151
A. 15    Wind Profiles, Run 8                                    154
A. 16    Sulfur Dioxide Distribution, Run 8                      155
A. 17    Wind Profiles, Run 9                                    158
A.18    Sulfur Dioxide Distribution, Run 9                      159
A. 19    Wind Profiles, Run 10                                   162
A.20    Sulfur Dioxide Distribution, Run 10                     163
A.21    Wind Profiles, Run 11                                   166
A.22    Sulfur Dioxide Distribution, Run 11                     167
A.23    Wind Profiles, Run 12                                   170
A.24    Sulfur Dioxide Distribution, Run 12                     171
A.25    Wind Profiles, Run 14                                   176
A.26    Sulfur Dioxide Distribution, Run 14                     177
A.27    Wind Profiles, Run 15                                   180
A.28    Sulfur Dioxide Distribution, Run 15                     181
A.29    Wind Profiles, Run 16                                   184
A. 30    Sulfur Dioxide Distribution, Run 16                     185
A.31    Wind Profiles, Run 18                                   188
A.32    Sulfur Dioxide Distribution, Run 18                     189

-------
A.33    Wind Profiles, Run 19                                   192
A.34    Sulfur Dioxide Distribution, Run 19                     193
A.35    Wind Profiles, Run 20                                   196
A. 36    Sulfur Dioxide Distribution, Run 20                     197
A.37    Wind Profiles, Run 21                                   200
A.38    Sulfur Dioxide Distribution, Run 21                     201

-------
                             LIST OF TABLES
                                                                Page
A.I     Summary of Field Period I—October-November, 1969       123
A.2     Sampling Data, Run 1, Arc A                             126
A. 3     Sampling Data, Run 1, Arc B                             127
A.4     Sampling Data, Run 2, Arc A                             130
A.5     Sampling Data, Run 2, Arc B                             131
A.6     .Sampling Data, Run 3, Arc A                             134
A.7     Sampling Data, Run 3, Arc B                             135
A.8     Sampling Data, Run 4, Arc A                             138
A.9     Sampling Data, Run 4, Arc B                             139
A. 10    Sampling Data, Run 5, Arc A                             142
A. 11    Sampling Data, Run 5, Arc B                             143
A.12    Summary of Field Period II—February, 1970              145
A.13    Sampling Data, Run 6, Arc A                             148
A. 14    Sampling Data, Run 6, Arc B                             149
A.15    Sampling Data, Run 7, Arc A                             152
A. 16    Sampling Data, Run 7, Arc B                             153
A.17    Sampling Data, Run 8, Arc A                             156
A.18    Sampling Data, Run 8, Arc B                             157
A.19    Sampling Data, Run 9, Arc A                             160
A.20    Sampling Data, Run 9, Arc B                             161
A.21    Sampling Data, Run 10, Arc A                            164
A.22    Sampling Data, Run 10, Arc B                            165
A.23    Sampling Data, Run 11, Arc A                            168
A.24    Sampling Data, Run 11, Arc B                            169
A.25    Sampling Data, Run 12, Arc B                            172
A. 26    Summary of Field Period Ill—April-May, 1970            173
A.27    Sampling Data, Run 13, Arc B                            175
A.28    Sampling Data, Run 14, Arc B                            178
A.29    Sampling Data, Run 15, Arc B                            182
A. 30    Sampling Data, Run 16, Arc C                            186
A. 31    Sampling Data, Run 17, Arc C                            187
A.32    Sampling Data, Run 18, Arc A and C                      190

-------
A.33    Sampling Data, Run 19, Arc A and C
A.34    Sampling Data, Run 20, Arc A and C
A.35    Sampling Data, Run 21, Arc A                             20?
A.36    Sampling Data, Run 21, Arc B                             203
A.37    Sampling Data, Run 22, Arc B                             205

-------
                                             TABLE A.I









1
I-"
NO
U>



SUMMARY OF
Run Wind1 S02
Date No. Velocity Emission
(deg/mph) (tons-hr-1)
10-20 1 230/125 13.6

11-1 2 150/8 12.2

11-1 3 145/6 12.1


11-2 4 162/6 11.7

11-2 5 145/6 10.8
-
FIELD

Arcs

A
B
A
B
A
B

A
B
A
B
PERIOD I —

Run Time2

2201-2236
2224-2253
0745-0848
0747-0846
0916-1050
0918-1042

0838-0854
0844-0906
0957-1055
1031-1106
OCTOBER-NOVEMBER, 1969
Sampling3
Duration
(min)
80
97
118
118
99
114

61
72
104
92
Precip.4
Rate
(mm-hr-1)
0.52
0.29
0.38
0.37
0.14
0.11

1.28
0.87
0.60
0.36
S02
Recovery
(ymoles)
4.21
3.46
1.93
5.07
0.42
0.48

16.93
16.79
11.77
4.71
S02 Concentration
Max
Mean
Min
(ymole-liter- ^T
7.1
9.8
8.8
16.7
9.0
5.8

12.3
10.4
10.8
8.8
5.8
7.7
1.5
3.1
2.1
2.6

7.2
7.2
6.7
4.1
0.1
1.8
0.0
0.2
0.2
0.1

2.0
2.2
1.4
0.4
1Wind velocity at stack exit, unless otherwise noted.
2Time of concurrent sampling by all collectors.
3Average duration of collector exposure.
4Average rate determined from average volume of precipitation collected  and average  duration of
 collector exposure.

-------
      RUN  NO. .1
DATE 10-20-69
TIME 2201^2253
-2160-
-1905-
-1650
-1390
-1130
- 865
- 600
- 335





n— TOP OF
I STACK

1
X
/



/
s














"— -^.











PITTSBURGH
D A u T M c n N n c
PITTSBURGH
RAWINSONDE


                         SPEED SCALE: 1/4" = 10 MPH
HEIGHT (FT)
2160
1905
1775
1650
1390
1130
865
775
600
335
300
TIME — 1900 2100
PIBAL
PIBAL
RAWIN
PIBAL
PIBAL
PIBAL
PIBAL
RAWIN
PIBAL
PIBAL
AERO


235/39




230/30













280/17






















NOTES:  1.  WIND ARROWS POINT DOWNWIND  TO  INDICATE  DIRECTION OF
          EFFLUENT TRAVEL.  COMPASS  ORIENTATION  IS CONVENTIONAL.
          CLOCKWISE F'ROM NORTH TO TOP.

       2.  PIBAL WIND VELOCITIES ARE BASED  ON  THEORETICAL  RISE
          RATE OF  A DRY, 30 gm PIBAL  WITH  71  gms  FREE LIFT.

       3.  THE AEROVANE WAS  MOUNTED  ABOUT 100  FT ABOVE GROUND
          ON A 200 FT HILL  ONE-HALF MILE SE OF  KEYSTONE.

       4.  PITTSBURGH RAWINSONDES  FOR THE  2000  AND 3000  MSL
          LEVELS ARE PLOTTED AT SAME  HEIGHTS  ABOVE SURFACE
          AS AT PITTSBURGH.
                  Figure A.I  Wind Profiles, Run 1

-------
                           4
                           4
                           1
                                    •
                         ."•      /
                                      ' «« 4
LEGENDS
  X.  RRC  R  STRTI0NS
  4  flRC  B  STRTI6NS
  *  flRC  C  STRTI0NS
  K DENOTES PLRNT L0CRTI0N
                                       N
SCRLES
        H= 1 MILE
        = 10 MICR0M0LES/L
         Figure A.2   Sulfur Dioxide Distribution,  Run 1

-------
               TABLE A. 2
SIM
  .U
10.U
11.U
12.U
13.p
14.U
lb.0
16.U
17.U
20.U
21.U

2b!u
26.0
              HUN
            'lOu
       HKOI*'     JO
                              —  KtYSTONE STUDY
                OCT. 2U»  19b9    ARC A
PHtClP
VOLUME

31.0
do.u
/9.b
7«+.U
7H.U
7i,U
67. U
bJ.U
lb'j.5
21. U
19. b
I/. 5
2<4.b


Sdii
2.b
5.6
5.9
6.1
6.b
5.1
6.3
6.3
7.1
.1
3.U
2.U
2.^
                        CONCENTRATIONS IN SAMPLES
                                            N02
                                           N03
:M 7
              u:ub
19.5   13.0
     VULUNt UNIT lb VL, SOii  CO|>,CLNmflT
                   bUL^ATEr  MTKITE
           AKL
                                   uMTS AR£
                                 NlTKATt CONCtNTKAT I Or-

-------
                           TABLE A.3


                       bCAVtlMGlNb  UATA — KtYSTO(Mt  STUUY

               HUK   i     OCT. 2u»  1954   AKC ti

                                                   IN
                To
lu.u  £*:ia  2,j:b,s    db.o    d.y
11,0  *<::iti  i:^:37    7b.5    9,U
ic.o  <^<::<£u   u:ui    ou.u    g.ts
IH.U  <:^:^b   u:ub    bo,o    7.H
ib.o  £i^:«!ti   u:o9    bu.u    t.d
ib.u  d^:^o   o;i^    *4o*o    5.1
17.U  c.b
is.o  <:*i:«4b   u:ki    lo.u    1,6
^U.b  tLdl^l   u:<^H     ^.U    b.ti
iix.u  ^^:<4<^   u:^7    i£*u    w.d
^H.U  ^ATLr MIHITL  ANu M fKATt  COKCtlN I Kfl r JOt
      UlMlTb A*t i'ILLlv;K«(V'S/u.

-------
      RUN NO.  2
DATE  11-1-69
TIME   0745-0846

~2160
1 905

~1 650
~1 390
~1 1 30
865
— £ nn
ouu
— t ") K
J Jo

V
\
\


v
\
•« TOP OF N
1 STACK

1
1
' \\ \\
\ V
u \
\1 V
^ ^
\1 0
\ \



u ^^.

\\
\\
v\
\\
\\
\
A
V

V














                                                        PITTSBURGH
                                                        RAWINSONDE
                                                        PITTSBURGH
                                                        RAWINSONDE
                                                        •AEROVANE
                         SPEED SCALE: 1/4" = 10 MPH
HEIGHT (FT)
2160
1905
1775
1650
1390
1130
865
775
600
335
300
TIME -— 0700 0725 0823 0930
PIBAL
PIBAL
RAWIN
PIBAL
PIBAL
PIBAL
PIBAL
RAWIN
PIBAL
PIBAL
AERO


160/34




145/32



166/33
165/32

170/30
165/27
165/27
165/22

166/14

155/12
162/30
162/28

160/28
152/21
150/22
147/21

144/16
121/11
110/11
158/35
151/27

153/28
147/30
147/26
143/15

135/21
123/15
120/13
NOTES:  1.  WIND ARROWS  POINT DOWNWIND  TO  INDICATE  DIRECTION  OF
          EFFLUENT TRAVEL.  COMPASS  ORIENTATION  IS CONVENTIONAL-
          CLOCKWISE F^OM NORTH  TO  TOP.

       2.  PIBAL WIND VELOCITIES ARE BASED  ON  THEORETICAL  RISE
          RATE OF  A DRY, 30 gm  PIBAL  WITH  71  gms  FREE  LIFT.

       3.  THE AEROVANE WAS  MOUNTED  ABOUT  100  FT ABOVE  GROUND
          ON  A 200 FT  HILL  ONE-HALF MILE  SE OF  KEYSTONE.

       4.  PITTSBURGH RAWINSONDES   FOR THE  2000  AND  3000 MSL
          LEVELS ARE PLOTTED AT SAME  HEIGHTS  ABOVE  SURFACE
          AS  AT PITTSBURGH.


