EPA-600/3-76-110
November 1976
Ecological Research Series
MIDWEST INTERSTATE SULFUR TRANSFORMATION
AND TRANSPORT PROJECT:
Aerial Measurements of Urban and
Power Plant Plumes, Summer 1974
Environmental Sciences Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
1
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into five series. These five broad
categories were established to facilitate further development and application of
environmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The five series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
This report has been assigned to the ENVIRONMENTAL MONITORING series.
This series describes research conducted to develop new or improved methods
and instrumentation for the identification and quantification of environmental
pollutants at the lowest conceivably significant concentrations. It also includes
studies to determine the ambient concentrations of pollutants in the environment
and/or the variance of pollutants as a function of time or meteorological factors.
This document is available to the public through the National Technical Informa-
tion Service. Springfield, Virginia 22161.
-------
EPA-600/3-76-110
November 1976
MIDWEST INTERSTATE SULFUR TRANSFORMATION AND TRANSPORT PROJECT:
Aerial Measurements of Urban and Power Plant Plumes, Summer 1974
by
W.H. White, J.A. Anderson, W.R. Knuth
D.L. Blumenthal and J.C. Hsiung
Meteorology Research, Inc.
Altadena, California 91001
R.B. Husar
Washington University
St. Louis, Missouri 63130
68-02-1919
Project Officer
William E. Wilson
Atmospheric Chemistry and Physics Division
Environmental Sciences Research Laboratory
Research Triangle Park, North Carolina 27711
ENVIRONMENTAL SCIENCES RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NORTH CAROLINA 27711
-------
DISCLAIMER
This report has been reviewed by the Environmental Sciences Research
Laboratory, U.S. Environmental Protection Agency, and approved for
publication. Approval does not signify that the contents necessarily
reflect the views and policies of the U.S. Environmental Protection Agency,
nor does mention of trade names or commercial products constitute endorsement
or recommendation for use.
11
-------
ABSTRACT
A portion of the research activities of the Midwest Interstate Sulfur
Transformation and Transport Project (Project MISTT) during the summer of
1974 is documented. Using a light plane equipped with instruments for mea-
suring air pollutants and meteorological parameters, investigators mapped
the three-dimensional distribution of aerosols and pollutant gases originating
in the St. Louis area. Each day's flight plan was designed to characterize
a large pollutant plume at discrete distances downwind from its source. The
plume from the coal-fired power plant at Labadie, Missouri was followed out
to 45 km. Secondary aerosol production within the plume was documented. The
estimated average conversion rate for sulfur dioxide to sulfate was about
three percent/hour at the distances sampled. The overall removal rate of S02
was too small to detect, and no net production of ozone was observed. Large
pollutant plumes were also identified downwind of central St. Louis and the
Wood River refineries. These urban-industrial plumes were followed out to
60-70 km, where they were characterized by elevated concentrations of ozone
and light-scattering aerosols.
iii
-------
CONTENTS
Abstract iii
List of Figures vi
List of Tables ix
Acknowledgement x
Sections
I Introduction 1
II Experimental Procedure 8
III Meteorology 31
IV Sulfur Chemistry 40
V Oxidant Chemistry 63
VI References 85
Appendices 89
-------
FIGURES
No. Page
1 Outline of SO3 Plume Downwind of Labadie Power Plant
on 14 August 1974 4
2 Horizontal Profiles of the SO2 Concentration and Light
Scattering Coefficient (bsca{;) Downwind of Greater St.
Louis Under Three Different Wind Regimes 5
3 Cessna 206 Sampling Aircraft 9
4 Installation of Electrical Aerosol Analyzer, Royco
Electronics, Multichannel Analyzer, and Multifilter
Sampling Manifold in MRI Aircraft 12
5 Installation of Grab Bag Sampling System in MRI Aircraft 13
6 Wind Pod Housing Optics for Royco Optical Particle
Counter 14
7 Installation of Sulfate Filter Holders in MRI Aircraft 15
8 Installation of Gas Monitors in MRI Aircraft 16
9 Corrected Transient Response of Theta Systems SOS
Monitor (Markowski, 1975). 18
10 Probes for MRI Airborne Instrument Package 19
11 Installation of Data Loggers in MRI Aircraft 21
12 Sample Intake Manifolds on Left-hand Side of Aircraft 23
13 Instrument Layout in MRI Aircraft 24
14 Sample Flow in MRI Aircraft 25
15 Instrument Schematic for First Data Logger 26
16 Instrument Schematic for Second Data Logger 27
17 Flight Plan for Spiral Soundings 29
vi
-------
FIGURES (continued)
Nq^ Page
18 Flight Plan for Horizontal Traverses 30
19 St. Louis Weather During Sampling Period 32
20 Weather Map for 0700 CDT on 1 August 1974 33
21 Weather Map for 0700 CDT on 5 August 1974 35
22 Weather Map for 0700 CDT on 12 August 1974 37
23 Weather Map for 0700 CDT on 15 August 1974 39
24 Winds Aloft on 14 August 1974, From Pilot Balloon
Measurements at Marthasville, Missouri 41
25 Location of Aircraft Sampling Routes and Pilot Balloon
Release Points 14 August 1974 43
26 Horizontal Plume Traverse 10 km Downwind of Labadie
Stacks on 14 August 1974 44
27 Horizontal Plume Traverse 21 km Downwind of Labadie
Stacks on 14 August 1974 45
28 Horizontal Plume Traverse 45 km Downwind of Labadie
Stacks on 14 August 1974 46
29 Composite Side View of Labadie Plume on 14 August 1974 48
30 Vertical Profile from Spiral Descent at Labadie on
14 August 1974, 1045-1121 CDT 49
31 Scatter Plot of Particulate Sulfur Concentrations (S^) vs.
Plume bscat and Background bscat for Labadie Samples 52
32 Scatter Plot of Particulate Sulfur Concentration (S ) vs.
Plume bscat and Background bgcat for Wood River Samples 53
33 Two Aerosol/Emissions Ratios in the Labadie Plume on
14 August 1974 56
VII
-------
FIGURES (continued)
No. Page
34 Horizontal Plume Traverse 21 km Downwind of Labadie
Stacks on 14 August 1974 59
35 Horizontal Profiles of Ozone and Oxidant Concentrations
Downwind of Greater St. Louis Under Three Different
Wind Regimes 64
36 Vertical Profiles of Ozone Concentration and Temperature
on 30 July 1974 65
37 Cross Section of Ozone Plume Downwind of Greater St.
Louis on the Afternoon of 30 July 1974 66
38 Horizontal Plume Traverse 10 km Downwind of Labadie
Stacks on 14 August 1974 69
39 Horizontal Plume Traverse 21 km Downwind of Labadie
Stacks on 14 August 1974 70
40 Horizontal Plume Traverse 45 km Downwind of Labadie
Stacks on 14 August 1974 71
41 Early-morning Horizontal Profiles of Ozone Concentra-
tion, Oxidant (O3 + NO2) Concentration, and Light-
scattering Coefficient (bscat) Downwind of Greater St.
Louis on 3 August 1974 72
42 Horizontal Profiles of Oxidant Concentration and Light-
scattering Coefficient (bscat) Downwind of Greater St.
Louis Under Three Different Wind Regimes 75
43 Horizontal Plume Traverse 21 km Downwind of Labadie
Stacks on 14 August 1974 80
44 Horizontal Plume Traverse 45 km Downwind of Labadie
Stacks on 14 August 1974 81
Vlll
-------
TABLES
No.
1 Summary of 1974 MISTT Aircraft Sampling Program 3
2 Instrumentation of MRI Aircraft for 1974 MISTT
Field Program 10
3 Aircraft Sampling Schedule, 14 August 1974 42
4 Particulate Sulfur in the Labadie and Wood River Plumes 51
5 Sampling Environment of Sulfate Filters 58
6 Parameters Used in Model Calculations 83
IX
-------
ACKNOWLEDGEMENTS
An experimental program of this size could not have been completed
without the coordinated efforts of many individuals. Key responsibili-
ties in the collection and processing of data were shared by S. Gregg,
S. Muller, and K. Lamb of MRI, J. Husar, K. Horton, and N. Gillani
of Washington University, and B. Cantrell and K. Whitby of the
University of Minnesota. Active support was given by many others.
Valuable comments on the interpretation of the data were received from
J. Ogren (MRI), W. Wilson (U. S. Environmental Protection Agency),
H. Walker (Monsanto Polymers and Petrochemicals Co.), and P.
McMurry (California Institute of Technology). Typing and illustrations
for this report were supervised by T. Herrick and C. Williams.
-------
SECTION I
INTRODUCTION
PROJECT MISTT AIRCRAFT SAMPLING PROGRAM
During July and August o£ 1973, 1974, and 1975, the U.S. Environmental
Protection Agency (EPA) funded a series of coordinated field studies in
the St. Louis area as part of Project MISTT (Midwest Interstate Sulfur
Transformation and Transport). Meteorology Research, Inc. (MRI) pro-
vided aircraft monitoring support for these studies, using an instrumented
light plane to map the three-dimensional distribution of aerosols and pollu-
tant gases in large plumes. This report presents a description of the 1974
aircraft sampling program, together with an analysis of the results. The
data from the program are presented in summary form here; with the
exception of particle size distributions, which will be reported separately
by the Particle Technology Laboratory of the University of Minnesota, a
complete listing of the 1974 aircraft data is available in White, et al (1975).
The results of the 1973 aircraft sampling program were reported by
Anderson and Blumenthal (1974).
A fixed-wing aircraft offered certain advantages as a sampling platform,
beyond its obvious three-dimensional mobility. The speed of the MRI
Cessna 206 enabled it to sample large areas in a short time interval,
making it possible to determine the large-scale geometry of pollutant
plumes. Flying a few hundred meters above the ground, monitoring instru-
ments were not "blinded" by small local sources as they sometimes are on
the ground. These features are well illustrated by Figures 1 and 2, which
show how clearly large plumes can be identified in the data taken by the
aircraft. Finally, as will be discussed later, knowledge of the vertical
distribution of pollutants was necessary for the calculation of pollutant
budgets.
The MRI Cessna 206 flown in the 1974 MISTT field program was equipped
for the continuous monitoring of gas concentrations (SO2, O3, NO, NOX,
CO), aerosol indices (light-scattering coefficient, condensation nuclei
count), and meteorological variables (temperature, relative humidity, tur-
bulent dissipation). Data from these instruments were recorded at 0.4-
second intervals on magnetic tape cartridges. An on-board electrical
mobility analyzer/optical particle counter/multichannel analyzer system
determined the particle size distributions of grab samples, and a multiple-
-------
filter sampling system collected aerosol for later sulfate analysis at
Washington University. A detailed description of the aircraft, instru-
mentation, and experimental procedure is given in Section II.
Intensive airborne sampling was conducted on 10 days of 1974, extending
from 30 July to 15 August (Table 1). The meteorology of the sampling
period is summarized in Section III. Each day's flight plan, outlined
in Appendix A, was designed to characterize a large pollutant plume at
discrete distances downwind from its source. Selected for study were
the plumes from three major sources: the coal-fired power plant at
Labadie, Mo., an elevated point source; the refinery complex at Wood
River, 111., a concentrated low-level area source; the central St. Louis
urban-industrial area including East St. Louis and Granite City, a
diffuse low-level area source. The plumes from all three sources are
identifiable in Figures 1 and 2, and are analyzed in Sections IV and V.
The pollutant flow downwind of the Labadie power plant was mapped by
the MRI aircraft on four days during the 1974 MISTT field program.
Shifting winds during the initial attempt on 31 July restricted sampling
to the immediate vicinity of the plant. On the morning of 1 August and
early evening of 5 August, the plume was characterized at three down-
wind distances, from 1 km to 16-18 km. Under nearly steady-state
wind conditions during the middle of the day on 14 August, the plume
was characterized at 10, 21, 32, and 45 km downwind of the plant
(Figure 1).
The St. Louis and Wood River plumes were usually mapped together
(Figure 2), and it is convenient to refer to the combined plumes as
"the urban-industrial plume. " This plume was mapped by the MRI
aircraft on six days during the 1974 MISTT field program. On most
of these days the plume was characterized in one set of passes
immediately downwind of the metropolitan area and again in a second
set of passes 30-60 km farther out. On 12 August it was characteri-
zed at three downwind distances.
RESULTS AND CONCLUSIONS OF 1974 PROGRAM
The Labadie Plume
Much of the analysis in this report is focused on the Labadie plume,
which possesses the advantages of relatively well-defined geometry
and chemistry, together with a minimal interaction with the ground.
Peak sulfur dioxide concentrations measured within the plume varied
from over 10 ppm at 1 km from the plant to about 0. 06 ppm at 45
km from the plant. Peak sulfate concentrations at 10 km from the
-------
Table 1. SUMMARY OF 1974 MISTT
AIRCRAFT SAMPLING PROGRAM
Date
30 July
31 July
1 August
Time
(CDT)
0900-1615
0645-0900
0800-1115
Plume
Urban-Industrial
Power Plant
Urban Industrial
3 August 0815-1615 Urban-Industrial
5 August 1730-2015 Power Plant
10 August 1500-1815 Urban-Industrial
11 August 0900-1300 Urban-Indus trial
12 August 1100-1500 Urban-Industrial
14 August 1045-1515 Power Plant
15 August 0945-1230 Urban-Industrial
Urban-industrial plume sampling was conducted downwind of greater
St. Louis area, including Wood River (Illinois), East St. Louis
(Illinois), and Granite City (Illinois); power plant plume sampling was
conducted downwind of coal-fired steam generating station at Labadie
(Missouri).
-------
Figure 1. Outline of SO 3 plume downwind of Labadie power plant
on 14 August 1974. Solid lines show sampling paths,
dashed curves indicate region in which SO3 concentra-
tions above 0. 01 ppm were measured between 455 and
610 m (1500 and 2000 ft) msl. Sampling began near the
plant at 1045 CDT, ended 45 km downwind at 1515 CDT.
-------
Figure 2. Horizontal profiles of the SOS concentration and light scatter-
ing coefficient (bscat) downwind of Greater St. Louis under
three different wind regimes. Profiles were recorded during
traverses of profile baselines at the following altitudes and
times: (counter-clockwise from lower right) 455 m (1500 ft)
msl, 0933-0952 CDT, 30 July 1974; 610 m (2000 ft) msl,
1145-1152 CDT, 12 August 1974;.455 m (1500 ft) msl,
1027-1048 CDT, 15 August 1974. Arrows show average
winds measured in mixing layer during sampling period;
their lengths equal distance covered in one hour at average
wind speed. Large aerosol plumes can be identified down-
wind of Wood River and Central St. Louis.
-------
plant are estimated to have been in excess of 75 /Ltg/m3 . A sulfur
balance on the 14 August plume indicates that at least 75 percent of the
sulfur dioxide measured in the plume 10 km downwind of the plant was
accounted for in the measurements at 45 km downwind of the plant.
Based on this finding, the average sulfur dioxide removal rate by all
sinks on 14 August is estimated to have been well below 11 percent/
hour.
The flow rate of light-scattering aerosols within the 14 August plume
increased with distance downwind of the plant. This increase is
attributed to secondary aerosol formation in the plume. Most of the
new aerosol can be accounted for by sulfates and associated water. It
is estimated that the average SO2 - SO^ oxidation rate in the 14 August
plume was about 3 percent/hour. Some homogeneous nucleation
apparently occurred in the 1 August and 14 August plumes.
In all traverses of the Labadie plume, ozone concentrations encountered
within the plume were depressed below their levels on either side. The
depth of this ozone deficit and the ratio of NO to NOX in the plume both
decreased with distance downwind of the plant. A model is presented in
Section V which accurately predicts the observed plume profiles of .
ozone, NO, and NOX, from the background conditions, plume geometry,
and basic photolytic cycle for NO, NO2, and ozone. It is concluded
that free-radical reactions within the plume did not significantly affect
oxidant concentrations.
The St. Louis and Wood River Plumes
An important result of the aircraft sampling program was the identifi-
cation and documentation of a large, well-defined pollutant plume down-
wind of greater St. Louis. The geometry of the urban-industrial plume
strongly depended on the prevailing wind direction. Under westerly
or easterly conditions, parallel plumes from Wood River and St. Louis
could be distinguished in the traverses immediately downwind of the
metropolitan area. The combined initial width of the two plumes was
about 50 km. On 12 August the two plumes were apparently super-
imposed by southerly flow, although this cannot be established with
certainty because of the variability and disorganization of the measured
winds. The single narrow plume mapped north of Wood River on this
day contained the highest light-scattering coefficients (over 10 X 10~4
m""1) and ozone concentrations (up to 0. 20 ppm) measured by the air-
craft during the 1974 program.
The photochemistry of the St. Louis/Wood River plume was markedly
different from that of the Labadie plume, as net production of oxidant
-------
was observed in both the St. Louis and Wood River plumes less than
15 km downwind of hydrocarbon source areas. Indeed, ozone concen-
trations above background levels were the most conspicuous indicator
of the downwind St. Louis/Wood River plume, and it is calculated that
roughly 100 tons/hour of ozone was generated in this plume on the after-
noon of 30 July. This is five times the estimated hourly rate of hydro-
carbon emissions in the St. Louis/Wood River urban-industrial area
(averaged over the year), and indicates that the o2one_yjeld (?fhydro-
carbons in the atmosphere may be significantly greater than is measured
&*m*eM*^at**uiiliiSH**a*>*aa>l*ai!!ifa^^
in smog chambers. High oxidant concentrations in the St. Louis/Wood
^iveT?°]plume were generally associated with high light-scattering co-
efficients, suggesting that much of the light-scattering aerosol was
produced photoehemically.
-------
SECTION II
EXPERIMENTAL PROCEDURE
AIRCRAFT
Meteorology Research, Inc. (MRI) employed an instrumented Cessna
206 as its sampling platform during the 1974 MISTT field program. This
aircraft is shown in Figure 3. It is equipped with Robertson STOL modi-
fication to permit slower sampling speeds, which were particularly advan-
tageous in the Labadie power plant plume. An added belly pod increased
the space available for instrumentation.
INSTRUMENTATION
Instrumentation of the aircraft took place early in 1974. The basic cri-
teria used for instrument selection were low weight and bulk and low
power requirements, in addition to the required sensitivity and a fast
time response. The specific instruments used are listed in Table 2, to-
gether with their ranges and time responses. The instruments can be
divided into four groups: aerosol monitors, gas monitors, meteorological
sensors, and position sensors.
Aerosol Monitors
The MRI aircraft carried an integrating nephelometer, a condensation
nuclei counter, an electrical mobility analyzer, an optical particle counter
and a multifilter sulfate sampling manifold. The integrating nephelometer
and condensation nuclei monitor are continuous instruments, well suited
to the airborne sampling environment because of their ruggedness and fast
time response. The mobility analyzer and optical counter were operated
in an intermittent mode, analyzing bag samples obtained at specific points
during each sampling pass. The multifilter sulfate sampling system
collected integrated aerosol samples for later sulfate analysis at Wabhington
University.
The MRI integrating nephelometer measures the total light scattering co-
efficient of ambient air by detecting the light scattered from an illuminated
air sample, as described by Ahlquist and Charlson (1968). The light
scattering coefficient is roughly proportional to the mass concentration of
aerosol particles in the 0.2 pm. - 2.0 jum diameter range.
-------
Figure 3. Cessna 206 sampling aircraft. Photograph shows added belly pod, exhaust Venturis, and
sample inlets.
-------
Table 2. INSTRUMENTATION OF MRI AIRCRAFT
FOR 1974 MISTT FIELD PROGRAM
Instrument
MRI 1550B Integrating Nephelpmeter
Environment One Condensation Nuclei
Monitor (Rich 100)
Thermo Systems 3030 Electrical
Aerosol Size Analyzer
Royco 218 Optical Particle Counter
University of Minnesota Multichannel
Analyzer
Theta Sensor LS-400 SOS Monitor
REM 612 Ozone Monitor
Monitor Labs 8440 NO/NOX Monitor
Andros 7000 CO Monitor
MRI Airborne Instrument Package:
Temperature
Humidity
Turbulent Dissipation
Altitude
Indicated Airspeed
Metrodata M/8 VOR Analog Converter
Two Metrodata 620 Data Loggers,
20 channels each
Power
Required
70 W
70 W
«125W
«50 W
«50 W
7 W
60 W
400 W
125 W
10W
35 W
(each)
Ranges Used
0-10 (10-* m'1)
0-100* (10a cm'3)
0-300 (103 cm"3 )
0-1* (ppm)
0-10 (ppm)
0-0.5 (ppm)
0-0.5* (ppm)
0-1 (ppm)
0-2 (ppm)
0-5 (ppm)
O-SO(ppm)
-5 to +45 (°C)
0-100 (%)
0-10 (cma/a sec"1)
0-10,000 (ft)
50-150 (mph)
Response
Time (to 90%)
1 sec
5 sec
I A* 2 min
f per
I analysis
15 sec(a)
5 sec
5 sec
5 sec
5 sec
30 sec
3 sec (to 60%)
1 sec
1 sec
1 sec
48 channels /sec
(each)
* Normal operating range.
(a) Improved to an effective response time of 3 sec through transient response corrections
applied in data processing.
(b) Fast time response by special order.
-------
The Environment One condensation nuclei monitor described by Rich
(1970) is sensitive to particles less than 0. 1 micron in diameter, and
thus complements the nephelometer. The instrument humidifies an air
sample and expands it in a cloud chamber, causing condensation on the
particles in this size range. The resulting cloud attenuates a light beam
which is focused on a light-sensitive element.
The Thermo Systems, Inc. (TSI), electrical aerosol analyzer and the
Royco optical particle counter, shown in Figure 4, together measure
the particle size distribution of the ambient aerosol. The electrical
aerosol analyzer sizes the smaller particles by their electrical mo-
bility, and the optical counter sizes the larger particles by their scat-
tering cross section. These two instruments, together with their
accompanying multichannel analyzer, were installed and maintained by
the University of Minnesota personnel associated with the project. De-
tails of their operation have been given by Sem (1975) and Liu et al. (1973).
The mobility analyzer was operated in conjunction with.a grab bag sam-
pling system, shown in Figure 5 . The normal mode of operation con-
sisted of rapidly filling the bag (less than 3 sec) and then starting the
analysis of its contents. This analysis took nearly two minutes. The
bag would then be completely emptied, flushed (filled and emptied), and
the system readied for the next sampling. Normally, only one sample
was obtained during a particular traverse due to the cycle time required
by the system.
The optics and sampling probe for the optical particle counter were
mounted in a protective pod (Figure 6 ) attached to the wing of the air-
plane. Vacuum and electrical lines were strung through the wing to the
interior of the aircraft.
The components of the multifilter sulfate sampling system are shown in
Figures 4 (control manifold) and 7 (filter holders). Operation of the
system was controlled from the filter manifold by the use of manual
valves and a method of determining the flow. A single filter was nor-
mally exposed during all of the plume traverses at a given distance from
the source, as indicated in Appendix A. The exposed filter was later
analyzed for sulfate at Washington University by a sensitive flash vapor-
ization/flame photometric technique described by Husar, et al (1975).
Gas Monitors
The MRI aircraft carried monitors for sulfur dioxide, NO and NOX,
ozone, and carbon monoxide. These instruments are shown in Figure 8.
11
-------
Figure 4. Installation of electrical aerosol analyzer, Royco electronics,
multichannel analyzer, and multifilter sampling manifold in
MRI aircraft.
12
-------
Figure 5. Installation of grab bag sampling system in MRI aircraft.
Bag supplied sample to electrical aerosol analyzer.
13
-------
t
Figure 6. Wing pod housing optics for Royco optical particle counter. Sampling probe (not shown) was
inserted in center orifice.
-------
Ul
Figure 7. Installation of sulfate filter holders in MRI aircraft.
-------
D
Figure 8. Installation of gas monitors in MRI aircraft.
-------
The Theta Sensor SOg monitor utilizes an electrochemical cell to
measure the concentration of SOS , This instrument was the slowest
responding of the continuous monitors, with a response time to 90 per-
cent of 15 seconds. Through the incorporation of a transient response
correction subroutine into the data reduction software, as described by
Markowski (1975Vthe effective response time of this instrument was cut
to only 3 seconds (Figure 9 ). The small size and limited power
demands of the instrument are its principal advantages.
The REM ozone monitor is a chemiluminescent device that measures the
amount of light produced by the reaction of ozone and ethylene. A small
lecture bottle mounted behind the pilot's seat was the source of the
necessary ethylene.
The Monitor Labs NO/NOX monitor has dual reaction chambers for the
simultaneous chemilumine scent detection of NO and NOX (NOX = NO
+ NO 2 ). The concentration of NO is determined from the light produced
by its reactions with internally generated ozone. The NOX concentration
is measured by first passing the sample through a catalytic converter
which reduces NO2 to NO.
The Andros 7000 CO Monitor employs the dual isotope fluorescent source
nondispersive infrared absorption technique described by Link, et al (1971).
The principal advantage of the technique is the elimination of interference
by water vapor and carbon dioxide.
Meteorological Sensors
Ambient temperature, relative humidity, and turbulent dissipation were
measured in the aircraft as part of the MRI Airborne Instrument Package
(Figure 10). The ambient temperature outside the aircraft is measured
by a vortex temperature sensor using a thermister for the sensing element.
The relative humidity sensor uses a strain gage to measure the stress on
a cellulose fiber, the accuracy of the measurement being about ±5 percent
RH. Turbulence is measured by an MRI Universal Indicated Turbulence
System described by MacCready, et al (1965). The sensor is a standard
pitot-static probe connected to a Validyne P-24, 0. 7 psid differential
pressure transducer. The instrument measures pressure fluctuations in
the 2 to 40 Hz range and the signal is processed to give an output pro-
portional to the energy dissipation rate to the 1/3 power. The scale of
0 to 10 cms/ s sec'1 spans a range from calm air to what is considered
severe turbulence for a light plane. The output of the instrument is inde-
pendent of the speed of the aircraft for air speeds above 50 mph.
17
-------
1.00
0.90
0.80
0.70
0 60
E 0.50
0.
0.
<* 0.40
o
in
0.30
0.20
0.10
0.00
-o.io
-n 90
w. £ \J .
*
-
"
"
-
-
-
-
-
t
1
1
-.4-4
-
)
3-
P V
' ' '* " * " "k
,* * '--' -( *
-' ,
' *
*A*^i'
' ,A
r A" ,'
/
fT A '
'
A
A
j /
|B 4
, A
i A
A
A
A
4'
A
, j
4 8
TIM
- TRUE CONCENTRATION
A INDICATED CONCENTRATION
CORRECTED CONCENTRA-
TION USING 1st AND 2nd
DERIVATIVES
» CORRECTED CONCENTRA"
A TION USING 1st DERIVA-
% TIVE ONLY
A
A
, A
A
' '
i At
A
A
-
, ' ^A
1 ', X * -
t i "*
£m: _..-^--;ii,:L^-^«
'*&'
1 i i i
12 16 20 24
E, sec 7
76-131
10 second input pulse
Figure 9. Corrected transient response of Theta
Systems SO2 monitor (Markowski,
1975). Corrections were incorporated
in data reduction software to speed
effective time response of instrument.
18
-------
I
Figure 10. Probes for MRI airborne instrument package.
-------
Position Sensors
The position instrumentation in the aircraft utilized the standard aircraft
VOR and DME systems, which detect the radial and distance from a known
ground omnidirectional radio transmitter. The aircraft VOR and DME
were interfaced to a Metrodata M/8 package to convert their signals to
an analog output. Altitude was measured by a Validyne P-24 15 psi
absolute pressure sensor in the MRI Airborne Instrument Package, and
indicated airspeed was measured by a pressure transducer associated
with the turbulence probe (Figure 10).
Data Recording System
The output of the above instruments was recorded on two Metrodata 620
digital data loggers (Figure 11). The first data logger recorded twenty
channels of data, including the position information, all meteorological
parameters, all gas measurements, and the light scattering coefficient
and condensation nuclei count. The second data logger was dedicated to
recording the particle size distribution measurements of the electrical
aerosol analyzer/optical particle counter/multichannel analyzer system.
The outputs of the nephelometer, condensation nuclei monitor, and SO2
NO, NOX, and CO monitors were recorded on this second data logger
as well.
Both data loggers recorded at a rate of 48 channels per second, and inter-
nal digital clocks were simultaneously set for the two units. The clocks
provided time in hours, minutes, and seconds to each unit. A manually
operated Mode/Event switch (ten different modes with ON/OFF positions)
was used to mark events during the flight on one of the channels, and
another channel was labeled with the tape code as a backup identification
system.
Support System
All instruments except the multichannel analyzer, temperature, turbulence,
indicated airspeed, ^altitude, and relative humidity were powered by a 1 kw-
110 v inverter run off the aircraft power system, which was modified to a
24 volt system. The above exceptions were operated on available aircraft
DC power. The vacuum source for the sulfate sampler was provided by
two venturi mounted on the left wing of the aircraft. The vacuum source
for the bag system was provided by a third venturi mounted on the left
engine cowling.
20
-------
Figure 11. Installation of data loggers in MRI aircraft.
-------
CONFIGURATION OF SAMPLING SYSTEM
The sample inlet tubes shown in Figure 12 were mounted in a dummy
window panel on the left-hand side of the aircraft. Flight tests by MRI
and normally recorded data both indicate no interference between the
sample inlets and the engine exhaust located under the right engine
cowling. The effects of propwash and slipstream on the sample have
been reported by Adams and Koppe (1969) to be unimportant.
Ram pressure was used to move air through the inlet tubes during flight.
Since there were few flow restrictions, the velocity in the sample tube was
reasonably close to the airspeed of the aircraft. The gas monitors were
connected to the inlet manifold by short lengths of 1/4 inch OD Teflon
tubing. Static sampling ports in the manifold prevented changes in flight
speed from affecting the output of the instruments.
Figure 13 shows the arrangement of the instruments in the aircraft cabin.
The inverter, not shown in the figure, was installed in the belly pod. The
sample flow system is shown in Figure 14.
Figures 15 and 16 are schematic diagrams of the instrument wiring used for
the data loggers. The output of each instrument was directed through the
junction box to the input of the appropriate data logger. All instruments
were plugged into the junction box with banana plugs, thus facilitating any
necessary wiring changes.
FLIGHT AND SAMPLING PROCEDURES
The first flight of each sampling day was preceded by a 1-1/2 hour warm-
up period with the instruments connected to 110 v ground power. When
several flights were made during the day, the instruments were also run
on ground power between flights. A standard preflight procedure was used
to check out the instruments, the data acquisition system, and the sampling
system; the gas and aerosol monitors were zeroed and, where possible,
spanned electronically.
The sampling patterns flown during the 1974 MISTT field program are
summarized in Appendix A. Each flight consisted of spiral soundings,
which measured the vertical structure of the atmosphere above a given
point, and constant altitude traverses, which mapped the horizontal distri-
bution of pollutant. The flight plans were designed to characterize a given
plume at discrete distances downwind from its source.