                Figure  A.3  Wind Profiles, Run 2

-------
                           4
                           1
LEGENDS
  5K  RRC  fl  STHTI8NS
  <  flRC  B  STHTI0NS
  O  flRC  C  STRTI0NS
  K DENOTES PLflNT L0CflTI8N
                                        N
SCflLES
  I	1   1 MILE
        - 10  f1ICR0M0LES/L
         Figure  A,4  Sulfur Dioxide Distribution, Run 2

-------
                   TABLE A.4
HKtLlHU«lluN
  .0
44.U

4b . U
4/.0
4to.U
4V.U
bl.U

5J.U
b4,u
bb.u
bo.U
b/.U
bb.u

6U.U
bl.U

bJ.U
(\\l
        re
       •a : J
                     v/OLUi'E

                      bU.U
                      o/.U
                      /l.U

                      7b.O
                      7U.O
                      bu.U
                      o/.U

                      bl.U

                      54,U
                      4o.U
                      tU. U
                                      — KhYSTONL  STUbY

                               1» 19b<3   AKc A

                                       I HAT IONS  IN SAMPLES
SU2
 .U
                        .1

                       1.1

                        .b
                        .si
                        .o
                       4.9
                       d.ti
                             4.M
      J,8

      4,1
      4.4
      J.4
      J.4


      J.4
                                                            Pt-,
                                                    4.7U
OIL: VOLI>L
    AHL
                i  ib  I'L.  Soc.  GChceMuM ION
                  . SLLt-A(t»  Miwlit
                                      ITS

-------
                         TABLE A.5
STA.
 63.0
 6,4.0
 6b.O
 67.0
 ba.O
 69.0
 7U.O
 7i.O
 72.0
 73.0
 74.0
 7b.O
 7o.u
 77.0
 7b.O
 79.0
 bu.u
 bl.o
 8*.o
 83.U
 84.0
 80. u
 87.0
bb.O
89.0
90.0
       7:00
       7:0*1
       7:ob
       7:07
       7:11
       7Uo
       7:19
       7:20
       7:27
       7:29
                     11.2
       /:**?
.2
.2
.2
.2
2.6
.2
.2
.2
1.8
.3
• b
1.3
3.3
l.b
3.2
4.b
* -z
•3 • O
b.3
8,4
11.4
8.7
9.1
10. b
lb.7
10.8
4.3
4.2
4.4
3. ft
4.1
4.1
4.3
3.4
4.ri
4.8
b.O
4.4
4.«
4.8
b.b
b.fe
bj
. £
4.7
4.4
4.7
b.ti
b.d
b.o
fc.U
b.8
                      SCAVENGING  UA1A  — KtYSTONE STUuY

               HUN  2    NOV.   1,  1969   AKC B

                      KLCIP       CCKCENHATIONS IN SAMPLES
             1CJL1     VOLUME
 8:49   7«i.O
 8:bl   79.0
 8lb4   7u.O
 8:bo   70.0
 CJ59   oo.O
 9:01   7o.o
 S:U3   74,0
 9:ob   7o.o
 9:u7   b9.0
 9Ub   09.0
 9:18   ?U,U
 9:^1   7J.O
 9:23   bb.O
 9:25   bo.O
 9:27    bo.O
 9:30    bb.O
 9:34    bb.O
 9:3b    b4.0
 9:37    o^.O
 9:3fa    O/.O
 9:-4«i    43.0
 «s:44    4o.O
9:47    40.U
    VOLUf"t
      KL. SO*
     bUI>ATE»
                                     flNu NlfKATL
                                                      1 H. 1 I( c,

-------
      RUN  NO.
DATE  11-1-69
TIME   0916-1042
                               V
-2160 	 r-
1 905 \
1 650
1 390
V
•1 130 \
865 n— TOP OF >
rnn I STACK
11 r 1 \ , .. , ,
11
\\
v\
\\
\\
\
\
>

V
*^^























PITTSBURGH
RAWINSONDE
PITTSBURGH


                         SPEED SCALE: 1/4" = 10 MPH
HEIGHT (FT)
2160
1905
1775
1650
1390
1130
865
775
600
335
300
TIME — 0700 0930
PIBAL
PIBAL
RAWIN
PIBAL
PIBAL
PIBAL
PIBAL
RAN IN
PIBAL
PIBAL
AERO


160/34




145/32



158/35
151/27

153/28
147/30
147/26
143/15

135/21
123/15
120/13













,








NOTES:  1.  WIND ARROWS  POINT  DOWNWIND  TO  INDICATE  DIRECTION  OF
          EFFLUENT TRAVEL. COMPASS  ORIENTATION  IS CONVENTIONAL.
          CLOCKWISE FROM  NORTH  TO TOP.

       2.  PIBAL WIND VELOCITIES ARE BASED  ON  THEORETICAL  RISE
          RATE OF  A DRY,  30  gm  PIBAL  WITH  71  gms  FREE  LIFT.
       3.  THE  AEROVANE  WAS MOUNTED  ABOUT  100  FT ABOVE  GROUND
          ON A 200 FT  HILL ONE-HALF MILE  SE OF  KEYSTONE.
       4.  PITTSBURGH RAWINSONDES FOR THE  2000 AND  3000 MSL
          LEVELS ARE PLOTTED  AT SAME  HEIGHTS  ABOVE  SURFACE
          AS AT PITTSBURGH.
                Figure A.5  Wind Profiles, Run 3

-------
                                           V
                                       N
LEGENDS
  *  RRC R STRTI0NS
  <  RRC B STRTI0NS
  0  RRC C STRTI0NS
  K DEN0TES  PLRNT L0CRTI0N
SCRLES
  I	1= 1 MILE

  I	1-  10  MICR0M0LES/L
           Figure A.6  Sulfur  Dioxide Distribution, Run 3

-------
                           TABLE A.6
                  ION SCMVEkGlNb  UATA — KEYSTONt  STUUY
bl«.    bA(vt-LlKt>    PKLCIH       LOfxCEMKATUWS  IN  bA^PLhb
         HtHiOu     VULu^t
       hKUl*     TO             bo2   bC1*     hOJd     N03     PH
  .u   «:u7   iu:^u   2s*,u     i.^   4.fa                    u.bi

-------
                          TABLE A. 7
S1A.
b4.U
b<+.U
fob.U
b7.u
bb.U
69.U
7U.U
71.U
7*.U
73,0
7H.U
7b.U
       ti:5b
       a:b9
       y:ui
       9:u3
       9:ob
       9;u7
       9;ib
               rtUN
                re
              iu;
              iu
              lu;
              iu;
              lu;
              iu;
              lu
              iu;
              iu;
              11!
              11
49
bl
t>3
bb
b?
b9
U2
UH
                         NOV.   1»
    PKtClP
              ix;ub
<:c,u
2ti.U
iiO.U

-------
      RUN  NO.   4
DATE  11-2-69
TIME    0838-0906
-Z160-
~1 905 '
— 1 f r r\
1 650
~1 390
~1 130
• O £ C
865
— a r\r\
— o o c
job
1
1 X
\




i-. -*•• TOP OF
1 STACK

1
1 y
\
\
^
\
I y
f x

^


























PITTSBURGH
RAWINSONDE


PITTSBURGH

KAwl iNbUNUt



                         SPEED SCALE: 1/4" = 10 MPH
HEIGHT (FT)
2160
1905
1775
1650
1390
1130
865
775
600
335
300
TIME — 0700 0830
PIBAL
PIBAL
RAW IN
PIBAL
PIBAL
PIBAL
PIBAL
RAW IN
PIBAL
PIBAL
AERO


'145/21




185/28


020/10
172/22
167/21

162/15
162/13
162/11
153/9

118/11
089/12
085/11






















NOTES.  1.  WIND ARROWS POINT DOWNWIND  TO  INDICATE  DIRECTION  OF
          EFFLUENT TRAVEL.  COMPASS  ORIENTATION  IS CONVENTIONAL'
          CLOCKWISE F^OM NORTH  TO  TOP.

       2.  PIBAL WIND VELOCITIES ARE BASED  ON  THEORETICAL  RISE
          RATE OF A DRY, 30 gm  PIBAL  WITH  71  gms  FREE  LIFT.

       3.  THE AEROVANE WAS  tlOUNTED  ABOUT 100  FT ABOVE  GROUND
          ON A 200 FT HILL  ONE-HALF MILE SE  OF  KEYSTONE.

       4.  PITTSBURGH RAWINSONDES  FOR THE-2000  AND 3000  MSL
          LEVELS ARE PLOTTED AT SAME  HEIGHTS  ABOVE SURFACE
          AS AT PITTSBURGH.
                 Figure A.7  Wind Profiles, Run 4

-------
LEGENDS
  *  RRC  fl STRTI0NS
  <  flRC  B STRTI8NS
  *  RRC  C STRTIQNS
  K DENOTES  PLRNT L0CRTI8N
                                       N
SCRLES
        1   1 MILE
        =  10  MICROM6LES/L
           Figure A.8  Sulfur Dioxide Distribution, Run 4

-------
                           TABLE A.8




               IAI ION  bCAVfcKGlNt. LA I « — KtYbTCKt  STULY

               HOK   <*     NCv/.  :u2    db.O    9.3   l.b
*»d.u   -d:o3    y:u3   iuu.u    9.H   i.d
4S«.U   H:UH    s:u<4    Sb.U    9.b   l.b
bu.u   «:ub    s:uc    92.0   lu.2   1.6
bi.u   u:ub    s;ii   uu.'o    9.b   1.3
bci.U   b:u/    9:23    9b.U    7.9   1.8
b3.U   «:^0    9:28   1<;1.0    3.0   1.8
bl.U   d221    KMMb/u.

-------
                             TABLE A.9
                       FHLCiP
  by.o   «:u2    q:ub  lub.o
  60.0   d:u<4    3;o9   d/.U
  6i.o   e:ub    y:n   yi.u
  6^.0   d:ud    y:u   by.o
  63.u   a:iu    ^:it   oy.d
  6*.o   d:i2    y:i«   tio.o
  63.0   d:m    y;«;u   7£.U
  bo.O   d:ib    y;^   7b.u
  67.o   d:i«    y:<:^   70,o
  6b.u   b:tu    y:«i^   du.u
  6s».u   o:«ii    ^iji    dA.u
  70.U   31*4    S»;jH    /t+.U
  71.u   t:2b    s;j/    Ssu.u
  7*:.o   0:^0    s»:jy    «4o,u
  7J.u   ^:<;0    s,;^^    ti4t(j
  7^.o   d:^s   S*:HH    db.u
  7b.u   c>:ji   s»:H7    ^.u
  7c.u   b:^<;   s:*y    c^+.u
  77.0   tt:j^   ^^.x    c,0,j
  7a.U   «:3b   s:h4    «3i.O
 eo.u    d:3b    s:bd  luo.o
 bi.o    e:*+o   iu:uu  Hb.u
 b^.u    dim   iu:j2  131.u
 b3.o    d:4j   iu:u**  i3o.u
 tiu.u    d;n«*   iu:u7
Jil\l3 I
. 2»
t
SU^!
7.2
7.7
b.t
61,
• o
10.4
y,2
9.b
y.o
9.1
d.b
/.^
7.b
7.b
b.b
7.b
7.ti
7.J
OJ
• c
b.y
3.H
2.H
*.2
3.1
b.2
ti.d
>AIA — KEYSTONE STUDY
1^69 ARC ti
-CKCENlHArioNS IN SAI^HI
bCH ( 02 N03
«i.3
2.6
2.6
J ._
e »b
2.0
2.1
«i.3
2.J
2.b
2.<4
3.U
2.U
^*U
l.u
1.3
• b
.4
fi
**+
i.y
3.0
H.O
3.3
3.b
3.0
d.6
,'Olt: VULUfh OMl  ib KL. SOd COwCEMHATiCN  i|.NITS
      ^'.iCKo^OLt
      UNITS /\Kt

-------
RUN NO.   5
                          DATE  11-2-69
TIME   0957-11Q6
-2160 	 —
1905 i \
1 650
A
\
\
\
1390 v
-,,-,,-, \f >
1130 / v
,rr 1 \
35j n— TOP OF
,nn \ STACK
GUI) / 1
/I

X


^




\
\



^
0















                                                         PITTSBURGH
                                                         RAWINSONDE
                                                        PITTSBURGH
                                                        RAWINSONDE
                                                        AEROVANE
                         SPEED SCALE  1/4"   10 MPH
HEIGHT (FT)
2160
1905
1775
1650
1390
1130
865
775
600
335
300
TH1E -*• 0700 1010 1110
PIBAL
PIBAL
RAW IN
PIBAL
PIBAL
PIBAL
PIBAL
RAWIN
PIBAL
PIBAL
AERO


145/21




185/28



157/26
154/21

153/20
147/17
144/15
141/15

126/18
102/8
080/10

•



150/13
150/11

115/8
081/4
055/4











NOTES.  1.  WIND ARROWS POINT DOWNWIND  TO INDICATE  DIRECTION OF
          EFFLUENT TRAVEL.  COMPASS  ORIENTATION  IS CONVENTIONAL-
          CLOCKWISE F^OM NORTH  TO  TOP.

       2.  PIBAL WIND VELOCITIES ARE BASED  ON  THEORETICAL RISE
          RATE OF A DRY, 30 gm  PIBAL  WITH  71  gms  FREE LIFT.