Spirals were flown at an ascent or descent rate of 500 feet per minute.
This rate was fairly critical; it was slow enough to allow the instruments to
22
-------
N
00
Figure 12. Sample intake manifolds on left-hand side of aircraft.
-------
SO Monitor
Tape Code Box
Data Logger
Ozone Monitor
Nephelometer
CO Monitor
M8
Sulfate Filter
Sampler
NO/NOx
Monitor
Condensation
Nuclei Monitor
Bag
Sampling System
Nephelometer
Tube
Multichannel
Analyzer
Electrical
Aerosol
Analyzer
Multifilter Sulfate
Sampling Manifold
Optical Particle
Counter
Observer Seat
Nephelometer
Blower Box
Airborne Instrument Package
Temperature
Humidity
Turbulence
Altitude
Indicated Air Speed
6-112
Figure 13. Instrument layout in MRI aircraft.
24
-------
External
Probes for
Temperature,
Humidity,
Turbulence,
and Indicated
Airspeed
Venturi
Exhausts
Roy co
Optics
and
Sampling
Probe
Inlet Tube
Nephelometer
1-3/4 inch O. D. Tube
Condensa-
Static Probes
tion Nuclei
Inlet Tube
1 /2 inch
O. D. Tube
Monitor
Cartridges
i Monitor f
|NO/NOX
Monitor f~
1 Monitor [
Electrical
Aerosol
Analyzer
(Wing-mounted)
Venturi
Exhaust
76-113
Figure 14. Sample flow in MRI aircraft.
25
-------
Junction Box
Cx)
Nephel ometer
Nephe 1 on
Range S»
Cohdensal i on
b 1 i j
Nuc 1 e i
Monl tor
Condensat i<
Nuclei Mon
Range Swit
CO
Monitor,
°3
Monitor
NO Monitor
NOX Monitor
S02
Monitor
0-5 V
(Range Selection)
leter +2.+4.+6V
"Tcn (|0 40.100) x 10*"^
Q-2Y . ^
(Range Selection)
:>n
tor r_. 0.1,2,3,4,5.6 V^
ch (1.3,10,30,100, ^
300,IOk)x|03
0-5V
-50 to + 50°C
0-5V
0-.5 ppm """
0-5V
0-.5 ppm
0-5V
0"" ^ t\ f\ tfi
0-5V
0-1 ppm
10
1 1
i > 7
12 '
8
13 C Q
' H 9
A
,4 N '»
1 4 "
L
S
1 5 20
If. \
O J
1 7 j
i / 4
i q
0-5V
^t , . - * -
-5-45°C
0-5V
^ 0-100?
^ 0-5V
0-10,000 ft
^ 0-5V
""* 50-150 mph
0-5V
0-10
0-5V
^ 0-360°
_ 0-4V
0-100 Mi les
0-1 0V
0-1 0V
_ Digital
Logger
A i rborne
nstrument
Package
Temperature
Relative
Hum! d i ty
A titude
Indicated Air
Speed
Turbu 1 ence
ONE
Mode/Event
"""" Switcn
Figure 15. Instrument schematic for first data logger.
-------
Junction Box
Nephelometer
Multichanne 1
Analyzer Flag
Condensation
Nuclei Monitor
NO Monitor
NOX Monitor
Electrical Aerosol
Analyzer Flag
»
10 5
11 6
C 8
12 H
A
N 9
N
16 E
* 3
17
4
19
«» -
, f^
- -»
Oj Monitor
Altitude
Multichannel Analyzer
Digital - Analog
Converter #1
Multichannel Analyzer
Digital - Analog
Converter #2
Electrical Aerosol
Analyzer Current
Mode /Event Switch
Digital
Data
Logger
76-401
Figure 16. Instrument schematic for second data logger.
-------
respond, but slower ascent/descent rates had to be avoided because of the
possibility of sampling aircraft exhaust from the previous turn of the spiral.
The radius of a spiral was typically 1/2 km. The bottom of the spiral was
usually as low as safety permitted, and the top was determined by meteor-
ological conditions. The event switch was used to mark the spiral on the
data tape, and the observer recorded the beginning and ending times and
altitudes (from the aircraft altimeter) of the spiral (see Figure 17).
For each set of horizontal traverses, recognizable ground landmarks were
chosen as end points in such a way that the traverse path was nearly per-
pendicular to the wind direction. The aircraft then made several flights
at different altitudes along the traverse path until the plume had been well
characterized at that downwind distance. New end points were then chosen
further downwind, and the procedure repeated (see Figure 18). The air-
speed was usually as slow as possible to allow for instruement response.
Each traverse was indicated on the magnetic tape by the event switch; for
additional identification, the times and altitudes of each traverse were
manually recorded by the observer on flight records.
CALIBRATION
University of Minnesota personnel performed calibrations of the electrical
aerosol analyzer, optical particle counter, and multichannel analyzer
before and after the St. Louis sampling program. In addition, two in-field
calibrations were performed during the sampling period.
Calibration of the other aircraft sensors were performed both before and
after the program. During the study period, the gas monitors were cali-
brated twice by Mr. R. Rhodes (QAS/QAEML) of the EPA as a part of
the "RAPS Cross-Comparison Program for Gaseous Pollutant Measure-
ments. " Additional field audits of the gas monitors were performed by
both EPA and MRI personnel using a commercially available (Monitor Labs
8500) calibrator.
DATA PROCESSING
All data from the aircraft were processed by Washington University, as
described by Gillani (1975).
PILOT BALLOON OBSERVATIONS
Observations of wind aloft were made hourly at up to three different loca-
tions during each intensive sampling period. Pilot balloons with standard
ascent rates were tracked to 1140 m agl by theodolite, and the winds at
110 m altitude intervals calculated by triangulation. Data collection and
reduction was performed by MRI personnel under a separate contract.
28
-------
76-132
Event on, time, altitude, &
location noted.
Event off, time, and altitude
noted.
Figure 17. Flight plan for spiral soundings.
29
-------
0 Event on, time, altitude,
and location noted.
Event off, time & focatfon
noted.
76-133
Figure 18. Flight plan for horizontal traverses. Points A,
B, C, and D are ground reference points selected
during flight.
30
-------
SECTION in
METEOROLOGY
INTRODUCTION
During the period from 30 July 1974 through 15 August 1974, two types
of synoptic weather regimes affected the St. Louis sampling area.
The more prevalent regime was one characterized by moderate to
weak anticyclonic surface circulation and weak or disorganized flow
patterns aloft. These anticyclonic periods alternated with transition
regimes of general cyclonic surface circulation, more clearly defined
upper air patterns, and unsettled local weather. The chronology of the
two types of regimes is outlined in Figure 19. The following paragraphs
describe the synoptic patterns in each period and relate them to the
local weather in the sampling area.
30 JULY 1974 - 1 AUGUST 1974 (ANTICYCLONIC)
After the passage of a midlatitude cyclone across the central Mississippi
Valley on 28 July, a large dry continental polar air mass moved into
the Great Plains Region during 29 and 30 July. By the morning of 31
July, the air mass had spread out and modified into a large flat warm
high pressure area occupying the entire eastern half of the country
(Figure 20). At the same time, a low center to the north had moved
into eastern Canada, leaving very weak westerly flow aloft south of
40° N latitude. As a result, the St. Louis sampling area experienced
rather stagnant weather conditions through 1 August, with large and
scattered low clouds.
High temperature readings at Lambert Field in St. Louis increased
from +30° C to +33° C during this period, while dew points advanced
from +13° C to 16° C. The increase in temperature and moisture
corresponded directly with the onset of light south-to-southeasterly
anticyclonic flow around the surface high pressure center in the
eastern part of the country. Aloft, a pronounced subsidence inversion
capped the surface mixing layer at around 2130 m (7000 ft) msl during
most of the period, inhibiting convective activity and dispersion above
the lowest layer of the atmosphere. On 1 August, however, the
establishment of neutral to unstable temperature lapse rates throughout
the lower 6400 m (20,000 ft) of the atmosphere signaled the beginning
of a cyclonic transitional period. By midafternoon on 1 August, a
broken ceiling of towering cumulus clouds dominated the St. Louis area.
31
-------
A
A
A
20
o
o
LiT
or
Z>
I-
<
a:
LJ
Q_
UJ
I-
0
A
\
/ SKY COVER
\
A
,'\
\
«
V
o
n
n
n
1.0
en
UJ
>
o
o
0-5
CO
30
Figure 19.
31 I 234 567 8 9 10 II 12 13 14 15 76_12l
JULY DATE AUGUST
St. Louis weather during sampling period. Curves show daily temperature (24-hour
average), dew point (24-hour average), and sky cover (sunrise to sunset average) at
Lambert Field, St. Louis (U.S. Department of Commerce, 1974). Type of synoptic
regime (C - cyclonic, A - anticyclonic) affecting sampling area is shown at upper
edge of graph. Temperature and dew point tended to increase during passage of
anticyclones. Sampling days are indicated at bottom of graph.
o
-------
w
LEGEND
H . L Surface pressure centers
'y "'^ Surface frimts
_ Surface isobars (every 4 mb)
Figure 20. Weather map for 0700 CDT on 1 August 1974. St. Louis sampling area
had been experiencing rather stagnant weather conditions within the
large flat high pressure area occupying the eastern half of the country.
-------
2 AUGUST 1974 - 4 AUGUST 1974 (CYCLONIC AND POSTCYCLONIC)
By the morning of 2 August, a surface low pressure system with a
trailing cold front had moved into northcentral Missouri. At the
same time, a cool, dry continental polar air mass was moving out
of southwestern Canada along the eastern lee of the Rockies. By the
time this cool, dry air arrived in the central Great Plains late on
3 August, the surface low had moved into northeastern Wisconsin,
while the trailing cold front had pushed eastward into Ohio, Kentucky,
and Tennessee. By the morning of 4 August, the center of the high
pressure ridge was in Nebraska and Kansas. However, the onset of
stable anticyclonic flow did not take place in the St. Louis sampling
area until 5 August, because of the slow movement of the surface
low in Wisconsin. As a result, conditions favorable to mixing and
pollutant dispersion persisted in the sampling area through 4 August,
even though the cloudiness and inclement weather associated with the
cold front had ended on 3 August.
Dew point and temperature readings at Lambert Field reached their
highs during the afternoon of 2 August, when the cold front was still
to the west. The lowest absolute moisture values and temperatures
were recorded on 4 August, when the cool, dry anticyclone was
approaching St. Louis from the west. Winds were southwesterly
before frontal passage, then northwesterly to northerly through 4
August.
5 AUGUST 1974 - 10 AUGUST 1974 (ANTICYCLONIC)
On the morning of 5 August, the dry continental air mass which had
moved down from Canada began to modify and spread out across the
eastern half of the country. The center of this air mass was marked
by a high pressure ridge located in northern Missouri and southern
Iowa (Figure 21). Aloft, the trough of low pressure which had been
associated with the surface cyclonic system of 2 and 3 August had
moved into northern Labrador, leaving light westerly winds in the
midwest and eastern sections of the country. By 6 August, the
surface anticyclone had modified into a moist, warm air mass, cen-
tered in Pennsylvania but extending westward to the Western Plains
states. During 7-9 August, this air mass continued to warm, while
additional moisture circulated into the central and eastern states
from the Gulf of Mexico. The center of the high pressure ridge
during that time drifted slowly from central Pennsylvania to Cape
Hatteras, finally dissipating over western South Carolina on 9 August.
34
-------
Ul
H,L
LEGEND
. Surface pressure centers
, Surface fronts
. Surface isobars (every 4 mb)
Figure 21. Weather map for 0700 CDT on 5 August 1974. St. Louis sampling area
experienced increasing temperature and humidity over next few days
as anticyclonic circulation brought in Gulf air.
-------
At that time, a new maritime polar air mass drifted southward from
Quebec, extending into the entire eastern half of the United States by
the morning of 10 August. Despite the change in high pressure centers,
on 9 August light, moist easterly to southeasterly air flow persisted
over the Mississippi and Ohio Valleys throughout the period. In the
upper atmosphere, weak ridging prevented the development of well-
established system path trajectories on the surface.
Skies during the entire five day period were generally cloudy with
early morning fog and hazy midday conditions. Some light rain
occurred on 10 August due to the development of a cyclonic system
to the west, but most of the clouds which persisted during the period
lacked enough vertical development to produce rain. Typical temper-
ature soundings during the period showed a neutral or slightly unstable
mixing layer below 2438 m (8000 ft) msl capped by a more stable layer
above that level. Only on 10 August did the lapse rate show unstable
characteristics above the elevated stable layer. Surface temperatures
showed daily increases through 8 August, with the daily maximum
rising from +26° C on 5 August to +31° C on 8 August. Increasing
cloudiness reduced maximum temperatures slightly on 9 and 10
August. Dew points also rose throughout the period, increasing from
a low of +14° C on 5 August to +22° C on 9 August. Since the low
level winds recorded after 5 August were consistently east to south-
easterly, there can be no doubt that the St. Louis sampling area was
receiving an extended exposure to advected maritime tropical air
circulating around the weak but persistent anticyclone to the east. It
wasn't until 10 August that this pattern began to change.
11 AUGUST 1974 - 12 AUGUST 1974 (CYCLONIC)
The broad ridge of high pressure aloft over the eastern half of the
continent finally moved to the Atlantic Coast on the morning of 11
August. Behind it, a rather shallow long wave trough aloft began
to sink southeastward into the Central Plains. This triggered the
formation of a weak surface low and trailing cold front. Because of
the weak and short-lived nature of the trough aloft, the surface cyclone
became rather poorly defined by the morning of 12 August, especially
in its southern half (Figure 22). As the surface low moved northward
out of the United States into Canada, it left only a series of weak,
quasi-stationary fronts in the middle of the country. As flow aloft
south of 45° N latitude continued to dissipate, a rather flat, stagnant
synoptic weather pattern began to reestablish itself late on 12 August.
36
-------
LEGEND '
H , L Surface pressure centers
'Y ** Surface frbnts
. Surface isobars (every 4 mb)
Figure 22. Weather map for 0700 CDT on 12 August 1974. St. Louis sampling
area experienced scattered thunderstorm activity from 10 August
to 13 August.
-------
Skies were broken to overcast during most of 11 August, with some
scattered thunderstorms in the St. Louis sampling area. Improved
sky conditions existed during the morning of 12 August, but towering
cumulus clouds and thundershowers developed again during the after-
noon. Temperature soundings during both days were relatively un-
stable above 2740 m (9000 ft) msl. Temperatures and dew points
were consistent with those recorded on 10 August, but wind directions
at the surface and aloft had become south to southwesterly. Local
readings of all meteorological parameters were inconsistent and
variable close to the persistent thunderstorm activity during the two
days.
13 AUGUST 1974 - 15 AUGUST 1974 (ANTICYCLONIC)
During this three-day period, a weak, disorganized flow pattern once
again prevailed aloft south of 40° N latitude. In that region, a flat,
featureless, synoptic surface pattern produced stagnant, motionless,
warm, humid weather. At the same time, relatively strong west to
east air movement existed to the north. As a result, a weak surface
high pressure ridge slid across the continent along the Canadian
border, bringing modified continental air and anticyclonic circulation
to the region north of 40° N latitude. The St. Louis sampling area
was situated at the transition between the two air masses. By morn-
ing of 15 August, the northern ridge was centered in Quebec with the
anticyclonic circulation extending southward as far as North Carolina,
Tennessee, and southern Missouri (Figure 23). Flow aloft continued
very weak and disorganized throughout the eastern half of the country.