       3.  THE AEROVANE WAS  ilOUNTED  ABOUT 100  FT ABOVE GROUND
          ON A 200 FT HILL  ONE-HALF MILE SE OF  KEYSTONE.

       4.  PITTSBURGH RAWINSONDEG  FOR THE  2000  AND 3000  MSL
          LEVELS  ARE PLOTTED AT onME  HEIGHTS  ABOVE SURFACE
          AS AT PITTSBURGH.
              Figure A.9  Wind Profiles, Run 5

-------
                                              '    •
)tf
                                         ]/
                                        ..••••
                                                ,"."'   ;
                                                      t
LEGENDS
  *  flRC  R  STRTI0NS
  <  RRC  B  STRTIQNS
. *  RRC  C  STRTIQNS
  K DENOTES PLRNT LOCRTI0N
                                        N
        SCRLES
                 - 1 MILE
                = 10 MICR0I10LES/L
        Figure A.10  Sulfur  Dioxide Distribution,  Run 5

-------
                           TABLE A.10
                     bCAVENGlNt>  UA1A  — KEYSTONE STUDY

              HUN  b    NOV.   
-------
                           TABLE A. 11
                                  LAIA  —  KhYSTGlNfc

                          NOV.  <;, i^ec    AKC  b
 S7A.    -«,.,-,.*,,,o     n
-------
                                                   TABLE A.12
Ui
SUMMARY OF FIELD PERIOD

Date

2-3

2-9

2-9

2-10

2-10

2-10

2-14
Run
No.

6

7

8

9

10

11

12
Wind1
Velocity
(deg/mph)
345/S5

145/9

149/13

290/106

290/106

290/106

146/8
so2
Emission
(tons-hr-1)
12.3

12.2

11.1

11.6

11.6

11.6

11.8

Arcs

A
B
A
B
A
B
A
B
A
B
A
B
B

Run Time2

1229-1430
1307-1455
1029-1135
1054-1207
1201-1315
1238-1340
1012-1115
1048-1138
1133-1247
1218-1312
1309-1444
1357-1506
2015-2124
II — FEBRUARY, 1970,
Sampling3
Duration
(min)
139
131
87
96
95
92
79
86
95
97
115
119
149
Precipi1*
'Rate i
(mm-hr-1)
0.31
0.24
1.63
1.30
0.56
0.15
0.61
0.76
0.84
0.79
0.42
0.41
0.06
so2
Recovery
(limbles)
0.98
1.34
32.22
10.44
8.47
2.85
18.66
19.12
8.93
14.28
7.75
10.19
1.21
S02 Concentration
Max.
Mean
Min
(pmole- liter-1)
1.9
4.9
12.8
7.2
9.5
12.1
13.4
11.8
8.8
9.8
7.8
11.1
3.5
0.8
1.4
7.6
2.5
5.3
6.3
8.3
8.2
3.6
5.2
5.1
5.9
1.1
0.1
0.1
3.8
0.2
1.8
1.3
2.7
4.0
1.1
2.0
2.7
1.9
0.1
            velocity  at  stack exit,  unless  otherwise noted
      2Time of  concurrent  sampling by  all collectors
      3Average  duration  of collector exposure
      ^Average  rate determined from  average volume  of .precipitation collected and average duration of
        collector  exposure.
      50700 EST Pittsburgh rawinsonde.

-------
      RUN .NO.  6
DATE  2-3-70
TIME  1229-1455
-21 60 	 — j 	 	
1 905 '
1650
1390 ' '
1 130
865 n— T°p OF
rnn I STACK






V
- *
1 1












^











PITTSBURGH
RAWINSONDE
PITTSBURGH
RAWINSONDE
^^ HtKUVnNt
                          SPEED  SCALE:  1/4"  =  10  MPH
HEIGHT (FT)
2160
1905
1775
1650
1390
1130
865
775
600
335
300
TIME-* 0700 1230
PIBAL
PIBAL
RAWIN
PIBAL
PIBAL
PIBAL
PIBAL
RAM IN
PIBAL
PIBAL
AERO

'
350/17




345/20













290/u






















NOTES:  1.  WIND ARROWS  POINT DOWNWIND  TO  INDICATE  DIRECTION  OF
          EFFLUENT TRAVEL.  COMPASS  ORIENTATION  IS CONVENTIONAL-
          CLOCKWISE F^OM NORTH  TO  TOP.

       2.  PIBAL WIND VELOCITIES ARE BASED  ON  THEORETICAL  RISE
          RATE OF  A DRY, 30 gm  PIBAL  WITH  71  gms  FREE-LIFT.
       3.  THE AEROVANE WAS  MOUNTED  ABOUT 100  FT ABOVE  GROUND
          ON A 200 FT  HILL  ONE-HALF MILE SE OF  KEYSTONE.
       4.  PITTSBURGH RAWINSONDES   FOR THE  2000  AND  3000 MSL
          LEVELS ARE PLOTTED AT SAME  HEIGHTS  ABOVE  SURFACE
          AS AT. PITTSBURGH.


               Figure  A.11   Wind Profiles, Run 6

-------
                                              44*
                                                   4

                                                    4\
                                    /     \
                                                 •    4
                                 II  •
                                   *
                                4    •
                                     'M  \
LEGENDS
  *  RRC  R  STPTI0NS
  <  RRC  B  STRTIQNS
  *  RRC  C  STRTI6NS
  K qENOTES PLHNT L8CHTI0N
                                       N
SCflLES
  I	1= 1 MILE
        = 10  MICR0M8LES/L
       Figure A.12  Sulfur  Dioxide Distribution, Run 6

-------
                            TABLE A.13





                       bCAVfclvGlNu  L.ATA  — KfcYSTONt

                          Ftu.  >ir 1*70   AKC A
 SI*.    SdiVt-HisO     H
 11.U  12:21   14:42    eo.U    -.U                  l.nb     b.bb
 U.b  JL^:^^   I4:nb    tsl.O    1.2    3,U             .^1
 Ib.U  1^:10   J.4;JJ    bl.U     .'j     .7
 17.U  J.2:17   A4IJO    ^y.O     ,1     ,7            l.Ub     4.7b
 l«a.U  12:«:0   i^IJti    01.0     *3    .l6
 20.u  i«::2H   m:^*;    a^.o     .«    
-------
                           TABLE A.14
                      SCAVENGING LiATA — *trSTOlNt <

                         Ftu.   o,  197()   AKC

SIA.     SAM
                                                 IN SAMPLES

                               S>02    SC4     f>02    i\uJ     J-H
11.0   i*:bi  m:bb   4*.u      .1    2.2            .91    4.7q
1<:.0   !2:bJ  l4:b/   4b.O      .1    2.9
I4.o   i2:bb  i-,;uu   bo.u     2.^    2.^            .b9    4.fo,
Ib.O   1^:57  I3;.jj   7t.O      .J    z.u                      X
Ib.o   1^:39  I3:ub   ob.o      .1    i.j            ,eu    u b7
20.0   u:oi  lb:utj   4C.o      .1    2.b
2*;.o   A^:O^  ib:iu   m.o      .2    .3.2             90    u ss
24.u   12:44  i4:3e   /o.o     j.i*
2b.O   i                  i.ub    4.5^
3ti.O   ij:u/  Ib:l9   44.0     4.1    2.0           i,U    4.56
40.0   i2:4b  ib:uo   bo.u     2.b    l.b
4^.0   i.o   i*::b4  ib:i2   o^.u      ,j    2.0
bo.o   I2:b7  iij:ib   ^o.o      ,j    2.5           1.20    4.6u
5<:.0   I«i:b9  A5:i9   Ob.O      .6    2.b
54.u   u:ui  ls:2J   «rb.u      ,3    2.b           l.in    4.be.
     VOLOh't LMT  ib  f^i-«  S0
-------
      RUN NO.
DATE  2-9-70
TIME   1029-1207
,,6U ,
•1 905 • • \
'1 650
1 390
1 1 30
nrr >>
5Gj n— TOP OF >
rnn I STACK
11 r 1 \ 	

' 11
\\
\\
M
, \| ' V
\
\
\
\


\
V

N













PITTSBURGH
RAVJINSONDE
' PITTSBURGH

AEROVANE
                         SPEED SCALE: 1/4" = 10 MPH
HEIGHT (FT)
2160
1905
1775
1650
1390
1130
865
775
600
335
300
TIME-* 0700 . 0915 1119
PIBAL
PIBAL
RAWIN
PIBAL
PIBAL
PIBAL
PIBAL
RAWI.'J
PIBAL
PIBAL
AERO


160/21




13-5/20



-


178/30
171/28
166/23
154/17

143/16
138/16







145/21

138/19
132/15












NOTES:  1.  WIND ARROWS  POINT  DOWNWIND  TO  INDICATE  DIRECTION  OF
          EFFLUENT  TRAVEL.  COMPASS  ORIENTATION  IS CONVENTIONAL-
          CLOCKWISE F^OM  NORTH  TO TOP.

       2.  PIBAL WIND VELOCITIES ARE BASED  ON  THEORETICAL  RISE
          RATE OF A DRY,  30  gm  PIBAL  WITH  71  gms  FREE  LIFT.
       3.  THE  AEROVANE WAS  MOUNTED  ABOUT  100  FT ABOVE  GROUND
          ON A 200  FT  HILL  ONE-HALF MILE  SE OF  KEYSTONE.
       4.  PITTSBURGH RAWINSONDES FOR THE  2000  AND 3000 MSL
          LEVELS ARE PLOTTED AT SAME  HEIGHTS  ABOVE SURFACE
          AS AT PITTSBURGH.
             Figure A.13  Wind Profiles, Run 7

-------
                                 ^>
V
                             « 4
LEGENDS
  *  RRC  R STflTIQNS
  <  RRC  B STRTIONS
  «  RRC  C STflTIQNS
  K DENOTES PLRNT L0CRTI0N
                                       N
SCRLES
        -\= 1 MILE
        = 10  MICR0M0LES/L
        Figure A.14  Sulfur Dioxide Distribution, Run 7

-------
                            TABLE A.15
       HKLCIHIIATION SCAVENGING UATA  —  KfcySTONt
SlA.    bAmLlKb     HHLCIP      CONCEM»AUuNS IN
                TO             So2   bC4      l\02    (\03     HH
 *.U  1U:19   li:4b  13H.5   12.8   9.3
 4.U  1U:21   li:4b  lbb.5   1U.7   3.9             .66    4,61
 D.U  iu:2«d   li:48  174.b    b.3   b.3
 e.U  iu:«i3   li:bl  220.U    6.4   b.3             .88    b.lQ
Ib.U  iu:2b   li:5/  19b.b    9.1   3,3
1*.U  lu:*>o   Ii:b4  ..J7U.U    9.9   .3.2             .60    4.bo
14.U  iu:«;/   Ii:b9  J/4.U    9.8   4.b
lo.U  JLu:^   l^:ul  49b.U   1U.U   3.8             «a2    4.b6
30.U  101^2   li:Jts  241.U    4.0   3.J             .bl    4.2(J
4«i.u  iu:il   Ii:j8  1/u.b    b.8   3.b
44.u  Iu:i3   li:41  lou.b    3.rt   3.b             .42    4.91
4c.O  iu:i4   li:-43  19b.b    5.«i   3.b
4t<.U  iu:io   Ji:4S  104.U    9.4   fa.9             .76    4.50
bii.U  iu:i9   li:49  ^Ub.U    3.9  16,b             .43    4.88
b4.0  10:^4   lllbl   3u.U    4.U   9.7
bb.U  Ld'.ti   li:sb  
-------
                           TABLE A.16
                                  UA1A — K-EYSTONt.  STObY

                                 •  1970   AKC B
SU.     SmLiivb    KHLCIP       V.CKO-MKATICNS  IN  SAMPLES
          PtKlUu     VOLUME.
        FKo"    |(;             bU2   SC«*     rOk     hQ3      hh
 c.u   xu:b3  i<::3i  
-------
      RUN'NO.  8
DATE 2-9-70
TIME  1201-1340
-21 60 	 T-
1 905 )
1 650
1 390
1 1 30
i-i-*~TOP OF >
,nn 1 STACK

11




V
\
^k
\
X


\\
\\
\\
\\
\\
\\
\
v\
\













PITTSBURGH
RAWINSONDE
PITTSBURGH


                          SPEED  SCALE:  1/4"  =  10  MPH
HEIGHT (FT)
2160
1905
1775
1650
1390
1130
865
775
600
335
300
TIME-* 0700 1119 1310
PIBAL
PIBAL
RAWIN
PIBAL
PIBAL
PIBAL
PIBAL
RAWIN
PIBAL
PIBAL
AERO


160/21




135/20









145/21

138/19
132/15


156/25

154/28
151/30
150/27
149/29

152/27
153/20












NOTES:  1.  WIND ARROWS POINT DOWNWIND TO INDICATE  DIRECTION  OF
          EFFLUENT TRAVEL.  COMPASS  ORIENTATION IS CONVENTIONAL-
          CLOCKWISE F-*OM NORTH TO TOP.