Skies in the St. Louis sampling area were cloudy with scattered
thunderstorm activity early on 13 August, but conditions began to
improve during the afternoon and early evening. By the morning of
14 August, skies had become nearly clear, with fog and haze near
the surface. Sunny, hazy conditions prevailed until 15 August, with
night and early morning fog continuing. A broken layer of high level
clouds, along with some cumulus development, appeared during the
afternoon of 15 August as a new cyclonic system approached St. Lovr.s
from the west. Temperature soundings throughout the period were
stable to neutral above 910 m (3000 ft) msl. High temperatures at
Lambert Field ranged from +27° C on 13 August to +31° C on 14
and 15 August. Dew point readings near +20° C persisted throughout
the three days. Wind directions remained east to southeasterly due
to the anticyclonic circulation around the high to the east.
38
-------
LEGEND :
H , L Surface pressure centers
jr~-^~ Surface fronts
- Surface isobars (every 4 mb)
Figure 23. Weather map for 0700 CDT on 15 August 1974. St. Louis sampling
area was situated near transition between two air masses.
-------
SECTION IV
SULFUR CHEMISTRY
INTRODUCTION
The combustion of fossil fuels releases SO3 to the atmosphere,
where it then oxidizes to form sulfate aerosols. These sulfate
aerosols have been implicated as causative agents for many of the
adverse effects of air pollution. Further, particulate sulfur appears
to persist in the troposphere for much longer periods than gas-phase
sulfur does. It is thus of great importance to be able to predict the
effects, on ambient sulfate aerosol concentrations, of projected in-
creases in national SO2 emissions.
In the neighborhood of St. Louis, a coal-fired steam generating plant
at Labadie (Mo. ) and a group of refineries at Wood River (HI. ) are
major sources of SO2. On several days during the 1974 MISTT field
program, the MRI aircraft made detailed measurements of the SO2
concentration, particulate sulfur concentration, and aerosol light-
scattering coefficient downwind of these two source areas. In this
chapter, the data from these flights are discussed. Particular atten-
tion will be paid to the rates at which SO2 was oxidized to SO4 and
otherwise removed from the atmosphere, as these rates are critical
in determining the relationship between SOS emissions and ambient
sulfate concentrations.
AN SO 2 BALANCE FOR THE LABADIE PLUME
An unusually complete record of the Labadie plume was obtained
during the middle of the day on 14 August 1974. Meteorological con-
ditions at this time were well suited to the characterization of pollu-
tant flow, with a steady easterly flow of air throughout the lowest
kilometer of the atmosphere (Figure 24). During the period from
1100 to 1500 CDT, the MRI aircraft flew a total of 19 sampling passes,
mapping the plume in considerable detail out to a distance of 45 km
from the plant (Table 3 , Figure 25). Figures 26 - 28 show the
profiles of SO2 and bscat measured on representative horizontal
traverses of the plume.
The rate at which SO2 was lost from the plume through all mecha-
nisms can be estimated by comparing the mass flow rates of SO2 at
increasing distances downwind of the plant. These flow rates, which
are not affected by cross-plume and vertical dispersion, are
40
-------
1000
E
E
500
3.6
3.1
3.5 -"
5 c*
4.6
4.1
3.5
2.0
2.7
2.8
2.9
2.0 1.7
4.7
4.9
4.0
5,5
4.0
2.9-
3.0
4'8
3.1
3.8
3.2X
5.0
4.8
3.0
2.9
2.7
4.5
4.5
4.5.
4.5,
4«°
0900 1000 1100 1200 1300
TIME, CDT
1400
76-049
Figure 24. Winds aloft on 14 August 1974, from pilot balloon measurements at Marthasville,
Missouri (see Figure Z5 for location).
-------
Table 3. AIRCRAFT SAMPLING SCHEDULE,
14 AUGUST 1974
Pass No.
Spiral
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Spiral
Route
(See Figure 25)
Labadie
A -
B -
A -
B -
E -
F -
E -
F -
H -
J -
K -
L -
M -
N -
O -
N -
P -
RohlfLng
B
A
B
A
F
E
F
G
J
H
L
K
N
O
N
P
N
Airport
Start
Time
(CDT)
1045
1129
1137
1147
1157
1210
1218
1225
1233
1251
1302
1316
1326
1351
1407
1416
1425
1438
1501
Stop
Time
(CDT)
1121
1135
1144
1154
1204
1215
1223
1231
1243
1259
1310
1323
1332
1405
1414
1423
1436
1451
1518
Altitude Filter
(m msl) Number
3505-153
457
610
762
914
914
762
610
457
457
610
762
914
1220
914
762
610
457
1829-153
1
1
1
1
2
2
2
2
3
3
3
3
4
4
4
4
4
42
-------
04
Figure 25. Location of aircraft sampling routes and pilot balloon release points (underlined)
14 August 1974.
-------
E
ex
Q.
CM
O
0.5
0.4
0-3
0.2
O.I
SO
2 .'
1142
TIME ,CDT
o
CO
.a
76-037
140
Figure
10 km
26. Horizontal plume traverse 10 km downwind of Labadie stacks
on 14 August 1974. The SO2 concentration and light- scattering
coefficient (bscat) were usually well correlated in the Labadie
plume, as here. The altitude of this traverse was 610 m
(2000 ft) msl.
44
-------
Q.
CL
*
( ! ! !
1240 1238 1236
TIME, CDT
I 1
10 km
Figure 27. Horizontal plume traverse 21 km downwind of Labadie stacks
on 14 August 1974. The elevated light-scattering coefficients
outside the SO 2 plume represent the contributions of back-
ground aerosols. This traverse was made at 455 m (1500 ft) msl.
45
-------
0.5
0.4
E
o.
Q.
O
CO
0.2
O.I
'SCAT
SO2
4_
i
3 O
2 <
o
CO
I <
1436
» ( ! ! .1 '
1434 1432 1430
TIME, CDT
76-036
! ! !
1428 1426
10 km
Figure 28. Horizontal plume traverse 45 km downwind of Labadie
stacks on 14 August 1974. Compared to traverses
nearer the stacks, plume is broader, peak SO 2 con-
centrations are lower, and ratio of excess bsca(- to
SO 3 in plume is greater. The altitude of this traverse
was 610 m (2000 ft) msl.
46
-------
calculated from the following integral:
F (x) = ux (x) lie (x, y, z) dy dz,
where
F = the mass flow rate of SOS
x = distance downwind of the plant,
y, z = cross-plume arid vertical coordinates, extending
beyond plume boundaries,
i^ = wind speed, and
c = the concentration of SO2.
Figure 29 displays the inner integrals
I c (x, y, z). dy
calculated from the individual plume traverses. These values, together
with an average wind speed of 4 m/sec, yield the following flow rates
for SO3 when substituted in the above formula:
Distance from plant (km) 10 21 32 45
Flow rate of SO2 within
455 m - 915 m layer (kg/sec) 2.0 1.9 1.7 1.5
These flow rates take into account only the SOS within the horizontal
layer between 455 m (1500 ft) and 915 m (3000 ft) msl, where most
plume traverses were made. Over the 35 km distance spanned by the
plume measurements, the flow rate of SO2 within the layer declined
by 25 percent, or about 11 percent/hour.
The above estimate for the rate of SO3 loss from the sampling layer
of course represents not only the SOa removed from the atmosphere
by deposition and oxidation but also the SO2 which simply diffused
out of the layer into the air above and below. The changing vertical
gradients in Figure 29 make it clear that a substantial fraction of the
decrease in SO2 flow was in fact due to diffusion out of the layer.
The vertical profiles of temperature and turbulent dissipation support
this conclusion, indicating vigorous mixing through the lowest kilo-
meter of the atmosphere (Figure 30). The value of 11 percent/hour
47
-------
SAMPLING
TIME (CDT)
1129-1205
1210-1243
1251-1332
1351-1451
430
1,000
CO
E
oo
500
1,360
1,280
1,290
^ _ 1000
1,080
1,240
990
990
1,180
I
///////////////////// / / / / / // / /// // III III II I I I I
JL.
J 76-041
10
20 30
DISTANCE FROM STACKS,km
40
50
Figure 29.
Composite side view of LAbadie plume on 14 August 1974. Numbers give cross-wind integral,
f [SOS] dy» of SO 3 concentration measured on each traverse. Relatively slow decline in SOs
integrals with increasing distance from the stacks indicates that broadening of the plume caused
most of the decrease in peak SO s concentrations seen in Figures 26, 27, and 28.
^
1
-------
E
E
LJ
Q
ID
-I
<
<
£
/*
»
£.
^..^ TURBULENCE
TEMPERATURE
10 20 30
TEMPERATURE. °C
TURBULENT DISSIPATION , cm2/3 sec
40 76-035
fo
Figure 30. Vertical profile from spiral descent at Labadie
on 14 August 1974, 1045-1121 CDT. Good mixing
is indicated in the lowest kilometer of the atmosphere.
49
-------
must therefore be understood as a very conservative (i. e., high) upper
bound for the rate at which SO2 was lost from the plume. This bound
is consistent with the results of English workers on deposition velocity
(Owens and Powell, 1974, Shepherd, 1974). Their estimates of about
1 cm/sec for the dry deposition velocity of SO2 correspond to a re-
moval rate of 3-4 percent/hour for a plume well mixed through the low-
est kilometer of the atmosphere. Together with an SO2 - 804 con-
version rate of 3 percent/hour, this would give a combined SOS loss
rate of about 6 percent/hour.
PARTICULATE SULFUR IN THE LABADIE AND WOOD RIVER PLUMES
Table 4 shows the particulate sulfur concentrations, SOS concentra-
tions, and light-scattering coefficients measured downwind of the Labadie
power plant (5 August, 14 August) and the Wood River industrial area
(12 August 15 August) during the 1974 program. In order to facilitate
the comparison of continuous measurements with filter samples, the SO3
concentrations and the light-scattering coefficients have been averaged
numerically over the sampling periods of the filters.
Much of the aerosol collected on each sampling pass came from sources
other than the one under study. Each particulate sulfur concentration in
Table 4 thus represents an average of particulate sulfur concentrations
within the plume and particulate sulfur concentrations in the background
air. The contribution of background aerosols to the light-scattering co-
efficient is evident in Figures 26 - 28, which show bscat levels of
1.8 - 2.5 X 10~4 m~x outside the SO3 plume from the Labadie plant.
To aid in estimating the relative contributions of plume and background
to the particulate sulfur concentration, each average light-scattering
coefficient in Table 4 has been divided into the portion contributed by
the plume under study and the portion contributed by the background.
These divisions were based on regression analysis of the bscat and
SOS profiles, as described in the appendix at the end of this section.
Figures 31 and 32 show the particulate sulfur concentration plotted
against background bscaj. and plume bscat. It is clear from these
figures that the particulate sulfur concentration is more sensitive to
changes in plume bscaj. than it is to changes in background b fc. If
it is assumed that bscaj- is proportional to the submicron aerosol mass
concentration, an interpretation which is supported by the uniformity of
measured particle size distributions (Whitby et al., 1976), then it can be
inferred that the plume aerosols contain higher fractions of sulfur than
the background aerosols do.
50
-------
Table 4. PARTICULATE SULFUR IN THE LABADIE
AND WOOD RIVER PLUMES
(a)
(a)
(a)
Background Plume
Filter
8/05
8/12
8/14
8/15
(a)
1
2
3
4
1
2
3
1
2
3
4
1
2
Description
Labadie (1 km)
Labadie (4 km)
Labadie (18 km)
Wood River (12 km)
Wood River (25 km)
Wood River (50 km)
Labadie (10 km)
Labadie (21 km)
Labadie (32 km)
Labadie (45 km)
Wood River (10 km)
Wood River (28 km)
Average over filter sample
Dscat
1.54
1.45
2.06
1.83
1. 60
1.64
1.15
2.01
1.96
2;13
2.52
2.64
2.50
period.
bscat
(lO-'m-1)
0
0.43
0.14
0.14
1.41
1.58
0.68
0.27
0.35
0.33
0.39
0.51
0.86
7m3
2.52
6.09
5.53
4.33
7.45
8.63
3.68
6.20
5.53
5.48
6.10
5.73
4.96
(a,b)
Plume
S
0
3.30
1.18
1.01
5.81
6.85
2.59
2.17
2.30
2.10
2.33
2.50
2.87
(a)
SO a
(ppm)
0
0.171
0.060
0.031
0.055
0.050
0.023
0.042
0.040
0.026
0.014
0.044
0.036
(c)
Plume
spsT
0.015
0.015
0.024
0.075
0.094
0.079
0.038
0.042
0.059
0.114
0.042
0.058
/ 4 plume bscat \
(c)
Plume
*" ^background bscat + 4 plume bscat / *"
= plume Sp/(plume Sp + Sg); Sg = SOa as sulfur, averaged over filter sample period.
-------
3.0
2.5
o
(/)
XI 2.0
or
e>
*:
o
<
CD
I .5
1.0
0.5
0
O Sp (jjg m3) on 8/14/74
-f- Sp (pg m3) on 8/05/74
O 6-1
05.5
0-1
0-2
0-3
0-4 0-5
PLUME bSCAT,ICT4 m'1
76-042
Figure 31. Scatter plot of particulate sulfur concentration (Sp) vis.
plume b and background b (note different scales)
SCctE S CcLu
for Labadie samples. Sloping lines indicate best-fit
linear relationship: Sp = 7. 18 plume b +1.77 back-
ground b
scat
52
-------
3.0
E
T
b
2.0
o
w
.£>
Q
:z5
o
o:
o
<
CO
1.5
1.0
0.5
O
'0
on 8/12/74
on 8/15/74
0-5
PLUME b
SCAT
.0
ICT4m-
1.5 76-043
Figure 32. Scatter plot of particulate sulfur concentration (S ) vs. plume
b and background b , (note different scales ; for Wood
scat & scat
River samples. Sloping lines indicate best-fit linear relation-
ship: S =4.04 plume b . + 1.04 background b
r p r scat scat
53
-------
If the sulfur fraction of the plume aerosol and the sulfur fraction of the
background aerosol are each fairly constant, then the particulate sulfur
concentration is described by the following approximate relationship:
S = A plume bscat 4- B background bscat .
The sloping lines in Figures 31 and 32 are the contours of constant S
for the best-fit relationships of this form:
Labadie (5 August, 14 August)
Sp (jUg/m3) = 7. 18 x Plume bscat (KT m" )
+ * 77 do--?7-! ) background bscat (lO"4 m"1 ) (1)
Urban-Industrial (12 August, 15 August)
Sp(Mg/m3) = 4.04 (^-4^-1) plume bscat (ID"4 m'1)
+ 1.04 -i background bgcat (10 ~4 m'1) (2)
According to formula 1, the particulate sulfur concentration reached
26jxg/m3, equivalent to 80 /jg/m3 of H3SO4, during the 14 August
1974 traverse 10 km downwind of Labadie shown in Figure 26. By
the same reckoning, peak particulate sulfur concentrations in excess
of 50 fjig/m.3 , equivalent to more than 150 jug/m3 of sulfate, were
encountered during the traverses 1 km from the plant on 5 August 1974.
Common to the fits of both data sets is the feature that plume bsca£
is weighted about four times as heavily as background bscat in deter-
mining the particulate sulfur concentrations. Accordingly, the portion
of Sp contributed by either type of plume is estimated in Table 4 by
the formula:
4 plume b
scat
4 plume bscat + background b t
54
-------
In the last column of Table 4, this is expressed as a fraction of ST, the
total (gas and particulate) sulfur concentration in the plume. The
calculated fraction of plume sulfur in the particulate phase ranges from
1. 5 percent in the Labadie plume 1 km downwind of the stacks, which is
about what would be expected in the absence of significant atmospheric
conversion, to 11.4 percent in the Labadie plume 45 km downwind of
the stacks.