       2.  PIBAL WIND VELOCITIES ARE BASED ON THEORETICAL RISE
          RATE OF A DRY, 30 gm PIBAL WITH 71  gms  FREE LIFT.

       3.  THE AEROVANE WAS  MOUNTED  ABOUT 100 FT ABOVE GROUND
          ON A 200 FT HILL  ONE-HALF MILE SE OF KEYSTONE.

       4.  PITTSBURGH RAWINSONDES  FOR THE 2000 AND 3000  MSL
          LEVELS ARE PLOTTED AT SAME HEIGHTS ABOVE SURFACE
          AS AT.PITTSBURGH.


               Figure A.15  Wind Profiles, Run 8

-------
                            \\l
                             r//      '
                             v*     y *     *
'/,

                                      4
                                     4

LEGENDS
 X  flRC P STRTI0NS
 <  flRC B STRTI8NS
 *  flRC C STHTI0NS
  K DEN0TES  PLflNT L0CRTI0N
                                N
 SCflLES
   I	1= 1 MILE
       = 10 MICR0M0LES/L
       Figure A. 16  Sulfur Dioxide Distribution,  Run 8

-------
                              TABLE A.17
                IrtTlCN bL«VE.is  lo:k!4    /S^.O    d.4                  l.iJtf     4,bG
  b.O  li:4o  l^l^b    ti^.U    6.3
  o.O  n:bl  13:    bb.o    6.7
 l«:.o  11:^4  I3:ou    /4,u    b.l                  1.72     4,3u
 14.0  ii:t>y  u:3<:    b^.o    4.0
 lo.u  1^:01  ij;34    b^.o    4,d
 3o.o  ii:<4u  14:^:0   ii/.o    4.4                    ,bt?     4.ib
 4<^.u  ii:je  io:ib    «b,u    9.4
 44.0  ii:m  u;i/    9«i.o    b.b
 4c.o  11:43  io;.!,1?    ^4,0    4,1
 4t!,0  IlUa  13:^2   11-s.O    b.3                  1.29     3,9u
 bO,0  li:47  13:<;4    04.0    4.7
 b2.0  11:49  I3:«ifc>    94.0    *i.4                    .99     3.9b
 b4.o  n:bi  io:«!7    /o.o    i.o
 bo.o  ii:bb  13!3u    Di.O    4.b                    ,ri^     4.1h
 bO.O  J.i:bb  13:31    b4.o    4.b
 60.0  li:3b  I3:ib    bo.O    b.4                  I,«i2     4,lu
 b*:.0  ii:od  I3:iu    Dtj.o    b.2
 64.0  11:40  13:19    b4.o    6.4                  l.lb     4.5(1
 bb.O  li:4
-------
                          TABLE A.18




                      SLAVtivjGlNb bATA — KtYbTONL STUbY

               HUN   6     Fto.   9» 197IJ
SlA.    bAKl-LlNb     PKtUP      COKCEfgTHAriOlMb IN
         HLRIUU      tfCLUIvt
       F-KUI*    TC             SO?   SC4     i 02    N03     PH
 ^.0  12:^1   U:3tj    10.0    1.9
 4.0  12:30   14:00    lb.0    4.3                  ,db
 fc.O  12:3b   14:03     b.O    J.b
 b.O  i«::3d   I4:ob     b.o    1.6                 \. 51
bb.O  12:07   lo:-+3    27.0   12.1                 1.^4    3.91
bO.O  12:io   I3:4b    2b.O    7.b
b^.O  I2:i*i   13:48    2b.O    7.ti                 i.ll    J.go
b4.0  12:ib   13:50    «J7.0    8.3
bb.O  12Ub   13:b4    2o.O    8.0                  ,90    3.9o
bb.O  12:19   13:5b    27.0    b.o
70.0  12:22   I3:b9    27.0    7.3                 I.Ob    3.90
7^.o  12:24   i4:m    20.0    3.1
74.0  12:27   I4:ub    20.0    2,b                 1.21    3.50
7b.o  12:30   m:ud    21.0    4.5
7b.O  12:32   14:10    lb.0    3.9                 1.4b
60.0  12:10   13:40    2^.0    4.1
62.0  12:14   13:41    
-------
      RUN.NO.  9
DATE 2-10-70
TIME  1012-1138
-2160 	
1 905
"1 650
1 390
1 1 30
865 n— TOP OF
rnn I STACK
000 / 1
.. i t r / 1 .,., 	 	
11









^

'








%_












PITTSBURGH
1 RAW1NSONDE
PITTSBURGH

AEROVANE
                         SPEED SCALE: 1/4" = 10 MPH
HEIGHT (FT)
2160
1905
1775
1650
1390
1130
865
775
600
335
300
TIME — •- 0930 1040
PIBAL
PIBAL
RAW IN
PIBAL
PIBAL
PIBAL
PIBAL
RAN IN
PIBAL
PIBAL
AERO









282/6.0










293/5.0
280/8






















NOTES.  1.  WIND ARROWS  POINT  DOWNWIND  TO  INDICATE  DIRECTION  OF
          EFFLUENT TRAVEL. COMPASS  ORIENTATION  IS  CONVENTIONAL-
          CLOCKWISE F^OM  NORTH  TO TOP-

       2.  PIBAL WIND VELOCITIES ARE BASED  ON  THEORETICAL  RISE
          RATE OF  A DRY,  30  gm  PIBAL  WITH  71  gms  FREE  LIFT.
       3.  THE  AEROVANE WAS MOUNTED  ABOUT  100  FT ABOVE  GROUND
          ON A 200 FT  HILL ONE-HALF MILE  SE OF  KEYSTONE.
       4.  PITTSBURGH RAWINSONDES  FOR  THE  2000  AND  3000 MSL
          LEVELS ARE PLOTTED AT SAME  HEIGHTS  ABOVE  SURFACE
          AS AT PITTSBURGH.

              Figure A.17 Wind Profiles,  Run 9

-------
LEGENDS
  *  flRC fl STflTIQNS
  <  flRC B STflTIONS
  *  flRC C STflTIONS
  K DEN0TES PLflNT LBCflTIQN
                                       N
SCflLES
  I	1= 1  MILE
        =  10  MICR0M0LES/L
       Figure A.18  Sulfur Dioxide Distribution,  Run 9

-------
                          TABLE A.19



                                      — KE.YSTONF  STUUY

                                    70   AKC  A

blA,    bAitfHLlNia     HKtuClP      CCKCEfMIK^ Tli;NS  1^

                                            i02     K04
 *,U  lo:o«i   li:20    b^.O    b.l
 4,0  10:04   11:21    4 b,o    7,b
 b,0  10:04   £i:23    «i9»0    7,2
 b,o  io:ob   n:2b    7o.o  .  4.4
10.0  I0:i.o   H:2b    bo.U    b.7

14.0  iu:tiy   IxI^S    ob.U    b.b
ib,u  10:10   li:4i    b^.u    ft.b
lb.0   t,:t:o   ii:ib  .  *b.O    3,b
20,0   S»:b/   11U7,   Sb.O    b.4

24.0  I0:u2   li:21    44.0    7.0
2b.O  10:L4   li:22    bO.O    6,2
2b,0  J.o:ob   li:24    «£,U    7*4
40.0  10:ub   li:2c    79.0    <:,7
4^.u  iu:uti   11:27    ti^.o    7.1
34,0  !0:uS>   li:2^    bb.U   10,4
3b.O  I0:i0   li:JO    bb,0    7,7
4o.O  Io:i2   li:44    4>b,0   13.4
4U.O   SJIbb   li:ib    Vb.O   11.b
44,0   y:b7   11:17    /o,o   10,5
4b.O   9:b9   li:20    bc.O   10.«
4ti.O  iuUiO   li:22    7o.O   12.4
bo.o  iu:u<^   11:24    /c?.o   14.2
b<:.0  I0:u4   li:24    74.0    9*4
b4.0  10:04   li:2b    db.O    b.O
bb.O  I0:ob   li:2ti    /b..O   11,b
bb.O  10:0b   li:29    7o,0    9,4
bO.O  10:07   li;4l    bV.O    b.b
b
-------
                          TABLE A.20
               HUN  9    FEM. iu» I97fi   ARC B

                                                IN
                !<*            S02   bC4     '.02    N03
 t.u   AUi40   l*;ud   03.0   10.0   3,1
 H.U   iu:4b   l£:ub   b9.0    s.2
 b.u   10:47   i«>:u7   su.o    8.2   3.1
 c.o   iu:wd   1^:10   ou.o    4.3
lu.o   iu:io   n:3d   9i.u    b.b   2.9
1^.0   10:1*9   li:4U   eb.O    fa.*:
1H.O   10:2u   li:43  llb.U    b.fa   3.2
ifa.o   10:23   li:4b   db.o    b.3
lb.0   10:25   11^7   db.O   10.2   2.2
2U.O   10:itJ   li:42  1U4.U   10.4
22.0   10:21   11:44   db.O    4.0   S,3
24.0   10:2*   11U/  1U3.0    4.b
26.0   10:
-------
      RUN  NO.  10
DATE  2-10-70
TIME  1133-1312
-2160 	 -
1 905
1 650
1 390
1 1 30
365 n-^TOP OF
rnn I STACK
n r 1 1 ,
11









|L










^•- ^--












PITTSBURGH
RAWINSONDE
PITTSBURGH
1 	 RAWINSONDE

                         SPEED SCALE: 1/4"   10 MPH
HEIGHT (FT)
2160
1905
1775
1650
1390
1130
865
775
600
335
300
TIME — 1040 1400
PIBAL
PIBAL
RAWIN
PIBAL
PIBAL
PIBAL
PIBAL
RAWIN
PIBAL
PIBAL
AERO









293/5
280/8









287/10
270/17






















NOTES:  1.  WIND ARROWS  POINT  DOWNWIND  TO  INDICATE  DIRECTION  OF
          EFFLUENT TRAVEL.  COMPASS  ORIENTATION  IS  CONVENTIONAL-
          CLOCKWISE F^OM  NORTH  TO TOP.

       2.  PIBAL WIND VELOCITIES ARE BASED  ON  THEORETICAL  RISE
          RATE OF  A DRY,  30  gm  PIBAL  WITH  71  gms  FREE  LIFT.
       3.  THE  AEROVANE  WAS  MOUNTED  ABOUT 100  FT ABOVE  GROUND
          ON A 200 FT  HILL  ONE-HALF MILE SE OF  KEYSTONE.
       4.  PITTSBURGH RAWINSONDES FOR THE  2000  AND  3000 MSL
          LEVELS ARE PLOTTED AT SAME  HEIGHTS  ABOVE  SURFACE
          AS AT PITTSBURGH.
               Figure A. 19  Wind  Profiles, Run  10

-------
4

4
4
                         4
                         4
                         i
                                    ^
                                /
                                     \\
                                                  /   .
                                                   //
LEGENDS
  *  RRC fl STHTI0NS
  <  flRC B STflTIflNS
  *  flRC C STflTIONS
  K DENOTES  PLflNT  LBCflTIGN
                                       N
                 SCflLES
                   I	1= 1 MILE
                         - 10  MICR0M0L.ES/L
        Figure A.20  Sulfur Dioxide Distribution,  Run 10

-------
                          TABLE A.21
        HKtClHl l«IiUM bCAVEhGllNO LATa — KtYSTONt STUbY


               KUK 1U    Fttt.  lu» 197D   A«C  A


SIM.     b/Aff'LlKt;    HHLClP       COlsCtM »/U.. SULf-ATLr l\IlKllt ANiu NlfHAlt  CQNCtKl h: 1 10-
     UMTb

-------
                            TABLE A.22
                                  UAU — KtYbTONt  STOUY

               HOK  1U     FEb. lu» 1970   ARC b
 Sl«.    SA^LlNfa     HKLCiP      COKCEM KATIES  IN
          HtHloL      VOLUIvt
                Tc             S02   SC4     r.oa
                                                             Hh

  4.u  i2:ob  13:3«    97.u     H."<*                          4'lu
  e.u  12:07  Uim   1^7.0     3.3                          ,  car,
  b.O  i2;iu  13:43    /b.O     ^.u
 10.U  li:3ti  13:12   13b.U     3.0                          4  7o
 14.u   li:<43  13:1^  Ibo.U
 Ib.U
 Iti.U
 2U.O

       11:44  I3:  • j . < .       .     -.                          O.OU
 fcT . U   11 . 4 /
 2b.u   iiiau  n^te.i  XHO.U     ^.^                          4.In
^>D.O
141.0
ibb.U
l^o.O
tl^.O
140.U
ll/.U
iut>.u
l«ib.U
o/.O
lof .U
1 1 0 . U
llD.O
luu.U
llc.J
^o.u
b«••   l^.-^o  xuu.u    3.^                          3.90


                                                           4.in
4/.U  	   	            " -
     VOLUKt OMT  is  I"L.  SO^ CCuCbMHaTiON UhlTb
                   SoLHAThf MlHlTt ANu MTHATt COIMLLN|K(.