SULFUR CONVERSION IN THE LABADIE PLUME
In the Labadie plume of 14 August 1974, the fraction of plume sulfur in
the particulate phase grew with distance downwind of the stacks (Table 4,
Figure 33. Over the 35 km interval spanned by sampling, the in-
crease in plume Sp/ST was over 200 percent. According to the esti-
mates made from the sulfur balance earlier in this section, less than
one-sixth of this increase can be attributed to loss of sulfur from the
plume. The remainder of the growth in the particulate sulfur fraction
was presumably due to the oxidation of SO2 to SO 4 . The average
pseudo-first-order rate constant required to produce this degree of
conversion would be about 3 percent/hour, although it is clear from
Figure 33 that the rate is not constant,
A second indicator of the rate of aerosol production in the 14 August
1974 plume was the changing ratio of light-scattering aerosols to SO3.
The ratio plume bscat/SO2 shown in Figure 33 is calculated from the
continuous bscat and SO3 record by regression analysis as described
in the appendix at the end of this section. It is computationally in-
dependent of the ratio plume Sp/S-p which is calculated from filter
measurements and the background and average light-scattering coef-
ficient and SOS concentration. Figure 33 shows that plume bscat/SO2
displayed a pattern of downwind growth similar to that of plume Sp/S
on 14 August 1974. Since loss of SO3 from the plume cannot account
for more than a small fraction of the change, the increase in plume
bscat/SO2 confirms that light-scatter ing aerosols were being generated
in the plume.
Some insight into the composition of the Labadie plume aerosol can be
gained by examining the ratios Sp/bscat and bscat/V, where V (^m3 /
cm3 ) is aerosol volume concentration. In the preceding subsection,
the ratio Sp/bscat was estimated to be about 7 in the Labadie plume,
based on the observed variation of Sp with plume bscat and background
bscat (Formula 1 ). The ratio bscat/V can be calculated theoretically
from the average particle size distribution measured in the plume by the
University of Minnesota on 14 August 1974 (Whitby et al, 1976); based on
Mie theory for spherical particles of refractive index 1. 5 (White et al,
1975), this ratio was about 0.043.
55
-------
ESTIMATED TRAVEL TIME,hrs
25
20
E
Q.
OL
\ .,
o
v
o*
V)
-Q
10
Sp/S
to
10 20 30 40
DISTANCE FROM STACKS, km
50
76-048
Figure 33. Two aerosol/emissions ratios in the Labadie plume on
14 August 1974. The downwind increase in both ratios
reflects growth of aerosol within the plume. Lower
scale shows actual distance downwind of stacks, upper
scale shows equivalent travel time for average wind of
4 m/sec. .
56
-------
a
The product of the two ratios, (Sp/bscat) (bscat/V) + S_/V, is _
quantity which can be calculated directly for liquid particles of known
composition. The empirical ratio, Sp/V = 7 X 0.043 = 0.30, falls
between the theoretical ratios for pure sulfuric acid (S^/V = 0.60)
and sulfuric acid in equilibrium with water vapor at the ambient
humidity (Sp/V = 0.24) (Whitby et al, 1976). This observation indi-
cates that the observed light-scattering coefficient in the Labadie
plume can be completely accounted for by sulfate aerosol and associ-
ated water.
A clue to the mechanism of aerosol production in the Labadie plume
is contained in a CN profile recorded 21 km downwind of the stacks
on 14 August 1974 (Figure 34). The sharp peaks in the nuclei count
encountered at the edges of the plume on this traverse indicate that
homogeneous nucleation probably was occurring there. Well-defined
CN peaks of this magnitude were not recorded on other traverses,
although sampling in the Labadie plume on 1 August 1974 showed CN
counts which increased with distance downwind of the stacks. The
occurrence of homogeneous nucleation implies that material of very
low vapor pressure, presumably a sulfate compound, was being pro-
duced by reactions in the gas phase. These gas phase reactions may
or may not have been the dominant mechanism for the conversion of
SO 3 to SO £ throughout the Labadie plume. Calculations for such a
gas phase system, based on the average SOS oxidation rate estimated
above, the average aerosol size distribution measured in the plume
(Whitby et al, 1976), and Fuch's expression for diffusional transport
to a sphere (Husar, 1971), indicate that the average pseudo-steady-
state concentration of gas phase sulfate compounds in the 14 August
1974 Labadie plume would have been on the order of 6 X 10~6 ppm.
This average concentration is roughly one-twentieth of the concen-
tration of H2SO4 calculated to be necessary for homogeneous nucle-
ation in a pure H2SO4 - H2O system at 75 percent relative humidity
(Mirabel and Katz, 1974). The CN peaks of Figure 34 could be due
to locally high concentrations of H2SO4, or to the production of a
compound of lower vapor pressure, such as (NH4 )2SO4.
SAMPLING ENVIRONMENT OF SULFATE FILTERS
Table 5 summarizes the temperature and humidity of the air sampled
by each of the filters. The dependence of ambient temperature on
altitude has been eliminated by converting measured temperature to
potential temperature, the temperature a given parcel of air would
have when transported adiabatically to the 1000 mb level. For com-
putational purposes the adiabadic lapse 'rate was taken to be 1° C/
57
-------
Table 5. SAMPLING ENVIRONMENT OF
SULFATE FILTERS
Filter
5 Aug
12 Aug
14 Aug
15 Aug
(a)
Potential
Description Temperature
(°C)
2
3
4
1
2
3
1
2
3
4
1
2
Labadie {1 km)
Labadie (4 km)
Labadie (18 km)
Wood River (12 km)
Wood River (25 km)
Wood River (50 km)
Labadie (10 km)
Labadie (21 km)
Labadie (32 km)
Labadie (45 km)
Wood River (10 km)
W ood River (28 km)
26
26
25
29
29
29
27
28
28
29
25
26
(a)
Dew Point
(°C)
7
8
7
18
19
19
17
16
15
12
16
17
Plume
sp/sT
(%)
1.5
1.5
2.4
7.5
9.4
7.9
3.8
4.2
5.9
11.4
4.2
5.8
(a) Average over filter sample period.
58
-------
0.5r
*
T240
i i i
1238 1236
TIME,CDT
76-033
1234
10 km
Figure 34. Horizontal plume traverse 21 km downwind of Labadie
stacks on 14 August 1974. Condensation nuclei peaks
at edge of plume indicate probable homogeneous
nucleation. (The same traverse is shown in Figure 27.)
59
-------
100 m, with sea level at 1013.25 mb. Similarly, the dependence of
relative humidity on temperature, and hence on altitude, has been
eliminated by converting measured humidity to dew point. This con-
version was based on an empirical formula derived from the relation-
ship of the saturation vapor pressure to the temperature.
The fraction of plume sulfur in the particulate phase, from Table 4,
is repeated in the right-hand column for comparison with the environ-
mental parameters. Because of the many uncontrolled variables in
the experimental design, the volume of data in Table 5 is not adequate
to establish a relationship between reaction rates and ambient tempera-
ture and humidity.
60
-------
APPENDIX TO SECTION 4
DETERMINATION OF BACKGROUND bscat AND
PLUME bscat/S03 BY REGRESSION ANALYSIS
When studying the pollutant plume from a specific source, it is often
desirable to separate the measured concentration of a pollutant into the
plume concentration, which is the excess contributed by the plume, and
the background concentration, which is the concentration which would be
present in the absence of the plume. For the Labadie plume, this is
conveniently done by regression analysis, using SO2 as a tracer for the
plume.
Figures 26, 27, and 28 show that the light scattering coefficient corre-
lated well with the SOS concentration during individual traverses of the
14 August 1974 Labadie plume. This means that on each traverse there
was an approximate linear relationship between the two parameters:
'scat
^ A [SO2] + B .
As part of the data screening process, the coefficients A and B which
minimize the mean square error in this relationship were determined
for each traverse, using standard computational methods (Bevington,
1969). For the traverses shown in Figures 26 - 28, the results were
as follows:
Regression Coefficients for 14 August 1974
Labadie Plume Traverses
o
Figure Pass A B P
Number Number (10~4 m"1 /ppm) (10~4 rrf1 ) Percent
26 3 4.9 1.8 85
27 9 7.8 2.0 91
28 17 25.0 2.5 87
The column ps gives the percent of the variance of bscat which is
"explained" by the linear relationship with SOS.
The constant term B gives the relationship's prediction for the light-
scattering coefficient when no SO2 is present. Since the SO3 concen-
tration outside the plume was generally below the detection threshold of
the instrument, B is interpreted as the average light-scattering coef-
ficient of the background aerosol. This interpretation is supported by
61
-------
comparison of the values for B given above with the bscat levels seen
outside the plume in Figures 26-28. (The coefficients A and B were
calculated for the full traverses, which included more sampling outside
the plume than is shown in the figures. )
The physical significance of the coefficient A can be seen by rewriting
the regression relationship in the following way: A - (bscaf;- B)/[SO2].
The slope of the relationship thus represents the ratio between plume
bscat and SO2, averaged across the plume. Computation of the ratio
in this way eliminates the variability associated with ratios calculated
from a single data point.
62
-------
SECTION V
OXIDANT CHEMISTRY
INTRODUCTION
The polluted St. Louis atmosphere was generally an oxidizing environ-
ment during the 1974 MISTT field program. Ozone concentrations
exceeding the 8 pphm Federal Ambient Standard were recorded on
most sampling days, and concentrations as high as 20 pphm were
measured near the city on 12 August 1974 (Figure 35).
Elevated concentrations were not confined to the metropolitan area,
but were regularly carried well downwind in an identifiable plume.
On the afternoon of 30 July 1974, for example, ozone concentrations
of 13 pphm were found over rural Illinois, 60 km downwind of the
Arch (Figures 36 and 37). These high concentrations were restricted to
the region directly downwind of Greater St. Louis, and ozone concen-
trations on either side were uniformly about 6 pphm.
The overall contribution of the metropolitan area to regional ozone
levels can be judged by the flow rate of ozone within the downwind
plume. This flow rate is calculated from the measured ozone cross
section of the plume and the pilot balloon observations of winds aloft:
Flow Rate = u f/([OaJ - [Oa], . ,) dy dz,
XJJ\L 01 L -"background
where u is the wind speed and y and z are the crosswind and
vertical coordinates. On the afternoon of 30 July 1974, the calculated
ozone flow rate in the St. Louis/Wood River plume was about 100
tons/hour. Wind speeds during the day were fairly steady, so this
figure can be interpreted as the approximate net rate at which ozone
was generated from the emissions of the urban industrial area in the
first 60 km (roughly 3 hours travel time) downwind.
The elevated concentrations of ozone and other oxidizing gases present
in the sampling area are of interest both for their direct impact on
human health (Hackney, 1975) and for their promotion of sulfur oxida-
tion and secondary aerosol formation (Wilson et al, 1972, Roberts, 1975).
This section discusses the observations of the Oa /NO/NO8 system made
by tho Mid. aircraft.
63
-------
76-135
Figure 35. Horizontal profiles of ozone and oxidant concentrations
downwind of Greater St. Louis under three different
wind regimes. Profiles were recorded during traverses
of profile baselines at the following altitudes and times:
(counter-clockwise from lower right) 455 m (1500 ft) msl,
0933-0952 CDT, 30 July 1974; 610 m (2000 ft) msl, 1145-
1152 CDT, 12 August 1974; 455 m (1500 ft) msl, 1027-1048
CDT, 15 August 1974. Arrows show average winds measured
in mixing layer during sampling period; their lengths equal
distance covered in one hour at average wind speed. Ozone
and oxidant concentrations substantially above background
levels formed downwind of Wood River and central St. Louis
area on most sampling days.
64
-------
(A)
(B)
E
E
bl
o
3
'OZONE
TEMPERATURE
5 2
E
UJ
o
r
OZONE
TEMPERATURE
%
<.
IO 20 30
TEMPERATURE, *C
1
40
76-117
10 20 30 40
TEMPERATURE, °C
76-113
OZONE,pphm
to
0
h
10
OZONE ,pphm
Figure 36. Vertical profiles of ozone concentration and temperature on 30 July 1974. Morning
profile (A), recorded during an 0859-0920 CDT spiral ascent near Lambert Field,
shows a temperature inversion at about 2400 m msl, together with an elevated layer
of ozone which was presumably formed during the previous day. Afternoon profile
(B), recorded during a 1604-1616 CDT spiral descent over Oakdale (111. ), shows
temperature inversion remaining at about 2400 m msl, with uniformly high ozone
concentrations in mixed layer below. Locations of spirals are shown in Figure 37(B).
-------
(A)
5000
4000
3000
UJ 2000
O
1000
13
II
1500
1000
E
LU
O
13
500 H-
CENTRALIA VOR
SPARTA f
76-120
76-134
Figure 37. Cross section of ozone plume downwind of Greater St.
Louis on the afternoon of 30 July 1974. Isopleths (A)
of ozone concentration (in pphrn) are drawn from records
of horizontal cross-plume traverses between 1356 and
1542 CDT. Vertical sounding over Oakdale (Figure
showed that plume extended to 2400 m msl at 16QO CDT.
Map (B) shows route of traverses and locations of spiral
soundings presented in Figure 36. Arrow shows average
wind measured in mixing layer at 1400 and 1500 CDT;
its length equals distance covered in one hour at average
wind speed.
66
-------
OXIDANT PROFILES
In terms of total concentration, the most important oxidizing species
normally present in polluted urban air are ozone and nitrogen dioxide.
Both of these species are of secondary origin. Ozone is formed by
the photolysis of nitrogen dioxide, followed by the reaction of oxygen
atoms with molecular oxygen:
NO3 + hv * NO + O /!/
O + Oa+M-»O3+.M. /2/
Nitrogen dioxide is a product of the reaction of ozone with nitric
oxide:
NO + Oa * NOS + O3. /3/
These three reactions form a closed cycle, leaving invariant the total
concentration of oxidizing species:
dt
{[O] + [03] + [NOa]}= (ri - ia) + (ra - r3) + (r3 - rx) = 0.
*
In the absence of competing reactions, the oxidant level would thus
be constant.
During the daylight hours in polluted urban air, Reactions 1-3
proceed faster than do any other known reactions involving the same
species (Leighton, 1961). Even relatively slow competing reactions
can affect oxidant levels, however. An example is the oxidation of
nitric oxide by reaction with a peroxy radical:
NO + ROO« -* NO2 + RO- /4/
Reactions of this sort destabilize the oxidant concentration, and allow
it to rise:
*The term "oxidant" is used in different senses by different authors.
In this report it is used to refer to any of the oxidizing gases, and the
oxidant concentration is taken to be the sum of the concentrations of
atomic oxygen, ozone, and nitrogen dioxide.
67
-------
+ [03]
(rjL - ra) + (ra - ra) + (r3 - ri + r4) = k* [NO] [ROO-] .
During the 1974 MISTT field program, the MRI aircraft monitored
Oa, NO, and NOX concentrations with continuous chemiluminescent
analyzers. Figures 35 and 38 - 41 show oxidant profiles recorded
on several traverses. (The oxidant concentrations given are derived
from the measured concentrations as follows: [O] + [Oa] + [NO3] ="
[Oa] + [NO2] = [Oa] + [NOX] - [NO]. This formula is based on the
observations that atomic oxygen contributes a negligible fraction
of the oxidant concentration and that the low-temperature molybdenum
converter in the NOx analyzer is designed to reduce only NOs )
The oxidant profiles of the Labadie and St. Louis/Wood River plumes
were distinctly different. The Labadie plume had little effect on
ambient oxidant concentrations at the distances sampled, from 1 to
45 km downwind of the plant (Figures 39 and 40). It appears that the
competing reactions which lead to the accumulation of oxidant in
urban photochemical smog were of minor importance in the somewhat
simpler chemical system of the Labadie plume. The St. Louis and
Wood River plumes, in contrast, were characterized by oxidant
concentrations well above background levels (Figures 35 and 41 ).
Net production of oxidant within these urban-industrial plumes was
apparent less than 15 km downwind of hydrocarbon source areas.
In such hydrocarbon-rich environments, reactions other than 1-3
quickly destroyed the conservation of oxidant.
OZONE IN THE LABADIE PLUME
The oxidant chemistry of power plant plumes has recently come
under close scrutiny, following the reported observation by Davis
et al (1974) of elevated ozone concentrations downwind of a coal-
fired power plant in Maryland. No such ozone "bulge" was observed
in the Labadie plume on any of the sampling days.
The sulfur dioxide plume from the coal-fired, steam-generating
plant at Labadie was followed for 45 km on 14 August 1974. There
was little loss of sulfur dioxide over the sampling distance, so this
primary pollutant can be used as a tracer for the effluent of the plant.
Ozone concentrations outside the sulfur dioxide plume on 14 August
68
-------
20r
15
E
.c
Q.
Q.
9 10
i-
<
o:
UJ
o
z
o
o
.-mm*
i
i
i i
j
/V MEASURED 03 !
1
S\'' MEASURED S02 j
j
0° 0 ° CALCULATED 03
i
j
1
i
i
i
i
i
i
i
i
i
i
i
OCXTN^ !
"^ oB>eO*bCfv XSJ
-1
10 km
XCOO 00000
O I3*& ^-+*^~*>
J
rv
1143
1142
1141
1140
TIME, CDT
76-089
Figure 38.
Horizontal plume traverse 10 km downwind of Labadie
stacks on 14 August 1974. Ozone concentrations within
plume were depressed below background levels on either
side. Measured ozone profile agreed well with ozone
profile calculated from measured SO 2 profile and oxidant-
conserving photochemistry. The altitude of this traverse
was 610 m (2000 ft) msl.
69
-------
- /'\/ MEASURED 034-N02 | \
CL
cx
O
H
10
O
2
O
u
10 km
MEASURED
II
O O
" MEASURED
CALCULATED
0
V
/"-
1240
1238 1236
TIME, CDT
1234
76-090
Figure 39. Horizontal plume traverse 21 km downwind of Labadie stacks
on 14 August 1974. Ambient ozone concentrations outside
plume were slightly higher than for earlier traverse of
Figure 38. Oxidant (Oa +NO3) concentrations within plume
were consistent with background levels on either side. This
traverse was flown at 455 m (1500 ft) msl.
70
-------
MEASURED 03
' MEASURED 63
,/' MEASURED S02
0°0° CALCULATED 0*
10km
10
A
^7 vs r-
E
JC
Q.
0.
O
V
LU
(J
2
O
o
1436
1434
1432 1430
TIME,CDT
1428
1426
76-085
Figure 40.
Horizontal plume traverse 45 km downwind of Labadie stacks on 14 August 1974. Ozone
concentrations within the plume at this distance were still depressed below background levels
on either side, although not as strongly as on traverses nearer the stacks. Measured ozone
and oxidant profiles both show large-scale concentration gradient, possible due to urban
plume from Greater St. Louis. The altitude of this traverse was 610 m (2000 ft) msl.
-------
76-136
Figure 41. Early-morning horizontal profiles of ozone concentra-
tion, oxidant (Oa + NO 3) concentration, and light-
scattering coefficient (bscat) downwind of Greater
St. Louis on 3 August 1974. The profiles were
recorded during a horizontal traverse of their com-
mon baseline at 457 m (1500 ft) msl between 0835
and 0856 CDT. The arrow shows the average wind
measured in the mixing layer during the sampling
period; its length equals the distance covered in one
hour at the average wind speed. Narrow plumes of
oxidant and light-scattering aerosols formed immedi-
ately downwind of Wood River refinery zone and
downtown St. Louis.
72
-------
1974 were elevated, generally exceeding 6 pphm. Within the plume,
ozone concentrations were depressed below these levels (Figures
39 - 40), due to the reduction of ozone by the nitric oxide in the
plant effluent.
While ozone concentrations within the plume were depressed relative
to those outside at all distances, the absolute ozone concentrations in
the center of the plume increased with distance downwind, from
1 pphm at 10 km to 5 pphm at 45 km. In principle, such a rise could
be due either to the simple dilution of the plume by background air,
or to the net production of oxidant by chemical reactions which would
lead eventually to excess ozone levels within the plume. These two
effects can be distinguished by quantitatively analyzing the relation-
ship of ozone to plume geometry. Figures 38 - 40 compare measured
ozone profiles of the plume with ozone profiles calculated from
oxidant-conserving photochemistry and the observed plume geometry,
as described in the appendix at the end ofthis section. These calcula-
tions use sulfur dioxide as a tracer for the plume and are based on
the photolytic cycle governed by Reactions 1-3.
The correspondence between calculated and measured concentrations
is quite good at 10 km and 21 km from the plant. There is a sugges-
tion in Figure 40 that the agreement was beginning to break down
45 km from the plant, where the observed ozone deficit was narrower
and deeper than that calculated from the geometry of the SO 2 plume.
The origin of this difference is not known. Possible explanations
include the following: (a) ultraviolet radiation was significantly
attenuated within the core of the plume due to scattering by aerosols,
depressing the rate of nitrogen dioxide photolysis; (b) reactions
involving hydrocarbon products were important where the plume
mixed into the background air, destabilizing the oxidant concentra-
tion; (c) the profile of the background oxidant concentration was
somewhat irregular. Overall, the agreement between calculated
and measured concentrations indicates that any net production of
oxidant was insignificant at the distances sampled, and that the
observed downwind decrease in the plume ozone deficit was due to
the spreading and dilution of the plume.
OXIDANT AND LIGHT-SCATTERING AEROSOLS IN THE ST. LOUIS
AND WOOD RIVER PLUMES
Oxidant concentrations in the urban-industrial plume are established
by a complex chemical system involving many reactants which were
73
-------
not monitored by the MRI aircraft. It is nevertheless possible to
deduce some of the bulk characteristics of the reaction scheme from
the profiles of the species which were measured.
Figures 35, 41, and 42 show oxidant profiles observed immediately
downwind of the metropolitan area under three different wind condi-
tions: northwesterly (30 July and 3 August), southerly (12 August),
and southeasterly (15 August). All four traverses were flown in the
morning, at altitudes from 457 m (1500 ft) to 610 m (2000 ft) msl.
The profiles from all four traverses displayed certain regularities;
the highest oxidant concentrations were measured downwind of down-
town St. Louis and the Wood River refineries, and were associated
with high light-scattering coefficients.
The oxidant profiles of 30 July and 15 August, measured on oppostie
sides of the city, are almost mirror images. Each shows two
distinct oxidant plumes, one downwind of St. Louis/East St. Louis/
Granite City and the other downwind of Alton/Wood River. The
apparent superposition of these two plumes, by a southerly wind on
12 August, produced the highest concentrations measured during the
17-day sampling period. The narrow oxidant plumes observed before
0900 CDT on 3 August appear to pinpoint the central business district
of St. Louis and the refinery complex of Wood River as the source areas.
The automotive traffic of downtown St. Louis and the refinery opera-
tions at Wood River are the most important sources of reactive hydro-
carbons in the sampling area, and their emissions evidently played
a critical role in the production of oxidant. It is worth noting that
the 100 ton/hour estimated net rate of ozone production in the 30 July
St. Louis/Wood River plume greatly exceeds the 20 ton/hour estimated
annual average rate of hydrocarbon emissions in the St. Louis/Wood
River urban-industrial area (U.S. EPA, Illinois EPA). In contrast,
the maximum ozone concentrations produced by the irradiation of
hydrocarbon mixtures in smog chambers are typically less than the
initial concentrations of hydrocarbons (e.g., Wilson, et al, 1972).
Possible explanations for the apparent high ozone yield of hydro-
carbons in the urban-industrial plume have been discussed by Walker
(1976).
In most traverses, the light-scattering coefficient correlated well
with oxidant concentrations within a given plume, although the
relative concentrations of light-scattering aerosols and oxidant
varied from plume to plume. Moreover, the narrow aerosol and
oxidant plumes of 3 August and 12 August are remarkably coincident.
74
-------
76-137
Figure 42.
Horizontal profiles of oxidant concentration and light-
scattering coefficient (bscat) downwind of Greater St.
Louis under three different wind regimes. Traverses
and wind arrows are as in Figure 35. High oxidant
concentrations were generally associated with high
light-scattering coefficients.
75
-------
These observations can, in principle, be explained by either of two
hypotheses; the geographical distribution of hydrocarbons and primary
particulate sources could be similar, or oxidant and secondary aerosols
could both be products of the same system of reactions. Preliminary
analysis of the data from the 1975 MISTT field program (Appendix B)
shows that most of the light-scattering aerosol contributed by the
emissions of the metropolitan area was definitely of secondary origin.
76
-------
APPENDIX TO SECTION 5
A MODEL FOR THE OZONE DEFICIT IN POWER PLANT PLUMES
In the following paragraphs, a model is presented which predicts
the ozone profile of the plume from, the observed sulfur dioxide
profile, based on the fundamental photolytic Oa/NO/NO8 cycle:
NO 2 + h V + NO + O
121
O 3 + NO +- NO 2 + O 2
In the atmosphere, molecular oxygen and the third body M of Reaction
2 are present at high concentrations which are not disturbed by trace
chemistry. The state of the chemical system participating in Reac-
tions 1 - 3 is thus characterized at any point by four variables: the
local concentrations of ozone, atomic oxygen, nitric oxide, and
nitrogen dioxide. Through a change of coordinates, this state can
be expressed in terms of other variables as well. It is particularly
convenient to express the state of the system in terms of variables
which are invariants of Reactions 1-3 and depend only on the
environment and geometry of the plume. In these coordinates, the
problem reduces to that of the physical dispersion of an inert
pollutant.
Since the system under consideration has four degrees of freedom,
four chemical invariants must be identified:
In the atmosphere, atomic oxygen tends to a stationary
state in which it is destroyed by Reaction 2 at about
the same rate at which it is formed by Reaction 1. In
this situation
77
-------
or
[NOa] = ks [O] [03] [M],
[0]/[N08] -
-------
It was observed in the introduction that the total oxidant
concentration is a stoichiometric invariant of Reactions
1 - 3:
[O] + [Os] + [NO 2] = [oxidant]. (5)
Unlike the relationships in Equations 3 and 4, which
could be perturbed only by competing reactions with
speeds comparable to those of Reactions 1 - 3 , the
conservation of oxidant can be perturbed over a period
of time even by comparatively slow competing reactions.
The background oxidant level in the sampling area is in
fact established by the interplay of such competing
reactions, but it does not appear to be perturbed by the
Labadie plume at the distances sampled. Very little
oxidant is emitted by the plant, and the concentration of
oxidant within the plume was consistent with the back-
ground levels on either side (Figures 39 and 40).
Like the concentration of oxidant, the concentration of
nitrogen oxides is a stoichiometric invariant of Reactions
1-3
[NO] + [NO2] = [NOx]
Unlike the previous three chemical invariants, which are
essentially unaffected by the emissions of the Labadie
plant, the concentration of nitrogen oxides is significantly
higher within the plume than it is on either side, due to
the nitric oxide emitted by the plant. The spatial distribu
tion of nitrogen oxides is easy to model, however, since
NOX can be treated as a chemically inert contaminant.
In Figures 43 and 44, SO2 is used as a tracer for the
plume, and the expected NOX concentration is calculated
from the formula
[NOX] = 0.3 [SOa] + B ,
*
with the background, B, equal to 26 ppb at 21 km and
30 ppb at 45 km. The correspondence between calculated
and measured concentration is good. (As with Relation 5,
79
-------
io
o
H
<
£T
H
Z.
UJ
o
z:
o
o
10 km
MEASURED NO x
o°0° CALCULATED NO
/^S MEASURED NO
. % * CALCULATED NO
X
1240
1238 1236
TIME , CDT
1234
76-119
Figure 43. Horizontal plume traverse 21 km downwind of Labadie stacks on 14
August 1974. Most of the NO emitted by the power plant had
already been oxidized to NO3. Measured NOX profile agreed
well with NOX profile scaled from measured SOS concentrations.
Measured NO profile agreed well with NO profile calculated
from measured SO2 profile and oxidant-preserving photochemistry.
(The same traverse is shown in Figure 39).
80
-------
oo
io
CL
Q.
O
h-
<
er
h-
^
LU
O
~ZL
a
o
10 km
MEASURED NOX
CALCULATED NO
MEASURED NO
CALCULATED NO
A
/ \
436
1434
1432 1430
TIME,CDT
1428
1426
76-118
Figure 44. Horizontal plume traverse 45 km downwind of Labadie stacks on 14 August 1974. Nitric oxide
concentrations were near the detection limit of the monitor, and very little excess within the
plume was observed or predicted. (The same traverse is shown in Figure 40 . )
-------
comparatively slow competing reactions may, over a
period of time, erode the conservation of NOX in
Relation 6. This effect does not appear to have been
important at the distances sampled. )
The preceding paragraphs have identified four quantities, [O]/[NOg],
[Oa] [NO]/[NO J, [O] + [03] + [NO 3], and [NO] + [NO2], which
are invariants of Reactions 1-3 under the photostationary condi-
tions which prevail in the atmosphere. Together, they characterize
the state of this simple chemical system, and elementary algebraic
manipulation of Equations 3 to 6 gives:
[O] =~ ftpkx/ks + [oxidant] + O [NOX] - X), (7)
[03] = j (-pki/ks + [oxidant] - [NOX ] + X), (8)
[NO] =y^ (-
-------
Table 6 . PARAMETERS USED IN MODEL, CALCULATIONS
Figure
Number
38
-------
(d) The traverses at 10 km from the plant were flown with
the nitrogen oxides monitor on the 0-1 ppm scale, which
was not separately calibrated. Nitric oxide, NO x» and
oxidant profiles are therefore not shown for this distance.
84
-------
SECTION VI
REFERENCES
Adams, D. F. , and R. K. Koppe. Instrumenting Light Aircraft for
Air Pollution Research. Journal Air Pollution Control Assoc., 19.
pp 410-415. 1969.
Ahlquist, N. C., and R. J. Charlson. Measurement of the Vertical
and Horizontal Profile of Aerosol Concentration in Urban Air with the
Integrating Nephelometer. Environ. Sci. Tech., !2. pp 363-366. 1968.
Anderson, J. A., and D. L. Blumenthal. Aircraft Monitoring Support
for an Aerosol Characterization Study in St. Louis. MRI Final Report
No. MRI 74 FR-1146. 1974.
Bevington, P. R. Data Reduction and Error Analysis for the Physical
Sciences. McGraw-Hill, New York. 1969.
Davis, D. D., G. Smith, G. Klauber. Trace Gas Analysis of Power
Plant Plumes via Aircraft Measurement: Os, NOX, and SO2
Chemistry. Science, 186. pp 733-736. 1974.
Gillani, N. V. Data Processing, Management and Analysis Related
to Project MISTT, 1975. Report of the Air Pollution Laboratory,
Washington University, St. Louis, Missouri. 1975.