-------
      RUN NO. 11
DATE  2-10-70
TIME  1309-1506
-2160 	
1 905
1 650
1 390
11 30 '
865 n— TOP OF
rnn 1 1 STACK
t)UU / 1
n 11 r I 1 	 	 ,,. -,
11









•^ «







^••^
^^*
^^
^^^-












PITTSBURGH
1 RAWlNSUNUt
PITTSBURGH


                         SPEED SCALE: 1/4"   10 MPH
HEIGHT (FT)
2160
1905
1775
1650
1390
1130
865
775
600
335
300
TIME-* 1400 1515
PIBAL
PIBAL
RAW IN
PIBAL
PIBAL
PIBAL
PIBAL
RAW IN
PIBAL
PIBAL
AERO









287/10
270/17






290/24

289/22
282/16























NOTES:  1.  WIND ARROWS  POINT  DOWNWIND  TO  INDICATE  DIRECTION  OF
          EFFLUENT' TRAVEL. COMPASS  ORIENTATION  IS CONVENTIONAL'
          CLOCKWISE F^OM  NORTH  TO TOP.

       2.  PIBAL WIND VELOCITIES  ARE BASED  ON  THEORETICAL  RISE
          RATE OF  A DRY,  30  gm  PIBAL  WITH  71  gms  FREE  LIFT.

       3.  THE  AEROVANE  WAS MOUNTED  ABOUT 100  FT ABOVE  GROUND
          ON A 200 FT  HILL ONE-HALF MILE SE OF  KEYSTONE.

       4.  PITTSBURGH RAWINSONDES FOR THE  2000  AND  3000 MSL
          LEVELS ARE PLOTTED AT  SAME  HEIGHTS  ABOVE  SURFACE
          AS AT PITTSBURGH.
              Figure A.21 Wind Profiles, Run 11

-------
                        -«
                        «
                        4
                        4
                        i
-•~l
                                       1
                                                    s
                                         V
LEGENDS
  * flRC fl STflTIONS
  < flRC B STflTIQNS
  » flRC C STflTIQNS
  K DEN0TES PLflNT L0CRTI0N
                                     N
                                     A
           SCflLES
             I	1   1 MILE
                  = 10 MICR0M0LES/L
       Figure A.22   Sulfur Dioxide Distribution, Run 11

-------
                            TABLE A.23
                                  UA1A — KEYSTONE  STUUY

                   il    FEo.  HJr  1970   AKC A
ST«.     SAKhLlNto    PKtClP       CCNCEMKATIONS  IN
          HtHiCb     VOLUME
        HKUK     1C             bu2   bC4     r,02
 £.0   i«>:b4   14.'49   /O.O     b.9   3.4
 4.U   i*:bb   I4;b0   73.U     3,9   4.0
 o.O   i^:b7   I4:b*!   7o.U     4,b   3.4
 e.O   I2:b9   14:b3   ou.O     4.4   3.b
10.u   13:00   ii*:bb   7b.o     b.o   **•&
l^.U   13:U<»   I4:bb   btt.O     b.4   4.U
14.U   13:U3   l4:bti   bO.U     4.4   b.4
Ic.U   A3:ub   I4:b9   c/.U     b.9   H.ft
lo.U   i«i:bU   14:44   o«+.U     «i.7   b.9
2U.U   iNu NITKAFt CGKCtMrtMIC'

-------
                             TABLE A. 24
 SlA.
 b.O
 ti.O
lU.O
Ic.u
Ic.o
2U.U
22.0
24. U
.iH.U
4U.U
               HUN
                TV;
   bLAVEUfilNb  DATA --

U    FEo.  lu»  19/0   AKC b

               CCUCtMKAlIUNS IN

                          I'02
                                                    STUb#Y
              IblOe
              ib:iu
                      I-KLCIP
                      VOLUME
   41,U
   bc.U
   J2.U
   bH.O
   51.U
   bo.U
   HO.U
   bc.U
   oV.U
   bl.u
   7b.U
   Jo.O
   CD* U
   SiH.U
  10 0 . 0
                     1 4 J , U
                            4.9
                            b.7
                            «+.<4
                            J.8
                            <4.<4
                            b.7
                            J.y
                            b.3
                            b.g
                            7.1
                            b.7
                            5.4
                            7.u
                            b.7
                            btd
                            b.9
                            b.U
                            b,4
                            b.ti
                            7.7
                          11.1
4.«
J.b
J,4
2.a
d.2
J.O
J.7
4.0
2.b
H.U
J.U
2.B
J.«
                                     0,7
                                     J.U
                       H.Uu

                       ^.90
4.UO
3.90
5.7u

J.90
                                                            <4.0U
                                        3.9U

                                        J.bO
              la: 27
:  vULUiv'h bf-jiT
 nL,
                             COiMCtMK/illCN  t,(Ilb

-------
      RUN Nfl.  12
DATE 2-14-70
TIME 2015-2124
-2160 	
'1905 ^
'1650
1 390
1130
003 !-•••- TOP Of**
rnn 11 STACK
•>•>[-/ 1 	

II
\
y
\
\
X
^

^*s^
^



1
I
Hj
J
\
^
X
>.








PITTSBURGH
' RAWINSONDt
PITTSBURGH

AEROVANE
                         SPEED SCALE: 1/4" = 10 MPH
HEIGHT (FT)
2160
1905
1775
1650
1390
1130
865
775
600
335
300
TIME-* 1900 1725 2100
PIBAL
PIBAL
RAWIN
PIBAL
PIBAL
PIBAL
PIBAL
RAW IN
PIBAL
PIBAL
AERO


160/16




110/11


090/u
152/12
152/14

151/16
138/18
123/20
118/20

115/19
111/15

213/12
183/11

177/10
173/12
161/19
146/19

130/16
121/12












NOTES:  1.  WIND ARROWS  POINT  DOWNWIND  TO  INDICATE  DIRECTION  OF
          EFFLUENT  TRAVEL. COMPASS  ORIENTATION  IS CONVENTIONAL-
          CLOCKWISE FROM  NORTH  TO TOP.

       2.  PIBAL WIND VELOCITIES  ARE BASED  ON  THEORETICAL  RISE
          RATE OF A DRY,  30  gm  PIBAL  WITH  71  gms  FREE  LIFT.
       3.  THE  AEROVANE  WAS MOUNTED  ABOUT  100  FT ABOVE  GROUND
          ON A 200  FT  HILL ONE-HALF MILE  SE OF  KEYSTONE.

       4.  PITTSBURGH RAWINSONDES FOR THE  2000  AND  3000 MSL
          LEVELS ARE PLOTTED AT  SAME  HEIGHTS  ABOVE  SURFACE
          AS AT PITTSBURGH.
              Figure A.23  Wind Profiles, Run 12

-------
                                     m
                                     m  •
                                   M

                               /
                                •I
                           ,  'ft
    4

   4
LEGENDS
  * flRC fl STRTI6NS
  < flRC 8 STflTIQNS
  » flRC C STRTI0NS
  K DENQTES  PLflNT LQCflTIQN
                                      N
SCflLES
  I	1=  1 MILE
       -  10 MICRQM0LES/L
         Figure A.24  Sulfur Dioxide Distribution,  Run 12

-------
                            TABLE A.25
         SCAVENGiNo  bATA

  KUi\ 12    FEU.  !•*»  1970
      i u : 4 b
SIM.
 4.U
 b.U
 fc.O
10.0
1^.0
14,0
Ib.U
le.O
24,o   19:07
2b.u   xy:os
2e,o   19:1*1
30.0   19:14
3«;.u   iy:ib
34.0   iy:io
3b,o   if:4t)
3b.u   iti:4b
4U.U   ItiZbU
H«i.u   id:b<:
44.u   J.b:b4
4b.U   XbCbb
Hb.u   ib;b9
bu.o   xy:ui
b^.u   xy:u3
b4.o   xy:u7
bb.U   I9:ub
bb.o   19:10
60.0   iy:i3
6<:,0   I9:ib
bH.U   i9:^U
bo.U   X9:^4
bb.u   xy:^b
7o.u   xy:ub
7TlON

-------
                                             TABLE A.26
Date
4-13
4-14
4-14
4-19
4-23
4-24
4-24
4-24
4-29
5-2
Run
No.
13
14
15
16
17
18
19
20
21
22
Wind1
Velocity
(deg/mph)
151/4
095/8
090/8
110/10
218/18
234/12
235/10
325/185
270/116
315/97
SUMMARY
so2
Emission
(tons-hr-1)
12.3
11.7
11.7
10.9
23.3
19.5
18.1
18.0
12.4
24.1
OF FIELD PERIOD III — APRIL-MAY, 1970
Arcs
B
B
B
C
C
C
C
C
A
B
B
Run Time2
1435-1632
0923-1029
1107-1225
1349-1640
1025-1521
0912-0958
1038-1443
1554-1630
1048-1440
1115-1445
1518-1833
Sampling3
Duration
(min)
158
102
132
209
327
81
286
100
255
255
215
Precip.4
Rate
(mm-hr"4-)
0.004
0.66
1.03
0.04
Dry
1.00
0.19
1.44
Dry
0.35
so2
Recovery
(ymoles)
3.10
149.65
95.10
4.89

2.00
7.04
58.20

3.46
S02 Concentration
Max
Mean
Min
(pmoles-liter"1 )
83.8 21.0 1.7
21.2
20.9
27.8

1.7
8.7
29.1

0.6
17.1
13.6
15.6

0.5
2.7
8.8

0.3
13.3
10.0
0.3

0.2
0.3
4.7

0.2
1Wind velocity at stack exit, unless otherwise noted
2Time of concurrent sampling by all collectors.
3Average duration of collector exposure
4Average rate determined from average volume of precipitation collected and average duration of
 collector exposure.
51900 EST Pittsburgh rawinsonde.
60700 EST Pittsburgh rawinsonde.

-------
                          TABLE A.27
SlA<
  .0
 4,0
 b.O
 b.u
10,U
Ic.U
14.U
lb.0
Ib.U
20.0
22.0
24. U
30.U
bb.U
bb.U
70.0
7  CCNCtN THAT lON uuITS ARE
                 . SOL^ATE.  M1RITE  ANu  MTHATt

-------
      RUN  NO.14
DATE  4-14-70
TIME   0923-1029
                                                        PITTSBURGH
                                                        RAWIiiSONDE
                                                        PITTSBURGH
                                                        RAWINSONDE
                                                        AEROVANE
                         SPEED SCALE: 1/4" = 10 MPH
HEIGHT (FT)
2160
1905
1775
1650
1390
1130
865
775
600
335
300
TIME — 0700 0910 1040
PIBAL
PIBAL
RAWIN
PIBAL
PIBAL
PIBAL
PIBAL
RAWL'J
PIBAL
PIBAL
AERO


135/17




125/17




110/36

111/36
110/32
103/30
101/19

107/16
105/14
080/15

101/32

098/24
094/32
090/23
090/20

090/13
090/12
080/14











NOTES:  1.  WIND  ARROWS  POINT  DOWNWIND  TO  INDICATE  DIRECTION  OF
          EFFLUENT  TRAVEL. COMPASS  ORIENTATION  IS CONVENTIONAL-
          CLOCKWISE FROM  NORTH  TO TOP.
       2.  PIBAL WIND VELOCITIES ARE BASED  ON  THEORETICAL  RISE
          RATE  OF A DRY,  30  gm  PIBAL  WITH  71  gms  FREE  LIFT.
       3.  THE AEROVANE WAS MOUNTED  ABOUT  100  FT ABOVE  GROUND
          ON A  200  FT  HILL ONE-HALF MILE  SE OF  KEYSTONE.

       4.  PITTSBURGH RAWINSONDES FOR THE  2000  AND 3000 MSL
          LEVELS ARE PLOTTED AT SAME  HEIGHTS  ABOVE SURFACE
          AS AT PITTSBURGH.
             Figure A.25  Wind Profiles, Run 14

-------
LEGENDS
  *  flRC  fl  STHTI0NS
  <  flRC  B  STPTI0NS
  »  HRC  C  STHTI0NS
  K DENOTES PLflNT LOCUTION
                                       N
SCflLES
  I	1= 1 MILE
        = 10  MICR0MOLES/L
          Figure A.26  Sulfur Dioxide Distribution, Run 14

-------
                        TABLE A.28




                       bCAVEisGlKb UAlA  —  KtYSTONt STUUY


                  K  A4    APHiL  !«*» 197.)    AKC B


 bl/4.    b«|v--Hi\l3
          HC.H1L.U
        f-HON    1C             bu^   bC4      102    NU3     f-h
                          0    Ib.U
                          U    Ib.b1
  /.u   Sfijb  u:bu   «mo.o    14.2
  O.U   9:^0  Ijib2   234.0    14.IJ
  9.0   9:3^  IJIbb       la.u
 3b.U   «:b9  Iul4l   ^:4/.U    ln.1
 4U.U   hlUU  lu:>+4   
-------

-------
      RUN  NO.  15
DATE  4-14-70
TIME  1107-1225
             — TOP OF
               STACK
                                                        PITTSBURGH
                                                        RAWINSONDE
                               PITTSBURGH
                             -RAWINSONDE
                                                        •AEROVANE
                         SPEED SCALE: 1/4" = 10 MPH
HEIGHT (FT)
2160
1905
1775
1650
1390
1130
865
775
600
335
300
TIME-* 0700 1040 1140
PIBAL
PIBAL
RAWIN
PIBAL
PIBAL
PIBAL
PIBAL
RAW IN
PIBAL
PIBAL
AERO


135/17




125/17




101/32

098/24
094/32
090/23
090/13

090/12

080/14



098/19
093/20
087/20
080/14

079/16

065/10











NOTES:  1.  WIND  ARROWS  POINT  DOWNWIND  TO  INDICATE  DIRECTION  OF
          EFFLUENT  TRAVEL. COMPASS  ORIENTATION  IS CONVENTIONAL-
          CLOCKWISE F^?OM  NORTH  TO TOP.

       2.  PIBAL WIND VELOCITIES ARE BASED  ON  THEORETICAL  RISE
          RATE  OF A DRY,  30  gm  PIBAL  WITH  71  gms  FREE  LIFT.
       3.  THE AEROVANE WAS MOUNTED  ABOUT 100  FT ABOVE  GROUND
          ON A  200  FT  HILL ONE-HALF MILE SE OF  KEYSTONE.

       4.  PITTSBURGH RAWINSONDES FOR THE  2000  AND  3000 MSL
          LEVELS ARE PLOTTED AT SAME  HEIGHTS  ABOVE  SURFACE
          AS AT PITTSBURGH.
             Figure A.27  Wind Profiles,  Run 15

-------
LEGENDS
  *  ARC  fl  STflTIQNS
  «  RRC  B  STHTI0NS
  *  RRC  C  STflTIQNS
  K DENOTES PLflNT L0CRTI0N
                                       N
SCflLES
  I	1- 1 MILE
        = 10  MICR8M0LES/L
         Figure A.28  Sulfur Dioxide Distribution,  Run 15

-------
                         TABLE A. 29
SIA
  .0
2b.U
«itt.U
3u.u
34.
3b.
3c.
40.
4*:,
44.
4o.
4ti,
bu.
b*:.
bn.
bb.
bb.
bb.
bd.
t)4.
bt.
6h.
7U.
7k.
7»*.
7o.
7o.
ttU.
tic.
64.
ht .
Vu.u
        HKLCIH1TAUUM

                    Ab
                                                     STljUY
             :AUL
                 ft.
                      ^uu.o
       JL u :
              !«; : ou
                     lo/.U
                     i b /. o
                     I'+b.U
                     AHA.O
       i i : u u
       lu:
       iu:mi
                      lio.H
                      lbb.0
                      t>,U
                     ^4/,U
                       fc^.U
                       Ab.O
                               !«*»  1S»7(J
                              13. b
                              U.b
lH.7
Ib.b
11.b
Ib.u
11.u
lu.u
11.*!
11.0
13.1
lb.3
1^.4
14.1
1H.1
13.7
14.1
14.1
14.1
14.U
13.U
ID,4
lb.7
                              13.b
                              12.7
                                                b
                                      /.u
                                      o.l
                                      9.9
                                      b.tf
                                      -4.6
             LMl
                     »L.
                                             ,,',ITb
                                           M
                                                              r](>

-------

-------
      RUN  NO.  16
DATE  4-19-70
TIME   1349-1640
                                                        PITTSBURGH
                                                        RAWINSONDE
                                                        PITTSBURGH
                                                        RAWINSONDE
                                                        AEROVANE
HEIGHT (FT)
2160
1905
1775
1650
1390
1130
865
775
600
335
300
SPEED SCALE: 1/4" = 10 MPH
TIME-*- 1200 1348 1627 1900
PIBAL
PIBAL
RAW IN
PIBAL
PIBAL
PIBAL
PIBAL
RAW IN
PIBAL
PIBAL
AERO
152/31
150/27

144/25
135/24
127/19
113/18

110/15
101/12
090/15
139/17
139/21

127/25
121/23
113/17
100/13

096/12
090/10
090/16
168/12
165/15

162/14
152/10
121/20
119/33

119/31
121/25



140/26




120/24

120/14

NOTES:  1.  WIND  ARROWS  POINT  DOWNWIND  TO  INDICATE  DIRECTION  OF
          EFFLUENT  TRAVEL. COMPASS  ORIENTATION  IS CONVENTIONAL'
          CLOCKWISE F?OM  NORTH  TO TOP-

       2.  PIBAL  WIND VELOCITIES ARE BASED  ON  THEORETICAL  RISE
          RATE  OF A DRY,  30  gm  PIBAL  WITH  71  gms  FREE  LIFT.

       3.  THE AEROVANE  WAS MOUNTED  ABOUT  100  FT ABOVE  GROUND
          ON A  200  FT  HILL ONE-HALF MILE  SE OF  KEYSTONE.

       4.  PITTSBURGH RAWINSONDES FOR THE  2000  AND  3000 MSL
          LEVELS ARE PLOTTED AT SAME  HEIGHTS  ABOVE  SURFACE
          AS AT  PITTSBURGH.
             Figure A.29   Wind Profiles, Run 16

-------
                           \
                 •  \'
                X
                          4
                          4
                          4
                          4
4
4
                                     •
                                     f
                                                       4

                                                      4
                                              -•
                                         ..—        '
                                \f"

LEGENDS
  *  flRC  fl STflTIGNS
  «3  flRC-  B STflTIGNS
  *  RRC  C STflTI8NS
  K DENQTES PLRNT LQCflTIQN
                   SCflLES
                           H= 1 MILE
                           - 10 MICR0M0LES/L
         Figure A.30   Sulfur Dioxide Distribution, Run  16

-------
                             TABLE A.30



                                    UATA —  KEYSTONt STUOY

                   \  ib   AHKii  l^r  1S7'I
 S I A .     b/i|«(- LiiNib     HHC.C.IP       CClSLtiN 1 rtATlulNiS IN
           Her- i')u      VOLUl^t
         HKi>     10             bU2   bC^      '.02    I\C3
   l.U   lo:J^  X/:u3    i/.U      .3

   b.U   i3:iS  lc:4fa    2b.b    9,b

   /.U   13:^U  Jlcib^    la,U   10.7
   p.u   i3:ob  l/:ob     b.u   io.o

 13. U   13US  lo:Hb    2M. I I    Vl  i-l
 £. r • v   X w • C i7  X C • w O    JL O 0 *J    O • ^
 2**.b   io;3i  17:00    11.b    7.u
 2b.U   13:<3&  1?:OH     b.b    t>.3

 27.0   l^;4fa  I/.lo    H,U   11.1
 2u.b   10:
-------
                            TABLE A.31
                                    LAI* -
               L. t     H«U  luii//   i^:.<^      ^D
. U
.»'      .u
  /.u  iu:i,3   iszjb      ,u      .u
 10.u  iu:a      ,u      .^
 l b. u  i u: u s   i * i * ^
        .u * u•»   i^; ^o      .u      .b
Ih.O   lfj:il   IbiJt*      . U      . U
it.o   iu:i/   13:^3      .u      ,u
ko.o   iu:£j   it;:bH      .0     i.l
21.0   iu:^b   lb:tir      .u      ,u
it.O   10:00   I'j:«U   iu:«:a   lalbb      .U     l.n
             OM i  ib  •• L.  Sv>c' COi.CLM,
-------
      RUN  NO. 18
DATE  4-24-70
TIME   0912-0958
-2160 	
1 905
1650
•1390
1130 \

865 n— TOP OF ]
fnn 1 STACK

11

1









//
/ y
/ /
/ /
/
^

X
4*

















                                                        PITTSBURGH
                                                        'RAWINSONDE
                                                        PITTSBURGH
                                                        •RAWINSONDE
                                                        •AEROVANE
                         SPEED SCALE: 1/4" = 10 MPH
HEIGHT (FT)
2160
1905
1775
1650
1390
1130
865
775
600
335
300
TIME-* 0700 1002
PIBAL
PIBAL
RAW IN
PIBAL
PIBAL
PIBAL
PIBAL
RAW IN
PIBAL
PIBAL
AERO


195/44




165/29



225/32
225/38

220/34
222/36
226/34
234/26

238/16
246/11
210/10






















NOTES-  1.  WIND ARROWS  POINT  DOWNWIND  TO  INDICATE  DIRECTION  OF
          EFFLUENT  TRAVEL. COMPASS  ORIENTATION  IS  CONVENTIONAL-
          CLOCKWISE FROM  NORTH  TO  TOP.

       2.  PIBAL WIND VELOCITIES ARE BASED  ON  THEORETICAL  RISE
          RATE OF A DRY,  30  gm  PIBAL  WITH  71  gms  FREE  LIFT.
       3.  THE  AEROVANE WAS MOUNTED  ABOUT 100  FT ABOVE  GROUND
          ON A 200  FT  HILL ONE-HALF MILE SE OF  KEYSTONE.

       4.  PITTSBURGH RAWINSONDES  FOR THE  2000  AND  3000 MSL
          LEVELS ARE PLOTTED AT SAME  HEIGHTS  ABOVE  SURFACE
          AS AT PITTSBURGH.
             Figure A.31  Wind Profiles,  Run 18

-------

                           4
:•      .r      *
4       •
              ^il

        «««
                                                       4
                                                        4
                        4
                        4
                        4
                         4
LEGENDS
  *  RRC  fl  STHTI0NS
  
-------
                        TABLE A. 32
STA.
 0.0
 b.O
  PEKiOb
FKUN    TO
a:b9  io:i9
9:02  10:22
                SCAVENGING OATA — KEYSTONE. 5TUUY

         hUN  18    APHiL  2«*» 1970   ARC A

                            COlNCEN THAT IONS  IN  SAMPLES

                         SU2   bC4     N02     N03      PH
                          .3
                          .9
PKtLlP
VULUIVE

 82.0
                      iL  24,  1970   ARC C
SfA.


.0
1.0
2.0
3.0
4.0
b.O
6.0
7.0
b.O
9.0
li.O
13.0
14.0
Ib.o
16.0
17.0
Ib.o
19.0
20. U
21.0
22.0
<>3.0
24.0
2b.O
2b.U
27. U
27. b
2k. U
29.0
2y.b
30*0
30. b
SAILING
PEhUQiJ
FRCC 1C
a:4b io:ob
9:02 10:24
9:08 10:31
9:12 lo:38
a:43 9:bb
a:4b io;o3
a:bi 10:09
8:58 lUllb
9:03 10:23
9:0b lo:2b
a:4i iu;oo
a:4b iu:o3
a: u? 10:0?
8:49 iu:i3
a:b3 iu:i7
a:57 iu:2i
9:00 iu:27
9:u3 iu:32
9:ub io:3b
a:4b 10:00
0:49 io:ub
a:b3 io:io
a;bb iu:i4
a:ba 10:19
9:uo lu:24
9:04 iu:28
9:u^ io:3i
a:4b iu:>ju
b:4b iu:u4
ttlbc; luUO
a:bb iu:ib
a: be lullq
PKtClP
VOLUME

92.0
12^.0
102.0
93.0
87.0
4/.0
29.0
14.0
10.0
*:.o
10.0
13.0
7.0
b.O
b7.0
37.0
3b.O
97,0
217.0
23 7.0
190.0
177.0
19^.0
213.0
21*1.0
21*:. o
2bel.O
*:17.0
26<;.0
tct.U
titt.U
197.0
CONCENTHATIONS IN SAMPLES

S02
.3
.3
.3
.3
.6
.4
*b
1.7
.6
.9
.2
.7
.5
1.1
.3
.4
1.0
.3
.3
.3
.4
1.1
.3
.3
.4
.3
.3
.4
1.6
.6
.6
.3

SC4

16.5


12.7

13.4
2b.O
lb.5

14.8
13.1




11.0



11.6
13. b

11. u


a.o

9.7
8.9
10. b


N02 N03
.0540 3.9b
.0270 4.26


,0470 b.08


.U670




.0870



.0500 4,60



.0300 3.60


.0300 3.44


.0^50 2.6b


.1)170 3.18



PH
3.8n
3.72


4.00











3.9p



3.93


4.0Q


4. Ob


4.02


VOLUMt UMT Is I^L. S02  COjMCtMKATiGN UMTb
MCR(^OLtb/L. SbLhATE»  NIFKlTE ANU NlTKATt COMCtN IH/\1 ION
UMTS AHfc MLulGHAKS/L.