Hackney, J. D. Testimony given to the Permanent Subcommittee on
Air Quality of the California State Assembly. 26 September 1975.
Husar, R. B. Coagulation of Knudsen Aerosols. Ph.D. Thesis,
University of Minnesota, Minneapolis, Minnesota. 1971.
Husar, J. D., R. B. Husar, E. S. Macias, W. E. Wilson, J. L.
Durham, W. K. Shepherd, and J. A. Anderson. Particulate Sulfur
Analysis: Application to High Time Resolution Aircraft Sampling
in Plumes. To be published. 1975.
f
Jackson, J. O. Ambient Kinetics of the NO2 Photolysis Rate and
Verification of the Atmospheric NO/NO3 /O3 Photo stationary State.
Ph.D. Thesis, University of Michigan, Ann Arbor, Michigan. 1974.
85
-------
REFERENCES (continued)
Leighton, P. A. Photochemistry of Air Pollution. Academic Press,
New York. 1961.
Link, W. T., D. A. McClatchie, A. B. Watson, and A. Compher.
A Fluorescent Source NDIR Carbon Monoxide Analyzer. Joint
Conference on Sensing of Environmental Pollutants, Palo Alto,
California. 8-10 November 1971.
MacCready, P. B., Jr., R. E. Williamson, S. Berman, and A.
Webster. Operational Application of a Universal Turbulence
Measuring System. MRI Final Report No. MRI 65 FR-301 to NASA
Edw, Contract NASA-784. 1965.
Markowski, G. R. A Useful Method for Transient Response Analysis
and Data Correction Applied to Aircraft Plume Sampling. 68th Annual
Meeting of the Air Pollution Control Assoc. Paper No. 75-40. 3.
June 1975.
Mirabel, P. and J. L. Katy. Binary Homogeneous Nucleation as a
Mechanism of the Formation of Aerosols. J. Chem. Phys., 60.
pp 1138-1144. 1974.
Owens, M. J. and A. W. Powell. Deposition Velocity of Sulfur
Dioxide on Land and Water Surfaces Using a SB S Tracer Method.
Atmos. Environ., 8^. pp 63-67. 1974.
Rich, T. A. Particles in Air Pollution. Colloquium on the Physics
and Chemistry of Aerosols, Fontenay-Aux-Roses, France. September
1970.
Roberts, P. T. Gas-to-Particle Conversion: Sulfur Dioxide in a
Photochemically Reactive System. Ph.D. Thesis, California Institute
of Technology, Pasadena, California. 1975.
Schofield, K. An Evaluation of Kinetic Rate Data for Reactions of
Neutrals of Atmospheric Interest. Planetary and Space Sciences, 15.
pp 643-670. 1967.
Shepherd, J. G. Measurements of the Direct Deposition of Sulfur
Dioxide onto Grass and Water by the Profile Method. Atmos. Environ.,
8.. pp 69-74. 1974.
86
-------
REFERENCES (continued)
Walker, H. M. Some Considerations Regarding Rural Ozone. Air
Pollution Control Assoc. Conference on Ozone/Oxidants, Dallas,
Texas. March 11-12, 1976.
Whitby, K. T. , B. K. Cantrell, R. B. Husar, N. V. Gillani, J. A.
Anderson, D. L. Blumenthal, W. E. Wilson, Jr. Aerosol Formation
in a Coal Fired Power Plant Plume. For presentation before the
Division of Environmental Chemistry, Americal Chemical Society,
New York. April 4-9, 1976.
White, W. H., P. T. Roberts, and S. K. Friedlander. On the Nature
and Origins of Visibility-Reducing Aerosols in the Los Angeles Air
Basin. In the Characteristics and Origins of Smog Aerosol, G. M.
Hidy and P. K. Mueller, Editors. To Appear. 1975.
Wilson, W. E. Jr., A. Levy, and D. B. Wimmer. A Study of Sulfur
Dioxide in Photochemical Smog II. Effect of Sulfur Dioxide on Oxidant
Formation in Photochemical Smog. J. Air Poll. Control Assoc., 22.
pp 27-32. 1972.
87
-------
APPENDIX A
SAMPLING PATTERNS
88
-------
Fig. 45- LOCATION OF AIRCRAFT SAMPLING ROUTES,
MORNING, JULY 30, 1974
89
-------
Fig. 46. LOCATION OF AIRCRAFT SAMPLING ROUTES,
AFTERNOON, JULY 30, 1974
90
-------
TABLE 7.
Flight Outline
Date: July 30, 1974 (Morning)
Cartridge No. 036
Pass No.
Route
Spiral East of Lambert
2
3
4
5
6
7
New Hanover -Hamel
Hamel-Valmeyer
Valmeyer -Hamel
Hamel-New Hanover
New Hanover-Hamel
Hamel-Bi-State
Start Stop
Time Time
(GOT) (CDT)
Altitude
(mrnsl)
Spiral Bi-State Airport
0859 0920 180-2896
0933 0952 457
0958 1021 610
1025 1047 914
1052 1111 1219
1115 1134 1524
1142 1152 2134
1153 1204 2134-126
Sulfate
Filter
1
1
1
1
1
1
91
-------
TABLE 8.
Flight Outline
Date: July 30, 1974 (Afternoon)
Cartridge No. 027
Start Stop
Time Time Altitude Sulfate
Pass No. Route (CDT) (CDT) (m msl) Filter
Spiral Bi-State Airport 1323 1341 126-2591
2 Sparta-Centralia 1356 1411 305 2
3 Centralia-Walsh 1415 1437 610 2
4 Walsh-Centralia 1440 1500 914 2
5 Centralia-Sparta 1504 1522 1211 2
6 Sparta-Centralia 1525 1542 1524 2
Spiral Oakdale 1604 1616 2591-183
92
-------
vO
Fig. 47. LOCATION OF AIRCRAFT SAMPLING ROUTES, JULY 31, 1974
-------
Date: July 31, 1974
Cartridge No. 033
Pass No. Route
Spiral Labadie
1 K - L
2
3
4
5
6
7
8
9
10
11
12
13
14
15 A - M
16 I
17
18 ^
Spiral Labadie
TABLE 9.
Flight Outline
Start Stop
Time Time Altitude Sulfate
(CDT) (CDT) (m msl) Filter
0650 0659 1370-305
0707
0711
0715
0719
0723
0728
0733
0738
0745
0750
0753
0757
0804
0808
0708
0712
0716
0722
0725
0730
0734
0739
0747
0751
0755
0759
0806
0810
305
365
425
487
548
610
670
730
792
853
914
975
1036
1097
0811
0816
0821
0826
0814
0819
0823
0828
1158
1219
1280
1158
0842 0854 1829-152
94
-------
Fig. 48. LOCATION OF AIRCRAFT SAMPLING ROUTES, AUGUST 1, 1974
-------
TABLE 10.
Flight Outline
Date: August 1, 1974
Cartridge No. 037
Pass No.
Route
Spiral West of JLabadie
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
A
A
A
A
A
C
D
C
C
E
F
E
F
E
H
E
- B
- B
- B
- B
- B
- D
- C
- D
- D
- F
- E
- F
- E
- H
- E
- H
Start Stop
Time Time Altitude Sulfate
(CDT) (CDT) (m msl) Filter
Spiral
^^i
0808
0821
0825
0829
0833
0837
0850
0858
0905
0912
0927
0938
0950
1004
1020
1030
1040
1100
W^^^B
08
0824
0828
0832
0836
0840
0855
0902
0909
0918
0937
0948
1000
1014
1028
1036
1048
1112
1371-213
457
426
488
549
518
427
488
549
610
427
488
365
549
488
365
610
1676-183
96
-------
Fig. 49. LOCATION OF AIRCRAFT SAMPLING ROUTES,
MORNING, AUGUST 3, 1974
-------
vO
00
Fig. 50. LOCATION OF AIRCRAFT SAMPLING ROUTES,
AFTERNOON, AUGUST 3, 1974
-------
TABLE 11.
Flight Outline
Date: August 3, 1974 (Morning)
Cartridge No. 034
Pass No.
Route
Start Stop
Time Time
(CDT) (CDT)
Altitude
(m msl)
Spiral Lambert
0810 0821 180-1524
2
3
4
5
6
7
Prairietown-Water loo
Waterloo -Prairietown
Prairietown-Waterloo
Waterloo -Prairietown
Prairietown-Waterloo
Waterloo -Prairietown
0835
0859
0923
0945
1013
1040
0856
0920
0943
1007
1035
1100
457
762
914
610
1067
1372
Sulfate
Filter
1
1
1
1
Spiral Nichols Airstrip
1128 1146 2897-128
99
-------
TABLE 12.
Flight Outline
Date: August 3, 1974 (Afternoon)
Cartridge No. 747
Pass No.
Spiral
2
3
4
5
6
7
8
9
10
11
12
Spiral
Route
A
New Douglas -Highland
Highland -New Douglas
New Douglas -Highland
Highland-E
E -Highland
Highland-E
E -Highland
Highland-E
E -Highland
Highland-E
E -Highland
A
Start
Time
(CDT)
1327
1341
1352
1402
1413
1427
1439
1452
1504
1517
1529
1540
1556
Stop
Time
(CDT)
1336
1350
1400
1410
1422
1436
1448
1501
1513
1527
1537
1550
1610
Altitude
(m msl)
1981-152
305
457
610
762
914
1067
1219
1372
1524
1676
1829
2438-183
Sulfate
Filter
2
2
2
2
2
2
2
2
2
2
2
100
-------
Fig. 51. LOCATION OF AIRCRAFT SAMPLING ROUTES, AFTERNOON, AUGUST 5, 1974
-------
TABLE 13.
Flight Outline
Date: August 5, 1974
Cartridge No. 788
Pass No.
Route
Start Stop
Time Time Altitude Sulfate
(CDT) (CDT) (m msl) Filter
Spiral
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Labadie
A -
C -
B
Augusta
Augusta - C
C -
Augusta
Augusta - C
C
E -
F -
E -
F -
E -
- Augusta
F
E
F
E
F
1724
1748
1751
1754
1759
1804
1807
1810
1814
1825
1830
1835
1841
1848
1856
1905
1918
1928
1940
1744 2590-183
1749
1753
1756
1800
1806
1809
1812
1815
1827
1832
1837
1843
1850
1902
1911
1924
1935
1947
365
427
488
549
610
762
884
686
1676
1372
1067
762
457
762
396
1067
1372
1676
1
2
2
2
2
2
2
2
2
4
4
4
4
4
3
3
3
3
3
Spiral H
1954 2014 2590-213
102
-------
o
CO
Fig. 52. LOCATION OF AIRCRAFT SAMPLING ROUTES, AUGUST 10, 1974
-------
TABLE 14.
Flight Outline
Date: August 10, 1974
Cartridge No. 795
Pass No.
Spiral
2
3
4
5
Spiral
Route
Start Stop
Time Time Altitude Sulfate
(CDT) (CDT) (m msl) Filter
Z 1505 1532 2743-153
Cedar Hill-New Athens 1547 1603 457
New Athens-Cedar Hill 1607 1629 1067
Labadie-St. Louis VOR 1637 1652 457
Winfield Dam-Highland 1700 1729 457
Prairietown 1751 1820 2743-153
3
4
104
-------
o
Ol
Fig. 53. LOCATION OF AIRCRAFT SAMPLING ROUTES, AUGUST 11, 1974
-------
Date: August 11, 1974
Cartridge No. 755
Pass No.
2
3
5
6
7
8
Route
Spiral Augusta
TABLE 15.
Flight Outline
Augusta-Festus
Festus -Augusta
Labadie-Troy
Lebanon-Brighton
Brighton-Lebanon
0942
1009
B - Litchfield
LitchfLeld - Gr eenville
Spiral D
Start stop
Time Time Altitude Sulfate
(CDT) (CDT) (m msl) Filter
1
1
0904 0934 2896-153
1000
1028
1030 1054
457
914
610
1100
1123
1158
1224
1118
1140
1219
1236
457
914
914
457
2
2
3
3
1244 1253 152-1067
106
-------
Fig. 54. LOCATION OF AIRCRAFT SAMPLING ROUTES, AUGUST 12, 1974
-------
TABLE 16.
Flight Outline
Date: August 12, 1974
Cartridge No. 809
Pass No.
Spiral
2
3
4
5
6
7
8
9
10
11
12
13
Route
Civic MemorialAirpt.
Wordon - A
A - B
B - A
Brighton-Bunker Hill
Bunker Hill- Brighton
Brighton-Bunker Hill
Carlinville - C
C - Carlinville
Carlinville - C
C - Carlinville
1055
1134
1145
1158
1213
1220
1229
1242
1253
1304
1315
1125
1143
1152
1205
1218
1225
1234
1250
1302
1312
1323
Start Stop
Time Time Altitude Sulfate
(CDT) (CDT) (m msl) Filter
Morrisonville .-
Jacksonville VOR 1345
Jacksonville VOR -
Morrisonville 1412
457
610
762
457
610
762
457
610
762
457
1408 457
1434 762
1
1
1
2
2
2
3
3
3
3
4
4
Spiral Ldtchfield
1443 1503 1981-GR
108
-------
o
NO
Fig. 55. LOCATION OF AIRCRAFT SAMPLING ROUTES, AUGUST 14, 1974
-------
TABLE 17.
Flight Outline
Date: August 14, 1974
Cartridge No. 885
Pass No.
Route
Spiral Labadie
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
A
B
A
B
E
F
E
F
H
J
K
L
M
N
0
N
P
- B
- A
- B
-A
- F
- E
- F
- G
- J
- H
- L
- K
- N
- O
- N
- P
- N
Start Stop
Time Time Altitude Sulfate
(CDT) (CDT) (m msl) Filter
1
1
1
1
2
2
2
2
3
3
3
3
4
4
4
4
4
1045
1129
1137
1147
1157
1210
1218
1225
1233
1251
1302
1316
1326
1351
1407
1416
1425
1438
1121
1135
1144
1154
1204
1215
1223
1231
1243
1259
1310
1323
1332
1405
1414
1423
1436
1451
3505-153
457
610
762
914
914
762
610-
457
457
610
762
914
1220
914
762
610
457
Spiral RohlfLng Airport
1501 1518 1829-153
110
-------
Fig. 56. LOCATION OF AIRCRAFT SAMPLING ROUTES, AUGUST 15, 1974
-------
TABLE 18.
Flight Outline
Date: August 15, 1974
Cartridge No. 824
Pass No.
Spiral
2
3
4
5
6
7
8
Spiral
Route
Hammet Airport
Fosterburg-W. Alton
W. Alton-Fosterburg
Fosterburg-W. Alton
Fidelity - C
C - Medora
Medora - C
C - Medora
St. Charles Airport
Start
Time
(CDT)
0912
0946
0953
1007
1027
1050
1114
1139
1215
Stop
Time
(CDT)
0940
0951
1000
1012
1048
1111
1136
1200
1236
Altitude
(m msl)
2896-153
457
549
640
457
549
670
793
1981-137
Sulfate
Filter
1
1
1
2 & 3
2 & 3
2 & 3
2 & 3
112
-------
APPENDIX B
FORMATION AND TRANSPORT OF SECONDARY
AIR POLLUTANTS: OZONE AND AEROSOLS
IN THE ST. LOUIS URBAN PLUME
MRI 76 Pa-1392 (revised), accepted for publication in SCIENCE.
113
-------
ABSTRACT
Emissions from metropolitan St. Louis caused reduced visibilities and viola-
tions of the Federal Ambient Standard for ozone 160 km or more downwind of
the city on July 18, 1975. Atmospheric production of ozone and visibility-
reducing aerosols continues long after their primary precursors have been
diluted to low concentrations.