-------

-------
      RUN NO.  19
DATE   4-24-70
                                             TIME  1038-1443
                                                        PITTSBURGH
                                                        RAWINDSONDE
                                                        PITTSBURGH
                                                        RAW 11
                                                        AEROVANE
                         SPEED SCALE:
             1/4"  =  10 MPH
HEIGHT (FT)
2160
1905
1775
1650
1390
1130
865
775
600
335
300
TIME— 0700 1105 1405
PIBAL
PIBAL
RAWIN
PIBAL
PIBAL
PIBAL
PIBAL
RAW IN
PIBAL
PIBAL
AERO


195/44




165/29



226/40
225/46

225/37
221/29
224/25
230/23

238/12
244/18
230/8
249/36
251/38

254/34
251/30
246/26
242/23

239/20
238/21
220/20











NOTES-  1.  WIND ARROWS  POINT  DOWNWIND  TO  INDICATE  DIRECTION  OF
          EFFLUENT  TRAVEL. COMPASS  ORIENTATION  IS  CONVENTIONAL-
          CLOCKWISE F^OM  NORTH  TO TOP.

       2.  PIBAL WIND VELOCITIES  ARE BASED  ON  THEORETICAL  RISE
          RATE OF A DRY,  30  gm  PIBAL  WITH  71  gms  FREE  LIFT.

       3.  THE  AEROVANE  WAS MOUNTED  ABOUT 100  FT ABOVE  GROUND
          ON A 200  FT  HILL ONE-HALF MILE SE OF  KEYSTONE.

       4.  PITTSBURGH RAWJNSONDES FOR THE  2000  AND  3000 MSL
          LEVELS ARE PLOTTED AT  SAME  HEIGHTS  ABOVE  SURFACE
          AS AT PITTSBURGH.
            Figure A.33  Wind Profiles,  Run 19

-------
                                            4«\    /
                          4
                          4
                          4
                          i
                          4
                          4
                        4
                        4
                               V
    V
               4
               4
                                         4
LEGENDS
  * flRC fl STflTI6NS
  < flRC B STflTIQNS
  * flRC C STflTI0NS
  K DENOTES  PLRNT L8CHTI0N
N
A
                                        '     \
     SCflLES
       I	1=  1 MILE
            = 10 MICRQM0LES/L
       Figure A.34  Sulfur Dioxide Distribution, Run 19

-------
                            TABLE A.33
blA.
       PRtClHlTATluN  SCAVENGING  LATA — KfcYSTONE STUbY

              n.U\  i9    APRIL  2*,  197H   ARC A

                                 COKCENlRATlONS IN SAMPLES

                                    SC4     h02    N03     HH
 SIA.
          Hfc.h>iOL
        FRtP     To
  3.0  10;20   I4'.b2
  b.O  10:20   l4:bb
 39.0  10:11   i4:4<+
PrttCIP
VOLUME
72.0
44.0

.a
2,b
1.3
                   19   APRll -2 Hi

                     HKtClP
CONCENfRATIONS  IN SAMPLtS


1
^
0
4
0
b
/
b
9
10
11
14
10
14
Ib
Ib
I/
lo
19
20
21
24
20
24
2b
2b
2/
27
2ti
29
29
30
31

.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
• u
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.b
.0
.0
.b
.0
.b
h
10
10
10
10
9
lu
lu
IU
IU
lu
10
10
10
10
10
IU
10
lu
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
Ptk
Roiv
:ob
:24
:3i
: 3ti
:bu
:oo
:us
:ib
:23
:2b
:oi
:34
:ou
:o3
:u7
:i3
: 1 1
:2i
:27
:32
:3b
:oo
:uo
:io
:i4
:i9
:24
:2e
:3i
:uu
:(J4
:io
:ib
:i9
AOL,
14
Ib
is
13
14
14
14
Ib
la
Ib
Ib
Ib
14
14
14
lb
Ib
la
lb
lb
Ib
14
14
14
14
lb
lb
lb
Ib
14
14
14
14
la
TC
:
;
;
;
:
*
m
t
*
f
t
f
:
:
*
*
*
t
:
t
;
t
?
*
*
;
•
,
t
t
t
•
•
:
*
t
*
t
4b
Oo
12
j. ?
43
47
b3
L'O
35
43
bO
b4
45
b3
bti
iJ2
ub
10
15
20
24
4b
4d
b3
bb
00
04
07
lb
4b
4ti
b3
bb
01
37
229
104
y (
*
.
.
.
ob.
He:
3^
2b
10
47
3/
bo
32
34
17
44
bf
VI
U/
14t
14/
112
9/
U2
£l
b4
47
b£>
177
44
172
104
b7
104
,
,
,
,
•
.
.
.
.
.
,
*
,
.
.
.
.
.
.
,
.
,
,
*
,
,
,
.
.
0
0
0
0
0
0
0
0
0
0
0
0
0
u
0
0
0
0
0
u
0
0
0
0
o
0
0
0
0
0
0
0
0
0
S02

b
1
1



1
22
3
3
B

4
2
4
1
1
3
2
3
3
1
3
5
2


1
1
2
1
1

.9
.5
.4
.3
.b
.7
.3
,8
.4
.1
• U
.7
,7
.b
.4
.4
.6
.b
,7
.b
.b
.7
.9
.4
.b
.9
.b
.9
,b
.2
.9
,1
,7
.3
                                   SCH
                                  Id.i
                                  lb.9
                                   9.9
                                   9.1
                                   11.1
                                             U02
                                            .U210
                                            .U10U  ?.b8
i-Olt: VoLUivt CM f

      UMTb AKh
                           PH
                          3.7b
                          3.94
                          4,06
                          4.00
                          4.06
                                                           4.23
                                                           4.20
                                           .1)280
                                           .0370
                                           .U200   4.33
                                           .OlbO  3.12    4,13
                                           .0320  5.42    4.13
                                           .UlHU  2.bb    4.18
                                           .0210  4.34    4.00
                    i*L. bO.d COi\CtMRAHON  UMTS ARt
                   bOLl-ATE» MjRlTL  ANU  NlTKATt CONCENTRATION

-------

-------
      RUN  NO.  20
DATE   5-24-70
TIME   1554-1630
-2160- 	
1 905
1650
1390 	 	
1130' 'A
865 n— TOP OF *
rnn I STACK

335 / 1 	 ' '
/
r

















X
*r











"^
PITTSBURGH
r1 RAWINSONDL
PITTSBURGH
i RAUTN^nMnF

AtKOVANE
HEIGHT (FT)
2160
1905
1775
1650
1390
1130
865
775
600
335
300
SPEED SCALE: 1/4" = 10 MPH
TIME-* 0700 1545 1615 1900
PIBAL
PIBAL
RAWIN
PIBAL
PIBAL
PIBAL
PIBAL
RAM IN
PIBAL
PIBAL
AERO


195/44




165/29













230/19










290/30


330/23




325/18


'
NOTES:  1.  WIND  ARROWS  POINT  DOWNWIND  TO  INDICATE  DIRECTION  OF
          EFFLUENT  TRAVEL. COMPASS  ORIENTATION  IS  CONVENTIONAL'
          CLOCKWISE FROM  NORTH  TO TOP.

       2.  PIBAL WIND VELOCITIES ARE BASED  ON  THEORETICAL  RISE
          RATE  OF A DRY,  30  gm  PIBAL  WITH  71  gms  FREE  LIFT.

       3.  THE AEROVANE  WAS MOUNTED  ABOUT 100  FT ABOVE  GROUND
          ON A  200  FT  HILL ONE-HALF MILE SE OF  KEYSTONE.

       4.  PITTSBURGH RAWINSONDES  FOR THE  2000  AND  3000 MSL
          LEVELS ARE PLOTTED AT SAME  HEIGHTS  ABOVE  SURFACE
          AS AT PITTSBURGH.
             Figure A.35  Wind Profiles,  Run 20

-------
                             \
1     I
                    \
                          4"
                          4
                         4
                         4
                         t
                                      •   ''• /
                                     I      \
    ..-"•

LEGENDS
  X  RRC  R  STflTIQNS
  <  RRC  B  STRTI0NS
  «  RRC  C  STflTIBNS
  K DENOTES PLRNT L0CRTI0N
         SCRLES
                  = 1 MILE
                 =  10  MICR0M0LES/L
        Figure A.36  Sulfur Dioxide Distribution,  Run
                    20

-------
                       TABLE A.34




  PRECIPITATION SCAVENGING UATA — KEYSTONE STUDY

         HbN 
-------

-------
      RUN  NO.  21
DATE  4-29-70
TIME   1048-1445
-2160 	
1 905
1 650
1 390
1 1 30
865 r,— TOP OF
rnn I STACK
ODD | 1
T T rL 1 1





















r










^

PITTSBURGH
KAWlNbUNUL
PITTSBURGH
— RAWINSONDE
AtKOVANE
                         SPEED SCALE: 1/4" = 10 MPH
HEIGHT (FT)
2160
1905
1775
1650
1390
1130
865
775
600
335
300
TIME -— 0700 1100 1300
PIBAL
PIBAL
RAWIN
PIBAL
PIBAL
PIBAL
PIBAL
RAWIN
PIBAL
PIBAL
AERO


280/14




270/11













215/6










240/7











NOTES-  1.  WIND ARROWS  POINT  DOWNWIND  TO  INDICATE  DIRECTION  OF
          EFFLUENT  TRAVEL.  COMPASS  ORIENTATION  IS CONVENTIONAL-
          CLOCKWISE F?OM  NORTH  TO  TOP.

       2.  PIBAL WIND VELOCITIES ARE BASED  ON  THEORETICAL  RISE
          RATE OF A DRY,  30  gm  PIBAL  WITH  71  gms  FREE  LIFT.
       3.  THE  AEROVANE WAS  flOUNTED  ABOUT 100  FT ABOVE  GROUND
          ON A 200  FT  HILL  ONE-HALF MILE SE OF  KEYSTONE.

       4.  PITTSBURGH RAWINSONDES   FOR THE  2000  AND  3000 MSL
          LEVELS ARE PLOTTED AT SAME  HEIGHTS  ABOVE  SURFACE
          AS AT PITTSBURGH.
            Figure A.37   Wind Profiles, Run 21

-------

LEGENDS
  *  RRC fl STflTIQNS
  <  flRC B STflTIQNS
  *  flRC C STflTIQNS
  K DENOTES  PLflNT  LOCRTI0N
                                      N
SCRLES
        1= 1  MILE
          H   10 MICROM0LES/L
          Figure A.38   Sulfur Dioxide Distribution,  Run 21

-------
                          TABLE A.35



                      SCAVENGING DATA  —  KEYSTONE STUDY

               rtUN 21   APRIL 2y»  1970   A«c  A

 SlA,    SWH-li\ti    PHtClP      CONCENTRATIONS IN SAMPLES
          HtKlCU     VOLUME
        FKoi"    TC            S02    SC4     N02    N03     pH
   .u  10:29  14:40     .0     *ti
  i.o  10:30  I4:4b     tO     .0
  2.0  10:31  14:47     .0     .7
  3.0  10:3^  14:49     .0     .1
  4,0  10:33  14:50     «o     .6
  b.o  io:3b  m:b2     .0    2.6
  o.O  iu:3o  14:5*4     .U     .3
  7.U  10:37  l<+:b6     .0    2.3
  ti.u  io:3ti  m:b/     .0    i.a
  9.0  iu:m  i-4:bd     tO     .1
 lo.o  io:*4&  ib:uu     *o    1.3
 li.o  10:30  i*4:*45     .0    1.1

 lo.o  10:33  JLH:na     .0    b.l

 ib.o  io:3b  14:51     .0    1,2
 17.0  10:37  14:53     .0    2.4
 lti.0  10:39  14:55     .0    3.3
 IV.0  10:41  I4:b7     .0    4.4
 20.0  10:42  14:59     .0     .b


l\01t: V/OLUlvt UMT ib PL. S02 CONCtNTKflT ION  yMTS APt
                    SULHATtf NITHirt AND  NljkATE CONCtfM
      UNITS

-------
                            TABLE A.36




                       bLAVfci\,GlNo UAIA —  ^YSTONt

                             L 2y, ly/it    AKC  ti
SlA.     SfcM-HUCj     H*tUP      CCNCtMKAUoNS IN $Al*HLtS
          f-trUOu      VULUh't
        *"*W    TL             bu£   bCU      1^02    NOi
 l.u   iu:j*i  i4;qb      .u     ,ti
 
-------
                            TABLE A.37
                   iOK  bLnVklMfiiMs  LAM — Kfc.YS1Ur,c  STUUY


                    c.*-      VAY  c.»  I97u   AHC «


              Ii\u    PKtLiP       LClsCth lHATluMS  IN


                TC
ii.u   ibiLj  ir:.sc   iUo.u      ,i
<:-.u   ib:oo  io:»4u   liu.u      »>3
«i*+tu   ib:uf  13:4,5   ilo.o      .h
iJa.u   ib:iu  icJ4b   im.u      .4
2c.u   1^:1*;  i«:Hd   luu.u      .3
«;7.u   ib:ij  irfZbu   IUH.U      .4
kv.u   ib:io  iiiib*;    yo.n      ,j
Jl.u   lb:in  IdJ^b   llo.Li      ,H
7j.u   xb:jib  1^:12
             OKJ.I  ib  |VL.  SOt: COuctNTHflTlCN lir-ITS
                   . bOLHATt. NiihflTt  /\Ku ^IlKATt CQKCtMKM

-------
                               APPENDIX B

     REVERSIBLE WASHOUT — AN EXAMINATION OF THE CONSEQUENCES OF
     LINEARITY. NONLINEARITY. AND WASHOUT THROUGH DROPLET CAPTURE

Chapter III emphasizes the distinction between processes of reversible
and irreversible washout.  Additional examination of the equations demon-
strates the utility of dividing  the study of reversible washout into the
two following subcatagories:  linear and nonlinear.  Linear washout
processes are ones that satisfy  the following requirement:
            NAo = " Ky  (yAb " HxAb)  ;  Ky  ' H  constant     ,        (B.I)

 (cf. Equation 3.5).  Nonlinear systems, conversely, are ones that do not
 satisfy Equation B.I.   One should note here that "first-order" systems
 are not necessarily linear, unless Henry's law is obeyed.

 The distinction between linear and nonlinear washout is somewhat arti-
 ficial, since no washout process can be expected to conform exactly to
 the conditions of linearity.  Some systems should approach linearity to
 a fairly high degree, however, and the use of  this idealization is
 reasonably valid.  Because of its blatant disobeyance of Henry's law the
 sulfur dioxide-water system must be considered nonlinear for most cases
wherein washout is of interest; however, linear theory is still useful
 to some extent here, as will be shown in the following development.

 To illustrate the properties associated with linear and nonlinear washout
 it is convenient to visualize a single drop of radius a falling through
 polluted air of concentration yAb.  Performing a material balance on the
 drop and combining with Equation B.I provides

-------
                   dc.,    3K
                   _Ak	Z.  (y   . H' c., )     ,                 (B.2)
                    dz    v a   JAb       Ab
where c ,  is the average concentration of pollutant in the drop and H1 is
the Henry's-law constant modified to units of c...  For the corresponding
case wherein Henry's law is not obeyed one may write
                    dc.,    3K
g' denoting the appropriate nonlinear functional equilibrium relationship
for y.  in terms of c,, .
     Ae              Ab

Now consider a number of drops of radius a falling steadily through a
stationary pollutant plume into a fixed collector at ground level.  The
plume concentration at any point (x,y,z) in space will fluctuate with
time; however, the average concentration at that point may be expressed
by the integral
        i  r
>v»z)  ~"7  /
                                                  dt
where T is some appropriately- large averaging period.

For linear conditions Equation B.2 may be solved to obtain the ground
level concentration of any given drop
                       3K
                          'I        I ~ v   I
                                                                    (B.5)

-------
where ZQ is some datum point high above the plume.  Since the plume is
fluctuating with time, drops entering the collector at different times
will have varying concentrations.  The average concentration of steadily
falling drops of radius a collected at times between t=o and t-T is
given by
                            1  f
                            T   /
                               ^O
CAvg -        C(t)  dt
                            ^_   r
                            TaVt  Jz
       3K
         T   I         I ~ fT   \
                                               (B.6)
                             3K    ff
                             -*  I
                             avt Jz
showing that the time-averaged  concentration in the collected sample
should be related simply  to  the time-averaged concentration in the plume,
whenever linear conditions prevail.

This same type of analysis can  be applied  to irreversible linear washout
theory to produce a similar  result.  For nonlinear systems, however,
analytical integration of Equations B.2 or B.3  (if possible at all) tends
to weight disproportionately the influence of gas-phase fluctuations.
Therefore, c.   cannot be expressed simply in terms of y.  ; because of
            Avg                                         Avg
this much more must be known about the plume before washout can be deter-
mined, and the overall problem  is complicated enormously.

It should be noted that this analysis is somewhat incomplete in that it
implies that all drops of radius a that enter the collector have fallen

-------
through the same trajectory - a situation not expected to occur  in nature
owing to time-fluctuations in wind velocity.  Further analysis of this
question and of the consequences of employing average plume concentrations
to nonlinear washout problems will be the subject of a forthcoming publi-
cation'68'.
An extended analysis of Equation B.2 can be employed to investigate the
significance of the mechanism of gas washout through pickup of suspended,
gas-laden fog droplets.  (Mechanisms 2-3-4-10 in Figure 1.1.)  If one
assumes  (conservatively) that the rate of water-mass pickup by the falling
drop is negligible compared to the mass of the drop itself, then from
aerosol-washout theory
                      (53)
      one may write
       Gain of concentration in drop through mechanism 2-3-4-10 with
                    dc
       distance  =
                      'Ab
                     dz
                                              m
                           2-3-4-10
                       4 a Hf
                                              (B.7)
where m is the mass concentration of saturated, suspended, droplets,
E(a) is the droplet washout efficiency, and p  is the density of water.
Assuming additivity of mechanisms, this may be combined with Equation
B.2 to give
         dc
          _Ab
          dz
3E(a) m p
         w
3K
                       4 a H1
              v a
               z
       'Ab
3K  H'
  y
 v a
  z
                                              (B.8)
which becomes, upon integration,

-------

Inspection of Equation B.9 shows that the terms for pollutant-gas washout
from the mechanisms of fog-droplet capture and simple gas absorption are
additive; their relative effects, therefore, can be assessed simply by com-
paring the bracketted terms.  Even though Equation B.9 pertains strictly to
linear systems only, this equation can be utilized for order-of-magnitude
comparisons for nonlinear washout simply by substituting constant equi-
librium and mass-transfer coefficient values corresponding to the conditions
of interest.

It is of interest to perform such a comparison for the sulfur dioxide-water
system.  Consider, for instance, a point in a plume wherein the ambient sul-
fur dioxide concentration is 1.0 part per million, corresponding to a
linearized H' of about 20 cm3/mole.  The following table provides a comparison
of the relative importances of the mechanisms for drops of various sizes
falling through this point.  These calculations are based on the following values:
E(a) = 1, m = 3 gm/m3, K  given by the lowest  values appearing in Table 3.2.


                                 TABLE  B.I
               COMPARISON OF SULFUR DIOXIDE WASHOUT MECHANISMS

                                    3E(a)m          _ 3 Ky
               Drop Diameter       4 a pw H'          vt a
               	cm	       moles /cm1*        moles/cm1*
                   0.03            7.5 x 10-6      4.2 x lO"1*
                   0.10            2.3 x 10~6      1.5 x 10~5
                   0.30            7.5 x 10-7      9.0 x 10~7


From Table B.I it is obvious that the fog-capture mechanism should rarely be
a significant factor in sulfur dioxide washout by rain.  This is especially
apparent when one notes that the values employed in calculating the groups


-------
in the table were chosen so as to accentuate,  as much as is realistically
                      •
possible, the relative effects of the fog-capture  mechanism.  One should note
that at lower ambient concentrations of sulfur-dioxide,  the relative effect
of the fog-capture mechanism should be enhanced, this arising from a lowering
of H1.  This enhancement, however, is still insufficient to render the fog-
capture mechanism of any importance under circumstances  of practical meteor-
ological interest.

-------
         3E(a) m p      3K
CA  =	w     _v_
 A°        4 a H'
                                ]r
Inspection of Equation B.9 shows that the terms for pollutant-gas washout
from the mechanisms of fog-droplet capture and simple gas absorption are
additive; their relative effects, therefore, can be assessed simply by
comparing the bracketted terms.  Even though Equation B.9 pertains
strictly to linear systems only, this equation can be utilized for order-
of-magnitude comparisons for nonlinear washout simply by substituting
constant equilibrium  and mass-transfer coefficient values corresponding
to the conditions of interest.

It is of interest to perform such a comparison for the sulfur dioxide-
water system.  Consider, for instance, a point in a plume wherein the
ambient sulfur dioxide concentration is 0.5 parts per million, corre-
sponding to a linearized H' of about 600 cm3 /mole.  The following table
provides a comparison of the relative importances of the mechanisms for
drops of various sizes falling through this point.  These calculations
are based on the following values:  E(a) = 1, m = 3 gm/m3, K  given by
lowest values appearing in Table 3.2.

                               TABLE B.I
            COMPARISON OF SULFUR DIOXIDE WASHOUT MECHANISMS

                              3E(a) m P
                                4aH'                   V
          Drop                        .                        u
       Diameter,  cm           moles /cm1*               moles/cm1*
          Oi03                2.5 x 10~7              4.2 x 10~4
          0>10                7.5 x 10~8              1.5 x 10~5
          Oi30                2.5 x 10~8              9.0 x 10~7

-------
From Table B.I it Is obvious that the fog-capture mechanism should never
be a significant* factor in sulfur dioxide washout by rain under these
conditions.  This is especially apparent when one notes that the values
employed in calculating the groups in the table were chosen so as to
accentuate, as much as is realistically possible, the relative effect of
the fog-capture mechanism.  One should note that at lower ambient concen-
trations of sulfur dioxide, the relative effect of the fog-capture
mechanism should be enhanced, this arising from a lowering of H1.  This
enhancement, however, is still insufficient to render the fog-capture
mechanism of any importance under circumstances of practical interest.

-------
                              APPENDIX C


                      SIGN-X ANALYZER CALIBRATION


Laboratory calibrations of the Sign-X analyzer were performed using sul-

fur dioxide from permeation-tube and syringe-pump sources as shown in

Figure C.I.  Pure nitrogen was used as the diluent in all calibrations.

Carbon dioxide was admitted to the system directly from a gas cylinder

and controlled with a fine needle value. The results are given in Table

C.I.


                              TABLE C.I


                      SIGN-X ANALYZER CALIBRATION

          0 ppm CO- Added                     256 ppm C0» Added
S02 ConcentrationSign-X Reading*  S02 ConcentrationSign-X Readings'

     (ppm)           (mv scale/10)        (ppm)           (mv scale/10)

     0.110              0.077             0.000              0.053
     0.143              0.097             0.118              0.131
     0.147              0.100             0.143              0.159
     0.198              0.143             0.146              0.160
     0.202              0.145             0.180              0.182
     0.250              0.181             0.191              0.193
     0.252              0.195             0.235              0.227
     0.368              0.300             0.265              0.237
     0.419              0.361             0.306              0.278
     0.580              0.430             0.383              0.345
     0.910              0.610             0.594              0.476
     0.945              0.735             0.925              0.760
     2.00               1.09              2.00               0.985
     4.00               2.30              2.00               1.09
     8.00               4.26              4.00               2.06
                                          8.00               3.99
                                          8.00               4.15
 Corrected  for  zero and reagent conductivity, read as minimum value
 of  fluctuating curve, filter out.

-------
                                                        MIXING VESSEL
N>
                            PERMEATION-TUBE  HOLDER
                            (OR SYRINGE  PUMP)>
           N2  SUPPLY
                    GAS METER
           CONSTANT-TEMPERATURE BATH
DILUENT PREHEATER
                                                                                 C02 SOURCE
                                                                                SIGN-X
                                                                               ANALYZER

-------