114
-------
Metropolitan St. Louis is a major urban-industrial center, encompass-
ing coal-fired power plants with a combined capacity of 4600 MW, oil refineries
with a combined capacity of 4.4 x 10 bbl/day, various other industry, and a pop-
ulation of about two million. It is surrounded by flat, predominantly agricultural
terrain, the nearest neighboring city of 50, 000 or more people being 135 km dis-
tant. Due to its isolation, the impact of St. Louis on ambient air quality is rela-
tively easy to identify; air which has been modified by the aggregate emissions of
the metropolitan area forms an "urban plume" downwind. The Fate of Atmo -
1-4
spheric Pollutants Study (FAPS) has shown that this plume is often identifiable
80 or 120 km from the city.
c c
As part of Project MISTT ' (Midwest Interstate Sulfur Transformation and
Transport), pilot balloons and instrumented aircraft were used during the summer
of 1975 to quantify the three-dimensional flow of aerosols and trace gases in the
St. Louis urban plume. The plume was successfully tracked out to 240 km, and
»
was mapped in detail out to 160 km (distances are measured from the Gateway
Arch). At these distances the plume was still well defined, and on the order of
50 km wide. Our observations indicate that much of the pollutant background in
the eastern United States may be due to the combined emissions of metropolitan
areas far upwind.
Ozone was a conspicuous indicator of the St. Louis urban plume during the
summer sampling period. Daytime ozone concentrations within the plume gen-
erally exceeded the 0. 08 ppm Federal Ambient Standard, even on the most distant
sampling runs. Peak concentrations in the plume were typically twice those in
115
-------
W. H. White, 4
the unmodified background air, and surpassed 0.15 ppm on most sampling days.
It is estimated that one to two parts ozone were produced for each part (C) non-
methane hydrocarbons emitted by the city.
Reduced visibilities were a second characteristic feature of the St. Louis
urban plume. Most of the light-scattering aerosol responsible for this reduction
was of secondary origin, and its production was observed farther downwind and
later in the day than that of ozone. At 100 km or more downwind, the average
7
light-scattering coefficient (b ) within the plume was typically 50 percent
S Cctt
above the background levels on either side, with sulfate compounds accounting
for most of the excess.
The results reported here generally corroborate those of FAPS on the
identifiability and geometry of the St. Louis urban plume. The primary contri-
bution of Project MISTT was to quantify the flow of material at increasing down-
wind distances. This made possible the study of transformations undergone by
pollutants in the atmosphere, at dilutions and time scales which are difficult to
simulate in the laboratory.
8
Figure I shows ozone and b profiles recorded by the sampling aircraft
SCd>t*
during selected cross-wind traverses downwind of St. Louis on July 18, 1975.
Each traverse path was flown at three different altitudes, as indicated in Fig-
ure 2a, starting just downwind of St. Louis at 0900 CDT and finishing near ,
Decatur, Illinois, at 1900 CDT. These horizontal traverses, together with
vertical soundings from 1500 m msl to near ground level, documented a broad,
shallow pollutant plume extending from St. Louis out past Decatur, 170 km to
116
-------
the northeast. The polluted airmass was about 50 km wide and deepened during
the day, the mixing depth exceeding 1 km in the afternoon.
Ozone and light- scattering aerosols were the most prominant indicators of
the aged plume 50 km. or more downwind of St. Louis. Outside the plume, back-
ground concentrations of these contaminants were fairly uniform on July 18, al-
though some large-scale gradients were noted (cf. ref . 2); ozone concentrations were
generally below the 0. 08 ppm Federal Ambient Standard, scattering coefficients
-4 -1
were below 2.8X10 m , and particulate sulfur concentrations were below
3. 6 A*g/m in the few passes made completely outside the plume. Within the
plume, the measured ozone concentration peaked at 0. 17 ppm 66 km out, the
.4 -1
measured scattering coefficient reached 4. 9 X 10 m at 125 km,, and the par-
ticulate sulfur concentration, averaged over an entire traverse, measured as
2
high as 7. 5 fig/m at 125 km, corresponding to an estimated average in-plume
concentration of about 15 uglm. . In contrast, SO and NO^ concentrations
fell rapidly to within 0. 03 ppm of background, away from major point sources.
The July 18 plume was mapped under hazy skies, with some scattered
thunder showers appearing toward evening. The continued presence of a low
pressure trough over the western plains produced decreasing stability and
strong southwesterly flow over the Missouri and upper Mississippi valleys,
which was documented in the sampling area by a total of 36 pilot balloon obser-
vations carried out as part of Project ME5TT. The mean trans port vector (cf. ref. 1)
lay between 230° and 243° during the middle of the day, at speeds of 20-36
km/hr. At these speeds, among the highest encountered during the program,
117
-------
emissions from the city would have aged roughly 5-7 hours by the time they
were sampled in the farthest passes.
The general direction of the airflow was corroborated by the alignment
of power plant plumes in successive traverses. All of the major (greater than
1,000 MW capacity) power plants lying in and immediately upwind of the sampling
area are noted in Figure 1. The plumes from some of these plants can be identi-
field in Figure 1 by their ozone "deficits", which result from the scavenging of
ambient ozone by plume NO. Th e Coffeen plant, 85 km WNW of St. Louis, is the
only major downwind source known to the authors which lay within the July 18
urban plume.
The MISTT plume mapping program was designed to characterize pollutant
concentrations and winds over cross sections of the plume at discrete distances
downwind from the sources. From these measurements, the horizontal mass flow
rates of pollutants at each distance can be calculated directly :
FLOW RATE = fu^x, z) f(C(x,y, z) - CQ(X, z))dydz ,
where x is distance downwind, y, z are cross-wind and vertical coordinates, u is
ji
wind speed (from pilot balloon observations), C is pollutant concentration (from air-
craft measurements), and C is average pollutant concentration outside the plume
(from aircraft measurements). Figure 2b shows the flow rates of ozone,
b and particulate sulfur on July 18. These flow rates represent the
SCcLX
118
-------
contribution of metropolitan St. Louis (augmented slightly by the Coffeen
power plant) to atmospheric loadings, and are unaffected by cross-plume
and vertical dispersion.
The flow rates of ozone, bgcat, and particulate sulfur all increased with
distance downwind of St. Louis on July 18, reflecting the secondary origin of
all of the ozone and most of the light-scattering aerosol. The ozone flow rate
leveled off at about 125 T/hr, which can be interpreted as the rate at which
ozone was produced within the urban plume. The estimated annual average
emissions of non-methane hydrocarbons in metropolitan St. Louis are about
1.6 X 10 T/yr(ref. 9); this corresponds to 18-36 T/hr, depending onwhether emis
sions are spread out through the day or concentrated in the daylight hours.
Comparison of these two figures shows that a net one to two parts ozone was
produced in the urban plume for each part (C) hydrocarbons emitted in the
metropolitan area . This is much more efficient production than is found in
most laboratory studies , which are conducted at higher reactant concentra-
tions.
Most of the increase in the b flow rate was observed downwind of
scat
the major increase in the ozone flow rate, which is consistent with the find-
ing in laboratory studies that aerosol production lags ozone production. The
calculated particulate sulfur flow rates are less accurate than those for ozone
and b particularly at the last two sampling distances, due to the more
scat r
limited data available on background concentrations. The ratio of the b
SC3>t
119
-------
and particulate sulfur flow rates indicates that sulfate compounds accounted
for most of the new aerosol.
The St. Louis urban plume was mapped on a total of eight days during
the July 15-August 15, 1975 MISTT experiment. The object of Project MISTT
was not to develop a climatology of the urban plume, but to measure it in de-
tail under conditions favorable for doing so, and sampling days were chosen
according to this criterion. A plume of polluted air was always found down-
wind of the city, regardless of wind direction , confirming our attribution
here of the excess concentrations to emissions from metroplitan St. Louis
rather than some upwind source. On August 11, 1975, the plume again stretched
out toward the northeast, and was mapped in detail out to 145 km and tracked out
to 240 km. Winds on this date were lighter than on July 18, but the plume was
as well defined, and concentrations and flow rates were similar.
W. H. White, J. A. Anderson and D. L. Blumenthal
Meteorology Research, Inc.
Altadena, California 91001
R. B. Husar, N. V. Gillani and J. D. Husar
Washington University
St. Louis, Missouri 63130
W. E. Wilson, Jr.
U. S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
120
-------
FOOTNOTES AND REFERENCES
1. P. L. Haagenaon and A. L. Morris, J. of Applied Meteorology _13>, 901 (1974).
2. J. F. Stampfer, Jr. and J. A. Anderson, Atmospheric Environment _9» 301
(1975).
3. R. J. Breeding, P. L. Haagenson, J. A. Anderson, J. P. Lodge, Jr., and
J. F. Stampfer, Jr., J. of Applied Meteorology _14, 204(1975).
4. R. J. Breeding, H. B. Klonis, J. P. Lodge, Jr., J. B. Pate, D. C.
Sheesley, T. R. Englert and D. R. Sears, Atmospheric Environment 10,
181 (1976).
5. R. B. Husar, J. D. Husar, N. V. Gillahi, S. B. Fuller, W. H. White,
W. M. Vaughan and W. E. Wilson, Jr., Proc. of the Div. Environmental
Chemistry, 171st National ACS Meeting, New York, April 1976.
6. K. T. Whitby, B. K. Cantrell, R. B. Husar, N. V. GiEani, J. A. Ander-
son, D. L. Blumenthal and W. E. Wilson, Jr. , ibid; W. E. Wilson, Jr.,
R. J. Charlson, R. B. Husar, K. T. Whitby, D. L. Blumenthal, Proc. 69th
Annual Meeting Air Pollution Control Assoc., paper #76-30-06, June 1976.
7. The light-scattering coefficient gives the rate (fraction per distance) at
which a beam of light is scattered in all directions, and is the main deter-
minant of visual range in the atmosphere. In polluted urban air, the light-
scattering coefficient is roughly proportional to the mass concentration of
aerosol particles in the 0. 2-2. 0 Mm diameter range, those inhaled most
deeply into the lungs. See R. J. Charlson, N. C. Ahlquist, H. Selvidge,
and P. B. MacCready, Jr., J. Air Pollution Control Assoc. 1^ 937 (1969).
8. The sampling aircraft was instrumented for O^, NO, NO_ and SO_ concen-
trations, aerosol size distribution, light scattering coefficient, electrical
charge acceptance and condensation nuclei count, temperature, relative
humidity, dew point and turbulent dissipation, as described by W. H.
\
White, J. A. Anderson, W. R. Knuth, D. L. Blumenthal, J. C. Hsiung
and R. B. Husar, Final Report No. MRI76-FR-1414 by Meteorology
121
-------
Research, Inc., Altadena, CA. Submicron aerosol samples were col-
lected for sulfate analysis as described by J. D. Husar, R. B. Husar,
and P. K. Stubits, Analytical Chemistry 47, 2062 (1975).
9. U.S. E.P.A., Illinois E.P.A.
10. This point was brought to the authors' attention by H. M. Walker, Proc.
of the Conf. on Ozone/Qxidants, Air Pollution Control Assoc., Dallas,
March 1976.
11. W. E. Wilson, Jr., D. F. Miller, A. Levy and R. K. Stone, J. Air Pol-
lution Control Assoc. 23, 949 (1973).
12. A. P. Waggoner, A. J. Vanderpol, R. J. Charlson, S. Larson, L.
Granat and C. Tragardh, Nature 26l_, 120 (1976); W. H. White, P. T.
Roberts and S. K. Friedlander, in the Character and Origins of Smog
Aerosol. G. M. Hidy and P. K. Mueller, Eds. to appear.
This research was supported by the U. S. Environmental Protection Agency,
Environmental Sciences Research Laboratory, Aerosol Research Branch.
This paper is MISTT Report Number 4.
122
-------
LEGENDS
Figure 1. Ozone concentration and light-scattering coefficient, b ,
scat
downwind of St. Louis on July 18, 1975. Data are taken
from horizontal traverses by instrumented aircraft, at alti-
tudes indicated in Figure 2a. Graph baselines show sampling
paths; note that baseline concentrations are not zero.
Figure 2. Traverse altitudes and pollutant flow rates in St. Louis urban
plume on July 18, 1975. Data are plotted against distance
downwind of the St. Louis Gateway Arch, (a) Location of
horizontal traverses; solid dots correspond to traverses shown
in Figure 1. Mixing heights were determined from aircraft sound-
ings. Approximate time (CDT)of sampling is shown at bottom.
(b) Flow rates (in excess of background) of ozone (O ), light-
scattering coefficient (b ), and particulate sulfur (S ).
o CSLt TL)
123
-------
'!
I1
NJ
OZONE
bSCAT
CHAMPAIGN
WINDS
URBANA
MISSOURI
ILLINOIS
50
A POWER PLANT
REFINERY
KILOMETERS
71-4IS
Figure 1
-------
20
Q.
C/)
B
0
DISTANCE DOWNWIND, km
50 100 150
1
A
iM
W^_ -
f"
0915
1
«5
G - ""
- ""**
o
1230
i
\G^_ """"
-~ "
0
o
1400
_ --
0.0
0 0
1645 1815
I
V7
E
1.0 uf
Q
0.5 t
i
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/5-76-110
2.
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
MIDWEST INTERSTATE SULFUR TRANSFORMATION AND TRANSPORT
PROJECT: Aerial Measurements of Urban and Power Plant
Plumes, Summer 1974
5. REPORT DATE
November 1976
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
W.H. White, J.A. Anderson, W.R. Knuth, D.L. Blumenthal,
J.C. Hsiung, and R.B. Husar.
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Meteorology Research, Inc.
464 West Woodbury Road
Box 637
Altadena, CA 91001
10. PROGRAM ELEMENT NO.
1AA603
11. CONTRACT/GRANT NO.
68-02-1919
12. SPONSORING AGENCY NAME AND ADDRESS
Environmental Sciences Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
13. TYPE OF REPORT AND PERIOD COVERED
FINAL 7/74-6/76
14. SPONSORING AGENCY CODE
EPA-ORD
15. SUPPLEMENTARY NOTES
16. ABSTRACT
A portion of the research activities of the Midwest Interstate Sulfur Transformation
and Transport Project (Project MISTT) during the summer of 1974 is documented.
Using a light plane equipped with instruments for measuring air pollutants and
meteorological parameters, investigators mapped the three-dimensional distribution
of aerosols and pollutant gases originating in the St. Louis area. Each day's
flight plan was designed to characterize a large pollutant plume at discrete
distances downwind from its source. The plume from the coal-fired power plant at
Labadie, Missouri was followed out to 45 km. Secondary aerosol production within
the plume was documented. The estimated average conversion rate for sulfur dioxide
to sulfate was about three percent/hour at the distances sampled. The overall
removal rate of S02 was too small to detect, and no net production of ozone was
observed. Large pollutant plumes were also identified downwind of central St. Louis
and the Wood River refineries. These urban-industrial plumes were followed out to
60-70 km, where they were characterized by elevated concentrations of ozone and
light-scattering aerosols.
7.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
*Air pollution
*Sulfur dioxide
*Sulfate
*Aerosols
Ozone
*Electric power plants
*Plumes
*Chemical Reactions
Transport properties
13B
07B
07D
10B
21B
07D
8. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (ThisReport)
UNCLASSIFIED
21. NO <">c PAGES
136
20. SECURITY CLASS (Thispage)
UNCLASSIFIED
22. PRICE
EPA Form 2230-1 (9-73)
126
------- |